NXG 2023 9 Issue-6

Volume 9, Number 6, December 2023
Neurology.org/NG
A peer-reviewed clinical and translational neurology open access journal
RESEARCH ARTICLE
mTOR Pathway Somatic Pathogenic Variants in Focal
Malformations of Cortical Development: Novel Variants,
Topographic Mapping, and Clinical Outcomes e200103
RESEARCH ARTICLE
Adult Phenotype of SYNGAP1-DEE e200105
RESEARCH ARTICLE
A Phenotypic Atlas for Huntington Disease Based on
Data From the Enroll-HD Cohort Study e200111
RESEARCH ARTICLE
Genetic Patterns of Selected Muscular Dystrophies
in the Muscular Dystrophy Surveillance, Tracking,
and Research Network e200113
Academy Officers Neurology® is a registered trademark of the American Academy of Neurology
(registration valid in the United States).
Carlayne E. Jackson, MD, FAAN, President
Neurology® Genetics (eISSN 2376-7839) is an open access journal published
Natalia S. Rost, MD, MPH, FAAN, FAHA, President Elect online for the American Academy of Neurology, 201 Chicago Avenue,
Janis M. Miyasaki, MD, MEd, FRCPC, FAAN, Vice President Minneapolis, MN 55415, by Wolters Kluwer Health, Inc. at 14700 Citicorp
Drive, Bldg. 3, Hagerstown, MD 21742. Business offices are located at Two
Sarah M. Benish, MD, FAAN, Secretary Commerce Square, 2001 Market Street, Philadelphia, PA 19103. Production
offices are located at 351 West Camden Street, Baltimore, MD 21201-2436.
Charles C. Flippen II, MD, FAAN, Treasurer © 2023 American Academy of Neurology.
Orly Avitzur, MD, MBA, FAAN, Past President Neurology® Genetics is an official journal of the American Academy of
Neurology. Journal website: Neurology.org/ng, AAN website: AAN.com
CEO, American Academy of Neurology Copyright and Permission Information: Please go to the journal website
(www.neurology.org/ng) and click the Permissions tab for the relevant
Mary E. Post, MBA, CAE article. Alternatively, send an email to healthpermissions@wolterskluwer.com.
Chief Executive Officer General information about permissions can be found here: https://shop.lww.com/
journal-permission.
20l Chicago Ave Disclaimer: Opinions expressed by the authors and advertisers are not
Minneapolis, MN 55415 necessarily those of the American Academy of Neurology, its affiliates, or of
the Publisher. The American Academy of Neurology, its affiliates, and the
Tel: 612-928-6000 Publisher disclaim any liability to any party for the accuracy, completeness,
efficacy, or availability of the material contained in this publication
(including drug dosages) or for any damages arising out of the use
Editorial Office or non-use of any of the material contained in this publication.
Patricia K. Baskin, MS, Senior Director, Publications and Executive Editor Advertising Sales Representatives: Wolters Kluwer, 333 Seventh Avenue,
Rachel A. Anderson, Senior Publications Assistant New York, NY 10001. Contacts: Eileen Henry, tel: 732-778-2261, fax: 973-215-
2485, eileen.henry@wolterskluwer.com and in Europe: Craig Silver, tel: +44
Adria Gottesman-Davis, PhD, Manager, Journal Editorial & Content Development 7855 062 550 or e-mail: craig.silver@wolterskluwer.com.
Careers & Events: Monique McLaughlin, Wolters Kluwer, Two Commerce
Morgan S. Sorenson, Managing Editor Square, 2001 Market Street, Philadelphia, PA 19103, tel: 215-521-8468, fax: 215-
521-8801; monique.mclaughlin@wolterskluwer.com.
Neurology® Neuroimmunology & Neuroinflammation
Reprints: Meredith Edelman, Commercial Reprint Sales, Wolters Kluwer, Two
Neurology® Genetics Commerce Square, 2001 Market Street, Philadelphia, PA 19103, tel: 215-356-2721;
Neurology® Education meredith.edelman@wolterskluwer.com; reprintsolutions@wolterskluwer.com.
Special Projects: US & Canada: Alan Moore, Wolters Kluwer, Two
Kathleen M. Pieper, Senior Managing Editor, Neurology® Commerce Square, 2001 Market Street, Philadelphia, PA 19103, tel:
Karen Skaja, Senior Editorial Coordinator 215-521-8638, alan.moore@wolterskluwer.com. International: Andrew
Wible, Senior Manager, Rights, Licensing, and Partnerships, Wolters Kluwer;
Skyler M. Kane, Assistant Managing Editor translationrights@wolterskluwer.com.
Sarabeth Ng, Editorial Coordinator
Lee Ann Kleffman, Managing Editor, Neurology® Clinical Practice
Andrea Rahkola, ELS, Production Editor
Kristen Swendsrud, Production Coordinator
Aubrey Zalewski, Production Coordinator
Sharon L. Quimby, Digital Managing Editor

Brooke Eikenberry, Digital Marketing Communications Administrator
Kaitlyn Aman Ramm, Senior Digital Multimedia/Graphics Coordinator
Anna Rehbein, Digital Multimedia/Podcast Coordinator
Publisher
Wolters Kluwer
Baltimore, MD
Publishing Staff
Thomas Pacific, Lead Publisher
Jessica Heise, Production Team Leader, Neurology Journals
Emily Moore, Senior Production Editor
Steve Rose, Editorial Assistant
Stacy Drossner, Production Associate
Copyright ª 2023 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
A peer-reviewed clinical and translational neurology open access journal Neurology.org/NG
Neurology ® Genetics
Editor
Stefan M. Pulst, MD, Dr med, FAAN Vision Neurology®: Genetics will be the premier peer-
reviewed journal in the field of neurogenetics.
Deputy Editor
Massimo Pandolfo, MD, FAAN
Associate Editors Mission Neurology: Genetics will provide neurologists
Alexandra Durr, MD, PhD and clinical research scientists with
Suman Jayadev, MD outstanding peer-reviewed articles,
Raymond P. Roos, MD, FAAN editorials, and reviews to elucidate the role
Antonella Spinazzola, MD
of genetic and epigenetic variations in
Editorial Board diseases and biological traits of the central
Danielle M. Andrade, MD, MSc, and peripheral nervous systems.
FRCPC, CSCN (EEG)
Geneviève Bernard, MD, MSc, FRCPc Caterina Mariotti, MD
Elizabeth E. Blue, PhD Paolo Moretti, MD
Stefan Nicolau, MD Editorial Tel: 612-928-6400
Francois Bolduc, MD, FRCPC, PhD
Davide Pareyson, MD Inquiries Toll-free: 800-957-3182 (US)
Joshua Bonkowsky, MD, PhD
Giovanni Coppola, MD Gerald Pfeffer, MD, PhD, FRCPC Fax: 612-454-2748
Louise A. Corben, PhD Catarina M. Quinzii, MD ngjournal@neurology.org
M. Elizabeth Ross, MD, PhD, FANA
Chantal Depondt, MD, PhD
Barbara Scelsa, MD
AntoineDuquette,MD,MSc,FRCP(C)
Susanne A. Schneider, MD, PhD
Brent L. Fogel, MD, PhD, FAAN Stay facebook.com/NeurologyGenetics
Joshua M. Shulman, MD, PhD, FAAN
Alica M. Goldman, MD, PhD, MS, FAES Connected
Shoji Tsuji, MD, PhD twitter.com/greenjournal
Anthony J. Griswold, PhD
Paul N. Valdmanis, PhD
Andrea L. Gropman, MD, FAAP,
David Viskochil, MD, PhD youtube.com/user/NeurologyJournal
FACMG, FANA
Orhun H. Kantarci, MD Juliane Winkelmann, MD
Juan I. Young, PhD instagram.com/aanbrain
Julie R. Korenberg, PhD, MD
Carla Marini, MD, PhD
Neurology ® Journals
Editor-in-Chief Classification of Evidence

José G. Merino, MD, MPhil, FAAN Review Team
Deputy Editor Malik M. Adil, MD

Olga Ciccarelli, MD, PhD, FRCP Gregory S. Day, MD, MSc, MSCI,
FAAN
Section Editors Guillermo Delgado-Garcı́a, MD, MSc
Associate Editor, Methodology and Statistics James Dorman, MD, FAAN, FACP
Andrea L.C. Schneider, MD, PhD RichardM.Dubinsky,MD,MPH,FAAN
Classification of Evidence Evaluations
Jeffrey J. Fletcher, MD, MSc
Gary S. Gronseth, MD, FAAN Gary M. Franklin, MD, MPH, FAAN
David S. Gloss II, MD, MPH&TM
Diversity, Equity, & Inclusion John J. Halperin, MD, FAAN
Nicole Gonzales, MD Jason Lazarou, MSc, MD
Roy H. Hamilton, MD, MS, FAAN Steven R. Messé, MD, FAAN
H. E. Hinson, MD, MCR, FAAN Zurab Nadareishvili, MD, PhD
Podcasts Pushpa Narayanaswami, MBBS, DM,
Stacey L. Clardy, MD, PhD, FAAN FAAN
Jeffrey B. Ratliff, MD, Deputy Editor Sonja Potrebic, MD, PhD, FAAN
Scientific Integrity Advisor
David S. Knopman, MD, FAAN
Ombudsperson
Lisa DeAngelis, MD, FAAN
Copyright © 2023 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
TABLE OF CONTENTS Volume 9, Number 6, December 2023 Neurology.org/NG
Research Articles e200113 Genetic Patterns of Selected Muscular Dystrophies

in the Muscular Dystrophy Surveillance,
e200096 IRF2BPL Causes Mild Intellectual Disability
Tracking, and Research Network
Followed by Late-Onset Ataxia
P.B. Kang, M. Jorand-Fletcher, W. Zhang, S.W. McDermott,
S. Heide, C.-S. Davoine, P. Cunha, C. Scherer-Gagou, B. Keren, R. Berry, C. Chambers, K.N. Wong, Y. Mohamed, S. Thomas,
G. Stevanin, P. Charles, D. Heron, A. Brice, and A. Durr Y.S. Venkatesh, C. Westfield, N. Whitehead, and N.E. Johnson,
Open Access for the Muscular Dystrophy Surveillance, Tracking, and Research
Network (MD STARnet)
e200103 mTOR Pathway Somatic Pathogenic Variants in Open Access
Focal Malformations of Cortical Development:
Clinical/Scientific Notes
Novel Variants, Topographic Mapping, and
Clinical Outcomes e200098 Expanding the Clinical Spectrum of UBTF-Related
E. Krochmalnek, A. Accogli, J. St-Onge, N. Addour-Boudrahem, Neurodevelopmental Disorder
G. Prakash, S.-H. Kim, T. Brunette-Clement, G. Alhajaj,
L. Mougharbel, E. Bruneau, K.A. Myers, F. Dubeau, A. Pietra, F. Palombo, M. Giannotta, M. Maffei, C. Fiorini, R. Costa,
J. Karamchandani, J.-P. Farmer, J. Atkinson, J. Hall, G. Cenacchi, V. Carelli, D.M. Cordelli, A. Pini, and C. Garone
C. Chantal Poulin, B. Rosenblatt, J. Lafond-Lapalme, A. Weil, Open Access Video
C. Fallet-Bianco, S. Albrecht, N. Sonenberg, J.-B. Riviere,
R.W. Dudley, and M. Srour
e200100 Ataxia Syndrome With Hearing Loss and
Open Access
Nephronophthisis Associated With a Novel
e200105 Adult Phenotype of SYNGAP1-DEE Homozygous Variant in XPNPEP3
I. Ben-Shabat, M. Kvarnung, W. Sperker, H. Bruhn, A. Wredenberg,
M. Rong, T. Benke, Q. Zulfiqar Ali, Á. Aledo-Serrano, A. Bayat,
R. Wibom, I. Nennesmo, M. Engvall, and M. Paucar
A. Rossi, O. Devinsky, F. Qaiser, A.S. Ali, A. Fasano,
A.S. Bassett, and D.M. Andrade Open Access Video
Open Access
e200101 Novel SLC13A3 Variants and Cases of Acute
e200107 Molecular Diagnosis of Facioscapulohumeral Reversible Leukoencephalopathy and
Muscular Dystrophy in Patients Clinically α-Ketoglutarate Accumulation and
Suspected of FSHD Using Optical Literature Review
Genome Mapping K.N. Wong, L.D. Botto, M. He, P.R. Baker II, A.L. Vanderver, and
J.L. Bonkowsky
N.M. Guruju, V. Jump, R. Lemmers, S. Van Der Maarel, R. Liu,
B.R. Nallamilli, S. Shenoy, A. Chaubey, P. Koppikar, R. Rose, Open Access
S. Khadilkar, and M. Hegde
Open Access e200102 Agenesis of Pectoralis Major Muscle in Late-Onset
GFPT1-Related Congenital Myasthenic
e200109 Estimated Familial Amyotrophic Lateral Sclerosis Syndrome: A Case Report
Proportion: A Literature Review and E.K. Williams, C. Shea, and P. Gonzalez-Perez
Meta-analysis Open Access
J. Barberio, C. Lally, V. Kupelian, O. Hardiman, and W.D. Flanders
Open Access e200106 PMPCA-Related Encephalopathy: Novel Variants,
Phenotype Extension, and Mitochondrial
e200111 A Phenotypic Atlas for Huntington Disease Based Morphology
on Data From the Enroll-HD Cohort Study V. Rambani, M. Kolnikova, M. Cagalinec, M. Skopkova, and
D.R. Langbehn, S.S. Sathe, C. Loy, C. Sampaio, and E.A. Mccusker D. Gasperikova
Open Access Open Access
TABLE OF CONTENTS Volume 9, Number 6, December 2023 Neurology.org/NG
NeuroImages
e200104 FOLR1 Gene Variation With Adult-Onset Cerebral

Folate Deficiency and Stable Clinical and MRI
Features up to 2 Years
C. Manco, R. Cortese, M. Alberti, S. Bianchi, L. Monti,
N. De Stefano, and C. Battisti
Open Access
Cover image
Baseline diffusion tensor imaging analysis comparing a patient with
FOLR1 gene mutation with adult-onset cerebral folate deficiency to a sex-
and age-matched healthy control patient. Stylized by Kaitlyn Aman
Ramm, Senior Digital Multimedia/Graphics Coordinator.
See page e200104
RESEARCH ARTICLE OPEN ACCESS
IRF2BPL Causes Mild Intellectual Disability Followed

by Late-Onset Ataxia
Solveig Heide, MD, Claire-Sophie Davoine, BS, Paulina Cunha, MD, Clarisse Scherer-Gagou, MD, Correspondence
Dr. Durr
Boris Keren, MD, PhD, Giovanni Stevanin, PhD, Perrine Charles, MD, PhD, Delphine Heron, MD,
alexandra.durr@icm-institute.org
Alexis Brice, MD, PhD, and Alexandra Durr, MD, PhD
Neurol Genet 2023;9:e200096. doi:10.1212/NXG.0000000000200096
Abstract
Background and Objectives
Neurodevelopmental and neurodegenerative disorders have long been considered as dif-
ferent clinical and molecular entities, and only a few genes are known to be involved in both
processes. The IRF2BPL (interferon regulatory factor 2 binding protein like) gene was
implicated in a severe pediatric phenotype characterized by developmental and epileptic
encephalopathy and early regression. In parallel, inherited IRF2BPL variants have been
reported in cohorts of patients with late-onset progressive dystonic and ataxic syndrome with
few information about the neurodevelopment of these patients. This study aimed to describe
both neurodevelopmental and neurodegenerative aspects of the phenotype in adults with
IRF2BPL pathogenic variant.
Methods
We report here the clinical and molecular data of 18 individuals carrying truncating IRF2BPL
variants (identified by either exome or genome sequencing), including a large pedigree of
16 patients presenting with a neurodevelopmental disorder (NDD) associated with late-onset
cerebellar ataxia and atrophy.
Results
Genome sequencing identified the p.(Gln117*) variant in a large family first assessed for
familial ataxia, with multiple individuals presenting with NDD. The p.(Ser313*) variant was
identified by exome sequencing in a second family with a young adult patient with NDD
without ataxia which was inherited from her asymptomatic mother, suggesting incomplete
penetrance of IRF2BPL-linked disorders.
Discussion
This study illustrates the importance of neurologic evaluation of adult patients initially
diagnosed with NDD to detect a late-onset neurodegenerative condition. Two different
disorders may be clinically diagnosed in the same family, when not considering that NDD
and late cerebellar changes may be part of the same molecular spectrum such as for
IRF2BPL.
From the Genetic Department (S.H., B.K., P. Charles, D.H., A.D.), Assistance Publique-Hôpitaux de Paris (AP-HP) Pitié-Salpêtrière; Reference Center for Rare Diseases « Intellectual
disabilites of rare causes » « Déficiences Intellectuelles de Causes Rares » (S.H., P. Charles, D.H.), Pitié-Salpêtrière Hospital; Sorbonne Université (C.-S.D., P. Cunha, G.S., A.B., A.D.), Paris
Brain Institute (ICM Institut du Cerveau), INSERM, CNRS, Assistance Publique-Hôpitaux de Paris (AP-HP); Department of Neurology (C.S.-G.), University Hospital d’Angers; and INCIA
(G.S.), EPHE, Université de Bordeaux, France.
Go to Neurology.org/NG for full disclosures. Funding information is provided at the end of the article.
The Article Processing Charge was funded by the authors.

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading
and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.
Copyright © 2023 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1
Glossary
ID = intellectual disability; NDD = neurodevelopmental disorder.
Introduction samples was >96%, >89%, and >93% at a 20× depth threshold,
for SAL-394-10, 13, and IIA patients, respectively. Variants
Neurodevelopmental disorders (NDDs) are defined by im- were annotated with SnpEff 4.3, dbNSFP 2.9.3, gnomAD,
pairments in cognition, communication, behavior, and/or ClinVar, HGMD, and OMIM. Filtering was performed with
motor skills resulting from abnormal brain development, criteria based on the consequence on the protein and fre-
manifesting either in utero or during early postnatal life. More quency in gnomAD.
than 1,000 genes have been implicated in NDD, mostly highly
penetrant and evolutionarily constrained fetal brain-expressed All candidate variants and their segregation within pedigrees
genes.1,2 Neurodegenerative disorders are characterized by were further confirmed by Sanger sequencing. We used the
progressive neurodegeneration which results in progressive NM_024496.3 transcript as the reference sequence.
decline variably affecting cognition and behavior, motor and/
or sensory functions, and presenting mostly in adulthood. The Standard Protocol Approvals, Registrations,
developmental and degenerative processes have long been and Patient Consents
considered as different clinical and biological entities. More All procedures followed were in accordance with the ethical
recently, some common denominators and interactions be- standards in accordance with local French legislation (ap-
tween neurodevelopmental and neurodegenerative disorders proval from local ethics committees on December 19, 1990
have emerged suggesting that proteins implicated in neuro- and November 10, 1992). Written informed consent was
degenerative disorders play important roles in brain de- obtained from all patients and/or their legal representatives.
velopment. For example, pathogenic variants in the RAB39B
and WDR45 genes are responsible for phenotypes character- Data Availability
ized by early neurodevelopmental disorder with intellectual The data that support the findings of this study are available
disability (ID) and secondary parkinsonism.3 Severe infantile from the corresponding author on reasonable request.
onset developmental and epileptic encephalopathy are caused
by mutations in the autophagy gene WDR45.4
Results
We identified an IRF2BPL (interferon regulatory factor 2 bind- Family SAL-394
ing protein like) variant segregating in a previously unreported This index case, SAL-394-013, experienced the onset of un-
large pedigree of 16 patients presenting with NDD associated
steady gait due to cerebellar ataxia at age 30 years and had a
with cerebellar ataxia which appeared later in life and in a spo- progressive worsening of her ability to walk making wheel-
radic case with NDD inherited from her asymptomatic mother. chair use necessary at age 40 years (Figure 1). Reflexes were
increased in all limbs with unilateral extensor plantar reflex
and Hoffman signs, mild proximal weakness but no wasting.
Methods Both arms showed dystonic postures, and there was a mild
Both families have been examined at the Pitié-Salpêtrière loss of facial mimicry. Eye movements were abnormal because
University Hospital 28 years apart. of the presence of gaze-evoked nystagmus and a limited up-
ward gaze. She complained of swallowing difficulties but not
Exome and Genome Sequencing of urinary problems. Cognitive impairment was clinically
Two individuals (SAL-394-10 and 13) underwent genome suspected. Cerebral MRI showed mild global cortical atrophy,
sequencing on a HiSeq X Five (Illumina). Patient IIA had trio moderate cerebellar atrophy with normal brainstem volume,
exome sequencing on a NextSeq 500 Sequencing System and no white matter changes (Figure 2). Nerve conduction
(Illumina, San Diego, CA), with a 2 × 150 bp high output velocities were normal as was the muscle biopsy. Visual-
sequencing kit after a 12-plex enrichment with the SeqCap EZ evoked potentials showed normal optic nerve conduction
MedExome kit (Roche, Basel, Switzerland), according to the time; auditory-evoked potentials were abnormal with delayed
manufacturer’s specifications. bulbar and brainstem latencies. Somatosensory-evoked po-
tentials were evocative of abnormal bilateral thalamocortical
For all patients, sequence quality was assessed with FastQC connections and impaired bilateral lemniscus fibers. This was
0.11.5, then the reads were mapped using BWA-MEM (ver- also reflected by decreased vibration detection at the ankles.
sion 0.7.13), sorted and indexed in a bam file (samtools 1.4.1),
duplicates were flagged (sambamba 0.6.6), and coverage was Family history revealed several other affected members, and
calculated (picard tools 2.10.10). Variant calling was per- the family had been seen at their homes by AD and AB (see
formed with GATK 3.7 Haplotype Caller. Coverage for these Table). Ages at examination ranged from 11 to 74 years. Ages
2 Neurology: Genetics | Volume 9, Number 6 | December 2023 Neurology.org/NG

Figure 1 Family Structure of SAL-394
Bold symbols indicate that individuals had cerebellar and pyramidal signs; small squares indicate intellectual disability only. Genotypes are indicated;
heterozygous carriers are +/−. The arrow indicates the index case. Deceased individuals are crossed out.
at death ranged from 58 to 77 (n = 5) and ataxia durations were also noted without specificities. A clinical geneticist
from 21 up to 43 years. Patients presented with variable specialized in neurodevelopmental disorders (SH) contacted
combinations of intellectual disability and/or a cerebellar several members of the family to better delineate the neuro-
ataxia with pyramidal signs. Half of the patients (8/16) had developmental trajectory of the affected members (psycho-
ataxic features with onset between ages 21 and 53 years. Seven motor development, scholarship, and acquisition of writing
individuals (021, 035, 040, 044, 045, 047, and 048) had very and reading). No formal IQ scores were available for these
slight difficulties with sway in the upright position with feet patients, but it seems that all affected members presented with
together or in tandem walking or isolated mild dysarthria. mild-to-moderate ID. This study indicates that all variant
These very slight signs were confirmed in 3 (040, 045, and carriers examined after the age of 35 years had clinical signs of
047) seen first in their twenties, with evident cerebellar ataxia cerebellar ataxia and pyramidal signs. Those examined at a
in their thirties or even fifties, in addition to their mild in- younger age had intellectual difficulties, and several already
tellectual difficulties since school. Reflexes were increased in had increased reflexes (4/9) and/or minimal cerebellar signs.
most (11/16), while plantar reflexes were extensor in 5/16. We could not reach 4 patients from the initial family.
ID was present in all individuals evaluated. Evaluations were
not available for the oldest patients (004, 005, 010, 012, 013, Family II
and 040) who all had severe cerebellar signs and no speech for Patient IIA was the third child of nonconsanguineous healthy
3. Neurodevelopmental difficulties in most affected members parents. She was born eutrophic at term after an uneventful
Figure 2 Cerebral MRI From SAL-394-048 After 10 Years of Ataxia Duration
T1-weighted MPRAGE sagittal (A) and axial (B) views. Mild but
visible cerebellar (vermian), mesencephalic and lower
brainstem atrophy, as well as general cortical thinning.
Neurology.org/NG Neurology: Genetics | Volume 9, Number 6 | December 2023 3

4
Table Clinical Characteristics of 2 Families (SAL-394 and Family II) Including 18 Patients Carrying the IRF2BPL Variant p.Gln117Ter and p.(Ser313*), Respectively, Listed
According to Age at Examination
Developmental
delay
Motor Ataxia/SPATAX
(age Work in a Reading and Dysarthria disability score Reflexes,
when Language School protected writing (onset age (onset age knee and Cognitive
ID walking) (age) performance environment abilities years) years) plantar Oculomotor signs Extrapyramidal signs/other decline
FAMILY
SAL-394
004 NA NA NA NA Normal Severe (38, no Severe/7 (38) Abolished, No saccades, limited vertical Normal/general wasting, Probably
speech since age 64) extensor and horizontal gaze swallowing
005 NA NA NA NA NA Severe (38, no Severe/7 (38) Increased, Slow saccades Dystonic postures, chorea/ Probably
speech at age 70) extensor Limited upward gaze swallowing
010 NA NA NA Yes NA Moderate (53) Moderate/4 (53) Increased, Limited upward gaze Normal/swallowing, No
extensor decreased sense of vibration
at ankles 5/8
012 NA NA NA NA NA Severe (37) Moderate/4 (37) Normal Slow saccades Facial masking/decreased No
Limited upward gaze sense of vibration 5/8/axonal
Neurology: Genetics | Volume 9, Number 6 | December 2023

neuropathy, urinary problems
013 NA NA NA Yes Nl (help Moderate (33) Severe/6 (30) Increased, Nystagmus, limited upward Dystonia UL, facial masking Yes
(wheelchair at needed) extensor gaze
age 40)
021 NA NA NA NA Normal Mild No Normal extensor Not testable No
033 No Yes (no No work No writing, no No speech No Increased NA No NA

(15 mo) speech) possible reading
034 No No Adapted Yes (gardener) Only his name No No Increased Normal No NA

(15 mo) school indifferent
040 NA NA Adapted Yes (legally Difficulties No No (severe/6 at Increased No NA

school protected) age 50)
035 Yes Yes Adapted Yes No writing, no No Cannot walk on Normal No cataract No NA
(18 mo) school reading a line/0
044 NA NA Normal NA Normal No Mild sway at age Increased, Normal No No

24 indifferent
045 No NA Difficulties, Yes Difficulties 53 Mild/0 (53) Normal at age 26 Normal at age 26 NA NA
left at age 13
047 Yes A Adapted Yes Difficulties No and mild (38) No and mild/ Increased, at age Normal/pes cavus, EMG No No
(20 mo) school (help needed) 0 (38) 48 spastic gait normal
Continued
Neurology.org/NG
Table Clinical Characteristics of 2 Families (SAL-394 and Family II) Including 18 Patients Carrying the IRF2BPL Variant p.Gln117Ter and p.(Ser313*), Respectively, Listed
According to Age at Examination (continued)
Developmental
delay
Neurology.org/NG
Motor Ataxia/SPATAX
(age Work in a Reading and Dysarthria disability score Reflexes,
when Language School protected writing (onset age (onset age knee and Cognitive
ID walking) (age) performance environment abilities years) years) plantar Oculomotor signs Extrapyramidal signs/other decline
048 Yes Yes Difficult No Difficult (help No Mild sway at age Increased, flexor Normal No/scoliosis No
needed) 21
049 Yes Yes Left at age 14 Yes (legally Difficult No No Increased Normal No NA
Adapted protected)
school
050 No Yes Adapted Yes (legally Difficult No No Increased Normal No No

(17 mo) school protected)
Isolated
case
IIA Yes Yes Adapted Yes (legally No reading, No No Normal at age 25 Normal at age 25 No No
(25 mo) school protected) No writing
IIB No No Difficult No Normal No NA NA NA NA NA
Index cases are in bold. SPATAX disability score (0: no functional handicap; 1: no functional handicap but signs at examination; 2: able to run, walking unlimited; 3: unable to run, limited walking without aid; 4: walking with one
cane; 5: walking with 2 canes; 6: unable to walk, requiring wheelchair; 7: confined to bed).
Abbreviations: ext plantar = extensor plantar reflex (Babinski sign); NA = not assessed.

5
pregnancy. Her first months of life were normal. She pre- ubiquitin-dependent degradation of target proteins and is
sented with a global developmental delay with unsupported ubiquitously expressed in human tissues, including the brain.
sitting acquired after age 9 months, independent walking ac- This protein plays a role in the development of the CNS and in
quired at 25 months, and a language delay with first words neuronal maintenance through Wnt (Wingless/integrated)
emerging around 4 years. She went to a special school from signaling. The Wnt family of ligand glycoproteins acts as key
the age of 8 because of learning difficulties and had speech and regulators of the development especially in the CNS by regu-
psychomotor therapies during childhood. She had no history lating cell proliferation, migration, differentiation, and synapse
of epilepsy. At the last evaluation at the age of 25 years, she development.5,6
had moderate intellectual disability, was not able to read or
write, and was under curatorship. Her neurologic examination Heterozygous loss of function variants in the IRF2BPL gene had
was normal. She had moderate obesity (body mass index first been reported in individuals with a neurodevelopmental
30.9) without compulsive eating behavior. Her brain MRI was disorder characterized by initial normal or subnormal de-
normal. velopment and early neurologic regression with epilepsy prior
to the age of 7 in most patients.7,8 Neurologic features include
Genetic Analyses severe tetraparesis and cerebellar syndrome with ataxia, dysar-
In family SAL-394, previous screening of the repeat expan- thria, and nystagmus, associated with inconstant cerebellar
sions in ATXN1, 2, and 3; CACNA1A; ATXN7; ATXN10; atrophy. Some patients did not show signs of neurologic re-
PPP2R2B; TBP; and ATN1 was negative as was a search for gression and instead presented with global mild to moderate
point mutations by targeted sequencing in most known ataxia- developmental delay. In these pediatric patients, the reported
related genes. Genome sequencing in 2 individuals later IRF2BPL variants arose de novo. In parallel, IRF2BPL variants
identified variant c.349C>T, p(Gln117*) in a polymorphic have been identified in patients with adult-onset dystonia in 2
CAG repeat region of the IRF2BPL gene absent from the out of 8 NDD patients with cerebellar ataxia and pyramidal signs
GnomAD database (v2.1.1 and v3.1.2) and segregating in all as well as dystonic features.9-11 One of the first descriptions of a
affected (either with DI and/or ataxia) individuals. link between neurodegenerative and neurodevelopmental dis-
orders was for Rett syndrome, characterized by motor degra-
In family II, exome sequencing identified the IRF2BPL variant dation and attributed to a disturbance of BDNF transport
c.938C>A, p.(Ser313*), absent from the GnomAD database throughout the corticostriatal pathway.12
(v2.1.1 and v3.1.2), and inherited from the healthy mother
who had no evidence of mosaicism (variant allelic frequency = Similarly, pathogenic variants in the WDR45 or RAB39B
0.45). The mother reported no developmental delay and went genes have been reported to induce learning difficulties and
to a normal school until the age of 14 years but did not work. ID with progression towards a neurodegenerative parkin-
She was autonomous as an adult. She died of a domestic sonism phenotype in adulthood.3,4
accident at the age of 45 years. No neurologic abnormalities
were reported. A segregation study in family II showed that In this study, patient IIA inherited the IRF2BPL variant from
the healthy sister of the patient did not carry the IRF2BPL her asymptomatic mother who died in an accident at age 45,
variant. suggesting incomplete penetrance. At the last evaluation after
identification of the IRF2BPL variant, patient IIA presented
no neurologic signs, too young at the time.
Discussion
We are reporting on 2 families carrying pathogenic IRF2BPL Neurodegenerative aspects in NDD are very likely under-
truncating variants, one large family with 27 sampled individuals estimated as young adult patients with NDD are often lost to
including 16 patients and a second family with a mother-child specialized follow-up. This study illustrates the importance
dyad. Of interest, affected members in the large family exhibited of regular evaluation of patients with NDD during adulthood
2 phenotypes, not believed to be related at first: intellectual to better delineate the natural history of the disease. Un-
disability and late-onset cerebellar ataxia with pyramidal signs. derstanding the relationships between nervous system de-
This prevented us from identifying a common cause through velopment and degeneration is essential for early detection
genome sequencing because the phenotypes were treated as 2 and prevention of neurodegenerative disease. Looking at pre-
different traits. The large family was seen 28 years ago, and thus, manifest phases of dominantly inherited diseases has allowed
we contacted members of the younger generation to gather us to show neurodevelopment changes in Huntington dis-
information about their outcomes. Three individuals seen in ease.13 Human fetal tissues which carried an expansion in the
their twenties developed cerebellar signs in their thirties and HTT gene responsible for late-onset disease showed abnor-
fifties. This was in addition to mild intellectual difficulties pre- malities in the developing cortex, including abnormal locali-
sent since beginning school linked to the full phenotype of zation of mutant huntingtin and junction complex proteins,
IRF2BPL in adults. This shows that variants in IRF2BPL are defects in polarity and differentiation of neural precursors,
responsible for both a neurodevelopmental and a neurodegen- abnormal ciliogenesis, and changes in mitosis and cell cycle
erative aspect of the disease. The IRF2BPL gene encodes for an progression. These abnormalities disrupt the “division-differ-
E3 ubiquitin protein ligase involved in the proteasome-mediated entiation” balance of progenitors. This work not only provides

the first direct evidence from human fetuses that brain de-
velopment is impaired in a neurodegenerative disease with Appendix (continued)
delayed onset but also clearly demonstrate that molecular Name Location Contribution
changes even in adult-onset neurologic diseases occur very
early on. These early changes may prime specific neuronal Perrine Genetic Department, Assistance Drafting a significant
Charles, Publique-Hôpitaux de Paris (AP- portion of the
populations for neurodegeneration occurring much later. MD, PhD HP) Pitié-Salpêtrière; Reference manuscript or figures
Center for Rare Diseases «
Intellectual disabilites of rare
Acknowledgment causes » « Déficiences
The authors are deeply indebted to all family members for Intellectuelles de Causes Rares »,
Pitié-Salpêtrière Hospital, Paris,
their patience and participation. Special thanks to Bertrand France
Fontaine, Fausto Viader, and Soraya Medjbeur for referral and
Delphine Genetic Department, Assistance Acquisition and analysis
initial neurologic examination. Heron, MD Publique-Hôpitaux de Paris (AP- of data
HP) Pitié-Salpêtrière; Reference
Center for Rare Diseases «
Study Funding Intellectual disabilites of rare
The authors report no targeted funding. causes » « Déficiences
Intellectuelles de Causes Rares »,
Pitié-Salpêtrière Hospital, Paris,
Disclosure France
The authors report no relevant disclosures. Go to Neurology. Alexis Brice, Sorbonne Université, Paris Brain Conception and design of
org/NG for full disclosures. MD, PhD Institute (ICM Institut du the study; acquisition and
Cerveau), INSERM, CNRS, analysis of data; drafting
Assistance Publique-Hôpitaux de a significant portion of
Publication History Paris (AP-HP), France the manuscript or figures
Received by Neurology: Genetics February 1, 2023. Accepted in final form
Alexandra Genetic Department, Assistance Conception and design of
August 4, 2023. Submitted and externally peer reviewed. The handling Durr, MD, Publique-Hôpitaux de Paris (AP- the study; acquisition and
editor was Deputy Editor Massimo Pandolfo, MD, FAAN. PhD HP) Pitié-Salpêtrière; Sorbonne analysis of data; drafting
Université, Paris Brain Institute a significant portion of
(ICM Institut du Cerveau), the manuscript or figures
INSERM, CNRS, Assistance
Publique-Hôpitaux de Paris (AP-
Appendix Authors HP), France
Name Location Contribution
Solveig Genetic Department, Assistance Conception and design of

Heide, MD Publique-Hôpitaux de Paris (AP- the study; acquisition and References
HP) Pitié-Salpêtrière; Reference analysis of data; drafting 1. Hoischen A, Krumm N, Eichler EE. Prioritization of neurodevelopmental disease
Center for Rare Diseases « a significant portion of genes by discovery of new mutations. Nat Neurosci. 2014;17(6):764-772. doi:
Intellectual disabilites of rare the manuscript or figures 10.1038/nn.3703
causes » « Déficiences 2. Samocha KE, Robinson EB, Sanders SJ, et al. A framework for the interpretation of de
Intellectuelles de Causes Rares », novo mutation in human disease. Nat Genet. 2014;46(9):944-950. doi:10.1038/ng.3050
Pitié-Salpêtrière Hospital, Paris, 3. Wilson GR, Sim JCH, McLean C, et al. Mutations in RAB39B cause X-linked in-
France tellectual disability and early-onset Parkinson disease with α-synuclein pathology. Am
J Hum Genet. 2014;95(6):729-735. doi:10.1016/j.ajhg.2014.10.015
Claire- Sorbonne Université, Paris Brain Conception and design of 4. Carvill GL, Liu A, Mandelstam S, et al. Severe infantile onset developmental and
Sophie Institute (ICM Institut du the study; acquisition and epileptic encephalopathy caused by mutations in autophagy gene WDR45. Epilepsia.
Davoine, BS Cerveau), INSERM, CNRS, analysis of data; drafting 2018;59(1):e5-e13. doi:10.1111/epi.13957
Assistance Publique-Hôpitaux de a significant portion of 5. Marcogliese PC, Dutta D, Ray SS, et al. Loss of IRF2BPL impairs neuronal maintenance
Paris (AP-HP), France the manuscript or figures through excess Wnt signaling. Sci Adv. 2022;8(3):eabl5613. doi:10.1126/sciadv.abl5613
6. Higashimori A, Dong Y, Zhang Y, et al. Forkhead box F2 suppresses gastric cancer
Paulina Sorbonne Université, Paris Brain Acquisition and analysis through a novel FOXF2-IRF2BPL- β -Catenin signaling axis. Cancer Res. 2018;78(7):
Cunha, MD Institute (ICM Institut du of data 1643-1656. doi:10.1158/0008-5472.can-17-2403
Cerveau), INSERM, CNRS, 7. Tran Mau-Them F, Guibaud L, Duplomb L, et al. De novo truncating variants in the
Assistance Publique-Hôpitaux de intronless IRF2BPL are responsible for developmental epileptic encephalopathy.
Paris (AP-HP), France Genet Med. 2019;21(4):1008-1014. doi:10.1038/s41436-018-0143-0
8. Marcogliese PC, Shashi V, Spillmann RC, et al. IRF2BPL is associated with neurological
Clarisse Department of Neurology, Acquisition and analysis phenotypes. Am J Hum Genet. 2018;103(3):456-460. doi:10.1016/j.ajhg.2018.08.010
Scherer- University Hospital d’Angers, of data 9. Antonelli F, Grieco G, Cavallieri F, Casella A, Valente EM. Adult onset familiar dystonia-
Gagou, MD France plus syndrome: a novel presentation of IRF2BPL-associated neurodegeneration. Par-
kinsonism Relat Disord. 2022;94:22-24. doi:10.1016/j.parkreldis.2021.10.033
Boris Keren, Genetic Department, Assistance Acquisition and analysis 10. Ganos C, Zittel S, Hidding U, Funke C, Biskup S, Bhatia KP. IRF2BPL mutations
MD, PhD Publique-Hôpitaux de Paris (AP- of data cause autosomal dominant dystonia with anarthria, slow saccades and seizures. Par-
HP) Pitié-Salpêtrière, France kinsonism Relat Disord. 2019;68:57-59. doi:10.1016/j.parkreldis.2019.09.020
11. Prilop L, Buchert R, Woerz S, Gerloff C, Haack TB, Zittel S. IRF2BPL mutation
Giovanni Sorbonne Université, Paris Brain Drafting a significant causes nigrostriatal degeneration presenting with dystonia, spasticity and keratoconus.
Stevanin, Institute (ICM Institut du portion of the Parkinsonism Relat Disord. 2020;79:141-143. doi:10.1016/j.parkreldis.2020.03.030
PhD Cerveau), INSERM, CNRS, manuscript or figures 12. Pejhan S, Rastegar M. Role of DNA Methyl-CpG-binding protein MeCP2 in Rett
Assistance Publique-Hôpitaux de syndrome pathobiology and mechanism of disease. Biomolecules. 2021;11(1):75. doi:
Paris (AP-HP); INCIA, EPHE, 10.3390/biom11010075
Université de Bordeaux, France 13. Barnat M, Capizzi M, Aparicio E, et al. Huntington’s disease alters human neuro-
development. Science. 2020;369(6505):787-793. doi:10.1126/science.aax3338

mTOR Pathway Somatic Pathogenic Variants in Focal

Malformations of Cortical Development
Novel Variants, Topographic Mapping, and Clinical Outcomes
Eric Krochmalnek, MSc,* Andrea Accogli, MD,* Judith St-Onge, DEC, Nassima Addour-Boudrahem, PhD, Correspondence
Dr. Srour
Gyan Prakash, MSc, Sung-Hoon Kim, PhD, Tristan Brunette-Clement, MD, Ghadd Alhajaj, MD,
myriam.srour@mcgill.ca
Lina Mougharbel, PhD, Elena Bruneau, BSc, Kenneth A. Myers, MD, PhD, Francois Dubeau, MD,
Jason Karamchandani, MD, Jean-Pierre Farmer, MDCM, Jeffrey Atkinson, MD, Jeffrey Hall, MD,
Chantal Chantal Poulin, MD, Bernard Rosenblatt, MDCM, Joel Lafond-Lapalme, MSc, Alexander Weil, MD,
Catherine Fallet-Bianco, MD, Steffen Albrecht, MD,† Nahum Sonenberg, PhD, Jean-Baptiste Riviere, PhD,
Roy W. Dudley, MD, MSc, PhD, and Myriam Srour, MDCM, PhD
Abstract
Somatic and germline pathogenic variants in genes of the mammalian target of rapamycin
(mTOR) signaling pathway are a common mechanism underlying a subset of focal malfor-
mations of cortical development (FMCDs) referred to as mTORopathies, which include focal
cortical dysplasia (FCD) type II, subtypes of polymicrogyria, and hemimegalencephaly. Our
objective is to screen resected FMCD specimens with mTORopathy features on histology for
causal somatic variants in mTOR pathway genes, describe novel pathogenic variants, and
examine the variant distribution in relation to neuroimaging, histopathologic classification, and
clinical outcomes.
Methods
We performed ultra-deep sequencing using a custom HaloPlexHS Target Enrichment kit in
DNA from 21 resected fresh-frozen histologically confirmed FCD type II, tuberous sclerosis
complex, or hemimegalencephaly specimens. We mapped the variant alternative allele fre-
quency (AAF) across the resected brain using targeted ultra-deep sequencing in multiple
formalin-fixed paraffin-embedded tissue blocks. We also functionally validated 2 candidate
somatic MTOR variants and performed targeted RNA sequencing to validate a splicing defect
associated with a novel DEPDC5 variant.
Results
We identified causal mTOR pathway gene variants in 66.7% (14/21) of patients, of which 13
were somatic with AAF ranging between 0.6% and 12.0%. Moreover, the AAF did not predict
balloon cell presence. Favorable seizure outcomes were associated with genetically clear re-
section borders. Individuals in whom a causal somatic variant was undetected had excellent
postsurgical outcomes. In addition, we demonstrate pathogenicity of the novel c.4373_
4375dupATG and candidate c.7499T>A MTOR variants in vitro. We also identified a novel
germline aberrant splice site variant in DEPDC5 (c.2802-1G>C).
*These authors contributed equally to this work as cofirst authors.

†Retired since 2021.
From the Research Institute of the McGill University Health Centre (E.K., J.S.-O., N.A.-B., L.M., E.B., K.A.M., J.L.-L., J.-B.R., R.W.D., M.S.); Integrated Program in Neuroscience (E.K.), McGill
University; Department of Specialized Medicine (A.A.), McGill University Health Centre; Department of Human Genetics (A.A., J.-B.R.), Faculty of Medicine; Goodman Cancer Centre
(G.P., S.-H.K., N.S.), Department of Biochemistry, McGill University; Department of Pediatric Neurosurgery (T.B.-C., A.W.), Centre Hospitalier Universitaire Sainte-Justine, University of
Montreal; Division of Pediatric Neurology (G.A., K.A.M., C.C.P., M.S.), Department of Pediatrics, McGill University, Montreal, Quebec, Canada; Department of Pediatrics (G.A.), Unaizah
College of Medicine and Medical Sciences, Qassim University, Saudi Arabia; Department of Neurology and Neurosurgery (K.A.M., F.D., J.H., C.C.P., M.S.), McGill University Health
Centre; Department of Pathology (J.K., S.A.), McGill University; Division of Neurosurgery (J.-P.F., J.A., R.W.D.), Department of Pediatric Surgery, McGill University Health Center; McGill
University (B.R.); Department of Pathology (C.F.-B.), Centre Hospitalier Universitaire Sainte-Justine, University of Montreal, Quebec, Canada.
The Article Processing Charge was funded by authors.

Glossary
AAF = alternate allele frequency; CNV = copy number variant; EC = Engel classification; FCD = focal cortical dysplasia;
FMCDs = focal malformations of cortical dysplasia; FFPE = formalin-fixed paraffin-embedded; HMEG = hemimegalencephaly;
mTOR = mammalian target of rapamycin; PMG = polymicrogyria; TSC = tuberous sclerosis complex.
Discussion
The AAF of somatic pathogenic variants correlated with the topographic distribution, histopathology, and postsurgical out-
comes. Moreover, cortical regions with absent histologic FCD features had negligible or undetectable pathogenic variant loads.
By contrast, specimens with frank histologic abnormalities had detectable pathogenic variant loads, which raises important
questions as to whether there is a tolerable variant threshold and whether surgical margins should be clean, as performed in
tumor resections. In addition, we describe 2 novel pathogenic variants, expanding the mTORopathy genetic spectrum. Although
most pathogenic somatic variants are located at mutation hotspots, screening the full-coding gene sequence remains necessary in
a subset of patients.
Introduction gene.27-31 To date, the underlying cause of FMCDs is found in

15.6–63% of surgical samples, at alternate allele frequencies
Focal malformations of cortical development (FMCDs) (AAFs) ranging from 0.14 to 33%.31,32
comprise a spectrum of developmental disorders ranging
from focal cortical dysplasia (FCD) to hemimegalencephaly. This study aims to examine the sequence variant load of
The common denominator of these conditions is a disruption mTOR pathway pathogenic variants and their topographic
of the normal cytoarchitecture of the cerebral cortex, fre- distribution in relation to neuroimaging, histopathologic
quently resulting in medication-resistant epilepsy that usually classification, and clinical outcomes; assess the diagnostic
requires surgical treatment. FCDs are classified into different yield of screening mTOR pathway genes in FMCD specimens
neuropathologic subtypes (type Ia, Ib, Ic, IIa, IIb, IIIa, IIIb, showing histologic features of mTORopathy; and characterize
IIIc, and IIId) based on the severity of cytoarchitectural dis- previously unreported mTOR pathway variants.
ruption.1 FCD type II is by far the most common in patients
who undergo epilepsy surgery and is characterized on his-
tology by loss of cortical lamination and large dysmorphic
neurons without (type IIa) or with (type IIb) balloon cells.2
Methods
Patient Cohort and Specimens
Recent studies have highlighted the role of hyperactivation of As part of an ongoing epilepsy surgery biobanking program,
the mechanistic target of rapamycin (mTOR) pathway in a we collect fresh-frozen and formalin-fixed and paraffin-
subset of FMCDs, which includes FCD type II, subtypes of embedded (FFPE) brain specimens and blood and/or saliva
polymicrogyria, and hemimegalencephaly.3-7 These disorders from patients undergoing epilepsy surgery at the Montreal
are now considered part of the same disease spectrum, termed Children’s Hospital, Montreal Neurologic Institute, and CHU
“mTORopathies”, and share indistinguishable histopatho- Sainte-Justine Hospital. We collect between 1 to 7 fresh brain
logic features, namely disrupted cytoarchitecture and dys- specimens per patient undergoing epilepsy surgery, which are
morphic neurons with or without balloon cells. The mTOR snap-frozen on dry ice. For each of these specimens, the ad-
pathway has an important role in cell growth, maturation, jacent tissue is FFPE and analyzed by the neuropathology
proliferation, and energy metabolism.8 Both somatic gain-of- department. Additional FFPE blocks are also available for
function variants in activator genes (such as MTOR,9-12 each patient. Neuropathologic diagnoses are performed
AKT313, RHEB14, and PIK3CA15) and loss-of-function according to the ILAE guidelines.33
germline variants in negative regulator genes (such as TSC1,
TSC2,16 DEPDC5,15,17-20 PTEN,21,22 NPRL223, and All patients who underwent epilepsy surgery between January
NPRL323,24) lead to upregulation of mTOR signaling, 2016 and June 2019 in whom histopathology of the resected
resulting in cellular overgrowth and abnormal migration.25 lesion confirmed an mTORopathy based on cortical dyslami-
Although pathogenic variants in a single allele of activator nation with dysmorphic neurons with or without balloon cells
genes are sufficient to hyperactivate mTOR signaling, a two- were included in this study. Presurgical evaluation and surgical
hit allelic variant mechanism has been suggested for at least procedures were performed as previously reported.34 All pa-
some negative regulator genes, such as TSC1/216,26 and tients underwent a 3T MRI. Clinical information, including age
DEPDC5.19,26,27 More recent studies also suggest that there at seizure onset, seizure localization, developmental milestones,
may be a synergistic second hit in another mTOR pathway and postsurgical outcomes, was collected from medical records.

Standard Protocol Approvals, Registrations, of DNA extracted from multiple FFPE brain tissue blocks
and Patient Consents (including those from previous surgeries). The region span-
This multicentric study had ethical approval from the research ning the variant was amplified using custom intronic primers.
ethics committee of the McGill University Health Center (13- Libraries were prepared using Nextera XT DNA Sample
244-PED). All participants or their parents gave written in- Preparation kit (Illumina) and sequenced on a MiSeq plat-
formed consent. form using paired end 150-bp reads. The variant was present
in a specimen if it was found in ≥0.5% of reads.
Data Availability
All generated data are included in this article and associated Functional Validation of Candidate mTOR
Supplementary material. The raw data generated and/or an- Pathway Variants
alyzed during the study are available from the corresponding To assess whether variants p.Asp1458dup and p.Ile2500Asn
author on request. resulted in upregulation of the mTOR pathway, an in vitro
transfection assay was used to probe for downstream
Screening of mTOR Pathway Genes hyperphosphorylation of P70-S6K1, a well-described marker
For each patient, genomic DNA was extracted from fresh- for mTOR pathway hyperactivation.8 Candidate MTOR
frozen (Qiagen, QIAamp Fast DNA Tissue Kit), FFPE brain variants were cloned separately into a pcDNA3-Flag MTOR
sections (Qiagen QIAamp DNA FFPE Tissue Kit) and pe- wild-type plasmids obtained from Addgene (Plasmid #26603),
ripheral blood or saliva (Qiagen, Puregene and DNA Geno- using QuikChange Lightning site-directed mutagenesis kit
tek, PrepIt) using standard methods. We designed a custom (Agilent). Mutant constructs were transiently cotransfected
panel of 13 genes (AKT1, AKT2, AKT3, CCND2, DEPDC5, with P70-S6K1 in HEK293T cells to probe for hyper-
MTOR, NPRL2, NPRL3, PIK3CA, PIK3R2, PTEN, TSC1, and phosphorylation of P70-S6K1 at threonine 389. Four in-
TSC2) belonging to the mTOR pathway using a HaloPlexHS dependent western blot repeats were performed. ImageJ
Target Enrichment kit (Agilent Technologies). This capture analysis (Version 1.53j) was conducted for quantification of
method allows the identification of low allele frequency var- the signal intensity of the bands. Statistical tests applied were
iants through the attachment of a unique barcode that permits one-way analysis of variance (ANOVA), followed by the
the tracking of individual DNA molecules, thus avoiding en- Dunnett post hoc test.
richment bias and minimizing sequencing errors to allow a
more accurate estimation of the level of mosaicism. Libraries Investigation of Aberrant Splicing and Search
were prepared according to the manufacturer’s protocol from for a Potential Second Hit in DEPDC5
50 ng of DNA extracted from the most histologically abnor- Total RNA was extracted from the fresh-frozen brain of pa-
mal fresh-frozen brain specimen of each patient, except in- tient 14 and blood from the patient and her mother, and
dividual 12. Because brain tissue was unavailable in this cDNA was synthesized according to standard protocols. To
individual, DNA was extracted from saliva and scraping of his verify the aberrant splicing of DEPDC5, cDNA was amplified
hypertrophic tongue. Deep sequencing (approximately 1,500 using a set of primers between exons 28 and exon 32 of
reads) was performed on a MiSeq platform using paired end DEPDC5 (eFigure 1, links.lww.com/NXG/A637). We also
150-bp reads. Sequenced reads were aligned to the human amplified full-length DEPDC5 mRNA by long-range PCR
genome reference sequence (hg19) using BWA. A coverage of to look for a second pathogenic variant. The statistical
99.2% of targeted bases was obtained by at least 100 reads in framework mixture-of-isoforms (MISO) software was used
all samples. Variant calling was performed using a publicly to estimate the expression and effect of alternatively spliced
available analytic pipeline (DnaSeq high Coverage Pipe- exons and isoforms.39 Finally, we searched for the presence
line).35 Candidate variants were retained if supported by at of a somatic copy number variant (CNV) involving
least 3 nonreference reads with a base quality threshold of 30 DEPDC5 in the brain. DNA derived from patient 14 (fresh
and an AAF of ≥0.5%. Heterozygous coding and splice site brain and blood) and her mother (blood) was genotyped
variants were retained if absent in gnomAD36 and in-house using high-resolution SNP array CytoScan HD (Affyme-
controls. Variants were prioritized if previously reported in trix, Santa Clara, CA) at the McGill Genome Innovation
the literature or reported in COSMIC.37 The pathogenicity of Center querying a total of 750,000 SNPs and 2.67 million
variants was determined based on the American College of CNV markers.
Medical Genetics and Genomics (ACMG) Classification
Guidelines.38 Genetic, Radiologic, and Clinical Correlations
We further delineated whether there was a relationship
Validation and Topographic Mapping of between pathogenic variant burden (AAF) and its topo-
Pathogenic Variants graphic distribution, neuroimaging, histopathologic classi-
Candidate variants were confirmed by targeted ultra-deep fication, clinical features, and postsurgical epilepsy
sequencing (approximately 100,000-fold) in DNA extracted outcomes. Chi-square, Fischer exact test, or t test was used
from fresh-frozen brain, blood, and/or saliva. In addition, we to compare outcomes and clinical features between patient
investigated the distribution of pathogenic variant loads groups. Two-sided tests with p-values below 0.05 were
across the resected brain by performing targeted sequencing considered statistically significant.

Results testing on blood was negative, and we did not identify a
second TSC2 variant). Diagnostic yield was similar whether
Patients’ Characteristics the FMCD was apparent on brain imaging (abnormal imaging
A total of 47 individuals undergoing epilepsy were recruited in 10/14 with positive genetic yield vs 4/7 with negative
between January 2016 and June 2019, and of them, 21 indi- genetic yield, p = 0.64).
viduals had histologic mTORopathy features on resected
brain (i.e., cortical dyslamination, dysmorphic neurons with Two Candidate MTOR Variants Result in mTOR
or without balloon cells) and were included in this study. A Pathway Upregulation
summary of the clinical, radiologic, and histologic features and We identified a novel somatic variant in MTOR (NM_
postsurgical seizure outcomes are described in Tables 1 and 2 004958.4), not previously associated with FCD: c.4373_
and eTable 2 (links.lww.com/NXG/A640). Our cohort 4375dupATG (p.Asp1458dup). In addition, variant
comprises 11 male and 10 female patients. The average age c.7499T>A (p.Ile2500Asn) was recently associated with FCD
was 19.5 years (range: 2–60 years, median: 14 years), at sei- type II in a single patient and had not undergone functional
zure onset was 4.8 years (range: 1 day-20 years, median: 3 studies.7,41 Both variants are absent in gnomAD and affect
years) and at surgery was 15.5 years (range: 9 months-58 highly conserved residues. The c.4373_4375dupATG results
years, median: 11 years). All patients had drug-resistant focal in an in-frame single amino acid duplication of 55 amino acids
epilepsy; 16 had FCD, 3 had HMEG (including one patient upstream from the FAT domain, in close proximity to several
with congenital lipomatous overgrowth, vascular malforma- other pathogenic variants (Figure 2). The p.Ile2500Asn
tions, epidermal nevi, and scoliosis/skeletal/spinal substitution is located 31 amino acids before the FATC do-
(CLOVES) syndrome and 1 with hypomelanosis of Ito), main (Figure 2). A different amino acid substitution at the
one had polymicrogyria, and one had tuberous sclerosis same position, p.Ile2500Phe, has been reported in 2 patients
complex (TSC). Fourteen individuals underwent a single with FCD type II31 and 2 patients with hemimegalencephaly.7,41,45
surgery, and 7 had multiple surgeries. Histology was consis- Both p.Ile2500Asn and p.Ile2500Phe are linked in COSMIC to
tent with FCD type IIa in 14 individuals and FCD type IIb in 7 several samples of different carcinoma types (COSV63869065,
individuals. FMCD was located in a single lobe in 14 of the 21 COSM1730782, eTable 1, links.lww.com/NXG/A639) and as-
(67%) individuals and involved the frontal lobe in 9 of 14 sociated with low-grade oncocytic renal tumors.46 The c.4373_
(64%), temporal lobe in 4 of 14 (29%), and cingulate cortex in 4375dupATG and c.7499T>A variants result in a 7-fold and 8-fold
1 of 14 (7%). 3T brain MRI was considered normal in 7 increase in P70-S6K1 phosphorylation (p ≤ 0.0001 and p ≤ 0.001),
individuals. respectively, in vitro, demonstrating that they cause mTOR path-
way upregulation (Figure 3).7
Diagnostic Yield of mTOR Gene Panel Screen
A total of 131 samples, including 37 fresh-frozen, 73 FFPE The c.7499T>A (p.Ile2500Asn) variant was detected at an
brain specimens, and 21 blood/saliva specimens, were col- AAF ranging between 1.2% and 7.6% (Figure 4A.c) in the
lected from the 21 patients with histologic features of brain specimen from patient 2, who developed focal seizures
mTORopathy. at age 16 months. Her initial 3T brain MRI was normal.
Seizures were initially controlled with levetiracetam; however,
We identified disease-causing variants in 14 patients, repre- the patient presented at 23 months of age in super-refractory
senting a diagnostic yield of 66.7% (Table 1, Figure 1). status epilepticus, unresponsive to standard antiseizure
Pathogenic variants were somatic in 13 patients and germline medications and anesthetics.24 An occipital brain biopsy
in one patient (Patient 14). Patient 14 carried a heterozygous revealed FCD type IIa. She subsequently underwent a right
germline splice site variant in DEPDC5 (NM_001242896.1: hemispherectomy, and histology revealed FCD type IIa fea-
c.2802-1G>C). Nine patients had somatic variants in MTOR, tures in all specimens examined. Similarly, the c.7499T>A
2 in PIK3CA, one in TSC2, and one in AKT3. We detected variant was identified in all tested brain specimens (eTable 1,
variants with an AAF as low as 0.6%. Notably, 43% (9/21) of links.lww.com/NXG/A639). She died at age 24 months after
specimens screened with the Haloplex panel had an AAF of 41 days of status epilepticus.
<5%. All variants were validated with targeted ultra-deep se-
quencing. In 11 patients, the somatic variants were previously The c.4373_4375dupATG (p.Asp1458dup) variant was
reported and shown to result in hyperactivation of the mTOR found at an AAF between 1.3% and 3.1% (Figure 4B.b) in
pathway and thus deemed pathogenic. We were only able to brain specimens from patient 3, a 31-year-old woman with
detect the somatic pathogenic variant in blood or saliva/ childhood-onset drug-resistant right frontal seizures. Her
buccal swabs in 2 individuals, both of whom had evidence of brain MRI was normal, and histology revealed FCD type IIa.
extracerebral involvement: an individual with hemi- She continues to have seizures despite 2 epilepsy surgeries.
megalencephaly as a manifestation of his CLOVES syndrome
(individual 12) and one individual with a clinical diagnosis of Detection of Novel Germline Splice Site Variant
TSC (individual 10 has hypomelanotic macules, facial in DEPDC5
angiofibromas, bilateral angiomyolipomas, subependymal We identified a novel germline variant affecting a canonical
nodules, and cortical/subcortical tubers; his previous clinical splice site (NM_001242896.1, c.2802-1G>C) of DEPDC5 in

Table 1 Clinical Characteristics and Genetic Findings in Individuals With Identified Pathogenic mTOR Pathway Variants
Variant allele frequency
Current age Age at sz onset, Total # Sz outcomea Nucleotide, protein Novel vs prev.
Patient# (age Dcd), sex last surgery Clinical diagnosis Development 3T brain MRI Histology surgeries (f/u yrs) Gene change reported Brainb Blood Saliva
12,40
Neurology.org/NG
1 43 y, M 9 y/41 y R hemisphere DRE, R Learning R HMEG HMEG/ 3 ECIV (2 y) MTOR c.4448G>A Prev. rep. 8.5% NA Not
HMEG and OVG disability FCD IIb p.Cys1483Tyr [3.0–11.6%] detected
2 (2 y-Dcd), F 16 m/2 y R hemisphere DRE Normal Normal* FCD IIa 2 ECIV (Dcd) MTOR c.7499T>A, Prev rep41 3.5% Not NA
p.Ile2500Asn [1.2–7.6%] detected
3 32 y, F 12 y/29 y R posterior cingulate Learning Normal FCD IIa 2 ECIV (2 y) MTOR c.4373_4375dupATG, Novel 3.1% Not Not
DRE disability p.Asp1458dup [1.3–3.1%] detected detected
4 21 y, M 3 y/15 y R frontal lobe DRE Normal R frontoparietal FCD, R subcortical FCD IIb 3 ECII (6 y) MTOR c.4447T>C Prev. rep.12 2.6% Not Not
parieto-occipital cysts p.Cys1483Arg [0.6–8.8%*] detected detected
5 8 y, M 18 m/5 y R lobe focal DRE Normal R frontal FCD FCD IIb 1 ECII (1.25 y) MTOR c.6644C>A, Prev. rep.42 0.9% Not NA
p.Ser2215Tyr [0.9–1.9%*] detected
6 8 y, F 2 m/5 y Focal left temporal Normal Normal FCD IIa 3 ECIV (1.8 y) MTOR c.6644C>T, Prev. rep.42 0.9% NA NA
DRE p.Ser2215Phe [2.1–5.4%*]
7 9 y, M 3 y, 3 y Left frontal DRE Normal L frontal FCD FCD IIb 1 ECI (6 y) MTOR c.5930C>A Prev. rep.12 0.8% Not NA
p.Thr1977Lys [0.8–3.5%*] detected
8 45 y, F 9 y, 42 y R frontal DRE Normal Normal FCD IIa 1 ECI (1 y) MTOR c.6644C>T, Prev. rep.42 0.7% [0.7*%] Not NA
p.Ser2215Phe detected
9 14 y, F 7 y, 10 y L parietotemporal Normal L supramarginal gyrus FCD FCD IIb 1 ECI (1.5 y) MTOR c.5930C>A, Prev. rep.12 0.6% [0.6*%] Not NA
DRE p.Thr1977Lys detected
10 5 y, M 3 m, 4 y TSC, left hemispheric GDD, ID Multiple R>L tubers and FCD IIb 2 ECIII (3.75 y) TSC2 c.2356-1G>A, p.? Novel 9.4% 6.3% 3.6%
DRE subependymal nodules [6.3–9.4%]
11 8 y, M 1 d, 9 m L hemispheric DRE, L GDD, ID L HMEG HMEG/ 1 ECIV (3.92 y) AKT3 c.49G>A, p.Glu17Lys Prev. rep.31 4.9% NA Not
HMEG FCD IIa [1.3–11.0%] detected
12 19 y, M 1 d, 1 y CLOVES syndrome, R Severe ID, ASD R HMEG HMEG/ 1 ECI (na) PIK3CA c.1624G>A, Prev. rep.32 NA Not 4.85–10.32%c
HMEG FCD IIa p.Glu542Lys detected
13 14 y, M 6 y, 7 y Right frontal DRE GDD R frontal PMG PMG/FCD x ECII (6 y) PIK3CA c.1624G>A, Prev. rep.32 12.0% NA NA
IIa p.Glu542Lys [5.1–22.7%]
14 19 y, F 1d, 13 y Right frontal lobe Severe ID R frontal FCD FCD IIa 3 ECIII (3 y) DEPDC5 c.2802-1G>C Novel 33.0% 50.8% NA
DREd [33.0–59.8%]
Abbreviations: ASD = autism spectrum disorder; CLOVES = congenital lipomatous overgrowth, vascular malformations, epidermal nevi, and scoliosis/skeletal/spinal syndrome; d = day; Dcd = deceased; DRE = drug-resistant
epilepsy; f/u = follow-up; F = female; FCD = focal cortical dysplasia; GDD = global developmental delay; HMEG = hemimegalencephaly; ID = intellectual disability; L = left; M = male; m = months; NA = not available; OVG =
overgrowth; PMG = polymicrogyria; Prev. rep. = previously reported; R = right; sz = seizure; TSC = tuberous sclerosis complex; y = years.
a
Seizure outcome according to Engel classification (Engel 1993).
b
Variant allele frequency obtained from DNA extracted from fresh-frozen resected brain tissue. The range of allele frequencies across all samples tested are provided in brackets. * indicates that the pathogenic variant is
undetectable in some specimens.
c
Buccal swab of hemihypertrophic tongue.
d
This patient also harbors a likely pathogenic heterozygous variant in EBF3 (c.431A>G, p.Gln144Arg) that likely underlies her severe ID and facial dysmorphism. The Engel Epilepsy Surgery Outcome Scale was used to classify
postsurgical outcomes (Engel Class I: freedom from disabling seizures; Class II: rare disabling seizures (almost seizure free); Class III: worthwhile seizure reduction; Class IV: no worthwhile improvement).43,44

5
Table 2 Clinical Characteristics of FCD Individuals With Negative mTOR Pathway Genetic Screening
Sz
Patient Current Age at sz onset, at Clinical diagnosis Total number outcome
# age, sex last surgery (duration of epilepsy) Development 3T brain MRI Histology of surgeries (f/u)
15 20 y, F 8 y, 14 Right frontal DRE (6 y) Normal R frontal FCD FCD type 1 ECI (6 y)

IIa
16 11 y, M 3 y, 7 y Left Fronto-central- Normal Normal FCD type 1 ECI (2 mo)

parietal DRE epilepsy (4 y) IIa
17 14 y, F 2 y, 11 y Left SMA DRE (9 y) Normal Normal FCD type 2 ECI (4 y)

IIa
18 14 y, F 10 y, 11 y Right SMA DRE (1 y) Normal Normal FCD type 1 ECI (3 y)

language IIa
disorder
19 11 y, F 4 y,4 y Left temporal DRE (4 m) Normal Left parieto-temporal FCD type 1 ECI (7 y)
FCD IIb
20 60 y, M 20 y, 58 y Left temporal DRE (38 y) Normal Left hippocampal FCD type 1 ECI (2 y)
sclerosis IIa
21 32 y, M 10 m Left temporal DRE (28 y) Normal Left hippocampal FLAIR FCD type 1 ECI (3 y)
signal abnormality IIa
Abbreviations: DRE = drug-resistant epilepsy; FCD = focal cortical dysplasia; ECI = Engel class I; f/u = follow-up; m = months; SMA = supplementary motor area;
sz = seizure; y = year.
patient 14, inherited from her asymptomatic mother. This c.2802-1G>C variant results in aberrant splicing and retention
variant has not been previously reported, is absent in control of intron 29 (eFigure 1A, links.lww.com/NXG/A637) pre-
databases (gnomAD), and is classified as pathogenic based on dicted to shift the reading frame. Moreover, the probabilistic
ACMG Guidelines. Targeted sequencing of DEPDC5 cDNA model of RNA-seq obtained with MISO and displayed with
derived from the patient’s blood and brain revealed that the Sashimi revealed the presence of extra read densities between
Figure 1 Pathogenic Variants in mTOR Pathway Genes Identified in Our FMCD Cohort
Representation of 14 variants detected with their corresponding patient number and location. Variants in MTOR, PIK3CA, and AKT3 are somatic gain-of-
function variants in positive regulators of the mTOR pathway. DEPDC5 and TSC2 are loss-of-function variants in negative regulators of the mTOR pathway.
FMCD = focal malformations of cortical development; mTOR = mammalian target of rapamycin.

Figure 2 Distribution of MTOR Pathogenic Variants Associated With FMCDs
Previously reported (in black) and novel (in red) pathogenic variants associated with FMCDs are indicated. Bolded substitutions were also found in our cohort.
The MTOR protein contains 20 tandem HEAT repeats that provide protein-protein interactions with the mTOR regulatory proteins Raptor and Rictor, the FAT
modulatory domain, the FKBP12-rapamycin binding domain (FRB), the Ser/Thr kinase domain, and the FATC modulatory domain. There is a clustering of
variants between the HEAT repeats and FAT domain, as well as within and close to the kinase domain. FMCDs = focal malformations of cortical development;
mTOR = mammalian target of rapamycin.
exon 29 and 30, demonstrating intron 29 retention (eFigure 1B). All solved FCDs in our cohort had somatic pathogenic
We searched for a somatic variant or CNV involving DEPDC5; MTOR variants. Pathogenic somatic variants were found in
however, a second hit was not identified after sequencing of full- AKT3 and PIK3CA in larger cerebral lesions, namely hem-
length DEPDC5 cDNA and whole-genome SNP array. imegalencephaly and polymicrogyria.
Variant Load, Topographic Distribution, In general, the load and topographic distribution of the somatic
Histology, and Clinical Outcomes pathogenic variants correlated with the size of the FMCD on
For the 14 individuals in whom we identified a causal somatic MRI and based on the distribution of histopathologic abnor-
mTOR pathway variant, we further studied a total of 103 brain malities: more extensive lesions were associated with higher
specimens, including 73 FFPE specimens, to assess the AAF maximal AAFs (eFigure 2, links.lww.com/NXG/A638 and
and distribution of the variants across multiple brain regions eTable 2, links.lww.com/NXG/A640). For example, patients
(average of 7.35 brain specimens per patient). A summary of with hemimegalencephaly (individuals 1, 11, and 12) and ex-
the topographic distribution of the variants for each patient is tensive polymicrogyria (individual 13) had the highest maximal
depicted in eFigure 2 (links.lww.com/NXG/A638). AAF (maximum AAF ranges 10.3%–22.7%). By contrast,
Figure 3 Functional Validation of Candidate MTOR Variants
(A) Western blots of flag-tagged mutant

MTOR plasmid constructs cotransfected
with HA-p70-S6K1 in HEK293T cells
to probe for hyperphosphorylation of
p70-S6K. Blotting for p-p70-S6 kinase at
threonine 389 reveals upregulation
of the mTOR pathway in both patient
variants Asp1458dup and Ile2500Asn.
Leu1460Pro and Ile2500Phe were used
as positive controls. HA and Flag tags in-
dicate equal cotransfection of p70-S6K1
and MTOR plasmids, respectively.
Β-actin is a cell lysate loading control. (B)
Quantification of western blot signal in-
tensity and one way ANOVA followed
by the Dunnett post hoc test reveals a
statistically significant increase in levels
of p70-S6K1 phosphorylation relative to
WT was observed for patient variants,
p.Asp1458dup and p.Ile2500Asn, as
well as a known upregulating mTOR
variants Leu1460Pro and Ile2500Phe.
Statistical values: * = p ≤ 0.05, *** =
p ≤ 0.001, **** = p ≤ 0.0001. mTOR =
mammalian target of rapamycin.

Figure 4 Examples of Mapping Somatic Variants and Their AAF Across Resected Brain
Preoperative (boxed) and postoperative brain MRIs of patients 2 (A), 3 (B), 4 (C), and 4 (D). Yellow lines indicate the outline of the resected brain. The AAF
frequency of the somatic pathogenic variants was obtained from targeted sequencing of DNA extracted from FFPE specimens corresponding to the indicated
regions. Preoperative MRIs were reported as normal in patients 2 (A) and 3 (B). Patient 4 had right parieto-occipital cystic lesions and an extensive FCD in the
right orbitofrontal lobe (dotted white circles, C), and patient 7 (D) had a right frontal bottom of sulcus FCD (dotted white circles, D). AAF = alternative allele
frequency; FCD = focal cortical dysplasia; FFPE = formalin-fixed paraffin-embedded.
patients with FCD had a lower maximum pathogenic variant our samples (average maximum AAF 8.4% in FCD2a vs 6.0%
load (maximum AAF range for FCD 0.6–8.88% and average in 2b, p = 0.5227). Patients with somatic MTOR pathogenic
maximum AAF 3.74% in FCD vs 15.1% in PMG/HMG, p = variants with similar AAF ranges could have either FCD type
0.0053); even within the FCD subgroup, those with histopatho- IIa or IIb on histology. Furthermore, when comparing the
logic extensive lesions (patients 2 and 4) had higher maximal AAF. AAF and histologic findings across multiple brain speci-
mens from the same patient, there was no relationship
In general, somatic pathogenic variants were detected in tissue between AAF and the presence of balloon cells. For ex-
specimens displaying histologic abnormalities (eTable 2, ample, in patient 1 with the hemimegalencephaly/MTOR
links.lww.com/NXG/A640). We always identified the so- variant, balloon cells were identified in only one of the 13
matic pathogenic variants in specimens that were frankly FFPE specimens with a variant load of 3.9%; all other
histologically abnormal and consistent with FCD type II. specimens showed the presence of dysmorphic neurons
Moreover, almost all histologically normal specimens showed without balloon cells, with variant loads ranging between
the absence of a causal variant. However, it is important to 3.3 and 11.6%. Similarly, in patient 4 with the MTOR
note that there were rare specimens considered histologically variant, FFPE specimens with balloon cells had an AAF at
normal in which we identified the presence of the somatic 1.1%–3.3%, and the specimen with the highest AAF at
pathogenic variant at low levels. For example, in patient 4, we 8.8% displayed no balloon cells.
detected the somatic pathogenic variant at an AAF of 0.6 and
1.2% in DNA extracted from FFPE blocks from the right Individuals with PIK3CA and AKT3 variants were more likely
occipital lobe that were considered normal; in this patient, the to have neonatal-onset seizures than the remainder of the
epicenter of the FCD and the epileptogenic zone was much cohort (2/3 vs 0/18, p = 0.0143). There was also a correlation
more anterior in the right frontal lobe where the histology was between neurodevelopmental outcome and causal gene: All
frankly abnormal, and the AAF was up to 8.8%. individuals with somatic MTOR variants had normal de-
velopment and intelligence, whereas those with AKT3 or
We did not find a significant correlation between maximum PIK3CA variants had global developmental delay and in-
variant load and histologic diagnosis of FCD type IIa vs IIb in tellectual disability (GDD/IDD in 0/9 vs 4/4, p = 0.0014).

Patients with good surgical outcomes tended to have lower delay/intellectual disability, whereas development was nor-
pathogenic variant loads and surgical margins with normal mal in all patients with MTOR variants. Note that poor
histology without the causal variant (Table 1). It is interesting neurodevelopmental outcome does not seem to be related
to note that all 7 individuals from our cohort in whom we were only to the lesion size because patient 1 with the hemi-
unable to identify the underlying genetic etiology were megalencephaly and MTOR variant had normal development
seizure-free postresection, with Engel classification I and intelligence, suggesting that poor cognitive outcome
(Table 2) (ECI-II in 7/13 with somatic variants identified vs is not only associated with the topographic extent of the
7/7 with no somatic variant identified, p = 0.0515). These cortical malformation. Our cohort also supports neonatal
patients also had histologically normal surgical margins. onset of seizures with PIK3CA or AKT3 compared with
MTOR.
Of interest all patients with no identified pathogenic variant

Discussion had good postsurgical epilepsy outcomes (Engel Class I, see
In this study, we genetically characterized a total of 131 Table 2). A possible explanation for this is that the AAF of the
specimens (including fresh-frozen and FFPE brain specimens, pathogenic variant was below our method’s detection threshold
blood, and saliva) from 21 patients with FMCD with histo- and that the DNA was extracted from sections that were not
logic features of mTORopathy. pathologic and did not contain mutant cells, implying that a smaller
FCD or one with a low pathogenic variant AAF is associated with
We demonstrate that systematic screening of abnormal ce- better postsurgical outcome. Other potential reasons for not
rebral specimens using an mTOR pathway gene panel has a identifying the causal pathogenic variant include the presence of a
high diagnostic yield of 66.7%. This yield is high compared pathogenic variant outside of the coding regions, deletions, struc-
with the range previously reported in the literature tural rearrangements, and genes absent from our panel.47
(15.6%–63%).31,32 A further breakdown of the diagnostic
yield shows that mTOR pathway variants underlie 86% (6/7) As illustrated in our study, the variable levels of pathogenic
of FCD type IIb patients but only 53% (8/15) of FCD type IIa variants across FCDs raise important questions as to whether
patients. Several factors may account for this high yield: We there is a tolerable variant level and whether the surgical
had access to fresh affected specimens of small size, which margins of the resected FCD should be clean, as performed in
gave us a high resolution of the variations within the patho- tumor resections. The relationship between variant load, ep-
logic tissue and allowed for the enhanced detection of somatic ileptogenic zone, and focus are still unknown, and larger-scale
variants at lower AAF; we chose the most abnormal speci- studies with combined intracranial recording, genetic, and
mens based on the histology of the adjacent tissue sections, histopathologic analysis and long-term seizure outcome are
and finally, we screened the full coding regions of the genes. needed to address this matter.
We confirmed many of the previously published observations, We describe 2 novel somatic pathogenic variants responsible
although our cohort included a modest number of patients. As for FMCDs and illustrate that, although most pathogenic
noted by Baldassari et al. (2019)30 and Pirozzi et al. variants are recurrent, they may be present outside of muta-
(2022),31,47 we found that the highest AAFs were usually, tion hotspots. Therefore, screening only for recurrent muta-
although not strictly, associated with more extensive cortical tions is insufficient to identify causal pathogenic variants in
lesions and that PIK3CA and AKT3 were associated with large patients with mTORopathies.
lesions such as hemimegalencephaly or polymicrogyria.
Similarly, AAF appeared to correlate with histologic findings: We performed functional validation of 2 variants in MTOR,
cortical regions with absent histologic FCD features had p.Ile2500Asn and p.Asp1458dup, in patients with FCD type
negligible or undetectable pathogenic variant loads, whereas IIa and demonstrated using an in vitro assay that these vari-
specimens with frank histologic abnormalities had detectable ants result in hyperphosphorylation of P70-S6K1, indicating
pathogenic variants. Our findings support the conclusions by they are pathogenic and cause mTOR pathway upregulation.
Lee et al. (2023) and Baldassari et al.that the density of the We also report a novel germline canonical splice site variant in
dysmorphic cells correlated with the AAF.31,48 Of note, we did DEPDC5 (c.2802-1G > C) and show that it results in aberrant
not observe any clear correlation between the histologic splicing leading to a frameshift. DEPDC5 encodes for DEP
subtype of FCD type II (i.e., IIa or IIb) and AAF because domain containing 5, a member of the GATOR1 complex
regions of similar variant load may demonstrate the presence (GAP activity toward Rags complex 1) and, along with
or absence of balloon cells; studies including a larger number NPRL2 and NPRL3, acts as a negative regulator of
of specimens will be required to confirm this observation. mTORC1.26 Variants in DEPDC5 are typically loss of func-
tion, with only a few recurrent variants reported. It has been
The findings from our cohort support the previously noted hypothesized that a second-hit mechanism may be required to
correlation between PIK3CA or AKT3 variants and poor generate FCDs, as previously observed in cancer48 and
neurodevelopmental outcome31 because all our patients with TSC.15,25 To date, this phenomenon has been demonstrated
PIK3CA or AKT3 variants have global developmental 6 times in GATOR1 genes for FCD.17,22,23,27,29-31 We did not

identify an additional pathogenic somatic variant in our surgery. In addition, our study demonstrates that screening
patient. fresh-frozen specimens using a custom HaloPlexHs mTOR
pathway gene panel results in a high diagnostic yield. We
Many of the somatic variants identified in our cohort were identify a novel somatic MTOR variant and provide in vitro
recurrent and involved substitution of the same amino acid. In evidence of pathogenicity, highlighting the importance of
MTOR, 2 patients had substitutions at p.Cys1483, 2 at screening the full coding regions in mTORopathy lesions. We
p.Thr1977, and 3 at p.Ser2215. The p.Cys1483 amino acid is also describe a novel germline DEPDC5 splice site mutation
located 30 amino acids upstream of the FAT catalytic domain, and show its impact on mRNA splicing. Routine molecular
where the N-terminal portion of the domain is required for testing and integration of genetic results into the classification
binding of the regulator proteins RAPTOR and RICTOR.49 and diagnosis of FMCDs will be key to enhancing the char-
The p.Thr1977 variant is located 38 amino acids upstream of acterization of FMCD cohorts, improving our understanding of
the rapamycin binding FRB domain and is thought to act as a underlying pathophysiology and allowing development of
gatekeeper at mTOR’s catalytic cleft.49 The p.Ser2215 variant novel targeted treatment options and personalized medicine.53
is located just outside the kα3 helix domain at the active site of
the mTOR kinase, and its substation has been previously Acknowledgment
demonstrated to upregulate mTOR through a gain-of-function The authors thank the patients and families for participating
sequence variant mechanism11,12,19,31,42 (Figure 2). Similarly, 2 in this study. E. Krochmalnek’s graduate training was
patients had the identical p.Glu542Lys substitution in PIK3CA, supported by the FRQS, McGill University’s Faculty of
which has previously been observed in hemimegalencephaly50 Medicine Neurology and Neurosurgery Excellence Award,
and is frequently observed in tumors (COSV55873227 in and MUHC Desjardins Studentship Award. Accogli’s fellow-
COSMIC). This substitution disrupts an inhibitory charge- ship training was supported by the Mel Hoppenheim Fund
charge interaction with the p85α regulatory subunit by affecting (Montreal Children’s Foundation) and the Rotary Founda-
the catalytic region of the PI3K helical domain.47 tion. The authors thank the Agilent team for their support in
SureCall analysis. M. Srour holds a Fonds de Recherche de
A few limitations of our study need to be mentioned. First, our Santé Quebec salary award.
study included a relatively small number of patients, which
limits our ability to find statistically significant differences Study Funding
between groups and may also increase our margin of error. This manuscript has been funded by Fondation Pierre Lavoie,
Nevertheless, our findings were in keeping with previous CIHR/Sick Kids New Investigator Operating Award (NI16-
studies. Second, potential disease-causing variants may be 028), Montreal Children’s Hospital Foundation (Husain
present below the minimal AAF detection threshold Family Endowment).
(i.e., <0.005) for HaloplexHS or lie in promotor regions or
introns not covered by our panel. Third, RHEB (MIM* Disclosure
601293)14 and RPS6 (MIM* 180460),28 which have recently The authors report no relevant disclosures. Go to Neurology.
been shown to be implicated in FCD pathogenesis, were not org/NG for full disclosures.
included in our gene panel. Fourth, we have investigated so-
matic CNV only in the individual harboring the DEPDC5 Publication History
variant but not in other patients.51 Finally, determination of Received by Neurology: Genetics April 21, 2023. Accepted in final form
the mosaic gradient of the somatic variants was performed on September 6, 2023. Submitted and externally peer reviewed. The
handling editor was Editor Stefan M. Pulst, MD, Dr med, FAAN.
DNA extracted from FFPE and not fresh-frozen tissue as we
did not have access to many fresh-frozen specimens per pa-
tient; although there is concern that fixing introduces DNA
artifacts and may affect variant calling, it has been shown that Appendix Authors
sequencing of somatic variants in FFPE tissue is highly con- Name Location Contribution
cordant with results from fresh tissue.52
Eric Research Institute of the Drafting first version of
Krochmalnek, McGill University Health manuscript/revision of the
In summary, through the study of 103 specimens, we provide a MSc Centre; Integrated manuscript for content,
Program in Neuroscience, major role in the
compelling demonstration of the mosaic pattern of the somatic McGill University, acquisition of data and
pathogenic variants in mTORopathies and show an association Montreal, Quebec, analysis or interpretation
Canada of data
between the level of mosaicism, histopathologic findings, and
clinical outcomes, concordant with previous studies. Cortical Andrea Accogli, Department of Specialized Drafting first version of
MD Medicine, Division of manuscript/revision of the
regions without histologic FCD features had negligible or un- Medical Genetics, McGill manuscript for content,
detectable pathogenic variant loads, whereas specimens with University Health Centre; major role in the
Department of Human acquisition of data and
frank histologic abnormalities had detectable pathogenic vari- Genetics, Faculty of analysis or interpretation
ants. Seizure outcomes were favorable when resection borders Medicine, McGill of data
were genetically clear, and individuals without an identified University, Montreal,
Quebec, Canada.
causal somatic variant had excellent postsurgical outcomes after

Appendix (continued) Appendix (continued)
Name Location Contribution Name Location Contribution
Judith St-Onge, Research Institute of the Revision of the Jean-Pierre Division of Neurosurgery, Revision of the
DEC McGill University Health manuscript, major role in Farmer, MDCM, Department of Pediatric manuscript, collection of
Centre, Montreal, Quebec, experimentation and FRCSC Surgery, McGill University specimens, contribution
Canada acquisition of data Health Center, Montreal, of patients and clinical
Quebec, Canada data
Nassima Addour- Research Institute of the Drafting/revision of the
Boudrahem, PhD McGill University Health manuscript/acquisition Jeffrey Atkinson, Division of Neurosurgery, Revision of the
Centre, Montreal, Quebec, and interpretation of data/ MD Department of Pediatric manuscript, collection of
Canada recruitment of patients/ Surgery, McGill University specimens, contribution
administrative support Health Center, Montreal, of patients and clinical
Quebec, Canada data
Gyan Prakash, Goodman Cancer Centre, Revision of the
MSc Department of manuscript, role in Jeffery Hall, MD Department of Neurology Revision of the
Biochemistry, McGill experimentation and FRCSC and Neurosurgery, McGill manuscript, collection of
University, Montreal, acquisition of data University Health Centre, specimens, contribution
Quebec, Canada Montreal, Quebec, of patients and clinical
Canada data
Sung-Hoon Kim, Goodman Cancer Centre, Revision of the
PhD Department of manuscript, role in Chantal Poulin, Division of Pediatric Revision of the
Biochemistry, McGill experimentation and MD Neurology, Department manuscript, contribution
University, Montreal, acquisition of data of Pediatrics, McGill of patients and clinical
Quebec, Canada University; Department data
of Neurology and
Tristan Department of Pediatric Revision of the Neurosurgery,
Brunette- Neurosurgery, Centre manuscript, acquisition of McGill University
Clement, MD Hospitalier Universitaire data Health Centre, Montreal,
Sainte-Justine, University Quebec, Canada
of Montreal, Montreal,
Quebec, Canada
Bernard Division of Pediatric Revision of the
Rosenblatt, Neurology, Department of manuscript, contribution
Ghadd Alhajaj, Division of Pediatric Revision of the MDCM Pediatrics, McGill of patients and clinical
MD Neurology, Department of manuscript, acquisition of University; Department of data
Pediatrics, McGill data Neurology and
University, Montreal, Neurosurgery, McGill
Quebec, Canada; University Health Centre,
Department of Pediatrics, Montreal, Quebec,
Unaizah College of Canada
Medicine and Medical
Sciences, Qassim
Joël Lafond Research Institute of the Revision of the
University, Qassim, Saudi
Lapalme, MSc McGill University Health manuscript, bioinformatic
Arabia
Centre, Montreal, Quebec, analysis of data
Canada
Lina Research Institute of the Revision of the
Mougharbel, McGill University Health manuscript, role in
PhD Centre, Montreal, Quebec, experimentation, Alexander G. Department of Pediatric Revision of the
Canada acquisition and analysis of Weil, MD Neurosurgery, Centre manuscript, collection of
data Hospitalier Universitaire specimens, contribution
Sainte-Justine, University of patients and clinical
of Montreal, Montreal, data
Elena Bruneau, Research Institute of the Revision of the
Quebec, Canada
BSc McGill University Health manuscript, role in
Centre, Montreal, Quebec, experimentation
Canada Catherine Fallet- Department of Revision of the
Bianco, MD Pathology, Centre manuscript, collection,
Kenneth A. Research Institute of the Revision of the Hospitalier Universitaire analysis and
Myers, MD, PhD, McGill University Health manuscript, contribution Sainte-Justine, University interpretation of data
CCSN Centre; Division of of patients and clinical of Montreal, Montreal,
Pediatric Neurology, data Quebec, Canada
Department of Pediatrics,
McGill University; Steffen Albrecht, Department of Revision of the
Department of Neurology MD Pathology, McGill manuscript, collection,
and Neurosurgery, McGill University, analysis and
University Health Centre, Montreal, interpretation of data
Montreal, Quebec, Quebec, Canada
Canada
Nahum Goodman Cancer Revision of the
François Department of Neurology Revision of the Sonenberg, PhD Centre, Department manuscript, analysis and
Dubeau, MD and Neurosurgery, McGill manuscript, contribution of Biochemistry, interpretation of data,
University Health Centre, of patients and clinical McGill University, supervision
Montreal, Quebec, data Montreal, Quebec,
Canada Canada
Jason Department of Pathology, Revision of the Jean-Baptiste Research Institute of the Revision of the
Karamchandani, McGill University, manuscript, contribution Riviere, PhD McGill University Health manuscript, bioinformatic
MD Montreal, Quebec, of clinical data Centre, Montreal, Quebec, analysis of data
Canada Canada
Continued

19. Mirzaa GM, Campbell CD, Solovieff N, et al. Association of MTOR mutations with
developmental brain disorders, including megalencephaly, focal cortical dysplasia, and pig-
Appendix (continued)
mentary mosaicism. JAMA Neurol. 2016;73(7):836-845. doi:10.1001/jamaneurol.2016.0363
20. Scheffer IE, Heron SE, Regan BM, et al. Mutations in mammalian target of rapamycin
regulator DEPDC5 cause focal epilepsy with brain malformations. Ann Neurol 2014;
75(5):782-787. doi:10.1002/ana.24126
Roy W. Dudley, Research Institute of the Revision of the 21. Child ND, Cascino GD. Mystery case: Cowden syndrome presenting with partial
MD, MSc, PhD McGill University Health manuscript, collection of epilepsy related to focal cortical dysplasia. Neurology. 2013;81(13):e98-e99. doi:
Centre; Division of specimens, analysis and 10.1212/WNL.0b013e3182a55ef0
Neurosurgery, interpretation of data 22. Adachi T, Takigawa H, Nomura T, Watanabe Y, Kowa H. Cowden syndrome with a
Department of Pediatric Novel PTEN mutation presenting with partial epilepsy related to focal cortical dys-
Surgery, McGill University plasia. Intern Med. 2018;57(1):97-99. doi:10.2169/internalmedicine.9052-17
Health Center, Montreal, 23. Baldassari S, Licchetta L, Tinuper P, Bisulli F, Pippucci T. GATOR1 complex: the
Quebec, Canada common genetic actor in focal epilepsies. J Med Genet. 2016;53(8):503-510. doi:
10.1136/jmedgenet-2016-103883
Myriam Srour, Research Institute of the Conception and design of 24. Sim JC, Scerri T, Fanjul-Fernandez M, et al. Familial cortical dysplasia caused by
MDCM, PhD McGill University Health the study, supervision, mutation in the mammalian target of rapamycin regulator NPRL3. Ann Neurol. 2016;
Centre; Division of drafting and revision of 79(1):132-137. doi:10.1002/ana.24502
Pediatric Neurology, the manuscript; analysis 25. Iffland PH 2nd, Crino PB. Focal cortical dysplasia: gene mutations, cell signaling, and
Department of Pediatrics, and interpretation of data therapeutic implications. Annu Rev Pathol. 2017;12:547-571. doi:10.1146/annurev-
McGill University; pathol-052016-100138
Department of Neurology 26. Qin W, Chan JA, Vinters HV, et al. Analysis of TSC cortical tubers by deep sequencing
and Neurosurgery, McGill of TSC1, TSC2 and KRAS demonstrates that small second-hit mutations in these
University Health Centre, genes are rare events. Brain Pathol. 2010;20(6):1096-1105. doi:10.1111/j.1750-
Montreal, Quebec, 3639.2010.00416.x
Canada 27. Baulac S, Ishida S, Marsan E, et al. Familial focal epilepsy with focal cortical dysplasia due to
DEPDC5 mutations. Ann Neurol. 2015;77(4):675-683. doi:10.1002/ana.24368
28. Pelorosso C, Watrin F, Conti V, et al. Somatic double-hit in MTOR and RPS6 in
hemimegalencephaly with intractable epilepsy. Hum Mol Genet. 2019;28(22):
3755-3765. doi:10.1093/hmg/ddz194
References 29. Sim NS, Ko A, Kim WK, et al. Precise detection of low-level somatic mutation in
1. Blumcke I, Spreafico R. An international consensus classification for focal cortical resected epilepsy brain tissue. Acta Neuropathol. 2019;138(6):901-912. doi:10.1007/
dysplasias. Lancet Neurol. 2011;10(1):26-27. doi:10.1016/S1474-4422(10)70225-8 s00401-019-02052-6
2. Blumcke I, Spreafico R, Haaker G, et al. Histopathological findings in brain tissue 30. Ribierre T, Deleuze C, Bacq A, et al. Second-hit mosaic mutation in mTORC1
obtained during epilepsy surgery. N Engl J Med. 2017;377(17):1648-1656. doi: repressor DEPDC5 causes focal cortical dysplasia-associated epilepsy. J Clin Invest.
10.1056/NEJMoa1703784 2018;128(6):2452-2458. doi:10.1172/JCI99384
3. Baybis M, Yu J, Lee A, et al. mTOR cascade activation distinguishes tubers from focal 31. Baldassari S, Ribierre T, Marsan E, et al. Dissecting the genetic basis of focal cortical
cortical dysplasia. Ann Neurol. 2004;56(4):478-487. doi:10.1002/ana.20211 dysplasia: a large cohort study. Acta Neuropathol. 2019;138(6):885-900. doi:10.1007/
4. Crino PB. mTOR: a pathogenic signaling pathway in developmental brain malfor- s00401-019-02061-5
mations. Trends Mol Med. 2011;17(12):734-742. doi:10.1016/j.molmed.2011.07.008 32. Marsan E, Baulac S. Review: mechanistic target of rapamycin (mTOR) pathway, focal
5. Lee WS, Baldassari S, Stephenson SEM, Lockhart PJ, Baulac S, Leventer RJ. Cortical cortical dysplasia and epilepsy. Neuropathol Appl Neurobiol. 2018;44(1):6-17. doi:
dysplasia and the mTOR pathway: how the study of human brain tissue has led to 10.1111/nan.12463
insights into epileptogenesis. Int J Mol Sci. 2022;23(3):1344. doi:10.3390/ 33. Blumcke I, Thom M, Aronica E, et al. The clinicopathologic spectrum of focal cortical
ijms23031344 dysplasias: a consensus classification proposed by an ad hoc Task Force of the ILAE
6. Crino PB. Molecular pathogenesis of focal cortical dysplasia and hemi- Diagnostic Methods Commission. Epilepsia. 2011;52(1):158-174. doi:10.1111/
megalencephaly. J Child Neurol. 2005;20(4):330-336. doi:10.1177/ j.1528-1167.2010.02777.x
08830738050200041101 34. Schur S, Moreau JT, Khoo HM, et al. New interinstitutional, multimodal presurgical
7. Gerasimenko A, Baldassari S, Baulac S. mTOR pathway: insights into an established evaluation protocol associated with improved seizure freedom for poorly defined
pathway for brain mosaicism in epilepsy. Neurobiol Dis. 2023;182:106144. doi: cases of focal epilepsy in children. J Neurosurg Pediatr. 2022;29(1):74-82. doi:
10.1016/j.nbd.2023.106144 10.3171/2021.6.PEDS218
8. Saxton RA, Sabatini DM. mTOR signaling in growth, metabolism, and disease. Cell. 35. Assessed July 20, 2023. bitbucket.org/mugqic/mugqic_pipelines
2017;168(6):960-976. doi:10.1016/j.cell.2017.02.004 36. Assessed July 20, 2023. gnomad.broadinstitute.org/
9. Lee JH, Huynh M, Silhavy JL, et al. De novo somatic mutations in components of the 37. Assessed July 20, 2023. cancer.sanger.ac.uk/cosmic
PI3K-AKT3-mTOR pathway cause hemimegalencephaly. Nat Genet. 2012;44(8): 38. Richards S, Aziz N, Bale S, et al. Standards and guidelines for the interpretation of
941-945. doi:10.1038/ng.2329 sequence variants: a joint consensus recommendation of the American College of
10. Leventer RJ, Scerri T, Marsh AP, et al. Hemispheric cortical dysplasia secondary to a Medical Genetics and Genomics and the Association for Molecular Pathology. Genet
mosaic somatic mutation in MTOR. Neurology. 2015;84(20):2029-2032. doi: Med 2015;17(5):405-424. doi:10.1038/gim.2015.30
10.1212/WNL.0000000000001594 39. Katz Y, Wang ET, Airoldi EM, Burge CB. Analysis and design of RNA sequencing
11. Nakashima M, Saitsu H, Takei N, et al. Somatic mutations in the MTOR gene cause experiments for identifying isoform regulation. Nat Methods. 2010;7(12):1009-1015.
focal cortical dysplasia type IIb. Ann Neurol. 2015;78(3):375-386. doi:10.1002/ doi:10.1038/nmeth.1528
ana.24444 40. Kim JK, Cho J, Kim SH, et al. Brain somatic mutations in MTOR reveal translational
12. Lim JS, Kim WI, Kang HC, et al. Brain somatic mutations in MTOR cause focal dysregulations underlying intractable focal epilepsy. J Clin Invest. 2019;129(10):
cortical dysplasia type II leading to intractable epilepsy. Nat Med. 2015;21(4): 4207-4223. doi:10.1172/JCI127032
395-400. doi:10.1038/nm.3824 41. Lopez-Rivera JA, Leu C, Macnee M, et al. The genomic landscape across 474 surgi-
13. Poduri A, Evrony GD, Cai X, et al. Somatic activation of AKT3 causes hemispheric cally accessible epileptogenic human brain lesions. Brain. 2023;146(4):1342-1356.
developmental brain malformations. Neuron. 2012;74(1):41-48. doi:10.1016/ doi:10.1093/brain/awac376
j.neuron.2012.03.010 42. Moller RS, Weckhuysen S, Chipaux M, et al. Germline and somatic mutations in the
14. Salinas V, Vega P, Piccirilli MV, et al. Identification of a somatic mutation in the RHEB MTOR gene in focal cortical dysplasia and epilepsy. Neurol Genet. 2016;2(6):e118.
gene through high depth and ultra-high depth next generation sequencing in a patient doi:10.1212/NXG.0000000000000118
with hemimegalencephaly and drug resistant Epilepsy. Eur J Med Genet. 2019;62(11): 43. Wieser HG, Blume WT, Fish D, et al. Proposal for a new classification of outcome
103571. doi:10.1016/j.ejmg.2018.11.005 with respect to epileptic seizures following epilepsy surgery. Epilepsia. 2001;42(s2):
15. D’Gama AM, Geng Y, Couto JA, et al. Mammalian target of rapamycin pathway 282-286. doi:10.1046/j.1528-1157.2001.4220282.x
mutations cause hemimegalencephaly and focal cortical dysplasia. Ann Neurol. 2015; 44. Engel J Jr, Van Ness PC, Rasmussen TB, Ojemann LM. Outcome with respect to epileptic
77(4):720-725. doi:10.1002/ana.24357 seizure. In: Jr IEJ, ed. Surgical Treat-Ment of the Epilepsies. Raven Press; 1993:609-621.
16. Lim JS, Gopalappa R, Kim SH, et al. Somatic mutations in TSC1 and TSC2 cause 45. Chung C, Yang X, Bae T, et al. Comprehensive multi-omic profiling of somatic
focal cortical dysplasia. Am J Hum Genet. 2017;100(3):454-472. doi:10.1016/ mutations in malformations of cortical development. Nat Genet. 2023;55(2):209-220.
j.ajhg.2017.01.030 doi:10.1038/s41588-022-01276-9
17. D’Gama AM, Woodworth MB, Hossain AA, et al. Somatic mutations activating the 46. Morini A, Drossart T, Timsit MO, et al. Low-grade oncocytic renal tumor (LOT):
mTOR pathway in dorsal telencephalic progenitors cause a continuum of cortical mutations in mTOR pathway genes and low expression of FOXI1. Mod Pathol. 2022;
dysplasias. Cell Rep. 2017;21(13):3754-3766. doi:10.1016/j.celrep.2017.11.106 35(3):352-360. doi:10.1038/s41379-021-00906-7
18. Jamuar SS, Schmitz-Abe K, D’Gama AM, et al. Biallelic mutations in human DCC 47. Pirozzi F, Berkseth M, Shear R, et al. Profiling PI3K-AKT-MTOR variants in focal
cause developmental split-brain syndrome. Nat Genet. 2017;49(4):606-612. doi: brain malformations reveals new insights for diagnostic care. Brain. 2022;145(3):
10.1038/ng.3804 925-938. doi:10.1093/brain/awab376

48. Lee WS, Stephenson SEM, Howell KB, et al. Second-hit DEPDC5 mutation is limited 51. Moloney PB, Cavalleri GL, Delanty N. Epilepsy in the mTORopathies: opportunities
to dysmorphic neurons in cortical dysplasia type IIA. Ann Clin Transl Neurol. 2019; for precision medicine. Brain Commun. 2021;3(4):fcab222. doi:10.1093/brain-
6(7):1338-1344. doi:10.1002/acn3.50815 comms/fcab222
49. Yang H, Rudge DG, Koos JD, Vaidialingam B, Yang HJ, Pavletich NP. mTOR kinase 52. Gao XH, Li J, Gong HF, et al. Comparison of fresh frozen tissue with formalin-fixed
structure, mechanism and regulation. Nature. 2013;497(7448):217-223. doi:10.1038/ paraffin-embedded tissue for mutation analysis using a multi-gene panel in patients
nature12122 with colorectal cancer. Front Oncol. 2020;10:310. doi:10.3389/fonc.2020.00310
50. Tripolszki K, Knox R, Parker V, et al. Somatic mosaicism of the PIK3CA gene 53. Arai H, Akagi K, Nakagawa A, et al. Clinical and genetic diagnosis of Cowden syn-
identified in a Hungarian girl with macrodactyly and syndactyly. Eur J Med Genet. drome: a case report of a rare PTEN germline variant and diverse clinical presentation.
2016;59(4):223-226. doi:10.1016/j.ejmg.2016.02.002 Medicine. 2023;102(1):e32572. doi:10.1097/MD.0000000000032572

Adult Phenotype of SYNGAP1-DEE

Marlene Rong, MSc, Tim Benke, MD, PhD, Quratulain Zulfiqar Ali, MD, Ángel Aledo-Serrano, MD, PhD, Correspondence
Dr. Andrade
Allan Bayat, MD, PhD, Alessandra Rossi, MD, Orrin Devinsky, MD, Farah Qaiser, MSc, Anum S. Ali, BScN, MSc,
Danielle.Andrade@uhn.ca
Alfonso Fasano, MD, PhD, FAAN, Anne S. Bassett, MD, FRCPC, and Danielle M. Andrade, MD, MSc
Abstract
SYNGAP1 variants are associated with rare developmental and epileptic encephalopathies
(DEEs). Although SYNGAP1-related childhood phenotypes are well characterized, the adult
phenotype remains ill-defined. We sought to investigate phenotypes and outcomes in adults
with SYNGAP1 variants and epilepsy.
Methods
Patients 18 years or older with DEE carrying likely pathogenic and pathogenic (LP/P) SYN-
GAP1 variants were recruited through physicians’ practices and patient organization groups.
We used standardized questionnaires to evaluate current seizures, medication use, sleep, gas-
trointestinal symptoms, pain response, gait, social communication disorder and adaptive skills
of patients. We also assessed caregiver burden.
Results
Fourteen unrelated adult patients (median: 21 years, range: 18–65 years) with SYNGAP1-DEE
were identified, 11 with novel and 3 with known LP/P SYNGAP1 de novo variants. One patient
with a partial exon 3 deletion had greater daily living skills and social skills than others with
single-nucleotide variants. Ten of 14 (71%) patients had drug-resistant seizures, treated with a
median of 2 antiseizure medications. All patients (100%) had abnormal pain processing. Sleep
disturbances, social communication disorders, and aggressive/self-injurious behaviors were
each reported in 86% of patients. Only half of adults could walk with minimal or no assistance.
Toileting was normal in 29%, and 71% had constipation. No adult patients could read or
understand verbal material at a sixth-grade level or higher. Aggressive/self-injurious behaviors
were leading cause of caregiver burden. The oldest patient was aged 65 years; although non-
ambulant, she had walked independently when younger.
Discussion
Seventy-one percent of patients with SYNGAP1-DEEs continue to have seizures when adults.
Nonseizure comorbidities, especially aggression and self-injurious behaviors, are major man-
agement challenges in adults with SYNGAP1-DEE. Only 50% of adults can ambulate with
minimal or no assistance. Almost all adult patients depend on caregivers for many activities of
daily living. Prompt diagnostic genetic testing of adults with DEE can inform clinical care and
guide outcomes of precision therapies.
From the Institute of Medical Science (M.R.), University of Toronto; Adult Genetic Epilepsy (AGE) Program (M.R., Q.Z.A., F.Q., A.S.A., D.M.A.), Krembil Neurosciences Institute, Toronto
Western Hospital, University Health Network, Ontario, Canada; Department of Pediatrics, Neurology, Pharmacology and Otolaryngology (T.B.), University of Colorado School of
Medicine and Children’s Hospital Colorado, Aurora; Epilepsy and Neurogenetics Program (A.A.-S.), Neurology Department, Ruber Internacional Hospital, and Initiative for Neuro-
science (INCE) Foundation, Madrid, Spain; Department of Drug Design and Pharmacology (A. Bayat), University of Copenhagen; Department for Genetics and Personalized Medicine (A.
Bayat), Danish Epilepsy Centre, Dianalund; Institute for Regional Health Services (A. Bayat), University of Southern Denmark, Odense; Department of Epilepsy Genetics and Per-
sonalized Medicine (A.R.), Danish Epilepsy Centre, Dianalund, Denmark; Pediatric Clinic (A.R.), IRCCS San Matteo Hospital Foundation, University of Pavia, Italy; NYU Langone Epilepsy
Center (O.D.), NY; Edmond J. Safra Program in Parkinson’s Disease (A.F.), Morton and Gloria Shulman Movement Disorders Clinic, Toronto Western Hospital; Division of Neurology
(A.F.), University of Toronto; Krembil Brain Institute (A.F.); Clinical Genetics Research Program (A.S.B.), Centre for Addiction and Mental Health; The Dalglish Family 22q Clinic (A.S.B.),
Toronto General Hospital, University Health Network; Department of Psychiatry (A.S.B.), University of Toronto; Toronto Congenital Cardiac Centre for Adults (A.S.B.), Division of
Cardiology, Department of Medicine, and Department of Psychiatry, University Health Network; Toronto General Hospital Research Institute and Campbell Family Mental Health
Research Institute (A.S.B.); Division of Neurology (D.M.A.), Department of Medicine, University of Toronto, Ontario, Canada.

Glossary
ASD = autism spectrum disorder; ASMs = antiseizure medications; DBS = deep brain stimulation; DEEs = developmental and
epileptic encephalopathies; ID = intellectual disability; LP/P = likely pathogenic or pathogenic; RNS = responsive
neurostimulation; SCQ = Social Communication Questionnaire; VNS = vagus nerve stimulation.
Introduction was obtained from the substitute decision makers of all

patients.
The human SYNGAP1 gene on chromosome 6p21.3 encodes
the synaptic RAS-GTPase-activating protein 1 (SynGAP), Clinical data were collected using these validated assessments
which regulates cell growth and synaptic plasticity.1,2 As a tools.
highly enriched protein in excitatory glutamatergic synapses,
SynGAP is necessary for normal brain function by maintain- Severity of Clinical Outcomes
ing a balance between excitation and inhibition.3 A modified version of the Severity Assessment17 tool was used
to determine a composite score of disease severity in patients.
SYNGAP1 variants were initially associated with nonsyndromic The final severity assessment comprised 51 items that com-
intellectual disability (ID) (OMIM#306684).4-7 The link be- prehensively describe seizures, treatment usage, gait, pain re-
tween developmental and epileptic encephalopathies (DEEs) sponsiveness, toileting, reflux, and abnormal sleep. Each item
and SYNGAP1 was first observed in 2013, when 7 unrelated consisted of a Likert scale rating. Ratings of therapy effective-
patients presented with epilepsy, delayed development in the ness and the patient’s overall condition were also included. A
first years of life, autism spectrum disorder (ASD), and varying maximum score of 129/129 is indicative of the most severe
degrees of ID.8 Most patients with SYNGAP1-related DEE clinical phenotype. Seizures were classified according to the
harbor de novo heterozygous variants and present with varying 2017 International League Against Epilepsy guidelines.18
degrees of clinical severity.
Social Communication Skills
In patients with ID without epilepsy, SYNGAP1 pathogenic The Social Communication Questionnaire (SCQ) (Lifetime
variants have a prevalence of roughly 1:1,000 to 1:10,000, version) was used to screen for the presence of communica-
comprising approximately 1% of all ID cases.6,7 Similarly, tion deficits suggestive of ASD. The Lifetime version yields a
when SYNGAP1 variants were first linked to DEE, these Total Score that is interpreted with reference to cut-off scores.
variants comprised approximately 1% of patients with DEE.8 Scores above the cutoff of 15 suggest that the patient is likely
Currently, only 13 adult patients, from 9 different studies, to be on the autism spectrum and that a more extended
have been reported in the literature (median: 25 years, evaluation should be undertaken.
range: 18–33 years).8-16
Adaptive Behavioral Abilities
We evaluated an international group of adults with DEE and The Vineland Adaptive Behavior Scales 3 was used to de-
likely pathogenic or pathogenic (LP/P) SYNGAP1 variants. termine adaptive behavior abilities in the following domains:
Specifically, we investigated seizures, adaptive skills, social communication (receptive, expressive, and written), daily living
communication skills, sleep, gait and gastrointestinal distur- skills (personal, domestic, and community), socialization (in-
bances, pain tolerance, aggression, and behavior. Finally, we terpersonal relationships, play/leisure, and coping skills), mo-
evaluated the functional dependence of adult patients and tor ability (fine and gross motor), and maladaptive behaviors.
caregiver burden. Raw scores were extracted from each domain, which were se-
lected due to floor effects found in age-wise comparisons. An
overall rating was given based on the patient’s results compared
Methods with those of a norm sample, which is a representative group of
patients the same age from across the United States.
Study Population and Data Collection
We recruited patients aged 18 years or older with P/LP Statistical Analyses
SYNGAP1 variants to participate in our study. Enrollment Descriptive statistics were used to summarize clinical features
was performed between July 2021 and February 2022. Pa- and characteristics of the patients in the study. The Kruskal-
tients were recruited through the investigator’s institutions, Wallis test was used to compare assessment scores with var-
and through the SYNGAP1 Research Fund, across the fol- iant types (missense, nonsense, deletion, splice acceptor, and
lowing countries: Canada, the United States, Spain, and frameshift where applicable) and affected functional domains.
Germany. Study materials were translated from English to Statistical significance was set to p < 0.05, and statistical values
Spanish and German. The study was approved by the Re- were not reported for nonstatistically significant findings.
search Ethics Board with the University Health Network Analyses were conducted using R, and figures were created
(protocol 21-5009) in Toronto, Canada. Informed consent using GraphPad PRISM.

Genotyping stimulation (VNS), deep brain stimulation (DBS), or re-
The pathogenicity of SYNGAP1 variants was interpreted sponsive neurostimulation (RNS) as treatment options, al-
according to the American College of Medical Genetics and though 1 patient had used ketogenic diet in the past. Thirteen
Genomics guidelines.19 Each of the SYNGAP1 variants was patients (93%) reported no use of rescue medications or
also queried using the ClinVar database and GnomAD hospital visits for prolonged seizures. Four patients reported
browser database.20,21 no current usage of ASMs; however, only 2 of those patients
were seizure free without ASMs.
Data Availability
Deidentified data may be provided on reasonable request. Comorbidities
Constipation, pain responsiveness, sleep disturbances, dis-
ruptive daytime sleepiness, gait, toileting, and reflux were
Results evaluated (Figure 1). Four of 14 patients (29%) had no
constipation; 9 patients (64%) had constipation that was
Description of Participants
controlled either with or without medication over the pre-
Fourteen unrelated adult patients (9 female patients) partic-
vious 12 months. One patient had uncontrolled constipation
ipated in the study, with a median age of 21 years (range:
that was reported to be a key component to the patient’s
18–65) (Table 1). Five patients had a family history of epi-
quality of life.
lepsy. All patients came from nonconsanguineous families. A
full table of genotypic and phenotypic data from previously
Reflux was absent in 9 (64%) patients; 5 (36%) patients had
reported patients in the literature and patients from this study
reflux which was controlled with daily medication in 4/5 pa-
is presented in eTable 1 (links.lww.com/NXG/A647).
tients. Toileting was normal in 4 (29%) patients, timed in 5
Genotypic Spectrum (36%), and 5 (36%) patients required the use of diapers.
All patients had LP/P SYNGAP1 variants. Eleven are new, pre-
viously unreported variants. Three variants have been previously Pain responsiveness was abnormal in all patients; 4 (29%) had
reported in the literature.10,14,22,23 Eight of 14 patients had delayed reactions to minor pain, 5 (36%) had delayed reac-
frameshift variants. Five of 8 frameshift variants led to protein tions to major pain, whereas 5 (36%) had no response to
truncation (nonsense). Another 3 patients had nonsense vari- minor pain.
ants, one had a missense variant, and one had a splice acceptor
variant leading to exon skipping. One patient had an indel in Sleep disturbances were present in 12 (86%) patients—only 2
exon 3 of SYNGAP1. In all patients, the transcript analyzed was (14%) patients had normal sleep with no issues. Daytime
the longest one: NM006772.2. Together, these 14 patients sleepiness varied across the cohort, with 11/14 (79%) dis-
harbored 11 novel (likely) pathogenic variants (Table 1). playing some form of sleepiness and fatigue that was disrup-
tive during the day.
Phenotypic Spectrum
Scoliosis and Walking Abilities
Seizures Mild scoliosis not requiring treatment was reported in 8/14
Overall, 10/14 (71%) patients had at least one type of seizure (57%) patients, with no patients requiring braces or surgery
in the past 12 months. Seven of 14 (50%) patients had on- (Figure 2). Walking abilities varied as 3 (21%) walked in-
going nonconvulsive seizures (including eyelid myoclonia dependently with minimal assistance and 4 (29%) walked
with absence): yearly in one patient, monthly in another pa- community distances with minimal assistance. Two (14%)
tient, weekly in 2 patients, and daily in 3 patients. walked without assistance only on even surfaces. Another 2
patients (14%) could walk outdoors, but typically used
Four (29%) patients had convulsive seizures in the past 12 wheelchairs. One patient walked only indoors, while another
months. Of these patients, 3 had monthly convulsive seizures could only take some steps. The oldest patient (patient #1)
and 1 patient had daily convulsive seizures. was nonambulant, although she was able to walk in-
dependently when she was younger.
Three (21%) adult patients had prolonged convulsive seizures
lasting over 5 minutes in the previous 12 months. One patient Autism Spectrum Disorder
had seizures associated with hyperventilation. Twelve of 14 (86%) patients scored above the threshold on
the SCQ, indicating further evaluation for autism is required
Disruptive isolated epileptic spasms in the past 12 months (eTable 2, links.lww.com/NXG/A648). In fact, 11 of these 12
were reported by 4 (26%) patients. patients already had a previous formal diagnosis of ASD, given
by a physician or health care provider. Specific characteristics
In this cohort of adult patients, there was a median lifetime of in this cohort included ritual-like behaviors (11/14, 79%) and
4 antiseizure medications (ASMs) usage (range: 1–5+) and specific interests with high intensity (11/14, 79%). Of rele-
median current usage of 2 (range: 1–5+) ASMs. There was vance, 12/14 (86%) patients showed self-injurious behavior,
no reported current usage of ketogenic diet, vagus nerve e.g., head banging and biting oneself.

4
Table 1 Demographics of 14 Adult Patients With SYNGAP1 Variants and Their Genetic Profiles (n = 14)
Patient Age ASD diagnosis Family Exon Protein ACMG Variant previously
# Sex (y) (age of diagnosis) Ethnicity history # cDNA consequence Variant classification Inheritance interpretation reported?
1 F 65 Yes White No Exon 8 c.1329_ p.Lys444Glyfs*27 Frameshift leading to Unknown Pathogenic No

1333delCAAGG truncation
2 M 20 Yes (2y) White No Exon 8 c.781_784delGACA p.Asp261Metfs*3 Frameshift leading to De novo Pathogenic No
truncation
3 F 20 No White, Jewish Yes, sister Exon c.3295delT p.Tyr1099Metfs* Frameshift leading to De novo Pathogenic No
15 truncation
4 F 48 Yes Other No Exon 8 c.870del p.Tyr291fs Frameshift De novo Pathogenic No
5 F 24 Yes (5y) White Yes, half sister Exon c.2019delA p.Thr674Profs*36 Frameshift leading to De novo Pathogenic No
12 truncation
6 F 19 Yes South Asian, Latino/ Yes, maternal Exon 8 c.1167_1168del p.Gly391fs Frameshift De novo Pathogenic Jimenez-Gomez et al.
Hispanic uncle (2019)
7 M 23 Yes White N/A Exon c.3233_ p.Val1078fs* Frameshift leading to De novo Pathogenic No
15 3236delTCAG truncation
8 F 20 Yes Latino/Hispanic Yes, Exon c.2526dup p.Met843Hisfs*7 Frameshift leading to De novo Likely No

grandmother 15 truncation pathogenic
9 F 23 Yes White, Latino/Hispanic Yes, cousin Exon c.4006G>A p.Glu1336Lys Missense De novo Likely No
19 pathogenic
10 F 20 No White No Exon 5 c.403C>T p.Arg135* Nonsense De novo Pathogenic Mignot et al. (2016)
11 F 22 Yes Latino/Hispanic No Exon c.1861C>T p.Arg621* Nonsense De novo Pathogenic Aguilera et al. (2021)
11 Verma et al. (2020)
12 M 20 Yes White No Exon c.3227delT p.Leu1076* Nonsense De novo Pathogenic No

15
13 M 18 No White No Exon 3 c.190-15_ — Insertion/deletion (indel) De novo Pathogenic No

206delins28
14 M 22 Yes White, Jewish N/A — c.1532-1G>C — Splice acceptor leading to De novo Likely No
exon 10 skipping pathogenic
Abbreviations: ASD = autism spectrum disorder; F=Female; M = Male; N/A = not available.
Sex, age, ASD diagnosis, family history of epilepsy, and ethnicity are listed when provided. Information on the genetic variant, molecular consequences, inheritance, and interpretation from a patient’s genetic report is provided.
The zygosity of all variants was heterozygous.
Neurology.org/NG
Figure 1 Summary Graphs of Various Clinical Features in Adults With SYNGAP1
Severity assessment results regarding: (A) constipation, (B) pain responsiveness, (C) sleep disturbances, (D) daytime sleepiness, (E) toileting, and
(F) reflux (n = 14).
Adaptive Behavioral Abilities With respect to gross and fine motor skills, 11 (79%) patients
Adaptive behavioral abilities were varied among SYNGAP1- were able to sit unsupported for at least 10 minutes, 12 (86%)
DEE patients. Domain level scores are presented in eTable 2 could walk upstairs, and 11 (79%) could walk downstairs.
(links.lww.com/NXG/A648). There were no statistically Seven (50%) could jump off the ground with both feet.
significant differences between clinical findings and geno- However, no patient had the ability to manipulate very small
types, except for one patient, a 19-year-old man carrying an objects.
indel of SYNGAP1 exon 3 (c.190-15_206delins28). This pa-
tient demonstrated an elevated ability to perform daily living Regarding language and learning abilities, 7 (50%) patients
skills. He also exhibited stronger social skills and abilities to could talk using short phrases or sentences. However, all
pursue relationships, compared with the rest of the cohort. patients were responsive to caregivers, could recognize their
Although his overall summary score for adaptive behaviors own names, and respond to one-word actions. Eight patients
was moderately low for his age, all other patients in this cohort (57%) could identify all letters of the alphabet, but only 2
had lower scores compared with normative data. Nine pa- (14%) could sometimes write at least 10 simple words from
tients (64%) were able to feed themselves with a fork and memory. Although 7 patients (50%) could read at least 10
spoon. Twelve (86%) were cooperative in personal activities, words, only 5 (36%) could read simple sentences out loud and
such as undressing, dressing, and washing of the hands and just 3 (21%) could read simple stories out loud. No adult
face. Of the 9 patients who were able to dress themselves, patients were able to read or understand material at a sixth-
none could use zippers (Table 2). grade level or higher.

Figure 2 Walking Abilities of Adult Patients With SYNGAP1-DEE
Maladaptive Behaviors This patient experienced her first seizure at age 7 months. She
Physical aggression, temper tantrums, and neediness were developed absence seizures and generalized tonic-clonic sei-
observed in 11 (79%) patients. Ten patients (71%) disobeyed zures that were drug resistant. At age 18 years (several years
those in authority and had eating problems, such as a refusal to before receiving the SYNGAP1 genetic diagnosis), she un-
eat or overeating. Loss of awareness regarding surroundings derwent a frontal lobe resection for the treatment of seizures.
was identified by caregivers in 12 (86%) patients. Unfortunately, the surgery was unsuccessful. She has never
had the ketogenic diet, VNS, or DBS/RNS. By age 65 years,
Negative findings included no lying and breaking rules/laws her caregiver reported daily absences with eyelid myoclonia
because of peer pressure, no harming animals, or interest in induced by sounds and lights and daily isolated epileptic
extreme violence. Furthermore, there were no reports about spasms that are disruptive to the patient and/or family. Other
holding untrue beliefs or talks about auditory/visual halluci- convulsive seizures had not occurred in over a year, and no
nations. One patient expressed feelings of helplessness/ prolonged seizures lasting more than 5 minutes were reported
hopelessness, and another patient has threatened to hurt/kill in the previous 6 months. The caregiver’s impression of sei-
someone in the past. zures in the past 12 months was of worsening seizures, with
daily seizures that are disruptive to daily life.
Longevity
This study features the oldest SYNGAP1-DEE patient cur- This patient has moderate intellectual disability and has re-
rently reported in the literature, a 65-year-old White woman ceived a formal diagnosis of autism spectrum disorder. She is
carrying a pathogenic frameshift variant (p.Lys444Glyfs*27) at present wheelchair-bound, has feeding/swallowing issues,
of unknown inheritance. and is unable to consume previously enjoyed foods due to
choking hazards. Other clinical features of concern include
uncontrolled constipation and toileting accidents. Daytime
sleepiness is disruptive throughout most of the week, and the
Table 2 Comparison of Daily Living Abilities Between
patient often arouses from sleep more than once per week.
Pediatric Patients and Adult Patients With
SYNGAP1 Variants Overall, the caregiver reported a “really worse” patient con-
dition compared with the first 10 years of life, particularly
SYNGAP1 pertaining to motor ability. Some key maladaptive behaviors
pediatric Current study
patients (SYNGAP1 adult reported included a tendency to harm herself, frequent threats
Daily living ability (n = 13)a patients) (n = 14) to hurt or kill someone, lose awareness of surrounding, and
Speak in short phrases or sentences 39% 50% fixation on a specific topic.
Eating independently 62% 64%
Collaborative during personal hygiene 40% 86% Discussion

Simple dressing unassisted 39% 64% In this study, we present molecular and clinical information
Adults may be more independent than pediatric patients but are still low
on 14 adults with SYNGAP1-DEE, the largest such cohort yet
functioning for their age. to be reported. Of the 14 adult patients included in this study,
a
Findings for SYNGAP1 pediatric patients were extracted from Lo Barco
et al.24
11 had previously unreported likely pathogenic or pathogenic
SYNGAP1 variants. One patient with an indel of SYNGAP1

exon 3 had stronger daily living skills and social skills in- as stereotypies.9,13,14 Other behavioral issues, such as ag-
cluding abilities to pursue relationships, compared with the gression, temper tantrums, and obsessive behaviors, align with
rest of the cohort who had frameshift, missense, nonsense, previous reports in both children and adults.11,14-16
and splice acceptor site variants affecting exons 3 to 19.
The daily living abilities of adult SYNGAP1-DEE patients
Half of the adult patients were free of convulsive seizures. have not been previously explored. We found that all adults in
Four patients were free of all types of seizures in the previous our cohort had below average adaptive skills, at similar levels
12 months, and 2 of those 4 were off ASMs. In previous to pediatric SYNGAP1 patients.30 While direct comparisons
reports, only 19% of patients 7 years and older were seizure- cannot be made due to different assessment tools used in each
free.16 On one end of the spectrum, 28% of adult patients in study, Table 2 presents a brief summary of the daily living
our study are seizure-free; on the other end, 21% of these abilities of pediatric patients (range: 3.7–17.7 years) as
adults still have prolonged convulsive seizures, which repre- reported by Lo Barco et al. and the adults of this study.30
sent a significant morbidity.9,10 When compared with the pediatric sample, 25% more adults
were able to dress themselves with assistance, pointing toward
Abnormal pain responsiveness was observed in 100% of adults, possible improved daily living abilities across the life course.
making it more common than previously reported in chil- The proportion of patients able to feed themselves was similar
dren.16 Although it is possible that the pain threshold may in both pediatric and adult patients (;60–70%), suggesting
change over time, methodological differences could also have that these skills may be preserved into middle age. Of interest,
led to this finding. For example, previous studies evaluated a slightly higher proportion of adults could speak using short
severely abnormal pain responsiveness (e.g., to broken phrases or sentences compared with pediatric patients. This is
bones),16 while we also asked about mild nociception abnor- in contrast to adults previously presented in the literature,
malities. The precise mechanism leading to abnormal noci- where speech was either absent or, at most, of the ability of a
ception is unclear, but Syngap1 mouse models reveal that touch 1-year-old.10,11,14
is weakly encoded in upper-lamina neurons in the somato-
sensory cortex, leading to improper sensory processing.24 The exact reason for the improvement in daily living abilities
compared with pediatric patients is unknown. However, in
Self-injury behaviors were previously observed in children other DEEs, such as Dravet syndrome, seizure freedom has
with SYNGAP1.11,14,16 However, our findings showed a been associated with improved everyday executive function-
whopping 86% prevalence of self-injury in adults with this ing of children and young adults.32,33 Although direct com-
condition. Self-injury has been noted in patients with DEE as a parisons cannot be made, it is possible that this may be the
whole, but it seems to be less prevalent compared with case in some adults with SYNGAP1. Longitudinal studies
SYNGAP1-DEE. For example, only 28% of patients with Rett would be required to confirm this relationship.
syndrome may have externalizing behavioral issues, such as
self-injury.25,26 Similarly, one cohort of adult patients with Regarding sleep, our findings align with previous research
Dravet syndrome saw self-injurious behaviors in 31% of pa- observing sleep disturbances in SYNGAP1-DEE pediatric
tients.27 Self-injury and aggression are rarely observed in pa- patients, particularly difficulties initiating and maintaining
tients with CDKL5 variants.17 Finally, an evaluation of adults sleep.16,34 Other studies of adult cases in the literature report
with KCNQ2-DEEs found that 31% of patients demonstrated insomnia and night-time awakenings.11 As such, sleep dis-
self-injurious behaviors.28 turbances and daytime sleepiness may be important features
that warrant continuous monitoring and treatment as patients
It is unclear why most adults with SYNGAP1-DEE show self- age. The reasons for sleep problems are unclear, but studies of
injurious behavior. One potential mechanism could include a adult mice have shown that SYNGAP1’s abnormal interictal
relationship to abnormal sensory perception of self-harming epileptiform discharges can increase during sleep and in-
behavior as a form of self-stimulation. While ASMs are gener- terrupt sleep architecture.35 It is possible that a similar path-
ally well tolerated in several forms of DEEs,29 one patient in our ophysiology underlies the sleep problems in adult patients.
study showed some behavioral improvement after discontinu-
ing an ASM. Similarly, one adult in the literature discontinued As seen in other DEEs, walking may be worse in adults.36 In
levetiracetam at age 13 years due to behavioral issues.12 Given this cohort, only 50% of patients could walk with minimal or
the caregiver burden associated with neurobehavioural chal- no assistance, and the oldest patient (age 65 years) is
lenges in patients with DEEs, considerations for behavioral wheelchair-bound, although she had been able to walk when
changes when selecting an ASM may be worthwhile.30,31 younger.
The prevalence of ASD diagnoses is greater in adults (79%) Most patients with SYNGAP1 LP/P variants diagnosed today
with SYNGAP1-DEE than previously reported for children are children. This is in part due to the recency of our
(53%).16 The reasons for this discrepancy are unclear. Re- knowledge of this gene as a cause of DEE. As such, when
gardless, ASD emerges as a key finding in adults, aligning with parents of newly diagnosed SYNGAP1-DEE children ask
other adults in the literature exhibiting autistic features, such about longevity, there are no definitive answers. Here, we

describe the natural history of the oldest patient with SYN- Q. Zulfiqar Ali reports no disclosures relevant to the manu-
GAP1-DEE reported so far, aged 65 years. script; A. Aledo‐Serrano reports no disclosures relevant to
the manuscript; A. Bayat reports no disclosures relevant to the
Although this is the largest cohort of adult patients with SYN- manuscript; A. Rossi reports no disclosures relevant to the
GAP1-DEE yet studied, the sample size is small, and only 2 manuscript; O. Devinsky receives grant support from NINDS,
patients were older than 24 years. This might be in part due to NIMH, MURI, CDC and NSF. He has equity and/or com-
the relatively recent recognition of SYNGAP1 gene variants as a pensation from the following companies: Ajna Biosciences,
cause of DEEs and adults not receiving up-to-date genetic test- Tilray, Receptor Life Sciences, Hitch Biosciences, Tevard
ing.5 As with any caregiver reported outcome, findings may be Biosciences, Regel Biosciences, Script Biosciences, Actio Bio-
subject to recall bias. This study did not track specific ASM sciences, Empatica, SilverSpike, and California Cannabis En-
usage. Studies are needed to examine the efficacy of drugs, as terprises (CCE). He has received consulting fees from Zogenix,
there is no standardized ASM management of SYNGAP1-DEE Ultragenyx, BridgeBio, GeneMedicine and Marinus. He holds
patients. Conclusions about adults with SYNGAP1-DEE will patents for the use of cannabidiol in treating neurologic dis-
require large numbers and detailed longitudinal follow-up data, orders which are owned by GW Pharmaceuticals for which he
ideally in a prospective study. waived financial interests. He holds other patents in molecular
biology. He is the managing partner of the PhiFund Ventures;
This is the study of adult patients with SYNGAP1-DEE. The F. Qaiser reports no disclosures relevant to the manuscript; A.
detailed characterization of the natural history of this condi- Ali reports no disclosures relevant to the manuscript; A. Fasano
tion, including seizures, communication skills, pain re- reports no disclosures relevant to the manuscript; A.S. Bassett
sponsiveness, sleep, digestive issues, gait abnormalities, and reports no disclosures relevant to the manuscript; D.M.
other comorbidities, may help to identifying adults with DEE Andrade receives grant support from McLaughlin Foundation,
who so far lack genetic diagnosis. We also report the adaptive UHN Foundation, Dravet Syndrome Foundation. She also
behavioral abilities of adults, allowing caregivers an opportu- received consulting fees from UCB, Biocodex, Paladin, Eisai.
nity to help plan for future care as SYNGAP1-DEE patients Finally, she receives royalties from UpToDate. Go to Neurol-
enter adulthood. ogy.org/NG for full disclosures.
In this study, we also report the oldest SYNGAP1-DEE patient Publication History
in the literature and the first view into possible longevity Received by Neurology: Genetics May 16, 2023. Accepted in final form
issues. Further studies in larger groups of adults are still September 20, 2023. Submitted and externally peer reviewed. The
necessary to have a more comprehensive view of the natural handling editor was Massimo Pandolfo, MD, FAAN.
history of this condition. Encouraging genetic (re)testing of
adults with undiagnosed epilepsies may contribute to these
efforts.
Appendix Authors
Acknowledgment Name Location Contribution
The authors acknowledge the participating patients and families
for their time, especially the SYNGAP1 Research Fund. Marlene Institute of Medical Science, Drafting/revision of the
Rong, MSc University of Toronto; Adult manuscript for content,
Genetic Epilepsy (AGE) including medical writing
Program, Krembil for content; major role in
Study Funding Neurosciences Institute, the acquisition of data;
MR received unrestricted educational funding from Biocodex. Toronto Western Hospital, study concept or design;
AB is funded by a BRIDGE - Translational Excellence Pro- University Health Network, analysis or interpretation
Ontario, Canada of data
gramme grant funded by the Novo Nordisk Foundation, grant
agreement number: NNF20SA0064340. ASB holds the Tim Benke, Department of Pediatrics, Drafting/revision of the
MD, PhD Neurology, Pharmacology and manuscript for content,
Dalglish Chair in 22q11.2 Deletion Syndrome at the Uni- Otolaryngology, University of including medical writing
versity Health Network and University of Toronto. DMA Colorado School of Medicine for content; major role in
and Children’s Hospital the acquisition of data;
received grant support from Ontario Brain Institute and Colorado, Aurora study concept or design
McLaughlin Foundation for this study.
Quratulain Adult Genetic Epilepsy (AGE) Drafting/revision of the
Zulfiqar Ali, Program, Krembil manuscript for content,
Disclosure MD Neurosciences Institute, including medical writing
Toronto Western Hospital, for content; major role in
M. Rong reports no disclosures relevant to the manuscript; University Health Network, the acquisition of data;
T.A. Benke receives grant support from NINDS, NIDCD, NIA, Ontario, Canada study concept or design
Simons Foundation and IRSF. He performs consultancy for Ángel Aledo- Epilepsy and Neurogenetics Drafting/revision of the
AveXis, Ovid, GW Pharmaceuticals, International Rett Syn- Serrano, MD, Program, Neurology manuscript for content,
PhD Department, Ruber including medical writing
drome Foundation, Takeda, Taysha, CureGRIN, GRIN Internacional Hospital; for content; major role in
Therapeutics, Alcyone, Neurogene, and Marinus; Clinical Initiative for Neuroscience the acquisition of data
(INCE) Foundation,
Trials with Acadia, Ovid, GW Pharmaceuticals, Marinus and Madrid, Spain
RSRT; all remuneration has been made to his department;

References
Appendix (continued) 1. Hae Kim J, Liao D, Lau LF, Huganir RL. SynGAP: A Synaptic RasGAP That Associates
with the PSD-95/SAP90 Protein Family, Vol 20; 1998:683-691.
Name Location Contribution 2. Ligeti E, Welti S, Scheffzek K. Inhibition and termination of physiological re-
sponses by GTPase activating proteins. Physiol Rev. 2012;92(1):237-272. doi:
Allan Bayat, Department of Drug Design Drafting/revision of the 10.1152/physrev.00045.2010
MD, PhD and Pharmacology, University manuscript for content, 3. Clement JP, Aceti M, Creson TK, et al. Pathogenic SYNGAP1 mutations impair
of Copenhagen; Department including medical writing cognitive development by disrupting maturation of dendritic spine synapses. Cell.
for Genetics and Personalized for content; major role in 2012;151(4):709-723. doi:10.1016/j.cell.2012.08.045
Medicine, Danish Epilepsy the acquisition of data 4. Hamdan FF, Gauthier J, Spiegelman D, et al. Mutations in SYNGAP1 in autosomal
Centre, Dianalund; Institute nonsyndromic mental retardation. N Engl J Med. 2009;360(6):599-605. doi:10.1056/
for Regional Health Services, NEJMoa0805392
University of Southern 5. Berryer MH, Hamdan FF, Klitten LL, et al. Mutations in SYNGAP1 cause intellectual
Denmark, Odense, Denmark disability, autism, and a specific form of epilepsy by inducing haploinsufficiency. Hum
Mutat. 2013;34(2):385-394. doi:10.1002/humu.22248
Alessandra Department of Epilepsy Major role in the 6. McRae JF, Clayton S, Fitzgerald TW, et al. Prevalence and architecture of de
Rossi, MD Genetics and Personalized acquisition of data novo mutations in developmental disorders. Nature. 2017;542(7642):433-438. doi:
Medicine, Danish Epilepsy 10.1038/nature21062
Centre, Dianalund; Pediatric 7. Deciphering Developmental Disorders Study. Large-scale discovery of novel genetic
Clinic, IRCCS San Matteo causes of developmental disorders. Nature. 2015;519(7542):223-228. doi:10.1038/
Hospital Foundation, nature14135
University of Pavia, Italy 8. Carvill GL, Heavin SB, Yendle SC, et al. Targeted resequencing in epileptic en-
cephalopathies identifies de novo mutations in CHD2 and SYNGAP1. Nat Genet.
Orrin NYU Langone Epilepsy Center, Drafting/revision of the 2013;45(7):825-830. doi:10.1038/ng.2646
Devinsky, NY manuscript for content, 9. Klitten LL, Møller RS, Nikanorova M, Silahtaroglu A, Hjalgrim H, Tommerup N. A
MD including medical writing balanced translocation disrupts SYNGAP1 in a patient with intellectual disability,
for content; major role in speech impairment, and epilepsy with myoclonic absences (EMA). Epilepsia. 2011;
the acquisition of data 52(12):e190-e193. doi:10.1111/j.1528-1167.2011.03304.x
10. Mignot C, von Stülpnagel C, Nava C, et al. Genetic and neurodevelopmental spec-
Farah Adult Genetic Epilepsy (AGE) Drafting/revision of the trum of SYNGAP1 -associated intellectual disability and epilepsy. J Med Genet. 2016;
Qaiser, MSc Program, Krembil manuscript for content, 53(8):511-522. doi:10.1136/jmedgenet-2015-103451
Neurosciences Institute, including medical writing 11. Prchalova D, Havlovicova M, Sterbova K, Stranecky V, Hancarova M, Sedlacek Z. Analysis
Toronto Western Hospital, for content; analysis or of 31-year-old patient with SYNGAP1 gene defect points to importance of variants in
University Health Network, interpretation of data broader splice regions and reveals developmental trajectory of SYNGAP1-associated phe-
Ontario, Canada notype: case report. BMC Med Genet. 2017;18(1):62. doi:10.1186/s12881-017-0425-4
12. Kluger G, von Stülpnagel-Steinbeis C, Arnold S, Eschermann K, Hartlieb T. Positive
Anum S. Ali, Adult Genetic Epilepsy (AGE) Drafting/revision of the short-term effect of low-dose rosuvastatin in a patient with SYNGAP1-associated
BScN, MSc Program, Krembil manuscript for content, epilepsy. Neuropediatrics. 2019;50(4):266-267. doi:10.1055/s-0039-1681066
Neurosciences Institute, including medical writing 13. Vidal S, Brandi N, Pacheco P, et al. The most recurrent monogenic disorders that
Toronto Western Hospital, for content; analysis or overlap with the phenotype of Rett syndrome. Eur J Paediatr Neurol. 2019;23(4):
University Health Network, interpretation of data 609-620. doi:10.1016/j.ejpn.2019.04.006
Ontario, Canada 14. Aguilera C, Gabau E, Ramirez-Mallafré A, et al. New genes involved in Angelman
syndrome-like: expanding the genetic spectrum. PLoS ONE. 2021;16(10):e0258766.
Alfonso Edmond J. Safra Program in Drafting/revision of the doi:10.1371/journal.pone.0258766
Fasano, MD, Parkinson’s Disease, Morton manuscript for content, 15. Cook EH, Masaki JT, Guter SJ, Najjar F. Lovastatin treatment of a patient with a De
PhD, FAAN and Gloria Shulman including medical writing Novo SYNGAP1 protein truncating variant. J Child Adolesc Psychopharmacol. 2019;
Movement Disorders Clinic, for content; study concept 29(4):321-322. doi:10.1089/cap.2018.0159
Toronto Western Hospital; or design 16. Vlaskamp DRM, Shaw BJ, Burgess R, et al. SYNGAP1 encephalopathy: a distinctive
Division of Neurology, generalized developmental and epileptic encephalopathy. Neurology. 2019;92(2):
University of Toronto; Krembil e96-e107. doi:10.1212/WNL.0000000000006729
Brain Institute, Ontario, 17. Saldaris J, Weisenberg J, Pestana-Knight E, et al. Content validation of clinician-
Canada reported items for a severity measure for CDKL5 deficiency disorder. J Child Neurol.
2021;36(11):998-1006. doi:10.1177/08830738211019576
Anne S. Clinical Genetics Research Drafting/revision of the 18. Fisher RS, Cross JH, French JA, et al. Operational classification of seizure types by the
Bassett, MD, Program, Centre for Addiction manuscript for content, International League Against Epilepsy: Position Paper of the ILAE Commission for
FRCPC and Mental Health; The including medical writing Classification and Terminology. Epilepsia. 2017;58(4):522-530. doi:10.1111/epi.13670
Dalglish Family 22q Clinic, for content; study concept 19. Richards S, Aziz N, Bale S, et al. Standards and guidelines for the interpretation of
Toronto General Hospital, or design sequence variants: a joint consensus recommendation of the American College of
University Health Network; Medical Genetics and Genomics and the Association for Molecular Pathology. Genet
Department of Psychiatry, Med. 2015;17(5):405-424. doi:10.1038/gim.2015.30
University of Toronto; Toronto 20. Landrum MJ, Lee JM, Benson M, et al. ClinVar: Improving access to variant inter-
Congenital Cardiac Centre for pretations and supporting evidence. Nucleic Acids Res. 2018;46(D1):D1062-D1067.
Adults, Division of Cardiology, doi:10.1093/nar/gkx1153
Department of Medicine, and 21. Karczewski KJ, Francioli LC, Tiao G, et al. The mutational constraint spectrum
Department of Psychiatry, quantified from variation in 141,456 humans. Nature. 2020;581(7809):434-443. doi:
University Health Network; 10.1038/s41586-020-2308-7
Toronto General Hospital 22. Verma V, Kumar MJV, Sharma K, et al. Pharmacological intervention in young
Research Institute and adolescents rescues synaptic physiology and behavioural deficits in Syngap1+/− mice.
Campbell Family Mental Exp Brain Res. 2022;240(1):289-309. doi:10.1007/s00221-021-06254-x
Health Research Institute, 23. Jimenez-Gomez A, Niu S, Andujar-Perez F, et al. Phenotypic characterization of
Ontario, Canada individuals with SYNGAP1 pathogenic variants reveals a potential correlation be-
tween posterior dominant rhythm and developmental progression. J Neurodev Disord.
Danielle M. Adult Genetic Epilepsy (AGE) Drafting/revision of the 2019;11(1):18-18. doi:10.1186/s11689-019-9276-y
Andrade, Program, Krembil manuscript for content, 24. Michaelson SD, Ozkan ED, Aceti M, et al. SYNGAP1 heterozygosity disrupts sensory
MD, MSc Neurosciences Institute, Toronto including medical writing processing by reducing touch-related activity within somatosensory cortex circuits.
Western Hospital, University for content; major role in Nat Neurosci. 2018;21(12):1-13. doi:10.1038/s41593-018-0268-0
Health Network; Division of the acquisition of data; 25. Buchanan CB, Stallworth JL, Scott AE, et al. Behavioral profiles in Rett syndrome:
Neurology, Department of study concept or design; data from the natural history study. Brain Dev. 2019;41(2):123-134. doi:10.1016/
Medicine, University of Toronto, analysis or interpretation j.braindev.2018.08.008
Ontario, Canada of data 26. Cianfaglione R, Clarke A, Kerr M, Hastings RP, Oliver C, Felce D. Ageing in Rett
syndrome. J Intellect Disabil Res. 2016;60(2):182-190. doi:10.1111/jir.12228

27. Berkvens JJL, Veugen I, Veendrick-Meekes MJBM, et al. Autism and behavior in adult patients phase 3 long-term extension study in children/young adults with Dravet syndrome.
with Dravet syndrome (DS). Epilepsy Behav. 2015;47:11-16. doi:10.1016/j.yebeh.2015.04.057 Epilepsy Behav. 2021;121(Pt A):108024. doi:10.1016/j.yebeh.2021.108024
28. Boets S, Johannesen KM, Destree A, et al. Adult phenotype of KCNQ2 encepha- 33. Catarino CB, Liu JYW, Liagkouras I, et al. Dravet syndrome as epileptic encepha-
lopathy. J Med Genet. 2022;59(6):528-535. doi:10.1136/jmedgenet-2020-107449 lopathy: evidence from long-term course and neuropathology. Brain J Neurol. 2011;
29. Perucca P, Mula M. Antiepileptic drug effects on mood and behavior: molecular 134(Pt 10):2982-3010. doi:10.1093/brain/awr129
targets. Epilepsy Behav. 2013;26(3):440-449. doi:10.1016/j.yebeh.2012.09.018 34. Smith-Hicks C, Wright D, Kenny A, et al. Sleep abnormalities in the synaptopathies—
30. Lo Barco T, De Gaetano L, Santangelo E, et al. SYNGAP1-related developmental and SYNGAP1-related intellectual disability and Phelan–McDermid syndrome. Brain Sci.
epileptic encephalopathy: the impact on daily life. Epilepsy Behav. 2022;127:108500. 2021;11(9):1229-1229. doi:10.3390/brainsci11091229
doi:10.1016/j.yebeh.2021.108500 35. Creson TK, Rojas C, Hwaun E, et al. Re-expression of SynGAP protein in adulthood
31. Gallop K, Lloyd AJ, Olt J, Marshall J. Impact of developmental and epileptic en- improves translatable measures of brain function and behavior. eLife. 2019;8:e46752.
cephalopathies on caregivers: a literature review. Epilepsy Behav. 2021;124:108324. doi:10.7554/eLife.46752
doi:10.1016/j.yebeh.2021.108324 36. Selvarajah A, Gorodetsky C, Marques P, et al. Progressive worsening of gait and motor
32. Bishop KI, Isquith PK, Gioia GA, et al. Improved everyday executive functioning abnormalities in older adults with Dravet syndrome. Neurology. 2022;98(22):
following profound reduction in seizure frequency with fenfluramine: analysis from a e2204-e2210. doi:10.1212/WNL.0000000000200341

Molecular Diagnosis of Facioscapulohumeral Muscular

Dystrophy in Patients Clinically Suspected of FSHD
Using Optical Genome Mapping
Naga M. Guruju, PhD, Vanessa Jump, BS, Richard Lemmers, PhD, Silvere Van Der Maarel, PhD, Ruby Liu, PhD, Correspondence
Dr. Guruju
Babi R. Nallamilli, PhD, Suresh Shenoy, PhD, Alka Chaubey, PhD, Pratik Koppikar, BS, Rajiv Rose, PhD,
naga.guruju@revvity.com
Satish Khadilkar, MD, and Madhuri Hegde, PhD
Abstract
Facioscapulohumeral muscular dystrophy (FSHD) represents the third most common mus-
cular dystrophy in the general population and is characterized by progressive and often
asymmetric muscle weakness of the face, upper extremities, arms, lower leg, and hip girdle. In
FSHD type 1, contraction of the number of D4Z4 repeats to 1–10 on the chromosome
4–permissive allele (4qA) results in abnormal epigenetic derepression of the DUX4 gene in
skeletal muscle. In FSHD type 2, epigenetic derepression of the DUX4 gene on the permissive
allele (4qA) with normal-sized D4Z4 repeats (mostly 8–20) is caused by heterozygous path-
ogenic variants in chromatin modifier genes such as SMCHD1, DNMT3B, or LRIF1. We present
validation of the optical genome mapping (OGM) platform for accurate mapping of the D4Z4
repeat size, followed by diagnostic testing of 547 cases with a suspected clinical diagnosis of FSHD
and next-generation sequencing (NGS) of the SMCHD1 gene to identify cases with FSHD2.
Methods
OGM with Bionano Genomics Saphyr and EnFocus FSHD analysis software was used to
identify FSHD haplotypes and D4Z4 repeat number and compared with the gold standard of
Southern blot–based diagnosis. A custom Agilent SureSelect enrichment kit was used to enrich
SMCHD1, followed by NGS on an Illumina system with 100-bp paired-end reads. Copy
number variants were assessed using NxClinical software.
Results
We performed OGM for the diagnosis of FSHD in 547 patients suspected of FSHD between
December 2019 and December 2022, including 301 male (55%) and 246 female patients (45%).
Overall, 308 of the referred patients were positive for D4Z4 contraction on a permissive haplotype,
resulting in a diagnosis of FSHD1. A total of 252 of 547 patients were referred for concurrent testing
for FSHD1 and FSHD2. This resulted in the identification of FSHD2 in 9/252 (3.6%) patients. In
our FSHD2 cohort, the 4qA allele size ranged from 8 to 18 repeats. Among FSHD1-positive cases,
2 patients had biallelic contraction and 4 patients had homozygous contraction and showed early
onset of clinical features. Nine of the 308 patients (3%) positive for 4qA contraction had mosaic 4q
alleles with contraction on at least one 4qA allele. The overall diagnostic yield in our cohort was 58%.
Discussion
A combination of OGM to identify the FSHD haplotype and D4Z4 repeat number and NGS to
identify sequence and copy number variants in the SMCHD1 gene is a practical and cost-
effective option with increased precision for accurate diagnosis of FSHD types 1 and 2.
From the Revvity Omics (N.M.G., V.J., Ruby Liu, B.R.N., S.S., R.R., M.H.), Pittsburgh, PA; Leiden University Medical Centre (Richard Lemmers, S.V.D.M.), Netherlands; Bionano Genomics
(A.C.), San Diego, CA; UT Dallas (P.K.), TX; Bombay Hospital (S.K.), Mumbai, India.

Glossary
CNV = copy number variant; FSHD = facioscapulohumeral muscular dystrophy; NGS = next-generation sequencing; OGM =
optical genome mapping; PFGE = pulsed-field gel electrophoresis; RU = repeat units; VOUS = variant of unknown significance.
Introduction repeats in the 4q35.2 region are critical in the diagnosis of

both FSHD1 and FSHD2.
Facioscapulohumeral muscular dystrophy (FSHD, MIM#158900)
is the third most common muscular dystrophy in the general The most common method of diagnosing FSHD is Southern
population after myotonic dystrophy and dystrophinopathy, with blot by double digestion of genomic DNA by EcoRI alone
an estimated prevalence of 1:8,500.1,2 FSHD is an autosomal and double digestion by EcoRI/BlnI and/or EcoRI/XapI,
dominant disorder characterized by progressive and often followed by linear or pulsed-field gel electrophoresis and
asymmetric muscle weakness in the face, scapular stabilizers, hybridization with a p13E-11 probe. Multiple restriction
shoulders, arms, lower leg, and hip girdle. Approximately 20% enzymes and probes are required to discriminate between
of patients are wheelchair bound. The age at onset is variable 4qA, 4qB, and 10q haplotypes by Southern blot analysis.15-17
and ranges from infancy to adulthood. Two types of FSHD Southern blot has been the gold standard method for the
have been described: FSHD1 (MIM#158900) and FSHD2 diagnosis of FSHD1; however, this method of diagnosis
(MIM#158901). FSHD1 is the most common form and is ob- has limitations. It is time-consuming, the number of D4Z4
served in >95% of individuals diagnosed with FSHD, repeats are calculated by band size, and the method may
whereas FSHD2 is observed in the remaining <5% of cases require radioactive material. In addition, most diagnostic
with FSHD. Both FSHD1 and FSHD2 are caused by ab- laboratories use linear gel electrophoresis for EcoRI frag-
normal epigenetic derepression of the double homeobox 4 ment separation, which presents technical difficulties in
gene (DUX4, MIM#606009), a cleavage stage and germline identifying somatic mosaics in D4Z4 units; however, this
transcription factor that is normally expressed during em- can be overcome by using a pulsed-field gel electrophoresis
bryogenesis and silenced during development due to (PFGE) separation method.17 Moreover, the detection of
chromatin condensation.3,4 a deletion or structural rearrangement upstream of the
D4Z4 repeats that disrupts the p13E-11 probe-binding site
The DUX4 gene is embedded within a 3.3 kilobase (kb) re- requires multiple rounds of hybridization with different
peat unit (RU) of D4Z4 macrosatellite repeats on chromo- probes. Recently, alternative methods, such as molecular
some 4 in the q35.2 region.5,6 Similar D4Z4 repeats are combing,18,19 single-molecule real-time sequencing tech-
located on chromosome 10q26.3 that exhibit approximately nology (SMRT),20 Nanopore CRISPR-/Cas9-targeted
98% sequence identity, adding complexity to the testing.7 resequencing (nCATS),21 and a qPCR-based approach,22
Two sequence variants of the 4q subtelomere region distal to have been developed to determine the number of macro-
the last repeat have been observed: the permissive allele, 4qA, satellite (D4Z4) repeats. Two studies have shown that
and the nonpermissive allele, 4qB. In the general population, optical genome mapping (OGM) has been evaluated for
the D4Z4 repeat number ranges from 11 to 100 on chro- clinical utility in detecting D4Z4 repeats at the 4q region to
mosome 4. However, 8–10 repeats are observed in 2% of the diagnose FSHD.23,24
European population.8 FSHD1 is caused by contraction of the
number of repeats to 1–10 on 4qA, resulting in local chro- In this study, we are presenting the validation of optical ge-
matin relaxation evidenced by, e.g., hypomethylation of D4Z4 nome mapping (OGM) technology in the diagnosis of FSHD
repeats, leading to abnormal toxic expression of the DUX4 and OGM data from 547 cases clinically suspected of FSHD.
gene in skeletal muscle.9-11 The phenotypic severity of 317 positive samples were detected with a 58% diagnostic
FSHD1 roughly correlates with D4Z4 repeat size on 4qA with yield, demonstrating the power and efficacy of OGM in
a lower number of repeats corresponding with earlier age at detecting the number of D4Z4 repeat units and haplotypes in
onset and rapid progression. The repeat region on 4qB and the diagnosis of FSHD.
10qA do not have a somatic polyadenylation signal; therefore,
the DUX4 RNA is not stable and does not cause FSHD.
FSHD2 is clinically identical to FSHD1 but has a different
genetic cause. In FSHD2, hypomethylation of the D4Z4 re-
Methods
peat on 4qA is caused by pathogenic variants in chromatin Sample Preparation
modifier genes, such as SMCHD1, DNMT3B, and LRIF1, Genomic DNA is initially isolated from fresh or frozen pe-
leading to global hypomethylation at the D4Z4 repeats on ripheral blood using the Bionano DNA isolation kit (Bionano
chromosomes 4 and 10.12-14 Discrimination between the Genomics, San Diego).25 The labeling and staining portion of
chromosome 4q35 and 10q26 D4Z4 repeats and the 4qA and the protocol allows the gDNA to be visualized and mapped
4qB haplotypes and identification of the number of D4Z4 once inserted into the Bionano Saphyr instrument.

Optical Genome Mapping Results
In brief, OGM was performed using Bionano Genomics
Saphyr with subsequent analysis by Bionano Enfocus Validation of OGM in Identifying D4Z4 Repeat
FSHD analysis software (Bionano, San Diego, CA) to Size and Haplotype
identify the FSHD haplotype and D4Z4 repeat number in Initial optimization of OGM in identifying D4Z4 repeat con-
patients suspected of FSHD. Molecules aligning to the traction was performed on 6 cell lines derived from patients
D4Z4 repeat regions on chromosome 4 using human with FSHD that were previously confirmed by Southern blot at
genome reference build GRCh38 are distinguished from Coriell Institute (Camden, NJ). All 6 samples were processed
regions of high homology on chromosome 10 based on at our laboratory (Revvity Omics, Pittsburgh) and at Bionano
the fluorescent pattern of markers proximal to the D4Z4 Genomics (3 independent runs). D4Z4 repeat number from
repeat region. The permissive (4qA) and nonpermissive our laboratory and 3 independent runs performed at Bionano
(4qB) alleles were assigned using the dynamic pro- Genomics showed concordance with the previous Southern
gramming algorithm included in the Enfocus FSHD blot results (eTable 1, links.lww.com/NXG/A652).
analysis pipeline. The OGM de novo assembly was used to
detect any large structural changes in the SMCHD1 locus Next, we evaluated the sensitivity of OGM in 14 individuals
on chromosome 18.25 with well-characterized clinical features of FSHD by com-
paring OGM results obtained at our laboratory to standard
Southern Blot Southern blotting performed at Leiden University Medical
Southern blot analysis was performed at the Van der Maarel Center in the Netherlands (eTable 2, links.lww.com/NXG/
laboratory (Leiden University Medical Center, the Nether- A653, eFigure, links.lww.com/NXG/A651). Twelve of the 14
lands) using genomic DNA embedded in agarose plugs and individuals had D4Z4 repeat contraction of the 4qA allele,
PFGE for the separation of the DNA.26 whereas 2 individuals had no contraction of D4Z4 according
to both OGM and Southern blot analyses. Two of the positive
cases (772 and 830) had a mosaic 4qA allele according to
SMCHD1 Gene Sequencing
OGM, which was comparable with the Southern blot results.
A custom Agilent SureSelect enrichment kit was used to
This confirms the ability of OGM to detect mosaic alleles. In
enrich the SMCHD1 gene, followed by next-generation
addition, the repeat size and haplotype highly correlated with
sequencing (NGS) on an Illumina system with 100-bp
Southern blot analysis [p < 0.001]; however, a repeat size
paired-end reads. The analyzed regions include the cod-
difference of 1 unit was observed between OGM and
ing exons and 50 bp of flanking intronic regions on both
Southern blot analysis due to the differences in the calculation
sides of each exon. Copy number variation was assessed
methods used to determine the size of the repeats.
using Bionano Genomics’ NxClinical software (Bionano,
San Diego, CA). Variants were evaluated by their reported
To further validate OGM in identifying D4Z4 repeats in
frequency in databases, including the Genome Aggrega-
healthy participants, we tested whole blood samples from
tion Database (gnomAD), Human Gene Mutation Data-
normal human volunteers by OGM at Revvity Omics, Pitts-
base (HGMD), ClinVar, and other disease-specific
burgh, and another site (Bionano Genomics, San Diego, CA).
databases when applicable. Variants that have a pop-
These samples were predicted to show no contraction in
ulation frequency greater than expected given the preva-
D4Z4 repeats on a 4qA allele. As expected, none of the 6
lence of the disease in the general population were
normal human volunteers had contracted D4Z4 repeats
considered to be benign. All variants including VOUS
(i.e., all had >10 RU, eTable 3, links.lww.com/NXG/A654).
were evaluated up to ±3, and variants in the exon and
Five of the volunteers had identical haplotype and repeat unit
intron boundaries were reported. Only pathogenic vari-
measurements at the 2 laboratories. Volunteer 4 had an allele
ants are reported in the region between ±3 and ±50 base
with 69 D4Z4 RU and 4qA haplotype but was measured as
pairs.
>20 units with an unknown haplotype at Bionano Genomics.
This discrepancy is likely because at Revvity Omics, Pitts-
Standard Protocol Approvals, Registrations, burgh, we produced a longer molecule reaching the end of
and Patient Consents D4Z4 including the haplotype, whereas the same molecule at
Optical genome mapping and/or SMCHD1 gene sequencing Bionano Genomics could not reach up to the distal region
presented in this study is in line with the original request containing the haplotype and were represented as >20 repeats
for diagnostic testing, and therefore, no additional informed and unknown haplotype (eTable 3, links.lww.com/NXG/
consent is needed. A654).
Data Availability OGM Analysis of Patients Suspected of FSHD

Data presented in this article cannot be made publicly avail- After thorough validation of OGM, we performed FSHD
able because they consider patient information. To protect testing using OGM, followed by sequencing of the SMCHD1
patient privacy, access to the data can only be made by request gene for the diagnosis of FSHD2. OGM can identify both the
from the corresponding author. D4Z4 allele size as well as A or B haplotype (shown in

Figure 1 Case Examples Processed by OGM Differentiating 4qA and 4qB Haplotypes and D4Z4 Repeats
(A) The reference is shown in green with a graphical representation of both haplotype patterns (A and B) shown on the same molecule for comparison. The
patient alleles are shown in blue. (A) D4Z4 contraction of 2 RU was detected on the 4qA (permissive) haplotype. A 2nd 4qB (nonpermissive) allele with 22 RU
was detected. (B) A D4Z4 repeat contraction of 1 and 8 units on the 4qA (permissive) haplotype. An additional allele with a repeat count of 12 was detected on
the 4qA haplotype indicating mosaicism. (C) A biallelic D4Z4 repeat contraction of 6 and 9 units on the 4qA (permissive) haplotype. (D) A D4Z4 repeat
contraction of 4 units on the 4qA (permissive) haplotype in cis with a duplication (red arrows) that caused this allele to be masked in the FSHD output. A second
4qB (nonpermissive) allele with 45 repeat units was detected in this patient.
Figure 1A). OGM was performed for 547 patients suspected 547 referred patients, 308 were positive for a D4Z4 contrac-
of FSHD, including 301 male (55%) and 246 female patients tion on a 4qA allele, resulting in a diagnosis of FSHD1, and 9
(45%), between December 2019 and December 2022. Of the cases were positive for FSHD2 (Table 1). The overall

diagnostic yield in our cohort, including both FSHD1 and 4qB alleles, the median was 23 RU and range was 5–84
FSHD2, was 58%. Among the 308 cases positive for FSHD1, RU. The median allele size of all 10qB was 23 RU (range
mosaic alleles with at least 1 contracted 4qA allele were ob- 6–71 RU). The frequency of 10qA and 10qB observed in
served in 9 cases (3%, Figure 1B), biallelic contraction of 4qA our cohort of cases suspected of FSHD was 94.2% and
were observed in 2 cases (0.6%, 1 case with A1/A10 and 1 5.8%, respectively.
with A6/A9, Figure 1C), and 4 patients had an apparently
homozygous contraction (A5/A5, A8/A8, and 2 patients with 295/547 cases were referred for only FSHD1 (OGM);
A7/A7). In cases with an apparently homozygous contraction, therefore, no SMCHD1 sequencing was performed. How-
a possible distal deletion within the repeats at 4qter could not ever, 252/547 cases were referred for testing for both
be completely ruled out. Upon manual observation of the FSHD1 (OGM) and FSHD2 (NGS analysis of SMCHD1);
molecules, we were able to identify a proximal duplication in in these cases, the FSHD2 testing was performed concur-
cis with D4Z4 repeat contraction on 4qA in 3 cases (example rently regardless of the FSHD1 result. This resulted in
in Figure 1D). In our cohort, among 317 cases positive for the identification of 130 (of 252) cases positive for FSHD1
FSHD, 117 (37%) have contraction on 10qA, and among 230 and 9 patients positive for FSHD2. All cases with FSHD2
cases negative for FSHD, 78 (34%) have contraction on 10qA showed intermediate repeats of 8–18 RU on the 4qA
(regardless of 4q haplotype) and 54 (23.5%) have at least 1 haplotype. The overall frequency of cases with FSHD2
normal 4qA allele and contraction on 10qA. among patients screened for both FSHD1 and 2 was 3.6%.
Pathogenic or likely pathogenic SMCHD1 variants are
The allele frequency of 4qA and 4qB in cases negative for reported in Table 2. No copy number variants were ob-
FSHD cases are equally distributed (4qA = 49.1% and 4qB = served in the SMCHD1 gene by NGS-based deletion/
50.9%), similar to the frequency observed in the general duplication analysis. In addition, 3 patients had a variant of
population27,28; however, in our cases positive for FSHD, the unknown significance in SMCHD1 in combination with
frequency is skewed with A being 76.9% and B being 23.1%. 4qA (<20 RU). Of 295 cases referred for FSHD1 testing
Figure 2 shows the distribution of 4qA and 4qB D4Z4 repeats, only, 178 were positive for FSHD1; therefore, no follow-up
as well as 10qA and 10qB. In cases with FSHD2, the 4qA FSHD2 testing was required. Of the 117/295 cases nega-
repeats ranged from 8 to 18 RU (1case had contraction to tive for FSHD1, 88 carried at least 1 4qA allele with >10
8 as well as pathogenic variant in SMCHD1 with age at repeats; 25 of these cases had between 11 and 20 repeats in
onset in teens) with a median 13 RU, comparable with 4qA. In these cases, the possibility of FSHD2 could not be
previous reports of OGM data. 23,24 In cases negative for completely ruled out because these patients were referred
FSHD, the overall median including both short and long only for FSHD1 testing, and therefore, SMCHD1 gene
alleles on 4qA was 32 RU (range 11–90 RU), whereas the sequencing was not performed.
median of the shortest 4qA allele was 29 RU (range
11–77). Regarding chromosome 10, the median shortest
10qA allele size in FSHD1-positive cases was 14 RU
(range 1–55 RU), and the median shortest 10qA allele
Discussion
size in FSHD1-negative cases was 14, range (1–71). There FSHD is one of the genetic disorders associated with re-
was no significant difference in shortest 10qA allele size petitive regions caused by contraction of the macrosatellite
between FSHD-negative and FSHD-positive cases. For all region (D4Z4) on chromosome 4, and the macrosatellite
Table 1 Result Summary of Patients With FSHD

D4Z4 number 547 Total patients tested
Disease Association Haplotype of repeats SMCHD1 (overall diagnostic yield = 58%)
FSHD type 1 4qA (Permissive) 1–10 N/A 308 patients (56%)
FSHD type 2 4qA (Permissive) 8–18 Pathogenic/LP variant 9 patients (2%, see Table 2)
FSHD 1 and 2 4qA (Permissive) 1–10 Pathogenic variant 1 patient
Mosaic FSHD1 4qA (Permissive) 1–10 N/A 9 patients
cis duplication of region proximal to D4Z4 repeats 4qA (Permissive) 1–10 N/A 3 patients
Biallelic contraction 4qA (Permissive) 1–10 N/A 2 patients
Homozygous contraction 4qA (Permissive) 1–10 N/A 4 patients
Abbreviation: FSHD = facioscapulohumeral muscular dystrophy.

The case with FSHD 1 and 2 is included in both the FSHD1 category and the FSHD2 category. Mosaic FSHD1, proximal cis duplication, bialleleic contraction, and
homozygous contraction are included in the FSHD1 category.

Figure 2 Distribution of D4Z4 Repeats in 4q and 10q Haplotypes
The median RU size is represented by the horizontal gray

lines. (A) Shortest 4qA allele RU in all FSHD-negative cases;
RU median: 29; Range: 11–77. (B) Shortest 4qA allele RU in
FSHD1-positive cases; RU median: 5; range: 1–10. (C)
Shortest 4qA allele RU in FSHD2-positive cases; RU median:
13; Range: 8–18. (D) Shortest 10qA allele RU in FSHD1-posi-
tive cases; RU median: 14; range 1–55. (E) Shortest 10qA
allele RU in FSHD-negative cases; RU median 14; range 1–71.
(F) 4qB alleles in all cases combined; RU median: 23; range:
5–84. (G) All 10qB alleles; RU median 23; range 6–71.
region has high homology with other regions of the genome.29 assembling NGS data. Though long read sequencing can
Because this disease is complex with repeat contraction, address these limitations, clinical adoption of long read is cost
rearrangements within the repeat sequences, translocation prohibitive for clinical laboratories at this time. Southern blot
between 4qA and 10qA repeats, duplication of repeat se- is widely used as the gold standard for identifying D4Z4
quences, and variants in chromatin modifier genes (e.g., repetitive regions and haplotypes. Molecular combing and,
SMCHD1, DNMT3B, and LRIF1), diagnosis is difficult when more recently, nCATS18-20 have been developed for FSHD
using a single technique. Despite significant improvements in diagnosis by the identification of repeat contraction. In ad-
NGS technology over the past decade, some limitations exist dition, OGM has been validated in the diagnosis of FSHD.23,24
in terms of the sensitivity of poorly covered and uncovered
regions. In particular, repetitive DNA sequences such as the The definitive diagnosis of FSHD is important for effective
D4Z4 repeat pose major obstacles to accurate analysis by disease management in patients and for appropriate genetic
creating uncertainty in the processes of aligning and counseling. In general, the Southern blot technique has been
Table 2 Variants Detected in the SMCHD1 Gene in Patients Positive for FSHD2
Variant Position Variant type ACMG classification 4q35 allele 1 4q35 allele 2
SMCHD1 c.2071_2075del Exon 16 Deletion Pathogenic A13 A19
SMCHD1 c.3276+4_3276+7del Intron 25 59 splice site Pathogenic A13 B24
SMCHD1 c.4566G>A (p.Thr1522=) Exon 36 59 splice site Pathogenic A15 A29
SMCHD1 c.1186C>T (p.Gln396Ter) Exon 10 Nonsense Pathogenic A23a A42
SMCHD1 c.2176_2179del Exon 17 Deletion Likely Pathogenic A14 B66
SMCHD1 c.35_45dup Exon 1 Duplication Likely Pathogenic A18 B22
SMCHD1 c.3938C>G (p.Ser1313Ter) Exon 31 Nonsense Likely Pathogenic A8 A55
SMCHD1 c.5286dup Exon 42 Duplication Likely Pathogenic A11 A19
SMCHD1 c.5720-2A>C Intron 45 39 splice site Likely Pathogenic A12 A18
a
Patient is asymptomatic, family studies showed 4qA/11 in combination with variant is associated with FSHD2 (see Figure 4).

replaced by alternate techniques for routine molecular diag- OGM data reported a median of 6, 7, and 5, respectively24,30
nostics because it is technically challenging and requires high (Figure 3). In cases with FSHD2, 4qA repeats ranged from 8
technical expertise. Because most cases with FSHD (95%) are to 18 and median 13 RU, comparable with previous reports of
due to contraction of the 4q permissive allele to <10 D4Z4 OGM data.24 Among cases with FSHD1 with contraction of
RU, OGM can identify the haplotype corresponding to the the D4Z4 allele, we identified somatic mosaic cases, biallelic
D4Z4 repeats on both copies of chromosome 4 as well as contraction, and homozygous contraction. However, FSHD
chromosome 10. Identification of the haplotype and repeat testing is challenging due to the presence of common trans-
size is also important in FSHD2 because an intermediate- locations between the 4q and 10q arrays, duplication of D4Z4
sized (8–20 RU) 4qA allele along with a pathogenic variant in alleles, and somatic mosaicism and other rearrangements.
the SMCHD1 gene causes most cases with FSHD2.
Due to high homology (approximately 98%) between the 4q
In this study, we used OGM in combination with SMCHD1 and 10q regions, complex rearrangements between these re-
gene sequencing for the diagnosis of FSHD1 and FSHD2. gions lead to a hybrid haplotype consisting of 4qA and 10q-
The validation studies demonstrate 100% analytical accuracy like repeats. These hybrid alleles, specifically the presence of
and precision of this assay using FSHD-positive Coriell cell 4qA D4Z4 repeat at the distal end of the 10qA repeats on
lines, with an accuracy of ± 1 repeat. Normal male and female chromosome 10, can cause FSHD.31 Hybrid alleles are
control samples revealed that the D4Z4 repeats were within detected in 0.5% and 14% of the European and Asian control
normal range (15–69 D4Z4 repeats of either the 4qA or 4qB populations, respectively. In this population of cases sus-
haplotype). In 14 clinically diagnosed cases with FSHD, 12 pected of FSHD and in which FSHD1 is negative, then the
cases were positive for repeat contractions (ranging from 2 to 10qA allele repeat size may be carefully evaluated.32 OGM
8 repeats) and were reproducible across intrasite, intersite, gives information on the contraction of D4Z4 repeats on 4qA
interinstrument, and intermethod comparisons. Two of the and 10qA, but it cannot differentiate 4qA repeats within 10qA
positive cases also had a mosaic pattern of the contracted (hybrid pattern). One study recently identified D4Z4 repeats
allele. Chromosome 10q and the normal 4qB allele were also of 4qA within 10q repeats in approximately 6.7% of cases in
deemed highly reproducible across different runs at 2 sites. which contraction of 10qA was observed by OGM.24 No
significant difference for contracted 10qA was observed in our
OGM combined with NGS sequencing of the SMCHD1 gene cohort between the FSHD-positive group and the FSHD-
would be able to diagnose most cases of FSHD though this negative group. Patients with a clinical indication of FSHD
does not overrule clinical diagnosis indicating the possibility with contraction of 10qA repeats (<10 RU) observed by
of yet undiscovered loci. In our cohort, we observed an overall OGM require further testing by Southern blot analysis, which
diagnostic yield of 58%, including FSHD1 (56%) and FSHD2 can detect hybrid 4qA/10qA alleles using different restriction
(the frequency of FSHD2 is 3.6% among 252 patients enzymes.
screened for both FSHD1 and FSHD2). The median number
of D4Z4 repeats in our cohort of cases with FSHD1 was 5, and NGS sequencing of SMCHD1 could identify both single-
the most frequent repeat number was 5 (22%), followed by 4 nucleotide variants and CNVs (deletions/duplications).
(17%), whereas the previously published Southern blot and Among 252 cases that were screened, 9 were positive for
Figure 3 4qA Contracted Allele Size Distribution Among 308 Patients Positive for FSHD1

either “likely pathogenic” or “pathogenic” variants in combi- however, upon manual inspection, we identified the duplica-
nation with an intermediate repeat size on the 4qA haplotype. tion and repeat size corresponding to the haplotype. In ad-
D4Z4 methylation analysis could provide further evidence for dition, we also identified a small inversion immediately
pathogenesis.33 No deletions within the SMCHD1 gene were proximal to the D4Z4 region in 1 case that was not initially
identified in our cohort. We observed repeat contraction to 8 detected in the FSHD output by the software (data not
RU on 4qA and a pathogenic variant in SMCHD1 in a patient shown). For these reasons, we manually inspect molecules
with severe clinical features. Contraction of D4Z4 on 4qA where a single chromosome 4 allele is detected by the FSHD
with a pathogenic variant in SMCHD1 has been reported with output to ensure there are no rearrangements that may be
an earlier disease onset and more rapid progression.34 Indi- masking a contracted allele. Improvements in the bio-
viduals with an SMCHD1 variant alone may not develop informatic tools will help in cases of duplication/inversion
FSHD2 because the pathogenic variant may not segregate immediately proximal to the D4Z4 repeat region. Apart from
with the 4qA allele. In a family with 2 generations, the father the identification of repeat sequences at the telomere region
(I:1) and his 3 brothers were affected with FSHD with early of chromosomes 4 and 10, using OGM de novo analysis, we
age at onset in their early teens. The individual (I:1) had 2 can identify large deletions of the 18p region. Individuals
4qA alleles with 11 and 42 D4Z4 RU, in combination with a with 18p microdeletions or loss of short arm of chromosome
heterozygous nonsense variant, SMCHD1.c.1186>T. We 18 (18p syndrome) including the SMCHD1 gene in
performed OGM and sequenced SMCHD1 on a 24-year-old combination with permissive 4qA may also show clinical
index case (II:3) with a family history of FSHD who was features of FSHD along with other clinical features un-
healthy with no clinical symptoms of FSHD during testing. related to FSHD.33,35 In addition, OGM can identify
OGM identified 23 and 42 RU on 2 4qA alleles and a het- genome-wide CNVs (deletions/duplications), insertions,
erozygous variant, SMCHD1 c.1186>T, as observed in his and other rearrangements that may be related to the
father and 2 sisters who are affected with FSHD. The father (I: phenotype when FSHD diagnosis is negative.
1) and sisters (II:1 and II:2) of the index case have a 4qA allele
with 11 RU in combination with the SMCHD1 pathogenic Based on our experience with OGM, we developed a testing
variant, causing FSHD2. However, although individual (II:3) algorithm (Figure 5) for the diagnosis of FSHD1 and 2. Ad-
has the familial pathogenic variant in SMCHD1, this in- ditional testing is suggested in cases in which the clinician’s
dividual received 23 RU from his mother and 42 RU from his expectation is that the patient has true FSHD, and no con-
father, and the SMCHD1 variant did not cosegregate with 23 traction of 4qA or SMCHD1 variant is observed. If one of
repeats; thus, no phenotype developed. Follow-up studies are the alleles is 4qA with RU between 10 and 20, and rarely >20,
required to monitor for late age at onset or slow progression then this may cause FSHD2; therefore, methylation analysis is
of FSHD. This confirms that the combination of 11 RU and recommended to identify hypomethylation of the D4Z4 that
SMCHD1 variant caused FSHD2 in this family (Figure 4). might be caused by other epigenetic modifier genes that cause
FSHD2. If there is a contraction on 10qA with clinical in-
We also identified duplication immediately proximal to the dication of FSHD and normal repeats observed in chromo-
D4Z4 repeats in 3 cases (example shown in Figure 1D). The some 4, Southern blot (i.e., pulsed-field gel electrophoresis) is
software masked the duplication allele in the FSHD output; recommended.
Figure 4 Segregation Analysis of a Nonsense Variant in the SMCHD1 Gene

Figure 5 Testing Algorithm for the Diagnosis of FSHD1 and FSHD2
*In case of clear FSHD phenotype,

methylation testing is recommended
to rule out hypomethylation that may
be caused by complex D4Z4 rear-
rangements or unusual alleles
(p13E11 deletion, D4Z4 proximal ex-
tended deletions, or consider PFGE/
Southern blot to identify hybrid alleles
when 1–4 RU are observed on 10qA).
Routine use of conventional Southern blot technique may not

identify the repeat contraction and corresponding haplotype to- Appendix Authors
gether and may not identify mosaic and complex rearrangements
compared with OGM or PFGE. Many laboratories do not have
technical expertise in processing and analysis of PFGE Southern Naga M. Revvity Omics Drafting/revision of the article for
Guruju, PhD content, including medical writing for
blot. OGM requires less training to process the sample and requires content; major role in the acquisition of
lesser amounts of DNA compared with Southern blot analysis. data; study concept or design; and
analysis or interpretation of data
OGM has advantages of identifying repeat sizing on 4q and 10q
together with haplotyping in a single run and can identify mosaicism. Vanessa Revvity Omics Drafting/revision of the article for
Jump, BS content, including medical writing for
Recent advances in understanding FSHD disease mechanisms have content; major role in the acquisition of
helped in developing targeted treatment for FSHD. New cost- data; and analysis or interpretation of
data
effective and faster turnaround technologies such as OGM help in
early diagnosis that improve early intervention leading to better Richard Leiden Drafting/revision of the article for
clinical outcomes. The results of our study demonstrate 100% Lemmers, University content, including medical writing for
PhD Medical Centre content; analysis or interpretation of
reproducibility and precision of the samples used for the data
LDT evaluation of Bionano’s Saphyr ® genome imaging
Silvere Van Leiden Drafting/revision of the article for
platform as a high-resolution, high-throughput, and cost- Der Maarel, University content, including medical writing for
effective method for the diagnosis of FSHD. PhD Medical Centre content; analysis or interpretation of
data
Ruby Liu, PhD Revvity Omics Analysis or interpretation of data;

Study Funding drafting/revision of the article for
The authors report no targeted funding. content, including medical writing for
content
Disclosure Babi R. Revvity Omics Drafting/revision of the article for

Nallamilli, content, including medical writing for
The authors report no relevant disclosures. Go to Neurology. PhD content
org/NG for full disclosures.
Suresh Revvity Omics Drafting/revision of the article for content,
Shenoy, PhD including medical writing for content
Publication History Alka Bionano Drafting/revision of the article for

Received by Neurology: Genetics May 19, 2023. Accepted in final form Chaubey, PhD Genomics content, including medical writing for
content
September 18, 2023. Submitted and externally peer reviewed. The
handling editor was Antonella Spinazzola, MD. Continued

15. Lemmers RJ, Osborn M, Haaf T, et al. D4F104S1 deletion in facioscapulohumeral
muscular dystrophy: phenotype, size, and detection. Neurology. 2003;61(2):178-183.
Appendix (continued)
doi:10.1212/01.wnl.0000078889.51444.81
16. Deak KL, Lemmers RJ, Stajich JM, et al. Genotype-phenotype study in an FSHD
family with a proximal deletion encompassing p13E-11 and D4Z4. Neurology. 2007;
68(8):578-582. doi:10.1212/01.wnl.0000254991.21818.f3
Pratik UT Dallas Drafting/revision of the article for 17. Lemmers RJ, van der Wielen MJ, Bakker E, Padberg GW, Frants RR, van der Maarel
Koppikar, BS content, including medical writing for SM. Somatic mosaicism in FSHD often goes undetected. Ann Neurol. 2004;55(6):
content 845-850. doi:10.1002/ana.20106
18. Vasale J, Boyar F, Jocson M, et al. Molecular combing compared to Southern blot for
Rajiv Rose, Revvity Omics Drafting/revision of the article for measuring D4Z4 contractions in FSHD. Neuromuscul Disord. 2015;25(12):945-951.
PhD content, including medical writing for doi:10.1016/j.nmd.2015.08.008
content 19. Nguyen K, Puppo F, Roche S, et al. Molecular combing reveals complex 4q35 rear-
rangements in Facioscapulohumeral dystrophy. Hum Mutat. 2017;38(10):1432-1441.
Satish Bombay Drafting/revision of the article for doi:10.1002/humu.23304
Khadilkar, MD Hospital content, including medical writing for 20. Morioka MS, Kitazume M, Osaki K, Wood J, Tanaka Y. Filling in the gap of human
content chromosome 4: single molecule real time sequencing of macrosatellite repeats in the
facioscapulohumeral muscular dystrophy locus. PLoS One. 2016;11(3):e0151963.
Madhuri Revvity Omics Drafting/revision of the article for doi:10.1371/journal.pone.0151963
Hegde, PhD content, including medical writing for 21. Hiramuki Y, Kure Y, Saito Y, et al. Simultaneous measurement of the size and methyl-
content; major role in the acquisition of ation of chromosome 4qA-D4Z4 repeats in facioscapulohumeral muscular dystrophy by
data; study concept or design; and long-read sequencing. J Transl Med. 2022;20(1):517. doi:10.1186/s12967-022-03743-7
analysis or interpretation of data 22. Zernov NV, Guskova AA, Skoblov MY. FSHD1 diagnosis in a Russian population
using a qPCR-based approach. Diagnostics (Basel). 2021;11(6):982. doi:10.3390/
diagnostics11060982
23. Dai Y, Li P, Wang Z, et al. Single-molecule optical mapping enables quantitative
References measurement of D4Z4 repeats in facioscapulohumeral muscular dystrophy (FSHD).
1. Deenen JC, Arnts H, van der Maarel SM, et al. Population-based incidence and J Med Genet. 2020;57(2):109-120. doi:10.1136/jmedgenet-2019-106078
prevalence of facioscapulohumeral dystrophy. Neurology. 2014;83(12):1056-1059. 24. Stence AA, Thomason JG, Pruessner JA, et al. Validation of optical genome mapping
doi:10.1212/WNL.0000000000000797 for the molecular diagnosis of facioscapulohumeral muscular dystrophy. J Mol Diagn.
2. LaPelusa A, Kentris M. Muscular Dystrophy. StatPearls; 2023. 2021;23(11):1506-1514. doi:10.1016/j.jmoldx.2021.07.021
3. Dixit M, Ansseau E, Tassin A, et al. DUX4, a candidate gene of facioscapulohumeral 25. Koppikar P, Shenoy S, Guruju N, Hegde M. Testing for facioscapulohumeral mus-
muscular dystrophy, encodes a transcriptional activator of PITX1. Proc Natl Acad Sci cular dystrophy with optical genome mapping. Curr Protoc. 2023;3(1):e629. doi:
USA. 2007;104(46):18157-18162. doi:10.1073/pnas.0708659104 10.1002/cpz1.629
4. Snider L, Geng LN, Lemmers RJ, et al. Facioscapulohumeral dystrophy: incomplete 26. Lemmers RJ. Analyzing copy number variation using pulsed-field gel electrophoresis:
suppression of a retrotransposed gene. PLoS Genet. 2010;6(10):e1001181. doi: providing a genetic diagnosis for FSHD1. Methods Mol Biol. 2017;1492:107-125. doi:
10.1371/journal.pgen.1001181 10.1007/978-1-4939-6442-0_7
5. Gabriëls J, Beckers MC, Ding H, et al. Nucleotide sequence of the partially deleted 27. Thomas NS, Wiseman K, Spurlock G, MacDonald M, Ustek D, Upadhyaya M. A large
D4Z4 locus in a patient with FSHD identifies a putative gene within each 3.3 kb patient study confirming that facioscapulohumeral muscular dystrophy (FSHD)
element. Gene. 1999;236(1):25-32. doi:10.1016/s0378-1119(99)00267-x disease expression is almost exclusively associated with an FSHD locus located on a
6. van Deutekom JC, Wijmenga C, van Tienhoven EA, et al. FSHD associated DNA 4qA-defined 4qter subtelomere. J Med Genet. 2007;44(3):215-218. doi:10.1136/
rearrangements are due to deletions of integral copies of a 3.2 kb tandemly repeated jmg.2006.042804
unit. Hum Mol Genet. 1993;2(12):2037-2042. doi:10.1093/hmg/2.12.2037 28. Lemmers R, van der Vliet PJ, Vreijling JP, et al. Cis D4Z4 repeat duplications asso-
7. Bakker E, Wijmenga C, Vossen RH, et al. The FSHD-linked locus D4F104S1 (p13E- ciated with facioscapulohumeral muscular dystrophy type 2. Hum Mol Genet. 2018;
11) on 4q35 has a homologue on 10qter. Muscle Nerve. 1995;18(S13):S39-S44. doi: 27(20):3488-3497. doi:10.1093/hmg/ddy236
10.1002/mus.880181309 29. Lyle R, Wright TJ, Clark LN, Hewitt JE. The FSHD-associated repeat, D4Z4, is a
8. Scionti I, Fabbri G, Fiorillo C, et al. Facioscapulohumeral muscular dystrophy: new member of a dispersed family of homeobox-containing repeats, subsets of which are
insights from compound heterozygotes and implication for prenatal genetic clustered on the short arms of the acrocentric chromosomes. Genomics. 1995;28(3):
counselling. J Med Genet. 2012;49(3):171-178. doi:10.1136/jmedgenet-2011- 389-397. doi:10.1006/geno.1995.1166
100454 30. Wang Z, Qiu L, Lin M, et al. Prevalence and disease progression of genetically-
9. Tassin A, Laoudj-Chenivesse D, Vanderplanck C, et al. DUX4 expression in FSHD confirmed facioscapulohumeral muscular dystrophy type 1 (FSHD1) in China be-
muscle cells: how could such a rare protein cause a myopathy? J Cell Mol Med. 2013; tween 2001 and 2020: a nationwide population-based study. Lancet Reg Health West
17(1):76-89. doi:10.1111/j.1582-4934.2012.01647.x Pac. 2022;18:100323. doi:10.1016/j.lanwpc.2021.100323
10. Lemmers RJ, Wohlgemuth M, van der Gaag KJ, et al. Specific sequence variations 31. Lemmers R, van der Vliet PJ, Blatnik A, et al. Chromosome 10q-linked FSHD
within the 4q35 region are associated with facioscapulohumeral muscular dystrophy. identifies DUX4 as principal disease gene. J Med Genet. 2022;59(2):180-188. doi:
Am J Hum Genet. 2007;81(5):884-894. doi:10.1086/521986 10.1136/jmedgenet-2020-107041
11. Lemmers RJ, van der Vliet PJ, Klooster R, et al. A unifying genetic model for 32. Lemmers RJ, van der Vliet PJ, van der Gaag KJ, et al. Worldwide population analysis of
facioscapulohumeral muscular dystrophy. Science. 2010;329(5999):1650-1653. the 4q and 10q subtelomeres identifies only four discrete interchromosomal sequence
doi:10.1126/science.1189044 transfers in human evolution. Am J Hum Genet. 2010;86(3):364-377. doi:10.1016/
12. Hamanaka K, Šikrová D, Mitsuhashi S, et al. Homozygous nonsense variant in LRIF1 j.ajhg.2010.01.035
associated with facioscapulohumeral muscular dystrophy. Neurology. 2020;94(23): 33. Lemmers RJ, van den Boogaard ML, van der Vliet PJ, et al. Hemizygosity for
e2441-e2447. doi:10.1212/WNL.0000000000009617 SMCHD1 in facioscapulohumeral muscular dystrophy type 2: consequences for 18p
13. van den Boogaard ML, Lemmers R, Balog J, et al. Mutations in DNMT3B modify deletion syndrome. Hum Mutat. 2015;36(7):679-683. doi:10.1002/humu.22792
epigenetic repression of the D4Z4 repeat and the penetrance of facioscapulo- 34. Lemmers RJ, van der Vliet PJ, Balog J, et al. Deep characterization of a common D4Z4
humeral dystrophy. Am J Hum Genet. 2016;98(5):1020-1029. doi:10.1016/ variant identifies biallelic DUX4 expression as a modifier for disease penetrance in
j.ajhg.2016.03.013 FSHD2. Eur J Hum Genet. 2018;26(1):94-106. doi:10.1038/s41431-017-0015-0
14. Lemmers RJ, Tawil R, Petek LM, et al. Digenic inheritance of an SMCHD1 mutation 35. Renard D, Taieb G, Garibaldi M, et al. Inflammatory facioscapulohumeral muscular
and an FSHD-permissive D4Z4 allele causes facioscapulohumeral muscular dystrophy dystrophy type 2 in 18p deletion syndrome. Am J Med Genet A. 2018;176(8):
type 2. Nat Genet. 2012;44(12):1370-1374. doi:10.1038/ng.2454 1760-1763. doi:10.1002/ajmg.a.38843

Estimated Familial Amyotrophic Lateral Sclerosis

Proportion
A Literature Review and Meta-analysis
Julie Barberio, PhD, Cathy Lally, MSPH, Varant Kupelian, PhD, Orla Hardiman, MD, and Correspondence
Dr. Kupelian
W. Dana Flanders, MD, DSc
varant.kupelian@biogen.com
Abstract
Amyotrophic lateral sclerosis (ALS) is a rare neurodegenerative disorder. Familial (fALS) cases
are usually reported to constitute 5%–10% of all ALS cases; however, no recent literature review
or meta-analysis of this proportion (referred to throughout as “proportion fALS”) has been
conducted. Our objective was to estimate the proportion fALS by geographic region and to
assess the effect of study characteristics on the estimates.
Methods
A comprehensive literature review was performed to identify all original studies reporting the
number of fALS cases in an ALS cohort. The results were stratified by geographic region, study
design (case series or population-based), and decade of study publication. Subgroup analyses
were conducted according to family history criteria used to define fALS. We report pooled
estimates of the proportion fALS from random-effects meta-analyses when >2 studies are
available and I2 is < 90%; weighted averages and ranges are otherwise presented.
Results
The overall pooled proportion fALS based on a total 165 studies was 8% (0%, 71%). The
proportion fALS was 9% (0%, 71%) among 107 case series and 5% (4%, 6%) among 58
population-based studies. Among population-based studies, proportion fALS by geographic
region was 6% (5%, 7%; N = 37) for Europe, 5% (3%, 7%; N = 5) for Latin America, and 5%
(4%, 7%; N = 12) for North America. Criteria used to define fALS were reported by 21
population-based studies (36%), and proportion fALS was 5% (4%, 5%; N = 9) for first-degree
relative, 7% (4%, 11%; N = 4) for first or second-degree relative, and 11% (N = 1) for more
distant ALS family history. Population-based studies published in the 2000s or earlier generated
a lower pooled proportion fALS than studies published in the 2010s or later.
Discussion
The results suggest that variability in the reported proportion fALS in the literature may be, in
part, due to the differences in geography, study design, fALS definition, and decade of case
ascertainment. Few studies outside of European ancestral populations were available. The
proportion fALS was marginally higher among case series compared with population-based
studies, likely because of referral bias. Criteria used to define fALS were largely unreported.
Consensus criteria for fALS and additional population-based studies in non-European ancestral
populations are needed.
From the Epidemiologic Research and Methods LLC (J.B., C.L., W.D.F.); Rollins School of Public Health (J.B., W.D.F.), Emory University, Atlanta, GA; Biogen (V.K.), Cambridge, MA; and
Trinity Biomedical Sciences Institute (O.H.), Dublin, Ireland.
The Article Processing Charge was funded by Biogen.

Glossary
ALS = amyotrophic lateral sclerosis; fALS = familial amyotrophic lateral sclerosis.
Introduction considered with at least 2 affected family members or clear

evidence of genetic inheritance, reserving “probable fALS” for
Amyotrophic lateral sclerosis (ALS) is a neurodegenerative those with one affected first or second-degree relative.12
disorder characterized by the progressive deterioration of “Possible fALS” allows for ALS family history in more distant
motor neurons, which affects voluntary movement. Based on relatives.12 Notwithstanding, there is currently no consensus
summary meta-analyses across the literature, ALS is estimated regarding a preferred definition of fALS for use in epidemio-
to be newly diagnosed in 2.02 (95% confidence interval [CI] logic studies.21 As a result, studies may define fALS based on
1.76–2.31) cases per 100,000 person-years.1 The incidence family history of ALS only or may allow for family history of
rate (per 100,000 person-years) varies substantially according alternative neurodegenerative disorders or incorporate genetic
to region: 2.35 (95% CI 1.75–3.15) for North America, 2.31 testing, introducing difficulties for synthesizing the literature.
(95% CI 2.08–2.55) for Europe, 1.25 (95% CI 0.54–2.89) for Moreover, it has been reported that less than 10% of studies
Latin America, and 0.93 (95% CI 0.57–1.51) for Asia.1 reporting the proportion fALS provide a definition of the cri-
teria used to differentiate fALS and sALS.3 It is vital to consider
Most ALS cases (90%–95%) are believed to occur sporadically whether studies provide clear definitions of fALS and what
(sporadic ALS [sALS]) while 5%–10% report known family those definitions are, to better understand the true population
history of the disease (familial ALS [fALS]).2 Extensive het- fALS proportion.
erogeneity in study designs across the literature has lessened
the ability to confidently reach conclusions about the pro- In addition to discrepancies in how fALS is defined, the
portion fALS among patients with ALS. A 2011 meta-analysis estimated fALS proportions in the literature may also vary
of 34 studies found the pooled proportion fALS to be 4.6% because of study design features. The proportion fALS
(95% CI 3.9–5.5).3 Despite important variation in ALS in- from meta-analytic estimates has been reported to be
cidence according to geographic region, pooled estimates of higher among studies with recruitment based on a se-
fALS proportion were not presented according to geographic quential series of cases (5.1%, 95% CI 3.4–7.1) compared
region. In the decade since the publication of this meta-analysis, with population-based studies (4.5%, 95% CI 3.8–5.3).3 It
studies have reported higher proportions of fALS (7%–17%), has been suggested that cohorts of patients with ALS drawn
emphasizing the need for further research.4-11 Understanding from hospital databases may be subject to referral bias
the true proportion of ALS cases with reported family histories, because patients in these settings are more likely to have
and the reasons for variation in such estimates reported in the affected family members than the general population.22
literature, may contribute to an improved understanding of the Consideration of patient recruitment methods is essential
genetic etiology of ALS and the usefulness of the fALS vs for understanding heterogeneity in estimates of proportion
sALS classification system. Distinguishing fALS vs sALS is fALS.
not possible based on clinical presentation alone and instead
relies on detailed family history. Even with this information, There may also be a temporal trend in the estimated fALS
false reporting can occur for several reasons, including proportions in the literature, for example, due to improve-
misdiagnosis and early death of relatives who would have ments in ALS case ascertainment, diagnostic criteria, and
developed ALS.12-14 Moreover, the binary classification of fALS classification over time.14,23-25 In addition, distributions
fALS vs sALS ignores the complexities of ALS genetics.15 For of population age and environmental risk factors and average
example, low gene penetrance and recessive transmission that family size may have changed over time, which are expected to
may occur in some ALS-associated mutations can result in the affect observed fALS proportion.17,23-25 These considerations
apparent lack of family history.12,13,15,16 Furthermore, the suggest the importance of considering the period that a
probability of a mutation-carrying family having only one or study’s fALS proportion represents.
none of its members affected, given less than complete pene-
trance, is dependent on family size.17 These situations may The objective of this comprehensive literature review and
result in erroneous labeling of patients as sporadic, when family meta-analysis was to estimate the proportion of ALS cases that
inheritance occurred. Alternatively, classification of cases as are familial (henceforth referred to as “proportion fALS”) by
familial in the absence of genetic inheritance can occur if geographic region. We explore variation in the proportion
multiple family members are affected because of random fALS because of study design (population-based registry or
chance or shared environmental risk factors (e.g., heavy met- case series), fALS definition (family history of ALS in a first-
als), although this evidence remains inconclusive.18-20 For this degree, second-degree, or more distant relative), and publi-
reason, it has been proposed that “definite fALS” should be cation decade.

Methods If a region included 3 or more eligible studies reporting the
proportion fALS, a meta-analytic fALS summary proportion
Literature Review and the corresponding 95% confidence interval were calcu-
We initiated our sample by including all studies identified lated. Meta-analyses were conducted according to random-
in the 2011 meta-analysis.3 These studies were identified effects models, which were fit using the R package “metafor”
through a MEDLINE search from 1966 to October 2009 with with restricted maximum-likelihood estimators. Random-effect
the following MeSH terms: ‘ALS,’ ‘amyotrophic lateral scle- meta-analytic estimates are presented for regions for which
rosis,’ ‘fALS,’ ‘familial amyotrophic lateral sclerosis,’ ‘familial the I2 heterogeneity statistic value was found to be <90%;
motor neuron(e) disease,’ ‘motor neuron(e) disease,’ ‘MND,’ otherwise, weighted averages and ranges are presented as a
‘incidence,’ ‘prevalence,’ and ‘mortality.’ To update this study descriptive summary because of substantial heterogeneity.26 If a
list, PubMed and EMBASE were searched from January 1990 region included 2 eligible studies reporting the proportion
to August 2021 by a medical librarian using identical search fALS in a defined cohort, a weighted average of the estimates
terms. Additional sources within the same scope were sought was calculated for the pooled estimate and the range is pre-
from references of the identified articles. sented as the interval. Otherwise, if a given region only had a
single eligible study, the single study’s estimate is presented
Studies eligible for inclusion were those presenting original data without an interval.
on a defined ALS cohort that adequately described enrollment
methods and included enough information, published in the Because we suspected that important study characteristics
main text or supplemental, to calculate the proportion fALS at might affect the estimated proportion fALS, we stratified our
the individual level (i.e., studies presenting only counts of fALS results based on study design (population-based or case se-
pedigrees in a population were excluded). Estimation of the ries), family history criteria used to define fALS, and publi-
proportion fALS did not necessarily need to be the objective cation decade. A study was considered population-based if the
of the study. Study-level ALS diagnostic criteria may have procedures were expected to capture all ALS cases over a
included other motor neuron diseases (e.g., spinal muscular specified period in a defined population. Studies that other-
atrophy, progressive muscular atrophy, and primary lateral wise recruited a collection of cases, such as a clinic or hospital-
sclerosis); sensitivity analyses in which these studies were ex- based series of cases, were classified as case series. Studies that
cluded are described in more detail below. Abstracts and un- did not explicitly report the criteria used to define fALS, in-
published studies were not included. Articles published in cluding those that stated a requirement for “family history”
languages other than English were translated. Contact with without any further detail regarding the degree of relatives
corresponding authors was attempted for articles with ambi- considered, were considered not clear and were excluded from
guity in the proportion fALS that were otherwise eligible for the family history–defined subgroup analysis. The remaining
inclusion. studies were grouped according to whether the fALS defi-
nition was based on family history of ALS only, allowed for
Studies were excluded for the following reasons: (1) They family history of alternative neurodegenerative disorders
lacked sufficient description of participant enrollment to allow (e.g., frontotemporal dementia), or incorporated confirmed
for determination of whether enrollment was population-based genetic diagnoses. Studies that operationalized fALS based
or plausibly based on a series of clinic or hospital-based cases, on family history of ALS alone were further divided based on
(2) they based enrollment on selection of a special group of the degree of ALS-affected relatives: (1) first-degree, (2) first
patients with ALS (e.g., patients with fALS, juvenile ALS, and or second-degree, or (3) first or second-degree or more
genetic mutations), (3) they followed separate procedures to distant. First-degree relatives include parents, full siblings,
recruit fALS and sALS cases and would not be expected to and children. Second-degree relatives include grandparents,
represent an accurate ratio of fALS to sALS in an underlying grandchildren, uncles, aunts, nephews, nieces, and half sib-
source population, and (4) they obtained data on ALS cases lings. Subgroup pooled estimates are only presented for the
from a biobank because it is expected that the fALS distribution definitions based on degree of family history of ALS, partially
may not be representative of the population-level distribution. because of substantial variability in estimates among the
other subgroups. Publication decade was used as a proxy for
Analytic Approach the time of case diagnosis. Studies published before 1990
Details regarding case ascertainment, diagnostic criteria, region, were included in a single group because of the small number
and fALS definition were extracted by a single researcher and of studies from each prior decade.
independently reviewed by a second researcher; discrepancies
were resolved by consensus. In each study, the proportion fALS Meta-regression
is reported as the number of fALS cases among all ALS cases. To further explore the potential sources of heterogeneity
Studies were grouped according to region to produce pooled described above, we conducted meta-regression analyses.
estimates of the proportion fALS. We report a summary for Meta-regression applies regression techniques to study-level
each region detailing the included studies with their countries data to discern whether a linear relationship exists between
of origin, total number of ALS cases, and total number of fALS the reported outcome measure and the covariate(s) of in-
cases in eTable 1 (links.lww.com/NXG/A646). terest. Meta-regressions were conducted using generalized

linear mixed-effects models, using logit transformation to model studies not reporting data for fALS, and case studies. The
proportion fALS with a random intercept for each study. The remaining 366 articles underwent in-depth full-text review. The
models were fit using the R package “lme4.” Analyses were final analytic sample included 165 study estimates (from 160
conducted among all studies and then repeated among articles), 58 (35%) of which identified cases based on
population-based studies. Univariate models separately in- population-based methods and 107 (65%) based on case series.
cluded study type, region, family history criteria used to Details of the included studies are included in eTable 1.
define fALS, and publication decade as fixed effects. We
additionally explored average family size as a fixed predictor, Of the 56 studies that reported a clear definition of their
which was estimated based on average fertility rate (number criteria for fALS, 13 (23%) restricted family history of ALS
of children per woman) in the country of publication 20 to first-degree relatives only, 14 (25%) restricted to first or
years before the end of study data collection.27 Multivariate second-degree relatives, and 8 (14%) allowed for more
models included multiple fixed effects. We quantitatively distant generations. In addition, 8 studies (14%) allowed pa-
described the variance explained by the fixed predictor(s) by tients with confirmed genetic mutations to be categorized
calculating the percent change in tau2 estimate of between- as fALS and 1 study (2%) required family history of ALS in
study variance when adding the predictor(s) to the model multiple relatives. Finally, 12 studies (21%) also considered
with no covariates. family history of other neurologic diseases in the categorization
of fALS.
Sensitivity Analysis
In the main analyses, ALS diagnostic criteria may include other Meta-analysis
motor neuron diseases (e.g., spinal muscular atrophy, pro- The details regarding number of studies by region and the
gressive muscular atrophy, primary lateral sclerosis, and pro- pooled estimate with its corresponding interval (lower limit and
gressive supranuclear palsy), cases were required to be reported upper limit from random-effects meta-analysis when >2 studies
as either fALS or sALS, and fALS classification criteria may are available and I2 < 90%; otherwise minimum and maximum)
include confirmed ALS-associated genetic variants. We con- are provided in Table 1. All meta-analytic estimates and I2 values
ducted a sensitivity analysis in which more stringent criteria to are reported in eTable 2 (links.lww.com/NXG/A646), and
define ALS were applied, such that analyses were restricted to forest plots are available in eFigure 1 (case series) and eFigure 2
studies only including “pure ALS” cases (i.e., no diagnoses for (population-based). The overall summary proportion fALS
alternative motor neuron diseases). A second sensitivity analysis across all 165 studies was 0.08 (interval 0.00, 0.71). The pro-
was conducted in which study-level estimates were altered such portion fALS varied according to region. Europe had the highest
that all cases of ALS without reported fALS or sALS status were proportion of fALS (0.09, interval 0.00, 0.71), followed by the
included as sALS cases. A third sensitivity analysis was con- Middle East and North Africa (0.09, interval 0.01, 0.40) and
ducted in which we excluded studies allowing patients with Australasia (0.08, interval 0.04, 0.16). The lowest proportion
confirmed genetic mutations to be categorized as fALS given the fALS was observed in Asia (0.04, interval 0.03, 0.06) and Sub-
shared genetic architecture of fALS and sALS.28,29 Saharan Africa (0.05, interval 0.03, 0.07). Substantial variability
in estimates of proportion fALS remained at the regional level,
Data Availability with most I2 values being >90%.
Data not provided in the article because of space limitations
may be shared at the request of any qualified investigator for When stratified by study design, the proportion fALS was
purposes of replicating procedures and results. higher among case series (0.09, interval 0.00, 0.71) compared
with population-based studies (0.05, interval 0.04, 0.06). This
observation was consistent at the regional level (Table 1). For
Results most regions, a meaningful difference in the I2 value from case
series vs population-based studies was not observed. However,
Literature Review substantially greater variability was observed in case series vs
Byrne et al. described 34 studies eligible for inclusion and population-based studies from North America (88% vs 74%)
published between 1966 and October 2009.3 We included 31 and from Latin America (88% vs 18%).
of these studies in our analytic sample. The 3 remaining studies
were excluded because of lack of information regarding family Owing to the observed variability in the proportion fALS
history or overlap with data from a more recent study.30-32 according to study design, subgroup estimates according to
family history criteria used to define fALS were only computed
Our comprehensive literature search identified 6,816 articles. among population-based studies (Table 2). As expected,
An additional 150 sources were identified from review of population-based studies defining fALS according to first or
identified articles for a total of 6,966 studies. Those deemed second-degree or more distant family history of ALS generated
ineligible based on a review of titles and abstracts were ex- a higher pooled proportion fALS (0.11, based on a single
cluded (N = 6,450). Exclusions at this stage included review study) compared with studies restricting to family history
articles, studies not involving human subjects and/or ALS within the first or second degree (0.07, interval 0.04, 0.11) or
populations, studies restricted to either fALS or sALS cases, only the first degree (0.05, interval 0.04, 0.05). We present

Table 1 Pooled Proportion of ALS Cases That Are Familial (fALS), According to Geographic Region and Study Type
Region Studies fALS cases Total ALS cases Estimate (lower limit, upper limit)
Asia 24 332 7,748 0.04 (0.03, 0.06)a
Case series 23 331 7,664 0.05 (0.03, 0.06)a
Population-based 1 1 84 0.01 (NA, NA)
Australasia 4 30 432 0.08 (0.04, 0.16)a
Case series 3 20 188 0.12 (0.08, 0.17)a
Europe 77 2,783 30,553 0.09 (0.00, 0.71)
Case series 40 1,848 16,730 0.11 (0.00, 0.71)
Population-based 37 935 13,823 0.06 (0.05, 0.07)a
Latin America 17 275 2,518 0.06 (0.04, 0.09)a
Case series 12 237 1,708 0.07 (0.05, 0.12)a
Population-based 5 38 810 0.05 (0.03, 0.07)a
Middle East and North Africa 10 159 1,832 0.09 (0.01, 0.40)
Case series 9 154 1,750 0.09 (0.01, 0.40)
North America 28 975 16,989 0.06 (0.05, 0.08)a
Case series 16 479 6,705 0.08 (0.06, 0.10)a
Population-based 12 496 10,284 0.05 (0.04, 0.06)a
Sub-Saharan Africa 5 23 514 0.05 (0.03, 0.07)a
Case series 4 15 329 0.05 (0.03, 0.09)a
Overall 165 4,577 60,586 0.08 (0.00, 0.71)
Case series 107 3,084 35,074 0.09 (0.00, 0.71)
Population-based 58 1,493 25,512 0.05 (0.04, 0.06)a
Abbreviation: ALS = amyotrophic lateral sclerosis; NA = not applicable.

a
Random-effects meta-analytic estimates and corresponding 95% confidence intervals are presented (>2 and I2 < 90%). Otherwise, weighted averages and
ranges are presented.
regional-level fALS proportions according to family history proportions according to publication decade for Europe and
criteria used to define fALS for Europe and North America only North America only because these were the only regions with
because these were the only regions with population-based population-based studies in all decades. Heterogeneity was
studies in the 3 categories of degree of family history. Grouping substantially reduced when studies were grouped by publication
studies by family history criteria used to define fALS resulted in decade, as evidenced by most I2 values being <65%. Hetero-
substantially reduced heterogeneity, as evidenced by all I2 val- geneity remained considerable among studies published in the
ues being <50%. 2010s, which was the decade during which most population-
based studies (50%) were published.
Subgroup estimates according to publication decade were only
computed among population-based studies because of ob- Meta-regression
served variability in proportion fALS according to study design Detailed results of the meta-regression analyses are presented
(Table 3). Population-based studies published in the 2000s or in eTable 3 (links.lww.com/NXG/A646) (all studies) and
earlier generated a lower pooled proportion fALS than studies eTable 4 (population-based). Study type, region, family his-
published in the 2010s or later (1990s: 0.03, interval 0.01, 0.07; tory definition, publication decade, and average family size
2020s: 0.09, interval 0.06, 0.12). We present regional-level fALS each explained <10% of between-study heterogeneity among

Table 2 Pooled Proportion of ALS Cases That Are Familial (fALS), According to Geographic Region and Family History
Criteria Used to Define fALS, Among Population-Based Studies
Degree of ALS-affected relative used to define fALS Population-based studies fALS cases Total ALS Estimate (lower limit, upper limit)
Europe
First-degree 4 18 424 0.05 (0.03, 0.07)a
First or second-degree 3 16 194 0.09 (0.05, 0.15)a
First or second-degree or more distant 1 50 444 0.11 (NA, NA)
North America
First-degree 4 324 6,985 0.05 (0.04, 0.05)a
First or second-degree 1 49 946 0.05 (NA, NA)
First or second-degree or more distant 0 NA NA NA
Overall
First-degree 9 344 7,512 0.05 (0.04, 0.05)a
First or second-degree 4 65 1,140 0.07 (0.04, 0.11)a
First or second-degree or more distant 1 50 444 0.11 (NA, NA)

a
Random-effects meta-analytic estimates and corresponding 95% confidence intervals are presented (>2 and I2 < 90%). Otherwise, weighted averages and
all studies. Among population-based studies, family history was observed to be 9% according to studies in which partic-
definition and publication decade together explained sub- ipant recruitment was based on a clinic or hospital-based se-
stantial between-study heterogeneity, in that the tau2 was re- ries of cases, but was only 5% according to population-based
duced by 24%. studies. Notably, these overall pooled results are driven by the
large number of publications (47%) with data derived from
Sensitivity Analysis Europe. Notwithstanding, when examining pooled estimates
In the main analysis, ALS cases with missing/unknown family at the regional level, a higher proportion fALS among case
history information were excluded from 10 studies. Results series vs population-based studies was consistently observed.
from a sensitivity analysis in which these cases were reincluded Meaningful difference in meta-analytic estimates of the pro-
as sALS were consistent with the main analysis. Similarly, a portion fALS according to study design has previously been
sensitivity analysis that excluded studies in which ALS di- described, although the reported difference (5.1% for case
agnostic criteria allowed for diagnosis of alternative motor series vs 4.5% for population-based) was not as pronounced as
neuron diseases also produced results consistent with the main we found.3 Population-based studies (which make up the mi-
analysis. Finally, a sensitivity analysis in which 8 studies nority of this literature) are expected to more accurately capture
allowing patients with confirmed genetic mutations to be the true proportion of ALS cases that are familial, given that a
classified as fALS were excluded produced slightly lower overall series of cases from clinics or hospitals may be inadvertently
proportion fALS estimates among all studies (0.07 vs 0.08) and enriched by fALS cases.22 Evidence from population-based
case series (0.08 vs 0.09), driven by the change in European registers, however, should not be used without careful consid-
studies (0.08 vs 0.09 for all studies; 0.10 vs 0.11 for case series); eration of potential biases (e.g., shifts in demography, increased
heterogeneity remained >90%. Population-based and other awareness, “startup bias” in newly established registers, and
regional results remained unchanged. “information creep” in registers of longer duration).33,34
Our study also explores global geographic variability in the

proportion of fALS. Substantial variability across regions was
Discussion observed, providing support for potential differences in un-
The proportion of ALS cases that are of familial, rather than derlying genetic structure, distribution of environmental fac-
sporadic, origin is commonly cited as 5%–10%. Our analysis tors, clinical practices related to fALS assessment, and average
suggests that observed variability in the reported proportion family size in these populations.17,35,36 Among population-
fALS in the literature may be, in part, due to differences in based studies, the proportion fALS was highest for Europe
region, study design, definition of fALS, and decade of case (6%) and the Middle East and North Africa (6%), followed by
ascertainment. The pooled proportion fALS among ALS cases North America (5%) and Latin America (5%). It is important

Table 3 Pooled Proportion of ALS Cases That Are Familial (fALS), According to Geographic Region and Publication
Decade, Among Population-Based Studies
Publication decade Population-based studies fALS cases Total ALS cases Estimate (lower limit, upper limit)
Europe
Pre-1990s 4 22 415 0.06 (0.04, 0.09)a
1990s 0 NA NA NA
2000s 5 98 2,134 0.04 (0.02, 0.06)a
2010s 21 732 10,502 0.07 (0.00, 0.23)
2020s 6 82 714 0.10 (0.07, 0.15)a
North America
Pre-1990s 1 6 139 0.04 (NA, NA)
1990s 3 58 936 0.04 (0.01, 0.10)a
2000s 1 1 21 0.05 (NA, NA)
2010s 6 409 8,950 0.05 (0.04, 0.05)a
2020s 1 22 238 0.09 (NA, NA)
Overall
Pre-1990s 5 28 554 0.05 (0.04, 0.08)a
1990s 5 60 1,078 0.03 (0.01, 0.07)a
2000s 10 145 3,074 0.05 (0.04, 0.06)a
2010s 29 1,143 19,587 0.06 (0.00, 0.23)
2020s 9 117 1,219 0.09 (0.06, 0.12)a

a
Random-effects meta-analytic estimates and corresponding 95% confidence intervals are presented (>2 and I2<90%). Otherwise, weighted averages and
to note that our population-based estimate for Europe is based members or clear evidence of genetic inheritance.12,21 “Pos-
on more available literature (37 studies) than Latin American sible fALS” may also incorporate cases with a first-degree
(5 studies) and the Middle East and North Africa (1 studies). relative with frontotemporal dementia because of the overlap
Furthermore, the Latin American estimates may be affected by in phenotype and genotype of these disorders.12,43-45 Addi-
founder effects, which have been described in the literature in tional neuropsychiatric disorders (e.g., all-type dementia and
this region for various neurodegenerative disorders.37-39 The schizophrenia) are also genetically linked to ALS, suggesting
lowest pooled, population-based proportion fALS was from that incorporation of these disorders in an extended fALS
Asia (1%), although only one study was available. Lower in- definition may be important for capturing familial aggregation
cidence and prevalence rates of ALS in Asia compared with related to ALS.14,46 Moreover, it has been suggested that the
Europe and North America have previously been reported, binary classification of ALS cases as fALS vs sALS is an “over-
which may affect ALS incidence within families.1,40-42 Fur- simplification” because of the complexities of genetic pleiotropy,
thermore, because the proportion fALS is dependent on family as well as oligogenic and polygenic inheritance patterns that have
size, it is worth noting that this population-based estimate is been documented in ALS, including in apparently sporadic
derived from China and expected to be affected by China’s cases.47 Even in cases with familial inheritance, incomplete gene
historic one-child policy.17 penetrance and recessive transmission may result in the apparent
lack of family history.12,13,16 We observed that the minority
Our analysis also demonstrates that variation in the reported (34%) of studies provided a clear fALS. Approximately 60% of
proportion fALS is partly attributable to study-level differ- these studies based their definition on family history of ALS
ences in the operational definition of fALS. There is currently within an explicitly stated number of generations while the
no consensus regarding a preferred definition of fALS among remaining allowed for family history of other neurologic diseases
clinicians, although it has been suggested that the optimal or confirmed genetic mutations. To our knowledge, this is the
classification system should reserve naming a “definite” fALS first study to comprehensively examine the proportion fALS
case based on the presence of at least 2 affected family according to a gradient of family history criteria used to define

fALS in the literature. Although it may have been preferable to disorders in the cohort on the proportion fALS, we conducted a
examine the proportion fALS according to the “definite,” sensitivity analysis in which we restricted our analysis to studies
“probable,” or “possible” categorizations described above, the that explicitly reported the exclusion of all other diagnoses from
level of detail provided in the literature did not allow for this the ALS case group. Although we did not observe a meaningful
analysis. We observed, as expected, that the pooled proportion of difference in the results, we recognize that detailed information
fALS among population-based studies using the most stringent on ALS diagnostic criteria was not always reported by authors,
family history criteria was substantially lower compared with and therefore, discrimination of studies based on this feature
those using more lenient family history criteria. We observed, was not always possible.
importantly, that I2 heterogeneity statistic values were sub-
stantially reduced for subgroups based on the family history We also observed a lack of clarity regarding inclusion of
criteria compared with all population-based studies, suggesting multiple cases from a given fALS pedigree. In some studies,
this as a potentially important source of heterogeneity. the authors referred to fALS cases as being “unrelated,” but it
was not necessarily clear whether this was an intentional
Our analysis also demonstrates variation in the proportion feature of recruitment or an incidental occurrence. Similarly,
fALS in the literature according to publication decade, which some studies referred to cases as “index cases,” a term typically
has not previously been described. We observed a positive used in genetics literature to refer to the first affected case in a
temporal trend in proportion fALS, likely due to changes in family, but it is unclear whether use of the term was consistent
ALS case ascertainment and diagnostic criteria, enhanced dis- with this meaning. We found that most studies (60%) did not
ease understanding, fALS classification (including changes the comment on whether fALS cases were related. We acknowl-
classification criteria themselves, as well as changes in incidence edge this as a potentially meaningful source of variability in
and recognition of family histories of ALS-related phenotypes), the literature and expect that those studies that restrict to one
distribution of population age and environmental risk factors, individual with fALS per family would underestimate the true,
and average family size over time14,17,23-25,34 Collapsing across individual-level population proportion fALS.
several decades of published studies on fALS proportion to
create a single summary estimate may, therefore, not be ap- In addition to these concerns, additional unrecognized
propriate.14 It is important to note that publication decade was study flaws could have biased or affected representativeness
used as a proxy for period of case ascertainment because many of study-level estimates of the proportion fALS, which may
studies did not provide these details. For studies that did pro- have affected our summary estimates. For the purposes of
vide information on case ascertainment years, case diagnosis this study, we included all studies that did not explicitly
often spanned multiple decades, but fALS proportion in- report any method that would affect representativeness rel-
formation was not presented with the level of granularity that ative to the total population of patients with ALS. It is
would allow stratification by time. possible, however, that some studies may have failed to
disclose certain recruitment features that hinder represen-
This analysis has several limitations. We observed meaningful tativeness. It is also important to note that current un-
variability within regions, even after accounting for study design derstanding of fALS is dependent on current practices of
methods, as evidenced by many I2 values being greater than reporting and genetic testing, which may change over time.
90%. Random-effects meta-analyses and weighted averages
were calculated as central values, causing us to collapse across Despite these recognized limitations, this study contributes to
potentially meaningful within-region variability. It is important an improved understanding of factors that affect variability in
to note that, in the presence of substantial between-study reports of the proportion of ALS cases that are of familial vs
heterogeneity, weighted averages are presented as a descriptive sporadic origin in the epidemiologic literature, namely geo-
summary of our findings and do not necessarily represent graphic region, study design, operational definition of fALS,
any population-based measure. Further variations in study and publication decade. Future identification of ALS cases,
features, beyond study design and fALS definition features especially fALS cases, with an underlying genetic etiology may
examined here, may have contributed to observed variation benefit from increased genetic testing to address limitations in
in the proportion fALS. Unfortunately, we were limited by estimating proportion fALS based on unstandardized family
the shortcomings of the literature, which included a lack of history information alone.
detailed reporting on additional features that may have af-
fected proportion fALS variability. For example, diagnostic Study Funding
criteria varied widely across studies, such that some studies Biogen.
defined ALS cases according to the El Escorial criteria, to
varying degrees (i.e., definite only; definite or probable; defi- Disclosure
nite, probable, or suspect), while others used the Awaji criteria J. Barberio: employee of Epidemiologic Research & Methods,
or their own institution-defined criteria.48-50 In addition, we LLC, doctoral stipend and tuition are supported by an award
observed inconsistency in whether investigators excluded all, or to Emory University from Amgen Inc; C. Lally: employee of
some subset of, other neuromuscular disorders. To investigate Epidemiologic Research and Methods; V. Kuplian: employee
the effect of including patients with other neuromuscular of and holds stock/stock options in Biogen; O. Hardiman:

Science Foundation Ireland grants SFI 16/RC/3948 and 11. Tarlarini C, Lunetta C, Mosca L, et al. Novel FUS mutations identified through
molecular screening in a large cohort of familial and sporadic amyotrophic lateral
20/SP/8953, consulting fees from Cytokinetics, Wave Phar- sclerosis. Eur J Neurol. 2015;22(11):1474-1481. doi:10.1111/ene.12772
maceuticals, Orion, Biogen, Denali, Novartis and Accelsior, 12. Byrne S, Bede P, Elamin M, et al. Proposed criteria for familial amyotrophic lateral sclerosis.
Amyotroph Lateral Scler. 2011;12(3):157-159. doi:10.3109/17482968.2010.545420
Editor-in-Chief of the journal ALS and Frontotemporal 13. Belzil VV, Rouleau GA. Familial ALS: less common than we think? J Neurol Neurosurg
Degeneration; W.D. Flanders: employee of Epidemiologic Psychiatry. 2012;83(12):1133-1133. doi:10.1136/jnnp-2012-303127
14. Ryan M, Heverin M, Doherty MA, et al. Determining the incidence of familiality in
Research & Methods, LLC, employee of Rollins School of ALS: a study of temporal trends in Ireland from 1994 to 2016. Neurol Genet. 2018;
Public Health. Go to Neurology.org/NG for full disclosures. 4(3):e239. doi:10.1212/nxg.0000000000000239
15. Goutman SA, Hardiman O, Al-Chalabi A, et al. Emerging insights into the complex
genetics and pathophysiology of amyotrophic lateral sclerosis. Lancet Neurol. 2022;
Publication History 21(5):465-479. doi:10.1016/s1474-4422(21)00414-2
Received by Neurology: Genetics May 8, 2023. Accepted in final form 16. Camu W, Khoris J, Moulard B, et al. Genetics of familial ALS and consequences for
diagnosis. French ALS Research Group. J Neurol Sci. 1999;165(suppl 1):S21-S26. doi:
September 22, 2023. Submitted and externally peer reviewed. The 10.1016/S0022-510X(99)00022-2
handling editor was Associate Editor Raymond P. Roos, MD, FAAN. 17. Al-Chalabi A, Lewis CM. Modelling the effects of penetrance and family size on rates
of sporadic and familial disease. Hum Hered. 2011;71(4):281-288. doi:10.1159/
000330167
18. Belbasis L, Bellou V, Evangelou E. Environmental risk factors and amyotrophic lateral
sclerosis: an umbrella review and critical assessment of current evidence from sys-
tematic reviews and meta-analyses of observational studies. Neuroepidemiology. 2016;
46(2):96-105. doi:10.1159/000443146
Appendix Authors 19. Farace C, Fenu G, Lintas S, et al. Amyotrophic lateral sclerosis and lead: a systematic
update. Neurotoxicology. 2020;81:80-88. doi:10.1016/j.neuro.2020.09.003
Name Location Contribution 20. Al-Chalabi A, Hardiman O. The epidemiology of ALS: a conspiracy of genes, environ-
ment and time. Nat Rev Neurol. 2013;9(11):617-628. doi:10.1038/nrneurol.2013.203
Julie Epidemiologic Research and Major role in the acquisition 21. Byrne S, Elamin M, Bede P, Hardiman O. Absence of consensus in diagnostic criteria
Barberio, Methods LLC; Rollins School of data; study concept or for familial neurodegenerative diseases. J Neurol Neurosurg Psychiatry. 2012;83(4):
PhD of Public Health, Emory design; analysis or 365-367. doi:10.1136/jnnp-2011-301530
University, Atlanta, GA interpretation of data 22. Logroscino G, Marin B, Piccininni M, et al. Referral bias in ALS epidemiological
studies. PLoS One. 2018;13(4):e0195821. doi:10.1371/journal.pone.0195821
Cathy Epidemiologic Research and Study concept or design; 23. Arthur KC, Calvo A, Price TR, Geiger JT, Chiò A, Traynor BJ. Projected increase in
Lally, Methods LLC, Atlanta, GA analysis or interpretation of amyotrophic lateral sclerosis from 2015 to 2040. Nat Commun. 2016;7(1):12408. doi:
MSPH data 10.1038/ncomms12408
24. Fontana A, Marin B, Luna J, et al. Time-trend evolution and determinants of sex ratio
Varant Biogen, Cambridge, MA Study concept or design; in Amyotrophic Lateral Sclerosis: a dose–response meta-analysis. J Neurol. 2021;
Kupelian, analysis or interpretation of 268(8):2973-2984. doi:10.1007/s00415-021-10464-2
PhD data 25. Tobin K, Gilthorpe MS, Rooney J, et al. Age-period-cohort analysis of trends in
amyotrophic lateral sclerosis incidence. J Neurol. 2016;263(10):1919-1926. doi:
Orla Trinity Biomedical Sciences Drafting/revision of the 10.1007/s00415-016-8215-z
Hardiman, Institute, Dublin, Ireland manuscript for content, 26. Deeks JJ, Higgins JP, Altman DG. 9.5.2 Identifying and measuring heterogeneity. In:
MD including medical writing for Cochrane Handbook Syst Rev Interventions Version 2011;5.
content 27. World Bank. World Development Indicators, Fertility rate, total (births per woman).
Accessed January 15, 2022. data.worldbank.org/indicator/SP.DYN.TFRT.IN.
W. Dana Epidemiologic Research and Study concept or design; 28. Renton AE, Chiò A, Traynor BJ. State of play in amyotrophic lateral sclerosis genetics.
Flanders, Methods LLC; Rollins School analysis or interpretation of Nat Neurosci. 2014;17(1):17-23. doi:10.1038/nn.3584
MD, DSc of Public Health, Emory data 29. Chia R, Chiò A, Traynor BJ. Novel genes associated with amyotrophic lateral scle-
University, Atlanta, GA rosis: diagnostic and clinical implications. Lancet Neurol. 2018;17(1):94-102. doi:
10.1016/s1474-4422(17)30401-5
30. McGuire V, Longstreth W, Koepsell TD, van Belle G. Incidence of amyotrophic
lateral sclerosis in three counties in western Washington state. Neurology. 1996;47(2):
References 571-573. doi:10.1212/wnl.47.2.571
1. Brown CA, Lally C, Kupelian V, Flanders WD. Estimated prevalence and incidence of 31. Chiò A, Mora G, Calvo A, Mazzini L, Bottacchi E, Mutani R, PARALS. Epidemiology
amyotrophic lateral sclerosis and SOD1 and C9orf72 genetic variants. Neuro- of ALS in Italy: a 10-year prospective population-based study. Neurology. 2009;72(8):
epidemiology. 2021;55(5):342-353. doi:10.1159/000516752 725-731. doi:10.1212/01.wnl.0000343008.26874.d1
2. Kiernan MC, Vucic S, Cheah BC, et al. Amyotrophic lateral sclerosis. Lancet. 2011; 32. Del Aguila M, Longstreth W, McGuire V, Koepsell T, Van Belle G. Prognosis in
377(9769):942-955. doi:10.1016/s0140-6736(10)61156-7 amyotrophic lateral sclerosis: a population-based study. Neurology. 2003;60(5):
3. Byrne S, Walsh C, Lynch C, et al. Rate of familial amyotrophic lateral sclerosis: a 813-819. doi:10.1212/01.wnl.0000049472.47709.3b
systematic review and meta-analysis. J Neurol Neurosurg Psychiatry. 2011;82(6): 33. Hardiman O, Al-Chalabi A, Brayne C, et al. The changing picture of amyotrophic
623-627. doi:10.1136/jnnp.2010.224501 lateral sclerosis: lessons from European registers. J Neurol Neurosurg Psychiatry. 2017;
4. Andrew AS, Pioro EP, Li MF, et al. The incidence of amyotrophic lateral sclerosis in 88(7):557-563. doi:10.1136/jnnp-2016-314495
Ohio 2016-2018: the Ohio population-based ALS registry. Neuroepidemiology. 2021; 34. Rooney JPK, Brayne C, Tobin K, Logroscino G, Glymour MM, Hardiman O. Ben-
55(3):196-205. doi:10.1159/000515103 efits, pitfalls, and future design of population-based registers in neurodegenerative
5. Marjanović IV, Selak-Djokić B, Perić S, et al. Comparison of the clinical and cognitive disease. Neurology. 2017;88(24):2321-2329. doi:10.1212/wnl.0000000000004038
features of genetically positive ALS patients from the largest tertiary center in Serbia. 35. Vajda A, McLaughlin RL, Heverin M, et al. Genetic testing in ALS: a survey of current
J Neurol. 2017;264(6):1091-1098. doi:10.1007/s00415-017-8495-y practices. Neurology. 2017;88(10):991-999. doi:10.1212/wnl.0000000000003686
6. Bartoletti-Stella A, Vacchiano V, De Pasqua S, et al. Targeted sequencing panels in 36. Ryan M, Zaldı́var Vaillant T, McLaughlin RL, et al. Comparison of the clinical and
Italian ALS patients support different etiologies in the ALS/FTD continuum. J Neurol. genetic features of amyotrophic lateral sclerosis across Cuban, Uruguayan and Irish
2021;268(10):3766-3776. doi:10.1007/s00415-021-10521-w clinic-based populations. J Neurol Neurosurg Psychiatry. 2019;90(6):659-665. doi:
7. Jericó I, Elizalde-Beiras I, Pagola I, et al. Clinical features and incidence trends of 10.1136/jnnp-2018-319838
amyotrophic lateral sclerosis in Navarre, Spain, 2007-2018: a population-based study. 37. Pineda-Trujillo N, Apergi M, Moreno S, et al. A genetic cluster of early onset Par-
Amyotroph Lateral Scler Frontotemporal Degener. 2021;22(5-6):401-409. doi:10.1080/ kinson’s disease in a Colombian population. Am J Med Genet B Neuropsychiatr Genet.
21678421.2021.1891249 2006;141B(8):885-889. doi:10.1002/ajmg.b.30375
8. Cady J, Allred P, Bali T, et al. Amyotrophic lateral sclerosis onset is influenced by the 38. Acosta-Uribe J, Aguillón D, Cochran JN, et al. A neurodegenerative disease landscape
burden of rare variants in known amyotrophic lateral sclerosis genes. Ann Neurol. of rare mutations in Colombia due to founder effects. Genome Med. 2022;14(1):27.
2015;77(1):100-113. doi:10.1002/ana.24306 doi:10.1186/s13073-022-01035-9
9. Borg R, Farrugia Wismayer M, Bonavia K, et al. Genetic analysis of ALS cases in the 39. Rodrı́guez-Labrada R, Martins AC, Magaña JJ, et al. Founder effects of Spinocerebellar
isolated island population of Malta. Eur J Hum Genet. 2021;29(4):604-614. doi: Ataxias in the American continents and the Caribbean. Cerebellum. 2020;19(3):
10.1038/s41431-020-00767-9 446-458. doi:10.1007/s12311-020-01109-7
10. Scialò C, Novi G, Bandettini di Poggio M, et al. Clinical epidemiology of amyotrophic 40. Chiò A, Logroscino G, Traynor BJ, et al. Global epidemiology of amyotrophic lateral
lateral sclerosis in Liguria, Italy: an update of LIGALS register. Amyotroph Lateral Scler sclerosis: a systematic review of the published literature. Neuroepidemiology. 2013;
Frontotemporal Degener. 2016;17(7-8):535-542. doi:10.1080/21678421.2016.1197942 41(2):118-130. doi:10.1159/000351153

41. Marin B, Boumédiene F, Logroscino G, et al. Variation in worldwide incidence of Genetic correlation between amyotrophic lateral sclerosis and schizophrenia. Nat
amyotrophic lateral sclerosis: a meta-analysis. Int J Epidemiol 2017;46(1):57-74. Commun. 2017;8:14774. doi:10.1038/ncomms14774
doi:10.1093/ije/dyw061 47. Hardiman O, Al-Chalabi A, Chio A, et al. Amyotrophic lateral sclerosis. Nat Rev Dis
42. GBD 2016 Motor Neuron Disease Collaborators. Global, regional, and national burden of Primers. 2017;3:17071. doi:10.1038/nrdp.2017.71
motor neuron diseases 1990-2016: a systematic analysis for the Global Burden of Disease 48. Brooks BR. El Escorial World Federation of Neurology criteria for the diagnosis of
Study 2016. Lancet Neurol. 2018;17(12):1083-1097. doi:10.1016/s1474-4422(18)30404-6 amyotrophic lateral sclerosis. Subcommittee on Motor Neuron Diseases/Amyotrophic
43. Mackenzie IR, Rademakers R, Neumann M. TDP-43 and FUS in amyotrophic lateral Lateral Sclerosis of the World Federation of Neurology Research Group on Neuro-
sclerosis and frontotemporal dementia. Lancet Neurol. 2010;9(10):995-1007. doi: muscular Diseases and the El Escorial "Clinical limits of amyotrophic lateral sclerosis"
10.1016/s1474-4422(10)70195-2 workshop contributors. J Neurol Sci. 1994;124(Suppl):96-107. doi:10.1016/0022-
44. Vance C, Al-Chalabi A, Ruddy D, et al. Familial amyotrophic lateral sclerosis with 510x(94)90191-0
frontotemporal dementia is linked to a locus on chromosome 9p13.2-21.3. Brain. 49. Brooks BR, Miller RG, Swash M, Munsat TL, World Federation of Neurology Re-
2006;129(Pt 4):868-876. doi:10.1093/brain/awl030 search Group on Motor Neuron Diseases. El Escorial revisited: revised criteria for the
45. Renton AE, Majounie E, Waite A, et al. A hexanucleotide repeat expansion in diagnosis of amyotrophic lateral sclerosis. Amyotroph Lateral Scler Other Mot Neuron
C9ORF72 is the cause of chromosome 9p21-linked ALS-FTD. Neuron. 2011;72(2): Disord. 2000;1(5):293-299. doi:10.1080/146608200300079536
257-268. doi:10.1016/j.neuron.2011.09.010 50. Costa J, Swash M, de Carvalho M. Awaji criteria for the diagnosis of amyotrophic
46. McLaughlin RL, Schijven D, van Rheenen W, et al.; Project MinE GWAS Consor- lateral sclerosis:a systematic review. Arch Neurol. 2012;69(11):1410-1416. doi:
tium, Schizophrenia Working Group of the Psychiatric Genomics Consortium. 10.1001/archneurol.2012.254

A Phenotypic Atlas for Huntington Disease Based on

Data From the Enroll-HD Cohort Study
Douglas R. Langbehn, MD, PhD, Swati S. Sathe, MD, MS, Clement Loy, PhD, FRACP, MBA, Correspondence
Dr. Langbehn
Cristina Sampaio, MD, PhD, and Elizabeth A. Mccusker, MB BS (Hons), MBioeth, FRACP
douglas-langbehn@uiowa.edu
Abstract
The variable CAG repeat expansion in the huntingtin gene and its inverse relationship to motor
dysfunction onset are fundamental features of Huntington disease (HD). However, the wider
phenotype (including non-motor features) at particular CAG lengths, ages, and functional
levels is less well-characterized. The large number of participants in the Enroll-HD observa-
tional study enables the development of a phenotype atlas that summarizes the range and
distribution of HD phenotypes, including outliers and possible clusters, with respect to various
CAG repeat lengths, age ranges, and declining functional levels.
Methods
Enroll-HD is an ongoing prospective longitudinal observational study that collects natural
history data, releasing periodic data sets, in people with HD (PwHD) and controls. Core
assessments at annual visits focus on behavioral, cognitive, motor, and functional status. Pe-
riodic data set 5, used for the development of the first iteration of the Enroll-HD Phenotype
Atlas (EHDPA), included all eligible data collected through October 31, 2020. The atlas is
based on subsets (cells) of descriptive data for all motor, cognitive, psychiatric, and functional
measures that are routinely collected at most Enroll-HD sites, analyzed by single CAG lengths
and 5-year age blocks.
Results
Data from 42,840 visits from 15,982 unique PwHD were available for analysis. At baseline,
participants had a mean ± SD age of 48.9 ± 13.9 years and CAG repeat length of 43.4 ± 3.6 and
54.1% were female. The EHDPA includes 223 age-by-CAG subsets for CAG repeats between
36 and 69 with five-year age brackets starting from 20–24 years up to 85–89 years. The atlas can
be browsed at enroll-hd.org/for-researchers/atlas-of-hd-phenotype/.
Discussion
The EHDPA summarizes the spectrum and distribution of HD phenotypes, including outliers
and possible clusters, in all domains of disease involvement for the range of CAG repeat lengths,
ages, and functional levels. Its availability in an easy-to-use online format will assist clinicians in
tracking disease progression in PwHD by identifying phenotypic features most associated with
loss of function and enabling conversations related to prognosis. The observable patterns in the
EHDPA should also catalyze more formal multidomain characterization of motor, cognitive,
and psychiatric progression and their relationships to functional decline and disease modifiers.
Trial Registration Information

Enroll-HD is registered with clinicaltrials.gov: NCT01574053.
From the Departments of Psychiatry (D.R.L.), Biostatistics, University of Iowa, Iowa City; CHDI Management/CHDI Foundation (S.S.S., C.S.), Princeton, NJ; Macquarie Medical School
(C.L.), Macquarie University; and Department of Neurology (Huntington disease Service) (E.A.M.), Westmead Hospital, University of Sydney, Australia.
The Article Processing Charge was funded by CHDI.

Glossary
EHDPA = Enroll-HD Phenotype Atlas; HD = Huntington disease; PCA = principal component analysis; PDS5 = periodic data
set 5; PwHD = people with HD; SDMT = symbol digit modality test; TMS = total motor score; UHDRS = Unified
Huntington’s Disease Rating Scale.
Introduction Development of the Enroll-HD Phenotype Atlas

The EHDPA was purposefully designed for presentation as
Huntington disease (HD) is an inherited autosomal dominant an interactive website [enroll-hd.org/for-researchers/atlas-
neurodegenerative disease caused by an unstable expansion of-hd-phenotype/] where plots and tables can be easily
of CAG repeats in exon 1 of the huntingtin gene (HTT) on browsed. PDF reports containing the same information
chromosome 4 that encodes the huntingtin protein (HTT). and table summaries of the underlying data are also available
The variable CAG repeat expansion in the huntingtin gene and for download through the site. eFigure 1 (links.lww.com/
its inverse relationship to motor dysfunction onset and survival NXG/A655) provides a case example of the clinical appli-
are fundamental features of the disease.1,2 However, much of cation of the atlas.
the literature addressing the influence of CAG length in HD
has focused on the concepts of age at motor diagnosis3 or age of Analyses for the EHDPA were limited to PwHD; community
motor onset,4 and the wider phenotype (including non-motor controls were not included. The Table lists Enroll-HD as-
features) at particular CAG lengths, ages, and functional levels sessments that are included in the atlas. The EHDPA is based
is less well-characterized. on subsets of data containing single CAG lengths and five-
year age blocks. A minimum of 6 observations were required
Enroll-HD is a clinical research platform and the world’s to generate statistics and associated plots for each age range
largest observational study for families affected by HD.5 and CAG combination. Because of deidentification risk, re-
Others have previously used this platform to develop a sults for sparse combinations with 5 or fewer observations
clinical dashboard that allows comparison of total motor were suppressed. Furthermore, such sparse cells may not re-
score (TMS), total functional capacity, and symbol digit liably represent measurement distributions. We report on the
modality test (SDMT) score of an individual against a de- 223 unique age-CAG combination cells with at least 6 ob-
fined Enroll-HD cohort, controlling for age and CAG repeat servations available.
length.6 We report here the development of a new Enroll-HD
Phenotype Atlas (EHDPA) that expands upon this concept Statistical Analysis
to allow similar comparisons for a wider range of motor, To maximize the number of observations available for gen-
cognitive, and behavioral aspects of HD that have all been eration of descriptive statistics, data from all available visits for
measured as part of the Enroll-HD study.5 The online atlas every eligible participant were reported. Missing data were
has been developed to assist clinicians to better visualize ignored in these descriptive analyses, except for cognitive data
(through tables, charts, and/or illustrations) and understand for creating the cognitive principal component (PC) scores
the relationships between age; CAG repeat length in the
HTT gene; and key markers of phenotypic onset, function,
and progression in HD.
Table Measurements Used to Build the HD Phenotypic
Atlas
Methods Domain Scale/scores used
Motor UHDRS total motor score

Enroll-HD Periodic Data Set 5 Each motor subscore
Enroll-HD (NCT01574053) is a prospective longitudinal
Cognitive Symbol digit modality test
observational study that collects natural history data in Verbal fluency (letter and category)
PwHD and community controls (18 years and older).5,7 The Stroop Color and Word Test (the Word Condition)
study began in 2012 and is ongoing in 23 countries at 155 Trails A and B Tests
Mini-Mental Status Examination
sites across 4 continents. Core assessments at annual visits
focus on behavioral, cognitive, motor, and functional status Behavioral Problem Behavior Assessment (PBA) scores
HADS depression/anxiety and Snaith irritability scales
conducted using a battery of validated and widely accepted Additional depression and irritability scales
assessments, e.g., the Unified Huntington’s Disease Rating Three items from suicidal thought instrument
Scale (UHDRS).8 Periodic data set 5 (PDS5), used for the Functional Total Functional Capacity (plus all subscales)
development of this atlas, contained data from 21,116 Independence Scale
Functional Assessment (FAS)
Enroll-HD participants (16,120 PwP and 4,996 community
controls) from 71,682 visits with an average longitudinal Abbreviation: UHDRS = Unified Huntington’s Disease Rating Scale.
follow-up of 2.3 years.

(described below). The frequency of partially missing data is willingness to be contacted regarding participation in future
included with the age and CAG descriptive reports within the studies.5,7
atlas.
Data Availability
Standardized scores were also included in the age and CAG Periodic data set 5 is accessible through the Enroll-HD
descriptive reports to facilitate comparisons of these mea- website (enroll-hd.org/) to those with data security and pri-
sures. These standardized scores (Z scores) are expressed in vacy measures meeting standards described on the website.
standard deviaitons from the mean Enroll HD gene-expanded Access to nontransformed, nonaggregated, or suppressed data
outcome values. may be obtained through request, subject to approval by the
Scientific Review Committee that weighs the scientific merit
When the entirety of the HD gene–expanded data is analyzed, of the proposed project against the increased risk of partici-
all cognitive scores used in the atlas (with the partial exception pant identification.
of the Mini-Mental Status Examination [MMSE]) are highly
and similarly correlated. This suggests that the aspect(s) of
cognition relevant to HD are common to all of them and can Results
be expressed most precisely using a single composite of these
scores (the PC score). As part of the atlas development Description of the Data Set Used
process, we performed a principal component analysis (PCA) Data from 42,840 visits from 15,982 unique participants were
that demonstrated that 84% of the total variance in the unavailable for analysis. At baseline, participants had a mean ±
derlying cognitive scores can be accounted by this common SD age of 48.9 ± 13.9 years and CAG repeat length of 43.4 ±
aspect of cognition (eTable 1, links.lww.com/NXG/A655). 3.6 and 54.1% were female. Using this data set, we generated
As a convenient summary of measured cognition, we report 223 age-by-CAG subsets for each CAG repeat number be-
this PC score in several of the plots and report tables. As with tween 36 and 69 with 5-year age brackets starting from 20–24
standardized versions of the individual outcomes, the PC years to 85–89 years. The numbers of observations available
score is scaled such that the mean is 0 and the SD is 1 when for each age-by-CAG cell are listed in eTable 2 (links.lww.
jointly considering all observations used in the analysis. com/NXG/A655).
PCA requires complete data, and some participants were Overview of the Enroll-HD Phenotype Atlas
missing one or more cognitive scores. This was most fre- The full online atlas contains the following categories of plots
quently due to nonadministration of some measures at some and reports:
sites because the protocol considered those assessments
optional (extended). In the context of repeated measures per 1. Box plot series summarizing age-related trends for a
participant, we performed multiple imputation of these specific CAG length and assessment.
missing data using multilevel predictive means matching—a 2. Box plot series summarizing CAG-length–related
technique that combines related information from both the trends for a specific age range and assessment.
same visit, and also the participant’s other visits, to impute 3. Heat maps illustrating patterns of mean and median
plausible values for the missing data.9 The PCA was per- scores across all possible age and CAG length
formed after pooling 10 imputations of the missing data. combinations for a specific assessment.
4. Domain correlations illustrating the inter-relationship
There was no generation of hypothesis-testing p-values or among assessment measures within a specific domain
confidence intervals. Furthermore, aside from the cognitive for a specific CAG length and age range combination.
PC score, there was no modeling or smoothing of the data. All For motor, cognitive, and behavioral domains, the
plots and reports were generated using R 4.0.2. We used the R pairwise scatterplots of all assessments are illustrated
packages mice 3.11.0 and miceads 3.10–28 to perform multiple along with corresponding correlation coefficients.
imputation for the PC analysis. There are also cross-domain plots containing key
measurements from each of these domains plus the
Standard Protocol Approvals, Registrations, UHDRS Independence Scale.
and Patient Consents 5. Descriptive statistics reports for each assessment
The Enroll-HD study is performed in accordance with the containing all age and CAG length box plots for
Declaration of Helsinki. All participating sites received in- a specific assessment. These reports also include
stitutional review board approval, and all participants pro- tables of the statistical values (medians, quartiles,
vided written informed consent to take part in the study outlier boundaries) that are graphically displayed in
(including consent for research genotyping). Additional op- the box plots, as well as auxiliary tables indicating
tional components that require participant consent include age and CAG distributions available for each box
biosampling for banking purposes, family history assessment, plot.
linking of clinical information collected in other studies, and 6. Descriptive statistical reports for each age and CAG
length combination, available as downloadable PDF

files. These contain a comprehensive overview of all principal component score. By contrast, Figure 2C illustrates
assessment measures for a specific CAG length and age the CAG pattern of chorea severity for this age range. The
range combination. These reports contain descriptive mean total chorea score measured in individuals aged 40–44
statistics tables, bar plots of mean and median years predictably increases with increasing CAG length up to
assessment scores, plus domain correlation plots. 50 repeats, beyond which the mean total chorea score appears
to be stable or possibly even trend downward. Bearing in mind
To facilitate comparison among assessments, detailed reports that more severe chorea is represented by higher scores, this
in PDF form can also be viewed and downloaded within the plot is nearly a mirror image of the cognition plots. We must
atlas. There are reports available for each assessment measure caution against overinterpretation of apparent trend reversals
and across all measures for each specific CAG-age combina- for the lowest and highest CAG lengths. The widths of the
tion. Each figure is supported by tables of the underlying boxes in Figure 2 are proportional to the sample sizes in the
summary statistics; there are tables of summary statistics for age-CAG cells. There is often sparse representation of CAG
each measure and for standardized (z-score) versions of the lengths below 40 or above 49. Furthermore, participants
measures. The standardized scores allow easy comparison of within these cells may be a biased representation of the un-
severity across measures within the selected CAG-age cate- derlying population. HD is only partially penetrant in CAG
gory. The means and medians for these measures are also lengths less than 40, and the Enroll-HD data over-represent
illustrated with accompanying bar plots. those with penetrance. Reversed or stable trends at high CAG
lengths may also represent selection bias relative to the
Visualization of Trends Within the Atlas population based on ability and willingness to participate in
For most measures, visualization of age-related trends for a Enroll-HD visits.
specific CAG length revealed a consistent increase beginning in
the age range 40–44 years, with skewed scores for the 55–59 Heat maps illustrating patterns of mean scores across all
years and younger age ranges. Figure 1 illustrates a representative possible age and CAG length combinations are shown for
box plot series for UHDRS motor scores across age for CAG = mean UHDRS Independence Scale (Figure 3A) and UHDRS
42. Although unusual, motor scores elevated substantially above Chorea Score (Figure 3B). Compared to the Independence
the age and CAG norm do occur throughout the age range. Scale, the choreas vs CAG relationship is less variable for ages
30 and above. This suggests that chorea may not be the
Figure 2 illustrates box plot series summarizing CAG- predominant motor manifestation in adult-onset HD with
length–related trends for participants aged 45–49 years. higher CAG repeats. After mean chorea scores of approxi-
Cognitive function as assessed by the principal component mately 8 are reached, further increases in mean scores are less
composite (Figure 2A) and SDMT (Figure 2B) clearly de- clearly dependent on CAG length or age (this suggests that
clined with higher CAG lengths. The similarity of the 2 series chorea may not be as predominant within the motor mani-
is a result of the very high correlations among the cognitive festation as the disease progresses in severity). Analogous heat
scores (including the symbol digit test) that contribute to the map plots are also available for medians.
Figure 1 Representative Box Plot Series for UHDRS Motor Scores Across Age for CAG = 42
The proportion of the data in each age group is represented

by the width of the corresponding box.

Figure 2 Representative Box Plot Series Summarizing CAG-Length–Related Trends for Participants Aged 45–49 Years
(A) Principal component composite, (B) symbol digit modality test, (C) chorea severity. The proportion of the data in each age group is represented by the
width of the corresponding box.
Domain correlations illustrating the inter-relationship among ages, and functional levels. Examination of the nature and fre-
assessment measures for participants aged 35–39 years with a quency of outliers in these profiles may help identify pheno-
CAG length of 45 are shown for psychiatric/behavioral do- typic variation reflecting the effect of secondary genetic,
main (Figure 4) and across motor (TMS), cognitive (prin- comorbid, and environmental influences on disease pro-
cipal component score), and daily function (UHDRS gression. The atlas is available on the Enroll-HD website as a
Independence Scale) domains (Figure 5). Figure 5 clearly user-friendly tool for clinicians, researchers, and health care
illustrates notable and similar correlations among the UHDRS professionals. The data aid in understanding the age-dependent
TMS, cognitive principal component score, and UHDRS features of HD as CAG lengths vary. The EHDPA readily
Independence Scale. For the psychiatric/behavioral domain, illustrates whether the typical range varies widely or not for a
no single measure is a good summary of overall severity. given CAG length and age range, allowing judgment of the
However, irritability and apathy scores from the Problem degree to which an individual’s phenotype is atypical. The huge
Behavior Assessment for Huntington Disease (PBA)10 show sample size allows detailed observational summaries by 5-year
the highest association with other measures of overall HD age range for each CAG length as well as enabling the de-
severity, including functional measures. Nonetheless, these velopment of descriptive plots relating specific measurements
behavioral measures have weaker correlations with the other to age and CAG.
measures and with each other. Similar patterns are seen for
most CAG-age combinations available in the atlas. There are some limitations in interpretation of the EHDPA
that users should understand. To maximize the number of
observations available, data from all available visits for every
eligible participant were reported. There is, therefore, a degree
Discussion of nonindependence between repeated annual observations
The EHDPA summarizes the spectrum and distribution of HD from the same participants. Such nonindependence of obser-
phenotypes, including outliers and possible clusters, in all do- vations would need to be accounted for in potential future
mains of disease involvement for the range of CAG lengths, hypothesis-driven testing and modeling of these data. Although

Figure 3 Representative Heat Maps Illustrating Patterns of Mean Scores Across All Possible Age and CAG Length
Combinations
(A) UHDRS Independence Scale, (B) UHDRS Chorea scores.

The scales at the right of the plots convert heat map colors to
raw scores of the corresponding measures.
it is based on the largest observational HD database ever col- illness. These potential biases may also distort these age-
lected, this database is not a random sample of the entire dependent cross-sectional patterns if we interpret them as
population at risk. For example, the EHDPA illustrates phe- typical longitudinal progression for an individual. Un-
notypes and the degree of variation in the assessments for the fortunately, substantial ideal data—repeated systematic
incompletely penetrant CAG repeat lengths of 36–39; for this measurements of the same people across several decades—
range, there is the important caveat that the available sample is, simply do not exist, and the potential biases for non-
at best, representative of the population in this age range that participation may not be improved by increasing sample sizes
comes to clinical attention, but because of the partial pene- unless the sources of study recruitment evolve substantially.
trance, the EHDPA probably does not represent typical pat- As the Enroll-HD study continues, it will be important to use
terns for the whole population of individuals who have these future periodic data sets to compare longitudinal within-
repeat lengths. person data with the disease course suggested by the atlas.
Clinical features affecting participation in the Enroll-HD Future work could expand on the observable patterns in the
study may also bias the data, particularly at the severe end of atlas to produce a more formal multidomain characterization
the illness spectrum; for instance, the apparent stability of of motor, cognitive, and psychiatric progression and the re-
some features in advanced HD may instead be attributable to lationship to functional decline and disease modifiers.11,12
nonparticipation or drop-out among those with more severe Inclusion of age-specific control data would help the clinician

Figure 4 Psychiatric/Behavioral Domain Correlations Illustrating the Inter-Relationship Among Assessment Measures for
Participants Aged 35–39 Years With a CAG Length of 45
Scatterplots of individual participant data for all pairwise combinations of measures are displayed beneath the diagonal. For scales with a limited number of
values such as the PBA Apathy, a small amount of random noise is added so that the density of various score combinations is illustrated. The points within the
scatterplots are coded to distinguish whether the participant has been given a clinical motor diagnosis of manifest HD by virtue of the highest possible score
of 4 on the UHDRS clinician diagnostic confidence limit rating scale (DCL). This is meant to provide some sense of the degree to which the severity of
combinations of measures separates so-called motor manifest vs premanifest HD. Absolute values of the Pearson correlation coefficients for each mea-
surement pair are displayed above the diagonal with font sizes roughly proportional to their magnitude. The diagonal cells contain histograms of the
individual distributions of the measures. hadsAnx = HADS Anxiety Scale, hadsDep = HADS Depression Scale, Sn_Irrit = Snaith Irritability Scale, dep = PBA
Depression, irrit = PBA Irritability, psychosis = PBA Psychosis, apathy = PBA Apathy, exec = PBA Executive Function, DCL = UHDRS Diagnostic Confidence Score.
to understand how PwHD are different from people without phenotypes and assist clinicians in tracking disease progression
the HD CAG expansion. Relevant phenotype definition is in an individual by identifying phenotypic features most asso-
critical to the success of gene-discovery studies. The atlas will ciated with loss of function. It will also assist in determining
facilitate definition of more detailed and possibly more sen- whether a suspected deviation from the likely course is truly
sitive CAG-adjusted phenotypes for such studies.13 For ex- unusual. The work done to develop this atlas has potential as a
ample, variability not well-explained by age and CAG might prototype for initiatives in other trinucleotide repeat disorders
be used as the phenotypic outcome measure in studies if large databases like Enroll-HD can be created.
searching for additional HD-modifying genes. The EHDPA
may also be a useful tool in assessing whether future thera- Acknowledgment
peutic agents have a differential effect on motor, cognitive, Data used in this work were generously provided by the
and psychiatric aspects of HD. participants in the Enroll-HD study and made available by
CHDI Foundation, Inc. Enroll-HD is a clinical research
Clinicians are often faced with the challenge of providing platform and longitudinal observational study for Huntington
prognosis for an individual, which could help PwHD and disease families intended to accelerate progress toward
families with their professional and financial plans as well as for therapeutics; it is sponsored by CHDI Foundation, a nonprofit
future care needs. The EHDPA has been developed to provide biomedical research organization exclusively dedicated to
a graphical multidimensional representation of the range of HD collaboratively developing therapeutics for HD. Enroll-HD

Figure 5 Cross-Domain (Motor, Cognitive, Psychiatric/Behavioral, Functional Independence) Correlations Illustrating the
Inter-Relationship Among Assessment Measures for Participants Aged 35–39 Years With a CAG Length of 45
Please see the Figure 4 legend for further details. Motor =

UHDRS Total Motor Score, Cog PC1 = Cognitive PC1, PBA
Exec = PBA Executive function, PBA Apath = PBA Apathy,
Indep Scale = UHDRS Independence Scale, DCL = UHDRS
Diagnostic Confidence Score.
would not be possible without the vital contribution of the

research participants and their families. The individuals who Appendix Authors
contributed to the collection of the Enroll-HD data are also Name Location Contribution
gratefully acknowledged; see enroll-hd.org/enrollhd_docu-
ments/2020-10-R1/Enroll-HD-Acknowledgement-list-2020- Douglas R. Departments of Drafting/revision of the
Langbehn, MD, Psychiatry, Biostatistics, manuscript for content,
10-R1.pdf. Medical writing (editing and final styling) assistance PhD University of Iowa, Iowa including medical writing for
was provided by Anita Chadha-Patel (ACP Clinical Commu- City content; major role in the
acquisition of data; study
nications, Ltd) and was funded by CHDI. concept or design; analysis
or interpretation of data
Study Funding Swati S. Sathe, CHDI Management/ Drafting/revision of the

MD, MS CHDI Foundation, manuscript for content,
This work was funded by CHDI Foundation, Inc. Princeton, NJ including medical writing for
content; analysis or
interpretation of data
Disclosure
D.R. Langbehn reports personal consulting fees and non- Clement Loy, Macquarie Medical Drafting/revision of the
PhD, FRACP, MBA School, Macquarie manuscript for content,
financial support from Voyager Therapeutics, personal con- University, Sydney, including medical writing for
sulting fees from Novartis, personal consulting fees from Australia content; major role in the
acquisition of data; analysis
uniQure, personal consulting fees from Takeda, personal or interpretation of data
consulting fees from AskBio, and personal consulting fees
Cristina CHDI Management/ Drafting/revision of the
from Guidepoint Consultants, all outside the submitted work; Sampaio, MD, CHDI Foundation, manuscript for content,
Clement Loy is supported by the Australian National Medical PhD Princeton, NJ including medical writing for
content; major role in the
and Health Research Council; Elizabeth McCusker has acquisition of data; study
nothing to report; Swati Sathe and Cristina Sampaio are concept or design; analysis
employed by CHDI Management as advisors to CHDI or interpretation of data
Foundation. Go to Neurology.org/NG for full disclosures. Elizabeth A. Department of Drafting/revision of the

Mccusker, MB BS Neurology (Huntington manuscript for content,
(Hons), MBioeth, disease Service), including medical writing for
Publication History FRACP Westmead Hospital, content; major role in the
Received by Neurology: Genetics May 31, 2023. Accepted in final form University of Sydney, acquisition of data; study
Australia concept or design; analysis
October 4, 2023. Submitted and externally peer reviewed. The handling or interpretation of data
editor was Editor Stefan M. Pulst, MD, Dr med, FAAN.

References 7. Sathe S, Ware J, Levey J, et al. Enroll-HD: an integrated clinical research platform and
1. Langbehn DR, Stout JC, Gregory S, et al. Association of CAG repeats with long-term worldwide observational study for Huntington’s disease. Front Neurol. 2021;12:
progression in Huntington disease. JAMA Neurol. 2019;76(11):1375-1385. doi: 667420. doi:10.3389/fneur.2021.667420
10.1001/jamaneurol.2019.2368 8. Huntington Study Group. Unified Huntington’s disease rating scale: reliability and
2. Langbehn DR. Longer CAG repeat length is associated with shorter survival after disease onset consistency. Mov Disord. 1996;11(2):136-142. doi:10.1002/mds.870110204
in Huntington disease. Am J Hum Gen. 2022;109(1):172-179. doi:10.1016/j.ajhg.2021.12.002 9. Burren SV. Flexible Imputation of Missing Data. 2nd ed. CRC Press; 2018.
3. Chen T, Wang Y, Ma Y, Marder K, Langbehn DR. Predicting disease onset from 10. Craufurd D, Thompson JC, Snowden JS. Behavioral changes in Huntington disease.
mutation status using proband and relative data with applications to Huntington’s Neuropsychiatry Neuropsychol Behav Neurol. 2001;14(4):219-226.
disease. J Probab Stat. 2012;2012:375935. doi:10.1155/2012/375935 11. McAllister B, Gusella JF, Landwehrmeyer GB, et al. Timing and impact of psychiatric,
4. Lee JM, Ramos EM, Lee JH, et al. CAG repeat expansion in Huntington disease cognitive, and motor abnormalities in Huntington disease. Neurology. 2021;96(19):
determines age at onset in a fully dominant fashion. Neurology. 2012;78(10):690-695. e2395-e2406. doi:10.1212/wnl.0000000000011893
doi:10.1212/WNL.0b013e318249f683 12. Genetic Modifiers of Huntington’s Disease (GeM-HD) Consortium. Identification of
5. Landwehrmeyer GB, Fitzer-Attas CJ, Giuliano JD, et al. Data analytics from enroll- genetic factors that modify clinical onset of Huntington’s disease. Cell. 2015;162(3):
HD, a global clinical research platform for Huntington’s disease. Mov Disord Clin 516-526. doi:10.1016/j.cell.2015.07.003
Pract. 2017;4(2):212-224. doi:10.1002/mdc3.12388 13. Moss DJH, Pardiñas AF, Langbehn D, et al. Identification of genetic variants
6. Walker T, Ghosh B, Kipps C. Assessing decline: visualising progression in Hun- associated with Huntington’s disease progression: a genome-wide associa-
tington’s disease using a clinical dashboard with enroll-HD data. J Huntingtons Dis. tion study. Lancet Neurol. 2017;16(9):701-711. doi:10.1016/s1474-4422(17)
2017;6(2):139-147. doi:10.3233/jhd-170234 30161-8

Genetic Patterns of Selected Muscular Dystrophies in

the Muscular Dystrophy Surveillance, Tracking, and
Research Network
Peter B. Kang, MD, Magali Jorand-Fletcher, MPH, Wanfang Zhang, MS, Suzanne W. McDermott, PhD, Correspondence
Dr. Kang
Reba Berry, RN, Chelsea Chambers, MS, CGC, Kristen N. Wong, MS, CGC, Yara Mohamed, MD,
pkang@umn.edu
Shiny Thomas, MBBS, MPH, Y Swamy Venkatesh, MD, Christina Westfield, BSN, Nedra Whitehead, MS, PhD, and
Nicholas E. Johnson, MD, for the Muscular Dystrophy Surveillance, Tracking, and Research Network (MD STARnet)
Abstract
To report the genetic etiologies of Emery-Dreifuss muscular dystrophy (EDMD), limb-girdle
muscular dystrophy (LGMD), congenital muscular dystrophy (CMD), and distal muscular
dystrophy (DD) in 6 geographically defined areas of the United States.
Methods
This was a cross-sectional, population-based study in which we studied the genes and variants
associated with muscular dystrophy in individuals who were diagnosed with and received care
for EDMD, LGMD, CMD, and DD from January 1, 2008, through December 31, 2016, in the 6
areas of the United States covered by the Muscular Dystrophy Surveillance, Tracking, and
Research Network (MD STARnet). Variants of unknown significance (VUSs) from the original
genetic test reports were reanalyzed for changes in interpretation.
Results
Among 243 individuals with definite or probable muscular dystrophy, LGMD was the most
common diagnosis (138 cases), followed by CMD (62 cases), DD (22 cases), and EDMD (21
cases). There was a higher proportion of male individuals compared with female individuals,
which persisted after excluding X-linked genes (EMD) and autosomal genes reported to have
skewed gender ratios (ANO5, CAV3, and LMNA). The most common associated genes were
FKRP, CAPN3, ANO5, and DYSF. Reanalysis yielded more definitive variant interpretations for
60 of 144 VUSs, with a mean interval between the original clinical genetic test of 8.11 years for
all 144 VUSs and 8.62 years for the 60 reclassified variants. Ten individuals were found to have
monoallelic pathogenic variants in genes known to be primarily recessive.
Discussion
This study is distinct for being an examination of 4 types of muscular dystrophies in selected
geographic areas of the United States. The striking proportion of resolved VUSs demonstrates
the value of periodic re-examinations of these variants. Such re-examinations will resolve some
genetic diagnostic ambiguities before initiating repeat testing or more invasive diagnostic
procedures such as muscle biopsy. The presence of monoallelic pathogenic variants in recessive
genes in our cohort indicates that some individuals with muscular dystrophy continue to face
From the Paul & Sheila Wellstone Muscular Dystrophy Center (P.B.K.), Department of Neurology, and Institute for Translational Neuroscience, University of Minnesota, Minneapolis;
Department of Pediatrics (M.J.-F., Y.M.), University of Florida College of Medicine, Gainesville; Department of Epidemiology and Biostatistics (W.Z.), University of South Carolina,
Columbia; Department of Environmental, Occupational, and Geospatial Health Sciences (S.W.M.), Graduate School of Public Health and Health Policy, City University of New York;
Division of Population Health Surveillance (R.B., C.W.), Bureau of Maternal and Child Health, South Carolina Department of Health and Environmental Control, Columbia; Department
of Human and Molecular Genetics (C.C.), Virginia Commonwealth University, Richmond; Department of Pediatrics (K.N.W.), University of Utah, Salt Lake City; New York State
Department of Health (S.T.), Albany; Department of Neurology (Y.S.V.), University of South Carolina, Columbia; RTI International (N.W.), Research Triangle Park, NC; and Department of
Neurology (N.E.J.), Virginia Commonwealth University, Richmond.
Funding information and disclosures are provided at the end of the article. Full disclosure form information provided by the authors is available with the full text of this article at
Neurology.org/NG.

Glossary
BMD = Becker muscular dystrophy; CMD = congenital muscular dystrophy; DD = distal muscular dystrophy; DMD =
Duchenne muscular dystrophy; EDMD = Emery-Dreifuss muscular dystrophy; LGMD = limb-girdle muscular dystrophy;
VUS = variants of unknown significance.
incomplete genetic diagnoses; further refinements in genetic knowledge and diagnostic approaches will optimize diagnostic
information for these individuals.
Introduction Study Population and Data Sources (Standard

MD STARnet Methodology)
Four classic but less common forms of muscular dystrophy are Individuals with EDMD, LGMD, CMD, and DD were iden-
Emery-Dreifuss muscular dystrophy (EDMD), limb-girdle tified through MD STARnet surveillance using previously
muscular dystrophy (LGMD), congenital muscular dys- described methods.15,17 MD STARnet is a multisite, population-
trophy (CMD), and distal muscular dystrophy (DD). based muscular dystrophy surveillance system in the United
These disorders share some overlapping associated genes States that currently identifies individuals who were diagnosed
and some phenotypic features. A number of epidemiologic with one of 8 muscular dystrophies (Becker [BMD], CMD, DD,
studies that include various combinations of these muscular Duchenne [DMD], EDMD, facioscapulohumeral [FSHD],
dystrophies have been published over the years, ranging LGMD, and myotonic [DM]). Cohort eligibility for case ab-
from broad-based reports1 and a genetic database mining straction included meeting the following criteria from January 1,
study of LGMD2 to focused population studies on in- 2008, to December 31, 2016: clinical diagnosis of an eligible MD
dividual genes or even individual variants.3-8 Several geo- and receipt of clinical care and residency in Colorado (CO),
graphically defined population-based studies have been Iowa (IA), South Carolina (SC), the Piedmont region of North
conducted for one or more of these muscular dystrophies, Carolina (NC), a 21-county area in Western New York State
primarily outside the United States.9-14 Variants of un- (wNY), or Utah (UT/NV) (Figure).
known significance (VUSs) often complicate the in-
terpretation of genetic test reports, either when they are the Case Abstraction (Standard MD
primary findings or when they are secondary findings in STARnet Methodology)
addition to pathogenic or likely pathogenic variants. The Potential cases were identified by International Classification
presence of such VUSs often leads to ambiguous conclu- of Disease codes (ICD-9-CM: 359.0, 359.1, 359.21; ICD-10:
sions from genetic test reports. G71.0, G71.1)18 from clinic and administrative data and
screened for cohort eligibility. Data abstraction of medical
The Muscular Dystrophy Surveillance, Tracking, and Re- records for eligible cases began in 2016 by trained abstractors
search Network (MD STARnet) has reported general who reviewed and abstracted medical records for clinical data
sociodemographic and clinical characteristics of individ- during the follow-up period. Collected data included de-
uals with these MDs in specific regions of the United mographic characteristics, medical history (including earliest
States15,16; however, MD STARnet has not previously signs and symptoms), diagnostic testing (including genetic
analyzed detailed patterns of genes and variants associated testing, muscle biopsy immunostaining, skin biopsy immu-
with the 4 less common MD types. In this report, we nostaining, diagnostic Western blot, and/or diagnostic MRI),
characterize such patterns of genetic test results for the 4 clinical care, and family history of muscular dystrophy.
MDs under investigation. Our primary analysis is followed
by a review of information in publicly available databases
Case Review (Standard MD
that can be used to inform interpretation and classification
STARnet Methodology)
of VUSs.
Abstracted clinical data for each eligible case were reviewed by
a panel of MD STARnet neuromuscular physicians who
assigned clinical MD type and a case classification of definite,
Methods probable, possible, asymptomatic, or not an eligible MD
Standard Protocol Approvals, Registrations, (eAppendix 1, links.lww.com/NXG/A649). MD type was
and Patient Consents assigned by the panel based on defined patterns of clinical
Study activities were conducted under protocols that were signs and symptoms for each type, as well as diagnostic
approved by the institutional review board and/or public findings. A case was categorized as definite if there were
health authority for surveillance at each MD STARnet documented clinical symptoms referable to one of the MD
site. These activities qualified for waivers of consent at types, a genetic report of DNA analysis with the identification
each site. of pathogenic findings in the patient or a family history of

Figure Analytic Flowchart
Flowchart of the analytic process from

the total initial cohort to final genetic
analysis of variants, including exclusions.
genetically confirmed case status in a family member showing EDMD, LGMD, CMD, and DD. From the pooled data, we
a recognizable inheritance pattern, and other confirmatory excluded asymptomatic individuals, individuals whose genetic
testing. Probable cases were defined by documented clinical tests showed benign genes, individuals whose abstracted ge-
symptoms referable to an MD type and supported by family netic test results lacked sufficient details for analysis, and in-
history and laboratory results referable to one of the selected dividuals who resided in Nevada and were ascertained under
MDs, but without meeting the criteria for a definite case. UT authority (cases residing in UT were included in the
Asymptomatic cases were those who had positive genetic test analysis), the latter due to a small, unrepresentative subgroup
results for an associated gene but showed no signs or symp- (Figure). The asymptomatic cases were excluded because the
toms of muscular dystrophy. clinical diagnosis assignment could not be confirmed.
Data Pooling Variables

A pooled, analytic data set was created and included clinical We studied the ages when the diagnosis was confirmed by
data from each MD STARnet site for individuals classified as genetic testing or the first abnormal neuromuscular diagnostic
having definite, probable, and asymptomatic diagnoses of test (serum creatine kinase [CK], EMG, and/or muscle

biopsy). Racial and ethnic classifications were recorded when pathogenic allele was unaccompanied by a second pathogenic
available. Family history status was defined as yes, no, and allele for the same gene (“single hit”).
unknown.
Missing Variants
Statistical Analysis At the time of abstraction, pathogenic variant information
Frequencies and proportions were calculated for categorical from genetic test reports was entered into the MD STARnet
variables; mean and SD or range were used for continuous database. However, some variants did not have nucleotide or
variables. The distribution of age in years when 50% (25%, 75%) amino acid positions entered, rendering them impossible to
of the study group had each outcome was estimated using the characterize or analyze further. In some cases where a VUS
Kaplan-Meier estimator. A minimum of 10 individuals per table test result was missing cDNA or amino acid change in-
cell were required to report numbers and percentages for de- formation in the structured part of the MD STARnet database,
mographic data to avoid potentially compromising patient pri- this information was found in the free text “description” field
vacy, including racial and ethnic identities. of the database. In those cases, study authors filled in the
missing data manually.
Pathogenicity of Variants and Reanalysis
of VUSs Data Availability
The lists of pathogenic variants and VUSs were obtained from Owing to privacy concerns, data from MD STARnet are not
the pooled MD STARnet database. The information in these publicly available. Researchers interested in MD STARnet
fields was supplemented by additional variants identified in a data can contact MDSTARnet@cdc.gov.
free text “description” field of the database. Duplicate entries
were deleted, and typographical errors were corrected. The
original variant classification categories in the database were Results
pathogenic, VUS, normal, and unknown, with “normal” cor-
Demographics
responding to the currently accepted variant classification
We first examined key demographic data among the 243 in-
“benign” and “unknown” corresponding to a variant classifi-
dividuals in our cohort to characterize basic information
cation that could not be confirmed based on the abstracted
(Table 1). 64.6% of our cohort was male and 35.4% was
information. In light of the possibility that some of the VUSs
female. The unexpectedly higher proportion of male indi-
and unassigned (“unknown”) variants may be subject to re-
viduals compared with female individuals was present in all 4
interpretation, we included the unassigned variants in the
diagnostic categories, despite the presence of only one
general category of VUS and re-evaluated this combined
X-linked gene (EMD) among the commonly associated
variant list, with each variant assessed by a pair of authors who
genes. Regarding other aspects of our cohort, 44.4% had no
determined the current ACMG classification19 and Revel
known family history of the disease in question. The mean age
score20 (franklin.genoox.com), along with ClinVar21 in-
at diagnosis was 27.2 years, and the median age at diagnosis
terpretation and gnomAD22 allele frequencies. When a pair of
was 22.1 years, and as noted above, the diagnosis had to be
authors did not agree on the current ACMG classification of a
made in the 2008–2016 period for inclusion. The mean age at
particular variant, they reviewed their findings with each other
the last abstracted clinic visit was 37 years, and the median age
to reach consensus, consulting with the lead author (P.B.K.)
at the last clinic visit was 35 years. These mean and median
when needed. Using information from these online databases,
ages were younger in the CMD group (20.7 and 18 years,
we then determined whether a VUS would still be classified as
respectively) compared with the other groups.
a VUS or be reclassified for the purposes of this analysis as
pathogenic, likely pathogenic, likely benign, or benign. For Category Distributions and Genetic Findings
those variants that could be reclassified for this analysis, we As 4 major categories of muscular dystrophy were represented
determined whether the change would either alter the original in our cohort, we examined the distribution of individuals
overall interpretation of the primary genetic test finding or among these categories. LGMD was the most common clin-
eliminate a secondary finding. A secondary finding is defined ical diagnosis, followed by CMD, with DD and EDMD being
as a VUS that is noted in a genetic test report in the presence the least common and nearly equivalent to each other nu-
of a pathogenic or likely pathogenic variant in a different gene. merically (Table 2). Overall, 60.1% of the cohort had a defi-
We also recorded the intervals between the original clinical nite classification, with LGMD having the lowest proportion
genetic test date and the date of completion of our reanalysis (Table 2). As originally characterized by the clinical genetic
(December 13, 2022) and calculated mean and median in- test reports, the associated genes with the highest overall
tervals for each diagnosis (EDMD, LGMD, CMD, and DD). occurrence of pathogenic variants, excluding VUSs, in the
cohort were FKRP, CAPN3, ANO5, and COL6A1 (eTable 1,
Monoallelic Pathogenic Variants in Genes With links.lww.com/NXG/A650), all of which are most commonly
Recessive Inheritance (“Single Hits”) found in LGMD, except COL6A1, which is typically found in
We reviewed all pathogenic variants of muscular dystrophy– CMD (eTable 2). In EDMD, the genes with the most fre-
associated genes that are known to have recessive or primarily quent pathogenic variants were EMD and LMNA. In CMD,
recessive patterns of inheritance and noted when one the genes with the most frequent pathogenic variants were

male and female individuals were analyzed in aggregate, both
Table 1 Demographic Features of 243 Individuals With before and after exclusion of the X-linked gene EMD and
EDMD, LGMD, CMD, and DD, With Comparisons autosomal genes previously found in a higher proportion of
Among the 4 Muscular Dystrophy Categories for male individuals (ANO5, CAV3, and LMNA)23 (eTable 4).
Each Set of Variables
Total (N = 243) p Valuea Pathogenicity of Variants and Reanalysis of
VUSs and Unassigned Variants
Sites, n (%) 0.0628
Given the high number of VUSs that appear in clinical genetic
Colorado 34 (14.0) testing, we asked whether genetic findings could be refined by
Iowa 45 (18.5) reanalysis using online databases. Before data cleaning, we
identified 162 pathogenic variants, 169 VUSs, and 10 un-
North Carolina (Piedmont region) 28 (11.5)
assigned variants. After data cleaning and review of the initial
New York (Western 21 counties) 54 (22.2) abstraction results, we identified 162 pathogenic variants and
South Carolina 32 (13.2)
179 VUSs for a total of 341 variants (Figure). Among the 179
VUS results reviewed, 35 were classified as having missing
Utah 50 (20.6) data with the following breakdown: 26 were missing cDNA
No known family history, n (%) 0.0786 change or amino acid change information, 4 had unverifiable
information, 4 only listed intervening sequence information,
No 135 (55.6)
and 1 had mtDNA information, yielding 144 VUSs with ad-
Yes 108 (44.4) equate information for analysis (eTable 5, links.lww.com/
Sex, n (%) 0.0777
NXG/A650). Our review of those 144 VUSs using currently
accepted standard classification systems and databases yielded
Female 86 (35.4)
reclassification of 23 variants to pathogenic or likely patho-
Male 157 (64.6) genic and 37 variants to benign or likely benign. Eighty-four
VUSs remained unchanged (Table 3). The reclassifications
Raceb, n (%) 0.00571
changed the interpretations of primary genetic test findings
White 189 (77.8) for 35 variants and eliminated secondary findings for 23 var-
Others/Unknown 54 (22.2) iants (Table 4 and eTables 6 and 7). Of note, there were
individuals with multiple VUSs, thus the number of individ-
Ethnicity, n (%) 0.0939
uals with reinterpretations of primarily genetic findings was 28
Hispanic 17 (7.0) and the number of individuals with elimination of secondary
Non-Hispanic 211 (86.8)
findings was 18 (Table 4). The mean intervals between the
original clinical genetic test report and the time of VUS
Unknown 15 (6.2)
reanalysis was 8.11 years for all 144 VUSs analyzed (Table 5)
Age at first diagnosis <0.001 and 8.62 years for the 60 reclassified VUSs (Table 6).
Mean (SD) 27.2 (20.7)
Monoallelic Pathogenic Variants in Recessive
Median [Min, Max] 22.1 [0, 79.1] Genes (“Single Hits”)
Projected age on June 15, 2022 <0.001
The problem of VUSs is often accompanied by the dilemma
presented by single pathogenic variants identified in genes
Mean (SD) 37.0 (20.5) that are known to be recessive, leaving diagnostic uncertainty.
Median [Min, Max] 35.0 [6.00, 90.0] We identified 10 individuals with such monoallelic pathogenic
variants in genes with recessive or primarily recessive patterns
Abbreviations: CMD = congenital muscular dystrophy; DD = distal muscular of inheritance (“single hits”). These individuals have ambig-
dystrophy; EDMD = Emery-Dreifuss muscular dystrophy; LGMD = limb-gir-
dle muscular dystrophy. uous genetic diagnoses despite the presence of pathogenic
a
p-values are for the association between categorical variables and the 4
categories of muscular dystrophy, examined with a χ2 test, and for the as-
variants. These single hits were most common for LGMD
sociation between continuous variables and the 4 categories of muscular (eTable 8, links.lww.com/NXG/A650), and after removing
dystrophy, examined with an ANOVA.
b
Others/Unknown category includes American Indian or Alaska Native, individuals with duplicate pathogenic genes, the most com-
Black or African American, Asian, Multiple, Other and Unknown. Owing to mon gene in which this phenomenon was observed was FKRP
small subgroup sizes, individual categories are not enumerated.
(eTable 9). For all 10 affected symptomatic individuals, there
were no other findings in the original genetic test report to
indicate the presence of alternative genetic diagnoses.
COL6A1, LAMA2, COL6A3, COL6A2, and FKRP. For DD,
the genes with the most frequent pathogenic variants were Cases With Pathogenic Variants in
DYSF and GNE. Most of the variants were pathogenic for Multiple Genes
CMD and LGMD, whereas VUSs were proportionately more The possibility of digenic Mendelian inheritance as an ex-
frequent for EDMD and DD (eTable 3). The proportions of planation for some genetically unsolved cases of muscular

Table 2 Clinical Classification of the 243 Individuals With Certainty of Diagnoses
Case status EDMD (N = 21) LGMD (N = 138) CMD (N = 62) DD (N = 22) Total (N = 243)
Definite, n (%) 14 (66.7) 67 (48.6) 52 (83.9) 13 (59.1) 146 (60.1)
Probable, n (%) 7 (33.3) 71 (51.4) 10 (16.1) 9 (40.9) 97 (39.9)
Abbreviations: CMD = congenital muscular dystrophy; DD = distal muscular dystrophy; EDMD = Emery-Dreifuss muscular dystrophy; LGMD = limb-girdle
muscular dystrophy.
dystrophy has been raised. In our cohort, 3 individuals were Although the available information does not enable us to draw
found to have pathogenic or likely pathogenic alleles in 2 definitive conclusions about the origins of this sex distribu-
different genes: (1) DNAJB6 c.271T>G (p.F91V) paired with tion, it is plausible that female individuals affected by these
GAA c.546G>A (p.T183 = ); (2) TTN c.70493dupA paired categories of muscular dystrophy were diagnosed at lower
with FKRP c.826C>A (p.L2761); and (3) ANO5 c.191dupA rates or sought specialty care less often than affected male
paired with TTN c.85692_85696delAGCTT. individuals during this more recent surveillance period.
Milder manifestations in female individuals could account for
either explanation. As the most widely known muscular dys-
Discussion trophy, Duchenne muscular dystrophy (DMD) is X-linked
Prior geographically defined studies of EDMD, LGMD, and almost exclusively affects male individuals; there may be a
CMD, and DD in the United States consist principally of 2 misperception that muscular dystrophy of all kinds does not
MD STARnet reports that did not include the genetic analysis tend to affect female individuals. It is thus important that
presented here.15,16 In other countries, epidemiologic studies outreach efforts for the medical community and for the general
focusing on LGMD have been published from Austria,9 public emphasize that both female and male individuals can be
Chile,10 Italy,11,23 the Netherlands,12 and Spain13 and CMD affected by many of the subtypes of muscular dystrophy.
from Italy.14 These and other studies from around the world
that covered these diagnoses provide valuable information but The excess of individuals in our cohort with probable rather
had differences in scope from our study because they did not than definite diagnoses indicates that a gap in confirmatory
include genetic subtype information,24 were broad-based gen- genetic diagnosis persists in these categories of muscular
eral studies of muscle diseases or neuromuscular disorders,25-35 dystrophy. Of note, the case definitions (eAppendix 1, links.
or focused on genetic subsets of one of these muscular lww.com/NXG/A649) classify affected individuals with a
dystrophies.36-41 Our findings are consistent with prior studies family history of genetic confirmation as definite cases. The
for the relative frequency of the 4 MD types and the common percentage of probable cases is highest for LGMD, similar to
genes identified in our cohort. the high unsolved rates for this type of muscular dystrophy
found on both clinical genetic testing and research-based
The skewed sex ratio is striking and cannot be explained by genomic analyses.43-45 This may in part be due to uneven
expected genetic distributions, given that only one major access to genetic testing in some populations.
gene, EMD, is X-linked.42 There are several autosomal genes
that have previously been associated with male-predominant The common occurrence of VUSs in clinical genetic test re-
ratios, including ANO5, CAV3, and LMNA.23 A study of ports and the unexpectedly high rate of reclassification on
EDMD, LGMD, CMD, and DD from the prior MD STARnet reanalysis of these VUSs in this study indicate that further
cycle only detected a skewed male/female ratio in EDMD.16 advances are needed in genetic diagnostic technology and
Table 3 Clinical MD Types and ACMG Classifications for VUSs

ACMG classification EDMD (N = 13) LGMD (N = 73) CMD (N = 44) DD (N = 14) Total (N = 144)
Benign, n (%) 1 (7.7) 5 (6.8) 11 (25.0) 6 (42.9) 23 (16.0)
Likely benign, n (%) 1 (7.7) 9 (12.3) 3 (6.8) 1 (7.1) 14 (9.7)
VUS, n (%) 11 (84.6) 48 (65.8) 21 (47.7) 4 (28.6) 84 (58.3)
Likely pathogenic, n (%) 0 0.00 4 (5.5) 5 (11.4) 1 (7.1) 10 (6.9)
Pathogenic, n (%) 0 (0.0) 7 (9.6) 4 (9.1) 2 (14.3) 13 (9.0)
muscular dystrophy.

Table 4 Changes in Interpretation and Elimination of Secondary Findings for VUSs by Clinical MD Type
By variants
EDMD (N = 23) LGMD (N = 84) CMD (N = 50) DD (N = 22) Total (N = 179)
Change in genetic diagnosis, n (%)
No 22 (95.7) 69 (82.1) 39 (78.0) 14 (63.6) 144 (80.4)
Yes 1 (4.3) 15 (17.9) 11 (22.0) 8 (36.4) 35 (19.6)
Eliminated secondary finding, n (%)
No 22 (95.7) 74 (88.1) 40 (80.0) 20 (90.9) 156 (87.2)
Yes 1 (4.3) 10 (11.9) 10 (20.0) 2 (9.1) 23 (12.8)
By individuals
EDMD (N = 11) LGMD (N = 49) CMD (N = 29) DD (N = 8) Total (N = 97)
Change in genetic diagnosis, n (%)
No 10 (90.9) 36 (73.5) 18 (62.1) 5 (62.5) 69 (71.1)
Yes 1 (9.1) 13 (26.5) 11 (37.9) 3 (37.5) 28 (28.9)
Eliminated secondary finding, n (%)
No 10 (90.9) 40 (81.6) 22 (75.9) 7 (87.5) 79 (81.4)
Yes 1 (9.1) 9 (18.4) 7 (24.1) 1 (12.5) 18 (18.6)
muscular dystrophy.
interpretation to improve the accuracy and detection rate of Cases in which a monoallelic pathogenic variant (“single hit”)
genetic testing. At the very least, a basic scan of VUSs iden- is unaccompanied by a pathogenic variant in the same gene on
tified on clinical genetic testing using online databases is the other allele, for genes known to have recessive or primarily
warranted when the original genetic test is more than a few recessive inheritance, are frustrating for the patients and cli-
years old. In light of the frequent posting of new online re- nicians involved because it leaves the genetic diagnosis with-
sources for variant interpretation, we recommend consulting out a full resolution. Approaches that promise to improve the
with a neuromuscular neurologist, geneticist, or genetic genetic diagnosis of muscular dystrophies include tran-
counselor with expertise in these resources to determine scriptome analysis (RNAseq), computational reanalysis to
which ones to use at a given time. detect more subtle changes such as splice variants, and long
Table 5 Time Intervals Between Clinical Genetic Tests and Reanalysis of All 144 VUSs That Were Reanalyzed (y)
CMD (N = 44) DD (N = 14) EDMD (N = 13) LGMD (N = 73) Total (N = 144)
Intervals
Mean (SD) 8.22 (1.99) 7.00 (0) 7.14 (1.21) 8.31 (2.18) 8.11 (2.00)
Median [Min, Max] 8.00 [6.00, 13.0] 7.00 [7.00, 7.00] 7.00 [6.00, 9.00] 7.50 [6.00, 14.0] 7.00 [6.00, 14.0]
Missing 17 (38.6%) 11 (78.6%) 6 (46.2%) 37 (50.7%) 71 (49.3%)
ACMG classification, n (%)
Benign 11 (25.0) 6 (42.9) 1 (7.7) 5 (6.8) 23 (16.0)
Likely benign 3 (6.8) 1 (7.1) 1 (7.7) 9 (12.3) 14 (9.7)
Likely pathogenic 5 (11.4) 1 (7.1) 0 (0) 4 (5.5) 10 (6.9)
Pathogenic 4 (9.1) 2 (14.3) 0 (0) 7 (9.6) 13 (9.0)
VUS 21 (47.7) 4 (28.6) 11 (84.6) 48 (65.8) 84 (58.3)
The date used for reanalysis was December 13, 2022, the date when the VUS analysis was completed.

Table 6 Time Intervals Between Clinical Genetic Tests and Reanalysis of 60 VUSs That Were Reclassified (y)
CMD (N = 23) DD (N = 10) EDMD (N = 2) LGMD (N = 25) Total (N = 60)
Lag years
Mean (SD) 8.29 (1.86) 7.00 (NA) NA (NA) 9.18 (2.79) 8.62 (2.28)
Median [Min, Max] 8.00 [6.00, 13.0] 7.00 [7.00, 7.00] NA [NA, NA] 8.00 [6.00, 14.0] 8.00 [6.00, 14.0]
Missing, n (%) 9 (39.1) 9 (90.0) 2 (100) 14 (56.0) 34 (56.7)
ACMG classification, n (%)
Benign 11 (47.8) 6 (60.0) 1 (50.0) 5 (20.0) 23 (38.3)
Likely benign 3 (13.0) 1 (10.0) 1 (50.0) 9 (36.0) 14 (23.3)
Likely pathogenic 5 (21.7) 1 (10.0) 0 (0) 4 (16.0) 10 (16.7)
Pathogenic 4 (17.4) 2 (20.0) 0 (0) 7 (28.0) 13 (21.7)
The date used for reanalysis was December 13, 2022, the date when the VUS analysis was completed.
read sequencing. Long read sequencing in particular holds transcriptome sequencing (RNAseq), as well as review of
promise to find the “second hits” for those individuals with muscle imaging studies, could yield additional meaningful
monoallelic pathogenic variants in genes with recessive diagnostic information. However, MD STARnet does not
inheritance.46 collect raw sequence data, genomic DNA samples, specimens
from muscle biopsies, or images from muscle ultrasound and
It has been postulated that there may be rare cases in which MRI studies; thus those types of investigations are beyond the
variants at 2 different loci may together cause disease. For scope of this study.
muscular dystrophy, this is best documented for facioscapu-
lohumeral muscular dystrophy type 2, caused by variants in EDMD, LGMD, CMD, and DD collectively comprise a sig-
SMCHD1 paired with a D4Z4 allele harboring a poly- nificant portion of the muscular dystrophy population. Their
adenylation signal.47,48 There are sparse reports of compound genetic heterogeneity, compared with more common mus-
pathogenic variants in different genes potentially causing cular dystrophies such as dystrophinopathies (DMD and
muscular dystrophy, including SCGB paired with SCGD49 and BMD), facioscapulohumeral muscular dystrophy (FSHD),
COL6A1 paired with COL6A2.50 We found only 3 cases of and myotonic dystrophy (DM1 and DM2), leads to distinct
potential digenic inheritance in our cohort; more extensive challenges in diagnosis, prognosis, and management. Our
studies are required to determine whether both variants are findings indicate that periodic reanalysis of VUSs using pub-
necessary and sufficient to cause disease in these circumstances. licly available databases will at times yield new information. It
will be important to continue characterizing these MDs to
Our study has some limitations. The subgroups for EDMD optimize genetic diagnosis, clinical management, and research
and DD were small, although the presence of expected studies that will help lead to novel therapies. Encouragingly,
common genes in those subgroups indicates that they were to investigational therapies are already undergoing human clin-
some extent representative of broader populations with these ical trials for some of these muscular dystrophies, providing a
disease categories. The absence of ANO5 in the DD group great deal of hope for the future.
was likely because of the small cohort size. The numbers for
some genetic subtypes did not meet the MD STARnet Acknowledgment
reporting threshold of at least 10 cases. Thus, we were not able The authors thank Kristin M. Conway, PhD, at the
to present details of the distributions of certain variables Department of Epidemiology, College of Public Health,
within these subtypes such as sex ratios. As genetic testing was University of Iowa for assistance with this project. In-
performed at different times at different diagnostic facilities, termountain Healthcare was a source for some of the data
variant interpretation practices likely varied throughout the from the Utah site for this study.
cohort, although all clinical genetic diagnostic test facilities in
the United States are required to qualify for and maintain Study Funding
Clinical Laboratory Improvement Amendment certification, This publication was supported by the Cooperative Agree-
providing some standardization in variant interpretation ment numbers DD001126, DD001119, DD001123,
practices over time. Beyond variant reanalysis, reanalysis of DD001116, DD001117, DD001108, DD001120, DD001054,
raw sequence data and the use of newer technologies such as DD001242, DD001243, DD001245, DD001248, DD001249,
nanopore whole-genome long read sequencing46 and whole- DD001252, and DD001255, funded by the Centers for

Disease Control and Prevention (CDC). The findings and
conclusions in this report are those of the authors and do not Appendix (continued)
necessarily represent the official position of the Centers for Name Location Contribution
Disease Control and Prevention. The Iowa site was addi-
tionally supported by DD001247. Partial support for all data Suzanne W. Department of Drafting/revision of the
McDermott, Environmental, manuscript for content,
sets with the Utah Population Database (UPDB) was pro- PhD Occupational, and including medical writing for
vided by the University of Utah Huntsman Cancer Institute Geospatial Health Sciences, content; study concept or
Graduate School of Public design; analysis or
and the Huntsman Cancer Institute Cancer Center Support Health and Health Policy, City interpretation of data
Grant, P30 CA042014, from the National Cancer Institute. University of New York
Reba Berry, Division of Population Health Analysis or interpretation of

Disclosure RN Surveillance, Bureau of data
Maternal and Child Health,
P.B. Kang has served on advisory boards for Sarepta Thera- South Carolina Department
peutics, NS Pharma, and Teneofour; and has served as a of Health and Environmental
Control, Columbia
consultant for Novartis and Neurogene. M. Jorand-Fletcher
reports no disclosures. W. Zhang reports no disclosures. S.W. Chelsea Department of Human and Analysis or interpretation of
McDermott reports no disclosures. R. Berry reports no dis- Chambers, Molecular Genetics, Virginia data
MS, CGC Commonwealth University,
closures. C. Chambers reports no disclosures. K.N. Wong Richmond
reports no disclosures. Y. Mohamed reports no disclosures.
Kristen N. Department of Pediatrics, Analysis or interpretation of
S. Thomas reports no disclosures. Y.S. Venkatesh reports no Wong, MS, University of Utah, Salt Lake data
disclosures. C. Westfield reports no disclosures. N. Whitehead CGC City
reports no disclosures. N.E. Johnson has received grant funding Yara Department of Pediatrics, Analysis or interpretation of
from the NIH (R01 NS104010 and R21 TR003184), CDC Mohamed, University of Florida College data
MD of Medicine, Gainesville
(U01 DD001242), and the FDA (R01 FD006071). He receives
research funds from Dyne, AveXis, Vertex Pharmaceuticals, Shiny New York State Department Analysis or interpretation of
Thomas, of Health, Albany data
Fulcrum Therapeutics, ML Bio, Sarepta, Triplet Therapeutics, MBBS, MPH
Avidity Biosciences, and AMO Pharma. He has provided con-
Y. Swamy Department of Neurology, Analysis or interpretation of
sultation for AMO Pharma, AveXis, Fulcrum Therapeutics, Venkatesh, University of South Carolina, data
Dyne, Avidity, Vertex, and Entrada. He receives licensing fees MD Columbia
from the University of Rochester for the CCMDHI and CMTHI.
Christina Division of Population Health Analysis or interpretation of
Full disclosure form information provided by the authors is Westfield, Surveillance, Bureau of data
available with the full text of this article at Neurology.org/NG. BSN Maternal and Child Health,
South Carolina Department
of Health and Environmental
Publication History Control, Columbia
Received by Neurology: Genetics June 22, 2023. Accepted in final form Nedra RTI International, Research Analysis or interpretation of
September 29, 2023. Submitted and externally peer reviewed. The Whitehead, Triangle Park, NC data
handling editor was Associate Editor Antonella Spinazzola, MD. MS, PhD
Nicholas E. Department of Neurology, Drafting/revision of the

Johnson, MD Virginia Commonwealth manuscript for content,
University, Richmond including medical writing for
content; analysis or
interpretation of data
Appendix Authors
Peter B. Paul & Sheila Wellstone Drafting/revision of the References

Kang, MD Muscular Dystrophy Center, manuscript for content, 1. Mah JK, Korngut L, Fiest KM, et al. A systematic review and meta-analysis on the
Department of Neurology, including medical epidemiology of the muscular dystrophies. Can J Neurol Sci. 2016;43(1):163-177. doi:
and Institute for writing for content; 10.1017/cjn.2015.311
Translational Neuroscience, study concept or 2. Liu W, Pajusalu S, Lake NJ, et al. Estimating prevalence for limb-girdle muscular
University of Minnesota, design; analysis or dystrophy based on public sequencing databases. Genet Med. 2019;21(11):2512-2520.
Minneapolis interpretation of data doi:10.1038/s41436-019-0544-8
3. Polavarapu K, Mathur A, Joshi A, et al. A founder mutation in the GMPPB gene
Magali Department of Pediatrics, Drafting/revision of [c.1000G > A (p.Asp334Asn)] causes a mild form of limb-girdle muscular dystrophy/
Jorand- University of Florida College the manuscript for congenital myasthenic syndrome (LGMD/CMS) in South Indian patients. Neuro-
Fletcher, of Medicine, Gainesville content, including genetics. 2021;22(4):271-285. doi:10.1007/s10048-021-00658-1
MPH medical writing for 4. Pantoja-Melendez CA, Miranda-Duarte A, Roque-Ramirez B, Zenteno JC. Epide-
content; analysis or miological and molecular characterization of a mexican population isolate with high
interpretation of data prevalence of limb-girdle muscular dystrophy type 2A due to a novel calpain-3 mu-
tation. PLoS One. 2017;12(1):e0170280. doi:10.1371/journal.pone.0170280
Wanfang Department of Epidemiology Drafting/revision of 5. Tétreault M, Srour M, Allyson J, et al. Founder mutation for α-sarcoglycan-LGMD2D
Zhang, MS and Biostatistics, University the manuscript for in a Magdalen Islands Acadian cluster. Can J Neurol Sci. 2011;38(5):747-752. doi:
of South Carolina, Columbia content, including 10.1017/s0317167100054135
medical writing for 6. Frosk P, Greenberg CR, Tennese AAP, et al. The most common mutation in FKRP
content; analysis or causing limb girdle muscular dystrophy type 2I (LGMD2I) may have occurred only
interpretation of data once and is present in Hutterites and other populations. Hum Mutat. 2005;25(1):
38-44. doi:10.1002/humu.20110

7. Merlini L, Kaplan JC, Navarro C, et al. Homogeneous phenotype of the gypsy limb- 29. Lefter S, Hardiman O, Ryan AM. A population-based epidemiologic study of adult
girdle MD with the gamma-sarcoglycan C283Y mutation. Neurology. 2000;54(5): neuromuscular disease in the Republic of Ireland. Neurology. 2017;88(3):304-313.
1075-1079. doi:10.1212/wnl.54.5.1075 doi:10.1212/WNL.0000000000003504
8. Al-Zaidy SA, Malik V, Kneile K, et al. A slowly progressive form of limb-girdle muscular 30. Chung B, Wong V, Ip P. Prevalence of neuromuscular diseases in Chinese children: a
dystrophy type 2C associated with founder mutation in the SGCG gene in Puerto Rican study in southern China. J Child Neurol. 2003;18(3):217-219. doi:10.1177/
Hispanics. Mol Genet Genomic Med. 2015;3(2):92-98. doi:10.1002/mgg3.125 08830738030180030201
9. Krenn M, Tomschik M, Wagner M, et al. Clinico-genetic spectrum of limb-girdle 31. El-Tallawy HN, Khedr EM, Qayed MH, Helliwell TR, Kamel NF. Epidemiological
muscular weakness in Austria: a multicentre cohort study. Eur J Neurol. 2022;29(6): study of muscular disorders in Assiut, Egypt. Neuroepidemiology. 2005;25(4):205-211.
1815-1824. doi:10.1111/ene.15306 doi:10.1159/000088674
10. Cerino M, González-Hormazábal P, Abaji M, et al. Genetic profile of patients with 32. Hughes MI, Hicks EM, Nevin NC, Patterson VH. The prevalence of inherited neu-
limb-girdle muscle weakness in the chilean population. Genes (Basel). 2022;13(6): romuscular disease in Northern Ireland. Neuromuscul Disord. 1996;6(1):69-73. doi:
1076. doi:10.3390/genes13061076 10.1016/0960-8966(94)00017-4
11. Guglieri M, Magri F, D’Angelo MG, et al. Clinical, molecular, and protein correlations 33. Nakagawa M, Nakahara K, Yoshidome H, et al. Epidemiology of progressive muscular
in a large sample of genetically diagnosed Italian limb girdle muscular dystrophy dystrophy in Okinawa, Japan. Classification with molecular biological techniques.
patients. Hum Mutat. 2008;29(2):258-266. doi:10.1002/humu.20642 Neuroepidemiology. 1991;10(4):185-191. doi:10.1159/000110268
12. van der Kooi AJ, Frankhuizen WS, Barth PG, et al. Limb-girdle muscular dystrophy in 34. Darin N, Tulinius M. Neuromuscular disorders in childhood: a descriptive epide-
The Netherlands: gene defect identified in half the families. Neurology. 2007;68(24): miological study from western Sweden. Neuromuscul Disord. 2000;10(1):1-9. doi:
2125-2128. doi:10.1212/01.wnl.0000264853.40735.3b 10.1016/s0960-8966(99)00055-3
13. Urtasun M, Sáenz A, Roudaut C, et al. Limb-girdle muscular dystrophy in Guipúzcoa 35. Theadom A, Rodrigues M, Poke G, et al. A nationwide, population-based prevalence
(Basque Country, Spain). Brain. 1998;121(Pt 9):1735-1747. doi:10.1093/brain/121.9.1735 study of genetic muscle disorders. Neuroepidemiology. 2019;52(3-4):128-135. doi:
14. Mostacciuolo ML, Miorin M, Martinello F, Angelini C, Perini P, Trevisan CP. Genetic 10.1159/000494115
epidemiology of congenital muscular dystrophy in a sample from north-east Italy. 36. Guimarães-Costa R, Fernández-Eulate G, Wahbi K, et al. Clinical correlations and
Hum Genet. 1996;97(3):277-279. doi:10.1007/BF02185752 long-term follow-up in 100 patients with sarcoglycanopathies. Eur J Neurol. 2021;
15. Do TN, Street N, Donnelly J, et al. Muscular dystrophy surveillance, tracking, and research 28(2):660-669. doi:10.1111/ene.14592
network pilot: population-based surveillance of major muscular dystrophies at four U.S. sites, 37. Alonso-Pérez J, González-Quereda L, Bello L, et al. New genotype-phenotype cor-
2007-2011. Birth Defects Res. 2018;110(19):1404-1411. doi:10.1002/bdr2.1371 relations in a large European cohort of patients with sarcoglycanopathy. Brain. 2020;
16. Wallace B, Smith KT, Thomas S, et al. Characterization of individuals with selected 143(9):2696-2708. doi:10.1093/brain/awaa228
muscular dystrophies from the expanded pilot of the Muscular Dystrophy Surveil- 38. Bardhan M, Anjanappa RM, Polavarapu K, et al. Clinical, genetic profile and disease
lance, Tracking and Research Network (MD STARnet) in the United States. Birth progression of sarcoglycanopathies in a large cohort from India: high prevalence of
Defects Res. 2021;113(7):560-569. doi:10.1002/bdr2.1764 SGCB c.544A>C. Neurogenetics. 2022;23(3):187-202. doi:10.1007/s10048-022-
17. Miller LA, Romitti PA, Cunniff C, et al. The muscular dystrophy surveillance tracking 00690-9
and research network (MD STARnet): surveillance methodology. Birth Defects Res A 39. Fanin M, Duggan DJ, Mostacciuolo ML, et al. Genetic epidemiology of muscular
Clin Mol Teratol. 2006;76(11):793-797. doi:10.1002/bdra.20279 dystrophies resulting from sarcoglycan gene mutations. J Med Genet. 1997;34(12):
18. World Health Organization, ed. International Statistical Classification of Diseases and 973-977. doi:10.1136/jmg.34.12.973
Related Health Problems, 10th revision, 2nd ed. World Health Organization; 2004. 40. Fanin M, Nascimbeni AC, Fulizio L, Angelini C. The frequency of limb girdle mus-
19. Richards S, Aziz N, Bale S, et al.; ACMG Laboratory Quality Assurance Committee. Stan- cular dystrophy 2A in northeastern Italy. Neuromuscul Disord. 2005;15(3):218-224.
dards and guidelines for the interpretation of sequence variants: a joint consensus recom- doi:10.1016/j.nmd.2004.11.003
mendation of the American College of Medical Genetics and Genomics and the Association 41. Stensland E, Lindal S, Jonsrud C, et al. Prevalence, mutation spectrum and phenotypic
for Molecular Pathology. Genet Med. 2015;17(5):405-425. doi:10.1038/gim.2015.30 variability in Norwegian patients with Limb Girdle Muscular Dystrophy 2I. Neuro-
20. Ioannidis NM, Rothstein JH, Pejaver V, et al. REVEL: an ensemble method for muscul Disord. 2011;21(1):41-46. doi:10.1016/j.nmd.2010.08.008
predicting the pathogenicity of rare missense variants. Am J Hum Genet. 2016;99(4): 42. Heller SA, Shih R, Kalra R, Kang PB. Emery-Dreifuss muscular dystrophy. Muscle
877-885. doi:10.1016/j.ajhg.2016.08.016 Nerve. 2020;61(4):436-448. doi:10.1002/mus.26782
21. Landrum MJ, Chitipiralla S, Brown GR, et al. ClinVar: improvements to accessing 43. Ghaoui R, Cooper ST, Lek M, et al. Use of whole-exome sequencing for diagnosis of
data. Nucleic Acids Res. 2020;48(D1):D835–D844. doi:10.1093/nar/gkz972 limb-girdle muscular dystrophy: outcomes and lessons learned. JAMA Neurol. 2015;
22. Karczewski KJ, Francioli LC, Tiao G, et al. The mutational constraint spectrum 72(12):1424-1432. doi:10.1001/jamaneurol.2015.2274
quantified from variation in 141,456 humans. Nature. 2020;581(7809):434-443. doi: 44. Reddy HM, Cho KA, Lek M, et al. The sensitivity of exome sequencing in identifying
10.1038/s41586-020-2308-7 pathogenic mutations for LGMD in the United States. J Hum Genet. 2017;62(2):
23. Magri F, Nigro V, Angelini C, et al. The Italian limb girdle muscular dystrophy 243-252. doi:10.1038/jhg.2016.116
registry: relative frequency, clinical features, and differential diagnosis. Muscle Nerve. 45. Saha M, Reddy HM, Salih MA, et al. Impact of PYROXD1 deficiency on cellular
2017;55(1):55-68. doi:10.1002/mus.25192 respiration and correlations with genetic analyses of limb-girdle muscular dystrophy in
24. Deenen JCW, van Doorn PA, Faber CG, et al. The epidemiology of neuromuscular Saudi Arabia and Sudan. Physiol Genomics. 2018;50(11):929-939. doi:10.1152/
disorders: age at onset and gender in The Netherlands. Neuromuscul Disord. 2016; physiolgenomics.00036.2018
26(7):447-452. doi:10.1016/j.nmd.2016.04.011 46. Bruels CC, Littel HR, Daugherty AL, et al. Diagnostic capabilities of nanopore long-
25. Norwood FLM, Harling C, Chinnery PF, Eagle M, Bushby K, Straub V. Prevalence of read sequencing in muscular dystrophy. Ann Clin Transl Neurol. 2022;9(8):
genetic muscle disease in Northern England: in-depth analysis of a muscle clinic 1302-1309. doi:10.1002/acn3.51612
population. Brain. 2009;132(Pt 11):3175-3186. doi:10.1093/brain/awp236 47. Lemmers RJLF, Tawil R, Petek LM, et al. Digenic inheritance of an SMCHD1
26. Megarbane A, Bizzari S, Deepthi A, et al. A 20-year clinical and genetic neuromuscular mutation and an FSHD-permissive D4Z4 allele causes facioscapulohumeral muscular
cohort analysis in Lebanon: an international effort. J Neuromuscul Dis. 2022;9(1): dystrophy type 2. Nat Genet. 2012;44(12):1370-1374. doi:10.1038/ng.2454
193-210. doi:10.3233/JND-210652 48. Mitsuhashi S, Boyden SE, Estrella EA, et al. Exome sequencing identifies a novel
27. Müller KI, Ghelue MV, Lund I, Jonsrud C, Arntzen KA. The prevalence of hereditary SMCHD1 mutation in facioscapulohumeral muscular dystrophy 2. Neuromuscul
neuromuscular disorders in Northern Norway. Brain Behav. 2021;11(1):e01948. doi: Disord. 2013;23(12):975-980. doi:10.1016/j.nmd.2013.08.009
10.1002/brb3.1948 49. Trabelsi M, Kavian N, Daoud F, et al. Revised spectrum of mutations in sarcoglyca-
28. Pagola-Lorz I, Vicente E, Ibáñez B, et al. Epidemiological study and genetic charac- nopathies. Eur J Hum Genet. 2008;16(7):793-803. doi:10.1038/ejhg.2008.9
terization of inherited muscle diseases in a northern Spanish region. Orphanet J Rare 50. Nadeau A, Kinali M, Main M, et al. Natural history of Ullrich congenital muscular
Dis. 2019;14(1):276. doi:10.1186/s13023-019-1227-x dystrophy. Neurology. 2009;73(1):25-31. doi:10.1212/WNL.0b013e3181aae851

CLINICAL/SCIENTIFIC NOTE OPEN ACCESS
Expanding the Clinical Spectrum of UBTF-Related

Neurodevelopmental Disorder
Andrea Pietra, MD,* Flavia Palombo, PhD,* Melania Giannotta, MD, Monica Maffei, MD, Claudio Fiorini, PhD, Correspondence
Dr. Garone
Roberta Costa, PhD, Giovanna Cenacchi, MD, PhD, Valerio Carelli, MD, PhD, Duccio Maria Cordelli, MD, PhD,
caterina.garone@unibo.it
Antonella Pini, PhD, and Caterina Garone, MD, PhD
Abstract MORE ONLINE
Video
Objectives
UBTF1 gene encodes for Upstream Binding Transcription Factor, an essential protein for RNA
metabolism. A recurrent de novo variant (c.628G>A; p.Glu210Lys) has recently been asso-
ciated with a childhood-onset neurodegenerative disorder characterized by motor and language
regression, ataxia, dystonia, and acquired microcephaly. In this study, we report the clinical,
metabolic, molecular genetics and neuroimaging findings and histologic, histochemical, and
electron microscopy studies in muscle samples of 2 patients from unrelated families with a
neurodevelopmental disorder.
Methods
Data were retrospectively analyzed by medical charts revision.
Results
Patient 1, a 16-year-old boy, presented a childhood-onset slowly progressive neurodegenerative
disorder mainly affecting language skills, behavior, and motor coordination. Patient 2, a 22-year-
old woman, presented with a severe and rapidly progressive disease with dystonic tetra paresis,
acquired microcephaly, and severe cognitive deficit complicated by pseudobulbar syndrome
characterized by involuntary and uncontrollable outbursts of laughing, dysphagia requiring tube
feeding, and nocturnal hypoventilation treated with noninvasive ventilation. Both patients carried
the recurrent previously described UBTF1 de novo variant and had signs of mitochondrial
dysfunction at muscle biopsy. The metabolic profile of patient 2 also revealed a decrease in CSF
biopterin.
Discussion
These case reports add new insights to the UBTF1 disease spectrum instrumental to improving
the diagnostic rate in neurodevelopmental disorders.
Introduction
Upstream Binding Transcription Factor, OMIM *600673 (UBTF) gene encodes for 2 isoforms
of the upstream binding factor (UBF), UBTF1 and UBTF2, able to form homodimers and
heterodimers and plays an essential role in RNA transcription within the nucleolus.1 Specifi-
cally, UBTF1 regulates ribosomal RNA transcription by RNA polymerase I, whereas UBTF2
regulates mRNA transcription by RNA polymerase II. A recurrent monoallelic de novo
*These authors contributed equally to the manuscript as first authors.

From the UO Genetica Medica (Andrea Pietra), IRCCS Azienda Ospedaliero-Universitaria di Bologna; Department of Medical and Surgical Sciences (Andrea Pietra, C.G.), Alma Mater
Studiorum University of Bologna; IRCCS Istituto delle Scienze Neurologiche di Bologna (F.P.), Programma di Neurogenetica; IRCCS Istituto delle Scienze Neurologiche di Bologna
(M.G., Antonella Pini, C.G.), UOC Neuropsichiatria dell’età Pediatrica; IRCCS Istituto delle Scienze Neurologiche di Bologna (M.M., C.F.), Programma di neuroradiologia con tecniche ad
elevata complessità; Department of Biomedical and Neuromotor Sciences (R.C., G.C., V.C., D.M.C.), Alma Mater Studiorum University of Bologna; and UO Anatomia (G.C.), Istologia
Patologica, IRCCS Azienda Ospedaliero-Universitaria di Bologna, Italy.
The Article Processing Charge was funded by Italian Minister of Research.

missense variant c.628G>A (p.Glu210Lys) causes a neuro- atrophy (Figure 1, A and B). A follow-up brain MRI at 16 years
developmental disorder characterized by motor, speech, of age showed an increase in hyperintensity in T2-weighted
language, and cognitive regression in early childhood pro- white matter with moderate worsening in brain atrophy
gressively leading to severe cognitive deficit, loss of mile- (Figure 1, C and D). Metabolic analyses of arylsulfatase A, CSF
stones, pyramidal and extrapyramidal signs, and behavioral and plasma lactic acid, plasma, and urinary amino acids, acyl-
dysfunction.2-4 This variant pathogenicity is due to the UBF carnitine, and urinary organic acid were normal. Histologic
gain of function responsible for a marked increase in the and histochemical study of muscle biopsy specimen displayed
expression of pre-RNA and 18S rRNA.2 a few eosinophilic subsarcolemmal accumulations, mild di-
ameter variability of the fibers and isolated vesicular nuclei at
In this study, we report clinical, metabolic, and neuroradiologic hematoxylin-eosin staining (data not shown), a focal increase in
findings of 2 additional cases from unrelated families, which oil red O staining (data not shown), and several fibers with a
expand the clinical phenotype of UBTF1-related disease. subsarcolemmal increase in cytochrome c oxidase (COX) re-
action (Figure 1A). The ultrastructural analysis revealed mi-
tochondrial alterations compatible with the light microscopy
Methods findings: accumulation of abnormal large swollen mitochon-
dria, irregularly shaped with hypodense matrix and aberrant
We retrospectively revised the following: clinical features;
residual cristae (eFigure 1, links.lww.com/NXG/A642).
neuroimaging studies; metabolic investigations including
Mitochondrial respiratory chain enzyme activities in muscle
urine, plasma, CSF; histologic, histochemical, and electron
homogenate were within normal ranges (data not shown).
microscopy studies in muscle biopsy samples5; and the
molecular genetics data including exome sequencing of
Patient 2 is a 22-year-old woman, born at term after an un-
the 2 patients and segregation analysis in the 2 unrelated
eventful pregnancy and cesarian delivery. No consanguinity
families.
nor family history for neurodevelopmental disorders was
reported. After nonspecific febrile illness, a global neuro-
development regression occurred in the second year of life
Results with arrest in motor and language skills, abnormal behavior
Patient 1 is a 16-year-old boy, born at term after an uneventful with irritability, gait unbalance with difficulty climbing stairs,
pregnancy and spontaneous delivery. No consanguinity nor frequent falls, and cranial circumference growth delay. At 6
family history for neurodevelopmental disorders was repor- years of age, the clinical course was complicated by axial
ted. Psychomotor development was normal up to 5 years of asymmetric recurrent dystonic episodes lasting some days
age when he presented with attention deficit, difficulties in compromising the ability to walk and sialorrhea. At 9 years of
relationships with pairs, and coarse movements. At 6 years of age, she developed tetraparesis with parkinsonian rigidity and
age, he had a progressive global regression in cognitive, be- limb dystonia and pseudobulbar syndrome characterized by
havioral, and motor skills. He progressively developed a involuntary and uncontrollable outbursts of laughing lasting
complex clinical picture characterized by ataxia, nystagmus, several minutes. She lost the ability to walk at 10 years.
oculomotor apraxia, severe cognitive deficit, dysarthria, stut- Dysphagia was progressive and required percutaneous gas-
tering, self-injurious episodes, dysmetria, and dysgraphia. A trostomy (PEG) tube feeding at 18 years of age. She also
decline in cognitive functions, more prominent in the lan- presented with nocturnal hypoventilation treated with non-
guage and speech skills, and a worsening of the extrapyramidal invasive ventilation. Awake and sleep EEG recordings showed
symptoms’ severity were observed during the follow-up. At 15 no epileptiform or periodic discharges and nerves’ conduc-
years of age, he also presented with episodes of obsessive- tions were normal. At 22 years of age, she completely lost
compulsive behavior. At the last evaluation at 17 years of age, verbal communication, but she partially reacted to stimuli in
he was still able to walk with support but with an ataxic gait; he the family context (Video 2). A first brain MRI, when she was
had severe sialorrhea but was still able to orally feed himself; 4 years of age, showed signs of modest cerebral atrophy and
he presented with difficulties articulating every single word hyperintensity in T2-weighted images in the periventricular
because of verbal fluency impairment; upper limb movements supratentorial and deep white matter and thin corpus cal-
were functionally impaired by severe dysmetria; he also losum (Figure 1, E and F). Follow-up brain MRI at the age of
presented with subacute onset of additional symptoms of 13 years showed signs of progressive cortical and subcortical
movement disorders with dystonia and parkinsonism for atrophies, initial signs of cerebellar atrophy, and brainstem
which he started treatment with levodopa with a definitive atrophy, with further thinning of the corpus callosum and
improvement at 4 months of follow-up (Video 1); IQ assessed higher hyperintensity of supratentorial white matter and
with Wechsler Intelligence Scale for Children–type IV thalami in T2-weighted images (Figure 1, G and H). Clinical
(WISC-IV) scale confirmed a severe cognitive deficit. Signs or and neuroimaging findings were suggestive of metabolic
symptoms in other organs/apparatus were not reported. A neurodegeneration. Metabolic analyses showed a decrease in
first brain MRI was performed at 8 years of age showing in- biopterin in the CSF and a slight increase in plasma lactic (2.4;
creased T2 signals in the periventricular supratentorial and n.v. 0.3–1.3 mmol/L) and pyruvic acid (0.11; n.v. 0.03–0.08
deep white matter, thin corpus callosum, and supratentorial mmol/L) levels, while white cell enzymes activities, plasma

Figure 1 Brain MRI Findings in Patient 1 (A–D) and Patient 2 (E–H)
Patient 1: Brain MRI at the first evalu-

ation at 8 years of age showing in mid-
sagittal T1-weighted (A) and axial T2-
weighted (B) images, increased T2
signals in the periventricular supra-
tentorial and deep white matter, thin
corpus callosum, and supratentorial
atrophy. Follow-up brain MRI at 16
years of age showing in mid-sagittal
T1-weighted (C) and axial T2-weighted
(D) images an increase in hyper-
intensity in T2-weighted white matter
with moderate worsening in brain at-
rophy. Patient 2: Brain MRI at the first
evaluation at 4 years of age showing in
mid-sagittal T1-weighted (E) and axial
T2-weighted (F) images modest cere-
bral atrophy and hyperintensity in T2-
weighted images in the periven-
tricular supratentorial and deep white
matter and thin corpus callosum; Fol-
low-up brain MRI at 13 years of age
showing in mid-sagittal T1-weighted
(G) and axial T2-weighted (H) images
progressive cortical and subcortical
atrophies, initial signs of cerebellar
atrophy, and brainstem atrophy, with
further thinning of the corpus cal-
losum and higher hyperintensity of
supratentorial white matter signal in
T2-weighted images and thalami sig-
nal hyperintensity in T2-weighted
images.
sialotransferrin and vitamin E, plasma and urinary amino acids missense variant, c.628G>A [p.Glu210Lys], in the UBTF gene
and organic acids, urinary purine, pyrimidine, and mucopoly- (NM_014233.4). According to the American College of Med-
saccharides were in the normal range. Muscle biopsy showed a ical Genetics classification, the c.628G>A was classified as
modest subsarcolemmal increase at Gomori trichrome staining pathogenic with the PM2, PP2, PP3, and PP5 (PM = moderate
in numerous fibers, observed also with succinate dehydrogenase evidence of pathogenicity; PP = supporting evidence of patho-
staining (data not shown), displaying a normal COX activity genicity) criteria. Segregation analysis showed that the c.628G>A
(Figure 2B). Mitochondrial respiratory chain activity in muscle have arisen de novo in both families, definitely confirming the
homogenate detected a partial reduction in complex I activity diagnosis of a UBTF1-related disorder.
(10.3; n.v. 13-24).
After inconclusive diagnostic analysis with a next-generation Discussion

multigene panel, the genetic investigation was expanded to WES,
and an in silico panel analysis of genes associated with severe Neurodevelopmental disorders (NDD) are a heterogeneous
pediatric disorders revealed the presence of a heterozygous group of disorders including many neurodegenerative and
Figure 2 Cytochrome Histochemical Activity in Patient 1 (A) and Patient 2 (B) Muscle Specimens Revealing Subsarcolemmal
Rims

neurometabolic diseases and presenting with delay or loss of adulthood.2 Pathogenicity of the recurrent human variant
acquired milestones based on the underlying pathogenetic p.Glu210Lys in the UBTF1 gene was demonstrated because
mechanism. Currently, WES with trio analysis increasingly of an overexpression of the UBF protein. Of interest the
reveals pathogenic recurrent de novo variants in genes knockdown of UBTF in 3T3 cells was associated with the
encoding proteins responsible for the early commitment or upregulation of peroxisome proliferator-activated receptor
maturation of the neural lineage further expanding the genetic gamma coactivator 1-alpha (PPARGC1A), a master regu-
complexity of NDD.6 lator of mitochondrial biogenesis. Similarly, PPARGC1A
was upregulated in UBTF1 p.Glu210Lys human fibro-
In 2017, one of these recurrent de novo variants was recog- blasts.2 We speculate that the UBTF1 disease mechanism
nized in the UBTF1 gene as a single heterozygous change affects mitochondrial function and leads to activation of
(c.628G>A) in a highly conserved amino acid (p.Glu210Lys) compensatory upregulation of PPARGC1A and increased
located within the second HMG-box. (High-Mobility mitochondrial biogenesis, as seen in muscle biopsies from
Group box) homology box domain was first demonstrated our patients. The role of this rewiring of mitochondrial
to cause UBF gain-of function activity with a consequent function and biogenesis requires further investigations but
aberrant rRNA metabolism. This variant was associated possibly fits with the spectrum of clinical features of these
with childhood-onset neurodegenerative disorders.7 Since then, patients.
14 patients have been reported whose main clinical features are
summarized in eTable 1 (links.lww.com/NXG/A643).3,4,7-9 Our In conclusion, our study expands the clinical, metabolic, and
additional case reports contribute to defining UBTF1-related neuroradiologic findings of UBTF1 NDD contributing to
NDD as a clinical spectrum ranging from a milder (patient 1) to improving its clinical definition, suggesting mitochondrial
a more severe (patient 2) phenotype. Patient 1 presented with dysfunction as a new direction for further investigations and
motor, behavioral, and language regression, but he was still able as a potential therapeutic target.
to walk and orally feed himself at the last examination (16 years
of age). Patient 2 presented instead with a more severe rapidly Acknowledgment
progressive clinical deterioration with highly disabling dystonic F. Palombo and V. Carelli were supported by the Italian
tetraparesis and severe language and cognitive impairment Ministry of Health with the "Ricerca Corrente" funding; C.
complicated by the need for PEG tube feeding and nocturnal Garone was supported by the Minister of Italian Research
ventilation. Moreover, she presented with pseudobulbar syn- with the Rita Levi Montalcini Award; C. Garone, V. Carelli,
drome that was never reported in the UBTF1 NDD patients’ and D.M. Cordelli were supported by #NEXTGENERA-
cohort. MRI confirmed that UBTF1 NDD involve gray and TIONEU (NGEU) and funded by the Ministry of University
white matter including the cerebellum and basal ganglia. Brain- and Research (MUR), National Recovery and Resilience
stem atrophy, never reported in patients with de novo patho- Plan (NRRP), project MNESYS (PE0000006) - A Multi-
genetic variants in UBTF1, also occurred in patient 2, responsible scale integrated approach to the study of the nervous system
for severe dysphagia. The metabolic profile showed a decreased in health and disease (DN. 1553 11.10.2022); M. Maffei, M.
level in biopterin in patient 2 as also reported by Ikeda et al.,4 Giannotta, V. Carelli, and A. Pietra are part of the European
qualifying this as a potential biomarker of UBTF1 NDD. A Reference Network for neuromuscular diseases; The authors
metabolomic analysis of all UBTF1 NDD patients’ cohort may also thank Dr. Luca Soliani and Ilaria Pettenuzzo for kindly
be appropriate to fully disclose the underlying profile. We also providing patients’ videos.
showed for the first time that mitochondrial metabolism might
be also involved because the muscle biopsy of both patients Study Funding
displayed signs of compensatory mitochondrial proliferation Ricerca corrente of Italian Minister of Health- Rita Levi
with subsarcolemmal rims at cytochrome histochemical activity, Montalcini Award from the Italian Minister of Research-
abnormal ultrastructure in 1 case, as well as slightly reduced #NEXTGENERATIONEU (NGEU) funded by the Ministry
respiratory complex I activity. This suggests a potential con- of University and Research (MUR), National Recovery and
tributing role of mitochondrial dysfunction in the UBTF1 NDD Resilience Plan (NRRP), project MNESYS (PE0000006) - A
mechanism. Multiscale integrated approach to the study of the nervous
system in health and disease (DN. 1553 11.10.2022).
UBF plays an essential role in the early stage of neuro-
development. Either overexpression or abolished activity Disclosure
of UBF is responsible for a neurodegenerative process as The authors report no relevant disclosures. Go to Neurology.
demonstrated in vivo and in vitro models. Neuronal ex- org/NG for full disclosures.
pression of human UBTF1 was lethal in Drosophila spp.
Whereas tissue-specific expression in the eye caused a small- Publication History
eye phenotype with loss of photoreceptor development. Received by Neurology: Genetics April 3, 2023. Accepted in final form
Similarly, Ubtf −/− mouse is embryonic lethal while Ubtf +/− August 3, 2023. Submitted and externally peer reviewed. The handling
displayed only mild motor and behavioral dysfunction in editor was Alexandra Durr, MD, PhD.

Appendix Authors Appendix (continued)
Andrea UO Genetica Medica, IRCCS Drafting/revision of the Valerio Department of Biomedical Drafting/revision of the
Pietra, MD Azienda Ospedaliero- article for content, including Carelli, MD, and Neuromotor Sciences, article for content, including
Universitaria di Bologna; medical writing for content; PhD Alma Mater Studiorum medical writing for content;
Department of Medical and major role in the acquisition University of Bologna, Italy analysis or interpretation of
Surgical Sciences, Alma Mater of data data
Studiorum University of
Bologna, Italy Duccio Department of Biomedical Drafting/revision of the
Maria and Neuromotor article for content, including
Flavia IRCCS Istituto delle Scienze Major role in the acquisition Cordelli, Sciences, Alma Mater medical writing for content
Palombo, Neurologiche di Bologna, of data; analysis or MD, PhD Studiorum University of
PhD Programma di interpretation of data Bologna, Italy
Neurogenetica, Italy
Antonella IRCCS Istituto delle Scienze Major role in the acquisition
Melania IRCCS Istituto delle Scienze Major role in the acquisition Pini, PhD Neurologiche di Bologna, UOC of data
Giannotta, Neurologiche di Bologna, UOC of data Neuropsichiatria dell’età
MD Neuropsichiatria dell’età Pediatrica, Italy
Pediatrica, Italy
Caterina Department of Medical and Drafting/revision of the
Monica IRCCS Istituto delle Scienze Major role in the acquisition Garone Surgical Sciences, Alma Mater article for content, including
Maffei, MD Neurologiche di Bologna, of data Studiorum University of medical writing for content;
Programma di Bologna; IRCCS Istituto delle major role in the acquisition
neuroradiologia con tecniche Scienze Neurologiche di of data; study concept or
ad elevata complessità, Italy Bologna, UOC design; and analysis or
Neuropsichiatria dell’età interpretation of data
Claudio IRCCS Istituto delle Scienze Major role in the acquisition Pediatrica, Italy
Fiorini, PhD Neurologiche di Bologna, of data
Programma di
neuroradiologia con tecniche
ad elevata complessità, Italy
2. Toro C, Hori RT, Malicdan MCV, et al. A recurrent de novo missense mutation in
UBTF causes developmental neuroregression. Hum Mol Genet. 2018;27(4):691-705.
Roberta Department of Biomedical Major role in the acquisition
doi:10.1093/hmg/ddx435
Costa, PhD and Neuromotor Sciences, of data
3. Bastos F, Quinodoz M, Addor MC, et al. Childhood neurodegeneration associated
Alma Mater Studiorum
with a specific UBTF variant: a new case report and review of the literature. BMC
University of Bologna, Italy
Neurol. 2020;20(1):17. doi:10.1186/s12883-019-1586-x
4. Ikeda C, Kawarai T, Setoyama C, Orlacchio A, Imamura H. Recurrent de novo
Giovanna Department of Biomedical Analysis or interpretation of
missense variant E210K in UBTF causes juvenile dystonia-parkinsonism. Neurol Sci.
Cenacchi and Neuromotor Sciences, data
2021;42(3):1217-1219. doi:10.1007/s10072-020-04758-y
Alma Mater Studiorum
5. Cenacchi G, Peterle E, Fanin M, Papa V, Salaroli R, Angelini C. Ultrastructural
University of Bologna; UO
changes in LGMD1F. Neuropathology. 2013;33(3):276-280. doi:10.1111/neup.12003
Anatomia, Istologia
6. Wilfert AB, Sulovari A, Turner TN, Coe BP, Eichler EE. Recurrent de novo mutations
Patologica, IRCCS Azienda
in neurodevelopmental disorders: properties and clinical implications. Genome Med.
Ospedaliero-Universitaria di
2017;9(1):101. doi:10.1186/s13073-017-0498-x
Bologna, Italy
7. Edvardson S, Nicolae CM, Agrawal PB, et al. Heterozygous de novo UBTF gain-of-
function variant is associated with neurodegeneration in childhood. Am J Hum Genet.
2017;101(2):267-273. doi:10.1016/j.ajhg.2017.07.002
8. Agostini M, Romeo F, Inoue S, et al. Metabolic reprogramming during neuronal
References differentiation. Cell Death Differ. 2016;23(9):1502-1514. doi:10.1038/cdd.2016.36
1. Sanij E, Hannan RD. The role of UBF in regulating the structure and dynamics of 9. Sedláčková L, Laššuthová P, Štěrbová K, et al. UBTF mutation causes complex
transcriptionally active rDNA chromatin. Epigenetics. 2009;4(6):374-382. doi: phenotype of neurodegeneration and severe epilepsy in childhood. Neuropediatrics.
10.4161/epi.4.6.9449 2019;50(1):57-60. doi:10.1055/s-0038-1676288

Ataxia Syndrome With Hearing Loss and

Nephronophthisis Associated With a Novel
Homozygous Variant in XPNPEP3
Ilan Ben-Shabat, MD, Malin Kvarnung, MD, PhD, Wolfgang Sperker, MD, Helene Bruhn, MSc, Correspondence
Dr. Paucar
Anna Wredenberg, MD, PhD, Rolf Wibom, PhD, Inger Nennesmo, MD, PhD, Martin Engvall, MD, PhD, and
martin.paucar-arce@sll.se
Martin Paucar, MD, PhD
Abstract MORE ONLINE
Video
Objectives
Biallelic variants in XPNPEP3 are associated with a rare mitochondrial syndrome characterized
by nephronophthisis leading to kidney failure, essential tremor, hearing loss, seizures, and
intellectual disability. Only 2 publications on this condition are available. We report a man with
a complex ataxia syndrome, hearing loss, and kidney failure associated with a new biallelic
variant in XPNPEP3.
Methods
Clinical evaluation, neuroimaging studies, a kidney biopsy, and whole genome sequencing
(WGS) were applied. Since the phenotype was compatible with a mitochondrial disease, a muscle
biopsy with morphological and mitochondrial biochemical investigations was performed.
Results
Axial ataxia, cerebellar atrophy, hearing loss, myopathy, ptosis, supranuclear palsy, and kidney
failure because of nephronophthisis were the prominent features in this case. WGS revealed
the novel biallelic variant c.766C>T (p.Gln256*) in XPNPEP3. A muscle biopsy revealed COX
negative fibers, a few ragged red fibers, and ultrastructural mitochondrial changes. Enzyme
activity in respiratory chain complex IV was reduced in muscle and fibroblasts.
Discussion
This is the first report of a slowly progressive cerebellar ataxia associated with a novel biallelic
variant in XPNPEP3. Abnormalities typical for mitochondrial disease and the slow progression of
kidney disease are also striking. Our report expands the spectrum of XPNPEP3-related diseases.
Introduction
The protean symptoms and signs in mitochondrial disease include variable neurologic features. The
X-prolyl aminopeptidase 3 (XPNPEP3) gene encodes a mitochondrial aminopeptidase involved in
cleavage of matrix proteins.1 Biallelic pathogenic variants in XPNPEP3 are associated with
nephronophthisis-like nephropathy 1 (OMIM # 613159). O’Toole et al.2 reported 2 families (5
patients) featuring nephronophthisis and variable neurologic signs, whereas isolated early-onset
nephronophthisis was reported once.3 Associated symptoms include tremor, sensorineural hearing
loss, seizures, intellectual disability (ID), cardiomyopathy, and pancreatic cysts.2 We report a man
presenting with ataxia, hearing loss, myopathy, and chronic kidney failure associated with a novel
homozygous truncating variant in XPNPEP3.
From the Departments of Neurology (I.B.-S.) and Internal Medicine (W.S.), Sunderby Hospital, Luleå; Umeå University (I.B.-S.); Department of Clinical Genetics (M.K.), Centre for
Inherited Metabolic Diseases (H.B., A.W., R.W., M.E.), Karolinska University Hospital, Stockholm; Departments of Medical Biochemistry and Biophysics (A.W.), Oncology and Pathology
(I.N.), Molecular Medicine and Surgery (M.E.), and Neurology (M.P.), Karolinska Institutet, Stockholm, Sweden.

Methods Muscle tonus was normal, but strength was reduced in both
hands and legs (eAppendix 1, links.lww.com/NXG/A630),
Details provided in eAppendix 1 (links.lww.com/NXG/A630). reflexes were symmetrical, and plantar response was flexion.
Treatment with clonazepam reduced his myoclonic jerks.
Results Neurography demonstrated a moderate axonal sensorimotor
neuropathy and increased thresholds for temperature. Brain
Case Presentation MRI demonstrated severe and progressive cerebellar atrophy
A 53-year-old man from the Norrbotnian region in Sweden and global cortical atrophy grade 1–2 more pronounced in the
with a history of chronic kidney disease and gout presented frontoparietal lobes (Figure 1, A and B). Extended laboratory
with action and postural tremor, involuntary jerks, gait diffi- work-up for ataxias, including for neurometabolic disorders, did
culties, falls, dysarthria, and loss of sensation in his feet. Onset not provide any diagnostic clues other than mild CK and
was insidious and occurred during his late teens; this disorder neurofilament elevation in plasma. A recent echocardiography
was progressive and motivated the use of a walker a few months ruled out cardiomyopathy.
before his last visit. He was diagnosed with bilateral sensori-
neural hearing loss and stuttering speech at age 4 years; he A psychometric evaluation demonstrated mild impairments,
reached his developmental milestones as expected and atten- the patient obtained 76–90 index points on the working
ded a regular school but had trouble acquiring motor skills (e.g., memory part (12th percentile) and 74–88 points on the
ride a bicycle and doing winter sports). His mother and de- perceptual function part (9th percentile) of WAIS-IV. The
ceased father have ancestry in the Finnish side of the Tornio patient demonstrated perseveration, reduced concentration
valley, and none of them suffered from neurologic disease capacity and slower processing speed. Chronic kidney failure
(Pedigree shown in eFigure 1, links.lww.com/NXG/A631). At was diagnosed at age of 23 years. A kidney biopsy displayed
age 23 years, the patient was evaluated for mitochondrial dis- features interpreted as glomerulonephritis (eAppendix 1,
ease; however, a muscle biopsy and analyses of respiratory links.lww.com/NXG/A630). Estimated glomerular filtration
chain complexes were interpreted as normal then. EMG rate was 40 mL/min/1.73 m2 at age 23 years and slowly
demonstrated mild myopathic abnormalities. Protein electro- declined to 17 mL/min/1.73 m2 at his latest visit and urea was
phoresis, cytology analysis, lactate, and protein levels in the 35.6 mmol/L (normal value: 3.5–8.2 mmol/L) (Table 1).
CSF were normal. EEG demonstrated bilateral synchronous Alport’s disease was considered at this point, but no variants in
slow wave activity. Targeted analysis of common disease- COL4A5 were identified. A re-evaluation of earlier kidney
causing variants in mitochondrial DNA (mtDNA) did not biopsy was pursued based on the genetic findings. Indeed, the
detect m.3243A>G, m.8344A>G, or m.8993T>G/C. Exami- biopsy demonstrated abnormalities compatible with neph-
nation at age 53 years demonstrated predominant axial ataxia, ronophthisis. Looming dialysis has raised kidney transplantation
dysmetria, dysarthria, postural and action tremor, slow vertical into consideration.
saccades, and ptosis were also found (Video 1 and eFigure 2,
links.lww.com/NXG/A632). Examination with scale and rat- Genetic Analysis
ing of ataxia (SARA) yielded 17 points. In addition, the WGS of DNA derived from blood detected a novel homo-
Romberg test was abnormal, pinprick sensation was reduced in zygous nonsense variant, c.766C>T, (p.Gln256*), in exon 4 of
both feet, and vibratory sensation was absent in the left lateral XPNPEP3 (NM_022098.4). His mother is heterozygous
malleolus. During the examination, intermittent dystonic pos- carrier of the variant. No DNA from the father was available
tures (feet and neck) and myoclonic jerks were also found. for analysis. The variant is rare with a frequency of 24/251140
Figure 1 Neuroimaging of a Man With Ataxia Associated With a Homozygous XPNPEP3 Variant
(A) Midsagittal T1-weighted image shows severe cerebellar

atrophy. (B) Axial T1-weighted image shows global atrophy
more pronounced in the frontoparietal lobes.

isolated from fibroblasts showed decreased activity for CIV and
Table 1 Laboratory Values for a 53-Year-Old Man With a an activity in the lower normal range for complex I (CI)
Complex Neuro-renal Syndrome Associated (eFigure 3C). Coenzyme Q10 in the muscle specimen was 1.43
With a Homozygous Variant in XPNPEP3 ng/unit CS (Normal value >1.18), treatment with coenzyme
Parameter Value Normal reference value Q10 was started despite the normal value. Western blot analysis
b
confirmed loss of XPNPEP3 protein in both cultured fibroblasts
GFR 17 >60 mL/min/1.73 m2
and muscle tissue from the patient (eFigure 4, A and B, links.
Sodium 141 137–145 mmol/L lww.com/NXG/A634). In healthy controls, protein expression
Potassium 4.2 3.6–4.6 mmol/L was much lower in skeletal muscle compared to cultured fi-
broblasts. This finding is somewhat in line with previous work.
Urate 454 230–480 μmol/L
We applied the same antibody for XPNPEP3 reported in
Ionic calcium 1.18 1.15–1.33 mmol/L the Human Protein Atlas (Tissue expression of XPNPEP3 -
Phosphate 1.77b 0.75–1.40 mmol/L
Staining in skeletal muscle - The Human Protein Atlas, pro-
teinatlas.org/). In this atlas no expression of XPNPEP3 is found;
b
Hemoglobin 130 134–170 g/L however, we found a low degree of expression in skeletal muscle
MCV a
99 b
82-98 fL (eFigure 4B).
Platelet count 208 145-348 109/L
White blood cell count 7.3 3.5–8.8 109/L

Discussion
Albumin 39 36–45 g/L
This is the first report of cerebellar ataxia and cerebellar atro-
PTHb 32b 1.8–11 pmol/L
phy, myopathy and long survival in association with a biallelic
Urea 35.6b 3.5–8.2 mmol/L variant in XPNPEP3. In addition, morphological hallmarks for
Cholesterol 3.67 b
3.9–7.8 mmol/L
mitochondrial disease including ultrastructural abnormalities,
COX negative fibers and RRF, are described for the first time in
Triglycerides 2 0–2.6 mmol/L association with XPNPEP3 variants.
HDL cholesterol 1.1 0.8–2.1 mmol/L
b
The combined findings of a homozygous nonsense variant, loss of
LDL cholesterol 1.7 2.0–5.3 mmol/L
protein expression in fibroblasts and skeletal muscle, clinical
CK 13.5b 0–4.7 μkat/L presentation and abnormalities in the muscle biopsy support
pathogenicity for the c.766C>T variant in XPNPEP3. The crystal
Abbreviations: GFR = glomerular filtration rate; MCV = mean corpuscular
value; PTH = parathyroid hormone. structure of other reported variants in XPNPEP3, c.931_934del
a
MCV has been between 95-99 fL during the last year. PTH level went up to (leading to frameshift and a stop codon at amino acid position
101 pmol/L a few months later.
b
Indicates abnormal value. 316) and c.1357G>T (leading to aberrant splicing, subsequent
frameshift, and a stop codon at amino acid position 470), supports
that both variants lead to structural collapse.4 The nonsense var-
alleles in the general population and no homozygous indi- iant c.766C>T we report here leads to a stop codon and trun-
viduals have been found, according to GnomAD (v2.1.1). The cation of the protein at position 256 (p.Gln256*). The assumption
analysis detected also the heterozygous variant c.1997C>T, that mRNAs, with such early stop codons, are degraded by
(p.Ala666Val), in PUM1 (NM_001020658.2) inherited from nonsense mediated decay, is supported by loss of protein ex-
the healthy mother, which was interpreted as a variant of pression in fibroblasts and skeletal muscle from the patient.
uncertain significance (eAppendix 1, links.lww.com/NXG/ Normal findings in the muscle biopsy at age 23 years may suggest
A630). WGS of DNA extracted from muscle did not detect any that abnormalities compatible with mitochondrial disease may
pathogenic variants in mtDNA. appear after long disease duration. The protein expression of
XPNPEP3 is absent in both skeletal muscle and fibroblasts in the
Biochemical and Morphological Analysis patient and is low, but not absent, in skeletal muscle in healthy
A skin biopsy and a second muscle biopsy were obtained at age controls that could explain the muscle phenotype. The impor-
53 years. COX negative fibers and a few ragged red fibers (RRF) tance of XPNPEP3 for mitochondrial function is supported by the
were found under light microscopy, whereas electron micros- fact that deleting the orthologue gene icp55 in yeast leads to
copy demonstrated paracrystalline “parking lot” inclusions and decreased oxygen consumption and ATP synthase complex as-
rounded electron dense structures within the mitochondria as sembly.5 Signs of mitochondrial defect have also been seen in
well as abnormal cristae (Figure 2, A–D). In mitochondria other cases of XPNPEP3-associated disease. O’Toole et al.
isolated from muscle, ATP production using the complex IV reported decreased CI activity in muscle of one patient and in
dependent substrate combination TMPD+Ascorbate was re- fibroblasts from the other patient in a Turkish sibling pair har-
duced (eFigure 3A, links.lww.com/NXG/A633). Also, activity boring the homozygous c.931_934del variant. Full examination of
for the respiratory chain complex IV (CIV) was decreased the complexes of the respiratory chain has, however, not been
(eFigure 3B). Respiratory chain enzymes in mitochondria carried out in these cases, e.g., only CI was analyzed in fibroblasts.2

Figure 2 Muscle Biopsy From a Man With Ataxia Associated With a Homozygous XPNPEP3 Variant
(A) COX-SDH reaction showing fibers (blue) lacking COX-activity,

bar 200 μm. (B) A ragged red fiber is shown, Gomori trichrome,
bar 50 μm. (C) Electron microscopy showing mitochondria with
abnormal cristae. In one of these mitochondria, there is a large
dense rounded inclusion, bar 1.0 μm. (D) Mitochondria con-
taining parking lot inclusions, bar 0.5 μm.
O’Toole et al. reported variable neurologic features in asso- Other ataxia cases associated with XPNPEP3 are required to
ciation with nephronophthisis.2 In one family from northern delineate an association with this gene. Because ET, reported
Finland with 3 affected siblings, essential tremor (ET) was in previous cases, has a cerebellar generator, we suggest that
found in all, whereas 2 siblings had sensorineural hearing loss. long disease duration leading to ataxia as in our case may be
In addition, a Turkish family with 2 affected siblings suffered part of the natural history of neurologic features associated
from severe intellectual disability (ID), seizures, cardiomy- with XPNPEP3.
opathy, and pancreatic cysts.2 Neuroimaging data, which
demonstrated arachnoid cysts, was provided only for one of Acknowledgment
the affected Finnish siblings. Myoclonus, as in our patient, can The Promobilia Foundation, Region Stockholm.
be a manifestation seen in uremia. Striking findings in this case
are the slow progression of both neurologic and renal features, Study Funding
late-onset complex movement disorder, cerebellar atrophy, The authors report no targeted funding.
myopathy, clear morphological abnormalities in muscle as-
sociated with mitochondrial disease, and reduced activity in Disclosure
the CIV in muscle and fibroblasts. The rate of kidney failure The authors report no relevant disclosures. Go to Neurology.
was also slower in this case compared with previously org/NG for full disclosures.
reported who had an early need for dialysis.2,3 Taken together,
our findings expand the spectrum of disorders associated with Publication History
variants in XPNPEP3 (eTable 1, links.lww.com/NXG/A635). Received by Neurology: Genetics June 7, 2023. Accepted in final form
Early onset, intellectual disability, and cardiomyopathy were August 4, 2023. Submitted and externally peer reviewed. The handling
additional features in 2 Turkish siblings harboring a frame editor was Editor Stefan M. Pulst, MD, Dr med, FAAN.
shift variant in XPNPEP3 and deficits in CI.2 However,
genotype-phenotype correlations are not possible to establish
because of scarcity of cases. In addition, the neuropathology of Appendix Authors
this disease remains to be studied. The pattern of manifesta- Name Location Contribution
tions is striking, considering the ubiquitous pattern of ex-
Ilan Ben- Department of Major role in the acquisition
pression for XPNPEP3, including the brain6. The main Shabat, MD Neurology, Sunderby of data and analysis or
differential diagnosis includes Kearns-Sayre syndrome and Hospital, Luleå, interpretation of data
Sweden and Umeå
epilepsy, and progressive myoclonic 4, with or without renal University, Sweden
failure (OMIM # 254900).7,8

Appendix (continued) Appendix (continued)
Malin Department of Clinical Analysis or interpretation of Martin Department of Major role in the acquisition
Kvarnung, Genetics, Karolinska data Paucar, MD, Clinical Neuroscience, of data; study concept or
MD, PhD University Hospital, PhD Karolinska Institutet, design; and analysis or
Stockholm, Sweden Stockholm, Sweden interpretation of data
and Department
Wolfgang Department of Internal Analysis or interpretation of of Neurology,
Sperker, MD Medicine, Sunderby data Karolinska
Hospital, Luleå, Sweden University Hospital,
Stockholm, Sweden
Helene Centre for Inherited Analysis or interpretation of
Bruhn, MSc Metabolic Diseases, data
Karolinska University,
Stockholm, Sweden
References
Anna Centre for Inherited Analysis or interpretation of 1. Wachoski-Dark E, Zhao T, Khan A, Shutt TE, Greenway SC. Mitochondrial protein
Wredenberg, Metabolic Diseases, data homeostasis and cardiomyopathy. Int J Mol Sci. 2022;23(6):3353. doi:10.3390/
MD, PhD Karolinska University, ijms23063353
Stockholm, Sweden and 2. O’Toole JF, Liu Y, Davis EE, et al. Individuals with mutations in XPNPEP3, which
Department Medical encodes a mitochondrial protein, develop a nephronophthisis-like nephropathy. J Clin
Biochemistry and Biophysics, Invest. 2010;120(3):791-802. doi:10.1172/jci40076
Karolinska Institutet, 3. Alizadeh R, Jamshidi S, Keramatipour M, et al. Whole exome sequencing reveals a
Stockholm, Sweden XPNPEP3 novel mutation causing nephronophthisis in a pediatric patient. Iran
Biomed J. 2020;24(6):405-408. doi:10.29252/ibj.24.6.400
Rolf Wibom, Centre for Inherited Major role in the acquisition 4. Singh R, Jamdar SN, Goyal VD, Kumar A, Ghosh B, Makde RD. Structure of the
PhD Metabolic Diseases, of data human aminopeptidase XPNPEP3 and comparison of its in vitro activity with Icp55
Karolinska University, orthologs: insights into diverse cellular processes. J Biol Chem. 2017;292(24):
Stockholm, Sweden 10035-10047. doi:10.1074/jbc.m117.783357
5. Stames EM, O’Toole JF. Mitochondrial aminopeptidase deletion increases
Inger Department Oncology and Major role in the acquisition chronological lifespan and oxidative stress resistance while decreasing re-
Nennesmo, Pathology, Karolinska of data and analysis or spiratory metabolism in S. cerevisiae. PLoS One. 2013;8(10):e77234. doi:
MD, PhD Institutet, Stockholm, Sweden interpretation of data 10.1371/journal.pone.0077234
6. Ersahin C, Szpaderska AM, Orawski AT, Simmons WH. Aminopeptidase P
Martin Centre for Inherited Major role in the acquisition isozyme expression in human tissues and peripheral blood mononuclear
Engvall, MD, Metabolic Diseases, of data and analysis or cell fractions. Arch Biochem Biophys. 2005;435(2):303-310. doi:10.1016/
PhD Karolinska University, interpretation of data j.abb.2004.12.023
Stockholm, Sweden and 7. Badhwar A, Berkovic SF, Dowling JP, et al. Action myoclonus-renal failure syndrome:
Department of Molecular characterization of a unique cerebro-renal disorder. Brain. 2004;127(Pt 10):
Medicine and Surgery, 2173-2182. doi:10.1093/brain/awh263
Karolinska Institutet, 8. Govers LP, Toka HR, Hariri A, Walsh SB, Bockenhauer D. Mitochondrial DNA
Stockholm, Sweden mutations in renal disease: an overview. Pediatr Nephrol. 2021;36(1):9-17. doi:
10.1007/s00467-019-04404-6

Novel SLC13A3 Variants and Cases of Acute Reversible

Leukoencephalopathy and α-Ketoglutarate
Accumulation and Literature Review
Kristen N. Wong, MS, CGC, Lorenzo D. Botto, MD, Miao He, PhD, Peter R. Baker II, MD, Correspondence
Dr. Bonkowsky
Adeline L. Vanderver, MD, and Joshua L. Bonkowsky, MD, PhD
joshua.bonkowsky@hsc.utah.edu
Abstract
Objectives
Acute reversible leukoencephalopathy with increased urinary alpha-ketoglutarate (ARLIAK) is
a recently described autosomal recessive leukoencephalopathy caused by pathogenic variants in
the SLC13A3 gene. ARLIAK is characterized by acute neurologic involvement, often pre-
cipitated by febrile illness, with largely reversible clinical symptoms and imaging findings. Three
patients have been reported in the literature to date. Our objective is to report newly identified
patients and their genetic variants and phenotypes and review published literature on ARLIAK.
Methods
This report contributes 4 additional patients to the literature; describes novel variants in
SLC13A3; and reviews genetic, biochemical, clinical, and radiologic features of all published
patients with ARLIAK.
Results
We provide additional genetic, imaging, and laboratory insights into ARLIAK, an atypical
leukodystrophy with clinical and radiologic findings that can normalize.
Discussion
Our case series highlights the importance of reanalysis of next-generation sequencing in the
diagnostic workup.
Introduction
Leukodystrophies are heterogeneous conditions affecting the white matter of the brain,
variable in age at onset, severity, progression, and genetic etiology.1,2 Acute reversible
leukoencephalopathy with increased urinary alpha-ketoglutarate (ARLIAK) is character-
ized by neurologic involvement precipitated by febrile illness. Features include transient
leukoencephalopathy, dysarthria, altered mental status, and ataxia and increased urinary
excretion of dicarboxylic acids including alpha-ketoglutarate. Patients recover clinically
with concomitant amelioration of white matter abnormalities, whereas biochemical ab-
normalities persist.3,4
ARLIAK is autosomal recessive, caused by pathogenic variants in SLC13A3 encoding the plasma
membrane Na+/dicarboxylate cotransporter 3. Three patients are reported to date.3-5 We report
4 additional patients with novel variants in SLC13A3 and review features of all published patients.
From the Division of Pediatric Neurology, Department of Pediatrics (KNW, JLB) and Division of Genetics, Department of Pediatrics (LDB), University of Utah School of Medicine, Salt Lake City;
Division of Laboratory Medicine, Department of Pathology and Laboratory Medicine (MH), Children’s Hospital of Philadelphia, PA; Division of Clinical Genetics and Metabolism, Department
of Pediatrics (PRB), University of Colorado School of Medicine, Aurora; Department of Neurology (ALV), Perelman School of Medicine, University of Pennsylvania, Philadelphia; Division of
Neurology (ALV), Children’s Hospital of Philadelphia, PA; Center for Personalized Medicine (JLB), Primary Children’s Hospital, Salt Lake City, UT.
The Article Processing charge was funded by the authors.

Methods leukodystrophy panel, SNP microarray, mitochondrial ge-
nome, and whole-exome sequencing (WES).
This retrospective review was completed via a University of
Utah IRB–approved protocol. Consent to disclose was After discharge, she returned to neurologic baseline, though
obtained. We also reviewed published literature. Anonymized she had school difficulties. Brain MRI (Figure, D) 10 months
data not published within this article will be made available by later showed improved white matter findings with resolution
request from any qualified investigator. of diffusion restriction abnormalities.
At age 12 years, WES reanalysis detected compound het-

Results erozygous variants in SLC13A3 (Table): an intronic variant
Patients 1 and 2 c.1016+3A>G, previously reported in patients with ARLIAK;
Patient 1 is a 12-year-old girl with normal development. At age and a second variant, c.1167_1169delGTT (p.Leu389del),
3 years, she had concern for febrile seizure. Brain MRI was not previously reported in patients with ARLIAK or pop-
normal (Figure, A); EEG was abnormal with midline central ulation databases. Repeat urine organic acid testing showed
spikes. She started levetiracetam, and after 1 year of seizure persistently elevated alpha-ketoglutarate without other
freedom, parents discontinued medication. abnormalities.
At age 5 years, she presented with acute onset of fluctuating Patient 2 is patient 1’s full sister. She had mild motor delays
mental status. She had slurring of speech, partial aphasia, upper and failure to thrive with growth at first percentile for height,
extremity weakness, and absent reflexes. Parents noted she had weight, and head circumference. Testing for Russell Silver
an upper respiratory illness and fevers the week prior. Lumbar Syndrome Panel (H19 methylation and UPD7 analysis) and
puncture was normal. EEG showed slowing. Brain MRI SNP microarray were normal.
(Figure, B and C) demonstrated extensive confluent abnor-
malities of bright T2/low T1 signal involving the white matter Patient 2 was a comparator for her sister’s WES, and the same
and accompanying diffusion restriction. biallelic variants in SLC13A3 were found. After her sister’s
diagnosis, brain MRI completed at 8 years of age was normal
Testing during hospitalization and after discharge was (Figure, F). Urine organic acid testing showed elevated alpha-
normal including leukocyte lysosomal enzymes, multigene ketoglutarate without other abnormalities. She has had no
Figure MR Images
A–E, Patient 1; F, Patient 2; G–H, Pa-

tient 3; I–L, Patient 4. (A) Normal T2
image, age 2.8 years. (B) Age 5 years,
T2 FLAIR image, shows hyperintensity
in corpus callosum (arrow). (C) ADC
map, shows corresponding diffusion
restriction. (D) Age 6 years, T2 FLAIR
is normal. (E) Age 8 years, T2 FLAIR is
normal. (F) T2 FLAIR at age 8 years is
normal. (G, H) T2 FLAIR and corre-
sponding ADC map, age 21 years,
show diffusion restriction. (I, J) Age 7
years, T2 FLAIR and ADC show hyper-
intensity in the corpus callosum and
corresponding restricted diffusion.
(K, L) T2 FLAIR and ADC show
normalization.

Table Summary of Clinical, Biochemical, Imaging, and Genetic Features of New and Previously Reported Patients With
ARLIAK
Patient 1 2 3 4 5 6 7
Reference This report This report This report This report Dewulf et al.3,5 Dewulf et al.3 Kang et al.4
Episode features N/A
Febrile + + +/- + + +
Drowsiness - - + + + +
Dysarthria + + - + + -
Ataxia - + + + + +
Altered mental status + + + - + +
Weakness + - + - - -
Abnormal movements - + + - + -
Agitation - - + - - +
Hypotonia - - - - + -
Clinical symptoms reversible + + + + + +
Recurrent - + - + + +
Time between initial episode and N/A N/A 11 y N/A 12 y 6y 2y

recurrence
Urine α-ketoglutarate mmol/ 405 332 174 311 863 592 Elevated
molCr (normal <150)
Urine N-acetylaspartate Normal N/A Normal Increased Increased Increased Normal
Urine succinate Normal Normal Normal Normal Increased Normal Normal
Urine fumarate Normal Normal Normal Normal Increased Normal Normal
Treatment
First episode Acyclovir N/A Unknown Lorazepam, Levocarnitine Intravenous glucose Intravenous acyclovir, Acyclovir and
and IV fluids ceftriaxone, and intravenous
methylprednisolone cefotaxime
Additional episode N/A N/A IV fluids N/A IV fluids Intravenous Acyclovir and
ceftriaxone and intravenous
acyclovir cefotaxime
Brain MRI findings
During episode Confluent restricted diffusion N/A Confluent, Confluent restricted diffusion Bilateral symmetric Bilateral symmetric Bilateral symmetric
and T2 hyperintensity symmetric and T2 hyperintensity signal abnormalities signal abnormalities of signal abnormalities
throughout periventricular and restricted diffusion throughout periventricular and of white matter in the the white matter in of the white matter in
deep frontal and parietal white in the white matter deep frontal and parietal white periventricular periventricular periventricular
matter, with involvement of the of the frontal matter, with involvement of regions, centrum regions, centrum regions, centrum
genu of the corpus callosum parietal lobes and genu and splenium of corpus semiovale, and corpus semiovale, and corpus semiovale, and corpus
in corpus callosum callosum callosum callosum callosum
At follow-up Almost complete regression of N/A N/A Normal Almost complete Almost complete Almost complete
white matter abnormalities regression of white regression of white regression of white
matter abnormalities matter abnormalities matter abnormalities
Reversible + N/A Unknown + + + +
Genetics
Variant 1 c.1167_1169delGTT c.1167_ c.1016+3A>G, c.1016+3A>G c.761C>A c.1642G>A c.185C>T (p.T62M)

p.Leu389del, maternally 1169delGTT maternally (p.Ala254Asp) (p.Gly548Ser)
inherited p.Leu389del, inherited
maternally
inherited
Variant 2 c.1016+3A>G, paternally c.1016+3A>G, c.1033_1035del c.80T>G (p.Leu27Arg) c.761C>A c.1016+3A>G c.331C>T (p.R111*)
inherited paternally (p.Val345del), (p.Ala254Asp)
inherited paternally inherited
Other neurologic features/history
Seizure + - - + - + +
Developmental delay - +, FTT and - - - - -

mild motor
delays
Persistent cerebellar signs - - - - - + -
Abbreviations: +, presence; -, absence; FTT, failure to thrive; N/A, not applicable.

acute events or neurologic declines to date. Parents had second episode. Time between events ranged from 2 to 12
normal urinary alpha-ketoglutarate years. All but one patient (who has not yet had any acute
events) were reported to have normal development.
Patient 3
Patient 3 is a 22-year old woman referred after hospitalization All patients showed elevated urine alpha-ketoglutarate, de-
for acute onset of dysarthria, confusion, and difficulty with tectable by urine organic acid analysis, both during acute epi-
ambulation in the setting of COVID-19 with fevers. While sodes as well as between events when neurologically normal. The
hospitalized, she had normal routine laboratory values, drug 6 patients who experienced acute events had reversible white
screening, CT head, CT angiogram, and chest x-ray. Brain matter changes compared with initial presentation.
MRI revealed confluent restricted diffusion in the white
matter of the frontal parietal lobes and corpus callosum Five of the 7 patients (Table) had the previously reported
(Figure, G and H). After discharge, she returned to neurologic intronic variant c.1016+3A>G. Patients 1, 2, and 3 had single
baseline. On further review, patient reported a similar episode amino acid deletions. Patient 4 had a novel missense mutation.
at 10 years of age accompanying influenza, with slurred speech Compound heterozygous variants were confirmed in trans by
and ataxia. Genetic panel testing showed the SLC13A3 variant parental studies.
c.1016+3A>G; and a second variant, c.1033_1035del
(p.Val345del), not previously reported in population data-
bases or patient populations. Urine organic acid testing showed Discussion
elevated alpha-ketoglutarate without other abnormalities.
This study reports 4 additional cases to the previously iden-
Patient 4 tified patients with ARLIAK and expands our understanding
Patient 4 is an 11-year-old girl with normal development. of this condition. All 4 new patients had the previously
She had a single unprovoked tonic-clonic seizure at 3 years reported intronic variant in SLC13A3, c.1016+3A>G.3 This
of age. At 7 years of age, during a gastrointestinal illness with variant is present in the gnomAD population database (total
fever, she became confused, weak, ataxic, and unable to allele frequency 0.0008240) with one reported homozygous
walk. She was hospitalized and admitted to the intensive individual. All other entries in gnomAD are heterozygous.
care unit. On examination, she was somnolent, irritable, and Molecular characterization of this variant demonstrated that it
aggressive when aroused, with brisk patellar reflexes. Lum- results in 2 aberrant splicing transcripts, one lacking exon 7
bar puncture and testing for infectious and inflammatory and one lacking exons 7 and 8, and which may cause a portion
etiologies showed negative results. EEG showed diffuse of the transmembrane domain to be deleted.3 In 3 of our
background slowing. Brain MRI demonstrated extensive patients, the intronic variant was in trans with variants
confluent restricted diffusion and T2 hyperintensity of the predicted to cause a single amino acid deletion, in exons 8
white matter (Figure, I and J). and 9. Patient 4 had a previously unreported missense
variant in exon 1, a highly conserved region of the gene not
Patient 4’s symptoms resolved approximately 36 hours after found in population databases. Together with urine or-
admission. She was discharged with no residual neurologic ganic acid analysis and similarities in clinical neuro-
findings. Brain MRI repeated 6 weeks after discharge was radiologic presentations, these novel variants support a
normal (Figure, K and L). loss-of-function mechanism underlying the pathogenicity
of SLC13A3 variants.
Outpatient follow-up included whole-genome sequencing
(WGS). Two variants were found in SLC13A3: c.1016 + 3A > Four patients had seizures, both febrile and afebrile, with
G and c.80T > G (p.Leu27Arg). Urine organic acids dem- onset prior to age 5 years. To date, the seizures appear to have
onstrated elevated alpha-ketoglutarate with no additional resolved.3,4 Notably, urinary organic acids in parents of pa-
abnormalities. To date, she has not had another episode de- tients 1 and 2 (siblings) were normal, indicating that carriers
spite several febrile illnesses. She has no neurologic sequelae, do not exhibit the biochemical abnormality seen in ARLIAK.
but persistently elevated urine alpha-ketoglutarate. Patient 4 had several sets of qualitative urine organic acids
during and after her episode that did not detect alpha-
Review of Previous Cases ketoglutarate, emphasizing the importance of quantitative
Clinical, laboratory, and imaging findings of these 4 patients testing.
and the other 3 reported patients3-5 are summarized in the
Table. Five patients experienced acute onset of neurologic Our study also illustrates the value of WES/WGS reanalysis
symptoms in the setting of fever; patient 4 had a fever 2 days after nondiagnostic testing in cases with objective neurologic
prior to her episode. Common neurologic symptoms included and imaging findings6-8 and may provide important in-
drowsiness or altered mental status, dysarthria, and ataxia. formation to improve management of leukodystrophies.
Clinical signs and symptoms were reversible in all patients,
though one patient was noted to have persistent cerebellar Study Funding
signs following his initial episode. Four patients experienced a NIH, U54 NS115052.

Disclosure
The authors report no relevant disclosures. Go to Neurology. Appendix (continued)
org/NG for full disclosures. Name Location Contribution
Publication History Adeline L. Department of Neurology, Drafting/revision of the

Vanderver, Perelman School of article for content, including
Received by Neurology: Genetics May 3, 2023. Accepted in final form MD Medicine, University of medical writing for content;
August 18, 2023. Submitted and externally peer reviewed. The handling Pennsylvania; Division major role in the acquisition
of Neurology, Children’s of data
editor was Associate Editor Raymond P. Roos, MD, FAAN. Hospital of Philadelphia,
Philadelphia, PA
Joshua L. Division of Pediatric Drafting/revision of the

Bonkowsky, Neurology, Department of article for content, including
Appendix Authors MD, PhD Pediatrics, University of Utah medical writing for content;
School of Medicine; Center major role in the acquisition
Name Location Contribution for Personalized Medicine, of data; study concept or
Primary Children’s Hospital, design; and analysis or
Kristen N. Division of Pediatric Drafting/revision of the Salt Lake City, UT interpretation of data
Wong, MS, Neurology, Department of article for content, including
CGC Pediatrics, University of Utah medical writing for content;
School of Medicine, Salt Lake major role in the acquisition
City of data; and analysis or References
interpretation of data 1. Bonkowsky JL, Keller S. Leukodystrophies in children: diagnosis, care, and treatment.
Pediatrics. 2021;148(3):e2021053126. doi:10.1542/peds.2021-053126
Lorenzo D. Division of Genetics, Drafting/revision of the 2. Mahdieh N, Soveizi M, Tavasoli AR, et al. Genetic testing of leukodystrophies unraveling
Botto, MD Department of Pediatrics, article for content, including extensive heterogeneity in a large cohort and report of five common diseases and 38 novel
University of Utah School of medical writing for content; variants. Sci Rep. 2021;11(1):3231. doi:10.1038/s41598-021-82778-0
Medicine, Salt Lake City major role in the acquisition 3. Dewulf JP, Wiame E, Dorboz I, et al. SLC13A3 variants cause acute reversible leu-
of data koencephalopathy and α-ketoglutarate accumulation. Ann Neurol. 2019;85(3):
385-395. doi:10.1002/ana.25412
Miao He, Division of Laboratory Drafting/revision of the 4. Kang Q, Yang L, Liao H, et al. Case report: compound heterozygous variants of SLC13A3
PhD Medicine, Department of article for content, including identified in a Chinese patient with acute reversible leukoencephalopathy and α-keto-
Pathology and Laboratory medical writing for content; glutarate accumulation. Front Pediatr. 2021;9:801719. doi:10.3389/fped.2021.801719
Medicine, Children’s Hospital major role in the acquisition 5. Imbard A, Pernet J, Tarrano C, Lacroix D, Elmaleh-Bergès M, Schiff M. Covid-19:
of Philadelphia, PA of data; and analysis or possible trigger of SLC13A3 reversible leukoencephalopathy relapse?. Mol Genet
interpretation of data Metab. 2022;136(2):83-84. doi:10.1016/j.ymgme.2022.04.007
6. Schobers G, Schieving JH, Yntema HG, et al. Reanalysis of exome negative patients
Peter R. Division of Clinical Genetics Drafting/revision of the with rare disease: a pragmatic workflow for diagnostic applications. Genome Med.
Baker II, MD and Metabolism, article for content, including 2022;14(1):66. doi:10.1186/s13073-022-01069-z
Department of Pediatrics, medical writing for content; 7. Schmitz-Abe K, Li Q, Rosen SM, et al. Unique bioinformatic approach and com-
University of Colorado major role in the acquisition prehensive reanalysis improve diagnostic yield of clinical exomes. Eur J Hum Genet.
School of Medicine, Aurora of data; and analysis or 2019;27(9):1398-1405. doi:10.1038/s41431-019-0401-x
interpretation of data 8. Baker SW, Murrell JR, Nesbitt AI, et al. Automated clinical exome reanalysis reveals
novel diagnoses. J Mol Diagn. 2019;21(1):38-48. doi:10.1016/j.jmoldx.2018.07.008

Agenesis of Pectoralis Major Muscle in Late-Onset

GFPT1-Related Congenital Myasthenic Syndrome
A Case Report
Erika K. Williams, MD, PhD, Cristina Shea, MD, and Paloma Gonzalez-Perez, MD, PhD Correspondence
Dr. Gonzalez-Perez
Neurol Genet 2023;9:e200102. doi:10.1212/NXG.0000000000200102 pgonzalezperez@partners.org
Abstract
Objectives
The objective of this study was to expand the phenotypic spectrum of glutamine-fructose-
6-phosphate transaminase 1 (GFPT1)–related congenital myasthenia syndrome (CMS).
Methods
A 61-year-old man with agenesis of the left pectoralis major muscle presented with progressive
muscle weakness for a decade that transiently improved after exertion.
Results
His examination revealed proximal and distal muscle weakness in upper extremities and proximal
muscle weakness in lower extremities. Muscle enzymes were elevated. An electromyogram
revealed a myopathic pattern; however, a muscle biopsy of deltoid muscle and genetic testing for
limb-girdle muscular dystrophies were nondiagnostic. A 3-Hz repetitive nerve stimulation of the
spinal accessory nerve recording from trapezius muscle demonstrated a >20% drop in amplitude
of the 5th compound motor action potential relative to 1st at both baseline and after 45-second
exercise. Acetylcholine receptor binding, lipoprotein-related protein 4, muscle-specific kinase, and
voltage-gated calcium channel P/Q antibodies were negative. Genetic testing targeting CMS
revealed 2 likely pathogenic variants within GFPT1: novel c.7+2T>G (intron 1) that was pre-
dicted to result in a null allele and known c*22 C>A (exon 19) associated with reduced GFPT1
expression. His muscle strength dramatically improved after pyridostigmine initiation.
Discussion
In addition to other reported neurodevelopmental abnormalities, pectoralis major muscle
agenesis (or Poland syndrome) may be a clinical manifestation of GFPT1-related CMS.
Introduction
Congenital myasthenic syndromes (CMS) are inherited and frequently treatable neuromus-
cular junction (NMJ) disorders that are often misdiagnosed as seronegative myasthenia or
myopathy, which may delay the initiation of effective therapies.1 CMS can be caused by
defective genes encoding presynaptic, synaptic, or postsynaptic NMJ components or enzymes
that have an important role in the development and maintenance of the NMJ.1,2 Glutamine-
fructose-6-phosphate transaminase 1 (GFPT1) is a rate-limiting enzyme in the hexosamine
biosynthetic pathway responsible for the correct glycosylation of lipids and proteins of the
NMJ, which is key for its successful development and maintenance.3 GFPT1-related CMS is an
autosomal recessive disorder typically characterized by a limb-girdle muscle weakness distri-
bution that frequently presents within the first 2 decades of life and is responsive to
From the Department of Neurology (E.K.W., C.S., P.G.-P.), Massachusetts General Hospital; and Department of Neurology (E.K.W., C.S.), Brigham Women’s Hospital, Harvard Medical
School, Boston, MA.
The Article Processing charge was funded by NIH, Cleveland Clinic, Genzyme.
pyridostigmine.4 In this study, we present a patient with channel P/Q antibodies, normal activity of alpha-glucosidase,
congenital agenesis of the pectoralis major muscle and and a normal MRI of the right thigh muscles. An electromyo-
GFPT1-related CMS. gram revealed early recruitment of small motor unit potentials in
the right biceps, deltoid, infraspinatus, and iliopsoas muscles
without abnormal spontaneous activity. A muscle biopsy of the
Case Report left deltoid muscle showed mild fiber size variation and scattered
atrophic fibers as the only findings (Figure 1). He also un-
A 61-year-old athletic man reported progressive muscle derwent genetic testing for limb-girdle muscular dystrophies that
weakness for a decade. He was an avid water skier, hockey showed a variant of uncertain significance in COL12A1 (exon 18,
player, and was able to bench press 200 lbs of weight. How- c.3593>G, p.Ala1198Gly) that was unlikely to account for his
ever, in his early 50s, he experienced increasing difficulty muscle weakness.
standing up while water skiing, a slow and weak hockey shot,
and reduced grip strength. His bench press dropped to ap- At first evaluation with us, we considered the possibility of an
proximately 85 lbs. He also had to use his arms to lift his legs NMJ disorder based on the fluctuation of muscle weakness
to get in and out of his car. He noticed that his muscle with reported improvement in muscle strength after physical
weakness appeared to improve after exertion. activity. We then performed a 3-Hz repetitive nerve stimula-
tion (RNS) study of the right spinal accessory nerve recording
He denied ptosis, diplopia, speech changes, dysphagia, dyspnea, from trapezius muscle that demonstrated a significant decre-
episodes of dark urines, pain, numbness, or tingling. He was ment (>20%) in the amplitude of the fifth compound motor
born without a left pectoralis major muscle for which he un- action potential (CMAP) relative to the first CMAP at both
derwent surgery for cosmetic purposes. He had normal motor baseline and after a 45 second (postexhaustion) exercise.
development as a child and excelled athletically. He was not Electrical facilitation after a 10-second exercise was not ob-
taking any medications when he developed muscle weakness. served. A 3-Hz RNS of the right ulnar nerve recording from
His medical history included hyperlipidemia, hypertension, the abductor digiti minimi muscle was normal (Figure 2).
surgery of his cervical and lumbar spine, and thalassemia. There In view of this electrical postsynaptic dysfunction pattern,
was no family history of consanguinity or weakness. repeated acetylcholine receptor–binding antibodies and
lipoprotein-related protein 4 (LRP4) and muscle-specific ki-
His examination revealed normal cognition, language, speech, nase (MUSK) antibodies were tested and showed negative
and cranial nerves. He had normal muscle bulk and tone. There results. We then performed sequencing and deletion/
were no fasciculations or scapular winging. There was no action duplication assay of 19 genes known to cause CMS, which
or percussion myotonia or paramyotonia. Manual muscle test- revealed 2 likely pathogenic variants within GFPT1 gene: a
ing revealed the following MRC grades (R/L where applicable): novel c.7+2T>G (intron 1) canonical splice site variant that
neck flexion 5, neck extension 5, shoulder abduction 4/4, was predicted to result in a null allele and a known c*22 C>A
shoulder flexion 4+/4+, elbow flexors and extensors 5/5, finger in exon 19 (rs199678034) that had been associated with re-
extensors 4/4-, abductor digiti minimi 4+/4+, wrist extensors duced GFPT1 expression via an aberrant microRNA binding
and flexors 5/5, finger flexors 5/5, hip flexors 4-/4-, hip ab- site.3-5 Both patient’s parents were deceased and segregation
ductors 4+/4+, hip extensors 4+/4+, knee flexors 5/5, and ankle of these variants could not be investigated. We reviewed his
plantar and dorsal flexors 5/5. Although one would expect lack deltoid muscle biopsy, but tubular aggregates on electron
of adduction of the left abducted arm and lack of internal ro- microscopy were not seen.
tation of the left shoulder because of the absence of sternal and
clavicular head of pectoralis major muscle, respectively, we He was started on 60 mg of pyridostigmine 3 times a day after
suspect that he was able to perform these actions due to the which he reported dramatic improvement in his muscle
compensation of pectoralis minor muscle that was preserved. strength. He was able to shoot a hockey puck out of the rink,
Likewise, although pectoralis major muscle is the main driver of his bench press increased up to 150 lbs, and he no longer
shoulder flexion, compensation of coracobrachalis and deltoid required the use of his arms to lift his legs.
muscles probably accounted for the lack of differences in muscle
weakness between both sides. Deep tendon reflexes were 0 at
the biceps, triceps, brachioradialis, and ankles bilaterally, 2+ at
the left patella, and 1+ at the right patella, which may be at least
Discussion
partially explained by his history of cervical and lumbar spine We describe a patient with late-onset GFPT1-related CMS who
disease. Plantar responses were flexor. There was no clonus. was born with an absent left pectoralis major muscle (Poland
Coordination, sensation, and gait were normal. syndrome). Although an incidental coexistence of both condi-
tions is plausible, their rarity prompts consideration of a common
Ancillary investigations included creatine kinase 732–1262 IU/L pathogenic mechanism; impaired glycosylation due to a defective
(ref = 30–194 IU/L), aldolase 11.1 U/L (ref = <8.1 IU/L), GFPT1 (with one of the likely pathogenic variants being novel
negative antinuclear, 3-hydroxy-3-methylglutaryl-CoA reduc- and predicted to cause a null allele) may have contributed to the
tase, acetylcholine receptor–binding and voltage-gated calcium lack of appropriate pectoralis major development.

Figure 1 Left Deltoid Muscle Biopsy
Mild variation in fiber size and rare atrophic fibers

(arrows) were seen in sections stained with hema-
toxylin and eosin (A), ATPase at pH 4.3 (E), and
ATPase at pH 9.4 (F). Tubular aggregates were not
seen on trichrome (B) or electron microscopy (not
shown). Glycogen and lipid content in muscle fibers
were normal on periodic acid Schiff (C) and Oil Red
O (D), respectively. Scale bar, 250 μm.
Although GFPT1-related CMS has been associated with neu- neuron that results in increased acetylcholine release to the
rodevelopmental abnormalities such as cranial synostosis, synaptic cleft, after which, postsynaptic muscle membrane
camptodactyly, and leukoencephalopathy,4,6-8 skeletal muscle depolarization occurs.11 This effect of exercise on the NMJ
agenesis has not been reported to date. On the contrary, agenesis likely accounts for the transient improvement in strength
of unilateral pectoralis major muscle is a cardinal feature man- that the patient reported. It is plausible that this patient
datory for the diagnosis of Poland syndrome.9 The cause of could not have become symptomatic if he had had a sed-
Poland syndrome is unknown; a vascular insult during early entary life. On the contrary, and although rare, a late-onset
embryologic stages and/or as yet unidentified genetic defects GFPT1-associated CMS has been previously described
have been postulated as potential etiologies. Whereas pectoralis (Table)6,12-15 and whether the level of patients’ physical
major agenesis may be the only clinical feature of Poland syn- activity affects the age at symptom onset is uncertain.
drome (as in this patient), other congenital anomalies involv- Fortunately, he responded well to pyridostigmine, which
ing the ipsilateral thoracic wall and upper limb may occur reduces the clearance of acetylcholine from the synap-
(i.e., hypoplastic hand, symbrachydactyly, and high scapula).9,10 tic cleft and is the treatment of choice in this CMS
form.4 Other agents such as 3,4 diaminopyridine and
Exercise is known to contribute to NMJ integrity and in- salbutamol have also been tried in patients reported in
duce increased calcium influx to the presynaptic motor literature.6,8,13,15
Figure 2 Repetitive Nerve Stimulation Studies (RNS)
(A) A 3-Hz RNS of the right spinal accessory nerve recording from trapezius muscle demonstrated a 31% decrement in the amplitude of the fifth CMAP relative
to the first at baseline (left and train 1 on the right) that remained present after a 45-second exercise (trains 3 to 11 on the right). No electrical facilitation was
observed after a 10-second exercise (train 2 on the right). (B) A 3-Hz RNS of the right ulnar nerve recording from abductor digiti minimi (ADM) muscle was
normal.

Table Late-Onset (Third Decade or Later) CMS Associated With GFPT1 Pathogenic Variants
3-Hz RNS
Age at Muscle/ GFPT1
symptom Symptoms decrement Single Concentric Muscle pathogenic
onset (yr) at onset Sex CK (%) fiber needle EMG biopsy variants Treatment
This case report 51 Limb-girdle M 5X UNL Trapezius: — Myopathic pattern Deltoid: c. 7+2 T>G Pyridostigmine
muscle 31% in biceps, deltoid, minor (null allele) beneficial
weakness infraspinatus and nonspecific c. *22C>A
iliopsoas without myopathic (39UTR)
abnormal features
spontaneous No tubular
activity aggregates
were seen
on light
microscopy
or EM.
El-Wahsh et al.13 69 Limb-girdle F Normal Trapezius: EDC Myopathic pattern Deltoid: Homozygous Pyridostigmine:
muscle 35% muscle: in deltoid, EDC, minor c.1526T>C no benefit
weakness Increased and VM without nonspecific (p.Met509Thr) 3,4 DAP: partial
MCD and abnormal myopathic benefit
30% block spontaneous features Salbutamol was
activity EM: tubular planned
aggregates
Bauché et al.6 24 Limb-girdle — 4X UNL Trapezius: — Myopathic pattern — c.332 G>A Pyridostigmine
muscle 54% in deltoid, biceps, (p.Arg111His) and 3,4-DAP
weakness Anconeus: and/or quads c. *22 C>A beneficial
44% (39UTR)
Bauché et al.6 22 Limb-girdle — Normal Trapezius: — — — Homozygous Pyridostigmine

muscle 20% c.44 C>T and 3,4-DAP
weakness Anconeus: (p.Thr15Met) beneficial
14%
Natera-de 40 Limb-girdle F — — — — — c.1526T>C Pyridostigmine

Benito et al.14 muscle (p.Met509Thr) beneficial
weakness c. *22C>A
(39UTR)
Maselli et al.15 30 (sibling Limb-girdle M — — — — — IVS 7–8A>G Pyridostigmine

of muscle c. *22C>A and/or 3,4-DAP
proband) weakness (39UTR) beneficial
Guergueltcheva 40 Limb-girdle F Normal Deltoid: Abnormal — — (p. M492T) Did not receive
et al.12 muscle 12% c. *22C>A treatment
weakness (39UTR)
Abbreviations: ADM = abductor digiti minimi; CK = creatine kinase; 3,4 DAP = 3,4-diaminopyridine; EDC = extensor digitorum communis; EM = electronic
microscopy; MCD = mean consecutive difference; UNL = upper normal limit; VM = vastus medialis; — = not performed or not reported.
RNS of spinal accessory nerve suggested a postsynaptic Agenesia of skeletal muscles in patients with muscle weakness
NMJ disorder. Although mainly considered postsynaptic, an should prompt suspicion for CMS. Furthermore, GFPT1
impairment of the presynaptic NMJ components has been should be considered a candidate gene to screen in patients
described in GFPT1 mouse models.16,17 Furthermore, a with Poland syndrome.
superimposed myopathy is not uncommon in patients with
CMS; elevated muscle enzymes and myopathic pattern on Study Funding
needle electromyogram (as seen in this patient) and myo- Dr. Gonzalez-Perez is funded by the NIH/NINDS
pathologic features in biopsy (i.e., tubular aggregates) can (K23NS118048).
be seen in GFPT1-related CMS.8,12,18 It is plausible that
eventual involvement of presynaptic and skeletal muscle Disclosure
components contributed to the occurrence of patient’s The authors report no relevant disclosures. Go to Neurology.
symptoms later in life. Prompt CMS recognition might org/NG for full disclosures.
help defining disease-modifying effects that available
treatments may have if early initiated. Thus, it is possible Publication History
that development of extrasynaptic pathology accounts for Received by Neurology: Genetics May 17, 2023. Accepted in final form
resistance to acetylcholinesterase inhibitor therapy as pre- August 22, 2023. Submitted and externally peer reviewed. The handling
viously suggested.8 editor was Editor Stefan M. Pulst, MD, Dr med, FAAN.

5. Dusl M, Senderek J, Müller JS, et al. A 3’-UTR mutation creates a microRNA target
site in the GFPT1 gene of patients with congenital myasthenic syndrome. Hum Mol
Appendix Authors Genet. 2015;24(12):3418-3426. doi:10.1093/hmg/ddv090
6. Bauché S, Vellieux G, Sternberg D, et al. Mutations in GFPT1-related congenital
myasthenic syndromes are associated with synaptic morphological defects and un-
derlie a tubular aggregate myopathy with synaptopathy. J Neurol. 2017;264(8):
Erika K. Department of Neurology, Major role in the acquisition 1791-1803. doi:10.1007/s00415-017-8569-x
Williams, Massachusetts General of data; analysis or 7. Helman G, Sharma S, Crawford J, et al. Leukoencephalopathy due to variants in
MD, PhD Hospital; Department of interpretation of data GFPT1-associated congenital myasthenic syndrome. Neurology. 2019;92(6):
Neurology, Brigham e587-e593. doi:10.1212/wnl.0000000000006886
Women’s Hospital, Harvard 8. Jiang K, Zheng Y, Lin J, et al. Diverse myopathological features in the congenital
Medical School, Boston, MA myasthenia syndrome with GFPT1 mutation. Brain Behav. 2022;12(2):e2469. doi:
10.1002/brb3.2469
Cristina Department of Neurology, Major role in the acquisition 9. Baldelli I, Baccarani A, Barone C, et al. Consensus based recommendations for di-
Shea, MD Massachusetts General of data; analysis or agnosis and medical management of Poland syndrome (sequence). Orphanet J Rare
Hospital; Department of interpretation of data Dis. 2020;15(1):201. doi:10.1186/s13023-020-01481-x
Neurology, Brigham 10. Romanini MV, Calevo MG, Puliti A, et al. Poland syndrome: a proposed classification
Women’s Hospital, system and perspectives on diagnosis and treatment. Semin Pediatr Surg. 2018;27(3):
Harvard Medical School, 189-199. doi:10.1053/j.sempedsurg.2018.05.007
Boston, MA 11. Nishimune H, Stanford JA, Mori Y. Role of exercise in maintaining the integrity of the
neuromuscular junction. Muscle Nerve. 2014;49(3):315-324. doi:10.1002/mus.24095
Paloma Department of Neurology, Major role in the acquisition 12. Guergueltcheva V, Müller JS, Dusl M, et al. Congenital myasthenic syndrome with
Gonzalez- Massachusetts General of data; study concept or tubular aggregates caused by GFPT1 mutations. J Neurol. 2012;259(5):838-850. doi:
Perez, MD, Hospital, Harvard Medical design; and analysis or 10.1007/s00415-011-6262-z
PhD School, Boston, MA interpretation of data 13. El-Wahsh S, Wijesinghe R, Qiu J, Heard R, Stoll M, Reddel S. Very late-onset limb-
girdle congenital myasthenic syndrome due to GFPT1 mutation. Muscle Nerve. 2023;
68(3):E32-E34. doi:10.1002/mus.27842
14. Natera-de Benito D, Töpf A, Vilchez JJ, et al. Molecular characterization of congenital
References myasthenic syndromes in Spain. Neuromuscul Disord. 2017;27(12):1087-1098. doi:
1. Kao JC, Milone M, Selcen D, Shen X-M, Engel AG, Liewluck T. Congenital myas- 10.1016/j.nmd.2017.08.003
thenic syndromes in adult neurology clinic: a long road to diagnosis and therapy. 15. Maselli RA, Arredondo J, Nguyen J, et al. Exome sequencing detection of two un-
Neurology. 2018;91(19):e1770-e1777. doi:10.1212/wnl.0000000000006478 translated GFPT1 mutations in a family with limb-girdle myasthenia. Clin Genet.
2. Engel AG, Shen X-M, Selcen D, Sine SM. Congenital myasthenic syndromes: path- 2014;85(2):166-171. doi:10.1111/cge.12118
ogenesis, diagnosis, and treatment. Lancet Neurol. 2015;14(4):420-434. doi:10.1016/ 16. Zoltowska K, Webster R, Finlayson S, et al. Mutations in GFPT1 that underlie limb-
s1474-4422(14)70201-7 girdle congenital myasthenic syndrome result in reduced cell-surface expression of
3. Senderek J, Müller JS, Dusl M, et al. Hexosamine biosynthetic pathway mutations muscle AChR. Hum Mol Genet. 2013;22(14):2905-2913. doi:10.1093/hmg/ddt145
cause neuromuscular transmission defect. Am J Hum Genet. 2011;88(2):162-172. doi: 17. Issop Y, Hathazi D, Khan MM, et al. GFPT1 deficiency in muscle leads to myasthenia and
10.1016/j.ajhg.2011.01.008 myopathy in mice. Hum Mol Genet. 2018;27(18):3218-3232. doi:10.1093/hmg/ddy225
4. Selcen D, Shen X-M, Milone M, et al. GFPT1-myasthenia: clinical, structural, and 18. Luo H-Y, Zhao L, Mao C-Y, et al. Novel compound heterozygous GFPT1 mutations
electrophysiologic heterogeneity. Neurology. 2013;81(4):370-378. doi:10.1212/ in a family with limb-girdle myasthenia with tubular aggregates. Neuromuscul Disord.
wnl.0b013e31829c5e9c 2019;29(7):549-553. doi:10.1016/j.nmd.2019.05.008

PMPCA-Related Encephalopathy
Novel Variants, Phenotype Extension, and Mitochondrial Morphology
Vibhuti Rambani, MSc, Miriam Kolnikova, MD, PhD, Michal Cagalinec, PhD, Martina Skopkova, PhD, and Correspondence
Dr. Gasperikova
Daniela Gasperikova, PhD, DSc
daniela.gasperikova@savba.sk
Abstract
Objectives
The PMPCA gene encodes the α-subunit of mitochondrial processing peptidase (α-MPP), an
enzyme responsible for cleavage of nuclear-encoded mitochondrial precursor proteins after
their import into mitochondria. Mutations in this gene have been described in patients with
nonprogressive or slow progressive cerebellar ataxia, with variable age at onset and severity.
Cerebellar atrophy and striatum changes were found in severe cases.
Methods
The patient was diagnosed using whole exome sequencing. Skin fibroblasts were used for
confirmation of α-MPP levels using western blot and mitochondrial morphology assessment of
immunofluorescent confocal microscopy images.
Results
Two novel compound heterozygous variants in the PMPCA gene (p.Tyr241Ser and
p.Met251Val) were identified in an 8-year-old proband with progressive spastic quadriparesis,
delayed psychomotor development, and intellectual disability, with onset at 13 months. The
brain imaging showed cortical and cerebellar atrophy, reduced volume of basal ganglia with
striatum hyperintensity, and periventricular white matter changes. The patient’s fibroblasts
showed a decreased α-MPP level and reduced and fragmented mitochondria.
Discussion
The described case contributes to the number of patients with progressive PMPCA-related
disease with a severe intermediate phenotype. Moreover, we extend the phenotype to Leigh-
like white matter changes that have not been described in previously reported cases.
Introduction
Mitochondrial processing peptidase (MPP) is a heterodimeric enzyme responsible for proteolytic
cleavage of targeting presequences of nuclear-encoded mitochondrial precursor proteins after their
import into mitochondria.1 The PMPCA gene (9q34.3, MIM#613036) encodes the α-subunit of
MPP that is important for substrate recognition.2 In total, 9 different recessive variants have been
described to date in 24 patients from 9 families. They have led to a disorder with a spectrum
of symptoms from ataxia to multisystemic involvement.1,3-8 Initially, the PMPCA gene mutations
were reported to cause nonprogressive autosomal recessive cerebellar ataxia syndrome (SCAR2,
MIM#213200), with cerebellar atrophy in 17 patients from 4 families.8 Subsequently, 2 patients
from 1 family were reported with a progressive and extremely severe clinical course, with generalized
cerebral and cerebellar atrophy, profound developmental delay, optic atrophy, liver failure, re-
spiratory insufficiency, and cardiomyopathy.1 Furthermore, 5 patients with intermediate severity
(progressive but without extra-neurological symptoms) were described.3-6 Three of these patients
From the Institute of Experimental Endocrinology (V.R., M.C., M.S., D.G.), Biomedical Reserach Center, Slovak Academy of Sciences; Medical Faculty of Comenius University and
National Institute of Childern’s Diseases (M.K.); Centre of Excellence for Advanced Material Application (M.C.), Slovak Academy of Sciences, Bratislava, Slovakia.
The Article Processing Charge was funded by Biomedical Research Center SAS (APVV-22-0257).
had a combination of cerebellar atrophy and Leigh-like striatum Case Description
changes in the basal ganglia, which was proposed to represent a
hallmark of the PMPCA-associated intermediate phenotype.5 The proband is a boy from Slovakia who is the third child of
healthy nonconsanguineous parents born after an uneventful
Here we report a boy with 2 novel mutations in the PMPCA pregnancy and perinatal period. He was sitting at 6 months and
gene causing a decreased level of α-MPP and fragmentation of walking with support from 12 to 16 months. Regress in skills was
mitochondria. The patient had the intermediate phenotype noted after overcoming gastroenteritis followed by vaccination
with a severe course, and the brain imaging also included at the age of 16 months, and axial hypotony and delayed motoric
changes in the periventricular white matter, which has not development were noted. At the age of 21 months, the neuro-
been previously reported and would thus extend the clinical logic examination revealed psychomotor delay and development
picture of PMPCA-related encephalopathy. of spastic quadriparesis with no independent walking. On the
MRI at the age of 2.5 years, cerebellar atrophy and nonspecific
peritrigonal leukoencephalopathy without acute changes was
Methods visible (Figure 1A). At the age of 8 years, his state worsened after
febrile illness, and the boy could not talk or sit autonomously.
DNA Analysis Novel MRI showed significant cortical and cerebellar atrophy, a
Whole-exome sequencing (Theragen Etex, South Korea) was hyperintense signal and reduced basal ganglia volume and per-
performed using SureSelect XT V6 for library preparation and a iventricular leukoencephalopathy (Figure 1B). Metabolic in-
HiSeq 2000 Sequencer (Illumina). Candidate variants in exon 7 vestigations during life showed increased lactate in the plasma
of the PMPCA gene (NM_015160.3) were verified by Sanger (8.0 mmol/L) only once during acute deterioration.
sequencing (primers F: 59GAGAACACAGTTGGCCTCCA39
and R: 59TTCCCGCTACTTCACCTTGG39).
Western Blot
Results
50 ug of whole-cell lysates were separated using SDS-PAGE and Exome sequencing revealed the presence of 2 novel variants in the
transferred to a PVDF membrane. Rabbit anti-PMPCA primary PMPCA gene (NM_015160.3) in the proband—c.722A>C,
(NBP1-89126, Novus Biologicals, 1:1000 dilution) and anti-rabbit p.(Tyr241Ser) and c.751A>G, p.(Met251Val). Sanger sequenc-
IRDye 680LT secondary (926-68023, LI-COR, 1: 20,000 di- ing of the mother’s DNA confirmed the presence of heterozygous
lution) antibodies were used. Signals were detected and quantified single variant p.(Met251Val) in mother. Visualization of in-
using the OdysseyXF system and ImageStudioLite (LI-COR). dividual reads performed using IGV (Integrative Genomics
The α-MPP protein levels were normalized to total protein Viewer) software confirmed that variants are located in trans
staining (REVERT 700 Total Protein Stain, LI-COR). The sta- (eFigure 1, links.lww.com/NXG/A641). To evaluate the effects
tistical differences were determined using one-sample t test. of these PMPCA mutations, skin fibroblasts were established
from the patient and controls. Western blot revealed significantly
Immunostaining decreased level of α-MPP in the patient’s fibroblasts (Figure 2).
The patient and control fibroblasts were fixed with 4% Immunofluorescent labeling confirmed correct localization but a
paraformaldehyde and double-stained using a rabbit anti- decreased level of α-MPP in mitochondria and showed frag-
PMPCA antibody (NBP1-89126, Novus Biologicals, 1:200 mentation of the mitochondrial network (Figure 3A). Mito-
dilution) and mouse Total OXPHOS Rodent antibody chondrial morphometry measurements confirmed the decreased
cocktail (ab110413, Abcam, 1:200 dilution) as primary an- area of the mitochondrial network, a higher mean number of
tibodies; anti-rabbit DyLight™ 488 (35553, Invitrogen, 1:250 mitochondria per cell, a lower mean number and length of the
dilution) and anti-mouse DyLight™ 550 (SA5-10173, Invi- branches, as well as a lower mean form factor (measure of shape
trogen, 1:500 dilution) were used as secondary antibodies. complexity) and aspect ratio (measure of length-to-width ratio)
Nuclei were stained with DAPI. (Figure 3B).
Mitochondrial Morphology
A macro for ImageJ9 software provided by Merrille et al.10 was Discussion
adapted to measure and count multiple morphological pa-
rameters. We analyzed 200 immunostained cells from 3 in- Our study provides information about a patient with a PMPCA-
dependent experiments for each patient and a control sample. related disorder, thus increasing the number of reported pa-
The plots and t test were performed using the R Statistical tients to 25 (from 10 families). So far, 3 severity grades have
Software11 and the rstatix (0.7.0) and ggplot2 (3.3.6) packages. been described. The milder nonprogressive ataxia in 17 patients
from 4 families,8 extremely severe progressive mitochondrial
Ethics Approval and Consent to Participate encephalopathy with multisystemic involvement in 2 siblings,1
The study was conducted according to the guidelines of the and an intermediate phenotype of progressive encephalopathy
Declaration of Helsinki. All individuals or their legal repre- with psychomotor regression, intellectual disability, and spastic
sentatives (in participants younger than 18 years) signed in- ataxia without extra-neurological signs in 8 patients from
formed consent to genetic and fibroblasts studies. 6 families,3-6 including this study. However, the course of the

Figure 1 Brain MRI of the Patient
(A) At the age of 3 years—coronal and

sagittal T1-weighted images show a re-
duced infratentorial volume of cerebellar
hemispheres and cerebellar atrophy with
accentuated gyrification (arrows). On ax-
ial FLAIR sequences, increased bilateral
peritrigonal white matter changes (dou-
ble arrow) without acute manifestations
were visible. No structural signal changes
of the putamen were present. (B) At the
age of 8 years—coronal and sagittal T1-
weighted images show reduced volume
of the vermis and significant atrophy of
the cerebellar hemispheres. Axial FLAIR
images show persistent increase in
the periventricular signal (double arrow)
and newly revealed bilateral symmetric
hyperintensity in the caput nuclei caudati
and putamen. The main radiologic dif-
ference at follow-up at the age of 8 years
is the increase in the striatal signal ab-
normality and increasing atrophy of the
cerebellar hemispheres.
disease in our patient is more severe than that of the most can also be seen in Leigh syndrome cases.12 Hence, our data
patients labeled as intermediate, similar to the patient described support the suggestion that cerebellar atrophy and Leigh-like
in a study,6 who never achieved independent walking. It is basal ganglia involvement are the hallmarks of the intermediate
probable that patients with a clinical picture across this wide PMPCA phenotype and further propose that white matter
spectrum of severity will be identified in the future. changes may be part of this picture as well.
The authors of a study5 pointed to a specific combination of Mutations in the PMPCA gene often result in a decreased level
signs on brain imaging, including cerebellar atrophy and Leigh- of α-MPP protein,1,5,8 but increased5 or unaffected levels were
like basal ganglia changes, which was present in 3 patients seen.3 We confirmed a decreased level of α-MPP in patient
with the intermediate phenotype.4-6 In agreement with this, we fibroblasts in whole-cell lysate (Figure 2), and it was also ap-
observed hyperintensities and reduced volume of basal ganglia parent in the immunostaining images (Figure 3A). In addition,
in our patient, as well, but only after the disease had progressed mitochondria in our patient’s fibroblasts showed significant
in his eighth year of life. Furthermore, the brain MRI showed fragmentation (Figure 3, A and B), which has not been reported
hyperintensities in the periventricular white matter, which in previous studies. Thus far, none of the studies with patients
Figure 2 Western Blot Assessment of α-MPP Levels
α-MPP level is decreased in patient fibroblasts (P) compared with control samples (C1–C6, description in eTable 1, links.lww.com/NXG/A641). A representative image
of the western blot membranes is given along with statistics from 3 technical replicates plotted as the ratio of the intensity of the PMPCA signal (green) normalized to
total protein (red) and relative to sample C1. Error bars represent SD, **p < 0.01, 1-sample t test. α-MPP = α-subunit of mitochondrial processing peptidase.

Figure 3 Immunofluorescence Staining
(A) Immunofluorescence staining shows fragmentation of mitochondria in the patient fibroblasts. The α-MPP levels are lower in the patient, but analysis of
fluorescence profiles (white line in the merged images) shows the correct localization of α-MPP (green) in mitochondria (red). (B) Mitochondrial morphometry
measurements confirm statistically significant differences in the mitochondrial morphology of the patient and control fibroblasts. Each dot represents the
average value for a single measured cell. α-MPP = α-subunit of mitochondrial processing peptidase.
with the intermediate form performed immunostaining. Pre- of these particular proteins. This is, however, in agreement with
viously, swollen mitochondria with α-MPP accumulation were the authors of a study8 who showed that only 1 of 4 tested
reported in fibroblasts from a patient with the extremely severe targeted proteins, frataxin, revealed presence of unprocessed
form of the disease,1 and no morphological changes or accu- forms. Moreover, the initial cleavage step from precursor FXN1-
mulation were seen in a patient with the milder nonprogressive 210 to intermediate FXN42-210 appeared to be intact, and only
ataxia.8 In agreement with this, patients with the milder form do the subsequent cleavage to FXN81-210 was impaired. Given the
not show any typical mitochondrial signs, while patients with the vital biological function of MPP, it is unlikely that mutations that
intermediate or more severe forms present various degrees of would cause complete loss of MPP would be compatible with life.
increased lactate and progressive course of the disease.1,5 It is also possible that the reduced MPP function may not man-
ifest itself at the steady-state level, but only under stress condi-
To assess potential functional effect of variants, we added analyses tions, as indicated by the patient’s clinical course with worsening
of the steady-state levels of the mature forms of 2 nuclear-encoded of symptoms after intercurrent illnesses.
mitochondrial proteins whose processing requires MPP: heat-
shock protein 60 (HSP60) and the mitochondrial transcription The described case extends the number of patients with pro-
factor A (TFAM) in patient and controls. In addition, we assessed gressive PMPCA-related disease and the severe intermediate
also levels of VDAC1 that does not undergo cleavage of mito- form, where a patient is unable to walk independently but has no
chondrial targeting sequence by MPP. The levels of HSP60 and extra-neurological signs present in the extremely severe form. We
TFAM appeared similar in patient fibroblasts compared with show that apart from the typical cerebellar atrophy and Leigh-like
controls (eFigure 2, links.lww.com/NXG/A641), suggesting that striatum changes, the brain imaging can also include white matter
the variants do not alter the steady-state levels of processed forms changes in patients with progressive PMPCA-related disorder.

Acknowledgment
The authors thank the patient’s family for their kind collaboration. Appendix (continued)
The authors thank Dr. Eva Kutejova for kindly providing the Name Location Contribution
HSP60 and TFAM antibodies and control proteins.
Martina Institute of Experimental Drafting/revision of the
Skopkova, Endocrinology, Biomedical manuscript for content,
Study Funding PhD Reserach Center, Slovak including medical writing for
This work was supported by APVV-0296-17, VEGA 1/0572/21, Academy of Sciences, content; study concept or
Bratislava, Slovakia design; analysis or
APVV-22-0257, ITMS: 313021BZC9, ITMS: 313021T081. interpretation of data
Daniela Institute of Experimental Drafting/revision of the

Disclosure Gasperikova, Endocrinology, Biomedical manuscript for content,
The authors report no relevant disclosures. Go to Neurology. PhD, DSc Reserach Center, Slovak including medical writing for
Academy of Sciences, content; study concept or
org/NG for full disclosures. Bratislava, Slovakia design
Publication History
Received by Neurology: Genetics May 25, 2023. Accepted in final form
References
August 28, 2023. Submitted and externally peer reviewed. The handling 1. Joshi M, Anselm I, Shi J, et al. Mutations in the substrate binding glycine-rich loop of the
editor was Editor Stefan M. Pulst, MD, Dr med, FAAN. mitochondrial processing peptidase-α protein (PMPCA) cause a severe mitochondrial
disease. Cold Spring Harb Mol Case Stud. 2016;2(3):a000786. doi:10.1101/mcs.a000786
2. Gakh O, Obsil T, Adamec J, et al. Substrate binding changes conformation of the α-,
but not the β-subunit of mitochondrial processing peptidase. Arch Biochem Biophys.
2001;385(2):392-396. doi:10.1006/abbi.2000.2167
Appendix Authors 3. Choquet K, Zurita-Rendón O, La Piana R, et al. Autosomal recessive cerebellar ataxia
caused by a homozygous mutation in PMPCA. Brain. 2016;139(3):e19. doi:10.1093/
Name Location Contribution brain/awv362
4. Rubegni A, Pasquariello R, Dosi C, et al. Teaching NeuroImages: Leigh-like features
Vibhuti Institute of Experimental Drafting/revision of the expand the picture of PMPCA-related disorders. Neurology. 2019;92(2):e168-e169.
Rambani, Endocrinology, Biomedical manuscript for content, doi:10.1212/WNL.0000000000006740
MSc Reserach Center, Slovak including medical 5. Serpieri V, Biagini T, Mazzotta C, et al. Phenotypic definition and genotype-phenotype
Academy of Sciences, writing for content; major correlates in PMPCA-related disease. Appl Sci. 2021;11(2):748. doi:10.3390/app11020748
Bratislava, Slovakia role in the acquisition 6. Takahashi Y, Kubota M, Kosaki R, Kosaki K, Ishiguro A. A severe form of autosomal
of data; analysis or recessive spinocerebellar ataxia associated with novel PMPCA variants. Brain Dev.
interpretation of data 2021;43(3):464-469. doi:10.1016/j.braindev.2020.11.008
7. Yoon G, Delague V, Mégarbané A, Isaya G. Reply: autosomal recessive cerebellar
Miriam Medical Faculty of Comenius Drafting/revision of the ataxia caused by a homozygous mutation in PMPCA. Brain. 2016;139(Pt 3):e20. doi:
Kolnikova, University and National manuscript for content, 10.1093/brain/awv363
MD, PhD Institute of Childern’s including medical writing 8. Jobling RK, Assoum M, Gakh O, et al. PMPCA mutations cause abnormal mito-
Diseases, Bratislava, for content; major role in chondrial protein processing in patients with non-progressive cerebellar ataxia. Brain.
Slovakia the acquisition of data; 2015;138(Pt 6):1505-1517. doi:10.1093/brain/awv057
analysis or interpretation 9. Valente AJ, Maddalena LA, Robb EL, Moradi F, Stuart JA. A simple ImageJ macro tool
of data for analyzing mitochondrial network morphology in mammalian cell culture. Acta
Histochem. 2017;119(3):315-326. doi:10.1016/j.acthis.2017.03.001
Michal Institute of Experimental Analysis or interpretation of 10. Merrill RA, Flippo KH, Strack S. Measuring mitochondrial shape with ImageJ. In:
Cagalinec, Endocrinology, Biomedical data Techniques to Investigate Mitochondrial Function in Neurons, Neuromethods, Vol 123.
PhD Reserach Center; Centre of Springer Protocols. 2017. doi:10.1007/978-1-4939-6890-9_2
Excellence for Advanced 11. Team RC. R: A Language and Environment for Statistical Computing. R Foundation for
Material Application, Slovak Statistical Computing. R-project.org/
Academy of Sciences, 12. Topçu M, Saatci I, Apak RA, Söylemezoglu F, Akçören Z. Leigh syndrome in a 3-year-
Bratislava, Slovakia old boy with unusual brain MR imaging and pathologic findings. AJNR Am J Neu-
roradiol. 2000;21(1):224-227.

NEUROIMAGE OPEN ACCESS
FOLR1 Gene Variation With Adult-Onset Cerebral

Folate Deficiency and Stable Clinical and MRI Features
up to 2 Years
Carlo Manco, MD, Rosa Cortese, MD, PhD, Manfredi Alberti, Dr, Silvia Bianchi, PhD, Lucia Monti, MD, Correspondence
Dr. Cortese
Nicola De Stefano, MD, PhD, and Carla Battisti, MD, PhD
rosa.cortese@unisi.it
Abstract
Objectives
The objective of this case report was to describe the first report of FOLR1 variants associated
with adult-onset paucisymptomatic leukoencephalopathy associated with cerebral folate de-
ficiency (CFD).
Methods
Considering the patient’s symptoms, a nonprogressive leukoencephalopathy was suspected.
CSF 5-methyltetrahydrofolate levels were low (10 nmol/L, normal range 41–117). With no
other identifiable causes, a genetic analysis was conducted, revealing a compound heterozygous
FOLR1 variation (c.45G>T and c. 493+2T>C).
Results
A 47-year-old man with a history of drug and alcohol abuse was admitted to the hospital for
double vision and postural instability. MRI of the brain was performed, which showed bilateral
leukoencephalopathy. Diffusion tensor imaging revealed a diffuse reduction in fractional an-
isotropy, suggesting microstructural changes. MRI of the brain and overall clinical picture were
stable on subsequent serial examinations.
Discussion
Scientific evidence supports the deleterious effect of c.45G>T and c.493+2T>C variations on
the folate receptor-α (FRα) protein structure and function. The weakness of the expression and
function of FRα without elimination of its function caused by specific compound heterozygous
variations may explain the atypical features observed in our patient. Although rare, CFD should
be considered in paucisymptomatic adult patients with stable diffuse MRI white matter changes.
Introduction
FOLR1 gene variations are commonly associated with cerebral folate deficiency (CFD), a rare
neurologic syndrome characterized by low CSF concentration of 5-methyltetrahydrofolate (5-
MTHF) despite normal peripheral folate metabolism.1 CFD typically manifests in early infancy
with symptoms, such as irritability, sleep disturbances, and subsequently progresses to severe
epilepsy, cerebellar ataxia, and psychomotor retardation.1 MRI of the brain usually shows
diffuse, leukodystrophy-like, white matter changes. In children, treatment with oral calcium
folinate or folinic acid has shown improvement in clinical symptoms, as well as MRI and EEG
abnormalities.1 To date, clinical and imaging features associated with adult-onset CFD and
FOLR1 gene variations have not been described.
From the Centre for Precision and Translational Medicine (C.M., R.C., S.B., N.D.S., C.B.), Department of Medicine, Surgery and Neuroscience, University of Siena; Neurology Unit
(M.A.), Department of Neurology and Human Movement Sciences, University Hospital of Siena; Department of Medical, Surgical and Neurological Science (M.A.), University of Siena;
and Diagnostic and Functional Neuroimaging Unit (L.M.), Department of Neurology and Human Movement Sciences, University Hospital of Siena, Italy.

Glossary
5-MTHF = 5-methyltetrahydrofolate; CFD = cerebral folate deficiency; DTI = diffusion tensor imaging; FRα = folate receptor-α.
Methods and Results performed were stable (Figure, B) than a nonprogressive leu-
We report on a 47-year-old man who presented at the emer- koencephalopathy was hypothesized.2 A further CSF analysis
gency department with a four-day history of slight double vi- showed reduced 5-MTHF levels (10 nmol/L, normal values:
sion and postural instability. In his medical history, he reported 41–117), and genetic testing with next-generation sequencing
psoriasis, drug abuse up to age 21 years and alcohol addiction of selected genes revealed the heterozygous variation of the
for many years. The neurologic examination showed walking FOLR1 double gene (c.45G>T and c.493+2T>C); therefore,
ataxia and vertical diplopia in the left/upper left gaze position. A CFD was diagnosed. The patient’s father, mother, and 2
first brain MRI showed diffuse, bilateral, and symmetric daughters were also assessed. All were symptom-free, had
supratentorial hyperintensity on FLAIR images, without en- normal neurological examination, and unremarkable blood
hancement after gadolinium injection (Figure, A). MRI of the tests. MRI results showed no abnormalities. The genetic
spinal cord was unremarkable. Diffusion-tensor imaging (DTI) analysis was performed also in the asymptomatic father,
was also acquired and showed a lower fractional anisotropy in mother, and 2 daughters, with the father being a carrier of the
both abnormal and normal-appearing white matter in our pa- variant c.45G > T and 2 daughters and mother being carriers of
tient (Figure, C) when compared with the same brain regions the c.493+2T>C variation. Written informed consent was
of a sex-matched and age-matched healthy control (Figure, D). obtained from the patient for the case presentation.
During the hospitalization, the patient was tested for blood

and urine biochemical routine, autoimmunity screening, the Discussion
dosage of folate and lysosomal enzymes, β-galactocere-
brosidase, pyruvate and lactate, and the assay of serum heavy FOLR1 gene variations are associated with CFD, an autoso-
metal: All measures were within normal ranges. mal recessive disorder characterized by low CSF 5-MTHF
concentrations, normal plasma folate values, and late infantile
CSF analyses revealed a mild increase in protein levels onset with severe developmental regression, epilepsy, and
(51,70 mg/dL, n.v. 20–40) with normal cell count. At 2 years of leukodystrophy.3,4 Folic acid and folates are essential for
follow-up, the neurologic examination and brain MRI neurodevelopment and myelin formation5,6 after intestinal
Figure MRI FLAIR Axial Images Performed at Symptoms Onset and 2 Years Later and Correspondent Baseline Diffusion
Tensor Imaging (DTI) Analysis Comparing the Patient With a Sex-Matched and Age-Matched Healthy Control (HC)
Subject
MRI of the brain showed diffuse, bilateral, and symmetric supratentorial hyperintensities on FLAIR images, without enhancement after gadolinium injection
(A) stable 2 years later (B). DTI revealed a diffuse decrease in fractional anisotropy in the CFD patient (C) when compared with a HC (D), suggesting a
widespread microstructural damage beyond the lesional tissue.

absorption and are reduced, methylated, and enters the cir- presenting with neurologic symptoms and exhibiting a leukodys-
culation as 5-MTHF. The FOLR1 gene encodes for the folate trophy pattern on MRI. This case underscores the need for in-
receptor-α (FRα), which is essential in allowing 5-MTHF to creased awareness and recognition of CFD in the adult population
cross the blood-brain barrier.7 to ensure appropriate diagnosis and management of this condition.
Our patient is the first reported case in the literature of a Study Funding
compound heterozygous variation of FOLR1 associated with The authors report no targeted funding.
adult-onset leukoencephalopathy and clinical paucisympto-
matic picture. Based on the available data, patient’s conditions Disclosure
have remained stable over time in the absence of treatment. The authors report no relevant disclosures. Go to Neurology.
Despite the mild clinical involvement, MRI changes were org/NG for full disclosures.
diffuse and DTI analysis showed the presence of diffuse mi-
crostructural changes beyond lesional white matter. Publication History
Received by Neurology: Genetics July 21, 2023. Accepted in final form
From a genetic standpoint, in our patient, we identified 2 September 1, 2023. Submitted and externally peer reviewed. The
genetic variations. The first variation is the c.45G>T variation, handling editor was Editor Stefan M. Pulst, MD, Dr med, FAAN.
which is categorized as a missense variation. The second
variation is the c.493+2T>C variation, affecting the donor site
of the splice. In the ClinVar database, the c.45G>T variation Appendix Authors
has been reported with supporting evidence indicating its
deleterious effects on protein structure and function, as sug- Name Location Contribution
gested by silicon analysis. In addition, Grapp et al. conducted a Carlo Centre for Precision and Drafting/revision of the
study using FRα expression model cell systems and fibroblasts Manco, MD Translational Medicine, manuscript for content,
Department of Medicine, including medical writing for
from a cohort of patients with missense variations in the Surgery and Neuroscience, content; major role in the
FOLR1 gene. Their findings revealed that although FRα was University of Siena, Italy acquisition of data; study
concept or design; analysis
expressed, it was not properly localized to the cell membrane. or interpretation of data
Instead, it was misdirected to various intracellular compart-
Rosa Centre for Precision and Drafting/revision of the
ments, leading to a reduction in folic acid binding, thus Cortese, Translational Medicine, manuscript for content,
compromising its primary target.7 MD, PhD Department of Medicine, including medical writing for
Surgery and Neuroscience, content; analysis or
University of Siena, Italy interpretation of data
The c.493+2T>C variation, despite having a total frequency
Manfredi Neurology Unit, Department of Major role in the acquisition
of 0.3123% in the general population according to the ge- Alberti, Dr Neurology and Human of data, analysis or
nome aggregation database, could potentially lead to the loss Movement Sciences, University interpretation of data
of protein function and contribute to the development of the Hospital of Siena; Department
of Medical, Surgical and
disease. However, conflicting data regarding the true patho- Neurological Science,
genicity of this variation have been reported in the University of Siena, Italy
literature.8,9 Prediction software for splice alterations, such as Silvia Centre for Precision and Major role in the acquisition
BDGP and ESEfinder, suggests that the c.493+2T>C varia- Bianchi, Translational Medicine, of data; analysis or
PhD Department of Medicine, interpretation of data
tion may disrupt or weaken the native splice donor site. Surgery and Neuroscience,
University of Siena, Italy
Therefore, one possible explanation for the atypical features Lucia Diagnostic and Functional Analysis or interpretation of
observed in our patient could be that the presence of these specific Monti, MD Neuroimaging Unit, data
Department of Neurology and
compound heterozygous variations weakens the expression and Human Movement Sciences,
function of FRα without completely eliminating its function. This University Hospital of Siena, Italy
partial impairment could result in a milder phenotype with late- Nicola De Centre for Precision and Drafting/revision of the
onset symptoms and a relatively limited number of symptoms Stefano, Translational Medicine, manuscript for content,
MD, PhD Department of Medicine, including medical writing for
(pauci-symptomatic phenotype). The variations may cause a Surgery and Neuroscience, content; study concept or
blurred picture of the disease presentation, deviating from the University of Siena, Italy design
typical pattern seen in cases with a complete loss of function. This
Carla Centre for Precision and Drafting/revision of the
phenomenon highlights the complexity of genotype-phenotype Battisti, Translational Medicine, manuscript for content,
correlations and suggests that variations in the degree of protein MD, PhD Department of Medicine, including medical writing for
Surgery and Neuroscience, content; study concept or
impairment can lead to diverse clinical manifestations. University of Siena, Italy design
In conclusion, this clinical case is an example of how variations in

the FOLR1 gene can be present in adults with limited clinical
findings and a stable disease course. Thus, it highlights the im- References
1. Hyland K, Shoffner J, Heales SJ. Cerebral folate deficiency. J Inherit Metab Dis. 2010;
portance of considering the diagnosis of CFD in adult patients 33(5):563-570. doi:10.1007/s10545-010-9159-6

2. Nicolai J, van Kempen MJ, Postma AA. Teaching Neuro Images: white matter 6. Grapp M, Just IA, Linnankivi T, et al. Molecular characterization of folate receptor 1
hypomyelination and progressive calcifications in cerebral folate deficiency. Neurology. variations delineates cerebral folate transport deficiency. Brain. 2012;135(Pt 7):
2016;87(1):e4-e5. doi:10.1212/WNL.0000000000002805 2022-2031. doi:10.1093/brain/aws122
3. Steinfeld R, Grapp M, Kraetzner R, et al. Folate receptor alpha defect causes cerebral folate 7. Spector R, Johanson CE. Vectorial ligand transport through mammalian choroid
transport deficiency: a treatable neurodegenerative disorder associated with disturbed plexus. Pharm Res. 2010;27(10):2054-2062. doi:10.1007/S11095-010-0162-2
myelin metabolism. Am J Hum Genet. 2009;85(3):354-363. doi:10.1016/j.ajhg.2009.08.005 8. Najmabadi H, Hu H, Garshasbi M, et al. Deep sequencing reveals 50 novel genes
4. Molero-Luis M, Serrano M, O’Callaghan MM, et al. Clinical, etiological and thera- for recessive cognitive disorders. Nature. 2011;478(7367):57-63. doi:10.1038/
peutic aspects of cerebral folate deficiency. Expert Rev Neurother. 2015;15(7):793-802. nature10423
doi:10.1586/14737175.2015.1055322 9. Ramaekers VT, Segers K, Sequeira JM, et al. Genetic assessment and folate receptor
5. Ramaekers VT, Quadros EV. Cerebral folate deficiency syndrome: early diagnosis, in- autoantibodies in infantile-onset cerebral folate deficiency (CFD) syndrome. Mol
tervention and treatment strategies. Nutrients. 2022;14(15):3096. doi:10.3390/nu14153096 Genet Metab. 2018;124(1):87-93. doi:10.1016/J.YMGME.2018.03.001

NXG 2023 9 Issue-6

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

NXG 2023 9 Issue-6

Uploaded by

Copyright:

Available Formats

Volume 9, Number 6, December 2023

A peer-reviewed clinical and translational neurology open access journal

Sharon L. Quimby, Digital Managing Editor

Editor-in-Chief Classification of Evidence

Deputy Editor Malik M. Adil, MD

Research Articles e200113 Genetic Patterns of Selected Muscular Dystrophies

e200104 FOLR1 Gene Variation With Adult-Onset Cerebral

IRF2BPL Causes Mild Intellectual Disability Followed

Neurol Genet 2023;9:e200096. doi:10.1212/NXG.0000000000200096

The Article Processing Charge was funded by the authors.

2 Neurology: Genetics | Volume 9, Number 6 | December 2023 Neurology.org/NG

Figure 2 Cerebral MRI From SAL-394-048 After 10 Years of Ataxia Duration

Neurology.org/NG Neurology: Genetics | Volume 9, Number 6 | December 2023 3

Neurology: Genetics | Volume 9, Number 6 | December 2023

021 NA NA NA NA Normal Mild No Normal extensor Not testable No

033 No Yes (no No work No writing, no No speech No Increased NA No NA

034 No No Adapted Yes (gardener) Only his name No No Increased Normal No NA

040 NA NA Adapted Yes (legally Difficulties No No (severe/6 at Increased No NA

044 NA NA Normal NA Normal No Mild sway at age Increased, Normal No No

050 No Yes Adapted Yes (legally Difficult No No Increased Normal No No

IIB No No Difficult No Normal No NA NA NA NA NA

Neurology: Genetics | Volume 9, Number 6 | December 2023

6 Neurology: Genetics | Volume 9, Number 6 | December 2023 Neurology.org/NG

Name Location Contribution

Solveig Genetic Department, Assistance Conception and design of

Neurology.org/NG Neurology: Genetics | Volume 9, Number 6 | December 2023 7

mTOR Pathway Somatic Pathogenic Variants in Focal

Neurol Genet 2023;9:e200103. doi:10.1212/NXG.0000000000200103

*These authors contributed equally to this work as cofirst authors.

The Article Processing Charge was funded by authors.

Introduction gene.27-31 To date, the underlying cause of FMCDs is found in

2 Neurology: Genetics | Volume 9, Number 6 | December 2023 Neurology.org/NG

Neurology.org/NG Neurology: Genetics | Volume 9, Number 6 | December 2023 3

4 Neurology: Genetics | Volume 9, Number 6 | December 2023 Neurology.org/NG

Neurology: Genetics | Volume 9, Number 6 | December 2023

15 20 y, F 8 y, 14 Right frontal DRE (6 y) Normal R frontal FCD FCD type 1 ECI (6 y)

16 11 y, M 3 y, 7 y Left Fronto-central- Normal Normal FCD type 1 ECI (2 mo)

17 14 y, F 2 y, 11 y Left SMA DRE (9 y) Normal Normal FCD type 2 ECI (4 y)

18 14 y, F 10 y, 11 y Right SMA DRE (1 y) Normal Normal FCD type 1 ECI (3 y)

6 Neurology: Genetics | Volume 9, Number 6 | December 2023 Neurology.org/NG

Figure 3 Functional Validation of Candidate MTOR Variants

(A) Western blots of flag-tagged mutant

Neurology.org/NG Neurology: Genetics | Volume 9, Number 6 | December 2023 7

8 Neurology: Genetics | Volume 9, Number 6 | December 2023 Neurology.org/NG

Of interest all patients with no identiﬁed pathogenic variant

Neurology.org/NG Neurology: Genetics | Volume 9, Number 6 | December 2023 9

10 Neurology: Genetics | Volume 9, Number 6 | December 2023 Neurology.org/NG

Name Location Contribution Name Location Contribution

Neurology.org/NG Neurology: Genetics | Volume 9, Number 6 | December 2023 11

12 Neurology: Genetics | Volume 9, Number 6 | December 2023 Neurology.org/NG

Neurology.org/NG Neurology: Genetics | Volume 9, Number 6 | December 2023 13

Adult Phenotype of SYNGAP1-DEE

Neurol Genet 2023;9:e200105. doi:10.1212/NXG.0000000000200105

The Article Processing Charge was funded by the authors.

Introduction was obtained from the substitute decision makers of all

2 Neurology: Genetics | Volume 9, Number 6 | December 2023 Neurology.org/NG

Neurology.org/NG Neurology: Genetics | Volume 9, Number 6 | December 2023 3

1 F 65 Yes White No Exon 8 c.1329_ p.Lys444Glyfs*27 Frameshift leading to Unknown Pathogenic No

4 F 48 Yes Other No Exon 8 c.870del p.Tyr291fs Frameshift De novo Pathogenic No

Neurology: Genetics | Volume 9, Number 6 | December 2023

12 M 20 Yes White No Exon c.3227delT p.Leu1076* Nonsense De novo Pathogenic No

13 M 18 No White No Exon 3 c.190-15_ — Insertion/deletion (indel) De novo Pathogenic No

Neurology.org/NG Neurology: Genetics | Volume 9, Number 6 | December 2023 5