Partial biotinidase deficiency' Clinical

and biochemical features

Juiie R. Secor McVoy, BS, Harvey L. Levy, MD, Michael Lawler, BA,
M!chael A. Schmidt, MD, PhD, Douglas D. Ebers, MD, P. Suzanne Hart, BS,
Denise Dove Pettit, BS, Miriam G. Blitzer, PhD, a n d Barry Wolf, MD, PhD
From the Departments of Human Genetics and Pediatrics, Medical College oi Virginia, Rich-
mond; the State Laboratory Institute, Massachusetts Department of Public Health; Massachu-
setts General Hospital, Harvard Medical School~ Boston; the Lincoln Cliriic and the Lincoln Pe-
diatric Group, Lincoln, Nebraska; and the Division of Human Genetics, University of Maryland
School of Medicine, Baltimore

Neonatal screening f o r profound biotinidase deficiency (<10% of the mean

normal activity level) has identified a group of children with partial biotinidase
deficiency (10% to 30% of mean normal activity). Because partial biotinidase
deficiency may result in clinical consequences that may be prevented by
treatment with biotin, we evaluated such individuals and their family members
(I) to determine whether partial biotinidase deficiency is associated with symp-
toms and (2) to determine the inheritance pattern. We quantified serum biotini-
dase activity levels and obtained medical histories of probands, their parents
and siblings, and additional family members. All children with partial deficiency
were healthy at the time of diagnosis. One child, who was not initially treated
with biotin, later developed hypotonia, hair loss, and skin rash, which resolved
with biotin therapy. Four adults and three children with partial biotinidase defi-
ciency were identified among family members of infants identified by neonatal
screening. All these individuals were healthy, although one sibling had elevated
urinary lactate excretion. A fifth adult with partial deficiency, found among
clinically normal adult volunteers, later showed minor symptoms that resolved
after biotin therapy. Like children with profound biotinidase deficiency, children
with partial biotinidase deficiency are symptom free at birth. However, the sub-
sequent occurrence of symptoms of profound biotinidase deficiency in some
persons with partial deficiency suggests that biotin therapy for this condition
may be warranted. (J PEDIATR1990;116:78-83)

Deficient activity of the enzyme biotinidase (EC 3.5.1.12)
is the primary enzymatic defect in most children with
biotin-responsive, late-onset multiple carboxylase deficien-
cy. 1 Symptoms of this autosomal recessively inherited dis-
order include seizures, alopecia, skin rash, hypotonia,
ataxia, hearing loss, and developmental delay, often ac-
companied by lactic acidosis and organic aciduria. 2, 3 If the
disease is untreated, symptoms usually become progres-
sively worse; coma and death may occur. With the excep-

SNK Student-Newman-Keuls multiple t test

Supported by grants AM 33022 and HD 23233 from the National
Institutes Of Health.
Submitted for publication April 19, 1988; accepted July 5, 1989.
Reprint requests: Barry Wolf, MD, PhD, Department of Human
Genetics, Medical College of Virginia, PO Box 33, MCV Station,
Richmond, VA 23298-0033.
9/20/15111
LSD Least significant difference multiple t test
tion of irreversible neurologic damage, symptoms of bio-
tinidase deficiency resolve rapidly after treatment with
biotin.3,4 If the disorder is detected and treated early, the
clinical features of the disorder may be prevented. 4 After

?8
Volume 116
Number 1
the development of a simple method of determining biotini-
dase activity in dried whole blood filter-paper spots identi-
Cal to those used to screen newborn infants for other met-
abolic disorders, 5 several states and countries have begun
newborn screening for this disorder. 6, 7 Since the imple-
mentation of the first neonatal screening program for bio-
tinidase deficiency in 1984, a total of 14 children with pro-
found biotinidase deficiency have been identified in the
United States. In addition, 16 children with partial biotini-
dase deficiency have been detected. The purposes of this
study were to assess the clinical and biochemical features of
persons with partial biotinidase deficiency, to determine the
inheritance pattern of this condition, and to determine
whether those with partial enzyme activity are at risk of ac-
quiring symPtoms and would benefit from biotin therapy.
METHODS
Biotinidase activity in serum was measured as described
previously3 This method measures the quantity of p-ami-
nobenzoic acid released from the artificial substrate, bio-
tinyl-p-aminobenzoic acid. The coefficient of variation of
the assay is 27.1%. 8
Subjects in this study were divided into six groups for
comparison and statistical analysis. The first two groups
were composed of children identified by neonatal screening
for biotinidase deficiency who had tess than 30% of the mean
normal activity. An activity corresponding to 2 SD greater
than the log-transformed mean activity in clinically ascer-
tained children with symptoms was chosen as the boundary
between these two groups. This upper limit for the first
group and lower limit for the second group is equal to ap-
proximately 10% of the mean normal activity. Children
with 10% to 30% of the mean normal activity were desig-
nated as having "partial" biotinidase deficiency, and chil-
dren with less than 10% mean normal activity were said to
have "profound" biotinidase deficiency. The third group
consisted of children who were ascertained clinically. These
children had multiple symptoms of biotinidase deficiency
with or without organic aciduria. Deficient activity levels of
the biotin-dependent carboxylases in peripheral blood leu-
kocytes or low plasma biotin concentrations, or both, had
previously been found in some of these patients. The chil-
dren in this group were designated as "symptomatic." The
final three groups consisted of the parents of children in the
each of the first three groups, respectively. In addition, each
group was subdivided into male and female subjects. Bio-
tinidase activity levels were determined in siblings and var-
ious Other family members of children in the first three
groups; hOwever, these persons were not included in the
statistical analyses. Full adult biotinidase activity is reached
by 1 month of age9; all subjects in this study were 1 month
of age or older at the time serum samples were obtained.
Partial biotinidase deficiency 79
The Tukey studentized range, Student-Newman-Keuls,
Duncan multiple range, and the least significant difference
multiple t tests were performed on the biotinidase activity
data from the six groups and from the 12 subgroups.
Medical histories of patients, their siblings, and their
parents were obtained either from questionnaires completed
by the child's parents, physician, or both, or from patient
information compiled by ~he various screening centers.
RESULTS
The distributions of biotinidase activity levels in the six
groups described previously are shown in Fig. 1. The mean
normal activity level and range, computed from data on 521
normal adults, are also shown. The activity levels in the lat-
ter group approximate a normal distribution, with slight
positive skewness. Twelve boys and four girls identified by
newborn screening had Partial bi0tinidase deficiency. Three
additional children (2 girls) with partial deficiency were
found among the siblings of the children in this group.
Fourteen children (five girls) ascertained by neonatal
screening had profound biotinidase deficiency. One boy and
one girl with profound deficiency were found among the
older siblings of these children. In 24 clinically ascertained
children with symptoms (11 girls), biotinidase activity lev-
els were less than 10% of the mean normal level.
All t6 children with partial biotinidase deficiency iden-
tiffed by neonatal screening were symptom free at the time
of diagnosis. Nine of these children were treated with biotin,
whereas the parents of the remaining infants elected not to
treat their children. In one child with 30% of the mean nor-
mal biotinidase activity level, who was not initially treated
with biotin, hypotonia and occipital hair loss developed, and
a dark rash was manifested at 6 months of age. Conse-
quently, he received 10 mg of orally administered biotin per
day. The biotin therapy was suspended after 2 days because
of gastroenteritis (assumed to be viral) but was resumed 5
days later. After 10 more days of continuous biotin therapy,
the rash disappeared and muscle tone improved markedly.
Before treatment with biotin the infant had been unable to
sit without support, but after about 2 weeks of biotin ther-
apy he began to crawl and sit alone. After 4 weeks of ther-
apy he was able to pull himself to a standing position, and
after 2 months of treatment with biotin his hair was notice-
ably thicker. Biotin therapy has been continued, and at 1
year of age he is physically and developmentally normal
This child has a symptom-free 5-year-old brother with 27~
of the mean normal activity. Two other siblings of neona-
tally ascertained children with partial biotinidase defi-
ciency, both aged 2 years, also had activity in the partially
deficient range. Both were healthy, although one had
slightly elevated urinary lactate excretion as determined by
quantitative gas-liquid chromatography.
80 Secor McVoy et al. The Journal of Pediatrics
January 1990

10-

9_

6_
O
o
5_ o oo

"~ 4-

"uI 3 t
, 1[
.~ 2
m
1_ to
a ~o

0 I ~ I ] [ J
A B C D E F G
Group

Fig. 1. Distribution of biotinidase activity in biotinidase-deficient children, their parents, and normal subjects. Bar indi-
cates overall mean for each group. Group A: Children with partially deficient biotinidase activity, ascertained by newborn
screening (mean = 1.61 _+ 0.39 [nmol/min/ml serum _+ SD], n = 16). Group B: Children with neonatally ascertained
profound biotinidase deficiency (mean = 0.24 _+ 0.14, n = 14). Group C. Children with clinically ascertained profound bi-
otinidase deficiency (mean = 0.12 _+ 0.19, n = 24). Group D: Parents of children from group A, with partial deficiency
(overall mean = 4.10 _+ 1.40, n = 28; mean for men = 4.64 _+ 1.51, n = 13; mean for women = 3.64 _+ 1.16, n = 15).
Group E: Parents of children from group B, with profound deficiency (overall mean = 3.49 _+ 0.75, n = 24; mean for
men = 3.86 _+ 0.58, n = 12; mean for women = 3.12 _+ 0.73, n = 12). Group F: Parents of children from group C, who
were ascertained clinically (overall mean = 3.48 _+ 0.70, n = 40; mean for men = 3.73 _+ 0.71, n = 19; mean for
women = 3.25 _+ 0.62, n = 21). Group G: Normal range, computed from biotinidase activities of 521 adults (4.4 to 10,
mean = 7.1).

In the course of testing family m e m b e r s for carrier sta-
tus, we identified four healthy adults with biotinidase activ-
ity in the partial range. One is the m o t h e r of three children
with profound biotinidase deficiency, and the remaining
three are the mother, m a t e r n a l aunt, and m a t e r n a l grand-
mother of a child with partial deficiency. The pedigrees of
these family members are presented in Fig. 2 (pedigree A:
individual 11-9; pedigree B: individuals II-1, 1I-4, and 1-2,
respectively). Also depicted in Fig. 2 is the pedigree of a fifth
woman (pedigree C, individual I1-3) who has partial
biotinidase deficiency. This woman was identified during
the course of determining the normal range of biotinidase
activity in a large population. During pregnancy she expe-
rienced hair loss t h a t stopped after biotin treatment. N o n e
of the other adults with partial biotinidase deficiency has
had any of the symptoms t h a t are usually associated with
profound biotinidase deficiency. Pedigree D (Fig. 2) depicts
a family in which one p a r e n t has normal activity. The
proband, who has activity in the 10% to 30% range, appears
to have inherited a deficient allele from only one parent and
therefore is probably a heterozygote. In the r e m a i n d e r of
the families of children with clinically or neonatally ascer-
tained profound or partial biotinidase deficiency, both par-
ents had biotinidase activity levels t h a t were i n t e r m e d i a t e
between t h a t of the proband a n d those of n o r m a l individu-
als (pedigrees not shown).
T h e mean biotinidase activity level in the parents of chil-
dren with partial biotinidase deficiency (Fig. 1) was slightly
higher than t h a t of the other two groups of parents. This
difference was significant by the S N K , LSD, a n d D u n c a n
multiple range analyses ( a = 0.05), but was not significant
when the Tukey studentized range analysis was used. This
difference was no longer significant in the S N K analysis
when the tests were repeated with the biotinidase activity
data on individuals in pedigree D removed. T h e m e a n ac-
tivity for men was higher t h a n t h a t for women in all three
groups of parents as well as a m o n g the children with par-
tial deficiency. T h e difference between biotinidase activity
in fathers and t h a t in mothers of both neonatally ascer-
tained groups of children was significant (c~ = 0.05) in the
Volume 116 Partial biotinidase deficiency 81
Number I

II.

IlL

Biiil

lI.

IIL

Fig. 2. Pedigreesof subjects with biotinidase activity levelsin the partially deficientrange. Pedigree A: Family in which
both partial biotinidase deficiency (individual II-9) and profound biotinidase deficiency (individuals III-5,6 and 7) are
present. Individual I1-7 is g normal, healthy adult with partial deficiency.Pedigree B: Family with partial biotinidase de-
ficiencyin three generations. Individuals I-2, II-1, and II-3, are normal, healthy adults. Pedigree C." Family of woman with
partial biotinidasedeficiency(II-3). This individualexperiencedhair loss that ceased after treatment with biotin. Pedigree
D: Family in which only one parent of child with partial biotinidase deficiencyhas activity in heterozygous range, suggest-
ing that child (II-1) is a heterozygote with low activity.

LSD and Duncan multiple range tests but was not signif-
icant when analyzed by SNK or by the Tukey studentized
range tests. The mean biotinidase activity levels of the two
sexes for the remaining groups did not differ significantly.
The observation of higher biotinidase activity levels in male
subjects has been noted previously in normal adolescents
and adults. 1~ It The Michaelis constant (Kin) values of bi-
otinyl-p-aminobenzoicacid for biotinidase in the sera from
children with partial biotinidase deficiency and from their
parents were normal (data not shown).
DISCUSSION
In addition to identifying children with profound biotini-
dase deficiency, newborn screening for biotinidase defi-
ciency has revealed what may be a variant form of biotini-
dase. Clinically, partial biotinidase deficiency, like pro-
found biotinidase deficiency, appears to be inherited as an
autosomal recessive trait in the majority of cases. Biochem-
ically, the alleles that encode enzymes with normal, par-
tially deficient, and profoundly deficient activity may be
codominant. Children with partial deficiency may be ho-
mozygous for the variant allele, or they may be compound
heterozygotes, with one "variant" and one "deficient"
allele. Alternatively, as in pedigree D and in several fami-
lies ascertained in a neonatal screening program in
Quebec] 2 children with activity in the 10% to 30% range
may be heterozygotes with very low activity levels. Classi-
fication of children into one of these three genotypes simply
on the basis of biotinidase activity is not possible. Measure-
ment of biotinidase activity in parents, and occasionally in
the siblings, grandparents, or other relatives, may help to
determine the correct genotypic classification.
Accurate assignment of genotype to the individuals in a
family is essential for appropriate genetic counseling re-
garding recurrence risk and may also be important when
one is making decisions regarding treatment. However, be-
cause the ranges of biotinidase activity for heterozygotes
and for normal individuals are not discrete, and because the
ranges for "normal-partial" heterozygotes and "normal-
profound" heterozygotes almost totally overlap, it may be
difficult to determine the genotype of individual members of
a particular family even if activity levels for the parents and
grandparents are known. It is possible that several different
biotinidase alleles exist in the general population, each of
which codes for an enzyme with slightly different activity.
In addition, biotinidase activity correlates with several fac-
82 Secor McVoy et al.
tors, including age, gender, and hepatic protein synthetic
status.ll, 13 The genotype assignments in the pedigrees de-
scribed here were based on the biotinidase activity in the
children and their parents or other relatives, and no attempt
was made to correct for gender, age or liver function. Be-
cause of these uncertainties, we have limited the number of
possible alleles to three, which code for enzymes with either
normal, deficient, or partial biotinidase activity. More ac-
curate assignments should be possible once the biochemical
and molecular bases of the variant enzymes are determined.
In the parents of children with partial deficiency, the
presence of at least one parent with normal biotinidase ac-
tivity is responsible in part for the difference between this
group and the two other groups. When the activity levels of
both parents from this family are removed from the anal-
yses, this difference in means is still significant by the LSD
and Duncan multiple range analyses, but it is no longer sig-
nificant by the S N K analysis. If partial biotinidase defi-
ciency is due either to homozygosity for partial deficiency
alleles or to compound heterozygosity, with one partial and
one profound deficiency allele, then some of the parents
should carry one partial deficiency allele and one normal
allele. The presence of an allele for partial deficiency in
some of these parents may contribute to this difference in
mean activity levels. On the basis of the low enzyme activ-
ity level observed in homozygous or compound heterozy-
gous children, and because of the large range of "normal"
activities, the difficulty in distinguishing profound defi-
ciency carriers from partial deficiency carriers is not
surprising. In a single heterozygous individual, the contri-
bution to activity made by the defective allele may be
masked by the variable activity that is possible for the nor-
mal allele. When parents of children with partial deficiency
are viewed as a group, however, the tendency toward
enzyme activitylevels that are slightly higher than those in
parents of children with profound deficiency becomes evi-
dent. The discrepancies among the four statistical analyses
may be resolved with a larger sample size.
Before the implementation of neonatal screening for bi-
otinidase deficiency, partial biotinidase deficiency was un-
known. When the first children with partial deficiency were
found, no information was available on which to base a
prognosis. The observation that none of the children who
were clinically ascertained had activity in the partially de-
ficient range suggested that children with this level of
activity may remain symptom free. Because of this infor-
mation, some parents of children with partial deficiency
have elected not to treat their children. We described a pa-
tient with partial biotinidase deficiency who was not initially
treated with biotin and in whom symptoms of biotinidase
deficiency later developed, and then resolved after biotin
therapy. With the exception of this child, the subjects with
The Journal of Pediatrics
January 1990
partial biotinidase deficiency in this study were symptom
free or had relatively mild and nonspecific symptoms; thus
the clinical diagnosis of biotinidase deficiency probably
would not have been made. The manifestation of symptoms
of biotinidase deficiency is not dependent simply on the de-
gree of biotinidase deficiency but may also depend on the
interaction of reduced biotinidase activity with other fac-
tors. One of these factors may be the availability of exog-
enous biotin; another factor may be alterations in the met-
abolic demand for the vitamin. We have observed that the
development of symptoms in children with profound bio-
tinidase deficiency is sometimes precipitated by an infection
or other unrelated illness, and it is possible that the body's
requirements for biotin may increase during times of stress.
The reason that partial biotinidase deficiency has not been
ascertained clinically in the past may be the infrequent oc-
currence of the combination of factors that lead to the de-
velopment of symptoms. Because potentially serious conse-
quences may arise, it is important to identify children with
partial biotinidase deficiency.
Children with profound biotinidase deficiency are effec-
tively treated with the water-soluble vitamin biotin.4 No ill
effects have been associated with the administration of
pharmacologic doses in normal persons; however, biotin
metabolites may accumulate to higher than normal con-
centrations in persons with biotinidase deficiency. It has not
been established whether some symptoms of biotinidase de-
ficiency, particularly those which are not alleviated by bi-
otin therapy, may be due to the buildup of biotin metabo-
lites, rather than to the deficiency of biotin in its free, bio-
available form. The most severe and life-threatening
symptoms of the disorder are effectively and rapidly re-
solved by biotin therapy, and selection of an appropriate
dosage of biotin can lower the chance that symptoms may
arise from the buildup of biotin metabolites.
The possibility that depletion of free biotin may occur
earlier in the brain than in other tissues of biotinidase-de-
ficient individuals has prompted the suggestion that chil-
dren with less than 15% activity be treated with biotin. 14 It
now appears that this recommendation may be too conser-
vative. The demonstration of clinical features of profound
biotinidase deficiency in a child with 30% activity suggests
that children with 10% to 30% activity may be at risk of ac-
quiring symptoms, particularly if stressed by an infection or
dietary indiscretion, and may benefit from prophylactic bi-
otin therapy. Lower doses of biotin than the 5 to 10 mg/day
currently used to treat profound biotinidase deficiency may
be sufficient for the treatment of children with partial de-
ficiency. Continued follow-up of treated and untreated
children with partial biotinidase deficiency, as well as the
identification and evaluation of additional older children
and adults with partial deficiency, will help determine the
Volume 116 Partial biotinidase deficiency 83
Number 1

value and need for biotin supplementation for persons with
partial biotinidase deficiency.
We thank Drs. R. J. Allen, M. R. Lohff, S. E. Snyderman, A.
Slonim, N. R. M. Buist, and K. L. Sasser for generously providing
serum from some of the patients reported in this study.
REFERENCES
1. Wolf B, Grief RE, Allen R J, Goodman SI, Kein CL. Biotini-
dase deficiency: the enzymatic defect in late-onset multiple
carboxylase deficiency. Clin Chim Acta 1983;131:273-81.
2. Wolf B, Heard GS, Weissbecker KA, Secor McVoy JR, Grier
RE, Leshner RT. Biotinidase deficiency: initial clinical fea-
tures and rapid diagnosis. Ann Neurol 1985;18:614-7.
3. Wolf G, Grier RE, Allen R J, et al. Phenotypic variation in bi-
otinidase deficiency. J PEDIATR 1983;103:233-7.
4. Wolf B, Grier RE, Secor McVoy JR, Heard GS. Biotinidase
deficiency: a novel vitamin recycling defect. J Inherited Metab
Dis 1985;8(suppl 1):53-8.
5. Heard GS, McVoy JS, Wolf B. A screening method for bio-
tinidase deficiency in newborns. Clin Chem 1984;30:125-7.
6. Wolf B, Heard GS, Jefferson L J, Proud VK, Nance WE,
Weissbecker KA. Clinical findings in four children with
biotinicase deficiency detected through a statewide neonatal
screening program. N Engl J Med 1985;313:16-9.
7. Wolf B, Heard GS. Worldwide experience in newborn screen-
ing for biotinidase deficiency. Pediatrics (in press).
8. Weissbecker KA, Gruemer HD, Heard GS, Miller WG,
Nance WE, Wolf B. An automated procedure for measuring
biotinidase activity in serum. Clin Chem 1989;35:831-3.
9. Sourmala T, Wick H, Baumgartner ER. Low biotinidase ac-
tivity in plasma of some pre-term infants: possible source of
false-positive screening resuits. Eur J Pediatr i988;147:478-
90.
10. Weissbeeker KA, Wolf B, Piussan C, Nance W. Detection of
heterozygotes for biotinidase deficiency [Abstract]. Am J
Hum Genet 1985;37:A81.
11. Weissbeeker K. A genetic and biochemical study of variation
of normal and abnormal biotinidase activity [Doctoral Thesis].
Richmond, Va.: Medical College of Virginia, 1987.
12. Dunkel G, Scriver CR, Clow CL, et al. Prospective ascertain-
ment of complete and partial serum, biotinidase deficiency in
the newborn. J Inherited Metab Dis 1989;12:131-8.
13. Grier RE. Studies of human biotinidase and biotinidase activ-
ity [Doctoral Thesis]. Richmond, Va.: Medical College of
Virginia, 1985.
14. Sourmala TS, Baumgartner ER, Wick H, et al. Biotinidase
deficiency with residual biotinidase activity. Society for the
Study of Inborn Errors of Metabolism, 26th Annual Sympo-
sium, Glasgow, Scotland, 1988:89.

BOUND VOLUMES AVAILABLE TO SUBSCRIBERS

Bound volumes of the 1990 issues of THE JOURNAL OF PEDIATRICS are available to subscribers
(only) from the Publisher, at a cost of $48.00 ($64.00 international) for Vol. 116 (January-June)
and Vol. 117 (July-December), shipping charges included. Each bound volume contains subject
and author indexes, and all advertising is removed. Copies are shipped within 60 days after
publication of the last issue in the volume. The binding is durable buckram, with the Journal
name, volume number, and year stamped in gold on the spine. Payment must accompany all
orders. Contact The C.V. Mosby Company, Circulation Department, 11830 Westline Industrial
Dr., St. Louis, M O 63146-3318, USA/800-325-4177, ext. 351.
Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular
Journal subscription.