Journal of the Neurological Sciences 375 (2017) 294–298

Contents lists available at ScienceDirect

Journal of the Neurological Sciences

journal homepage: www.elsevier.com/locate/jns

Electrographic patterns in patients with posterior reversible

encephalopathy syndrome and seizures
Carlos Kamiya-Matsuoka ⁎, Sudhakar Tummala
Department of Neuro-Oncology, The University of Texas MD Anderson Cancer (CK-M, ST), 1400 Holcombe Blvd, Room FC7.3000, Unit 431, Houston, TX 77030, United States

a r t i c l e i n f o a b s t r a c t

Article history: Introduction: Posterior reversible encephalopathy syndrome (PRES) is a neurotoxic encephalopathic state associ-
Received 17 December 2016 ated with reversible cerebral vasogenic edema. Seizures are a common clinical presentation in PRES, however its
Received in revised form 4 February 2017 electroencephalographic and radiologic pattern correlation is limited in this subset of patients. The aim of this
Accepted 7 February 2017 study is to analyze the origin of electrographic dysfunction according to the radiologic pattern in patients with
Available online 8 February 2017
PRES and seizures.
Methods: We retrospectively identified 46 cancer patients who developed PRES and seizures at The University of
Keywords:
Posterior reversible encephalopathy syndrome
Texas MD Anderson Cancer Center between January 2006 and June 2012. Clinical, radiographic and electroen-
Seizures cephalographic data were abstracted from their records and reviewed for our analysis.
Status epilepticus Results: The average age at presentation was 49.9 ± 19.7 years. Thirty-four (73.9%) patients were women. Twen-
Neurotoxicity ty-two (47.8%) patients had a primary hematological malignancy whereas the rest had a solid tumor. Thirty-
Electroencephalogram three (71.7%) patients had received some form of chemotherapy. The mean systolic blood pressure (SBP) varia-
Chemotherapy tion was 23.7 ± 16.4 mmHg at onset of symptoms. On brain MRI, 32 (69.6%) patients had typical pattern while 14
Cancer (30.4%) had an atypical pattern. Thirty-seven (80.4%) patients had scalp electroencephalogram (EEG) evaluation.
Tacrolimus
Thirty-three (89.2%) had abnormal EEG findings: diffuse theta/delta slowing (N = 12, 36.4%), followed by diffuse
Bone marrow transplant
slowing with focal dysfunction (N = 8, 24.2%), focal dysfunction with epileptiform discharges (N = 4, 12.1%),
non-convulsive status epilepticus (N = 4, 12.1%), focal seizure activity and burst suppression (N = 2, 6.1%
each). Lateralized Periodic Discharges (LPDs) were recorded in 1 case. Four patients had focal dysfunction local-
ized to areas without conventional MRI signal changes. Four patients had recurrent seizures, of which 3 had an
atypical PRES pattern.
Conclusion: PRES appears to be a diffuse neurotoxic encephalopathic state. Origin of seizures seen on scalp EEG
may not correlate with the location of vasogenic edema/MRI signal changes raising the possibility of greater de-
gree of dysfunction which may exist beyond those areas.
© 2017 Elsevier B.V. All rights reserved.

1. Introduction vasogenic edema seems to be associated with loss of integrity of the

Posterior reversible encephalopathy syndrome (PRES) is a neurotox-
ic encephalopathic state associated with reversible cerebral vasogenic
edema that was first described in 1996 [1]. The clinical presentation
ranges from headache to confusion or frank encephalopathy. Addition-
ally, patients commonly have visual disturbances, and seizures [2]. Brain
imaging usually demonstrates vasogenic edema preferentially in the bi-
lateral parietal-occipital lobes, which might be related to the lower con-
centrations of sympathetic innervation of the posterior intracranial
arteries in comparison with other cerebral regions resulting in lower
autoregulatory capacity in these vessels [3,4]. There is still controversy
of the pathophysiologic trigger; however the mechanism that produces
⁎ Corresponding author.
E-mail address: carloskammat@live.com (C. Kamiya-Matsuoka).
blood-brain barrier. Recognized risk factors that are commonly found
at the onset of the PRES syndrome include hypertension, preeclamp-
sia/eclampsia, autoimmune disease, renal disease, infections and multi-
ple drugs especially immunosuppressive agents. Moderate-to-severe
hypertension is seen in approximately 75% of patients with PRES [5].
Seizures are commonly encountered in these patients and are gener-
ally responsible for severe morbidity and mortality. Although there are
several single case reports or small case series documenting the associ-
ation between PRES and seizures, there is still limited data eluding
towards the exact pathophysiology of seizures, seizure semiology,
electroencephalographic patterns and its correlation with radiologic
findings. Furthermore, there has not been any study focused on the
correlation between PRES and seizures in the oncologic population,
although this population has a high incidence of PRES. The aim of this
study is to analyze electrographic findings on routine scalp EEG in
patients with PRES and seizures.

http://dx.doi.org/10.1016/j.jns.2017.02.017
0022-510X/© 2017 Elsevier B.V. All rights reserved.
 C. Kamiya-Matsuoka, S. Tummala / Journal of the Neurological Sciences 375 (2017) 294–298 295

2. Methods 3. Results

We reviewed the records of all cancer patients who developed PRES
manifested with seizures at The University of Texas MD Anderson
Cancer Center between January 2006 and June 2012. The study was ap-
proved by the Institutional Review Board, and consent for retrospective
review of patient charts was waived.
The diagnosis of PRES in all patients was confirmed by brain MRI
using standard MRI sequences, which included axial T2 and FLAIR se-
quences; T1 axial pre and post-contrast; T1 post contrast in the coronal
and sagittal planes; T2-Gradient-echo axial, DWI, and ADC map. The ra-
diologic patterns were classified into typical and atypical according to
the lesion distribution after Bartynski and Boardman 2007 [6]. The typ-
ical pattern includes the involvement of occipital, parietal, frontal and
temporal lobes as well as the cerebellum. The atypical cases include
the distinct involvement of basal ganglia, thalamus, corpus callosum
and the periventricular white matter in addition to partial expression
of the typical pattern including the lack of involvement in the parietal
or occipital lobes. We collected and analyzed clinical, radiographic and
electroencephalographic data from the institutional database. The diag-
nosis of seizures was established by a neuro-oncologist. Electroenceph-
alograms (EEG) were recorded digitally using standard 10–20 system.
EEGs (20-min routine or extended) were interpreted by one reviewer
(ST). Resolution of PRES was defined as radiologic resolution of
vasogenic edema. Death was confirmed by review of medical records,
death certificate, and/or the Social Security Index.
We identified 46 patients with diagnosis of PRES and seizures. All pa-
tients had diagnosis of cancer, with the average age at presentation of
49.9 ± 19.7 years. Thirty-four (73.9%) patients were women. Twenty-
two (47.8%) patients had a primary hematological malignancy whereas
the rest had a solid tumor. Thirty-three (71.7%) patients had received
some form of chemotherapy (the most common used at the onset of
PRES were cyclophosphamide (N = 6), vincristine (N = 5) and doxoru-
bicin (N = 4)). Cyclophosphamide was the most frequently received in
patients with atypical pattern (N = 5) otherwise chemotherapy agents
were equally distributed in both groups. Elevated blood pressure (sys-
tolic blood pressure or diastolic blood pressure more than 140 mmHg
or 90 mmHg, respectively) was observed in all the patients. The mean
systolic blood pressure (SBP) variation before and at onset of symptoms
was 23.7 ± 16.4 mmHg. On brain MRI, 32 (69.6%) patients had typical
pattern (Fig. 2C and E) while 14 (30.4%) had an atypical pattern (Fig.
1C and 2A). There were no specific cerebrospinal fluid changes in
these patients.
All patients had at least seizure episode either witnessed or detected
on EEG. Twenty-nine (63%) patients had seizures and altered mental
status (AMS). Thirty-seven (80.4%) patients had an EEG evaluation.
Thirty-three (89.2%) of those had abnormal EEG findings. Seizures
focus and seizures were identified on regular 20-min routine EEG
(N = 7) and extended EEG (N = 3). As expected, seizures were mostly
(50%) of occipital origin (Fig. 1B and 2B). Generalized seizures were

Fig. 1. Radiologic and electrographic features of typical PRES. (A) MRI-Brain shows symmetric bilateral posterior parietal and occipital area signal changes compatible with typical PRES. (B)
Referential montage EEG shows focal epileptogenicity (left occipito-parietal repetitive spikes -blue arrows-, and independent lower amplitude right occipito-parietal region -red arrows-)
that corresponds with areas of edema. (C) MRI-Brain in shows also symmetric signal abnormality involving predominantly posterior brain and thalamus. (D) Longitudinal bipolar EEG
shows bihemispheric cortical dysfunction. Lateralized Periodic Discharges (LPDs) are occipital predominant with higher amplitude on the left -blue arrows- (D).
296 C. Kamiya-Matsuoka, S. Tummala / Journal of the Neurological Sciences 375 (2017) 294–298

Fig. 2. Non-correspondence between radiologic and electrographic features of PRES. (A) MRI-Brain shows diffuse patchy subcortical vasogenic edema including the basal ganglia,
thalamus, cerebellum, pons and both frontal lobes. (B) Longitudinal bipolar EEG shows electrographic seizure with repetitive fast spikes localized in the right occipital region (red
arrows). (C) MRI-Brain shows symmetrical T2-weighted hyperintensity involving only the occipital lobes, however the longitudinal bipolar EEG (D) shows symmetric generalized
rhythmic discharges (non-convulsive status epilepticus). Similarly, generalized seizure activity (F) may also be seen beyond the areas of vasogenic edema (E).

recorded in four of the 10 patients. Interestingly generalized seizures
and non-convulsive status epilepticus (NCSE) were also seen in four pa-
tients; three of these patients had leukemia. Seizures were controlled in
six patients with a single antiepileptic agent. Of the four patients who
experienced refractory seizures, two patients required two agents, and
only one patient needed four agents to control seizures. Refractory sei-
zures were associated with thrombocytopenia (less than 50,000/mm3)
in all the patients and with intracranial hemorrhage in one patient. Con-
tinuous infusion of midazolam or frequent doses of benzodiazepines
were required for four patients.
Most common EEG findings in both atypical and typical PRES pat-
terns are detailed in the Table 1. Lateralized Periodic Discharges
(LPDs) were recorded in only 1 case (Fig. 1D). Four patients had dys-
function beyond the areas of vasogenic edema (Fig. 2D and F). Focal
seizure activity was seen only in atypical PRES pattern (Fig. 2B) whereas
NCSE in both patterns. Otherwise, there was no noticeable difference
between the typical and atypical radiologic PRES patterns.
Complications of PRES such as recurrent seizures and death were ob-
served in 9 patients. Four patients developed recurrent seizures, of
those 3 (75%) had atypical PRES pattern. The five patients who died dur-
ing their hospitalization had either electrographic seizures or NCSE.
Four of them had the atypical PRES pattern.
4. Discussion
PRES is being increasingly recognized in cancer patients [7–12]. This
recognition is due to an earlier detection from increased MRI imaging
 C. Kamiya-Matsuoka, S. Tummala / Journal of the Neurological Sciences 375 (2017) 294–298
Table 1
Most common electrographic and radiographic patterns of 33 cancer patients with poste-
rior reversible encephalopathy syndrome who presented with seizures at The University
of Texas MD Anderson Cancer Center between January 2006 and June 2012.
PRES MRI pattern
Total N = 33 Atypical Typical
EEG pattern (100%) (N = 21) (N = 12)
Diffuse theta/delta slowing 12 (36.4) 6 6
Diffuse slowing with focal 8 (24.2) 6 2
dysfunction
Focal dysfunction with 4 (12.1) 3 1
epileptiform discharges
Non-convulsive status epilepticus 4 (12.1) 2 2
Focal seizure activity 2 (6.1) 2 –
Burst suppression 2 (6.1) 2 –
Lateralized Periodic Discharges 1 (3.0) – 1
Abbreviations: EEG, electroencephalogram; MRI, magnetic resonance imaging; N, number
of cases, PRES, posterior reversible encephalopathy syndrome.
use, although newer and more intensive treatments may also be con-
tributing to the rise.
The pathophysiology of PRES is still controversial and there are two
main hypotheses. One hypothesis involves impaired cerebral autoregu-
lation responsible for an increase in cerebral blood flow (hyperperfu-
sion) that may explain the changes that occur in hypertension. A
second hypothesis involves endothelial dysfunction that is thought to
represent those cases with normal or minimally elevated blood pres-
sure [13]. Endothelial dysfunction is believed to be caused by various cy-
totoxic agents and in autoimmune diseases by circulating cytokines.
Global dysfunction seen on scalp EEG and seizures originating from
areas without apparent neuroimaging correlation could favor cyto-
kine-related endothelial dysfunction.
Seizures are well recognized manifestation of PRES, however the
precise mechanism of ictogenesis is not known and suggests that PRES
is not just a subcortical pathology. Cortical irritation resulting from the
adjacent vasogenic edema that follows disruption of the blood brain
barrier is the most probable implicating factor. Seizures due to PRES
are usually a clinically isolated event and may be generalized or partial
[14]. EEG findings are not well described, however, common patterns
were reported and include generalized and posterior focal slowing
with epileptiform discharges [15] as found in our series. Other reports
describe status epilepticus with electrographic seizures only in the
parieto-occipital region [16].
The presence of LPDs suggests cortical neural dysfunction. Its associ-
ation with PRES is uncommon [17,18], but when it occurs, it may be clin-
ical significant provoking epilepsy [14]. Interestingly, we observed a
patient with bilateral independent LPDs (posterior dominant) correlat-
ing with the typical PRES pattern, but with asymmetric amplitude (mis-
match from the degree of MRI abnormalities), and associated with poor
outcome (Fig. 1D). It has to be acknowledged that standard EEG record-
ing compared to continuous EEG may not reveal the clear evaluation or
evolution of the described EEG findings over time.
PRES also manifests as status epilepticus in children and adults. Be-
sides the known factors; genetic and environmental factors may predis-
pose to the development of PRES in healthy children [19]. EEG pattern of
NCSE has been reported as occipital seizures corresponding with
vasogenic edema [16]. Another author reported partial status epilepti-
cus in both temporo-parietal-occipital regions. The seizure activity orig-
inated almost simultaneously from two independent epileptic foci
located in the occipital area of each hemisphere that were likely related
to the parenchymal abnormalities of PRES [20]. Apparently, there is
neuronal damage due to PRES and/or seizures as there may be residual
epilepsy despite neuroradiologic normalization [21,22]. Four of our pa-
tients persisted with seizures after resolution of PRES on brain MRI. The
radiologic PRES pattern of these patients was mainly atypical.
Altered sensorium is a common clinically isolated presentation in
patients with PRES. There is a higher incidence of patients with altered
297
sensorium who were found to have seizures in our study population.
Urgent EEG evaluation is recommended as initial diagnostic test in
these patients.
As previously mentioned, the location of radiologic lesions correlat-
ed well with the region of focal dysfunction on the routine scalp EEG.
We observed exceptions in 4 patients with atypical MRI pattern. It raises
suspicion of a greater degree of dysfunction which may exist beyond the
visible areas. In addition, NCSE does not necessarily present with more
severe or larger lesions on MRI, or correlate with seizure origin.
Although patients with NCSE in the setting of PRES carry a favorable
prognosis [16], we found that the presence of seizures or NCSE reflected
poor prognosis in 5 patients who died during hospitalization. According
to the literature, mortality rates and the incidence of residual deficits are
higher in cancer patients than in non-cancer patients [23]. Mortality
rates and residual deficits in cancer patients who develop PRES can be
higher than non-cancer patients owing to other comorbidities related
to cancer or to treatment, such as thrombocytopenia, renal failure, or
medication-induced hypertension.
5. Conclusion
PRES appears to be a diffuse neurotoxic encephalopathic state. Ori-
gin of seizures seen on scalp EEG may not correlate with the location
of vasogenic edema raising the possibility of greater degree of dysfunc-
tion which may exist beyond those areas. EEG evaluation is mandatory
in any patient with changes in mental status.
Disclosure
The authors report no conflicts of interest.
This study did not use any grant funds.
References
[1] J. Hinchey, C. Chaves, B. Appignani, et al., A reversible posterior
leukoencephalopathy syndrome, N. Engl. J. Med. 334 (8) (1996) 494–500.
[2] S.O. Casey, R.C. Sampaio, E. Michel, C.L. Truwit, Posterior reversible encephalopathy
syndrome: utility of fluid-attenuated inversion recovery MR imaging in the detec-
tion of cortical and subcortical lesions, AJNR Am. J. Neuroradiol. 21 (7) (2000)
1199–1206.
[3] W.S. Bartynski, J.F. Boardman, Distinct imaging patterns and lesion distribution in
posterior reversible encephalopathy syndrome, AJNR Am. J. Neuroradiol. 28
(2007) 1320–1327.
[4] W.S. Bartynski, B.C. Grabb, Z. Zeigler, et al., Watershed imaging features and clinical
vascular injury in cyclosporin A neurotoxicity, J. Comput. Assist. Tomogr. 21 (1997)
872–880.
[5] E.J. Kwon, S.W. Kim, K.K. Kim, H.S. Seo, Y. Kim Do, A case of gemcitabine and cisplat-
in associated posterior reversible encephalopathy syndrome, Cancer Res. Treat. 41
(1) (2009) 53–55.
[6] W.S. Bartynski, Z. Zeigler, M.P. Spearman, et al., Etiology of cortical and white matter
lesions in cyclosporin-a and FK-506 neurotoxicity, AJNR Am. J. Neuroradiol. 22
(2001) 1901–1914.
[7] R.C.S. Seet, A.A. Rabinstein, Clinical features and outcomes of posterior reversible en-
cephalopathy syndrome following bevacizumab treatment, Q. J. Med. 105 (2012)
69–75.
[8] Y.H. Chen, C.H. Huang, Reversible posterior leukoencephalopathy syndrome in-
duced by vinorelbine, Clin. Breast Cancer 12 (3) (2012) 222–225.
[9] U. Sharief, D.J. Perry, Delayed reversible posterior encephalopathy syndrome follow-
ing chemotherapy with oxaliplatin, Clin. Colorectal Cancer 8 (3) (2009) 163–165.
[10] C. Ozcan, S.J. Wong, P. Hari, Reversible posterior leukoencephalopathy syndrome
and bevacizumab, N. Engl. J. Med. 354 (9) (2006) 980–982.
[11] S. Kwon, J. Koo, S. Lee, Clinical spectrum of reversible posterior
leukoencephalopathy syndrome, Pediatr. Neurol. 24 (5) (2001) 361–364.
[12] J. Ocvirk, M. Boc, M. Rebersek, T. Ros, Cisplatin-induced non-convulsive posterior re-
versible encephalopathy syndrome in a 41-year-old woman with metastatic malig-
nant melanoma, Radiol. Oncol. 43 (2) (2009) 120–125.
[13] C. Kamiya-Matsuoka, D. Cachia, A. Olar, T.S. Armstrong, M.R. Gilbert, Primary brain
tumors and posterior reversible encephalopathy syndrome, Neurooncol. Pract. 1
(4) (2014) 184–190.
[14] V. Skiba, M. Etienne, J.A. Miller, Development of chronic epilepsy after recurrent ep-
isodes of posterior reversible encephalopathy syndrome associated with periodic
lateralized epileptiform discharges, Seizure 20 (1) (2011) 93–95.
[15] J. Natsume, A. Sofue, A. Yamada, K. Kato, Electroencephalographic (EEG) findings in
posterior reversible encephalopathy associated with immunosuppressants, J. Child
Neurol. 21 (7) (2006) 620–623.
298 C. Kamiya-Matsuoka, S. Tummala / Journal of the Neurological Sciences 375 (2017) 294–298
[16] O.S. Kozak, E.F. Wijdicks, E.M. Manno, J.T. Miley, A.A. Rabinstein, Status epilepticus as
initial manifestation of posterior reversible encephalopathy syndrome, Neurology
69 (9) (2007) 894–897.
[17] W. Fitzpatrick, N. Lowry, PLEDS: clinical correlates, Can. J. Neurol. Sci. 34 (4) (2007)
443–450.
[18] A. Bhatt, M.U. Farooq, S. Bhatt, A. Majid, M.Y. Kassab, Periodic lateralized epilepti-
form discharges: an initial electrographic pattern in reversible posterior
leukoencephalopathy syndrome, Neurol. Neurochir. Pol. 42 (1) (2008) 55–59.
[19] C. Mameli, D. Dilillo, D. Spiri, C. Cerini, S. Fasan, G.V. Zuccotti, Status epilepticus as
manifestation of posterior reversible encephalopathy syndrome in a healthy child,
Pediatr. Neurol. 48 (5) (2013) 418–420.
[20] R. Rossi, M.V. Saddi, A. Ticca, S.B. Murgia, Partial status epilepticus related to inde-
pendent occipital foci in posterior reversible encephalopathy syndrome (PRES),
Neurol. Sci. 29 (6) (2008) 455–458.
[21] M. Baldini, E. Bartolini, S. Gori, et al., Epilepsy after neuroimaging normalization in a
woman with tacrolimus-related posterior reversible encephalopathy syndrome, Ep-
ilepsy Behav. 17 (4) (2010) 558–560.
[22] Z. Sha, B.P. Moran, A.M. McKinney, T.R. Henry, Seizure outcomes of posterior revers-
ible encephalopathy syndrome and correlations with electroencephalographic
changes, Epilepsy Behav. 48 (2015) 70–74.
[23] J. Munoz, V. Kumar, J. Hamilton, et al., Posterior reversible encephalopathy syn-
drome: more than meets the eye, J. Clin. Oncol. 31 (20) (2013) 360–363.