Published Ahead of Print on September 19, 2018 as 10.1212/WNL.

0000000000006367
ARTICLE

Primary angiitis of the CNS and reversible

cerebral vasoconstriction syndrome
A comparative study
Hubert de Boysson, MD, MSc, Jean-Jacques Parienti, PhD, Jérôme Mawet, MD, Caroline Arquizan, MD, Correspondence
Grégoire Boulouis, MD, MSc, Cécilia Burcin, MD, Olivier Naggara, MD, PhD, Mathieu Zuber, MD, PhD, Dr. de Boysson
Emmanuel Touzé, MD, PhD, Achille Aouba, MD, PhD, Marie-Germaine Bousser, MD, PhD, [email protected]
Christian Pagnoux, MD, MSc, and Anne Ducros, MD, PhD

®
Neurology 2018;00:e1-e11. doi:10.1212/WNL.0000000000006367

Abstract
Objectives
To further improve the distinction between primary angiitis of the CNS (PACNS) and re-
versible cerebral vasoconstriction syndrome (RCVS).

Methods
We compared 2 large French cohorts of patients with PACNS (n = 110, retrospectively and
prospectively enrolled) and RCVS (n = 173, prospectively enrolled).

Results
Patients with RCVS were predominantly female (p < 0.0001), with migraines (p < 0.0001), and
were more often exposed to vasoactive substances (p < 0.0001) or postpartum (p = 0.002) than
patients with PACNS. Headache, especially thunderclap headache, was more frequent in RCVS
(both p < 0.0001). Thunderclap headache was absent in only 6% of patients with RCVS and was
mainly recurrent (87%) and provoked (77%) mostly by sexual intercourse, exertion, or emo-
tion. All other neurologic symptoms (motor deficit, seizure, cognitive disorder, or vigilance
impairment, all p < 0.0001) were more frequent in PACNS. At admission, brain CT or MRI was
abnormal in all patients with PACNS and in 31% of patients with RCVS (p < 0.0001). Acute
ischemic stroke was more frequent in PACNS than in RCVS (p < 0.0001). Although in-
tracerebral hemorrhage was more frequent in PACNS (p = 0.006), subarachnoid hemorrhage
and vasogenic edema predominated in RCVS (p = 0.04 and p = 0.01, respectively). Multiple
small deep infarcts, extensive deep white matter lesions, tumor-like lesions, or multiple
gadolinium-enhanced lesions were observed only in PACNS, whereas cervical artery dissection
was found only in RCVS.

Conclusions
Our study confirms that careful analysis of clinical context, headache features, and patterns of
brain lesions can distinguish PACNS and RCVS within the first few days of admission in most
cases. However, diagnosis remains challenging in a few cases.

From the Department of Internal Medicine (H.d.B., A.A.), Caen University Hospital; University of Caen-Normandie (H.d.B., E.T., A.A); Biostatistics and Clinical Research Unit (J.-J.P.),
Caen University Hospital; Emergency Headache Centre (J.M., C.B.), Lariboisière Hospital, Assistance Publique des Hôpitaux de Paris; Department of Neurology (C.A., A.D.), Montpellier
University Hospital; Department of Neuroradiology (G.B., O.N.), Sainte-Anne Hospital, Paris-Descartes University; INSERM UMR 894 (O.N.); Department of Vascular Neurology (M.Z.),
Saint Joseph Hospital; Université Paris-Descartes (M.Z.), INSERM UMR S 919, Paris; Department of Neurology (E.T.), Caen University Hospital; University of Caen-Normandie (E.T.),
INSERM U919; Department of Neurology (M.-G.B.), Lariboisière Hospital, Assistance Publique des Hôpitaux de Paris, France; and Division of Rheumatology (C.P.), Vasculitis Clinic,
Mount Sinai Hospital, Toronto, Ontario, Canada.

Go to Neurology.org/N for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article.

Copyright © 2018 American Academy of Neurology e1

Copyright ª 2018 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
 Glossary
CTA = CT angiography; CYC = cyclophosphamide; DSA = digital subtraction angiography; GC = glucocorticoid; MRA = MR
angiography; mRS = modified Rankin scale; NPV = negative predictive value; PACNS = primary angiitis of the CNS; PPV =
positive predictive value; RCVS = reversible cerebral vasoconstriction syndrome; SSRI = selective serotonine recapture
inhibitor; TCH = thunderclap headache.

Distinguishing primary angiitis of the CNS (PACNS) and re- suggestive of PACNS, with a combination of headache, focal
versible cerebral vasoconstriction syndrome (RCVS) can be deficits, seizures, or encephalopathy; (2) abnormalities of ce-
a challenge in clinical practice,1–4 although much progress has rebral arteries evidenced on biopsy (n = 32), digital subtraction
been made over the past decade in the early identification of the angiography (DSA; n = 68), or MR angiography (MRA; n =
2 diseases. Both conditions have polymorphic manifestations 10); (3) exclusion of other conditions and secondary CNS
with many overlapping features including headaches, focal neu- arteritis by a complete workup performed at diagnosis, includ-
rologic deficits, seizures, and strokes, but their natural history and ing infective serologies (tuberculosis, HIV, hepatitis B virus,
outcomes completely differ.5–10 In the absence of prompt di- and hepatitis C virus), autoimmunity serologies (antinuclear
agnosis and immunosuppressive therapy, progressive in- antibodies and antineutrophil cytoplasmic antibodies),
flammation in PACNS results in an important morbidity and echocardiography, and whole-body imaging (CT and/or
mortality.5,10–12 When brain biopsy is unavailable or negative, fluorodeoxyglucose-PET); and (4) a follow-up period of at
diagnosis of PACNS relies on the combination of clinical fea- least 6 months after diagnosis (unless the patient died before
tures, brain imaging, and cerebral angiography, together with the these 6 months from biopsy-proven PACNS) to exclude
exclusion of other more common diseases affecting CNS vessels, other differential diagnoses, such as RCVS, in which arterial
such as systemic vasculitides or other vasculopathies, including abnormalities are expected to reverse within the first months,
mainly RCVS.4,13,14 In the past 10 years, this noninflammatory or atherosclerosis in which clinical and radiologic manifes-
nonatheromatous self-limited vasculopathy has been recognized tations are not expected to improve under treatment. In all
as the main differential diagnosis of PACNS. In RCVS, removal patients, repeat MRI and MRA were performed 4–6 months
of precipitating vasoactive substances, rest, and empiric therapy after induction treatment and at least once a year thereafter.
with calcium channel blockers are associated in most patients
with an excellent clinical outcome and a complete normalization Before inclusion in the COVAC register, each case was reviewed
of brain arteries within 3 months.6–9,15,16 However, steroids may by a multidisciplinary panel of experts, including neurologists,
be deleterious in RCVS, further underscoring the need for internal medicine physicians, and neuroradiologists.
proper early distinction between RCVS and PACNS.17
Patients with RCVS
Few data exist comparing these 2 diseases.1–3 In 2016, Singhal We prospectively included 173 patients with RCVS from the
et al.3 compared 47 patients with PACNS and 159 patients Lariboisière Hospital study in Paris, a monocentric cohort
with RCVS and showed that recurrent thunderclap headache initiated in 2004, which recruited all consecutive patients
(TCH) and single TCH combined with either normal brain meeting the following 3 criteria: (1) unusual, acute, and severe
imaging, border-zone cerebral infarcts, or vasogenic edema headaches with or without focal neurologic deficit and/or
have a 100% positive predictive value (PPV) for the diagnosis seizures; (2) cerebral vasoconstriction demonstrated on DSA,
of RCVS. With the exception of recurrent TCH, all other MRA, or CT angiography (CTA), with involvement of at least
variables had a lower sensitivity for PACNS diagnosis. To 2 different cerebral arteries; and (3) disappearance of arterial
challenge or confirm these findings, we conducted a compar- abnormalities within the first 3 months. From 2004 to 2011,
ative study of 2 large French cohorts of PACNS (n = 110) and 121 patients were recruited from the emergency headache
RCVS (n = 173). clinic and 52 from the stroke unit. Previous publications about
this cohort have described the first 89 patients6,21 and the 20
patients with RCVS associated with a cervical artery dissec-
tion.22 RCVS was qualified as secondary in the presence of
Methods a known potential precipitating factor.6,23 For the purpose of
Patients with PACNS the present study, diagnosis of RCVS was retrospectively
We included 110 patients with PACNS from the ongoing reassessed in each of the 173 patients by the same neurologist
“COVAC” cohort (Cohort of Primary Cerebral Vasculitis [in (A.D.). Reversibility of cerebral vasoconstriction was proven
French: “cohorte de vascularite cérébrale primitive”]), a multi- in 170 patients by at least 1 normalized investigation (DSA in
centric French registry initiated in 2010. Most patients have 17 patients, MRA or CTA in 152 patients, and transcranial
been reported in previous publications.5,18–20 Data were ret- Doppler in 87 patients). The remaining 3 patients had purely
rospectively retrieved for patients diagnosed with PACNS be- cephalalgic RCVS with recurrent TCHs, normal brain MRI
fore 2010 and prospectively collected after 2010. All patients and angiographic narrowing and a self-limited clinical course
satisfied the following 4 criteria: (1) clinical presentation in the absence of glucocorticoid (GC) or immunosuppressant

e2 Neurology | Volume , Number  | Month 0, 2018 Neurology.org/N

Copyright ª 2018 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
treatment, and no recurrence during the long-term follow-up compared using Wilcoxon matched-pairs signed-rank test for
of 68, 115, and 152 months, respectively. quantitative variables.

Standard protocol approvals, registrations, With the aim of a validation study, we calculated in our
and patient consents cohorts the diagnostic sensitivity, specificity, PPV, and nega-
All patients were informed of their inclusion in the respective tive predictive value (NPV) of the variables that were found
COVAC and RCVS Lariboisière registers with the possibility significantly different between RCVS and PACNS in the study
to decline the inclusion, resulting in a deletion of their data. by Singhal et al.3
The COVAC cohort was approved by the Paris-Cochin in-
stitutional review board (CPPRB) (No. 12541). All patients Statistical analyses were computed using JMP 9.0.1 (SAS In-
from the RCVS cohort signed written informed consent. stitute Inc., Cary, NC). p < 0.05 defined statistical significance.
Study parameters and definitions Data availability
A data set was created for the purpose of this study, including all Anonymized data not published within this article will be
common information available in both databases. We thus made available by request from any qualified investigator.
extracted from the PACNS and RCVS databases the following
findings in each patient: demographics, medical history (in-
cluding the history of hypertension, tobacco use, diabetes mel- Results
litus, dyslipidemia, migraine, and depression), delivery within the
month before clinical onset, use of vasoactive drugs or medi- Demographics, medical history, and
cations within the 15 days before clinical onset, clinical mani- precipitating factors
festations at onset, laboratory tests (including CSF analysis), Compared with patients with PACNS, patients with RCVS
brain CT and/or MRI and neurovascular imaging, results of CNS were more likely to be women (71% vs 47% in patients with
biopsy when performed, administered treatment, and outcomes. PACNS, p < 0.0001) with a history of migraine (32% vs 7%,
p < 0.0001) (table 1). At least one precipitating factor was
CSF analysis was considered normal when showing a white identified in 121 patients with RCVS (70%) including early
blood cell count of <5/mm3 and/or a protein concentration postpartum in 14 women, and current exposure to 1 or several
of <0.45 g/L. vasoactive substances in 85 patients, the 3 commonest being
cannabis (in 35), selective serotonine recapture inhibitors
A TCH was defined as a headache reaching a maximum in- (SSRIs in 23), and nasal decongestants (in 16), followed by
tensity above 7/10, on an 11-point scale from 0 (no pain) to binge alcohol drinking (8), steroids (7), ergots (6), triptans
10 (worst possible pain), in less than 1 minute. (5), cocaine (5), nicotine patches (4), noradrenergic and
selective serotonergic antidepressants (3), epinephrine (3),
Functional status during follow-up was evaluated with the interferon alpha (2), cyclosporine (1), and sulprostone (1).
modified Rankin scale (mRS score, calculated on the basis of
medical charts when the data were unavailable: 0 = no neurologic Conversely, none of the patients with PACNS was in the
deficits; 1 = no important disability, despite symptoms; 2 = slight postpartum period. Twelve patients with PACNS had a long
disability; 3 = moderate disability; 4 = moderately severe dis- history of vasoactive substance use (cannabis in 3, SSRI in 8,
ability; 5 = severe disability; and 6 = death). The mRS score was and cannabis and SSRI in 1). Among these 12 patients,
recorded at different time points, especially at months 3, 6, and PACNS diagnosis was biopsy proven in 2. The 10 other
12 and then every year until final follow-up. patients presented with headaches (not thunderclap) and
motor deficits associated with vigilance impairment in 3 and
Relapse of PACNS was defined as a new clinical CNS mani- seizures in 3. MRI showed fluid-attenuated inversion recovery
festation (e.g., focal deficit) associated with at least one new white matter hyperintensities and multiple acute infarctions in
notable radiologic abnormality (cerebral infarct, extension of all, gadolinium enhancements in 5, intracerebral hemorrhage
white matter lesion, appearance of gadolinium enhancements, in 1, and subarachnoid hemorrhage in 2. All had elevated CSF
and worsening of arterial stenosis), leading to intensification protein levels >1.1 g/L and narrowings of at least 2 different
of treatment by the treating physician. Relapse of RCVS was arterial territories on DSA. Control angiography at >6 months
defined as a new severe headache with new vasoconstriction after onset showed persistent abnormalities in all of them.
on cerebral angiography. Despite treatment with GCs in all, combined with cyclo-
phosphamide (CYC) in 8, they had heavy sequelae with an
Statistics mRS at 3 (2–5) and 2 (1–5) at months 6 and 12, respectively.
Categorical variables are expressed as numbers (%) and
quantitative variables as medians (range). Categorical varia- Clinical manifestations
bles were analyzed using the χ 2 or Fisher exact test and Table 1 shows the different clinical manifestations at diagnosis
quantitative variables using the Wilcoxon rank-sum test. The between the 2 groups. Higher systolic and diastolic pressures were
repetitive mRS score over time among the same patients was observed in patients with RCVS at admission (both p < 0.0001).

Neurology.org/N Neurology | Volume , Number  | Month 0, 2018 e3

Copyright ª 2018 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
 Headaches were more frequent in RCVS (99% vs 54% in
Table 1 Anamnestic and clinical characteristics at patients with PACNS, p < 0.0001), especially TCHs (81% vs
diagnosis of patients with PACNS and RCVS 3% in patients with PACNS, p < 0.0001). All neurologic
PACNS RCVS symptoms (neurologic deficits, speech disorders, seizures,
cohort cohort p cognitive disorder, vigilance impairment, and cerebellar
(n = 110) (n = 173) Value
ataxia), except vision disorders, were significantly more fre-
Demographics quent in PACNS (all with a p value < 0.0001).
Age 46.5 (18–81) 44 (17–85) 0.08
Headache was present at onset in 171 patients with RCVS
Female/male 52/58 122/51 <0.0001
(99%), including 140 (81%) whose first headache was a TCH.
Medical history Within the first week, 163 patients with RCVS (94%) expe-
rienced at least one TCH (median number 4, range 1–18) and
Hypertension 27 (25) 26 (15) 0.05
151 patients (87%) had recurrent TCHs. At least 1 headache
Depression 15 (14) 29 (17) 0.48 had an immediate trigger in 77% of the patients with RCVS,
Migraine 8 (7) 56 (32) <0.0001 mostly sexual intercourse (in 46), exertion (in 38), and acute
emotional stress (in 32). The 2 patients with RCVS without
Tobacco use 40 (36) 62 (36) 0.93
headache at onset presented initially with seizures and focal
Precipitating factor deficits and developed headaches within the following 48
Postpartum 0 14 (8) 0.002
hours. RCVS started after a surgical procedure with traumatic
tracheal intubation in 1 and in the setting of cannabis con-
Any vasoactive drug or 12 (11) 85 (49) <0.0001 sumption in the other. Both had strokes and diffuse arterial
medication
narrowing on DSA, combined with a self-limited clinical
Drug (cocaine, cannabis) 4 (4) 35 (20) <0.0001 course and a complete normalization of arteries on control
SSRI 9 (8) 23 (13) 0.19 angiography at 3 months in the absence of GC and immu-
nosuppressant treatment. They both recovered completely.
Nasal decongestants 0 16 (9) 0.0003

Clinical manifestations Fifty-nine patients with PACNS (54%) experienced headaches

Systolic pressure 130 (90–230) 147 (90–230) <0.0001
at onset. None of the patients with PACNS had recurrent
TCHs. Three patients had a single TCH at onset, revealing
Diastolic pressure 70 (50–120) 82 (50–160) <0.0001
a convexity subarachnoid hemorrhage in 2 and a parenchymal
Headache at onset 59 (54) 171 (99) <0.0001 hemorrhage in one. One had a biopsy-proven diagnosis. The
other 2 showed abnormal angiography with bilateral arterial
TCH at onset 3 (3) 140 (81) <0.0001
narrowing, neurologic deficits, multiple brain infarctions, and
Any TCH 3 (3) 163 (94) <0.0001 abnormal CSF, combined with a partial recovery under treat-
Single TCH 3 (3) 12 (7) <0.0001 ment with GC and CYC. They eventually relapsed with new
ischemic lesions on control MRI and persistence of arterial
Recurrent TCH 0 151 (87) <0.0001
irregularities at 8 and 15 months after diagnosis.
Any neurologic deficit 97 (88) 41 (24) <0.0001

Transitory deficit 27 (25) 25 (14) <0.0001

Brain imaging and CSF analysis
Characteristics of imaging are shown in table 2. All patients
Persistent deficit 72 (66) 19 (11) <0.0001 underwent a brain imaging by CT or MRI at admission, which
Motor deficit 82 (75) 10 (6) <0.0001 was abnormal in all patients with PACNS and in 53 patients
with RCVS (31%, p < 0.0001). Repeated brain imaging in the
Sensory disorder 39 (35) 17 (10) <0.0001
following days still showed normal images in 107 patients with
Speech disorder 56 (51) 11 (6) <0.0001 RCVS (62%) and demonstrated new lesions in 13 patients
Seizures 35 (32) 9 (5) <0.0001 (8%) who initially had a normal brain scan. Acute ischemic
stroke was more frequent in PACNS than in RCVS (76% vs
Cognitive disorder 52 (47) 8 (5) <0.0001
8%, respectively, p < 0.0001). Although intracerebral hem-
Vigilance impairment 30 (27) 1 (1) <0.0001 orrhage was more frequent in PACNS (20% in PACNS vs 9%
Cerebellar ataxia 33 (30) 2 (1) <0.0001
in RCVS, p = 0.006), subarachnoid hemorrhage (27% in
RCVS vs 16% in PACNS, p = 0.04) and reversible vasogenic
Vision disorder 20 (18) 23 (13) 0.26 edema (8% in RCVS vs 1% in PACNS, p = 0.01) predomi-
Abbreviations: PACNS = primary angiitis of the CNS; RCVS = reversible ce-
nated in RCVS. Some brain lesions were observed only in
rebral vasoconstriction syndrome; SSRI = selective serotonin reuptake in- patients with PACNS, namely multiple diffuse small deep
hibitor; TCH = thunderclap headache.
Values are displayed as absolute number (%) or median (range).
infarcts (in 36 patients), extensive deep white matter lesions
(in 66 patients), tumor-like lesions (in 11 patients), or

e4 Neurology | Volume , Number  | Month 0, 2018 Neurology.org/N

Copyright ª 2018 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
 multiple gadolinium-enhanced lesions (in 44 of 91 patients
Table 2 Paraclinical characteristics at diagnosis of who underwent gadolinium injection).
patients with PACNS and RCVS
PACNS RCVS Cervical artery dissection was found in 12% of the patients
cohort cohort with RCVS but not in patients with PACNS.
(n = 110) (n = 173) p Value

C-reactive protein, 6 (0–95) 4 (4–58) 0.98 Regarding CSF analysis, higher white blood cell count and
mg/L
protein levels were observed in patients with PACNS (p <
Abnormal CSF analysis 71/102 (70) 72/140 (51) 0.005 0.0001 for both). A strictly normal CSF analysis (white blood
White blood cell 9 (0–425) 1 (0–35) <0.0001
cell count <5/mm3 and protein level <0.45 g/L) was more
count common in patients with RCVS than in patients with PACNS
(66% vs 32%, p < 0.0001). Moreover, striking CSF abnor-
Protein level, g/L 0.70 (0.1–4.1) 0.40 (0.18–1.01) <0.0001
malities with a white blood cell count of >35 mm3 and
Normal initial brain 0 120 (69) <0.0001 a protein level of >1.1 g/L were found in 37 patients with
imaginga
PACNS (36%) but in none of the patients with RCVS. No
Any abnormal brain 110 (100) 66 (38) <0.000.1 brain biopsy was performed in patients with RCVS.
imagingb

Acute ischemia on CNS 84 (76) 13 (8) <0.0001 Comparison of subgroups of PACNS and RCVS
imaging Table 3 shows the comparison of the 78 patients with
Single CNS infarction 20/84 (24) 6/13 (46) 0.09 angiography-diagnosed PACNS with the 173 patients with
RCVS. Significantly more neurologic symptoms (neurologic
Multiple CNS 64/84 (76) 7/13 (54) 0.09
infarctions
deficits, speech disorder, seizures, cognitive disorder, vigilance
impairment, and cerebellar ataxia), higher inflammation levels in
Border-zone 7/84 (8) 7/13 (54) <0.0001
infarction
CSF, and more ischemic and hemorrhagic parenchymal strokes
were observed in patients with angiography-diagnosed PACNS.
Deep infarction 81/84 (96) 1/13 (8) <0.0001 Conversely, more migraine, headaches, especially TCHs, and
Vertebrobasilar 21/84 (25) 2/13 (15) 0.73 subarachnoid hemorrhage were observed in patients with RCVS.
infarction

Microinfarctions 46/84 (55) 5/13 (38) 0.37 Table 4 compares the 32 patients with biopsy-proven PACNS
with the 173 patients with RCVS. The same differences pre-
Microbleeds 5/101 (5) 0/167 (0) 0.007
viously described were observed between the 2 groups, except
Parenchymal 22 (20) 15 (9) 0.006 rates of hypertension, transient neurologic deficits and sub-
hemorrhage
arachnoid hemorrhage on CNS imaging were no longer sig-
Lobar 12 (11) 10 (6) 0.12 nificantly different in the 2 groups.
Deep 14 (13) 6 (3) 0.003
Tables 5 and 6 compared the 110 patients with PACNS with
Subarachnoid 18 (16) 47 (27) 0.04 the 53 patients with RCVS with abnormal brain imaging at
hemorrhage
diagnosis. The same differences were observed, regarding
Ventricular hemorrhage 2 (2) 3 (2) 0.96 precipitating factors, clinical manifestations, and imaging
Subdural hematoma 2 (2) 2 (1) 0.65 abnormalities.
Posterior reversible 1 (1) 13 (8) 0.01 Treatment and outcomes
encephalopathy
syndrome Two patients with RCVS received GC: one presented with
TCHs, motor deficit, multiple brain infarctions, and abnormal
Abnormal angiography 84/109 (77) 173 (100) <0.0001
CSF analysis and received 3 months of GC; the other presented
Abnormal CTA/MRA 67/101 (66) 142/160 (89) <0.0001 with TCHs without additional neurologic symptoms but ab-
Abnormal DSA 74/94 (79) 77/82 (94) 0.004
normal CSF analysis and received a 1-month course of GC. Both
had a complete reversal of arterial narrowing on control angi-
Aneurysms on DSA or 10/109 (9) 14/173 (8) 0.75
MRA
ography. No patient with RCVS received immunosuppressant
therapy. Conversely, 107 patients with PACNS (97%) received
Cervical artery 0/109 (0) 20/173 (12) 0.0009 GC, combined with immunosuppressant in 91 (85%). The 3
dissection
patients who did not receive GC were administered IV CYC.
Abbreviations: CTA = CT angiography; DSA = digital subtraction angiography;
MRA = MR angiography; PACNS = primary angiitis of the CNS; RCVS = re-
versible cerebral vasoconstriction syndrome.
Calcium channel blockers were administered to 165 patients
Values are displayed as absolute number (%) or median (range).
a
with RCVS (95%) for a median of 40 (1–100) days. Most
CT or MRI at admission.
b
On repeat CT/MRI done within the first 2 weeks of admission. patients with RCVS (11 of 13) who initially presented with
a normal brain scan and showed further deterioration in the few

Neurology.org/N Neurology | Volume , Number  | Month 0, 2018 e5

Copyright ª 2018 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
 Table 3 Characteristics of patients with PACNS diagnosed Table 3 Characteristics of patients with PACNS diagnosed
on neurovascular imaging and RCVS on neurovascular imaging and RCVS (continued)
Patients Patients
with with
DSA/MRA- Patients DSA/MRA- Patients
diagnosed with diagnosed with
PACNS RCVS PACNS RCVS
(n = 78) (n = 173) p Value (n = 78) (n = 173) p Value

Demographics Abnormal CSF analysis 47/74 (64) 72/140 (51) 0.09

Age 49 (19–81) 44 (17–85) 0.008 White blood cell 5 (0–425) 1 (0–35) <0.0001
count
Female/male 37/41 122/51 0.0004
Protein level, g/L 0.6 (0.20–4.1) 0.40 (0.18–1.01) <0.0001
Medical history
Normal initial brain 0 120 (69) <0.0001
Hypertension 23 (29) 26 (15) 0.008 imaginga

Depression 13 (17) 29 (17) 0.98 Any brain imaging 78 (100) 66 (38) <0.000.1
abnormalb
Migraine 6 (8) 56 (32) <0.0001
Acute ischemia on CNS 71 (91) 13 (8) <0.0001
Tobacco use 32 (41) 62 (36) 0.43 imaging

Precipitating factor Border-zone 4/71 (6) 7/13 (54) <0.0001

infarction
Postpartum period 0 14 (8) 0.01
Multiple CNS 55/71 (77) 7/13 (54) 0.08
Any vasoactive drug 10 (13) 85 (49) <0.0001 infarctions
or medication
Parenchymal 14 (18) 15 (9) 0.03
Drug (cocaine, 3 (4) 35 (20) 0.0005 hemorrhage
cannabis)
Subarachnoid 11 (14) 47 (27) 0.02
SSRI 8 (10) 23 (13) 0.54 hemorrhage

Nasal decongestants 0 16 (9) 0.004 Abnormal angiography 78 (100) 173 (100) —

Clinical manifestations Abnormal CTA/MRA 62/72 (86) 142/160 (89) 0.57

Systolic pressure 130 (96–200) 147 (90–230) <0.0001 Abnormal DSA 68/69 (99) 77/82 (94) 0.15

Diastolic pressure 70 (50–120) 82 (50–160) <0.0001

Abbreviations: CTA = CT angiography; DSA = digital subtraction angiography;
MRA = MR angiography; PACNS = primary angiitis of the CNS; RCVS = re-
Headaches at onset 47 (60) 171 (99) <0.0001 versible cerebral vasoconstriction syndrome; SSRI = selective serotonin
reuptake inhibitors; TCH = thunderclap headache.
TCHs at onset 2 (3) 140 (81) <0.0001 Values are displayed as absolute number (%) or median (range).
a
CT or MRI at admission.
b
Any TCH 2 (3) 163 (94) <0.0001 On repeat CT/MRI done within the first 2 weeks of admission.

Single TCH 2 (3) 12 (7) <0.0001

first days were discharged from other institutions with a mis-
Recurrent TCH 0 151 (87) <0.0001
diagnosis of benign headache and were only started on calcium
Any neurologic 71 (91) 41 (24) <0.0001 channel blockers when they presented to the emergency
deficit
headache center.
Transitory deficit 23 (29) 25 (14) 0.005

Persistent deficit 59 (76) 19 (11) <0.0001

At month 3, 11/105 surviving patients with PACNS (10%)
and 167/173 patients with RCVS (97%) had an mRS score of
Motor deficit 68 (87) 10 (6) <0.0001 ≤1. At months 6 and 12, the rates were 28% vs 97% and 31%
Sensory disorder 31 (40) 17 (10) <0.0001 vs 98%, respectively (p < 0.0001 at months 3, 6, and 12).
Speech disorder 38 (49) 11 (6) <0.0001
Reversibility of cerebral vasoconstriction was proven in 170 of
Seizures 14 (18) 9 (5) 0.001 the 173 patients with RCVS by at least one follow-up neu-
Cognitive disorder 28 (36) 8 (5) <0.0001 rovascular assessment. The 3 other patients with RCVS made
a complete clinical recovery without relapse. Among the 84
Vigilance impairment 19 (24) 1 (1) <0.0001
patients with PACNS who initially had abnormal MRA or
Cerebellar ataxia 20 (26) 2 (1) <0.0001 DSA, 72 underwent a follow-up neurovascular assessment
Vision disorder 14 (18) 23 (13) 0.34
that showed fewer lesions in 64 patients, worsened lesions in
5, and disappearance of narrowings in 3.

e6 Neurology | Volume , Number  | Month 0, 2018 Neurology.org/N

Copyright ª 2018 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
Table 4 Characteristics of biopsy-proven PACNS and Table 4 Characteristics of biopsy-proven PACNS and RCVS
RCVS (continued)

Patients Patients
with with
biopsy- biopsy-
proven Patients proven Patients
PACNS with RCVS PACNS with RCVS
(n = 32) (n = 173) p Value (n = 32) (n = 173) p Value

Demographics White blood cell count 22 (1–350) 1 (0–35) <0.0001

Age 42 (18–61) 44 (17–85) 0.48 Protein level, g/L 0.95 0.40 <0.0001
(0.1–2.6) (0.18–1.01)
Female/male 15/17 122/51 0.009
Normal initial brain 0 120 (69) <0.0001
Medical history imaginga

Hypertension 4 (13) 26 (15) 0.71 Any abnormal brain 32 (100) 66 (38) <0.0001
imagingb
Depression 2 (6) 29 (17) 0.13
Acute ischemia on first CNS 13 (41) 13 (8) <0.0001
Migraine 2 (6) 56 (32) 0.003 imaging

Tobacco use 8 (25) 62 (36) 0.23 Border-zone infarction 3/13 (23) 7/13 (54) 0.11

Precipitating factor Multiple CNS infarctions 9/13 (69) 7/13 (54) 0.42

Postpartum period 0 14 (8) 0.1 Parenchymal hemorrhage 8 (25) 15 (9) 0.007

Any vasoactive drug or 2 (6) 85 (49) <0.0001 Meningeal hemorrhage 7 (22) 47 (27) 0.53
medication
Abnormal CTA/MRA/DSA 6/31 (19) 173 (100) <0.0001
Drug (cocaine, cannabis) 1 (3) 35 (20) 0.02
Abnormal MRA 5/29 (17) 142/160 (89) <0.0001
SSRI 1 (3) 23 (13) 0.10
Abnormal DSA 6/25 (24) 77/82 (94) <0.0001
Nasal decongestants 0 16 (9) 0.08
Abbreviations: CTA = CT angiography; DSA = digital subtraction angiography;
Clinical manifestations
MRA = MR angiography; PACNS = primary angiitis of the CNS; RCVS = re-
versible cerebral vasoconstriction syndrome; SSRI = selective serotonin
Systolic pressure 120 (90–230) 147 (90–230) <0.0001 reuptake inhibitor; TCH = thunderclap headache.
Values are displayed as absolute number (%) or median (range).
Diastolic pressure 70 (50–100) 82 (50–160) <0.0001 a
CT or MRI at admission.
b
On repeat CT/MRI performed within the first 2 weeks of admission.
Headaches at onset 12 (38) 171 (99) <0.0001

TCHs at onset 1 (3) 140 (81) <0.0001

Any TCHs 1 (3) 163 (94) <0.0001 The median follow-up period was 55 (0–192) months in
PACNS (4 patients with biopsy-proven PACNS died within
Single TCHs 1 (100) 12 (7) 0.0006
the first 6 months) and 72 (2–154) months in RCVS. A
Recurrence of TCHs 0 151 (87) <0.0001 relapse occurred in 34 patients with PACNS (31%) at 16
Any neurologic deficit 26 (81) 41 (24) <0.0001 (5–145) months from diagnosis, whereas 1 patient with
RCVS had an angiographically proven relapse with isolated
Transitory deficit 4 (13) 25 (14) 0.77
headache. Altogether, 8 patients with PACNS died, whereas
Persistent deficit 13 (41) 19 (11) <0.0001 none of the patients with RCVS died.
Motor deficit 14 (44) 10 (6) <0.0001
Improving the distinction between RCVS
Sensory disorder 8 (25) 17 (10) 0.02
and PACNS
Speech disorder 18 (56) 11 (6) <0.0001 We calculated the diagnostic sensitivity, specificity, PPV, and
NPV of the variables that were found significantly different
Seizures 21 (66) 9 (5) <0.0001
between RCVS and PACNS in the comparative study by
Cognitive disorder 24 (75) 8 (5) <0.0001 Singhal et al.3 Results are shown in table 7. We found that
Vigilance impairment 11 (34) 1 (1) <0.0001 recurrent TCH, or the combination of single TCH with
normal brain imaging or cortical-only infarctions or vasogenic
Cerebellar ataxia 13 (41) 2 (1) <0.0001
edema, reached a PPV of 100% for the diagnosis of RCVS. In
Vision disorder 6 (19) 23 (13) 0.42 patients with abnormal angiogram and without TCH, the
Abnormal CSF analysis 24/28 (86) 72/140 (51) 0.0008
combination of deep cerebral infarctions and abnormal CSF
analysis had a PPV of 98% for the diagnosis of PACNS.

Neurology.org/N Neurology | Volume , Number  | Month 0, 2018 e7

Copyright ª 2018 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
 Discussion
Table 5 Anamnestic and clinical characteristics at
diagnosis of patients with PACNS and RCVS with Only 2 comparative studies between RCVS and PACNS have
abnormal imaging at initial presentation been published so far, the first comparing 13 patients with RCVS
PACNS RCVS with 8 patients with PACNS, and the second, 159 patients with
cohort cohort RCVS with 47 patients with PACNS.2,3 The present study,
(n = 110) (n = 53) p Value
comparing 173 patients with RCVS enrolled in a single center
Demographics with 110 patients with PACNS enrolled in a multicenter register,
Age 46.5 (18–81) 47 (24–70) 0.61
is thus the largest reported so far.

Female/male 52/58 47/6 <0.0001 In all our patients with PACNS, a complete workup excluded
Medical history other conditions that may affect CNS vessels. In patients with
angiography-proven PACNS, confidence in the vasculitis di-
Hypertension 27 (25) 10 (19) 0.42
agnosis was strengthened by the long follow-up (>6 months
Depression 15 (14) 11 (21) 0.25 in all and 5 years in median) without complete clinical and
Migraine 8 (7) 25 (47) <0.0001 radiologic reversibility. In addition, the initial presence of
white matter lesions on brain MRI in 96% of patients along
Tobacco use 40 (36) 16 (30) 0.44
with multiple deep ischemic lesions in >80% of patients with
Precipitating factor angiography-proven PACNS is more likely secondary to
Postpartum period 0 9 (17) <0.0001
vasculitis rather than RCVS or other vasculopathies. In ad-
dition, as each patient with a PACNS diagnosis enrolled in the
Any vasoactive 12 (11) 22 (42) 0.09 register has a longitudinal follow-up, we only kept in the
medication
register patients with a definite diagnosis of PACNS.
Drug (cocaine, cannabis) 4 (4) 21 (40) <0.0001

SSRI 9 (8) 10 (19) 0.05 Our comparative study found several clinical and radiologic
features that help distinguish RCVS and PACNS within the
Nasal decongestants 0 1 (2) 0.33
first few days of presentation, especially in patients with
Clinical manifestations challenging presentations such as multiple strokes on imaging,
Systolic pressure 130 (96–210) 150 (110–230) <0.0001
and strengthens the results previously reported by Singhal
et al.3 First, patients with RCVS were more commonly
Diastolic pressure 70 (50–120) 88 (50–112) <0.0001
women and migrainers. Precipitating factors such as vasoac-
Headaches at onset 59 (54) 51 (96) <0.0001 tive drug exposure, emotional stress, or postpartum were
found by careful questioning in more than two-third of
TCHs at onset 3 (3) 37 (70) <0.0001
patients with RCVS but were very rarely seen in patients with
Any TCHs 3 (3) 45 (85) <0.0001 PACNS. Second, headache was almost twice as common in
Single TCHs 3 (100) 11/45 (24) 0.005 RCVS than in PACNS, with distinctive features. TCHs were
rare (3%) and never recurrent in our patients with PACNS.
Recurrence of TCHs 0 34/45 (76) <0.0001
Conversely, they were almost omnipresent in patients with
Any neurologic deficit 97 (88) 24 (45) <0.0001 RCVS, of whom >85% had the characteristic pattern of re-
Transitory deficit 27 (25) 11 (21) 0.59
current TCH over 1–2 weeks. Third, focal neurologic deficits
were found in >85% of patients with PACNS often associated
Persistent deficit 72 (66) 16 (30) <0.0001
with multiple bilateral and deep ischemic lesions, whereas
Motor deficit 82 (75) 7 (13) <0.0001 in RCVS, focal deficits affected only a quarter of patients
and were secondary to acute infarctions in less than 10%.
Sensory disorder 39 (35) 10 (19) 0.03
Three-fourths of patients with PACNS had a motor deficit
Speech disorder 56 (51) 6 (11) <0.0001 compared with only 6% of patients with RCVS. The rate of
Seizures 35 (32) 7 (13) 0.01 focal deficits in our patients with RCVS is lower than the 43%
observed in the American RCVS cohort, which were linked to
Cognitive disorder 52 (47) 7 (13) <0.0001
brain infarction in 39%.3 The frequency of focal deficits in
Vigilance impairment 30 (27) 0 <0.0001 a series of RCVS is known to vary according to patterns of
Cerebellar ataxia 33 (30) 2 (4) <0.0001
recruitment.6–8,24 Our large RCVS recruitment through an
emergency headache center probably favored inclusion of
Vision disorder 20 (18) 11 (21) 0.70
purely cephalalgic RCVS cases, whereas a multicentric re-
Abbreviations: PACNS = primary angiitis of the CNS; RCVS = reversible ce-
cruitment through stroke centers would have increased the
rebral vasoconstriction syndrome; SSRI = selective serotonin reuptake in- proportion of severe RCVS cases with strokes, as in the
hibitor; TCH = thunderclap headache.
Values are displayed as absolute number (%) or median (range). American cohort.3,7 However, whatever the reported cohorts,
multiple focal neurologic deficits are the rule in PACNS and

e8 Neurology | Volume , Number  | Month 0, 2018 Neurology.org/N

Copyright ª 2018 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
 affect twice more patients than in RCVS. Fourth, the first
Table 6 Paraclinical characteristics at diagnosis of brain imaging performed at hospital admission was constantly
patients with PACNS and RCVS with abnormal abnormal in PACNS, while 69% of patients with RCVS
imaging at initial presentation showed no lesion on first imaging and 62% still had normal
PACNS RCVS neuroimaging a few days later. However, a significant minority
cohort cohort of patients with RCVS (8% in our cohort) can present with
(n = 110) (n = 53) p Value
isolated headaches and initial normal brain scan and show
C-reactive protein, 6 (0–95) 4 (4–58) 0.11 a neurologic and radiologic deterioration within the first few
mg/L
days, while already treated with calcium channel blockers.
Abnormal CSF 71/102 (70) 29/40 (73) 0.73 This stepwise evolution of RCVS has already been described7
analysis
and should not lead to GC treatment initiation, which might
White blood cell 9 (0–425) 3 (0–34) 0.0001 be deleterious.17 Among patients with RCVS with abnormal
count
imaging, cortical subarachnoid hemorrhage was the most
Protein level, g/L 0.70 (0.1–4.1) 0.41 (0.23–0.92) 0.0001 frequent finding. Several studies have shown that in patients
Acute ischemia on 84 (76) 10 (19) <0.0001
combining cortical subarachnoid hemorrhage, TCHs, and
first CNS imaging arterial irregularities on cerebral angiography, RCVS is the
Single CNS 20/84 (24) 5/10 (50) 0.08
most likely diagnosis.21,25 However, we reported in this study
infarction 3 patients with PACNS showing TCHs, associated with
cortical subarachnoid hemorrhage and abnormal angiography
Multiple CNS 64/84 (76) 5/10 (50) 0.08
infarctions in 2. Such patients with overlapping diseases have been
reported before3 and represent a diagnostic and therapeutic
Border-zone 7/84 (8) 7/10 (50) 0.0002
infarction challenge as GC should be started as soon as possible in
PACNS and might be avoided in RCVS. These cases also
Deep infarction 81/84 (96) 1/10 (10) <0.0001
remind that cortical subarachnoid hemorrhage is a major
Vertebrobasilar 21/84 (25) 2/10 (20) 0.73 cause of TCHs, independently of RCVS. Fifth, we found
infarction
distinct CSF results in our comparative study with in-
Parenchymal 22 (20) 11 (21) 0.91 flammatory features in the setting of PACNS. A CSF protein
hemorrhage
level above 1.1 g/L should exclude RCVS.
Lobar 12 (11) 6 (11) 0.94
Altogether, our study confirms the findings observed by
Deep 14 (13) 6 (13) 0.80
Singhal et al.3 on the high PPV for RCVS diagnosis of re-
Meningeal 18 (16) 35 (66) <0.0001 current TCH or of the combination of single TCH with either
hemorrhage
normal brain imaging or cortical-only infarctions or vasogenic
Subarachnoid 18 (16) 34 (64) <0.0001 edema. Conversely, we also confirm that the combination of
hemorrhage
abnormal angiogram in a patient without TCH but with deep
Ventricular 2 (2) 1 (2) 0.98 brain infarctions and abnormal CSF is highly predictive of
hemorrhage
PACNS. However, this combination showed either poor
Subdural 2 (2) 1 (2) 0.98 sensitivity or poor specificity.
hematoma

Posterior 1 (1) 11 (21) <0.001 The present comparison between these 2 conditions is limited
reversible by 2 important points. First, PACNS is not a homogeneous
encephalopathy
syndrome condition, and its presentation is widely variable. Different
disease subgroups have been identified, some of them linked
Abnormal CTA/ 84/109 (77) 53 (100) <0.0001
MRA/DSA to the size of involved vessels.18,26–28 As compared with
patients with large-vessel PACNS, patients with small-vessel
Abnormal MRA 67/101 (66) 42/48 (88) 0.006
PACNS often show more encephalopathic presentation,
Abnormal DSA 74/94 (79) 36/40 (90) 0.12 smaller acute brain infarctions on imaging, and more abnor-
Aneurysms on 10/109 (9) 5 (8) 0.81
mal CSF. Angiography is more likely negative in these
DSA or MRA patients, but biopsy often leads to diagnosis. In these patients,
Intracranial 1/109 (1) 13 (21) <0.0001
the diagnosis of RCVS is not evoked. Conversely, large-vessel
vascular PACNS is more often responsible for large and multiple brain
dissection infarctions leading to focal neurologic deficits associated with
Abbreviations: CTA = CT angiography; DSA = digital subtraction angiography;
angiographic abnormalities.18,26 In these patients, RCVS must
MRA = MR angiography; PACNS = primary angiitis of the CNS; RCVS = re- be ruled out. It has been shown in these patients that biopsy is
versible cerebral vasoconstriction syndrome.
Values are displayed as absolute number (%) or median (range). less frequently useful, often providing negative results.29,30
The present study offers new insights to distinguish PACNS

Neurology.org/N Neurology | Volume , Number  | Month 0, 2018 e9

Copyright ª 2018 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
 Table 7 Distinction of RCVS from PACNS using the criteria from Singhal et al.3
Variables Sensitivity Specificity PPV NPV

Recurrent TCH 87 (81–92) 100 (97–100) 100 (95–100) 83 (77–88)

Single TCH 7 (4–12) 97 (92–99) 80 (54–93) 34 (39–41)

+ Normal brain imaging 1 (0–3) 100 (97–100) 100 (47–100) 36 (38.7–39.3)

+ Cortical-only infarction 1 (0–4) 100 (97–100) 100 (55–100) 39 (38.8–39.5)

+ Vasogenic edema (PRES) 2 (0–5) 100 (97–100) 100 (65–100) 39 (38.8–39.8)

+ ICH or cSAH 4 (2–8) 97 (92–99) 70 (38–90) 39 (38–40)

No TCH; positive angiogram

For the diagnosis of RCVS 6 (3–10) 25 (18–35) 11 (6–18) 15 (11–19)

+ Normal brain imaging 1 (0–4) 100 (97–100) 100 (24–100) 39 (38.8–39.5)

+ Cortical-only infarction 1 (0–3) 100 (97–100) 100 (30–100) 39 (38.8–39.3)

+ Vasogenic edema (PRES) 2 (1–5) 99 (95–100) 75 (24–97) 39 (24–97)

+ ICH or cSAH 3 (1–7) 80 (71–87) 19 (8–37) 34 (32–37)

For the diagnosis of PACNS 75 (65–82) 94 (90–97) 89 (82–94) 85 (81–89)

+ Deep or brainstem infarcts 65 (55–73) 99 (97–100) 99 (91–100) 82 (77–85)

+ Abnormal CSF 45 (35–54) 96 (92–98) 88 (77–94) 73 (70–76)

+ Both of the above 41 (31–51) 99 (97–100) 98 (85–100) 74 (71–77)

Abbreviations: cSAH = cortical subarachnoid hemorrhage; ICH = intracerebral hemorrhage; NPV = negative predictive value; PACNS = primary angiitis of the
CNS; PPV = positive predictive value; PRES = posterior reversible encephalopathy syndrome; RCVS = reversible cerebral vasoconstriction syndrome; TCH =
thunderclap headache.
Adapted from reference 3.

from RCVS, in the absence of histology, by careful analysis of patients with RCVS in our study underwent biopsy or was
anamnestic characteristics, clinical and radiologic findings, treated by immunosuppressive drugs. Finally, the initial dem-
and outcomes. Second, the important differences we observed onstration of multiple and bilateral abnormalities on angio-
among our cohorts have to be downplayed by the less severe graphic imaging does not help distinguish both diseases.
presentation of our patients with RCVS. In the study by However, Singhal et al.3 observed some differences on vascular
Singhal et al.,3 patients with RCVS presented with nearly morphology. Patients with RCVS were more likely to develop
twice as many neurologic symptoms as ours (43% vs 24% in symmetric, concentric, and smooth taper lesions, with segmental
our cohort), and 17% of them underwent brain biopsy. As dilation, than patients with PACNS. Conversely, the later may
a result, half the patients also received GCs. Distinction be- show more eccentric narrowings than patients with RCVS. We
tween PACNS and RCVS is thus probably more challenging did not conduct similar analyzes in our study precluding any
in cases of severe RCVS. However, in the comparative study confirmation, but further studies are planned. A future pro-
of Singhal et al.,3 similar distinctions regarding clinical and spective study including and analyzing all consecutive patients
radiologic presentations were demonstrated. with neurovascular abnormalities is thus required. A few small
studies have suggested that MR vessel wall imaging might help
Some other limitations have to be highlighted. First, the retro- differentiate PACNS and RCVS, but the sensitivity and speci-
spective design of our study limits complete data retrieval. ficity of this imaging technique have not been studied.31–34
However, we made sure to include in the database of the present
study only data available in both cohorts. Second, as previously Finally, our study confirms that careful analysis of clinical
stated, our 2 cohorts showed some differences with other pub- context, headache features, and patterns of brain lesions can
lished series, limiting the extrapolation of our results. Further distinguish PACNS and RCVS within the first few days of
comparative works with other non-French cohorts are required admission in most cases. However, diagnosis remains chal-
to confirm our findings. Third, our study does not reflect real-life lenging in a few cases.
practice, as reported patients were selected. Physicians from an
emergency headache center are probably more prone to di- Author contributions
agnose RCVS earlier than other physicians from general emer- H. de Boysson: concepted the study, acquired and interpreted
gency departments. This may also explain why none of the data, and drafted the manuscript. J.-J. Parienti: interpreted data,

e10 Neurology | Volume , Number  | Month 0, 2018 Neurology.org/N

Copyright ª 2018 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
performed statistical analysis, and read the manuscript. 12. Calabrese LH, Mallek JA. Primary angiitis of the central nervous system: report of 8
new cases, review of the literature, and proposal for diagnostic criteria. Medicine 1988;
J. Mawet, C. Arquizan, G. Boulouis, C. Burcin, O. Naggara, 67:20–39.
M. Zuber, E. Touzé, A. Aouba, M.-G. Bousser, and C. Pagnoux: 13. Hajj-Ali RA, Singhal AB, Benseler S, Molloy E, Calabrese LH. Primary angiitis of the
CNS. Lancet Neurol 2011;10:561–572.
acquired data and revised the manuscript. A. Ducros: con- 14. Salvarani C, Brown RD Jr, Hunder GG. Adult primary central nervous system vasculitis.
cepted the study, acquired and interpreted data, and drafted the Lancet 2012;380:767–777.
15. Ducros A. Reversible cerebral vasoconstriction syndrome. Lancet Neurol 2012;11:
manuscript. 906–917.
16. Ducros A, Hajj-Ali RA, Singhal AB, Wang SJ. Reversible cerebral vasoconstriction
syndrome. JAMA Neurol 2014;71:368.
Study funding 17. Singhal AB, Topcuoglu MA. Glucocorticoid-associated worsening in reversible ce-
No targeted funding reported. rebral vasoconstriction syndrome. Neurology 2017;88:228–236.
18. de Boysson H, Boulouis G, Aouba A, et al. Adult primary angiitis of the central nervous
system: isolated small-vessel vasculitis represents distinct disease pattern. Rheuma-
Disclosure tology 2017;56:439–444.
H. de Boysson, J.-J. Parienti, J. Mawet, C. Arquizan, G. Bou- 19. de Boysson H, Boulouis G, Dequatre N, et al. Tumor-like presentation of primary
angiitis of the central nervous system. Stroke 2016;47:2401–2404.
louis, C. Burcin, O. Naggara, M. Zuber, E. Touzé, A. Aouba, 20. de Boysson H, Parienti JJ, Arquizan C, et al. Maintenance therapy is associated with
and M.-G. Bousser report no disclosures relevant to the better long-term outcomes in adult patients with primary angiitis of the central ner-
vous system. Rheumatology 2017;56:1684–1693.
manuscript. C. Pagnoux reports having received speaker fees 21. Ducros A, Fiedler U, Porcher R, Boukobza M, Stapf C, Bousser MG. Hemorrhagic
from Roche. A. Ducros reports no disclosures relevant to the manifestations of reversible cerebral vasoconstriction syndrome: frequency, features,
and risk factors. Stroke 2010;41:2505–2511.
manuscript. Go to Neurology.org/N for full disclosures. 22. Mawet J, Boukobza M, Franc J, et al. Reversible cerebral vasoconstriction syndrome
and cervical artery dissection in 20 patients. Neurology 2013;81:821–824.
Received April 24, 2018. Accepted in final form July 13, 2018. 23. Choi HA, Lee MJ, Choi H, Chung CS. Characteristics and demographics of reversible
cerebral vasoconstriction syndrome: a large prospective series of Korean patients.
Cephalalgia 2017;38:765–775.
References 24. Chen SP, Fuh JL, Wang SJ, et al. Magnetic resonance angiography in reversible
1. Hammad TA, Hajj-Ali RA. Primary angiitis of the central nervous system and re- cerebral vasoconstriction syndromes. Ann Neurol 2010;67:648–656.
versible cerebral vasoconstriction syndrome. Curr Atheroscler Rep 2013;15:346. 25. Muehlschlegel S, Kursun O, Topcuoglu MA, Fok J, Singhal AB. Differentiating re-
2. Néel A, Auffray-Calvier E, Guillon B, et al. Challenging the diagnosis of primary versible cerebral vasoconstriction syndrome with subarachnoid hemorrhage from
angiitis of the central nervous system: a single-center retrospective study. J Rheumatol other causes of subarachnoid hemorrhage. JAMA Neurol 2013;70:1254–1260.
2012;39:1026–1034. 26. Schuster S, Bachmann H, Thom V, et al. Subtypes of primary angiitis of the CNS
3. Singhal AB, Topcuoglu MA, Fok JW, et al. Reversible cerebral vasoconstriction identified by MRI patterns reflect the size of affected vessels. J Neurol Neurosurg
syndromes and primary angiitis of the central nervous system: clinical, imaging, and Psychiatry 2017;88:749–755.
angiographic comparison. Ann Neurol 2016;79:882–894. 27. Cellucci T, Tyrrell PN, Sheikh S, Benseler SM. Childhood primary angiitis of the
4. Birnbaum J, Hellmann DB. Primary angiitis of the central nervous system. Arch central nervous system: identifying disease trajectories and early risk factors for
Neurol 2009;66:704–709. persistently higher disease activity. Arthritis Rheumatol 2012;64:1665–1672.
5. de Boysson H, Zuber M, Naggara O, et al. Primary angiitis of the central nervous 28. Salvarani C, Brown RD Jr, Calamia KT, et al. Angiography-negative primary central
system: description of the first fifty-two adults enrolled in the French cohort of nervous system vasculitis: a syndrome involving small cerebral vessels. Medicine
patients with primary vasculitis of the central nervous system. Arthritis Rheumatol 2008;87:264–271.
2014;66:1315–1326. 29. de Boysson H, Pagnoux C. Letter by de Boysson and Pagnoux regarding article,
6. Ducros A, Boukobza M, Porcher R, Sarov M, Valade D, Bousser MG. The clinical and “Diagnostic yield and safety of brain biopsy for suspected primary central nervous
radiological spectrum of reversible cerebral vasoconstriction syndrome: a prospective system angiitis”. Stroke 2016;47:e256.
series of 67 patients. Brain 2007;130:3091–3101. 30. Torres J, Loomis C, Cucchiara B, Smith M, Messe S. Diagnostic yield and safety of
7. Singhal AB, Hajj-Ali RA, Topcuoglu MA, et al. Reversible cerebral vasoconstriction brain biopsy for suspected primary central nervous system angiitis. Stroke 2016;47:
syndromes: analysis of 139 cases. Arch Neurol 2011;68:1005–1012. 2127–2129.
8. Chen SP, Fuh JL, Lirng JF, Wang YF, Wang SJ. Recurrence of reversible cerebral 31. Chen SP, Fuh JL, Lirng JF, Wang SJ. Hyperintense vessels on flair imaging in re-
vasoconstriction syndrome: a long-term follow-up study. Neurology 2015;84: versible cerebral vasoconstriction syndrome. Cephalalgia 2012;32:271–278.
1552–1558. 32. Chen SP, Wang SJ. Hyperintense vessels: an early MRI marker of reversible cerebral
9. John S, Singhal AB, Calabrese L, et al. Long-term outcomes after reversible cerebral vasoconstriction syndrome? Cephalalgia 2014;34:1038–1039.
vasoconstriction syndrome. Cephalalgia 2016;36:387–394. 33. Mandell DM, Matouk CC, Farb RI, et al. Vessel wall MRI to differentiate between
10. Salvarani C, Brown RD Jr, Christianson TJ, et al. Adult primary central nervous reversible cerebral vasoconstriction syndrome and central nervous system vasculitis:
system vasculitis treatment and course: analysis of one hundred sixty-three patients. preliminary results. Stroke 2012;43:860–862.
Arthritis Rheumatol 2015;67:1637–1645. 34. Mossa-Basha M, Shibata DK, Hallam DK, et al. Added value of vessel wall magnetic
11. Benseler SM, Silverman E, Aviv RI, et al. Primary central nervous system vasculitis in resonance imaging for differentiation of nonocclusive intracranial vasculopathies.
children. Arthritis Rheumatol 2006;54:1291–1297. Stroke 2017;48:3026–3033.

Neurology.org/N Neurology | Volume , Number  | Month 0, 2018 e11

Copyright ª 2018 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
 Primary angiitis of the CNS and reversible cerebral vasoconstriction syndrome: A
comparative study
Hubert de Boysson, Jean-Jacques Parienti, Jérôme Mawet, et al.
Neurology published online September 19, 2018
DOI 10.1212/WNL.0000000000006367

This information is current as of September 19, 2018

Updated Information & including high resolution figures, can be found at:
Services http://n.neurology.org/content/early/2018/09/19/WNL.0000000000006
367.full
Permissions & Licensing Information about reproducing this article in parts (figures,tables) or in
its entirety can be found online at:
http://www.neurology.org/about/about_the_journal#permissions
Reprints Information about ordering reprints can be found online:
http://n.neurology.org/subscribers/advertise

Neurology ® is the official journal of the American Academy of Neurology. Published continuously since
1951, it is now a weekly with 48 issues per year. Copyright © 2018 American Academy of Neurology. All
rights reserved. Print ISSN: 0028-3878. Online ISSN: 1526-632X.