You are here

Article Text

Article menu

Download PDFPDF

Cerebrovascular disease
Research paper
Subtypes of primary angiitis of the CNS identified by MRI patterns reflect the size of affected vessels
  1. Simon Schuster1,
  2. Henrike Bachmann1,
  3. Vivien Thom1,
  4. Ann-Katrin Kaufmann-Buehler2,
  5. Jakob Matschke3,
  6. Susanne Siemonsen2,
  7. Markus Glatzel3,
  8. Jens Fiehler2,
  9. Christian Gerloff1,
  10. Tim Magnus1,
  11. Götz Thomalla1
  1. 1 Department of Neurology, University Hospital Hamburg-Eppendorf, Germany
  2. 2 Department of Diagnostic and Interventional Neuroradiology, University Hospital Hamburg-Eppendorf, Germany
  3. 3 Institute of Neuropathology, University Hospital Hamburg-Eppendorf, Germany
  1. Correspondence to Simon Schuster, Department of Neurology, University Hospital Hamburg-Eppendorf, Martinistraße 52, 20246 Hamburg, Germany; s.schuster{at}uke.de

Abstract

Objective To describe patterns of diagnostic findings, and identify subgroups of primary angiitis of the central nervous system (PACNS).

Methods We retrospectively analysed 31 patients with PACNS. Cases were selected by predetermined diagnostic criteria and stratified into biopsy-proven and imaging-based PACNS. We compared clinical characteristics, cerebrospinal fluid (CSF) findings and imaging results including high-resolution vessel wall MRI between groups.

Results There were 31 cases of PACNS (mean age 45.6 years, 58.1% female), of whom 17 (55%) were biopsy-proven, 14 (45%) were based on imaging findings. Patients with a positive biopsy had fewer infarcts (29.4% vs 85.7%, p=0.003), were more likely to have meningeal and parenchymal contrast enhancement (76.5% vs 28.6%, p=0.012), were less likely to have abnormal MR angiography (11.8% vs 100%, p<0.001) and did not show vessel wall enhancement at the time of diagnosis (0% vs 76.9%, p<0.001). In contrast, patients with imaging-based diagnosis showed more frequently multiple infarcts and vessel abnormalities, with vessel wall enhancement in most of the cases. Clinical characteristics and CSF analysis did not reveal marked differences between groups.

Interpretation Multi-parametric MRI distinguishes two subtypes of PACNS that most likely differ concerning the affected vessel size. Biopsy-proven PACNS primarily involves smaller vessels beyond the resolution of vascular imaging, while imaging-based PACNS affects predominantly medium-sized vessels leading to false-negative biopsy results. Using distinct MRI patterns may be helpful for selecting patients for appropriate invasive diagnostic modalities.

https://doi.org/10.1136/jnnp-2017-315691

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Primary angiitis of the central nervous system (PACNS) is characterised by transmural inflammation of vessels exclusively in the brain, spinal cord and leptomeninges. Although the disease is rare, it is frequently considered in the differential diagnosis of cryptogenic stroke or aseptic meningoencephalitis. Distinction from other intracranial stenosing vasculopathies including atherosclerotic disease or reversible cerebral vasoconstriction syndrome (RCVS) is particularly challenging.1 Specific diagnostic markers are lacking and an invasive approach is required to confirm the diagnosis.

    The diagnostic criteria proposed by Calabrese and Malleck in 1988 demand an acquired neurological deficit otherwise unexplained and either angiographic changes consistent with vasculitis or a positive CNS biopsy.2 Of note angiographic changes have a low specificity and can be due to other non-inflammatory vasculopathies.3 4 The yield of brain biopsy depends on whether or not an affected brain area is hit and may be false negative.5 Therefore, different attempts in the last years aimed at increasing diagnostic accuracy of PACNS in the absence of histopathological confirmation. Birnbaum and Hellmann emphasised a probable diagnosis of PACNS based on the synopsis of epidemiological, clinical, laboratory and radiological findings.6 They favoured to include MRI and cerebrospinal fluid (CSF) findings since the combination of normal MRI and CSF has a high negative predictive value for the diagnosis of PACNS. Beyond the interplay of multiple established diagnostic modalities to create a signature for PACNS, new MRI vessel wall imaging techniques emerged in the last years to improve the distinction between intracranial vessel changes due to vasculitis, atherosclerosis and RCVS by different contrast enhancement patterns of the vessel wall.7 8 Recently, a distinct pattern of clinical and neuroimaging findings has been suggested to discriminate between RCVS and PACNS.9

    Furthermore, a major diagnostic challenge is the heterogeneity of PACNS.10–12 Notably, certain PACNS subtypes differ with regard to the size of affected vessels. Among others, the spectrum encompasses biopsy-proven PACNS with normal angiogram due to small vessel involvement beyond the resolution of cerebrovascular studies, but also angiographically defined PACNS where CNS biopsy can be normal due to predominantly affected medium-sized vessels not accessible by biopsy. Choosing a tailored diagnostic approach in order to recognise a certain PACNS subtype is—besides clear-cut discrimination between PACNS and its mimics—a major challenge in the diagnosis of PACNS.

    In this study, we report 31 patients diagnosed with PACNS either proven by biopsy or based on imaging criteria without histopathological proof. We describe the relevant diagnostic results and compare patterns between the two subgroups with a focus on MRI findings including new vessel wall imaging techniques.

    Methods

    Patient data collection

    In a retrospective analysis, we screened our institution’s neuroradiological and neuropathological medical reports for the term ‘cerebral vasculitis’, to identify patients in whom the diagnosis of PACNS was discussed during the diagnostic work-up from 1 January 2008 through 1 June 2014. Subsequently, we reviewed the medical reports of all identified patients and excluded those who finally did not meet the diagnostic criteria for PACNS and were discharged with an alternative diagnosis. The remaining patients were reviewed again by two neurology consultants. Patients were only included in the study if there was a consensus concerning the diagnosis of PACNS. Data collection included epidemiological data, information about the clinical manifestation, laboratory investigations (blood and CSF), radiological studies with MRI including high-resolution vessel wall imaging, MR angiography (MRA), conventional digital subtraction angiography (DSA) and detailed histopathology of biopsy specimen. We did not seek ethics approval for our retrospective study; the diagnostic and therapeutic protocol that we evaluated was according to national guidelines and standards of current practice.

    Diagnostic criteria

    For diagnosis of PACNS, we generally followed the diagnostic criteria originally proposed by Calabrese2 with slight modifications: either positive CNS biopsy was demanded for definite diagnosis (biopsy-proven group) or the combination of abnormal findings of the vessels and brain parenchyma on neuroimaging consistent with the diagnosis of PACNS and thorough exclusion of other vasculopathies were obligatory to establish a probable diagnosis (imaging-based group).

    A biopsy was considered positive if vascular inflammation of intracranial arterioles was present and neuropathologist’s overall evaluation resulted in the diagnosis of CNS vasculitis. The different histopathological patterns of inflammation were recorded as previously described and included granulomatous with/without amyloid deposition, lymphocytic and necrotising vasculitic changes.5

    In case of a lacking or non-conclusive tissue confirmation, intracranial vessel abnormalities established by MRA or conventional DSA were demanded. Findings in favour of the diagnosis of PACNS included alternating areas of narrowing or multilocular occlusions of intracranial vessels in the absence of stenosis of the extracranial vessels. In addition, the exclusion of other possible aetiologies by comprehensive diagnostic work-up, and the lack of major vascular risk factors and atherosclerotic disease, was required before establishing the diagnosis of PACNS in cases without biopsy-proven diagnosis. CSF analysis was used to rule out infectious causes or an angiocentric CNS lymphoma. We also took care not to include patients with features related to RCVS such as thunderclap headache, a history of exposure to vasoactive substances or an occurrence in the postpartum period. Finally, the clinical course was taken into account, and only patients with disease progression or relapsing episodes over a period of >12 weeks were diagnosed with PACNS in order to exclude the diagnosis of RCVS.

    MR protocol and sequences

    MR scans were obtained on clinical MR scanners (3 T Skyra Magnetom, 1.5 T Avanto, 1.5 T Symphony or 1.5 T Sonata; all Siemens, Erlangen, Germany). Among other standard MR sequences such as fluid-attenuated-inversion-recovery (FLAIR) and T2-weighted (w) images, the protocol included a standard multislab three-dimensional-time-of-flight (TOF) angiography centred on the circle of Willis, and a coronal high-resolution T1-weighted spin-echo-sequence before and after contrast administration (Gadovist; Bayer Schering Pharma, Berlin, Germany). MRI was performed during the initial work-up and in the time course either when a clinical deterioration occurred or during routine follow-up appointments.

    Statistical analysis

    Two-sided Fisher’s exact test was used for the comparison of categorical variables. Mann-Whitney U test was used to compare the median of continuous variables. Data analysis was exploratory, and no correction for multiple tests was applied. Group difference with p<0.05 was considered significant. IBM SPSS software was used for statistical analysis.

    Results

    During the study period, 31 patients (18 females and 13 males) were diagnosed with PACNS. Brain biopsy was performed in 26 patients in which 17 biopsies were positive for PACNS (65%). Biopsy was non-conclusive in 9 of 26 cases. Either the biopsy was non-specific (5/9) or showed no pathology (4/9). Five patients did not undergo biopsy; in those cases, either the lesions were not easily accessible or the patients refused brain biopsy; there were no differences in symptom presentation, laboratory or imaging findings between patients with or without biopsy within the imaging-based PACNS group. All 14 patients without or non-conclusive biopsy were diagnosed based on the synopsis of MRI lesions in combination with intracranial vessel abnormalities (detected by DSA in 8/14 patients and by MRA in 6/14 patients), mostly pathological CSF findings and a response to immunosuppressive treatment. We compared clinical characteristics, CSF findings and imaging results including high-resolution vessel wall MRI between patients with biopsy-proven PACNS and imaging-based PACNS. Figure 1 illustrates how the two subgroups were composed. Table 1 summarises the diagnostic findings of each subgroup.

    Figure 1
    Figure 1

    Diagnostic test findings of biopsy-proven (n=17) and imaging-based primary angiitis of the central nervous system (PACNS) patients (n=14). DSA, digital subtraction angiography; MRA, MR angiography.

    View this table:
    Table 1

    Diagnostic findings in the biopsy-proven and imaging-based primary angiitis of the central nervous system (PACNS) patients

    Clinical data

    Mean age at diagnosis was comparable between groups (44.5 vs 47 years). Overall, symptom onset was acute in 67.7% and chronic-progressive in 32.3%. The most frequent symptoms were focal neurological deficits (83.9%), headache (64.5%) and cognitive deficits (35.5%). Neurological deficits included motor symptoms (45.2%), sensory deficits (38.7%), aphasia (32.3%), ataxia (29.0%), visual field defects (16.1%) and dysarthria (12.9%). Seizures occurred in 25.8%. Clinical characteristics were comparable between groups. Clinical outcome assessed by the Modified Rankin Scale at the end of the observation period (median 46 months, range 5–110 months) did not differ between the two subgroups (data not shown).

    Laboratory findings

    Abnormal CSF was found in 29 of 31 cases (93.5%) with no significant differences between groups. An elevated leucocyte count was found in 7 of 17 (41.2%) patients with biopsy-proven PACNS and in 4 of 14 (28.6%) patients with imaging-based PACNS (p=0.707). An increased protein level was present in 12 of 17 (70.6%) and 7 of 14 (50%) patients, respectively (p=0.288). Isolated CSF oligoclonal bands were seen in 6 of 17 (35.3%) patients with biopsy-proven PACNS and 2 of 14 (14.3%) patients with imaging-based PACNS (p=0.24). Blood results were mostly normal including antibody screening for rheumatological diseases except of slightly elevated antinuclear antibody titres (all <1:640) in 12 patients. Antineutrophil cytoplasmic antibodies (p-ANCA and c-ANCA) were studied in 29/31 patients and were negative in all cases. In the two cases without information on ANCA, PACNS was proven by biopsy and follow-up of >5 years did not reveal other than CNS involvement.

    MRI findings

    All patients showed abnormal findings on MRI (n=31).

    Parenchymal imaging

    Focal ischaemic infarcts occurred significantly less often in the biopsy-proven PACNS patients compared with the imaging-based PACNS patients at the time of diagnosis (5/17 vs. 12/14, 29.4% vs. 85.7%, p=0.003). Within the follow-up period the appearance of infarcts increased in the biopsy-proven PACNS patients (9/17, 52.9%). Overall, ischaemic infarcts occurred in 21 of 31 patients (67.8%). Gadolinium-enhanced lesions (parenchymal and meningeal) were seen in 17 of 31 patients (54.8%). MRI of the biopsy-proven PACNS patients was characterised by significantly more frequent parenchymal and leptomeningeal contrast enhancement compared with the imaging-based PACNS patients (13/17 vs 4/14, 76.5% vs 28.6%, p=0.012). Mass lesions were seen in two patients. Meningeal contrast enhancement (CE) and mass lesions were only present in the biopsy-proven PACNS group. Non-specific T2-hyperintensities were detected in all patients except one patient with a biopsy-proven diagnosis. No parenchymal abnormality was present in this patient, but only severe brain atrophy; biopsy revealed transmural inflammatory changes only in the very small vessels. Haemorrhagic changes occurred in 5 of 31 (16,1%) patients, of which 3 were microbleeds, 1 patient had haemorrhagic changes of a mass lesion and 1 patient showed mild sulcal subarachnoid bleeding, with no difference between the groups. In addition, haemorrhagic transformation secondary to ischaemic lesions was observed in two patients with large territorial infarcts.

    Vessel imaging

    Intracranial vessel abnormalities on MRA were seen in 2 of 17 patients with biopsy-proven PACNS (11.8%) at the time of diagnosis while present in all imaging-based PACNS patients (p<0.001). Later in the course of disease, MRA—performed during follow-up appointments or after new onset of signs and symptoms—showed intracranial vessel irregularities in 10/17 biopsy-proven PACNS patients (58.8%). If vessel abnormalities were present, mostly multiple vessel irregularities could be detected (21/24, 87.5%). High-resolution MR vessel wall imaging initially revealed no CE of the vessel wall in the biopsy-proven PACNS patients, but in 10 of 13 imaging-based PACNS patients (76.9%, p<0.001). Within the follow-up period, CE of the vessel wall was evident in 3 of 14 biopsy-proven PACNS patients (21.4%, p=0.007). Figure 2 showcases the MRI findings of two patients—one with predominantly small and the other with medium-sized vessel involvement.

    Figure 2
    Figure 2

    MRI findings of patients with predominantly small vessel involvement (patient 1) and medium-sized vessel involvement (patient 2). Patient 1: (A) T2w-FLAIR periventricular hyperintensities (B) T1w contrast enhancement (CE) of periventricular lesion, (C) no DWI lesion, (D) no CE of the vessel wall in T1w CE images and (E) normal TOF angiography. Patient 2: (A) right parietal T2w-FLAIR lesion, (B) DWI lesions with (C) ADC correlation, (D, E) CE of the proximal vessel wall in T1w CE images and (F) digital subtraction angiography with characteristic vessel wall irregularities. FLAIR, fluid-attenuated-inversion-recovery; CE, contrast enhancement; DWI, diffusion-weighted-imaging; TOF, time-of-flight; ADC, apparent-diffusion-coefficient.

    Histopathology

    We identified different histopathological patterns in our biopsy-proven PACNS cohort: two patients had a granulomatous PACNS, further two patients had a granulomatous vasculitis with β-amyloid deposition (amyloid-beta-related angiitis), six patients had a lymphocytic PACNS and two patients had a necrotising vasculitis in the biopsy specimen. In five patients, the pattern could not be classified in a specific subtype, transmural or perivascular T-cell infiltration was present and neuropathologist’s overall evaluation resulted in the diagnosis of CNS vasculitis.

    Discussion

    PACNS is a diagnostic challenge: clinical presentation is variable, specific diagnostic markers are lacking and usually an invasive diagnostic approach (ie, CNS biopsy) is considered necessary to confirm the diagnosis. However, brain biopsy may be false negative or not feasible. As a consequence, diagnostic criteria were established that allow for diagnosing PACNS based on angiographic findings in the absence of an alternative aetiology. As of yet, there are little data that may guide the choice of the optimal diagnostic approach in suspected PACNS. The systematic retrospective analysis of PACNS cases diagnosed in our hospital extends this knowledge and may serve as a basis for the guidance of diagnostic decisions in PACNS. The patterns of biopsy results and imaging findings in our cohort of PACNS suggest that there are two subtypes of PACNS that differ with regard to the type of vessels predominantly affected (ie, small vs medium to large vessel) as recently suggested.13

    Angiography-negative, biopsy-proven PACNS appears to be a distinct subtype of PACNS with primarily involvement of small intracranial vessels.14–16 In our cohort, only about 1 in 10 patients with biopsy-proven PACNS had abnormal findings of intracranial arteries on initial MRA or DSA. In other series of biopsy-proven PACNS, the frequency of brain vessel abnormalities was slightly higher than in our cohort, but presence of abnormal vessels still had a rather low sensitivity in the diagnosis of biopsy-proven PACNS. It was reported that only 27% of angiograms performed in biopsy-proven granulomatous PACNS cases were considered to show changes typical for vasculitis.17 In another study, only 5 of 14 (35.7%) biopsy-proven PACNS patients showed vessel irregularities on DSA.18 In a cohort study of Salvarani et al of 24 patients with a biopsy-proven diagnosis who additionally underwent DSA, only 8 had abnormal findings in DSA (33.3%).13 Recently, de Boysson et al described 26 patients with biopsy-proven PACNS who did not have stenosis on either MR or conventional angiograpy.16 Thus, normal angiography findings are rather expected if disease is limited to smaller vessels since the resolution of DSA is limited in vessels <500 μm.1

    In contrast, angiographically defined PACNS with negative CNS biopsy seems to be another subtype of the PACNS spectrum. The affected vessels are usually medium-sized and therefore not accessible by biopsy. Also, 9 of 14 patients of our imaging-based PACNS group underwent biopsy, none revealed pathology characteristic for PACNS, that is, has to be considered false negative. In the largest published cohort of 163 PACNS patients, 31 patients with abnormal DSA additionally underwent biopsy, the yield of positive biopsy in this subgroup was only 25.8% (8/31).13 In another study of 46 patients, where biopsy was performed to confirm the diagnosis of PACNS, overall sensitivity of biopsy was 63%. However, when only PACNS patients with angiographic findings suggesting vasculitis were considered, biopsy showed typical signs of PACNS only in 22.7% (5/22).5 The frequent finding of false-negative biopsies in angiography-positive PACNS may be explained by vasculitic changes limited to proximal or medium-sized intracranial vessels which are not covered by brain biopsy. A single case, where postmortem pathology showed lymphocytic vessel wall infiltration restricted to the proximal cerebral arteries while smaller arteries were not involved, is in line with the assumption that a subtype of PACNS may only affect primarily larger vessels.19

    There is not a strict dichotomy of vasculitic involvement of either proximal or smaller vessels as cases of biopsy and angiography positive PACNS are well described. Predominantly small or larger vessel PACNS are rather the end points of a spectrum of PACNS subtypes. In our cohort, 4 of 17 biopsy-proven PACNS patients showed vessel irregularities on MRA or DSA at the time of diagnosis. Of note, the detection of intracranial vessel abnormalities on neuroimaging increased during the time course in biopsy-proven PACNS patients to 58.8% (10/17), which might reflect a spatial dispersion from initial involvement of small vessels beyond the resolution of imaging to larger vessels during the course of the disease.

    Most of the published PACNS cases in the literature are purely imaging-based without histopathological proof including 105 of 163 patients (64.4%) in the Salvarani cohort.13 We do not know if these patients truly have a vasculitis or rather non-inflammatory vasculopathies. New high-resolution vessel wall MRI techniques emerged in the last years in clinical practice in order to distinguish vasculitic from atherosclerotic changes or those associated with RCVS in the absence of histopathological proof. In several recently performed studies, CE of the vessel wall is a frequent finding in patients with active PACNS,8 but could also be detected in patients with atherosclerotic disease and lacks therefore specificity.20 21 Despite the missing selectivity towards atherosclerotic changes, vessel wall imaging can be a useful tool to differentiate between RCVS and PACNS because lesions due to RCVS show no to mild CE of the vessel wall.7 22 23 In our study, 85% of patients with an imaging-based PACNS had CE of the vessel wall which was valued in favour of PACNS and against the competing diagnosis of RCVS. Vessel wall imaging techniques might help to reach selectivity towards patients with RCVS; however, prospective studies are necessary to confirm the diagnostic relevance. A recently published study by Singhal et al points out that clear-cut distinction between RCVS and PACNS is also possible considering a certain pattern of clinical and neuroimaging characteristics.9 Recently, concentric CE of the intracranial vessel walls was also evident in patients with intravascular CNS lymphoma24 which illustrates the rather low specificity of the investigation; in the cases reported in our study, intravascular CNS lymphoma as a mimic of PACNS can be ruled out due to survival during a median follow-up period of 41.5 months in our imaging-based cohort.

    In contrast to the imaging-based PACNS group, we could not detect CE of the vessel wall in our biopsy-proven PACNS patients at the time of diagnosis; even in the further course of the disease, CE of the vessel wall was only evident in 3 of 14 patients. These findings additionally reflect that mainly inflammation of small-sized vessels beyond the resolution of vessel wall imaging studies is present in our biopsy-proven group.

    Besides vascular imaging, parenchymal MRI is the main initial non-invasive imaging modality for the work-up of patients with suspected PACNS with a sensitivity of about 90%–100%.6 25 In line with previous reports, MRI was abnormal in all our cases of PACNS. T2-hyperintensities were non-specific findings in all patients except one, where no parenchymal lesion but only severe brain atrophy was present. In this patient, biopsy revealed involvement of only very small vessels with transmural infiltrates on biopsy, suggesting that parenchymal lesions may have been below the resolution of MRI.26

    Patterns of parenchymal lesions differed significantly between imaging-based and biopsy-proven cases as recently described.13 16 While more focal infarcts were present in imaging-based PACNS patients, parenchymal and leptomeningeal contrast enhancement was seen more frequently in biopsy-proven PACNS patients. Mass lesions were only seen in the biopsy-proven PACNS cohort. Of note, the results may be biased due to the selection of subgroups: PACNS patients with meningeal contrast enhancement or mass lesions are promising candidates for a brain biopsy in the first place; therefore, is is expected that they only occur in the biopsy-proven subgroup. The almost complete lack of abnormal MRA findings in combination with meningeal contrast enhancement or mass lesions in the biopsy-proven subgroup can mislead the diagnostician towards other inflammatory or neoplastic processes. In those cases, we consider biopsy as obligatory to confirm the diagnosis. On the contrary, MRI patterns that come along more like a typical vascular process with multiple infarcts may not be promising for brain biopsy: all performed nine biopsies in our imaging-based cohort showing multiple infarct lesions on MRI were non-conclusive. It is undoubtable that histopathological proof is the gold standard in the diagnostic work-up of PACNS patients, but the yield of positive biopsy might be lower in patients that present with recurrent strokes and vessel abnormalities on neuroimaging presumably caused by more proximal vasculitic processes which are not accessible by biopsy. In those cases, a careful selection of patients for brain biopsy is advisable especially with regard to the findings of a recently published study showing a relatively high risk of biopsy (16%) with a low yield of histopathological proof of PACNS (11%) in 79 patients who underwent biopsy for suspected PACNS.27 An alternative diagnosis could be established in 30% of cases. Of note, the cohort had differing characteristics compared with our PACNS cohorts and a classification with regard to the affected vessel size was not provided. Nonetheless in certain cases, biopsy remains important in order to allow a clear-cut discrimination between PACNS and its mimics. For instance, discrimination of CNS lymphoma can be challenging as the sensitivity of CSF studies including cytology and flow cytometry is rather low.28 With respect to CSF analysis in PACNS, it mainly serves to exclude infectious or malignant diseases mimicking PACNS. CSF findings are not specific and include either mild pleocytosis <20 cells/microliter or an elevated protein concentration or both, occasionally intrathecal IgG synthesis is present.1

    The strength of our study is the careful selection of PACNS patients with a relatively high number of biopsy-proven PACNS patients compared with previous studies, and the availability of new vessel wall MRI in most of the patients. We carefully excluded patients with features related to RCVS; the follow-up time of at least 6 months (if not deceased earlier) helped to distinguish between RCVS and PACNS. Some patients stabilised or improved under immunosuppressive treatment (steroid pulse therapy, followed by oral tapering of steroids and intermittent intravenous cyclophosphamide pulse therapy), which confirmed our suspicion of a vasculitic process in the imaging-based cohort. However, our retrospective study has also limitations mainly due to missing data, and selection and referral bias that we cannot control.

    In conclusion, MRI with parenchymal and vessel imaging identifies two subtypes of PACNS that most likely differ in the affected vessel size. This simplified diagnostic approach does not fully consider the spectrum of PACNS subtypes, but might help to select patients for appropriate invasive diagnostic modalities. Figure 3 illustrates a possible diagnostic algorithm. A positive biopsy is more likely expected in small vessel PACNS presenting with parenchymal and leptomeningeal contrast enhancement or mass lesions on neuroimaging, vascular irregularities or multiple infarct lesions may not be in the spotlight; in contrast, biopsy may be less seminal in larger vessel PACNS that comes along more like a typical vascular process, benefit and risk should be carefully considered and other non-invasive modalities such as vessel wall imaging techniques might be of preferential use. Future research will have to elucidate the relevance of MRI patterns in PACNS.

    Figure 3
    Figure 3

    Possible diagnostic algorithm for primary angiitis of the central nervous system (PACNS). MRI findings represent the end points of a spectrum of PACNS subtypes. CSF, cerebrospinal fluid.

    References

    1. ↵
      2. Hajj-Ali RA ,
      3. Singhal AB ,
      4. Benseler S , et al
      . Primary angiitis of the CNS. Lancet Neurol 2011;10:561–72.doi:10.1016/S1474-4422(11)70081-3
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      2. Calabrese LH ,
      3. 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;67:20–39.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      2. Duna GF ,
      3. Calabrese LH
      . Limitations of invasive modalities in the diagnosis of primary angiitis of the central nervous system. J Rheumatol 1995;22:662–7.
      OpenUrlPubMedWeb of Science
    4. ↵
      2. Kadkhodayan Y ,
      3. Alreshaid A ,
      4. Moran CJ , et al
      . Primary angiitis of the central nervous system at conventional angiography. Radiology 2004;233:878–82.doi:10.1148/radiol.2333031621
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      2. Miller DV ,
      3. Salvarani C ,
      4. Hunder GG , et al
      . Biopsy findings in primary angiitis of the central nervous system. Am J Surg Pathol 2009;33:35–43.doi:10.1097/PAS.0b013e318181e097
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      2. Birnbaum J ,
      3. Hellmann DB
      . Primary angiitis of the central nervous system. Arch Neurol 2009;66:704–9.doi:10.1001/archneurol.2009.76
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      2. Mossa-Basha M ,
      3. Hwang WD ,
      4. De Havenon A , et al
      . Multicontrast high-resolution vessel wall magnetic resonance imaging and its value in differentiating intracranial vasculopathic processes. Stroke 2015;46:1567–73.doi:10.1161/STROKEAHA.115.009037
      OpenUrlAbstract/FREE Full Text
    8. ↵
      2. Küker W ,
      3. Gaertner S ,
      4. Nagele T , et al
      . Vessel wall contrast enhancement: a diagnostic sign of cerebral vasculitis. Cerebrovasc Dis 2008;26:23–9.doi:10.1159/000135649
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      2. Singhal AB ,
      3. Topcuoglu MA ,
      4. Fok JW , et al
      . Reversible cerebral vasoconstriction syndromes and primary angiitis of the central nervous system: clinical, imaging, and angiographic comparison. Ann Neurol 2016;79:882–94.doi:10.1002/ana.24652
      OpenUrlCrossRefPubMed
    10. ↵
      2. Salvarani C ,
      3. Brown RD ,
      4. Hunder GG
      . Adult primary central nervous system vasculitis. Lancet 2012;380:767–77.doi:10.1016/S0140-6736(12)60069-5
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      2. Salvarani C ,
      3. Brown RD ,
      4. Hunder GG
      . Adult primary central nervous system vasculitis: an update. Curr Opin Rheumatol 2012;24:46–52.doi:10.1097/BOR.0b013e32834d6d76
      OpenUrlPubMed
    12. ↵
      2. MacLaren K ,
      3. Gillespie J ,
      4. Shrestha S , et al
      . Primary angiitis of the central nervous system: emerging variants. QJM 2005;98:643–54.doi:10.1093/qjmed/hci098
      OpenUrlCrossRefPubMed
    13. ↵
      2. Salvarani C ,
      3. Brown RD ,
      4. Christianson T , et al
      . An update of the Mayo Clinic cohort of patients with adult primary central nervous system vasculitis: description of 163 patients. Medicine 2015;94:e738.doi:10.1097/MD.0000000000000738
    14. ↵
      2. Benseler SM ,
      3. deVeber G ,
      4. Hawkins C , et al
      . Angiography-negative primary central nervous system vasculitis in children: a newly recognized inflammatory central nervous system disease. Arthritis Rheum 2005;52:2159–67.doi:10.1002/art.21144
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      2. Salvarani C ,
      3. Brown RD ,
      4. Calamia KT , et al
      . Angiography-negative primary central nervous system vasculitis: a syndrome involving small cerebral vessels. Medicine 2008;87:264–71.doi:10.1097/MD.0b013e31818896e1
      OpenUrlCrossRefPubMed
    16. ↵
      2. de Boysson H ,
      3. Boulouis G ,
      4. Aouba A , et al
      . Adult primary angiitis of the central nervous system: isolated small-vessel vasculitis represents distinct disease pattern. Rheumatology 2017;56:439–44.doi:10.1093/rheumatology/kew434
      OpenUrl
    17. ↵
      2. Vollmer TL ,
      3. Guarnaccia J ,
      4. Harrington W , et al
      . Idiopathic granulomatous angiitis of the central nervous system. diagnostic challenges. Arch Neurol 1993;50:925–30.
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      2. Alrawi A ,
      3. Trobe JD ,
      4. Blaivas M , et al
      . Brain biopsy in primary central nervous system vasculitis. Neurology 1999;53:852–5.
      OpenUrlCrossRef
    19. ↵
      2. Pfefferkorn T ,
      3. Schüller U ,
      4. Cyran C , et al
      . Giant cell arteritis of the basal cerebral arteries: correlation of MRI, dsa, and histopathology. Neurology 2010;74:1651–3.doi:10.1212/WNL.0b013e3181df0a09
      OpenUrlCrossRefPubMed
    20. ↵
      2. Swartz RH ,
      3. Bhuta SS ,
      4. Farb RI , et al
      . Intracranial arterial wall imaging using high-resolution 3-tesla contrast-enhanced MRI. Neurology 2009;72:627–34.doi:10.1212/01.wnl.0000342470.69739.b3
      OpenUrlCrossRefPubMed
    21. ↵
      2. Pfefferkorn T ,
      3. Linn J ,
      4. Habs M , et al
      . Black blood MRI in suspected large artery primary angiitis of the central nervous system. J Neuroimaging 2013;23:379–83.doi:10.1111/j.1552-6569.2012.00743.x
      OpenUrlCrossRefPubMed
    22. ↵
      2. Obusez EC ,
      3. Hui F ,
      4. Hajj-Ali RA , et al
      . High-resolution MRI vessel wall imaging: spatial and temporal patterns of reversible cerebral vasoconstriction syndrome and central nervous system vasculitis. AJNR Am J Neuroradiol 2014;35:1527–32.doi:10.3174/ajnr.A3909
      OpenUrlAbstract/FREE Full Text
    23. ↵
      2. Mandell DM ,
      3. Matouk CC ,
      4. Farb RI , et al
      . Vessel wall MRI to differentiate between reversible cerebral vasoconstriction syndrome and central nervous system vasculitis: preliminary results. Stroke 2012;43:860–2.doi:10.1161/STROKEAHA.111.626184
      OpenUrlAbstract/FREE Full Text
    24. ↵
      2. Schaafsma JD ,
      3. Hui F ,
      4. Wisco D , et al
      . High-Resolution Vessel Wall MRI. Clin Neuroradiol 2017;27:105–8.doi:10.1007/s00062-016-0529-9
      OpenUrl
    25. ↵
      2. Salvarani C ,
      3. Brown RD ,
      4. Calamia KT , et al
      . Primary central nervous system vasculitis: analysis of 101 patients. Ann Neurol 2007;62:442–51.doi:10.1002/ana.21226
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      2. Bönstrup M ,
      3. Ott K ,
      4. Glatzel M , et al
      . Frontal lobe dementia syndrome as a first manifestation of primary angiitis of the central nervous system (PACNS). Clin Neurol Neurosurg 2016;141:92–4.doi:10.1016/j.clineuro.2015.12.016
      OpenUrl
    27. ↵
      2. Torres J ,
      3. Loomis C ,
      4. Cucchiara B , et al
      . Diagnostic yield and safety of Brain Biopsy for Suspected Primary Central Nervous System Angiitis. Stroke 2016;47:2127–9.doi:10.1161/STROKEAHA.116.013874
      OpenUrlAbstract/FREE Full Text
    28. ↵
      2. Scott BJ ,
      3. Douglas VC ,
      4. Tihan T , et al
      . A systematic approach to the diagnosis of suspected central nervous system lymphoma. JAMA Neurol 2013;70:311–9.doi:10.1001/jamaneurol.2013.606
      OpenUrl

    Footnotes

    • Contributors SS and HB: study design, data acquisition, analysis and interpretation of data, and drafting the manuscript. TV and JF: revising the manuscript. A-K K-B and SS: data acquisition and revising the manuscript. JM: data acquisition, analysis and interpretation of data, and revising the manuscript. MG: analysis and interpretation of data, and revising the manuscript. CG, TM and GT: study concept and design, analysis and interpretation of data, and revising the manuscript.

    • Competing interests None declared.

    • Provenance and peer review Not commissioned; externally peer reviewed.