Primary central nervous system vasculitis: comparison of patients with and without cerebral amyloid angiopathy

Transcription

1 Rheumatology 2008;47: Advance Access publication 27 August 2008 doi: /rheumatology/ken328 Primary central nervous system vasculitis: comparison of patients with and without cerebral amyloid angiopathy C. Salvarani 1, R. D. Brown Jr 2, K. T. Calamia 3, T. J. H. Christianson 4, J. Huston III 5, J. F. Meschia 6, C. Giannini 7, D. V. Miller 7 and G. G. Hunder 8 Objectives. To describe the clinical features and outcomes of patients with primary central nervous system vasculitis (PCNSV) and cerebral amyloid angiopathy (CAA) from a large cohort of consecutive patients with PCNSV treated at a single institution. Methods. We identified 101 consecutive patients with PCNSV admitted between January 1983 and December PCNSV diagnoses were based on findings from a central nervous system (CNS) biopsy (n ¼ 31) and conventional angiography (n ¼ 70). CNS tissue specimens from 49 cases were examined histologically, and 49 were stained for amyloid deposits. Those with vascular amyloid deposits (CAA) were compared with those without histological evidence of amyloid deposition. Results. Eight cases (26%) with CNS biopsy specimens positive for PCNSV also showed findings of CAA. Compared with patients with PCNSV only, these patients were older at diagnosis, predominantly male, had a more acute onset, a higher frequency of cognitive dysfunction and showed prominent gadolinium-enhanced leptomeningeal lesions with MRI. Histologically, all had a granulomatous vascular inflammatory pattern. Six patients responded promptly to therapy. Outcomes at last follow-up were similar in the two groups. Conclusions. PCNSV with CAA appears to form a clinical subset of PCNSV. The vasculitis influences the clinical findings to a greater degree than the presence of amyloid deposits in the vessels. KEY WORDS: Cerebral amyloid angiopathy, Cerebral angiography, Cerebral biopsy, Primary central nervous system vasculitis, Therapy. Introduction Primary central nervous system vasculitis (PCNSV) is an uncommon condition in which vascular inflammatory lesions are limited to the brain and spinal cord [1 4]. Diagnostic criteria include a newly acquired neurological deficit that is unexplained by other processes and angiographic or central nervous system (CNS) biopsy features of vasculitis [2]. Cerebral amyloid angiopathy (CAA) is a disorder characterized by the deposition of amyloid (A) peptide in the media and adventitia of small- to mediumsized arteries in the cerebral cortex and leptomeninges [5 9]. Deposition may lead to vessel fragility, rupture and subsequent intracerebral haemorrhage [5, 6], or it may cause generalized symptoms such as cognitive impairment [7, 8]. Although vascular inflammation generally is not observed in CAA, coexistence of sporadic CAA and PCNSV has been reported [10 13]. Causal relationships between CAA and vasculitis have not been determined. Recently, Scolding and coworkers [13] suggested that Arelated PCNSV was a distinct clinicopathological entity; they termed it A-related angiitis. Their conclusions were based on the combined findings of nine new cases and individual case reports already in the medical literature. Because of the relatively small number of cases studied and the lack of follow-up, the clinical spectrum of A-related PCNSV, its response to treatment and its long-term outcome is still uncertain. We recently reviewed the medical records of 101 patients with PCNSV at Mayo Clinic (Rochester, MN) over a 21-yr period [4]. CNS biopsies showed the presence of vasculitis in 31 patients. 1 Department of Neurology, Mayo Clinic, Rochester, MN, USA, 2 Department of Neurology, Mayo Clinic, Rochester, MN, 3 Division of Rheumatology, Mayo Clinic, Jacksonville, FL, 4 Division of Biostatistics, 5 Department of Radiology, Mayo Clinic, Rochester, MN, 6 Department of Neurology, Mayo Clinic, Jacksonville, FL, 7 Division of Anatomic Pathology and 8 Division of Rheumatology, Mayo Clinic, Rochester, MN, USA. In this report, we describe the clinical characteristics of a subset of 8 of the 31 biopsy-positive patients who also had vascular deposits of -amyloid peptide that were consistent with CAA. Patients and methods Patients We used the Mayo Clinic medical records linkage system to obtain a list of all patients with a possible diagnosis of CNS vasculitis between 1 January 1983 and 31 December 2003 [4]. Patients were given a diagnosis of PCNSV if a brain or spinal cord biopsy specimen showed vasculitis (transmural destructive inflammatory infiltrate) or if angiograms showed changes that were highly suggestive of vasculitis (segmental narrowing, dilatation or occlusion affecting multiple cerebral arteries in the absence of proximal vessel changes of atherosclerosis). We excluded patients with vasculitis in organs other than the CNS and those with evidence of other diseases such as SLE and infection. During the period reviewed, 101 patients at Mayo Clinic (Rochester, MN, USA) fulfilled the diagnostic criteria for PCNSV [4]. The Mayo Clinic Institutional Review Board approved this study. Biopsy specimens were reviewed by two pathologists (D.V.M. and C.G.), and angiograms were reviewed by a neuroradiologist. In cases with an uncertain initial diagnosis, the complete medical record was reviewed again by two rheumatologists (C.S. and G.G.H.) and one neurologist (R.D.B.) to reach a consensus. A standard data collection form was completed for all cases and included information about clinical manifestations at presentation and during follow-up, other medical conditions, findings from laboratory investigations and radiological imaging, response to treatment, number of relapses, follow-up functional status and cause of death. To assess the effect of treatment, we used the treating physician s global opinion about the response to therapy. The modified Rankin scale [14] was used to evaluate the functional status at presentation and at the last visit. It is a standardized and commonly used method of evaluating stroke victims by measuring disability or dependence in activities of daily living. The scale consists of seven grades: 0 indicates no signs or symptoms of disability; 1 indicates no significant disability, despite symptoms; 2 indicates slight disability; 3 indicates Submitted 19 October 2007; revised version accepted 11 July Correspondence to: R. D. Brown Jr, Department of Neurology, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA. brown@mayo.edu Present address: C. Salvarani, Unitá Operativa di Reumatologia, Arcispedale S. Maria Nuova, Reggio Emilia, Italy ß The Author Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org

2 1672 C. Salvarani et al. moderate disability; 4 indicates moderately severe disability; 5 indicates severe disability; and 6 indicates death. CNS tissue was examined histologically in 49 of the 101 cases (D.V. Miller, C. Salvarani, G.G. Hunder et al., unpublished data). Vasculitis was identified in 31 of 49 (63%) histologically examined specimens. In addition to routine staining, specimens were examined for amyloid deposits by conventional staining methods and immunostained for A peptide (immunoperoxidase stain for A4 amyloid: clone 6F3D, 1 : 20 dilution with 88% formic acid pre-treatment; Novocastra Laboratories Ltd., Newcastle upon Tyne, UK). Statistical analysis Differences among patients with A-related PCNSV and those with PCNSV but no histological evidence of CAA were tested with two-sided Wilcoxon rank sum tests for numeric characteristics and the Fisher s exact test for categorical characteristics. Survival among patients was estimated with the Kaplan Meier method, and a one-sample log-rank test was used to compare survival of patients with A-related PCNSV with survival of patients with PCNSV only. Results Demographic and clinical features A peptide deposition was identified in 9 of the 49 patients (18.4%) whose specimens were histologically examined. In 1 of the 9 with CAA, the biopsy findings for vasculitis were negative (diagnosis was made by angiography). Therefore, 8 of the 31 patients (25.8%) with histologically confirmed PCNSV also showed amyloid deposition that was consistent with CAA. The median age at diagnosis of these 31 patients was 46 yrs (range yrs). At diagnosis, 11 of the 31 patients (35.5%) were >60 yrs and 7 (22.6%) were 65 yrs. The median age at diagnosis of all 101 patients was 47 yrs (range yrs) and 24 (23.8%) were over 60 yrs at diagnosis. Table 1 shows the demographic characteristics and the clinical symptoms of the eight patients with PCNSV and CAA. Six patients were men. The median age at diagnosis was 63 yrs (range yrs). At diagnosis, five were 65 yrs. The time from onset of symptoms to diagnosis was 1 month in four of the eight patients (median time for all patients, 39 days; range days). Focal manifestations and a cognitive disorder were the most common symptoms at presentation. Constitutional symptoms such as fever were present in only one patient. Laboratory investigations Table 2 shows results of cerebrospinal fluid (CSF) examinations and ESRs at diagnosis. Results of CSF examinations were abnormal in all eight patients. Seven had CSF protein levels >70 mg/dl (median 111 mg/dl; range mg/dl). Five patients had an elevated CSF white blood cell count of at least 10 cells/mm 3 (median count 21 cells/mm 3 ; range 1 68 cells/mm 3 ). CSF white blood cells consisted mostly of lymphocytes. ESRs were normal in seven of eight patients. The median rate was 14 mm/h (range 2 34 mm/h). Radiological imaging Table 1 shows the results of MRI, cerebral angiography and magnetic resonance angiography at presentation. Brain MRI examinations were performed without and with contrast enhancement, and abnormalities were observed in all eight patients. Patchy or confluent T 2 -weighted signal abnormalities in the white matter were present in six patients. Contrast-enhanced lesions were observed in five patients. Four patients had leptomeningeal enhancement (three diffuse, one linear), and one had multiple, nodular, parenchymal, contrast-enhanced lesions. Multiple infarcts were observed in three patients. One patient had an intracerebral haemorrhage. Conventional cerebral angiography was performed in five patients, and features of vasculitis were present in two. Brain magnetic resonance angiography was performed in four patients, but none showed evidence of vasculitis. One patient with positive results after conventional cerebral angiography had negative results with brain magnetic resonance angiography. Biopsy results The 31 patients with histologically proven PCNSV had the following histological patterns: 18 (58.1%) had a granulomatous inflammatory pattern, 8 (25.8%) had a lymphocytic pattern and 5 (16.1%) had an acute necrotizing pattern. Table 2 shows the results of pathology examinations in the eight with histologically confirmed PCNSV and vascular deposits of A that were consistent with CAA. A granulomatous vasculitis histological pattern was present in all eight patients (Fig. 1). The vasculitis was characterized by transmural mononuclear inflammation with well-formed granulomas and multinucleated giant cells that resulted in vessel wall destruction. In six patients, leptomeningeal and parenchymal involvement was observed; the other two had only leptomeningeal involvement. Infarcts were present in two patients. Treatment and outcome Table 3 shows details of treatment, follow-up MRI examinations and patient status at the last visit. The median duration of followup was 24 months (range months). All patients were treated with prednisone. The median initial dosage of oral prednisone was 60 mg/day (range mg/day). In three patients (Cases 2, 3 and 8), oral prednisone was preceded by a course of intravenous methylprednisolone pulse therapy (two patients received 1 g/day for 3 days; 1 patient received 500 mg/day for 9 days). The median duration of oral prednisone therapy was 6 months (range 1 19 months). Three patients (Cases 2, 4 and 7) initially were treated with corticosteroids only, and five were treated with prednisone and cyclophosphamide (Cases 1, 3, 5, 6 and 8). One patient (Case 6) received monthly pulse intravenous injections of cyclophosphamide, and four patients (Cases 1, 3, 5 and 8) received daily oral doses of cyclophosphamide. The median duration of treatment with cyclophosphamide was 12 months (range 1 17 months). Two patients had a relapse. Case 4 was treated with oral prednisone (80 mg/day) and improved rapidly. One month later, her neurological examination findings were normal. Eleven months later, the prednisone therapy was discontinued. Six years later, she had a recurrence that was characterized by headache, confusion and expressive aphasia. At the time of recurrence, an MRI examination of the brain showed reappearance of contrast-enhanced meningeal lesions. She was treated again with prednisone (60 mg/day) and was in complete remission 1 month later (neurological and MRI findings were normal). Case 5 initially was treated with prednisone and cyclophosphamide, and he showed marked improvement in the neurological status during an examination 3 months later. Four months later, prednisone therapy was suspended; 5 months later, his neurological examination showed normal findings, and cyclophosphamide was discontinued at that time. Fourteen months later, he presented with an increased level of confusion. Axial CT images (without contrast enhancement) showed a right-sided frontal intracranial haemorrhage that was considered secondary to CAA. The haematoma was evacuated by craniotomy, and he had a complete neurological recovery. Immunosuppressive therapy was not initiated. Ten months later, his neurological examination had normal findings, and he had no symptoms. At that time, an MRI examination of the brain showed nearly complete resolution of confluent T 2 -weighted signal abnormalities in the white matter of both cerebral hemispheres. Multiple small deposits of haemosiderin within the

3 TABLE 1. Clinical characteristics and neuroimaging findings at presentation Case Age (yrs) Sex Clinical findings Brain MRI 1 66 F Focal TIAs Haematoma in the inferolateral aspect of the right frontal lobe, multiple zones of ischaemia and secondary haemorrhage of the subcortical white matter of the left frontal lobe and left paracentral gyrus 2 75 M Subacute dementia, personality change, focal TIAs, aphasia 3 42 M Headache, paraparesis, sudden change in level of consciousness, aphasia, headache, right homonymous hemianopsia, fever, vomiting 4 63 F Confusion, cognitive decline, headache, aphasia, visual field defect Subcortical infarcts, foci of nodular enhancement present either within the sulci or adjacent cortex T 2 W signal hyperintensity in the sulci of the left frontal, bilateral temporal, parietal and occipital regions. T 1 W signal hyperintensity in the left frontal lope, generalilzed prominence of the vessels within abnormal sulci in both hemispheres in contrast-enhanced, T 1 W images Multiple infarcts involving both cerebral hemispheres and cerebellum Diffuse leptomeningeal enhancement in the left cerebral hemisphere and right frontal and temporal regions 5 60 M Aphasia with alexia T 2 W signal abnormality involving the subcortical white matter of both cerebral hemispheres, diffuse leptomeningeal enhancement (greater in the right hemisphere) 6 64 M Confusion, cognitive decline, personality change, headache, ataxia 7 57 M Confusion, cognitive decline, amnestic syndrome, aphasia, headache, papilloedema 8 84 M Drowsiness, confusion, cognitive decline, agitated behaviour F: female; M: male; MRA: magnetic resonance angiography; TIA: transient ischaemic attack; W: weighted. Diffuse bilateral leptomeningeal enhancement involving the cerebrum and the cerebellum, multiple infarcts, patchy T 2 W white matter signal abnormality Confluent T 2 W signal abnormality of the white matter, bilateral linear leptomeningeal enhancement of the cerebral hemispheres, basal ganglia, and pons Extensive confluent T 2 W signal abnormality throughout the frontal and temporal lobes Cerebral angiography Subtle areas of bilateral vascular narrowing affecting small branch vessels of the anterior and middle cerebral arteries Brain MRA Normal Time between symptom onset and diagnosis 2.5 months Normal Not done 2 months Alternating areas of vasoconstriction and normalcalibre vessels, involving numerous large and medium-sized cerebral arteries Not done 21 days Normal Normal 17 days Not done Not done 16 days Normal Not done 1.5 months Not done Normal 14 months Not done Normal 1 month PCNSV with cerebral amyloid angiopathy 1673 Downloaded from at Pennsylvania State University on February 27, 2014

4 1674 C. Salvarani et al. TABLE 2. Laboratory and pathology findings CSF Case WBC count (cells/mm 3 ) Protein (mg/dl) ESR (mm/h) Pathology findings a 1 1 (50% lymphocytes, 26% monocytes, 24% neutrophils) Granulomatous, leptomeningeal and intraparenchymal inflammation 2 20 (95% lymphocytes, 5% monocytes) Granulomatous and leptomeningeal marked angiocentric inflammation 3 68 (78% lymphocytes, 22% monocytes) 77 8 First biopsy: multiple infarcts Second biopsy: granulomatous and leptomeningeal marked angiocentric information; infarcts 4 68 (79% lymphocytes, 17% monocytes, 4% neutrophils) Granulomatous, leptomeningeal and intraparenchymal marked angiocentric inflammation 5 2 (69% neutrophils, 30% lymphocytes) First biopsy: granulomatous, leptomeningeal and intraparenchymal marked angiocentric inflammation Second biopsy: same findings as the first biopsy 6 23 (90% lymphocytes, 10% monocytes) 97 2 Granulomatous, leptomeningeal and intraparenchymal marked angiocentric inflammation; infarcts 7 61 (96% lymphocytes, 4% monocytes) Granulomatous, leptomeningeal and intraparenchymal marked angiocentric inflammation 8 2 (84% lymphocytes, 16% monocytes) Granulomatous, leptomeningeal and intraparenchymal marked angiocentric inflammation All biopsy specimens showed Ab peptide deposits. a All patients underwent an open brain biopsy. CSF: cerebrospinal fluid; WBC: white blood cell. subcortical and deep white matter of both cerebral hemispheres and the left cerebellum were observed. Altogether, six of eight patients responded well to therapy. Response was rapid and generally occurred in the first 2 3 weeks. A rapid response to therapy was also noted in the patients with a recurrence, as described above. Follow-up MRI studies were available in six patients. Gadolinium-enhanced lesions were not detectable after therapy. T 2 -weighted signal abnormalities in the white matter also showed nearly complete resolution. The outcome was poor for two patients. One patient (Case 3) died after having multiple strokes 1.5 months after the initial presentation. Another patient (Case 2) had a progressive cognitive decline with serious disability at the last follow-up. The median modified Rankin disability score at the last visit was 1 (range 0 6); it was lower than the score at presentation (median 2; range 0 5). Comparison of patients with and without CAA Table 4 compares the clinical features, treatment, pathology findings and MRI findings of the eight patients with histological evidence of PCNSV and CAA, and the 40 patients with biopsies that did not show CAA. The median age of patients with CAA was significantly higher at diagnosis (P ¼ 0.002), and gadoliniumenhanced meningeal lesions occurred more frequently (P ¼ 0.05). More patients with CAA showed a granulomatous histopathological pattern (P ¼ 0.01). Other differences in those with CAA that were not statistically significant included a greater number of men, more acute onset, shorter median period between onset of symptoms to diagnosis, a higher frequency of cognitive disorders at presentation, and fewer relapses or recurrences. More were treated initially with cyclophosphamide or AZA and less frequently with prednisone only. Other similarities between the groups included the Rankin disability score at last follow-up, the therapy duration, and the frequency of patients who had discontinued therapy by the last follow-up visit. Figure 2 shows the similar survival curves of patients with and without CAA (P ¼ 0.60). Discussion This study was part of an analysis of 101 patients with PCNSV who were seen at our institution over a 21-yr period [4]. We noted that 8 of the 31 patients who had biopsy specimens showing vasculitis also had vascular deposits of A peptide. To determine whether these eight patients had different clinical characteristics, we compared them to 40 other patients with histological specimens that did not show evidence of CAA. The results indicate that PCNSV is a heterogeneous condition and suggest that the eight cases with CAA may represent a distinct subgroup in the clinical spectrum of PCNSV. When compared with the patients with PCNSV only, the eight patients with vascular deposits of A were older at diagnosis, had an increased frequency of gadolinium-enhanced lesions with MRI and had granulomatous histological patterns in the CNS biopsy specimens. Other differences were also noted but were not statistically significant, perhaps because of small numbers. The extent of disability, survival rates and frequency of suspending therapy during follow-up were similar in the patients with and without evidence of CAA. No differences in laboratory test findings were observed among the two groups, and CSF findings were generally mild and non-specific. MRI findings were abnormal in all eight cases of PCNSV with CAA. Findings included infarctions, haemorrhage and gadolinium-enhanced lesions. To further evaluate the possible clinical significance of prominent gadolinium enhancement of the leptomeninges, we have performed a separate analysis of all patients with this finding from the overall cohort of 101 patients [15]. Intracerebral haemorrhage occurred with similar frequencies in both groups.

5 PCNSV with cerebral amyloid angiopathy 1675 FIG. 1. Leptomeningeal vasculitis with amyloid angiopathy. (A) Destructive vasculitis with well-formed granulomas involves the small vessels. Others show wall thickening with eosinophilic material (arrows). (Haematoxylin and eosin; 100.) (B) All vessels show Ab deposits. (Immunoperoxidase staining for ba4 amyloid; 100.) TABLE 3. Treatment, outcome and duration of follow-up Case Treatment Follow-up MRI 1 Oral prednisone (initial dosage, 60 mg/day) for 11 months Oral cyclophosphamide (100 mg/day) for 17 months 2 IV methylprednisolone (1 g/day) for 3 days, then oral prednisone (initial dosage, 60 mg/day) for 4 months 3 IV methylprednisolone (500 mg/day) for 9 days, then oral prednisone (initial dosage, 150 mg/day) for 1 month Oral cyclophosphamide (150 mg/day) for 1 month 4 Oral prednisone (initial dosage, 80 mg/day) for 13 months 5 Oral prednisone (initial dosage, 60 mg/day) for 7 months Oral cyclophosphamide (150 mg/day) for 12 months 6 Oral prednisone (initial dosage, 30 mg/day) for 2 months IV cyclophosphamide (1.7 g/months) for 14 months 7 Oral prednisone (initial dosage, 60 mg/day) for 19 months 8 IV methylprednisolone (1 g/day) for 3 days, then oral prednisone (initial dosage, 60 mg/day) for 5 months Oral cyclophosphamide (150 mg/day) for 4 months Status at last follow-up (modified Rankin score a ) Follow-up period b (months) Partially resolved haematoma, enhancement nearly completely resolved, T 2 W signal abnormalities involving left frontal lobe and left paracentral gyrus nearly completely resolved Improved (0) 112 Not available Stable (3) 7 New ischaemic lesions, particularly involving the right frontal lobe and left occipital lobe Resolution of leptomeningeal enhancement Subacute intracerebral haemorrhage in the right frontal lobe, decreased T 2 W signal abnormality in the white matter of both cerebral hemispheres, multiple small deposits of hemosiderin in the white matter of both cerebral hemispheres and the left cerebellum Resolution of leptomeningeal enhancement, no new infarcts Stroke-related fatality (6) 1.5 Improved (1) 87 Improved (0) 37 Improved (1) 15 Resolution of initial changes Improved (1) 33 Not available Improved (0) 6 a The modified Rankin score is defined as follows: 0: no signs or symptoms of disability; 1: no significant disability, despite symptoms; 2: slight disability; 3: moderate disability; 4: moderately severe disability; 5: severe disability; 6: death. b From the onset of symptoms. W, weighted. Like the findings of Scolding et al. [13], our data indicate that the clinical manifestations of PCNSV with CAA more closely resemble those of PCNSV only and are less similar to those of CAA only. Although impaired cognition has been associated with CAA [7, 8], the acute onset, radiological findings and response to immunosuppressive therapy observed in our patients are not typical of sporadic CAA. This suggests that vascular inflammation has a major influence in determining the disease manifestations. The absence of an increased frequency of intracerebral haemorrhage, which frequently is observed in patients with CAA only [5, 6], suggests a less critical role for vascular A deposition in the pathogenesis of this syndrome. Because reports of A-related PCNSV in the medical literature consist mainly of individual cases, the frequency of vascular deposits of A in PCNSV is uncertain. Our data, derived from the largest cohort reported to date of PCNSV patients from a single institution, showed A peptide deposition in 9 of the 49 patients (18.4%) whose specimens were histologically examined and in 8 of the 31 patients (25.8%) who had biopsy-proven PCNSV. CAA without vasculitis is common in the elderly, and its prevalence increases with age. Evidence of CAA has been identified in 45% of autopsies performed in individuals aged yrs, but it was observed in only % of autopsies in individuals aged yrs [16 19]. At diagnosis, 24 of our 31 patients with

6 1676 C. Salvarani et al. TABLE 4. Findings of patients with primary CNS vasculitis, with or without CAA (n ¼ 48) a histological evidence of PCNSV and five of the eight patients with CAA and PCNSV were younger than 65 yrs. Therefore, the proportion of patients with CAA who were younger than 65 yrs at diagnosis was greater than that observed in general autopsy studies as above (5 of 24; 21%). The median age at diagnosis in our series of patients with PCNSV was 47 yrs [4]. The patients with PCNSV and CAA had an older age at diagnosis (median 63 yrs), but this was considerably lower than that reported for patients with sporadic CAA. The prevalence of CAA observed in our series of PCNSV patients was much higher than we expected, given the age of the patients. This finding strongly argues against the possibility that the association of these two conditions is coincidental. Patients with CAA (n ¼ 8) Patients without CAA (n ¼ 40) P value Males, n (%) 6 (75.0) 19 (47.5) 0.25 Age at diagnosis, yrs b 63 (42 84) 42.5 (25 79) Time from symptom onset to diagnosis, days b 39 (16 426) 74 (2 1913) 0.48 Headache, n (%) 4 (50.0) 28 (70.0) 0.41 Cognitive disorder, n (%) 6 (75.0) 23 (57.5) 0.45 Focal manifestations, n (%) 7 (87.5) 39 (97.5) 0.31 Systemic manifestations, n (%) c 1 (12.5) 7 (17.5) >0.99 CSF abnormality, n (%) Protein 45 mg/dl or WBC count 5/mm 3 8 (100.0) 26 (81.3) d 0.32 Protein 70 mg/dl or WBC count 10/mm 3 7 (87.5) 22 (68.8) d 0.41 ESR, mm/h b 14.5 ( ) 7.0 ( ) 0.50 Modified Rankin disability score at last follow-up 0 3, n (%) 6 (75.0) 32 (80.0) , n (%) 2 (25.0) 8 (20.0) Patients without relapse or recurrence, n (%) 7 (87.5) 25 (62.5) 0.24 Initial treatment, n (%) Prednisone only 2 (25) 23 (58) 0.12 e Prednisone and cyclophosphamide (oral or intravenous pulse) 5 (63) 13 (33) Prednisone and AZA 0 (0) 1 (3) Prednisone and other therapy 1 (13) 1 (3) Other therapy 0 (0) 1 (3) No therapy 0 (0) 1 (3) Duration of therapy, months b Prednisone 6 (1 19) 9 (1 107) 0.46 Cyclophosphamide 8 (1 17) 6 (1 20) >0.99 Patients not requiring therapy at last follow-up, n (%) 2 (25) 9 (23) f >0.99 Brain biopsy specimen pattern, n (%) Lymphocytic 0 (0) 8 (38) g Granulomatous 8 (100) 9 (43) g 0.01 Necrotizing 0 (0) 4 (19) g MRI findings at presentation, n (%) Minimal or age-consistent white matter changes 0 (0) 1 (3) h >0.99 Any infarct present 3 (38) 15 (44) h >0.99 Intracerebral haemorrhage 1 (13) 2 (6) h 0.48 Gadolinium-enhanced lesions (all) 5 (63) 16 (47) h 0.70 Gadolinium-enhanced intracerebral lesions 1 (13) 11 (32) h 0.40 Gadolinium-enhanced meningeal lesions 4 (50) 5 (15) h 0.05 a All patients underwent a CNS biopsy. We excluded one patient with negative biopsy but with A peptide vascular deposition. b Data are shown as median (range). c Presence of at least one of the following systemic manifestations: fatigue, anorexia, weight loss or fever. d Eight patients had protein levels <45 mg/dl (n ¼ 32). e Comparison was between patients receiving any immunosuppressive therapy vs patients receiving prednisone only. f One patient without cerebral amyloid angiopathy had no therapy (n ¼ 39). g Nineteen patients did not undergo a brain biopsy (n ¼ 21). h Six patients did not undergo MRI (n ¼ 34). FIG. 2. Survival of patients with PCNSV, with and without CAA. Both groups had similar survival curves (P ¼ 0.60). Vascular A peptide deposition may trigger a variable inflammatory response. Yamada et al. [20] examined brain specimens of patients with CAA and reported an increased number of perivascular monocytes. However, in most cases, it was so slight that it could be recognized only by immunohistochemical analysis. In a report of 42 cases of pathologically diagnosed CAA, Eng et al. [21] noted seven patients with perivascular inflammation and multinucleated giant cells. This group of seven patients appeared to be pathologically different from our patients, who presented destructive vascular inflammation, and from other patients in published case reports of A-related PCNSV. However, Eng and coworkers patients did have a clinical syndrome that resembled PCNSV more than it resembled noninflammatory CAA. Their patients were younger at presentation when compared with patients with CAA only, the clinical symptoms were subacute cognitive decline or seizure rather than haemorrhagic stroke, and the clinical symptoms and radiological findings improved with immunosuppressive treatment. A variable inflammatory response may account for the clinical features observed in patients with CAA-related PCNSV and CAA-related perivascular inflammation. The inflammatory response to vascular amyloid observed in a transgenic mouse model that develops prominent CAA [22] supports amyloid deposition as the likely trigger of vascular or perivascular inflammation. The strength of this study was the identification of patients with A-related PCNSV from the largest group of consecutive patients reported to date from a single institution. In this cohort, case ascertainment was well defined and uniform, and clinical followup data were available. The retrospective nature of the study and potential selection bias are limitations. By restricting the

7 PCNSV with cerebral amyloid angiopathy 1677 present study to cases with pathology samples, we also may have introduced a potential bias towards selection of patients with atypical clinical presentations or more severe disease (such patients may have been more likely to undergo a biopsy). Our study did not elucidate the pathogenesis of PCNSV or CAA. To summarize, we described a group of patients with PCNSV and CAA. The diagnosis was confirmed by brain biopsy specimens that showed granulomatous vasculitis and vascular deposits of A peptide. When compared with patients who had PCNSV only, those with PCNSV and CAA were older at presentation but younger than patients with CAA and no inflammation. They showed a trend of more acute clinical onset and higher frequency of cognitive dysfunction. MRI findings showed contrast-enhanced meningeal lesions more frequently, and biopsy specimens typically showed granulomatous vasculitis. The eight patients tended to have a monophasic disease course and generally responded well to immunosuppressive treatment. Rheumatology key messages The combination of PCNSV and CAA may be more common than previously thought. When these conditions occur together, the course and outcome resembles vasculitis to a greater degree than isolated amyloid angiopathy. Acknowledgements Editing, proofreading, and reference verification were provided by the Section of Scientific Publications, Mayo Clinic. Disclosure statement: G.H. is a member of a speaker s bureau for Merck and Abbott. All other authors have declared no conflicts of interest. References 1 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;67: Calabrese LH, Duna GF, Lie JT. Vasculitis in the central nervous system. Arthritis Rheum 1997;40: Moore PM. Diagnosis and management of isolated angiitis of the central nervous system. Neurology 1989;39: Salvarani C, Brown RD Jr, Calamia KT et al. Primary central nervous system vasculitis: analysis of 101 patients. Ann Neurol 2007;62: Greenberg SM. Cerebral amyloid angiopathy: prospects for clinical diagnosis and treatment. Neurology 1998;51: Vinters HV. Cerebral amyloid angiopathy: a critical review. Stroke 1987;18: Greenberg SM, Vonsattel JP, Stakes JW, Gruber M, Finklestein SP. The clinical spectrum of cerebral amyloid angiopathy: presentations without lobar hemorrhage. Neurology 1993;43: Greenberg SM, Gurol ME, Rosand J, Smith EE. Amyloid angiopathy-related vascular cognitive impairment. Stroke 2004;35(Suppl. I): Greenberg SM, Vonsattel J-PG. Diagnosis of cerebral amyloid angiopathy: sensitivity and specificity of cortical biopsy. Stroke 1997;28: Ginsberg L, Geddes J, Valentine A. Amyloid angiopathy and granulomatous angiitis of the central nervous system: a case responding to corticosteroid treatment. J Neurol 1988;235: Fountain NB, Eberhard DA. Primary angiitis of the central nervous system associated with cerebral amyloid angiopathy: report of two cases and review of the literature. Neurology 1996;46: Schwab P, Lidov HGW, Schwartz RB, Anderson RJ. Cerebral amyloid angiopathy associated with primary angiitis of the central nervous system: report of 2 cases and review of the literature. Arthritis Rheum 2003;49: Scolding NJ, Joseph F, Kirby PA et al. A-related angiitis: primary angiitis of the central nervous system associated with cerebral amyloid angiopathy. Brain 2005;128: Sulter G, Steen C, De Keyser J. Use of the Barthel index and modified Rankin scale in acute stroke trials. Stroke 1999;30: Salvarani C, Brown RD Jr, Calamia KT et al. Primary central nervous system vasculitis with prominent leptomeningeal enhancement: a subset with benign outcome. Arthritis Rheum 2008;58: Vinters HV, Gilbert JJ. Cerebral amyloid angiopathy: incidence and complications in the aging brain. II. The distribution of amyloid vascular changes. Stroke 1983; 14: Tomonaga M. Cerebral amyloid angiopathy in the elderly. J Am Geriatr Soc 1981;29: Masuda J, Tanaka K, Ueda K, Omae T. Autopsy study of incidence and distribution of cerebral amyloid angiopathy in Hisayama, Japan. Stroke 1988;19: Neuropathology Group of the Medical Research Council Cognitive Function and Ageing Study (MRC CFAS). Pathological correlates of late-onset dementia in a multicentre, community-based population in England and Wales. Lancet 2001; 357: Yamada M, Itoh Y, Shintaku M et al. Immune reactions associated with cerebral amyloid angiopathy. Stroke 1996;27: Eng JA, Frosch MP, Choi K, Rebeck GW, Greenberg SM. Clinical manifestations of cerebral amyloid angiopathy-related inflammation. Ann Neurol 2004;55: Winkler DT, Bondolfi L, Herzig MC et al. Spontaneous hemorrhagic stroke in a mouse model of cerebral amyloid angiopathy. J Neurosci 2001;21: