Treatment of small vessel primary CNS vasculitis in children: an open-label cohort study

Transcription

1 Treatment of small vessel primary CNS vasculitis in children: an open-label cohort study Clare Hutchinson, Jorina Elbers, William Halliday, Helen Branson, Suzanne Laughlin, Derek Armstrong, Cynthia Hawkins, Robyn Westmacott, Susanne M Benseler Summary Background There is no treatment protocol or standardised documentation of neurological outcome for patients with small vessel childhood primary angiitis of the CNS, a rare inflammatory brain disease. We aimed to assess a treatment regimen and describe long-term neurological outcomes in a cohort of children with this disorder. Methods We did a single-centre open-label cohort study in children with small vessel childhood primary angiitis of the CNS who were less than 18 years old at diagnosis. The treatment protocol consisted of induction therapy with steroids and pulses of intravenous cyclophosphamide followed by maintenance therapy with either azathioprine or mycophenolate mofetil. Clinical and neurological assessments, quality of life measures, and laboratory markers were done at baseline, 3, 6, 9, 12, 18, and 24 months, and every year thereafter. Brain imaging was done at baseline, 6, 12, 18, and 24 months. The primary outcome was the paediatric stroke outcome measure (PSOM) score at 24 months. Findings From January, 2002, to December, 2009, 127 patients were enrolled, 19 of whom met the inclusion criteria and were given induction therapy. Median age at diagnosis was 9 8 years (range ) and median follow-up was 33 months (range 1 86). 14 patients completed induction and received maintenance therapy with azathioprine (n=9) or mycophenolate mofetil (n=5). 13 patients completed 24 months follow-up, nine of whom had a good neurological outcome by PSOM. Eight of 19 patients experienced disease flares. Four patients achieved remission of disease off medication. Interpretation This treatment protocol of immunosuppressive therapy may improve long-term neurological outcome in children with small vessel childhood primary angiitis of the CNS. Identification and appropriate diagnosis of children with the disorder is crucial because with standardised treatment good neurological outcome is a realistic goal. Funding None. Introduction Childhood primary angiitis of the CNS is a recently recognised rare inflammatory disease that causes severe neurological deficits and unexplained neurological deterioration in previously healthy children. Children with the disorder present with a range of neurological symptoms including intractable seizures, hemiparesis, cranial nerve deficits, severe cognitive deficits, and decreased consciousness. 1 3 There are two types of childhood primary angiitis of the CNS: medium-large vessel and small vessel vasculitis. 2 4 Medium-large vessel disease affects arteries that are large enough to be differentiated by conventional angiography. In patients with small vessel childhood primary angiitis of the CNS, angiography findings are typically negative and thus diagnosis must be confirmed by brain biopsy. Small vessel childhood primary angiitis of the CNS has been described in a small number of case reports and case series, with little information known about its incidence and age distribution. 1,2,5 7 Treatments have included corticosteroids, methotrexate, intravenous gammaglobulin, oral or intravenous cyclophosphamide, and azathioprine, all at varying doses and for different durations. The length of follow-up in these cases was variable and there was no standardised method of assessing clinical outcome. Neurological outcome in patients with small vessel childhood primary angiitis of the CNS can be devastating and resulted in death secondary to refractory status epilepticus in one child whose diagnosis was confirmed on autopsy. 7 However, some children with small vessel disease have shown neurological recovery after immunosuppressive treatment, suggesting that the neurological deficits caused by brain inflammation are reversible. 1,2,5 This recovery contrasts with the irreversible damage caused by acute ischaemia in paediatric patients with medium-large vessel disease. 3 We aimed to describe a cohort of children with small vessel childhood primary angiitis of the CNS, report the efficacy and safety of a treatment regimen, and describe long-term neurological outcomes. Methods Patients We did an open-label cohort study at The Hospital for Sick Children, ON, Canada, a tertiary care centre, from January, 2002, to December, Patients with childhood primary angiitis of the CNS who were less than 18 years old at Published Online October 4, 2010 DOI: /S (10)70243-X See Online/Reflection and Reaction DOI: /S (10) Division of Rheumatology, Department of Pediatrics (C Hutchinson MDCM, S M Benseler MD), Division of Neurology, Department of Pediatrics (J Elbers MD), Department of Pediatric Laboratory Medicine (Prof W Halliday MD, C Hawkins MD), Department of Diagnostic Imaging (H Branson MD, S Laughlin MD, D Armstrong MD), and Department of Psychology (R Westmacott PhD), The Hospital for Sick Children, Toronto, ON, Canada Correspondence to: Dr Susanne M Benseler, Division of Rheumatology, Department of Pediatrics, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario M5G 1X8, Canada susanne.benseler@sickkids.ca Published online October 4, 2010 DOI: /S (10)70243-X 1

2 9 had azathioprine maintenance therapy 7 switched to mycophenolate mofetil 5 had azathioprine treatment failure 2 had azathioprine treatment intolerance Figure 1: Study profile diagnosis were included. Childhood primary angiitis of the CNS was defined according to Calabrese criteria: 8 a newly acquired neurological deficit plus angiographic or Panel 1: Treatment protocol Induction therapy: 6 months Seven pulses of mg/m² intravenous cyclophosphamide every 4 weeks with co-trimoxazole prophylaxis 2 mg/kg prednisone daily (maximum 60 mg daily) weaned every 4 weeks to 50, 40, 30, 25, and 20 mg daily and then by 2 5 mg every 4 weeks until completed. Supplementary calcium and vitamin D also given during prednisone treatment Anticonvulsants and antipsychotics as needed Maintenance therapy: 18 months mg/m² mycophenolate mofetil per day (maximum 2000 mg daily) or 2 3 mg/kg azathioprine per day (maximum 150 mg daily) Prednisone weaned every 4 weeks. Supplementary calcium and vitamin D also given during prednisone treatment Anticonvulsants and antipsychotics as needed 27 had negative angiography 24 had confirmatory biopsy 19 started induction therapy 14 completed induction therapy 127 patients enrolled 3 did not have biopsy 5 did not receive protocol treatment 2 died before treatment 2 lost to follow-up 1 had different treatment regimen (4 years of prednisone and 6 years of azathioprine) 2 failed induction 3 induction ongoing 5 had mycophenolate mofetil maintenance therapy 100 had positive angiography histological features consistent with CNS vasculitis and absence of evidence of an underlying condition that could explain these findings. Exclusion criteria were diagnosis of medium-large vessel childhood primary angiitis of the CNS by abnormal CNS angiography, absence of brain biopsy to confirm the diagnosis, treatment initiation that did not conform to the treatment protocol, and age less than 1 month old. Patients who required additional immunosuppression that was not detailed in the treatment protocol were included in the study analysis. Children and their parents or legal guardians provided written informed consent at the time of diagnosis. Ethics approval was provided by The Hospital for Sick Children local research ethics board. Procedures Patient information, including sex, age at diagnosis, and duration of symptoms before diagnosis were obtained at baseline. Standardised assessments, including clinical and neurological examination, quality of life measures, and laboratory markers were done in the CNS vasculitis clinic at baseline, 3, 6, 9, 12, 18, and 24 months, and yearly thereafter until patients reached 18 years of age. Brain imaging was done at baseline, 6, 12, 18, and 24 months. Clinical neurological examinations included assessments of consciousness, behaviour, mental status, language deficits, cranial nerve function, motor skills, tendon reflexes, fine motor coordination, sensory function, gait, and occurrence of seizures and electroencephalography (EEG) findings. Presence of systemic symptoms including fever, weight loss or fatigue, and manifestations of rheumatic disease such as arthritis or rash were documented on clinical examination. Laboratory values, including complete blood count; erythrocyte sedimentation rate; concentrations of c-reactive protein, complement C3, von Willebrand factor antigen, and IgG; and presence of lupus anticoagulant, antinuclear antibody, antineutrophil cytoplasmic antibody, and anticardiolipin antibody, were taken at diagnosis and at follow-up visits. We did a standardised series of infectious assessments including cerebrospinal fluid (CSF) bacterial, viral cell, and fungal culture; Mycobacterium tuberculosis culture and direct amplification; PCR for herpes viruses, Mycoplasma pneumoniae, enterovirus, West Nile virus, Japanese encephalitis, and St Louis encephalitis; and serum serology for herpes viruses, Borrelia burgdorferi, mumps, measles, parvovirus B19, influenza viruses, arboviruses, Bartonella henselae, and Mycoplasma pneumoniae. Additional CSF findings included opening pressure, cell count, protein concentration, glucose concentration, presence of oligoclonal banding, IgG synthesis, and IgG index. We reviewed radiographic data including CT scan; T1-weighted, T2-weighted, fluid attenuated inversion recovery, diffusion-weighted, and gadolinium-enhanced MRI; magnetic resonance angiography and venography; and conventional angiography. Lesion location, 2 Published online October 4, 2010 DOI: /S (10)70243-X

3 gadolinium enhancement, diffusion-restriction, and meningeal enhancement were recorded. Brain biopsies were taken at the time of diagnosis from MRI-detected lesions when the area involved was accessible and was not an eloquent region of the brain. Non-lesional biopsies were taken from the non-dominant frontal lobe. All specimens included the meninges and grey and white matter. We examined brain biopsy samples with haematoxylin and eosin staining; immunohistochemistry for lymphocytes, polymorphonuclear cells, eosinophils, and macrophages; electronmicroscopy; and did infectious studies including viral inclusions, bacterial culture, fungal culture, silver stain, and acid-fast bacillus smear and culture. Brain biopsies were reviewed for adequacy of specimen sampling, presence of intramural or perivascular cellular infiltrate, localisation of involved vessels, identification of inflammatory cell type, and associated findings such as demyelination, infarction, and gliosis. We reviewed the paediatric quality of life inventory (PedsQL) 9 child self-report and parent proxy report from the final follow-up. The PedsQL is a quality of life measure that includes 23 items covering four domains physical, emotional, social, and school functioning. Physical health summary score, psychosocial health summary score, and an overall total score are calculated and reported on a scale from 0 (worst health) to 100 (best health). Neurocognitive testing was done using age-appropriate Wechsler scales of intellectual ability the Wechsler Intelligence Scale for Children, 4th edition (WISC-IV) 10 for children 6 16 years old and the Wechsler Adult Intelligence Scale, 3rd edition (WAIS-III) 11 for children over 16 years of age. We examined the following domains: full-scale IQ, verbal comprehension, perceptual reasoning, working memory, and processing speed. Abnormal scores were defined as any index score that was one or more SDs below the mean of the normative sample. Neurological outcome was documented using the paediatric stroke outcome measure (PSOM), which has been validated as a measure of neurological outcome in children with stroke. 12,13 The PSOM consists of five different spheres: sensorimotor deficits on the right and left, deficits in language production and comprehension, and deficits in cognition or behaviour. Every domain is assigned a score between 0 (normal) and 2 (severe deficit). An overall deficit score is assigned on the basis of the number and severity of affected domains, as described by deveber and colleagues, 13 Patients are categorised as having normal neurological functioning or mild (not interfering with function), moderate (decreased function), or severe (loss of function) neurological deficit. Patients received a treatment protocol that was based in part on previous reports of treatment of childhood primary angiitis of the CNS and that consisted of 6 months of induction therapy with steroids and pulses of intravenous cyclophosphamide followed by 18 months of maintenance therapy with either azathioprine or mycophenolate mofetil at the discretion of the treating physician (panel 1). 4,14 In patients on azathioprine, genetic testing for the thiopurine S-methyltransferase gene was done to look for polymorphisms associated with altered metabolism and potential toxicity. Hepatic enzymes were measured regularly to monitor for azathioprine toxicity. Mycophenolate mofetil kinetics were assessed in patients receiving this drug. The primary outcome was the PSOM score at 24 months. Neurological outcome was categorised as either good (normal functioning or mild deficit) or poor (moderate or severe deficit). 13 Secondary outcomes were categorised into treatment efficacy and safety. Treatment efficacy secondary outcomes were PSOM score at 12 months and final follow-up visit, presence of disease flare, and time to discontinuation of immunosuppressive drugs. Remission was defined as complete absence of disease activity in clinical symptoms, examination findings, laboratory markers, and imaging for at least 3 months. Treatment safety secondary outcomes were mortality; serious infection requiring hospital admission; and presence of cataracts, avascular necrosis, vertebral fractures, or type 2 diabetes mellitus. Statistical analysis We did descriptive statistics on all independent variables. Primary and secondary outcomes were measured as the number of patients reaching the outcome. Role of the funding source There was no funding source for this study. All authors had full access to all the data in the study and SMB had final responsibility for the decision to submit for publication. Results From January, 2002, to December, 2009, 127 patients were enrolled (figure 1), 19 of whom met the inclusion criteria. Most patients were girls, the median age at diagnosis was about 10 years, and the median symptom duration before diagnosis was about 6 weeks (table 1). Patients were followed up for a median of 33 months (range 1 86). Age at diagnosis (years) 9 8 ( ) Girls 15 Duration of symptoms before diagnosis (weeks) 6 (1 208) Biopsy specimen Lesional 12/17 Non-lesional 5/17 Biopsy histology Lymphocytic vasculitis 18/18 Granulomatous changes 0/18 Data are median (range), number, or n/n. Table 1: Patient characteristics at baseline Published online October 4, 2010 DOI: /S (10)70243-X 3

4 All 19 patients had cognitive or behavioural deficits and most patients had sensorimotor deficits (table 2). Headache was reported in most patients, and psychiatric manifestations and meningitis-like presentation were each reported in about 30% of patients. None of the patients had rash, arthritis, or other signs of systemic inflammatory disease. Seizures were present in about 80% of patients. No patients had rapidly changing neurological deficits at presentation, and recovery of neurological function took several months for all patients. EEG was available for detailed review in 13 of 19 patients, and most showed abnormalities. 13 of 17 patients had at least one abnormal inflammatory marker, 14 of 17 had at least one complete blood count abnormality, and 15 of 16 had abnormal lumbar puncture at diagnosis (table 3). No patients had positive antineutrophil cytoplasmic antibody at diagnosis. Antinuclear antibody was negative in most patients and extractable nuclear Cognitive or behavioural deficits Cognitive deficits 19 Personality or behaviour changes 18 Decreased level of consciousness 9 None 0 Sensorimotor deficits Sensory deficits 9 Gross motor deficits 7 Cranial nerve deficits 5 Movement disorder 3 None 7 Language production deficits 10 Language comprehension deficits 6 Other neurological manifestations Headache 17 Psychiatric manifestations 6 Meningitis-like presentation 6 Seizures Generalised 7 Focal 3 Focal with secondary generalisation 2 Status epilepticus 3 None 4 Electroencephalography Abnormal 12/13 Diffuse background slowing 6/13 Focal epileptiform discharges 4/13 Focal slowing 3/13 Generalised epileptiform discharges 2/13 Frontal intermittent rhythmic delta activity 2/13 Periodic lateralising epileptiform discharges 1/13 Electrographic seizures 1/13 Data are number or n/n. Table 2: Clinical features at baseline antigens were negative. Anticardiolipin antibody and lupus anticoagulant were negative in most patients. Microbiology testing did not reveal an infectious cause in any patient. At diagnosis, about half of patients had normal head CT scans but about 90% had abnormal brain MRI (table 4); the patients with normal MRI presented with status epilepticus. MRI lesions showed increased signal on T2-weighted imaging or fluid attenuated inversion recovery in most patients and restriction on diffusionweighted imaging in few. Almost 50% of patients had parenchymal gadolinium enhancement, and about a third had gadolinium enhancement of the meninges. Lesion location was unilateral in about 20% of patients, bilateral symmetric in 5%, and bilateral asymmetric in about 60%. Half of patients had supratentorial involvement alone, around a third had both supratentorial and infratentorial lesions, and 5% had infratentorial involvement alone. The most common areas of involvement were subcortical white matter and cortical grey matter. About 20% of patients had abnormal spinal MRI; abnormalities included increased signal on T2-weighted imaging in intra medullary lesions in the cervical and thoracic spinal cord as well as enhancement of the distal dorsal nerve roots. All patients had normal angiography. At last follow-up, 18 patients had a repeat brain MRI, 16 of which showed residual abnormalities already present on initial studies. At diagnosis, 70% of children had lesions accessible for biopsy and thus had lesional biopsies; the remaining five patients had non-lesional brain biopsies (table 1). All specimens were suitable for histological study. Two patients had biopsies at an outside institution, which were read by an external expert neuropathologist as small vessel childhood primary angiitis of the CNS before transfer to the study site. As a consequence, complete results on biopsy location (two patients) and histology (one patient) were not available. 13 patients had a vascular intramural lymphocytic infiltrate (figure 2); in five the infiltrate was predominantly perivascular. No granulomatous changes were noted. Electronmicroscopy identified endothelial cell activation in seven and tubuloreticular inclusions in three of 12 patients. PedsQL child self-report was available for 17 children at final follow-up: median physical health summary score was 78 (range 0 100), median psychosocial health summary score was 73 (27 100), and median total score was 73 (20 100). PedsQL parent proxy report was available for 15 patients at final follow-up. The median physical health summary score was 66 (range 9 100), median psychosocial health summary score 63 (37 100), and median total score 67 (32 100). Detailed results of WISC-IV and WAIS-III neurocognitive testing were available for review in ten patients. The median time of assessment was 14 5 months from diagnosis (range 2 32 months). All patients had at least one score that was abnormal. Full scale IQ was abnormal in eight of ten patients, working memory in all ten, verbal 4 Published online October 4, 2010 DOI: /S (10)70243-X

5 Inflammatory markers Erythrocyte sedimentation rate (mm/h) 32 (1 117) Erythrocyte sedimentation rate >10 mm/h 11/17 C-reactive protein (mg/l) 21 ( ) C-reactive protein >8 mg/l 8/17 Complement C3 (g/l) 1 48 ( ) Complement C3 >1 43 g/l 8/17 von Willebrand factor (IU/mL) 1 73 ( ) von Willebrand factor >1 92 IU/mL 5/12 Complete blood count White blood cell count (10⁹ cells per L) 11 9 ( ) White blood cell count > ⁹ cells per L 12/17 Haemoglobin (g/l) 124 ( ) Haemoglobin <120 g/l 5/17 Platelet count (10⁹/L) 360 (87 573) Platelet count >400 10⁹/L 6/17 Rheumatology laboratory tests IgG (g/l) 10 1 ( ) Antineutrophil cytoplasmic antibody negative 16/16 Antinuclear antibody negative 13/17 Anticardiolipin antibody negative 13/15 Lupus anticoagulant negative 7/9 Lumbar puncture Abnormal 15/16 Opening pressure >20 cm H 2 0 5/9 White blood cell count (10⁶ cells per L) 28 9 (0 71 0) White blood cell count >5 0 10⁶ cells per L 13/16 Protein (g/l) 0 6 ( ) Protein >0 4 g/l 10/16 Glucose (mmol/l) 3 3 ( ) Glucose <2 1 mmol/l 0/15 Oligoclonal bands 3/9 IgG index >0 25 3/9 IgG synthesis increased (>0 06 g/l) 3/9 Data are n/n or mean (range). Table 3: Laboratory features at baseline comprehension in seven, perceptual reasoning in six, and processing speed in five. About 70% of patients treated with standardised therapy completed the induction protocol (figure 1). Two patients experienced poor control of CNS symptoms during induction and required additional treatment that was not included in the protocol. Nine patients went on to maintenance therapy with azathioprine and five received mycophenolate mofetil at the discretion of their treating physician. Most patients treated with azathioprine switched to mycophenolate mofetil; five had azathioprine treatment failure and two had intolerance, one with lymphopenia and one with Abnormal brain CT 9/16 Abnormal brain MRI 17/19 Increased signal on T2-weighted imaging or FLAIR 17/19 Restriction on diffusion-weighted imaging 2/19 Gadolinium enhancement (parenchyma) 8/19 Gadolinium enhancement (meninges) 6/19 Unilateral lesion 4/19 Bilateral symmetric lesion 1/19 Bilateral asymmetric lesion 12/19 Supratentorial lesion only 10/19 Supratentorial and infratentorial lesion 6/19 Infratentorial lesion only 1/19 Subcortical white matter lesion 14/19 Periventricular white matter lesion 5/19 Deep white matter lesion 7/19 Cortical grey matter lesion 14/19 Basal ganglia lesion 8/19 Cerebellar white matter lesion 1/19 Abnormal spine MRI 3/17 Abnormal angiography 0/19 Data are n/n. FLAIR=fluid attenuated inversion recovery. Table 4: Imaging features at baseline pancytopenia with infection. No patients experienced hepatic toxicity related to azathioprine. Five patients completed maintenance and came off medication. 13 of 19 patients reached the 24-month outcome and were included in the primary analysis, 70% of whom had a good neurological outcome by PSOM (figure 3). Two patients who had a disease flare during induction or maintenance continued to have a severe deficit, as assessed by PSOM. At 12 months, 50% of patients had a good neurological outcome (figure 3). At final follow-up nine of 19 patients had achieved a good neurological outcome by PSOM. One patient who had achieved normal PSOM at 24 months had a moderate PSOM at final followup, which was caused by a recent flare in symptoms. Five patients who switched from azathioprine to mycophenolate mofetil during maintenance did so because of flare of their CNS symptoms at 1, 6, 8, 9, and 16 months into the maintenance period (table 5); no patients flared on mycophenolate mofetil. All four patients with optic neuritis (as diagnosed by a paediatric ophthalmologist) experienced flare of their eye disease during induction (two patients who received additional intravenous gammaglobulin and infliximab) or maintenance (two patients who switched from azathioprine to mycophenolate mofetil). Four patients remained in remission after discontinuation of drugs at 26, 31, 39, and 57 months after diagnosis. One patient experienced a flare after 14 months off treatment. After brain biopsy, one child developed a wound infection that required hospital admission and one patient had an Published online October 4, 2010 DOI: /S (10)70243-X 5

6 Figure 2: Inflammatory findings in a brain biopsy from a patient with small vessel childhood primary angiitis of the CNS Movat s stain of a histology section of the white matter of the patient shows transmural vessel inflammation with a mixed cell population including lymphocytes and macrophages, reactive endothelial cell hypertrophy, and adjacent reactive astrocytes in the surrounding white matter. Patients μm Severe deficit Moderate deficit Mild deficit Normal 12 months 24 months Figure 3: Neurological outcome in children with small vessel childhood primary angiitis of the CNS, as measured by the paediatric stroke outcome measure score episode of pancytopenia during maintenance with azathioprine; she developed a calf abscess and required hospital admission. Thiopurine S-methyltransferase genetic testing was done and revealed a normal azathioprine metabolism variant. Although all patients experienced weight gain associated with prednisone treatment, no patients had type 2 diabetes mellitus, and there were no deaths. Discussion In this relatively large cohort of children with small vessel childhood primary angiitis of the CNS, we show the potential efficacy of an immunosuppressive protocol. The treatment protocol was effective in controlling the Disease flares On induction therapy 2/19 On maintenance therapy 5/14 Off medication 1/5 Total 8/19 Remission off medication 4/19 Serious infection requiring hospital admission During induction therapy 1/19 During maintenance therapy 1/14 Complications Cataracts 1/19 Avascular necrosis 1/19 Vertebral fractures 3/19 Type 2 diabetes mellitus 0/19 Death 0/19 Data are n/n. Table 5: Secondary outcomes severe neurological manifestations of CNS vasculitis in the majority of our patients. Our treatment regimen was well tolerated and effective in reversing neurological deficits. We therefore recommend this protocol for treatment of patients with small vessel childhood primary angiitis of the CNS (panel 2). A high proportion of patients showed flare of their neurological symptoms or had a significant adverse event (lymphopenia, pancytopenia, and infection) while on azathioprine, which was resolved after switching to mycophenolate mofetil. Thus, we suggest that mycophenolate mofetil should be the preferred maintenance steroid-sparing therapy. We no longer use azathioprine as a maintenance drug in patients at our centre; all patients now receive mycophenolate mofetil as maintenance medication. There are a number of important limitations to our study. The number of patients was small; the requirement of a brain biopsy to confirm the diagnosis of small vessel childhood primary angiitis of the CNS limited the study size. However, brain biopsy ensured that there was no confusion with diseases such as multiple sclerosis or acute disseminated encephalomyelitis, which can have similar clinical and imaging features but a different pathological target. Despite the small sample size, this is the largest reported cohort of patients with this condition; larger studies should be possible in the future when recognition and diagnosis of CNS vasculitis in children increases. Statistical analysis to identify risk factors present at diagnosis that affect long-term outcome will be possible in these larger studies. Although the PSOM is a useful measure for documenting neurological deficits on the basis of clinical findings, it does not account for other neurological sequelae, such as academic impairments revealed in neurocognitive testing, psychiatric 6 Published online October 4, 2010 DOI: /S (10)70243-X

7 Panel 2: Research in context Systematic review We did a PubMed search between 1990 and 2009 for articles that reported cases of primary CNS vasculitis in children. Particular attention was paid to descriptions of CNS vasculitis that involved the small vessels of the brain and that were diagnosed on brain biopsy or autopsy. Previous case reports and case series have described a wide variety of treatments in patients with small vessel childhood primary angiitis of the CNS; 1,2,5 7 however, there has been no uniform treatment protocol described for this group of patients, and neurological outcome has been variable. Interpretation We report findings from the largest cohort of children followed up prospectively with small vessel childhood primary angiitis of the CNS; this is the first study to implement and show the potential efficacy of an immunosuppressive protocol for these children. Our treatment regimen was well tolerated and effective in reversing neurological deficits. We therefore recommend that this protocol should be considered when a patient is diagnosed with small vessel childhood primary angiitis of the CNS. manifestations, and ongoing seizure disorder. Additionally, the PSOM does not include any measure of quality of life; the patients and parents in our series reported a broad range of total, physical, and psychosocial quality of life, which shows their varied experience of disease and treatment. However, the PSOM was the most effective and practical means for the objective measurement of neurological outcome in these patients. Future work to modify the PSOM to capture the complete spectrum of deficits seen in these patients is warranted. Identification and appropriate diagnosis of children affected by small vessel childhood primary angiitis of the CNS is crucial because with standardised treatment and measurement of efficacy and safety good neurological outcome is a realistic goal. Contributors All authors were involved in the study concept and design, data collection, standardised re-assessment of patient information, study conduct, data summary and analysis, literature search, and writing of the report. Conflicts of interest We declare that we have no conflicts of interest. References 1 Yaari R, Anselm IA, Szer IS, Malicki DM, Nespeca MP, Gleeson JG. Childhood primary angiitis of the central nervous system: two biopsy-proven cases. J Pediatr 2004; 145: Benseler SM, deveber G, Hawkins C, et al. Angiography-negative primary central nervous system vasculitis in children: a newly recognised inflammatory central nervous system disease. Arthritis Rheum 2005; 52: Benseler SM, Silverman E, Aviv RI, et al. Primary central nervous system vasculitis in children. Arthritis Rheum 2006; 54: Barron TF, Ostrov BE, Zimmerman RA, Packer RJ. Isolated angiitis of CNS: treatment with pulse cyclophosphamide. Pediatr Neurol 1993; 9: Bitter KJ, Epstein LG, Melin-Aldana H, Curran JG, Miller ML. Cyclophosphamide treatment of primary angiitis of the central nervous system in children: report of 2 cases. J Rheumatol 2006; 33: Lanthier S, Lortie A, Michaud J, Laxer R, Jay V, deveber G. Isolated angiitis of the CNS in children. Neurology 2001; 56: Matsell DG, Keene DL, Jimenez C, Humphreys P. Isolated angiitis of the central nervous system in childhood. Can J Neurol Sci 1990; 17: 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: Varni JW, Seid M, Rode CA. The Peds QL: measurement model for the pediatric quality of life inventory. Med Care 1999; 37: Wechsler D. Wechsler intelligence scale for children, 4th edn. New York: Psychological Corporation, Wechsler D. Wechsler adult intelligence scale, 3rd edn. New York: Psychological Corporation, Kitchen L, Anderson PE, Friefeld S, et al. A validation study of the pediatric stroke outcome measure. Stroke 2003; 34: deveber GA, MacGregor D, Curtis R, Mayank S. Neurologic outcome in survivors of childhood arterial ischemic stroke and sinovenous thrombosis. J Child Neurol 2000; 15: Gallagher KT, Shaham B, Reiff A, et al. Primary angiitis of the central nervous system in children: 5 cases. J Rheumatol 2001; 28: Published online October 4, 2010 DOI: /S (10)70243-X 7