Treatment of vasculitic peripheral neuropathy: a retrospective analysis of outcome
From the 1Department of Neurology, Radcliffe Infirmary, Oxford, 2Department of Clinical Neurology, University of Oxford, Oxford, and 3Department of Neuropathology, Frenchay Hospital, Bristol, UK
Address correspondence to Dr Michael Donaghy, Department of Clinical Neurology, Radcliffe Infirmary, Woodstock Road, Oxford OX2 6HE, UK. email: joanna.wilkinson{at}clneuro.ox.ac.uk
Received 20 April 2006 and in revised form 16 August 2006
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Background: Vasculitis of the peripheral nervous system (PNS) is rare. There are no controlled treatment trials, and clinical practice is guided by experience from case series and indirectly by analogy with systemic vasculitis.
Methods: We identified patients (n = 212) with possible vasculitic peripheral neuropathy (VPN) from the neuropathology and neurophysiology records of two centres over 28 years. Case-notes were available for 181, from which, 106 cases of clinicopathological VPN were identified. Adequate treatment data were available in 100; follow-up data, in 93.
Results: Of 106 cases, 95 had systemic vasculitis and 11 had vasculitis confined to the PNS. Pharmacological treatment (94/100 cases) was corticosteroid-based, and included cyclophosphamide in 54; 17 received additional agents. Initial stabilization was achieved in all but six. One-year survival was 90.3%. Of the nine who died in the first years (mean age 73 years), seven had received cyclophosphamide, and all but two had severe, multisystem vasculitis. The neurological relapse rate was 10%. Only one relapse occurred after cyclophosphamide treatment. Outcome was reported as good in 72% (78% in those who relapsed).
Discussion: Death and relapse were infrequent in treated patients. Relapse occurred almost exclusively in patients treated with prednisolone alone. Aggressive early treatment with cyclophosphamide may prevent relapse. The current management approach to VPN appears largely effective, especially if cyclophosphamide is used.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Vasculitic peripheral neuropathy (VPN) is an aetiologically heterogeneous condition characterized by necrotizing inflammation and luminal narrowing of the vasa nervorum, resulting in ischaemic lesions of the peripheral nerve. This may occur as part of systemic vasculitis, or as an isolated condition of the peripheral nervous system (PNS). Unsurprisingly, there have been no randomized controlled treatment trials for this rare and varied condition. However, neuropathy is a common feature of the vasculitides, where it causes significant disability but is often strikingly responsive to immunosuppressive therapy.1 At present, clinical practice is guided by experience from case studies and extrapolation from clinical data from systemic vasculitis.
Classification of primary vasculitides (those of unknown aetiology) has been notoriously difficult. Two schemes predominate, though neither is designed for diagnosis of individual cases. The American College of Rheumatology (ACR) provides criteria for seven subtypes: polyarteritis nodosa (PAN), Churg-Strauss syndrome (CSS), Wegener's granulomatosis (WG), hypersensitivity vasculitis, Henoch-Schönlein purpura (HSP), giant cell (temporal) arteritis, and Takayasu's arteritis.2 The Chapel Hill Consensus Conference (CHCC)3 distinguishes vasculitides by vessel size and histopathology, hence requiring diagnostic biopsy. By this classification, temporal and Takayasu's arteritis are large-vessel vasculitides, classic PAN and Kawasaki disease are medium-vessel, and WG, microscopic polyangiitis (MPA), CSS, HSP, essential cryoglobulinaemic vasculitis and cutaneous leukocytoclastic angiitis are small-vessel.
VPN is a common feature of vasculitides affecting blood vessels of small-to-medium calibre, because the size of these vessels corresponds to that of the vasa nervorum. It is an ACR diagnostic criterion in PAN and CSS, where PNS involvement occurs in 
50–78%.4–7 The reported incidence of VPN in primary vasculitides has varied from 20% to 80%,8 reflecting the clinicopathological overlap between individual syndromes and associated nosological uncertainty.
VPN is also a well-recognized feature of the secondary vasculitides. A systemic vasculitis, histologically indistinguishable from PAN, occurs in 5–15% of patients with rheumatoid arthritis (RA), of whom 40–50% will develop a clinically apparent VPN.9 VPN is seen in 45% of hepatitis-C-infected patients with mixed cryoglobulinaemia and 9–10% of those without,10 and in 0.3–1% of HIV-infected subjects.11 It may also be a paraneoplastic phenomenon, particularly associated with solid tumours, such as small cell lung cancer, and lymphoproliferative disorders.
Vasculitis restricted to the PNS was first described in 1938.12 In the last 20 years, non-systemic vasculitic neuropathy (NSVN) has emerged as distinctive clinical entity, thought to account for 25–33% of cases of VPN.13–18
In the absence of controlled clinical trials on VPN treatment, or any likelihood that they will be conducted, we wished to assess the effectiveness of current approaches to treatment as a pragmatic guide to physicians who encounter the condition in their clinical practice. This study reviews 106 cases of VPN seen in two centres over 28 years, with particular reference to treatment, outcome and relapse.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Case ascertainment
Cases were initially identified from three sources. The majority were from review of the neuropathology records of two neurology tertiary referral centres (Radcliffe Infirmary, Oxford, and Frenchay Hospital, Bristol) between 1976 and 2004: all nerve biopsies performed because of a clinical suspicion of vasculitis and reported by a consultant neuropathologist as being consistent with vasculitis were considered. These included those with definite vasculitis (at least one blood vessel infiltrated by inflammatory cells in association with signs of vascular injury) and those with more minor changes, predominantly axonal loss, the severity of which varied (sometimes markedly) between fascicles or parts of fascicles. Other cases were identified from neurophysiology coding, as mononeuritis multiplex (Radcliffe Infirmary). The remaining cases came from a tertiary referral peripheral nerve specialist's database (MD). Case-note review was undertaken following ethical approval by the relevant Local Research Ethics Committee.
Patient selection
Patients for whom case-notes were unavailable (lost or untraceable by the medical records department) were not included in further analysis. Patients for whom case-notes were available were included, following critical analysis of each case from their notes by the authors, only if a diagnosis of VPN had been made by the physician responsible for the patient's care at the time of the illness. In all cases, the physician was a neurology, nephrology or rheumatology consultant with expertise in vasculitis, including its manifestations in the PNS. The diagnosis was based on the combination of supportive clinical, pathological and laboratory features in each patient, requiring the presence of a neuropathy unexplained by an aetiology other than vasculitis with at least one of the following features: (i) evidence of a systemic vasculitis on laboratory assays, and/or (ii) positive diagnostic nerve or other tissue biopsy (see later definitions).
Data collection
Diagnostic, demographic, clinical, laboratory, neurophysiological, histopathological and initial treatment data were collected. We accepted the diagnostic classification given by the treating physician at the time of diagnosis. No attempt was made to retrospectively reclassify the cases according to modern criteria, as it was felt that this was likely to be misleading given the evolution of laboratory assays and clinical knowledge during the years encompassed by this study. However, readers should be alert to the significant changes that have occurred in the classification schemes of the primary vasculitides. In particular, prior to the emergence of MPA as a distinct subtype in 1994, it is likely that these cases were included in those diagnosed as PAN, as the presence of a glomerulonephritis was not considered incompatible with the diagnosis as it is now, according to the CHCC.
For inclusion in the outcome and relapse analysis, a minimum of 1 year (or until death, if sooner) of adequately documented follow-up was required. To be considered ‘adequately documented’, case-notes with clear descriptions of the patient's clinical status (including management changes, relapses, and subjective treatment responses) must have been available. In every case, efforts were made to extend the follow-up period. In the majority, this entailed review of case-notes from referring district general hospitals and (less often) communication with general practitioners. Patients were not contacted directly or examined by the authors specifically for the purpose of this study. Statistical analysis was performed using the independent-samples t test.
Definitions
Remission includes those who had partial or complete recovery of their peripheral neuropathy regardless of treatment, being defined as the absence of symptoms of VPN, or a stable state consistent with scarring. To be considered as stable, there must have been no neurological deterioration for a minimum period of four weeks. For this study, only neurological relapses were considered. A relapse was defined as the occurrence of a new manifestation of VPN (i.e. new peripheral nerve territory lesion on examination) or recurrence or worsening of neurological symptoms attributable to vasculitis, including pain, following remission or stabilization for a minimum of 4 weeks. As such, a relapse could occur either spontaneously or following a change in medication (either tapering or withdrawal), but could not occur if a period of remission had not first been achieved. The outcome was classified as ‘good’ if the notes included details of independent functioning in the activities of daily living. The outcome was ‘poor’ when significant handicap due to neurological impairment or symptoms was documented. In either case, this reflected objective measures, where available, or the patient's subjective assessment of their functioning in the activities of daily living, as recorded in out-patient follow-up letters. The outcome was also considered poor if death occurred within 1 year of diagnosis. Given the nature of this study, with cases collated from disparate sources over many years, no attempt has been made to retrospectively apply more rigorous outcome definitions than those outlined above. To guide readers, however, it may be helpful to consider a ‘good’ outcome as equating with a modified Rankin score of 0–2 (i.e. at worst, unable to carry out all previous activities but able to look after own affairs without assistance), and a ‘poor’ outcome being equivalent to a score of 3–6 (i.e. at best, moderately disabled, requiring some help but able to walk without assistance).
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Clinical features
Between 1976 and 2004, 212 possible cases were identified. Of these, notes were unavailable for 31. Of the remaining 181 patients, 106 fulfilled the inclusion criteria (Table 1): 54 (50.9%) were female, and 52 (49.1%) were male. Mean age at presentation was 58.7 years (range 24–83) for the total sample: 56.5 (range 24–83) for females and 61.5 (range 24–80) for males. There were three females and eight males with NSVN, with the mean age in this group being 59.9 years (range 38–75). Mean age in the systemic vasculitis group was 58.2 (range 24–83). Overall, 85 patients (80.2%) presented with a clinical picture of mononeuritis multiplex, 18 (17.0%) with distal symmetrical polyneuropathy and 3 (2.8%) with a mononeuropathy.
|
Biopsies
There were 122 biopsies, performed on 98 patients (Table 2): 63 patients had a nerve biopsy alone; 20 had a nerve biopsy and one other non-neural biopsy; one had a nerve biopsy and two other non-neural biopsies; and 14 had only non-neural biopsies (12 had a single biopsy and 2 had two).
|
A total of 84 nerve biopsies were performed (82 sural, 2 superficial radial). We could not ascertain the exact proportion of the nerves biopsied that had been involved either clinically or electrophysiologically. Forty-nine biopsies confirmed VPN (as evidenced by transmural or perivascular inflammation in association with vascular injury, including vascular thickening, fibrinoid necrosis, endothelial cell disruption, luminal obliteration, fragmentation of the internal elastic lamina, haemorrhage, acute thrombus or recanalization, or epineurial neovascularization) and 31 were ‘consistent’ with VPN (asymmetric acute axonal loss with differential fasicular involvement with or without active Wallerian degeneration and/or regenerating fibres in the absence of inflammatory infiltration). The sural nerve biopsy was normal in four patients, only one of whom (who had a painful mononeuritis multiplex in the context of proven meningococcal sepsis with typical vasculitic skin rash) was included in the study without other positive tissue biopsy.
Of the 22 additional tissue biopsies performed in those undergoing nerve biopsy (9 renal, 7 muscle, 5 skin, 1 GI tract), 3 were done in patients with normal nerve biopsies (all positive), 8 were done in patients with ‘consistent’ nerve biopsies (6 positive, 1 (muscle) showing acute asymmetrical axonal denervation), and 11 in patients with positive nerve biopsies (7 positive, 2 (muscle) showing acute asymmetrical axonal denervation).
Of those not having a nerve biopsy, five had renal biopsies (4 positive for vasculitis); four had skin biopsies (3 positive); six had muscle biopsies (3 showed acute axonal asymmetrical denervation and atrophy, but only 1 was positive); and one had a GI tract biopsy (positive). One patient with long-standing rheumatoid arthritis who developed skin lesions and an acute painful mononeuritis multiplex, had two normal non-neural biopsies (skin and muscle).
A total of 12 patients were included on the basis of their clinical and laboratory features without confirmatory biopsy (8 who did not have a biopsy and 4 with normal or inconclusive histology). All 12 cases in this category had multiorgan systemic vasculitis, with a minimum of three organs clinically involved (with the exception of the meningococcal case described above, who had only skin and nerve disease) in addition to the presence of markers of systemic vasculitis on laboratory assays (a minimum of two of: markedly raised ESR, eosinophilia, ANA, ENA or ANCA positivity).
Other investigations
ANCA test were done in 78 patients and were positive in 37. Subtyping for MPO or PR3 was available in 23 (14 and 9, respectively). No attempts has been made to correlate this to diagnosis as both the assays and the diagnostic classification schemes have changed during the period studied. None of the patients with NSVN had a positive ANCA. Complete neurophysiology data were often unavailable: of the 78 reviewed, 58 had results consistent with axonal sensorimotor mononeuritis multiplex, 15 had a distal symmetrical polyneuropathy (of which two were predominantly sensory), three mononeuropathy and two, radiculopathy. Given the limited and non-standardized data available, we have not attempted further analysis.
Diagnostic classification
VPN was part of a systemic vasculitis in 84 patients (79.2%): 66 (62.3%) were classified as having primary systemic vasculitis and 23 (22.0%), secondary. The commonest diagnosis was Churg-Strauss syndrome (22, 20.8%), followed by PAN (19, 17.9%), WG (14, 13.2%), MPA (8, 7.5%), essential cryoglobulinaemia (2, 1.9%) and Henoch-Schönlein purpura (1, 0.9%). Of the 23 cases attributable to a secondary vasculitis, 16 (15.1%) occurred in association with connective tissue disease (12 rheumatoid arthritis, 4 systemic lupus erythematosis), 3 (2.8%) were infection-related (2 hepatitis-C-associated cryoglobulinaemia, 1 meningococcal septicaemia), and 4 (3.8%) were paraneoplastic (2 non-Hodgkin's lymphoma, 1 small-cell lung carcinoma and 1 unspecified GI malignancy). There were 11 (10.4%) cases of non-systemic vasculitic neuropathy. In 6 (5.7%), there was multiorgan involvement by an unclassified systemic vasculitis (USV).
Treatment
Of the 106 cases, initial treatment was known in 100 (Figure 1); the other six were excluded from subsequent analysis. In total, 54 received a regimen including cyclophosphamide, with 40 having a corticosteroid-based regimen without cyclophosphamide. Of these 40, 37 received oral prednisolone alone (initial doses of 40–100 mg/day, median 60), two had intravenous methylprednisolone (IVMP) followed by oral prednisolone, and one had IVMP alone. Of the 54 receiving a regimen including cyclophosphamide, 35 had cyclophosphamide and oral prednisolone, and 19 had IVMP followed by cyclophosphamide and oral prednisolone. Single courses of IVMP were given, with doses from 1 g for 5 days (although this dose was more commonly given for only 3 days) to 500 mg for 5 days. Specific data regarding the use of either pulsed intravenous or oral cyclophosphamide were not collected as, while patients may have been given one or more IV pulses early in the disease as an inpatient, many were subsequently managed on daily oral regimens and thus dichotomized classification into either treatment regimen was not clear cut. As a general observation, however, there was an increasing trend towards the use of pulsed intravenous therapy and away from daily oral regimens, presumably following the publication of several systematic randomized controlled trials in this area.19,20 Seventeen received other drugs in addition to the above combinations (including doxorubicin, vincristine, chlorambucil, methotrexate, anti-virals and prostacyclin). Six received no pharmacological treatment.
|
Follow-up, remission and relapse
There were sufficient follow-up data for further analysis in 93 of the 100 cases in whom treatment was known (Figure 2). Overall duration of follow-up ranged from 13 months (or death, whichever was the sooner) to 13 years (mean 43.8 months, median 34 months). In the NSVN group, mean follow-up was 37.7 months (median 35) with the mean duration of follow-up being 43.9 months (median 34) in the systemic vasculitides group. Remission was achieved in all but six cases. Five received treatment for their VPN but died within the first year, three within the first 2 months of diagnosis. In the remaining case in which remission was not achieved, the patient chose not to be treated despite recommendation. Of the 87 cases in which remission was achieved, nine subsequently relapsed: four had CSS, two RA, two NSVN and one USV. The patient with USV was the only one who relapsed after treatment with cyclophosphamide; the others received prednisolone alone.
|
Mortality
There were 16 deaths during follow-up, 9 within the first year of diagnosis.
In the 9 deaths occurring within 1 year of diagnosis, non-actuarial one-year survival was 90.3%. The mean age of those dying within 1 year was significantly higher than that of the group as a whole (73.1 years, p = 0.006). Although five of these deaths occurred in patients in whom remission of the VPN had not occurred, the remaining four were in cases in which remission had been achieved with treatment: two deaths from neutropenic sepsis (both had received cyclophosphamide); one from the underlying lung carcinoma (causing a paraneoplastic VPN); and one from an unknown cause, in a nursing home 9 months after the diagnosis of PAN had been made. Seven had severe multisystem vasculitis causing multiorgan failure (3 PAN, 1 CSS, 1 WG, 1 MPA, 1 RA): one died from small-cell lung carcinoma (as above), and one from ischaemic heart disease having never achieved remission (essential cryoglobulinaemic vasculitis). Seven of those dying within 1 year had received treatment including cyclophosphamide. In five of these, the clinical picture at death included neutropenic sepsis. The early deaths in patients not receiving cyclophosphamide were in the two dying without multisystem vasculitis (essential cryoglobulinaemia and paraneoplastic VPN).
In the 7 patients who died during follow-up but after the first year, two deaths were attributed to renal complications of their vasculitis, one to underlying GI malignancy, one to bone-marrow suppression secondary to long-term cyclophosphamide, and two to unrelated causes. In one death, the cause was unclear.
Outcome
Of the 93 patients included in outcome analysis, a total of 67 (72.0%) reported a good outcome (this includes the one case in whom treatment was refused, despite symptomatic remission not being achieved, because the patient reported independent functioning). Three of the patients who had a good outcome died during the period of follow up, but this was a minimum of 5 years after diagnosis, and in two was unrelated to their vasculitis. Seven (10.6%) of the 66 patients in who remission was achieved who had a good outcome experienced a relapse during follow-up, but all subsequently improved on escalation of therapy.
The outcome was poor in a total of 21 patients (22.6%) (by definition, this includes the 9 who died within 1 year of diagnosis). Of the 12 who survived the first year but had a poor outcome, three died within the period of follow-up (two within 3 years of diagnosis, and two from causes related to their vasculitis or its treatment). Of the remaining nine, only two had documented neurological relapses, but neither of these responded to therapy escalation. In four of the others, loss of manual dexterity or impaired mobility were prominent symptoms, often with pain. In five (5.4%) of the 93 patients, there were insufficient details for outcome analysis (Tables 3 and 4).
|
|
Subset analysis of those with positive nerve biopsies
Of the 49 patients with a positive nerve biopsy, 28 were male, 21 female. Mean age was 62.5 years. Forty-two had a systemic vasculitis (13 PAN, 9 CSS, 6 WG, 3 MPA, 1 EC, 6 connective tissue disease, 2 paraneoplastic, 1 associated with hepatitis C infection, 1 USV). There were 7 with NSVN. Treatment and follow-up data were available in 46 of these patients (mean follow-up 42.7 months, median 41). Twenty-six (including two with NSVN) were treated with a regimen including cyclophosphamide, 18 (4 with NSVN) received a corticosteroid-dominant regime without cyclophosphamide. and two (1 with NSVN) did not receive pharmacological treatment. Remission was achieved in all but two patients (one with EC and one with CSS), both of whom died within a year of diagnosis.
There were four deaths in the first year (mean age 73 years): three had severe multisystem disease and were treated with cyclophosphamide, with the picture at death including multiorgan failure and neutropenic sepsis in two; the fourth, without multisystem disease or treatment with cyclophosphamide, was the patient with EC who died of causes unrelated to her VPN (described above). Four further deaths occurred during follow-up: one from the GI malignancy causing a paraneoplastic VPN, one from bone-marrow suppression, and two from unrelated causes.
Two relapses occurred in the 46 patients in this subset; both had CSS and had been treated with oral prednisolone alone. Both subsequently reported a good overall outcome.
Overall outcome was good in 35/46 patients (76.1%), poor in 8/46 (17.4%) (including, by definition, the 4 early deaths) and unknown in 3/46 (6.5%). Of the 24 patients with systemic vasculitis treated with cyclophosphamide, 17 had a good outcome and five a poor one, with two unknown. Both patients with NSVN treated with cyclophosphamide had a good outcome. Of the 14 with systemic vasculitis treated with a corticosteroid-based regimen, 11 had a good outcome and three a poor one (all four NSVN patients had a good outcome with corticosteroid treatment). The outcome was also good in the NSVN patient who received no treatment (outcome unknown in the other case).
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
VPN is rare, and treatment remains largely based on intuitive clinical knowledge, guided by experience from case series. This review of 106 cases gathered over 28 years is, to our knowledge, the largest series currently available. The majority of cases occurred in the context of a primary systemic vasculitis (62%). The proportion of cases occurring as a manifestation of secondary vasculitis was surprisingly low (22%) given the prevalence of these disorders; for example, RA affects
1% of the population, of whom 2.5–7.5% would be expected to develop VPN during the course of their disease,10 and yet only 12 such cases were included. This may be because patients with known primary rheumatological disease are less likely to be referred to a neurologist or to undergo neurophysiology or biopsy. In addition, the presence of neuropathy is often overlooked in severe multisystem vasculitis. In our experience, many cases excluded following review of the notes were renal patients in whom there was passing reference to neurological symptoms but inadequate supportive data to diagnose VPN. Hence, our study could not include all the cases of VPN occurring in two catchment areas over 28 years, and no conclusions can be drawn with regard to its incidence. Given our case ascertainment sources, it was perhaps surprising that the proportion of cases due to NSVN was only 10%, lower than the 25–33% reported by other studies.14,15 However, of our 11 cases, nine were diagnosed after 1994, suggesting that earlier cases may have been overlooked. The overall neurological outcome was good: 72% of these patients reported minimal impairment at follow-up, and only 13% of those surviving the first year reported a poor outcome. This concurs with other series, in which final neurological outcome has been described as ‘surprisingly good‘, with ‘improvement ... to be expected‘.2,21–23 Given the limitations of retrospective analysis of routine clinical records, we have only attempted a broad classification of outcomes. By comparison, in prospective trials in systemic vasculitides, outcome measures combining disease activity, damage indices and patient function are widely used. The collaborative European Vasculitis Study Group (EUVAS) uses a combination of the Birmingham Vasculitis Activity Score, the Vasculitis Damage Index and the Short-Form 36 (a well-validated questionnaire of subjective perception of disease impact).24 Similar approaches will be required for more rigorous outcome data to be ascertained in VPN.
Deriving quantitative outcome data retrospectively from case-notes is inherently difficult. However, the presence or absence of neurological relapse is usually unequivocal and clearly documented. In the 93 cases with adequate follow-up in this study, there were only nine relapses. This is comparable with other studies with similarly strict definitions of relapse: Hawke reported no relapses in 17 surviving patients with VPN (mean follow-up 36 months),23 and only two of Hattori's 28 patients with CSS-associated neuropathy developed relapses (mean follow-up 4.2 years).25 However, Collins reported a relatively high relapse rate of 46% in NSVN.17 This may reflect their longer duration of follow-up (median 63 months, range 6–260) as significantly delayed relapse has been observed in some patients with NSVN and may occur following several years of apparent quiescence. However, Collins’ study also includes non-specific symptoms, such as malaise and fever, in addition to neurological lesions in their definition of relapse, and this is also likely to be contributing to their overlap with the relapse rates observed in systemic vasculitides, which range from 11–60%26 when all manifestations are included.
A striking finding in this series is that of the nine relapses, only one occurred in a patient treated with cyclophosphamide: all the others were initially treated with prednisolone alone. This tendency for a reduced relapse rate in vasculitis patients treated more aggressively at the outset has been observed by other groups. In 1981, Moore and Fauci commented that ‘virtually no extension or progression of disease activity was seen in any patient [with neurological manifestations of systemic vasculitis] receiving an adequate course of cyclophosphamide‘.2 Similarly, Collins and Periquet noted that in NSVN, combination therapy (generally with cyclophosphamide and corticosteroids) was ‘more effective than corticosteroid monotherapy in inducing remission and improving disability, with trends towards a reduced relapse rate‘: 29% of their 20 patients receiving combination therapy relapsed, compared with 59% of the 28 receiving corticosteroid monotherapy. Interestingly, they also found that patients with mild neuropathies were no more likely to respond to corticosteroid monotherapy than those with severe neuropathies.17 Supportive data also come from the systemic vasculitides, where reduced rates of mortality, morbidity and relapse in those treated with cyclophosphamide have been reproducibly demonstrated:27,28 cyclophosphamide is now considered integral to the standard regimen in the majority of cases, with more recent studies focusing on the comparative efficacy of pulsed intravenous vs. oral cyclophosphamide.19,20 From our study, it is difficult to give categorical advice with respect to the use of cyclophosphamide in NSVN. While aggressive early treatment may prevent relapse, the neuropathy appears to follow a relatively benign course, with relapses responding to appropriate therapy escalation such that all our cases reported a good eventual outcome. More detailed indices of outcome incorporating measures of treatment-associated morbidity will be of assistance to the managing physician.
Our one-year survival was 90.3%, with death occurring predominantly in older patients with severe multisystem disease. This concurs with data from the systemic vasculitides.29–31 Most of those who died in the first year had been treated with cyclophosphamide. This apparent excess of deaths in part reflects severe underlying disease, which instinctively warrants more aggressive treatment. However, in five patients, the clinical picture at death included neutropenic sepsis, and although (given our average duration of follow-up) it is not possible to assess the late mortality and morbidity attributable to cyclophosphamide in this study, at least one delayed death was related to bone-marrow suppression secondary to its use. As such, the apparent beneficial effects of cyclophosphamide in controlling the underlying vasculitic process and limiting subsequent relapse must clearly be tempered with caution with respect to its toxicity, in both the short and long term. However, overall, the number of delayed deaths in our study was small and, contrary to some earlier studies, we did not find an increased mortality in patients with VPN: in those surviving the first year, our mortality rate was 8.3%, with a mean follow-up period of 43.8 months, whereas Hawke, in 1991, reported 50% mortality at a mean follow-up of 36 months.23
In this study, CSS was the largest diagnostic subgroup, accounting for 22 cases (21%). This is perhaps not surprising, as neuropathy is a dominant feature of CSS: Lane found that only 11% of patients with a systemic vasculitis were first referred to a neurologist, and all of these had CSS.32 Only one of our CSS patients died, and a good outcome was reported in 84%, including all 12 receiving predominantly glucocorticoid treatment. A similarly benign outcome was seen in the glucocorticoid monotherapy cohort in Hattori's study of CSS-associated neuropathy. These findings are in agreement with the systemic vasculitis literature, where a consensus of less aggressive treatment is emerging for CSS, with glucocorticoid monotherapy being advocated by many except in the context of life-threatening disease.33,34
VPN is a rare condition, but one which may cause considerable morbidity if untreated. We believe that the intuitive extension of more evidence-based practice in systemic vasculitis to the management of VPN is justified, and results in a positive neurological outcome. The efficacy of cyclophosphamide in severe vasculitis is indisputable, where its introduction has revolutionized the previously dismal prognosis.35 However, its role in the initial and safe treatment of relatively benign disease remains unclear, given its associated toxicity. Although relapse occurred almost exclusively in patients treated with prednisolone alone, the effect of this on long-term neurological outcome remains uncertain from this study.
|
|
|
Acknowledgements |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
We are grateful to the physicians for consenting to the review of their patients’ case-notes. KT is funded by an MRC Clinician Scientist Fellowship. None of the authors have a conflicting interest.
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
1. Kieseier BC, Kiefer R, Gold R, Hemmer B, Willison HJ, Hartung HP. (2004) Advances in understanding and treatment of immune-mediated disorders of the peripheral nervous system. Muscle Nerve 30 131–56.[CrossRef][Web of Science][Medline]
2. Moore PM and Fauci AS. (1981) Neurologic manifestations of systemic vasculitis. A retrospective and prospective study of the clinicopathologic features and responses to therapy in 25 patients. Am J Med 71 517–24.[CrossRef][Web of Science][Medline]
3. Hunder GG, Arend WP, Bloch DA, et al. (1990) The American College of Rheumatology 1990 criteria for the classification of vasculitis. Arthritis Rheum 33 1065–7.[Web of Science][Medline]
4. Jennette JC, Falk RJ, Andrassy K, et al. (1994) Nomenclature of systemic vasculitides. Proposal of an international consensus conference. Arthritis Rheum 37 187–92.[Web of Science][Medline]
5. Frohnert PP and Sheps SG. (1967) Long-term follow-up study of periarteritis nodosa. Am J Med 43 8–14.[CrossRef][Web of Science][Medline]
6. Moore PM and Cupps TR. (1983) Neurological complications of vasculitis. Ann Neurol 14 155–67.[CrossRef][Web of Science][Medline]
7. Guillevin L, Cohen P, Gayraud M, Lhote F, Jarrousse B, Casassus P. (1999) Churg-Strauss syndrome. Clinical study and long-term follow-up of 96 patients. Medicine 78 26–37.[CrossRef][Medline]
8. Sehgal M, Swanson JW, DeRemee RA, Colby TV. (1995) Neurologic manifestations of Churg-Strauss syndrome. Mayo Clin Proc 70 337–41.[Abstract]
9. Said G and Lacroix C. (2005) Primary and secondary vasculitic neuropathy. J Neurol 252 633–41.[CrossRef][Web of Science][Medline]
10. Puechal X, Said G, Hilliquin P, Coste J, Job-Deslandre C, Lacroix C, Menkes CJ. (1995) Peripheral neuropathy with necrotizing vasculitis in rheumatoid arthritis. A clinicopathologic and prognostic study of thirty-two patients. Arthritis Rheum 38 1618–29.[Web of Science][Medline]
11. Nemni R, Sanvito L, Quattrini A, Santuccio G, Camerlingo M, Canal N. (2003) Peripheral neuropathy in hepatitis C virus infection with and without cryoglobulinaemia. J Neurol Neurosurg Psychiatry 74 1267–71.
12. Mahadevan A, Gayathri N, Taly AB, Santosh V, Yasha TC, Shankar SK. (2001) Vasculitic neuropathy in HIV infection: a clinicopathological study. Neurol India 49 277–83.[Web of Science][Medline]
13. Kernohan JW and Woltman HW. (1938) Periarteritis nodosa: a clinicopathologic study with special reference to the nervous system. Arch Neurol Psychiatry 39 655–86.
14. Kissel JT, Slivka AP, Warmolts JR, Mendell JR. (1985) The clinical spectrum of necrotizing angiopathy of the peripheral nervous system. Ann Neurol 18 251–7.[CrossRef][Web of Science][Medline]
15. Dyck PJ, Benstead TJ, Conn DL, Stevens JC, Windebank AJ, Low PA. (1987) Nonsystemic vasculitic neuropathy. Brain 110 843–53.
16. Davies L, Spies JM, Pollard JD, McLeod JG. (1996) Vasculitis confined to peripheral nerves. Brain 119 1441–8.
17. Collins MP, Periquet MI, Mendell JR, Sahenk Z, Nagaraja HN, Kissel JT. (2003) Nonsystemic vasculitic neuropathy: insights from a clinical cohort. Neurology 61 623–30.
18. Kararizou E, Davaki P, Karandreas N, Davou R, Vassilopoulos D. (2005) Nonsystemic vasculitic neuropathy: a clinicopathological study of 22 cases. J Rheumatol 32 853–8.
19. Guillevin L, Cordier JF, Lhote F, et al. (1997) A prospective, multicenter, randomized trial comparing steroids and pulse cyclophosphamide versus steroids and oral cyclophosphamide in the treatment of generalized Wegener's granulomatosis. Arthritis Rheum 40 2187–98.[Web of Science][Medline]
20. Haubitz M, Schellong S, Gobel U, Schurek HJ, Schaumann D, Koch KM, Brunkhorst R. (1998) Intravenous pulse administration of cyclophosphamide versus daily oral treatment in patients with antineutrophil cytoplasmic antibody-associated vasculitis and renal involvement: a prospective, randomized study. Arthritis Rheum 41 1835–44.[CrossRef][Web of Science][Medline]
21. Pagnoux C and Guillevin L. (2005) Peripheral neuropathy in systemic vasculitides. Curr Opin Rheumatol 17 41–8.[CrossRef][Web of Science][Medline]
22. Collins MP and Periquet MI. (2004) Non-systemic vasculitic neuropathy. Curr Opin Neurol 17 587–98.[CrossRef][Web of Science][Medline]
23. Hawke SH, Davies L, Pamphlett R, Guo YP, Pollard JD, McLeod JG. (1991) Vasculitic neuropathy. A clinical and pathological study. Brain 114 2175–90.
24. Carruthers D and Bacon P. (2001) Activity, damage and outcome in systemic vasculitis. Best Pract Res Clin Rheumatol 15 225–38.[CrossRef][Medline]
25. Hattori N, Ichimura M, Nagamatsu M, Li M, Yamamoto K, Kumazawa K, Mitsuma T, Sobue G. (1999) Clinicopathological features of Churg-Strauss syndrome-associated neuropathy. Brain 122 427–39.
26. Jayne D. (2000) Evidence-based treatment of systemic vasculitis. Rheumatology 39 585–95.
27. Guillevin L, Lhote F, Jarrousse B, Fain O. (1992) Treatment of polyarteritis nodosa and Churg-Strauss syndrome. A meta-analysis of 3 prospective controlled trials including 182 patients over 12 years. Ann Med Interne (Paris) 143 405–16.[Medline]
28. Hogan SL, Nachman PH, Wilkman AS, Jennette JC, Falk RJ. (1996) Prognostic markers in patients with antineutrophil cytoplasmic autoantibody-associated microscopic polyangiitis and glomerulonephritis. J Am Soc Nephrol 7 23–32.[Abstract]
29. Guillevin L, Le Thi Huong D, Godeau P, Jais P, Wechsler B. (1988) Clinical findings and prognosis of polyarteritis nodosa and Churg-Strauss angiitis: a study in 165 patients. Br J Rheumatol 27 258–64.
30. Exley AR, Bacon PA, Luqmani RA, Kitas GD, Carruthers DM, Moots R. (1998) Examination of disease severity in systemic vasculitis from the novel perspective of damage using the vasculitis damage index (VDI). Br J Rheumatol 37 57–63.
31. Mahr A, Girard T, Agher R, Guillevin L. (2001) Analysis of factors predictive of survival based on 49 patients with systemic Wegener's granulomatosis and prospective follow-up. Rheumatology 40 492–8.
32. Lane SE, Watts RA, Shepstone L, Scott DG. (2005) Primary systemic vasculitis: clinical features and mortality. Q J Med 98 97–111.[Web of Science]
33. Solans R, Bosch JA, Perez-Bocanegra C, Selva A, Huguet P, Alijotas J, Orriols R, Armadans L, Vilardell M. (2001) Churg-Strauss syndrome: outcome and long-term follow-up of 32 patients. Rheumatology 40 763–71.
34. Langford CA. (2003) Vasculitis. J Allergy Clin Immunol 111 S602–12.[CrossRef][Medline]
35. Fauci A, Hayes B, Katz P, Wolff S. (1983) Wegeners's granulomatosis: prospective clinical and therapeutic experience with 85 patients for 21 years. Annals of Internal Medicine 98 76–85.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M P T Lunn and H J Willison Diagnosis and treatment in inflammatory neuropathies J. Neurol. Neurosurg. Psychiatry, March 1, 2009; 80(3): 249 - 258. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||