Article Text
Abstract
Background: Results of uncontrolled studies have suggested that infliximab is efficacious against systemic necrotising vasculitides (SNV) refractory to conventional treatment. However, its safety and ability to induce and maintain remission over the long term remain unknown.
Objectives: To report the use of infliximab to treat refractory SNV, focusing on patients’ longer-term outcomes.
Methods: The medical charts of patients given adjunctive infliximab for refractory SNV ⩾2 years before this evaluation were reviewed retrospectively.
Results: The 15 patients (median age 46 (range 20–69) years, median follow-up 35 (24–41) months) included 10 with Wegener’s granulomatosis, 1 microscopic polyangiitis, 3 rheumatoid arthritis-associated and 1 cryoglobulinaemia-related vasculitides. Infliximab was taken for a median time of 8 (2–31) months; 2 patients are still being treated. By day 45, 11 patients had entered remission (Birmingham Vasculitis Activity Score (BVAS) = 0) and 4 others had responded (BVAS decrease ⩾50%). Five patients achieved sustained remissions (⩾6 months, corticosteroids ⩽7.5 mg/day). Thirteen stopped infliximab because of loss of efficacy (n = 4), remission (n = 6) or non-compliance, chest tightness or side effect (1 each). Ten patients relapsed (median interval 13 months), 3 while still receiving infliximab; 2 were successfully re-treated with infliximab.
Conclusion: These observations highlight infliximab as a potentially useful and safe salvage treatment for patients with refractory SNV.
Statistics from Altmetric.com
For most patients with systemic necrotising vasculitides (SNV), conventional treatment combines corticosteroids (CS) and immunosuppressant drugs. Although attenuation is usually obtained, relapses are common, occurring in 49% of patients with Wegener’s granulomatosis (WG)1 and 35% of those with microscopic polyangiitis.2 However, SNV in some patients remain refractory to all conventional treatments and this highlights an unmet need for alternative treatments.
Results of uncontrolled studies suggested that infliximab, a monoclonal antibody blocking the tumour necrosis factor α (TNFα) receptor, had some efficacy against refractory SNV in the short term.3–5 In contrast, the authors of a large randomised, placebo-controlled trial reported that systematic adjunction of etanercept, a soluble TNFα receptor, to standard treatment provided no benefit as maintenance treatment for patients with WG, newly diagnosed or with a flare.6 The findings of that trial raised the question of whether its negative results apply only to etanercept or to the entire family of TNFα blockers, and whether they can be extrapolated to the subset of patients with refractory SNV.
Herein, we describe our experience using infliximab to treat refractory SNV, focusing on the patients’ longer-term outcomes.
PATIENTS AND METHODS
Patient selection
All patients with SNV refractory to conventional regimens, who had received off-label infliximab between 2000 and 2004, and for whom at least 2 years had elapsed between infliximab initiation and this evaluation, were considered for the study. SNV diagnoses satisfied the vasculitis criteria defined by the Chapel Hill Nomenclature7 or the American College of Rheumatology classification criteria for WG.8 Refractory SNV was defined as unchanged or heightened disease activity and at least one of the following: (a) CS and cyclophosphamide ⩾8 weeks; (b) CS and azathioprine, mycophenolate mofetil and/or methotrexate ⩾8 weeks with previous cyclophosphamide administration during the same WG flare and/or with high cumulative exposure to cyclophosphamide or (c) prescribing or resuming immunosuppression judged to be contraindicated.
Methods
The medical charts of eligible patients were reviewed retrospectively for duration of infliximab treatment, combined treatments (all patients were taking at least CS), subsequent outcome and reasons for infliximab withdrawal. Treatment response was assessed using the Birmingham Vasculitis Activity Score (BVAS), a validated instrument to measure disease activity in vasculitides,9 on day 45 (D45), then every 6 months. Remission was defined as no disease activity, corroborated by BVAS = 0. For patients not achieving remission, response to treatment was defined as BVAS reduction ⩾50%. Sustained remission was defined as BVAS = 0 for ⩾6 months with daily CS ⩽7.5 mg. Relapse was diagnosed when new/worsening clinical manifestations occurred after an initial response.10 Neither infliximab levels nor anti-infliximab antibodies were monitored.
RESULTS
Patient characteristics at infliximab initiation
The charts of 15 patients (nine men/six women; median age 46 (range 20–69) years, including nine of the 10 previously published cases3) diagnosed with WG (n = 10), microscopic polyangiitis (n = 1), rheumatoid arthritis-associated vasculitis (n = 3) or multivisceral hepatitis C virus-related type II mixed cryoglobulinaemia-associated vasculitis (henceforth referred to as MC; n = 1), were analysed. Vasculitis was biopsy proven for 12 patients, two others had clinical evidence of WG and were proteinase-3 antineutrophil cytoplasm antibody positive, and one had MC and an hepatitis C virus-positive serology.
Infliximab was prescribed to 14 patients for refractory SNV, despite CS and immunosuppressant drugs. For the patient with MC, infliximab was given after successful plasma exchanges had been stopped because of poor venous access. Table 1 gives details of the individual data.
At the onset of anti-TNFα therapy, SNV symptoms were constitutional (weight loss, malaise, fever, headache, myalgias and/or arthralgias), n = 7; cutaneous (ulcers, extensive pyoderma gangrenosum), n = 6; ophthalmological (uveitis, scleritis, retro-orbital pseudotumour), n = 3; ENT (nasal discharge, hearing loss), n = 6; pulmonary (infiltrate, nodules, alveolar haemorrhage, respiratory failure), n = 6; nervous system (mononeuritis multiplex, dura mater infiltration with hypophysial involvement and diabetes insipidus, pachymeningitis, stroke), n = 7; bloody diarrhoea, n = 1; glomerulonephritis relapse, n = 1. Median initial BVAS before infliximab was 10 (range 4–17).
Treatments
Infliximab (5 mg/kg) was infused on D0, D15 and D45, then every 4–6 weeks depending on the clinical response. CS were maintained at the same dose as before infliximab onset (median dose 26 (range 5–80) mg/day), then tapered as soon as possible, according to the clinical response, reaching a median of 11 (5–25) mg/day after 6 months of treatment.
Immunosuppressant dose or type was changed after the third infliximab infusion (D45) for seven patients. Patients 3, 8, 12 and 13 started a new immunosuppressant drug and/or intravenous CS within 2 weeks of infliximab onset. Treatments were unchanged for patients 9, 10 and 11. Relapsing WG in patient 7 was treated with no adjunctive agent other than infliximab.
Outcomes
Median follow-up was 35 (range 24–41) months. Median duration of infliximab administration was 8 (range 2–31) months, and is continuing for patients 8 and 12.
Figure 1 shows BVAS and CS-dose evolutions since starting infliximab, and the flow chart (fig 2) summarises patient outcomes. On D45 after infliximab initiation, 11 patients were in remission and BVAS had decreased ⩾50% for the four others (overall median BVAS = 0 (range 0–4)). During follow-up, patients 4, 7, 8, 12 and 13 achieved sustained remissions, and patient 1 was in remission ⩾6 months but still required prednisone (15 mg/day).
Ten patients relapsed at a median of 13 (range 6–32) months in previously affected sites (n = 9) or others (n = 1), including seven after stopping infliximab, and three while still receiving infliximab. The relapses of four patients were re-treated with infliximab: the early cutaneous relapse of patient 14 with rheumatoid arthritis-associated vasculitis, the cutaneous MC flare of patient 6, and patients 5 and 15 with WG. Of these, patient 5 again entered remission (but subsequently relapsed) and patient 14 achieved a sustained remission. The six other relapses were treated with other agents; among these six, patients 3 and 6 died of active SNV.
Adverse events
Infliximab was well tolerated by most patients, with side effects seen in three. Patient 2 had chest tightness during infusion with no evidence of cardiac involvement based on specific investigations, including cardiac MRI; etanercept replaced infliximab after the second infusion and was stopped after 10 months for inefficacy. Patient 6 with MC experienced an anaphylactoid reaction when infliximab was reintroduced 1 year after the first five infusions because of worsening cutaneous ulcers. Patient 11 developed severe pneumococcal pneumonia. No cancer was diagnosed during follow-up. Infliximab was stopped for patients 2 and 11 because of side effects.
DISCUSSION
The results of this retrospective study on 15 patients with different refractory SNV support the efficacy of infliximab, which rapidly obtained disease remissions in 11 patients and therapeutic responses in the four others. Despite the heterogeneity of treatments, infliximab was the only new drug for 11 of the patients until D45, thereby suggesting that the clinical improvement was attributable to it. This 73% remission rate agrees with 83–100% reported for other trials that used infliximab to treat refractory SNV.4 5 11
Although immunosuppressive therapy was maintained (in all but one patient), 10 of our 15 patients relapsed while receiving infliximab or after its withdrawal, followed by its successful reintroduction for two patients. Among the four patients in sustained remission who did not relapse, two are still receiving infliximab. These observations could evoke a “holding” effect of this TNFα blocker on SNV or, alternatively, reflect the particularly aggressive disease course in this subset of patients. The relapses in our patients might have been delayed or prevented by continuing infliximab administration, but the increased incidence of solid tumours in WGET trial patients treated with etanercept for a median duration of 25 months6 raised questions about the safety of long-term TNFα blockage in immunosuppressed patients. However, for other chronic inflammatory diseases—for example, rheumatoid arthritis,12 it has not yet been clearly demonstrated that use of these agents puts patients at risk for malignancies.
The different responses of Crohn’s disease to the soluble TNFα receptor etanercept and the monoclonal anti-TNFα antibodies—for example, infliximab, indicated the possibility of their differential efficacies.13 Unlike etanercept, infliximab can bind to circulating and transmembrane receptor-bound TNFα, and has better affinity and avidity, resulting in longer binding to TNFα.14 Moreover, infliximab linkage triggers apoptosis of cells in established granulomas, thereby explaining the higher risk of tuberculosis reactivation with infliximab use.15
As in our previous report,3 we evaluated infliximab efficacy using BVAS9 as an outcome measure. According to the original BVAS version used, only clinical features which either newly develop or worsen during the month before the assessment are supposed to be rated, thereby potentially underlining the presence of persistent, “grumbling” disease manifestations.16 Therefore, our findings should be viewed while keeping in mind the fact that the observed effect size may have been overestimated with respect to persistent low-grade SNV manifestations—for example, ear, nose and throat signs, which did not disappear under infliximab but did not require a change of treatment. Another shortcoming of our study lies in its retrospective design and the heterogeneity of the vasculitides investigated which might have diversely influenced the responses to infliximab seen herein.
To conclude, based on ⩾2 years of follow-up after starting infliximab in patients with refractory SNV, these observations highlight the potential usefulness of this biological agent to induce disease remission. Salvage treatment with monoclonal anti-TNFα antibodies should be further evaluated in this subset of patients with difficult-to-treat SNV.
REFERENCES
Footnotes
Competing interests: None.