Introduction

Cyclophosphamide (CYC) and glucocorticoids have been the standard therapy for remission induction for anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV) for many years. Before this treatment, patients with granulomatosis with polyangiitis (GPA) had a median survival of 5 months, with a 1-year mortality of 82% and a 2-year mortality of 90% [1, 2]. With the use of CYC and glucocorticoids along with adjunctive treatments such as anti-hypertensive drugs, and renal replacement therapy, the outcome of AAV has improved dramatically with survival rates of approximately 90% at 1 year [3, 4]. However, not all patients respond to these drugs and disease flares necessitating intensification of treatment even among responders are not rare. Furthermore, serious adverse effects of CYC, such as cytopenia, infertility, infections, bladder injury and cancer are major obstacles for the long-term use of this regimen [1]. Among these adverse effects, infections are responsible for approximately half of the mortality [5, 6]. Thus, attention has focused on alternative treatment modalities with better efficacy and less toxicity.

Rituximab (RTX), an anti-CD20 monoclonal antibody, has emerged as a valuable treatment modality for AAV. The results of two randomized control studies (RCTs) [Rituximab versus Cyclophosphamide for ANCA-Associated Vasculitis (RAVE) and Rituximab versus Cyclophosphamide in ANCA-Associated Renal Vasculitis (RITUXVAS)] showed that RTX was not inferior to CYC for remission induction in newly diagnosed patients [7] or in patients with a severe disease relapse with a similar safety profile [8]. Moreover, RTX seemed to be superior to CYC in patients with severe relapsing disease [8]. However, these studies have some limitations to extrapolate the results to all AAV patients. First, newly diagnosed patients constituted the entire population of RITUXVAS trial and 48% of RAVE trial and both studies excluded refractory patients. Second, RITUXVAS has evaluated the efficacy of RTX only in patients with severe renal disease and RAVE has excluded patients with severe alveolar hemorrhage and renal disease. Third, in RCTs exclusion criteria are used to exclude high-risk patients such as those with comorbidities and poorer prognosis which makes the findings of RCTs difficult to generalize to wide patient populations [9]. Several reports on real life data obtained from observational studies have shed light on the efficacy of RTX in AAV patients with different types of involvement and especially those refractory to conventional treatment modalities [10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33].

In this study, we report our experience with RTX in the management of patients with AAV refractory to conventional treatment. Furthermore, to more clearly understand the real life experience with RTX, we conducted a systematic literature review of non-randomized studies with RTX for remission induction in patients with AAV.

Methods

Observational study

We searched our biologic database and selected AAV patients who were treated with RTX in our clinic between June 2011 and October 2015. Among our 49 AAV patients treated with RTX, 24 had received RTX as first line therapy. Thus, we surveyed the remaining 25 AAV patients who were prescribed RTX due to inadequate response to conventional immunosuppressives. We used a standard form to recover all data on demographic and clinical characteristics of the patients, ANCA status, previous immunosuppressive treatment, concomitant immunosuppressive agents during remission induction and maintenance, corticosteroid dose, adverse events and response to RTX treatment from the charts.

The primary outcome was the proportion of patients achieving complete remission at month 6 which was defined as no disease activity with a daily corticosteroid dose of ≤ 10 mg. Secondary outcomes were corticosteroid dose reduction, improvement of individual organ manifestations and relapse rates during follow-up. Treatment response was also assessed with Birmingham Vasculitis Activity score (BVAS version 3) at the final visit [34]. Relapse was defined as the emergence of new disease activity during remission and requirement of an increase of corticosteroid dose (minor relapse) or addition of another immunosuppressive therapy (major relapse). Improvement of the renal function was assessed with the decrease of creatinine and decrease of proteinuria measured by spot urine or 24-h urine collection method. Additionally, infusion reactions, severe adverse events requiring hospitalization or intravenous antibiotic treatment, occurrence of cancer and minor adverse events were noted.

Systematic literature review

We searched PubMed with the keyword combination “(granulomatosis with polyangiitis OR Wegener OR microscopic polyangiitis OR eosinophilic granulomatosis with polyangiitis OR Churg Strauss OR ANCA-associated vasculitis) AND rituximab” from inception up to June 2017. Observational studies evaluating the efficacy of RTX treatment for remission induction in AAV patients were eligible. RCTs, articles reporting the outcome of RTX treatment for maintenance of remission, articles including pediatric age population, studies reporting on less than five patients and articles in languages other than English were excluded. Two independent reviewers (SNE and GA) screened titles and abstracts. The full text of articles that potentially met eligibility criteria at the first screening stage were assessed for inclusion criteria. Disagreements were solved by a senior author (GH). Data extraction was also done by two reviewers. The quality of the included studies was assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach [35].

Ethical approval

This study was conducted in accordance with the Declaration of Helsinki. The ethics committee of Cerrahpasa Medical Faculty approved the study (174711/2017).

Statistics

Statistical analysis was done using SPSS v.22.0. We used normality tests such as visual (histograms, probability plots) and analytical (Kolmogorov–Smirnov/Shapiro–Wilk’s test) methods to determine whether the data were normally distributed. Descriptive statistics were presented using median (interquartile range) for the non-normally distributed variables. The Wilcoxon signed rank test was used to compare the change in prednisolone dose, creatinine level and the amount of proteinuria between baseline and the end of the follow-up. A p value of less than 0.05 was considered statistically significant.

For the systematic literature review, we retrieved the data that the authors presented in their articles. We did not pool the data for an efficacy analysis due to heterogeneity of the included studies regarding the characteristics of patient populations, concomitant immunosuppressive therapies, RTX protocols, outcome measures, length of the follow-up time, and remission and relapse definitions.

Results

Clinical characteristics of patients

Twenty-five patients (21 GPA, four unclassified AAV) received RTX after an inadequate response to conventional immunosuppressives and comprised the study cohort. All patients met the 2012 Revised Chapel Hill Consensus nomenclature for AAV [36]. The baseline demographic and clinical characteristics of the patients are shown in Table 1.

Table 1 The baseline demographic and clinical characteristics of 25 AAV patients
Full size table

Treatment protocol

RTX was given either intravenously in two weekly doses of 1000 mg (rheumatologic protocol; n = 21) or 375 mg/m2/week for 4 weeks (oncologic protocol; n = 4). The same cycles were repeated at fixed intervals every 6 months thereafter. All patients received premedication with intravenous methylprednisolone 100 mg, oral acetaminophen 1 g, and intravenous pheniramine maleate 45.5 mg before RTX infusions. Additionally, 19 patients (76%) received high-dose steroids concurrently with RTX. This was in the form of iv methylprednisolone 1 gm for 3 days in 5 patients and oral prednisolone 1 mg/kg/day in 14 patients. The remaining six patients received oral prednisolone between 10 and 20 mg daily.

Thirteen patients (52%) were given trimethoprim–sulfamethoxazole for pneumocystis jiroveci pneumonia prophylaxis. Three patients (12%) were vaccinated against pneumococcal bacteria and influenza virus. All patients were screened for hepatitis serology before the initiation of RTX. Two of them had positive anti-HBc IgG but they did not receive further investigation or intervention for this.

RTX was given for a median of two courses (IQR 1–3) during a median follow-up of 13 months (IQR 9–22). The concomitant drugs during remission induction and maintenance of remission are given in Table 1.

Outcome

At month 6, 18 (72%) patients had achieved complete remission, 6 had ongoing disease activity and 1 had died at month 4 due to respiratory failure. Four of the six patients with ongoing disease activity gradually achieved remission at month 12. Thus, there were 22 (88%) patients achieving remission at month 12. The remaining two patients were considered refractory to RTX and received high-dose corticosteroids with cyclophosphamide. One of them died with acute renal failure 14 months later and the second developed end-stage renal disease. Three patients experienced minor relapses after a median follow-up of 15 months (range 14–16) from baseline.

Final visit

After a median duration of 24 months (IQR 17–50), 23 patients were under follow-up, 2 patients had died, 1 due to renal failure and the other due to respiratory failure, as described above. BVAS score was 0 in 18 patients, whereas the median BVAS score was four (IQR 3–6) in the remaining five patients. The median prednisolone dose was 40 mg/day (IQR 25–60) at the beginning and 5 mg/day (IQR 0–25) at the final assessment (p = 0.004). Among the patents with BVAS score of 0, 7 were free of corticosteroids and 11 patients were using less than 10 mg/day.

Renal involvement was found to be stable in six patients, improved in six patients and worse in two patients. The median creatinine level remained stable at the initiation of RTX treatment and at the end of follow-up [1.98 mg/dL (IQR 1.22–3.35) vs 1.47 mg/dL (IQR 0.91–2.65); p = 0.35]. The amount of proteinuria was significantly lower at the end of follow-up [1333 mg/day (IQR 537–2300) vs 611 mg/day (150–1321); p = 0.016). One of the two patients who had been undergoing hemodialysis at RTX initiation had no need for this after RTX.

The last evaluation of patients with other types of organ involvement showed good clinical response in 11 of the 16 patients with lung involvement, 9 of the 10 patients with upper respiratory system involvement, all 2 patients with eye involvement, all 2 patients with peripheral nervous system involvement, 1 of the 2 patients with ear involvement, and both of the patients with cutaneous involvement and central nervous system involvement, respectively. One patient with lung involvement had persistent wheezing and another patient with upper respiratory system involvement had tenderness over paranasal sinuses. One patient had persistent sensorineural hearing loss and four patients with lung involvement showed signs of progression on CT scans.

Safety

RTX was terminated in two patients due to infusion reactions on first and fifth courses, respectively. Four patients were hospitalized while being on RTX, one due to pneumonia and the other three due to disease worsening. The pneumonia occurred after 15 days of RTX and the patient recovered with antibiotics.

Systematic literature review

The systematic literature search yielded 546 articles of which 458 were excluded after reading the titles and abstracts. Fifty-six articles were included after full text review [10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33, 37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68] (Fig. 1). Two articles reported the long-term followup of two previous studies [23, 37]. Data obtained from these 56 studies regarding the demographic characteristics, baseline vasculitis activity score, definition of remission and remission rates with induction, concomitant immunosuppressive therapies other than glucocorticoids and adverse events during follow-up are shown in Table 2.

Fig. 1
figure 1

Flow chart of the study selection process

Full size image
Table 2 Demographic characteristics and outcome of RTX remission induction therapy of the included 56 studies
Full size table

We also retrieved the data on the definition of AAV, the disease status (newly diagnosed, relapsing and refractory), type of RTX regimen, RTX treatment for maintaining remission, definition of relapse, relapse rates, outcome after remission induction and follow-up time in each study and provided the quality of the included studies (see online supplementary Table S).

Overall, 56 studies (46 retrospective and 10 prospective) reported 1422 AAV patients (1098 GPA, 70 EGPA, 191 MPA, 11 renal-limited vasculitis, four unclassified, 48 type of AAV not provided) treated with RTX [10,11,12,13,14,15,16,17,18,19,20,21,22, 24,25,26,27,28,29,30,31,32,33, 37, 39,40,41,42, 44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68]. The majority of the patients (86%) had refractory or relapsing disease whereas 195 patients (14%) had new diagnosis. In 15 studies, RTX was prescribed for a specific manifestation such as diffuse alveolar hemorrhage (n = 1) [43]. ocular involvement (n = 8) [10,11,12, 15, 21, 27, 48, 64], severe renal disease (n = 2) [42, 45], pulmonary nodules (n = 1) [13], orbital mass (n = 1) [19], head and neck involvement (n = 1) [26], or granulomatous lesions (n = 1) [30]. Patients with renal transplantation [59] and elderly [44] patients are evaluated in two retrospective studies. The most common vasculitis assessment tools were BVAS.V.3.0 (18 studies), followed by BVAS/WG (11 studies) and BVAS (8 studies). Remission was defined as a BVAS score of 0 in most of the studies (n = 33) or as no disease activity (n = 15) with variable doses of glucocorticoids. Overall, 753 (53%) patients received immunosuppressive therapy in addition to RTX. Among the studies reporting the details of concomitant therapy, CYC (35%) was the most frequently used drug followed by mycophenolate mofetil (MMF) (16%), azathioprine (AZA) (12.5%), and methotrexate (MTX) (6%). RTX was given alone in 10 studies (147 patients). Among the 1283 patients with available data on RTX regimen, 640 (50%) patients received rheumatologic and 593 (46%) patients received oncologic protocol. Two studies on eye involvement used Foster regimen (eight doses of 375 mg/m2 weekly) in three patients [10, 11] and one study used 4 + 2 regimen (a dose of 375 mg/m2 on days 1, 8, 15 and 22 and two more doses 1 and 2 months after the last administration) in 11 patients [65]. The efficacy of low dose (13 patients) and a single dose of RTX (23 patients) was evaluated in one [50] and two [37, 52] studies, respectively.

The initial remission rates ranged between 7 and 100% but the definition of remission was not uniform across studies. When only complete or partial remission was considered, the remission rates were > 80% in most of the studies with the lowest rate being 37.5%. Regarding specific manifestations or conditions, remission rates were between 87.5–100% in scleritis, 86.5–100% in severe renal disease, 60% in pulmonary nodules, 94% in diffuse alveolar hemorrhage, 88% in head and neck involvement, 37.5% in granulomatous lesions. Remission rates were 97% in the elderly population, 80% in patients with renal transplantation and a study on orbital masses (ten patients) reported 100% improvement in visual acuity and 80% size reduction in mass [19]. However, it should be noted that remission rates were derived from one study each. After remission induction, 395 patients continued to receive maintenance therapy with RTX either according to a fixed schedule (11 studies) or when needed (10 studies). The mean follow-up time was longer than 12 months in 43 studies. Overall, 428 (30%) relapses occurred during the follow-up. Among 1216 adverse events, infections (69%) were the most frequent followed by infusion reactions (5%), cytopenias (4%) and malignancies (3%). There were 57 (4%) deaths with the main reasons being infections (n = 25), cardiovascular complications (n = 5), and refractory disease (n = 4).

Discussion

In the observational study, 72% of our 25 patients with refractory AAV achieved complete remission with RTX at month 6. At month 12, the remission rate reached 88% after a second course of RTX. Two patients died due to refractory disease. One infection requiring hospitalization and two infusion reactions occurred. During a median follow-up of 24 months, only three patients experienced minor relapses. These results are similar to the beneficial results reported in most of the studies summarized in the systematic review.

One unresolved issue is the additional benefit of immunosuppressive therapy such as CYC, AZA, MTX and MMF added to RTX during remission induction. Concomitant immunosuppressives were used in 50% of patients in the studies underlining the uncertainty of this issue. Another uncertainty is the optimal protocol for RTX. The systematic review showed that oncologic and rheumatologic regimens are preferred nearly equally. A retrospective study comparing these two RTX regimens did not find a difference regarding the duration of B-cell depletion and treatment response [67]. A single dose of RTX was evaluated in two small-sized studies, reporting beneficial result in one [37, 52]. Data on the Foster regimen is insufficient since there were only three patients with ocular involvement treated with this protocol [10, 11]. Finally, one single study with low-dose RTX reported 92% complete remission in 12 patients during a median of 7.5-month follow-up [50].

Early case series have suggested that granulomatous manifestations of AAV are less likely to respond to RTX [30, 69, 70]. A recent study also reported nearly double remission rates for vasculitis manifestations compared to granulomatous manifestations [20]. The authors have speculated that the lower response for granulomatous lesions may be due to the predominance of CD4 + T cell-mediated inflammation. On the other hand, there are also other studies reporting beneficial results with RTX in treating retro-orbital masses, head and neck involvement, and pulmonary manifestations [13, 27, 29]. Moreover, Joshi and colleagues compared the response rates between patients with orbital granulomatous disease and those with scleritis, and reported similar results among the groups [48]. These conflicting results indicate the need for further studies to better understand the value of RTX in treating granulomatous lesions especially in refractory patients.

Infections were the most common adverse events in our study and in the systematic review. Although RTX was thought to be safer than CYC, the RAVE trial reported similar safety profiles among CYC and RTX groups [8]. Similarly, severe adverse events were one of the primary outcomes in the RITUXVAS trial and safety profiles were similar between RTX versus CYC-based regimens [7]. However, both studies included complications that were also attributable to the disease, glucocorticoid exposure and study medications [7, 8]. Furthermore, both studies had insufficient duration of follow-up to show a difference regarding malignancy and infertility two well-known long-term adverse events of CYC. A retrospective study including 323 AAV patients with a mean follow-up of 5.6 years looked at whether RTX is a safer option with regard to malignancy. The malignancy risk was 3.10-fold higher in CYC-treated patients compared to the general population whereas RTX therapy was not associated with malignancy [71]. Longer follow-up results of the RAVE and RITUXVAS will throw more light on this subject. On the other hand, there is no doubt about the safety of RTX over CYC in preserving fertility.

The main limitations of our study are its retrospective design and small patient numbers. The limitations of the SLR are the heterogeneity of the included studies regarding the characteristics of patient populations, concomitant immunosuppressive therapies, RTX protocols, outcome measures, and remission definitions.

In conclusion, our experience and data coming from observational studies suggest that RTX may be a good option in patients with refractory disease. We still have few data on the long-term outcome of this treatment and on its efficacy on individual organ manifestations. In addition, more data on different patient populations are needed that would lead to the development of consensus on endpoints, outcome measures and treatment strategies.