Safety and efficacy of rituximab in steroid-refractory chronic GVHD

Chronic GVHD (cGVHD) is the most important determinant of late morbidity and mortality in long-term survivors after allo-SCT. Because of the increasing age of recipients, the use of alternative donors and the more frequent use of G-CSF-mobilized PBSCs as stem cell source, cGVHD is increasingly becoming a matter of concern. cGVHD patients who fail first-line therapy have a poor outcome. Recent data suggested that B lymphocytes may have a major role in cGVHD pathophysiology.1 Several authors reported promising findings with the use of rituximab for salvage therapy of cGVHD.2 However, late onset neutropenia (LON) has been suggested by some investigators as a potential complication associated with rituximab therapy, especially in the context of allo-SCT.3 This retrospective study aimed to report the safety features of rituximab salvage therapy in a series of 18 consecutive patients with refractory cGVHD treated between 2008 and 2010 at the University Hospital of Nantes (CHU de Nantes, France). Patients, transplant, graft and GVHD characteristics are summarized in Table 1. Written informed consent was obtained from each patient and donor. The study was performed according to the principles of the declaration of Helsinki. All allogeneic grafts were obtained from HLA-A, HLA-B, HLA-C, HLA-DR and HLA-DQ-matched donors. A single HLA mismatch of 10 (at HLA-Cw) was allowed at the allelic level. HLA-DP typing was not routinely performed at the time of this analysis. Supportive care and antimicrobial prophylaxis were reported previously.4, 5 For GVHD prophylaxis, patients in this cohort received either CsA alone (n=5; 28%) or CsA and mycophenolate mofetil (MMF, n=9; 50%), or CsA and MTX (n=4; 22%). CsA was administered at a dose of 3 mg/kg/day by continuous i.v. infusion starting from day 3 or 2 before graft infusion. MMF was given at a fixed oral dose of 2 g/day. MMF was tapered over 4 weeks starting from day +60 and CsA from day +90 if no GVHD appeared. During the whole study period, supportive care and CMV infection management was the same. Acute and cGVHD were evaluated according to the Seattle standard criteria. Data were computed using the Prism 5.0 (Graphpad Software Inc., La Jolla, CA, USA). In this cohort, the median age was 54 years (range 21–68) with 61% males, and 50% of patients having a myeloid malignancy. At the time of rituximab initiation, all patients presented with extensive and refractory cGVHD, which received and failed (in addition to corticosteroids) several lines of prior immunosuppressive therapy: CYA (n=14), MMF (n=6), imatinib (n=2), dasatinib (n=1), low-dose MTX (n=2) or PUVA therapy (n=3). The median time between cGVHD diagnosis and rituximab therapy was 5 (range, 0.5–38) months. On an intention to treat, patients were scheduled to receive rituximab 375 mg/m2/week for 4 consecutive weeks. All patients received paracetamol plus antihistaminic plus corticosteroids as a pre-medication before therapy to minimize infusion-related manifestations. The median number of rituximab infusions was 4 (range, 1–16). All patients, but four, received the scheduled four infusions of rituximab. One patient (5.5%) had to discontinue treatment after one infusion because of worsening cGVHD, whereas two patients (11%) with a partial response received a second cycle of rituximab according to the attending physician's decision. Finally, one patient (5.5%) completed the four courses treatment and received subsequent maintenance therapy for 6 months (one rituximab infusion per month). In all, we observed a 55% (n=11) overall response rate with one patient achieving a complete response. cGVHD responses were as follows: 8 out of 18 skin, 6 out of 11 oral mucosa, 2 out of 5 musculoskeletal, 1 out of 3 liver, 1 out of 10 gastrointestinal and 1 out of 2 lung responses. Furthermore, 10 patients could reduce (two patients) or discontinue steroids (eight patients). However, four out of these eight patients had to resume corticosteroids within 3 months. At last follow-up, four patients (22%) had died, of whom three had (16%) infections and one (5.5%) died of unknown cause. In this series, with a median follow-up after rituximab therapy of 19 (range, 2.4–37.7) months, four patients (22%) developed grade IV LON, which appeared at a median of 4 months after the last rituximab infusion (range, 1.6–4.9). LON lasted for a median of 14 (range, 6–28) days, but none of these patients developed any serious or clinically significant infection that could be attributed to this LON. The median number of peripheral blood neutrophils (ANC) at the beginning of rituximab therapy was 8365/μL (range, 1500–13860) and 4320/μL (range, 880–13660) at time of last follow-up (Figure 1). Moreover, no other toxicities that could be attributed to rituximab were observed, suggesting that rituximab is an efficient and safe salvage treatment in patients with steroids-refractory cGVHD. The pathophysiology of cGVHD is still poorly understood. As already observed in some autoimmune disorders, the beneficial effect of B-cell depletion with rituximab highlights the potential primary pathologic role of B lymphocytes in the development of cGVHD.6 Furthermore, in some studies, circulating autoantibodies could be observed and were associated with different forms and activity of cGVHD.1 In the light of this altered B-cell homeostasis in cGVHD, the role of rituximab in controlling the clinical manifestations of cGVHD has been studied.7 In our study, an overall response rate of 55% was observed, especially at the skin, oral mucosa and musculoskeletal levels. Also, one of the two patients with a pulmonary cGVHD achieved a significant improvement of his pulmonary function as shown by the improvement of his lung function tests. Moreover, 71% of the patients could reduce or discontinue steroids, though four out of these eight patients who discontinued steroids had to resume corticosteroids therapy within 3 months in line with previous findings from Zaja et al.6 As patients after allo-SCT often have long-lasting B lymphopenia, Bonin et al.8 preferred to administer low-dose rituximab (50 mg/m2) and were still able to observe responses in steroid-refractory cGVHD. In our series, despite the use of a full dose of rituximab, only 22% of patients developed grade IV LON, but none of these patients developed any serious infection that could be attributed to LON. The incidence of LON described in different studies was generally in the range of 3–27%. Many studies demonstrated that rituximab use after allo-SCT was an independent risk factor of LON, and LON correlated with increased occurrence of non-severe infectious events after allo-SCT.9 In contrast, McIver et al.10 found prolonged and life-threatening cytopenia occurred when rituximab was infused within 6 months of T-cell-depleted allo-SCT. In summary, this series suggest that rituximab used as salvage therapy for refractory cGVHD has a promising efficacy and an acceptable safety profiles without excessive LON, warranting further prospective assessments. Mohamad Mohty and Steven Le Gouill received lectures honoraria and research support from Roche whose product is discussed in this study. The other authors report conflict of interests that have a direct link with the subject matter or products discussed in this study. We would like to thank the nursing staff for providing excellent care for our patients. We also thank the ‘Région Pays de Loire’, the ‘Association pour la Recherche sur le Cancer (ARC)’, the ‘Fondation de France’, the ‘Fondation contre la Leucémie’, the ‘Agence de Biomédecine’, the ‘Association Cent pour Sang la Vie’ and the ‘Association Laurette Fuguain’, and the IRGHET for their generous and continuous support for our clinical and basic research work.