Concomitant cyclophosphamide and oral immunosuppressants with rituximab for systemic lupus erythematosus.

SIR, Turner-Stokes et al. [1] recently published their experience using repeat cycles of rituximab in SLE. We welcome this addition to the evidence base for rituximab. We recently published our experience using rituximab in SLE in 41 patients [2]. In general, we agree with the observations of the University College London (UCL) group in terms of response rates and time to relapse. However, there are two important differences in our treatment protocol: (i) we did not routinely use concomitant CYC and (ii) we did not discontinue concomitant oral immunosuppressants if taken at baseline. Comparison of the findings of our studies is therefore of interest. We used the original BILAG and the scoring A = 9, B = 3, C = 1 and the UCL group used British Isles Lupus Activity Group (BILAG)-2004 and A = 12, B = 5, C = 1. However, definitions of major and partial response in our article are identical to the definitions of complete and partial remission used by the UCL group (complete remission required reduction of all active domains to BILAG C/D, and partial remission permitted one persistent A/B, provided other domains had improved). Our standard protocol is 100 mg methylprednisolone + 1000 mg rituximab given on days 1 and 15 with a course of prednisolone 60 mg daily on days 1 7 and 30 mg daily on days 8 14. We used CYC in only three patients. Despite using very little CYC, response rates in our study appear as good as those reported by TurnerStokes et al. [1]. Of 38 evaluable patients, 18 were receiving a concomitant immunosuppressant (6 AZA, 8 MMF, 4 MTX), providing the opportunity to compare efficacy and safety with or without these agents (Table 1). Confounding by indication is obviously a major limitation of this comparison, especially given the baseline difference in disease activity. However, there was no evidence of opportunistic or more frequent infections when rituximab was used in combination with a concomitant immunosuppressant in this series. In addition, we noted in our previous report [2] that plasmablast numbers were lower before and after rituximab when concomitant immunosuppressants were used, which was associated with better clinical response and is similar to a previous observation with MTX or LEF in RA [3]. Additionally, a non-serious opportunistic infection (herpes zoster) was observed in a patient shortly after receiving two pulses of CYC with her rituximab therapy and also continued MMF. We understand the reason that concomitant immunosuppressants were discontinued in the UCL series was because of the observation of opportunistic infection in patients who had received concomitant immunosuppressants. We also note the low rates of opportunistic infection (and infection in general) in patients who received rituximab with concomitant immunosuppressants, but not CYC in the Ex-ploratory Phase II/III SLE Evaluation of Rituximab (EXPLORER) trial [4]. Collectively, these observations suggest that the combination of rituximab with both CYC and concomitant immunosuppressants is to blame for opportunistic infection rather than rituximab with concomitant immunosuppressants alone. Controlled trials are required to conclusively determine the best treatment regimen. However, at present rituximab is widely used for SLE based on case series evidence and comparisons of series therefore is of value. We suggest that based on present evidence, rituximab may be used without CYC in many patients. Further, when CYC is not used, continuation of concomitant immunosuppressants is not harmful and may be beneficial.