First-line treatment of indolent B-cell lymphomas (including follicular[FL], marginal zone [MZL], lymphoplasmacytic [LPL]), and mantle-cell lymphoma (MCL) has undergone a shift after two randomized trials (the Study group indolent Lymphomas [StiL], and the BRIGHT trial) demonstrated noninferiority of bendamustine-rituximab (BR) compared with RCHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone) or RCVP (rituximab, cyclophosphamide, vincristine, and prednisone) chemotherapy. 1,2 However, updates of these trials, as well as some observational data, 3 raised questions about potentially increased risk of secondary malignancies (SM) and late infections after BR. Flinn et al. reported that BR (compared with RCHOP/RCVP) was associated with more profound lymphocytopenia, a significantly higher rate of SM (19% versus 11%, P 5 .022), and a numerically higher mortality from infections, particularly pneumonias. 4 In contrast, no difference in SM between BR or RCHOP (17% and 20%, respectively) was reported in the StiL trial. Furthermore, in the GALLIUM study, which randomized FL patients to rituximab or obinutuzumab in combination with bendamustine or CHOP/CVP, patients receiving bendamustine experienced decreased T-cell counts, more fatal adverse events (5% versus 2%), grade 3-4 infections (23% versus 12%), and SM (9% versus 4%), particularly during postinduction maintenance therapy. 5 Our objective was to examine the incidence of hospitalizations, infections, and SM among patients treated with BR or RCHOP/ RCVP using population-based data from the US Surveillance, Figure S1A). Patients treated with BR were on average older (P 5 .001), and more often had MCL (31% versus 24%, P 5 .003), but there were no significant differences in performance status, comorbidities, or time from diagnosis to treatment between BR and RCHOP/RCVP groups.The incidence of hospitalizations, infections, and pneumonias was lower for BR during the first 6 months of observation (Supporting Information Figure S1B-D). Subsequently, the incidence decreased slower for BR than for RCHOP/RCVP, and consequently patients receiving BR experienced 27% higher risk of hospitalization, 54%higher risk of infections, and 93% higher risk of pneumonia throughout the 2nd year of follow-up (Table 1). The absolute increase in incidence was very small: IR difference during the second year of follow-up was 0.9 per 100 person-months for hospitalizations (95%CI, 0.2-1.6), 0.9 for infections (95%CI, 0.5-1.4), and 0.6 for pneumonia (95%CI, 0.3-1.0).
AJH AJHThe differences disappeared in the third year of follow-up, and were not statistically significant when BR was compared with RCHOP only.With a median follow-up of 2.8 years and total observation time of 3779 person-years, 95 nonlymphoid SMs were observed, in 4% of patients receiving BR, and 6% of those receiving RCHOP/RCVP (Supporting Information Figure S1E). A majority of SMs were carcinomas, with <10 secondary myeloid malignancies. In a multivariable model we found no significant difference in t...