In this issue of the Journal of Clinical Oncology, three groups report their results with rituximab monotherapy in B-cell lymphoma. Following the approximately 50% response rate in recurrent or refractory indolent lymphoma reported in the pivotal trial, investigators have studied alternate dosing schedules in untreated and previously treated patients and have proposed both host and tumor characteristics that may predict the quality of response. How are the reports in the current issue of the Journal informative for current practice and clinical investigations? Witzig et al administered four weekly rituximab infusions to 36 previously untreated, follicular grade 1 lymphoma patients who, in their judgment, required no immediate therapy. Although the primary objective was to estimate the response rate, an important secondary end point was the time to initiating chemotherapy after rituximab treatment. The alternative of rituximab to the conservative approach of observation alone as initial management for selected patients may be considered advantageous if quality of life is improved or if the time to requiring chemotherapy is prolonged. The overall response rate in the North Central Cancer Treatment Group study was 72%, with a median time to progression of 2.2 years from registration—longer than previously reported in treatment-naive patients. The time to subsequent chemotherapy was, however, unexpectedly brief, at just 2.3 years after rituximab or approximately 2.4 years after registration. In two prior randomized phase III studies comparing observation to immediate therapy, time to initiating treatment was 2 and 2.4 years, respectively, after initial observation. In the study by Brice et al, the time to treatment failure was 3.6 years following 3 months of initial interferon alfa. Clearly, it can be hazardous to draw conclusions from a small patient sample, as reported by Witzig et al, in a disease as heterogeneous as follicular lymphoma. Perhaps these patients had more biologically aggressive disease than predicted at diagnosis, or perhaps the treating physicians had a lower threshold for administering chemotherapy, as the North Central Cancer Treatment Group did not establish standard criteria for initiating chemotherapy. It is also important to recall that the time to progression or time to initiating chemotherapy may be as much a function of the underlying biology of indolent lymphoma as the effect of prior treatment. Finally, the potential therapeutic benefit of rituximab was not maximized by the study design used by Witzig et al, which did not include re-treatment of responding patients. Previously, a 40% response rate was reported in follicular lymphoma patients with disease progression after initial response to rituximab and were subsequently re-treated weekly for 4 weeks. The goal, therefore, of rituximab monotherapy might then be best expressed as prolonging the time to requiring alternate therapy, allowing for rituximab re-treatment. Hainsworth et al studied this end point in a prospective trial in previously treated, but rituximab-naive, indolent B-cell lymphoma patients. The trial was a randomized phase II study, in which patients received four weekly rituximab infusions, with assessment of response at 6 weeks and random assignment of stable or responsive patients to maintenance rituximab delivered as four weekly treatments every 6 months, for a maximum of four courses or re-treatment (standard 4-week course) on disease progression. The purpose of this type of study is to evaluate various approaches so that the most promising can be selected for further study. Randomized phase II studies do not have adequate statistical power for definitive treatment comparisons, but seek to estimate treatment effect and toxicity in each arm, without patient selection bias. The primary study end point, duration of rituximab benefit as defined by the administration of alternative treatment, was not different between the two approaches in this trial. However, these results are only estimates, and imbalances in histology and the number of patients progressing before either maintenance or re-treatment may have influenced the results. In addition, rituximab benefit was based on a subjective end point—the requirement for alternate therapy—rather than an objective definition of response or duration of response. As in the Witzig et al JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 23 NUMBER 6 FEBRUARY 2