Rituximab was approved by the US Food and Drug Administration largely on the basis of the pivotal trial conducted in 166 patients with follicular/low-grade NHL who had failed a median of two prior chemotherapy regimens (McLaughlin et al: J Clin Oncol 16:2825-2833, 1998). The response rate was 48%, including 6% complete remissions (CR), lasting a median of about a year. Attempts to improve on these results by using rituximab as initial therapy have not provided convincing evidence for improved patient outcome (Hainsworth et al: Blood 95:3052-3056, 2000; Colombat et al: Blood 97:101-106, 2001). The high response rates in the French trial by Colombat and co-workers are related in large part to the favorable nature of the patient population. Moreover, median duration of response was only about a year. In the study from Hainsworth et al mentioned above, response rates when calculated by standard methods were comparable to those seen in relapsed and refractory patients.
A more promising direction has involved the development of combinations incorporating rituximab with chemotherapy or other biological agents. This concept is supported by in vitro data suggesting that anti-CD20 antibodies may potentiate the activity of chemotherapy or biotherapy (Demidem et al: Cancer Biochem Radiopharm 12:177-186, 1997; Johnson and Press: Int J Cancer 85:104-112, 2000; Alas et al: Clin Cancer Res 7:709-723, 2001; Smith et al: Blood 96:338a[abstract 1458], 2000).
One of the first attempts at combining rituximab with chemotherapy was published by Czuczman et al (J Clin Oncol 17:268-276, 1999), in which the antibody was added to CHOP (cyclophosphamide [Cytoxan, Neosar], doxorubicin HCl, vincristine [Oncovin], prednisone). Patients with follicular/ low-grade NHL received six courses of CHOP with two infusions of antibody before course 1 and two following the sixth cycle of chemotherapy, as well as one infusion prior to courses 3 and 5. At ASH, these investigators presented the impressive 6-year follow-up of their 40 patients, 31 of whom were previously untreated (abstract #2519). The response rate was 100%, including 53% complete remissions. The median duration of response was 63.6+ months and the progression-free survival had not been reached, with 21 patients still in remission.
Maloney et al (abstract #3502) from the Southwest Oncology Group (SWOG) conducted a phase II study with a different schedule of CHOP plus rituximab, in which patients with follicular NHL first received six cycles of CHOP. Those who did not progress were treated with four weekly infusions of rituximab. The overall response rate was 72%, with a 54% CR rate. Of note was that the quality of response following CHOP improved in 19% of patients after the ensuing antibody therapy. The overall survival of 95% and progression-free survival of 76% at 2 years were considered encouraging. However, there are no immediate plans to pursue this sequence approach.
CHOP is only one of many regimens with which rituximab is being combined in indolent NHL. Czuczman et al (abstract #2518) also updated their results with fludarabine plus rituximab. The response rate of 90% with 83% complete remissions and responses ongoing in 30 of 36 evaluable patients appear to be comparable to their own CHOP/rituximab data (see abstract #2519). A phase III comparison of CHOP/rituximab with fludarabine/rituximab would be needed to determine which regimen is preferable.
The encouraging phase II data with chemotherapy/rituximab regimens have led to an increasing proliferation of randomized phase III trials. SWOG, in collaboration with the Cancer and Leukemia Group B (CALGB), is conducting a phase III Intergroup study in the United States in which previously untreated patients with advanced-stage follicular NHL are randomized to either CHOP, CHOP/rituximab (Czuczman et al: J Clin Oncol 17:268-276, 1999), or CHOP followed by the radioimmunoconjugate (RIC) tositumomab/iodine-131 tositumomab (Bexxar). This study is critical to determining the optimal up-front therapy for patients with follicular NHL.
Zinzani et al (abstract #3500) presented an interim analysis of patients with polymerase chain reaction (PCR)-positive follicular NHL who were randomized to six cycles of either FM (fludarabine plus mitoxantrone [Novantrone]) or CHOP. Those patients who achieved a CR or PR but were still PCR-positive then went on to receive rituximab. FM appeared to be associated with a higher response rate (95% vs 70%) and complete remission rate (64% vs 42%). Moreover, more patients on the FM/rituximab arm became PCR-negative in the bone marrow and/or peripheral blood. Whether or not these results will translate into prolonged survival will require more patients and longer follow-up.
Hiddemann and coworkers from the German Low Grade Study Group reported on their experience with 147 patients with follicular NHL or mantle cell lymphoma who failed CHOP chemotherapy and then were randomized to four cycles of FCM (the combination of fludarabine, cyclophosphamide, and mitoxantrone), with or without rituximab (abstract #3507). The antibody was administered the day before the chemotherapy. In the interim analysis of the ongoing trial including 80 of the 147 randomized patients, the FCM/rituximab regimen induced more responses (89% vs 53%) and complete remissions (36% vs 15%). Complete and overall response rates were higher in the follicular NHL patients compared with the mantle cell patients, yet the apparent effect of the antibody was seen in both histologies. At this point, there were 5% treatment-related deaths.
Despite impressive response rates, no regimen has yet demonstrated a survival advantage in patients with follicular NHL. Results from the large number of regimens being tested will be difficult to interpret because of differences in patient populations, the chemotherapy regimens, and the manner in which rituximab is incorporated into those regimens. Therefore, after all the patients have been accrued, the question will remain as to which regimen is the best, and whether the combination of chemotherapy plus rituximab is preferable to the sequence of the two modalities. Fewer, larger collaborative trials are needed to address these important questions.