Fludarabine and Cyclophosphamide Produces a Higher Complete Response Rate and More Durable Remissions than Fludarabine in Patients with Previously Untreated CLL: Intergroup Trial E2997.

The combination of fludarabine and cyclophosphamide (FC) has been noted to produce a high complete response rates in previously untreated patients with CLL. However, it may be accompanied by increased toxicity. To further evaluate its efficacy and toxicity, a phase III randomized study of FC versus Fludarabine (F) was conducted in patients with previously untreated CLL. The study, which was open to accrual from December 23, 1999 to March 19, 2004, closed with 278 patients enrolled, 141 on the FC arm and 137 on the F arm. Four patients declined to receive protocol treatment, including one who was later found to be ineligible. Five additional patients were also deemed ineligible. All patients with data are included in this analysis (intent to treat). Patients on the FC arm received C 600 mg/m² iv day 1 and F 20 mg/m² iv days 1 through 5, followed by filgrastim 5 mg/kg SC starting approximately day 8. Patients randomized to the F arm received F 25 mg/m² iv days 1 through 5. In April 2004, the ECOG Data Monitoring Committee conducted a planned review at 76% information, and determined that the null hypothesis of no difference in CR rates could be rejected. The Data Monitoring Committee gave permission for the submission of abstracts to ASH. The median age of patients was 62 years (34–86), and the median performance status was 1 (0 to 2). As is expected in CLL, 70% of patients were male (194) and 30% were female (83). At study entry, 56% of cases were Rai stages, 0,1 or 2, while 44% were in stage 3 or 4. 57% of patients received the maximum of 6 cycles of therapy. Toxicity data is available on 127 CF and 125 F patients. There were two deaths due to infection with grade 3 or 4 neutropenia (one in each arm). In the CF arm, 17% of patients suffered grade 4 or higher non-hematologic toxicities, while in the F alone arm, 13% had higher grade toxicities (p= 0.48). Additionally, 17% of patients in the FC arm suffered infections versus 11% in the F alone arm (p= 0.21). Response data was available on 246 of the 278 patients. In the FC arm (125 cases), 28 patients achieved CR (22.4%), 60 patients achieved PR (48.0%) for a total of 88 objective responses (70.0%). In contrast, on the F alone arm (121 cases), there were 7 CRs (5.8%), 53 PRs (43.8%) for a total of 60 objective responses (49.6%). The Fisher exact test for the difference in CR rates gives a p-value of 0.0002, while the test for difference in OR rate was 0.001. Currently, 235 patients are alive and 42 have died. Among the 229 patients with information on time to progression, 131 are alive without progression, 78 have progression, and 20 have died without progression. 58 of the 78 with progressive disease remain alive. The preliminary estimates of the median progression free survival time are 41.0 months for the FC arm, and 17.7 months for the F alone arm ( p <0.001). It is noteworthy that the CR rate of F alone is similar to that reported by the GCLLSG (Eichhorst #243 ASH 2003). In summary, FC is a highly effective and tolerable regimen that produces more durable remissions than F in patients with previously untreated CLL.