Randomized phase III U.S./Canadian intergroup trial (SWOG S9704) comparing CHOP ± R for eight cycles to CHOP ± R for six cycles followed by autotransplant for patients with high-intermediate (H-Int) or high IPI grade diffuse aggressive non-Hodgkin lymphoma (NHL).

Abstract

8001 Background: This SWOG led intergroup trial investigated the benefit of autotransplant in 1st remission patients with bulky stage II, III, and IV H-Int/High IPI diffuse NHL after 5 cycles of CHOP±R, as previously suggested by European studies. Methods: 397 patients (pts) up to age 65 from 40 sites were enrolled between 9/97-12/07 and induced with 5 cycles of CHOP (n=215) or CHOP-R (after 3/03; n=182). Pts with ≥ PR were randomized between CHOP±R x 1 followed by autotransplant using TBI or BCNU based regimens, vs CHOP±R x 3. Primary endpoints were toxicity, and 2 yr PFS and OS for randomized pts with sample size chosen to detect a hazard ratio between arms 1.50 with 82% power. Results: Of 370 induction eligible pts 253 were randomized to standard (128) or transplant (125) therapies; 35% had High IPI . Reasons for non-randomization included progression (66), and refusal (23). On the transplant arm, 9 never received transplant and 6 had fatal toxicities (lung-3). On the standard arm, 3 did not receive CHOP±R and 3 had fatal toxicities. Forty-five transplant and 67 standard therapy pts have progressed or died: 2 yr PFS was 69 and 56% respectively [hazard ratio: 1.72 (95% CI: 1.18-2.51) p = .005]. Thirty-seven transplant and 47 standard therapy pts have died: 2 yr OS was 74 and 71% respectively [hazard ratio 1.24 (95% CI: 0.81-1.91) p = .32). Based on exploratory analyses, there was a significant differential treatment effect between the H-Int and High IPI pts for both PFS (interaction p value = 0.02) and OS (interaction p value = 0.01): for transplant vs standard arms, the 2 yr PFS was 66% vs 63% for the H-Int IPI pts and 75% vs. 41% for the High IPI pts; the 2 yr OS was 70% vs 75% for the H-Int IPI pts and 82% vs 64% for the high IPI pts. There was no differential treatment interaction by histology (B vs T) for either PFS (p=.43) or OS (p=.53), or for B-cell pts by induction (CHOP-R vs CHOP) for PFS (p=.35) or OS (p=.29). Conclusions: For advanced stage H-Int and High IPI diffuse aggressive NHL, early autotransplant improves PFS for responders, including those induced with CHOP-R, with a stronger outcome seen for those with High IPI grade.