Introduction. HDS with autografting has proved to be effective (Corradini et al, J Clin Oncol 2004) and feasible at the multicenter level in FL (Ladetto et al, Blood 2002); in addition, it can be successfully combined with Rituximab (R-HDS) (Magni et al, Blood 2000). Aim of this trial was to compare R-HDS with Rituximab-supplemented CHOP (R-CHOP) in poor risk FL patients (pts) younger than 60 years.

Patients and methods. Between January 2000 and March 2005, 136 patients have been randomized (68 in each arm). Patients were stratified according to histology (grade I or II 101, grade III 35). Patients were eligible if they had an age-adjusted IPI ≥2 (125 pts) or, in the absence of this criterion if they had three or more adverse parameters according to the Italian Lymphoma Intergroup score (11 pts). Clinical features were: median age 50 yrs. (22–60), stage III-IV 98%, elevated LDH 78%, bulky disease 66%, B symptoms 47%, extranodal disease (other than BM) 45%, leukemic disease 11%, ECOG PS >1 47%. R-HDS consisted of: i. 2 APO and 2 DHAP courses; ii. sequential administration of Etoposide (2 g/sqm), 2 Rituximab, Cyclophosphamide (Cy) (7g/sqm) with PBPC collection (in vivo-purged with two Rituximab on day 1 and 11 post Cy); iii. Mitoxantrone (60mg/sqm) + L-Pam (180mg/sqm) with autografting (5–8x106 CD34+ cells/kg). R-CHOP consisted of 6 CHOP courses followed by 4–6 doses of Rituximab as originally reported (Rambaldi et al, Blood 2002). Cross-over was allowed for patients failing R-CHOP. Minimal residual disease analysis with the bcl-2/IgH was planned on BM cells. Patients were assessed on an “intention to treat” basis.

Results. The two treatments arms were well balanced for all the previously described clinical parameters. 68% patients were able to conclude R-CHOP (failure due to progression 29% and toxicity 3%) and 78% R-HDS (failure due to progression 9%, toxicity 5%, poor mobilization 4 %, refusal 4%). Toxic deaths were 4 (2 in each arms); in addition 1 gastric cancer and 1 AML occurred in the R-HDS group. Progressions and non-responders were 35% in the R-CHOP arm and 13% in the R-HDS (p<0.05) with 53% and 82% CR rates, respectively. Current median follow-up is 24 months. Event-free survival (EFS) at 24 months is 41% for the R-CHOP arm and 66% for the R-HDS arm (p< .001). At present there is no difference in terms of OS. We currently have data on salvage treatment in 22 patients failing R-CHOP: 16 of them were treated with R-HDS, with 10 achieving CR. A molecular marker was available in 72% of patients. PCR results at follow-up are available in 27 patients. Molecular remission, defined as two PCR-negative BM samples taken at six months intervals, was observed in 54% of R-CHOP and 77% of R-HDS patients.

Conclusions. R-HDS induces a greater number of CR and ensures a better EFS compared to R-CHOP in this rare and extremely aggressive population of high-risk FL patients. It is unknown if these results reflect a peculiar behaviour of high-risk patients or can be relevant for all younger FCL patients with advanced disease requiring cytoreductive treatments.

Author notes

Corresponding author