Prospective Randomized Comparison of Reduced Intensity (FLU-BU-ATG) and Non-Myeloablative (FLU-TBI) Conditioning for Genoidentical Allo-SCT: a Clinical and economical Multicenter Itac Study.

Abstract 193

No definitive data exist defining the optimal myeloablative and/or immunosuppressive association of Reduced Intensity Conditioning (RIC) for allo-SCT. In this perspective, we report the first prospective comparison between 2 widely used conditioning regimens based on reduced intensity or non-myeloablative approaches. Pts were randomized between FBA (Study A) (Fludarabine (30mg/m²/5 days)+Oral Busulfan (4 mg/kg/d over 2 days)+Thymoglobuline (2.5 mg/kg/1day)) (Post graft immunosuppression (IS): CSA) and FTBI (Study B) (Fludarabine (25mg/m²/day over 3 days)+2 Gy TBI) (Post Graft IS: CSA+MMF). Primary endpoint was one-year overall survival (OS). Inclusion criteria were: hematological malignancies, pts non eligible for myeloablative allo SCT, age between 18 and 65, suitable HLA identical sibling, written informed consent. 139 pts were randomized and treated between 2003 and 2008 (Group FBA: N=69; group FTBI: N=70) at 4 transplant centers. The 2 groups were comparable in term of pts characteristics; Median age 54 (21–65); Male gender: 65%; Diagnosis: acute leukemia 18%; NHL 23%; MM 39% others 20%; Disease status: only 32% of the 139 pts were in CR while 68% had measurable disease (PR and stable disease=60%; refractory disease: 8%). Graft failure was documented in 4 pts (6%) in FTBI group. Cumulative incidences (CI) of grade >= 2 aGVHD and cGVHD were respectively: 37% (Group FBA 51%; Group FTBI 26%; p=.003) and 77% (Group FBA 79%; Group FTBI 76%: p=NS). At 1 year, PFS differed (Group FBA 0.68 [0.56 – 0.78]; Group FTBI 0.51 [0.39 – 0.62]; p=0.048) while OS was similar (Group FBA 0.75 [0.63 – 0.84]; Group FTBI 0.74 [0.62 – 0.83]; p=NS). With a median follow-up of 39 months (3–71), 72 pts were alive (Group FBA: 35; Group FTBI: 37: p=NS) with a 5 year OS probability estimate of 0.45 [0.31– 0.57] and 0.49 [0.35– 0.61] for groups FBA and FTBI respectively (p=NS). 53 pts were progression free with a 5 year PFS probability estimate of 0.35 [0.22– 0.48] for group FBA ,and 0.23 [0.10– 0.38] for Group FTBI (p=NS). Median PFS were 26.3 (IC95%:13.6 – 47.3) and 13.1 (IC95% : 7.4 – 25.6) months (mths) in groups FBA and FTBI respectively. More relapses/progressions occurred in group FTBI (p=.005) with a 5 year relapse/progression cumulative incidence (CI) of 0.28 [0.16– 0.40] for group FBA and 0.50 [0.39– 0.60] for Group FTBI. Three pts died from secondary cancers (Group FBA: 1; Group FTBI: 2) and 38 from transplant related causes with a 5 year TRM CI of 0.37 [0.25– 0.49] for group FBA and 0.24 [0.14– 0.34] for Group FTBI (p=0.199). QOL was assessed over a 1-year period with the EORTC QLQ-C30 questionnaire. FBA regimen had a stronger negative impact on patients' QOL during the treatment administration which resolved 80 days after the SCT. Detailed economic analysis was included in the clinical trial. Preliminary evaluation of medical direct costs (conditioning regimen, transfusions, hospitalisations and anti-infectious drugs consumption) demonstrated a crude advantage for the FTBI group (66,711€ vs 42,080€ for the FBA and FTBI groups respectively, p<0.001). The cost-effectiveness ratio using PFS as endpoint was 22,392 € per year of life free of relapse gained using FBA conditioning regimen when compared to FTBI.

In conclusion, this study establishes that, these 2 regimens produce similar 1 year OS. However, FBA is associated with better 1 year PFS and socially acceptable cost-effectiveness ratio but worse early QOL. FBA is also associated with better long term disease control, whereas FTBI tends to produce lower TRM and higher rejection rates. At 5 years, both OS and PFS appear to be similar in this population of rather old patients with advanced hematological malignancies. Cost-effectiveness analysis using OS as effectiveness criterion and including the cost evaluation of relapse treatment is ongoing. Overall, these results contribute to clarify some previously unanswered issues. Clinical data might help designing individual and optimal strategies for each candidate patient, based on factors that predict the probability of relapse and TRM while economical data may help hospitals to tailor their transplant program, depending on the patient population that they care for.