A Randomised Comparison of Clofarabine Versus Low Dose Ara-C As First Line Treatment for Older Patients with AML

There is no adequate standard of care for older patients with AML who are not considered fit for intensive treatment. In recent years a number of potential novel agents have been tested in single arm studies, but have not been validated in a randomised assessment. Clofarabine is a novel nucleoside analogue which has shown encouraging activity in this patient group in single arm studies [1,2]. As part of our “Pick a Winner” (PaW) trial programme, Clofarabine (CLO) 20mg/m² IV days 1 to 5 was compared to Low Dose Ara-C (LDAC) 20mg bid SC days 1 to 10, each for up to 4 courses, with patients deriving benefit allowed to continue on treatment. The rules of PaW [3] required a confidential interim analysis after 50 & 100 randomised patients per arm to confirm that the aim of doubling the CR rate to >30% was likely to be achieved, in order to proceed to phase 3 with overall survival as the primary endpoint.

Between August 2006 and April 2011, 406 patients entered the randomised comparison from 109 centres in the UK, Denmark and Australia with a median age of 74 yrs (range 51–90). Sixty-two per cent had de novo disease, 24% secondary AML and 14% high risk MDS (>10% blasts): 2% had favourable cytogenetics, 72% intermediate and 26% adverse. By Wheatley Score [4], 3% were good; 46% standard and 51% poor risk; 13% had WHO performance score >2.

Overall 28% entered CR/CRi and 28% and 13% were alive at 12 and 24 months respectively. The median number of courses given was 2.0 (range 0–8) in both arms. Of the remitters 34% are alive at 2 years compared with 4% of non-remitters. At the two interim analyses 34% and 41% were in CR in the CLO arm compared with 15% and 21% in LDAC. The trial therefore recruited to full accrual. In the final analysis the CLO arm had a superior response rate (CR + CRi) of 38% vs 20% (CR 22% vs 12%, CRi 16% vs 8%; OR 0.41 (0.26-0.62); p<0.0001). The 30 and 60 day all-cause mortality was higher in the CLO arm (18 vs 13% and 32% vs 26%). RFS (2 yrs) was non-significantly better in the CLO arm (20% vs 8%, HR 0.76 (0.49-1.19), p=0.2) but for those achieving CR the survival was non-significantly better in the LDAC arm (44%vs 26% HR 1.19 (0.74-1.91), p=0.5), which is partly explained by a significantly superior survival from relapse in the LDAC arm (8% vs 0%, HR 1.91 (1.10-3.31), p=0.02). The lack of OS benefit in the CLO arm is also partially explained by a better survival for those in the LDAC arm who did not achieve initial CR (HR 1.37 (1.06-1.76), p=0.02).

The CLO arm had more reported grade 3/4 toxicities of nausea, diarrhoea and liver biochemistry in course 1 & 2, while LDAC had more reports of grade 3/4 cardiac toxicities. The increased myelosuppression in the CLO arm was reflected in an increased use of RBC and platelet transfusions, and more days on antibiotics and in hospital.

In an analysis stratified by demographic, cytogenetic and molecular (FLT3/NPM1) subgroups, clofarabine consistently resulted in a superior remission rate. However, with respect to overall survival, there was no evidence of any subgroup benefit.

Clofarabine was shown to double the response rate in older patients compared with LDAC, but did not result in an improvement in survival overall, or in any demographic or risk subgroup. CR may not serve as a reliable surrogate for overall survival in this patient group

This study received research support from Genzyme and Cancer Research UK.

Burnett:Genzyme: Membership on an entity's Board of Directors or advisory committees. Off Label Use: Clofarabine in treatment of AML. Hills:Genzyme: Consultancy.