Outcome of Adult AML at First Relapse Following a Risk-Oriented Strategy: An Update of the Northern Italy Leukemia Group (NILG) Experience

INTRODUCTION The impact on post-relapse survival of selected prognostic factors and salvage therapy (finalized to perform an allo-SCT) was retrospectively analyzed in 172 patients (patients) with relapsed non-APL AML, who had been initially treated with standard induction and risk-adapatiented consolidation. The aim was to identify factors associated with a better outcome at first relapse.

METHODS All 172 patients were at first recurrence following consolidation of CR1 with

• high-dose Ara-C (HiDAC) multicycle therapy supported by blood stem cells (standard risk, as defined by mixed clinical-cytogenetic criteria) or

• allo-SCT in case of high-risk prognostic profile.

Median age at relapse was 55 y (range 21–70). CR1 duration was <6 months in 50 patients (29%), ranging from 0.6 to 52,7 mo (median 9,1). High risk patients were 128/172 (74%) and 43/172 patients (25%) had an unfavourable cytogenetics (CG). One hundred-eleven patients (64%) received HiDAC and 24 (14%) an allo-SCT according to study design.

RESULTS 140 patients (81%) received salvage treatment. The remaining 32 patients (19%) received palliation and all of them died. The median OS was 17.1 mo, with a 2yOS of 34%. Favorable prognostic factors identified by univariate analisys were: favourable or intermediate CG (p=0,007), standard risk category according to first line protocol (p=0.004), availibility of a HLA matched donor (p= 0.048), achievement of an early CR1(p=0,000), HiDAC as first line therapy(p=0,000), alloHSCT perfomed at relapse (p=0,000) and a DFS from CR1>12 mo (p=0,000). In multivariate analysis favourable or intermediate CG and DFS >12 mo were confirmed as independent prognostic factors (p=0,036 and p=0,001 respectively). Among the 140 patients, 50 received an allo-SCT following relapse (36%, group 1), and the remaining 90 (64%, group 2) received high dose chemotherapy alone (85), autologous SCT (2), or DLI (3, in case of previous alloSCT). Both groups were comparable regarding age >55 y, prior allo-SCT and risk class at diagnosis. After salvage therapy, 44 patients(88%) in the group 1 achieved CR2, compared to 26 patients (29%) in the group 2. The median duration of CR2 was 9 mo (range 2–64) and 3 mo (range 1–34) in group 1 and 2 respectively. NRM was 17/140: 12 patients (24%) in the allo-SCT group and 5 (6%) in group 2. The 2yOS was 57% and 23% respectively (p=0,000). Moreover, among 50 alloSCT patients, survival was affected by risk category at diagnosis: 2yOS of 19 (38%) standard risk patients was 83% compared to 42% in 31 high risk patients (62%) (p=0.01). This risk stratification has no impact on OS in the group 2.

CONCLUSIONS DFS > 12 mo and standard risk category at diagnosis, according to NILG protocol, are the most important independent positive prognostic factors impacting OS of AML relapsed patients. The availibility of a HLA matched donor and a subsequent intensification with alloSCT may offer substantial salvage rates and its outcome is affected by the risk stratification at diagnosis. Nevertheless, high risk patients could benefit from alloSCT, reaching an 2yOS of 42%.