Introduction

Gemtuzumab ozogamicin (GO, Mylotarg) in AML treatment remains controversial: some studies report significant survival benefits either overall or in patient subgroups, but one study showed no, with an excess of early mortality in the GO arm. The five trials so far reported had the common aim of augmenting standard induction chemotherapy which lends itself to a meta-analysis, to determine whether the totality of the evidence demonstrates any benefit for GO, and/or in which specific subgroups of patients. We therefore performed an individual patient data (IPD) meta-analysis of these trials. An IPD meta-analysis has considerable advantages over a published data meta-analysis. Analyses can take into account updated data, unpublished trials, and also look at stratified analyses to identify subgroups of patients.

Methods and Materials

The 5 trials combining GO with the first course of standard induction chemotherapy were identified, the SWOG-0106 study (ClinicalTrials.gov Identifier: NCT00085709), the ALFA-0701 trial, the UK MRC/NCRI AML15 & 16 trials, and the GOELAMS AML2006IR trial. The last of these is so far unpublished. Updated datasets were requested from all trialists. A total of 3339 patients, ranging in age from 0 to 84 were recruited. Information was requested on age, sex, de Novo/Secondary disease, performance status, cytogenetics, FLT-3 ITD, NPM1 status, survival, relapse, remission, and stem cell transplant. Outcomes were assessed using standard meta-analytic techniques, where patients within the same trial were compared and then summary statistics combined to obtain overall estimates. Heterogeneity between trials, and interactions between treatment and baseline covariates were explored using standard heterogeneity and trend statistics.

Results

The addition of GO did not change remission rate (OR 0.91, 95% CI 0.77-1.07, p=0.3). However, there was a suggestion of slightly greater early mortality (30-day mortality HR 1.28 (0.97-1.70) p=0.08. Overall survival was improved (HR 0.89 (0.82-0.97) p=0.01), due to a significant reduction is relapse (HR 0.80 (0.72-0.89) p=0.00006), leading to significantly improved survival from remission (HR 0.84 (0.76-0.94). p=0.001). In all cases there was no significant heterogeneity between trials, although there was a suggestion that the excess early mortality was greater in patients given higher dose of GO. In analyses stratified by baseline covariates, there was no evidence of significant interaction between either age or FLT-3 ITD status and the effect of GO on survival (p=0.9 for trend by age, p=0.9 for heterogeneity by ITD mutation status). However, there was a highly significant interaction between cytogenetic risk group, as defined by each cooperative group, and treatment (p=0.003 for trend, p=0.006 for heterogeneity). There was highly significant benefit for patients with favourable (HR 0.50 (0.33-0.77) p=0.001) and intermediate (HR 0.85 (0.76-0.96) p=0.007) risk patients, but no evidence of benefit for patients with adverse risk cytogenetics (HR 1.04 (0.86-1.25) p=0.7).

Conclusions

The combined evidence from 5 randomised trials of GO given with course 1 of induction chemotherapy shows a significant benefit in survival which more than outweighs any possible increase in early mortality. However, the benefit appears to be restricted to patients who have favourable or intermediate risk cytogenetics, with the greatest benefit in patients with a core binding factor leukaemia, although survival is still highly significantly improved in intermediate risk patients. On the basis of this IPD meta-analysis, adding GO to induction improves survival in the majority of patients with AML with no restriction by age of FLT3 ITD mutation status.

On behalf of the GO trialists collaborative group.

The SWOG-0106 Study was supported in part by the following PHS Cooperative Agreement grant numbers awarded by the National Cancer Institute, DHHS: CA32102, CA38926 and in part by Pfizer Pharmaceuticals.

Disclosures:

Off Label Use: Gemtuzumab Ozogamicin in AML.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution