Increasing The Dose Of AraC In Consolidation Therapy Does Not Lead To Higher Overall Survival Or Improved Relapse Incidence In Patients With AML Over The Age Of 60 Years

The optimal consolidation chemotherapy in AML patients >60 years has yet to be defined in detail. Although age-adjusted induction chemotherapy results in CR rates comparable to those in younger patients, relapse remains the major hurdle to successful treatment. While the role of stem cell transplantation (HSCT) in elderly patients is currently being evaluated in randomized studies, we focus here on the intensity of consolidation chemotherapy.

Patients data from the elderly AML trials OSHO 1997 (n=410) and OSHO 2004 (n=733) were pooled and analyzed. These protocols have identical inclusion/exclusion criteria and induction chemotherapy, but differ in the intensity of consolidation therapy. In the OSHO 1997 trial, Ara-C 120 mg/m² bid was given from day 1-5 and mitoxantrone 10 mg/m² from day 1-2 as consolidation. In the OSHO 2004 an intensified consolidation using Ara-C 500 mg/m² bid on day 1/3/5 was applied together with mitoxantrone as used in the OSHO 1997 study. Of the 1143 patients, 689 entered CR (60% in the OSHO 1997 and 61% in the OSHO 2004) and 536 (OSHO 1997, n=242, OSHO 2004, n=294) did not receive HSCT as consolidation. The analysis concentrated on the dose of AraC used in the consolidation for this elderly population and on the cycles of consolidation applied. Patient characteristics were compared using chi-square test for categorical data and Wilcoxon rank sum test for continuous data. OS was analyzed using the Kaplan-Meier method, and univariate comparisons were made by means of the log-rank test. Cox regression was used to find any association between consolidation chemotherapy considered as a time-dependent covariate on Overall Survival (OS) or Relapse Incidence (RI). RI and Non Relapse Mortality (NRM) were calculated using the competing risk method, and the Gray test was applied to compare differences. Multivariate modeling was performed by Cox regression analyses with a forward selection method.

Median ages in the AML studies were 66 (60-81) years and 69 (60-85) years for the OSHO 1997 and OSHO 2004, respectively. Patients characteristics were balanced except for age and Karnofsky score (p< .0005) and a trend towards more intermediate and high risk karyotypes, more female and less WBC in the OSHO 2004 compared to the OSHO 1997 study (p=0.06). OS at 15 years was 14±2% in all patients with no difference between the two consolidations, but strong dependence on cytogenetic risk factors. In multivariate analyses risk factors for survival were high/intermediate risk karyotypes, male gender, non de-novo AML and less than two consolidations. Patients with two consolidations had better OS than patients with one or no consolidations in the pooled group and in each of the two protocols with no difference between OSHO 1997 and OSHO 2004. Relapse incidence amounted to 79±2% and NRM 10±04% at 15 years with no difference between the two protocols. Relapse incidence was dependent upon cytogenetic risk and the number of consolidations applied in a multivariate model. There were no risk factors predicting TRM in multivariate analysis. Our analysis of patient characteristics according to the number of consolidations showed the distribution of consolidation therapies to be 15.2%, 28.0%, 56.6% and 14.2%, 32.3% and 53.4% for 0, 1 and 2 consolidations in the OSHO 1997 and OSHO 2004 respectively (n.s.). Higher age, higher risk cytogenetics, non-de novo AML type, less CR after one induction cycle and lower WBC count at diagnosis were characteristic of patients receiving none or one as compared to two consolidation therapies. The multivariate analysis revealed cytogenetics and gender as independent risk factors, but not the application of one as opposed to two consolidation treatments.

The increase of AraC dose in the OSHO 2004 was unable to either increase survival or improve relapse incidence in the cohort of elderly patients. TRM was not different between the OSHO 1997 and 2004 studies. However, the application of one or two consolidation cycles had a significant impact on survival that was not due to decreased relapse incidence after normalization for risk factors. Interestingly, just above 50% of patients received 2 consolidations as proposed in the protocol with no statistically significant difference between OSHO 1997 and OSHO 2004. Patients receiving fewer consolidation therapy cycles are older, have more non-de novo AML and lower WBC count.