The treatment of elderly patients (pts) with AML remains challenging. High treatment associated mortality using protocols developed for younger patients and high relapse rates for pts reaching CR are frequent causes of failure, while many pts are assessed as ineligible for intensive chemotherapy. Patient registration at diagnosis to check for patient allocation or the use of age-adjusted induction protocols to reduce treatment related mortality may improve the management of these pts. In a prospective German Intergroup Study for patients ≥ 60 years, comparable to a completed study for patients < 60 years (Büchner JCO 2012 in press), the outcomes from two study groups using specific induction and consolidation protocols were compared to a common standard arm (CSA).
By October 2011, 1041 pts had been randomized to the study-specific regimens or CSA in a 9:1 ratio. Eighty four patients (8%) were excluded due to incorrect diagnosis, secondary neoplasias or other reasons. Treatment in the CSA consisted of araC [100 mg/m2 continuous infusion (c.i.) d1-7] and daunorubicin (60 mg/m2 i.v. on d3- 5). A second induction was given if marrow blasts ≥5% on d15. Pts in CR received two consolidations with araC (1 g/m2 i.v. bid on d1, 3 und 5). The OSHO study group (group A) investigated araC (1 g/m2 i.v. bid d1, 3, 5) plus mitoxantrone (10 mg/m2 d1-3) for induction and araC (0.5 g/m2 i.v. bid d1, 3, 5) plus mitoxantrone (10 mg/m2 d1-2) for consolidation, while the AMLCG (group B) analyzed TAD (ara-C 100 mg/m2 c.i. d1,2; ara-C 100 mg/m2 bid i.v. d3-8)-HAM (ara-C 1g/m2bid i.v. d1-3) vs HAM-HAM ± G-CSF in pts with ≥5% blasts and TAD as consolidation followed by maintenance.
Of 957 eligible pts, the median age was 69 (range: 60–87) years (68, 70 and 67 years for A, B and CSA, respectively; p<0.03), 45% were female (with no imbalance between groups) and 61% had de novo AML. Significantly more secondary AML were present in group A than in group B or CSA (A 43%, B 28%, CSA 37%, p<0.0001). Risk factors were unevenly distributed with significantly more favorable cytogenetics in group A (15%) than in group B (7%; p=0.0139). There were fewer patients with favorable molecular markers (NPM1 mut/FLT3 wt) in group B than in group A or the CSA (CSA 36%, A 29%, B 16%, p=0.04). No difference was detected in baseline white blood cell counts (WBC) between the three arms, but there was a trend to a higher serum LDH in group A (p=0.06).
Induction therapy led to CR in 71% and 68% of pts in the standard and study arms respectively with early death rates of 20% and 21%. Nine percent of pts in the CSA and 6% in the study group arms had persistent AML. The results after 90 days are available for 743 patients with a CR rate of 56% in the study arms and 50% in the CSA. At 90 days, 156 patients had died with no difference between CSA and study groups (22.0 vs. 21.0% respectively). Persistent AML was present in 21% of the patients in the CSA, but in only 16% of the study arms.
Univariate (Χ2and Mann-Whitney U-test) and multivariate analyses (logistic regression, Wald test) were performed to identify risk factors.
CR after 90 days was more frequent in pts with de novo AML than in those with secondary AML (60.7% vs. 47.9%; p=0.0007) and also higher in pts with favorable as compared with intermediate and unfavorable cytogenetics (68.1% vs 55.0% vs 48.4%; p=0.0107). Pts in CR after 90 days were younger (mean [95% CI]: 68.3 years [67.9; 68.8] vs 69.4 years [68.8; 70.0]; p=0.0067) and had a lower WBC than pts without CR (27.5 per μL [22.6; 32.3] vs 36.1 per μL [29.7; 42.6]; p=0.0077). LDH was higher in pts without CR after 90 days (641.0 U/l [537.1; 744.8] vs 536.0 U/l [461.3; 610.8]; p=0.0041). The percentage of bone marrow blasts, treatment groups, sex, FAB and NPM1/FLT3 mutation status had no significant influence on treatment outcome at 90 days.
AML diagnosis (de novo or secondary; p=0.0002), cytogenetic risk (p=0.0114), age (p=0.0069) and WBC (p=0.0025) were independent factors influencing the CR rate. Adjusted overall survival (OS) and event free survival (EFS) showed no significant differences between the groups after a median follow up of 33 months.
In conclusion, high CR rates can be achieved in elderly patients with AML. The CR-rate is dependent upon the type of AML (de novo or secondary), cytogenetic risk, age and WBC at diagnosis in a multivariate analysis. No differences have been detected in the CR rates between the three arms to date. Further follow up is needed to detect differences in OS and EFS.
Hoffmann:Novartis Pharma: Research Funding.
This icon denotes a clinically relevant abstract
Asterisk with author names denotes non-ASH members.