Abstract
Among the entire patients with AML the majority is 60 years of age or older. In present German multicenter AML Cooperative Group (AMLCG) trial the proportion of these older patients amounts to 54% of all 2734 patients enrolled and receiving intensive chemotherapy. While older age AML is increasingly recognized as a main challenge the therapeutic outcome unlike that in younger patients has remained constantly poor. Thus, the patients of ≥ 60y show an overall survival (OS) of 13% and a relapse rate (RR) of 82% at 5y versus 40% and 52% in younger patients. Age related differences in treatment and in risk profiles are commonly used to explain the differences in outcome. In the AMLCG 99 trial including 2734 patients 16 to 85 (median 61) years of age we investigated factors determining the disease biology and outcome. For induction treatment patients received standard dose TAD and high-dose AraC 3 (age < 60y) and 1 (≥ 60y) g/m² × 6/mitoxantrone (HAM) or randomly HAM-HAM, for consolidation TAD, and for maintenance monthly reduced TAD randomized (in patients < 60y) against autologous SCT. When compared with patients younger than 60y older patients had more frequent secondary AML (29% vs 17%, p< 0.0001), unfavorable cytogenetics (29% vs 23%, p= 0.0004), less frequent favorable cytogenetics (4% vs 12%, p< 0.0001), and NPM1mut/FLT3-ITDneg status (26% vs 34%, p< 0.009) in those with normal karyotype, and overall even lower median WBC (7.360 vs 12.600/μl, p< 0.0001) and LDH (340 vs 413 U/l, p< 0.0001). A multivariate analysis identified independent risk factors determining therapeutic endpoints such as CR rate, OS, RR, and RFS. With similar results across all endpoints, risk factors for OS were age ≥ 60y (HR 1.96, 95% CI 1.75–2.17), AML secondary to MDS or cytotoxic treatment (1.28, 1.14–1.45), unfavorable karyotype (2.17, 1.92– 2.44), WBC > 20×10³/μl (1.15, 1.02– 1.30), LDH > 700U/L (1.32, 1.15– 1.52), favorable karyotype (0.49, 0.38– 0.63) and female gender (0.90, 0.81– 0.99). In the 891 patients with normal karyotype and complete mutation status risk factors for OS were age ≥ 60y (2.00, 1.64– 2.44), and NPM1mut/FLT3-ITDneg (0.39, 0.30– 0.49). Risk factors for RR overall were age ≥ 60y (2.04, 1.75– 2.38), unfavorable karyotype (2.08, 1.47– 2.13), LDH (1.41, 1.16– 1.72) and favorable karyotype (0.40, 0.29– 0.56). In patients with normal karyotype and complete mutation status risk factors for RR were age ≥ 60y (2.00, 1.56– 2.63), and NPM1mut/FLT3-ITDneg (0.32, 0.23– 0.43). Testing the role of older age in favorable subgroups, the 198 patients with CBF leukemia show an OS at 5 years of 27.5 (95% CI 12.0– 43.0) % in the older versus 69.4 (60.7– 78.2) % in the younger age group, and a RR of 56.6 (35.7– 77.3) % versus 25.0 (15.6– 34.4) %. Comparatively, the 264 patients with a normal karyotype and NPM1mut/FLT3-ITDneg show an OS of 37.1 (26.6– 47.5) % in the older versus 71.9 (63.4– 80.4) % in the younger age group, and a RR of 61.0 (47.8– 74.2) % versus 23.0 (14.0– 32.0) %. There was no influence by randomized treatment variables on any therapeutic endpoint.
Conclusion: Considering the prognostic spectrum of all major historic or genetic subgroups older age maintains its dominant role not explained by age related differences in risk profiles. Even within CBF leukemias and sole NPM1 mutation as the best prognostic categories older age predicts for markedly shorter OS and higher RR. Thus, understanding older age AML requires further genetic and epigenetic work.