Combining the HCT-CI, G8, and AML-Score for Fitness Evaluation of Elderly Patients with Acute Myeloid Leukemia: A Single Center Analysis

Simple Summary In elderly patients with acute myeloid leukemia (AML), non-relapse mortality (NRM) after intensive chemotherapy and, if necessary, allogeneic hematopoietic stem cell transplantation (HSCT) depends on an accurate initial patient fitness assessment. We determined three scores in 197 patients at diagnosis and subsequently investigated the patients' respective overall survival with each fitness score assignment. The G8-score, the HCT-CI and the AML-score for complete remission (CR) were able to significantly separate "fit" from "unfit" patients, p < 0.001, p = 0.008 and p = 0.001, respectively. Their predictive power was statistically confirmed by univariate analysis and was especially enhanced when combining them. Thus, in the present study, all three scores, simple, fast and, if necessary, also without cytogenetic information, represent significant tools for fitness evaluation before intensive treatment of elderly AML. Background: Accurate assessment of elderly acute myeloid leukemia (AML) patients is essential before intensive induction chemotherapy and subsequent allogeneic hematopoietic stem cell transplantation. In this context, we investigated the capacity of three scores for frailty prediction. Methods: At diagnosis, 197 patients were clinically evaluated for appropriate treatment intensity. In parallel and independently, the G8-score, the Hematopoietic Stem Cell Index (HCT-CI) and the AML-score for CR were determined for each patient and analyzed with respect to overall survival (OS). Results: The G8-score and the HCT-CI were able to significantly separate "fit" from "unfit" patients, p = 0.008. In univariate Cox models, the predictive role for OS was confirmed: for the G8-score (HR: 2.35, 95% CI 1.53-3.60, p < 0.001), the HCT-CI (HR: 1.91, 95% CI 1.17-3.11, p = 0.009) and the AML-score (HR: 5.59, 95% CI 2.04-15.31, p = 0.001), the latter was subsequently used to verify the cohort. In the multivariate Cox model, the results were confirmed for the G8- (HR: 2.03, p < 0.001) and AML-score (HR: 3.27, p = 0.001). Of interest, when combining the scores, their prediction capacity was significantly enhanced, p < 0.001. Conclusions: The G8-, the HCTCI and the AML-score represent valid tools in the frailty assessment of elderly AML patients at diagnosis.