A Prospective, Multicenter Study Assessing Functional and Psychosocial Parameters in AML Patients ≥ 60 Years Receiving Induction Chemotherapy

Background

Older adults affected by acute myeloid leukemia (AML) have mostly dismal outcome. They represent an extremely heterogeneous group, both in terms of fitness and disease biology. Hence, treatment decision-making is complex (Deschler et al., Haematologica 2006), and to date no standardized assessment tool including non-disease variables is available for this often vulnerable patient group (Bhatt et al., Blood 2024).We investigated the prevalence and prognostic value of patient-related symptoms and performance parameters in older AML patients (pts) receiving induction chemotherapy (IC) for overall survival (OS).

Study design

In this prospective study we accrued newly diagnosed AML pts aged 60 years and older receiving IC. Therapy-associated (t) AML, secondary (s) AML and prior MDS treatment were allowed. We performed a battery of assessment tools capturing geriatric symptoms (G8), health-related (HR) quality of life (QoL, EORTC-QLQ C30), psychological resilience (RS-11), anxiety and depression (HADS-A/-D), performance status (Karnofsky index, KI), activities of daily living (Barthel Index) and comorbidity burden (HCT-CI); common disease-related characteristics were also captured. Between May 2012 and December 2020, 82 pts were included, with sequential assessments at baseline (BL), 3 (T1), 6 (T2) and 12 (T3) months. Primary endpoint: overall survival (OS), secondary endpoints: changes in QOL and psychological distress. Biometric analyses included descriptive analyses of assessment changes over time, and uni- and multivariate analyses of BL parameters with Cox proportional hazards models.

Results

Of 82 pts enrolled, 44 were female (54%), their median age was 68 years (range 60-81), median Hb was 8.1 g/dL (range 5.9-15), median platelet counts 46 x10⁹/L (range 4-248), median leukocyte counts 22.4 x10⁹/L (range 0.65-269) with a median of 35% peripheral blood blasts (range 6-97). The median KI at BL was 80% (range 40-100). According to ELN 2010 risk stratification, 29% had favorable, 32% int-I, 17% int-II and 18% adverse risk. Six (7%) pts had sAML and 8 (10%) had tAML. Pts received “7+3” (43%), MICE (33%) or other IC regimes (24%). Almost two-thirds of pts (62%) received allogeneic hematopoietic stem cell transplantation after IC. The median OS of all pts was 22.1 months.

Over time, most of the QoL scales measured by EORTC QLQ-C30 worsened at first compared to BL, but then improved again, e.g. physical functioning (BL score 64.7, T1 -12.0, T2 -9.0, T3 -0.9), role functioning (BL score 54.0, T1 -17.5, T2 -12.9, T3 -7.0) and social functioning (BL score 56.1, T1 -14.0, T2 -13.4, T3 +1.6) as well as for the symptoms nausea and vomiting (BL score 9.6, T1 +9.1, T2 +7.3, T3 -1.2) and appetite loss (BL score 39.6, T1 +15.8, T2 -1.2, T3 -14.0). Global QoL also improved incrementally compared to BL (BL score 45.8, T1 +6.1, T2 +9.4, T3 +11.7). In contrast, persistent and clinically meaningful deterioration was noted regarding cognitive function (BL score 77.2, T1 -8.0, T2 –7.4, T3 -8.7) and resilience (BL score 65.3, T1 -5.5, T2 -7.2, T3 -7.0). Overall, these results are in line with the HRQoL analyses of AML pts ≥ 60 years treated on the EORTC LCG AML21 phase III trial (Efficace et al. Blood 2024). Using the HADS-A/D for evaluation, improvements of anxiety and depression over time were noted (not statistically significant, McNemar test P=0.13). Regarding BL parameters, in multivariate analyses, only role functioning played a relevant role in predicting OS (hazard ratio [HR] 2.90, 95% confidence interval [CI] 0.95-8.91, P=0.062). Regarding disease-related factors, high-risk cytogenetics (HR 2.46, 95% CI 1.06-5.74, P =0.037), tAML (HR 2.48, 95% CI 1.12-5.49, P=0.026) and elevated creatinine (HR 2.62, 95% CI 1.18-5.81, P=0.018) showed an independent association with OS.

Conclusions

Deterioration in QoL scales following IC were reversible at the 12 months timepoint in all parameters except cognitive function and resilience, which indicates a need for specific supportive measures. Among BL QoL parameters, only role functioning appeared to be relevant for outcome, along with several disease-related parameters (cytogenetics, AML type, renal function). This trial was registered at the German Clinical trial register (DRKS00003601) and funded by the German José Carreras Leukemia Foundation (R11/07).