Identifying patients with indolent Acute Myeloid Leukemia: Long term survivors after best supportive care only, results from the pethema registry

Abstract

Abstract Introduction: Most patients with acute myeloid leukemia (AML) who receive palliative best supportive care (BSC) alone experience poor outcomes, with reported median overall survival (OS) below two months. However, a subset of these patients not deemed for active therapy may survive significantly longer (i.e, more than those receiving non intensive chemotherapy approaches), suggesting the presence of a clinically indolent AML (iAML) phenotype. Identifying such patients could have meaningful implications for therapeutic decision-making and trial design. Objectives: We analyze clinical and biological characteristics of patients aged ≥60 years with indolent AML (i.e, receiving BSC only and surviving ≥9 months), comparing them with other AML patients aged ≥60 years receiving BSC only and surviving <9 months. Methods: We conducted a retrospective analysis of the PETHEMA epidemiological AML Registry (NCT02607059), including patients aged ≥60 years diagnosed between 1990 and 2024 who received BSC only. BSC included transfusions and other supportive measures, with or without oral cytoreductive agents (e.g., hydroxyurea, melphalan, mercaptopurine or thioguanine). iAML was defined as OS ≥9 months. Patients alive at last follow-up with <9 months of observation were excluded. Baseline characteristics and outcomes were compared between iAML and non-iAML groups (i.e, those aged ≥60 years receiving BSC only and surviving <9 months). A multivariate logistic regression model was used to identify independent predictors of iAML. Results: Of 2,506 eligible patients (i.e, aged ≥60 years receiving BSC only and follow-up ≥9 months), 192 (7.6%) met criteria for iAML (survival ≥9 months). Median OS was 14.8 months (95% CI, 13.2–16.4) in iAML vs. 0.85 months (95% CI, 0.78–0.93) in non-iAML patients (p<0.001). Compared with non-iAML patients, the iAML group had lower leukocyte counts (median 4.4 vs. 12.5 ×10⁹/L, p<0.001), higher platelet counts (median 82 vs. 49 ×10⁹/L, p<0.001), lower peripheral blood blast percentage (median 14% vs. 29%, p<0.001) and lower bone marrow blast percentage (median 35% vs. 56%, p<0.001). They also showed lower lactate dehydrogenase (LDH) levels (median 360 vs 535 U/L, p<0.001) and higher serum albumin levels (median 3.72 vs. 3.30 g/dL, p<0.001), reflecting a less proliferative disease profile. Performance status was better among iAML patients, with 52.9% having ECOG <2 compared to 32.6% in the non-iAML (p<0.001). Secondary and/or therapy-related AML was more frequent in the iAML group (48.7% vs. 36.7%, p=0.001). Fever at diagnosis, was more frequent among non-iAML patients (37.0% vs 18.4%, p<0,001). No significant differences were observed in sex distribution, creatinine levels, or bilirubin levels. Baseline karyotype was available in 68.5% of iAML patients and 63.5% of non-iAML patients, with no significant difference in the frequency of normal karyotypes (51.1% vs. 60.8% among evaluable patients; p=0.063). Among patients with available molecular data, NPM1 mutations were less frequent in the iAML group (11.9% vs. 23.7%; p=0.078), and FLT3-ITD positivity was low and similar between groups (8.7% vs. 14.6%; p=0.268). In multivariate analysis the following variables were independently associated with iAML: bone marrow blasts <50% (OR 3.36; 95% CI, 1.60–7.06, p=0.001), platelets ≥50 ×10⁹/L (OR 2.64; 95% CI, 1.31–5.32, p=0.007), and albumin ≥3.5 g/dL (OR 2.21; 95% CI, 1.15–4.28, p=0.018). Conclusions:Approximately 8% of elderly AML patients treated exclusively with BSC show unexpectedly long survival, consistent with an indolent clinical course, reflecting that in the absence of more efficacious therapeutic regimens, some patients could benefit from less invasive approaches. Easily measurable clinical parameters such as bone marrow blasts, platelet count, and serum albumin may help identify these patients. Further prospective and molecular studies are warranted to validate these findings and elucidate the biological underpinnings of iAML.