Indolent Acute Myeloid Leukemia with Long Survival in Patients Treated with Best Supportive Care Only: A Pethema Registry Study

Abstract

INTRODUCTION Unfit AML patients who receive supportive care (SC) only have dismal prognosis with median overall survival (OS) of less than 2 months. However, a low proportion of these patients receiving a palliative approach may have similar or even better survival than those receiving upfront active therapies, suggesting that "indolent” forms of AML exists. Therefore, identification of such "indolent” AML patients could be relevant, since they could even benefit only from observation and supportive care instead of active treatment, as the median OS in older treated patients is usually less than 9 months. The aim of this study is to identify and characterize "long-survivors” with AML treated only with best supportive care in a large cohort of patients included in the PETHEMA epidemiological Registry (NCT02607059). METHODS Patient records from the PETHEMA AML Registry were used in this retrospective survival study. All AML adult patients ≥ 60 year-old diagnosed between 2000 and October 2021 and treated with SC only were included. Indolent AML (iAML) was defined as AML treated with SC and an OS ≥ 9 months (based on median OS with hypomethilating agents (Saiz-Rodriguez, et al. Cancers 2021). SC included patients receiving transfusions and other supportive measures, with or without oral agents to control high white blood cell (WBC) counts (i.e, hydroxyurea, melphalan, mercaptopurine or thioguanine). Those SC patients who were alive at last visit, but with a follow-up less than 9 months were excluded. RESULTS 963 AML patients treated with SC only were included in this study, 511 men and 452 women, with a median age of 79 years (range: 60-99). The main characteristics of the included patients are shown in Table 1. 79 out of 963 patients included (8.2%) had an iAML (≥9 months OS), 47 men (59.5%) and 32 women (40.5%). Median age was 79 years (range, 64-93), median white blood cell count (WBC) was 4.5 x109/L (range, 0.7-131.7) and median bone marrow blast was 35% (range, 5-100). iAML patients has a better ECOG PS compared with those non-iAML patients, 56.7% of iAML had an ECOG<2 vs 34.8% of non-iAML (p=0.001). A lower median leukocyte counts (4.5 vs. 10.9 x109/L, p=0.001), higher median platelet count (90 vs 53 x109/L, p=0.000), higher median albumin (3.66 vs. 3.40 g/dL, p=0.005) and lower median bone marrow blast count (35% vs. 56% p=0.000). In a multivariate regression model, including leukocytes < 10 x109/L, platelet count > 75 x109/L and BM blast < 50%, all three variables retained its statistical significance. The OR for leukocytes < 10 x109/L was 2.181 (95% CI: 1.141-4.1694, p=0.018), 2.523 for platelet count > 75 x109/L (95% CI: 1.410-4.517, p=0.002) and 2.002 for BM blast < 50% (95% CI: 1.102-3.636, p=0.023). Median OS of iAML was 14 months (95% CI 12.57-15.43), statistically higher than non-iAML patients (<1 month), p=0.000 (Figure 1). CONCLUSIONS We characterize 8.2% of iAML among unfit patients treated with SC only, with a similar median OS, or even better, than those actively treated. ECOG, WBC, platelet and bone marrow blast count could help us to identify these patients. New diagnostic techniques, such as NGS, could improve the identification of these patients and to understand underlying biology. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal