Abstract
Background Treatment response in acute myeloid leukemia (AML) is influenced by multiple biological and molecular mechanisms. High-mobility group box 1 (HMGB-1) is involved in inflammation, autophagy, apoptosis, and immunogenic cell death. This study evaluated associations of baseline circulating miR-181a-3p, miR-181b-5p, ATM, ATR, and HMGB-1 with successful induction response in adults with newly diagnosed AML. Methods This prospective cohort study was conducted at two referral hospitals in Jakarta, Indonesia. Adults with newly diagnosed AML receiving standard D3A7 induction chemotherapy were enrolled consecutively. Baseline miR-181a-3p and miR-181b-5p were quantified using digital polymerase chain reaction, whereas ATM, ATR, and HMGB-1 were measured using enzyme-linked immunosorbent assays before chemotherapy. Successful induction response was assessed by clinical and bone marrow evaluation after induction. Receiver operating characteristic analysis determined an exploratory HMGB-1 cut-off. Associations were evaluated using relative risks (RRs) with 95% confidence intervals (CIs). Results Among 102 patients initiating induction chemotherapy, 25 died during induction or early post-induction care. The biomarker-evaluable cohort comprised 71 patients, of whom 40 (56.3%) achieved a successful response. Baseline miR-181a-3p, miR-181b-5p, ATM, and ATR were not significantly associated with response. HMGB-1 yielded an area under the curve of 0.615 (95% CI, 0.480–0.749). The exploratory cut-off was 28.59 pg/mL, with 65.0% sensitivity and 61.3% specificity. Low HMGB-1 was associated with a lower probability of successful response (RR, 0.620; 95% CI, 0.394–0.975; p = 0.038). In the final clinical-biomolecular model, high HMGB-1 remained associated with successful response after adjustment for platelet count and ATR (adjusted RR, 1.666; 95% CI, 1.092–2.541; p = 0.018). Conclusion Low baseline circulating HMGB-1 was associated with a lower probability of successful induction response in the biomarker-evaluable adult AML cohort. Given its modest discriminatory performance and internally derived cut-off, HMGB-1 should be considered an exploratory complementary biomarker requiring external validation and serial evaluation.