**Background:** The maintenance phase of acute lymphoblastic leukemia (ALL) treatment is the final and longest stage, focused on antimetabolite therapy to eliminate residual leukemic clones and prevent relapse. However, dose-limiting hematotoxicity is a major problem, often resulting in dose reduction or treatment discontinuation. While well investigated in high-income countries, data on incidence and risk factors for chemotherapy-induced hematotoxicity in low- and middle-income countries (LMICs) like Ethiopia are limited.
**Methods:** This cohort study was conducted at Tikur Anbessa specialized hospital (TASH) in Addis Ababa, Ethiopia, from 2019 to 2021. A total of 160 pediatric ALL patients were enrolled; 142 had sufficient data for analysis. Patients with renal disease, liver disease, or heart failure were excluded. Patients were stratified into standard risk, intermediate risk, and high risk groups. The maintenance phase was initiated with 75 mg/m² of 6-MP. Trimethoprim/sulfamethoxazole was given for Pneumocystis jirovecii prophylaxis. Complete blood counts were performed at 4-week intervals. Hematologic toxicity was graded per CTCAE version 4.0, with grade 4 defined as WBC <1000/mm³, ANC <500/mm³, anemia <6.5 g/dl, and thrombocytopenia <25,000/mm³. The primary outcome was grade 4 neutropenia. Bivariable followed by multivariable Cox regression analysis was performed. Two multivariable models were used to avoid interaction between WBC and ANC.
**Key Results:** Of the 142 participants, 92 (64.8%) were boys and 50 (35.2%) were girls. The mean age was 6.2±3.1 years, with 55.6% under age 6. Most (57.7%) were from urban areas. Based on WHO Z-scores, 14.3% were wasted, 22.1% underweight, 24.6% stunted, and 21.8% thin. Medication adherence was high in 78.2% and medium in 21.8%. During the first six months of maintenance therapy, grade 4 neutropenia occurred in 52.8% of patients, leukopenia in 31.7%, anemia in 13.4%, and thrombocytopenia in 5.6%. Treatment interruption occurred in 58.5% of patients, and 45.8% required emergency admission, primarily due to neutropenic fever (31%). Early-onset leukopenia (within 60 days) was observed in 19.7% and early-onset neutropenia in 32.4%. Multivariable analysis showed that children aged ≤6 years had approximately twice the risk of developing grade 4 neutropenia compared to those >6 years (Model 1: AHR 2.189, 95% CI 1.351–3.548, p=0.001; Model 2: AHR 2.09, 95% CI 1.291–3.383, p=0.003). Children with day 1 maintenance WBC <4500/mm³ had a 2.477 times higher risk of grade 4 neutropenia (AHR 2.477, 95% CI 1.461–4.200, p=0.001). Similarly, day 1 maintenance ANC <2500/mm³ was associated with a 2.11 times higher risk (AHR 2.11, 95% CI 1.105–4.029, p=0.024). Child’s age ≤6 years and low day 1 WBC/ANC were also independent predictors of treatment interruption and early-onset leukopenia/neutropenia. Caregiver-reported side effects included fever/flu-like symptoms (66.9%), itching/skin rash (57.8%), decreased appetite (50%), and behavior alterations (43%).
**Clinical Implications:** This study demonstrates a high incidence of grade 4 neutropenia (52.8%) in Ethiopian children with ALL receiving 6-MP-based maintenance therapy, higher than reports from high-income countries. Younger age (≤6 years) and low baseline WBC/ANC counts at the start of maintenance therapy are significant, independent risk factors for severe hematotoxicity and treatment interruption. These findings suggest that closer monitoring of WBC and ANC counts during maintenance therapy is essential, particularly in younger children and those with low baseline counts, to enable early detection and management of hematotoxicity. The high rates of treatment interruption (58.5%) and neutropenic fever (31%) underscore the clinical burden of toxicity in this setting. Multi-centered studies with larger sample sizes are needed to validate these findings and develop targeted interventions to reduce toxicity while maintaining treatment intensity.