Metabolic and genetic determinants of severe asparaginase toxicities in adolescent and young adult acute lymphoblastic leukemia.

Abstract Introduction. Asparaginase is a key component in the treatment of Philadelphia chromosome–negative acute lymphoblastic leukemia (ALL). In adolescent and young adult (AYA) patients, its use is complicated by underlying metabolic comorbidities. While its toxicities are well documented in pediatric populations, data in AYAs remain limited. Genetic polymorphisms may influence asparaginase-related toxicity through effects on hepatic lipid metabolism, oxidative stress, and inflammation. The PNPLA3 rs738409 variant, linked to impaired triglyceride hydrolysis, has been associated with hypertriglyceridemia. Variants in APOA5, APOE, and APOC3 have been implicated in dyslipidemia and drug-induced liver injury. This study aimed to identify risk factors for grade 3–4 (G3–4) toxicities in AYA ALL patients treated with pediatric-inspired regimens, and to explore the influence of selected genetic variants. Methods. We conducted a retrospective single-center study of patients aged >15 years treated with a modified CALGB10403 regimen (mCALGB) between February 2017 and 2025. E. coli L-asparaginase (6000 IU/m²) or pegaspargase (2500 IU/m²) were administered intramuscularly or intravenously, with or without premedication. Toxicities were graded using CTCAE v5.0. Clinical and biochemical predictors were evaluated using chi-square tests, odds ratios (ORs), and logistic regression. Mixed-effects models accounted for repeated cycles per patient. Survival was assessed via 6-month landmark analysis using Kaplan–Meier and log-rank tests. Four polymorphisms (PNPLA3, APOA5, GCKR, AKR1C4) were genotyped and analyzed under three penetrance models. Results. A total of 533 asparaginase cycles were analyzed in 111 patients (median age: 27 years; 42.3% female; 91% B-ALL). The median number of planned cycles was 7; the median administered was 6 (IQR 3–7). E Coli L-asparaginase accounted for 92.9% of doses. G3–4 toxicities included hypertriglyceridemia (27.8%), hepatotoxicity (25.7%), hyperglycemia (13.1%), and thrombosis (2.3%). Anaphylaxis was rare (3 cases), all occurring after IM administration of E. coli L-asparaginase despite premedication. Independent risk factors identified in multivariate analysis were: Hypertriglyceridemia: BMI >24 kg/m² (OR 2.51, p<0.001), history of dyslipidemia (OR 3.24, p<0.001) Hepatotoxicity: BMI >28 kg/m² (OR 1.13, p<0.001), induction phase (OR 5.84, p=0.031) Hyperglycemia: IV route (OR 2.54, p=0.011), diabetes (OR 5.80, p<0.001), BMI >25 kg/m² (OR 3.92, p<0.001) Thrombosis: concomitant glucocorticoids (OR 3.70, p=0.035) There were no significant toxicity differences between E. coli L-asparaginase and pegaspargase. To account for multiple cycles per patient, we used mixed-effects logistic regression. These models confirmed the same independent risk factors as in the standard analysis. Additionally, significant between-patient variability was observed for hypertriglyceridemia and hepatotoxicity, supporting the use of mixed modeling in this context. Genetic findings (n=73 patients, 370 cycles): APOA5 rs662799 was protective against hepatotoxicity (dominant model OR 0.51 (95% CI 0.30-0.88), p=0.015; codominant OR 0.59 (95% CI 0.36-0.95), p=0.033). PNPLA3 rs738409 was associated with higher hypertriglyceridemia risk (dominant OR 2.06 (95% CI 1.3-3.3), p=0.002; codominant OR 1.77 (95% CI 1.26-2.51), p=0.001; recessive OR 2.1(95% CI 1.04-4.18), p=0.035). Survival analysis (6-month landmark): Hypertriglyceridemia was associated with improved 3-year OS (100% vs. 62%, p=0.023) and EFS (95% vs. 50%; p=0.017) Hepatotoxicity and hyperglycemia had no significant impact on survival Conclusions. IV administration was associated with higher rates of metabolic and hepatic toxicity. Overweight status and metabolic comorbidities were consistent risk factors for G3–4 toxicities. PNPLA3 rs738409 increased the risk of hypertriglyceridemia, while APOA5 rs662799 appeared protective against hepatotoxicity, suggesting gene-specific effects on lipid metabolism. These findings underscore the need for personalized metabolic monitoring and supportive care strategies in AYA patients treated with asparaginase.