Optimizing high-dose methotrexate treatment regimens to improve exposure and mitigate toxicity in infant patients with acute lymphoblastic leukemia

Abstract

Abstract Introduction: High-dose methotrexate (HDMTX) is a key component of therapy for high risk acute lymphoblastic leukemia (ALL). Typical dosing regimens infuse 5 g/m2 over 24 hours in pediatric and adolescent young adult (AYA) patients without Down Syndrome. Empiric dose reductions to 3.5-4.5 g/m2 are used in infant patients to compensate for physiological maturation that slows the elimination of HDMTX. Despite dose reduction, infants have more methotrexate (MTX) adverse events (AEs) compared to non-infants, suggesting further dose optimization is needed. The objectives of this study were to (1) describe rates of MTX AEs in infants with ALL, (2) analyze HDMTX pharmacokinetics (PK), (3) determine the relationship between HDMTX and AEs and (4) perform modeling and simulation to understand optimal dosing strategies of HDMTX for infants with ALL. Methods: PK and AE data were retrospectively abstracted from infant, pediatric, and AYA patients with ALL after HDMTX (3.5–5 g/m² over 24 hours) at 3 large children's hospitals. Infants included patients aged ≤1.25 years at time of HDMTX infusion. AEs were graded by manual chart abstraction following a priori Common Terminology Criteria for Adverse Events v5 definitions for mucositis (grades 0-4), neutropenia (grade 4), thrombocytopenia (grades 3-4), and serum creatinine (SCr) increased (grade 2+). Chi-squared tests were used to compare the frequency of AEs for infants and non-infants. A PK model was developed and describes a triphasic elimination of HDMTX. Covariates on MTX elimination include body surface area, baseline SCr, time-dependent changes in SCr from baseline, and infant age groups (<6 months, 6-12 months, and 12-15 months). Model building evaluated changes to the log-likelihood and goodness-of-fit plots. The final model underwent internal (visual predictive checks; bootstrapping) and external validation (bias; accuracy) from data obtained at three additional pediatric institutions. Logistic regression with a binomial link determined the relation between MTX exposure and presence of severe mucositis (grades 3-4). The final regression model underwent internal (bootstrap; k-fold cross-validation) and external validation (discrimination; calibration). Dosing simulations using a virtual patient population (N=1000) were integrated with the logistic regression to determine the probability of severe mucositis for each simulation scenario and age group. Area under the concentration-time curves (AUC) was analyzed for the 24-hour (AUC24) exposure. Results: The cohort included 619 patients who received 2259 HDMTX administrations. This included 50 (8%) infants who received 163 (7%) administrations. Despite receiving a lower dose (g/m2), infants had higher rates of severe mucositis (36% vs 10%, p<0.001), neutropenia (53% vs 21%, p<0.001), thrombocytopenia (38% vs 15%, p<0.001), and an increase in SCr (13% vs 7.5%, p=0.015) compared to non-infants. The PK model characterized the effects of infant age groups (<6 months, 6-12 months, 12-15 months) on MTX elimination compared to non-infants, with reductions of 25%, 18%, and 9%, respectively. The 42-hour MTX concentrations were lower in infants compared to non-infants (0.44 vs 0.56, p<0.001) and were associated in a regression analysis with severe mucositis for both age groups. However, when infant status was included as a binary variable and analyses controlled for differences in 42-hour concentration, infants had a 5.9x greater odds of mucositis. Simulations using doses of 3-5 g/m2 revealed that non-infants receiving 5 g/m2 over 24 hours generated a median AUC24 of 1928 µM (1730-2134). Infants (<6 months and 6-12 months) matched the non-infant AUC using infusions of 4.0-4.25 and 4.5 g/m2 over 24 hours, respectively. Infants with similar MTX exposure to non-infants demonstrated a 40% probability of experiencing severe mucositis compared to a 10% probability for non-infants. Conclusion: Infants have higher rates of MTX AEs despite current empiric dose reductions. Our data support prior studies that HDMTX infusion of 4-4.5 g/m2 in infants generates an equivalent exposure to non-infants receiving 5 g/m2; however, infants are still expected to experience more toxicity compared to non-infants, even with similar exposure. Future studies should evaluate if alternative dosing strategies or supportive care interventions could ameliorate the high rates of AEs in infants with ALL without loss of efficacy of HDMTX as a key component of ALL therapy.