High-Dose Methotrexate in Patients With Lymphoma: Predictors of a Complicated Course

High-dose methotrexate (HDMTX; > 500 mg/m²) is an important component of lymphoma therapy. Serum MTX monitoring at 48 hours is the standard approach to identify those at increased risk of developing MTX toxicity. Our aim was to characterize the incidence of complications and their association with MTX levels.

A retrospective review of our institutional electronic medical record was conducted to identify patients with lymphoma who received HDMTX between January 1, 2002, and December 31, 2018. We characterized the incidence of acute kidney injury (AKI), intensive care unit (ICU) admission, length of hospital stay (LOS), and 30-day mortality across 48-hour MTX levels. To establish an association between 48-hour MTX levels and the complications listed, we performed chi-square analysis for dichotomous variables and Kruskal-Wallis for nonparametric data. Receiver operator characteristic curve analysis was performed to identify the MTX level where AKI grade ≥ 2 was more likely. Multivariate logistic regression analysis was performed to identify risk factors for this MTX level.

We identified 642 patients with 2,804 cycles of HDMTX. The incidence of AKI was 19.1% with AKI grade ≥ 2 making up 21% of cases. Rates of AKI, ICU admission, and 30-day mortality are associated with elevated 48-hour MTX levels. There was a significant increase in median LOS with elevated MTX levels (P < .001). Receiver operator characteristic curve analysis for AKI grade ≥ 2 demonstrated a 48-hour MTX level threshold of 1.28 μmol/L. Multivariate logistic regression analysis revealed age, male sex, elevated body surface area, higher MTX dose, monotherapy, and first cycle as independent factors.

Elevated MTX levels are associated with a significant increased rate of AKI, ICU admission, prolonged LOS, and 30-day mortality. Elevated 48-hour MTX levels, particularly > 1.28 μmol/L, should alert clinicians for complications and to initiate measures to reduce MTX levels.