Early methotrexate level thresholds predictive of acute kidney injury in patients with b cell lymphomas receiving high-dose methotrexate

Abstract Background: Methotrexate (MTX) induced acute kidney injury (AKI) occurs in 9% of patients (pts) receiving high dose (HD) methotrexate for B cell malignancies and is associated with increased mortality. Retrospective studies have identified MTX levels 48 hour (hr) after initiation of HD-MTX infusion to be predictive of MTX complications. However, the value of this time point may not be as clinically useful if a patient warrants earlier intervention of supportive strategies. For example, the optimal window for glucarpidase administration is within 48 to 60 hrs. The data evaluating prognostic value of 24 hr MTX values is mixed with some studies showing no correlation to MTX complications and others showing higher levels associated with AKI and receipt of glucarpidase. In this study we describe rates of AKI according to the presence of early supratherapeutic MTX levels in a cohort of pts receiving HD-MTX (MTX infusion ~4-6 hr) for B cell lymphoma. Methods: We retrospectively reviewed pts treated at our institution from Feb 2014 to May 2023 who received HD-MTX (> 1 g/m2) for an indication of B cell lymphoma including central nervous system (CNS) lymphoma or for CNS prophylaxis . We excluded MTX infusion for other cancer indications (e.g. acute lymphocytic leukemia and solid tumors). Per institutional protocol, MTX is given over a course of 4 hours and first MTX level is collected with morning labs the day after start of MTX infusion and then daily subsequently till MTX level < 0.1 uM. To examine early MTX levels and AKI, we excluded pts without MTX level within 36 hours post HD-MTX infusion initiation (n=5). In this study, we defined early delayed clearance (EDC) as supratherapeutic MTX at 36 hrs or earlier post MTX infusion (<24 hr level > 150 uM, 24-36 hr level > 15 uM). Late delayed clearance (LDC) was defined as those without EDC but detection of supratherapeutic levels after 36 hrs (36-60 hr level > 1.5 µM, 60-84 hr level > 0.15 µM, > 84 hr level > 0.1 µM). AKI was defined as 1.5x fold or higher increase in creatinine over baseline and grade >2 AKI was defined as > 2x increase. We used logistic regression to adjust for covariates and within pt variability. Results: Sixty-three pts receiving a total of 150 cycles of HD-MTX were included who had their first MTX level measured at ≤36 hrs post HD-MTX infusion initiation. Twenty-nine pts (46%) received HD-MTX for CNS prophylaxis and the others had primary or secondary CNS lymphoma. 70% were male and median age was 63 yrs. 24 pts (38%) received 1 cycle, 30% received 2 cycles, and 33% received 3+ cycles. Of all 150 cycles, AKI was seen in 22% of cycles, and grade 2 or higher AKI was seen in 10% of cycles. EDC was seen in 23 cycles (15%) cycles, LDC in 57 cycles (38%), and 70 cycles (47%) had no delay in MTX clearance; the rates of any AKI for EDC, LDC and no delay were 48% vs 30% vs 7%, respectively (p < 0.001). The rate of grade >2 AKI was 35%, 12%, and 0%, respectively (p < 0.001). Amongst pts who received multiple cycles of MTX with available early MTX level data, cycles with EDC (n = 14) also had a higher rate of both AKI and grade >2 AKI compared to LDC only and those with no delayed MTX clearance (p= 0.007, and p = 0.002, respectively). The unadjusted odds ratio of AKI for pts with EDC vs no EDC was 4.38 (95% CI 1.71-11.18, p = 0.002); similarly, the odds ratio for grade >2 AKI was 9.01 (95% CI 2.9-28.81, p < 0.001). With adjustment for within-pt correlation, age and gender, the odds ratio for any AKI for those with EDC was 3.10 (95% CI 1.13-8.51, p = 0.028) and for grade >2 AKI was 10.53 (95% CI 2.85-38.93, p < 0.001). Conclusions: To our knowledge, this is the largest study to identify <36 hr post initiation MTX level thresholds as predictive of grade> 2 AKI in pts receiving HD-MTX for B cell lymphoma. This challenges the notion established by historical studies that time points prior 48 hours do not correlate with MTX-AKI. Use of higher MTX concentration thresholds and limiting cohort to those receiving similar durations of MTX infusion in our study likely contributed to this finding. Earlier assessments of MTX levels may help guide who most benefits from early supportive measures including more frequent Cr monitoring, adjustment of leucovorin dose, early glucarpidase administration, and/or MTX dose reduction for subsequent cycles.