546P - Early urinary potassium level is predictive of delayed methotrexate (MTX) elimination in patients with primary central nervous system lymphoma (PCNSL)

Background

First-line treatment of PCNSL includes high dose MTX. Nevertheless, MTX is associated with renal toxicity, requiring prolonged hospitalization and potentially compromising chemotherapy completion and so prognosis. Our objective was to identify early predictive factors of MTX delayed elimination.

Methods

We prospectively included all patients referred to our department for newly-diagnosed PCNSL. Daily serum and urinary creatinine and ionogram were collected from the day before MTX administration until elimination. Standard elimination delay of MTX was defined by plasmatic MTX level ≤ 0.05 μg/L before 72 hours. We conducted 2 independent training (TC) and confirmation (CC) cohorts.

Results

We included 64 cures (20 patients) in the TC and 59 cures (22 patients) in the CC between November 2020 to May 2021 and January 2022 to May 2022, respectively. At inclusion, median ages were 71 years (range 33-86) and 74 years (range 47- 87), median KPS were 70% (range 40-80%) and 80% (range 40-100%) and median elimination delay of MTX was 95 hours (range 47-205) and 96 hours (range 62-264) in the TC and the CC respectively. In the TC, median urinary potassium level at day 1 (uK+D1) was 23.9 mmol/L (range 5.5-64.1). In multivariate analysis, older age (p=0.004), low KPS (p=0.036) and high uK+D1 (p=0.001) were associated with delayed MTX elimination. Using a ROC analysis (p=0.008, AUC=0.712), an optimal cutoff for uK+D1 was defined at 17.1 mmol/L with a specificity of 81% and a positive predictive value of 84%. In the CC, we confirmed in multivariate analysis that older age (p=0.002), low KPS (p=0.007) and high uK+D1 (p=0.004) remained associated with delayed MTX elimination. After merging the cohorts, we were able to predict a standard MTX elimination probability based on age and uK+D1. Patients younger than 70 years had a standard MTX elimination probability of 84% versus 33%, depending on low and high uK+D1, respectively. Conversely, older patients had a probability of 26% and 14%, respectively.

Conclusions

uK+D1 may be predictive of delayed MTX elimination, opening personalized perspectives for young patient management. Its relevance as an element of decision making need to be validated in a larger prospective cohort.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

Has not received any funding.

Disclosure

All authors have declared no conflicts of interest.