713b A Multi-Systems Approach with Big Data from Adverse Event Reporting and Health Records, Clinical Biospecimens, and Experimental Animal Data Reveal That Retinoids Impact Asparaginase-Associated Pancreatitis Outcomes in Leukemic Patients

Abstract

Asparaginase is a crucial cornerstone chemotherapy for the most common cancer in children, acute lymphoblastic leukemia (ALL). However, a major concern is that 5-10% of all asparaginase users develop the painful, life-threatening complication of pancreatitis. Asparaginase-associated pancreatitis (AAP) is a more severe form of pancreatitis, and discontinuation of asparaginase is associated with inferior disease-free survival in high-risk cancer patients. Thus, there is a major unmet need to mitigate AAP. In this study, we took a multi-systems approach that combined adverse event reporting data and electronic health records (EHR), clinical biospecimens, and experimental data to identify factors that impact AAP. An analysis of 14 million records from the FDA Adverse Events Reporting System (FAERS) demonstrated the novel finding that vitamin A supplementation to asparaginase was correlated with reduction in the risk of AAP. To home in the role of systemic vitamin A exposure, versus topical agents, an analysis of over 64 million records from an EHR database (TriNetX Research) identified a protective effect of non-topical vitamin A supplementation on AAP. To pursue this further, we performed a case control study of 26 ALL patients who developed pancreatitis at a median of 3 months after the start of chemotherapy and compared them with 24 well-matched control ALL patients who did not develop pancreatitis over the course of leukemia therapy. We performed non-targeted metabolomics of plasma before and after induction chemotherapy. The striking difference between the two groups was that the precursors of vitamin A were reduced in the pancreatitis cases post-induction compared to the controls. In mice, we found that single agent asparaginase administration by itself led to reduce retinol levels in serum. Analysis of transcriptomic data from leukemic cells demonstrated that the transcriptional signature from asparaginase exposure was potentially reversed by retinoid interventions. The finding was similar in transcriptomic data from a cerulein model of pancreatitis. Taken together, these data suggest that reductions in retinol levels may be correlated with AAP, that asparaginase causes those reductions, and that retinoid supplementation may prevent AAP. The major novelty of the current work is that it provides crucial insight into mitigating the adverse event of AAP and elucidates a broader role for retinoids in preventing the sterile inflammation of the pancreas. Fig 2Vitamin A metabolite levels between pancreatitis patients and controls. The y-axis represents log2 post/pre. * q<0.1. View Large Image Figure Viewer Download Hi-res image