Disparities in pegasparagase-induced toxicity in Hispanic patients with acute lymphoblastic leukemia.

Abstract

e18590 Background: Hispanic individuals experience the highest incidence of acute lymphoblastic leukemia of any racial or ethnic group in the United States. While Hispanic individuals are known to have disproportionately increased mortality and higher therapy-related toxicity rates in ALL, there is a paucity of data regarding ethnic differences in rates of toxicity related to pegasparagase, a backbone of ALL treatment regimens. In our safety-net hospital system, we sought to evaluate disparities in rates of pegasparagase-induced hepatotoxicity between Hispanic and non-Hispanic patients. Methods: Pegasparagase infusions administered in adult ALL patients from 2011-2020 at Parkland Memorial Hospital in Dallas, Texas, were identified using the electronic medical record. Retrospective review of demographic, clinical, and socioeconomic data was conducted. For each infusion, dosing, timing, transaminase trends, and toxicity data were gathered. Statistical analysis was completed with Excel Analysis ToolPak and R Statistical Software. Accounting for baseline characteristics (age, BMI, sex, metabolic comorbidities, and baseline hepatic function), multiple linear regression models were constructed for peak transaminase and bilirubin levels. ANOVA testing, adjusted R2 values, and AIC were utilized to compare the contribution of race, ethnicity, and socioeconomic variables to modeling. Results: 47 individual patients received pegasparagase during the study period with 196 infusions evaluated. The mean age was 30 years, 29% were female and 73.4% white Hispanic. There were 64 grade 3-4 pegasparagase-related toxicities (54 hepatotoxicity, 4 pancreatitis, 6 thrombosis). 83.3% of grade 3-4 hepatotoxicity events were in Hispanic patients, and 16.7% in non-Hispanics. The median time from infusion to grade 3-4 toxicity was 18 days. Hispanic ethnicity was significantly associated with severe hepatoxicity (P = 0.0004). An interaction was identified with insurance status and primary language which significantly strengthened the impact of ethnicity and race on hepatotoxicity severity (P = 0.005, P = 0.02). Conclusions: In our safety-net population, Hispanic ethnicity was associated with significantly higher grades of hepatotoxicity when compared to non-Hispanic patients. Uninsured status and primary Spanish language use were not only independently associated with increased toxicity but also strengthened the effect of Hispanic ethnicity on the severity of liver dysfunction related to pegasparagase. These findings propose that Hispanic patients may benefit from closer monitoring during treatment with pegasparagase-containing regimens and highlight the need for further studies to improve risk-stratification of Hispanic patients when selecting ALL therapy.