Predicting hyperglycemia among patients receiving alpelisib plus fulvestrant for metastatic breast cancer.

Abstract

1060 Background: Although hyperglycemia is recognized as a common adverse event (AE) on alpelisib (ALP), this AE has been little studied outside clinical trials. We report the frequency of ALP-associated hyperglycemia in a real-world setting and evaluate proposed risk factors. Methods: We retrospectively identified patients with PIK3CA-mutated, hormone receptor-positive, metastatic breast cancer who initiated treatment with ALP+fulvestrant (FUL) between August 2019 and December 2021. Five primary characteristics (diabetes, prediabetes, body mass index (BMI), age, Asian ancestry) were evaluated as independent risk factors for ALP-associated hyperglycemia using ordinal logistic regression that considered 3 glycemic levels: normoglycemia, grade 2, and grade 3-4 hyperglycemia. Overall risk of error from multiple hypothesis testing was kept below 5% using the False Discovery Rate method. Results: The study included n = 92 subjects, all but 1 female, mean age 59.9 (+11.9) years, 13.0% with Asian ancestry. One third (33.7%) of patients had pre-existing diabetes, another 9.8% had pre-diabetes only. One third (32.6%) were obese, another third (31.5%) were overweight. Hypertension and hyperlipidemia were present in 53.3% and 41.3%, respectively. On ALP+FUL, 59 (64.1%) current subjects developed hyperglycemia of grade 1-4, a rate no different than the 181/284 (63.7%) reported in the ALP+FUL arm of the SOLAR-1 trial. Among our subjects, risk of grade 2-4 hyperglycemia was independently increased by 4 of 5 hypothesized risk factors, specifically pre-existing diabetes (Odds Ratio 3.75, 95% Confidence Interval: 1.40-10.01), pre-diabetes (6.22, 1.12-34.47), Asian ancestry (7.10, 1.75-28.84), each unit of BMI above 20 (1.17, 1.07-1.28), but not by additional year of age (1.01, 0.97-1.05). Exploratory analysis detected no association with pre-existing hypertension or hyperlipidemia. Conclusions: These findings suggest that Asian ancestry merits further study as a predisposing factor for ALP-associated hyperglycemia. Our study of this AE also demonstrates that pre-existing hyperglycemia and greater BMI are independent risk factors; diabetes and pre-diabetes confer similar degrees of risk; risk from BMI begins after BMI 20 and rises incrementally; and age is not a contributing factor.