A risk analysis of alpelisib-induced hyperglycemia in patients with advanced solid tumors and breast cancer

Abstract

Background: Hyperglycemia is an on-target effect of PI3Kα inhibitors. Early identification and intervention of treatment-induced hyperglycemia is important for improving management of patients receiving a PI3Kα inhibitor like alpelisib. Here, we characterize incidence of grade 3/4 alpelisib-related hyperglycemia, along with time to event, management, and outcomes using a machine learning model.

Methods: Data for the risk model were pooled from patients receiving alpelisib ± fulvestrant in the open-label, phase 1 X2101 trial and the randomized, double-blind, phase 3 SOLAR-1 trial. The pooled population (n = 505) included patients with advanced solid tumors (X2101, n = 221) or HR+/HER2- advanced breast cancer (SOLAR-1, n = 284). External validation was performed using BYLieve trial patient data (n = 340). Hyperglycemia incidence and management were analyzed for SOLAR-1.

Results: A random forest model identified 5 baseline characteristics most associated with risk of developing grade 3/4 hyperglycemia (fasting plasma glucose, body mass index, HbA_1c, monocytes, age). This model was used to derive a score to classify patients as high or low risk for developing grade 3/4 hyperglycemia. Applying the model to patients treated with alpelisib and fulvestrant in SOLAR-1 showed higher incidence of hyperglycemia (all grade and grade 3/4), increased use of antihyperglycemic medications, and more discontinuations due to hyperglycemia (16.7% vs. 2.6% of discontinuations) in the high- versus low-risk group. Among patients in SOLAR-1 (alpelisib + fulvestrant arm) with PIK3CA mutations, median progression-free survival was similar between the high- and low-risk groups (11.0 vs. 10.9 months). For external validation, the model was applied to the BYLieve trial, for which successful classification into high- and low-risk groups with shorter time to grade 3/4 hyperglycemia in the high-risk group was observed.

Conclusions: A risk model using 5 clinically relevant baseline characteristics was able to identify patients at higher or lower probability for developing alpelisib-induced hyperglycemia. Early identification of patients who may be at higher risk for hyperglycemia may improve management (including monitoring and early intervention) and potentially lead to improved outcomes.

Registration: ClinicalTrials.gov: NCT01219699 (registration date: October 13, 2010; retrospectively registered), ClinicalTrials.gov: NCT02437318 (registration date: May 7, 2015); ClinicalTrials.gov: NCT03056755 (registration date: February 17, 2017).

Keywords: Alpelisib; BYLieve; HR+/HER2− advanced breast cancer; Hyperglycemia; Machine learning; SOLAR-1.

Fig. 1

Cumulative incidence of all-grade hyperglycemic events among patients in SOLAR-1 treated with alpelisib + fulvestrant (A) or grade 3/4 hyperglycemic events among patients in SOLAR-1 treated with alpelisib + fulvestrant (B). Time to first medication vs first hyperglycemic event in patients treated with alpelisib from SOLAR-1 who had a hyperglycemia adverse event and was treated by antihyperglycemic medication (C) and subsequent grade of hyperglycemia among patients in SOLAR-1 with early (first quantile of time from hyperglycemia to treatment) and late (fourth quantile) treatment with antihyperglycemic medication (D). Cumulative incidence curves using Kaplan–Meier method. Hyperglycemic events by Standardized MedDRA query. AE, adverse event; G, grade

Fig. 1

Cumulative incidence of all-grade hyperglycemic events among patients in SOLAR-1 treated with alpelisib + fulvestrant (A) or grade 3/4 hyperglycemic events among patients in SOLAR-1 treated with alpelisib + fulvestrant (B). Time to first medication vs first hyperglycemic event in patients treated with alpelisib from SOLAR-1 who had a hyperglycemia adverse event and was treated by antihyperglycemic medication (C) and subsequent grade of hyperglycemia among patients in SOLAR-1 with early (first quantile of time from hyperglycemia to treatment) and late (fourth quantile) treatment with antihyperglycemic medication (D). Cumulative incidence curves using Kaplan–Meier method. Hyperglycemic events by Standardized MedDRA query. AE, adverse event; G, grade

Fig. 2

Variable importance of random forest (model 4) in the X2101 + SOLAR-1 training set (A) and prognostic association of risk status based on a random forest model with fasting plasma glucose, BMI, HbA_1c, monocytes, and age (model 7) with time to grade 3/4 hyperglycemia event in the X2101 + SOLAR-1 training set (B) and the X2101 + SOLAR-1 test set (C). A was calculated by conditional permutation importance, B was calculated by cumulative incidence curves, and C was calculated using the Kaplan–Meier method. bid, twice daily dosing; BMI, body mass index; DBP, diastolic blood pressure; FPG, fasting plasma glucose; HDL, high-density lipoprotein cholesterol; RBC, red blood cells; SBP, systolic blood pressure.

Fig. 3

Time to grade 3/4 hyperglycemia in patients classified as high or low risk by model 7 in BYLieve

Fig. 4

Progression-free survival for patients with PIK3CA mutations with high and low risk of hyperglycemia (model 7) treated with alpelisib in SOLAR-1