Abstract
Phosphatidylinositol 3-kinase (PI3K) inhibitors are a promising class of targeted oncologic therapies but metabolic toxicities, particularly hyperglycemia, limit their use. We conducted a meta-analysis of phase II/III randomized controlled trials to quantify the incidence and relative risk (RR) of hyperglycemia associated with PI3K inhibition in cancer patients. A systematic search of PubMed through August 2025 identified 16 eligible trials, comprising approximately 5,000 patients. Among patients treated with PI3K inhibitors, the summary incidence of all-grade hyperglycemia was 44.4% (95% CI: 33.3-55.5%), while high-grade hyperglycemia occurred in 18.6% (95% CI: 9.7-27.6%). PI3K inhibitor was associated with an increased risk of hyperglycemia with an RR of 3.43 (95% CI: 2.42-4.87, p p < 0.001) for high-grade events. Subgroup analysis showed that cancer type significantly moderated the incidence of high-grade hyperglycemia (p = 0.01). Subgroup analyses also showed significant variation among PI3K inhibitor subtypes (p < 0.01), with PI3Kα-selective inhibitors conferring the greatest relative risk, particularly for severe hyperglycemia, whereas dual PI3K/mTOR inhibitors showed comparatively lower risk. These findings have demonstrated that PI3K inhibitors are associated with a significantly increased risk of hyperglycemia. Proactive monitoring and early intervention are essential to mitigate adverse outcomes and optimize therapeutic benefit.