Managing the side effects of PI3K, AKT and mTOR inhibitors so breast cancer patients stay on treatment

The PI3K/AKT/mTOR pathway controls cell growth, survival and metabolism, and its dysregulation drives cancer development, treatment resistance and progression. About 35%-40% of hormone receptor (HR)-positive breast cancers carry activating PIK3CA mutations, a group with poor prognosis. Early drugs such as buparlisib and taselisib caused high rates of treatment discontinuation because of side effects. Newer, more selective agents are now approved, so clinicians need to know how to handle their toxicities.

This ESMO Open review by Jhaveri and colleagues summarizes safety data from the phase III trials behind four approved drugs. These are everolimus (BOLERO-2, with exemestane), alpelisib (SOLAR-1, with fulvestrant), capivasertib (CAPItello-291, with fulvestrant) and inavolisib (INAVO120, with palbociclib and fulvestrant). The authors focus on hyperglycemia, rash, stomatitis and diarrhea, covering incidence, severity, dose modifications and time to onset. They also give guidance on preparing and monitoring patients.

The review explains why these problems occur. Hyperglycemia likely arises because inhibiting AKT blocks GLUT4 translocation, which reduces glucose uptake into muscle and fat and increases glycogenolysis and gluconeogenesis. Patients with undiagnosed prediabetes or limited beta-cell reserve are most vulnerable. Insulin would correct the glucose but may blunt the drugs' effect on the tumor. Rash is attributed to impaired keratinocyte growth and differentiation. Stomatitis reflects epithelial cell apoptosis plus inflammatory signaling. Diarrhea reflects disrupted gut homeostasis, reduced enterocyte function and altered mucosal immunity.

The authors stress that cross-trial comparisons are risky. BOLERO-2 (CTCAE v3) and SOLAR-1 (v4) graded hyperglycemia by fasting blood glucose. CAPItello-291 and INAVO120 used v5, which grades by intervention. Grade 3 now requires hospitalization or intravenous therapy, and events managed with oral metformin or SGLT2 inhibitors count as grade 1-2. This produces fewer high-grade events without necessarily meaning a safer drug. Trials also differed in HbA1c and fasting glucose exclusion criteria and in how adverse event terms were grouped. Long-term INAVO120 safety data are not yet available, so the authors cite the phase I/Ib GO39374 study (HbA1c below 7%) for patients treated over a year.

Biomarker testing underpins appropriate use. Inavolisib is approved in the US and EU for PIK3CA-mutated disease, including recurrence on or after adjuvant endocrine therapy, and showed a significant overall survival benefit at final analysis. Capivasertib covers PIK3CA, AKT1 or PTEN alterations. PIK3CA mutations are truncal and detectable in tissue or plasma. PTEN testing is more reliable in tissue, and immunohistochemistry can mislead. Everolimus was approved without a biomarker requirement. The authors call for earlier testing to guide first-line decisions.

The central message is that proactive preparation, monitoring and management of toxicities can help patients tolerate higher dose intensity, which may translate into better outcomes. This is a narrative review, so it offers expert guidance rather than new trial data, and it depends on trials that are not directly comparable.