Research Summary:
Capivasertib is a potent, selective inhibitor of all three protein kinase B (AKT1/2/3) isoforms. The phosphoinositide 3- kinase/protein kinase B pathway is frequently overactivated in many solid tumors, including hormone receptor–positive (HR+), human epidermal growth factor receptor 2–negative (HER2–) breast cancer. In the Phase III CAPItello-291 trial, combination therapy with capivasertib 400mg (twice daily, 4 days on, 3 days off) and fulvestrant improved progression-free survival (PFS) versus placebo plus fulvestrant in patients with aromatase inhibitor-resistant, HR+, HER2– advanced breast cancer, with clinically meaningful PFS benefit in both the overall population (hazard ratio [HR]: 0.60, 95% confidence interval [CI]: 0.51–0.71, p<0.001) and the subgroup with PIK3CA/AKT1/PTEN-altered tumors (HR: 0.50, 95% CI: 0.38–0.65, p<0.001), and an improved objective response rate (ORR).
This study by Fernandez Teruel et al1 sought to expand the existing population pharmacokinetic (PopPK) model using pooled data from six clinical studies with 5,960 capivasertib plasma concentrations included in the analysis, spanning Phase I to III monotherapy and combination regimens, and evaluated exposure-response relationships for efficacy and safety for capivasertib 400mg twice daily on the four-days-on/three-days-off schedule in combination with fulvestrant. The PopPK analysis used nonlinear mixed-effects modeling, and exposure metrics (area under the concentration-time curve [AUC], maximum concentration [Cmax], and minimum concentration [Cmin]) were generated as empirical Bayes estimates for the final dosing day of the second week (252–264 hours) based on the planned 400mg dose.
The final PopPK model was a three-compartment structural model with parallel zero-order and first-order absorption, an absorption lag time, and time- and dose-dependent apparent clearance (CL/F). Initial CL/F was 59.6L/h with interindividual variability of 38.2 percent, and was modeled to decrease over time, with the time at which inhibition of CL/F was halved (T50) of 126 hours and a plateau in CL/F predicted after approximately 168 hours (Day 7) at 400mg twice daily. CL/F also decreased with increasing dose, with reductions of 11.2 percent and 29.2 percent after multiple dosing at 400mg and 640mg, respectively. The model showed that capivasertib disposition was adequately described across regions, with goodness-of-fit plots and prediction-corrected visual predictive checks supporting the model’s performance. Covariate analysis identified age and body weight as statistically significant predictors of CL/F at steady state, while sex, hepatic function, renal function, race, geographic region, smoking status, and concomitant fulvestrant were not statistically significant covariates.
Exposure-efficacy analyses were based on 798 patients from CAPItello-291. PFS Kaplan–Meier curves for patients treated with capivasertib stratified by quartiles of exposure metrics were all clearly separated above the placebo curve, indicating consistently improved PFS with capivasertib plus fulvestrant compared with placebo plus fulvestrant, but the curves for the four capivasertib exposure quartiles substantially overlapped with each other, therefore indicating no relationship between exposure and PFS. Cox proportional hazards models confirmed that none of the exposure metrics were significant predictors of PFS after likelihood ratio testing. Similar findings were observed in the PIK3CA/AKT1/PTEN-altered population. Additionally, logistic regression analyses of ORR also showed no significant associations between any exposure metrics and ORR, and no covariates were identified as significant predictors of ORR in the exposure-response analysis.
Exposure-safety analyses included 468 patients who received capivasertib 400mg twice daily four-days-on/three-days-off with fulvestrant. Across trials and cohorts, adverse events (AEs) were common (96.4% had at least one AE), but AE-related treatment discontinuations remained infrequent (11.3%), and median relative dose intensity exceeded 95 percent across all groups, indicating that the 400mg regimen was generally well tolerated. In the pooled dataset, 43.6 percent of patients experienced Grade 3 or greater AEs, 43.2 percent had an AE leading to dose modification (interruption and/or reduction), and 19.0 percent experienced serious AEs. Logistic regression modeling identified statistically significant relationships between capivasertib exposure and several AE endpoints, as higher AUC and Cmin were associated with increased probabilities of AE leading to dose modification and Grade 3 or greater AEs.
Overall, capivasertib demonstrated moderate PK variability, with no clinically meaningful impact of intrinsic factors such as race, age, body weight, or renal or hepatic function on exposure, and no evidence that higher exposure within the range achieved at 400mg twice daily on a four-days-on/three-days-off schedule improved PFS or ORR in patients with aromatase inhibitor-resistant, HR+, HER2– advanced breast cancer. Although higher exposure modestly increased the risk of certain AEs, most patients tolerated 400mg without permanent dose reduction or discontinuation. Taken together, the updated PopPK model and exposure-response analyses supported the capivasertib regimen in combination with fulvestrant as an appropriate fixed dose for this patient population.
References
- Fernandez Teruel C, Cullberg M, González-García I, et al. Population pharmacokinetics and exposure-response analyses for capivasertib in combination with fulvestrant in patients with breast cancer. Clin Transl Sci. 2025;18(7):e70286.