Phase I Dose-Escalation Study of Taselisib, an Oral PI3K Inhibitor, in Patients with Advanced Solid Tumors

Abstract

Taselisib is a potent and selective tumor growth inhibitor through PI3K pathway suppression. Thirty-four patients with locally advanced or metastatic solid tumors were treated (phase I study, modified 3+3 dose escalation; 5 cohorts; 3-16 mg taselisib once-daily capsule). Taselisib pharmacokinetics were dose-proportional; mean half-life was 40 hours. Frequent dose-dependent, treatment-related adverse events included diarrhea, hyperglycemia, decreased appetite, nausea, rash, stomatitis, and vomiting. At 12 and 16 mg dose levels, dose-limiting toxicities (DLT) were observed, with an accumulation of higher-grade adverse events after the cycle 1 DLT assessment window. Pharmacodynamic findings showed pathway inhibition at ≥3 mg in patient tumor samples, consistent with preclinical PIK3CA-mutant tumor xenograft models. Confirmed response rate was 36% for PIK3CA-mutant tumor patients with measurable disease [5/14: 4 breast cancer (3 patients at 12 mg); 1 non-small cell lung cancer], where responses started at 3 mg, and 0% in patients with tumors without known PIK3CA hotspot mutations (0/15).Significance: Preliminary data consistent with preclinical data indicate increased antitumor activity of taselisib in patients with PIK3CA-mutant tumors (in comparison with patients with tumors without known activating PIK3CA hotspot mutations) starting at the lowest dose tested of 3 mg, thereby supporting higher potency for taselisib against PIK3CA-mutant tumors. Cancer Discov; 7(7); 704-15. ©2017 AACR.See related commentary by Rodon and Tabernero, p. 666This article is highlighted in the In This Issue feature, p. 653.

Figure 1

In vivo efficacy of taselisib in the KPL-4 PIK3CA-mutant breast cancer xenograft model. Taselisib was dosed orally and daily at the doses indicated for 21 days as indicated by treatment period (Rx). Control tumor bearing mice were treated with 0.5% methylcellulose/0.2% Tween-80 (vehicle). Tumor volumes were measured and calculated as described in Materials and Methods.

Figure 2

Study design. (A) Phase Ia dose escalation with 5 cohort levels tested. (B) Schedule of assessments while on study.

Figure 3

Percentage change from baseline in target lesion by FDG-PET in patients in different dose cohorts.

Figure 4

(A) Best FDG-PET response (mean percentage change in SUV_max). Partial metabolic response was defined as greater than a 20% decrease in %ΔSUV_max. Patients with N/A (not applicable) did not have subsequent scan after starting treatment. All patient data arranged in (A) are in the same patient order as in as panels (B) and (C). (B) Best percent change from baseline in the sum of longest diameter (SLD) for target lesions via RECIST v1.1 available for 28 measurable patients with at least one post-baseline tumor assessment for target lesion (from 29 patients with baseline measureable disease) out of 34 enrolled patients. (C) Corresponding somatic mutation profiling in both tumor- and plasma-extracted DNA from enrolled from patients. The patient with PIK3CA mutation type “PIK3CA Other” had an R88Q mutation.

Figure 5

Pharmacodynamic modulation of the PI3K pathway. Needle core tumor biopsies obtained from patients at baseline and at steady state (cycle 1, between days 15-21) were fixed and evaluated by reverse phase protein array for PI3K-Akt pathway markers. Decreases of > 60% in pAkt and pS6, and up-phosphorylation of BIM (pro-apoptopic protein) were demonstrated in comparison to baseline for (A) patient 1 on 3 mg QD taselisib with paired biopsies from right endobronchial mass and (B) patient 2 on 16 mg QD taselisib with paired biopsies from right upper anterior thigh mass.