PIK3CA mutation H1047R is associated with response to PI3K/AKT/mTOR signaling pathway inhibitors in early-phase clinical trials

Abstract

PIK3CA mutations may predict response to PI3K/AKT/mTOR inhibitors in patients with advanced cancers, but the relevance of mutation subtype has not been investigated. Patients with diverse cancers referred to the Clinical Center for Targeted Therapy were analyzed for PIK3CA and, if possible, KRAS mutations. Patients with PIK3CA mutations were treated, whenever possible, with agents targeting the PI3K/AKT/mTOR pathway. Overall, 105 (10%) of 1,012 patients tested harbored PIK3CA mutations. Sixty-six (median 3 prior therapies) of the 105 PIK3CA-mutant patients, including 16 individuals (of 55 PIK3CA-mutant patients tested) with simultaneous KRAS mutations, were treated on a protocol that included a PI3K/AKT/mTOR pathway inhibitor; 17% (11/66) achieved a partial response (PR). Patients with a PIK3CA H1047R mutation compared with patients who had other PIK3CA mutations or patients with wild-type PIK3CA treated on the same protocols had a higher PR rate (6/16, 38% vs. 5/50; 10% vs. 23/174, 13%, respectively; all P ≤ 0.02). None of the 16 patients with coexisting PIK3CA and KRAS mutations in codon 12 or 13 attained a PR (0/16, 0%). Patients treated with combination therapy versus single-agent therapies had a higher PR rate (11/38, 29% vs. 0/28, 0%; P = 0.002). Multivariate analysis showed that H1047R was the only independent factor predicting response [OR 6.6, 95% confidence interval (CI), 1.02-43.0, P = 0.047). Our data suggest that interaction between PIK3CA mutation H1047R versus other aberrations and response to PI3K/AKT/mTOR axis inhibitors warrants further exploration.

Conflict of interest statement

Conflict of interest: Filip Janku has research support from Novartis. Razelle Kurzrock has research support from Novartis, Merck, and Bayer.

Figures

Figure 1. Waterfall plot shows best response for patients with PIK3CA mutations treated with PI3K/AKT/mTOR inhibitors

Of the 66 treated patients, 65 are depicted in the waterfall plot (one patient died of unrelated causes prior to her first restaging). A total of 11 PRs and 11 minor regressions less than PR were observed. The overall PR rate was 17%. Patients with a H1047R mutation (marked with *) had a PR rate of 37.5% (6/16 – denominator includes the patient, who is not depicted), whereas none of the patients with KRAS mutations in codons 12 and 13 (marked with #) had a PR. Two patients with KRAS mutation Q61H had a SD ≥ 6 months and a PR, respectively (marked with ‡).

Figure 2. Kaplan-Meier plot for progression-free survival (PFS)

Tick marks represent patients who were progression-free at last follow up and are censored at that point. A. Patients with a PIK3CA H1047R mutation (yellow) demonstrated a trend toward having a longer median PFS compared to patients with other PIK3CA mutations (blue) (5.7 months vs. 2 months; p=0.06). B. Patients with a PIK3CA E545K mutation (yellow) did not have a significantly different median PFS compared to patients with other PIK3CA mutations (blue) (3.1 months vs. 1.9 months; p=0.54). C. Patients with a PIK3CA E542K mutation (yellow) compared to patients with other PIK3CA mutations (blue) had a trend toward having a shorter median PFS (1.8 months vs. 2.6 months; p=0.06). D. Patients with PIK3CA and simultaneous KRAS mutations either in codon 12 or 13 (blue) had a shorter median PFS compared to patients with PIK3CA mutations and wt KRAS or codon 61 mutations (yellow) (1.8 months vs. 2.6 months; p=0.046).