PIK3CA mutations are associated with reduced pathological complete response rates in primary HER2-positive breast cancer: pooled analysis of 967 patients from five prospective trials investigating lapatinib and trastuzumab

PIK3CA mutation is associated with a lower pathological complete response (pCR) rate overall and in particular in HER2+ primary breast cancer treated with trastuzumab and lapatinib in addition to chemotherapy. Moreover, a significant interaction between the PIK3CA mutation status and the hormone-receptor status for pCR and DFS in patients who received the double blockade has been demonstrated.The predictive value of PIK3CA mutations in HER2 positive (HER2+) breast cancer treated with neoadjuvant anti-HER2 and chemotherapy has been reported, but the power for subgroup analyses was lacking. We combined individual patient data from five clinical trials evaluating PIK3CA mutations and associations with pathological complete response (pCR), disease-free survival (DFS) and overall survival (OS). Patients received either trastuzumab (T), lapatinib (L) or the combination T/L in addition to a taxane-based chemotherapy. PIK3CA was genotyped in tumour biopsies taken before therapy. A total of 967 patients were included in this analysis; the median follow-up is 47 months. Overall, the pCR rate was significantly lower in the PIK3CA mutant compared with the wild-type group (16.2% versus 29.6%; P < 0.001). Within the hormone-receptor positive (HR+) subgroup, the PIK3CA mutant group had a pCR rate of only 7.6% compared with 24.2% in the wild-type group (P < 0.001). In contrast, in the HER2+/HR- group, there was no difference in pCR (27.2% versus 36.4%; P = 0.125) according to PIK3CA mutation status (interaction test P = 0.036). According to treatment arm, the pCR rate for mutant versus wild-type was 20.3% versus 27.1% for T (P = 0.343), 11.3% versus 16.9% for L (P = 0.369) and 16.7% versus 39.1% for T/L (P < 0.001). In the HR+ T/L group, the pCR rate was 5.5% versus 33.9% (interaction between HR and PIK3CA genotype P = 0.008). DFS and OS were not significantly different by mutation status, though the incidence rate of events was low. However, HR+/PIK3CA mutant patients seemed to have significantly worse DFS {hazard ratio (HR) 1.56 [95% confidence interval (CI) 1.00-2.45], P = 0.050; P-interaction = 0.021}. T/L tended to improve DFS compared with T in the wild-type cohort, especially in the HR- group [HR 0.72, 95% CI (0.41-1.25), P = 0.242]. Overall PIK3CA mutant/HER2+ tumours had significantly lower pCR rates compared with wild-type tumours, however mainly confined to the HR+/PIK3CA mutant population. No definite conclusions can be drawn regarding survival.