Palbociclib plus letrozole versus weekly paclitaxel, both in combination with trastuzumab plus pertuzumab, as neoadjuvant treatment for patients with HR-positive/HER2-positive early breast cancer: primary results from the randomized phase II TOUCH trial (IBCSG 55-17).

BACKGROUND: Hormone receptor (HR)-positive/human epidermal growth factor receptor 2 (HER2)-positive breast cancer (BC) is a heterogeneous disease with low pathological complete response (pCR) to standard neoadjuvant treatment. Cyclin-dependent kinase 4 and 6 inhibitors with endocrine and anti-HER2 therapy have shown a potential for chemotherapy omission in this context. PATIENTS AND METHODS: TOUCH is an international, open-label, phase II trial for postmenopausal patients with cT >1 cm, cN0 or cN1, HR-positive/HER2-positive BC, randomly assigned to 16 weeks of neoadjuvant weekly paclitaxel or palbociclib and letrozole, both with trastuzumab + pertuzumab (HP). The primary objective investigated the interaction between a gene signature of E2F pathway activity (RBsig) and pCR (ypT0N0 or ypTisN0), hypothesizing higher pCR for RBsig-high tumors in the paclitaxel + HP group and for RBsig-low tumors in the palbociclib + letrozole + HP group. RBsig was assessed by RNA-sequencing from pre-treatment biopsies; intrinsic subtypes were estimated by absolute intrinsic molecular subtyping Treatment-by-biomarker interaction was estimated using logistic regression in 115 assessable patients. RESULTS: A total of 147 patients were randomly assigned (74 paclitaxel + HP, 73 palbociclib + letrozole + HP) and 145 constituted the treated population, with a median age of 69 years (interquartile range 63-73 years). More patients completed palbociclib versus paclitaxel (94.4% versus 79.5%). The most frequent grade 3-4 adverse events were neutropenia and diarrhea (6.9% versus 43.1% and 11% versus 8.3% in the paclitaxel + HP versus the palbociclib + letrozole + HP groups, respectively). The pCR rate was 32.9% [95% confidence interval (CI) 22.3% to 44.9%] in the paclitaxel + HP group and 33.3% (95% CI 22.7% to 45.4%) in the palbociclib + letrozole + HP group. No significant treatment-by-RBsig interaction was observed (P = 0.18): pCR in RBsig high versus low was 31.3% (95% CI 16.1% to 50.0%) versus 42.3% (95% CI 23.4% to 63.1%) in the paclitaxel + HP group, and 38.5% (95% CI 20.2% to 59.4%) versus 25.8% (95% CI 11.9% to 44.6%) in the palbociclib group. pCR was higher in non-luminal versus luminal subtypes (45.5% versus 18.4%), with no interaction with treatment. CONCLUSIONS: Although the primary hypothesis was not supported, TOUCH shows that dual anti-HER2 blockade with a chemotherapy-free backbone of palbociclib + letrozole yields pCR rates similar to paclitaxel, representing an attractive alternative treatment strategy.