TP53 genomics predict higher clinical and pathologic tumor response in operable early-stage breast cancer treated with docetaxel-capecitabine ± trastuzumab
- First Online:
- Cite this article as:
- Glück, S., Ross, J.S., Royce, M. et al. Breast Cancer Res Treat (2012) 132: 781. doi:10.1007/s10549-011-1412-7
- 732 Downloads
To determine rates of pathologic complete response (pCR) and near-complete response (npCR) in operable early-stage breast cancer using neoadjuvant capecitabine plus docetaxel, with or without trastuzumab, and investigate biomarkers of pathologic response. Women with operable early-stage breast cancer were enrolled in a multicenter study of neoadjuvant therapy for four 21-day cycles with capecitabine 825 mg/m2 plus docetaxel 75 mg/m2 if human epidermal growth factor receptor 2 (HER2)-negative, and additionally, a standard trastuzumab dose if HER2-positive. Primary endpoint was rate of pCR and npCR. Secondary endpoints were potential associations between response and TP53 mutational analysis using the AmpliChip TP53 assay or immunohistochemical (IHC) staining, and genomic subtyping using the PAM50 assay. In patients who completed treatment and surgery, pCR and npCR rates were 15.8% in patients with HER2-negative and 50% in patients with HER2-positive tumors. Stratified by genomic subtype, patients of HER2-enriched subtype had the best response (72.2%), and luminal A (9.1%) and B (4.8%) subtypes, the poorest. Of 147 patients tested for TP53 mutations using the AmpliChip assay, 78 variants were detected; 55 were missense. Response rate among TP53-mutated patients was 30%, significantly higher than TP53 wild-type patients (10%; P = 0.0032). Concordance between AmpliChip mutation status versus TP53 IHC staining was 65%, with AmpliChip status predictive of response and IHC status not predictive. Capecitabine plus docetaxel in HER2-negative, and with trastuzumab in HER2-positive patients, provided a good response rate with four cycles of non-anthracycline-containing therapy. TP53 mutational analysis and genomic subtyping were predictive.