Biomarker analysis of the NeoSphere study: pertuzumab, trastuzumab, and docetaxel versus trastuzumab plus docetaxel, pertuzumab plus trastuzumab, or pertuzumab plus docetaxel for the neoadjuvant treatment of HER2-positive breast cancer

Abstract

Background: NeoSphere showed significantly higher pathologic complete response (pCR) with neoadjuvant pertuzumab, trastuzumab, and docetaxel compared with trastuzumab plus docetaxel, pertuzumab plus trastuzumab, or pertuzumab plus docetaxel. We assessed associations between human epidermal growth factor receptor 2 (HER2) pathway-related biomarkers and clinical outcome in response to these regimens.

Methods: Tumor, serum, and whole blood samples were collected at baseline and post neoadjuvant treatment before surgery. Associations between biomarkers and pCR, and between biomarkers and clinical variables were assessed in the overall and estrogen receptor (ER)-positive and ER-negative populations. Changes in serum marker levels between baseline and post-neoadjuvant treatment were examined.

Results: No markers were associated with pCR across all groups; however, significant associations were observed for two markers in individual groups. High HER2 was significantly associated with higher pCR rates (P = 0.001) and a significant treatment interaction (P = 0.0236) with pertuzumab, trastuzumab, and docetaxel (odds ratio 2.07, P = 0.01). Low serum transforming growth factor alpha (TGFα) was associated with higher pCR rates with pertuzumab plus trastuzumab (P = 0.04) without a significant treatment interaction. Presence of truncated HER2 did not affect pCR. A non-significant decreased pCR benefit was observed consistently across groups in patients with mutated PIK3CA while the treatment benefit from pertuzumab was maintained when comparing the trastuzumab plus docetaxel and pertuzumab, trastuzumab, and docetaxel groups. Notably, PIK3CA exon 9 mutations were associated with residual disease (pooled groups), which was not found for exon 20 mutations. Serum HER2 extracellular domain levels were significantly increased between baseline and post-neoadjuvant treatment in the non-trastuzumab-treated group, and decreased in the trastuzumab-containing groups (likely due to trastuzumab's mechanism of action). Differences in biomarker profiles according to ER status were observed.

Conclusions: The observed associations of HER2 protein levels with sensitivity to pertuzumab, and of PIK3CA exon 9 mutation to lack of sensitivity to HER2-targeted monoclonal antibody treatment, warrant further investigation. Previously reported findings of truncated forms of HER2 as resistance markers to HER2-targeted treatment could not be confirmed in NeoSphere. Conventional HER2 assessment should continue and HER2 remains the only biomarker suitable for patient selection in this population.

Trial registration: Clinicaltrials.gov, NCT00545688 . Registered on 16 October 2007.

Keywords: Biomarker; Breast cancer; Docetaxel; HER2; Neoadjuvant; Pertuzumab; Trastuzumab.

Fig. 1

Relationship between biomarkers and pathologic complete response (pCR) by treatment group. CR concentration ratio, cyt cytoplasmic, EGF epidermal growth factor, EGFR epidermal growth factor receptor, IGF1R insulin-like growth factor 1 receptor, Mem membranous, Mut mutant, Nuc nuclear, PIK3CA gene encoding phosphoinositide 3-kinase catalytic subunit, PTEN phosphatase and tensin homolog, qRT-PCR quantitative reverse transcription polymerase chain reaction, sHER2 serum HER2 extracellular domain, TGF transforming growth factor, WT wild-type

Fig. 2

Human epidermal growth factor receptor 2 (HER2) membrane (Mem) analyses. a HER2 membrane scores versus outcome (pathologic complete response (pCR)) in all four treatment groups. b HER2 membrane staining intensity versus outcome in all four treatment groups (patients with missing pCR were excluded). c HER2 membrane score by outcome

Fig. 3

Human epidermal growth factor receptor 2 (HER2) extracellular domain (ECD)/intracellular domain (ICD) ratio. pCR pathologic complete response

Fig. 4

PIK3CA analyses. a PIK3CA status and relationship to outcome per treatment group. b Analysis of PIK3CA mutations by estrogen receptor (ER) status and relationship to outcome. Mut mutant, pCR pathologic complete response, PIK3CA gene encoding phosphoinositide 3-kinase, catalytic subunit, WT wild-type

Fig. 5

Matched pair analyses of serum human epidermal growth factor receptor 2 extracellular domain (sHER2) levels at baseline and at surgery (S). *P < 0.05 (exploratory analyses). BL baseline, pCR pathologic complete response

Fig. 6

Insulin-like growth factor 1 receptor (IGF1R) expression according to estrogen receptor (ER) status. a Low IGF1R level/ER-positive group. b High IGF1R level/ER-positive group. c Low IGF1R level/ER-negative group. d High IGF1R level/ER-negative group. pCR pathologic complete response