3439 - Early on-treatment vs pre-treatment tumor transcriptomes as predictors of response to neoadjuvant therapy for HER2-positive inflammatory breast cancer

Background

Inflammatory breast cancer (IBC) is an understudied form of breast cancer (BC). The incidence of HER2-positive disease in IBC is 2-fold greater than that in non-IBC. Several studies evaluating dual-HER2 blockade in HER2-positive BC have demonstrated an association of specific gene expression signatures with response. Those analyses were based on pretreatment tumor characteristics and did not focus specifically on IBC. We used RNA-Seq to elucidate the impact of short term neoadjuvant dual-HER2 blockade on IBC transcriptomic profile and to identify early predictors of response.

Methods

We analyzed fresh frozen tumor samples prospectively obtained from 23 patients (pts) with HER2-positive IBC, treated at a single institution in a phase 2 trial (NCT01796197) with neoadjuvant trastuzumab, pertuzumab (HP) and paclitaxel for 16 weeks (wk). Breast biopsies were performed at baseline (D1) and 1 wk later (D8) after a single dose of dual blockade with HP before adding paclitaxel. Primary endpoint was pathologic complete response (pCR) defined as no evidence of invasive disease in breast or lymph nodes. Tumors from D1 and D8 were used for RNA-Seq analysis and assessment of tumor-infiltrating lymphocytes (TILs) and CelTIL score.

Results

Paired breast tumor biopsies (D1, D8) were obtained in all pts; 2 pts did not have surgery. In the intent-to-treat analysis, 10/23 (43%) pts achieved a pCR. Across all metrics, D8 biopsies were significantly better predictors of response than D1 (p-value: 1.0X10^-15). Upregulation of immune signatures by RNA-Seq was observed at D1 and D8. D8 biopsies showed a greater upregulation of anti-tumor immunity and changes in multiple signalling pathways. Neither TILs nor CelTIL were associated with pCR.

Conclusions

We identified an accurate predictor of response based on transcriptomic profiling by RNAseq, following a single dose of neoadjuvant dual-HER2 blockade in HER2-positive IBC. It outperforms a similar predictor constructed on a pretreatment profile in the same cohort. Assessing early-changes in gene expression level by RNA-seq following one dose of treatment may provide insights for the molecular mechanisms underlying response or resistance to anti-HER2 therapy.

Clinical trial identification

NCT01796197.

Editorial acknowledgement

Legal entity responsible for the study

Dana-Farber Cancer Institute.

Funding

Genentech, Inc and the Inflammatory Breast Cancer (IBC) Network.

Disclosure

S. Pernas: Travel / Accommodation / Expenses: Roche; Advisory / Consultancy: Polyphor. S. Goel: Advisory / Consultancy, Research grant / Funding (institution): Eli Lilly; Advisory / Consultancy, Research grant / Funding (institution): Novartis; Advisory / Consultancy: G1 Therapeutics; Research grant / Funding (institution): Merck. J.L. Guerriero: Advisory / Consultancy, Research grant / Funding (institution): Glaxo-Smith Kline; Research grant / Funding (self): Eli Lilly. E.A. Mittendorf: Advisory / Consultancy: Amgen; Advisory / Consultancy, Research grant / Funding (institution): AstraZeneca; Advisory / Consultancy, Research grant / Funding (institution): Genentech; Advisory / Consultancy: Genomic Health; Advisory / Consultancy: Merck; Advisory / Consultancy: Peregrine Pharmaceuticals; Advisory / Consultancy, Research grant / Funding (institution): Sellas Lifesciences; Advisory / Consultancy: Tapimmune. B. Overmoyer: Research grant / Funding (institution): Genentech; Research grant / Funding (institution): Incyte; Research grant / Funding (institution): Eisai. All other authors have declared no conflicts of interest.