29P - Therapy-induced priming of natural killer cells predicts patient-specific tumor rejection in multiple breast cancer indications (29P)

Date 08 December 2017
Event ESMO Immuno-Oncology Congress 2017
Session Lunch & Poster Display session
Topics Cancer Immunology and Immunotherapy
Translational Research
Presenter Munisha Smalley
Citation Annals of Oncology (2017) 28 (suppl_11): xi6-xi29. 10.1093/annonc/mdx711
Authors M. Smalley1, B. Shanthappa2, H. Gertje2, M. Lawson2, B. Ulaganathan3, A. Thayakumar3, L. Maciejko3, P. Radhakrishnan4, M. Biswas5, S. Thiyagarajan6, B. Majumder6, K. Gopinath7, G. Babu8, A. Goldman1
  • 1Integrative Immuno-oncology Center, Mitra RxDx, 01801 - Woburn/US
  • 2Histology, Mitra RxDx, 01801 - Woburn/US
  • 3Cancer Biology, Mitra RxDx, 01801 - Woburn/US
  • 4R&d, Mitra RxDx, 01801 - Woburn/US
  • 5Pathology, Mitra RxDx, 560037 - Bangalore/IN
  • 6R&d, Mitra RxDx, 560037 - Bangalore/IN
  • 7Oncology, Bangalore Institute of Oncology, 560027 - Bangalore/IN
  • 8Oncology, Kidwai Memorial Institute of Oncology, 560029 - Bangalore/IN

Abstract

Background

Predicting patient-specific responses to anticancer therapy is the holy grail of therapy-selection. Response or resistance to therapy depends on the heterogeneous tumor microenvironment. This is particularly true for emerging anticancer drugs, such as immune checkpoint inhibitors, which recalibrate the body’s immune defense by modulating exhaustion of cytotoxic lymphocytes, T cells and natural killer (NK) cells. Clinical response to therapy varies greatly and there is a critical gap in our understanding of the mechanisms that drive response or resistance to therapy at the individual patient level.

Methods

We used a patient-autologous, clinically-validated ex-vivo tumor model that recreates the native tumor microenvironment (CANscriptTM) and uses an algorithm-driven method to predict clinical response to therapy (M-Score). Using tissue from luminal, HER2 positive, and triple-negative (ER- PR- HER2-) breast cancer patients (N = 10), we studied phenotypic alterations in the tumor-immune contexture under either standard-of-care regimens or immunotherapies, ex-vivo. We used a comprehensive panel of immunological assays to evaluate changes in cytotoxic lymphocytes by flow cytometry and multiplex immunohistochemistry.

Results

We identified that tumor response, predicted by M-Score, correlated with increased infiltration of NK cells and a pro-inflammatory cytokine signature from the tumor microenvironment. This coincided with induction of MICA/B, known to attract and recruit active NK cells. We also determined that therapy-induced protein biomarkers associated with NK cell exhaustion inversely correlated to the expression of cytotoxic granzyme B in the tumor microenvironment.

Conclusions

These data show that NK cells contribute to the antitumor effect of conventional and immuno-modulatory drugs and how a novel ex-vivo platform can be used to study the mechanisms of response and resistance, which wouldn’t be possible in a drug naïve state. Such an advance in our preclinical methods to study anticancer drugs at the individual patient level can help guide treatment decisions for clinicians while also functioning as a platform to study clinical efficacy of novel and emerging agents.

Clinical trial identification

Legal entity responsible for the study

Mitra RxDx

Funding

Mitra RxDx

Disclosure

M. Smalley, B. Shanthappa, H. Gertje, M. Lawson, B. Ulaganathan, A. Thayakumar, L. Maciejko, P. Radhakrishnan, M. Biswas, S. Thiyagarajan, B. Majumder, A. Goldman: Employee of Mitra RxDx. All other authors have declared no conflicts of interest.