Abstract P5-06-14: A novel combination of a 2 gene score & TIL as a predictive Biomarker for responders to novel therapies in Indian TNBC - A population with greater proportion of TNBCs

Abstract

Abstract Introduction: Triple-Negative Breast Cancer (TNBC) comprises approximately 30% of all breast cancers in Indian women. Given their aggressive nature, TNBCs have high rates of systemic metastasis and mortality with only chemotherapy available for treatment. Compelling evidence has demonstrated the prognostic value of tumor-infiltrating lymphocytes (TILs), in many cancers especially in Breast Cancer and play a critical role in tumour progression, response to therapeutics and prognosis. They are typically measured by H&E staining and immunohistochemistry for research purpose and degree of infiltrate is also known to differ among subtypes. Nonetheless, the literature regarding the types of immune cells characterizing TILs and their prognostic utility in TNBC has been conflicting for lack of accurate “functional” TIL assessments. Herein we have used a combination signature of TILs and a 2 gene immune function expression based signature to develop an unique classifier, to identify a subgroup within TNBC that has better clinical outcome, such as survival and response to treatment. Our findings also suggest a possible use of the score as a predictive biomarker for response to immune checkpoint therapy Patients and Methods: Surgically excised TNBC tumor specimens from 44 women from a single treating hospital in Bangalore, India were accessed under IERB approved protocols and assayed on nCounter® PanCancer Immune Profiling Panel which comprised of 740 genes and characterises 14 different immune cell types. Analysis using Non-negative Matrix factorization (NMF) based unsupervised clustering was done to arrive at stable subtypes characterized by different tumor microenvironments within TNBC. SSGSEA analysis was done to identify immune cell types. TILs were characterised by a pathologist on H&E sections according to guidelines from TIL working group. The Immune classifier was validated on Breast Cancer data sets from TCGA. Results: NMF followed by SAM and PAM analysis yielded 2 stable subtypes (ST1 and ST2) of proportions 59% and 41% respectively. TIL groups also were divided into Dense and Mild groups which were 52% and 48% respectively. Despite similar distributions of the groups, ST1 and ST2 had distinct survival with ST2 having poorer outcome (p=0.023) while the dense and mild TIL groups did not separate as distinct groups. On a closer examination using SSGSEA analysis with Rooney et al signature, ST1 was enriched with T-cells, Dendritic cells, MHC-class 1 and very significantly high cytolytic score (CYT), which was defined by Granzyme A and Perforin expressions, proteins secreted by cytotoxic T cells (p=0.0074). The CYT Score ranged from -35 to 170 and cutoff was based on 75th percentile. We next hypothesized that a combined score of TIL and CYT would represent a functional TIL group with high immune activity. We arrived at a CYT+TIL score based on 75th percentile cut off of CYT and 2 groups of TIL. Of the resultant four groups, TIL High-CYT high that constituted about 20% of all TNBCs indicated an elevated immune response because of their microenvironment constitution and this can be identified using our simple immune classifier. On the other hand, in the TIL-high and CYT-low group, despite lymphocyte migration, outcome was not favorable. The classifier was applied to TNBC from TCGA (n=96) and similar results seen, with ST1 and ST2 which had distinctly separate survival and had similar patterns on CYT high groups. Conclusion: We developed a simple immune classifier with 2 gene signature c and H&E slide TIL assessment for identifying a group with TNBC that can be better targeted with therapies. Our signature is predictive for selecting TNBCs which are potential responders to Immune therapies and has the potential for quick clinical adoption though it requires validation on a larger set. Citation Format: Aruna Korlimarla, Hari PS, Jyothi S Prabhu, Chantirika Ragulan, Ravi B Diwakar, Sandhya Apachu, Maggie Cheang, Rekha V Kumar, BS Srinath, TS Sridhar, Savitha Rajarajan, Annie Alexander, Anguraj Sadanandam. A novel combination of a 2 gene score & TIL as a predictive Biomarker for responders to novel therapies in Indian TNBC - A population with greater proportion of TNBCs [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P5-06-14.