Abstract PS2-07-19: Proteomic subtyping of stage II/III TNBC from the I-SPY2 TRIAL reveals a druggable steroid hormone receptor signature (SRS) that is associated with poor outcome and enriched in non-pCR patients’ residual disease

Abstract Background: I-SPY Triple Negative Breast Cancer (TNBC) Response Predictive Subtypes (RPS) predict for benefit from immunotherapy and DNA damaging agents in Immune+ and DNA Repair Deficient (DRD) subgroups. We explored TNBC subtyping based on functional protein/phosphoprotein architecture and previously described a subset of TNBC that expresses high relative levels of estrogen receptor alpha (ERα)1. These ERα “high” TNBC express a steroid hormone receptor signature (SRS) characterized by co-incidentally increased expression of ERα, activated glucocorticoid receptor (pGR S211) and total androgen receptor (AR), present in ∼30% of TNBC. We have now investigated the drug target activation architecture of SRS-positive (SRS+) TNBC and associations of SRS+ status with outcomes in I-SPY2 patients not responding to neoadjuvant therapy. Methods: Quantitative reverse phase protein array (RPPA) data from laser capture microdissection-enriched tumor epithelium comprised of 153 proteins/phosphoproteins, including ERα, AR, and pGR S211, were obtained from 253 I-SPY2 TNBC patients at pre-treatment baseline1 as well as in patient-matched surgical residual disease samples from 48 non-pCR TNBC patients. The SRS score was determined based on the equally weighted rank-sum addition of the 3 SRS analytes. SRS scores above and below the population median were considered SRS+ and SRS-negative (SRS-), respectively. Clinical information was provided by Quantum Leap Health Care Collaborative. For survival status and signaling pathway analyses, SRS scores were divided into tertiles and the top and bottom tertiles (N=84 each) were used as the SRS+ and SRS- populations. Statistical analyses were performed using GraphPad Prism v10 and JMP software v14. Uncorrected p-values 0.05 were used for determination of significance and results are descriptive. Results: SRS+ TNBC was found to be distinct from the luminal AR, Immune/DRD, PAM50, TNBC4, TNBC7, and MP1/2 subtypes. Analysis of the pre-treatment TNBC tumors revealed that patients with SRS+ tumors had worse survival status vs patients with SRS- tumors (p=0.05). Proteomic analysis revealed that SRS+ TNBC from both pre-treatment and residual disease surgical tumor samples were enriched for druggable targets including HARPS EGFR-HER21,2, AKT-mTOR and RAF-ERK activation signatures. Of interest, 58% (28/48) of the surgical samples from non-pCR patients were SRS+ based on the T0 population median cutpoint. SRS positivity in the pre-treatment samples did not correlate with worse outcomes from patients who achieved pCR. Conclusions: Proteomic data from I-SPY 2 showed that patients with SRS+ TNBC have a worse prognosis than SRS- TNBC. Surgical samples from non-pCR pts are enriched with SRS+ cancers. Proteomic analysis of SRS+ tumors obtained at pre-treatment baseline or in the residual disease surgical tumor samples from non-pCR SRS+ pts revealed activation of EGFR, HER2 and AKT-mTOR signaling networks. HER2/EGFR co-activated TNBC has been previously shown to have elevated levels of ER and were shown to be highly responsive to the EGFR-HER2 TKI inhibitor neratinib2. We hypothesize that inhibiting ER and AR, +/- GR, along with EGFR/HER2 may improve the poor prognosis of non-pCR SRS+ TNBC patients. 1.https://doi.org/10.1158/1538-7445.SABCS22-PD5-08 2. Wulfkuhle J et al. JCO Precis Oncol 2018 Citation Format: J. Wulfkuhle, P. Blas, H. Williams, G. Hirst, L. Brown-Swigart, R. I. Gallagher, M. Pierobon, C. Isaacs, L. van 't Veer, L. J. Esserman, J. O'Shaughnessy, E. F. Petricoin III. Proteomic subtyping of stage II/III TNBC from the I-SPY2 TRIAL reveals a druggable steroid hormone receptor signature (SRS) that is associated with poor outcome and enriched in non-pCR patients’ residual disease abstract. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS2-07-19.