Abstract P1-03-21: Improving gene signatures for predicting the progression of ductal carcinoma in situ

Diego M. Marzese, Miquel Ensenyat-Méndez, Sandra Íñiguez-Muñoz, Andrés F. Bedoya-López, Siri H. Strand, Robert West, Maggie L. DiNome, Eui-Hyoung HwangClinical 2025 · DOI

Abstract

Abstract Introduction: Ductal carcinoma in situ (DCIS) is a non-invasive form of breast cancer characterized by abnormal cells confined within the milk ducts. While DCIS is not life-threatening, its clinical significance lies in its potential to progress to invasive ductal carcinoma (IDC). This progression is driven by molecular imbalances from genetic and epigenetic alterations, including increased expression of matrix metalloproteinases (MMPs). Our recent study demonstrated that MMP1 expression modestly predicts DCIS progression to IDC, independent of tumor subtype (AUC = 0.76). In this study, we aimed to identify additional informative genes that could enhance the predictive potential of MMP1 based on molecular data from DCIS clinical studies. Methods: We analyzed molecular and clinical data from 380 DCIS specimens across three cohorts: the Sweden Cancerome Analysis Network - Breast (SCAN-B, n = 85), the Translational Breast Cancer Research Consortium (TBCRC, n = 216), and the Resource of Archival Breast Tissue (RAHBT, n = 79). Potential gene candidates were identified in TBCRC and RAHBT cohorts. First, DESeq2 v 1.40.2 was employed to select differentially expressed genes between cases with ipsilateral breast events (n = 121, considering both DCIS and IDC) and without reported events (n = 95) within five years in the TBCRC study. Additionally, we considered the genes encoding the 37 proteins analyzed by ion beam imaging by time of flight (MIBI-TOF) technology in the RAHBT cohort. Multiple score values based on the percentile expression of selected genes were evaluated in the SCAN-B cohort (n = 85). AUCs were computed using pROC version 1.1.8. Results: We identified 89 genes that improved the predictive power (AUC > 0.8) when combined with MMP1 on the TRCRC and RAHBT studies. These genes were combined into four-gene signatures using the formula: (MMP1+X)/(Y+Z), generating 92,070 unique combinations. 3,821 of these signatures improved the AUC of MMP1 alone. To avoid redundancy in the information, we restricted combinations to genes uncorrelated with MMP1 (28 dividend genes and 34 divisor genes) and tested the predictive performance in an independent cohort (SCAN-B). We identified that the ratio between the percentiles of (MMP1+MMP27)/(CD36+IGFBP6), showed a strong predictive performance in the SCAN-B cohort (AUC = 0.91). Other genes that enhanced the predictive signature included transcriptional regulators (SP1, SF3B4, and CTNNBIP), metabolic enzymes (GNMT, TKT, and PGD), and mitochondrial proteins (NDUFB7, MRPL9, and FH). Conclusions: The identified four-gene ratio signature significantly enhanced the prediction of DCIS progression to IDC across different study cohorts. We have designed and are developing an assay based on immunohistochemistry (IHC) to assess these proteins in FFPE tissues, enabling retrospective exploration of DCIS cohorts. Our current studies are focused on validating this signature in larger independent cohorts and exploring its utility in prospective clinical settings. This approach has the potential to refine diagnostic algorithms for DCIS patients, ultimately improving clinical decision-making and patient outcomes. Citation Format: Diego Marzese, Miquel Ensenyat-Méndez, Sandra Íñiguez-Muñoz, Andrés F. Bedoya-López, Siri Strand, Robert West, Maggie L DiNome, E Shelley Hwang, Diego M Marzese. Improving gene signatures for predicting the progression of ductal carcinoma in situ [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr P1-03-21.
BibTeX
@inproceedings{marzese2025abstract,
  author = {Diego M. Marzese and Miquel Ensenyat-Méndez and Sandra Íñiguez-Muñoz and Andrés F. Bedoya-López and Siri H. Strand and Robert West and Maggie L. DiNome and Eui-Hyoung Hwang and Diego M. Marzese},
  title = {{Abstract P1-03-21: Improving gene signatures for predicting the progression of ductal carcinoma in situ}},
  booktitle = {Clinical},
  year = {2025},
  doi = {10.1158/1557-3265.sabcs24-p1-03-21},
}

Loading related content...

Figure