Toward precision oncology: An integrative multi-omics approach for prognosis prediction and inferred immunotherapy responsiveness in breast cancer

Abstract

Breast cancer heterogeneity is still a primary concern, with a large variation in the prognosis necessitating the development of personalised treatment plans. Single biomarkers can't fully encompass the breast cancer complexity and variations, and therefore, multi-omics approaches offer a wide comprehension of the cancer-specific biology. In this study, we developed a multi-omics framework for the prediction of immune-active features framework integrating four-omics data, mainly genomic, proteomic and transcriptomic. By leveraging deep learning along with survival-based feature selection, we constructed an autoencoder that generated compressed multi-omics features to stratify the patients into two optimal immune subtypes with a significantly different overall survival (p < 0.002) and a high C-index of 0.74 (95% confidence interval: 0.62–0.83). An XGBoost classifier was trained to predict these subtypes by integrating all omics data as well as each omic individually, and was validated using both internal and external (the Cancer Genome Atlas and Gene Expression Omnibus) datasets. The integrated model achieved high predictive performance (ACC = 0.95). Omics-unique classifiers showed consistently strong validation on independent datasets, particularly the immunotherapy-treated cohort (GSE241876, p = 4.80×10⁻²). We further investigated the biological mechanisms across the clusters and discovered that the C2, low-risk cluster, exhibited an immune-active landscape, characterised by a high infiltration of cells and more immune-related pathways, making it a better candidate for a favourable immunotherapy response. On the other hand, the C1 cluster, the immune-cold group, displayed an immunosuppressive microenvironment and poor prognosis. This methodology demonstrated the promising potential of deep learning-driven multi-omics integration to support precision oncology by enhancing prognostic prediction and tailoring treatment decisions.

Keywords

breast cancer / computational analysis / immunotherapy / machine learning / multi-omics / prognostic model

References

[1] Tang L, Zhang Z, Fan J, et al. Comprehensively analysis of immunophenotyping signature in triple-negative breast cancer patients based on machine learning. Front Pharmacol. 2023; 14:1195864.

[2] Bendani H, Boumajdi N, Belyamani L, Ibrahimi A. A decision-aid model for predicting triple-negative breast cancer ICI response based on tumor mutation burden. BioMedInformatics. 2025; 5(1): 9.

[3] Wang X, Wang L, Bu H, et al. How can artificial intelligence models assist PD-L1 expression scoring in breast cancer: results of multi-institutional ring studies. Npj Breast Cancer. 2021; 7(1): 1-10.

[4] Bendani H, Boumajdi N, Belyamani L, Ibrahimi A. Revolutionizing breast cancer immunotherapy by integrating AI and nanotechnology approaches: review of current applications and future directions. Bioelectron Med. 2025; 11(1): 13.

[5] Lai J, Chen Z, Liu J, et al. A radiogenomic multimodal and whole-transcriptome sequencing for preoperative prediction of axillary lymph node metastasis and drug therapeutic response in breast cancer: a retrospective, machine learning and international multicohort study. Int J Surg. 2024; 110(4): 2162.

[6] Yu Y, Cai G, Lin R, et al. Multimodal data fusion AI model uncovers tumour microenvironment immunotyping heterogeneity and enhanced risk stratification of breast cancer. MedComm. 2024; 5(12):e70023.

[7] Gao Y, Ventura-Diaz S, Wang X, et al. An explainable longitudinal multi-modal fusion model for predicting neoadjuvant therapy response in women with breast cancer. Nat Commun. 2024; 15(1): 9613.

[8] Mehmood A, Li R, Kaushik AC, Wei DQ. Comparative analysis of the genomic and expression profiles of ANLN and KDR as prognostic markers in breast Cancer. Silico Pharmacol. 2025; 13(1): 15.

[9] Mayakonda A, Lin DC, Assenov Y, Plass C, Koeffler HP. Maftools: efficient and comprehensive analysis of somatic variants in cancer. Genome Res. 2018; 28(11): 1747-1756.

[10] Wilkerson AD, Parthasarathy PB, Stabellini N, et al. Phase II Clinical Trial of Pembrolizumab and Chemotherapy Reveals Distinct Transcriptomic Profiles by Radiologic Response in Metastatic Triple-Negative Breast Cancer. Clin Cancer Res Off J Am Assoc Cancer Res. 2024; 30(1): 82-93.

[11] Papanicolaou M, Parker AL, Yam M, et al. Temporal profiling of the breast tumour microenvironment reveals collagen XII as a driver of metastasis. Nat Commun. 2022; 13(1):4587.

[12] Barroso-Sousa R, Jain E, Cohen O, et al. Prevalence and mutational determinants of high tumour mutation burden in breast cancer. Ann Oncol. 2020; 31(3): 387-394.

[13] Hwang SH, Baek SH, Lee MJ, et al. Clinical relevance of TP53 mutation and its characteristics in breast cancer with long-term follow-up data. Cancers. 2024; 16(23): 3899.

[14] Jing X, Liang H, Hao C, Yang X, Cui X. Overexpression of MUC1 predicts poor prognosis in patients with breast cancer. Oncol Rep. 2019; 41(2): 801-810.

[15] Chen X, Zhao W, Huang Y, Luo S, Tang X, Yi Q. Association of GATA3 expression in triple-positive breast cancer with overall survival and immune cell infiltration. Sci Rep. 2024; 14(1):17795.

[16] Morrow E, Pennel K, Hatthakarnkul P, et al. High expression of STAT3 within the tumour-associated stroma predicts poor outcome in breast cancer patients. Cancer Med. 2023; 12(12): 13225-13240.

[17] Kim GC, Kwon HK, Lee CG, et al. Upregulation of Ets1 expression by NFATc2 and NFKB1/RELA promotes breast cancer cell invasiveness. Oncogenesis. 2018; 7(11): 91.

[18] Hyseni A, van der Groep P, van der Wall E, van Diest PJ. Subcellular FIH-1 expression patterns in invasive breast cancer in relation to HIF-1α expression. Cell Oncol. 2011; 34(6): 565-570.

[19] Arrouchi H, Lakhlili W, Ibrahimi A. A review on PIM kinases in tumours. Bioinformation. 2019; 15(1): 40-45.

[20] Wang M, Li J, Huang J, Luo M. The Predictive Role of PIK3CA Mutation Status on PI3K Inhibitors in HR+ Breast Cancer Therapy: a Systematic Review and Meta-Analysis. BioMed Res Int. 2020; 2020(1):1598037.

[21] Knappskog S, Helwa R, Rachakonda S, et al. Abstract PO5-25-12: cDH1 mutations predict resistance to neoadjuvant taxane therapy. Cancer Res. 2024; 84(9_Supplement):PO5-25-12.

[22] Tabassum S, Munir F, Awadh AAA, Anwar Z. Pathological significance of CDH1/E-cadherin germline sequence variants in breast cancer patients. Exp Oncol. 2023; 45(2): 170-179.

[23] Guillaudeux T, Ovechkina Y, Sridhar S, et al. Abstract 4261: cD27, a new immuno-oncology target shaping innate and adaptive anti-tumour immune responses. Cancer Res. 2022; 82(12_Supplement): 4261.

[24] Starzer AM, Berghoff AS. New emerging targets in cancer immunotherapy: cD27 (TNFRSF7). ESMO Open. 2019; 4:e000629.

[25] Palmieri D, Lockman PR, Thomas FC, et al. Vorinostat inhibits brain metastatic colonization in a model of triple-negative breast cancer and induces DNA double-strand breaks. Clin Cancer Res. 2009; 15(19): 6148-6157.

[26] Wojtukiewicz MZ, Rek MM, Karpowicz K, et al. Inhibitors of immune checkpoints—PD-1, PD-L1, CTLA-4—new opportunities for cancer patients and a new challenge for internists and general practitioners. Cancer Metastasis Rev. 2021; 40(3): 949-982.

[27] Hsu J, Sage J. Novel functions for the transcription factor E2F4 in development and disease. Cell Cycle. 2016; 15(23): 3183-3190.

[28] Kim J, Yim GW, Lee DW, Kim YT, Lee YJ, Rhee YJ. Knockdown of E2F4 suppresses the growth of ovarian cancer cells through the cell cycle pathway. Int J Clin Exp Pathol. 2021; 14(8): 866-874.

[29] Ballard JL, Wang Z, Li W, Shen L, Long Q. Deep learning-based approaches for multi-omics data integration and analysis. BioData Min. 2024; 17(1): 38.

[30] Huang L, Song M, Shen H, et al. Deep learning methods for omics data imputation. Biology. 2023; 12(10): 1313.

RIGHTS & PERMISSIONS

2026 The Author(s). Clinical and Translational Discovery published by John Wiley & Sons Australia, Ltd on behalf of Shanghai Institute of Clinical Bioinformatics.