Bendani Houda, Boumajdi Nasma, Belyamani Lahcen, Ibrahimi Azeddine
Laboratory of Biotechnology Lab (MedBiotech), Bioinova Research Center, Rabat Medical and Pharmacy School, Mohammed V University in Rabat, Rabat, Morocco.
Mohammed VI Center for Research and Innovation (CM6), Rabat, Morocco.
Bioelectron Med. 2025 May 30;11(1):13. doi: 10.1186/s42234-025-00173-w.
Breast cancer (BC) is still the most diagnosed cancer for females with an increased focus on immunotherapy as a promising precise treatment. Selecting appropriate patients and monitoring patient treatments are crucial to ensure higher response rates with low adverse events. Various biomarkers were proposed to predict immunotherapy response, including tumor mutation burden, immune cell, and tumor microenvironment expression. However, traditional methods for evaluating immunotherapy are invasive and inaccurate, and their assessments could be biased due to the variability in quantification techniques. Artificial intelligence (AI) has emerged as a powerful technology that addresses these challenges, handling heterogeneous data to identify complex patterns and offering accurate and non-invasive solutions. In this paper, we review emerging AI-based models for immunotherapy prediction in BC using diverse biomarkers. We first discussed the application of AI models for each biomarker, highlighting both direct prediction of immunotherapy response and prognosis, as well as indirect approaches via the identification of immune subtypes or specific predictive biomarkers. Then, we investigated the integration of all biomarkers in multi-modal AI approaches for a precise and personalized prediction of immunotherapy response. We have also addressed the implication of integrating AI in the healthcare ecosystem with other new technologies, including nanodevices, and wearable technologies. We further elucidated the role of AI and healthcare providers with this convergence of personalized medicine and demonstrated its role in enhancing population health management and supporting personalized patient care.
乳腺癌（BC）仍然是女性中诊断最多的癌症，免疫疗法作为一种有前景的精准治疗方法受到越来越多的关注。选择合适的患者并监测患者的治疗对于确保高反应率和低不良事件至关重要。人们提出了各种生物标志物来预测免疫疗法的反应，包括肿瘤突变负担、免疫细胞和肿瘤微环境表达。然而，评估免疫疗法的传统方法具有侵入性且不准确，并且由于量化技术的可变性，其评估可能存在偏差。人工智能（AI）已成为一种强大的技术，可应对这些挑战，处理异构数据以识别复杂模式并提供准确且非侵入性的解决方案。在本文中，我们回顾了使用多种生物标志物的基于人工智能的新兴模型，用于预测BC中的免疫疗法。我们首先讨论了人工智能模型对每种生物标志物的应用，重点介绍了对免疫疗法反应和预后的直接预测，以及通过识别免疫亚型或特定预测生物标志物的间接方法。然后，我们研究了在多模态人工智能方法中整合所有生物标志物，以精确和个性化地预测免疫疗法反应。我们还讨论了将人工智能与包括纳米设备和可穿戴技术在内的其他新技术整合到医疗保健生态系统中的意义。我们进一步阐明了人工智能和医疗保健提供者在个性化医疗融合中的作用，并展示了其在加强人群健康管理和支持个性化患者护理方面的作用。
