激素受体阳性（Hormone receptor positive）/人表皮生长因子受体2阴性（Human epidermal growth factor receptor 2 negative ）型乳腺癌（HR+/HER2-）在所有乳腺癌患者中占比超过70%，内分泌治疗对这一分型的疗效显著，但化疗敏感性却具有较强异质性。在临床工作中，该分型内部化疗获益人群的分层、化疗时机选择的问题存在争议。对拟接受新辅助化疗以期获得更佳手术机会的HR+/HER2-患者来说，化疗不敏感的肿瘤特性有可能会导致疾病进展，从而延误手术机会。而对已经完成手术治疗的HR+/HER2-患者来说，低获益人群接受辅助化疗对预后改善不明显，反而可能因化疗严重的副反应而降低生活质量。因此，如何筛选出化疗获益人群并挖掘化疗敏感性的分子机制，从而逆转肿瘤的耐药是临床关注的重点问题。

得益于测序技术的蓬勃发展，乳腺癌的分子图谱已经被逐渐解析。结合多组学分析手段，目前研究者可以从基因组学、转录组学、蛋白组学多个角度刻画乳腺癌的分子特征，并通过临床分组之间的差异分析挖掘耐药相关机制。

本研究通过结合基因组测序、转录组测序、蛋白组测序以及单细胞测序技术对HR+/HER2-乳腺癌进行了全方位、多角度的分析。在基因组突变层面，我们发现HR+/HER2-乳腺癌具有以PIK3CA和GATA3为主的突变特征。在转录组学层面，我们发现HR+/HER2-乳腺癌在药物代谢、离子运输相关通路显著上调，而抗肿瘤免疫相关通路则呈现抑制状态。在单细胞层面，我们发现HR+/HER2-型乳腺癌存在大量浸润的基质成分，但具有抗肿瘤功能的免疫细胞比例却较少，整体呈现“冷肿瘤”的特征。

在上述分析的基础上，我们希望进一步通过临床样本探究HR+/HER2-乳腺癌的内在异质性，从而优化化疗获益人群的筛选。在术前新辅助方面，我们通过既往维护的新辅助化疗临床队列，利用患者肿瘤石蜡组织切片进行了蛋白组测序，并根据新辅助治疗的疗效和预后进行了分型，鉴定出HR+/HER2-患者新辅助化疗疗效和预后高获益的亚型。同时为使术后低复发风险人群避免过度化疗，我们利用早期HR+/HER2-乳腺癌手术+辅助化疗队列，通过对术后肿瘤石蜡组织切片的转录组测序结果进行分析鉴定出复发相关基因，并通过log-rank检验、LASSO回归、多元Cox回归的方法构建了复发风险预测模型，准确评估患者术后复发风险，为术后化疗豁免人群的鉴别提供了有效方法。

为了在细胞层面深入研究化疗敏感性的关键分子，我们构建了HR+/HER2-乳腺癌细胞MCF7的蒽环类药物多柔比星（Doxorubicin，DOX）耐药株，通过转录组测序找出了介导耐药的关键分子PANX1。随后利用细胞和动物实验证实MCF7细胞系可通过增强PANX1通道蛋白的表达，从而促进炎症小体通路激活和IL1β释放的方式诱导细胞产生耐药性。而联合利用PANX1抑制剂甘珀酸（carbenoxolone，CBX）以及化疗药物则可有效逆转MCF7的耐药性。这一发现为临床耐药患者的化疗增敏提供了新的策略。

通过上述研究成果，我们围绕术前化疗、术后化疗、化疗耐药三个关键临床问题构建了疗效预测模型，挖掘并证实了分子层面的耐药机制，这一系列研究为进一步理解HR+/HER2-乳腺癌化疗耐药提供了新的证据，并为HR+/HER2-乳腺癌的精准治疗提供了新的策略和方向。

Hormone receptor-positive (HR+)/human epidermal growth factor receptor 2-negative (HER2-) breast cancer (HR+/HER2-) accounts for over 70% of all breast cancer cases. While endocrine therapy demonstrates significant efficacy in this subtype, chemosensitivity exhibits marked heterogeneity. In clinical practice, controversies persist regarding the stratification of chemotherapy-benefiting populations and optimal timing of chemotherapy within this subtype. For HR+/HER2- patients scheduled for neoadjuvant chemotherapy to improve surgical outcomes, chemoresistance may lead to disease progression and delayed surgery. Conversely, postoperative adjuvant chemotherapy offers limited prognostic benefits for low-benefit subgroups and may reduce quality of life due to severe adverse effects. Therefore, identifying chemotherapy-responsive populations, elucidating molecular mechanisms underlying chemosensitivity, and reversing drug resistance remain critical clinical priorities.

Advancements in sequencing technologies have progressively unraveled the molecular landscape of breast cancer. By integrating multi-omics approaches, researchers can now characterize molecular features from genomic, transcriptomic, and proteomic perspectives, enabling the identification of resistance-related mechanisms through comparative analyses across clinical subgroups.

In this study, we conducted comprehensive multi-dimensional analyses of HR+/HER2- breast cancer using genomic, transcriptomic, proteomic, and single-cell sequencing. At the genomic level, we identified predominant mutation signatures involving PIK3CA and GATA3. Transcriptomic profiling revealed significant upregulation of drug metabolism and ion transport pathways, alongside suppressed anti-tumor immune pathways. Single-cell sequencing demonstrated abundant stromal infiltration but scarce anti-tumor immune cells, collectively characterizing a "cold tumor" microenvironment.

Building upon these findings, we further investigated the intrinsic heterogeneity of HR+/HER2- breast cancer using clinical cohorts to optimize chemotherapy stratification. For neoadjuvant therapy, we performed proteomic sequencing on tumor paraffin sections from a retrospective neoadjuvant chemotherapy cohort, identifying distinct subtypes with high therapeutic response and favorable prognosis. To prevent overtreatment in postoperative low-risk patients, we analyzed transcriptomic data from surgical specimens of early-stage HR+/HER2- breast cancer patients receiving adjuvant chemotherapy. Through recurrence-associated gene identification and integration of log-rank tests, LASSO regression, and multivariate Cox regression, we established a recurrence risk prediction model to accurately guide chemotherapy de-escalation.

To mechanistically explore chemoresistance, we generated a doxorubicin (DOX)-resistant MCF7 cell line (HR+/HER2-) and identified PANX1 as a key resistance mediator via transcriptomic sequencing. Cellular and animal experiments confirmed that PANX1 upregulation promotes chemoresistance through inflammasome pathway activation and IL1β release. Notably, combining the PANX1 inhibitor carbenoxolone (CBX) with chemotherapy effectively reversed drug resistance in MCF7 cells, offering a novel therapeutic strategy for chemosensitization.

Collectively, our study addresses three critical clinical challenges-neoadjuvant chemotherapy optimization, postoperative risk stratification, and chemoresistance reversal-by establishing predictive models and validating molecular resistance mechanisms. These findings provide new insights into chemotherapy resistance in HR+/HER2- breast cancer and advance precision treatment strategies for this prevalent subtype.