Breast density and estrogen exposure are two major risk factors for breast cancer; however, the underlying biological mechanisms remain incompletely understood. Here, we investigated the extracellular proteomic landscape of normal human breast tissue in situ to define the microenvironment associated with these risk factors. Forty-two postmenopausal women with nondense or dense breasts and 19 premenopausal women underwent microdialysis. We quantified 461 inflammatory proteins in breast tissue and matched subcutaneous fat, enabling discrimination between local and systemic alterations. Breast density was assessed using magnetic resonance imaging. Dense breast tissue exhibited a distinct protein signature characterized by immune-related signaling, altered lipid metabolism, and the extracellular presence of intracellular proteins, consistent with cellular stress and immune modulation, with limited changes in angiogenic and extracellular matrix remodeling proteins. In contrast, estrogen-exposed breasts displayed a proteomic profile dominated by pro-inflammatory cytokines, angiogenic factors, extracellular matrix remodeling proteins, and complement activation, indicative of a dynamic and pro-tumorigenic microenvironment. These findings demonstrate that breast density and estrogen exposure are associated with fundamentally distinct breast microenvironments, both permissive for tumor progression. These breast-specific protein signatures provide mechanistic insight into how these risk factors contribute to cancer development and suggest that effective prevention strategies may require differential targeting.
This open-label phase III trial assessed a chemotherapy-free regimen for recurrent/metastatic triple-negative breast cancer (TNBC). Patients were 1:1 randomized to benmelstobart plus anlotinib or nab-paclitaxel monotherapy, stratified by prior taxane exposure, liver and brain metastases. Planned enrollment was 322, yet recruitment was prematurely stopped by COVID-19, leaving only 147 randomized participants (75 experimental, 72 control). All efficacy analyses are exploratory and require cautious interpretation. The primary endpoint, IRC-assessed progression-free survival (PFS), was not met. Investigator-assessed median PFS reached 7.85 months for the combination vs. 5.55 months for nab-paclitaxel (HR = 0.70, 95%CI 0.46–1.06, P = 0.1687). Median overall survival (OS) was 35.81 vs. 21.03 months (HR = 0.78, 95%CI 0.49–1.24, P = 0.2625). Grade ≥3 treatment-related adverse events affected 58.7% of combination patients and 36.6% of monotherapy patients. This regimen failed to deliver statistically superior clinical benefits over nab-paclitaxel, only showing a non-significant favorable trend. Further trials are warranted to validate this observation. Trial ID: NCT04405505, registered May 24, 2020 on ClinicalTrials.gov.
Tumors presumed to be metastatic triple-negative breast cancer (TNBC) may occasionally represent misdiagnoses of primary or metastatic tumors from other origins—particularly lung—due to overlapping clinical characteristics and lack of organ-specific morphological features. While immunohistochemical markers can resolve these discrepancies, they are often underutilized without prior clinical suspicion, leading to significant treatment delays. To address this, we employed GPSai, a tissue-of-origin artificial intelligence (AI) model integrated into routine molecular profiling at Caris Life Sciences, to systematically identify differential diagnoses among 2423 presumed TNBC cases. By integrating AI results with clinical and molecular evidence, we identified a misdiagnosis rate of 3.0%. While over half of these cases were reclassified as non-small cell lung cancer, the misdiagnosis cohort also included various other tumor types. These diagnostic shifts profoundly impact staging, prognosis, and therapeutic selection, highlighting the necessity of an unbiased, systematic approach to identifying misdiagnoses among presumed TNBC.
Abstract The majority of breast cancer patients have tumors expressing estrogen receptor α (ER) and receive endocrine therapy. However, around one-third relapse in their disease, predominantly with retained ER expression. Molecular alterations are proposed to be contributors to the resistance mechanisms. Patients with ER-positive, human epidermal growth factor receptor 2 (HER2)-negative primary breast cancer with an ER-positive relapse < 5 years of ongoing endocrine therapy were retrospectively assessed. Extracted DNA was analyzed through panel sequencing, and RNA by microarray, from patients’ primary (n = 58), and paired relapse tumors (n = 54), and tumor-free lymph nodes (DNA germline controls, n = 62). Several single-nucleotide variations and copy number variations showed nominal exploratory associations with worse overall survival. Copy number correlations with intrinsic subtypes and individual gene expression supported the findings. These results identify hypothesis-generating genomic and transcriptomic features, including potentially targetable alterations, in a clinically defined cohort of endocrine-resistant breast cancer patients.
Stromal tumor-infiltrating lymphocytes (sTILs) may hold clinical value for triple-negative breast cancer (TNBC) for outcome prediction. Manual scoring of sTILs suffers from inter-observer variability, limiting its clinical adoption. We developed an artificial intelligence (AI) workflow to automatically quantify sTILs from diagnostic biopsies and evaluated its predictive performance for pathological complete response (pCR) in patients with early-stage TNBC receiving neoadjuvant chemotherapy. The pipeline combined tissue segmentation with context-aware cell classification to derive sTILs scoring, AI-sTILdensity. Using pretreatment biopsies (n = 394) from the ARTEMIS trial (ClinicalTrials.gov: NCT02276443; October 21, 2014), AI-sTILdensity showed strong correlation with manual sTILs (Spearman rho: 0.686-0.739) and improved prediction of pCR (AUC: 0.702-0.747 vs. 0.692-0.713) as an independent feature (P = 0.016). Predictive performance was consistent in an external TNBC set (n = 64). These findings demonstrated that automated assessment of sTILs from biopsy specimens is reproducible and can modestly enhance pCR prediction, supporting the use of AI approaches to enable more objective and scalable evaluation of sTILs in TNBC.
In addition to delivering cytotoxic payloads, HER2-targeted antibody–drug conjugates (ADCs) can engage innate immune effector pathways (such as FcγR and complement) similar to monoclonal antibodies (mAbs). However, the relative contribution of innate pathways to therapeutic benefit remains unclear, especially in cases where ADCs are combined with mAbs. To address this question, we performed analysis of bulk clinical transcriptomic datasets that revealed higher baseline expression of C1q and FcγR genes, likely representing intratumoral macrophage, monocyte, and neutrophil transcription, was associated with greater benefit from HER2-targeted mAb and ADC regimens. Mechanistic studies validated these findings, demonstrating that trastuzumab deruxtecan (T-DXd) and pertuzumab each induced robust FcγR-dependent ADCC and ADCP, and together cooperatively activated C1q-dependent classical complement pathway. Consistent with these mechanisms, the efficacy of T-DXd plus pertuzumab was diminished in FcγR-deficient or C1q-deficient in vivo models. These findings implicate FcγR and classical complement as key contributors to T-DXd plus pertuzumab efficacy and suggest C1q/FcγR expression signatures as potential biomarkers to predict treatment response.
Advancements in transmission electron microscopy (TEM) have enabled in-depth studies of biological specimens, offering new avenues to large-scale imaging experiments with subcellular resolution. Mitochondrial morphology is of growing interest in cancer biology due to its crucial role in regulating the multi-faceted functions of mitochondria. We and others have established the crucial role of mitochondria in triple-negative breast cancer (TNBC), an aggressive subtype of breast cancer with limited therapeutic options. Building upon our previous work demonstrating the functional role of mitochondrial morphology dynamics in the metabolic adaptations and survival of chemotherapy-refractory TNBC cells, we sought to extend those findings to analysis of transmission electron micrographs. Here, we present a novel U-Net artificial intelligence (AI) model for automatic annotation and assessment of mitochondrial morphology and feature quantification. Our model was trained on 11,039 manually annotated mitochondria across 125 micrographs derived from a variety of orthotopic patient-derived xenograft (PDX) mouse model tumors and adherent cell cultures. The model achieves an F1 score of 0.85 on test micrographs at the pixel level. To validate the ability of our model to detect expected mitochondrial morphology changes, we utilized micrographs from mouse primary skeletal muscle cells genetically modified to lack Dynamin-related protein 1 (Drp1), a key mitochondrial fission protein. We subjected in vitro and in vivo TNBC models to conventional chemotherapy treatments commonly used for clinical management of TNBC, including doxorubicin, carboplatin, paclitaxel, and docetaxel. We found substantial within-sample heterogeneity of mitochondrial morphology in both in vitro and in vivo. In four of five PDX models, in vivo treatment with DTX elicited significant alteration in mitochondrial elongation and/or area. We went on to compare mammary tumors and matched lung metastases in a highly metastatic PDX model of TNBC, revealing altered mitochondrial elongation in metastatic lesions compared to their matched primary mammary tumor. The successful application of our AI model provides a framework for high-throughput quantitative analysis of mitochondrial morphology and enables future studies investigating how mitochondrial structural changes relate to chemotherapy response and mechanism of action. Our publicly available manually curated electron micrograph dataset serves as a unique resource for developing, benchmarking, and applying computational models, while further advancing investigations into mitochondrial morphology in breast cancer. This study provides proof of concept that mitochondrial structural remodeling is an additional layer of cellular reprogramming accompanying therapeutic resistance in TNBC that merits further investigation.
Postpartum breast cancer (PPBC), diagnosed within 10 years of the last childbirth, has an increased risk of liver metastasis. In murine models, tumor transplant studies reveal that weaning-induced liver involution is a key determinant of breast cancer metastasis to the liver. Here we explored potential mechanisms by which hepatic stellate cells (HSCs), the tissue-resident liver fibroblasts, contribute to pre-metastatic niche formation immediately following weaning. We characterized normal murine liver involution and breast cancer metastasis to the liver using mIHC, bulk RNAseq, and scRNAseq approaches. In non-tumor-bearing mice, post-wean HSCs activation was confirmed by increased expression of ECM (Col3, Col5, and fibronectin), and increased expression of immune regulatory gene signatures when compared to nulliparous mice. Mammary tumor cells injected into the portal vein of InvD2 mice developed liver lesions with increased stromal attributes, when compared to nulliparous hosts, consistent with fibroblast activation. When isolated HSCs were mixed with mammary tumor cells and subcutaneously injected, InvD4-HSC group tumors were larger, desmoplastic, and enriched for regulatory immune cells when compared to the nulliparous-HSC tumors. These findings illuminate how HSCs are activated during weaning-induced liver involution, consistent with the establishment of a pre-metastatic niche, and underscore the translational potential of targeting HSCs to improve outcomes for PPBC patients.
Metastasis remains the primary cause of breast cancer death, driven by organotropism to brain, lung, bone, and lymph nodes. Single-cell and spatial multi-omics reveal organotropism as an active ecological process orchestrated by crosstalk between tumor cells and stromal/immune compartments. We synthesize advances into a unified framework: conserved programs (hypoxic memory, stemness, immune evasion) operate alongside organ-specific adaptations—lipid rewiring via RARRES2-PTEN-mTOR in brain, DLAT-mediated fatty acid oxidation in lung, TDO2+ fibroblast-kynurenine ferroptosis resistance in lung, IL-34-CSF1R-ARG1+ macrophage immunosuppression in brain, and MAF-CD248 dormancy in bone. These vulnerabilities inform therapies: dual-payload ADCs, metabolic blockers, and engineered EVs. We address limitations such as spatial mapping, dormancy models, translation and outline a roadmap with AI, liquid biopsies, and spatial profiling for precision interception. This Review advocates shifting from cytotoxic to niche-directed therapies, framing metastasis as an ecosystem challenge.
Tumor-infiltrating lymphocytes (TILs) are established prognostic and predictive biomarkers in early-stage triple-negative breast cancer (eTNBC), yet their role in patients treated with the KEYNOTE-522 (KN522) regimen remains underexplored. The Neo-Real/GBECAM-0123 study is a real-world cohort of patients with early-stage TNBC treated with pembrolizumab plus neoadjuvant chemotherapy across Brazil and Argentina since 2020; we evaluated baseline stromal TILs (<30%, 30–49%, or ≥50%) as predictors of pathologic complete response (pCR, primary endpoint) and event-free survival (EFS, secondary endpoint). Among 248 patients, TILs were <30% in 72.6%, 30–49% in 12.9%, and ≥50% in 14.5%. pCR rose with increasing TILs: 59%, 65.6%, and 91.4% (P = 0.001). In multivariable analysis, TILs ≥50% (OR 6.96, 95% CI 1.89–25.65, P = 0.004) and Ki-67 ≥ 50% (OR 4.89, 95% CI 2.31–10.33, P < 0.001) independently predicted pCR; virtually all patients with both TILs ≥50% and Ki-67 ≥ 50% achieved pCR (n = 26/27, 96.3%). Among patients with pCR, 2-year EFS was uniformly high across TIL groups (97.0%, 95.2%, 93.2%; P = 0.828); among those with residual disease, EFS trended higher with increasing TILs (73.3%, 90.0%, 100%; P = 0.350), non-significant. Baseline TILs, especially with Ki-67, emerged as a strong predictor of pCR in this real-world eTNBC cohort treated with KN522, supporting their systematic assessment in future studies.
Oligoprogression poses a growing challenge in precision oncology. We report a case series of hormone receptor–positive/HER2-negative metastatic breast cancer patients treated with the selective PI3Kα inhibitor RLY-2608 who developed isolated progressive lesions. Integrated tissue and circulating tumor DNA analyses identified acquired resistance mechanisms, enabling local ablative therapy and continuation of targeted treatment beyond progression, with subsequent clearance of resistance mutations.
Special histologic subtypes of triple-negative breast cancer (TNBC) are biologically distinct, yet their outcomes in the immunotherapy era remain undefined. We analyzed 823 patients with early-stage TNBC treated with pembrolizumab plus chemotherapy in the multicenter real-world Neo-Real/GBECAM-0123 cohort. Histologies included invasive carcinoma of no special type (NST; n = 762, 92.6%), metaplastic carcinoma (n = 26, 3.2%), triple-negative lobular carcinoma (n = 17, 2.1%), and other rare variants (n = 18, 2.2%). Metaplastic carcinoma demonstrated significantly lower pathological complete response (pCR) rates compared with NST (17.4% vs 65.4%; OR 0.14, 95% CI 0.04–0.44; p = 0.001), suggesting relative resistance to chemo-immunotherapy, with a numerical trend toward worse event-free survival (EFS) (HR 2.07, 95% CI 0.88–4.88; p = 0.094). Triple-negative lobular carcinoma achieved pCR rates comparable to NST (56.2%) but showed poorer 2-year EFS, with a significant association in the pre-surgical multivariable model (HR 4.25, 95% CI 1.50–12.06; p = 0.006), and a persistent trend after adjustment for pCR. This highlights the biological heterogeneity of TNBC and supports the investigation of histology-driven therapeutic strategies.
Invasive lobular breast cancer (ILC), the most common special histologic subtype of breast cancer, may metastasize to atypical sites, including the peritoneum, gastrointestinal tract, ovaries, urinary tract, and orbit. Because metastatic ILC (mILC) is often less FDG-avid and may present after a long disease-free interval with nonspecific symptoms, diagnosis may be delayed or mistaken for other conditions. We developed and piloted a 45-question IRB-approved patient survey with breast cancer researchers, clinical oncologists, and patient advocates to evaluate patient-reported delayed diagnosis or misdiagnosis of mILC. Responses were analyzed in R using descriptive statistics after exclusion of incomplete or inconsistent responses. Among 525 completed surveys, 450 respondents had ILC and 321 had mILC; 107 (33.3%) presented with de novo mILC. Misdiagnosis was reported by 84 patients (26.2%), including 26 (31.0%) with ≥2 misdiagnoses. Nearly half (44.5%) waited ≥1 year for an accurate diagnosis, 49 received treatment for an incorrect diagnosis, and 6 received incorrect cancer-directed therapy. Symptoms before diagnosis were reported by 138 patients (42.9%), most commonly back pain, fatigue/malaise, gastrointestinal symptoms, bloating, and weight loss. These findings highlight the need for improved diagnostic strategies and clinician awareness to reduce diagnostic delays and errors in mILC.
Triple-negative breast cancer (TNBC) is commonly approached as a single clinical entity despite marked biological heterogeneity. A subset of cases corresponds to rare histologic subtypes that display distinct molecular drivers, immune microenvironments, and clinical behavior. In this Review, we summarize the clinicopathologic and molecular features of rare TNBC histologies and discuss emerging therapeutic vulnerabilities. Integrating histologic classification with molecular profiling may enable more precise risk stratification and guide biomarker-driven treatment strategies.
Metastatic colonization of distant organs remains the primary cause of breast-cancer-specific mortality, yet the underlying mechanisms governing organ-preferential colonization remain poorly understood. To better delineate the underlying mechanisms to establish successful metastatic niches, animal models recapitulating the pathogenesis of human breast cancer are necessary. Through iterative in vivo selection, we established a liver-enriched HER2+ murine breast cancer model, NT2.5LV, which achieved 95% hepatic penetrance within 10 weeks. Comparative whole-genome and bulk RNA sequencing of NT2.5LV, the lung-enriched counterpart NT2.5LM, and the parental primary breast tumor cell line NT2.5 revealed conserved genomic alterations, including Erbb2 amplification, and Cdkn2a/b deletion, despite their distinct organ-preferential phenotypes. NT2.5LM displayed transcriptional homologous-recombination-repair suppression and a type-I interferon-stimulated gene response coupled with immune-checkpoint remodeling, and lung-enriched integrin-mediated adhesions. Conversely, NT2.5LV activated proliferative and metabolic programs with MMP-mediated ECM remodeling suitable to the hepatic microenvironment. These organ-preferential strategies were validated in two independent clinical cohorts MET500 and AURORA. Integrated multi-omic and immunohistochemical analyses demonstrated that NT2.5LV recapitulates key features of human HER2+ breast tumors, including HER2 overexpression. NT2.5LM additionally exhibited a transcriptionally anti-apoptotic, BH3-primed state. The HER2-targeted tyrosine kinase inhibitor neratinib reduced primary tumor growth and the frequency of gross visceral metastases in vivo, establishing NT2.5LV as a clinically relevant preclinical model to assess HER2-targeted therapies.
Inflammatory breast cancer (IBC) is a rare, aggressive disease with poorly understood biology. To provide a comprehensive characterization of blood-based biomarkers in IBC, we retrospectively analyzed 341 metastatic breast cancer patients (83 IBC, 258 non-IBC). Circulating tumor cells (CTCs), CTC clusters, cytokeratin + /CD45+ cells (DPcells), and tumor-derived extracellular vesicles (tdEVs) were evaluated. Matched circulating tumor DNA (ctDNA) data were available for 59 IBC and 110 non-IBC. Analyte distributions and overall survival (OS) association were compared by IBC status. Circulating cellular analytes showed similar distributions, except for lower tdEVs in IBC. CTCs and tdEVs were associated with OS irrespective of IBC status, whereas DPcells only in IBC. TP53 (63%), PIK3CA (29%), and MYC (17%) were the most frequently altered genes in IBC. TP53 sequence variants (SVs; SNVs/indels) and CCNE1/MYC CNVs were enriched in IBC, but after adjusting for confounders, only PIK3CA SVs remained significantly less frequent. Network analysis identified TP53 SVs as key node in IBC, co-occurring with a CNV module (MYC/CCNE1/PIK3CA/FGFR1) with higher centrality in IBC than non-IBC. In multivariable analyses, PIK3CA and MYC CNVs were associated with worse OS. A clinical–ctDNA model showed strong prognostic performance, validated in an independent IBC cohort (n = 39). Overall, integrated liquid biopsy profiling unveiled distinct IBC features and identified prognostic biomarkers.
The preventive value of compression sleeves during docetaxel remains unclear, despite concerns about docetaxel-induced edema increasing breast cancer-related lymphedema (BCRL) risk. This randomized controlled trial evaluated whether prophylactic sleeve use reduces BCRL incidence and characterized docetaxel-induced edema trajectory and quality of life (QoL) impact. Eighty-three women were randomized to an intervention group (IG, n = 41) wearing 15–21 mmHg medical-grade sleeves during and 3 months after docetaxel chemotherapy, or to a usual-care group (UG, n = 42). The primary outcome was 2-year cumulative BCRL incidence, and secondary outcomes included the extracellular water-to-total body water (ECW/TBW) ratio and QoL (FACT-Taxane). Over 2 years, 14 participants (16.9%) developed BCRL with no significant between-group difference (HR 0.91; 95% CI 0.31–2.66; P = 0.869). Preventive sleeves significantly reduced the ECW/TBW ratio and arm volume (P = 0.010 and P = 0.008, respectively), but these reductions did not translate into a lower BCRL incidence. Notably, docetaxel-induced edema resolved spontaneously by 3 months post-chemotherapy in all participants, regardless of group. IG reported significantly worse QoL at 3 and 6 months (Δ = –8.47 and –9.81). The dissociation between edema reduction and lymphatic outcomes suggests that transient docetaxel-induced edema is unlikely to be a primary driver of chronic lymphedema, although a preventive benefit cannot be firmly excluded given the wide confidence interval. These findings do not support routine prophylactic sleeve use and highlight the need for targeted prevention in anatomically high-risk patients.
We developed and validated IICM+, a multimodal model integrating clinicopathologic variables, transcriptomic features, and multiscale histopathology-derived image representations to predict distant recurrence in hormone receptor-positive, HER2-negative, node-negative early breast cancer. Image features included tile-level embeddings and slide-level representations generated by a custom multimodal foundation model pretrained on paired histopathology images, RNA sequencing, and pathology reports. Using TAILORx specimens with long-term follow-up, models were trained in a development cohort with five-fold cross-validation (n = 2808) and evaluated in an independent institutional holdout validation set (n = 1621). In holdout validation, IICM+ showed strong prognostic discrimination for overall distant recurrence (C-index 0.735, 95% CI 0.681–0.782), early distant recurrence (0.791, 95% CI 0.715–0.858), and late distant recurrence (0.710, 95% CI 0.645–0.773). IICM+ separated high- versus low-risk groups for overall distant recurrence (HR 5.25, 95% CI 3.50–7.86; P < 0.001) and remained prognostic (HR 3.56, 95% CI 2.22–5.70, P < 0.001) after adjustment for clinicopathologic covariates and RS category. IICM+ also identified RS/IICM+ discordant groups with different observed recurrence risks, supporting additional prognostic stratification beyond RS.
Human epidermal growth factor receptor 2 positive (HER2+) and estrogen receptor positive (ER+) breast cancer represents a distinct subtype characterized by complex receptor crosstalk, which can promote therapeutic resistance. Dual targeting of HER2 and ER has shown potential both pre-clinically and in patients. However, investigations of optimal drug combinations remain limited. The advent of novel small molecule therapies and next-generation hormonal treatments suggests the need for additional combinatorial studies. In this report, we demonstrate that combining the pan-HER tyrosine kinase inhibitor neratinib with next-generation, orally available selective estrogen receptor degraders (SERDs), particularly camizestrant, elicits potent synergistic anti-tumor effects in HER2+/ER+ pre-clinical models. Neratinib and camizestrant demonstrated strong synergy in vitro across multiple cell lines, including HER2-low and anti-HER2 therapy-resistant models. The combination showed the greatest induction of apoptosis and inhibition of 3D spheroid growth compared to single agents and other neratinib plus SERD combinations. Neratinib treatment alone induced compensatory increases in ER-associated gene expression, which were reversed by camizestrant co-treatment. In vivo, neratinib plus camizestrant resulted in significant tumor regression, with all treated mice showing tumor shrinkage and marked suppression of pro-tumorigenic signaling. These findings support the clinical evaluation of neratinib and camizestrant as a promising new strategy for patients with HER2+/ER+ breast cancer.
To evaluate the prognostic role of patient-reported outcomes (PROs) in nonmetastatic breast cancer, we conducted secondary analysis of 2763 women in A011502, a randomized trial of adjuvant aspirin vs. placebo. High stress (Perceived Stress Scale) was associated with worse invasive disease-free survival (iDFS) (adjusted hazard ratio (aHR) 2.22) and distant recurrence-free survival (DRFS) (aHR 2.3). Moderate/severe pain (Brief Pain Inventory) predicted worse iDFS (aHR 1.67), DRFS (aHR 1.83), distant recurrence-free interval (aHR 1.53) and overall survival (aHR 2.21). PROs may enhance risk stratification in survivorship care. Trial Registration: ClinicalTrials.gov: NCT02927249.
