Disease recurrence following definitive therapy remains one of the primary drivers of morbidity and mortality in early-stage breast cancer, presenting an ongoing clinical obstacle for patients who experience relapse despite receiving standard interventions. Up to one-third of individuals who achieve disease-free status after initial treatment eventually develop recurrent tumors, and patients facing metastatic relapse experience a five-year survival rate of approximately thirty percent. Although cell surface–targeted therapeutics—such as antibody-drug conjugates, radiopharmaceutical agents, and immune-directed modalities—have demonstrated substantial clinical efficacy across multiple malignancies, current clinical options in breast cancer remain concentrated on a limited repertoire consisting predominantly of HER2 and TROP2. To address this unmet need, an international research team investigated seventy-eight cell surface targets currently undergoing evaluation in clinical trials, analyzing extensive patient cohorts to profile surface protein expression specifically within treatment-resistant early-stage tumors. Their systematic evaluation across diverse breast cancer molecular subtypes identifies candidate targets for therapeutic repositioning, establishing a foundational framework to guide treatment intensification.
Cell surface–targeted therapies provide a targeted anti-cancer approach by exploiting proteins differentially displayed on malignant cells, thereby maximizing tumor specificity while curtailing systemic toxicity compared to conventional cytotoxic regimens. This therapeutic class encompasses diverse platforms, including antibody-drug conjugates that deliver potent cytotoxic agents via monoclonal antibodies, radiopharmaceutical therapies delivering localized radionuclides, and cell-based or immune-engaging platforms such as chimeric antigen receptor T cells, natural killer cells, macrophages, and bispecific T-cell engagers. In breast oncology, four antibody-drug conjugates have secured regulatory authorization across localized and metastatic settings, all targeting either HER2 or TROP2. However, both HER2 and TROP2 exhibit broad baseline expression across healthy normal tissues, which imposes substantial constraints on therapeutic dosing, increases off-target adverse profiles, and complicates clinical treatment sequencing. Because dozens of alternative cell surface targets are currently under active evaluation in clinical trials for other solid tumors, repositioning these established drug candidates offers an attractive avenue to address treatment-resistant breast disease without requiring de novo molecular drug design.
To identify viable targets with strong translational potential, the investigators integrated high-throughput transcriptomic and proteomic data from several rigorously characterized clinical cohorts and biological model repositories. The primary investigative cohort drew from the population-based Sweden Cancerome Analysis Network–Breast initiative, encompassing nearly ten thousand patients with annotated long-term clinical trajectories. For evaluating responses to neoadjuvant interventions, the team gathered an aggregated cohort of nearly four thousand patients across multiple published neoadjuvant chemotherapy studies. Radiation-specific outcomes were evaluated using clinical and genomic records from the randomized phase III SweBCG91RT trial, which monitored seven hundred sixty-five early-stage breast cancer patients receiving breast-conserving management with or without adjuvant radiation. To determine whether transcript levels reliably reflect target protein availability, the authors evaluated paired RNA and protein measurements from primary tumors in the Clinical Proteomic Tumor Analysis Consortium and cell lines from the Sanger Cell Model Passports. Single-cell transcriptomic data from fifty-six primary breast tumors within the Curated Cancer Cell Atlas provided high-resolution assessments of intratumoral heterogeneity.
The analytical architecture utilized uniform computational standards to identify clinically meaningful expression levels across the distinct patient cohorts. Because universally standardized biological thresholds separating high from low surface expression are absent for most target genes in breast cancer, the investigators designated the upper quartile of normalized gene expression within each dataset as the high-expression cohort. Defining the top twenty-five percent as elevated expression ensured a sufficiently substantial subpopulation to support realistic clinical evaluation and future trial design. Furthermore, targets demonstrating minimal expression across most tumors were filtered out by mandating that the twenty-fifth percentile of target expression exceed the cohort-wide median of all profiled genes. Expression values were scaled from zero to one within each specimen, allowing direct cross-assay comparability and clear interpretation. Time-to-event outcomes, including recurrence-free survival, were evaluated through Cox proportional hazards regression modeling, whereas categorical therapeutic responses were analyzed using Fisher’s exact test. To control false-positive rates arising from multiple statistical evaluations, the researchers applied the Benjamini-Hochberg adjustment with a significance threshold of two-sided p-values below 0.05.
A central prerequisite for leveraging transcriptomic registries in surface target discovery is verifying that messenger RNA abundance accurately mirrors functional surface protein synthesis in clinical specimens. Evaluating the seventy-eight prioritized cell surface targets across tumor samples from the Clinical Proteomic Tumor Analysis Consortium demonstrated strong concordance between upper-quartile RNA abundance and upper-quartile protein detection, achieving a median predictive accuracy of eighty percent. Analysis of cancer cell lines from the Sanger Cell Model Passports corroborated these findings, yielding a median predictive accuracy of ninety-four percent across investigated surface targets. Statistical assessments also confirmed that expression profiles among these seventy-eight targets were generally not inter-correlated, demonstrating that individual markers represent discrete molecular subsets rather than uniform transcriptional modules. These robust concordances confirmed that high transcript expression serves as a dependable surrogate for elevated cell surface protein expression, justifying the application of large-scale clinical RNA datasets to uncover candidate surface receptors in treatment-refractory disease.
Within estrogen receptor–positive and HER2-negative tumors, which represent the largest clinical subgroup, adjuvant endocrine therapy substantially mitigates recurrence risk yet fails to prevent metastatic relapse in up to forty percent of high-risk patients over twenty years. Evaluating 5,645 endocrine-treated patients within the SCAN-B cohort revealed that elevated expression of ERBB2 and TACSTD2 was unexpectedly associated with superior recurrence-free survival rather than therapy failure. Instead, the analysis identified nine alternative surface targets whose high expression correlated significantly with compromised recurrence-free survival. Most of these candidates—including CEACAM5, CEACAM6, EPCAM, NECTIN4, CD276, FLT1, and FN1—retained statistically significant associations after adjusting for standard clinicopathologic factors such as patient age, lymph node status, tumor grade, pathological size, and PAM50 molecular intrinsic subtype. These findings align with prior biological literature documenting elevated protein levels of EPCAM, NECTIN4, and CEACAM5 in aggressive luminal tumors, while supporting experimental evidence implicating CEACAM6 and fibronectin in tamoxifen resistance and post-treatment relapse.
The study uniquely characterized surface targets associated with radiotherapy resistance by examining the randomized phase III SweBCG91RT clinical trial. Because very few patients in this trial received adjuvant endocrine therapy or systemic chemotherapy, the cohort offered an isolated window into radiation-specific outcomes without the confounding influence of systemic treatments. Focusing on 590 estrogen receptor–positive participants, the researchers tracked targets from the SCAN-B analysis whose elevated expression predicted inferior recurrence-free survival specifically following adjuvant radiation. Two targets emerged with clear treatment-specific associations: EPCAM and CEACAM5. High expression of both genes was associated with adverse survival outcomes solely in the irradiated arm, showing no prognostic detriment among patients managed without radiotherapy. Statistical modeling revealed a significant interaction between radiotherapy and EPCAM expression with an unadjusted p-value of 0.02 and a multivariate-adjusted p-value of 0.022. CEACAM5 displayed a borderline interaction with p-values of 0.073 unadjusted and 0.071 adjusted, indicating that elevated levels of these surface molecules mark tumors possessing intrinsic radioresistance.
In early-stage HER2-positive breast cancer, standard interventions combining anti-HER2 antibodies such as trastuzumab and pertuzumab with chemotherapy have markedly improved patient survival, yet residual disease following neoadjuvant therapy continues to carry relapse rates reaching thirty to forty percent. Among 834 HER2-positive patients receiving anti-HER2 regimens in the SCAN-B cohort, neither ERBB2 nor TACSTD2 predicted post-treatment recurrence. In contrast, elevated expression of FOLH1 and TNFRSF10B was strongly associated with shortened recurrence-free survival, retaining significance following adjustment for clinical grade, age, nodal involvement, tumor diameter, and PAM50 classification. FOLH1 encodes prostate-specific membrane antigen, an established therapeutic target in metastatic prostate cancer where the radiopharmaceutical lutetium-177-PSMA-617 and imaging agent gallium-68-PSMA-11 are clinically approved. While PSMA has been observed in breast tumor vasculature, this investigation provides the first evidence linking high FOLH1 expression to anti-HER2 therapeutic relapse. Meanwhile, TNFRSF10B encodes the apoptosis-inducing receptor TRAILR2, which preclinical investigations have previously combined with anti-HER2 antibodies to overcome trastuzumab resistance in murine models.
For patients diagnosed with triple-negative breast cancer, residual disease following neoadjuvant chemotherapy or chemoimmunotherapy heralds an aggressive disease course, with median survival following metastatic relapse falling below twenty-four months. Analyzing pre-treatment transcriptomic profiles from 1,686 patients with triple-negative disease revealed that CD274, ERBB2, and TACSTD2 expression correlated with pathologic complete response rather than treatment refractoriness. Conversely, the authors uncovered six alternative cell surface targets significantly enriched in pre-treatment biopsies from tumors that subsequently exhibited residual disease. Protein expression for five of these six candidates—including ALCAM, GRPR, STEAP1, ADAM9, and CEACAM6—has been observed in clinical triple-negative breast cancer specimens, though their prognostic association with residual disease had remained undefined. Notably, certain markers such as ADAM9 and CEACAM6 overlapped across both chemotherapy-resistant triple-negative disease and endocrine-resistant luminal tumors. This shared enrichment suggests conserved biological mechanisms mediating therapy resistance across disparate molecular subtypes, presenting coherent opportunities to deploy repositioned surface-targeted agents against persistent residual disease.
Single-cell transcriptomic profiling across fifty-six breast cancers demonstrated critical variations in intratumoral target distribution, providing practical guidance for payload selection in drug development. While EPCAM and TROP2 showed relatively uniform cellular expression across individual tumors, markers such as CEACAM5, CEACAM6, NECTIN4, FN1, FOLH1, STEAP1, and TNFRSF10B displayed marked intratumoral heterogeneity. Targets characterized by heterogeneous distribution may be ill-suited for immunomodulatory approaches requiring direct cellular contact, but they remain viable candidates for radiopharmaceuticals or antibody-drug conjugates equipped with cleavable linkers capable of exerting regional bystander killing. The authors noted key study limitations, including the reliance on transcriptomic surrogates rather than direct proteomic quantification and the uncertain threshold of surface protein required to elicit therapeutic responses. Because comprehensive immunohistochemical screening remains constrained by tissue availability and expense, commercial RNA-sequencing platforms may provide a scalable initial screen to identify candidate targets for subsequent proteomic validation, expanding therapeutic possibilities for patients facing resistant breast cancer.
Subject of Research: Cell surface target expression and therapeutic repositioning in treatment-resistant early-stage breast cancer
Article Title: The landscape of cell surface target expression in treatment resistant early-stage breast cancer
Article References: Karlsson, P., Sharifi, M. N., Holmberg, E., Bakhtiar, H., Niméus, E., Killander, F., Sjöström, M., & Zhao, S. G. (2026). The landscape of cell surface target expression in treatment resistant early-stage breast cancer. Breast Cancer Research and Treatment, 220(1), Article 34. https://doi.org/10.1007/s10549-026-08120-2
Image Credits: AI Generated
DOI: 10.1007/s10549-026-08120-2
Keywords: breast cancer, cell surface targets, antibody-drug conjugates, treatment resistance, oncology, genomics, radiotherapy resistance, targeted therapy, landscape, cell, surface, target
