A retrospective investigation examining 280 patients with pathologically confirmed urothelial carcinoma has identified significant variations in human epidermal growth factor receptor 2 protein levels across distinct tumor tissue phenotypes. Conducted at the Third Clinical Medical College of Xinjiang Medical University, the single-center research evaluated how membranous staining for the tyrosine kinase receptor correlated with patient age, histological grade, and cytokeratin markers. Across the entire clinical cohort, elevated protein levels characterized as immunohistochemical scores of 2+ or 3+ were identified in 191 cases, representing 68.2 percent of the analyzed specimens. Researchers observed that while elevated expression was widespread throughout the cohort, it appeared disproportionately among specific tumor categories defined by cytokeratin 5/6 and cytokeratin 20 staining profiles.
Urothelial carcinoma represents a common malignancy of the urinary tract, predominantly affecting the bladder in an estimated 90 to 95 percent of cases, while the remaining 5 to 10 percent arise within the upper urinary tract. In the introductory framework of the study, the investigators cited global estimates from 2022 documenting 614,298 new cases of bladder cancer alongside 220,596 associated fatalities. Clinicians face considerable therapeutic hurdles when tumors progress to muscle-invasive or metastatic stages, where traditional platinum-based systemic chemotherapy remains the established standard of care yet proves unsuitable or ineffective for a substantial proportion of patients. Because the human epidermal growth factor receptor 2 modulates cellular proliferation and survival pathways, its protein overexpression and underlying gene alterations have gained attention as actionable therapeutic targets, particularly given recent activity reported for receptor-targeted antibody-drug conjugates in advanced disease.
To investigate how receptor presentation interacts with broader tissue architecture, the pathology team categorized specimens into four distinct immunohistochemical phenotypes using cytoplasmic staining thresholds for cytokeratin 5/6 and cytokeratin 20. Staining in more than 20 percent of viable tumor cells was established as the criterion for positivity. Under this framework, tumors presenting cytokeratin 20 positivity without cytokeratin 5/6 were designated as luminal, while those displaying cytokeratin 5/6 without cytokeratin 20 were classified as basal. Specimens demonstrating reactivity for both cytokeratins were grouped as double-positive, and those showing neither marker were termed double-negative. This four-tier surrogate system offers an accessible, cost-effective alternative to comprehensive transcriptomic profiling, which categorizes molecular categories like luminal and basal types but remains difficult to implement within daily diagnostic routines.
The cohort enrolled consecutive eligible patients treated between January 2021 and July 2025, comprising 205 men and 75 women with an overall average age of 65.8 years and a median age of 66 years. Tissue specimens were procured through diagnostic biopsies, transurethral resections, or formal radical surgeries. After excluding 10 cases characterized by synchronous disease in both the bladder and the upper urinary tract, primary tumors originated in the urinary bladder in 219 patients, while 51 tumors developed in the upper tract. High-grade histology was identified in 234 patients, constituting 83.6 percent of the study group, whereas low-grade tumors occurred in 46 individuals. Among 58 patients who underwent radical resection and comprehensive pathological staging under the 2017 American Joint Committee on Cancer guidelines, 13 tumors were staged as T1, 11 as T2, 27 as T3, and 7 as T4.
Laboratory evaluation utilized formalin-fixed, paraffin-embedded tissue blocks sectioned at 4-micrometer thickness and processed on an automated staining platform with dedicated monoclonal antibodies. Two experienced pathologists evaluated membranous receptor staining while blinded to patient clinical backgrounds, resolving any scoring discrepancies through consensus review alongside a third pathologist. Membrane reactivity was graded using established scoring criteria: scores of 0 indicated either absent staining or faint, incomplete membrane coloration in fewer than 10 percent of cells, while 1+ designated faint, incomplete staining in at least 10 percent of cells. Tumors exhibiting weak-to-moderate staining in 10 percent or more of cells, or strong complete staining in under 10 percent, were assigned an intermediate score of 2+, whereas strong, complete membranous staining in at least 10 percent received a score of 3+.
Evaluation of the baseline scoring revealed that 32 tumors were scored as 0, 57 as 1+, 174 as 2+, and 17 as 3+. When assessing observer reproducibility across the four individual score tiers, the pathologists recorded an agreement rate of 77.9 percent, yielding a linearly weighted Cohen’s kappa coefficient of 0.70. When consolidating the scores into a binary clinical grouping of 0 or 1+ versus 2+ or 3+, agreement rose to 88.6 percent with an unweighted Cohen’s kappa of 0.74. Disagreements among observers occurred in 62 cases overall, though only the 32 cases of discordance between 1+ and 2+ influenced the binary categorization. Furthermore, among 50 tumors with an intermediate score of 2+ that had previously undergone dual-probe fluorescence in situ hybridization during clinical care, ERBB2 gene amplification was confirmed in 12 specimens, corresponding to a 24.0 percent positive amplification rate in that subset.
Marked disparities in receptor overexpression emerged when evaluating tumors by their cytokeratin phenotypes. Staining categorized as 2+ or 3+ was detected in 78.5 percent of luminal tumors and 84.8 percent of double-positive tumors, contrasting sharply with rates of 53.1 percent in the basal subgroup and 48.4 percent among double-negative cases. Univariable statistical testing revealed that elevated staining was also more prevalent among patients older than 60 years compared to those aged 60 or younger, showing rates of 74.2 percent versus 55.6 percent. High-grade tumors demonstrated elevated receptor levels more frequently than low-grade lesions, at 71.4 percent compared to 52.2 percent. Elevated receptor expression also correlated with higher cellular proliferative activity, marked by a Ki-67 index of 30 percent or greater, and an abnormal nuclear p53 staining pattern.
Multivariable logistic regression modeling was subsequently performed to isolate independent predictors of elevated receptor expression, controlling for potential confounding factors. Within this adjusted model, patient age greater than 60 years remained an independent predictor, carrying an odds ratio of 1.888 with a 95 percent confidence interval spanning 1.054 to 3.380. The cytokeratin classifications also retained independent significance when compared against the double-negative baseline: luminal tumors exhibited an odds ratio of 3.848 with a 95 percent confidence interval of 1.930 to 7.671, while double-positive tumors showed an odds ratio of 3.429 with a confidence interval of 1.204 to 9.762. In contrast, histological grade, abnormal p53 patterns, and elevated Ki-67 proliferative indices did not maintain independent statistical significance in the multivariable analysis.
The study authors highlighted that the anatomical distribution of elevated receptor scores did not directly parallel typical adverse histopathological markers across phenotypes. Basal tumors, while showing comparatively lower rates of elevated receptor expression, displayed the greatest proportion of invasive disease, accounting for muscle invasion in 93.8 percent of radical resection specimens within that subset and 6 of the 7 documented instances of perineural invasion. Conversely, double-positive tumors demonstrated a unique convergence, displaying both high rates of receptor overexpression and unfavorable histopathological traits, including high tumor grade and elevated Ki-67 indexes. This dual presentation may reflect mixed cellular populations expressing overlapping migratory and luminal features, underscoring the complex biological heterogeneity present within urothelial neoplasms.
The researchers noted several key constraints inherent to their investigation. Because data were gathered retrospectively from a single medical center, institutional referral patterns likely contributed to the high proportion of upper tract cases, which reached nearly 19 percent. In addition, a notable portion of tumor samples was obtained through superficial biopsy rather than complete radical excision, restricting comprehensive depth assessment and reducing statistical power for advanced pathological staging analyses. The authors also underscored that patient survival metrics and therapeutic follow-up were unrecorded, precluding direct conclusions regarding clinical prognosis or treatment efficacy. Nevertheless, the study demonstrates that pairing standard cytokeratin immunohistochemistry with receptor testing provides a practical, standardized approach for characterizing phenotypic diversity in clinical pathology.
Subject of Research: HER2 protein expression and clinicopathological correlates across immunohistochemical phenotypes in urothelial carcinoma
Article Title: HER2 expression and clinicopathological correlates across immunohistochemical phenotypes of urothelial carcinoma: a retrospective single-center study
Article References: Peng, Y., Zhang, Y., Miao, G., & Guo, Y. (2026). HER2 expression and clinicopathological correlates across immunohistochemical phenotypes of urothelial carcinoma: a retrospective single-center study. Frontiers in Medicine, 13, Article 2015460. https://doi.org/10.3389/fmed.2026.2015460
Image Credits: AI Generated
DOI: 10.3389/fmed.2026.2015460
Keywords: urothelial carcinoma, HER2 expression, immunohistochemistry, cytokeratin 20, cytokeratin 5/6, luminal phenotype, basal phenotype, double-positive phenotype, double-negative phenotype, ERBB2 amplification, fluorescence in situ hybridization, bladder cancer, upper tract urothelial carcinoma, Ki-67 index, p53 expression, antibody-drug conjugates, pathology
News Source: Nathaniel Bowman. (November 16, 2026). Study Links High HER2 Protein Expression Across Specific Urothelial Carcinoma Phenotypes. Scienmag.



