High prevalence of TROP2 expression identifies EBV-positive gastric cancer as a candidate for TROP2-directed antibody–drug conjugates

Introduction

Gastric cancer ranks as the fifth most common malignancy worldwide and the third leading cause of cancer-related death [1, 2]. For patients with unresectable advanced or recurrent disease, median overall survival remains limited to 13–20 months despite multimodal.

therapy [3, 4, 5]. Molecular profiling by The Cancer Genome Atlas (TCGA) has revealed substantial genomic heterogeneity within gastric adenocarcinoma, identifying four distinct subtypes: chromosomal instability (CIN), microsatellite instability (MSI), genomically stable (GS), and Epstein-Barr virus (EBV)-positive [6]. These molecular distinctions have informed biomarker-directed treatment strategies, including HER2-targeted therapy with trastuzumab [7],t claudin 18.2-targeted therapy with zolbetuximab [5, 8], and immune checkpoint inhibitors, whose benefits is greatest in MSI-high tumors [9] and tumors with high PD-L1 expressions [3, 4, 7, 10, 11].

Nevertheless, objective response rates with current regimens remain modest (40–60%) and median progression-free survival spans only 6.9–10.5 months [3, 12, 13]. These outcomes highlight a persistent unmet need to identify and validate novel therapeutic targets.

EBV-positive gastric cancer constitutes approximately 5–10% of gastric malignancies globally [14]. With nearly one million gastric cancers diagnosed worldwide each year, this amounts to an estimated 80,000 EBV-associated cases annually surpassing EBV-associated Hodgkin lymphoma and making it the most prevalent EBV-associated neoplasm [12, 14]. Clinically, these tumors are characterized by preferential localization to the proximal stomach, a male predominance, a younger age at presentation (< 60 years), and recurrent somatic mutations [13]. At the epigenetic level, EBV-positive gastric cancers exhibit the most extensive DNA hypermethylation patterns observed among solid malignancies [15]. Paradoxically, despite these aggressive molecular features, EBV-positive gastric cancers demonstrate a more favorable prognosis and a significantly lower frequency of lymph node metastasis compared to their EBV-negative counterparts, findings largely attributed to a highly immunogenic tumor microenvironment [8, 13].

Trophoblast cell surface antigen 2 (TROP2) encoded by the tumor-associated calcium signal transducer 2 (TACSTD2) has emerged as a promising therapeutic target across epithelial malignancies, including gastric adenocarcinoma [16, 17 ,18, 19]. TROP2 is a 36-kDa type I transmembrane glycoprotein that functions as a calcium signal transducer and modulates multiple oncogenic signaling cascades [20]. In gastric adenocarcinoma, TROP2 overexpression is observed in approximately 56% of cases and serves as an independent adverse prognostic marker for disease recurrence, correlating with decreased disease-free survival, particularly in intestinal-type histology, lymph node-positive disease, and advanced-stage tumors [21].

Despite the established significance of TROP2 as a therapeutic target in the broader cancer population [22, 23, 24, 25] and the distinctive molecular and immunological features of EBV-positive gastric cancer [6, 12, 15], a critical knowledge gap persists regarding TROP2 expression patterns, prognostic implications, and therapeutic relevance specifically within the EBV-positive molecular subtype. Several lines of evidence suggest that TROP2-directed antibody-drug conjugates (ADCs) may be particularly effective in this context. EBV-positive tumors are characterized by dense CD8 + T-cell infiltration, high tumor mutational burden, PD-L1/PD-L2 overexpression, and clinical responsiveness to immune checkpoint blockade [9, 26, 27] features that collectively define an immune-permissive microenvironment. TROP2-targeted ADCs can induce immunogenic cell death (ICD) through the release of damage-associated molecular patterns (DAMPs) [28, 29], a mechanism that may synergize with this pre-existing antitumor immune milieu to augment therapeutic efficacy beyond direct cytotoxicity. Furthermore, the bystander killing effect of TROP2-targeted ADCs may confer an additional advantage in EBV-positive tumors, extending cytotoxic activity to antigen-heterogeneous tumor cell.

populations [30]. Critically, TROP2 expression itself determines ADC efficacy. In esophageal adenocarcinoma, 88% of 598 tumors expressed TROP2 and complete absence of tumoral TROP2 was associated with a lack of response to the TROP2-directed ADC Sacituzumab Govitecan (SG) in vitro and in vivo, leading to advocate TROP2 testing at the protein level before therapy [31]. Consistently, in the phase III ASCENT trial in metastatic triple-negative breast cancer, response rates to SG declined with decreasing TROP2 expression [32]. Establishing the prevalence of TROP2 loss therefore defines the proportion of patients eligible for TROP2-directed ADCs. To address this gap, we investigated the relationship between EBV status and TROP2 expression in gastric adenocarcinoma and validated the expression pattern in the publicly available TCGA cohort and explored their combined prognostic significance.

Materials and methods

Results

Discussion

This study demonstrates that TROP2 expression varies significantly across TCGA molecular subtypes of gastric adenocarcinoma, with EBV-positive tumors exhibiting the highest median expression compared with CIN, GS, and MSI-H subtypes. EBV-positive tumors also demonstrated the most abundant expression of TROP2 with only 3.7% of the patients showing no expression at all. This subtype-specific pattern was reproduced at the mRNA level in the independent TCGA cohort, in which EBV-positive tumors were enriched among the highest TACSTD2-expressing cases. Within the EBV-positive subgroup, overall survival did not differ significantly by TROP2 status.

In the overall cohort, TROP2-high expression showed a trend toward shorter median OS (25.0 vs. 32.0 months), without reaching statistical significance, consistent with the heterogeneous prognostic significance reported for TROP2 in gastric cancer [17, 21, 31, 35]. Taken together, these data suggest that the clinical relevance of TROP2 in gastric cancer lies primarily in its role as a therapeutic target rather than as a prognostic marker.

Mühlmann et al. demonstrated that TROP2 overexpression was an independent predictor of disease recurrence in intestinal-type gastric cancer [21], while a meta-analysis by Zeng et al. confirmed the association between TROP2 overexpression and poor OS across solid tumors (pooled HR, 1.896; 95% CI, 1.60–2.25) [36]. More recently, Kim et al. reported that high TROP2 expression was significantly associated with EBV positivity, intestinal-type histology, and lymph node metastasis, and was an independent predictor of poor survival on multivariate analysis [35]. Our study extends these observations, which were derived from an Asian cohort, to a Western population and quantifies the frequency of complete TROP2 absence across molecular subtypes.

In contrast to the favorable prognosis of EBV-positive gastric cancer reported in the literature, median OS of EBV-positive tumors in our cohort was not longer than that of CIN tumors, which may reflect the small subgroup size and the characteristics of this Western, surgically treated cohort. EBV-positive gastric cancers are characterized by dense CD8 + T-cell infiltration, PD-L1/PD-L2 overexpression, and recurrent PIK3CA mutations, features that collectively define an immune-permissive microenvironment and likely account for the overall favorable prognosis of this subtype [9, 26, 34, 35]. Within the EBV-positive subgroup, overall survival did not differ significantly by TROP2 status. Although TROP2 has been linked to tumor-intrinsic oncogenic signaling, including activation of EGFR/AKT and disruption of desmosomal adhesion via DSG2 downregulation, our exploratory data do not allow conclusions on a prognostic role of TROP2 in this subtype [20, 35, 36, 37]. Notably, CD8 + T-cell infiltration has been reported to exert subtype-specific prognostic effects in gastric cancer. From a therapeutic perspective, the high prevalence of TROP2 expression in EBV-positive gastric cancer (96.3% positivity) positions this subtype as a candidate population for TROP2-directed ADCs. Also, the high prevalence of TROP2 in the overall cohort (approximately 80%) suggests that targeting TROP2 may represent a therapeutic option for gastric cancer patients. TROP2-targeted ADCs, including Sacituzumab Govitecan and Datopotamab deruxtecan, deliver topoisomerase I inhibitor payloads to TROP2-expressing tumor cells and exert additional antitumor activity through the bystander killing effect, whereby released payload diffuses to adjacent antigen-negative cells [18, 37, 38]. Notably, complete absence of TROP2, the expression pattern shown to abrogate SG efficacy in esophageal adenocarcinoma, was found in only 3.7% of EBV-positive tumors, indicating that nearly all patients within this subtype would be eligible for TROP2-directed ADC therapy. This distinguishes EBV-positive tumors from other subtypes: although TROP2 is expressed in the majority of gastric cancers, approximately 20% of CIN tumors, one quarter of GS tumors and more than one third of MSI-H tumors lack TROP2 entirely and would therefore be unlikely to benefit from TROP2-directed ADCs. The clinically relevant difference between subtypes thus lies less in the level of expression than in the proportion of patients who are potentially eligible for treatment. For MSI-H tumors, which appear least suited to TROP2-directed therapy, immune checkpoint inhibitors already represent an established option. The frequency of TROP2 expression observed here is comparable to that reported in esophageal adenocarcinoma, where 88% of 598 tumors were TROP2-positive [31]. This bystander effect may be particularly advantageous in EBV-positive tumors, which can exhibit heterogeneous TROP2 expression. Furthermore, TROP2-targeted ADCs induce immunogenic cell death by releasing damage-associated molecular patterns, including calreticulin, ATP, and HMGB1, thereby promoting dendritic cell maturation and cytotoxic T-lymphocyte activation [28, 29]. This mechanism may synergize with the pre-existing immune-permissive microenvironment of EBV-positive gastric cancer, potentially augmenting therapeutic efficacy beyond direct cytotoxicity. Preclinical evidence supports this rationale: ADC-mediated immunogenic cell death, combined with immune checkpoint blockade, has demonstrated synergistic antitumor responses across multiple tumor models [39]. Given that EBV-positive gastric cancers already exhibit high PD-L1 expression and clinical responsiveness to immune checkpoint inhibitors, combining TROP2-directed ADCs with checkpoint blockade warrants prospective investigation in this molecularly defined population [9, 26, 35].

With regard to immunotherapy, we envision two complementary scenarios. Although early data suggested high response rates to PD-1 blockade in EBV-positive gastric cancer, these observations are based on very small numbers, and EBV status has not been established as a predictive biomarker for immune checkpoint inhibitors [27, 40]. Patients who do not respond to or progress on immune checkpoint inhibitors therefore represent a clinically defined population with limited options. In the near term, TROP2-directed ADCs appear most relevant as a non-immunological treatment option in this setting, in line with the ongoing evaluation of TROP2-directed ADCs in pretreated gastroesophageal adenocarcinoma [41]. In the longer term, the capacity of ADCs to induce immunogenic cell death provides a rationale for combining TROP2-directed ADCs with checkpoint blockade in this immune-permissive subtype, a strategy that has recently improved progression-free survival compared with checkpoint blockade alone in first-line non-small-cell lung cancer [42]. Because nearly all EBV-positive tumors express TROP2, EBV status could be recorded or used as a stratification factor in clinical trials of TROP2-directed ADCs in gastric cancer to test these hypotheses prospectively.

Several limitations of this study merit consideration. First, this is a retrospective, single-center analysis, which introduces potential selection bias and limits generalizability. The molecular subtype distribution may not reflect that of other geographic populations; in particular, the proportion of EBV-positive tumors in our cohort (5.1%) lies at the lower end of the 5% to 18% reported worldwide. Second, the EBV-positive subgroup comprised only 27 patients, limiting statistical power and precluding multivariate adjustment for potential confounders, including stage, treatment modality, and histological subtype. .

Consequently, the survival analysis within the EBV-positive subgroup (n = 18 with follow-up; p = 0.123) is exploratory and underpowered, and the wide confidence intervals preclude conclusions on a prognostic role of TROP2 in this subtype. This limitation reflects the rarity of EBV-positive gastric cancer, particularly in Western populations; even the largest publicly available dataset, TCGA-STAD, includes only 30 EBV-positive tumors with survival data, in which TACSTD2 expression was likewise not associated with overall survival. Third, the TCGA validation is based on bulk mRNA expression, which does not capture membranous protein localization or complete protein absence and may underestimate epithelial TROP2 expression in lymphocyte-rich EBV-positive tumors. Fourth, the H-score cut-off of 100 used to define TROP2-high expression has not been validated for patient selection for TROP2-directed ADCs. Multicenter or consortium-based cohorts will be required to address these questions.

In conclusion, these findings support the assessment of TROP2 across molecular subtypes, are supported by independent TCGA data at the mRNA level and provide a rationale for investigating TROP2-directed ADCs in EBV-positive gastric cancer, particularly in patients who fail immune checkpoint inhibitors and, in the longer term, in combination with checkpoint blockade. Prospective, multicenter studies with larger EBV-positive cohorts are needed to validate these observations and to determine whether TROP2 expression has predictive value for ADC-based therapies in this molecularly defined population.

Supplementary information

Below is the link to the electronic supplementary material.

Author contributions

All authors contributed to the conception and overall design of the study. T. Wagner led the writing of the manuscript, planed the experimental read out and performed the statistical analysis. R. Weiten conducted the experiments and contributed to experimental read-outs. A. Pamuk contributed to data acquisition. A. Quaas and S.I. Lyu provided pathological expertise and contributed to data interpretation. C. J. Bruns and H. Alakus contributed to performing the surgical procedures and to providing surgical and clinical expertise. A. Heidenreich critically reviewed the manuscript. N. Schleussner critically revised the manuscript for important intellectual content and assisted with the statistical analysis. All authors reviewed, read, and approved the final version of the manuscript.

Funding

Open Access funding enabled and organized by Projekt DEAL. This work was supported by internal funds from the University Hospital Cologne. R. Weiten is supported by a Ferdinand Eisenberger grant from the Deutsche Gesellschaft für Urologie (German Society of Urology; grant ID WeR1/FE-22). T. Wagner is supported by the Walther Benjamin Program of the Deutsche Forschungsgemeinschaft (DFG; project number 572479455).

Data availability

The datasets generated and analysed during the current study are not publicly available due to data protection regulations and patient privacy but can be made available by the corresponding author upon reasonable request.

Declarations