Revisiting the pseudocapsule in pancreatic neuroendocrine tumors: a structural marker for surgical strategy and prognosis

Dear Editor, We read with great interest the study by Wang et al[1], titled “Evaluation of tumour pseudocapsule using computed tomography-based radiomics in pancreatic neuroendocrine tumours to predict prognosis and guide surgical strategy: a cohort study.” This study investigates the histopathological feature of the tumour pseudocapsule (PS) in pancreatic neuroendocrine tumours (PanNETs), and explores its potential role as a structural biomarker to guide surgical decision-making and prognostic stratification. The authors proposed an integrative approach combining clinical, pathological, radiological, and molecular data to assess PS status, and further developed a CT-based radiomics model to predict this feature preoperatively. From a methodological perspective, the study stands out for its multidimensional design. Clinical and pathological variables provided a baseline biological profile, while single-cell RNA sequencing (scRNA-seq) delineated immune microenvironmental differences associated with PS status. Radiomics analysis extended the assessment into the preoperative setting, enabling a non-invasive surrogate for structural prediction. Notably, the large cohort (n = 475) and a median follow-up of over 1200 days lend statistical robustness to the survival analyses. The integration of these complementary data sources enhances the interpretability of results and underscores the tumor’s phenotypic and microenvironmental heterogeneity. The study’s findings suggest that incomplete pseudocapsule is significantly associated with more aggressive pathological features—higher T stage, poorer tumor differentiation, perineural invasion—and correlates with worse recurrence-free survival and overall survival. The proposed radiomics model demonstrated promising performance in predicting PS status across tumor size subgroups (AUCs ranging from 0.72 to 0.75), highlighting its potential clinical utility. Most importantly, for small PanNETs (≤2 cm)—where the choice between observation and resection remains controversial—this structural indicator may offer a novel preoperative reference to support decisions around enucleation, helping to balance the preservation of pancreatic function against oncologic risk. Importantly, the authors’ use of scRNA-seq adds mechanistic insight into the biological underpinnings of PS integrity. Tumors lacking a complete pseudocapsule exhibited a more immunosuppressive microenvironment, with a relative depletion of T and B lymphocytes and enrichment of endothelial and cancer-associated fibroblast populations. These findings hint at a more invasive and immunologically inert phenotype. While causality remains to be fully elucidated, the study raises the hypothesis that pseudocapsule formation may reflect the outcome of tumor-stroma-immune interactions—a direction worthy of future mechanistic investigation. The clinical implications are also notable. Enucleation, though function-preserving, poses challenges in margin control and postoperative complications. A reliable, image-based marker of pseudocapsule integrity may thus assist surgeons in tailoring operative approaches. Embedding such radiomics tools into multidisciplinary team discussions could enhance the precision of surgical planning in PanNETs. However, there are limitations to be acknowledged. As a retrospective, single-center study, the model’s generalizability may be constrained by selection bias and institutional imaging protocols. Moreover, the current pipeline requires high-quality arterial phase CT images and manual ROI delineation, which could limit reproducibility across clinical environments. The absence of multi-modal imaging data and external validation further tempers the translational readiness of the model. Technically, the lack of automated segmentation and workflow standardization represents a practical barrier to real-world integration. Despite these caveats, the core concept—leveraging a structural histopathological trait to inform surgical strategy—is both innovative and clinically relevant. The authors have successfully repositioned the pseudocapsule from a descriptive anatomical feature to a prognostically and therapeutically meaningful marker. This reframing could inspire similar structural-functional investigations in other solid tumors, such as renal cell carcinoma, hepatic metastases, or meningiomas, where tumor margins hold parallel prognostic value. Based on the above content, we believe that future research could explore the following directions: prospective multicenter validation of the PS model across diverse platforms and imaging standards; integration of multi-modal data, such as PET/MRI or molecular biomarkers, to improve predictive accuracy; mechanistic dissection of the tumor-stroma-immune interplay underlying PS formation and disruption; clinical workflow automation to support real-time, point-of-care risk assessment; and long-term surveillance studies to evaluate the association of PS integrity with recurrence, metastasis, and treatment response. In conclusion, this study raises a clinically meaningful question, applies methodologically rigorous approaches, and presents results with translational implications. It revives interest in the tumor pseudocapsule as a potentially actionable anatomical marker, and sets a foundation for future investigations into personalized management of PanNETs and beyond.