Limited value of serum TPS/CY18 and CA125 for identifying hepatic and lymph-node metastases in pancreatic neuroendocrine tumors
Introduction: Circulating biomarkers that could complement histopathological and imaging assessment of metastatic pancreatic neuroendocrine neoplasms (PNENs) remain insufficiently characterized. This study evaluated serum tissue polypeptide-specific antigen/cytokeratin 18 (TPS/CY18) and cancer antigen 125 (CA125) in relation to hepatic metastases (HM) and regional or distant lymph-node metastases (LNM) in well-differentiated PNENs. Material and methods: This prospective, single-center study included 107 patients with well-differentiated PNENs: 60 with grade 1 neuroendocrine tumors (NET G1) and 47 with NET G2. The principal comparison was between patients with and without LNM and/or HM; skeletal and other distant metastases were not included in this definition. Biomarker concentrations were compared using the Mann–Whitney U test with Holm correction. Discriminatory performance was assessed using receiver operating characteristic curves, and factors associated with metastatic status were evaluated using multivariable logistic regression. Results: Lymph-node metastases and/or HM were present in 42 patients and absent in 65. Tissue polypeptide-specific antigen/cytokeratin 18 concentrations were numerically higher in patients with LNM/HM (median 79.96 vs. 51.64 U/L), but the difference was not significant (Holm-adjusted p = 0.140); the area under the curve (AUC) was 0.585 [95% confidence interval (CI) 0.469–0.700]. Median CA125 concentration was 6.95 vs. 5.16 U/mL (raw p = 0.042; Holm-adjusted p = 0.083), with an AUC of 0.617 (95% CI 0.502–0.726). After adjustment for age, sex, and grade, CA125 showed a weak association with LNM/HM [odds ratio (OR) per doubling 1.54; 95% CI 1.04–2.29; p = 0.032], whereas TPS/CY18 did not. Neuroendocrine tumor grade 2 was strongly associated with LNM/HM (OR 7.42; 95% CI 3.03–18.17; p < 0.001). Ki-67 and primary tumor size were independently associated with LNM and HM when analyzed separately, while neither biomarker independently contributed to the site-specific models. Conclusions: Cancer antigen 125 showed a weak and model-dependent association with LNM/HM but limited discriminatory performance. Tissue polypeptide-specific antigen/cytokeratin 18 provided no independent information. Tumor grade, Ki-67, and tumor size were more informative than the circulating biomarkers. These biomarkers should not replace histopathological or imaging assessment.
Authors
- Beata Kos‐Kudła (ORCID: https://orcid.org/0000-0003-3755-1722)
- Anna Zemczak (ORCID: https://orcid.org/0000-0001-6314-2661)
- Bogdan Marek (ORCID: https://orcid.org/0000-0001-5958-1338)
- Violetta Rosiek (ORCID: https://orcid.org/0000-0002-5238-9287)
- Magdalena Londzin-Olesik (ORCID: https://orcid.org/0000-0003-4178-6153)
Institutions
- Medical University of Silesia (PL)
Publication Details
- Journal
- Endokrynologia Polska
- Published
- 2026-11-17
- DOI
- https://doi.org/10.5603/ep.114445
- Primary Topic
- Neuroendocrine Tumor Research Advances
- Type
- article
- Field-Weighted Citation Impact
- 0.00