Photon-counting computed tomography in pancreatic ductal adenocarcinoma: correlation of iodine concentration with histopathological and clinical parameters

Objectives Photon-counting computed tomography (PCCT) enables quantitative assessment of tissue iodine distribution. This exploratory study investigated associations between PCCT-derived iodine parameters and clinical, serologic, and histopathologic characteristics of pancreatic ductal adenocarcinoma (PDAC). Materials and methods This retrospective single-center study included 27 patients with histologically confirmed PDAC who underwent pretreatment contrast-enhanced PCCT. Iodine concentration was measured independently by two blinded readers in arterial and portal-venous phases using a whole-tumor ROI on the largest axial section and a solid-component ROI. Absolute iodine concentration, tumor-to-aorta normalized iodine ratios, and phase-to-phase Δ iodine were evaluated. CA 19 − 9 was analyzed continuously using Spearman rank correlation; cutoff-based and clinicopathologic subgroup comparisons were considered secondary exploratory analyses and were re-analyzed using two-sided Wilcoxon rank-sum tests. All p-values are nominal. Results CA 19 − 9 was available in 24 patients (median 123.5 U/mL; IQR 55.3–1962). Higher CA 19 − 9 was associated with lower portal-venous absolute iodine concentration in the solid tumor component (Spearman ρ=-0.45, 95% CI -0.77 to -0.04; nominal p = 0.028) and with the arterial whole-tumor normalized iodine ratio (ρ=-0.58, 95% CI -0.84 to -0.14; nominal p = 0.010). At the 100 U/mL cutoff, portal-venous solid-component absolute iodine concentration was higher below the cutoff (1.77 ± 0.57 vs. 1.19 ± 0.50 mg/mL; nominal p = 0.017). After two-sided re-analysis, previously reported differences by T category, Ki-67, and tumor-cell count did not reach nominal significance. Interreader reliability was excellent across iodine parameters (ICC range 0.960–0.989). Conclusion In this small exploratory cohort, selected absolute and normalized iodine measurements showed nominal inverse associations with CA 19 − 9, whereas associations with tumor stage and histopathologic parameters were not consistent after conservative two-sided re-analysis. PCCT iodine measurements were highly reproducible between readers. These findings are hypothesis-generating and require validation in larger independent cohorts before any incremental clinical value can be inferred.