Key points are not available for this paper at this time.
Background and purpose Accurate preoperative assessment of peritoneal metastasis burden in ovarian cancer remains challenging as conventional CT lacks sensitivity for small-volume disease. Spectral CT iodine quantification provides objective metrics of tumor vascularity that may predict surgical resectability. This retrospective study evaluated iodine-based parameters for estimating surgical Peritoneal Cancer Index (PCI) and predicting complete cytoreduction (R0). Materials and methods We retrospectively identified 145 ovarian cancer patients who underwent preoperative triphasic spectral CT scans and subsequent cytoreductive surgery between June 2022 and December 2023. Two blinded radiologists quantified iodine concentration (IC), normalized iodine concentration (NIC), spectral curve slope (λHU), and effective atomic number (Zeff). Patients were stratified by surgical PCI into low (≤6), intermediate (7-15), and high (≥16) burden subgroups. Statistical analyses included correlation studies, ROC curves, and multivariable logistic regression. Results Iodine-based parameters demonstrated a strong correlation with intraoperative PCI (IC: r=0.85, NIC: r=0.74, both P0.001). NIC increased progressively across tumor burden categories (0.43, 0.57, 0.71; P0.001). For predicting R0 resection, NIC achieved optimal performance (AUC = 0.88, 95% CI: 0.81-0.94). The cutoff value ≤0.55 yielded sensitivity 84.5%, specificity 79.5%, and a negative predictive value of 94.8% for identifying patients who would achieve R0 resection (i.e., among patients with NIC ≤0.55, 94.8% achieved R0). A combined model (NIC + PCI + CA-125) achieved AUC = 0.93 (95% CI: 0.87-0.97) with excellent interobserver reproducibility (ICC 0.84). Conclusion Spectral CT iodine quantification demonstrates potential as a complementary tool for estimating peritoneal metastasis burden and predicting surgical resectability. The NIC threshold of 0.55 shows promise for stratifying patients, though external validation is needed before integration into preoperative staging protocols.
Yongfeng et al. (Thu,) studied this question.