Objective. Iodine uptake in tumor lesions can be quantitatively calculated in patients scanned with dual energy computed tomography (DECT). This study investigated the possibility of using DECT obtained iodine concentration as a prognostic factor for percutaneous hepatic perfusion (PHP) treatment response in patients with uveal melanoma metastases.

Methods. Thirty-eight patients with unresectable hepatic uveal melanoma metastases who underwent DECT before and after PHP were retrospectively enrolled. Regions of interest (ROI) were drawn for the healthy liver parenchyma and target lesions (TL). The average iodine concentration of each ROI was calculated. The TL concentration was normalized using the healthy liver concentration. The progression free survival (PFS), overall survival (OS) and hepatic progression free survival (hPFS) were determined for each patient. Univariate survival analyses using Cox proportional hazards model were performed to find significant prognostic factors. The analysis was performed for the following variable at baseline: normalized iodine concentration (BL NIC), diameter, lactate dehydrogenase, alkaline phosphatase, tumor load (≤5 or >5 hepatic lesions), sex, age, and co-morbidities (yes or no). Next, the absolute NIC difference (AD NIC), relative NIC difference (RD NIC), absolute diameter difference (AD diameter), and relative diameter difference (RD diameter) between baseline and follow-up were evaluated. Multivariate survival analyses were performed individually for BL NIC, AD NIC and RD NIC with sex, age, and tumor-load. Kaplan-Meier analyses were performed to visualize the results of the survival analyses. The Pearson and Spearman correlations were examined between the change in diameter and change in NIC.

Results. BL NIC was not a significant prognostic factor. Only the baseline diameter was significant for OS (p = 0,008). RD NIC was significant for PFS (p = 0,014/0,007) and hPFS (p = 0,004/0,002) in both the univariate and multivariate analysis. AD NIC was significant for PFS (p = 0,008) in the multivariate analysis. The multivariate hazard ratio (HR) for AD NIC was 1,119 (confident interval (CI): 1,048-4,528) for PFS. RD NIC had a HR of 1,064 (CI:1,017-1,112) for PFS and a HR of 1,084 (CI: 1,031-1,140) for hPFS. The results of the Kaplan-Meier analyses were on par with the survival analyses. No correlation was found between the change in diameter and change in NIC.

Conclusion. This study showed that RD NIC between baseline and the first follow-up after PHP treatment is a prognostic factor for PFS and hPFS, and that AD NIC is a prognostic factor for PFS in patients with hepatic uveal melanoma metastases. Patients who showed an increase in normalized iodine uptake between the baseline and follow-up scans had lower PFS and hPFS, suggesting that the lesion's vitality increased regardless of the PHP treatment received. Identifying these differences at the first follow-up after treatment can provide insight into the likelihood of a patient developing progression and can aid in further decision making.