Percutaneous isolated hepatic perfusion (chemosaturation) with melphalan following right hemihepatectomy in patients with cholangiocarcinoma and metastatic uveal melanoma: peri- and post-interventional adverse events and therapy response compared to a matched group without prior liver surgery.

To evaluate feasibility, frequency and severity of peri-procedural complications and post-procedural adverse events (AEs) in patients with advanced cholangiocarcinoma or liver metastasis of uveal melanoma and prior hemihepatectomy undergoing chemosaturation percutaneous hepatic perfusion (CS-PHP) and to analyze therapy response and overall survival compared to a matched group without prior surgery. CS-PHP performed between 10/2014 and 02/2018 were retrospectively assessed. To determine peri-procedural safety and post-procedural adverse events, hospital records and hematological, hepatic and biliary function were categorized using Common Terminology Criteria for Adverse Events (CTCAE) v5.0 (1-5; mild-death). Significance was tested using Wilcoxon signed-rank and Mann-Whitney U test. Kaplan-Meier estimation and log-rank test assessed survival. Overall 21 CS-PHP in seven patients (4/7 males; 52 ± 10 years) with hemihepatectomy (grouphemihep) and 22 CS-PHP in seven patients (3/7 males; 63 ± 12 years) without prior surgery (groupnoresection) were included. No complications occurred during the CS-PHP procedures. Transient changes (CTCAE grade 1-2) of liver enzymes and blood cells followed all procedures. In comparison, grouphemihep presented slightly more AEs grade 3-4 (e.g. thrombocytopenia in 57% (12/21) vs. 41% (9/22; p = 0.37)) 5-7 days after CS-PHP. These AEs were self-limiting or responsive to treatment (insignificant difference of pre-interventional to 21-45 days post-interventional values (p > 0.05)). One patient in grouphemihep with high tumor burden died eight days following CS-PHP. No deaths occurred in groupnoresection. In comparison, overall survival after first diagnosis was insignificantly shorter in groupnoresection (44.7(32-56.1) months) than in grouphemihep (48.3(34.6-72.8) months; p = 0.48). The severity of adverse events following CS-PHP in patients after hemihepatectomy was comparable to a matched group without prior liver surgery. Thus, the performance of CS-PHP is not substantially compromised by a prior hemihepatectomy.