Background: A growing number of centers offer hepatic artery infusion pump (HAIP) chemotherapy for advanced liver malignancies. While small series have demonstrated feasibility of robotic HAIP placement, comparison of outcomes with open placement is lacking. We compared outcomes after robotic versus open HAIP placement. Methods: We retrospectively reviewed HAIP placement without concurrent hepatectomy at Memorial Sloan Kettering Cancer Center from 1 January 2011 to 15 September 2022, and Duke Health from 1 November 2018 to 18 May 2023. Patients with prior liver surgery or who required catheterization of a non-standard vessel were excluded. Propensity score matching weights (PSMW) were calculated using age, sex, race, body mass index, American Society of Anesthesiologists class, neoadjuvant chemotherapy, colorectal procedure, and institution. Survey-weighted generalized linear models assessed the relationship between approach and outcomes. Results: Of 2002 consecutive HAIP placements, 819 (645 open/174 robotic) met the inclusion criteria. A higher proportion of open procedures involved combined colorectal procedures; other patient characteristics were similar. Overall, 15% of patients experienced an HAIP-specific complication and 12% required re-intervention; 2.7% had HAIP failure ≤ 90 days. After PSMW, the robotic approach had a longer operative time (β = 68 min, 95% confidence interval [CI] 55–81, p < 0.001) but shorter length of stay (β = − 1.8 days, 95% CI − 2.3 to 1.3, p < 0.001). The robotic approach was associated with increased HAIP-specific complications (odds ratio [OR] 1.72, p = 0.025) and re-intervention (OR 2.33, p < 0.001), with no difference in time to initiation of HAIP chemotherapy or HAIP failure. Conclusions: Robotic HAIP placement was associated with increased postoperative complications and significantly shorter length of stay, with similar time to initiation of HAIP therapy. There was no difference in the rate of early HAIP failure versus the open approach. These results suggest robotic HAIP placement is feasible and effective.

Objective: Describe the technical complications of hepatic artery infusion pump (HAIP) placement and characterize salvage rates. Summary Background Data: HAIP chemotherapy is an effective tool used in the treatment of hepatic malignancies and has become increasingly adopted at select centers. Much of the safety data is reported from small, newer programs or older, large series. Methods: Retrospective analysis of all patients who underwent HAIP placement at our institution between 2000 and 2022 was conducted. Pump salvage was defined as the ability to receive HAIP chemotherapy after a pump-associated complication. Pump failure was defined as an HAIP-related complication permanently stopping ongoing HAIP chemotherapy. Results: Over the study period, 2935 patients had an HAIP placed for colorectal liver metastases (n=2582, 88%), intrahepatic cholangiocarcinoma (n=294, 10%) or another malignancy (n=59, 2%). Of all patients, 759 (26%) had an HAIP-related complication, with 404/759 (53%) having their pump salvaged. The most common complications were abnormal perfusion (n=360 [12%], 285/360 [79%] salvaged), arterial pseudoaneurysm and/or hemorrhage (n=117 [4%], 7/117 [6%] salvaged), and pump-pocket infection (n=115 [4%], 60/115 [52%] salvaged). The complications with the highest salvage rates were pump-pocket seroma (n=15/16, 94%), pump-flipping (n=41/44, 93%), and pump-pocket hematoma (n=45/49, 92%). Unsalvaged complications were catheter dislodgement from vessel (n=0/47, 0%), catheter erosion into bowel (n=0/25, 0%), and arterial thrombosis (n=0/19, 0%). The cumulative incidence of pump failure at 1 year in patients receiving adjuvant therapy was 6.6%, and 4.5% among those receiving palliative therapy. Conclusions: HAIP-related complications occur in about one quarter of patients, predominantly perfusion and pump-pocket complications. These are salvageable in the majority of cases with multidisciplinary interventions.