Real-world outcomes of hepatic arterial infusion pump therapy.

Abstract

69 Background: Hepatic artery infusion (HAI) with floxuridine (FUDR) is a liver directed therapy for (pts) with unresectable colorectal liver metastases (CRLM) to facilitate downstaging to resection, as well as for resected CRLM and unresectable intrahepatic cholangiocarcinoma (ICC). HAI has expanded beyond specialized academic centers to community programs with appropriate expertise. We describe pt selection, outcomes and feasibility in our large community-based cancer center. Methods: We conducted a retrospective analysis of a prospectively maintained database of 40 consecutive pts selected by our multidisciplinary board for HAI. We evaluated pt demographics, perioperative and oncologic outcomes. Imaging was re-reviewed for objective hepatic response assessment using RECIST 1.1. Overall survival (OS) and progression-free survival (PFS) were calculated by Kaplan-Meier method. Results: From 03/2022 - 06/2025, 40 pts received HAI: 27 for unresectable CRLM, 9 adjuvant for resected CRLM, 4 for unresectable ICC. Median age was 54.5 years (35-81), 70% were males, 62% White. All CRC were MSI-S. KRas was mutated in 35% pts and 2 pts had BRAF mutation. 24/40 pts had mets to the liver and regional LN, 10/40 pts to the liver only and 6/40 pts had low volume extrahepatic mets at baseline. 97% had bilobar liver mets and 90% had >5 liver lesions. All pts received prior systemic therapy, 40% heavily treated with > 2 lines. Concurrently with pump placement, 12% had hepatic resection, 52% had primary tumor resection (17/40 colon, 4/40 rectum), 90% had cholecystectomy and portal LN resection. Median length of stay was 5 days. Common complications: 20% seroma/hematoma, 10% pocket infection, 10% pump thrombosis and 10% malfunction. FUDR was started at median 15 days after surgery. After a median follow-up of 345 days, pts received a median of 8.5 (3-22) cycles. 70% pts received concurrent systemic therapy. 45% pts developed transaminitis requiring dose reduction in 20% and temporary hold in 35%. 4 pts had HAI pump removed due to progression, at median 536 days after placement. Hepatic disease control was achieved in 65% at 3 months (14 partial response PR, 12 stable disease SD, 13 progressed PD), 58% at 6 months (8 PR, 9 SD, 12 PD), 30% at 9 months (2 PR, 3 SD, 11 PD), and 30% at 12 months (1 PR, 2 SD, 7 PD). Progression was extrahepatic in 6 patients, hepatic in 14 pts, and both hepatic and extrahepatic in 11 pts. 1 pt had liver transplant and another pt is being evaluated for transplant. Median OS was 9.4 months (95% CI 2.33–12.65). Overall, hepatic only and extrahepatic only PFS were 5.26 months (95% CI 2.63–7.66), 4.17 months (95% CI 2.33–8.11), and 4.17 months (95% CI 2.23–7.23), respectively. Conclusions: Our safety and survival data reflect real world outcomes of HAI pump therapy when integrated within a large community hospital. The HAI Consortium plays a pivotal role in standardizing protocols and guiding best practices for safe and effective program implementation.