BP20. Perioperative and volume-based outcomes for hepatic arterial infusion (HAI) pump chemotherapy: a multicenter analysis from the international HAI Consortium Research Network

Friday, March 6, 2026

1:51 PM - 1:58 PM MST

Location: 301A

Co-Author(s)

Ellery H. Reason, BS

Medical Student
Duke University School of Medicine
Durham, North Carolina, United States
Anthony T. Saxton, MD, MSc

Resident Physician
Duke University
Durham, North Carolina, United States
JP

James F. Pingpank, Jr., MD, FACS

Associate Professor of Surgery
University of Pittsburgh Medical Center
Pittsburgh, Pennsylvania, United States
Ajay V. Maker, MD, FACS

Professor of Surgery and Chief, Division of Surgical Oncology
University of California San Francisco
San Francisco, California, United States
PK

Paul J. Karanicolas, MD, PhD, FRCS(C), FACS

Professor of Surgery
University of Toronto
Toronto, Ontario, United States
AY

Adam C. Yopp, MD, FACS

Professor of Surgery and Chief, Division of Surgical Oncology
University of Texas Southwestern
Dallas, Texas, United States
Skye C. Mayo, MD, MPH, FACS (he/him/his)

Professor of Surgery
Oregon Health and Science University
Portland, Oregon, United States
GW

G. Paul Wright, MD, FACS, FSSO

Assistant Professor of Surgery
Michigan State University/Corewell Health West Cancer Center
Grand Rapids, Michigan, United States
Jashodeep Datta, MD, FACS (he/him/his)

Associate Professor of Surgery
University of Miami
Miami, Florida, United States
Michael J. Cavnar, MD, FACS (he/him/his)

Associate Professor of Surgery
University of Kentucky Markey Cancer Center
Lexington, Kentucky, United States
RM

Ryan P. Merkow, MD, MS, FACS

Associate Professor of Surgery
University of Chicago
Chicago, Illinois, United States
Michael I. D’Angelica, MD

Enid A. Haupt Endowed Chair in Surgery, Professor
Hepatopancreatobiliary Service, Department of Surgery, Memorial Sloan Kettering Cancer Center, United States
Michael E. Lidsky, MD, FACS

Associate Professor of Surgery
Department of Surgery, Duke University Health System
Durham, North Carolina, United States

Primary Presenter(s)

Brittany G. Sullivan, MD

Fellow Physician
Duke University
Durham, North Carolina, United States

Introduction: HAI chemotherapy is increasingly used for colorectal liver metastases (CRLM) and intrahepatic cholangiocarcinoma (ICC). Outcomes from high-volume centers have been reported, but real-world data are limited. The HAI Consortium Research Network (HCRN) established an international registry to enable multicenter analysis. This study evaluated perioperative and volume-based outcomes in patients treated with HAI.

Methods: The HCRN registry (2000–2024) was retrospectively reviewed. Demographics, clinical features, and perioperative outcomes were analyzed. Change-point analysis using polynomial regression identified thresholds in annual center volume associated with HAI-specific complication risk. Chi-square tests compared unadjusted complication rates across center- and surgeon-volume groups, and multivariable logistic regression estimated adjusted odds of postoperative complications.

Results: 1,480 patients underwent pump placement at 36 centers (median 6 cases/yr [1–43]). 1,375 (93%) had CRLM (957 (70%) unresectable; 418 (30%) adjuvant) and 105 (7%) ICC. Median time from diagnosis to HAI pump placement was 9 months; 94% had prior chemotherapy. Open implantation occurred in 81%; the gastroduodenal artery was cannulated in 99% and 28% required variant arterial ligation. Concomitant procedures included colorectal resection (36%), minor hepatectomy (34%), major hepatectomy (12%), and ablation (25%). Median length of stay was 5 days (IQR 4-7).
HAI-specific complications occurred in 460 patients (33%): 17% pocket (5% infection), 3% arterial (1% pseudoaneurysm), and 3% perfusion complications. Floxuridine (FUDR) was administered in 97% of patients. Biliary sclerosis occurred in 7%. 90-day non-HAI morbidity was 9%, reintervention 10% and mortality 2%.

Complication risk declined with increasing volume. A threshold of 8 cases/yr defined low (< 8) vs high (≥8) volume centers (Figure). High-volume centers had 32% lower odds of complications (OR 0.68, 95% CI 0.50–0.92, p=0.01); high-volume surgeons showed similar benefit (OR 0.68, 95% CI 0.53–0.87, p=0.003).

Conclusions: In the largest multicenter analysis of perioperative HAI outcomes, pumps were mainly performed via laparotomy for unresectable CRLM. The risk of complications was decreased when centers and surgeons performed ≥8 cases annually. As use of HAI expands, balancing accessibility with regionalized expertise is essential.

Learning Objectives:

Define the annual surgeon/center volume threshold for HAI pump placement associated with fewer HAI-specific complications.
Differentiate and select appropriate indications and timing for HAI pump placement.