Highlights
– In the multicenter nonrandomized PUMP‑2 phase II trial (Franssen et al., J Clin Oncol 2025), HAIP floxuridine combined with systemic gemcitabine–cisplatin produced a 1‑year overall survival (OS) of 80.0% (95% CI, 69.6%–91.9%) among patients with unresectable intrahepatic cholangiocarcinoma (iCCA) confined to the liver, significantly higher than a historical 47% control (P < .001).
– Median OS was 22.3 months (95% CI, 19.7–35.9 months); 3‑year OS was 31.5% (95% CI, 20.4%–48.6%). Objective partial response rate was 44% and disease control at 6 months was 84%.
– Technical feasibility was high: 96% of patients who underwent pump placement started HAIP therapy; 10% converted to resection, including one complete pathologic response.
Background and Unmet Need
Intrahepatic cholangiocarcinoma (iCCA) is an aggressive primary liver malignancy whose incidence has increased in many regions. Prognosis for unresectable disease remains poor. Systemic gemcitabine–cisplatin (gem‑cis) has been the established first‑line standard following the ABC‑02 trial, but outcomes are limited: median overall survival from historic first‑line trials approximates about 11–12 months across biliary tract cancers and long‑term survival is uncommon for liver‑limited disease. For patients with disease confined to the liver, achieving sustained hepatic control can meaningfully extend survival and, in a minority, allow conversion to surgical resection. Locoregional strategies that increase hepatic drug exposure while limiting systemic toxicity provide a biologic rationale for hepatic arterial infusion pump (HAIP) chemotherapy. Floxuridine (FUDR), a hepatic‑extraction–favored fluoropyrimidine, attains high liver tissue concentration when delivered intraarterially via implantable pumps and has been used in liver‑dominant metastatic disease; evidence in iCCA has been limited to single‑center series until now.
Study Design
The PUMP‑2 trial is a nonrandomized, multicenter phase II study performed in the Netherlands (Franssen et al., J Clin Oncol 2025). Key features:
- Population: Patients with unresectable iCCA confined to the liver. Both treatment‑naïve patients and those previously treated with systemic therapy were eligible.
- Intervention: Up to six cycles of HAIP floxuridine delivered via an implanted pump plus up to eight cycles of concurrent systemic gemcitabine and cisplatin for patients who had not received gem‑cis previously. Pump placement and arterial catheterization were performed at participating centers experienced with HAIP therapy.
- Primary endpoint: 1‑year overall survival compared with a historical cohort treated with systemic therapy alone.
- Secondary endpoints: Objective response rate (ORR), disease control at 6 months, conversion to resection, median OS, and safety/feasibility metrics.
- Consideration of HAIP therapy in multidisciplinary tumor boards for patients with liver‑dominant unresectable iCCA who are candidates for pump implantation and have adequate performance status and liver function.
- Integration of molecular profiling into patient selection to identify those who may benefit from combined locoregional and targeted systemic approaches.
- Need for randomized trials comparing HAIP + systemic therapy versus modern systemic therapy alone (including targeted agents where appropriate) to establish causality and quantify absolute benefit and harms.
- Standardization of implantation technique, postoperative management, and toxicity reporting to facilitate broader dissemination if randomized data confirm benefit.
Fifty patients had pumps placed between January 2020 and September 2022; 48 (96%) received at least one dose of HAIP FUDR.
Key Results
Survival and response outcomes were notable for a liver‑confined unresectable population.
Primary and overall survival
– Median overall survival: 22.3 months (95% CI, 19.7–35.9 months).
– 1‑year OS: 80.0% (95% CI, 69.6%–91.9%), significantly higher than the historical control of 47% (P < .001).
– 3‑year OS: 31.5% (95% CI, 20.4%–48.6%). These results compare favorably to outcomes typically seen with systemic therapy alone in historical series of liver‑limited iCCA.
Tumor response and conversion to resection
– Partial responses (PR) were observed in 22 of 50 patients (44%). Disease control at 6 months—defined as PR or stable disease—was 84% (42/50).
– Conversion to resection occurred in 5 patients (10%), and 1 patient achieved a complete pathologic response at surgery. Conversion to resection is an important outcome because complete resection offers the only potential for cure in iCCA.
Feasibility and technical outcomes
– Pump placement and initiation of therapy were feasible: 50 pumps placed, and 48 patients started HAIP chemotherapy.
– Two patients (4%) did not start HAIP FUDR: one died from COVID‑19 before therapy initiation and one had a hepatic arterial dissection during the implantation procedure.
Safety
Detailed grade‑specific adverse event (AE) data are not specified in the abstract. Reported implantation‑related complications included one arterial dissection that prevented HAIP initiation. Known toxicities associated with HAIP include catheter‑ and pump‑related complications (thrombosis, dislodgement, infection, arterial injury), regional hepatic toxicity (chemical hepatitis, biliary toxicity), and systemic AEs related to concurrent gem‑cis. Interpretation of safety from this report should await the full manuscript for AE incidence, severity, and management strategies employed across centers.
Expert Commentary and Interpretation
The PUMP‑2 results are striking in the context of liver‑limited unresectable iCCA. A 1‑year OS of 80% and median OS >22 months represent clinically meaningful improvements over historical systemic therapy cohorts. Several points merit emphasis:
Biologic plausibility
Delivering high concentrations of floxuridine directly to the hepatic arterial bed maximizes tumor exposure within the liver while limiting systemic exposure, a strategy that aligns with the predominantly arterial blood supply of many hepatic tumors. When combined with systemic gem‑cis, HAIP may provide robust local control while systemic therapy treats micrometastatic disease outside the liver or circulating tumor cells.
Strengths
– Multicenter design increases generalizability compared with single‑center series and demonstrates that centers in a national network can deliver HAIP reliably.
– Inclusion of both treatment‑naïve and previously treated patients better reflects real‑world practice.
Limitations
– Nonrandomized design and comparison with historical controls invites potential selection bias. Patients referred for HAIP may have more favorable disease biology (true liver confinement, preserved performance status, limited tumor burden) compared with unselected historic cohorts.
– Small sample size and limited AE reporting in the abstract constrain risk–benefit assessment. Detailed toxicity data, quality of life, and health‑economic analyses are critical for adoption.
– Expertise requirement: HAIP implantation and management require multidisciplinary teams (interventional radiology, hepatobiliary surgery, oncology, specialized nursing). Results may not be reproducible in centers lacking procedural experience.
Context with contemporary systemic and targeted therapy
Systemic therapy for biliary tract cancer is evolving; molecular profiling identifies actionable alterations (FGFR2 fusions, IDH1 mutations) for which targeted agents have regulatory indications in selected patients. The PUMP‑2 approach targets the dominant hepatic tumor burden and could be complementary to targeted systemic agents in selected molecular subgroups, but safety and efficacy of combining HAIP with targeted therapies remain to be determined.
Clinical Implications and Future Directions
PUMP‑2 provides compelling signal that intensified locoregional therapy using HAIP floxuridine can materially improve outcomes in liver‑confined unresectable iCCA. Practical implications and next steps include:
Conclusion
The PUMP‑2 trial reports substantially improved 1‑ and 3‑year survival for patients with liver‑confined unresectable iCCA treated with HAIP floxuridine combined with systemic gemcitabine–cisplatin, relative to historical gem‑cis cohorts. These results are promising and support further evaluation in randomized settings and collaboration across experienced centers. Until randomized evidence is available, careful patient selection in specialized centers and multidisciplinary management remain essential.
Funding and Trial Registration
Funding details and clinical trial registry are not provided in the abstract summary. Readers should consult the full J Clin Oncol article for funding statements and trial registration identifiers (Franssen S et al., J Clin Oncol. 2025 Oct 13:JCO2500923.).
References
1. Franssen S, Rousian M, Filipe WF, et al. Hepatic Arterial Infusion Pump Chemotherapy in Patients With Unresectable Intrahepatic Cholangiocarcinoma—PUMP‑2 Trial. J Clin Oncol. 2025 Oct 13:JCO2500923. doi:10.1200/JCO-25-00923.
2. Valle J, Wasan H, Palmer DH, et al. Cisplatin plus gemcitabine versus gemcitabine for biliary tract cancer. N Engl J Med. 2010;362(14):1273–1281. (ABC‑02 trial establishing gem‑cis standard.)