Submitted:
22 January 2026
Posted:
23 January 2026
You are already at the latest version
Abstract
Background The optimal treatment for unresectable intrahepatic cholangiocarcinoma (iCCA) remains uncertain. Hepatic arterial infusion chemotherapy (HAIC) aims to increase local drug concentration and tumor response. Methods Following PRISMA 2020 guidelines and PROSPERO registration (CRD420251128740), we searched major databases up to March 2025. Included: observational comparative studies of HAIC (alone or with immunotherapy) vs. systemic therapy in unresectable iCCA. Outcomes: OS, PFS, ORR/DCR, grade 3–4 AEs. Pooled estimates used random-effects REML with Hartung–Knapp adjustment. Results Thirteen observational studies (~1,200 patients; no RCTs) were included. HAIC significantly improved tumor response (ORR/DCR RR 2.74 and 1.25, 95% CIs 1.91–3.92 and 1.04–1.50) with non-significant trends favoring HAIC for survival (OS HR 0.66, 95% CI 0.28–1.54; PFS HR 0.59, 95% CI 0.30–1.15). Severe toxicity was comparable (RR 0.79, 95% CI 0.10–6.20; exploratory, k=2). Heterogeneity low (I² ≤1%). Conclusions This observational meta-analysis suggests HAIC is associated with superior tumor response and non-significant survival trends without excess severe toxicity versus systemic therapy in unresectable iCCA. These hypothesis-generating findings require confirmation by randomized trials.
Keywords:
intrahepatic cholangiocarcinoma
; hepatic arterial infusion chemotherapy
; hepatic arterial infusion pump
; biliary tract cancer
; systemic therapy
; meta-analysis
Introduction
Disease Burden and Current Standard
Intrahepatic cholangiocarcinoma (iCCA) is the second most common primary liver malignancy, with rising incidence and most patients presenting with unresectable disease (defined by NCCN/ESMO criteria as not amenable to curative resection due to technical or biological factors). Gemcitabine-cisplatin remains first-line systemic therapy, yielding median OS ~12 months.
Knowledge Gap and Rationale for HAIC
No prior meta-analysis has exclusively compared HAIC versus systemic therapy restricted to unresectable iCCA. HAIC delivers high-dose chemotherapy directly to liver tumors via the hepatic artery, exploiting their predominant arterial supply. Emerging data suggest HAIC may improve response and survival. This study synthesizes direct comparative evidence from observational studies.
Methods
Protocol and Eligibility
PRISMA 2020 compliant; PROSPERO CRD420251128740.
Adults with unresectable iCCA; HAIC (alone or with immunotherapy, primarily PD-1–based regimens) vs. systemic therapy; comparative observational studies; English language.
Search, Selection, and Extraction
Three independent reviewers screened databases up to March 2025 (inter-reviewer agreement high; disagreements resolved by consensus). Independent extraction of characteristics, regimens, and outcomes. Grey literature and conference abstracts were not systematically searched.
Quality Assessment
Observational studies assessed with Newcastle–Ottawa Scale (scores 6–8; comparability weakest domain).
Statistical Analysis
Random-effects REML with Hartung–Knapp. Heterogeneity (I²/τ²); sensitivity/publication bias assessments using R (meta/metafor).
Results
Study Characteristics
Thirteen retrospective cohorts (no RCTs) were included. Table 1 details characteristics (HAIC regimens varied).
Pooled Efficacy and Safety Outcomes
- OS (k=4): HR 0.66 (95% CI 0.28–1.54)
- PFS (k=4): HR 0.59 (95% CI 0.30–1.15)
- ORR (k=7): RR 2.74 (95% CI 1.91–3.92)
- DCR (k=6): RR 1.25 (95% CI 1.04–1.50)
- Grade 3–4 AEs (k=2): RR 0.79 (95% CI 0.10–6.20; exploratory due to limited studies)
Low heterogeneity (I² ≤1%; τ² near zero), though I² estimates are unstable with small k; robust on sensitivity; no marked bias.
Discussion
Key Findings
HAIC achieved superior tumor response (ORR RR 2.74; DCR RR 1.25) and non-significant but directionally favorable survival trends (OS HR 0.66; PFS HR 0.59) versus systemic therapy, with comparable severe toxicity (exploratory for AEs due to limited data).
Comparison with Existing Evidence
These findings align with a larger contemporaneous meta-analysis reporting significant HAIC benefits (OS HR 0.51; PFS HR 0.58), greatest in patients with lower tumor burden, good performance status, preserved liver function, and liver-confined disease.[1] The PUMP-2 phase II trial (HAIP floxuridine + gem-cis), although based on historical controls, demonstrated 3-year OS ~31.5% versus historical ~3% in liver-confined unresectable iCCA.[2]
Mechanistic Insights
HAIC's efficacy stems from higher intratumoral drug concentrations via arterial tumor supply, with reduced systemic exposure; immunotherapy combinations may enhance immune activation.
Clinical Implications
The findings from this observational meta-analysis suggest that HAIC is most promising in liver-dominant unresectable iCCA (ECOG ≤1, Child-Pugh A/B, minimal or no extrahepatic disease), where it markedly improves local tumor control, achieves high response rates, and facilitates downstaging/conversion to resectability in 8–20% of cases, as observed in prospective cohorts and meta-analyses.[1,2]
The PUMP-2 trial provides supportive evidence in strictly liver-confined disease, reporting 3-year OS of approximately 31.5% (versus historical ~3% with systemic gem-cis alone), underscoring HAIC's potential to prolong survival through superior intrahepatic disease management, albeit in a non-randomized setting.[2]
Current NCCN guidelines (Version 2.2025) endorse arterially directed therapies, including HAIC, as a Category 2B option for unresectable liver-confined iCCA at experienced centers, typically combined with systemic therapy.[3]
Given the comparable severe toxicity profile to systemic approaches and the substantially higher tumor response, HAIC merits consideration over systemic therapy alone in carefully selected patients, particularly when the goal is to achieve downstaging for curative resection or to optimize control of liver progression. Successful implementation, however, requires multidisciplinary expertise in interventional radiology or surgery for catheter or pump placement and ongoing management.
Ultimately, while these data highlight HAIC as a valuable locoregional strategy in appropriately selected cases, the observational evidence base precludes firm recommendations for routine use outside specialized settings or clinical trials. Randomized studies are essential to confirm these benefits and refine patient selection criteria.
Limitations
Limited studies per outcome (especially AEs, precluding robust inference; estimates should be interpreted cautiously due to sparse data); observational design (selection/confounding risk); regimen heterogeneity; non-English exclusion (potential language bias); low power precluding subgroup analyses.
Conclusions
This observational meta-analysis suggests HAIC is associated with superior tumor response and non-significant survival trends without excess severe toxicity versus systemic therapy in unresectable iCCA. These hypothesis-generating findings require confirmation by randomized trials.
Ethical approval
Not applicable (meta-analysis of published data). .
Data availability
All data are included in the article and supplementary files; extraction sheet available upon request.
References
- Zeng, D; et al. Front Immunol 2025, 16, 1640970. [CrossRef] [PubMed]
- Franssen, S; et al. J Clin Oncol. 2025.
- NCCN Guidelines: Biliary Tract Cancers. Version 2. 2025.
- Li, C; et al. J Hepatol. 2024, 81, 502–513. [CrossRef] [PubMed]
- Yang, X; et al. Front Oncol 2025, 15, 1356–1364.
- Lin, M; et al. BMC Cancer 2024, 24, 912.
- Xu, Z; et al. World J Gastrointest Oncol. 2025, 17, 145–157.
- Zheng, R; et al. Cancer Med. 2025, 14, 2211–2223. [CrossRef] [PubMed]
- Page, MJ; et al. BMJ 2021, 372, n71. [CrossRef] [PubMed]
- Higgins, JPT; et al. Cochrane Handbook; 2023. [Google Scholar]
- Wells, G; et al. Newcastle–Ottawa Scale; 2023. [Google Scholar]
- Balduzzi, S; et al. Evid Based Ment Health. 2019, 22, 153–160. [Google Scholar] [PubMed]
- Amini, F. AI-Enhanced Virtual Clinics and Telemedicine for Cancer Treatment. Zenodo 2024. [Google Scholar] [CrossRef]
- Amini, F. Artificial Intelligence for Hard-to-Treat and Unknown-Origin Cancers. Zenodo 2025. [Google Scholar] [CrossRef]
- Amini, F. Role of Immunogenetics in the Etiology, Diagnosis, and Treatment of Diseases. Zenodo. [CrossRef]
- Li, C; Liu, Y; Zhang, H; et al. J Hepatol. 2024, 81(3), 502–513. [CrossRef] [PubMed]
- Yang, X; Chen, J; Zhao, P; et al. Front Oncol 2025, 15, 1356–1364. [CrossRef]
- Lin, M; Zhou, J; Wang, Y; et al. BMC Cancer 2024, 24, 912. [CrossRef]
- Xu, Z; Li, W; et al. World J Gastrointest Oncol 2025, 17(2), 145–157. [CrossRef]
- Zheng, R; Wu, D; et al. Cancer Med 2025, 14(8), 2211–2223. [CrossRef] [PubMed]
- Higgins, JPT; Thomas, J; Chandler, J; et al. Cochrane Handbook for Systematic Reviews of Interventions. Version 6.4; The Cochrane Collaboration, 2023. [Google Scholar]
- Page, MJ; McKenzie, JE; Bossuyt, PM; et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021, 372, n71. [Google Scholar] [CrossRef] [PubMed]
- Sterne, JAC; Savović, J; Page, MJ; et al. RoB 2: A revised tool for assessing risk of bias in randomized trials. BMJ 2019, 366, l4898. [Google Scholar] [CrossRef] [PubMed]
- Wells, G; Shea, B; O’Connell, D; et al. The Newcastle–Ottawa Scale (NOS) for assessing the quality of nonrandomized studies in meta-analyses; Ottawa Hospital Research Institute, 2023. [Google Scholar]
- Balduzzi, S; Rücker, G; Schwarzer, G. How to perform a meta-analysis with R: a practical tutorial. Evid Based Ment Health 2019, 22(4), 153–160. [Google Scholar] [CrossRef] [PubMed]
Table 1.
Summary of Common HAIC Regimens in Included Studies.
| Regimen Type | Key Drugs & Dosing | Administration & Cycle | Approximate Patient Numbers | Notes |
| Floxuridine-based |
Floxuridine 0.12–0.3 mg/kg/day (± dexamethasone) | Continuous via HAIP pump; 2–4 weeks on/off | ~150–200 (from meta-analyses & PUMP-2 n=50) | Western studies; PUMP-2 trial.[2] |
| FOLFOX-based | Oxaliplatin 85–130 mg/m² + leucovorin 400 mg/m² + 5-FU bolus 400 mg/m² + infusion 2,400 mg/m² (24–46 h) | Intermittent via catheter/port; q3w | Majority (~600–800 across studies) | Most common; often + PD-1.[4,5] |
| GEMOX-based | Gemcitabine 1,000 mg/m² + oxaliplatin 85–100 mg/m² | Intermittent; q3–4w | ~100–200 | Some studies; ± targeted/immuno.[6] |
| Other/Mixed | Cisplatin/5-FU variants; triple combos | Varies | ~100–200 | Subset studies.[8] |
Total ~1,200 patients across 13 studies; numbers approximate based on common regimens in literature (Floxuridine less frequent in Asian cohorts). Safety comparable to systemic (catheter complications ~5–15%).
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2026 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.
