The addition of immune checkpoint inhibitors to gemcitabine–cisplatin chemotherapy has established a new first-line standard for advanced biliary tract cancer (BTC). However, disease progression remains common, and subsequent treatment options increasingly depend on molecular profiling and access to targeted therapies.
A new real-world study comparing patients treated in the United States and South Korea evaluated outcomes with first-line gemcitabine–platinum plus immunotherapy and examined the association between molecularly matched targeted therapy and survival following disease progression.
The study was published in ESMO Gastrointestinal Oncology, in its December 2026 issue, under the title: “Real-world outcomes of first-line gemcitabine–platinum plus an immune checkpoint inhibitor and subsequent treatment with matched targeted therapy in biliary tract cancer.”
Authors: H. Yoon, H. Jeong, K. P. Kim, I. Park, T. J. Song, D. Oh, G. Huh, S. H. Cho, Y. Lee, M. Nuh, M. Hsiang, Q. Kimana, M. M. Javle, S. Lee, and C. Yoo.
Study Design and Patient Population
This retrospective cohort study included 375 patients with advanced BTC who received first-line gemcitabine–platinum plus an immune checkpoint inhibitor (GP-ICI) between September 2021 and October 2025. Patients were treated at two tertiary cancer centers: MD Anderson Cancer Center (MDACC) in the United States (n=121) and Asan Medical Center (AMC) in South Korea (n=254).
Eligible patients had locally advanced unresectable, metastatic, or recurrent biliary tract adenocarcinoma. Durvalumab was administered in 86.9% of patients, while 13.1% received pembrolizumab. The median age was 66 years. Intrahepatic cholangiocarcinoma (iCCA) accounted for 51.7% of cases, extrahepatic cholangiocarcinoma (eCCA) for 31.7%, and gallbladder cancer for 16.5%.
Tumor distribution differed considerably between institutions. iCCA was more frequent at MDACC (71.9% versus 42.1%), whereas eCCA was more common at AMC (41.7% versus 10.7%). The investigators assessed objective response rate (ORR), progression-free survival (PFS), overall survival (OS), postprogression survival (PPS), molecular profiling, and subsequent treatment patterns.
First-Line Treatment Outcomes
At a median follow-up of 17.2 months, first-line GP-ICI demonstrated generally comparable effectiveness between the two cohorts. In the overall population, median PFS was 7.0 months (95% CI 6.2–7.8), while median OS was 15.6 months (95% CI 13.6–17.7).
Median PFS was 6.6 months at MDACC versus 7.2 months at AMC (P=0.050). Median OS was 19.5 versus 14.9 months, respectively (P=0.154). Among 344 patients evaluable for radiologic response, the overall ORR was 17.6%, with no significant difference between institutions (21.5% versus 15.7%; P=0.172).
Multivariable analyses also showed no independent association between treatment center and PFS (HR 0.77; 95% CI 0.54–1.11; P=0.165) or OS (HR 1.19; 95% CI 0.74–1.91; P=0.477), comparing AMC with MDACC. These findings suggest broadly consistent first-line outcomes across the two centers despite differences in tumor-site distribution and patient characteristics.
Survival Following Disease Progression
More substantial differences emerged after progression on first-line GP-ICI. Among patients with disease progression, median PPS was significantly longer at MDACC than at AMC: 13.1 versus 5.8 months (P=0.002).
This difference was particularly evident in iCCA, where median PPS was 13.1 versus 6.5 months (P=0.002). Subsequent treatment patterns also differed. Among 236 patients with documented progression, 148 (62.7%) received second-line systemic therapy. The proportion receiving second-line treatment was higher at MDACC than at AMC (88.3% versus 50.3%).
Among the 164 patients who received second-line systemic therapy, targeted therapies were used in 41.3% at MDACC compared with 14.6% at AMC. A similar difference was observed in the third-line setting (42.9% versus 6.5%). Among patients with documented progression on first-line GP-ICI, those receiving second-line targeted therapy had the longest median PPS, at 15.4 months.
Molecular Profiling and Matched Targeted Therapy
Molecular profiling was performed in 236 patients (62.9%), with substantial differences in testing rates between institutions: 95.9% at MDACC versus 47.2% at AMC (P<0.001). Actionable genomic alterations were identified in 72 of 236 profiled patients (30.5%). Their overall prevalence was comparable between cohorts, but the molecular profiles differed.
IDH1 mutations were more frequent at MDACC (14.7% versus 3.3%), as were FGFR2 fusions or rearrangements (8.8% versus 2.5%). Conversely, HER2 amplification was more common at AMC (14.2% versus 1.7%), along with HER2 mutations (6.7% versus 0%).
Among patients with actionable alterations, 30 of 72 (41.7%) received matched targeted therapy. Treatment utilization was substantially higher at MDACC than at AMC: 75.8% versus 12.8% (P<0.001). Importantly, the presence of an actionable alteration alone was not associated with PPS or OS. However, receipt of matched targeted therapy was significantly associated with longer PPS compared with patients who had actionable alterations but did not receive matched treatment (HR 0.15; 95% CI 0.06–0.41; P<0.001).
To address potential immortal time bias, investigators conducted a time-dependent Cox analysis accounting for when matched treatment began. The association with longer PPS remained significant (HR 0.33; 95% CI 0.12–0.87; P=0.025). A nonsignificant trend toward longer OS was also observed among patients receiving matched targeted therapy (HR 0.54; 95% CI 0.26–1.10; P=0.089). These results suggest that access to matched targeted therapy may contribute to improved outcomes after first-line treatment, although causality cannot be established from this retrospective analysis.
Clinical Implications and Conclusion
This study highlights an important distinction in advanced BTC management: while first-line immunochemotherapy achieved comparable outcomes across the US and South Korean cohorts, survival after progression differed substantially.
The investigators suggested that disparities in molecular testing, treatment availability, and reimbursement policies may partly explain these differences. However, the retrospective design, differences in patient populations, molecular profiling methods, and access to subsequent therapies introduce potential selection bias and confounding. The findings therefore require cautious interpretation.
The central finding is that receipt of molecularly matched targeted therapy was associated with significantly longer postprogression survival in advanced BTC, even after accounting for the timing of treatment initiation. Although the study cannot demonstrate a causal survival benefit, it reinforces the importance of timely molecular profiling and improved access to biomarker-directed therapies for patients whose disease progresses after first-line GP-ICI.
The full article is available in ESMO Gastrointestinal Oncology.
Find more videos on OncoDaily YouTube TV.
