I now have enough information to give a thorough, evidence-based answer. Here it is:
Gemcitabine + Lenvatinib vs. Gemcitabine + Cisplatin in Biliary Tract Cancer
These two regimens are used in different contexts and cannot be directly compared head-to-head in a simple "better/worse" way - their evidence bases and approved roles are different. Here is a breakdown:
1. Gemcitabine + Cisplatin (GemCis) - The Established Standard
Evidence base: Phase III, FDA/EMA approved
The ABC-02 trial (Valle et al., 2010) established GemCis as the standard of care for advanced/unresectable biliary tract cancer (BTC), including cholangiocarcinoma and gallbladder cancer. It demonstrated:
- Median OS: ~11.7 months (vs. 8.1 months with gemcitabine alone)
- Median PFS: ~8 months
- Main toxicity: hematologic (more in combination arm)
This result has been replicated globally. Both TOPAZ-1 and KEYNOTE-966 phase III trials then built upon GemCis as the backbone, adding immunotherapy:
| Trial | Addition to GemCis | OS HR | Status |
|---|
| TOPAZ-1 | Durvalumab | 0.80 | FDA approved 2022 |
| KEYNOTE-966 | Pembrolizumab | 0.83 | FDA approved 2023 |
The 3-year follow-up of TOPAZ-1 (July 2025, Journal of Hepatology) confirmed >2x more survivors at 3 years with durvalumab + GemCis vs. GemCis alone. GemCis (ideally with an immune checkpoint inhibitor) is now the global standard first-line regimen per NCCN 2025 and major international guidelines.
2. Gemcitabine + Lenvatinib - Investigational, Not a Standard
Evidence base: Phase II, single-center/small trials, mostly in intrahepatic cholangiocarcinoma (ICC)
Lenvatinib is a multi-target tyrosine kinase inhibitor (VEGFR1-3, FGFR1-4, PDGFRa, RET, KIT). It has been studied primarily in combination with chemotherapy AND immunotherapy (not gemcitabine alone) in BTC:
- GEMOX + lenvatinib + toripalimab (anti-PD-1) - Phase II single-arm (PMID 36928584): ORR 80%, median PFS 10.2 months, median OS 22.5 months in advanced ICC (n=30, single-center, China)
- GEMOX + lenvatinib vs. lenvatinib + toripalimab - Phase II, two-arm (PMID 41121202, BMC Med 2025): GEMOX+lenvatinib arm showed ORR 40%, median PFS 8.0 months, median OS 15.5 months (n=30)
- Note: These studies almost exclusively used GEMOX (gemcitabine + oxaliplatin), not gemcitabine + cisplatin. The "gemcitabine + lenvatinib" combination alone (without a PD-1 inhibitor) has very limited data.
There is no approved indication for lenvatinib in BTC as of 2026. No phase III trial has validated lenvatinib-based combinations as superior to the GemCis backbone.
Head-to-Head Comparison: Bottom Line
| Parameter | Gem + Cisplatin | Gem + Lenvatinib |
|---|
| Evidence level | Phase III RCT | Phase II only |
| Approval status | FDA/EMA approved | Not approved |
| Tumor type | All BTC (CCA + GBC) | Mainly ICC (intrahepatic) |
| Median OS | ~11.7 mo (GemCis alone) up to ~15-16 mo (with ICI) | ~15.5 mo (GEMOX+lenv, small trial) |
| Toxicity | Myelosuppression, nephrotoxicity | Myelosuppression + hypertension, proteinuria, hepatotoxicity |
| Guideline status | Standard of care (NCCN, ESMO) | Investigational |
Clinical Guidance
- For unresectable/advanced BTC in any patient eligible for cisplatin, the standard of care is GemCis + durvalumab (or pembrolizumab), based on phase III data and regulatory approval.
- Lenvatinib-based regimens (especially the GOLP regimen: GEMOX + lenvatinib + PD-1 inhibitor) are emerging in intrahepatic cholangiocarcinoma with promising phase II results, but are not yet validated in phase III for first-line use. A phase III trial (NMPA No. 2021LP01825) is ongoing.
- A notable 2026 NEJM paper (PMID 41780001) studied the neoadjuvant GOLP regimen in resectable high-risk ICC - this is a separate neoadjuvant context.
- In patients with FGFR2 fusion/rearrangement, IDH1 mutation, or HER2 amplification, targeted agents are preferred after (or instead of) standard chemotherapy.
In summary: Gemcitabine + cisplatin (optimally with durvalumab or pembrolizumab) is better supported by evidence and is the current standard. Gemcitabine/GEMOX + lenvatinib combinations are promising, particularly for ICC, but remain investigational and should ideally be offered in the context of a clinical trial.
Sources: Sleisenger & Fordtran's GI and Liver Disease; Goldman-Cecil Medicine; Sabiston Textbook of Surgery; ABC-02 trial; TOPAZ-1; KEYNOTE-966; PMID 36928584 (Signal Transduct Target Ther 2023); PMID 41121202 (BMC Med 2025); NCCN BTC Guidelines v2.2025