← Return to [[Biliary Tract Cancer Clinical Trials]]
----
Phase III | Global randomized double-blind placebo-controlled trial | n=1069 | Advanced biliary tract cancer | Lancet 2023
**Kelley et al.**
Published in The Lancet, 2023
Enrollment: 2019–2021
[PubMed Link](https://pubmed.ncbi.nlm.nih.gov/37075781/)
---
### Essential Takeaway
KEYNOTE-966 demonstrated that adding **pembrolizumab** to gemcitabine/cisplatin improved overall survival in advanced biliary tract cancer and further validated chemoimmunotherapy as the modern first-line treatment paradigm following TOPAZ-1.
The study confirmed that PD-1/PD-L1 checkpoint inhibition provides clinically meaningful survival benefit when combined with Gem/Cis, although the magnitude of benefit remained modest and biomarker selection remains poorly defined.
---
### Clinical Question
Does adding pembrolizumab to gemcitabine/cisplatin improve survival compared with chemotherapy alone in unresectable or metastatic biliary tract cancer?
---
### Population
Included:
- Unresectable locally advanced BTC
- Metastatic BTC
- Recurrent BTC
Tumor types:
- Intrahepatic cholangiocarcinoma
- Extrahepatic cholangiocarcinoma
- Gallbladder carcinoma
Excluded:
- Ampullary carcinoma
- Active autoimmune disease
Most patients had metastatic disease (~88%).
---
### Study Design
Randomized 1:1 double-blind phase III trial.
#### Arm 1 — Pembrolizumab + Gem/Cis
- Pembrolizumab 200 mg q3 weeks
- Gemcitabine 1000 mg/m²
- Cisplatin 25 mg/m²
- Cisplatin capped at 8 cycles
- Gemcitabine allowed beyond 8 cycles
#### Arm 2 — Placebo + Gem/Cis
Identical chemotherapy backbone with placebo substitution.
---
### Endpoints
#### Primary Endpoint
- Overall survival (OS)
#### Secondary Endpoints
- Progression-free survival (PFS)
- Objective response rate (ORR)
- Duration of response
- Safety
---
### Key Results
### Overall Survival
- 12.7 vs 10.9 months
- HR 0.83
- p = 0.0034
24-month OS:
- 25% vs 18%
Survival curves separated early and remained separated throughout follow-up.
---
### Progression-Free Survival
- 6.5 vs 5.6 months
- HR 0.86
PFS did not meet the prespecified significance boundary.
---
### Objective Response Rate
- 29% vs 29%
No major ORR difference was observed.
However, responses were more durable with pembrolizumab:
- median duration of response:
- 9.7 vs 6.9 months
---
### Subgroup Findings
Benefit appeared generally consistent across:
- ICC
- EHCC
- gallbladder carcinoma
- Asian and non-Asian populations
- PD-L1 low and PD-L1 high groups
Importantly:
- PD-L1 expression did not clearly predict benefit
similar to TOPAZ-1.
---
### Toxicity
#### [[CTCAE Grading|Grade 3-4 Adverse Events]]
- 79% vs 75%
#### Immune-Related Toxicity
More common with pembrolizumab:
- hypothyroidism
- pneumonitis
- immune-mediated adverse events
Overall safety profile remained manageable without major unexpected toxicity signals.
---
### Interpretation
KEYNOTE-966 was important less because it introduced a new concept and more because it independently validated the chemoimmunotherapy paradigm established by TOPAZ-1.
Together, TOPAZ-1 and KEYNOTE-966 strongly established:
PD-1/PD-L1 inhibition + Gem/Cis as the modern first-line standard in advanced BTC.
One important distinction from TOPAZ-1 was that gemcitabine continuation beyond 8 cycles was permitted in KEYNOTE-966, reflecting real-world practice variation and potentially contributing to the earlier separation of survival curves.
Unlike classic cytotoxic intensification strategies, benefit appeared driven primarily by:
- improved long-term disease control
- more durable responses
- prolonged survival tail
rather than dramatic response rate increases.
---
### Important Limitations
#### Modest Absolute Benefit
Median OS improvement remained relatively small:
- 12.7 vs 10.9 months
---
### Biomarker Uncertainty
No clear predictive role for:
- PD-L1
- MSI status
in most patients.
---
### Heterogeneous Disease Grouping
Combined:
- ICC
- EHCC
- gallbladder carcinoma
despite major biologic differences.
---
### Predominantly Metastatic Population
Limited applicability to localized unresectable disease.
---
### Relationship to TOPAZ-1
TOPAZ-1 first established:
- durvalumab + Gem/Cis
KEYNOTE-966 then confirmed the broader principle that checkpoint inhibition improves outcomes when added to Gem/Cis in BTC.
Together, these trials firmly established chemoimmunotherapy as standard first-line treatment for advanced BTC.
---
### Current Practice Relevance
KEYNOTE-966 supports:
- pembrolizumab + Gem/Cis
as a first-line option in advanced BTC.
More broadly, it reinforced:
- immunotherapy integration
- durable disease control strategies
- ongoing biomarker discovery efforts in BTC
and further validated the post-ABC-02 chemoimmunotherapy era.
---
### Practice Impact
KEYNOTE-966 confirmed that checkpoint inhibition meaningfully improves survival in advanced BTC when combined with Gem/Cis.
Alongside TOPAZ-1, it helped solidify chemoimmunotherapy as the modern global first-line treatment paradigm in biliary tract cancer.