Quick answer
Intrahepatic cholangiocarcinoma is the second commonest primary liver cancer, and even when it can be removed, more than half of patients recur and five-year survival is only 25–40%. For tumours that are large, multifocal or invading vessels, recurrence comes early. Fan Jia, Zhou Jian and Shi Guoming at Zhongshan Hospital tested giving treatment before surgery in exactly that group: GOLP — gemcitabine and oxaliplatin with the targeted drug lenvatinib and the PD-1 inhibitor toripalimab. Median event-free survival rose from 8.7 to 18.0 months, published in The New England Journal of Medicine on 5 March 2026 [1][2][3].
Key takeaways
- Design: investigator-initiated, multicentre, open-label, randomised phase II/III trial (NCT04669496) across 11 centres [4][5].
- Population: 178 patients with resectable intrahepatic cholangiocarcinoma carrying high-risk features [4].
- Regimen: GEMOX chemotherapy plus lenvatinib plus toripalimab, given before surgery [3].
- Primary endpoint met: median event-free survival 18.0 versus 8.7 months [2].
- Secondary outcomes: objective response rate 55%, R0 resection rate 95%, tolerable safety [4].
Why before surgery rather than after
Adjuvant chemotherapy after liver cancer surgery is poorly tolerated and frequently delayed by post-operative recovery, which is one reason trials of it have disappointed. Giving treatment first has three advantages: the tumour is measurable, so response can be assessed; the patient is fit enough to complete it; and downstaging can convert a marginal resection into a clear one [1][4].
The triple combination is deliberate: cholangiocarcinoma responds modestly to chemotherapy alone, has a fibroblast-rich stroma that lenvatinib may help normalise, and is being investigated as an immunotherapy-responsive tumour. Whether all three components are necessary is not answered by this trial [3][5].
How to read an 18-month event-free survival
Doubling event-free survival in a cancer with a 50% recurrence rate is a substantial result, and the 95% R0 rate suggests the surgery was not compromised by pre-treatment. But event-free survival is not overall survival, and the trial is reported before mature survival data [2][5].
With 178 patients across 11 centres, subgroup analyses will be underpowered, and an open-label design means the decision to operate and the timing of follow-up imaging were not blinded. Neoadjuvant treatment also delays surgery and can allow progression in the 10–20% who do not respond — the trial's response rate of 55% means a substantial minority gained nothing and lost time. Overall survival, the endpoint that matters most, requires longer follow-up [1][4].
The condition in context: intrahepatic cholangiocarcinoma
Incidence is rising in China, and most patients present too late for surgery.
| Region | What the data shows | Source |
|---|---|---|
| China | An estimated 39,000–58,000 new cases in 2018, with incidence rising [6]. | HEP journal cluster [6] |
| Prognosis | Post-resection recurrence exceeds 50%; five-year survival 25–40% [1]. | Fudan University [1] |
| Prior evidence | No internationally accepted neoadjuvant regimen existed before this trial [2]. | The Paper [2] |
Because high-risk features are common at presentation, a perioperative strategy that applies to them addresses most of the surgical population [1][2].
Why it matters
| Potential | What the report suggests | Source |
|---|---|---|
| New standard | First randomised evidence for neoadjuvant therapy in this disease [2][3]. | [2][3] |
| Surgical quality | R0 resection in 95% after pre-treatment [4]. | [4] |
| Caution | Open-label; overall survival immature [1][4]. | [1][4] |
Availability and limits
- This is an editorial summary of the publication, not a treatment recommendation [1][3].
- Eligibility requires resectable disease with defined high-risk features [4].
- Non-responders lose time before surgery [2].
- Immunotherapy and targeted therapy add toxicity and cost [5].
- Overall survival data are awaited [1].
What to ask before booking
| Question | Why it matters |
|---|---|
| Does my tumour meet the high-risk criteria used in the trial? | The benefit was shown in that defined group [4]. |
| What happens if I do not respond? | Roughly half respond; the rest lose time [2][4]. |
| Which department handles foreign patients, and in what language? | Zhongshan is a working hospital we do not book; confirm international-department coordination and language support. |
| How is resectability re-assessed after treatment? | Imaging response guides the decision to operate [1]. |
Questions patients ask
What did the trial test?
Neoadjuvant GOLP — gemcitabine and oxaliplatin plus lenvatinib and toripalimab — before surgery in resectable high-risk intrahepatic cholangiocarcinoma, versus upfront surgery [3].
What was the result?
Median event-free survival rose from 8.7 to 18.0 months; objective response rate was 55% and R0 resection rate 95% [2][4].
How many patients?
178 across 11 centres, in an investigator-initiated randomised phase II/III trial (NCT04669496) [4][5].
Where was it published?
The New England Journal of Medicine, online 5 March 2026 [1][2].
Is this relevant to international patients?
It is useful context for China's liver cancer care. Zhongshan is a working hospital we do not book — contact its international department directly [1].