Drug therapy before surgery delayed relapse in high-risk bile-duct cancer

Giving a course of chemotherapy plus the targeted drug lenvatinib and an immunotherapy antibody before liver surgery, rather than operating straight away, roughly doubled the time patients with high-risk bile-duct cancer stayed free of recurrence: in a Chinese phase 2–3 trial, median event-free survival was 18.0 months in the neoadjuvant group against 8.7 months with immediate surgery [s1]. A survival advantage pointed the same way but did not clear the trial's own bar for significance, so the case for the approach rests for now on the delay in relapse, not on proven longer life [s1].

Intrahepatic cholangiocarcinoma — a cancer arising in the bile ducts inside the liver — is usually treated, when it can be removed at all, by surgery followed by chemotherapy. Even so, tumours with high-risk features come back often and early, and until now no treatment given before surgery had been shown to help [s1]. The trial tested whether hitting the cancer up front, while it is still in place, could change that pattern.

What they did

The ZSAB-neoGOLP trial randomly assigned 178 patients with resectable high-risk intrahepatic cholangiocarcinoma in a 1:1 ratio: 88 to neoadjuvant treatment and 90 to a control group that went straight to surgery [s1][s2]. The neoadjuvant regimen, abbreviated GOLP, combined intravenous gemcitabine and oxaliplatin plus the PD-1 antibody toripalimab every three weeks for three cycles with the oral drug lenvatinib taken daily for nine weeks, after which patients had the operation [s1]. Every patient in both groups then received eight cycles of the chemotherapy drug capecitabine after surgery, so the two arms differed only in what came before the knife [s1]. The primary endpoint was event-free survival — time without recurrence, progression that blocked surgery, or death; overall survival and safety were secondary [s1].

What it showed

At an interim analysis with a median follow-up of 16.9 months, event-free survival was significantly longer with neoadjuvant GOLP: a median of 18.0 months (95% confidence interval, 13.8 to 27.6) against 8.7 months (95% confidence interval, 7.2 to 12.4) in the control group, with P<0.001 [s1]. Survival trended in the same direction — 79% of the neoadjuvant group were alive at 24 months (95% confidence interval, 70 to 90) versus 61% of the control group (95% confidence interval, 50 to 75), a hazard ratio for death of 0.43 (95% confidence interval, 0.23 to 0.79) [s1].

That survival result carries an important caveat the authors state plainly. Its P value of 0.005 did not meet the pre-specified significance threshold for this interim look, a demanding two-sided alpha of 0.0019 [s1]. In practice that means the overall-survival difference, though large, could still be a chance finding under the trial's own rules, and the benefit that is established here is the delay in recurrence, not a confirmed gain in length of life. This is the honest reading of an interim endpoint that has not yet matured.

The cost side

Adding drugs before surgery added side effects. Across all phases of treatment, adverse events occurred in 97% of the neoadjuvant group and 70% of the control group [s1]. During the neoadjuvant phase itself, grade 3 or higher adverse events occurred in 28% of patients and treatment-related grade 3 or higher events in 26%, but no treatment-related adverse event led to death [s1]. The regimen did not, on the evidence reported, stop patients from reaching their operation.

What to watch

Two limits frame the result. The trial was run at Chinese centres in a patient population where the biology and causes of bile-duct cancer can differ from those elsewhere, and one of its authors is affiliated with the maker of toripalimab, so independent replication matters [s1]. And the survival curves need longer follow-up before the endpoint that patients care about most can be called either way.

Even so, a first demonstration that pre-surgical therapy can push back recurrence in a cancer with few good options is a genuine step. It fits a wider shift toward giving drugs before surgery in solid tumours, seen in muscle-invasive bladder cancer and tested as adjuvant therapy after colon-cancer surgery. It also sits apart from the targeted route into this disease, where a drug such as zenocutuzumab treats the small subset with an NRG1 gene fusion, and from local treatments used when liver tumours cannot be removed at all.

Sources

• [s1] Shi GM, Huang XY, Liang F, et al. Neoadjuvant GOLP in Resectable High-Risk Intrahepatic Cholangiocarcinoma. New England Journal of Medicine. 5 March 2026 (published online 4 March 2026).
• [s2] ClinicalTrials.gov. Neoadjuvant GOLP for Resectable Intrahepatic Cholangiocarcinoma (ZSAB-neoGOLP, NCT04669496). U.S. National Library of Medicine.