Cholangiocarcinoma

In brief

What it is, and why its map was redrawn

Cholangiocarcinoma is a cancer of the cholangiocytes, the cells lining the bile ducts, which carry bile from the liver to the intestine ¹. Under the microscope it is an adenocarcinoma, a cancer of gland-forming cells, and it is strongly desmoplastic, meaning the tumor wraps itself in dense, scar-like tissue ¹. The European consensus describes it not as one disease but as a heterogeneous group of cancers that share features of bile duct cells ²³. It is the second most common primary liver cancer after hepatocellular carcinoma, the cancer of the liver's main cells, and accounts for about 15% of primary liver cancers, about 3% of gastrointestinal cancers and about 2% of cancer deaths worldwide each year ^2,23.

It is divided by where it begins, because each location has its own genetic changes, symptoms and treatment ¹. Intrahepatic cholangiocarcinoma arises inside the liver, between the smallest ductules and the second-order ducts, the branches two steps below the main right and left ducts, and European and international liver societies now treat it as a distinct entity with its own guideline ²⁴. Perihilar cholangiocarcinoma arises at the hilum, the gateway where the right and left hepatic ducts leave the liver and join; Klatskin's 1965 description of these tumors gave them the name Klatskin tumor ^25,26. Distal cholangiocarcinoma arises lower in the common bile duct, nearer the pancreas and intestine ¹. Perihilar and distal tumors lie outside the liver and together are called extrahepatic ²⁷.

The classification changed twice, for different reasons ^27,28. The first change was bookkeeping: the second edition of the International Classification of Diseases for Oncology gave Klatskin tumors their own code but cross-referenced it to the intrahepatic site, so perihilar cancers were counted as intrahepatic ²⁷. In United States registry data from 1992 to 2000, 91% of the 269 tumors carrying that code were filed as intrahepatic, which overstated intrahepatic incidence by 13% and understated extrahepatic incidence by 15% ²⁶. The second change was biological: the fifth World Health Organization classification splits intrahepatic tumors into a large-duct type and a small-duct type by location, appearance, protein markers and gene changes ²⁸. In one series of 190 resected tumors, every FGFR2 fusion was in the small-duct type ²⁸.

It is a quiet cancer ²⁴. Early disease usually causes no symptoms, and the symptoms that eventually come are nonspecific, so most patients are diagnosed when the cancer is locally advanced or has spread ^13,19. Pain and jaundice, the yellowing of skin and eyes when bile backs up into the blood, are two of the symptom scores on the disease-specific quality-of-life questionnaire that the KEYNOTE-966 analysis prespecified ²⁰. The silent presentation, the tumor's aggressiveness and its resistance to treatment together explain why mortality is so high ²⁴.

The whole history, including what was overturned

The history of this cancer runs on three tracks: a parasite story in Southeast Asia, a surgical story in Western hospitals, and a molecular story that arrived only in the past fifteen years ^31,33,37. The parasite track runs through the hamster: in 1983 Flavell and Lucas showed that Opisthorchis viverrini plus a single dose of a nitrosamine, one of a class of chemical carcinogens, produced bile duct cancers that neither caused alone ³¹, and by 1994 the International Agency for Research on Cancer had placed the fluke in Group 1 ⁹. Its East Asian relative, Clonorchis sinensis, followed in 2009 ⁹.

The surgical track is a story of a door that closed and reopened ^37,38. Klatskin's 1965 paper made perihilar tumors a recognized entity ²⁵, and in 1975 Bismuth and Corlette argued that the cholangiogram, an X-ray of dye in the ducts, read for which junctions of the ducts the tumor had invaded, should decide the operation ³⁹. Liver transplantation was tried and largely abandoned: in a registry of 207 patients, 51% had recurrence and five-year survival was 23%, and the registry's authors concluded in 2000 that transplantation should seldom be used ³⁷. Five years later the Mayo Clinic reported 82% five-year survival after chemoradiation followed by transplantation in selected perihilar patients, against 21% after conventional resection ³⁸.

The drug track had no randomized standard until 2010, when the ABC-02 trial made gemcitabine with cisplatin the standard by extending median survival from 8.1 to 11.7 months ⁴⁰. Immunotherapy was added to that pair in 2022 and 2023 ^41,42. From 2020 a series of drugs aimed at single gene changes was approved in the United States, and four of them were approved on tumor shrinkage in trials without a comparison group, one of which has since been withdrawn ^{15,16,17,43,44}.

The record contains several reversals, each marked in the timeline ^4,26. Part of the rise in intrahepatic cancer was a coding artifact, though a real rise survived correction ²⁶. The highest incidence ever recorded turned out to be falling, by about 3% a year ⁴. A third chemotherapy drug that looked like a breakthrough in a single-arm trial added nothing when randomized ^45,46. And the response rates on which several targeted drugs were approved turn out to predict survival in this disease almost not at all ⁴⁷.

How common it is, and where

Cholangiocarcinoma is uncommon in most high-income countries and a major health problem in endemic areas ¹⁹. The Global Burden of Disease study, which counts gallbladder and bile duct cancers together, estimated 216,768 new cases and 171,961 deaths worldwide in 2021, an age-standardized incidence of 2.6 and mortality of 2.0 per 100,000, with both rates falling slowly over three decades ⁵⁵. The best international comparison, using registries that record both the site and the tissue type of each tumor in 38 countries for intrahepatic and 33 for extrahepatic disease, found the highest age-standardized rates in Asia: South Korea at 2.80 per 100,000 for intrahepatic and 2.24 for extrahepatic, Thailand at 2.19 and 0.71, and Japan at 0.95 and 0.83 ⁵⁶. An age-standardized rate is adjusted so that populations with different age structures can be compared fairly. Between 1993 and 2012, incidence of both types rose in most of the countries studied ⁵⁶.

Northeast Thailand is the exception that defines the disease ³. In the Khon Kaen Cancer Registry, 10,731 of 18,589 liver cancers recorded from 1985 to 2009, 58%, were cholangiocarcinoma, the reverse of the usual pattern in which hepatocellular carcinoma dominates ³. The age-standardized rate over that period was 44.3 per 100,000 men and 17.6 per 100,000 women, with single years as high as 62.0 in men ³. The national Thai figure of 2.19 and Khon Kaen's 44.3 describe different populations, and the gap is the geography of the fluke, whose prevalence tracks cancer incidence in the northeast ^9,56. A program report quoted 135.4 per 100,000 for Khon Kaen men without giving its period or method; the registry's 44.3 is the measured long-run rate ^3,49. Only 10.8% of the registry's cases were confirmed by cytology or histology ³. The most recent national estimate, from four Thai registries covering 2012 to 2021, is 12.5 per 100,000 men and 5.9 per 100,000 women, 8.9 for both sexes together, with 19.2 in men in the northeast, and incidence is falling 7.2% a year in men and 5.8% in women, with 5.5 to 6.1 per 100,000 men projected for 2026 ⁵⁷.

That rate is now falling ⁴. Analyzed over 1989 to 2018 by three independent methods, incidence in Khon Kaen fell 3.1% a year in men and 2.4% a year in women ⁴. The fall follows birth cohorts: men born in 1998 had an incidence rate ratio of 0.09 against men born in 1966, about one-eleventh the risk, and the projected rates for 2028 are 7.6 per 100,000 men and 3.6 per 100,000 women ⁴. The earlier analysis of 1990 to 2009 had already reported a decline of 0.7% a year in men, though its confidence interval, the range within which the true value probably lies, ran from −2.1% to +0.8% and so included no change at all ³.

In the United States the trend runs the other way ⁵. In national registry data from 2001 to 2017, cholangiocarcinoma incidence rose 43.8%, from 3.08 to 4.43 per 100,000 person-years (a person-year is one person followed for one year); intrahepatic incidence rose 148.8%, from 0.80 to 1.99, and extrahepatic incidence rose 7.5% ⁵. The largest relative rise, 81.0%, was in adults aged 18 to 44 ⁵. Over the same years diagnoses of cancer of unknown primary site fell 54.4%, so some tumors once filed as unknown primary may now be called intrahepatic cholangiocarcinoma; the authors argue the rise is real because the absolute change in intrahepatic disease was larger ⁵. Three of that paper's authors listed Incyte Corporation, which markets pemigatinib for this cancer, as their affiliation ^5,15.

Where registries record subtype carefully, the proportions differ by place ^27,58. In the population registry of Burgundy, France, intrahepatic cholangiocarcinoma made up 40% of 714 biliary tract cancers recorded from 2012 to 2019, half of all patients were older than 75, and incidence was stable apart from a slight rise in intrahepatic disease in men ⁵⁸. In England and Wales from 1990 to 2008, the intrahepatic rate in men rose from 0.43 to 1.84 per 100,000 while the extrahepatic rate fell from 0.78 to 0.51, a pattern partly shaped by the same coding rules ²⁷.

What causes it, and how strong each cause is

Most cholangiocarcinomas have no identifiable cause ¹. The known risk factors vary by region and together account for a minority of cases ⁶. What they share is chronic injury to the bile ducts, from liver flukes in endemic regions and from stones, cysts or primary sclerosing cholangitis elsewhere ¹. That is why the disease clusters where it does: in northeast Thailand, 89.1% of the first people enrolled in screening had eaten uncooked fish ⁴⁹, while in Sweden the risk concentrates in people with primary sclerosing cholangitis ⁷.

The strongest associations are structural ⁶. A pooled analysis of 25 case-control studies from seven countries found that the strongest risk factors for both intrahepatic and extrahepatic disease were bile duct cysts and stones, cirrhosis, hepatitis B and hepatitis C ⁶. Choledochal cysts, congenital balloon-like widenings of the bile duct, carried the largest odds ratios, 26.71 for intrahepatic and 34.94 for extrahepatic cancer ⁶. An odds ratio is not the chance of getting the disease, only a comparison of how common the exposure is in each group. Among 1,337 Western adults with choledochal cysts, 10.9% had a malignancy, most often cholangiocarcinoma, and removing the cyst reduced but did not eliminate the risk ⁵⁹. Hepatolithiasis, stones in the ducts inside the liver, sits within the stone category of that pooled analysis ⁶; a separate meta-analysis of seven case-control studies put the odds ratio for intrahepatic cancer at 17.64 for stones in the bile ducts, 11.79 for stones in the main duct alone, and 2.00 for gallstones, which lie outside the ducts ⁶⁰.

Primary sclerosing cholangitis is an immune disease that inflames and scars the bile ducts; in a national Swedish cohort of 604 patients followed from 1970 to 1998, 79% also had inflammatory bowel disease ⁷. In that cohort 13.3% developed a bile duct or liver cancer, a risk 161 times that of the general population, and after the first year the rate held at about 1.5% a year ⁷. A 2025 meta-analysis of 51 studies and 26,482 patients put the incidence of cholangiocarcinoma in the disease at 9.31 per 1,000 person-years, with higher figures in smaller studies ⁶¹. Cholangiocarcinoma is the main cause of death in primary sclerosing cholangitis, and there is still no good tumor marker to screen for it ⁶².

Hepatitis B and C, the two viral infections that cause chronic liver inflammation, are among the strongest risk factors in the pooled data, and the association of hepatitis B with intrahepatic cancer differs between Eastern and Western studies ⁶. Metabolic dysfunction-associated steatotic liver disease, MASLD, the fatty liver disease that travels with obesity and diabetes, has a weaker association: a 2026 meta-analysis of 29 studies gave a hazard ratio of 1.26 for cholangiocarcinoma ⁶³. Earlier pooling of case-control studies found a stronger association, an odds ratio of about 2.2, confined to intrahepatic cancer with none for extrahepatic disease ⁶⁴; the true figure probably lies between, and either way fatty liver is a weak risk factor compared with cysts, stones or duct inflammation. A hazard ratio compares the rate at which an outcome occurs over time in exposed and unexposed people. In the case-control pooling, obesity and high blood pressure showed no significant association, while diabetes, a weaker factor, is becoming more common worldwide ⁶.

Two causes are historical or regional ^35,65. Thorotrast, a thorium-based contrast agent introduced in 1929 and used worldwide into the 1950s, is retained in the body for life ³⁵ and emits alpha particles, a heavy, short-range form of radiation ³⁶. It induces intrahepatic cholangiocarcinoma and angiosarcoma decades after injection, and Nordic data associate it with a 100-fold risk of liver cancer overall ^36,66. In the lower Mekong, liver flukes dominate the discussion, but a meta-analysis of 18 studies found that the combination of alcohol and smoking carried an odds ratio of 11.1, which its authors called a greater risk factor than exposure to Opisthorchis viverrini ⁶⁵.

How it develops: injury, a second hit, and the mutations that follow

No single mechanism explains every cholangiocarcinoma, but the best-studied path begins with chronic inflammation of the duct lining ¹. In the fluke belt three intertwined mechanisms have been proposed, worked out largely in the Syrian golden hamster: direct damage to the duct lining by the worms; oxidative stress, the damage done by reactive molecules that inflamed tissue releases; and proteins the parasite secretes that push host cells to divide and resist dying ⁹.

The secreted proteins are the most specific part of the story ²⁹. In 2009 a growth factor resembling human granulin, named Ov-GRN-1, was found to be the main growth-promoting protein in the fluke's secretions, active at nanomolar concentrations, that is, at very low doses, in cell culture ²⁹. In 2019 researchers used CRISPR, a gene-editing tool, to disable that gene in the fluke itself; the edited worms still matured in hamsters but caused less thickening and scarring of the ducts ⁵¹. In 2022 the same approach, with a dietary nitrosamine added, produced fewer TP53 mutations in duct cells and fewer hamsters with high-grade cancer ³⁰. Both experiments were done in hamsters by one collaborating group of laboratories ^30,51.

The second hit is chemical ³¹. In 1983, neither Opisthorchis viverrini nor a single oral dose of the nitrosamine N-nitrosodimethylamine caused malignant bile duct tumors in hamsters, but the two together did, in 10% and 20% of animals depending on the order in which they were given ³¹. In 1994 a single injected dose before infection produced cholangiocarcinoma in 44% of hamsters at 45 weeks, against none with the chemical alone; the often-quoted 100% figure applied only to continuous dosing ³². Nitrosamines are a class of chemical carcinogens, and reviews name them as a cofactor present in the undercooked and fermented fish dishes of the endemic region ⁹.

What the cancer becomes depends on its cause ^33,34. In 209 tumors from Asia and Europe, those related to Opisthorchis viverrini more often carried TP53 mutations, while those without the fluke more often carried mutations in BAP1, IDH1 and IDH2 ³³. In fluke-related tumors, TP53 was mutated in 44%, KRAS in 16.7% and SMAD4 in 16.7%, while BAP1 and IDH1/2 were each mutated in 2.8% ¹². A study of 489 tumors from 10 countries sorted them into four clusters that split by fluke status, with fluke-positive tumors enriched for extra copies, called amplification, of ERBB2, the gene for HER2, a growth-signal receptor on the cell surface that several drugs can block, and for TP53 mutation ³⁴.

Outside the fluke belt, intrahepatic tumors more often carry changes a drug can target ^33,67. In 32 intrahepatic tumors, almost half had an inactivating mutation in BAP1, ARID1A or PBRM1, genes that control how tightly DNA is packed ⁶⁸. FGFR2 fusions, in which part of a growth-factor receptor gene is joined to another gene and left switched on, were found in 13.6% of intrahepatic tumors and never alongside KRAS or BRAF mutations ¹⁰. In 260 biliary tract cancers, nearly 40% carried an alteration that a drug could target ⁶⁷. IDH1 mutations, changes in the gene for the metabolic enzyme isocitrate dehydrogenase 1, occur in about 13% of intrahepatic tumors by one estimate and up to about 20% by another ^11,69.

How it is found, and how it is missed

No single test settles the diagnosis: noninvasive approaches are not accurate enough on their own, so tissue confirmation is needed ². Imaging can do more than find a mass: on MRI, magnetic resonance imaging, features such as poor enhancement in the arterial phase and dilated ducts inside the liver distinguished the large-duct from the small-duct intrahepatic type with an area under the curve of 0.91 in 93 patients ⁷⁶. The area under the curve is a measure of how well a test separates two groups, where 1.0 is perfect and 0.5 is chance.

The most studied blood test is CA 19-9, a sugar-bearing molecule shed into the blood by some tumors and also by inflamed or blocked bile ducts, so a high value in a jaundiced patient can come from the blockage itself rather than from cancer: of 114 patients drained for obstructive jaundice, 59% of those whose disease proved benign had a raised value, but only 12% still did after drainage, against 63% of those with cancer ^77,78. Pooled across 31 studies of 1,264 patients, it had a sensitivity of 72% and a specificity of 84% for cholangiocarcinoma, with lower sensitivity in European patients ⁷⁸. Sensitivity is the share of people with the cancer whom a test catches; specificity is the share without it whom the test correctly clears. About 10% of people, called Lewis-negative, cannot make CA 19-9 at all, so in them the test is uninformative whatever the tumor is doing ⁷⁹. The marker says more about outlook than diagnosis: among patients starting second-line chemotherapy in ABC-06, those with a high baseline CA 19-9 lived a median of 4.4 months against 6.4 ⁸⁰.

For a narrowing of a bile duct, the long-standing next step is ERCP, endoscopic retrograde cholangiopancreatography, during which cells are brushed from the narrowing ⁸¹. In a 2026 network meta-analysis of 38 studies and 4,912 patients, a pooling that compares several tests at once even where no study compared them head to head, brush cytology had a pooled sensitivity of 45% and a specificity of 97%: it rarely calls a benign narrowing cancer, and it misses more than half of cancers ⁸¹. FISH, fluorescence in situ hybridization, adds a search for abnormal numbers of chromosomes in the brushed cells; combined with cytology it ranked second for accuracy, behind cholangioscopy, but the gap in sensitivity between the two was not statistically significant, so the authors call the combination a lower-risk alternative where a camera is unavailable ⁸¹. Earlier pooled data put FISH alone at a sensitivity of 54.2% ⁸².

Cholangioscopy puts a camera inside the duct so the narrowing can be seen and biopsied directly ⁸¹. Single-operator cholangioscopy had the highest pooled sensitivity, 82%, but a lower specificity, 89%, and more complications, 6.8% against 2.1% for cytology-based sampling ⁸¹. In primary sclerosing cholangitis, where the ducts are already scarred and narrowed, its pooled sensitivity was 65% ⁸³. Patients with that disease are watched with scheduled ERCP or yearly MRI, and in a 2025 comparison of 1,629 patients the cumulative incidence of cholangiocarcinoma was lower with scheduled ERCP but not significantly different from MRI surveillance ⁶².

Once advanced disease is confirmed, the tumor's genes become part of the diagnosis ¹³. European guidelines recommend next-generation sequencing, which reads many genes at once, when advanced disease is diagnosed, and a liquid biopsy, which reads tumor DNA shed into the blood, can be used when the tissue sample is too small ¹³. The FDA approved a companion test alongside pemigatinib to find the FGFR2 rearrangements it targets ¹⁵.

• CA 19-9 (blood)A carbohydrate antigen measured in blood, the most studied blood marker for this cancer; blocked or inflamed bile ducts raise it too ^77,78.To support a diagnosis and follow treatment; not as a screening test in primary sclerosing cholangitis, where no good tumor marker exists ⁶²Pooled sensitivity 72% and specificity 84% across 31 studies, with lower sensitivity in European patients ⁷⁸About 10% of people are Lewis-negative and cannot make CA 19-9 at all, so a normal or undetectable value in them says nothing about the tumor ⁷⁹
• ERCP brush cytologyCells brushed from a narrowed duct during endoscopic retrograde cholangiopancreatography, the long-standing way of sampling a bile duct stricture ⁸¹.First tissue test for a suspicious narrowing of the bile ductPooled sensitivity 45% and specificity 97% across 38 studies ⁸¹More than half of cancers; a negative brushing does not clear a stricture ⁸¹
• FISH on brushed cellsFluorescence in situ hybridization, which looks for abnormal numbers of chromosomes in the brushed cells, added to standard cytology ⁸¹.When cytology is negative or uncertain but suspicion remainsSensitivity 54.2% on its own in pooled data ⁸²; combined with cytology it ranked second for accuracy, behind cholangioscopy, though the gap in sensitivity between the two was not statistically significant ⁸¹Still close to half of cancers when used alone ⁸²
• Cholangioscopy with targeted biopsyA camera passed inside the bile duct so the narrowing can be seen and biopsied directly ⁸¹.Indeterminate strictures after brushing, where equipment and expertise allowPooled sensitivity 82% and specificity 89% ⁸¹; in primary sclerosing cholangitis, sensitivity 65% ⁸³More false positives than cytology and more complications, 6.8% against 2.1% ⁸¹
• Ultrasound screening (CASCAP)Abdominal ultrasound offered at least yearly to residents of northeast Thailand aged 40 and over, looking for masses, dilated ducts and scarring around the ducts ^49,100.Population screening in the fluke-endemic northeast of ThailandNo sensitivity for early cancer has been published; screen-detected patients had 53.9% five-year survival against 21.9% for walk-in patients, a retrospective comparison ¹⁰²Lead time: finding a cancer earlier lengthens survival counted from diagnosis even if death is not delayed, and the screened-versus-usual difference after surgery was not significant (p = 0.06) ⁴⁸

What is done about it

Surgery is the only treatment that can cure, and only about 20% of patients are eligible for resection with curative intent ⁵⁰. The best outcomes come from management by specialist teams, and the operation depends on where the tumor is, which may mean removing part of the liver, part of the pancreas or, less often, both ^19,50. Who reaches the operating room is not only a matter of anatomy: in a French population registry, only 15% of patients with non-metastatic intrahepatic cancer had a complete resection, and age, physical fitness and the type of hospital were all associated with whether surgery happened ⁵⁸. Before any of this, most patients with perihilar or distal tumors need the blocked duct opened, by a stent placed at ERCP or a drain passed through the skin, both to relieve jaundice and to make chemotherapy possible; whether to drain before surgery remains contested, and a 2025 pooling of 21 studies found less liver failure but more infection, more cholangitis and higher long-term mortality after drainage ⁸⁴. For tumors that cannot be removed, a 2026 pooling of 12 studies found that metal stents stayed open about twice as long as plastic ones, with less cholangitis and fewer repeat procedures, but survival was the same ⁸⁵. Treatment before surgery is new: in a 178-patient Chinese trial of resectable intrahepatic tumors at high risk of recurrence, three cycles of chemotherapy plus the immunotherapy drug toripalimab and the targeted drug lenvatinib before resection lengthened the time free of recurrence or death from 8.7 to 18.0 months, and 79% against 61% were alive at two years, a survival difference that did not meet the trial's own threshold for significance ⁸⁶.

After resection, capecitabine, an oral chemotherapy, is the longest-established follow-on treatment with trial evidence ⁵⁰. In the BILCAP trial of 447 patients, median survival was 51.1 months with capecitabine and 36.4 months with observation, but the primary intention-to-treat comparison, which counts every patient in the group they were assigned to, missed significance, with a hazard ratio of 0.81 and a p value of 0.097; the benefit reached significance only in the protocol-specified sensitivity analysis, which adjusts for factors chosen in advance, and in the per-protocol analysis, which counts only the patients who took the treatment as planned ⁵⁰. A hazard ratio below 1 favors the treatment, and a p value above 0.05 is conventionally treated as not significant. Before BILCAP, no study had shown a benefit from treatment after surgery in this disease ⁵⁰. Since then a Japanese trial of 440 patients found that S-1, an oral fluorouracil-type drug, raised three-year survival after resection from 67.6% to 77.1%, an adjusted hazard ratio of 0.69, a result its authors framed as a standard for Asian patients ⁵², and a 93-patient Chinese trial of chemoradiation plus the immunotherapy drug camrelizumab against observation reported a survival hazard ratio of 0.43, a result that needs confirmation against active treatment rather than against no treatment ⁸⁷.

For a narrow group of perihilar patients, liver transplantation after chemoradiation is an option ^22,38. The Mayo Clinic protocol, begun in 1993, combines radiation, a drug that sensitizes the tumor to radiation and transplantation for patients with early, node-negative (no spread to lymph nodes) perihilar tumors that cannot be resected or that arise in primary sclerosing cholangitis ³⁸. Across 12 United States centers and 287 patients, five-year survival counted from the start of therapy was 53%, and recurrence-free survival after transplantation was 65% ²². Selection is decisive: patients with a mass larger than 3 cm, a biopsy taken through the abdominal wall or spread beyond the liver did significantly worse, and 193 of the 287 patients came from one center ²². Transplantation for intrahepatic tumors, long refused, is being revisited for very small or chemotherapy-responsive cancers; a 2026 pooling of seven retrospective comparisons, 346 transplanted against 5,132 resected, found better five-year survival after transplantation, but the finding depended on a single study and no randomized trial exists ⁸⁸.

For unresectable or metastatic disease, the standard since 2010 has been gemcitabine with cisplatin, which in ABC-02 raised median survival from 8.1 to 11.7 months against gemcitabine alone ⁴⁰. Immunotherapy, drugs that release brakes the immune system places on itself, has since been added ^41,42. In TOPAZ-1, durvalumab, which blocks the brake protein PD-L1 ⁸⁹, improved survival with a hazard ratio of 0.80 in 685 patients ⁴¹, and at three years 14.6% of patients on durvalumab were alive against 6.9% on placebo ²¹, and at four years 11.8% against 4.3%, with median survival of 13.0 against 11.4 months ⁵⁴. In KEYNOTE-966, pembrolizumab, which blocks the receptor PD-1 ⁸⁹, raised median survival from 10.9 to 12.7 months in 1,069 patients ⁴². Both combinations are approved in the United States, durvalumab on September 2, 2022 and pembrolizumab on October 31, 2023, and both are recommended for first-line use ^13,89,90. TOPAZ-1 was funded by AstraZeneca and KEYNOTE-966 by Merck Sharp & Dohme, and company employees are among the authors of both ^41,42.

Not every addition has worked ⁴⁶. A three-drug regimen adding nab-paclitaxel produced a median survival of 19.2 months in a 60-patient single-arm trial, measured against historical controls rather than a randomized control arm ⁴⁵. In the randomized SWOG S1815 trial of 441 analyzable patients it gave 14.0 months against 13.6 months for gemcitabine and cisplatin, a difference that was not significant, with more toxicity ⁴⁶. Exploratory subgroup analyses favored the three drugs in locally advanced rather than metastatic disease and, for the time before the cancer grew, in gallbladder cancer, but subgroup findings of this kind generate hypotheses rather than settle them ⁴⁶. Two more phase 3 trials in 2025 point the same way: a second taxane added in Korea changed nothing, 12.0 against 11.1 months in 150 patients, and that trial too was stopped early ⁹¹; and a chemically modified gemcitabine designed to work better did worse than the drug it was meant to replace, 9.2 against 12.6 months in 773 patients, so the trial was stopped for futility ⁹².

After first-line treatment fails, FOLFOX chemotherapy, a combination of folinic acid, fluorouracil and oxaliplatin, extended median survival from 5.3 to 6.2 months against symptom control alone in ABC-06, and raised 12-month survival from 11.4% to 25.9% ⁹³. A second option, liposomal irinotecan with fluorouracil, gave conflicting results in a Korean and a German trial; pooled across 278 patients it delayed progression from 1.8 to 3.6 months and lengthened survival from 6.1 to 8.1 months, a difference that just missed significance ⁹⁴. For the minority whose tumors carry a targetable change, targeted drugs are recommended in the second or third line, and their evidence and approval status are set out below ¹³.

Where it leads: survival, counted honestly

The numbers are bleak and have moved slowly ^4,5. In United States registry data from 2001 to 2017, median overall survival from diagnosis was 8 months for all cholangiocarcinoma, 6 months for intrahepatic and 9 months for extrahepatic disease ⁵. Across all patients with biliary tract cancer, five-year survival is below 10% ⁵⁰. In Khon Kaen, five-year relative survival, the share alive compared with people of the same age and sex in the general population, was 10.9% ⁴.

Survival is set largely by whether the tumor can be removed ⁵⁰. In the Burgundy registry, only 15% of patients with non-metastatic intrahepatic cancer had a complete resection, one with no tumor left at the cut edges ⁵⁸. In a Khon Kaen surgical series of 1,091 patients, median survival after curative-intent resection rose from 14 months in 2002–2013 to 40 months in 2014–2021, as early-stage cases rose from 16.0% to 29.1% and positive margins fell from 53.7% to 40.0% ⁴⁸.

With modern treatment of advanced disease, a minority now live for years ²¹. In the three- and four-year updates of TOPAZ-1, 17.0% of patients on durvalumab were extended long-term survivors, alive at least 30 months after randomization, against 8.7% on placebo, and 11.8% against 4.3% were alive at four years ^21,54. After transplantation for selected perihilar disease, 65% were free of recurrence at five years ²². These figures describe selected patients in trials and specialist centers, not the disease as a whole ^21,22.

Screening where the flukes are

Population screening for this cancer began in one region ⁴⁹. The Cholangiocarcinoma Screening and Care Program, CASCAP, began in northeast Thailand in 2013 as the first program to screen the at-risk population at community level, offering ultrasound at least yearly ^48,49. It enrolls residents aged 40 and over and had enrolled 394,026 people by 2019 ¹⁰⁰. Among the first 85,927, 89.1% had eaten uncooked fish and 42.2% of those tested for liver fluke were infected ⁴⁹. Before CASCAP, control and prevention in the region from 1984 onward had focused on health education ⁴⁹.

Ultrasound screening looks for periductal fibrosis, scarring around the ducts, and for dilated ducts, which can signal a blockage ¹⁰⁰. In CASCAP, the most extensive grade of periductal fibrosis was associated with bile duct dilatation with an adjusted odds ratio of 5.74 ¹⁰⁰. Killing the worm does not reliably reverse that scarring: one year after praziquantel, 34.6% of infected people with fibrosis still had it, and a urine marker of oxidative DNA damage fell only in those who had no fibrosis to begin with ¹⁸. Nor does killing the worm keep it away: in a 2026 cohort of 612 villagers followed after selective praziquantel, reinfection measured by a urine antigen test ran at 64 per 100 person-years, ten times the rate stool microscopy detected, and people treated before were more likely to be infected again ¹⁰¹.

Screen-detected patients live longer, but how much of that gain is real is unsettled ^48,102. In 711 histologically proven cases, five-year survival was 53.9% in those found by ultrasound screening and 21.9% in those who walked in with symptoms, in a retrospective comparison ¹⁰². Because screening finds cancers earlier, survival counted from diagnosis lengthens even when death is not delayed, an effect called lead-time bias. In the surgical series, median survival after resection was 51 months for screened patients and 38 months for others, a difference that did not reach significance (p = 0.06), and the program's own surgeons wrote in 2024 that its clinical utility remains unclear ⁴⁸.

Elsewhere, surveillance is offered to people already known to be at high risk ⁶². Patients with primary sclerosing cholangitis are followed with scheduled ERCP or yearly MRI; the American liver society's guidance recommends imaging the ducts every year in adults with the disease, acknowledging that this rests on expert opinion rather than trials ¹⁰³; in a 2025 comparison of 1,629 such patients, the combined rate of bile duct or liver cancer, transplantation or liver-related death was lowest with scheduled ERCP, 14.1% against 35.0% with ERCP only on demand, and there is still no tumor marker good enough to rely on ⁶². Adults with choledochal cysts remain at some risk even after the cyst is removed ⁵⁹. In western Siberia, where Opisthorchis felineus is endemic, the authors of a 2023 case-control study called for screening guidelines to be developed ⁵³.

What was believed, and is not

Cholangiocarcinoma has collected confident statements that later evidence qualified ^26,47. Some were corrected by better analysis of the same registries that produced them, as when a falling rate was found inside the famous high one and a coding rule was found inside a famous rise ^4,26. Others were corrected by randomized trials that tested what single-arm studies had suggested ⁴⁶, or by a statistical look at whether the measures used for approval predict survival at all ⁴⁷. The cards below give what was said, what replaced it and when.

Two of them matter for how a listener should hear the rest of the season. Gene-editing results in the fluke are about the fluke's genes, not the patient's ⁵¹, and curing the infection has not been shown to undo the cancer risk it leaves behind ¹⁸. Neither correction weakens the parasite link; both make it more precise ⁸. The second also shapes practical advice: in endemic regions, control rests on stopping people from eating raw or fermented fish and on treating animal reservoirs, not on deworming after years of infection ⁹.

What is still unknown

Whether treating fluke infection prevents the cancer is the largest open question ^9,65. Praziquantel cures the infection, but mass deworming has struggled against reinfection, and repeated cycles of cure and reinfection may themselves promote carcinogenesis ⁹. The human data point in different directions: praziquantel treatment was associated with higher cancer risk in a meta-analysis that could not separate the drug from the exposure that prompts it ⁶⁵, while fibrosis persisted in a third of treated people with established scarring ¹⁸. The fall in Khon Kaen follows birth cohorts, but the registry analysis does not identify what changed ⁴.

It is not known whether several of the approved targeted drugs extend life ⁴⁷. Pemigatinib, futibatinib and zanidatamab were approved on response rates from single-arm trials ^15,16,17; the one randomized first-line test of pemigatinib since then lengthened the time before the cancer grew but did not lengthen life ⁹⁶, response rate barely tracks survival in this disease ⁴⁷, and the confirmatory trial for infigratinib closed after randomizing 48 patients because too few could be enrolled ⁴³. Its authors wrote that the episode shows how hard it is to power confirmatory trials in rare, biomarker-selected groups ⁴³.

Most cases outside the fluke belt have no known cause ¹. Whether the rise in intrahepatic cancer in the United States reflects metabolic disease, the reclassification of tumors once called cancers of unknown primary, or something unmeasured is not settled ^5,56. The association with MASLD is modest, a hazard ratio of 1.26, and comes from observational studies ⁶³. Diabetes is a weaker factor that is becoming more common worldwide, and the pooled analysis suggests it may be contributing to the rise without being able to show it ⁶.

The granulin mechanism is established in cells and hamsters, not in people, and the knockout experiments come from one collaborating group of laboratories ^29,30. Why so few helminths cause cancer while opisthorchiids do is itself unexplained; a 2026 review proposes that inflammation becomes carcinogenic only when intensity, duration, genotoxic stress, tissue vulnerability and environmental cofactors coincide ¹⁰⁴. Whether Opisthorchis felineus carries the same risk as its Asian relatives rests, in humans, on one study of 40 cases ⁵³, and the agency still lists it as not classifiable ¹⁰⁵.

Whether ultrasound screening saves lives has not been settled by the published comparisons, which are retrospective ^48,102; the study that would settle it would compare screened and unscreened communities on deaths from cholangiocarcinoma. Whether any blood marker can find early cancer in high-risk people is equally open, and in primary sclerosing cholangitis there is still no good one ⁶². One risk-mapping study of liver flukes in Surin Province, Thailand, has been retracted, and no figure on this page comes from it ¹⁰⁶.

How it connects to parasites and cancer, at its true strength

This is the strongest parasite-cancer link in the season ^8,9. Two liver flukes, Opisthorchis viverrini and Clonorchis sinensis, are classified by the International Agency for Research on Cancer as carcinogenic to humans, in 1994 and 2009 ⁹. The third human liver fluke, Opisthorchis felineus of Siberia and eastern Europe, remains in Group 3, meaning the agency found the evidence insufficient to classify it ¹⁰⁵. Pooled across 22 studies and 34,367 participants, infection with any of the three human liver flukes was associated with cholangiocarcinoma with an odds ratio of 4.24, and with a risk ratio of 10.43 whose confidence interval runs from 2.90 to 37.47 ⁸. An interval that wide means the estimate is imprecise. By species the odds ratio was 4.49 for Clonorchis sinensis and 3.69 for Opisthorchis viverrini, with too few studies to estimate Opisthorchis felineus separately ⁸. In South Korea, where Clonorchis is the fluke, roughly one cholangiocarcinoma in ten is attributed to it ¹⁰⁷.

A fourfold association is serious, but it does not mean that most infected people get cancer ^3,8. The three flukes infect about 25 million people ⁸, and even in Khon Kaen at its worst the cancer struck 44.3 per 100,000 men a year averaged over 25 years, and 62.0 in the highest single year ³. In the fluke belt other exposures matter as much or more, including alcohol with smoking ⁶⁵. The fluke raises the risk, and diet, chemicals, genes and time help decide who develops the cancer, which is the argument of a 2026 review that inflammation causes cancer only when several conditions coincide ¹⁰⁴.

The causal case rests on kinds of evidence that point the same way ^8,30. Human studies show association ⁸. Hamster studies show that the fluke promotes cancer started by a nitrosamine ^31,32. Gene-edited worms show that one parasite protein contributes to scarring, TP53 mutation and cancer in hamsters given a nitrosamine ³⁰. And the tumors themselves carry a different mutation pattern when a fluke is involved, which is what a distinct cause should leave behind ^33,34.

It is fair to call fluke-related cholangiocarcinoma a cancer with a known cause ⁹. It is not yet fair to call that cause treatable in the sense that treatment prevents the cancer: the infection is curable, but cure has not been shown to undo the scarring or the risk, and reinfection is common ^9,18. The honest description is a known, preventable cause whose prevention depends on stopping infection, not on curing it after the damage is done ⁹.

The season asks whether the cancers attributed to parasites are the real number or only the number anyone has looked for, and this disease answers in both directions. In the region with the most cases, only 10.8% of registered cholangiocarcinomas were confirmed under a microscope ³, and community screening began there only in 2013 ⁴⁸. In western Siberia, where Opisthorchis felineus is endemic, a 2023 case-control study of 40 cases found an odds ratio of 3.9, with a confidence interval from 1.4 to 10.8 ⁵³. Where nobody has looked, the parasite's share of the cancer is not zero; it is unmeasured.

Where it connects

Sources

107 sources, numbered as they are cited. Every one was checked against PubMed or its publisher before it was cited here; the note under each says what it shows and what it does not.

1. 1
   Brindley PJ, Bachini M, Ilyas SI, et al. Cholangiocarcinoma.Nat Rev Dis Primers · 2021 · 7(1):65doi:10.1038/s41572-021-00300-2 · PMID 34504109

   The disease primer: three anatomic subtypes with distinct genetics, presentation and treatment, and most cases without an identifiable cause; its statement that praziquantel lowers fluke-related cancer risk is not backed by a trial with cancer as the outcome.

2. 2
   Banales JM, Marin JJG, Lamarca A, et al. Cholangiocarcinoma 2020: the next horizon in mechanisms and management.Nat Rev Gastroenterol Hepatol · 2020 · 17(9):557-588doi:10.1038/s41575-020-0310-z · PMID 32606456

   Gives the approximate shares (about 15% of primary liver cancers, 3% of gastrointestinal cancers, 2% of cancer deaths) and states that noninvasive diagnosis is not accurate enough; the figures are rounded consensus estimates, not a new measurement.

3. 3
   Kamsa-ard S, Wiangnon S, Suwanrungruang K, et al. Trends in liver cancer incidence between 1985 and 2009, Khon Kaen, Thailand: cholangiocarcinoma.Asian Pac J Cancer Prev · 2011 · 12(9):2209-13PMID 22296358

   The registry rate behind the 'highest in the world' claim, 44.3 per 100,000 men, with only 10.8% of cases confirmed by cytology or histology; its reported decline had a confidence interval that included no change.

4. 4
   Kamsa-Ard S, Santong C, Kamsa-Ard S, et al. Decreasing trends in cholangiocarcinoma incidence and relative survival in Khon Kaen, Thailand: An updated, inclusive, population-based cancer registry analysis for 1989-2018.PLoS One · 2021 · 16(2):e0246490doi:10.1371/journal.pone.0246490 · PMID 33592053

   Shows the Khon Kaen rate falling 3.1% a year in men by birth cohort and gives five-year relative survival of 10.9%; it measures the decline but does not identify its cause.

5. 5
   Javle M, Lee S, Azad NS, et al. Temporal Changes in Cholangiocarcinoma Incidence and Mortality in the United States from 2001 to 2017.Oncologist · 2022 · 27(10):874-883doi:10.1093/oncolo/oyac150 · PMID 35972334

   SEER (the United States national cancer registry program) data showing a 43.8% rise in incidence and a 148.8% rise in intrahepatic disease, with median survival of 8 months; three authors are affiliated with Incyte, which markets pemigatinib.

6. 6
   Clements O, Eliahoo J, Kim JU, Taylor-Robinson SD, Khan SA. Risk factors for intrahepatic and extrahepatic cholangiocarcinoma: A systematic review and meta-analysis.J Hepatol · 2020 · 72(1):95-103doi:10.1016/j.jhep.2019.09.007 · PMID 31536748

   Pools 25 case-control studies: cysts and stones, cirrhosis and hepatitis B and C are the strongest factors, with choledochal cyst odds ratios of 26.71 and 34.94; case-control associations, and the abstract gives numbers only for cysts.

7. 7
   Bergquist A, Ekbom A, Olsson R, et al. Hepatic and extrahepatic malignancies in primary sclerosing cholangitis.J Hepatol · 2002 · 36(3):321-7doi:10.1016/s0168-8278(01)00288-4 · PMID 11867174

   A national Swedish cohort of 604 patients: 13.3% developed hepatobiliary cancer, a 161-fold risk, 37% of them within a year of the liver diagnosis; follow-up ended in 1998.

8. 8
   Huang YL, Zhang KY, Sun YL, Qian MB, Wang Z. The risk of hepatobiliary complications in Clonorchis and Opisthorchis infection: A systematic review and meta-analysis.Acta Trop · 2024 · 260:107457doi:10.1016/j.actatropica.2024.107457 · PMID 39521195

   The largest pooled human estimate, 22 studies and 34,367 participants: odds ratio 4.24 for cholangiocarcinoma; the risk ratio of 10.43 has a very wide interval and the studies are observational.

9. 9
   Buisson Y. [Control of Opisthorchis viverrini infection for cholangiocarcinoma prevention].Bull Soc Pathol Exot · 2017 · 110(1):61-67doi:10.1007/s13149-017-0544-8 · PMID 28105582

   A French-language review giving the IARC years (1994 for O. viverrini, 2009 for C. sinensis), the three proposed mechanisms, and the failure of mass deworming against reinfection; a review, not primary data.

10. 10
    Arai Y, Totoki Y, Hosoda F, et al. Fibroblast growth factor receptor 2 tyrosine kinase fusions define a unique molecular subtype of cholangiocarcinoma.Hepatology · 2014 · 59(4):1427-34doi:10.1002/hep.26890 · PMID 24122810

    FGFR2 fusions in 9 of 66 intrahepatic tumors (13.6%) and in no extrahepatic ones, mutually exclusive with KRAS and BRAF mutations, and transforming in mice; one Tokyo series.

11. 11
    Abou-Alfa GK, Macarulla T, Javle MM, et al. Ivosidenib in IDH1-mutant, chemotherapy-refractory cholangiocarcinoma (ClarIDHy): a multicentre, randomised, double-blind, placebo-controlled, phase 3 study.Lancet Oncol · 2020 · 21(6):796-807doi:10.1016/S1470-2045(20)30157-1 · PMID 32416072

    The only randomized, placebo-controlled phase 3 trial among the targeted drugs: progression-free survival 2.7 against 1.4 months; funded by Agios Pharmaceuticals, whose employees are among the authors.

12. 12
    Jusakul A, Kongpetch S, Teh BT. Genetics of Opisthorchis viverrini-related cholangiocarcinoma.Curr Opin Gastroenterol · 2015 · 31(3):258-63doi:10.1097/MOG.0000000000000162 · PMID 25693006

    Gives the mutation frequencies in fluke-related tumors (TP53 44%, KRAS 16.7%, SMAD4 16.7%, BAP1 and IDH1/2 2.8% each); a review of sequencing studies.

13. 13
    Macarulla T, Neuzillet C, Prager GW, Rimassa L, Bridgewater J. Opportunities and Approaches to Optimising Advanced Cholangiocarcinoma Outcomes in the Era of Targeted Therapies: A Narrative Review.Oncol Ther · 2025 · 13(4):939-962doi:10.1007/s40487-025-00370-2 · PMID 41060610

    Summarizes current recommendations: chemotherapy with durvalumab or pembrolizumab first, molecular testing by next-generation sequencing, and targeted drugs later; a narrative review from an educational meeting, not a guideline.

14. 14
    Casak SJ, Pradhan S, Fashoyin-Aje LA, et al. FDA Approval Summary: Ivosidenib for the Treatment of Patients with Advanced Unresectable or Metastatic, Chemotherapy Refractory Cholangiocarcinoma with an IDH1 Mutation.Clin Cancer Res · 2022 · 28(13):2733-2737doi:10.1158/1078-0432.CCR-21-4462 · PMID 35259259

    The FDA's account of the August 25, 2021 approval, noting that 70.5% of the placebo arm received ivosidenib after progression; it explains the regulator's reasoning, not new efficacy data.

15. 15
    Patel TH, Marcus L, Horiba MN, et al. FDA Approval Summary: Pemigatinib for Previously Treated, Unresectable Locally Advanced or Metastatic Cholangiocarcinoma with FGFR2 Fusion or Other Rearrangement.Clin Cancer Res · 2023 · 29(5):838-842doi:10.1158/1078-0432.CCR-22-2036 · PMID 36206041

    The FDA's account of the April 17, 2020 accelerated approval on a 36% response rate, with a companion diagnostic; accelerated approval is not proof of longer life.

16. 16
    Gandhy SU, Casak SJ, Mushti SL, et al. FDA Approval Summary: Futibatinib for Unresectable Advanced or Metastatic, Chemotherapy Refractory Intrahepatic Cholangiocarcinoma with FGFR2 Fusions or Other Rearrangements.Clin Cancer Res · 2023 · 29(20):4027-4031doi:10.1158/1078-0432.CCR-23-1042 · PMID 37289037

    The FDA's account of the September 30, 2022 accelerated approval on a 42% response rate, with ocular toxicity and high phosphate as key risks.

17. 17
    Yoon J, Oh DY. An evaluation of zanidatamab, a novel, anti-HER2 biparatopic antibody, for the treatment of biliary tract cancer.Expert Opin Biol Ther · 2025 · 25(9):935-946doi:10.1080/14712598.2025.2556903 · PMID 40892073

    Gives the regulatory dates for zanidatamab: FDA accelerated approval November 2024, China May 2025, European Medicines Agency June 2025; an expert review, not a trial.

18. 18
    Wangboon C, Yongvanit P, Loilome W, et al. Elevated Levels of Urinary 8-oxodG Correlate with Persistent Periductal Fibrosis after Praziquantel Treatment in Chronic Opisthorchiasis.Am J Trop Med Hyg · 2018 · 98(6):1763-1769doi:10.4269/ajtmh.17-0971 · PMID 29637887

    One year after praziquantel, 34.6% of infected people with periductal fibrosis still had it, and a urinary DNA-damage marker fell only in those without fibrosis; 66 infected participants, surrogate markers rather than cancer.

19. 19
    Valle JW, Kelley RK, Nervi B, et al. Biliary tract cancer.Lancet · 2021 · 397(10272):428-444doi:10.1016/S0140-6736(21)00153-7 · PMID 33516341

    The Lancet Seminar: low incidence in most high-income countries, a major problem in endemic areas, surgery as the only cure and most patients presenting late; a review, not new data.

20. 20
    Yoo C, Ueno M, Klümpen HJ, et al. Health-related quality of life in participants with advanced biliary tract cancer from the randomized phase III KEYNOTE-966 study.J Hepatol · 2025 · 83(3):692-700doi:10.1016/j.jhep.2025.03.019 · PMID 40154623

    Quality of life was maintained when pembrolizumab was added; cited for the disease-specific symptom domains (pain, jaundice) and the late, nonspecific presentation. Merck employees are among the authors.

21. 21
    Oh DY, He AR, Qin S, et al. Durvalumab plus chemotherapy in advanced biliary tract cancer: 3-year overall survival update from the phase III TOPAZ-1 study.J Hepatol · 2025 · 83(5):1092-1101doi:10.1016/j.jhep.2025.05.003 · PMID 40381735

    At 41.3 months' follow-up, 36-month survival was 14.6% against 6.9%; these updated analyses are exploratory, and AstraZeneca employees are among the authors.

22. 22
    Darwish Murad S, Kim WR, Harnois DM, et al. Efficacy of neoadjuvant chemoradiation, followed by liver transplantation, for perihilar cholangiocarcinoma at 12 US centers.Gastroenterology · 2012 · 143(1):88-98.e3doi:10.1053/j.gastro.2012.04.008 · PMID 22504095

    287 patients at 12 centers: intention-to-treat five-year survival 53% and post-transplant recurrence-free survival 65%; 193 patients came from one center, and patients outside the criteria did worse.

23. 23
    Banales JM, Cardinale V, Carpino G, et al. Expert consensus document: Cholangiocarcinoma: current knowledge and future perspectives consensus statement from the European Network for the Study of Cholangiocarcinoma (ENS-CCA).Nat Rev Gastroenterol Hepatol · 2016 · 13(5):261-80doi:10.1038/nrgastro.2016.51 · PMID 27095655

    The European consensus definition of a heterogeneous group of cancers and the second most common primary liver tumor; its abstract gives no subtype proportions or survival figures.

24. 24
    European Association for the Study of the Liver; International Liver Cancer Association. EASL-ILCA Clinical Practice Guidelines on the management of intrahepatic cholangiocarcinoma.J Hepatol · 2023 · 79(1):181-208doi:10.1016/j.jhep.2023.03.010 · PMID 37084797

    Defines intrahepatic cholangiocarcinoma as arising between the bile ductules and the second-order ducts and treats it as a distinct entity; the recommendations themselves are in the full text, not the abstract.

25. 25
    Klatskin G. Adenocarcinoma of the hepatic duct at its bifurcation within the porta hepatis. An unusual tumor with distinctive clinical and pathological features.Am J Med · 1965 · 38:241-56doi:10.1016/0002-9343(65)90178-6 · PMID 14256720

    The description that gave perihilar tumors their eponym; PubMed holds no abstract, so it is cited here only for its date and subject.

26. 26
    Welzel TM, McGlynn KA, Hsing AW, et al. Impact of classification of hilar cholangiocarcinomas (Klatskin tumors) on the incidence of intra- and extrahepatic cholangiocarcinoma in the United States.J Natl Cancer Inst · 2006 · 98(12):873-5doi:10.1093/jnci/djj234 · PMID 16788161

    Shows that 91% of coded Klatskin tumors in 1992–2000 SEER (United States national cancer registry) data were counted as intrahepatic, overstating intrahepatic incidence by 13%, and that a real rise (4% a year) remained after correction.

27. 27
    Khan SA, Emadossadaty S, Ladep NG, et al. Rising trends in cholangiocarcinoma: is the ICD classification system misleading us?J Hepatol · 2012 · 56(4):848-54doi:10.1016/j.jhep.2011.11.015 · PMID 22173164

    Traces the coding rule that cross-referenced Klatskin tumors to the intrahepatic site and the 2001 switch in United States data; it shows coding can skew site-specific trends, not that the overall rise is false.

28. 28
    Li Z, Huang N, Du Q, et al. Role of immunophenotypic characterisation in prognostic subtyping of intrahepatic cholangiocarcinoma.Pathology · 2023 · 55(7):979-988doi:10.1016/j.pathol.2023.07.008 · PMID 37858435

    Applies the fifth WHO split of intrahepatic tumors into large-duct and small-duct types to 190 resected cases and finds every FGFR2 fusion in the small-duct type; a single-center Chinese series.

29. 29
    Smout MJ, Laha T, Mulvenna J, et al. A granulin-like growth factor secreted by the carcinogenic liver fluke, Opisthorchis viverrini, promotes proliferation of host cells.PLoS Pathog · 2009 · 5(10):e1000611doi:10.1371/journal.ppat.1000611 · PMID 19816559

    Identifies Ov-GRN-1 as the main growth factor in the fluke's secretions, active at nanomolar concentrations; cell culture and hamster tissue only.

30. 30
    Chaiyadet S, Tangkawattana S, Smout MJ, et al. Knockout of liver fluke granulin, Ov-grn-1, impedes malignant transformation during chronic infection with Opisthorchis viverrini.PLoS Pathog · 2022 · 18(9):e1010839doi:10.1371/journal.ppat.1010839 · PMID 36137145

    With a dietary nitrosamine added, granulin-knockout flukes caused less fibrosis, fewer TP53 mutations and fewer high-grade cancers in hamsters; one collaborating group, animal evidence, not yet independently replicated.

31. 31
    Flavell DJ, Lucas SB. Promotion of N-nitrosodimethylamine-initiated bile duct carcinogenesis in the hamster by the human liver fluke, Opisthorchis viverrini.Carcinogenesis · 1983 · 4(7):927-30doi:10.1093/carcin/4.7.927 · PMID 6307539

    With one nitrosamine dose, fluke-infected hamsters developed bile duct cancer at 10% and 20%, and neither agent alone caused it; hamsters, not people.

32. 32
    Thamavit W, Pairojkul C, Tiwawech D, Shirai T, Ito N. Strong promoting effect of Opisthorchis viverrini infection on dimethylnitrosamine-initiated hamster liver.Cancer Lett · 1994 · 78(1-3):121-5doi:10.1016/0304-3835(94)90040-x · PMID 8180954

    Restates the 100% tumor rate as a product of continuous nitrosamine dosing and finds 44% with a single dose; animal evidence for co-carcinogenesis, not for human dose.

33. 33
    Chan-On W, Nairismägi ML, Ong CK, et al. Exome sequencing identifies distinct mutational patterns in liver fluke-related and non-infection-related bile duct cancers.Nat Genet · 2013 · 45(12):1474-8doi:10.1038/ng.2806 · PMID 24185513

    In 209 tumors, BAP1 and IDH1/2 mutations were more frequent without the fluke and TP53 mutations with it; shows that cause shapes the mutations, not that any one mutation is caused by the fluke.

34. 34
    Jusakul A, Cutcutache I, Yong CH, et al. Whole-Genome and Epigenomic Landscapes of Etiologically Distinct Subtypes of Cholangiocarcinoma.Cancer Discov · 2017 · 7(10):1116-1135doi:10.1158/2159-8290.CD-17-0368 · PMID 28667006

    Sorts 489 tumors from 10 countries into four clusters that split by fluke status; gene symbols are stripped from the PubMed abstract, and the ERBB2 and TP53 attribution follows the record's indexing.

35. 35
    Grosche B, Birschwilks M, Wesch H, Kaul A, van Kaick G. The German Thorotrast Cohort Study: a review and how to get access to the data.Radiat Environ Biophys · 2016 · 55(3):281-9doi:10.1007/s00411-016-0651-8 · PMID 27154786

    Describes Thorotrast, introduced in 1929 and used until the 1950s, its lifelong retention, and the 2,326-patient German cohort; the abstract does not give a cholangiocarcinoma-specific risk.

36. 36
    Yamamoto Y, Chikawa J, Uegaki Y, et al. Histological type of Thorotrast-induced liver tumors associated with the translocation of deposited radionuclides.Cancer Sci · 2010 · 101(2):336-40doi:10.1111/j.1349-7006.2009.01401.x · PMID 19917057

    States that Thorotrast induces intrahepatic cholangiocarcinoma and angiosarcoma decades after injection and studies why one or the other appears; autopsy tissue from exposed patients, not a risk estimate.

37. 37
    Meyer CG, Penn I, James L. Liver transplantation for cholangiocarcinoma: results in 207 patients.Transplantation · 2000 · 69(8):1633-7doi:10.1097/00007890-200004270-00019 · PMID 10836374

    A registry of unselected transplants with 51% recurrence and 23% five-year survival, concluding transplantation should seldom be used; the verdict predates neoadjuvant protocols.

38. 38
    Rea DJ, Heimbach JK, Rosen CB, et al. Liver transplantation with neoadjuvant chemoradiation is more effective than resection for hilar cholangiocarcinoma.Ann Surg · 2005 · 242(3):451-8doi:10.1097/01.sla.0000179678.13285.fa · PMID 16135931

    Mayo Clinic: 82% five-year survival after chemoradiation and transplantation against 21% after resection; non-randomized, and only 38 of 71 patients who entered the protocol reached transplant.

39. 39
    Bismuth H, Corlette MB. Intrahepatic cholangioenteric anastomosis in carcinoma of the hilus of the liver.Surg Gynecol Obstet · 1975 · 140(2):170-8PMID 1079096

    Argues that the cholangiogram, read for invasion of the primary and secondary duct confluences, should guide surgery for hilar tumors; the abstract does not itself set out the later staging scheme that carries the authors' names.

40. 40
    Valle J, Wasan H, Palmer DH, et al. Cisplatin plus gemcitabine versus gemcitabine for biliary tract cancer.N Engl J Med · 2010 · 362(14):1273-81doi:10.1056/NEJMoa0908721 · PMID 20375404

    ABC-02, 410 patients: median survival 11.7 against 8.1 months, hazard ratio 0.64; the trial that set the standard on which every later regimen is built.

41. 41
    Oh DY, Ruth He A, Qin S, et al. Durvalumab plus Gemcitabine and Cisplatin in Advanced Biliary Tract Cancer.NEJM Evid · 2022 · 1(8):EVIDoa2200015doi:10.1056/EVIDoa2200015 · PMID 38319896

    TOPAZ-1, 685 patients: overall survival hazard ratio 0.80 with durvalumab; funded by AstraZeneca, with company employees among the authors.

42. 42
    Kelley RK, Ueno M, Yoo C, et al. Pembrolizumab in combination with gemcitabine and cisplatin compared with gemcitabine and cisplatin alone for patients with advanced biliary tract cancer (KEYNOTE-966): a randomised, double-blind, placebo-controlled, phase 3 trial.Lancet · 2023 · 401(10391):1853-1865doi:10.1016/S0140-6736(23)00727-4 · PMID 37075781

    KEYNOTE-966, 1,069 patients: median survival 12.7 against 10.9 months, hazard ratio 0.83; funded by Merck Sharp & Dohme, with company employees among the authors.

43. 43
    Abou-Alfa GK, Borbath I, Roychowdhury S, et al. Phase III trial of infigratinib versus gemcitabine/cisplatin in adults with advanced cholangiocarcinoma with FGFR2 gene fusion or rearrangement: results and reflections on early termination of PROOF 301.ESMO Open · 2026 · 11(4):106306doi:10.1016/j.esmoop.2026.106306 · PMID 41850040

    The confirmatory trial for infigratinib closed for poor accrual after 48 of about 300 planned patients, so it answers nothing definitively; industry-employed authors from the sponsors.

44. 44
    Harding JJ, Fan J, Oh DY, et al. Zanidatamab for HER2-amplified, unresectable, locally advanced or metastatic biliary tract cancer (HERIZON-BTC-01): a multicentre, single-arm, phase 2b study.Lancet Oncol · 2023 · 24(7):772-782doi:10.1016/S1470-2045(23)00242-5 · PMID 37276871

    Confirmed response in 41.3% of 80 HER2-positive patients; single-arm, funded by Zymeworks, Jazz and BeiGene, the companies developing the drug.

45. 45
    Shroff RT, Javle MM, Xiao L, et al. Gemcitabine, Cisplatin, and nab-Paclitaxel for the Treatment of Advanced Biliary Tract Cancers: A Phase 2 Clinical Trial.JAMA Oncol · 2019 · 5(6):824-830doi:10.1001/jamaoncol.2019.0270 · PMID 30998813

    A 60-patient single-arm trial with median survival of 19.2 months against historical controls; the phase 3 trial that followed did not confirm it.

46. 46
    Shroff RT, King G, Colby S, et al. SWOG S1815: A Phase III Randomized Trial of Gemcitabine, Cisplatin, and Nab-Paclitaxel Versus Gemcitabine and Cisplatin in Newly Diagnosed, Advanced Biliary Tract Cancers.J Clin Oncol · 2025 · 43(5):536-544doi:10.1200/JCO-24-01383 · PMID 39671534

    Randomized, 441 analyzable patients: 14.0 against 13.6 months, hazard ratio 0.91, with more toxicity; the clearest recent example of a single-arm signal that did not survive randomization.

47. 47
    Castet F, Fabregat-Franco C, Bridgewater J, et al. Association of candidate surrogate endpoints with overall survival in advanced biliary tract cancer.J Hepatol · 2025 · 83(5):1102-1115doi:10.1016/j.jhep.2025.05.020 · PMID 40473034

    Across 41 trials, response rate explained almost none of the variation in overall survival (R² 0.01) while progression-free survival correlated moderately; it questions the measure, not any single drug.

48. 48
    Thanasukarn V, Srisuk T, Luvira V, et al. Improving postoperative survival in cholangiocarcinoma: development of surgical strategies with a screening program in the epidemic region.World J Surg Oncol · 2024 · 22(1):287doi:10.1186/s12957-024-03573-5 · PMID 39478620

    In 1,091 resections, median survival rose from 14 to 40 months between periods, but screened against usual presentation was 51 against 38 months, p = 0.06; the program's surgeons call its clinical utility unclear.

49. 49
    Khuntikeo N, Chamadol N, Yongvanit P, et al. Cohort profile: cholangiocarcinoma screening and care program (CASCAP).BMC Cancer · 2015 · 15:459doi:10.1186/s12885-015-1475-7 · PMID 26054405

    Describes the first community-level screening program and its early enrollees; the 135.4 per 100,000 figure it quotes for Khon Kaen men is given without period or method.

50. 50
    Primrose JN, Fox RP, Palmer DH, et al. Capecitabine compared with observation in resected biliary tract cancer (BILCAP): a randomised, controlled, multicentre, phase 3 study.Lancet Oncol · 2019 · 20(5):663-673doi:10.1016/S1470-2045(18)30915-X · PMID 30922733

    Median survival 51.1 against 36.4 months after resection, but the primary intention-to-treat analysis was not significant (p = 0.097); also the source for 20% resectability and under 10% five-year survival overall.

51. 51
    Arunsan P, Ittiprasert W, Smout MJ, et al. Programmed knockout mutation of liver fluke granulin attenuates virulence of infection-induced hepatobiliary morbidity.eLife · 2019 · 8:e41463doi:10.7554/eLife.41463 · PMID 30644359

    The first programmed gene editing in a parasitic flatworm: the fluke's own granulin gene was disabled and infected hamsters had less duct thickening and scarring; no carcinogen was given, so cancer was not measured.

52. 52
    Nakachi K, Ikeda M, Konishi M, et al. Adjuvant S-1 compared with observation in resected biliary tract cancer (JCOG1202, ASCOT): a multicentre, open-label, randomised, controlled, phase 3 trial.Lancet · 2023 · 401(10372):195-203doi:10.1016/S0140-6736(22)02038-4 · PMID 36681415

    Open-label, 440 patients at 38 Japanese hospitals: three-year overall survival 77.1% against 67.6%, adjusted hazard ratio 0.69; relapse-free survival was not significantly different, and the authors frame S-1 as a standard for Asian patients.

53. 53
    Fedorova OS, Kovshirina AE, Kovshirina YV, et al. Opisthorchis Felineus Infection is a Risk Factor for Cholangiocarcinoma in Western Siberia: A Hospital-based Case-control Study.Clin Infect Dis · 2023 · 76(3):e1392-e1398doi:10.1093/cid/ciac497 · PMID 35723279

    Links O. felineus to cholangiocarcinoma in 40 cases and 160 controls, odds ratio 3.9 (95% CI 1.4–10.8); small, hospital-based and partly reliant on self-report.

54. 54
    Oh DY, He AR, Qin S, et al. Durvalumab Plus Chemotherapy for Advanced Biliary Tract Cancer: A Post Hoc Analysis of the TOPAZ-1 Randomized Clinical Trial.JAMA Oncol · 2026 · 12(9):960-964doi:10.1001/jamaoncol.2026.2204 · PMID 42424063

    Four-year follow-up of TOPAZ-1, data cutoff February 28, 2025: median survival 13.0 against 11.4 months, hazard ratio 0.75, and 48-month survival 11.8% against 4.3%; a post hoc analysis, with AstraZeneca employees among the authors.

55. 55
    Zhao Z, Wu H, Han J, Jiang K. Global trends and disparities in gallbladder and biliary tract cancers: insights from the global burden of disease study 2021.Eur J Gastroenterol Hepatol · 2025 · 37(5):573-584doi:10.1097/MEG.0000000000002947 · PMID 39975993

    Modeled estimates that lump gallbladder cancer with bile duct cancer: 216,768 new cases and 171,961 deaths in 2021, age-standardized incidence 2.6 and mortality 2.0 per 100,000; useful for scale, not for cholangiocarcinoma alone.

56. 56
    Florio AA, Ferlay J, Znaor A, et al. Global trends in intrahepatic and extrahepatic cholangiocarcinoma incidence from 1993 to 2012.Cancer · 2020 · 126(11):2666-2678doi:10.1002/cncr.32803 · PMID 32129902

    The best international comparison, from Cancer Incidence in Five Continents Plus: highest rates in South Korea, Thailand and Japan, and rising incidence in most countries; national averages hide regional hot spots such as Khon Kaen.

57. 57
    Sahat O, Bilheem S, Lim A, et al. Updated cholangiocarcinoma incidence trends and projections in Thailand by region based on data from four population-based cancer registries.Lancet Reg Health Southeast Asia · 2025 · 35:100569doi:10.1016/j.lansea.2025.100569 · PMID 40230445

    The first nationwide analysis, four registries and 6,379 cases for 2012-2021: age-standardized rates of 12.5 per 100,000 men and 5.9 women nationally, 19.2 and 8.5 in the northeast, falling 7.20% a year in men and 5.81% in women; the central and southern trends could not be estimated for lack of cases, and only about 22% of cases were microscopically verified.

58. 58
    Ghiringhelli F, Jooste V, Manfredi S, et al. Biliary tract cancers have distinct epidemiological patterns and clinical characteristics according to tumour site.HPB (Oxford) · 2023 · 25(6):693-703doi:10.1016/j.hpb.2023.02.016 · PMID 36958986

    A French population registry: intrahepatic cancer was 40% of 714 biliary tract cancers, half the patients were over 75, and only 15% of non-metastatic intrahepatic cases had a complete resection; one region of one country.

59. 59
    Bloomfield GC, Nigam A, Calvo IG, et al. Characteristics and malignancy rates of adult patients diagnosed with choledochal cyst in the West: a systematic review.J Gastrointest Surg · 2024 · 28(1):77-87doi:10.1016/j.gassur.2023.11.007 · PMID 38353080

    In 1,337 Western adults with choledochal cysts, 10.9% had a malignancy, most often cholangiocarcinoma, and resection reduced but did not abolish risk; pooled case series from surgical centers, which concentrate sicker patients.

60. 60
    Cai H, Kong WT, Chen CB, et al. Cholelithiasis and the risk of intrahepatic cholangiocarcinoma: a meta-analysis of observational studies.BMC Cancer · 2015 · 15:831doi:10.1186/s12885-015-1870-0 · PMID 26526500

    Seven case-control studies, 123,771 participants and 4,763 cancers: bile duct stones odds ratio 17.64, choledocholithiasis alone 11.79, gallstones 2.00 with high heterogeneity; association, not cause.

61. 61
    Souza M, Lima LCV, Al-Sharif L, Huang DQ. Incidence of Hepatobiliary Malignancies in Primary Sclerosing Cholangitis: Systematic Review and Meta-analysis.Clin Gastroenterol Hepatol · 2025 · 23(10):1695-1709.e13doi:10.1016/j.cgh.2024.09.037 · PMID 39709139

    Pools 51 cohorts and 26,482 patients for a cholangiocarcinoma incidence of 9.31 per 1,000 person-years, with higher rates in smaller studies, which suggests referral bias inflates some estimates.

62. 62
    Barner-Rasmussen N, Molinaro A, Mol B, et al. Surveillance of primary sclerosing cholangitis - a comparison of scheduled or on-demand ERCP with annual MRI surveillance: a multicenter study.Endoscopy · 2025 · 57(5):431-440doi:10.1055/a-2511-3422 · PMID 39875118

    Compares three surveillance strategies in 1,629 patients and notes there is no good tumor marker for cholangiocarcinoma in this disease; observational cohorts from different countries, not a randomized comparison.

63. 63
    Duan S, Wei Y, Ding Z, et al. Occurrence and prognosis of metabolic dysfunction-associated steatosis liver disease and gastrointestinal tumors: a systematic review and meta-analysis.PeerJ · 2026 · 14:e20616doi:10.7717/peerj.20616 · PMID 41769402

    Pools 29 observational studies and finds MASLD associated with cholangiocarcinoma, hazard ratio 1.26; heterogeneity was substantial and the design cannot show cause.

64. 64
    Corrao S, Natoli G, Argano C. Nonalcoholic fatty liver disease is associated with intrahepatic cholangiocarcinoma and not with extrahepatic form: definitive evidence from meta-analysis and trial sequential analysis.Eur J Gastroenterol Hepatol · 2021 · 33(1):62-68doi:10.1097/MEG.0000000000001684 · PMID 32091438

    Odds ratio 2.19 for intrahepatic cancer and 1.48, not significant, for extrahepatic; case-control pooling under the older fatty-liver definition, so it is not directly comparable with the later cohort pooling.

65. 65
    Steele JA, Richter CH, Echaubard P, et al. Thinking beyond Opisthorchis viverrini for risk of cholangiocarcinoma in the lower Mekong region: a systematic review and meta-analysis.Infect Dis Poverty · 2018 · 7(1):44doi:10.1186/s40249-018-0434-3 · PMID 29769113

    Finds alcohol with smoking (odds ratio 11.1) more strongly associated than fluke exposure, and praziquantel treatment associated with higher risk; 18 heterogeneous studies, and the praziquantel signal is probably confounded by who gets treated.

66. 66
    Hemminki K, Tichanek F, Försti A, et al. Long-term incidence in hepatocellular carcinoma and intrahepatic bile duct cancer in Denmark, Finland, Norway and Sweden, role of Thorotrast?Int J Cancer · 2022 · 151(4):510-517doi:10.1002/ijc.34031 · PMID 35429352

    Cites a 100-fold liver cancer risk from Thorotrast and links a 1980 incidence peak in Denmark and Sweden to it; the authors call their own ecological finding hypothesis-generating.

67. 67
    Nakamura H, Arai Y, Totoki Y, et al. Genomic spectra of biliary tract cancer.Nat Genet · 2015 · 47(9):1003-10doi:10.1038/ng.3375 · PMID 26258846

    In 260 biliary tract cancers, nearly 40% carried a targetable alteration and FGFR2 fusions clustered in intrahepatic tumors; a Japanese cohort.

68. 68
    Jiao Y, Pawlik TM, Anders RA, et al. Exome sequencing identifies frequent inactivating mutations in BAP1, ARID1A and PBRM1 in intrahepatic cholangiocarcinomas.Nat Genet · 2013 · 45(12):1470-1473doi:10.1038/ng.2813 · PMID 24185509

    Almost half of 32 intrahepatic tumors carried an inactivating mutation in BAP1, ARID1A or PBRM1; a small discovery set.

69. 69
    Zhu AX, Macarulla T, Javle MM, et al. Final Overall Survival Efficacy Results of Ivosidenib for Patients With Advanced Cholangiocarcinoma With IDH1 Mutation: The Phase 3 Randomized Clinical ClarIDHy Trial.JAMA Oncol · 2021 · 7(11):1669-1677doi:10.1001/jamaoncol.2021.3836 · PMID 34554208

    Final overall survival 10.3 against 7.5 months, hazard ratio 0.79, not statistically significant; the crossover-adjusted estimate is favorable but model-dependent. Several authors were Agios employees.

70. 70
    Goyal L, DiToro D, Facchinetti F, et al. A model for decoding resistance in precision oncology: acquired resistance to FGFR inhibitors in cholangiocarcinoma.Ann Oncol · 2025 · 36(4):426-443doi:10.1016/j.annonc.2024.12.011 · PMID 39706336

    Seventy-seven patients followed through treatment with 486 samples: new FGFR2 kinase-domain mutations in 65% of those who had benefited against 10% of those who had not, and 26 distinct mutations; it explains why response to these drugs is temporary.

71. 71
    Javle M, Fountzilas C, Liao CY, et al. Tinengotinib for adults with advanced or metastatic cholangiocarcinoma: a multicentre, open-label, phase 2 trial.Lancet Gastroenterol Hepatol · 2026 · 11(2):137-149doi:10.1016/S2468-1253(25)00230-4 · PMID 41349554

    Fifty-five previously treated patients in four cohorts by FGFR status: response in 30.0% of those with acquired resistance to an earlier FGFR drug, 23.1% with other FGFR alterations, 6.3% with primary resistance and none without an FGFR alteration; single-arm, response rate as the endpoint, sponsor-funded.

72. 72
    Jeong H, Oh JH, Ahn HS, et al. Proteogenomic profiling predicts outcomes of adjuvant chemotherapy in extrahepatic cholangiocarcinoma.J Hepatol · 2025 · 84(1):122-134doi:10.1016/j.jhep.2025.07.031 · PMID 40803577

    Prespecified exploratory analysis of 89 resected tumors within the Korean STAMP trial: TP53 mutated in 63%, SMAD4 in 20% and KRAS in 18%, with an actionable alteration in 15%; exploratory and awaiting validation.

73. 73
    Li J, Deng T, Gu Y, et al. Efficacy and safety of glecirasib in solid tumors with KRAS G12C mutation: A pooled analysis of two phase I/II trials.Cancer Commun (Lond) · 2025 · 45(11):1500-1512doi:10.1002/cac2.70056 · PMID 41037823

    Fifty-four patients pooled from two early-phase trials, eight of them biliary: five of the seven evaluable biliary cancers responded; single-arm, no comparison group, with sponsor employees among the authors.

74. 74
    Angerilli V, Gasparello J, Niero M, et al. HER2 status in extrahepatic cholangiocarcinoma and gallbladder carcinoma: Concordance between immunohistochemistry and chromogenic in situ hybridization in 140 Cases.Hum Pathol · 2026 · 106234doi:10.1016/j.humpath.2026.106234 · PMID 42595189

    HER2 positivity in 9.2% of extrahepatic cholangiocarcinomas in one Italian series, with marked heterogeneity within tumors; it does not estimate frequency in intrahepatic disease.

75. 75
    Meric-Bernstam F, Makker V, Oaknin A, et al. Efficacy and Safety of Trastuzumab Deruxtecan in Patients With HER2-Expressing Solid Tumors: Primary Results From the DESTINY-PanTumor02 Phase II Trial.J Clin Oncol · 2024 · 42(1):47-58doi:10.1200/JCO.23.02005 · PMID 37870536

    A single-arm basket trial of 267 patients across seven tumor types including a biliary cohort: response in 37.1% overall and 61.3% of centrally confirmed IHC 3+ tumors, with drug-related lung inflammation in 10.5% and three deaths from it; the biliary cohort's own figures are in the full text, and AstraZeneca employees are among the authors.

76. 76
    Xiao Y, Zhou C, Ni X, et al. Preoperative subcategorization based on magnetic resonance imaging in intrahepatic cholangiocarcinoma.Cancer Imaging · 2023 · 23(1):15doi:10.1186/s40644-023-00533-2 · PMID 36782276

    MRI features predicted the large-duct subtype with an area under the curve of 0.91 in 93 patients; retrospective, single center, and about subtyping rather than first detection.

77. 77
    Kim MS, Jeon TJ, Park JY, et al. Clinical Interpretation of Elevated CA 19-9 Levels in Obstructive Jaundice Following Benign and Malignant Pancreatobiliary Disease.Korean J Gastroenterol · 2017 · 70(2):96-102doi:10.4166/kjg.2017.70.2.96 · PMID 28830135

    Of 114 patients drained for obstructive jaundice, 59% of those with benign disease had a raised CA 19-9 before drainage but only 12% afterwards, against 90% and 63% with cancer; a single-center retrospective series, and 12% of the benign group turned out to have cancer.

78. 78
    Liang B, Zhong L, He Q, et al. Diagnostic Accuracy of Serum CA19-9 in Patients with Cholangiocarcinoma: A Systematic Review and Meta-Analysis.Med Sci Monit · 2015 · 21:3555-63doi:10.12659/msm.895040 · PMID 26576628

    Pooled sensitivity 72% and specificity 84% across 31 studies, lower in European patients; mostly case-control designs, which flatter a test compared with use in undiagnosed patients.

79. 79
    Parra-Robert M, Santos VM, Canis SM, et al. Relationship Between CA 19.9 and the Lewis Phenotype: Options to Improve Diagnostic Efficiency.Anticancer Res · 2018 · 38(10):5883-5888doi:10.21873/anticanres.12931 · PMID 30275214

    About 10% of people are Lewis-negative and cannot make CA 19-9, so an undetectable result in them means nothing; a single-center laboratory study.

80. 80
    Lamarca A, Palmer DH, Wasan HS, et al. Prognostic markers, quality of life (QoL) and value of health (V-He) in advanced biliary cancers (ABC) treated with second-line active symptom control (ASC) alone or ASC with oxaliplatin-5-FU chemotherapy (ASC + FOLFOX) in the randomised phase III, multicentre, open-label ABC-06 clinical trial.ESMO Open · 2026 · 11(7):107777doi:10.1016/j.esmoop.2026.107777 · PMID 42349245

    A post hoc analysis of ABC-06: a high baseline CA 19-9 went with survival of 4.4 against 6.4 months in the chemotherapy arm, hazard ratio 1.97, in 135 of 162 patients with a baseline value; prognostic, not diagnostic.

81. 81
    Talpur AS, Urooj K, Talat F, et al. Comparative Diagnostic Accuracy of Cholangioscopy-Based Modalities for Indeterminate Biliary Strictures: A Systematic Review and Network Meta-Analysis.Dig Dis Sci · 2026 · 71(11):5480-5495doi:10.1007/s10620-026-10106-5 · PMID 42435259

    Across 38 studies, brush cytology had 45% sensitivity and 97% specificity, and cholangioscopy 82% and 89% with more complications; the rankings depend on pooled studies of varying quality.

82. 82
    Burnett AS, Bailey J, Oliver JB, Ahlawat S, Chokshi RJ. Sensitivity of alternative testing for pancreaticobiliary cancer: a 10-y review of the literature.J Surg Res · 2014 · 190(2):535-47doi:10.1016/j.jss.2014.04.014 · PMID 24969546

    Pooled sensitivities: ERCP cytology 41.6%, FISH 54.2%, serum CA 19-9 69.3%; it pools pancreatic and biliary cancers together.

83. 83
    Njei B, McCarty TR, Varadarajulu S, Navaneethan U. Systematic review with meta-analysis: endoscopic retrograde cholangiopancreatography-based modalities for the diagnosis of cholangiocarcinoma in primary sclerosing cholangitis.Aliment Pharmacol Ther · 2016 · 44(11-12):1139-1151doi:10.1111/apt.13817 · PMID 27696456

    In primary sclerosing cholangitis, cholangioscopy with targeted biopsy had pooled sensitivity of 65% and specificity of 97%; only four studies of cholangioscopy.

84. 84
    Chen X, Wei X, Yue L, et al. Efficacy and safety of preoperative biliary drainage in patients with Hilar Cholangiocarcinoma: a systematic review and meta-analysis.Int J Surg · 2025 · 111(5):3543-3560doi:10.1097/JS9.0000000000002324 · PMID 40072352

    Twenty-one studies and 3,059 patients to December 2024: drainage before surgery cut liver failure, odds ratio 0.38, but raised postoperative infection, cholangitis and long-term mortality, so routine drainage is not recommended; the pooled studies are retrospective.

85. 85
    Guo X, Xing J, Luo H, et al. The selection and strategy of stents in endoscopic drainage of unresectable perihilar cholangiocarcinoma: a meta-analysis.BMC Surg · 2026 · 26(1)doi:10.1186/s12893-026-03663-z · PMID 41947135

    Twelve studies and 1,405 patients: metal stents stayed open about twice as long as plastic ones with less cholangitis and fewer repeat procedures, but survival was the same; guidelines still lack consensus and the pooled studies are mostly cohorts.

86. 86
    Shi GM, Huang XY, Liang F, et al. Neoadjuvant GOLP in Resectable High-Risk Intrahepatic Cholangiocarcinoma.N Engl J Med · 2026 · 394(10):983-995doi:10.1056/NEJMoa2513918 · PMID 41780001

    Phase 2-3, 178 patients, interim analysis at a median 16.9 months: event-free survival 18.0 against 8.7 months, and 24-month survival 79% against 61%, a hazard ratio of 0.43 that missed the prespecified alpha of 0.0019; adverse events in 97% of the treated group.

87. 87
    Xiao H, Ji J, Li S, et al. Adjuvant Chemoradiation and Immunotherapy for Extrahepatic Cholangiocarcinoma and Gallbladder Cancer: A Randomized Clinical Trial.JAMA Oncol · 2025 · 11(9):1021-1029doi:10.1001/jamaoncol.2025.1926 · PMID 40638104

    ACCORD, 93 patients: camrelizumab with capecitabine chemoradiation against observation gave three-year survival of 58.2% against 30.5%, hazard ratio 0.43; small, single-country, and compared with no treatment rather than with chemotherapy, which its own authors call the next trial to do.

88. 88
    Kimura J, Asassfeh A, Tomosugi T, et al. Comparison of long-term outcomes between liver transplantation and liver resection for intrahepatic cholangiocarcinoma: An updated systematic review and meta-analysis.Surg Oncol · 2026 · 68:102504doi:10.1016/j.suronc.2026.102504 · PMID 42468108

    Seven retrospective studies, 346 transplanted against 5,132 resected: five-year survival favored transplantation with an odds ratio of 0.59, but leaving out one study removed the significance, and selection bias is unavoidable.

89. 89
    Das S. Comparison of Clinical Trial Results of the Recently Approved Immunotherapeutic Drugs for Advanced Biliary Tract Cancers.Rev Recent Clin Trials · 2024 · 19(2):81-90doi:10.2174/0115748871276666240123043710 · PMID 38288802

    A single-author perspective that records pembrolizumab and durvalumab as approved for advanced biliary tract cancer and compares the two trials; commentary, not data.

90. 90
    Casak SJ, Kumar V, Song C, et al. FDA Approval Summary: Durvalumab and Pembrolizumab, Immune Checkpoint Inhibitors for the Treatment of Biliary Tract Cancer.Clin Cancer Res · 2024 · 30(16):3371-3377doi:10.1158/1078-0432.CCR-24-0517 · PMID 38856639

    The FDA's own account of the two approvals: durvalumab with cisplatin and gemcitabine on September 2, 2022 and pembrolizumab on October 31, 2023, both on overall survival in a randomized placebo-controlled trial; it explains the regulator's reasoning, not new efficacy data.

91. 91
    Jeong H, Yoo C, Kim I, et al. Gemcitabine and Cisplatin Plus Polymeric Micellar Paclitaxel and Survival in Advanced Biliary Tract Cancer: A Randomized Clinical Trial.JAMA Netw Open · 2025 · 8(10):e2534560doi:10.1001/jamanetworkopen.2025.34560 · PMID 41032300

    Open-label, 150 patients at seven Korean centers, terminated early for slow accrual: median survival 12.0 against 11.1 months, hazard ratio 0.94, with no difference in progression or response.

92. 92
    Knox JJ, McNamara MG, Bazin IS, et al. A phase III randomized study of first-line NUC-1031/cisplatin vs. gemcitabine/cisplatin in advanced biliary tract cancer.J Hepatol · 2025 · 83(2):358-366doi:10.1016/j.jhep.2025.01.040 · PMID 39978598

    NuTide:121, 773 patients, stopped at the first interim analysis for futility: a modified gemcitabine gave a higher response rate, 18.7% against 12.4%, but shorter survival, 9.2 against 12.6 months; a clean example of response not predicting survival.

93. 93
    Lamarca A, Palmer DH, Wasan HS, et al. Second-line FOLFOX chemotherapy versus active symptom control for advanced biliary tract cancer (ABC-06): a phase 3, open-label, randomised, controlled trial.Lancet Oncol · 2021 · 22(5):690-701doi:10.1016/S1470-2045(21)00027-9 · PMID 33798493

    Second-line FOLFOX extended median survival from 5.3 to 6.2 months and doubled 12-month survival, at the cost of more grade 3–5 adverse events; open-label, 162 patients.

94. 94
    Yoo C, Saborowski A, Hyung J, et al. Liposomal irinotecan for previously treated patients with biliary tract cancer: A pooled analysis of NIFTY and NALIRICC trials.J Hepatol · 2025 · 83(4):909-916doi:10.1016/j.jhep.2025.03.013 · PMID 40147791

    Individual-patient pooling of two randomized trials that disagreed, 278 patients: progression-free survival hazard ratio 0.65 but overall survival 0.77 with p = 0.051; toxicity differed between Korea and Germany, and 31% of the treated Germans stopped without progression.

95. 95
    Abou-Alfa GK, Sahai V, Hollebecque A, et al. Pemigatinib for previously treated, locally advanced or metastatic cholangiocarcinoma: a multicentre, open-label, phase 2 study.Lancet Oncol · 2020 · 21(5):671-684doi:10.1016/S1470-2045(20)30109-1 · PMID 32203698

    FIGHT-202: response in 35.5% of 107 patients with FGFR2 fusions; single-arm, so it cannot show longer survival; funded by Incyte, which markets the drug.

96. 96
    Bekaii-Saab TS, Melisi D, Wilmink H, et al. Pemigatinib for Unresectable or Metastatic Cholangiocarcinoma With Fibroblast Growth Factor Receptor-2 Rearrangement: Results From the Phase III FIGHT-302 Trial.J Clin Oncol · 2026 · 44(23):2187-2197doi:10.1200/JCO-26-00788 · PMID 42223137

    The first randomized first-line test of an FGFR2 drug, 167 patients, closed early when the standard of care changed: progression-free survival 8.3 against 6.8 months and response 47% against 15%, but overall survival 24.4 against 25.0 months with crossover allowed; funded by Incyte, whose employees are among the authors.

97. 97
    Goyal L, Meric-Bernstam F, Hollebecque A, et al. Futibatinib for FGFR2-Rearranged Intrahepatic Cholangiocarcinoma.N Engl J Med · 2023 · 388(3):228-239doi:10.1056/NEJMoa2206834 · PMID 36652354

    FOENIX-CCA2: response in 42% of 103 patients and median overall survival of 21.7 months without a control arm; funded by Taiho Oncology and Taiho Pharmaceutical. The gene symbol is stripped from the PubMed title and restored here.

98. 98
    Subbiah V, Lassen U, Élez E, et al. Dabrafenib plus trametinib in patients with BRAF-mutated biliary tract cancer (ROAR): a phase 2, open-label, single-arm, multicentre basket trial.Lancet Oncol · 2020 · 21(9):1234-1243doi:10.1016/S1470-2045(20)30321-1 · PMID 32818466

    Forty-three previously treated patients with a BRAF mutation: response in 22 (51%) by investigator and 20 (47%) by independent review; single-arm, an interim analysis, funded by GlaxoSmithKline and Novartis.

99. 99
    Chen LT, Vogel A, Hsu C, et al. Pan-Asian adapted ESMO Clinical Practice Guidelines for the diagnosis, treatment and follow-up of patients with biliary tract cancer.ESMO Open · 2024 · 9(8):103647doi:10.1016/j.esmoop.2024.103647 · PMID 39232586

    Adapts the European guideline for Asia and records that drug approvals and reimbursement differ between Asian countries; the detailed recommendations are in the full text.

100. 100
     Chamadol N, Khuntikeo N, Thinkhamrop B, et al. Association between periductal fibrosis and bile duct dilatation among a population at high risk of cholangiocarcinoma: a cross-sectional study of cholangiocarcinoma screening in Northeast Thailand.BMJ Open · 2019 · 9(3):e023217doi:10.1136/bmjopen-2018-023217 · PMID 30898798

     Gives CASCAP's enrollment (394,026, residents aged 40 or over) and links the most extensive periductal fibrosis to duct dilatation (adjusted odds ratio 5.74); cross-sectional, so it cannot show progression to cancer.

101. 101
     Kopolrat KY, Boueroy P, Kammoolkon R, et al. Longitudinal changes and risk factors of Opisthorchis viverrini infection after selective praziquantel treatment: evidence from urine antigen assay and fecal examination in an endemic community in Northeast Thailand.PLoS One · 2026 · 21(7):e0352854doi:10.1371/journal.pone.0352854 · PMID 42406772

     A prospective 24-week cohort of 612 villagers: baseline prevalence 41.0% by urine antigen against 8.1% by stool microscopy, reinfection 63.7 per 100 person-years against 5.9 by stool, and raw fish within six months carried an adjusted risk ratio of 7.52.

102. 102
     Chamadol N, Laopaiboon V, Jareanrat A, et al. Improvement of survival outcomes of cholangiocarcinoma by ultrasonography surveillance: Multicenter retrospective cohorts.Heliyon · 2024 · 10(18):e38191doi:10.1016/j.heliyon.2024.e38191 · PMID 39381227

     Five-year survival 53.9% for screen-detected against 21.9% for walk-in patients; retrospective and open to lead-time and selection bias, so it does not prove screening saves lives.

103. 103
     Bowlus CL, Arrivé L, Bergquist A, et al. AASLD practice guidance on primary sclerosing cholangitis and cholangiocarcinoma.Hepatology · 2023 · 77(2):659-702doi:10.1002/hep.32771 · PMID 36083140

     The American reference guidance for surveillance in primary sclerosing cholangitis, which advises yearly imaging of the ducts by MRI with or without CA 19-9; PubMed holds no abstract, so the recommendation wording comes from the full text, and the guidance itself rests largely on expert opinion.

104. 104
     Heneberg P. From infection to cholangiocarcinoma: Why opisthorchiids break these rules.One Health · 2026 · 23:101517doi:10.1016/j.onehlt.2026.101517 · PMID 42500360

     Argues that chronic helminth inflammation causes cancer only when several conditions coincide, and that opisthorchiids are a rare exception; a conceptual review, not new data.

105. 105
     Lishai EA, Zaparina OG, Kapushchak YK, et al. Comparative liver transcriptome analysis in hamsters infected with food-borne trematodes Opisthorchis felineus, Opisthorchis viverrini, or Clonorchis sinensis.PLoS Negl Trop Dis · 2024 · 18(12):e0012685doi:10.1371/journal.pntd.0012685 · PMID 39652576

     States the IARC positions, O. viverrini and C. sinensis in Group 1 and O. felineus in Group 3, and finds species-specific liver responses in hamsters; animal data, cited here for the classification.

106. 106
     Rujirakul R, Ueng-arporn N, Kaewpitoon SJ, et al. Risk Areas of Liver Flukes in Surin Province of Thailand using Geographic Information System.J Med Assoc Thai · 2015 · 98 Suppl 4:S22-6PMID 26201130

     Retracted. PubMed flags this risk-mapping study as a Retracted Publication; it is listed only so a reader who meets its maps knows not to rely on them, and nothing on this page uses it.

107. 107
     Kim EM, Hong ST. Clonorchis sinensis and Cholangiocarcinoma.J Korean Med Sci · 2025 · 40(16):e145doi:10.3346/jkms.2025.40.e145 · PMID 40296827

     A Korean review: about 10% of Korean cholangiocarcinoma is attributed solely to clonorchiasis, odds ratio 4.7, and it repeats the hamster finding that neither the fluke nor the nitrosamine alone produced cancer; a review, not new data. The species name is stripped from the PubMed title and restored here.