Cholangiocarcinoma found earlier in patients already under liver follow-up

Design

retrospective international multicentre registry analysis with propensity score matching

Population

3,743 patients with cholangiocarcinoma diagnosed 2010–2024; 993 with chronic liver disease

Primary outcome

stage at diagnosis, curative-intent surgery and overall survival

Effect

localised disease 57% vs 43%; surgery 60% vs 48%; median OS 12.2 vs 11.1 months (HR 0.88, 95% CI 0.80–0.98)

This registry analysis covered 3,743 patients diagnosed with cholangiocarcinoma across international centres between 2010 and 2024, of whom 993 had documented chronic liver disease — primary sclerosing cholangitis, cirrhosis, viral hepatitis or another chronic disorder.

The patients with chronic liver disease were more often male (67% vs 53%) and younger (median 63 vs 66 years). They more often had intrahepatic tumours (64% vs 42%), better performance status (ECOG 0 in 53% vs 35%), lower CA19-9 (median 56 vs 135 U/mL) and earlier-stage disease (localised 57% vs 43%; metastatic 23% vs 31%). The stage difference held in propensity-score-matched analysis. Curative-intent surgery followed more often (60% vs 48%), and median overall survival was 12.2 against 11.1 months (hazard ratio 0.88, 95% CI 0.80 to 0.98), with five-year survival odds of 1.70 (1.37 to 2.11). The gap was larger in intrahepatic disease: 14.2 against 11.1 months (HR 0.77, 0.68 to 0.87), five-year survival odds 2.19 (1.60 to 3.01). Response to each treatment modality was comparable.

The mechanism is almost certainly detection rather than biology — these patients were already having imaging for something else. That makes the finding an argument about surveillance rather than about tumour behaviour, and it carries an obvious bias: earlier detection in a screened group lengthens measured survival even without changing the date of death. The higher rate of curative resection is the part that is harder to explain away.

• Maintain structured imaging surveillance in primary sclerosing cholangitis and cirrhosis
• Consider cholangiocarcinoma, not only hepatocellular carcinoma, when a surveillance scan is abnormal
• Do not use a normal CA19-9 to exclude it — median was 56 U/mL in the chronic liver disease group
• Note the survival difference is small in absolute terms: 12.2 against 11.1 months overall

Why it matters

The survival advantage may be an artefact of being watched, which is still an argument for watching.

Don't overread it

Retrospective and subject to lead-time bias — earlier diagnosis in a surveilled group lengthens measured survival by itself.

The statistics, in plain English

A hazard ratio of 0.88 with an interval of 0.80 to 0.98 barely excludes 1.0 and corresponds to about five extra weeks of median survival — statistically detectable in 3,743 patients, and modest for a patient. Lead-time bias is the central problem: finding a cancer earlier in people already being scanned extends survival measured from diagnosis whether or not anything changes. Propensity matching addresses measured differences between the groups, not the surveillance itself.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.