- Design
- Retrospective derivation cohort with bootstrap and external validation
- Population
- 598 patients with HCC and cACLD (Italy); 149 for validation (Vienna)
- Primary outcome
- Missed varices and endoscopies spared
- Effect
- Missed high-risk varices 3.0% (SAPHE-R) vs 6.0% (Baveno VI); 17.6% endoscopies spared
This study drew on the Italian Liver Cancer database to include 598 patients with hepatocellular carcinoma (HCC, BCLC 0-C) and compensated advanced chronic liver disease who had liver stiffness, endoscopy and blood tests within six months of diagnosis, with validation in 149 patients from Vienna. Published in Hepatology on 22 October 2026.
The Baveno VI criteria, widely used to avoid screening endoscopy in cirrhosis, missed 8.5% of varices and 6.0% of high-risk varices in these patients. The new SAPHE-R rule, liver stiffness of 16 kPa or less, platelets of 160 × 10⁹/L or more and ALBI grade 1, missed 4.0% of varices and 3.0% of high-risk varices, sparing 17.6% of endoscopies. In validation it missed 4.8% and 3.1%, sparing 14.8%. None of 22 rule-out patients with pressure measurements had clinically significant portal hypertension.
Patients with HCC often need endoscopy before systemic therapy, particularly anti-VEGF agents that raise bleeding risk. A rule that safely avoids some endoscopies, and is more reliable than Baveno VI in this group, saves time and procedures.
The rule spares fewer endoscopies than Baveno VI in exchange for missing fewer varices. Its validation cohort was modest, and it has not yet been tested prospectively.
- Do not rely on Baveno VI criteria alone to skip endoscopy in patients with HCC
- Consider skipping screening endoscopy when liver stiffness is 16 kPa or less, platelets 160 or more and ALBI grade 1
- Proceed to endoscopy for everyone else before starting anti-VEGF therapy
- Repeat assessment if liver function or platelets change during treatment
Why it matters
A rule designed for cirrhosis missed more varices in liver cancer than clinicians may assume.
Don't overread it
Retrospective derivation with modest external validation; it needs prospective testing.
The statistics, in plain English
A rule-out test is judged by how many cases it misses. SAPHE-R missed about 3% of high-risk varices, half the Baveno VI rate, but at the cost of avoiding fewer endoscopies.
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.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for gastroenterology & hepatology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free