Child-Pugh B, C scores predict failure in liver cancer surveillance
Child-Pugh B and C, and elevated alpha-fetoprotein, at diagnosis of hepatocellular carcinoma corresponded with a diagnosis outside Milan criteria and surveillance failure during a surveillance program of patients with cirrhosis.
“HCC surveillance in cirrhotic patients aims to detect tumors in the initial stages of the disease to offer treatments that increase patient survival,” Alejo Mancebo, MD, from the Hospital Universitario Central de Asturias, Spain, and colleagues wrote. “Although the usefulness of screening programs remains controversial, several studies have shown survival improvement in patients under surveillance.”
To assess the frequency and risk factors associated with HCC surveillance failure, the researchers prospectively followed 1,242 patients with cirrhosis enrolled in an HCC surveillance program with quarterly or semi-annual ultrasonography and alpha-fetoprotein (AFP) screening tests.
During a median follow-up of 48 months, 188 patients developed HCC who comprised the study. The patients were mostly men (85.1%) and developed cirrhosis due to alcohol use (50%), hepatitis C (39.8%), hepatitis B (6.9%) or other cause.
At the time of HCC diagnosis, median patient age was 62.6 years, most patients had Child-Pugh class A (69.7%) and the mean AFP level was 8.8 ng/mL (range, 1-18,430). Twenty-eight patients had an AFP level of 100 ng/mL or higher, 9 patients had an AFP level higher of 1,000 ng/mL or higher, and 50 of the 188 tumors were outside Milan criteria.
“HCC surveillance is not recommended for Child-Pugh class C patients, except for those included in a waiting list for liver transplantation,” Mancebo and colleagues wrote. “In our series, 12 patients were in this class at the time of tumor detection, but all of them had been in class A or B in the previous screening round, suggesting that the development of the tumor could have contributed to the liver function impairment.”
Patients with tumors outside Milan criteria had significantly more nodules at diagnosis (2.8 vs. 1.2; P < .001) than patients within Milan criteria, as well as higher diameters among the larger tumors (4.8 vs. 2.4 cm; P < .001), higher levels of AFP (659.4 vs. 48.5 ng/mL; P < .001) and shorter survival after diagnosis (7.6 vs. 33.9 months; P < .001).
Multivariate analysis showed that Child-Pugh B or C (HR = 3.15; 95% CI, 1.78-5.55) and AFP levels of 100 ng/mL or higher (HR = 2.8; 95% CI, 1.37-5.71) at HCC diagnosis correlated with surveillance failure.
“Variables collected before HCC diagnosis, such as Child-Pugh class B at inclusion or the development of portal hypertension-related complications during follow-up, are more useful in selecting patients with a higher risk of program failure,” the researchers concluded. “Current screening strategies are probably not suitable for these patients, and different modalities should be investigated.” – by Talitha Bennett
Disclosure: The authors report no relevant financial disclosures.