A meta-analysis «Singal AG, Zhang E, Narasimman M, ym. HCC surveill...»1 included 59 cohort studies from 2014 to 2020, with a total of 145396 patients with liver cirrhosis and hepatocellular carcinoma (HCC). 53 of the studies were retrospective. Included studies reported outcome measures for both patients undergoing surveillance (either ultrasound alone or ultrasound and alpha fetoprotein (AFP) measurement) and patients without surveillance. HCC was detected by surveillance in 28.2% of the cases. HCC surveillance was associated with improved early-stage detection (49 studies, RR 1.86, 95% CI 1.73-1.98; I² = 82%), curative treatment receipt (39 studies, RR 1.83, 95% CI 1.69-1.97; I² = 75%), and overall survival (12 studies adjusted for lead-time bias; hazard ratio 0.67, 95% CI 0.61-0.72; I² = 78%). However, authors reported notable heterogeneity (inconsistency index, I²) in all pooled estimates.

• Study quality: moderate
• Applicability: low
• Comment: Heterogeneity was high. Most study populations consisted of viral liver disease; the proportion of metabolic dysfunction-associated steatotic liver disease (MASLD) cirrhosis in the cohorts was generally low, with only two studies reporting >20% MASLD patients.

Another meta-analysis «Tzartzeva K, Obi J, Rich NE, ym. Surveillance Imag...»2 studied the sensitivity and specificity of different surveillance strategies. It included 32 cohort studies from 1990 to 2016, with a total of 13367 patients with cirrhosis. 23 of the studies were prospective. Ultrasound detected early-stage HCC with a fairly low 47% sensitivity (95% CI 33%-61%), while the 4 studies evaluating computed tomography or magnetic resonance imaging detected HCC with 84% sensitivity (95% CI 70%-92%). In studies comparing ultrasound with vs without AFP measurement, ultrasound detected early-stage HCC with a lower level of sensitivity than ultrasound plus AFP measurement (RR 0.81; 95% CI 0.71-0.93). 4 studies (2245 patients) reported accuracy for early HCC detection. The specificity for ultrasound alone was 92% (95% CI 85% – 96%) compared to 84% (95% CI 77% – 89%) for ultrasound plus AFP.

• Study quality: moderate
• Applicability: low
• Comment: Heterogeneity was high. Proportion of non-viral etiologies was not assessed, but is likely to be low.