Table 5.

Algorithmic approaches using the combination of serologic and clinical parameters for hepatocellular carcinoma risk prediction

Risk scores	Cohort: Patients/ratio of cirrhosis	Study population	Variables	External validation
GALAD score (Johnson et al[72], 2014)	HCC case: 670/90%; CLD control: 339/97%	Caucasian (England)	AFP; AFP-L3; DCP	Asian, Caucasian
Doylestown algorithm (Wang et al[76], 2016)	Training HCC case: 165/100%; CLD control: 195/100%; Validation 1 HCC case: 432/100%; CLD control: 438/100%; Validation 2 HCC case: 113/100%; CLD control: 586/100%; Validation 3 HCC case: 425/100%; CLD control: 804/100%	North America (United States)	Age; Gender; ALT; ALP; AFP; Fucosylated kininogen	North America (United States)
GALADUS score (Yang et al[75], 2019)	Training HCC case: 111/98%; CLD control: 180/85%; Validation HCC case: 233/100%; CLD control: 412/100%	North America (United States)	AFP; AFP-L3; DCP; Ultrasonography	North America (United States)
HES algorithm (Tayob et al[78,79], 2019)	HCC case: 4804/100%; CLD control: 33627/100%	North American (United States Veterans Administration)	Age; Rate of AFP change; ALT; Platelet count; Etiology of cirrhosis	-

AFP: Alpha-fetoprotein; AFP-L3: Alpha-fetoprotein-L3; ALP: Alkaline phosphatase; ALT: Alanine aminotransferase; DCP: Des-gamma-carboxy-prothrombin, HCC: Hepatocellular carcinoma; CLD: Chronic liver disease.