Premium
Ethnicity affects the diagnostic validity of alpha‐fetoprotein in hepatocellular carcinoma
Author(s) -
GAD Amal,
TANAKA Eiji,
MATSUMOTO Akihiro,
ELHAMID SERWAH Abd,
ATTIA Fawzy,
HASSAN Adel,
SANNY Ahmed,
ALI Khalil,
ABBAS Amro,
ELRAOOF ELDEEB Abd,
SUN Xiao Hong,
UMEMURA Takeji,
ICHIJO Tetsuya,
EHARA Takashi,
YOSHIZAWA Kaname,
KIYOSAWA Kendo
Publication year - 2005
Publication title -
asia‐pacific journal of clinical oncology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.73
H-Index - 29
eISSN - 1743-7563
pISSN - 1743-7555
DOI - 10.1111/j.1743-7563.2005.00017.x
Subject(s) - hepatocellular carcinoma , medicine , gastroenterology , alpha fetoprotein , hbsag , receiver operating characteristic , bilirubin , pathology , immunology , hepatitis b virus , virus
Hepatocellular carcinoma (HCC) is the fourth most common cancer worldwide with a high morbidity and mortality. Alpha‐fetoprotein (AFP) is considered the main tumor marker for HCC diagnosis, but the variation in its diagnostic validity among studies justifies further investigation of the underlying contributing factors. Ethnic difference could be one of the factors that has not been well studied. We aimed at investigating the ethnic difference in AFP validity between Egyptian (representing Arabic North African) and Japanese (representing Asian) for HCC diagnosis. Methods: Four cohorts with chronic liver diseases (CLD) were studied: 171 Egyptian (65 HCC/106 non‐HCC), and 173 Japanese (45 HCC/128 non‐HCC). Laboratory tests including serum AFP, protein‐induced vitamin K deficiency or absence (PIVKA‐II), alanine aminotransferase (ALT), total bilirubin, platelet count, HBsAg, anti‐HCV, and HCV core antigen were conducted using standard commercially available assays. Results: A significantly higher sensitivity of AFP in Egyptian in comparison with Japanese for HCC diagnosis (99 vs 67%, P < 0.001) was observed using an AFP cut‐off point of 10 ng/mL, with a comparable specificity (75 vs 82%) While a sensitivity of 98 versus 56%, P < 0.001 and a specificity of 83 versus 89% was found for AFP cut‐off point of 20 ng/mL, respectively. The area under the receiver operating characteristic curve (ROC) was found to be 0.98 (95%CI = 0.969–0.997) for Egyptian and 0.77 (95%CI = 0.686–0.864) for Japanese. The highest sensitivity for the former group occurred at AFP = 20.5 ng/mL and at AFP = 10.2 ng/mL for the latter. Univariate analysis showed no effect for age, sex, underlying liver disease, cirrhosis, Child's class or tumor characteristics (size, pathological grade) on AFP sensitivity, while race significantly contributed to the higher sensitivity among Egyptians in comparison with the Japanese. Using ROC analysis, the AFP cut‐off point for HCC detection in each subgroup of patients with and without each of the risk factors of interest was determined and the subgroups were again subclassified according to AFP positivity (< or ≥ the decided cut‐off point for each group). Logistic regression analysis of those factors combined showed that Egyptian ethnicity with an AFP level >20.5 ng/mL ( P = 0.007), older age (>50 years) with an AFP level >26 ng/mL ( P = 0.010), and cirrhosis with an AFP level >10.5 ng/mL ( P = 0.014) were the independent risk factors for HCC. Conclusion: There is an ethnic variation in AFP validity between Egyptian and Japanese patients with a significantly lower sensitivity in the latter. Alpha‐fetoprotein should not be the only marker used for screening HCC among Asian Japanese and younger age groups (<50 years) with CLD. In addition, an AFP cut‐off point of 20 ng/mL is recommended when screening patients of Asian origin for HCC.