Source Nobuhiro Aizawa, Hirayuki Enomoto, Hiroyasu Imanishi, Masaki Saito, Yoshinori Iwata, Hironori Tanaka, Naoto Ikeda, Yoshiyuki Sakai, Tomoyuki Takashima, Takashi Iwai, Ei-ichiro Moriwaki, Soji Shimomura, Hiroko Iijima, Hideji Nakamura, Shuhei Nishiguchi, Division of Hepatobiliary and Pancreatic Disease, Department of Internal Medicine, Hyogo College of Medicine, Mukogawa-cho 1-1, Nishinomiya, Hyogo 663-8501, Japan.
AIM: To analyze the relationship between the glycated albumin (GA) to glycated hemoglobin (HbA1c) ratio and the histological grading of liver fibrosis.
METHODS: The study retrospectively included consecutive hepatitis C virus positive chronic liver disease patients (n = 142) who had undergone percutaneous liver biopsy between January 2008 and March 2010 at our institution. The ratios of GA/HbA1c were calculated in all patients to investigate the relationship with the degree of the liver fibrosis. The values of the aspartate aminotransferase-to-platelet ratio index (APRI), an excellent marker for the evaluation of liver fibrosis, were also calculated. In addition, we combined the ratio of GA/HbA1c and the APRI in order to improve our ability to detect the presence of significant liver fibrosis.
RESULTS: Sixty-one (43%) patients had either no fibrosis or minimal fibrosis (METAVIR score: F0-F1), while 25 (17%) had intermediate fibrosis (F2). Fifty-six (39%) patients had severe fibrosis (F3-F4) and 27 of them had cirrhosis (F4). The mean values of the GA/HbA1c increased with the progression of the fibrosis (F0-1: 2.83 ± 0.24, F2: 2.85 ± 0.24, F3: 2.92 ± 0.35, F4: 3.14 ± 0.54). There was a significant difference between the F0-F1 vs F4, F2 vs F4, and F3 vs F4 groups (P < 0.01, P < 0.01, P < 0.01 and P < 0.05, respectively). The GA/HbA1c ratio was significantly higher in the patients with cirrhosis (F4) than in those without cirrhosis (F0-F3) (3.14 ± 0.54 vs 2.85 ± 0.28, P < 0.0001). The GA/HbA1c ratio was also significantly higher in the patients with severe fibrosis (F3-F4) than in those without severe liver fibrosis (F0-F2) (3.03 ± 0.41 vs 2.84 ± 0.24, P < 0.001). Furthermore, the GA/HbA1c ratio was also significantly higher in the patients with significant fibrosis (F2-F4) than in those without significant liver fibrosis (F0-F1) (2.98 ± 0.41 vs 2.83 ± 0.24, P < 0.001). The diagnostic performance of the increased GA/HbA1c ratio (> 3.0) was as follows: its sensitivity and specificity for the detection of liver cirrhosis (F4) were 59.3% and 70.4%, respectively and its sensitivity and specificity for the detection of severe liver fibrosis (F3-F4) were 50.0% and 74.4%, respectively. With regard to the detection of significant fibrosis (F2-F4), its sensitivity was 44.4% and its specificity was 77.0%. Although even the excellent marker APRI shows low sensitivity (25.9%) for distinguishing patients with or without significant fibrosis, the combination of the APRI and GA/HbA1c ratio increased the sensitivity up to 42.0%, with only a modest decrease in the specificity (from 90.2% to 83.6%).
CONCLUSION: The GA/HbA1c ratio increased in line with the histological severity of liver fibrosis, thus suggesting that this ratio is useful as a supportive index of liver fibrosis.