Serum albumin level is one of the important measures for Child-Pugh classification score that indicates liver insufficiency.1 Bromocresol green (BCG), which is used for conventional serum albumin measurement, reacts with proteins other than albumin. This results in the overestimation of albumin levels. The traditional bromocresol purple (BCP) method is highly specific for albumin, but reacts differently depending on the albumin form (e.g., mercaptalbumin, nonmercaptalbumin, albumin bound to bilirubin). On the other hand, the modified BCP methods are highly specific for albumin and are not affected by albumin form, thus allowing accurate and reproducible serum albumin measurement.2 In Japan, many laboratories have substituted the modified BCP method for the BCG method, and 45% of laboratories employed the modified BCP method in 2011. However, the modified BCP method generates lower values than does the BCG method. Thus, substituting the modified BCP method for the BCG method is likely to alter a patient's Child-Pugh class.
The objectives of the present study were (1) to compare serum albumin values that were determined by the BCG method and the modified BCP method in patients with liver cirrhosis (LC) and in patients with hepatocellular carcinoma (HCC) with underlying LC, and (2) to test whether the different reagents used to determine the serum albumin levels can alter the Child-Pugh classification. The serum albumin concentrations of 103 patients with LC or HCC were determined by immunonephelometry (N-Antiserum to Human Albumin; Siemens, Tokyo, Japan), the BCG method (ALB-A; Sysmex, Tokyo, Japan), and the modified BCP method (Albumin-II HA Test Wako; Wako Pure Chemicals Industries Ltd., Osaka, Japan). Patients provided informed consent.
Serum albumin levels measured by the modified BCP method were well correlated with the levels measured by immunonephelometry (gold standard) (Fig. 1). Serum albumin levels obtained by the BCG method were significantly higher than the levels measured by the modified BCP method (P = 0.031, Student t test). This overestimation of the albumin level by the BCG method resulted in a lower albumin score in the Child-Pugh classification in 11 of the 103 patients. Of 14 patients with an albumin score of 2 by using the BCG method, 2 patients were re-scored as 3 by the modified BCP method. Of 66 patients with an albumin score of 1 by using the BCG method, 9 patients were re-scored as 2 by the modified BCP method. This re-scoring resulted in a change in Child-Pugh class from A to B in another patient and from B to C in another patient when the modified BCP method was employed instead of the BCG method.
Thus, new criteria should be set in institutions that employ the modified BCP method. The threshold values for the scoring in the Child-Pugh classification were 28.0 g/L and 35.0 g/L. The threshold values for the modified BCP method were calculated as 25.3 g/L and 32.9 g/L from the regression equation (y = 1.076x − 4.8) between the BCG (x) and modified BCP (y) methods. Institutions should examine these criteria to set new criteria for the modified method. The method by which serum albumin is measured should be specified in both clinical and research settings.
In conclusion, the modified BCP method provided more accurate albumin measurements than did the BCG method. Overestimation of serum albumin levels by the BCG method can alter both the Child-Pugh score and thereby the Child-Pugh class in patients with LC and HCC.