OBJECTIVES: To clarify problems with the determination of serum albumin levels, the definition of hypoalbuminemia, and the implications of microheterogeneity of albumin, serum albumin was measured by using dye-binding methods and the authentic method (immunoassay) in patients with liver cirrhosis and healthy subjects.
METHODS: We enrolled 103 patients with liver cirrhosis and 36 healthy subjects. Serum albumin levels were analyzed by immunoassay and the bromcresol green and bromcresol purple methods. Oxidized albumin and glycoalbumin were determined by high-performance liquid chromatography.
RESULTS: In cirrhotic patients, serum albumin levels measured by the bromcresol green method was about 0.2 g/dL higher than that by immunoassay. Serum albumin levels measured by the bromcresol purple method also was higher in cirrhotic patients than those measured by immunoassay and varied widely. In addition, extensive variation was found across serum albumin levels determined by the bromcresol green method at individual institutions (five university hospitals) and those determined by immunoassay at a contract laboratory. The percentages of oxidized albumin and glycoalbumin within total serum albumin increased with progression of liver disease. Further, an increase in oxidized albumin led to an increase in the albumin level as measured by the bromcresol purple method.
CONCLUSION: These results show that adequate assessment of the pathophysiology and prognosis of patients with liver cirrhosis and the efficacy of treatment is not possible with dye-binding methods for determination of serum albumin. Further, the conventional definition of hypoalbuminemia as a serum albumin level of 3.5 g/dL or lower should be reconsidered, and the clinical implications of qualitative changes in albumin should be investigated in consideration of the microheterogeneity of albumin, such as oxidized albumin and glycoalbumin.