Assessment of urinary retinol-binding protein as an index of proximal tubular injury.

Abstract

Abstract The urinary excretion of retinol-binding protein (RBP), beta 2-microglobulin (beta 2-m), and beta-N-acetyl-D-glucosaminidase was monitored in patients with renal tubular damage secondary to multiple injuries, rhabdomyolysis, antibiotic treatment, or poisoning by various chemicals such as solvents, heavy metals, or pesticides. In almost all cases, RBP proved to be a more sensitive index of renal tubular damage than was beta-N-acetyl-D-glucosaminidase and, being more stable in acid urine, a more practical analyte to measure than was beta 2-m. We corroborated this finding by studying the relationships between these three analytes in more than 150 patients. On the average, an increase in the urinary excretion of beta-N-acetyl-D-glucosaminidase becomes detectable when urinary RBP already exceeds the normal value by 50- to 100-fold. In urines with pH greater than 6, RBP and beta 2-m concentrations are well correlated (r = 0.93, n = 150), beta 2-m tending to be more frequently positive (i.e., greater than 311 micrograms/L). But in urines with pH less than 6 (about 30-40% of the samples), the RBP/beta 2-m concentration ratio increases as pH decreases, up to 500 in some patients with massive tubular injury. Because the renal uptake of proteins involves a saturable process, the urinary excretion of RBP, like that of beta 2-m, specifically reflects the reabsorption capacity of proximal tubules only when the glomerular filtration rate is normal or slightly impaired (i.e., serum creatinine less than 20 mg/L). Under these conditions the determination of RBP protein in urine appears the most appropriate test when early detection of tubular injury is desirable.