URINARY ENZYMES IN ACUTE-RENAL-FAILURE

Intestinal-type alkaline phosphatase (IAP) has been localized to the S 3 segment of the renal tubule in previous studies, a site believed to be particularly vulnerable to toxic and ischaemic damage. During a 17- month period a pilot study of the value of urinary enzyme measurements (IAP and tissue non-specific alkaline phosphatase-TNAP, using monoclo nal antibody-based immunoassays, and N-acetyl-beta-glucosaminidase-NAG , using colorometric assay) in 50 prospectively followed cases of acut e renal failure (ARF) was performed. Urinary enzymes were measured at initial evaluation ('start') and then each day for 14 days, with the h ighest enzyme value ('peak') also used for analysis. Patients were div ided into prerenal (n = 16), renal (n = 28), postrenal (n = 6) categor ies according to standard criteria. Of the renal ARF patients 23 of 28 had acute tubular necrosis (ATN), 3 of 28 acute interstitial nephriti s (AIN), and 2 of 28 acute glomerulonephritis (AGN); 18 of 50 had a fa tal outcome and 1 of 50 was dialysis-dependent at discharge ('poor' pr ognosis group), while 31 of 50 survived hospital without becoming dial ysis-dependent ('good' prognosis group).Median enzyme concentration we re increased in 'poor' compared to 'good' prognosis patients: start IA P 3.2 versus 2.2 U/g creat (NS), start NAG 48.6 versus 13.7 (P<0.01), start TNAP 3.5 versus 0.9 (P<0.02). When renal ARF patients alone were analysed, only IAP (3.2 versus 1.3 U/g creat at start) and NAG (57.9 versus 7.8 U/g creat at start) were significantly increased in the poo r compared to the good prognosis group. Peak values showed similar tre nds. Of all patients, five with a start IAP > 12 U/g creat died, and a ll survivors had a start IAP < 12, but 14 of 19 poor prognosis patient s also had a start IAP < 12. All urinary enzymes were less in the post renal group, but only the IAP significantly so. None of the enzymes we re significantly different between prerenal and renal ARF groups. Urin ary enzymes IAP, NAG, TNAP appear to be unhelpful in determining the s ite of renal injury in ARF, except for postrenal cases, where IAP was significantly lower. There were too few patients with AGN or AIN to te st the hypothesis that the enzymes would be less in glomerular compare d to tubular pathologies. Despite a low sensitivity, the start IAP may be a marker of outcome in ARF if the high positive predictive value f or death is confirmed in larger studies.