Objective: Diagnosis of acute kidney injury (AKI) relies on measurement of serum creatinine concentration and urine flow, which change slowly and have low specificity and sensitivity. We investigated the potential of urinary levels of alpha-glutathione S-transferase (alpha-GST) and pi-GST - markers of proximal and distal renal tubule damage, respectively - to provide an earlier and more accurate indication of AKI in patients in the intensive care unit.
Design, setting and participants: Urine samples were collected over the 48 hours after ICU admission from 40 consecutive patients who were admitted with a diagnosis of sepsis between October 2007 and May 2008. AKI was diagnosed during the 48h after ICU admission with the criteria of the Acute Kidney Injury Network (AKIN). Urinary alpha-GST and pi-GST levels were measured with commercially available enzyme-linked immunosorbent assay (ELISA) kits. Serum creatinine concentration was also measured. Haemodynamic and resuscitation parameters were recorded, but managed independently by the ICU team.
Results: Urine samples were analysed from 38 patients (21 men, 17 women) with a median age of 54 years (interquartile range [IQR], 41-69 years), median APACHE II score of 13.3 (IQR, 8-17), and median ICU length of stay of 9 days (IQR, 3-19 days). Hospital mortality was 24%, and ICU mortality was 13%. Nineteen patients (50%) developed AKI, all within 24h of ICU admission. Urinary alpha-GST level was not increased in patients who developed AKI versus non-AKI patients. Median (IQR) urinary pi-GST level (microg/L) at ICU admission was 10.8 (4.7-22.65) in the non-AKI group, 19.3 (2.88-44) in those who developed Stage 1 AKI, and 27.4 (14.8-43.8) in those who developed Stage 3 AKI. Median urinary pi-GST level at ICU admission was higher in all groups than in healthy control subjects. The area under the receiver operating characteristics curve for urinary pi-GST level indicated that it was not a good predictor of AKI.
Conclusions: Urinary pi-GST is elevated early in all patients with sepsis syndrome, but is not predictive of AKI as defined by AKIN. This may indicate sensitive detection of an earlier phase of kidney injury, and suggests that sepsis-related renal injury affects the distal tubules, giving new insights into the pathophysiology of AKI in sepsis.