Diagnosis of DGF is usually made by elevated serum Cr, which is a late signal of kidney injury. Early predict and treatment is essential for better outcome. Therefore, it would be ideal to find a marker better than Cr to early predict the acute kidney injury (19). In the study, we compared the serum NGAL with serum Cr as predictive biomarkers for DGF following kidney transplantation. The study showed that values of serum NGAL after transplantation was more useful than absolute or percentage of serum Cr decrease in predicting DGF. In both ROC and multivariable analyses, serum NGAL was superior to serum Cr in predicting early DGF. We have shown that absolute values of serum NGAL on the first hours of post-transplantation correspond with DGF, whereas Serum Cr values at these times poorly differentiate between groups. In addition, the results revealed that better accuracy and predictive power of serum Cr occurred at 4 or more days after transplantation. The absolute value of serum Cr on the post-transplantation day 1 and 4 for predicting DGF resulted in AUCs was (0.607, 95%CI: 0.346–0.868, P = 0.398) and (0.770, 95%CI: 0.562–0.977, P = 0.029), respectively. Therefore, serum NGAL detects DGF 1 to 3 days earlier than serum Cr after transplantation. Furthermore, urine output at 24 hours after transplantation had modest diagnostic performance AUC (0.747, 95% CI: 0.555–0.939, P = 0.045).
The accuracy and predictive power of serum Cr was affected by many of parameters such as age, sex and muscle mass, protein metabolism, volume of distribution, and drugs (20). Furthermore, the ability of creatinine to detect functional impairment is less than 50% (21). Several studies have been evaluated the role of urine or serum NGAL in predicting DGF after kidney transplantation, compare with traditional biomarkers such as serum Cr and urine output (22–24). Similar to our study they showed that NGAL level was more useful than serum Cr in early predicting DGF. In contrast, Mahdavi-Mazdeh et al. (25) reported ROC curve and AUCs of serum NGAL and serum Cr levels on the first post-transplantation day had similar significance in predicting DGF. However they showed that the highest AUC (0.82) was attributed to serum NGAL at 24 hour (P = 0.002). Parikh et al. (23) in 53 transplant patients from living and deceased donor found a better AUC (0.9) for urine NGAL in prediction of DGF than serum Cr. While, Mojtahedzadeh eta al. (26) in 69 transplant patients from deceased donor found a relative fall in serum Cr had higher AUC (0.821 vs. 0.790) compared to urine NGAL at 24 hour post-transplantation in prediction of DGF. In contrast to our findings, some of previous studies showed that NGAL level in early hours of postoperative were not able to differentiate patients with DGF from those with non-DGF (25, 26). Lebkowska, et al. (27) found a significant decrease in serum NGAL as early as 24 hours after transplantation, before the serum Cr level decreases. Two studies demonstrated that 48 hours urine NGAL showed larger AUCs and better sensitivity and specificity than those of early hours after transplantation (28, 29).
Plasma and urine NGAL were evaluated and compare in some studies to investigate their performance to predict DGF (21, 22, 30). In some studies plasma NGAL showed an obviously higher sensitivity and a slightly higher specificity than those of urine NGAL, which supported the superiority of plasma NGAL over urine NGAL in predicting DGF within 24 hours following kidney transplantation (30). Hollmen et al. (21) found a higher AUC value of serum NGAL (AUC, 0.85) compared to urine NGAL (AUC, 0.74). However, Maier et al. (22) reported higher AUC (0.74) of urine NGAL compare to serum NGAL (0.73) at 24 hour post-transplantation to predict DGF. These contradictory findings may be due to differences in study designs or related to sample size. Therefore, large prospective kidney transplantation cohort with different sampling time points is required to illustrate the features of NGAL in predicting the risk of DGF.
The small sample size is the main limitation of this study, so further studies in larger cohorts is necessary to confirm our results. The small sample size also limited our ability to evaluate meaningfully between different variables and to examine important patient subgroups. In addition, due to the low sample size, a multivariate analysis of the combination of the two biomarkers was not possible. Another limitation of this study was the unavailability of cold ischemia time.
In conclusion, the study showed that the superior of serum NGAL to serum Cr as an early predictor of graft function and DGF. Serum NGAL at the early hours after transplantation also demonstrates excellent potential for predicting DGF with fair sensitivity and specificity. While, measuring serum Cr at 24 hour post-transplantation did not help in the diagnosis of DGF. Our findings emphasize the promising role that NGAL serum can play as an early indicator of DGF.