This study focused on sNGAL, uNGAL concentrations and uNGAL/creatinine ratio in dehydrated horses and compared these new markers with the traditional renal markers throughout the first 48 h of hospitalization time.
NGAL is rapidly expressed under various pathological conditions by several cell types, such as immune cells, hepatocytes and renal tubular cells [24]. There are two major forms of NGAL identified in humans and animals: a dimeric form which is released by neutrophils and a monomeric form that is secreted by tubular epithelial cells in the distal part of the nephron [11, 24, 25, 26]. Impaired reabsorption of NGAL in the proximal tubule, along with the rapid secretion of NGAL after kidney injury, could both lead to the elevation of sNGAL and uNGAL [11]. Several studies showed that both sNGAL and uNGAL can elevate rapidly within 2 h after renal damage and that both plasma and urinary NGAL correlated well with the creatinine concentration [13, 15, 27, 28]. In the present study, significant correlations between sNGAL, uNGAL, creatinine and BUN concentrations at admission were identified. Although uNGAL/creatinine ratio had a strong correlation with uNGAL, uNGAL/creatinine ratio correlated only significantly with BUN. The delayed concentrate ability of kidney after rapid dehydration which leads to inconsistent of the sampled urine to the real dehydration status [29, 30], for example a mix of normal urine and concentrated urine in the bladder at sampling time, might be the reason why uNGAL/creatinine ratio did not correlate with serum creatinine concentration. sNGAL and uNGAL, uNGAL/creatinine ratio correlated with each other throughout the 48 h study period. After T0, sNGAL, uNGAL and uNGAL/creatinine ratio did not show a correlation with either creatinine or BUN, which may be due to the different underlying diseases and different treatment regimes in our patients.
A precise cut-off value of NGAL for the usage of detecting AKI has not yet been defined, but in previous studies, sNGAL concentrations were significantly higher in horses with increased creatinine compared to control horses [31, 32]. Our study also showed similar results: dehydrated horses had higher mean and median sNGAL than our healthy controls, and what is more, healthy horses in our study all had sNGAL concentrations similar to the control group from the previous publication.
Unlike creatinine, which only increases when the renal damage has already advanced, NGAL tends to show the “active damage” in the kidney and, therefore, is suitable as a real-time indicator of AKI [12, 33, 34]. Except for one horse diagnosed with AKI during its hospitalization, creatinine and BUN concentrations decreased back to the reference range until T12 under rehydration therapy. By contrast, sNGAL, uNGAL and uNGAL/creatinine ratio remained elevated in most cases and never decreased to the concentrations measured in the healthy control horses (Fig. 3–7). sNGAL increased continuously in a total of 39.5% (15/38) of the dehydrated horses until T48 (all of them increased above 10 ng/mL), only three of these 15 horses had serum creatinine concentrations above the reference range at T0 and only one of these three horses had sustained high creatinine until T48. uNGAL increased sustainably in a total of 37.5% (9/24) of the patients until T48, 2 of these 9 horses had increased serum creatinine at T0 but both decreased between T12 and T24. Similarly, uNGAL/creatinine ratio kept increasing in 45.8% (11/24) of the horses throughout 48 h, only one horses among them had increased creatinine concentration until T12. The reason for that might be that our patients had sustained minor renal damage even after rehydration therapy that could not be detected by traditional renal markers. Markers for tubular injury, such as NGAL, could be more sensitive than GFR markers, such as creatinine, to detect prerenal AKI because the tubule is more susceptible to hypoperfusion and hypoxic damage than the glomerulus [27, 35, 36]. No significant difference of either sNGAL or uNGAL were identified for time-group interaction and among time points throughout 48 h in this study. There were lots of extrarenal factors affecting NGAL concentrations both in serum and urine because of the various primary disorders and therapy choices of each patient. These influences probably reduced the differences between the three dehydration groups (mild, moderate and severe). Regardless of the disease process and medication, however, sNGAL and uNGAL both varied significantly among three dehydration groups and healthy controls at T0 according to the dehydration status. This indicated that NGAL might have some ability to reflect the decreased GFR or minor renal damage caused by dehydration. uNGAL/creatinine ratio, however, showed also no association with the dehydration status at T0.
Although two studies in human cardiac patients indicated that plasma NGAL could be a predictive marker of mortality, neither sNGAL nor uNGAL or uNGAL/creatinine ratio were associated with the survival rate in other studies in dogs [36–38]. Although the one horse diagnosed with AKI in our study was euthanized, neither uNGAL or uNGAL/creatinine ratio nor sNGAL showed any significant difference between survivors and non-survivors. Therefore, even when NGAL might have shown early renal damage, it did not markedly affect the survival chance in our patient.
Several studies showed that correction of uNGAL with urinary creatinine yield similar test results in comparison with uncorrected uNGAL and uNGAL alone showed already good sensitivity and specificity to AKI [11, 39]. In this study, uNGAL shared similar value to short-term prognosis with uNGAL/creatinine ratio but not in dehydration status and correlations with traditional renal markers. The retention of the urine in the bladder might be one of the concern which might affect the concentration of urinary creatinine.
Since NGAL also originates from neutrophils, it has been discussed whether inflammation in the body could have an impact on this new marker. Studies described that NGAL could also increase in various infectious, malignant disorders and inflammations, even without kidney injury [24, 26, 40]. Research in small animals and humans indicated that there was a significant influence of SIRS on sNGAL, especially, and, therefore, limits its usage in patients with systemic inflammation and possible AKI [11, 37, 41]. This is further complicated by the lack of a uniform definition of the term SIRS in adult horses and the wide variety of existing scoring systems. Most scoring systems in horses resemble those from human medicine that were designed to ensure identification of at-risk patient. This leads to a high sensitivity and poor specificity and one should be careful when applying it to such heterogenous study group as ours. Urinary NGAL is thought to be more specific for diagnosing AKI than sNGAL, nevertheless, in cases with nephritis or pyuria, uNGAL still needs to be interpreted cautiously [24, 38]. However, neither increased urinary leukocytes nor leucocytosis in the blood made a difference between groups and they both did not affect the correlations of sNGAL, uNGAL and uNGAL/creatinine ratio with other biomarkers in this study. Because of a small number of these horses with inflammation, the interpretation of this results should be make cautiously. In this study, sNGAL, uNGAL and uNGAL/creatinine ratio all had significant differences among SIRS scores when SIRS score 1 was compared with score 3 or 4. However, since most of the SIRS criteria such as high heartbeat are unspecific, it could be caused by pain, stress or dehydration. When we selected horses with a SIRS score>2, increased lactate and abnormal leukocytes (all at admission), a significant difference of sNGAL has been found, while uNGAL and uNGAL/creatinine ratio showed no difference between these groups. This results indicate that sNGAL might be significantly influenced by inflammation.
Gentamicin and NSAIDs are known as potential nephrotoxic drugs and might cause transient renal damage through the clinical usage [42, 43]. However, we only had urine samples at all time points available in 7/28 horses in our study and, therefore, results after T0 should be interpreted cautiously. None of the chosen markers (sNGAL, uNGAL and uNGAL/creatinine ratio) and serum creatinine, BUN concentrations associated significantly with the usage of these two drugs. We supposed that there was no drug-related AKI developed in this group of horses or the minor damage was hard to detect by these markers.
The lack of urine samples at all time points and the heterogenous study group were the main limitations of this study. Potentially nephrotoxic medications such as Gentamicin and NSAIDs could not be avoided and were used in different regimes in individual patients.
In addition, sNGAL and uNGAL were measured in samples stored at -80 collected and retained for around one year. Validation of the ELISA Kit showed no influence on the NGAL measurement when -80 storage could be provided; the NGAL protein should be stable for at least 11 months under that condition [44]. However, we could not assure whether there were any differences between fresh and frozen samples.