Using NGAL as an Early Diagnostic Test of Acute Kidney Injury

Abstract

Neutrophil gelatinase-associated lipocalin (NGAL) has generated great interest as a novel biomarker for the timely detection of acute kidney injury (AKI). Despite the enthusiasm surrounding NGAL, the research so far details and attempts to minimize a host of limitations that substantially preclude its use as a valuable diagnostic biomarker to detect AKI and guide clinical treatment. In our review of the current research, obvious drawbacks such as variable sensitivity and specificity, even among similar patient populations were discovered. Furthermore, there are not well-defined cutoff values among various patient populations which would permit use of NGAL as a positive or negative diagnostic marker similar to troponin in cardiac injury. Moreover, due to the wide variation in baseline concentration of NGAL among patients, the added requirement of serial measurements, that may not even be accurate in at-risk or chronic kidney injury populations, further degrades the benefit of early detection.

• acute kidney injury
• marker
• neutrophil gelatinase-associated lipocalin
• early diagnosis

INTRODUCTION

Neutrophil gelatinase-associated lipocalin (NGAL) has been postulated as a new diagnostic biomarker for the early detection of acute kidney injury (AKI). Currently, the diagnostic tests used for the detection of AKI are serial serum creatinine measurements and monitoring of urine output.¹ The use of serum measurements of creatinine concentration as a functional marker of kidney excretory function was introduced to clinical practice in the 1960s. Serum creatinine as a marker of AKI has several limitations, including the lack of sensitivity, especially in patients with initially normal kidney function. There is also a lack of specificity due to factors such as ingestion of meat containing creatinine, its release by injured skeletal muscles, and delayed serum changes.² The treatments that have experimentally shown to reduce AKI have not been successful in clinical practice, perhaps due to delay in diagnosis. It is hypothesized that these treatments would be more clinically effective when initiated earlier.³ As such, this review of the literature was undertaken to attempt to correlate the timing, sensitivity, specificity, and utility of NGAL as a biomarker of AKI in order to determine if it can be practically applied to the clinical setting.

BACKGROUND

NGAL, also known as lipocalin-2 and oncogene 24p3, is a human protein encoded by the LCN2 gene and is expressed by multiple human cells including epithelial cells, neutrophils, as well as in various organs of gastrointestinal, respiratory, excretory, and reproductive systems.^4–6 NGAL is a 24 kDa protein rapidly eliminated from the circulation. T_1/2 is estimated at 10–20 min, depending on its form. The monomeric form is cleared twice as fast compared with the dimeric form.⁷ The precise mechanism of elimination is not known, but the molecule is too large to be eliminated by simple kidney excretion.

NGAL is thought to play several roles in regard to the kidney function. It has been experimentally demonstrated that NGAL may be involved in kidney development.⁸ NGAL has also been implicated in renal regeneration and repair after ischemic injury.⁹ A protective effect on the kidney from ischemia–reperfusion injury has also been demonstrated.¹⁰ These effects are thought to be linked to the role that NGAL plays in iron metabolism in growth and differentiation as well as prevention of cellular death.^8,11,12

Currently, at least one study demonstrates that NGAL is also a novel marker for progression of chronic kidney disease (CKD).¹³ Additionally, an inverse correlation between glomerular filtration rate (GFR) and serum NGAL (sNGAL) reflects the well-known theory that the progression of CKD is faster when there are lower numbers of glomeruli that are not damaged or destroyed.¹⁴ However, the diminished GFR seems to only partially explain the variability of sNGAL. Accordingly, in the “forest fire theory,” increased NGAL levels may reflect inflammatory damage to vital tubular cells rather than forms of acute injury. It was further postulated that sNGAL and urine NGAL (uNGAL) may be useful as sensitive markers of kidney impairment/injury in CKD patients.¹⁵

NGAL AS A BIOMARKER FOR AKI

If NGAL is to be demonstrated as effective in the early diagnosis of AKI and superior to the currently employed method of creatinine estimation, the optimal time of measurement and cutoff value must be established.

Currently, it is possible to measure NGAL in both urine and serum (uNGAL and sNGAL, respectively.)¹⁶ The increase of sNGAL does not reflect the decline of GFR like the increase of serum creatinine, but detects injury of kidney tissue, predominantly damage to tubular cells. Thus, its measurement both in serum and urine is expected to show the different aspects of AKI, like the markers of myocardial damage, that is, troponin and myoglobin. It should be stressed once again that NGAL is not an organ-specific marker. The significant limitations of sNGAL measurement as a structural marker of AKI are increased concentrations in patients with lower respiratory tract infections,¹⁷ systemic infections/sepsis,¹⁸ thrombocythemia, polycythemia,¹⁹ and cancers,²⁰ while uNGAL increases in patients with urinary tract infections²¹ as well being above individual baselines in CKDs.²² The implications of these limitations include false positive or negative results which could lead to failure of timely diagnosis, which of course is the stated purpose of using NGAL as a biomarker for AKI, demonstrating that the addition of quantitative measurement does not add value to clinical decision processes. The increased levels of NGAL in other disease processes also have the potential of causing misdiagnosis of AKI where none exists, as well as masking and delaying the detection and treatment of other pathologies. In order to mitigate these limitations performing both sNGAL and uNGAL measurements may be suggested to reconcile the differences in an effort to determine if the increase in NGAL is from renal or prerenal etiology, theoretically resulting in greater sensitivity for AKI and ruling out other causes of the increase. Additionally, in many of the above studies the authors have suggested a myriad of solutions in attempts to address the limitations of NGAL; however, no universal solution has been postulated as applicable to all patient populations.

Thus, uNGAL may be a better marker of AKI than sNGAL. Supporting this conclusion is a recent publication by Haase et al.²³ describing the renal origin of uNGAL in the distal tubule verified by photon emission detection in mouse models.

Multiple studies using sNGAL and uNGAL have attempted to determine if NGAL is both sensitive and specific in patients with AKI subsequent to various etiologies. Some populations were shown to be more sensitive to detection of sNGAL versus uNGAL and vice versa, requiring the ability to measure both body fluids. The use of photon emission detection, which is very accurate in the detection of NGAL in laboratory models, is both logistically and economically prohibitive to use clinically.

It has been shown that NGAL levels start to increase significantly as early as 1 h after kidney ischemia in cardiac surgery patients.²⁴ Numerous studies have demonstrated increased measurements of both urine and plasma NGAL 2 h after kidney injury caused by variety of etiological factors, where the baseline levels were known, for the purpose of AKI detection. The application of rapid detection tests for NGAL (based on turbidimetric method) further reduces the time of detection to approximately 15 min, which is at least 24 h earlier than serum creatinine measurements show a detectable change.³ Additionally, in the study of AKI post liver transplant, Wagener et al.²⁵ demonstrated an initial decrease in urine creatinine concentration post surgery for 3 h prior to an increase in patients with and without meeting RIFLE-R (risk of renal dysfunction) criteria while still registering an increase in uNGAL. In keeping with the hypothesis from Ronco³ that clinical intervention for AKI is not initiated early enough, due to the limitations of current detection standards, NGAL has been demonstrated as being detectable earlier as well as being more sensitive and specific than other potential AKI markers such as cystatin C (6 h post insult), IL-18 (12 h post insult), or kidney injury molecule-1 (12 h post insult)²⁶ in an attempt to detect AKI soon enough to allow for intervention.

As demonstrated in Table 1, there is a wide variability of baseline concentrations of NGAL in patients who do not develop AKI. This complicates efforts to determine early elevation from normal baseline. Thus, serial measurements of NGAL showing the dynamic changes seem to be more useful than just a single measurement. NGAL levels are demonstrated to stabilize between 24 and 48 h post insult. Most of the literature attempts to correlate NGAL detection of injury to the RIFLE score, which is based on creatinine and urine output. Studies demonstrate that the detectable levels of NGAL begin to stabilize about the same time as creatinine, with all of its limitations, begins to be detectable. This creates the limitation of needing creatinine measurements to definitively describe any increase in NGAL as diagnostic for AKI.^25,29,30 Moreover, should intervention be taken previously to prevent or respond to the potential of AKI in any subject, successful therapy may preclude the patient meeting criteria of the RIFLE score thus making the NGAL measurement superfluous.

Table 1. sNGAL and uNGAL concentrations in humans without acute kidney injury (AKI) (prior to injury).

Population	NGAL concentration	Reference
Noncardiac major surgery	93 ng/L ± 38 (serum)	Shavit et al.^22
Noncardiac major surgery	83 ng/L ± 234 (urine)	Shavit et al.^22
Critically ill children	14.2 ng/mL ± 27 (urine)	Zappitelli et al.^27
Critically ill adults	48.1 ng/mL (urine), IQR(12.7–179.3)	Siew et al.^28

Notes: Mean values (95% CI) ± SD. uNGAL, neutrophil gelatinase-associated lipocalin in urine; sNGAL, neutrophil gelatinase-associated lipocalin in serum; CI, confidence interval.

Sensitivity and specificity are also important factors in determining whether NGAL would make a superior biomarker to the current standard, especially when it will increase the cost by more than 10-fold (Table 2). Many similar studies that attempted to address the detection time of NGAL also looked at these factors. However, unlike the relatively uniform findings in detection time, a range of sensitivity and specificity was found in various studies. As was comprehensively detailed in the meta-analysis by Haase et al.³¹ the sensitivity and specificity vary at widely different cutoff values. Confounding the sensitivity and specificity of NGAL even further, different patient populations such as premature infants³² and those with existing risk factors for or actual CKD, ²² it is demonstrated that the sensitivity, specificity, and cutoff values for NGAL to detect AKI are specific to each patient population. It seems that while NGAL can add quantitative data, these data do not add anything to the clinical decision processes. Devarajan¹ recently published that even within similar patient populations, in this case cardiac surgery, sensitivity varied in two studies with a similar number of patients; sensitivity was 49% in one group with a specificity of 79%³³ and the sensitivity was 100% with a specificity of 98% in another.³⁴ Additionally, the use of NGAL may also be viable only in the absence of major risk factors or CKD.²² The increasing NGAL levels with CKD stage is another factor limiting usefulness in differentiating between AKI, acute-on-chronic, and CKD in patients with an unknown etiology of kidney injury and previously increased serum creatinine concentrations that went undetected prior to a suspected recent injury.³⁵

Table 2. Sensitivity and specificity of NGAL assessment for detection of AKI across patient populations.

Population	Size	Serum/uNGAL	Sensitivity (%)	Specificity (%)	Note	Study
Premature infants with oliguria	20	Urine	31	88	Birth weight and development were listed as confounding factors	Lavery et al.^32
Noncardiac major surgery	74	Urine	60	31	2 h post-op finding	Shavit et al.^22
		Serum	70	46	6 h finding had only marginal increase in sensitivity and 4–5% increase in specificity >90% of subjects had pre-existing CKD or risk factors for CKD
Liver	92	Urine	83.5	67.5	Maximized 3 h post-op by uNGAL/creatinine ratio	Wagener et al.^25
Critically ill multi-trauma	31	Urine	91	95		Makris et al.^29

Note: NGAL, neutrophil gelatinase-associated lipocalin; AKI, acute kidney injury; CKD, chronic kidney disease; uNGAL, neutrophil gelatinase-associated lipocalin in urine.

CONCLUSION

NGAL increase is definitively detectable much earlier than creatinine. The lack of a precise value at which AKI is determined across patient populations would reasonably decrease the initial enthusiasm and preclude its use for detection of AKI in a manner similar to using troponin to detect cardiac injury.

The wide range of normal baseline values confounds the ability to make clinical decisions in patients where the patient’s baseline NGAL measurement is not known. When compounding this with the limitations of patient populations and prior disease and/or risk states, there appears a limit to clinical usefulness. At the very least, it would require multiple measurements over a period of hours to establish any increase, significantly increasing both the cost and detection times.

Care must be exercised not to favorably prejudge the usefulness of NGAL measurements for the early diagnosis of AKI. Until its usefulness can be better validated, it would be better for the clinicians to assume the potential for AKI across all of the at-risk populations and to develop prophylactic procedures for patients at risk for AKI rather than expend time and expense trying to use NGAL measurements diagnostically to guide treatment.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.