Neutrophil gelatinase-associated lipocalin (NGAL) is a promising new biomarker of AKI.  While the cardiologists have troponins, and now highly sensitive troponins, nephrologists have been stuck using creatinine as a marker for AKI.  Creatinine has many flaws: at least 50% of kidney function must be lost before it even rises; it is highly dependent on age, gender, muscle mass, etc.; it does not distinguish between pre-renal, post-renal, and renal causes; and by the time it rises, the damage to the kidney is already done.  There is a lot of research devoted to finding a better marker, and maybe NGAL is it.

This review (and this more formal meta-analysis) discusses the evidence supporting the potential of NGAL as a marker for AKI.  In children, it appears to be a fantastic biomarker: whereas Cr bumped 1-3 days after cardiac surgery in kids, there was a 10-fold elevation in NGAL within 2-6 hours.  It was 100% sensitive and 98% specific.  It appears a bit less wonderful but still quite good in adults, with sensitivities generally ranging from the 70’s to the 90’s%, and specificities in roughly the same range.

There are a couple things to consider here: the diagnosis of AKI in these studies is made based on serum creatinine, which as we just discussed is a rather poor “gold standard.”  What if some of the “false-positives” were in fact not, and there was AKI without rise in Cr?  Or what if some of the “false-negatives” were in fact Cr elevations not due to AKI?  That’s the problem when you have a bad gold-standard.  It’s hard to really judge the new test.  I think the proof will be in the pudding: if NGAL elevations lead to more prompt treatment, or to novel treatments effective early in AKI but ineffective after Cr elevations, then clearly it would be a god-send.

Another comment in the article caught my attention: “[NGAL] performance compares favourably with that of troponin for the prediction of myocardial infarction during its clinical implementation period.”  I think looking for the perfect biomarker is probably to miss the point.  Troponin is one of the closest things we have, and that took years before it reached it’s current level of precision.  I think we’ll find that for AKI, this will be pretty similar: a good, but not perfect, biomarker that will be combined with some lesser biomarkers (analogous to CPK/CPK-MB) that will provide an answer.  Maybe not an irrefutable answer, but certainly a lot better than we have now.  And I suspect that much like troponins, with time NGAL (or some successor) will be refined to a point where we will finally have a gold-standard for detecting AKI.