Friday, May 28, 2010

Doc, I have kidney pain radiating to my left ureter....

The quest for biomarkers to accurately diagnose AKI is often analogized to ACS/myocardial infarction. Thus we have dubbed AKI biomarker investigation the "search for our troponin," with the goal of facilitating early and accurate diagnosis of AKI that will lead to successful interventions, the equivalent of PCI for STEMI. What remains unclear is what should trigger such markers to be tested. If the usual method of considering AKI via creatinine rise (of varying magnitudes) or fall in urine output is felt to be too late, a new set of earlier criteria must be used to trigger evaluation for AKI via the usual means (urinalysis, microscopy, urine electrolytes etc) and via biomarker testing. To that end, Goldstein and Chawla have tried to define "renal angina," the combination of risk factors, clinical criteria, and laboratory findings that are suggestive of evolving AKI and should trigger further exploration by available means, or, at some future point, with testing of proven biomarkers. They address both the adult and pediatric population but I will focus on the adults for now.

They note multiple known risk factors for AKI including:

  • Age > 65 years
  • Diabetes
  • CKD
  • Cirrhosis/hepatic failure
  • CHF
  • Sepsis
  • Volume depletion
  • Cardiopulmonary Bypass
  • Exposure to nephrotoxins (contrast, medications etc)
They then suggest three criteria that should constitute "renal angina" though they leave the exact definitions of each criterion somewhat open:
  1. oliguria for 1-5 hours

  2. any increase in serum creatinine from 0.1-0.4 mg/dL

  3. volume overload

What is somewhat dissatisfying is that their criteria are nearly equivalent to, and therefore no better than, current practice at detecting AKI to trigger further evaluation. Once volume overload, oliguria, or rising creatinine (however small) ensue, the damage may already be done. Further focus on the creatinine value, effectively a late marker, to define early AKI is seriously flawed. Setting aside the statistical pitfalls of comparing new markers to an imperfect gold standard, creatinine, no matter how small the changes considered, is still a poor biomarker of AKI - more analogous to AST or LDH in ACS/MI- and in the search for better alternatives to it, it cannot also be used to define the clinical syndrome it seeks to diagnose, any more than AST, LDH, or troponin can define angina. The goal is to define a high risk group in which to detect or evaluate for AKI before a significant decrement in GFR signified by a rising creatinine! And this is where I think the analogy falls apart.

Early or evolving AKI or risk for renal injury has no clinical signs or symptoms analogous to cardiac angina. If this is the case, what should trigger the putative biomarker testing? Seems that risk factors would be more appropriate. Maybe everyone should have AKI markers checked on admission to the medical ICU? Or maybe hospitalized patients with three or more risk factors, a renal TIMI score if you will, should have AKI markers checked regularly. As the editorialists suggest, the utility may be in the principle that we define an appropriate clinical scenario in which to screen for early AKI (the "angina") and not in the actual criteria proposed to define renal angina. Or maybe the analogy has, in the words of Hollywood, "jumped the shark" - that is, run its course?

No comments: