SCM14: Late-Breaking Abstract: CrCl Equation for AKI

Dr. John Mellas

Dr. John Mellas

Dr. John Mellas, from Saint Mary’s Health Center, Saint Louis, MO, discusses his abstract for the National Kidney Foundation’s 2014 Spring Clinical Meetings (SCM14), Description of Two Methods for Accurately Estimating Creatinine Clearance in Acute Kidney Injury, with Dr. Kenar Jhaveri (eAJKD), eAJKD Editor.

eAJKD: Why don’t you tell us a little about your research and abstract being presented at NKF 2014 Spring Meetings?

 JM: Having practiced nephrology for 28 years, I have been particularly interested in finding a method to quantify the degree of renal impairment in AKI. In that there is no accepted method for measuring GFR in the non-steady state, I determined that this was a much needed calculation in clinical nephrology. It is clear that the AKIN and RIFLE criteria estimate renal injury in a crude and inaccurate way. Furthermore, they are retrospective and not forward looking. Meanwhile, the optimism that we would find a kidney biomarker to estimate renal injury has still not panned out.

Looking back at the original GFR estimating equations, in AKI, by Jeliffe in 1972, along with Myer and Moran in 1985, it was clear that these methods were flawed, and based on the fact the GFR was dependent on a changing serum creatinine concentration, along with an estimate of the creatinine distribution volume. In that many patients with AKI receive large quantities of parenteral crystalloid and blood products, these equations, although mathematically correct, were not deemed accurate in estimating GFR in actual patients due to the problems involved in using serum creatinine measurements.

With this background, I decided to pursue estimating GFR in AKI by measuring urine creatinine. By this method, the creatinine excreted can be measured directly and accurately. If one knows the value for creatinine excretion (E) and can accurately estimate creatinine production P, (using creatinine production estimating equations) then it is apparent that the ratio of E/P will predict the direction of the serum creatinine trajectory from any time zero where: E/P<1 predicts a rise in creatinine, E/P>1 predicts a fall in creatinine, and E/P=1 predicts a stable creatinine concentration.

This method was evaluated in 250 patients where over 450 measurements were taken. The predictive accuracy of this method for determining the direction of the azotemia was between 85% and 100% with five different end points measured which included rising, falling, or stable serum creatinine, as well as predicting the eventual need for renal replacement therapy, or  predicting discontinuation of dialysis in advance.

We then reasoned that the ratio E/P multiplied by the estimated GFR at any time zero, i.e., the static GFR (Ks) would give an accurate estimate of actual GFR where K=Ks*E/P.

To show that this method was valid, we looked at 6 model patients with varying degrees of renal injury, as well as differences in age, sex, weight, and estimated P. Using the equation C(t)=P/K+[C0-P/K]*e^-Kt/V, we measured C(t) using known values for the other variables. From this, we calculated V*dC/dt where E=P-V*dC/dt. Using this value for E, we found that the value for K=Ks*E/P showed an excellent correlation with known measures of K, when using 4 hour urine collections. Hence, we derived Method 1 for estimating GFR in AKI where K=Ks*E/P.

Taking this logic one step further, it is apparent that V*dC/dt (which cannot be measured directly with accuracy) is equal to P-E. We next entered this value into the differential equation for creatinine clearance in the non-steady state; K=P/C-V*dC/dt/C and confirmed the accuracy of this method for estimating GFR in AKI.  This equation represented Method 2 for measuring GFR in AKI.

Finally, we changed several assumptions in the model patients where actual P was 20% below estimated P, and V was 50% body weight as opposed to 60% as in the original models. Much to our surprise, both equations still performed very well in estimating actual GFR in AKI.

eAJKD: Does the cause of AKI matter in this equation?

JM: By estimating GFR by either of these methods, one is able to determine the extent of renal injury regardless of etiology. When this is added to additional gathered clinical data, one can then determine the etiology of the renal injury, use appropriate drug dosing, and determine the renal prognosis in the short term.

In that AKI is often a dynamic process in unstable patients, one can take serial measurements of GFR to determine the course of the renal disease.

eAJKD: Where do you and your group go from here?

JM:  Our goal is to publish these methods, and ultimately have them verified by independent investigators. In the interim, we continue to use them at our institution and find it invaluable for estimating GFR in patients with AKI.

Click here for a full list of SCM14 abstracts of poster presentations.

Check out more eAJKD coverage of the NKF’s 2014 Spring Clinical Meetings!

 

 

 

 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s