Comparison of the diagnostic value of symmetric dimethylarginine, cystatin C, and creatinine for detection of decreased glomerular filtration rate in dogs

Abstract

Background: Early detection of decreased glomerular filtration rate (GFR) in dogs is challenging. Current methods are insensitive and new biomarkers are required.

Objective: To compare overall diagnostic performance of serum symmetric dimethylarginine (SDMA) and serum cystatin C to serum creatinine, for detection of decreased GFR in clinically stable dogs, with or without chronic kidney disease (CKD).

Animals: Ninety-seven client-owned dogs: 67 dogs with a diagnosis or suspicion of CKD and 30 healthy dogs were prospectively included.

Methods: Prospective diagnostic accuracy study. All dogs underwent physical examination, systemic arterial blood pressure measurement, urinalysis, hematology and blood biochemistry analysis, cardiac and urinary ultrasound examinations, and scintigraphy for estimation of glomerular filtration rate (mGFR). Frozen serum was used for batch analysis of SDMA and cystatin C.

Results: The area under the curve of creatinine, SDMA, and cystatin C for detection of an mGFR <30.8 mL/min/L was 0.98 (95% confidence interval [CI], 0.93-1.0), 0.96 (95% CI, 0.91-0.99), and 0.87 (95% CI, 0.79-0.93), respectively. The sensitivity of both creatinine and SDMA at their prespecified cutoffs (115 μmol/L [1.3 mg/dL] and 14 μg/dL) for detection of an abnormal mGFR was 90%. The specificity was 90% for creatinine and 87% for SDMA. When adjusting the cutoff for cystatin C to correspond to a diagnostic sensitivity of 90% (0.49 mg/L), specificity was lower (72%) than that of creatinine and SDMA.

Conclusions and clinical importance: Overall diagnostic performance of creatinine and SDMA for detection of decreased mGFR was similar. Overall diagnostic performance of cystatin C was inferior to both creatinine and SDMA.

Keywords: CKD; azotemia; biomarker; canine; renal.

Figure 1

Flowchart of the study including reasons for exclusion of 8 dogs. CKD, chronic kidney disease; mGFR, measured glomerular filtration rate; SDMA, symmetric dimethylarginine

Figure 2

Original plots of A, creatinine; B, symmetric dimethylarginine (SDMA); and C, cystatin C versus measured glomerular filtration rate (mGFR) in 97 dogs. Linear associations between D, log creatinine; E, log SDMA; and F, log cystatin C and mGFR in 97 dogs

Figure 3

A, Measured glomerular filtration rate (GFR); B, symmetric dimethylarginine (SDMA); and C, cystatin C concentrations in 30 healthy dogs and 50 dogs with chronic kidney disease, in international renal interest society (IRIS) stages 1‐3. The top, bottom, and middle lines through each box correspond to the 75th percentile, the median, and the 25th percentile, respectively. The whiskers extend from the bottom 10th percentile to the top 90th percentile. Asterisks indicate significant differences in biomarker concentrations between groups (P < .01)

Figure 4

Receiver operator characteristic curve analysis. Comparison of overall diagnostic value of creatinine, symmetric dimethylarginine (SDMA), and cystatin C for detection of A, measured glomerular filtration rate (mGFR) <30.8 mL/min/L and B, mGFR <37 mL/min/L in 97 dogs. There was no difference between overall performance of creatinine and SDMA at either cutoff for mGFR (P = .37 and .70, respectively)

Figure 5

Partition modelling. Resultant decision tree when every clinical variable was entered into the model to predict if mGFR was abnormal (< 30.8 mL/min/L) or not. An identical decision tree resulted if only creatinine, SDMA and cystatin C were entered into the model. Cystatin C units: mg/L. Creatinine units: μmol/L. The equivalent of 127 μmol/L is 1.4 mg/dL and the equivalent of 100 μmol/L is 1.1 mg/dL