Serum uromodulin in dogs with chronic kidney disease

Abstract

Background

Serum uromodulin concentration has been described as a novel biomarker of chronic kidney disease (CKD) in humans but not dogs.

Objective

To evaluate the serum uromodulin concentration in dogs with CKD and assess its diagnostic performance in distinguishing dogs with CKD from healthy dogs.

Animals

Forty‐nine dogs with CKD (International Renal Interest Society [IRIS] Stage 1, n = 23; Stage 2, n = 20; Stage 3‐4, n = 6) and 25 healthy controls.

Methods

Prospective, observational study. Serum uromodulin concentration was measured using a canine‐specific enzyme‐linked immunosorbent assay (ELISA), and its correlation with conventional renal markers was analyzed.

Results

Serum uromodulin concentrations were significantly lower in the CKD group than in the control group (P < .001), but no significant difference was observed among stages of CKD. A negative correlation was observed between serum uromodulin concentration and conventional renal markers (blood urea nitrogen concentration, r = −.60, P < .0001; serum creatinine concentration, r = −.46, P < .0001; serum symmetric dimethylarginine concentration [SDMA], r = −.65, P < .0001). In receiver operating characteristic analysis, the area under the curve (AUC) of uromodulin (AUC, 0.97; 95% confidence interval [CI], 0.94‐1.00) was higher than that of SDMA (AUC, 0.87; 95% CI, 0.79‐0.95) for CKD diagnosis (P = .01). The AUC of uromodulin (AUC, 0.95; 95% CI, 0.89‐1.00) also was higher than that of SDMA (AUC, 0.72; 95% CI, 0.58‐0.87) in distinguishing dogs with Stage 1 CKD from controls (P = .001).

Conclusions and Clinical Importance

Serum uromodulin concentration is decreased in dogs with CKD. Thus, serum uromodulin may be a valuable diagnostic marker for CKD in dogs, particularly in identifying early‐stage CKD.

Abbreviations

AUC

area under the curve

BUN

blood urea nitrogen

CKD

chronic kidney disease

CV

coefficient of variation

ELISA

enzyme‐linked immunosorbent assay

GFR

glomerular filtration rate

IRIS

International Renal Interest Society

ROC

receiver operating characteristic

SD

standard deviation

SDMA

symmetric dimethylarginine

TAL

thick ascending limb

1. INTRODUCTION

Chronic kidney disease (CKD) is common and mostly affects older dogs and results from an irreversible decrease in the nephron mass. ¹ Diagnosis is made when functional or structural changes in 1 or both kidneys are observed persistently for >3 months. ² Structural abnormalities of the kidney can be recognized using radiography or ultrasonography. However, changes in normal kidney structure are not always detected in these patients, and findings do not reflect the degree of functional impairment. ¹ Thus, when CKD is suspected, evaluation of kidney function is essential for confirmed diagnosis.

Measurement of glomerular filtration rate (GFR) is considered the gold standard for evaluating kidney function, with urinary inulin clearance used as the reference method. ³ However, inulin clearance is not often used in clinical practice, with endogenous renal markers being more commonly used for routine evaluation of the kidney function. Serum creatinine concentration is the most widely used marker in veterinary medicine, but it is insensitive in detecting mild decreases in GFR. ⁴ Therefore, recent studies have focused on searching for superior markers that enable detection of CKD at an early stage. ⁵

Uromodulin, also known as Tamm‐Horsfall protein, is the most abundant urinary protein under normal physiological conditions in dogs and humans. ⁶ It is a 95 kDa glycoprotein exclusively produced by the renal tubular epithelial cells lining the thick ascending limb (TAL) of the loop of Henle and by the early portion of the distal tubule to a lesser extent. ⁷ After production, most of the protein traffics toward the apical membrane of the TAL cell, which faces the tubular lumen, and is anchored by glycosylphosphtidylinositol at that site. ⁸ Membrane‐bound uromodulin is released into the lumen by proteolytic cleavage and excreted in urine. ⁹ However, a small portion of uromodulin also is detected in the serum of humans. Although the detailed mechanism is not fully understood, circulating uromodulin is thought to be derived from the basolateral release of uromodulin from the TAL cells, supported by immunolocalization studies showing the presence of uromodulin at the basolateral membrane of these cells. ¹⁰

Recent studies in human medicine have suggested that serum uromodulin concentration is a promising biomarker of CKD. Previous studies reported that serum uromodulin concentration was commonly decreased in human CKD patients. and was correlated with estimated GFR, which is the reference criterion for CKD staging in humans. ¹¹ , ¹² , ¹³ Moreover, serum uromodulin concentration has an advantage over urinary uromodulin concentration in terms of accurate measurement because the protein tends to form large aggregates in the urine. ¹⁴ Only a few studies have evaluated urinary uromodulin concentration in dogs, with decreased concentrations observed in dogs with CKD, ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ but its concentration in serum has not yet been studied.

We hypothesized that serum uromodulin concentrations would be lower in dogs with CKD and decrease with the disease severity. Our objectives were to assess serum uromodulin concentrations in dogs with CKD and compare them with those of healthy controls, to examine the relationship between uromodulin and other renal biomarkers used in veterinary medicine, and to evaluate the diagnostic performance of serum uromodulin concentration in dogs with CKD.

2. MATERIALS AND METHODS

2.1. Study population and selection criteria

Client‐owned dogs were prospectively recruited from those referred to Seoul National University Veterinary Medical Teaching Hospital between July 2021 and May 2022. Of the 116 dogs screened, 74 dogs were eligible for enrollment and were categorized as dogs with CKD or healthy controls. A flowchart of case enrollment is presented in Figure 1.

FIGURE 1.

Flowchart for enrollment of healthy and CKD dogs into the present study. CKD, chronic kidney disease; SDMA, symmetric dimethylarginine

Diagnosis of CKD was made based on the 2019 International Renal Interest Society (IRIS) CKD guidelines. ¹⁹ Patients persistently (≥3 months) having at least 1 of the following criteria were assigned to the CKD group: renal azotemia, renal proteinuria, or renal abnormalities on abdominal ultrasonography. Persistent renal azotemia was defined as serum creatinine concentration ≥1.4 mg/dL with repeated urine specific gravity <1.030 according to the patient's medical history. Persistent renal proteinuria was based on urine protein‐to‐creatinine ratio ≥0.5 documented on ≥2 occasions with ≥2 weeks. Proteinuria was considered renal in origin when there was no apparent evidence of pre‐renal or post‐renal causes. Ultrasonographic renal abnormalities indicative of CKD included small kidneys (kidney‐to‐aorta ratio <5.5), increased cortical echogenicity, decreased corticomedullary distinction, irregular renal contours, or some combination of these findings. Among the 3 diagnostic criteria, dogs assigned to the CKD group (49 dogs) met the diagnostic conditions as follows: persistent renal azotemia in 26 dogs, persistent renal proteinuria in 18 dogs, and ultrasonographic findings compatible with CKD in all 49 dogs. All 23 dogs in the Stage 1 CKD group had renal abnormalities on imaging analysis. Of those, 6 dogs also had persistent proteinuria.

Exclusion criteria for the CKD group were as follows: transient proteinuria (excluded based on persistence), pyuria or hematuria on microscopic sediment examination, positive results on urine culture, urogenital neoplasia, findings suggestive of acute kidney injury, acute worsening of CKD, and pre‐renal or post‐renal causes likely to increase serum creatinine concentration (Figure 1). Dogs with concurrent illnesses were not excluded when concurrent disorders were considered unlikely to influence uromodulin production or release into serum. These included gall bladder mucocele (n = 7), cholelithiasis (n = 4), myxomatous mitral valve disease (n = 19), upper respiratory disorders (n = 18), chronic pancreatitis (n = 13), gastric emptying disorder (n = 1), hypothyroidism (n = 2), hyperadrenocorticism (n = 6), diabetes mellitus (n = 2), idiopathic epilepsy (n = 1), cognitive disorder (n = 1), intervertebral disc disease (n = 3), pedal furunculosis (n = 1), perineal hernia (n = 1), and allergic dermatitis (n = 1). Dogs in the CKD group were further staged according to their serum creatinine concentrations based on the criteria determined by the IRIS guidelines. ¹⁹

The control group included dogs admitted to the Seoul National University Veterinary Medical Teaching Hospital for routine medical evaluations or client‐owned dogs enrolled in another study that was conducted separately. Dogs were considered healthy when there were no remarkable findings in their medical history, physical examination, or clinicopathologic analyses, including CBC, serum biochemistry, and urinalysis. Additionally, they did not have any morphological abnormalities in their kidneys on abdominal ultrasonography.

All experimental protocols related to the use of animals in this study were reviewed and approved by the Institutional Animal Care and Use Committee of Seoul National University (SNU‐201123‐3‐1).

2.2. Sample collection

Blood was collected as part of the patient's diagnostic or medical evaluation. The samples were collected in serum separation tubes. Serum was obtained by centrifugation at 4000 rpm for 4 min. After use in routine biochemistry tests, the remaining serum samples were aliquoted and stored at −80°C for uromodulin analysis. The maximum storage time of the serum samples was 8 weeks based on our preliminary results, in which the serum uromodulin stored at −80°C showed good stability for up to 8 weeks (Figure S1).

2.3. Uromodulin ELISA

Serum uromodulin concentrations were measured using a canine‐specific sandwich ELISA kit (Canine Uromodulin ELISA, BioVendor‐Laboratorni medicina a.s., Brno, Czech Republic). All analytical procedures were performed according to the manufacturer's instructions. Each sample was measured in triplicate. The provided standard material was diluted 2‐fold using a diluent buffer provided in the kit to obtain a standard curve (10, 5, 2.5, 1.25, 0.63, 0.31, 0.16, and 0 ng/mL). Frozen serum samples were thawed at room temperature before measurement. Samples were diluted (1:10 to 1:40) with the same diluent buffer used for the standard material to obtain a measurable concentration within the standard curve range (0‐10 ng/mL). The diluted samples were incubated with biotin‐labeled antibodies and conjugated with streptavidin‐horseradish peroxidase. After washing, the remaining conjugates were allowed to react with the substrate solution. The reaction was stopped by adding an acidic solution, and the absorbance of the wells was read using a microplate spectrophotometer (Epoch Microplate Spectrophotometer, BioTek Instruments, Winooski, Vermont, USA) at 450 nm.

The intra‐ and inter‐assay coefficients of variation (CVs) were calculated to evaluate the precision and reproducibility of the assay. Inter‐assay CVs were determined in canine samples with 3 different uromodulin concentrations (high, medium, and low), and the samples were analyzed in triplicate using 4 separately conducted assays. Intra‐assay CVs were calculated using the triplicate results from all examined serum samples.

2.4. Measurement of conventional renal markers

Blood urea nitrogen (BUN), creatinine, and SDMA concentrations were measured for all of the dogs at the time of serum collection. The BUN and serum creatinine concentrations were determined using an automated analyzer (Hitachi 7170 Chemistry Analyzer, Hitachi, Tokyo, Japan), according to the manufacturer's instructions. The SDMA concentration was measured using an enzyme immunoassay (Catalyst Dx Chemistry Analyzer, IDEXX, Westbrook, Maine, USA).

2.5. Statistical analysis

Statistical analyses were performed using GraphPad Prism (version 9.3.1, GraphPad Software Inc., San Diego, California, USA). Continuous data are presented as means with SD or medians with ranges, depending on normality as tested by the Shapiro‐Wilk test. Differences in continuous variables (age, body weight) between the 2 groups were analyzed using the Mann‐Whitney U‐test. Comparison of continuous variables (BUN, creatinine, SDMA, and serum uromodulin concentrations) among multiple groups was performed using the Kruskal‐Wallis test with post‐hoc comparisons using Dunn's multiple comparison test. Spearman's rank correlation coefficients were used to assess the correlation between serum uromodulin concentration and other biological variables. Receiver‐operating characteristic (ROC) analysis was performed to evaluate and compare the diagnostic performance of uromodulin and SDMA in differentiating CKD patients from the healthy controls (all CKD dogs vs controls, dogs with Stage 1 CKD vs controls). The AUC values are reported with 95% confidence intervals (CIs). DeLong's test using the pROC package ²⁰ in R (version 4.2.0, R Foundation for Statistical Computing, Vienna, Austria) was used to compare the AUCs of uromodulin and SDMA. For uromodulin, the optimal cut‐off value for discriminating Stage 1 CKD dogs from controls was determined at maximum Youden's index (sensitivity + specificity −1) and the sensitivity and specificity at the decided threshold are presented with 95% CIs.

3. RESULTS

3.1. Demographic characteristics of dogs

Our study included 74 dogs, with 25 in the healthy control group and 49 in the CKD group. The control group consisted of 15 castrated males, 1 intact male, 8 spayed females, and 1 intact female. The CKD group consisted of 26 castrated males, 1 intact male, 21 spayed females, and 1 intact female. The median age of the control and the CKD groups were 6 years (range, 1‐12 years) and 13 years (range, 4‐19 years), respectively, showing a significant difference between the groups (P < .001). The median body weight was 4.6 kg (range, 1.5‐33.5 kg) for the control group and 3.8 kg (range, 1.6‐22.0 kg) for the CKD group, and no significant difference in body weight was observed between groups (P = .24). Most of the included dogs belonged to small breeds (body weight <15 kg). Among them, Maltese (n = 17), Poodle (n = 12), and Shih tzu (n = 10) breeds were the most common in the overall study population. Breeds in the control group included Poodle (n = 7), Maltese (n = 3), mixed breed (n = 3), Bedlington Terrier (n = 2), Bichon Frise (n = 2), Dachshund (n = 2), Pomeranian (n = 2), and 1 of each of the following breeds: Chihuahua, Coton de Tulear, Golden Retriever, and Papillon. Breeds in the CKD group included Maltese (n = 14), Shih tzu (n = 10), Poodle (n = 5), Yorkshire Terrier (n = 5), Pomeranian (n = 3), Chihuahua (n = 2), Pekingese (n = 2), Schnauzer (n = 2), and 1 of each of the following breeds: Bichon Frise, Border Collie, Dachshund, mixed breed, Silky Terrier, and West Highland White Terrier.

3.2. Precision and reproducibility of ELISA for uromodulin

The precision and reproducibility of ELISA for serum uromodulin concentration were evaluated. The mean ± SD for intra‐assay CV was 4.16 ± 2.01%, with all CVs in the range of 0.28% to 9.04%, indicating good assay precision. Inter‐assay CVs were calculated using 3 samples with different uromodulin concentrations (Sample 1, low; Sample 2, medium; Sample 3, high). The CV values for Samples 1, 2, and 3 were 7.44%, 2.85%, and 2.38%, respectively. Therefore, the reproducibility of the assay was considered acceptable.

3.3. Serum uromodulin concentration and its association with conventional renal biomarkers

The median serum uromodulin concentrations of the control and CKD groups were 94.98 ng/mL (range, 51.31‐267.70 ng/mL) and 42.75 ng/mL (range, 3.51‐91.16 ng/mL), respectively (Figure 2). The dogs in the CKD group were further classified according to their serum creatinine concentrations, and all CKD stages showed significantly lower serum uromodulin concentrations than those of the healthy control group. Serum uromodulin concentration was significantly lower in dogs with Stage 1 CKD compared to controls (P < .001; Table 1). However, no significant differences were observed among the CKD stages. The serum uromodulin concentration showed significant but moderate negative correlations with BUN (r = −.60, P < .0001), creatinine (r = −.46, P < .0001), and SDMA (r = −.65, P < .0001; Figure 3A–C, respectively). The scatter plots of all 3 markers suggested an inverse hyperbolic relationship with serum uromodulin concentration.

FIGURE 2.

Box and whisker plot of the serum uromodulin concentrations in 25 healthy control dogs and 49 dogs with chronic kidney disease. The CKD dogs were staged according to their serum creatinine concentrations based on the criteria determined by the IRIS guidelines (Stage 1, n = 23; Stage 2, n = 20; Stage 3‐4, n = 6). The box extends from the 25th to 75th percentile, and the horizontal line is drawn at the median value. The whiskers represent the range of the measured concentrations. The adjusted P‐values corrected for multiple comparisons are shown. CKD, chronic kidney disease; IRIS, International Renal Interest Society; ns, not significant

TABLE 1.

Blood urea nitrogen, creatinine, symmetric dimethylarginine, and uromodulin concentrations in 25 healthy dogs and 49 dogs with chronic kidney disease

Variables	Control (n = 25)	CKD Stage 1 (n = 23)	CKD Stage 2 (n = 20)	CKD Stages 3‐4 (n = 6)
BUN (mg/dL)	16.4 (5.1‐25.9)	20.4 (12.8‐48.2)	34.4a (23.4‐54.3)	66.4a (48.3‐122.4)
Creatinine (mg/dL)	0.91 (0.65‐1.33)	0.90 (0.67‐1.30)	1.89a (1.42‐2.76)	4.20a (3.09‐5.59)
SDMA (μg/dL)	8 (5‐13)	10 (6‐13)	22a (13‐59)	53a (40‐69)
Uromodulin (ng/dL)	94.98 (51.31‐267.70)	49.22a (17.68‐91.16)	42.44a (3.51‐73.63)	33.58a (16.60‐51.08)

Note: All values are presented as medians and range in brackets. The variables were analyzed using the Kruskal‐Wallis test and post‐hoc Dunn's multiple comparison test. The superscript (a) indicates significant difference between the controls and dogs with Stage 1, 2, and 3‐4 CKD (P < .05).

Abbreviations: BUN, blood urea nitrogen; CKD, chronic kidney disease; SDMA, symmetric dimethylarginine.

FIGURE 3.

Scatterplot illustrating serum uromodulin concentrations according to the blood urea nitrogen (A), serum creatinine (B), and serum symmetric dimethylarginine (C) concentrations. The relationship between serum uromodulin concentrations and each renal parameter was analyzed using Spearman's rank correlation coefficient. Serum uromodulin showed a statistically significant but weak negative correlation with all the three markers (A, B, and C). r = Spearman's rank correlation coefficient. BUN, blood urea nitrogen; SDMA, symmetric dimethylarginine

3.4. Diagnostic performance of serum uromodulin concentration in differentiating between CKD dogs and healthy controls

The results of the ROC analysis for discriminating CKD dogs (Stages 1‐4) from healthy controls (Figure 4) demonstrated AUCs of 0.97 (95% CI, 0.94‐1.00) for serum uromodulin concentration and 0.87 (95% CI, 0.79‐0.95) for SDMA. The AUC of serum uromodulin concentration was significantly higher than that of SDMA (P = .01 by DeLong's test).

FIGURE 4.

Receiver operating characteristic curve for serum uromodulin and symmetric dimethylarginine to differentiate between CKD dogs (Stages 1‐4) and healthy controls. AUC of serum uromodulin = 0.97 (95% CI, 0.94‐1.00); AUC of SDMA = 0.87 (95% CI, 0.79‐0.95). AUC, area under the curve; CI, confidence interval; CKD, chronic kidney disease; SDMA, symmetric dimethylarginine

In the ROC curve in distinguishing dogs with Stage 1 CKD from the controls (Figure 5), the AUC was 0.95 (95% CI, 0.89‐1.00) for serum uromodulin concentration and 0.72 (95% CI, 0.58‐0.87) for SDMA, being significantly higher for serum uromodulin concentration (P = .001 by DeLong's test). With an optimal cut‐off of 72.77 ng/mL for uromodulin (Youden's index = 0.79), the sensitivity and specificity were 86.96% (95% CI, 66.41%‐97.22%) and 92.00% (95% CI, 73.97%‐99.02%), respectively.

FIGURE 5.

Receiver operating characteristic curve of serum uromodulin to differentiate between dogs with Stage 1 CKD and healthy controls. AUC of serum uromodulin = 0.95 (95% CI, 0.89‐1.00); AUC of SDMA = 0.72 (95% CI, 0.58‐0.87). Using a cut‐off of 72.77 ng/mL for serum uromodulin, the sensitivity was 86.96% (95% CI, 66.41%‐97.22%) and specificity was 92.00% (95% CI, 73.97%‐99.02%). AUC, area under the curve; CI, confidence interval; CKD, chronic kidney disease

4. DISCUSSION

Uromodulin or Tamm‐Horsfall protein is the most abundant protein in urine. ⁶ Small amounts of uromodulin also are detected in blood, and recent studies have suggested its value in CKD diagnosis in humans. ¹¹ , ¹² , ¹³ , ²¹ Our results illustrated that serum uromodulin concentrations were lower in dogs diagnosed with CKD than in the healthy controls. Moreover, the uromodulin concentrations negatively correlated with conventional renal markers such as BUN, creatinine, and SDMA. Our study also indicates that serum uromodulin concentration has good diagnostic potential in identifying early‐stage CKD.

We found that uromodulin is present in the serum of dogs as in humans. Circulating uromodulin concentrations were much lower than urine uromodulin concentrations as found in earlier investigations in dogs (i.e., the urine uromodulin concentrations were >1000 times higher than serum uromodulin concentrations). ¹⁵ A considerable difference between the urine and serum uromodulin concentrations also was observed in humans. ⁶ Immunolocalization studies in humans consistently indicated the presence of uromodulin in the basolateral membranes of the TAL cells, ¹³ , ²² , ²³ suggesting it as a route by which uromodulin enters serum. Moreover, the notion that uromodulin exists in monomers in serum ²⁴ while forming polymers of various sizes in urine corroborates that uromodulin in serum has an independent excretion route, rather than reuptake from urine. ²⁵ Uromodulin in canine serum also may come from the basolateral release of the protein from TAL cells. However, additional studies are warranted to confirm the presence of this excretory route, as well as its mechanism.

When comparing the control and CKD groups, serum uromodulin concentrations were lower in all CKD dogs (Stages 1‐4) as compared to controls. This finding corroborated the findings of previous studies that evaluated urinary uromodulin in dogs with CKD, in which the concentrations were low in dogs with CKD. ¹⁵ , ¹⁷ , ¹⁸ Alteration of uromodulin in dogs with CKD shows the same pattern in blood and urine, presumably because of decreased functional renal mass. The absolute production of uromodulin may decrease and result in a decrease in both apical and basolateral release of the protein, leading to lower uromodulin concentrations in urine and serum. Thus, a decrease in uromodulin may indirectly reflect renal tubulointerstitial abnormalities, and differ from conventional renal markers used for CKD (eg, creatinine, SDMA), which mainly reflect glomerular function.

According to a study in humans that measured serum and urinary uromodulin concentrations in the same population, serum concentrations showed stronger correlation with estimated GFR than did urine concentrations. This observation suggests that serum uromodulin concentration may be a more sensitive marker for kidney function than urinary uromodulin concentration. ²⁶ Furthermore, urinary uromodulin forms aggregates of different sizes in various urine compositions, ¹⁴ possibly hindering accurate measurement of its concentration. In our preliminary study, which tested the long‐term stability of urinary uromodulin stored at −80°C, uromodulin concentrations in stored urine samples showed significant differences from baseline concentrations, with no consistent pattern in the extent or increases or decreases in uromodulin among tested urine samples. This pattern of deviation from baseline partly may have been because of the aggregating characteristics of urinary uromodulin. That is, the aggregation of uromodulin may have changed its concentration to a different extent in each urine sample, according to changes in urine conditions (e.g., pH, ion composition; Figure S2).

In our study, a decrease in serum uromodulin concentration was observed in Stage 1 CKD patients when compared to the healthy controls. In addition, serum uromodulin concentration outperformed SDMA (AUC, 0.95 vs 0.72; P = .001 by DeLong's test; Figure 5) in distinguishing Stage 1 CKD patients from the healthy controls in the ROC analysis, overall suggesting its diagnostic advantage in detecting dogs with early‐stage CKD over currently used renal biomarkers.

In veterinary medicine, creatinine and SDMA generally are used to diagnose CKD and monitor its progression. ¹⁹ Both are surrogate markers of GFR, wherein their concentrations increase with decreased GFR. Symmetric dimethylarginine is considered to have a relative advantage in allowing earlier detection of deterioration in kidney function. ²⁷ , ²⁸ However, because SDMA has an exponential relationship with GFR, as does creatinine, it is suggested to have the same limitation in detecting early decreases in GFR. ²⁹ Additionally, the initial decrease in GFR caused by decreased nephron mass is offset by an increase in glomerular filtration for each nephron. Because of this compensatory effect, there is a limit to initial identification of renal damage using these GFR markers.

Similar results were reported in studies conducted in humans. Similar to our study, serum uromodulin concentration was the only marker that was significantly different in its concentration between non‐CKD and early‐stage CKD patients. ¹³ , ²¹ Our data illustrated that serum uromodulin concentration in dogs could assist in detecting dogs with early‐stage CKD as compared to the currently used diagnostic tests.

No significant difference was found in serum uromodulin concentrations among the dogs with CKD of different stages, which was in contrast to results in humans, where a gradual decrease in uromodulin concentration was observed with advanced stages of CKD. ¹¹ , ¹³ , ²¹ In part, our results may have resulted from possible discordance between the extent of tubular destruction and GFR reduction. In other words, the extent to which the production of uromodulin decreases may have varied among dogs presented with the same CKD stage. However, these conclusions should be carefully considered because of the small number of dogs with advanced‐stage CKD in our study, which limits evaluation. Thus, additional studies using a larger sample size are warranted.

Our study had some limitations. First, direct measurement of GFR was not performed in the CKD diagnosis and staging, which may have led to the inaccurate classification of some CKD dogs as controls or CKD cases. The relatively broad range of uromodulin concentrations in the control dogs could have been caused by such misclassification. However, despite the lack of GFR assessment, all of the CKD dogs had additional clinical evidence of CKD, such as persistent renal proteinuria or relevant ultrasonographic findings. Second, some of the dogs in the Stage 1 CKD group had only ultrasonographic findings, and others had both ultrasonographic findings and persistent proteinuria. Although ultrasonography is the recommended imaging test in veterinary medicine to find structural renal changes, diagnosing CKD solely based on abdominal ultrasonography has its limitations because of the subjective nature of ultrasonography. Therefore, the possibility of misclassification of the Stage 1 CKD group cannot be completely ruled out, and additional studies using other indicators to identify early kidney damage, such as GFR measurements and novel biomarkers, are needed. Third, some dogs in the CKD group had concurrent diseases that were not associated with the kidneys. We excluded dogs with disorders that could affect uromodulin production and release. Because production and release of uromodulin is limited to the kidneys, the effect of other conditions can be considered negligible. However, additional investigations are needed to evaluate whether non‐renal abnormalities affect serum uromodulin concentrations in dogs. Fourth, the control group was significantly younger than the CKD group, which will necessitate additional studies regarding the effect of age on serum uromodulin concentrations. Finally, our study comprised a relatively small number of dogs in each CKD stage, particularly in Stages 3 and 4. Therefore, additional studies with larger patient groups are warranted.

In conclusion, our study determined that serum uromodulin concentrations decreased in dogs with CKD. Our data also provided evidence that serum uromodulin may have an advantage over the conventional renal markers in detecting early‐stage CKD. The conventionally‐used biomarkers mainly reflect GFR as representative of renal function. On the other hand, serum uromodulin, because of its tubular origin, may provide a novel perspective in the evaluation of kidney function. However, studies on the effects of non‐renal factors on serum uromodulin concentrations and the direct link between tubulointerstitial abnormalities and circulating uromodulin concentrations are warranted.

CONFLICT OF INTEREST DECLARATION

Authors declare no conflict of interest.

OFF‐LABEL ANTIMICROBIAL DECLARATION

Authors declare no off‐label use of antimicrobials.

INSTITUTIONAL ANIMAL CARE AND USE COMMITTEE (IACUC) OR OTHER APPROVAL DECLARATION

This study was reviewed and approved by the IACUC of Seoul National University (SNU‐201123‐3‐1).

HUMAN ETHICS APPROVAL DECLARATION

Authors declare human ethics approval was not needed for this study.

Supporting information

Figure S1. Effect of sample storage on serum uromodulin concentration

Figure S2. Effect of sample storage on urinary uromodulin concentration

Seo D, Yang Y, Hwang S‐H, et al. Serum uromodulin in dogs with chronic kidney disease. J Vet Intern Med. 2022;36(6):2071‐2078. doi: 10.1111/jvim.16579