Abstract
Objectives
The aims of this study were to validate a semi-automated high-resolution electrophoretic technique to quantify urinary albumin in healthy and diseased cats, and to evaluate its diagnostic performance in cases of proteinuria and renal diseases.
Methods
Urine samples were collected from 88 cats (healthy; chronic kidney disease [CKD]; lower urinary tract disease [LUTD]; non-urinary tract diseases [OTHER]). Urine samples were routinely analysed and high-resolution electrophoresis (HRE) was performed. Within-assay and between-assay variability, linearity, accuracy, recovery and the lowest detectable and quantifiable bands were calculated. Receiver operating curve (ROC) analysis was also performed.
Results
All coefficients of variation were <10%, percentage recovery was between 97% and 109% with a high linearity (r = 0.99). HRE allowed the visualisation of a faint band of albumin and a diffused band between alpha and beta zones in healthy cats, while profiles from diseased cats were variable. Albumin (mg/dl) and urine albumin:creatinine ratio (UAC) were significantly (P <0.05) different between healthy and diseased cats. After ROC analysis, UAC values of 0.035 and 0.074 had a high sensitivity and high specificity, respectively, to classify proteinuria and identify borderline proteinuric cats. Moreover, a UAC of 0.017 had a high sensitivity in distinguishing between healthy and diseased cats. However, UAC was not able to distinguish between renal (CKD) and non-renal diseases (LUTD/OTHER), probably owing to the pathophysiology of CKD in cats, which is characterised by low-grade proteinuria and less glomerular involvement than in dogs.
Conclusions and relevance
HRE is an accurate and precise method that could be used to measure albuminuria in cats. UAC was useful to correctly classify proteinuria and to discriminate between healthy and diseased cats. HRE might also provide additional information on urine proteins with a profile of all proteins (albumin and globulins) to aid clinicians in the diagnosis of diseases characterised by proteinuria.
Introduction
Very low concentrations of proteins are physiologically present in the urine of healthy organisms, and all these proteins are collectively termed ‘proteinuria’. The increase of proteinuria is mainly a result of two different mechanisms: loss of selective glomerular filtration and impaired tubular resorption. Traces of albumin are physiologically present in the urine of healthy animals. In dogs, an albumin concentration >30 mg/dl is defined as overt albuminuria,1,2 while microalbuminuria is defined as a concentrat-ion of albumin in urine >1 mg/dl but below the limit of detection of semi-quantitative screening tests (30 mg/dl). Albuminuria can also be expressed as the urinary albumin:creatinine ratio (UAC) to adjust for differences in urine concentration and volume, as it is used for proteinuria (urine protein:creatinine ratio; UPC). In human medicine UAC values <0.03 are considered normal, microalbuminuria is defined between 0.03 and 0.3, and values >0.3 are considered as overt albuminuria. 3 Microalbuminuria is an important biomarker of glomerular damage and its role in predicting clinical endpoint (cardiovascular events, renal events and mortality) and monitoring diabetic nephropathy, cardiovascular disease and hypertension is well documented in people.3–6
In dogs and cats, microalbuminuria has been evaluated in correlation with different pathologies and clinical conditions, such as chronic kidney disease (CKD) and renal failure,7,8 leishmaniosis, 9 hypertension,10,11 hyperadrenocorticism,12,13 severe inflammatory response syndrome, 14 critically ill dogs in intensive care units,15,16 diabetes mellitus 17 and a variety of other systemic disorders.1,18 However, only Syme et al 8 reported the UAC calculated in a population of healthy cats and a reference interval for cats using an appropriate number of individuals has not been defined so far. In addition, the methods used to quantify albuminuria present some disadvantages. In particular, ELISA8,19,20 is time consuming and not applicable in routine clinical screening. The immunoturbidimetric assay21,22 is based on the interaction with an anti-human albumin polyclonal antibody, and the early renal damage (ERD) immunoassay (Healthscreen; Heska) is semi-quantitative.17,23,24
In this context, there is a need for a fast and reliable method applicable in routine clinical diagnostics. Therefore, the aims of this study were to validate a semiautomatic electrophoretic method on high-resolution agarose gel (high-reolution electrophoresis; HRE) to quantify albuminuria in cat urine and test its clinical applications in routine analysis.
Materials and methods
Study design
This study was performed using 88 urine samples submitted to the Veterinary Clinical Pathology Service of the Veterinary Teaching Hospital of the University of Bologna for a variety of clinical conditions.
The control group (H) comprised 22 cats considered healthy on the basis of history and physical examination and without any other abnormality in serum biochemistry and urinalysis. The diseased cats were assigned to one of three categories of disease based on history, clinical signs, clinicopathological and imaging data as follows: CKD (n = 32), lower urinary tract disease (LUTD) (n = 18), other disease without involvement of the urinary system (OTHER) (n = 16; inflammatory/infectious diseases n = 6; orthopaedic/neurological diseases n = 4; gastrointestinal diseases n = 3; neoplastic diseases n = 3). Cats with clinical and cytological evidence of urinary tract infection were included in the LUTD group. CKD cats were diagnosed and staged following International Renal Interest Society (IRIS) guidelines. 25
All experimental procedures were approved by the Institutional Scientific Ethical Committee for Animal Testing of the University of Bologna (approval number 8-72-2012; date of approval 1 October 2012) and the owners signed an informed consent form before inclusion of their animal in the study.
Urinalysis
Complete urinalysis, including measurement of the specific gravity by a refractometer and semi-quantitative dipstick test (Combur10Test; Roche Diagnostic), was performed on all samples. After centrifugation at 1500 × g for 10 mins and microscopic sediment analysis at low (× 100) and high (× 400) power fields, the supernatants were divided into two aliquots and stored at −80°C for subsequent analysis. Urine total proteins (UTP) and creatinine were determined using commercial kits (Urinary/CSF Protein, OSR6170, and Creatinine OSR6178 [Olympus/Beckman Coulter]) on an automated chemistry analyser (AU 400; Olympus/Beckman Coulter) allowing the calculation of the UPC.
HRE
All urine samples were analysed with HRE on 0.8% agarose gel at pH 8.6 (HydraGel HR 15; Sebia) in combination with the semi-automated system Hydrasys (Sebia), according to the manufacturer’s instructions. The gels were dried, stained with acid violet solution and band staining density acquired by the Epson Perfection V700 photo scanner/densitometer. The obtained pherograms were analysed by the Phoresis software (version 6.1.2). The relative percentage and absolute concentration of albumin were calculated based on the density determined by the densitometer. As reported by the manufacturer, the limit of detection of this technique for serum albumin is 1.5–2.0 mg/dl (0.15–0.2 μg/band), with concentrations linear up to at least 5.8 g/dl and optimal protein concentration of 200 mg/dl. Urine samples with higher protein concentrations were diluted in order to achieve the desired optimal concentration.
HRE validation
To validate the HRE method for albumin quantification in cat urine, precision (within and between assay variability), linearity, accuracy and analytical sensitivity were calculated. Within-assay variability was assessed on five cat urine samples with different albumin concentrations ranging from 7.4–67.7 mg/dl (7.4 mg/dl; 13.4 mg/dl; 23.5 mg/dl; 38.8 mg/dl; 67.7 mg/dl). For within-assay, samples were run six times on the same gel. Between-assay variability was evaluated measuring four samples (7.4 mg/dl; 13.4 mg/dl; 23.5 mg/dl; 67.7 mg/dl) in duplicate on five different days.
Linearity was tested by serial dilution of one sample (79 mg/dl) until reaching the expected albumin concentration under the limit of detection (LOD) defined by the manufacturer (1.5 mg/dl). All samples were analysed in duplicate. The analytical sensitivity was obtained by the definition of the lowest detectable band (the lowest concentration with weak or barely visible albumin band on the gel) and the lowest quantifiable band (the lowest concentration with a quantifiable peak in the pherogram).
To evaluate accuracy, in the absence of a reference method for cat urine, a percentage recovery study was made. Three solutions were prepared by adding to 100 µl cat urine, containing 24.5 mg/dl albumin, 100 µl saline solution (12.25 mg/dl of expected albumin) (L) and 100 µl of two cat urine samples reaching 32.9 mg/dl (M) and 53.8 mg/dl (H) of expected albumin. Five replicates were made for each solution.
Statistical analysis
Statistical analysis was performed with MedCalc 11.3.3.0. Data are reported as median (range). Precision was evaluated by the calculation of the coefficient of variation (CV) as follows: CV = (SD/mean) X 100. Correlation and regression analysis were performed on expected and observed values for albumin and on UPC and UAC. According to IRIS CKD staging guidelines, 25 cats with UPC >0.4, 0.2–0.4 and <0.2 were classified as proteinuric, borderline proteinuric or non-proteinuric, respectively.
Receiver operator curves (ROCs) were generated to evaluate sensitivity (SE), specificity (SP), positive predictive values (PPV) and negative predictive values (NPV), and positive and negative likelihood ratios (LR+; LR–) of UAC and dipstick to classify proteinuria, using as a reference cut-off UPC 0.2 (according to IRIS guidelines). ROC curves were also generated to evaluate SE, SP, PPV, NPV, LR+ and LR– of UAC, UPC and dipstick to classify patients as healthy or diseased.
Non-parametric statistics (Kruskal–Wallis with post-hoc test for pairwise comparison of groups) were performed to evaluate differences between groups (H, CKD, LUTD and OTHER) for UPC, UAC and age. Values of albumin <5 mg/dl (defined as lowest quantifiable band) were set at 2.5 mg/dl and UAC and albumin percentage values were calculated according to this value. Significance was set at P <0.05 unless otherwise specified.
Results
HRE validation and albumin quantification
HRE had within-assay CVs of 2.6% (67.7 mg/dl), 5.8% (38.8 mg/dl), 3.4% (23.5 mg/dl), 7.4% (13.4 mg/dl) and 7.6% (7.4 mg/dl), and between-assay CVs of 3.5% (67.7 mg/dl), 1.2% (23.5 mg/dl), 7.4% (13.4 mg/dl) and 9.0% (7.4 mg/dl). Percentage recoveries of 97%, 93% and 109% for high (H), medium (M) and low (L) concentrations, respectively, were calculated and a significant correlation was found between measured and expected albumin concentration (r = 0.99; P <0.001). The lowest detectable band of 1.25 mg/dl was set by visual inspection of the gels and the lowest quantifiable band of 5 mg/dl was determined by densitometric analysis.
After the validation, HRE was performed on the 88 urine samples. Nineteen of the 32 urine samples with UTP <40 mg/dl and 6/56 urine samples with UTP >40 had a barely visible protein band corresponding to an albumin concentration lower than the lowest quantifiable band.
Diagnostic performances of HRE and clinical application
Twenty-two cats were included in the study as the healthy group (Table 1 and Table 1 in the supplementary material). Median age was 22.5 (6–168) months and median UPC was 0.12 (0.06–0.32). HRE allowed the visualisation of a faint band of albumin and a diffused band between α and β zones (Figure 1a) in most samples. Only in nine samples (41%) was albumin concentration quantifiable, the median concentration was <5 mg/dl (<5–29.8) and the calculated median UAC was 0.011 (0.004–0.069). No differences were found between males and females.
Table 1.
Age and results of urine specific gravity (USG), proteinuria and albuminuria for the samples from healthy and diseased cats. Data are reported as median (minimum–maximum)
| Cats | Age (months) | sCREA (mg/dl) | sUREA (mg/dl) | USG (mg/dl) | UTP (mg/dl) | uCREA (mg/dl) | UPC | ALB (%) | ALB (mg/dl) | UAC |
|---|---|---|---|---|---|---|---|---|---|---|
| RI | 0.8–1.8 | 15–60 | >1.035 | <0.2 | ||||||
| Healthy (n = 22) |
22.5 (6–168) |
1.5 (1.03–1.8) |
48.4 (32–86) |
1.071 (1.036–1.090) |
35.0 (16–159.6) |
344 (162–766) |
0.12 (0.06–0.32) |
9.8 (5.5–42) |
2.5 (2.5–29.8) |
0.011 (0.004–0.069) |
| CKD (n = 32) |
163 (24–252) |
3.55 (1.33–12.7) |
169 (57.8–501.7) |
1.017 (1.006–1.068) |
75.4 (12.8–1956.3) |
92 (23–583) |
0.63 (0.10–16.15) |
22.3 (3.9–55.3) |
11.4 (2.5–962.0) |
0.117 (0.009–7.057) |
| LUTD (n = 18) |
132 (36–268) |
1.75 (1.37–10.78) |
108.9 (41.1–479.5) |
1.029 (1.008–1.072) |
68.3 (30.6–280.2) |
162 (91–672) |
0.70 (0.07–4.59) |
20.7 (5.3–59.1) |
20.2 (2.5–134.2) |
0.184 (0.010–1.487) |
| OTHER (n = 16) |
102 (24–192) |
1.59 (0.96–3.07) |
53.2 (32.3–199.2) |
1.058 (1.017–1.080) |
74.9 (27.1–881.8) |
338 (94–813) |
0.21 (0.09–1.52) |
20.7 (11.3–44.9) |
15.2 (6.6–147.2) |
0.052 (0.013–0.465) |
sCREA = serum creatinine; sUREA = serum urea; USG = urine specific gravity; UTP = urine total protein; UPC = urine protein:creatinine ratio; ALB = albumin; UAC = urine albumin:creatinine ratio; RI = laboratory reference intervals; CKD = chronic kidney disease; LUTD = lower urinary tract disease; OTHER = other diseases not involving the urinary system
Figure 1.
(a,b) Representative high-resolution electrophoresis (HRE) gels and (c) pherogram of urine samples from (a) healthy and (b,c) diseased cats. (a) HRE of healthy samples; (b) HRE of diseased samples (chronic kidney disease [CKD] lanes 3, 7; lower urinary tract disease lanes 5, 9–15; other diseases not involving the urinary system 1, 2, 4, 6) and a healthy sample (lane 8); (c) as an example, pherogram of lane 3 (CKD)
Sixty-six samples were included in the diseased group (Table 1 and Table 2 in the supplementary material) with a median age of 132 (24–268) months and median UPC of 0.51 (0.07–16.15). These samples had variable electrophoretic profiles (Figure 1b). In particular, the albumin band was evident and quantifiable in 54 samples (82%) and the median albumin concentration and UAC were 12 mg/dl (<5–962) and 0.111 (0.009–7.056), respectively, both significantly higher than in healthy cats (P <0.01). Moreover, it was also possible to separate and identify the α, β and γ globulin zones (Figure 1c).
Albuminuria vs proteinuria
Using the IRIS guidelines, the 88 samples were classified as non-proteinuric (UPC <0.2; n = 38), borderline proteinuric (UPC 0.2–0.4; n = 11) and proteinuric (UPC >0.4; n = 39). UAC was significantly correlated to UPC (r = 0.967; P <0.0001). ROC curves were generated to evaluate diagnostic performance of UAC and dipstick (considered as positive for trace results) to classify proteinuria correctly, considering as positive samples with UPC >0.2 (borderline proteinuric and proteinuric). Area under curve (AUC) for UAC (0.939) was significantly greater (P <0.01) than dipstick (0.537). Table 2 shows SE, SP, PPV and NPV for the best criteria calculated by the ROCs for UAC and dipstick. For UAC, two different values demonstrated high sensitivity (0.035) or high specificity (0.074) for overt or borderline proteinuria, while dipstick showed very low specificity, owing to the high number of false-positive results. Using these two cut-off values, as reported in Table 3, 10 non-proteinuric samples had abnormal UAC values.
Table 2.
Sensitivity, specificity, positive (PPV) and negative predictive values (NPV), and positive (LR+) and negative likelihood ratios (LR–) for the best criteria calculated for urine albumin:creatinine ratio (UAC; 0.035 and 0.07) and dipstick (30 mg/dl, lowest level of positivity) using urine protein:creatinine ratio 0.2 as reference cut-off
| UAC 0.035 |
UAC 0.074 |
Dipstick 30 |
|
|---|---|---|---|
| Sensitivity (%) | 94.00 | 76 | 82.00 |
| Specificity (%) | 76.30 | 100 | 23.70 |
| PPV (%) | 82.50 | 97.4 | 58.60 |
| NPV (%) | 90.30 | 74.0 | 50.00 |
| LR+ | 3.97 | – | 1.07 |
| LR– | 0.079 | 0.26 | 0.76 |
Table 3.
Sample distribution in the different classes of proteinuria, according to the International Renal Interest Society, and albuminuria, according to the values determined in the present study
| NA |
MA |
A |
Total | ||
|---|---|---|---|---|---|
| UAC <0.035 | 0.035 <UAC <0.074 | UAC >0.074 | |||
| NP | UPC<0.2 | 28 | 9 | 1 | 38 |
| BP | 0.2<UPC<0.4 | 2 | 4 | 5 | 11 |
| P | UPC>0.4 | 1 | 6 | 32 | 39 |
| Total | 31 | 19 | 38 | 88 | |
NA = non-albuminuric; MA = microalbuminuric; A = overt albuminuric; NP = non-proteinuric; BP = borderline proteinuric; P = proteinuric
Albuminuria vs diseases
According to serum creatinine concentration and IRIS guidelines, CKD cats were staged as follows: stage I (n = 3), stage II (n = 8), stage III (n = 9) and stage IV (n = 12). CKD cats were significantly older with a median age of 163 (24–252) months (P <0.01), and had significantly increased UPC (0.63 [0.10–16.15]), albumin concentration (11.4 [<5–962] mg/dl) and UAC (0.117 [0.009–7.057]) than the H group (P <0.01). LUTD cats were significantly older than H group cats with a median age of 132 (36–268) months (P <0.01), and had significantly increased UPC (0.7 [0.07–4.59]). Albumin concentration (20.2 [<5–134.2] mg/dl) and UAC (0.184 [0.010–1.487]) in LUTD cats were significantly higher than H group (P <0.01). OTHER cats were significantly older than H and younger than CKD groups with a median age of 102 (24–192) months (P <0.05), and had UPC (0.21 [0.09–1.52]) significantly higher than the H group (P <0.05). Albumin concentration (15.2 [6.6–147.2] mg/dl) and UAC (0.052 [0.013–0.465]) in this group were higher than in H group (P <0.05). Figure 2 shows the comparison of UAC among H, CKD, LUTD and OTHER groups.
Figure 2.
Urinary albumin:creatinine ratio (UAC) values for healthy and diseased groups. Different lower case letters indicate significant difference (P <0.05). UAC values >2 were excluded from graphic visualisation. UAC = urine albumin:creatinine ratio; CKD = chronic kidney disease; H = healthy; LUTD = lower urinary tract disease; OTHER = other diseases not involving the urinary system
ROCs were generated to evaluate and compare dipstick, UAC and UPC to classify cats correctly. The performance of UPC, UAC and dipstick is summarised in Table 4. ROC comparison for the classification of patients as healthy or diseased is reported in Figure 3a. AUCs for UPC (0.909), UAC (0.877) and urine specific gravity (USG; 0.865) were not significantly different from each other but significantly different from AUC for dipstick (0.631) (P <0.001). For UPC, the best criterion was set at 0.165, while for UAC two different values gave high sensitivity (0.017) or high specificity (0.074). Dipstick showed very low specificity. Lastly, ROC comparison was generated to discriminate cats as affected by CKD or LUTD/OTHER diseases (Figure 3b). AUCs for UPC (0.665) and USG (0.740) were not significantly different from each other, but were both significantly different from AUC for dipstick (0.525) and UAC (0.585) (P <0.05).
Table 4.
Sensitivity, specificity, positive (PPV) and negative predictive values (NPV) and positive (LR+) and negative likelihood ratios (LR–) for the best criteria calculated for urine albumin:creatinine ratio (UAC; 0.017 and 0.07), urine protein:creatinine ratio (UPC; 0.165 and 0.2) and dipstick (30 mg/dl, lowest level of positivity) to distinguish between healthy and diseased cats
| UPC >0.165 | UPC >0.2 | UAC >0.017 | UAC >0.074 | Dipstick >30 | |
|---|---|---|---|---|---|
| Sensitivity (%) | 84.8 | 72.7 | 87.9 | 59.1 | 87.9 |
| Specificity (%) | 86.4 | 90.9 | 63.6 | 100.0 | 22.7 |
| PPV (%) | 96.6 | 96.0 | 87.9 | 100.0 | 70.0 |
| NPV (%) | 66.7 | 52.6 | 63.6 | 44 | 5.6 |
| LR+ | 6.2 | 16.0 | 2.42 | – | 1.14 |
| LR– | 0.18 | 0.29 | 0.19 | 0.041 | 0.53 |
Figure 3.
Receiver operator curve (ROC) comparison for urine protein:creatinine ratio (UPC), urine albumin:creatinine ratio (UAC), dipstick (STICK) and urine specific gravity (USG) to classify cats (a) as healthy or diseased and (b) as renal (chronic kidney disease) or non-renal (lower urinary tract disease [LUTD]/other disease without involvement of the urinary system [OTHER])
Discussion
Several studies suggest that even mild proteinuria/albuminuria can be indicative of the severity of disease and predict poor outcome;8,19,26,27 in contrast, the difficulty of albumin quantification with dipsticks and the absence of a validated method for cat urine highlight the importance of a sensitive, accurate and precise analytical method for albumin quantification in urine. In this context, our work aimed to validate an electrophoretic method for albumin quantification and to evaluate its clinical application in routine urinalysis.
Validation of HRE
The electrophoretic method validated in the present study is a semi-automatic technique and the obtained CVs were <10%. As a general rule, CVs for automated assays should be <5% and for manual assays <10%. 28 Therefore, considering that HRE is a semi-automatic technique, the variability determined in the present study can be considered acceptable. The method was linear in the tested range of concentrations and the results of the recovery test can be considered acceptable for routine diagnostic laboratory use. 28 Regarding the LOD, it was difficult to quantify blank (saline solution, NaCl 0.9%) by densitometry. The scanner reads impurities on the gel generating irregular profiles, not related to the protein content of the samples; the same analytical problem was also reported by other authors. 29 As a consequence, the LOD was set at 1.25 mg/dl of expected albumin, as at this concentration, close to the LOD defined by the manufacturer, a faint band of the protein was optically visible on the gel, although not quantifiable by the software. Furthermore, the software allows the technician/clinician to reduce the interferences and evaluate/validate the correspondence between visual inspection of the gel and the densitometric profile. Therefore, the ease and rapidity of use, the high reproducibility and accuracy and wide range of linearity, combined with the final evaluation of the pherogram by the clinician/technician, allow us to conclude that HRE is a reliable method for quantifying albuminuria in cats.
Diagnostic performance of HRE and clinical application
HRE was useful to highlight that urine from healthy cats is characterised by very low albumin concentrations. This result is confirmatory of data reported by Ferlizza et al, 30 who pointed out that the most abundant proteins in urine from healthy cats were cauxin (at greater concentration in entire males) and uromodulin, both produced by healthy tubular cells and specifically secreted in urine. Diseased cats presented a greater variability in the electrophoretic profiles with increasing concentration/density of the albumin band and of α, β and γ zones. The quantification of the globulins was out of the scope of the present paper and they were not correlated with the presence/absence of CKD or other diseases. Nevertheless, Giori et al 29 correlated the reduced percentage of α1-globulins to the glomerular origin of proteinuria in dogs. These results associated with a clear difference between the pherograms from healthy and diseased cats, allowed us to hypothesise the potential usefulness of HRE to evaluate albumin, α, β and γ globulins to classify proteinuria correctly in both dogs and cats.
In the present study patients were arbitrarily classified according to IRIS guidelines for proteinuria. Although these guidelines have been created to classify patients with CKD, the authors considered the categories proposed as useful to classify proteinuria in cats, regardless of the presence of CKD, as cats frequently present low-grade proteinuria in the clinical practice.2,16,17 To evaluate the ability of HRE to classify proteinuria correctly, a comparison was made with dipstick, using an UPC value of 0.2 as a reference. Dipstick tests showed high sensitivity but very low specificity, owing to the high number of false-positive results. However, as suggested by Zatelli et al 31 for dogs, the dipsticks should always be interpreted in the light of the USG and, in the present study, most of the false-positive samples also had high USG values (>1.035). Moreover, healthy entire male cats secrete cauxin into urine, which has also been demonstrated to be a cause of false-positive proteinuria. 32 The lower UAC value, 0.035, was close to the cut-off used for human microalbuminuria (0.03) and using this criterion nine non-proteinuric samples (UPC <0.2) were classified as microalbuminuric. The second one, 0.074, is similar to data reported by Syme et al, 8 who reported an UAC of 0.08 as the upper limit of the reference interval in healthy cats, and lower than the value used in human medicine for overt albuminuria (0.3). Unlike the Syme et al report, the population of healthy cats included in the present study was not age-matched with those in the CKD and non-CKD. 8 However, the similarity between the data reported by Syme et al 8 and the cut-off reported in the present study, in spite of the different age of the healthy cats (median 12 vs 1.9 years) might be confirmatory of the clinical reliability of the analysis and suggests that UAC is not influenced by age alone. With regard to the differences between cats and humans, and also between cats and dogs, this could be explained by cats presenting with lower quantities of proteinuria/albuminuria due to a different pathophysiology in cases of renal diseases. These preliminary results, even if obtained on a limited number of cats, suggest that a UAC of 0.035 as a cut-off for microalbuminuria and a UAC of 0.074 as a cut-off for overt albuminuria, if confirmed by future studies with greater numbers of healthy cats, could be considered complementary to UPC in estimating and identifying pathological low-grade proteinuria.
With regard to the comparison among groups, both UPC and UAC were able to discriminate healthy cats from diseased ones. Interestingly, although they had renal and urinary tract involvement, CKD and LUTD cats did not show different proteinuria and albuminuria values than the OTHER group. In our study, diseases apparently not involving the kidney or the urinary tract were characterised by albuminuria higher than healthy cats. The pathogenesis of increased urine albumin concentration in these diseases has not been investigated in this study; however, different conditions such as hypertension, previous treatments or even subclinical renal involvement could be hypothesised. As previously suggested, proteinuria and albuminuria should be classified as renal or pre-/postrenal and LUTD is considered a cause of postrenal albuminuria in cats.2,33 To our knowledge, no data are present in the literature on albuminuria in cats affected by LUTD, and these are the first UAC values reported. The best criteria calculated with ROC analysis to distinguish between healthy and diseased cats were, for UPC, 0.165, close to the well-established value of 0.2, and for UAC, 0.017, lower than the previously calculated value of 0.035. However, after the subsequent ROC analysis, UAC was not able to distinguish between CKD and non-CKD diseases. It is well known that cats affected by CKD present with lower values of proteinuria and albuminuria than dogs,34,35 and that even low UPC and UAC ratios are correlated with poor outcome. 8 Using the calculated cut-offs (UPC 0.165 and UAC 0.017) within the different groups of diseases, the sensitivity to detect low levels of albuminuria/proteinuria is increased, suggesting that these values (UPC >0.165; UAC >0.017) could be reliable for early confirmation of disease with possible renal involvement. The low level of albuminuria/proteinuria in cats with CKD seems to be related to chronic tubulointerstitial nephropathy.30,36 Nevertheless, 8/32 CKD samples had a UPC >2 considered as indicative of prevalent glomerular diseases that is not the typical, but a possible, cause of CKD even in cats. Our results therefore suggest that in these patients histopathological characterisation of CKD should be performed to correlate urinary electrophoresis with histological findings.
Although the present study validated a fast and reliable technique to quantify albuminuria in cats producing preliminary results of UAC, our study has a few limitations. First, the limited number of healthy cats, as to construct a true reference interval a greater number of samples is needed. Second, HRE was less sensitive in detecting low concentrations of albumin (<5 mg/dl) than an ELISA (10–200 ng/ml) but fast and easy to use in the routine clinic. Third, UAC was not able to discriminate between CKD and other diseases characterised by proteinuria, suggesting its use as an unspecific marker of disease with possible renal involvement rather than a specific marker of CKD (as glomerular filtration rate could not be measured and used as the gold standard in this study). HRE was unable to separate albumin from cauxin, in particular in entire male healthy cats.
Conclusions
The electrophoretic method validated in this work is precise and accurate for measuring albuminuria in cats. UAC was useful to classify proteinuria correctly and to discriminate among healthy and diseased cats. Values for UAC of 0.035 and 0.074 are suggested as the thresholds of microalbuminuria and overt albuminuria, respectively, in cats, while a value of >0.017 could be indicative of disease with possible renal involvement. Moreover, we have reported the first data on albuminuria in cats affected by LUTD. In addition, HRE can provide information on urine proteins and the profile (including albumin and globulins) should be further analysed to aid clinicians in the diagnosis of kidney diseases.
Footnotes
Accepted: 14 July 2016
Supplementary material: The following files are available:
Table 1: Age, sex and results of urine specific gravity, dipstick, proteinuria and albuminuria for the healthy samples.
Table 2: Age, sex and results of urine specific gravity, dipstick, proteinuria and albuminuria samples of the diseased groups.
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: This study was supported by a grant from the University of Bologna (RFO) to GI. EF was supported by a grant from the University of Bologna (AA EF 2011) for a research period in Cambridge.
References
- 1. Whittemore JC, Gill VL, Jensen WA, et al. Evaluation of the association between microalbuminuria and the urine albumin-creatinine ratio and systemic disease in dogs. J Am Vet Med Assoc 2006; 229: 958–963. [DOI] [PubMed] [Google Scholar]
- 2. Grauer GF. Proteinuria: measurement and interpretation. Top Companion Anim Med 2011; 26: 121–127. [DOI] [PubMed] [Google Scholar]
- 3. Guh JY. Proteinuria versus albuminuria in chronic kidney disease. Nephrology 2010; 15: 53–56. [DOI] [PubMed] [Google Scholar]
- 4. de Jong PE, Gansevoort RT. Focus on microalbuminuria to improve cardiac and renal protection. Nephron Clin Pract 2009; 111: 204–211. [DOI] [PubMed] [Google Scholar]
- 5. Erdmann E. Microalbuminuria as a marker of cardiovascular risk in patients with type 2 diabetes. Int J Cardiol 2006; 107: 147–153. [DOI] [PubMed] [Google Scholar]
- 6. Dobre D, Nimade S, de Zeeuw D. Albuminuria in heart failure: what do we really know? Curr Opin Cardiol 2009; 24: 148–154. [DOI] [PubMed] [Google Scholar]
- 7. Smets PMY, Meyer E, Maddens BEJ, et al. Urinary markers in healthy young and aged dogs and dogs with chronic kidney disease. J Vet Intern Med 2010; 24: 65–72. [DOI] [PubMed] [Google Scholar]
- 8. Syme HM, Markwell PJ, Pfeiffer D, et al. Survival of cats with naturally occurring chronic renal failure is related to severity of proteinuria. J Vet Intern Med 2006; 20: 528–535. [DOI] [PubMed] [Google Scholar]
- 9. Cortadellas O, Fernández del Palacio MJ, Talavera J, et al. Glomerular filtration rate in dogs with leishmaniasis and chronic kidney disease. J Vet Intern Med 2008; 22: 293–300. [DOI] [PubMed] [Google Scholar]
- 10. Bacic A, Kogika MM, Barbaro KC, et al. Evaluation of albuminuria and its relationship with blood pressure in dogs with chronic kidney disease. Vet Clin Pathol 2010; 39: 203–209. [DOI] [PubMed] [Google Scholar]
- 11. Jepson RE, Elliott J, Brodbelt D, et al. Effect of control of systolic blood pressure on survival in cats with systemic hypertension. J Vet Intern Med 2007; 21: 402–409. [DOI] [PubMed] [Google Scholar]
- 12. Lien YH, Hsiang TY, Huang HP. Associations among systemic blood pressure, microalbuminuria and albuminuria in dogs affected with pituitary- and adrenal-dependent hyperadrenocorticism. Acta Vet Scand 2010; 52: 61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Smets PMY, Lefebvre HP, Meij BP, et al. Long-term follow-up of renal function in dogs after treatment for ACTH-dependent hyperadrenocorticism. J Vet Intern Med 2012; 26: 565–574. [DOI] [PubMed] [Google Scholar]
- 14. Schaefer H, Kohn B, Schweigert FJ, et al. Quantitative and qualitative urine protein excretion in dogs with severe inflammatory response syndrome. J Vet Intern Med 2011; 25: 1292–1297. [DOI] [PubMed] [Google Scholar]
- 15. Whittemore J, Marcum B, Mawby D, et al. Associations among albuminuria, C-reactive protein concentrations, survival predictor index scores, and survival in 78 critically ill dogs. J Vet Intern Med 2011; 25: 818–824. [DOI] [PubMed] [Google Scholar]
- 16. Vaden SL, Turman CA, Harris TL, et al. The prevalence of albuminuria in dogs and cats in an ICU or recovering from anesthesia. J Vet Emerg Crit Care 2010; 20: 479–487. [DOI] [PubMed] [Google Scholar]
- 17. Al-Ghazlat SA, Langston CE, Greco DS, et al. The prevalence of microalbuminuria and proteinuria in cats with diabetes mellitus. Top Companion Anim Med 2011; 26: 154–157. [DOI] [PubMed] [Google Scholar]
- 18. Whittemore J, Miyoshi Z, Jensen WA, et al. Association of microalbuminuria and the urine albumin-to-creatinine ratio with systemic disease in cats. J Am Vet Med Assoc 2007; 230: 1165–1169. [DOI] [PubMed] [Google Scholar]
- 19. Jepson R, Brodbelt D. Evaluation of predictors of the development of azotemia in cats. J Vet Intern Med 2009; 23: 806–813. [DOI] [PubMed] [Google Scholar]
- 20. Lyon S, Sanderson M, Vaden S, et al. Comparison of urine dipstick, sulfosalicylic acid, urine protein-to-creatinine ratio, and species-specific ELISA methods for detection of albumin in urine samples of cats. J Am Vet Med Assoc 2010; 236: 874–879. [DOI] [PubMed] [Google Scholar]
- 21. Kuwahara Y, Nishii N, Takasu M, et al. Use of urine albumin/creatinine ratio for estimation of proteinuria in cats and dogs. J Vet Med Sci 2008; 70: 865–867. [DOI] [PubMed] [Google Scholar]
- 22. Williams TL, Archer J. Evaluation of urinary biomarkers for azotemic chronic kidney disease in cats. J Small Anim Pract 2016; 57: 122–129. [DOI] [PubMed] [Google Scholar]
- 23. Mardell EJ, Sparkes AH. Evaluation of a commercial in-house test kit for the semi-quantitative assessment of microalbuminuria in cats. J Feline Med Surg 2006; 8: 269–278. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Hanzlicek AS, Roof CJ, Sanderson MW, et al. Comparison of urine dipstick, sulfosalicylic acid, urine protein-to-creatinine ratio and a feline-specific immunoassay for detection of albuminuria in cats with chronic kidney disease. J Feline Med Surg 2012; 14: 882–888. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. International Renal Interest Society. IRIS staging of CKD (modified 2015). http://iris-kidney.com/guidelines/staging.html (2015, accessed January 15, 2016)
- 26. King JN, Tasker S, Gunn-Moore DA, et al. Prognostic factors in cats with chronic kidney disease. J Vet Intern Med 2007; 21: 906–916. [PubMed] [Google Scholar]
- 27. Walker D, Syme H, Markwell P, et al. Predictors of survival in healthy, non-azotaemic cats [abstract]. J Vet Intern Med 2004; 18: 417. [Google Scholar]
- 28. Jensen AL, Kjelgaard-Hansen M. Diagnostic test validation. In: Weiss DJ, Wardrop JK. (eds). Schalm’s veterinary hematolgy. 6th ed. Ames, IA: Wiley-Blackwell, 2010, pp 1027–1033. [Google Scholar]
- 29. Giori L, Tricomi FM, Zatelli A, et al. High-resolution gel electrophoresis and sodium dodecyl sulphate-agarose gel electrophoresis on urine samples for qualitative analysis of proteinuria in dogs. J Vet Diagn Invest 2011; 23: 682–690. [DOI] [PubMed] [Google Scholar]
- 30. Ferlizza E, Campos A, Neagu A, et al. The effect of chronic kidney disease on the urine proteome in the domestic cat (Felis catus). Vet J 2015; 204: 73–81. [DOI] [PubMed] [Google Scholar]
- 31. Zatelli A, Paltrinieri S, Nizi F, et al. Evaluation of a urine dipstick test for confirmation or exclusion of proteinuria in dogs. Am J Vet Res 2010; 71: 235–240. [DOI] [PubMed] [Google Scholar]
- 32. Miyazaki M, Fujiwara K, Suzuta Y, et al. Screening for proteinuria in cats using a conventional dipstick test after removal of cauxin from urine with a Lens culinaris agglutinin lectin tip. Vet J 2011; 189: 312–317. [DOI] [PubMed] [Google Scholar]
- 33. Langston C. Microalbuminuria in cats. J Am Anim Hosp Assoc 2004; 40: 251–254. [DOI] [PubMed] [Google Scholar]
- 34. Grauer GF. Measurement, interpretation, and implications of proteinuria and albuminuria. Vet Clin North Am Small Anim Pract 2007; 37: 283–295. [DOI] [PubMed] [Google Scholar]
- 35. Harley L, Langston C. Proteinuria in dogs and cats. Can Vet J 2012; 53: 631–638. [PMC free article] [PubMed] [Google Scholar]
- 36. Syme HM. Proteinuria in cats. Prognostic marker or mediator? J Feline Med Surg 2009; 11: 211–218. [DOI] [PMC free article] [PubMed] [Google Scholar]