Abstract
Objectives
Hypertension is a common cause of proteinuria in HIV-infected people. In cats, feline immunodeficiency virus (FIV) infection appears to be associated with proteinuria. Therefore, the results from systolic blood pressure (SBP) measurements in naturally infected FIV-positive cats were reviewed to assess whether hypertension contributes to the observed proteinuria in these cats. Ultrasonographic findings in FIV-positive cats were reviewed to complete renal assessment and to extend the scant knowledge on renal ultrasonography in cats.
Methods
Data from client-owned, naturally infected FIV-positive cats were retrospectively reviewed. To obtain a control group, records were reviewed from age-matched, privately owned, FIV-negative cats.
Results
Data from 91 FIV-infected and 113 control cats were compared. FIV-infected cats showed a significantly lower SBP (P <0.0001) and significantly fewer FIV-infected cats were hypertensive (⩾160 mmHg) compared with control cats (P = 0.025). The prevalence of renal azotaemia did not significantly differ between groups, although FIV-infected cats had significantly lower urine specific gravity (USG) (P = 0.0273) and a higher incidence of USG below 1.035 (P = 0.043). Urinary protein:creatinine ratio (UPC) was significantly higher in FIV-infected cats (P = 0.0005) and proteinuria (UPC >0.4) occurred more frequently in FIV-infected cats (P <0.001). Renal ultrasonography showed abnormalities in 60/91 FIV-infected cats, with hyperechogenic cortices in 39/91 and enlarged kidneys in 31/91.
Conclusions and relevance
Hypertension can be excluded as a common cause of renal damage leading to proteinuria in FIV-infected cats. Proteinuria and poorly concentrated urine are common in naturally infected FIV-positive cats, in contrast to azotaemia. Clinicians should cautiously interpret ultrasonographic abnormalities as these occur in over half of FIV-infected cats.
Introduction
Kidney disease is one of the most common complications in HIV infection, with proteinuria in up to 30% of HIV-infected people. 1 The effect of HIV on kidney function is complex. 2 HIV-associated nephropathy (HIVAN) has been described,3,4 but comorbidities, such as hypertension, may also contribute to kidney damage. Hypertensive vascular disease appears the most common renal pathological finding in HIV-infected patients.4–7
Feline immunodeficiency virus (FIV) is closely related to HIV and often serves as a model for pathogenic research.8,9 FIV is a common pathogen in cats worldwide and has been associated with multiple disorders.10,11 Involvement of FIV in kidney disease appears controversial. Some authors report an association between FIV and azotaemia;12–16 others refute this theory.17–19 Studies reporting urinalysis in FIV-infected cats all describe an increased prevalence of proteinuria but without identifying a cause.14,16,18,20–22 Recently, our group reported a significantly higher systolic blood pressure (SBP) in a small number of FIV-infected cats when compared with FIV-negative cats. 23 Therefore, this study aimed to verify whether FIV-infected cats are, akin to HIV-infected people, at higher risk of developing hypertension.
To further elucidate the occurrence of kidney disease in this population, routine kidney variables were evaluated and results from renal ultrasound examinations were described. Ultrasonography may be of help when evaluating the pathogenesis of kidney disease, but renal ultrasonographic abnormalities appear common even in healthy cats. 24 Recent data on feline kidney ultrasonography are scarce, 24 and no data at all are available for FIV-infected cats. Therefore, information on renal ultrasonographic findings in FIV-infected cats is required to interpret ultrasonographic results.
Materials and methods
Study population
Records were reviewed from privately owned, naturally infected FIV-positive cats, presented to the Veterinary Teaching Hospital at the Faculty of Veterinary Medicine of Ghent University (Belgium). Cats were hospitalised for 1 day to perform pre-trial screening for a clinical study, approved by both local and national ethical committees (EC2010/067). To obtain a control group, records were reviewed from age-matched, privately owned, FIV-negative cats, presented to the Veterinary Teaching Hospital at the Faculty of Veterinary Medicine of Ghent University (Belgium). Records were selected from the data of two previous studies, approved by both local and national ethical committees (EC2010/27 and EC2011/197). 23 Only records from healthy control cats were included. Health was defined as clinically healthy for their owner and without significant abnormalities on physical examination, complete blood count and serum biochemistry profile, except for serum creatinine (sCreat) and urea (sUrea) concentrations. Cats with hyperthyroidism or feline leukaemia virus (FeLV) infection were excluded from both the FIV and the control group.
Procedures
SBP was measured using the Doppler ultrasonic technique, following the guidelines of the American College of Veterinary Internal Medicine (ACVIM) consensus statement. 25 Hypertension was defined as SBP ⩾160 mmHg and hypotension as SBP <80 mmHg.25–27
A standard physical examination was performed. Jugular vein blood was collected and cystocentesis was performed. Complete blood count (Advia 2120; Siemens) and serum biochemistry profile (Architect C16000; Abbott), automated urine chemistry, including urine protein:creatinine ratio (UPC) (iRICELL 3000; Instrumentation Laboratory) and urinary bacterial culture (VITEK 2 Systems; BioMerieux) were performed. sCreat was determined using a modified Jaffé method (reference interval [RI] 64.5–161.8 μmol/l). 28 sUrea was determined by enzymatic assay (RI 6.2–12.7 mmol/l). 28 The urinary sediment was evaluated as previously described.23,29 Urine specific gravity (USG) was determined using a manual refractometer. Renal azotaemia was defined as increased sCreat (sCreat >161.8 μmol/l) combined with USG ⩽1.035. In cats older than 6 years of age, total thyroxine (TT4) serum concentration was measured. In the FIV-infected group, serum amyloid A (SAA) (RI <4.0 mg/l) was measured. No SAA values were available for the control group. FIV and FeLV status was determined with a Witness FeLV–FIV test (Synbiotics) in the FIV-infected cats and a Snap Combo Plus (IDEXX) in the control cats. FIV infection was confirmed by reverse transcriptase quantitative PCR (Okapi Sciences NV).
In the FIV-infected cats, abdominal ultrasonography was performed by an ECVDI Diplomate or supervised Resident using a multifrequency (6–9 MHz) microconvex or multifrequency (7.5–12 MHz) linear transducer (Logic 7; GE Medical Systems). This study reviewed findings in the urinary system (kidneys, ureters and bladder). Based on current literature, a small kidney was defined as <3.2 cm, renomegaly as >4.4 cm and a large size difference between both kidneys as >0.7 cm.30–32 Abdominal ultrasonography was not performed in the control group.
Statistical analysis
All analyses were performed using the statistical software package SAS (version 9.4; SAS Institute). To compare continuous variables between the FIV-infected and the control group, a Student’s t-test was used. Binary outcome variables were compared using the Fisher’s exact test. A linear fixed effects model with normally distributed random error term was used to compare UPC values between three age groups within both populations, using Bonferroni’s technique to adjust for multiple comparisons. Tests were performed at a 5% significance level.
Results
In total, the records of 91 naturally FIV-infected cats were included: 77 males (13 intact, 64 neutered) and 14 females (three intact, 11 neutered). The control group consisted of 113 cats, 42 males (two intact, 40 neutered) and 71 females (17 intact, 54 neutered). Males were significantly more represented in the FIV-infected group (P <0.0001), but the proportion of intact vs neutered males was not significantly different between both groups (P = 0.0818). All FIV-infected cats were domestic short- or longhair cats. The control group consisted of 92 domestic short- or longhair cats and 21 purebred cats, including British Shorthair (n = 5), Persian (n = 4), Ragdoll (n = 3), Sphynx (n = 2), Peterbald (n = 2), Birman (n = 1), Maine Coon (n = 1), Norwegian Forest Cat (n = 1), Siamese (n = 1) and Oriental Shorthair (n = 1) cats. Owing to insufficient sample volume, TT4 concentration could not be determined in eight FIV-infected cats over 6 years of age. None of these cats had a palpable thyroid gland, nor did they show clinical signs consistent with hyperthyroidism. Urine sediment analysis was not performed in two control cats owing to lack of sample. These cats did not show proteinuria.
Descriptive statistics for age, body weight, SBP, sCreat, sUrea, USG and UPC of both groups are presented in Table 1. No significant differences were found for age, sCreat and sUrea. Control cats had a significantly higher body weight than FIV-infected cats (P = 0.042).
Descriptive statistics for continuous variables age, body weight, systolic blood pressure (SBP), serum creatinine concentration (sCreat), serum urea concentration (sUrea), urine specific gravity (USG) and urinary protein:creatinine ratio (UPC) for FIV-infected and FIV-non-infected cats
Significant difference between FIV-infected and FIV-non-infected cats for this parameter (P <0.05)
Reference interval (RI) based on Ghys et al 28
NA = not applicable
Cats with FIV had significantly lower SBP (P <0.0001) and a significantly lower prevalence of hypertension than the control cats (1.1% [1/91] of FIV-infected cats and 8.8% [10/113] of control cats; P = 0.025). No cats were hypotensive. Behaviour during the procedure was recorded as ‘completely at ease’ in 48/91 (52.7%) FIV-infected cats. The only hypertensive (SBP = 161 mmHg) FIV-infected cat performed behaviours indicating it was stressed during the procedure.
The frequency of cats with renal azotaemia did not differ significantly between groups (6.6% [6/91] of FIV-infected cats and 2.7% [3/113] of control cats; P = 0.1917).
FIV-infected cats had a significantly lower USG (P = 0.0273) and the proportion of FIV-infected cats with poorly concentrated urine was significantly higher when compared with control cats (35.2% [32/91] of FIV-infected cats and 22.1% [25/113] of control cats; P = 0.043).
UPC was significantly higher in the FIV-infected group than in the control cats (P = 0.0005) and a greater proportion of FIV-infected cats showed proteinuria (UPC >0.4) compared with control cats (27.5% [25/91] of FIV-infected cats and 1.8% [2/113] of control cats; P <0.001). The number of borderline proteinuric (UPC 0.2–0.4) cats did not differ significantly between both groups. Age did not significantly affect UPC. One proteinuric FIV-infected cat was diagnosed with immune-mediated haemolytic anaemia. Of the other FIV-infected cats with proteinuria, five cats had renal azotaemia, five cats showed microscopic haematuria, three cats showed crystalluria and one suffered from idiopathic cystitis. Increased SAA was found in 17/25 proteinuric (range 0.41–3.62) FIV cats. One of the two control cats with proteinuria showed microscopic haematuria.
Ultrasonography
Ultrasonography of the urinary tract was performed in all FIV-infected cats; findings are summarised in Tables 2 and 3. Renal abnormalities were found in 60/91 cats (65.9%).
Descriptive statistics for ultrasonographic kidney measurements in a population of naturally infected FIV-positive cats
Abnormalities observed on ultrasonography of the urinary tract in 91 naturally infected FIV-positive cats
Of the cats with hyperechoic cortices, 16/39 (41%) appeared proteinuric, two of them within nephrotic range (UPC ⩾2). 33 Renomegaly was found in 31 cats, bilaterally in 13 of them. In one cat, bilateral renomegaly was caused by polycystic kidney disease. Fine-needle aspiration cytology was available in 5/31 and autopsy was performed in 7/31, but no neoplastic changes were detected in these 12 samples. Telephone contact with owners of 17 of the 18 remaining cats with renomegaly revealed that six cats had died within 1 month, but 11 of them survived for 6 months or longer. No follow-up was available in one cat. Ten of 31 cats with renomegaly appeared proteinuric (range 0.41–1.93) and 8/10 of them had increased SAA levels.
Discussion
Hypertension is a common cause of renal disease in HIV-infected individuals and recently, 3/14 apparently healthy FIV-positive cats were found to be hypertensive.4–7,23 Hypertension may induce proteinuria in cats,25,34 and previous studies reported a significantly higher occurrence of proteinuria and a higher median UPC in FIV-infected cats when compared with a FIV-negative control group.16,18 However, SBP was not yet reported in a large number of FIV-infected cats.
The present results confirm a higher incidence of proteinuria in FIV-infected cats but show that hypertension can be excluded as a common cause for this phenomenon. In contrast, mean SBP and the proportion of hypertensive cats was significantly higher in the control group. The discrepancy in SBP between FIV-infected and control cats might be caused by the retrospective nature of this comparison. Although both groups were evaluated with respect to the ACVIM consensus guidelines, circumstances were different. Cats from the FIV-infected group were hospitalised for a day and had more time to acclimate. This resulted in completely tranquil behaviour in >50% of all FIV-infected cats during the procedure. Likewise, efforts were made to avoid stress-induced hypertension in the control group, but SBP measurements were unavoidably performed in more stressful situations, shortly after arrival at the hospital. This may have induced an anxiety-related artefactual increase of the SBP, the so-called ‘white coat’ effect. 35 A 5–10 min accommodation, as recommended in the ACVIM consensus statement, 25 may be insufficient for adequate acclimation.
FIV infection has been associated with azotaemia and chronic kidney disease (CKD),12,13,15,16 but other studies did not find significantly higher sCreat in naturally FIV-infected cats.17,18 In women with HIV, serum cystatin C (sCysC) estimates of glomerular filtration rate appeared more sensitive than sCreat in detecting reduced kidney function, 36 but other studies questioned the utility of sCysC in HIV-seropositive individuals.37,38 In FIV-infected cats, sCreat appeared higher, while sCysC did not differ significantly when compared with control cats. 22 However, recent data indicate that sCysC measured with human assays does not accurately reflect kidney function.39,40 The present study showed no significant difference in sCreat, sUrea or the prevalence of renal azotaemia in FIV-infected vs uninfected control cats, but a USG <1.035 was significantly more often observed in the FIV-infected group. Only one paper has studied USG in FIV-infected cats, describing a significantly lower USG, with USG below 1.035 in 44.2% of the FIV-infected cats. 18 Early kidney dysfunction in these cats cannot be excluded, but decreased USG as single abnormality does not necessarily indicate kidney dysfunction, as cats with normal glomerular filtration activity can also have a decreased USG.41,42
Defining proteinuria based on a single UPC measurement has its limitations, but the increased incidence of proteinuria in our FIV-positive population corroborates findings in previous studies.14,16,18,20–22 Haemolysis was the only prerenal cause that could be identified in one proteinuric cat and only a minority of proteinuric FIV-positive cats showed anomalies on urinalysis. Age has been associated with the level of proteinuria in cats, with a significantly higher UPC in cats >10 years of age compared with middle-aged cats (6–10 years),23,43 but although cats >10 years of age were well represented, increasing age did not affect the number of proteinuric cats in either of both study groups. A previous study in FIV-infected cats showed that intact males are at higher risk of proteinuria when compared with neutered males. 18 In the current study, there were too few intact males in the FIV-infected group to allow analysis of the association between neuter status and proteinuria. Although males were significantly more represented in the FIV-infected group, the proportion of neutered vs intact males did not differ significantly between groups. This suggests that the increased incidence of proteinuria in the FIV-infected group is unlikely to be sex-associated.
Proteinuria is associated with reduced survival time in cats with CKD, hypertension or hyperthyroidism, and non-azotaemic cats with borderline proteinuria also appear to have a decreased survival time.44–47 HIV-infected patients are at high risk of developing CKD once proteinuria is identified and therefore annual screening for renal disease, including urinalysis, is recommended.1,5,48 In cats, increased UPC may indicate the onset of renal disease and it has been reported that proteinuria may contribute to the progressive nature of CKD.49,50 Considering the fragile health status of FIV-infected cats, it is of major importance to detect, monitor and treat all possible health threats. Therefore, urinalysis, including UPC, should be part of the advised periodic monitoring of FIV-infected cats. 51
Over 60% of the examined cats showed abnormalities on renal ultrasonography. This is higher than the 40% renal ultrasonographic abnormalities reported by the same institution in healthy cats, but this healthy population was remarkably younger (mean ± SD age 2.7 ± 1.8 years, median 2.1) than the studied FIV-infected group (age 7.8 ± 3.6 years, median 7.9). 24 Various ultrasonographic changes were observed, with similar prevalences as previously described in healthy cats, except for renomegaly (34.1% vs 1.6%) and cortical echogenicity (39.7% vs 12.9%). 24
Enlargement of the kidneys and hyperechogenic cortices may be attributed to several causes.52,53 In patients with HIVAN, increased cortical echogenicity and decreased corticomedullary definition were identified as renal sonographic abnormalities due to glomerular and tubular lesions. Kidney size had no specific diagnostic utility.4,54–56 Renal biopsies in small groups of FIV-infected cats showed several histopathological kidney abnormalities similar to those observed in HIVAN patients,14,15,20,57 which might explain the ultrasonographic observations in the cats of our study cohort. However, HIVAN presents clinically as massive proteinuria,3,4 while in this study only a minority of the cats with hyperechoic cortices showed proteinuria and the UPC reached the nephrotic range in only two of them. 33
Uni- or bilateral renomegaly may be caused by neoplasia and FIV-infected cats are prone to lymphoma.11,53,58 However, fine-needle aspiration cytology or autopsy performed in approximately 40% of the cats with renomegaly did not show evidence of neoplasia. Survival time after chemotherapeutic treatment in cats with lymphoma, including renal lymphoma, varies from 46 days to 5.6 months.59–61 Follow-up of the studied FIV-infected cats with renomegaly showed survival >6 months in an additional 55% of them. Therefore, although lymphoma should not be disregarded in FIV-positive cats, it is an unlikely explanation for the commonly observed renomegaly in our study.
Increased SAA levels may lead to renal amyloid A (AA) amyloidosis, causing proteinuric kidney dysfunction, and the presence of enlarged kidneys on abdominal ultrasound.53,62,63 Recent studies reported increased SAA levels in diseased FIV-positive cats.21,64 Moreover, naturally FIV-infected cats appear to have a higher prevalence of renal AA amyloidosis than uninfected cats. 65 Interestingly, 68% of the proteinuric FIV-infected cats had increased SAA levels, combined with renomegaly in 47% of them. However, renal amyloidosis is frequently diagnosed with nephrotic range proteinuria, which was found in only two proteinuric FIV-infected cats with increased SAA levels. 66
The major limitations of this study are its retrospective nature and the use of a historical control group. In particular, a retrospective comparison of parameters subject to circumstantial variation, such as SBP measurement, can be difficult. However, although results suggested that this parameter was, indeed, affected by differences in the applied protocol, they undoubtedly served the main purpose of this study and excluded hypertension as a possible cause of proteinuria in FIV-infected cats. Furthermore, no control group was available from which to collect ultrasonographic data. However, cautious comparisons were possible using a control group from a study performed by the same institution, using the same infrastructure and methodology. These first analyses showed that a possible renal involvement of FIV may be visible on renal ultrasonography, but further studies are required to relate these to clinicopathological findings. Nevertheless, we acknowledge that a prospective study including a control group is preferable to a retrospective comparison. This would help to avoid bias due to different procedures (eg, SBP measurement) or due to incomplete datasets (eg, lack of SAA and ultrasonographic findings in the control group).
Conclusions
Hypertension occurs rarely and can be excluded as a common cause of renal damage in FIV-infected cats. Environmental circumstances may have a significant effect on mean SBP measurements, which emphasises the importance of stress reduction during blood pressure measurement of cats in a hospital setting. FIV-infected cats are more prone to proteinuria and less concentrated urine than control cats, but the cause of these anomalies is yet to be determined. Clinicians should remain attentive to possible kidney disease in FIV-infected cats and perform urinalysis, including UPC measurement, when monitoring these patients. Renal ultrasonographical abnormalities are very often observed in naturally infected FIV-positive cats, but the actual clinical relevance remains unclear. Clinicians should be aware of these ultrasonographic changes in FIV-infected cats, and abnormalities should be interpreted cautiously before initiating invasive diagnostic tests.
Footnotes
Acknowledgements
We thank Ms B Weyn for her excellent assistance.
The results of this study were presented, in part, at the 25th ECVIM-CA Congress, Mainz, 4–6 September 2014.
Conflict of interest
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
This work was funded by Okapi Sciences NV, Heverlee, Belgium.
