Renal Disease in Cats Infected with Feline Immunodeficiency Virus
Corresponding author: J. K. Levy, Maddie's Shelter Medicine Program, College of Veterinary Medicine, University of Florida, 2015 SW 16th Avenue, Gainesville, FL 32608; e-mail: Levyjk@ufl.edu
Feline immunodeficiency virus (FIV) and human immunodeficiency virus (HIV) infection cause similar clinical syndromes of immune dysregulation, opportunistic infections, inflammatory diseases, and neoplasia. Renal disease is the 4th most common cause of death associated with HIV infection.
To investigate the association between FIV infection and renal disease in cats.
Client-owned cats (153 FIV-infected, 306 FIV-noninfected) and specific-pathogen-free (SPF) research colony cats (95 FIV-infected, 98 FIV-noninfected).
A mixed retrospective/prospective cross-sectional study. Blood urea nitrogen (BUN), serum creatinine, urine specific gravity (USG), and urine protein:creatinine ratio (UPC) data were compared between FIV-infected and FIV-noninfected cats. In FIV-infected cats, total CD4+ and CD8+ T lymphocytes were measured using flow cytometry, and CD4+:CD8+ T lymphocyte ratio was calculated. Renal azotemia was defined as a serum creatinine ≥ 1.9 mg/dL with USG ≤ 1.035. Proteinuria was defined as a UPC > 0.4 with an inactive urine sediment.
Among the client-owned cats, no association was detected between FIV infection and renal azotemia (P = .24); however, a greater proportion of FIV-infected cats were proteinuric (25.0%, 16 of 64 cats) compared to FIV-noninfected cats (10.3%, 20 of 195 cats) (P < .01). Neither neuter status nor health status were risk factors for proteinuria in FIV-infected cats, but UPC was positively correlated with the CD4+:CD8+ T lymphocyte ratio (Spearman's rho = 0.37, P = .01). Among the SPF research colony cats, no association was detected between FIV infection and renal azotemia (P = .21) or proteinuria (P = .25).
Conclusions and Clinical Importance
Proteinuria but not azotemia was associated with natural FIV infection.
acquired immunodeficiency syndrome
blood urea nitrogen
feline leukemia virus
feline immunodeficiency virus
human immunodeficiency virus
human immunodeficiency virus – associated
North Carolina State University
specific pathogen free
urine protein:creatinine ratio
urine specific gravity
Veterinary Medical Teaching Hospital
Feline immunodeficiency virus (FIV) is a lentivirus of worldwide distribution. Seroprevalence in cats in North America is estimated at 2.5%, with highest prevalence in adult male, sick, and free-roaming cats. Infection is associated with progressive immune dysfunction characterized by decreasing CD4+ T lymphocyte count, CD4+:CD8+ T lymphocyte ratio, and lymphopenia.[2, 3] Over time, these alterations can predispose to morbidities such as chronic inflammatory, infectious, or neoplastic diseases, which can ultimately prove fatal.
Human immunodeficiency virus (HIV) is a related lentivirus and demonstrates marked similarities in pathogenesis to FIV. The clinical consequences of infection (termed acquired immunodeficiency syndrome, AIDS) are often a result of similar inflammatory, infectious, or neoplastic diseases. However, there are several complex sequelae of HIV infection that have not been thoroughly described in FIV-infected cats. These dissimilarities might be attributed to a lack of investigation in feline medicine, or might represent intrinsic differences in the viruses or species infected.
Renal disease has been documented to be the 4th most common cause of death of HIV-infected people. FIV-infected cats have a variable prevalence of azotemia, although the increased prevalence of proteinuria is unclear.[5-9] A study assessing serum biochemical data and urinalyses reported proteinuria in 14 of 15 naturally FIV-infected cats, but did not include a control group. The high prevalence of chronic kidney disease in cats, especially within the aged population, necessitates an age-matched population to accurately identify an increased prevalence of renal disease in FIV-infected cats.
We hypothesized that, relative to noninfected cats, cats infected with FIV are at increased risk of proteinuria, this proteinuria is associated with advanced FIV infection as demonstrated by alterations in the T lymphocyte subsets, and comorbidity factors contribute to the prevalence of renal disease. The objectives of this study were to determine the prevalence and risk factors of azotemia and proteinuria in cats with natural and experimental FIV infection.
Materials and Methods
A mixed retrospective/prospective cross-sectional study was designed. Medical records of cats referred to the Veterinary Medical Teaching Hospital (VMTH) at the College of Veterinary Medicine at North Carolina State University (NCSU) were examined retrospectively to identify cats that had been tested for retroviral infection, and had concurrent serum biochemistry and urinalysis data. Cats undergoing retroviral infection testing at the VMTH, and by veterinarians in the surrounding community after the initiation of the study, were also included prospectively. Data were collected for a total of 153 FIV-infected cats (63 retrospectively enrolled, 90 prospectively recruited). To create a comparison group of FIV-noninfected cats with similar sex and age characteristics, for each FIV-infected cat, 2 FIV-noninfected cats of the same sex and of similar age (within 3 years) were enrolled in the study. Thus, data were collected for a total of 306 FIV-noninfected cats (80 retrospectively enrolled, 226 prospectively recruited).
To investigate the effect of comorbidities found in cats exposed to the natural environment as compared to cats from specific-pathogen-free (SPF) barrier facilities, a 2nd group of 95 cats, housed in a SPF research colony within NCSU, and experimentally infected with FIV as part of a separate research investigation, was also studied. The duration of experimental infection was variable (median 2.2 years, range 0.3–4.9 years), and was recorded for each cat. A group of similarly housed 98 FIV-noninfected cats was included for comparison. No attempt was made to match these FIV-infected and FIV-noninfected cats by sex and age.
For each cat included in the study, the presence of FIV antibodies and feline leukemia virus (FeLV) antigen in the blood was analyzed using ELISA and the results of concurrent clinicopathologic testing were recorded. Clinicopathologic evaluation1 included serum creatinine concentration, urea concentration (BUN), urine specific gravity (USG), urine protein to creatinine ratio (UPC), and, for prospectively recruited FIV-infected cats, lymphocyte subset analysis. CD4+ and CD8+ T lymphocyte counts were measured using flow cytometric analysis. The presence of renal disease was defined in 2 ways: azotemic renal disease was defined by a serum creatinine above the reference range (ie, ≥ 1.9 mg/dL) with a concurrent USG of ≤ 1.035, while proteinuria was defined as a UPC > 0.4 with an inactive urine sediment. Those cats with a UPC > 0.4 together with an active sediment on urine sediment examination (9 client-owned FIV-infected cats, 12 client-owned FIV-noninfected cats, and 1 SPF FIV-noninfected cat) were excluded from further analysis with respect to proteinuria. An active sediment was defined as bacteriuria, pyuria (ie, a WBC count of > 5 cells/hpf), and/or hematuria (ie, a RBC count of > 5 cells/hpf). For each naturally FIV-infected cat, a clinician from the VMTH assessed the case and was asked to subjectively characterize each cat as “healthy” or “diseased,” based on history, physical examination findings, and bloodwork results, in addition to any other available diagnostic test results. Examples of the clinical findings and diagnoses that led to a categorization of a cat as “diseased” include dyspnea, stomatitis, anemia, heartworm disease, neoplasia, and hepatic lipidosis.
The study protocol was approved by the NCSU Institutional Animal Care and Use Committee, and owners of naturally infected cats provided informed consent before enrollment. The SPF cats were cared for in enriched group housing in accordance to the principles of the NIH Guide for the Care and Use of Laboratory Animals.
Statistical comparisons were performed using a standard software package.2 Serum creatinine, BUN, USG, and lymphocyte subset counts were not normally distributed; therefore comparison of the median values for FIV-infected and FIV-noninfected cats was performed using Mann-Whitney U-tests. The proportions of FIV-infected and FIV-noninfected cats with renal disease were compared using chi-squared tests; a Fisher's exact test was used when there were fewer than 5 cats in a comparison group. Risk factors determined to be significantly associated with the outcome of renal disease in univariate analyses were analyzed by means of multivariate logistic regression. Potential confounders were retained if they changed other factors’ odds ratios (ORs) by at least 10%. Spearman's rho was examined for non-parametric correlation between counts of T and B lymphocytes, CD4+ and CD8+ T lymphocytes, as well as the CD4+:CD8+ T lymphocyte ratio and UPC. Statistical significance was considered if P < .05.
Client-Owned FIV-Infected Cats
The age, sex, neuter, FeLV, and health status of the client-owned cats are described in Table 1, in addition to the median creatinine and BUN concentrations, USG, and UPC.
Table 1. Characteristics of 153 naturally FIV-infected and 306 FIV-noninfected client-owned cats. The FIV-infected and FIV-noninfected cats were recruited so that for every infected cat, 2 noninfected cats of the same sex and a similar age (within 3 years) were included.
|Age (years)|| || ||9 (0.5–18)|| || ||9 (1–18)||.90|
|< 11||70||66.0|| ||158||64.5|| ||.78|
|≥ 11||36||34.0|| ||87||35.5|| |
|Male||128||83.7|| ||256||83.7|| ||1.00|
|Female||25||16.3|| ||50||16.3|| |
|Altered||88||57.5|| ||254||83.0|| ||<.01|
|Intact/unknown:||65||42.5|| ||52||17.0|| |
|Infected||14||9.2|| ||22||7|| ||.48|
|Noninfected||139||90.8|| ||281||92.7|| |
|Sick||118||77.1|| ||206||67.3|| ||.03|
|Healthy||35||22.9|| ||100||32.7|| |
|Creatinine concentration (mg/dL)||151|| ||1.5 (0.8–14.0)||306|| ||1.5 (0.7–5.6)||.96|
|BUN concentration (mg/dL)||152|| ||26 (6–220)||306|| ||26 (8–147)||.29|
|USG||129|| ||1.039 (1.010–1.070)||305|| ||1.046 (1.000–1.080)||< .01|
|UPC||73|| ||0.26 (0.00–4.26)||195|| ||0.11 (0.00–2.55)||< .01|
There was no difference between the proportions of FIV-infected cats with a creatinine concentration above the upper limit of the reference range (≥ 1.9 mg/dL), compared to FIV-noninfected cats (20.5% [31 of 151 cats], and 17.0% [52 of 306 cats], respectively; P = .36); this was also true for BUN concentration (≥ 39 mg/dL, 15.8% [24 of 152 cats], and 10.8% [33 of 306 cats], respectively; P = .13). A greater proportion of FIV-infected cats (44.2%, 57 of 129 cats) had dilute urine as defined by a USG ≤ 1.035 compared to FIV-noninfected cats (24.6%, 75 of 305 cats) (P < .01).
There was no significant difference in the prevalence of azotemic renal disease (15.5% [22 of 142] of FIV-infected cats, 11.5% [33 of 287] of FIV-noninfected cats; P = .24). However, the prevalence of proteinuria in FIV-infected cats was more than double that in FIV-noninfected cats (25.0% [16 of 64 cats] and 10.3% [20 of 195 cats], respectively; P < .01).
Forty-nine FIV-infected cats had both UPC and T lymphocyte subset data available for analysis. There was no difference in the median CD4+ T lymphocyte counts between proteinuric FIV-infected cats (658 cells/μL, range 261–1308) and nonproteinuric FIV-infected cats (473 cells/μL, range 2–1254) (P = .18). The same was also true when comparing median CD8+ T lymphocyte counts between proteinuric and nonproteinuric FIV-infected cats (462 cells/μL, range 354–1106, and 559 cells/μL, range 11–2710, respectively; P = .57). However, the median CD4+:CD8+ T lymphocyte ratio was significantly higher in proteinuric FIV-infected cats (1.12, range 0.70–1.77) in comparison to nonproteinuric FIV-infected cats (0.70, range 0.18–1.53) (P < .01). A positive correlation was found between CD4+ T lymphocyte count and UPC (Spearman's rho = 0.37, P < .01), and between CD4+:CD8+ T lymphocyte ratio and UPC (Spearman's rho = 0.29, P = .03).
Health status was not a significant risk factor for proteinuria in FIV-infected cats (OR 0.36, 95% CI 0.11–1.16, P = .09); neither was neuter status (OR 2.83, 95% CI 0.89–9.03, P = .08). For FIV-infected cats, logistic analysis stratified on the basis of sex demonstrated that intact male FIV-infected cats were nearly 3.5 times more likely to be proteinuric compared with neutered male FIV-infected cats (OR = 3.46, 95% CI 1.02–11.74, P = .05); there were too few FIV-infected female cats to analyze the association between neuter status and proteinuria.
Experimentally FIV-Infected SPF Cats
The age, sex, and neuter status of the SPF cats are described in Table 2, in addition to the median creatinine and BUN concentrations, USG, and UPC. All cats were FeLV-noninfected and considered to be healthy based on historical information and physical examination.
Table 2. Characteristics of 95 experimentally FIV-infected and 98 FIV-noninfected SPF cats. The sex of 3 FIV-noninfected cats was not recorded.
|Age (years)||81|| ||2.2 (1.0–5.9)||81|| ||2.1 (1.0–5.2)||< .01|
|Male||37||38.9|| ||32||33.7|| ||.45|
|Female||58||61.1|| ||63||66.3|| |
|Altered||31||32.6|| ||38||38.8|| ||.37|
|Intact||64||67.4|| ||60||61.2|| |
|Creatinine concentration (mg/dL)||95|| ||1.5 (0.8–10.2)||98|| ||1.4 (0.9–2.5)||< .01|
|BUN concentration (mg/dL)||95|| ||24 (15–285)||98|| ||24 (13–41)||.92|
|USG||95|| ||1.056 (1.010–1.080)||98|| ||1.053 (1.020–1.080)||.02|
|UPC||95|| ||0.07 (0.02–5.01)||78|| ||0.08 (0.00–0.57)||.57|
Although the median creatinine concentration was significantly greater in FIV-infected cats compared to FIV-noninfected cats (Table 2), it is worth noting that 3 FIV-infected cats had marked increases in creatinine concentrations (10.2, 8.7, and 8.0 mg/dL) whereas the remainder all had creatinine concentrations less than 2.7 mg/dL. There was no statistically significant difference between the proportions of FIV-infected cats with a creatinine concentration above the upper limit of the reference range (≥ 1.9 mg/dL), compared to FIV-noninfected cats (10.5% [10 of 95 cats], and 3.1% [3 of 98 cats], respectively; P = .05); this was also true for BUN concentration (≥ 39 mg/dL) (4.2% [4 of 95 cats], and 1.0% [1 of 98 cats], respectively; P = .21).
A total of 6 FIV-infected cats and 6 FIV-noninfected cats had dilute urine as defined by a USG ≤ 1.035 (6.3% of FIV-infected, and 6.1% of FIV-noninfected cats, P = .96). A total of 4 of 95 FIV-infected cats and 1 of 98 FIV-noninfected cats had renal azotemia (P = .21). With respect to proteinuria, 3 of 95 FIV-infected cats and 0 of 77 FIV-noninfected cats met the definition (P = .25). There were an insufficient number of cats with renal disease to perform logistic regression.
Eighty-seven FIV-infected cats had both UPC and T lymphocyte subset data available for analysis. The 3 proteinuric FIV-infected cats had significantly lower median CD4+ T lymphocyte counts (32 cells/μL, range 7–96) compared to the nonproteinuric FIV-infected cats (541 cells/μL, range 38–1694) (P < .01). The same was also true for the median CD8+ T lymphocyte count (42 cells/μL, range 9–375, compared to 549 cells/μL, range 43–2661; P < .01). The CD4+:CD8+ T lymphocyte ratio was not different between proteinuric FIV-infected cats (0.78, range 0.09–2.29) and nonproteinuric FIV-infected cats (0.87, range 0.27–2.84) (P = .83). A negative correlation was found between CD4+ T lymphocyte count and UPC (Spearman's rho −0.23, P = .04).
The prevalence of proteinuria was higher in client-owned FIV-infected cats than in FIV-noninfected cats; however, there was no significant difference in the prevalence of azotemia. The results of this study are consistent with the results of 2 previous investigations.[5, 7] Other studies have documented an association between FIV infection and azotemia but prerenal azotemia could not be ruled out because of the lack of concurrent USG data.[6, 8-10] In an Australian study, young cats with chronic kidney disease (as defined by an increased creatinine concentration with concurrent low USG) were significantly more likely to be FIV-infected compared to young cats without renal disease, although FIV infection was not associated with renal disease in older cats. While the prevalence of azotemia was not significantly different between FIV-infected and FIV-noninfected cats in our study, the USG was significantly lower in infected cats. It is possible that this represents early renal insufficiency, but a longitudinal study would be required to confirm this.
Proteinuria has previously been reported in FIV-infected cats. Proteinuria without azotemia is commonly seen in renal disease diagnosed in HIV-infected people, particularly HIV-associated nephropathy (HIVAN). Histologic lesions similar to those that define HIVAN have been described in a small number of FIV-infected cats.[10, 13] It is therefore possible that FIV-infected cats may provide a model for HIVAN. However, HIVAN is often rapidly progressive and ultimately fatal and, while the course of proteinuria in FIV-infected cats was beyond the scope of this study, fulminant proteinuric renal disease leading to nephrotic syndrome appears to be rare in FIV-infected cats. Indeed, severe glomerular disease with nephrotic syndrome is rare in cats in general and might represent a species difference. The pathogenesis of HIVAN is also suspected to be caused by HIV genes (Vpr and Nef) which are not present in FIV. Longitudinal studies to examine the clinical course of proteinuria in FIV-infected cats and histopathologic studies to investigate the etiopathogenesis are warranted before speculation regarding similarities with HIVAN or other retrovirus-associated glomerulonephropathies can be confirmed in cats.
The identification of proteinuric renal disease in FIV-infected cats may have therapeutic implications. In HIV-infected people with proteinuric renal disease, angiotensin-converting enzyme inhibitors are the mainstay of treatment, as they are for protein-losing nephropathies in veterinary medicine. However, although controversial, there is some evidence that corticosteroids may result in significant improvements in proteinuria and renal function in people with HIVAN. Corticosteroid treatment has been associated with proteinuria in dogs,[17, 18] and therefore further characterization of FIV-associated proteinuria is required before specific therapeutic recommendations can be made. Proteinuric renal disease may also be prognostically significant. The presence of proteinuria in cats with chronic kidney disease (defined by an increased creatinine concentration and low USG) is an independent risk factor for the requirement of parenteral fluid treatment, death, or both. The effect of proteinuria on survival in FIV-infected cats was beyond the scope of this cross-sectional study.
Declining CD4+ T lymphocyte counts and CD4+:CD8+ T lymphocyte ratio are predictive of progression from asymptomatic clinical status to AIDS-related complex, and finally to AIDS in people. Although many studies have demonstrated that FIV-infected cats experience a similar gradual decline in CD4+ T lymphocyte counts and CD4+:CD8+ ratios over time,[6, 20, 21] there appears to be a weaker correlation with the onset of clinical disease in cats. Notwithstanding this, we were surprised to find a positive rather than negative correlation between the degree of proteinuria and the CD4+ T lymphocyte counts and CD4+:CD8+ T lymphocyte ratios in client-owned FIV-infected cats. This is a contradictory finding to that seen in HIVAN, in which albuminuria has been shown to be negatively correlated with CD4+ T lymphocyte counts. In contrast, the predicted negative correlation was noted in the experimentally FIV-infected SPF cats. The reason for the discrepancy is unknown; however, the small number of proteinuric cats within the experimentally FIV-infected group necessitates caution when interpreting these findings.
Logistic regression analysis failed to identify risk factors associated with proteinuria in client-owned FIV-infected cats. Neuter status, regardless of sex, was not a risk factor for proteinuria; however, although not statistically significant, intact male FIV-infected cats were more likely to be proteinuric compared to neutered male FIV-infected cats. The cause of this is unknown. A recent investigation into a renal tubular protein called cauxin demonstrated that the urinary cauxin concentration in intact male cats was significantly higher compared to that in neutered male cats, and might explain this finding. To our knowledge, urinary cauxin concentrations in FIV-infected cats have not been investigated. Female FIV-infected cats were underrepresented in our study, consistent with previous epidemiologic studies that have shown an increased risk of FIV infection in males, and thus it was impossible to analyze the effect of neutering on the FIV-infected female cat population. The small number of cats within these subgroups makes interpretation difficult and this is an area that would benefit from additional investigation.
Being “diseased” (as determined using our classification system) was not a risk factor for proteinuria in FIV-infected cats. The cause of the proteinuria in HIV-infected people remains uncertain although it appears to be most commonly related to a specific type of glomerulonephropathy, suggesting a single inciting factor, rather than secondary to other systemic inflammatory, infectious, or neoplastic diseases. Although the assessment of health status used in this study has limitations, it would suggest that the increased prevalence of proteinuria is not solely attributable to the secondary diseases that commonly accompany clinical FIV infection.
In a further effort to minimize the effect that concurrent infectious disease would have on the prevalence of renal disease, we studied a group of SPF cats. Although the median creatinine concentrations and median USGs were statistically different between experimentally FIV-infected and FIV-noninfected SPF cats, the difference is unlikely to be clinically relevant. The proportion of FIV-infected cats with a creatinine ≥ 1.9 mg/dL compared with FIV-noninfected cats was of borderline significance (P = .05), however, there was no difference in the prevalence of renal azotemia (ie, a high creatinine concentration together with dilute urine) between FIV-infected and FIV-noninfected cats. In contrast to the client-owned cats, there was also no difference in the median UPC or in the proportion of cats that were defined as being proteinuric. This suggests that the increased prevalence of proteinuria seen in client-owned animals may be partly caused by concomitant infectious diseases or environmental cofactors that are absent in SPF cats. However, the SPF cats were much younger than the client-owned cats, with FIV infections of presumptively shorter duration, and therefore it is difficult to directly compare these 2 populations.
There are some limitations to our study. FIV-noninfected cats were recruited to create an age- and sex-matched comparison population for FIV-infected cats, and thus it was impossible to examine those characteristics as risk factors for renal disease in the FIV-infected population. All cats were tested for FIV using a single ELISA test and ideally positive results would have been confirmed with additional testing to reduce the risk of false-positive results. Samples were collected before the release of the FIV vaccine, and therefore vaccine-associated false-positive FIV antibody test results were not considered an issue. The definition of renal disease is complex and therefore it is possible that some cats were miscategorized because of the limitations of using azotemia, USG, and proteinuria to define renal disease. Azotemia is an insensitive marker of mild renal disease, and the use of a combination of azotemia and dilute urine will have excluded those cats with renal disease that have maintained urine concentrating ability. Using the definition of proteinuria based on a single UPC measurement also has limitations, and from the data obtained it was not possible to definitively rule out pre- or postrenal causes of proteinuria. Blood pressure was not measured in the cats in this study and thus we are unable to rule out hypertension as a cause of prerenal proteinuria. Similarly, the lack of additional diagnostic tests (eg, imaging) of the lower urinary tract makes it impossible to completely exclude lower urinary tract disease as a source of postrenal proteinuria. Given that FIV-infected cats are prone to secondary inflammatory and neoplastic conditions, this may be a significant limitation; however, clinical illness, as determined by the examining veterinarian, was not a risk factor for the presence of proteinuria in FIV-infected cats. The health status classification system itself is also recognized as having significant limitations in that it was dependent on historical and physical examination findings from a number of attending referral clinicians and certain conditions may not have been recognized or recorded. However, this crude analysis would suggest that comorbidities are at least not the sole explanation for the increased prevalence of proteinuria in FIV-infected cats.
In conclusion, client-owned cats with naturally acquired FIV infection were at increased risk of proteinuria, but not renal azotemia. The etiopathogenesis of this proteinuria is uncertain but the potential homology to the glomerulonephropathies associated with HIV infection requires further investigation. Future studies regarding the histopathologic nature of this renal disease and the risk factors associated with its development and progression are warranted.
Study Funding: Study was funded by a grant from the University of Florida Resident Research Grant Fund and NIH RR-00124.
Monarch Analyzer, Instrumentation Laboratory Inc, Lexington, MA
SPSS 11.5 for Windows, SPSS Inc, Chicago, IL