Epidemiological, clinical, and laboratory factors associated with chronic kidney disease in Mexican HIV-infected patients

ABSTRACT Aim: To determine the prevalence of chronic kidney disease (CKD) and the epidemiological, clinical, and laboratory factors associated with CKD in Mexican HIV-infected patients. Methods: Cross-sectional study. We included 274 patients with HIV/AIDS. CKD was defined by the estimated glomerular filtration rate (eGFR < 60 mL/min/1.73 m2 assessed by CKD-EPI) and albuminuria criteria from KDIGO guidelines. Clinical, epidemiological, and laboratory characteristics were compared between patients with and without CKD. The factors associated with CKD were assessed by logistic regression analysis. Results: The mean age was 41±11 years, and 72.3% of the patients were men. The global prevalence of CKD was 11.7% (n = 32); 7.2% (n = 20) were defined by eGFR criterion; 7.6% (n = 21), by the albuminuria criterion; and 3.2% (n = 9), by both CKD criteria. The most frequently observed stages of CKD were KDIGO G3A1 stage with 4.7% (n = 13), KDIGO G1A2 stage with 3.6% (n = 10) and KDIGO G3A2 stage with 1.7% (n = 5). The factors associated with CKD were use of abacavir/lamivudine (OR 3.2; 95% CI 1.1-8.9; p = 0.03), a CD4 lymphocyte count < 400 cells/µL (OR 2.6; 95% 1.03-6.4, p = 0.04), age (OR 1.1; 95% CI 1.04-1.2, p = 0.001) and albuminuria (OR 19.98; 95% CI: 5.5-72.2; p < 0.001). Conclusions: CKD was a frequent complication in HIV-infected patients. These findings confirm the importance of screening and the early detection of CKD, as well as the importance of identifying and treating traditional and non-traditional risk factors associated with CKD.


IntRoductIon
Significant advances in antiretroviral therapy have deceased the progression of disease and improved survival of HIV-infected patients. In fact, chronic kidney disease (CKD) has arisen as one of the leading noninfectious conditions affecting HIV-infected persons. 1 The reported prevalence of CKD in patients infected with HIV in North America and Europe ranges from 4.7% to 9.7%, and higher rates have been reported when CKD is defined by either estimated glomerular filtration rate (eGFR) or proteinuria. [2][3][4] HIV infection is a well-established risk factor for CKD and endstage renal disease (ESRD).
The majority of CKD cases in HIV infection are due to HIV-associated nephropathy (HIV-AN). However, up to 50% of kidney diseases in HIV-infected persons result from a wide array of non-HIVAN pathology. 5 These patients may develop multiple glomerular nephropathies (IgA nephropathy, lupus-like glomerulonephritis, focal and segmental glomerulosclerosis, membranoproliferative glomerulonephritis with cryoglobulinemia, and membranous glomerulopathy) and vascular (thrombotic microangiopathy), tubulointerstitial (tubular nephropathy by drugs, tubulointerstitial immunoallergic nephritis, and Fanconi syndrome), and obstructive nephropathies (nephropathy due to crystal deposition) related to the virus itself, the drugs administered or the coinfections. 6 On the other hand, traditional risk factors for CKD are becoming increasingly prevalent in HIV-infected populations, including aging, diabetes mellitus, hypertension, cardiovascular disease, previous AKI and race/ ethnicity. 7 This has caused the risk factors for CKD in HIV-infected persons to be a combination of traditional and HIV-related factors, including low CD4 counts, high viral load, intravenous drug use, hepatitis C virus (HCV) coinfection, and use of specific antiretroviral drugs. [8][9][10][11][12] Finally, as in the general population, albuminuria and decreased kidney function in HIV-positive individuals have been associated with worse outcomes such as a progression to AIDS and death. 13,14 The aim of our study was to assess the prevalence of CKD and to determine epidemiological, clinical, and laboratory factors associated with CKD in Mexican HIV-infected patients. Descriptive statistics with means and standard deviations were used to describe continuous variables; frequencies and proportions were used to describe categorical variables. Comparisons between groups were performed using Student's t-test and one-way ANOVA for continuous variables and the χ 2 test for categorical variables. The study of factors associated with CKD was performed by multivariate logistic regression. Variables with p < 0.05 in the univariate analysis and those that have been consistently associated with CKD in the medical literature were included in the multivariate analysis. A value of p < 0.05 was considered statistically significant.

general pOpulatiOn characteristics
We included 274 patients with an average age of 41 ± 11 years; of these patients, 72.3% (n = 198) were males. The most frequently observed comorbidities were smoking in 28.5% (n = 78), drug abuse in 14.6% (n = 40), hypertension in 7.7% (n = 21), and diabetes mellitus in 6.9% (n = 19) of the patients. The main viral coinfection in our population was HCV in 2.9% (n = 8) of the patients.

cOmparisOn Of clinical, labOratOry, treatment, and hiV-infectiOn characteristics between patients with and withOut ckd
When comparing the general characteristics between CKD patients vs non-CKD patients, we observed that patients with albuminuria-CKD criterion were younger than patients with eGFR-CKD criterion (35 vs. 51 years, p = 0.001), eGFR + albuminuria CKD criteria (35 vs. 49, p = 0.001), and non-CKD patients (35 vs 40, p = 0.001). No statistically significant difference was observed in the other general characteristics shown in Table 2.
Comparing medical prescriptions between groups, we observed that the use of angiotensin II receptor antagonism (ARA) (11.1%, n = 1 vs. 0.8%, n = 2; p < 0.01) and antiretroviral treatment with saquinavir/ ritonavir (22.2%, n = 2 vs. 3.3%, n = 8; p = 0.03) were more frequent in patients with CKD by eGFR and /or albuminuria criteria than in patients without CKD. No statistically significant difference was found for the remaining medications shown in Table 3.

factOrs assOciated with ckd
In the multivariate logistic regression analysis, the factors associated with CKD defined by the eGFR and/or albuminuria in our HIV population were abacavir /lamivudine treatment, with an OR of 3.2 (95% CI of 1.1-8.9; p=0.03), and a CD4 lymphocyte count < 400 cells/μL, with an OR of 2.6 (95% CI of 1.03-6.4, p = 0.04). On the other hand, the factors associated with CKD defined only by the eGFR were age, with OR of 1.

dIscussIon
The global prevalence of CKD in our HIV population was 11.7% (n = 32); 7.2% (n = 20) of the patients were diagnosed by the eGFR criterion, and 7.6% (n = 21) of the patients were diagnosed by the albuminuria criterion. These findings are in line with previous reports by other researchers worldwide, ranging from 7.2% (USA) to 13.7% (China) applying the albuminuria criterion and from 3.5% (Europe) to 9.7% (USA) considering the eGFR criterion < 60 mL/min. On the other hand, the global prevalence of CKD (eGFR and   /or albuminuria) in our population was less than that reported in Japan (15.5%) and USA (23.7%), although globally, this varies from 2 to 30%. 18 Coinfection with HCV was 2.9% (n = 8), which was lower than previous reports (30%, with a range between 3-70%). 19 On the other hand, HBV coinfection was 0.7% (n = 2) in our population, which is below the prevalence described by other authors. 20,21 Raboni et al. conducted a cross-sectional study in southern Brazil in which the authors documented hepatitis/HIV coinfection in 6.6% of HIV-infected patients. The HCV/HIV coinfection was confirmed in 4.0% and HBV/HIV coinfection in 1.3% of HIVinfected patients. 22 Previous clinical studies have described multiple traditional risk factors (age, hypertension, diabetes, proteinuria, etc.) and non-traditional risk factors specific to HIV patients (elevated viral load, low CD4 cell count, coinfection with HBV and HCV, drugs, etc.) associated with the development of CKD. [7][8][9][10][11] Age was a factor associated with CKD in our population. This finding is similar to those reported by other studies, being that the relative risk increase for CKD (defined as eGFR or albuminuria) for every 10 years of age is from 1.5 to 5.5. When age was analyzed as a dichotomous variable by other authors, patients > 50 years presented a relative risk of 2.0 for progression to ESRD or an eGFR < 15 mL/min compared to patients < 30 years. 16 A CD4 lymphocyte count < 400 cells/μL was another factor associated with CKD (defined by GFR or albuminuria) in our population. This has been consistently demonstrated by other investigators, reporting a relative risk range for CKD (defined by GFR or albuminuria) from 1.1 to 1.25 for each decrease of 100 cells/μL or from 1.4 to 2.2 for patients with CD4 < 200 vs. > 201 cells/μL. On the other hand, the CD4 count has also been associated with the progression to ESRD or a GFR < 15 mL/min, with a relative risk of 1.7 per 100 cells/μL decrease and 1.4 to 2.7-fold increased risk for patients with CD4 < 200 vs. > 201 cells/μL. 16 Another study documented a decreased CKD risk in association with CD4 cell restoration and plasma viral load suppression during treatment with HAART. 23 In our HIV population, the presence of albuminuria was the main factor associated with CKD (defined by the eGFR criterion). The presence of albuminuria is frequent and is observed in up to a third of patients infected with HIV. Multiple observational studies in the general population and cohorts of HIVinfected patients demonstrate a strong and consistent association between the presence of albuminuria and clinical outcomes in these patients, such as progression to ESRD, cardiovascular disease, AIDS, and death. 14,[24][25][26][27] Treatment with abacavir/lamivudine was another factor associated with CKD in our population. Although abacavir and lamivudine have shown a low nephrotoxicity in some clinical studies, abacavir has been associated with tubular nephrotoxicity and acute tubulointerstitial nephritis and lamivudine with Fanconi syndrome and nephrogenic diabetes insipidus. 28,29 Takeshi Nishijima et al. compared the deterioration of eGFR (a 25% decrease in baseline renal function) between HIV-infected patients treated with an antiretroviral regimen based on a tenofovir-vs. an abacavir-based regimen in a cohort of 503 Japanese patients. Although impairment of renal function was greater in patients under the tenofovir regimen (22.1%), impairment of renal function was also observed in patients under the abacavir regimen (13.5%). 30 In the AIDS Clinical Trials Group Study 5202 (ACTG), the frequency of Fanconi syndrome, toxic nephropathy, proteinuria, and renal failure was similar between patients treated with tenofovir/emtricitabine and those treated with abacavir/ lamivudine. 31 These findings confirm that although abacavir has a lower nephrotoxicity profile, it is still a nephrotoxic antiretroviral drug.
Treatment with saquinavir/ritonavir (OR 3.3, 95% CI 0.8-13.3, p = 0.09) and tenofovir/emtricitabine (OR 2.2, 95% CI 0.9-5.6, p = 0.09) showed a statistical trend for CKD risk defined by eGFR and albuminuria. These findings have been confirmed by Mocroft et al., who studied the cumulative nephrotoxic effect of the major antiretroviral drugs used in a cohort of 23,005 HIV-infected patients in the United States, Europe, and Australia. The authors documented an increase in the risk of CKD (defined by eGFR < 60 mL/min) of 14, 20, and 11% for each year of exposure only for tenofovir, ritonavir/atazanavir and ritonavir/lopinavir, respectively. The increase in the incidence of CKD with chronic exposure to these antiretrovirals over the 6-year follow-up, indicates a cumulative nephrotoxic effect of tenofovir, ritonavir/atazinovir, and ritonavir/lopinavir. 32 However, saquinavir has been shown to be a safe antiretroviral drug from the point of view of nephrotoxicity, so the risk of CKD of the saquinavir/ritonavir combination observed in our population is probably due to ritonavir. 33 Regarding the weaknesses of our work, the retrospective nature of the study stands out with all the limitations that this entails for obtaining clinical, laboratory, and antiretroviral treatment information. It was not possible to assess the effect of coinfection with HBV and HCV in the logistic regression model due to the small number of patients coinfected with these viruses in our population. It was also not possible to assess the effect of other nephrotoxic drugs frequently used by HIV-infected patients, such as trimetropim/sulfamethoxazole, acyclovir, amphotericin B, and NSAIDs.

conclusIons
CKD was a frequent complication in the studied Mexican HIV-infected patients. The presence of albuminuria, age, CD4+ lymphocyte count, and abacavir/lamivudine treatment were the factors associated with CKD in our population. These findings confirm the importance of screening and early detection of CKD, as well as the identification and treatment of traditional and non-traditional risk factors associated with CKD. This could prevent or slow further decline in kidney function and improve outcomes in HIVinfected patients.