icon-    folder.gif   Conference Reports for NATAP  
  18th CROI
Conference on Retroviruses
and Opportunistic Infections
Boston, MA
February 27 - March 2, 2011
Back grey_arrow_rt.gif
The Kidney at CROI
  Christina M. Wyatt, MD
Assistant Professor, Medicine/ Nephrology
Mount Sinai School of Medicine
New York, NY
Chronic Kidney Disease in HIV
As in previous years, ongoing work presented at CROI has enhanced our understanding of chronic kidney disease (CKD) epidemiology in HIV-infected populations. Data from the UK Consumer Health Information Consortium (CHIC) demonstrated a strong "dose-response" relationship between decreased kidney function and mortality in a cohort of more than 20,000 HIV-infected individuals [abstract #836]. Kidney function was calculated using the CKD Epidemiology Collaboration (CKD-EPI) estimate of glomerular filtration rate (GFR) and stratified according to the National Kidney Foundation staging system. Compared to individuals with normal kidney function (GFR 90-104 mL/min/1.73m2), those with GFR 30-59 mL/min/1.73m2 had a 1 to 2-fold increase in risk of all-cause mortality, while GFR < 30 mL/min/1.73m2 was associated with a 3 to 4-fold increase in risk. The association between decreased GFR and mortality remained significant after adjustment for other important demographic and clinical characteristics. Somewhat surprisingly, there was also a small increase in mortality risk among individuals with estimated GFR >104 mL/min/1.73m2. The authors hypothesized that the highest GFR estimates may reflect low muscle mass or hyperfiltration, with the latter classically observed in early diabetic nephropathy. Decreased GFR was also a risk factor for progression to stage 4-5 CKD (GFR < 30 mL/min/1.73m2), a finding that remained significant in adjusted analysis. Black race was an important effect modifier, increasing the risk of both mortality and CKD progression among patients with decreased GFR at baseline.
Secondary analysis of data from the Strategies for Management of Antiretroviral Therapy (SMART) trial also identified decreased GFR and black race as important risk factors for the development of end-stage renal disease (ESRD) or renal death, although race was not significantly associated with less severe GFR decline estimated using the Modification of Diet in Renal Disease (MDRD) equation [abstract #837]. Similar to the results of a previous publication using cystatin C to estimate kidney function (Mocroft et al. AIDS 2009), this analysis demonstrated an increased risk of significant GFR decline among participants randomized to CD4-guided antiretroviral therapy (ART interruption). In multivariate analysis including only participants who were randomized to continuous antiretroviral therapy, other clinical characteristics associated with the risk of GFR decline included older age, hepatitis co-infection, diabetes, and hypertension.
Hepatitis C co-infection was also associated with more rapid GFR decline in a retrospective cohort study from Canada [abstract #929]. These investigators compared the annual rate of GFR decline between patients with and without HIV-Hepatitis C co-infection, adjusting for traditional CKD risk factors and for tenofovir or atazanavir use. During a median of 2 years of follow-up, the annual decline in MDRD GFR was > 3 mL/min/1.73m2 greater in HCV co-infected individuals. Traditional risk factors for CKD such as older age and hypertension, as well as the use of tenofovir or atazanavir, were associated with lower GFR, although the relationship with GFR decline was not studied. Interestingly, black race was associated with higher GFR at baseline and throughout the follow-up period. The prevalence of HCV co-infection was lower in these patients, and it is also possible that the MDRD GFR overestimates true in HIV-infected blacks.
In a separate analysis of data from the SMART trial, cystatin C was used as an alternative to creatinine-based GFR estimates. Abnormally elevated plasma cystatin C levels (≥ 1mg/L) were associated with abnormal lipid levels and higher levels of the inflammatory markers IL-6 and CRP [abstract #839], ["Dyslipidemia has been associated with an increased risk of developing renal impairment in the general population. Both renal dysfunction and abnormal lipid levels are common in HIV infection"]. Although the association with lipid levels and inflammatory markers was not observed when the MDRD GFR was used to dichotomize kidney function, the GFR cutoff of 104 mL/min/1.73m2 is well within the normal range. It is unclear whether the observed relationship between cystatin C and lipid abnormalities reflects increased sensitivity of cystatin C for the detection of early kidney disease, or if the laboratory abnormalities reflect a common inflammatory process. These intriguing data suggest the need for future studies to delineate the relationship between kidney dysfunction, inflammation, and lipid abnormalities, as well as studies to determine the most appropriate method to estimate kidney function in HIV-infected individuals.
The question of which GFR estimate to use is particularly relevant to studies performed in non-Western populations. The HIV Netherlands Australia Thailand Research Collaboration (HIV-NAT) reported a remarkably high prevalence of CKD based on MDRD GFR estimates [abstract #843]. In this cohort of 1317 HIV-infected Thais, nearly one-third had an estimated GFR < 60 mL/min/1.73m2 after a median of 8 years on antiretroviral therapy. Factors independently associated with decreased GFR included age over 50, comorbid diabetes, previous exposure to indinavir (HR 3.1), and current use of tenofovir (HR1.6). There was no significant association with the use of atazanavir in this cohort. Of note, creatinine-based GFR estimates have not been validated in this population, and it is possible that the true prevalence of decreased GFR is overestimated by the use of MDRD GFR. Nonetheless, this study may suggest a need for enhanced screening in this population prior to the initiation of potentially nephrotoxic antiretroviral agents.
Predicting Antiretroviral Nephrotoxicity
With increasing use of tenofovir in resource-limited settings, it is important to evaluate the potential for nephrotoxicity and the need for screening and monitoring of kidney function in different populations. According to current World Health Organization guidelines, lack of resources for such screening should not preclude the initiation of therapy; when possible, screening is encouraged in patients initiating tenofovir. The identification of patients at increased risk for baseline CKD or for nephrotoxicity may allow for more targeted and cost-effective laboratory testing.
Screening data from several large MTCT and HIV treatment trials were combined to evaluate the prevalence and predictors of decreased kidney function in treatment-naïve adults in Malawi [abstract #838]. In this study, 99% of patients had a Cockcroft-Gault creatinine clearance (CrCl) above the threshold for dose adjustment of renally cleared antiretroviral agents such as tenofovir (50 mL/min). Low body mass index (BMI) was associated with increased odds of decreased CrCl in adjusted analysis. While it is likely that sicker individuals with lower BMI are at increased risk for CKD, it is not clear how the inclusion of weight in the calculation of both BMI and CrCl may have influenced the observed relationship. While the results of this study are reassuring, it is also important to note that creatinine-based GFR estimates such as the Cockcroft-Gault CrCl may substantially overestimate true GFR in the setting of pregnancy.
Anticipating a significant increase in the use of tenofovir following a recent change in national treatment guidelines, investigators from South Africa evaluated the potential implications of widespread tenofovir use in that country [abstract #840]. In a retrospective cohort of 890 patients, the CrCl was between 30-59 mL/min in 5% and between 60-89 mL/min in 30% of patients prior to initiation of tenofovir. Decreased CrCl at baseline was associated with a significant increase in the risk of both mortality and subsequent kidney disease, defined as a CrCl <30mL/min or a clinical diagnosis of CKD or acute kidney injury. Other factors associated with the development of kidney disease during tenofovir treatment included older age, anemia, and more advanced HIV disease. Interestingly, antiretroviral-naïve patients were also at increased risk for subsequent kidney disease. Data on the etiology of kidney disease were not available, so it is not clear what proportion of cases were attributable to tenofovir (from Jules: abstract "Much of the incident renal dysfunction in TDF-treated patients is likely related to preexisting renal pathology, which may be exacerbated by TDF"). Nonetheless, the strong association with baseline CrCl does support screening for kidney disease prior to the initiation of tenofovir, at least when feasible.
The impact of tenofovir plasma concentration on the risk of nephrotoxicity was evaluated in a retrospective cohort of 163 patients who had at least one tenofovir trough (≥17 hours after last dose) performed as part of routine clinical care [abstract #842]. Comorbid risk factors for kidney disease, including HCV co-infection, diabetes, and hypertension, were common in this population, and 64% of patients were receiving concomitant boosted PI. The prevalence of comorbid conditions and PI use was similar among patients with low (<40ng/mL), therapeutic (40-90ng/mL) and high (>90ng/mL) tenofovir trough levels. In adjusted analysis, a high tenofovir trough level was significantly associated with decline in CrCl over one year among women, while duration of exposure to tenofovir was associated with decline in CrCl among men. Other risk factors for decline in CrCl included older age and decreased CrCl prior to tenofovir initiation.
Another small cohort study evaluated liver fibrosis as a potential risk factor for tenofovir toxicity in the setting of HIV-Hepatitis B virus (HBV) co-infection [abstract #977]. The investigators used a CKD-EPI GFR < 80 mL/min/1.73m2 as their primary outcome, and analyzed the association with a commercially available biomarker of hepatic fibrosis (FibroMeter®) in 137 patients with HIV-HBV co-infection. After adjusting for demographics, HIV viral load, duration of HBV co-infection, and concomitant PI use, higher fibrosis stage was independently associated with an increased risk of GFR decline. Although the clinical significance of GFR changes in the normal range is not clear, this study suggests the need for future studies to consider the potential influence of advanced liver disease in promoting tenofovir nephrotoxicity. If validated in other populations, this finding may also have implications for the large number of patients with liver disease related to HCV co-infection.
The impact of concomitant antiretroviral agents on the risk of GFR decline was evaluated in 322 antiretroviral-naïve patients randomized to tenofovir/ FTC in combination with efavirenz, boosted atazanavir, or zidovudine/abacavir [abstract #841]. There was a significant decline in GFR from baseline to 48 weeks in subjects randomized to tenofovir/ FTC in combination with boosted atazanavir (from Jules: this study did not evaluate other boosted PIs), with no change in GFR in the other treatment arms and no further decline between 48-96 weeks. The association between concomitant boosted atazanavir use and GFR decline remained significant in adjusted analysis. Of note, the observed GFR declines were mostly within the normal range, and there was no difference in the incidence of documented renal adverse events between the treatment arms. Although the clinical significance is unclear, these results suggest a need for further studies to evaluate the potential for boosted protease inhibitors, and atazanavir in particular, to promote GFR decline or tenofovir toxicity.