icon- folder.gif   Conference Reports for NATAP  
  16th CROI
Conference on Retroviruses and Opportunistic Infections Montreal, Canada
February 8-11, 2009
Back grey_arrow_rt.gif
Update on Kidney Disease in HIV Infection: CROI 2009
  CROI 2009 Feb 8-12 Montreal
Christina M. Wyatt and Paul E. Klotman, Mt Sinai Hospital, New York City
The growing relevance of non-AIDS conditions was again evident in this year's program, which included abstract sessions devoted to cardiovascular, hepatic, bone, and kidney disease, and concluded with a clinical symposium on non-AIDS complications. Interested clinicians are encouraged to refer to Lynda Szczech's presentation for a review of practical considerations in the management of acute and chronic kidney disease in the aging HIV population [abstract 180]. (from Jules: as the author's of this paper say at this year's CROI there was increased attention to so-called non-AIDS conditions, although they are tied to HIV in I think several ways. From looking at all the studies confirming non-AIDS events and deaths it appears evident that the ages of patients in the studies is young, in their 40s on average, which speaks volumes but is not said in a loud way. It appears that HIV plays a role in leading to premature aging of T-cells, upon initial infection HIV gets into various organs including the kidney, the heart, vascular system, the brain, the CNS, and more. HIV appears to cause immune activation and inflammation. So, although researchers are paying attention to increased morbidity and mortality due to non-AIDS researchers are not discussing understanding these underlying preocesses in such a way as to try to better understand it so we could perhaps address it and intervene, and slow aging and the development of non-AIDS events and death.)
Kidney Disease and Non-AIDS Morbidity
Investigators from several large cohort studies described the important contribution of non-AIDS conditions such as kidney disease to morbidity and mortality. Only half of all deaths in the ART Cohort Collaboration were considered AIDS-related [abstract 708]. While non-AIDS malignancy and infections were the most common causes of non-AIDS mortality, liver, cardiovascular, and renal disease were also important contributors. Death from renal disease was associated with advanced immunodeficiency, although it is unclear whether this reflects increased frequency and severity of kidney disease, in particular HIV-associated nephropathy (HIVAN), or a tendency to withhold dialysis in patients with advanced AIDS and severe comorbid disease.
The EuroSIDA investigators also reported more morbidity related to serious non-AIDS conditions than AIDS-defining illnesses [abstract 707]. The most common non-AIDS conditions were malignancy, cardiovascular disease, and liver failure, but end-stage renal disease was also an important contributor to morbidity in this cohort. The development of end-stage renal disease is of particular relevance in this largely Caucasian European cohort at relatively low risk for progressive kidney disease.
Prospective data from the Johns Hopkins HIV Clinical Cohort demonstrated a high hospitalization rate among ART-naïve patients within the first year of initiating therapy, with a similar proportion of those hospitalizations (approximately 4%) attributed to cardiovascular disease and renal or genitourinary disease [abstract 711]. Hospitalization rates for renal disease did not decline substantially in the first year following ART initiation. There was a modest decline in renal hospitalization rates among patients with a virologic response to therapy, which may reflect an improvement in kidney disease risk, particularly related to HIVAN.
The SMART investigators have previously described a higher incidence of severe renal events in patients assigned to CD4-guided interruption of ART, compared to those on continuous suppressive ART. In a biomarker substudy, markers of inflammation (IL-6, CRP, and D-dimer) and renal function (cystatin C) were compared between a subgroup of SMART participants and over 5,000 HIV-negative participants enrolled in the Multi-Ethnic Study of Atherosclerosis (MESA) trial [abstract 740]. The SMART participants had significantly higher median levels of all four biomarkers at study entry, which may predict an increased risk of cardiovascular disease and kidney disease compared to the general population.
Endothelial dysfunction has been proposed as a potential link between renal and cardiovascular disease, a hypothesis explored by Samir Gupta and colleagues [abstract 746]. As measured by flow-mediated dilatation of the brachial artery, endothelial dysfunction was not associated with estimated glomerular filtration rate (eGFR) or creatinine clearance in HIV-infected patients without traditional renal or cardiovascular risk factors, but was weakly associated with proteinuria or albuminuria. In adjusted analysis, endothelial dysfunction was no longer significantly associated with markers of kidney disease. Longitudinal studies such as WIHS have suggested a link between overt kidney disease and adverse events among HIV-infected patients, but further studies are needed to determine whether microalbuminuria and other markers of subclinical kidney disease are predictive of outcomes in this population.
Chronic Kidney Disease in HIV
Andy Choi and colleagues described an accelerated loss of kidney function among HIV-positive patients enrolled in the UCSF SCOPE Cohort, compared to general population estimates [abstract 38]. Over a median followup of 2.7 years, the decline in estimated glomerular filtration rate (eGFR) was -2.6mL/min/1.73m2 per year, a figure comparable to the rate of decline observed in many patients with established kidney disease. Of note, chronic kidney disease and traditional risk factors, such as hypertension and hyperlipidemia, were relatively common. Older age, female gender, diabetes, and hyperlipidemia were associated with eGFR decline, while HIV-related factors did not predict the rate of decline. In particular, there was no relationship between eGFR decline and the use of ART, although the study was not powered to look at the impact of specific drugs or drug classes. In a subgroup of 82 patients who initiated ART during the study period, the decline in eGFR was slower after the initiation of ART, suggesting a beneficial effect of virologic suppression. At the same time, a separate subgroup analysis revealed that patients who achieved virologic suppression on ART had more rapid eGFR decline than elite controllers. It is not clear whether this difference reflects a balance of benefit and harm associated with ART, or whether there are other unmeasured differences between these two populations that may impact kidney disease progression.
Plasma viral RNA may not accurately reflect viral replication in renal epithelial cells, which has been shown to induce local inflammation and apoptosis. The HIV genes responsible for these effects are not known, although both Vpr and Nef are known to be involved in the pathogenesis of HIVAN. Snyder and colleagues demonstrated a role for the ubiquitin-like protein FAT-10 in Vpr-induced apoptosis of renal tubular cells [abstract 232]. The implications of these findings could extend beyond HIVAN, and may provide additional insight into the deleterious effects of ART interruption.
Acute Renal Failure and Nucleoside Toxicity
In a prospective cohort study from Boston, Shikha Garg and colleagues reported a higher incidence of acute renal failure among patients with HIV-hepatitis C co-infection compared to patients with hepatitis C alone [abstract 822]. More than one-third of co-infected patients developed at least one episode of acute renal failure during the 7-year study period, for an incidence of 8.7 cases/100 person-years. Previous studies have demonstrated a similar increase in the incidence of acute renal failure among patients with HIV-hepatitis C co-infection compared to those with HIV alone, although the relative contribution of each viral infection is not known. Decompensated cirrhosis and cocaine use were also independently associated with acute renal failure, suggesting that factors associated with acquisition and complications of viral infection are important contributors to the risk of acute renal failure in this population.
Several abstracts addressed renal adverse events and tubular dysfunction in patients treated with tenofovir. The HEAT investigators have previously reported no difference in the incidence of proximal tubular dysfunction at 48 weeks between ART-naïve patients randomized to tenofovir/ FTC or abacavir/ 3TC in combination with boosted lopinavir. After 96 weeks of follow-up, there was no significant difference in the number of subjects experiencing any renal adverse event, although grade 3-4 renal events were more common in the tenofovir arm [abstract 744]. There was also a small increase in the number of subjects who progressed to a more advanced stage of chronic kidney disease in the tenofovir arm, but there was no decrease in mean eGFR. All 5 cases of study-defined tubular dysfunction and both study withdrawals resulting from a renal adverse event occurred in the tenofovir arm. The small but notable number of clinically relevant renal events observed in the tenofovir arm of HEAT is more consistent with observations in cohort studies than with pre-marketing randomized clinical trials. At the same time, there was actually a small increase in mean eGFR in the ART-naïve HEAT study population, consistent with improvements in eGFR observed in other populations initiating ART.
All subjects in HEAT received tenofovir in combination with boosted lopinavir. Plasma tenofovir levels have been shown to be higher in patients receiving some boosted PIs, and it has been suggested that concurrent boosted PI use may be a risk factor for tenofovir toxicity. Investigators from the Swiss HIV Cohort Study reported a synergistic effect of tenofovir and PI use on the prevalence of proximal tubular dysfunction, as defined by the abnormal urinary excretion of at least 3 of the following: phosphate, uric acid, protein, or glucose [abstract 743]. The most common urinary abnormality was a relative increase in the fractional excretion of phosphate, which was observed in 42% of patients receiving combination therapy with tenofovir and any PI, compared to only 6% of ART-naïve patients. Tubular dysfunction was identified in 12% of patients on concurrent tenofovir-PI therapy and in no ART-naïve patients. In addition to tenofovir use and PI use, older age and lower BMI were also independently associated with tubular dysfunction. In a subgroup of patients with repeat testing after a median of 10 months, at least one urinary parameter was persistently abnormal in 71% of those with 3 abnormalities at baseline and in 54% of those with only increased fractional excretion of phosphate at baseline.
Sonia Rodriguez Novoa and colleagues have also demonstrated an increased prevalence of tubular dysfunction among tenofovir-treated patients, defined by the abnormal urinary excretion of at least 2 of the following: phosphate, uric acid, glucose, amino acids, or beta-2 microglobulinuria [Labarga et al. AIDS 2009, in press]. Increased urinary phosphate excretion was also the most common urinary abnormality observed in this population. The longterm implications of increased urinary phosphate excretion were not addressed in either of these cross-sectional studies, and it is not known where this abnormality predicts progressive decline in renal function or bone density.
A pharmacogenomic substudy was performed in 115 tenofovir-treated subjects, including 19 with tubular dysfunction [abstract 37]. Although these investigators demonstrated an association with polymorphisms in the gene encoding MRP2 (ABCC2 positions 24, 1249, 3563, and 3972), there was no association with the CATC haplotype implicated in a previous pharmacogenomic study [Izzedine et al, JID 2006]. Instead, the TGTT haplotype was found at lower frequency among subjects with tubular dysfunction. The CC genotype at ABCC2-24 was independently associated with tubular dysfunction, although the relevance of individual SNP associations is controversial. The mechanism of the observed association with the ABCC2 gene is not known, since tenofovir is not a substrate for MRP2. There was no significant relationship between tubular dysfunction and polymorphisms in the genes encoding OAT-1 and MRP4, which are known to be involved in the tubular transport of tenofovir.
In a second substudy, Rodriguez Novoa and colleagues reported higher plasma levels of tenofovir in subjects who developed tubular dysfunction [abstract 745], with plasma levels > 160 ng/mL independently associated with the presence of tubular abnormalities. While the findings of pharmacokinetic and genetic association studies may provide insight into the pathogenesis of tenofovir-associated tubular dysfunction, the current findings do not have immediate implications for clinical management, and further studies are needed to determine the influence of genetic and pharmacokinetic factors on the risk of toxicity in individual patients As in the Swiss HIV Cohort Study, tubular dysfunction was also associated with older age and lower body weight. Unlike cumbersome and expensive pharmacogenomic testing and therapeutic drug monitoring, demographic and anthropomorphic data are routinely available even in resource-limited settings, and may be useful to identify patients at increased risk of toxicity who should be followed more closely.
Kidney Disease and ART in Resource-Limited Settings
With expanding access to ART in sub-Saharan Africa and other resource-poor regions, it is increasingly important to develop simple, inexpensive, and sensitive methods for the prevention and early recognition of ART toxicity. It is reassuring that the widespread introduction of tenofovir-containing ART in Zambia has been associated with low rates of severe renal insufficiency (creatinine clearance < 50 mL/min) comparable to that observed with alternative NRTI [abstract 142]. Longitudinal data from ART-naïve patients in Kenya suggest that baseline kidney function may predict progression of HIV disease and identify patients who would benefit from earlier initiation of ART [abstract 741], providing another potential rationale for the routine assessment of kidney function in this population. In a retrospective cohort of 8,737 patients with a CD4 above 200 enrolled in the USAID-AMPATH Partnership, a striking proportion of patients (almost 20%) had an estimated creatinine clearance below 60mL/min. Lower creatinine clearance was associated with more rapid progression to WHO stage 3-4 disease, CD4 below 200, and death. Data on progression of kidney disease were not reported.
While SMART established a clear benefit of continuous ART on the incidence of renal adverse events, data on kidney disease progression and change in kidney function were not prospectively collected. In a randomized trial of continuous versus interrupted ART in Uganda, structured interruption of ART (Combivir/ nevirapine) was associated with more rapid decline in eGFR, with the most dramatic decline in subjects randomized to 7 days on/ 7 days off [abstract 742]. Data on traditional risk factors for progressive kidney disease were not collected in this small study, and the impact of regimens including tenofovir or indinavir is not known. Nonetheless, these data are consistent with the rise in cystatin C noted in the treatment interruption arm of SMART [Mocroft et al, AIDS 2009], and suggest an important area for future study. Treatment interruption remained independently associated with eGFR decline after adjusting for HIV viral load, suggesting that systemic viral replication is not the only factor contributing to the observed association.
The high prevalence of decreased eGFR and the rapid rates of eGFR decline observed in these two African cohorts is particularly striking in light of the young age of both populations (median 35-39 years). These data are consistent with the accelerated eGFR decline observed in the UCSF SCOPE Cohort (mean age 45 years), and provide further support for the hypothesis that HIV infection may be associated with the earlier development of chronic diseases more commonly associated with aging.