|Home | About | Journals | Submit | Contact Us | Français|
Direct effects of HIV-1 infection on the kidney combine with immune and genetic factors, comorbidities, coinfections, and medication toxicities to induce a spectrum of kidney disorders in HIV disease. The most dramatic of these, HIV associated nephropathy (HIVAN), emerges almost exclusively in persons of African descent and is associated with rapid progression to end-stage renal disease in the absence of antiretroviral therapy (ART). ART modifies the natural history of HIVAN, but the renal benefits of ART may not be limited to HIVAN. ART is often under prescribed or incorrectly dosed in persons with kidney disease, and kidney disease is a marker of accelerated HIV disease progression. Vigilant attention to kidney function and an understanding of the complex associations involving the kidneys is necessary for optimal care of these patients.
HIV-1 viral replication and immune dysregulation combine with high rates of diabetes, hypertension, and hepatitis C coinfection in an increasingly older population of HIV infected patients portending a growing importance of kidney disease in these patients. Acute and chronic kidney disease predicts accelerated HIV disease progression and increased mortality [1–5]. Although kidney damage may contribute directly to these outcomes in some cases, markers of renal function also may represent markers of HIV disease activity.
Chronic kidney disease (CKD) is defined as renal damage that persists for ≥3 months . Renal damage is assessed by protein in the urine, and/or reduced renal function based on the creatinine clearance (CLCr), estimated by the Cockcroft-Gault equation, or estimated glomerular filtration rate (eGFR) by the Modification of Diet and in Renal Disease (MDRD) equation [Table 1]. The MDRD equation was developed in patients with CKD and may overestimate GFR in patients with normal renal function, however, and both equations are affected by changes in muscle mass. Cystatin-C is an alternative marker of kidney function that does not depend on muscle mass and is more sensitive for kidney disease than creatinine-based measures; concentrations > 1.0 mg/dL are associated with kidney disease .
Although none of these renal indices have been systematically validated in HIV infected patients, they were compared with direct GFR measurements using [125I]-Iothalamate (mGFR) in a small substudy of an antiretroviral treatment (ART) trial in subjects with predominately normal renal function, in which CLCr was a better estimate of mGFR (R2 between mGFR and estimates by Cockcroft-Gault, MDRD and cystatin-C: 0.74, 0.38, 0.31, respectively; mean mGFR minus these respective estimates: 0, −42, −41 ml/min) .
Microalbuminuria (defined as > 50 mg/L) and proteinuria (comprised of albumin and a mixture of other serum proteins in adults) are important markers of kidney damage. To overcome variability of urine concentrations, the ratio of urine protein-to-creatinine (P/Cr) is a reliable indicator of proteinuria, with ratios above 0.2 indicative of kidney damage .
Acute renal failure (ARF) complicated 6% of adult acute care hospitalizations of HIV infected patients in New York State, was three times more likely than in HIV uninfected patients, and was associated with a 27% in-hospital mortality . Acute tubular necrosis, pre-renal azotemia, and obstructive uropathy accounted for 43%, 39%, and 8% of ARF cases, respectively, among patients at the University of North Carolina . Medications contributed to 32% of these cases including antibiotics, non-steroidal anti-inflammatory agents, radiologic contrast dye and lithium.
In addition to traditional risk factors for kidney disease such as older age, diabetes, hepatitis C co-infection or liver disease, HIV associated risk factors for ARF also included an AIDS-defining illness, low CD4 cell counts (< 200/μL), and high plasma HIV RNA, and current or prior ART use [3, 9]. Indinivir, tenofovir and nevaripine were among the antiretroviral agents associated with ARF.
ARF has been reported in the setting of immune reconstitution inflammatory syndrome with tuberculosis and Mycobacterium avium-complex . Kidney biopsies revealed non-caseating granulomatous inflammation without demonstrable organisms, and kidney function normalized in all cases with the addition of prednisone.
As with ARF, HIV associated factors also combine with traditional risk factors for kidney disease, resulting in a spectrum of CKD in HIV infected patients. ART may mitigate, but it does not eliminate this risk.
The prevalence of Stage ≥3 CKD in resources-rich settings ranges from 3.5%, in the EuroSIDA cohort of predominately white Europeans , to 15.5%, in a large urban cohort with 90% African and Hispanic ethnic representation from the Mount Sinai AIDS Center in New York . HIV infected subjects had significantly higher cystatin-C levels compared to HIV-uninfected controls (31% vs. 4%, respectively with cystatin-C >1.0 mg/L) in the Fat Redistribution and Metabolic Change in HIV Infection cohort study . Included among the correlates of CKD were older age, African ancestry, injection drug use or HCV infection, and hypertension. HIV- associated factors included higher plasma HIV-1 RNA levels, lower CD4 cell counts, and in the EuroSIDA cohort, exposure to tenofovir or indinavir.
Renal replacement therapy for ESRD was initiated in nearly 1% of HIV infected persons each year, in a large prospective cohort of African-Americans in Baltimore (ALIVE) that included HIV uninfected, but at risk persons . In this cohort, HIV infection was associated with substantially higher rates of ESRD (standardized incidence rates 2.3, 6.9 and 16.1 for HIV uninfected, HIV infected without AIDS, and AIDS, respectively), and although the incidence of ≥ Stage 3 CRF was lower in the highly active antiretroviral therapy (HAART)-era compared to the pre HAART-era (adjusted relative risk 0.64), the incidence of ESRD did not differ between these intervals as a result of an increase in the prevalence of CRF due to improved survival with HAART. Notably, progression to ESRD from stage 3 CKD was almost exclusively observed among African-Americans (99 of 253 African Americans with CKD vs. 1 of 31 white subjects) in a second, Johns Hopkins HIV Clinical cohort .
Proteinuria or elevated serum creatinine was independently associated with increased mortality HIV Epidemiology Research Study (HERS), a multicenter prospective cohort of HIV-infected women, 78% of whom were African-American or Hispanic, despite adjusting for demographic, viral and immune factors, comorbidities, and ART use , and each of these renal indices was associated a greater risk of progression to AIDS or death in a second prospective, multicenter US cohort of HIV infected women (Women’s Interagency HIV Study) .
A clear picture of the burden of kidney disease in sub-Saharan Africa is yet to emerge, but reduced access to care and advanced HIV disease stage may contribute importantly to the risk of CKD in this region. The prevalence of ≥stage 3 CKD from large cohorts ranged from 7%, in participants of the randomized Development of Antiviral Therapy Trial (DART) in Uganda and Zimbabwe, where lower eGFR was observed in persons from rural, compared to the urban centers , to 26.5% in ART naïve adults enrolled in a large scale, urban ART initiation program in Zambia .
Baseline renal insufficiency as determined by CLCr was an independent predictor of early mortality (< 90 days) after ART initiation, in the Zambian observational cohort (adjusted hazard ratio of death compared to ≥ 90 ml/min: 1.7, 2.3, and 4.3 in association with baseline CLCr 60–89, 30–59 and < 30 ml/min, respectively) , and renal insufficiency predicted more rapid HIV disease progression and higher mortality in a large HIV-care program in western Kenya . Although these studies did not establish a causal relationship between renal dysfunction and mortality, they identify renal function as a marker of HIV disease activity.
There are several glomerular disorders associated with HIV infection including a thrombotic microangiopathy, a variety of immune complex glomerulonephritidies collectively known as HIV-immune complex kidney disease (HIVICK), and the more commonly occurring HIV-associated nephropathy (HIVAN), a collapsing glomerulopathy.
HIVAN is a form of focal segmental glomerulosclerosis (FSGS) that involves collapse of the glomerular tuft (Fig. 1). Unlike more typical forms of FSGS, the glomerular podocytes do not atrophy and die, but proliferate, hypertrophy and dedifferentiated. In addition to glomerular changes, HIVAN kidneys also have significant tubulointerstitial pathology. All segments of the tubules can become dilated or microcystic and may contain protein casts. There are varying degrees of immune cell infiltrates that are comprised of CD4+ T cells, CD8+ T cells and macrophages , as well as varying degrees of interstitial fibrosis and edema. HIVAN typically presents in advanced HIV disease with CD4 cell counts < 200/μL, and occurs almost exclusively occurs in Blacks of African descent suggesting a genetic basis of disease. Recent human genetic studies have linked FSGS development (including HIVAN patients) to the gene MYH9 which codes for a non-muscle myosin, however, the mechanism by which this gene causes FSGS remains elusive [17, 18]. For more information on the genetics of HIV-related kidney diseases, please see the recent in-depth review by Kiryluk et al. .
Glomerular injury in HIVICK is caused by immune mediated mechanisms and the spectrum of glomerulonephritidies includes; IgA nephropathy, membranoproliferative glomerulonephritis, membranous nephropathy, and a lupus-like glomerulonephritis . HIVICK occurs more frequently in Caucasians and is associated with concurrent hepatitis B or C infections without serological evidence of lupus [21, 22]. Subepithelial and/or subendothelial immune deposits are composed of HIV antigens, immunoglobulin complexes frequently of IgA, and complement. Unlike HIVAN, the glomerular lesions are typically characterized by mesangial expansion with crescents, and although tubulointerstitial inflammation and scarring is common, the immune cell infiltrates are predominantly B cells.
Thrombotic microangiopathy is believed to be caused by damage to the glomerular microvascular endothelium and typically occurs in the setting of acute renal failure associated with thrombocytopenia or hemolytic anemia. We will not discuss this rarely occurring disorder here but direct the reader to a recent review on the topic .
HIVAN and HIVICK accounted for 48%, and 31%, respectively of the renal pathology that was identified in 89 HIV infected patients from six US medical centers who underwent kidney biopsies for clinical indications , and for 83%, and 1%, respectively of 30 HIV-infected Black South Africans from a single center who underwent biopsy because of proteinuria or persistent microalbuminuria identified by screening .
The mechanism of HIVAN pathogenesis centers on the function of the viral accessory proteins Tat, Nef, Rev, Vif, Vpr, and Vpu, that are unique to HIV-1 and the other lentiviruses. Most studies investigating kidney pathogenesis have relied on transgenic mouse models expressing a combination of viral proteins. Because transgenic mice are not infected by HIV-1 nor do they develop AIDS, they demonstrate that HIVAN can occur independent of an active infection or immune suppression . What is required is the expression of the viral genes in kidney parenchymal cells. In fact, limiting the expression of HIV-1 genes exclusively to podocytes  recreates all the pathological findings of HIVAN, including the tubulo-interstitial abnormalities, suggesting that infection and injury of the podocyte is the initiating pathological event for the development of HIVAN. In humans, studies have confirmed that patients diagnosed with HIVAN have evidence of HIV-1 RNA and DNA in podocytes and tubular epithelial cells [28–30], and that these infected renal cells appear capable of supporting a productive infection . This corroborates the animal studies that direct infection of renal parenchymal cells occurs in HIVAN, although the in vivo mechanism by which human kidney cells are infected remains unknown.
The accessory protein Nef has consistently been implicated as a critical disease determinant, although the contribution of other accessory proteins to kidney pathogenesis is likely, most notably Vpr. Nef evolved for the virus to modulate the host’s immune system by controlling T cell receptor responses and downregulating the cell surface receptors CD4 and CD28. The mechanism by which Nef is pathogenic to the kidney is an area of active investigation. In cultured mouse podocytes, Nef expression underlies most of the previously described phenotypic changes including proliferation, changes in cell adhesion and loss of contact inhibition, cellular dedifferentiation, and cytoskeletal rearrangements [32–35]. Mechanistically, this appears to be related to Nef’s ability to intersect host signal transduction pathways that involve Src kinases and STAT3 . HIV-1/Nef also intersects the NF-κB signaling pathway via the classical activation pathway involving the phosphorylation of the key inhibitor, IκBα . This results in a persistent activation of NF-κB that has been associated with cell survival  and has the potential to exacerbate immune responses in the kidney microenvironment. Although the Src and NF-κB signaling pathways are both complex and share some signaling intermediaries, it has not been demonstrated that Src kinases have a direct role in activating NF-κB or vice versa . Overall, the disruption of several host signaling pathways in renal cells appears to be mediated by the viral protein Nef, and more comprehensive mechanistic studies will be required to determine their specific contributions to pathogenesis. Nevertheless, it seems logical that therapeutic strategies directed at neutralizing Nef function may be a more direct target to ameliorate many aspects of host pathogenesis.
Because of the extensive use of mouse models for host pathogenesis studies, it is easy to overlook the reality that humans do not live in pathogen, diet, and behavior controlled environments like laboratory mice. Consequently, the contribution of other environmental factors, especially infectious agents, may be underappreciated with regard to their role in influencing disease severity. Progressive HIV disease is associated with a heightened state of immune activation that is characterized by increased T cell turnover, and increased expression of T cell activation phenotypes and pro-inflammatory cytokines such as TNFα. As a result of the profound depletion of CCR5+ CD4+ cells in the gastrointestinal tract that occurs soon after an initial HIV-1 infection, it is possible that bacterial translocation may contribute to this generalized state of immune activation. This is supported by recent correlations between circulating levels of lipopolysaccaride (LPS) with progressive stages of HIV disease, and with heightened markers of innate and adaptive immunity . In addition to LPS, a ligand for Toll-like receptor (TLR)-4, new evidence indicates HIV-derived TLR-7 ligands (such as ssRNA) may also contribute to the immune activation in HIV infection resulting in the chronic production of pro-inflammatory cytokines regulated by both NF-κB (IL-6, TNFα) and IRF-7 (IFNα) [41, 42].
Immune responses are central to the development HIVICK, being associated with concurrent infections and the deposition of HIV antigen-antibody complexes within the glomerulus. In HIVAN, it is likely that immune responses are also contributors to disease severity or rate of progression, and it has been suggested that the pathogenic mechanism of the collapsing glomerulopathies may be directly link to maladaptive inflammatory cascades . In addition, it has recently been shown that human renal tubular epithelial cells infected in vitro respond with a battery of immune mediators, many regulated by NF-κB and IL-6 . Within the kidney microenvironment, renal cell produced TNFα and IL-6 were found to enhance viral replication of infected, infiltrating cells . Outside the kidney microenvironment, systemically high levels of circulating IL-6 and D-dimer have been strongly correlated with patient mortality in HIV infection  as well as the strong connection between circulating TNFα levels and AIDS progression . Taken together, this scenario of, 1) high levels of cytokines in both circulation and in the kidney microenvironment, 2) the HIV-induced persistent activation of NF-κB in renal cells, 3) the enhancement of HIV-1 and cytokine gene expression by a NF-κB, would create a feed forward loop that would establish a chronic immune activation cascade exacerbating both host pathogenesis and viral replication. Although direct infection of renal epithelial cells appears necessary for HIVAN, the involvement of both systemic and local immune responses of both the innate and adaptive immune systems are probably important accelerants to disease severity or progression.
Improvements in renal function with ART were suggested by case reports and case series in patients with HIVAN, but more direct evidence of such benefits are derived from observational cohort studies and clinical trials. In the Johns Hopkins HIV Cohort, ART use, compared to no ART, significantly reduced the risk of HIVAN (RR 0.4), and a significant decline in the incidence of HIVAN was observed in the early HAART-era compared to the pre HAART-era . In a case-controlled study of the multicenter, observational HIV Outpatient Study, ART use was associated with a lower risk of CRF, and this protective effect was significantly more pronounced among patients with nadir CD4 cell counts < 200/μL . Subjects who were randomized to interrupt ART in the multicenter Strategies for Management of Antiretroviral Therapy Study had a four-fold greater risk of chronic renal failure compared to those randomized to continuous ART . In a subset of these subjects, ART interruption was associated with sustained cystatin-C increases (by ≥ 0.15 mg/dL) that were evident within one month of randomization . Similarly, significant declines in renal function were observed over time in ART treated Ugandans who were randomized to weekly intermittent ART (7 days on/7 days off ART), compared with stable renal function in subjects randomized to a shorter treatment interruption (5 days on/2 days off ART), or continuous ART (average CLCr change/year: −12.4, +4.6, and +1.0 ml/min for each respective randomized arm) . These studies provide consistent evidence of renal function preservation with ART.
Evidence that ART improves kidney function in persons with renal impairment also is accumulating. The initiation of ART was associated with eGFR improvements in subjects with ≥ Stage 2 CKD in participants of the Longitudinal Linked Randomized Trials (ALLRT), a prospective cohort of HIV infected subjects randomized to ART or treatment strategies in the AIDS Clinical Trials Group (ACTG), where a greater magnitude of improvement was associated with greater baseline renal impairment . Renal function improvements were associated with suppression of plasma HIV-1 viremia, but this association was evident only in subjects with low baseline CD4 cell counts (< 200/μL); renal function improvements also were evident in subjects with higher CD4 cell counts but they were not associated with viral suppression. In the Ugandan Home-Based AIDS Care trial, the median CLCr rose from 63 to 76 ml/min after two years of ART and these improvements also were more pronounced in subjects with greater baseline renal impairment . In the DART study, eGFR improvements were similarly associated with greater baseline renal impairment and lower baseline CD4 cell counts . Although a small number of subjects progressed to ≥ Stage 3 CKD (1.9% and 1.6% of subjects in ALLRT and DART, respectively) these data constitute growing evidence of substantial kidney function improvement with ART, which may be mediated by viral suppression and other immune mechanisms. Because the renal benefits of ART in these studies were not limited to persons of African descent, a general effect of HIV on the kidney is implied that may not be limited to HIVAN.
In case reports and case series, tenofovir has been associated with renal failure, nephrogenic diabetes insipidus and Fanconi syndrome, a proximal renal tubular disorder that is characterized by increased excretion of protein, glucose, and electrolytes . In clinical trials and cohort studies, tenofovir was not associated with an increased risk of clinically significant kidney disease, however, but modest, non-progressive declines in renal function were consistently observed in tenofovir containing, compared to non-tenofovir containing ART regimens [56, 57]. Although tenofovir use was not associated with a greater risk of a clinically significant decline in eGFR in the DART study, renal function improvements were significantly attenuated in association with tenofovir, compared to other agents (mean eGFR changes from baseline to week 96: −3, +13, or +16 ml/min/1.73 m2 with initial randomized regimens that included tenofovir, abacavir, or nevaripine, respectively) .
ART is often under prescribed or incorrectly dosed in patients with ESRD undergoing renal replacement therapy, which in one study was directly associated with increased mortality . ART dose adjustments for renal insufficiency were recently summarized . With the exception of abacavir, dose adjustments are necessary for the nucleoside reverse transcriptase inhibitors. The safety of tenofovir has not been established in persons with CKD and some experts recommend prescribing alternative agents in patient with GFR < 60 mL/min , although the experience from the DART study suggests that tenofovir could be safely included as part of an initial ART regimen even with baseline renal function impairment.
Guidelines for the management of chronic kidney disease in HIV infected persons include recommendations to screen for kidney disease by urinalysis for proteinuria, and eGFR or CLCr, to be repeated annually for high risk patients (persons of African descent, CD4 cell counts < 200/mL, diabetes, hypertension, or hepatitis C); aggressive hypertension control ;the initiation of ART in children and adults with HIVAN, with the addition of ACE inhibitors, or angiotensin receptor blockers if renal function does not improve . Dialysis and the placement of arteriovenous fistulae should not be withheld solely on the basis of HIV infection, and kidney transplantation may be a viable option.
Transplantation of deceased donor and living donor kidneys is available in the US for HIV-infected patients through clinical trials or as part of routine care through several transplant centers.
Renal disease in HIV infected persons may arise from the combination of direct effects of HIV-1 replication, immune and genetic mechanisms, and comorbidities. Persons of African ancestry are at a significantly greater risk of ESRD, and kidney dysfunction is a marker for accelerated HIV disease progression. Improvements in survival of an increasingly aging HIV infected population necessitate vigilant screening for kidney disease and aggressive management of comorbidities. Appropriate, dose-adjusted ART use is essential in the optimal management of patients who exhibit this increasingly important complication.
L.A.B. is supported by NIH grant DK061395. L.A.B. and R.C.K. are investigators in the Case Center for AIDS Research supported by NIH grants AI36219 and AI25879