The nephrologist in the HAART era

Abstract

There is evidence that HIV-associated nephropathy (HIVAN) can be prevented and its progression slowed by HAART. More than 20 antiretroviral drugs and drug combinations are available, and life expectancy of HIV-infected individuals is now measured in decades. However, it is likely that more indolent forms of HIVAN remain common in the HAART era, predisposing patients to nephrotoxicity from HAART and related therapies. Indeed, the prevalence of acute renal failure and chronic kidney disease appears to be increasing among HIV-infected patients in the United States [1,2], and kidney disease has emerged as an important predictor of mortality [2,3]. Adverse effects of antiviral treatments should be considered, including their long-term renal toxicity and their role in renal scarring after acute adverse events. In addition, the burden of comorbid chronic kidney disease is likely to increase with ageing of the HIV-infected cohort, continued growth of the epidemic among susceptible minorities and increasing prevalence of HAART-related metabolic abnormalities. In this population of patients, evaluation of renal disorders and prevention of evolution toward chronic renal failure are a crucial challenge. Epidemiology of HIV-associated nephropathy and HIV-related diseases Most epidemiological data on HIVAN generated in the pre-HAART era are based on the US Renal Data System (USRDS). Initial clinical studies indicated that 10% of HIV-1-infected patients appeared to develop renal disease, of which 90% showed clinical and/or pathological features consistent with HIVAN [4,5]. Past postmortem studies also yielded a prevalence of HIVAN ranging between 1% and 15%, depending on the population [6,7]. The prevalence and distribution of HIVAN was associated strongly with African-American ancestry [8,9] as indicated by an USRDS-based report of 3653 patients with end-stage renal disease (ESRD) secondary to HIVAN during 1992 to 1997 [10]. Since the introduction of HAART, national epidemiological data show a reduction of incidence of ESRD associated with HIV-associated renal disease in the United States [11,12] contrasting with the steady increase in HIV/AIDS in the general population [13]. From a mathematical model using available epidemiological data on HIV-infected patients in the USRDS database and the Centers for Disease Control and Prevention data for HIV-seropositive patients, Blower et al.[14] suggested that HAART decreases the incidence of ESRD in patients with HIVAN and the mortality from HIV, with an overall efficacy of 23%. This trend has been attributed, in part, to beneficial effects of HAART, which was commenced in 1996. Preliminary retrospective series or case reports support the efficacy of HAART in improving outcome in HIVAN [15–19]. In a retrospective cohort study, Szczech et al.[20] reported that treatment with protease inhibitors (and prednisone) was associated with a slower decline in renal function in patients with HIVAN or other HIV-1-related renal diseases. Cosgrove et al.[18] reported another retrospective series of 23 patients with HIV-1-related nephropathies, including patients with HIVAN. Patients with HIVAN were treated with HAART and none doubled their serum creatinine. In the non-HAART group, all patients showed a doubling of serum creatinine, two patients died and eight required dialysis. One study [21], retrospectively comparing two cohorts of 102 and 33 patients with biopsy-proven-HIVAN in the pre-HAART and in the HAART eras, respectively, also argues for improvement of renal survival by HAART. However, a recent postmortem-based survey reported that 12% of African-American patients dying of HIV-1 infection have histologically confirmed HIVAN [22]. Even if HAART decreases the incidence of HIVAN in African-Americans, the prevalence of HIVAN may not change because of the improvement in the survival of these patients. The onset of HIVAN could, therefore, just be delayed. Indeed, we have recently reported a patient with biopsy-proven HIVAN despite the lack of any past or present AIDS-defining condition and HAART-controlled HIV-1 infection for at least 2 years [23] Schwartz et al.[1] developed a mathematical model of the dynamics of HIV infection in the ESRD population in order to assess the impact of HAART on the progression of patients with AIDS to the development of ESRD and to predict the prevalence of HIV-related ESRD through to 2020. The authors concluded that, despite the potential benefit of HAART, the prevalence of HIV-related ESRD in the United States would be expected to rise in the future as a result of the expansion of the number with AIDS among black individuals. Nonetheless, while prospective controlled trials evaluating HAART on HIVAN or other HIV-1-related nephropathies are not ethically acceptable, consensus guidelines recommend consideration of HAART in HIV-infected patients with chronic renal insufficiency [24]. Changes in clinicopathological presentation HIV-associated nephropathy HIVAN is an unusual form of poorly responsive glomerular disease characterized by nephrotic syndrome, focal segmental glomerulosclerosis and a rapid fulminant progression to ESRD. Proteinuria occurs in up to 30% of HIV-infected patients, but not all of these patients have HIVAN [25–27]. The true prevalence of HIVAN is not known. The geographic distribution of HIVAN is not uniform, and it depends on specific risk factors for HIV disease, including race, gender and drug use. There is a striking predilection for HIVAN among African-Americans, as is also true for focal segmental glomerulosclerosis associated with intravenous drug use [4]. HIVAN is 7–10 times more common in men than in women, and 30–60% of people with HIVAN have a history of intravenous drug use [7]. In the pre-HAART era, patients with HIVAN typically presented with a nephrotic syndrome consisting of nephrotic-range proteinuria classically without oedema despite severe hypoalbuminaemia. Urinalysis reveals microhaematuria. Patients with HIVAN are typically not hypertensive, even in the face of renal insufficiency, and their kidneys are usually normal to large in size and highly echogenic by ultrasonography. The most common histological light microscopy finding is a collapsing form of focal segmental glomerulosclerosis. Tubulo-interstitial scarring, atrophy and marked dilatation of the tubules (microcystic dilatations) are usually present. Immunofluorescent microscopy is usually negative. Electron microscopy reveals wrinkling of the basement membranes, epithelial cell proliferation and focal foot process effacement. Tubulo-reticular structures in the glomerular endothelial cells consisting of ribonucleoprotein and membrane, the synthesis of which is stimulated by alpha-interferon, is highly predictive of HIVAN. Risk factors for progressive renal disease include CD4 cell count < 200 cells/μl, detectable plasma HIV RNA, hypertension, low plasma albumin and elevated serum creatinine [28]. In our experience during the HAART era, where most patients have well-controlled HIV, a nephropathy typically presents as stable or slowly progressive renal failure, hypertension, glomerular proteinuria, not necessarily of nephrotic range, and a preserved rather than increased kidney size. Pathologically, simple tuft ischaemia tends to replace florid (?) glomerular collapse while mild interstitial infiltration and cystic tubular dilatation are seen. Paradoxically, we noted more frequent atherosclerotic vascular changes even in younger affected patients. Until recently, the clinical course of HIVAN was one of inexorable progression to ESRD in 6–12 months, with limited treatment options. More options are now available to patients; these include antiretroviral therapy, steroid treatment and angiotensin-converting enzyme inhibitors. Evidence exists that antiretroviral therapy can reverse or improve the progression of HIVAN [15,16,19]; however despite the widespread use of HAART, no prospective studies have demonstrated benefit in slowing the progression of HIVAN. The survival benefit from antiretroviral therapy is indisputable. HAART may also prevent the development of HIVAN in at-risk groups. Lucas et al.[29] evaluated a cohort of 3976 at-risk patients in the Johns Hopkins HIV clinic database from 1989 to 2001. They identified 135 cases of HIVAN based on either clinical or pathological criteria. There was a 50% decline in HIVAN incidence in 1998–2001 compared with 1995–1997, and HAART was associated with a 60% reduction in risk for developing HIVAN. Patients with AIDS developed HIVAN at a rate of 12.5/1000 person-years, compared with 3.1/1000 person-years in patients without AIDS (relative risk, 4.1). However, it is notable that in over 1071 person-years of follow-up, no patient developed HIVAN when HAART was initiated prior to the onset of AIDS. The rationale for treating HIVAN with corticosteroids is that steroids are the mainstay of treatment for idiopathic focal segmental glomerulosclerosis [30]. The first report of steroid treatment in four patients with HIVAN found a significant reduction in serum creatinine after a course of corticosteroids lasting 2–4 weeks [31]. The initial case report study has now been extended to include 20 patients, and these results confirm that prednisone at a dose of 60 mg daily for 2–11 weeks leads to a significant reduction in serum creatinine and in 24-hour urinary protein excretion [30]. The encouraging short-term results must be balanced by the findings during the follow-up period. During a median follow-up of 44 weeks, 8 of 20 patients required maintenance dialysis, 11 of 20 died of HIV disease after completing prednisone treatment, and 6 of 20 developed serious infections while receiving prednisone. Only 7 of 20 patients were alive and free from ESRD after a median of 25 weeks following initiation of prednisone. A more recent retrospective study reported the course of HIVAN in 13 patients treated with prednisone and a further eight patients not treated with prednisone. Even after controlling for baseline creatinine, proteinuria, and CD4 cell count, among other variables, the prednisone group had an 80% reduction in risk of progressive azotaemia after 3 months [32]. Prednisone may work by reducing the amount of interstitial inflammation [33]. In 1995, The AIDS Clinical Trials Group (ACTG) designed a phase II randomized, double-blind, placebo-controlled multicentre trial to determine the efficacy of prednisone therapy in HIVAN, but this trial was cancelled because of poor patient recruitment. The angiotensin-converting enzyme inhibitors captopril and fosinopril have also been studied as possible therapy for HIVAN [34]. In one study in which 18 patients were enrolled, nine were treated with captopril, 6.5–25 mg three times daily, and nine controls did not receive captopril. All patients had biopsy-proven HIVAN, and renal survival was defined as the time from initiation of captopril treatment to initiation of dialysis (ESRD). The initial mean serum creatinine concentration was 34 mg/l (±7.0) in the captopril group and 37 mg/l (±0.5) in the controls. A small renal survival advantage of approximately 8 weeks (median 83 versus 30 days), was seen in the captopril group [35]. Two non-randomized studies have investigated the effect of fosinopril on the progression of HIVAN. Both studies showed a significantly lower risk of reaching ESRD in the fosinopril group compared with non-treated controls [35,36]. Despite the limitations of these studies, they suggest that therapy with an angiotensin-converting enzyme inhibitor initiated early may offer renal survival benefits in HIVAN. The ACTG is currently developing a clinical trial (protocol A5179) to compare treatment with an angiotensin receptor blocker (valsartan) plus HAART with HAART alone in patients with HIVAN. Indications for kidney biopsy The decision to obtain a renal biopsy sample is somewhat controversial in the general medical community. Even if a patient presents with the classic clinical features of HIVAN, clinical presentation is predictive of the biopsy diagnosis in only 50–60% of patients. Furthermore, non-invasive tests or clinical markers to identify the precise renal lesion do not exist. Renal biopsy should be offered to patients because a variety of renal lesions occur in HIV-infected patients, and the treatment implications and prognosis vary according to the biopsy results. We, therefore, believe that to distinguish HIVAN from other forms of renal disease, and to redefine HIVAN pathological findings in the HAART era, HIV-positive patients who have unexplained renal abnormalities (i.e., kidney failure and/or daily protein excretion greater than 1 g and/or microscopic haematuria) should have a renal biopsy. HAART-related kidney disorders Electrolyte and acid-base profiles It is noteworthy that the biological profile in HIV-infected patients has dramatically changed in the HAART era. Previous studies stated that hyponatraemia, hyperkalaemia or hypokalaemia and acidosis were the main biological abnormalities in HIV-positive patients [37,38]. A prospective cross-sectional descriptive study (1219 HIV-infected patients over 3 months) undertaken to assess the prevalence of fluid electrolyte and acid–base disturbances showed hyperuricaemia and hypophosphataemia to be the most prevalent abnormalities [39]. Hyperuricaemia was detected in 140 (41.3%) out of the 339 patients tested. Among hyperuricaemic patients, only 47% were treated with didanosine. Multivariate analysis showed that patients not taking non-nucleoside transcriptase inhibitors (NNRTI) had a 1.8-fold risk [95%confidenceinterval(CI),1.1–2.9] of hyperuricaemia. With protease inhibitor treatment and male gender, the risk of hyperuricaemia rose to 4.4 (95% CI, 2.1–9.6). A plasma phosphate level below the normal range was observed in 63 (17.2%) of the patients tested for plasma phosphate. Multivariate analysis showed that patients taking an NNRTI regimen had a 1.9-fold increase in risk of hypophosphataemia (95% CI, 1.1–3.3), male patients had a 2.6-fold increase in risk (95% CI, 1.1–6.3). Bicarbonate plasma level below the normal range was observed in 112 patients (13.6%) out of the 824 patients tested. Multivariate analysis showed that patients taking HAART had a 4.4-fold increase in risk having a low plasma bicarbonate level (95% CI, 2.2–8.9), women had a 2.4-fold increased risk (95% CI, 1.5–3.8) and patients with CD4 cell count < 200 cells/μl had a 1.8-fold increased risk (95% CI, 1.2–2.9). Only 13 patients (3.1%) out of the 419 patients tested exhibited low calcium levels (n = 13). Factors significantly associated with low plasma calcium concentrations were NNRTI regimen and CD4 cell count. An absolute CD4 count < 200 cells/μl was associated with an increased risk of hypocalcaemia when compared with a cont of > 200 cells/μl. Lactic acidosis Approximately 20–30% of patients who are treated with nucleoside reverse transcriptase inhibitors (NRTI) can be found to have asymptomatic hyperlactataemia; this typically develops after several months of therapy and may be transient [40–42]. Severe lactic acidosis is much rarer, occurring in 1.5–2.5% of patients, is usually preceded by fatigue, nausea, vomiting, anorexia, abdominal pain and other systemic symptoms, and is associated with a mortality rate of approximately 80%. Lipoatrophy, myopathy, peripheral neuropathy and pancreatitis are more often observed in patients with symptomatic hyperlactataemia rather than in patients who have frank lactic acidosis. Risk factors include NRTI use, longer duration of treatment, older age, female gender, pregnancy, hypertriglyceridaemia, obesity, concomitant hepatitis C infection, use of ribavirin, impaired kidney function and alcohol ingestion [41–44]. Acute renal failure Before the HAART era, mild acute renal failure, defined as a peak serum creatinine ≥ 20 mg/l, was been reported to occur in up to 20% of hospitalized HIV-infected patients [45]. This is in comparison with an incidence rate of 1% in hospitalized non-HIV-infected patients [46]. The two major acute renal complications in HIV disease that resulted in potentially reversible failure were acute tubular necrosis and HIVAN. Sepsis contributed to the development of severe acute tubular necrosis, defined as a peak creatinine ≥ 60 mg/l, in up to 75% of cases [47]. A study of kidney biopsy specimens in HIV-infected patients with severe acute renal failure not thought to be from prerenal causes or acute tubular necrosis reported the following distribution of renal lesions: 53% haemolytic uraemic syndrome; 40% acute tubular necrosis, either of ischaemic–toxic origin or rhabdomyolysis; 26% obstructive renal failure, extrinsic, drug induced or secondary to paraprotein precipitation; 23% HIVAN; 3% acute interstitial nephritis; and 6% various glomerulonephritides [48]. In a recent cohort study of ambulatory HIV-infected patients, acute renal failure occurred in nearly 10% of patients, with an incidence rate of 5.9 episodes/100 person-years [49] Antiretroviral agents have been shown to have a range of nephrotoxic effects, including crystal-induced obstruction, tubular toxicity, interstitial nephritis and electrolyte abnormalities. Drugs are responsible for one-third of all acute renal failure events. Although only responsible for a few events, antiretroviral drugs cause two-thirds of all obstructive acute renal failure. Indinavir, tenofovir, and nevirapine were the antiretroviral drugs associated with acute renal failure in this cohort [49–52]. Iatrogenic acute renal failure can be dose dependent, manifesting as acute tubular necrosis, acute glomerulopathy, vascular disease or interstitial immunoallergic reactions. Table 1 summarizes the nephrotoxicities induced by antiretroviral drugs [53–77]. According to the renal syndrome, the clinician will attempt to distinguish between possible causes from each of the four aetiopathogenic groups of nephropathies defined above. Depending on the situation, renal biopsy may be indicated. Table 2 proposes a diagnostic flowchart applicable for a HAART- treated patient with renal abnormalities.Table 1: Antiretroviral drug-induced renal abnormalities.Table 2: HIV-infected patients with renal abnormalities: a diagnosis proposition.Chronic renal failure Chronic renal abnormalities are frequently observed in HIV-infected individuals. Despite drug adjustment, some HIV-infected patients experience progressive renal disease. Usually, the diagnosis of renal toxicity of antiretroviral treatment is considered when patients experience acute renal abnormalities. However, the potential insidious long-term renal toxicity of antiviral treatment may be underappreciated in the pathogenesis of nephropathies of chronic progression in HIV infected patients. In prospective studies evaluating the safety and efficacy of the protease inhibitor indinavir, or some reverse transcriptase inhibitors, a proportion of patients with treatment-related acute renal failure did not recover their baseline renal function and others remained hypertensive. These data underline the possibility of permanent renal damage after acute renal injury related to antiretroviral treatment. Given the large range of kidney diseases that may occur in HIV-positive patients taking HAART, each case should be analysed and discussed independently regarding the benefits and risks of drug withdrawal. The main challenge for the clinician is to determine the aetiology of the evolving nephropathy in order to initiate specific therapeutic intervention in addition to symptomatic measures [11]. Metabolic complications With the advent of potent HIV therapies, the complications associated with HIV infection have expanded to include changes in metabolism. HIV protease inhibitors acutely and reversibly inhibit the insulin-responsive glucose transporter GLUT4, leading to peripheral insulin resistance and impaired glucose tolerance [78]. Altered body fat distribution (lipodystrophy), elevated plasma triglycerides and cholesterol, bone demineralization and elevated plasma lactate have been reported in clinical trials and in observational databases. These changes in metabolism may be related to underlying HIV disease or may be a consequence of treatment with antiretroviral agents through mitochondrial dysfunction [79,80]. Likewise, in addition to direct nephrotoxicity, long-term antiretroviral therapy could itself promote the development of chronic kidney disease, through metabolic complications [81,82]. Indeed, black race [81] and indinavir [72,73] have been associated with the development of hypertension. Both protease inhibitors and NNRTI have been associated with insulin resistance, hyperlipidaemia and diabetes mellitus [83–86]. Several studies have shown plasma uric acid concentrations to be independently associated with increased mortality in multivariate analyses in hypertensive and/or older patients [87]. Evidence of hyperuricaemia in HIV-positive patients should lead to change in the management of such patients, such as avoidance of diuretics. The impact of HIV- and antiretroviral drug-associated metabolic disorders on the epidemiology of kidney or cardiovascular disease in patients with HIV may, therefore, become an important issue. Risk factors for nephrotoxicity While renal adverse events are not always predictable, it is possible to identify patients who may be at increased risk. Traditional risk factors for kidney disease, including older age, African-American origin, comorbid diabetes and preexisting kidney disease, appear to be associated with renal failure in HAART-treated patients [2,37]. In addition, liver disease and hepatitis virus coinfection have been consistently associated with increased risk for acute or chronic kidney disease in HIV-infected patients [2,20,37]. Baseline kidney function is one of the most powerful predictors of risk for acute or chronic kidney disease in HAART-treated patients [2,20,37]. The recently published consensus guidelines recommended screening for chronic kidney disease in all HIV-infected patients at the time of diagnosis, with frequent monitoring in the setting of reduced kidney function or other evidence of kidney disease [24]. Initial screening should include urinalysis and serum creatinine testing, with estimation of the glomerular filtration rate by either the Cockcroft–Gault [88] or Modification of Diet in Renal Disease equation [89]. While neither estimation equation has been well validated in patients with HIV, both methods are more sensitive than serum creatinine alone. Identification of patients with a glomerular filtration rate < 60 ml/min per m2 will allow for appropriate dose modification of medications cleared by the kidney, and should prompt more frequent monitoring of kidney function, toxicity and therapeutic efficacy. The pharmacokinetic modifications according to the glomerular filtration rate and the necessary adaptation of dose for each anti-HIV medication are detailed in Table 3[90–110]. The detection of proteinuria by dipstick or laboratory testing also indicates underlying chronic kidney disease, and may identify additional patients at increased risk for nephrotoxicity [24]. In addition to chronic kidney disease, other factors such as volume depletion, critical illness and concomitant administration of other nephrotoxic agents may also predispose patients to the nephrotoxic effects of HAART or related medications. HAART can also lead to hypertension. Indeed, studies conducted over the past 20 years, in numerous patient populations, have demonstrated a continuous relationship between elevations in blood pressure, cardiovascular disease and death [111]. Crane et al.[112] found a twofold increase in the risk of developing elevated blood pressure among patients receiving lopinavir/ritonavir compared with those receiving efavirenz-based regimens. In addition, the results suggested that the increased risk associated with lopinavir/ritonavir was mediated, at least in part, through an increase in body mass index.Table 3: Antiretroviral dose adjustment based on creatinine clearance.aTable 3: (continued)Special populations coinfection with HIV and hepatitis B or C virus The effect of coinfection with hepatitis viruses, in particular B and C (HBV, HCV), on disease progression and survival among patients infected with HIV has become an increasing concern. Ironically, this concern has arisen from the enormously positive effects of HAART; while this improves immune function and decreases the incidence of opportunistic infections in HIV-infected patients, it also increases the contribution of chronic liver disease to the overall morbidity and mortality of these individuals. Excessive use of alcohol is an additional factor influencing HIV/HCV coinfection in this population. Approximately one-quarter of HIV-infected patients are HCV seropositive [113,114]. Among intravenous drug users with HIV, this rate is at least 50% and can be up to 90% in many populations [114,115]. Coinfection with HBV and HIV is common, with 70–90% of HIV-infected individuals having evidence of past or active HBV infection [116–118]. The prevalence of chronic carriage of HBV surface antigen among HIV-infected individuals is 1.9–9% [119,120]. Among intravenous drug users, 90% of HIV-infected individuals have evidence of exposure to HBV (hepatitis B core antibody positivity) and 60% also have evidence of past infection with the presence of HBV surface antibody [118]. Prevalence of renal disorders has not been reported in HIV/hepatitis coinfected populations. However, proteinuria and haematuria have been reported frequently in HCV- or HBV-infected patients without HIV coinfection: proteinuria 12.4 [121] to 27.3% [122] and haematuria 9% [122,123] in HCV- infected patients; proteinuria 17% [124] to 30% [125] and haematuria 30% [126] in HBV-infected patients. HCV-associated cryoglobulinaemic glomerulonephritis has been well described in both HIV-coinfected and uninfected patients and is characterized by membranoproliferative glomerulonephritis, purpura, arthralgias and peripheral neuropathy [126,127]. The renal presentation of HCV in HIV/HCV coinfected patients may include renal insufficiency, proteinuria, haematuria, depressed complement levels, and circulating cryoglobulins. The course of HCV-associated renal disease is much more aggressive in the HIV-coinfected patient, with some progressing to ESRD in a matter of months [127,128]. However, while previous reports had shown associations between HIV and coinfection with HCV, Dezzutti et al.[129] did not find an association between HIV infection and cryoglobulinaemia. In patients not coinfected with HIV, the association of membranous nephropathy with HBV, malignancy and syphilis has also been well described, and the reports of membranous nephropathy in HIV-infected patients may be explained by the high incidence of HBV infection, malignancies and syphilis in this population [129–135]. The life expectancy of HIV-positive patients treated with HAART now approximates that of the general population. The challenge has become to avoid the cardiovascular and renal risks that can increase with longevity, more so in predisposed patients (e.g., hypertensive, type 2 diabetics, smokers and dyslipidaemic patients). These risk factors contribute to renal disorders independently from HIV infection and HAART. Consequently, it becomes more important to develop sound recommendations for treating these patients from the renal stand point. Recommendations Renal functions need to be monitored in patients with HIV/AIDS and potentially reversible factors need to be identified and managed. Adapting the Expert Guidelines Recommendations [24], we suggest 10 points concerning the antiretroviral drug use and kidney disease management in HIV-infected patients. All patients at the time of HIV diagnosis should be assessed for existing kidney disease with a screening urinalysis for proteinuria and a calculated estimate of renal function. If there is no evidence of proteinuria at initial evaluation, patients at high risk for the development of proteinuric renal disease (i.e., African-Americans, those with CD4 cell counts < 200 cells/μl or plasma HIV RNA > 4000 copies/ml, and those with diabetes mellitus, hypertension, or HCV coinfection) should undergo annual screening. Renal function should be estimated on a yearly basis to assess for changes over time. Additional evaluations included proteinuria, renal ultrasound and, potentially, renal biopsy. Patients with decreased glomerular filtration rate (< 60 ml/min per 1.73 m2) or with persistent proteinuria (grade > 1+ by dipstick analysis), or haematuria on urinalysis should be referred to a nephrologist for further evaluation and treatment. In HIV-infected patients with evidence of nephropathy, blood pressure should be controlled to a level no higher than 125/75 mmHg. Angiotensin-converting enzyme inhibitors or angiotensin receptor blockers are the drugs of first choice for those patients with proteinuria, and these drugs should also be considered in normotensive HIV-infected patients with renal disease. Calcium channel blockers should be avoided in patients receiving protease inhibitors. Patients with HIVAN should be treated with HAART at diagnosis. HAART should not be withheld from patients simply because of the severity of their renal dysfunction. The optimal antiretroviral regimen is a delicate balance between the benefits of virological suppression and the risk of adverse events, including nephrotoxicity. The degree of renal insufficiency should influence the choice and dose of the individual antiretroviral agents. Addition of prednisone should be considered in patients with HIVAN if HAART alone does not result in improvement of renal function or in patients with HIVAN whose renal failure is progressing rapidly. A discussion that includes the risks of short-term immunosuppression, the benefit of slowing the progression to ESRD and the patient's wishes should precede the final decision regarding corticosteroid therapy. Appropriate reduction of dosing for antiretroviral drugs that are primarily eliminated via the kidneys is warranted. With careful dose adjustment and close monitoring, renal insufficiency or ESRD is not an absolute contraindication to the use of any particular antiretroviral agent. For example, nucleoside analogues should not be withheld in patients with reduced renal function for fear of the development of lactic acidosis. Drugs that are readily removed by dialysis should be administered after haemodialysis session to avoid loss of efficacy. In HIV-positive renal transplant recipients, there are clinically important interactions between the calcineurin inhibitors and protease inhibitors, and the NNRTI drug efavirenz mycophenolate may increase intracellular levels of abacavir, didanosine and tenofovir, thus enhancing their toxicity. In the presence of protease inhibitors, a major reduction in the dosage of calcineurin inhibitors may be required. Although most antiretroviral agents are relatively free of renal toxicity, drug-related renal injury can occur and may need to be distinguished from progression of HIVAN or other HIV-related kidney diseases, kidney diseases caused by other infections (e.g., HCV) or lymphoproliferative diseases, or kidney diseases unrelated to HIV infection and its treatment. Metabolic disorders, kidney, and cardiovascular diseases have emerged as important complications in HAART-treated patients, overtaking opportunistic infections as leading causes of death among HIV-infected patients. For patients with preexisting cardiovascular risk factors, hyperlipidaemia or a family history of a lipid disorder, consideration should be given to initiating or switching to protease inhibitor-sparing antiretroviral regimens, with the use of triple NRTI regimens or those containing NRTI and NNRTI, particularly nevirapine. Early referral and close collaboration between infectious disease specialists, nephrologists and cardiologists will facilitate diagnosis and treatment of chronic kidney and vascular diseases with the goal of delaying progression to ESRD. Acknowledgements The authors thank Sandra Campin and Dr Colin Campin for their technical assistance. Note: The authors participated equally in the text elaboration. H. Izzedine had full access to all the data in the study and had final responsibility for the decision to submit to publication.