Overview of kidney disease in patients with HIV

INTRODUCTION — With dramatic improvements in survival and disease progression in the era of combination antiretroviral therapy (ART), complications such as kidney, liver, and cardiac disease have largely replaced opportunistic infections as the leading causes of mortality in the setting of HIV [1]. Patients with HIV are at risk for both acute kidney injury (AKI) and chronic kidney disease (CKD) secondary to medication nephrotoxicity, HIV-associated nephropathy (HIVAN) [2-5], and immune complex kidney diseases [5-9]. In addition, the aging cohort of patients with HIV is at increased risk for comorbid kidney disease related to traditional CKD risk factors including diabetes, hypertension, and obesity, as well as kidney disease due to hepatitis B virus (HBV) or hepatitis C virus (HCV) coinfection [5,10,11].

ACUTE KIDNEY INJURY

Epidemiology of AKI in patients with HIV

Incidence of AKI — The incidence of acute kidney injury (AKI) is increased in patients with HIV compared with patients without HIV. Although the overall incidence of AKI in patients with HIV appears to be decreasing in the antiretroviral therapy (ART) era, the incidence of severe, dialysis-requiring AKI continues to rise.

In a study of hospitalized adults in New York state that compared administrative data from 1995 (before the introduction of ART) with data from 2003 (after the introduction of ART) [12], AKI was documented in a significantly greater proportion of patients with HIV compared with patients without HIV, both in 1995 (2.9 versus 1 percent) and 2003 (6 versus 2.7 percent). Because this study relied on administrative data to identify AKI cases, it is likely that only more severe cases were included.

Other studies have also demonstrated an increasing incidence of severe, dialysis-requiring AKI among hospitalized patients with HIV. In a study including more than 56,000 United States military veterans, the incidence of dialysis-requiring AKI declined early in the ART era but then doubled between 2000 and 2006 [13]. A similar increase in the incidence of dialysis-requiring AKI among adults with HIV was also demonstrated in a nationally representative sample of United States hospital admissions [14].

Two single-center cohort studies have evaluated the incidence and etiology of AKI among ambulatory patients engaged in HIV care in the ART era, using clinical and laboratory data to identify AKI cases [15,16]. In a prospective study of 754 ambulatory patients with HIV followed at a single center in the United States, at least one episode of AKI occurred in 71 patients (9.4 percent) during a two-year period [15]. In a retrospective study of more than 2200 patients with HIV engaged in care at a single center in London, AKI occurred in 5.7 percent of patients and was most common within the first three months of initiating HIV care [16].

Risk factors for AKI — Some risk factors for acute kidney injury (AKI) among patients with HIV are similar to risk factors for AKI in the general population, such as older age, diabetes mellitus, preexisting chronic kidney disease (CKD), and acute or chronic liver disease [12,14]. However, some risk factors are specific to HIV. In studies that have included detailed data on HIV disease severity, other predictors of AKI included a diagnosis of AIDS, low CD4 count, high viral load, and coinfection with hepatitis C virus (HCV) [13,15,16].

Outcomes after AKI — As in the general population, the development of acute kidney injury (AKI) increases the risk of morbidity and mortality in patients with HIV. The following examples illustrate an association with both short-term and long-term outcomes:

●Inpatient mortality was significantly more frequent among patients with HIV who had AKI than among patients with HIV without AKI in a 2003 sample of hospitalized patients in New York state (27 versus 4 percent) [12]. Dialysis-requiring AKI was also associated with a significantly increased risk of in-hospital mortality among adults with HIV in a nationally representative sample of United States hospital admissions from 2010 (adjusted odds ratio [OR] 2.64, 95% CI 2.04-3.42) [14].

●Long-term mortality is also higher among individuals with HIV who develop AKI. In a national sample of United States military veterans from 1986 through 2006, 17,325 veterans with HIV were hospitalized and survived for at least 90 days after discharge [17]. AKI occurred in 3060 (18 percent); 334 of these patients with AKI also required dialysis. During an average follow-up of 5.7 years after hospital discharge, the patients with HIV who experienced AKI were significantly more likely to die (56 versus 47 percent). End-stage kidney disease (ESKD) and cardiovascular events were also significantly more common among patients with HIV with AKI.

Causes of AKI in patients with HIV — Prior to the introduction of combination ART, acute kidney injury (AKI) was commonly attributed to septicemia, volume depletion, or medication toxicity [18,19]. Although data on the etiology of AKI in the ART era are limited, available data suggest that sepsis remains a common contributor and that AKI is often multifactorial [14-16]. The most common types of AKI in patients with HIV infection, similar to patients without HIV, are prerenal states and acute tubular necrosis. Causes of AKI that are unique to patients with HIV are discussed in this section. (See "Etiology and diagnosis of prerenal disease and acute tubular necrosis in acute kidney injury in adults".)

Medication nephrotoxicity — Patients with HIV infection are at risk for nephrotoxicity from ART, as well as from medications used to treat opportunistic infections or hepatitis virus coinfection. In addition, ART-treated patients are often taking additional medications for the treatment of comorbid conditions. Medication nephrotoxicity may present with acute or chronic kidney injury or with acid-base and electrolyte disturbances. (See 'Chronic kidney disease' below and 'Electrolyte disorders' below.)

Some of the more commonly implicated agents include the following:

●Tenofovir disoproxil fumarate (TDF) – Tenofovir is a nucleotide reverse transcriptase inhibitor. TDF is a prodrug of tenofovir that can cause AKI, proximal tubular dysfunction, or both in combination. The use of TDF has also been associated with increased risk of CKD in large observational studies [20-22] (see 'Epidemiology of CKD in patients with HIV' below). In addition, the benefits of ART on proteinuria and albuminuria may be less substantial with regimens containing TDF as compared with the alternative nucleoside reverse transcriptase inhibitor, abacavir [23]. According to expert guidelines, TDF should be avoided, if feasible, in patients with an estimated glomerular filtration rate (eGFR) less than 60 mL/min per 1.73 m² and should be discontinued in patients who experience a 25 percent or greater decline to an eGFR less than 60 mL/min per 1.73 m² [24,25].

The risk of kidney toxicity with TDF has varied across different studies, with estimates ranging from less than 2 percent to as high as 10 percent in one study [26,27]. Although kidney biopsy may not be necessary to diagnose TDF toxicity in the setting of proximal tubular dysfunction, it can be useful in cases where the diagnosis is less clear or where there are compelling reasons not to discontinue TDF. As an example, in a French study of 222 patients with HIV infection who underwent kidney biopsy, 29 had tubulopathy (13 percent); TDF was considered a primary or secondary contributing factor in nearly two-thirds of these cases [28].

A newer prodrug, tenofovir alafenamide (TAF), was approved in 2015 based upon studies demonstrating the potential for an improved safety profile. TAF achieves similar antiviral efficacy at much lower plasma concentrations of tenofovir, and randomized trials have demonstrated a significantly smaller decline in eGFR over 96 weeks in patients treated with a TAF-containing regimen compared with those treated with a TDF-containing regimen [29]. Although TAF is expected to result in a lower risk of clinically relevant kidney toxicity compared with TDF-containing regimens, AKI has been reported with TAF use [30,31].

●Protease inhibitors – Indinavir and atazanavir are protease inhibitors that can cause crystalluria, nephrolithiasis, and AKI. The use of some ritonavir-boosted protease inhibitors has also been associated with increased risk of CKD in large observational studies [20-22]; of note, the newer protease inhibitor darunavir does not appear to be associated with an increased risk of CKD [32]. (See "Crystal-induced acute kidney injury", section on 'Protease inhibitors' and 'Epidemiology of CKD in patients with HIV' below.)

●Other antiviral agents – Acyclovir, foscarnet, and cidofovir are drugs used to treat herpes simplex virus or cytomegalovirus infection. Each of these agents can be associated with the development of AKI. (See "Crystal-induced acute kidney injury", section on 'Acyclovir' and "Foscarnet: An overview", section on 'Renal insufficiency' and "Cidofovir: An overview", section on 'Toxicity'.)

●Anti-Pneumocystis drugs – Trimethoprim-sulfamethoxazole and, less commonly, pentamidine are agents used to treat Pneumocystis infection. Trimethoprim-sulfamethoxazole can produce interstitial nephritis, while approximately 25 percent of patients treated with pentamidine develop reversible AKI that is likely due to nephrotoxic acute tubular necrosis [17,33-35]. (See "Clinical manifestations and diagnosis of acute interstitial nephritis", section on 'Drugs'.)

In addition, several antiretroviral agents can interfere with the tubular secretion of creatinine, producing an increase in serum creatinine (and decline in eGFR) without a true decline in kidney function. The most prominent effect has been observed with the pharmacoenhancer cobicistat and the integrase inhibitor dolutegravir; the increase in serum creatinine with these drugs occurs early and is typically in the range of 0.05 to 0.2 mg/dL [36,37].

HIV-associated kidney disease — HIV-associated nephropathy (HIVAN), HIV immune complex kidney disease (HIVICK), and HIV-associated thrombotic microangiopathy can all present with AKI (see "HIV-associated nephropathy (HIVAN)" and 'Causes of CKD in patients with HIV' below). While HIV-associated thrombotic microangiopathy can present with significant AKI [38,39], results of a large observational cohort study suggest that this is a rare complication of HIV infection in the ART era [40].

CHRONIC KIDNEY DISEASE

Epidemiology of CKD in patients with HIV — The prevalence and incidence of chronic kidney disease (CKD) are higher in individuals with HIV than in the general population. With earlier introduction of effective antiretroviral therapy (ART), the incidence of HIV-related end-stage kidney disease (ESKD) has decreased [41].

The prevalence of CKD associated with HIV infection varies according to the population studied; the following examples illustrate the range of findings:

●In a prospective cohort of 2059 women with HIV followed in the Women's Interagency HIV Study (WIHS) both before and after the introduction of combination ART, nearly one-third of 2059 subjects had persistent proteinuria at enrollment [34]. Advanced HIV infection was associated with both baseline proteinuria and the development of acute or chronic kidney disease in this cohort.

●In a prospective cohort of 22,156 military veterans with HIV followed for eight years, 27 patients (0.1 percent) had proteinuria greater than 300 mg/day at baseline, and 51 patients (0.2 percent) had an estimated glomerular filtration rate (eGFR) less than 60 mL/min per 1.73 m² at baseline [42].

●In a cross-sectional analysis of 783 men with HIV in the prospective Multicenter AIDS Cohort Study (MACS), 5 to 7 percent had an eGFR less than 60 mL/min per 1.73 m², depending upon the method used to estimate GFR [43].

●In a cross-sectional study of 300 patients with HIV from the East African Republic of Burundi, only 2 percent had an eGFR less than 60 mL/min per 1.73 m² [44], although persistent urinary abnormalities (principally, persistent pyuria) were much more common. By contrast, in a cross-sectional study of 400 adults with HIV from the West African nation of Nigeria, 38 percent had a persistently abnormal serum creatinine (>1.5 mg/dL) and/or proteinuria [45]. The prevalence of CKD among individuals with HIV varies significantly across different studies and different populations in sub-Saharan Africa, reflecting significant variability in genetic susceptibility to HIV-associated nephropathy (HIVAN) and other kidney diseases conferred by polymorphisms in the APOL1 gene [46]. (See "HIV-associated nephropathy (HIVAN)".)

●In a large, multinational, randomized trial enrolling ART-naïve participants with a CD4 cell count >500 cells/microL, the baseline prevalence of CKD, defined as an eGFR less than 60 mL/min per 1.73 m² or the presence of dipstick proteinuria, was only 6.2 percent [47]. The majority of participants with CKD had isolated proteinuria. Although the observed prevalence of CKD was low, it was higher than expected given the young age distribution of the patient population.

Risk factors for incident or progressive CKD in adults with HIV include hepatitis C virus (HCV) coinfection, low CD4 cell count, high HIV viral load, and traditional CKD risk factors such as diabetes and hypertension. Treatment with tenofovir disoproxil fumarate (TDF) and boosted protease inhibitors has also been associated with glomerular filtration rate (GFR) decline or decreased GFR in several studies, while ART in general appears to slow the rate of kidney function decline. The following studies provide examples:

●In the prospective cohort of 22,156 military veterans described above, ESKD developed in 366 patients (1.7 percent) [42]. Patients who developed ESKD were more likely to have hypertension, diabetes, cardiovascular disease, HCV coinfection, low CD4 cell count, and high HIV viral load.

●Among 8235 European adults with HIV enrolled in a prospective cohort, the incidence of CKD, defined as an eGFR persistently below 60 mL/min per 1.73 m², was 6 percent [48]. Older age, an AIDS diagnosis, and coinfection with HCV were independent predictors of CKD. HCV viremia was independently associated with CKD in this cohort and with CKD progression in a secondary analysis of two international HIV treatment trials [49]. By contrast, HCV antibody status was associated with CKD risk regardless of HCV viremia in an analysis from the North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD), a consortium of North American HIV cohorts [50].

●In a prospective cohort of 3329 patients who initiated ART under routine clinical care, ART initiation significantly slowed the rate of eGFR decline over a median follow-up of 4.8 years (eGFR loss of 1.4 after versus 2.2 mL/min per 1.73 m² per year before initiation) [51]. The incidence of CKD was 3.2 percent; the risk was higher among Black individuals and those with HCV coinfection, lower CD4 cell count, higher HIV viral load, an AIDS diagnosis, and hypertension. The combination of TDF with a boosted protease inhibitor was also associated with a significant increase in the risk of developing an eGFR below 60 mL/min per 1.73 m² but not below 30 mL/min per 1.73 m².

●Several other large cohort studies have also demonstrated an association between exposure to TDF and/or ritonavir-boosted protease inhibitors and moderately decreased GFR [20-22]. In a large European cohort, use of TDF or ritonavir-boosted atazanavir was associated with an eGFR below 70 mL/min per 1.73 m² but not with an eGFR below 30 mL/min per 1.73 m² [22]. The authors speculated that a high rate of TDF discontinuation among those with low GFR may have prevented further GFR decline. Of note, ritonavir-boosted darunavir, a newer protease inhibitor, does not appear to be associated with an increased risk of CKD [32]. Previous studies have not considered the use of cobicistat as an alternative pharmacoenhancer.

Kidney disease risk scores have been derived in United States military veterans with HIV and in a large European cohort of ART-treated individuals [52,53] and could be used to identify individuals at highest risk.

Causes of CKD in patients with HIV — The etiology of CKD in patients with HIV includes both traditional CKD risk factors such as hypertension, diabetes, obesity, and incomplete recovery from an episode of acute kidney injury (AKI), as well as HIV-related disorders, including HIVAN and immune complex kidney disease. In addition, individuals with viral hepatitis coinfection are also at risk for glomerulonephritis secondary to hepatitis B virus (HBV) or HCV.

HIV-associated nephropathy — The classic kidney disease of HIV infection, HIVAN, was first described in 1984 in patients with advanced HIV infection. (See "HIV-associated nephropathy (HIVAN)".)

HIVAN is a collapsing form of focal segmental glomerulosclerosis (FSGS) with associated tubular microcysts and interstitial inflammation. HIVAN classically presents with significant proteinuria and rapidly progressive kidney disease in the setting of normal blood pressure and normal to enlarged kidneys, although the presentation may be less dramatic in the ART era. The pathogenesis, clinical manifestations, diagnosis, and treatment of HIVAN are discussed in detail elsewhere. (See "HIV-associated nephropathy (HIVAN)".)

Immune complex kidney disease — A large proportion of patients with HIV who have proteinuric kidney disease and suspected HIVAN will have an alternative diagnosis on biopsy [5]. A number of immune complex kidney diseases have been reported in patients with HIV infection, including membranous nephropathy, membranoproliferative and mesangial proliferative glomerulonephritis, "lupus-like" proliferative glomerulonephritis, and immunoglobulin A (IgA) nephropathy [5-9,54]. In addition, a unique immune complex kidney disease with a characteristic "ball in cup" basement membrane reaction was described in a South African biopsy series [55]. The pathogenic relationship between HIV infection and the development of immune complex kidney diseases has not been extensively studied, although a role for viral antigens in circulating immune complexes and deposits has been suggested [7,56].

The natural history and response of immune complex kidney disease to ART have not been well described; the clinical course may vary depending upon the specific immune complex disease. As examples:

●In a case-control study including 751 patients with HIV, those with immune complex kidney disease were predominantly Black patients, and most had advanced HIV disease [57]. The incidence of ESKD in such patients was 32 percent at two years, much lower than that observed in the patients with HIVAN. ART was not associated with improved kidney disease outcomes in patients with immune complex kidney disease.

●In a small series of four patients with HIV-associated IgA nephropathy, the kidney disease followed a benign course similar to that observed in the absence of HIV [56].

●In a retrospective cohort including 42 patients with biopsy-proven HIVAN and 47 patients with alternative histologic diagnoses, Black patients and those with more advanced HIV infection had a higher risk for HIVAN. In addition, there was no evidence of a beneficial role for ART in non-HIVAN kidney diseases, which were associated with slower progression to ESKD regardless of therapy [58].

●A small retrospective study suggests a more aggressive course in patients with immune complex kidney disease resembling lupus glomerulonephritis [59]. In this cohort, 10 of 14 patients progressed to ESKD within one year.

Because the clinical course and response to ART appear to vary across different types of immune complex kidney disease in the setting of HIV infection, it is recommended that these cases be classified and managed based upon the primary histologic lesion [25].

Glomerulonephritis due to hepatitis C virus coinfection — HCV coinfection has been associated with the development of acute and chronic kidney disease in large cohort studies of patients with HIV [12,15,34,48,51]. However, the histologic patterns of kidney disease have not been studied in large populations of coinfected patients. Membranoproliferative glomerulonephritis (MPGN) is strongly associated with HCV infection in the general population and is a common alternative diagnosis in coinfected patients [5,10,54]. (See "Mixed cryoglobulinemia syndrome: Clinical manifestations and diagnosis" and "Overview of kidney disease associated with hepatitis C virus infection".)

Treatment of HCV-associated MPGN in the setting of HIV coinfection has not been rigorously studied. With the availability of effective and well-tolerated direct-acting antiviral therapy for HCV, treatment of the underlying HCV infection is now feasible in these cases. (See "Overview of kidney disease associated with hepatitis C virus infection" and "Mixed cryoglobulinemia syndrome: Treatment and prognosis" and "Treatment of chronic hepatitis C infection in adults with kidney function impairment".)

Diagnosis and management of CKD in patients with HIV — We agree with expert guidelines that recommend screening and early identification of chronic kidney disease (CKD) in patients with HIV [24,25]. Individuals with HIV should have their GFR estimated at least twice yearly and should have either a urinalysis or quantitative assessment of urine protein excretion at least once yearly in order to monitor for the development of kidney disease. Such patients whose eGFR has declined by 25 percent or more to a level below 60 mL/min per 1.73 m², or who have protein excretion greater than 300 mg/day, should be referred for nephrology evaluation.

GFR should be estimated using a creatinine-based estimate; available data do not support the use of cystatin C alone to estimate GFR in adults with HIV [60,61]. The Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) GFR estimates, either based on creatinine alone or creatinine plus cystatin C, are the most accurate for use in adults with HIV on stable ART [60,62], while the Cockcroft-Gault creatinine clearance remains the standard of care for drug dosing. CKD is identified by the presence of proteinuria or by an eGFR less than 60 mL/min per 1.73 m² for at least three months. (See "Assessment of kidney function" and "Definition and staging of chronic kidney disease in adults".)

Detailed discussions of monitoring patients with HIV are presented elsewhere:

●(See "Primary care of adults with HIV", section on 'Hematologic, renal, and hepatic toxicity'.)

●(See "Patient monitoring during HIV antiretroviral therapy", section on 'ART-associated toxicity'.)

Identification of CKD in a patient with HIV should prompt initiation of ART (if not already started) and tight control of comorbid diabetes and hypertension (if present). Medication doses should be adjusted for the calculated creatinine clearance, with particular attention to nucleoside and nucleotide reverse transcriptase inhibitors. (See "Overview of antiretroviral agents used to treat HIV".)

Patients with HIV who have CKD should have at least biannual monitoring of eGFR, with medication dose adjustments as needed. The frequency of monitoring should be increased in patients with mildly reduced eGFR (ie, less than 90 mL/min per 1.73 m²) who are taking medications that have potential nephrotoxic effects, including the nucleotide analog tenofovir and boosted protease inhibitors [24,25].

Early referral to a nephrologist is important to coordinate diagnosis and treatment of CKD in patients with HIV, including discussion of the need for dialysis and the timely placement of dialysis access in patients with progressive disease [24,25]. Analyses of data from the United States Renal Data System (USRDS) have demonstrated similar outcomes in ESKD patients with HIV treated with hemodialysis or peritoneal dialysis, as well as significant improvements in survival in the ART era. (See "Human immunodeficiency virus and dialysis".)

Kidney transplantation is a safe alternative in patients with preserved immune function and undetectable viral load. Based on promising results from South Africa, several transplant centers in the United States are now investigating the safety of kidney transplantation from donors with HIV [63]. (See "Kidney transplantation in adults: Kidney transplantation in patients with HIV".)

ELECTROLYTE DISORDERS — A variety of electrolyte disorders can occur in patients with HIV, which may be due to HIV-associated complications or medications used to treat HIV or its complications. These issues are discussed in detail elsewhere. (See "Electrolyte disturbances with HIV infection".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Glomerular disease in adults".)

SUMMARY AND RECOMMENDATIONS

●Patients with HIV are at risk for both acute kidney injury (AKI) and chronic kidney disease (CKD) secondary to medication nephrotoxicity, HIV-associated nephropathy (HIVAN), and immune complex kidney diseases. In addition, the aging cohort of patients with HIV may be at increased risk for kidney disease related to hepatitis B virus (HBV) or hepatitis C virus (HCV) coinfection and comorbid or treatment-related diabetes and hypertension. (See 'Introduction' above.)

●The incidence of AKI in patients with HIV is higher than it is in patients without HIV. Some risk factors for AKI among patients with HIV are similar to risk factors for AKI in the general population, such as older age, diabetes mellitus, preexisting CKD, and acute or chronic liver disease. However, some risk factors are specific to HIV. As in the general population, the development of AKI increases the risk morbidity and mortality in patients with HIV. (See 'Epidemiology of AKI in patients with HIV' above.)

●The most common types of AKI in patients with HIV are prerenal states and acute tubular necrosis, although other etiologies may also occur. (See 'Causes of AKI in patients with HIV' above.)

●Patients with HIV infection are at risk for nephrotoxicity from antiretroviral therapy (ART), as well as from medications used to treat opportunistic infections or hepatitis virus coinfection. Some of the more commonly implicated agents include the following (see 'Medication nephrotoxicity' above):

•Protease inhibitors, such as indinavir and atazanavir, can cause crystalluria and AKI. Use of some ritonavir-boosted protease inhibitors has also been associated with an increased risk of CKD. The newer protease inhibitor darunavir has not been associated with kidney disease risk.

•Tenofovir disoproxil fumarate (TDF), a nucleoside reverse transcriptase inhibitor, can cause AKI with or without proximal tubular dysfunction. The newer prodrug tenofovir alafenamide may be associated with a lower risk of kidney injury.

•Drugs used to treat opportunistic infections, including the antiviral agents acyclovir, foscarnet, and cidofovir, and the anti-Pneumocystis drugs trimethoprim-sulfamethoxazole and pentamidine, can cause AKI.

●The prevalence and incidence of HIV-related CKD and end-stage kidney disease (ESKD) have decreased with the widespread use of ART but remain higher than in the general population. The etiology of CKD in patients with HIV includes traditional causes of CKD (such as hypertension, diabetes, and incomplete recovery from an episode of AKI) and HIV-related disorders (such as HIVAN and immune complex kidney disease in the setting of HIV). (See 'Epidemiology of CKD in patients with HIV' above and 'Causes of CKD in patients with HIV' above.)

●All patients with HIV should be screened for proteinuria and reduced glomerular filtration rate (GFR). Identification of CKD in a patient with HIV should prompt nephrology referral and initiation of ART. Medication doses should be adjusted for the level of GFR, with particular attention to nucleoside and nucleotide reverse transcriptase inhibitors. Patients with CKD should have at least biannual monitoring of estimated GFR (eGFR), with medication dose adjustments as needed. The frequency of monitoring should be increased in patients who are taking medications that have potential nephrotoxic effects, including the nucleotide analog, tenofovir disoproxil fumarate, and some boosted protease inhibitors. (See 'Diagnosis and management of CKD in patients with HIV' above.)

●A variety of electrolyte disorders can occur in patients with HIV, which may be due to HIV-associated complications or medications used to treat HIV or its complications. (See "Electrolyte disturbances with HIV infection".)