Human immunodeficiency virus-1-associated nephropathy (HIVAN) is characterized by high-grade proteinuria and rapid progression to end-stage renal disease (ESRD) . ESRD incurs high cost , and has been associated with very high mortality in HIV-1-infected individuals in the past [3,4], although recent reports suggest improved survival in the era of highly active antiretroviral therapy (HAART) . HIVAN occurs overwhelmingly in individuals of African or Haitian descent  with advanced HIV-1 disease , and is the third leading cause of ESRD in African Americans between the ages of 20 and 64 years . As African Americans constitute only 13% of the US population, but accounted for an estimated 54% of new HIV infections in 2000 , some experts have described HIVAN as a looming urban epidemic .
Case reports and case series have suggested that HAART [11–14], corticosteroids [15,16], and angiotensin-converting enzyme inhibitors  may stabilize renal function and delay progression to ESRD in patients with HIVAN. However, we are aware of no longitudinal cohort studies evaluating changes in HIVAN incidence over the period of widespread dissemination of HAART. Such data have important implications for resource planning, research on screening strategies for detecting early stage HIVAN, and new treatments. We evaluated factors associated with development of HIVAN from 1989–2001 in a large, predominantly African American cohort of HIV-1-infected individuals in Baltimore, Maryland, USA.
Database and case definition
The Johns Hopkins HIV Clinic provides longitudinal care to a large number of HIV-1-infected individuals in the Baltimore area. All patients enrolling in the clinic undergo a detailed baseline assessment, with collection of demographic, social, behavioral, and clinical information. A comprehensive panel of blood tests and a urinalysis is obtained at clinic enrollment. Serum chemistries, renal function tests, hepatic function tests, complete blood count, CD4 cell counts, and HIV-1 RNA (after 1995) are routinely monitored at 3-month intervals (or more frequently if clinically indicated) following enrollment.
An observational clinical database comprising over 4000 patients who have been treated in the clinic was established in 1989 . Information from clinical records is reviewed and abstracted by professionally trained technicians onto structured data collection forms, then entered into an automated database. The clinic medical records, the main hospital medical record and various institutional automated databases (e.g., laboratory, radiology, pathology, discharge summaries) are abstracted. Comprehensive demographic, clinical, laboratory, pharmaceutical, and psychosocial data are collected at times corresponding to enrollment in the clinic and at 6-month intervals thereafter. The Johns Hopkins Institutional Review Board approved this clinical cohort study.
According to a protocol, consistently applied since cohort inception, data abstractors identified patients as having possible HIVAN if this diagnosis was considered at any time in the clinical or pathological records, or if patients had progressive renal disease of unknown etiology. Patients who underwent a renal biopsy were also identified from an automated institutional pathology database.
We subsequently conducted a detailed medical record review of all patients initially identified as having possible HIVAN. Information was recorded on structured data collection forms and included, relevant laboratory tests, renal imaging studies, use of nephrotoxic drugs, other diagnoses relevant to renal disease, and nephrology consultant findings. HIVAN cases were categorized as biopsy defined if characteristic histopathologic findings were documented . Because many patients with progressive renal disease never underwent renal biopsy, HIVAN cases were categorized as clinically defined if no biopsy had been performed and the patient had (1) proteinuria in excess of 3 g/24 h or a urine protein/creatinine concentration ratio above 2.5, and (2) a subacute progressive increase in serum creatinine. Clinically defined HIVAN cases were excluded by the presence of: (1) acute renal failure, with return to a stable baseline serum creatinine below 1.5 mg/dl; (2) history of collagen vascular disease, diabetes of more than 5 years duration, uncontrolled hypertension, or other underlying disease likely to cause chronic renal failure; or (3) detectable cryoglobulinemia, serum complement levels below the lower limit of normal, bilateral small kidneys on sonogram, or a urine sediment containing red blood cell casts.
Only incident HIVAN cases were included in the analysis. Patients who entered the cohort with a confirmed or possible diagnosis of HIVAN or advanced renal disease were excluded. The date of HIVAN onset was defined as the date that renal disease was first documented as a problem in patients’ clinical records.
Demographic, clinical, and laboratory data were compared in the general clinic population, patients with biopsy defined HIVAN, and patients with clinically defined HIVAN. Categorical variables were compared with Fisher's exact test and continuous variables with the Wilcoxon rank-sum test. HIVAN incidence rates were expressed as cases per 1000 person-years. Sex, race, and HIV-1 exposure group were included as fixed covariates. Age, calendar time, AIDS status, and use of antiretroviral therapy were included as time-dependent covariates. AIDS was defined according to the 1993 Centers for Disease Control and Prevention guidelines , which included CD4 cell count < 200 × 106 cells/l. Antiretroviral therapy was categorized hierarchically as none, nucleoside reverse transcriptase inhibitor (NRTI) only, or HAART. HAART was defined as use of a protease inhibitor, a non-nucleoside reverse transcriptase inhibitor, or abacavir, in a regimen including three or more antiretroviral agents. Once a given level of therapy had been used for more than 30 days, patients were categorized at that level. Factors associated with HIVAN incidence were evaluated in univariate and multivariate Poisson regression models. Potential interactions between variables were assessed by stratification and combining terms in multivariate models.
Over 11 732 person-years of follow-up from 1989 to 2001, 315 potential HIVAN cases were identified during cohort data abstraction, of which 313 (99%) had records available for review. Of those, eight (3%) had been miscoded, 29 (9%) had an acute renal condition that resolved completely, 31 (10%) were lost to follow-up or died before renal disease could be further characterized, 75 (24%) had biopsy or clinical evidence most consistent with a renal diagnosis other than HIVAN, 35 (11%) had renal disease of unknown etiology, but criteria for HIVAN were not met, 105 (34%) met criteria for clinically defined HIVAN, and 30 (10%) had biopsy-defined HIVAN. Of the 135 patients with HIVAN, 94 (23 biopsy defined and 71 clinically defined) developed renal disease following enrollment in the clinical cohort (incident cases) and were included in the analysis. Calendar period was not significantly associated with the proportions of reviewed cases that were ultimately categorized as biopsy defined HIVAN, clinically defined HIVAN, or excluded (data not shown). Of the 71 patients in the cohort who underwent a renal biopsy, the frequencies of histopathological findings were as follows: 30 (42%) showed HIVAN, 18 (25%) had an alternative form of glomerulonephritis other than HIVAN, eight (11%) were non-diagnostic, six (8%) had diabetic glomerulosclerosis, two biopsies showed hypertensive nephrosclerosis, two interstitial nephritis, two acute tubular necrosis, and one showed amyloid, one lymphoma and one obstructive uropathy.
In comparison with the general clinic population, baseline factors associated with development of HIVAN included, African American race, lower CD4 cell count, higher HIV-1 RNA level (in the subset followed after 1995, when this assay became available), history of injection drug use, and not being a man who has sex with men (Table 1). Hepatitis C antibody positive status (in the subset of 2960 patients in whom results from this assay were available) was associated with HIVAN in univariate analysis (Table 1), but did not remain significantly associated with HIVAN after adjustment for race [rate ratio (RR) for hepatitis C antibody positive versus negative, 1.5; 95% confidence interval (CI), 0.9–2.4]. There were no statistically significant differences in covariates between patients with biopsy defined and clinically defined HIVAN (Table 1).
The HIVAN incidence rates were 10.1 and 1.1 per 1000 person-years in African American and white patients, respectively (RR, 9.0; 95% CI, 2.8–28.4). Patients with AIDS developed HIVAN at a rate of 12.5 compared to 3.1 per 1000 person-years in patients without AIDS (RR, 4.1; 95% CI, 2.4–6.9). Compared with prior calendar periods, the incidence of HIVAN declined in 1998 through 2001, in both AIDS-defined and non-AIDS defined individuals (Fig. 1). In a Poisson regression model, controlling for AIDS status, the incidence of HIVAN was similar in the 1995–1997 period and the 1989–1994 period (RR, 1.2; 95% CI, 0.8–2.0). However, HIVAN incidence declined significantly in the 1998–2001 period compared to the 1995–1997 period (RR, 0.5; 95% CI, 0.3–0.9).
Figure 2 shows the association of antiretroviral use with subsequent HIVAN incidence. The risk of HIVAN was low in patients who had not developed AIDS, and it is notable that no HIVAN cases developed when HAART was used and AIDS had not occurred (1071 person-years of follow-up). Among individuals with AIDS, lower rates of HIVAN were strongly associated with NRTI use and HAART use, compared with no antiretroviral therapy use (P < 0.001 for trend, Fig. 2).
In a multivariate Poisson regression model African American race (RR, 7.8; 95% CI, 2.5–24.6), the presence of AIDS (RR, 5.0; 95% CI, 2.8–8.7), and antiretroviral treatment status (RR, 0.8; 95% CI, 0.5–1.3 for NRTI compared with no therapy; RR, 0.4; 95% CI, 0.2–0.7 for HAART compared with no therapy) remained significantly associated with HIVAN. Calendar time, injection drug use, and men having sex with men exposure were not significantly associated with HIVAN after adjusting for other variables and were not included in the final model. Similar trends were observed in an analysis including only biopsy-defined HIVAN cases. For example, compared with not using antiretroviral therapy, the RR of HIVAN was 1.1 (95% CI, 0.4–3.3) in those using NRTI therapy, and 0.1 (95% CI, 0.01–0.8) in those using HAART, in a Poisson model adjusting for race and AIDS status.
To evaluate the associations of immunologic factors, virologic factors and HIVAN incidence, CD4 cell count and HIV-1 RNA values were included as time-dependent covariates (updated at 6-month intervals) in a Poisson regression model. Because HIV-1 RNA levels were not available in clinical practice until 1995, this analysis included a subset of 2716 individuals in whom at least one CD4 cell count and one HIV-1 RNA value had been obtained. CD4 cell count was categorized as greater than or less than 200 × 106 cells/l, and HIV-1 RNA was categorized as greater than or less than 100 000 copies/ml. Table 2 shows the associations of proximate CD4 cell count, HIV-1 RNA level and the risk of HIVAN. Compared with having a CD4 cell count ≥ 200 × 106 cells/l, a CD4 cell count < 200 × 106 cells/l was strongly associated with increased HIVAN risk at both low (RR, 3.5; 95% CI, 1.8–7.0) and high HIV-1 RNA levels (RR, 3.0; 95% CI, 1.1–8.7). In contrast, compared with having an HIV-1 RNA < 100 000 copies/ml, having an HIV-1 RNA ≥ 100 000 copies/ml was not significantly associated with HIVAN risk at either high (RR, 2.0; 95% CI, 0.7–6.1) or low CD4 cell counts (RR, 1.7; 95% CI, 0.9–3.2).
In this first longitudinal study of factors associated with HIVAN incidence, we found a HIVAN rate of approximately 1% per year in African Americans. The presence of AIDS conferred a five-fold increased risk. There was a 50% decline in HIVAN incidence in 1998–2001 compared with 1995–1997. This temporal trend was explained by HAART use, which was associated with a 60% reduction in HIVAN risk. The effect of HAART was most striking in patients with AIDS (Fig. 2), where HIVAN burden was the greatest. 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 temporal decrease in HIVAN incidence that we observed occurred during a time when the incidence of ESRD was increasing in the general population. For example, in the state of Maryland, the age-, gender- and race-adjusted rate of ESRD in non-diabetics increased from 141 per million in 1991 to 210 per million in 2000 , making it very unlikely that the trend we observed was explained by secular trends in kidney disease in the general population.
The nearly exclusive occurrence of HIVAN in African Americans  and the strong correlation between HIVAN and advanced HIV-1 disease  have been reported by others. A murine model of HIVAN has suggested that expression of HIV-1 regulatory and envelope genes, and not systemic perturbations in cytokine regulation, is sufficient to cause HIVAN . Additionally, renal epithelial cells have been found to be a reservoir for HIV-1, even with the effective use of HAART . Our study strongly suggests that HAART reduces the risk of developing HIVAN in susceptible individuals; however, the mechanism of this effect remains unclear. Our data indicate that a proximate CD4 cell count < 200 × 106 cells/l is a stronger risk factor for HIVAN than a high viral load. This may suggest that the development of HIVAN requires advanced immunosuppression in addition to exposure to viral antigens. A second possibility, which could not be specifically addressed in our cohort, is that low CD4 cell counts are a marker for prolonged duration of HIV infection, which may be the primary risk factor for HIVAN.
The major limitation of our study was that all HIVAN diagnoses were not biopsy confirmed. A broad range of renal pathology has been reported in HIV-1-infected patients with renal disease, including HIVAN, classical focal segmental glomerulosclerosis, acute tubular necrosis, interstitial nephritis, thrombotic micro-angiopathies, and immune complex renal disease . However, as biopsies were not performed in the majority of patients in the cohort with clinically significant renal disease, including only biopsy-proven cases would substantially underestimate the true burden of disease. Our goal in defining clinical criteria for HIVAN, in patients in whom a renal biopsy was not performed, was to design, a reasonable, evidence-based categorization scheme, which could be applied uniformly over time in our clinical cohort. Possible HIVAN cases were identified during prospective cohort data collection, according to sensitive criteria, which were not altered during the 12-year study. We retrospectively adjudicated clinically defined HIVAN cases using a conservative and uniformly applied coding protocol. While misclassifications may have occurred with our clinical coding protocol, such misclassification would tend to bias toward the null hypothesis of no association between HAART use and HIVAN incidence. Patients with clinically defined HIVAN were similar to patients with biopsy-defined disease across a wide range of demographic and clinical variables, including dialysis-free survival time (Table 1). Finally, a statistically significant protective association between HAART and HIVAN incidence was observed in a subgroup analysis including only biopsy-defined HIVAN cases.
Our study implies that the appropriate use of HAART may attenuate the growing epidemic of HIVAN, which disproportionately affects HIV-1-infected patients of African descent. Longer-term follow-up will be needed to determine if renal damage in susceptible individuals is halted or merely slowed by HAART, particularly when control of viremia is incomplete or intermittent. Research is needed to determine whether early markers of glomerular injury , family history of renal disease , or genetic markers may be used to predict predisposition to HIVAN in HIV-1-infected individuals. As guidelines for HAART initiation have become more conservative , identification of increased HIVAN risk may inform decisions about initiation of HAART in individuals with asymptomatic HIV-1 disease.
This research was supported by the National Institute on Drug Abuse (R01-DA11602, K23-DA015616, and K24-DA00432).
Conflict of interest statement: The authors have no conflicts of interest, potential or actual, with the data analysis reported in this manuscript.
1. Rao TKS. Clinical features of human immunodeficiency virus associated nephropathy.Kidney Int
(Suppl. 35): S13–S18.
2. Lee H, Manns B, Taub K, Ghali WA, Dean S, Johnson D, et al
. Cost analysis of ongoing care of patients with end-stage renal disease: the impact of dialysis modality and dialysis access.Am J Kidney Dis
3. Dave MB, Shabih K, Blum S. Maintenance hemodialysis in patients with HIV-associated nephropathy.Clin Nephrol
4. Rao TK, Friedman EA, Nicastri AD. The types of renal disease in the acquired immunodeficiency syndrome.N Engl J Med
5. Ahuja TS, Grady J, Khan S. Changing trends in the survival of dialysis patients with human immunodeficiency virus in the United States.J Am Soc Nephrol
6. Shahinian V, Rajaraman S, Borucki M, Grady J, Hollander WM, Ahuja TS. Prevalence of HIV-associated nephropathy in autopsies of HIV-infected patients.Am J Kidney Dis
7. Winston JA, Klotman ME, Klotman PE. HIV-associated nephropathy is a late, not early, manifestation of HIV-1 infection.Kidney Int
8. US Renal Data System (USRDS). USRDS 2000 Annual Data Report.
Bethesda, MD: National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health; 2000.
9. UNAIDS. Report on the Global HIV/AIDS Epidemic, July 2002.
Available at http://www.unaids.org
. Last accessed December 2002.
10. Monahan M, Tanji N, Klotman PE. HIV-associated nephropathy: an urban epidemic.Semin Nephrol
11. Wali RK, Drachenberg CI, Papadimitriou JC, Keay S, Ramos E. HIV-1-associated nephropathy and response to highly-active antiretroviral therapy.Lancet
12. Szczech LA, Edwards LJ, Sanders LL, van der HC, Bartlett JA, Heald AE, et al
. Protease inhibitors are associated with a slowed progression of HIV-related renal diseases.Clin Nephrol
13. Winston JA, Bruggeman LA, Ross MD, Jacobson J, Ross L, D'Agati VD, et al
. Nephropathy and establishment of a renal reservoir of HIV type 1 during primary infection.N Engl J Med
14. Cosgrove CJ, Abu-Alfa AK, Perazella MA. Observations on HIV-associated renal disease in the era of highly active antiretroviral therapy.Am J Med Sci
15. Eustace JA, Nuermberger E, Choi M, Scheel PJ, Jr., Moore R, Briggs WA. Cohort study of the treatment of severe HIV-associated nephropathy with corticosteroids.Kidney Int
16. Smith MC, Austen JL, Carey JT, Emancipator SN, Herbener T, Gripshover B, et al
. Prednisone improves renal function and proteinuria in human immunodeficiency virus-associated nephropathy.Am J Med
17. Burns GC, Paul SK, Toth IR, Sivak SL. Effect of angiotensin-converting enzyme inhibition in HIV-associated nephropathy.J Am Soc Nephrol
18. Moore RD. Understanding the clinical and economic outcomes of HIV therapy: The Johns Hopkins HIV clinical practice cohort.J Acquir Immune Defic Syndr
19. D'Agati V, Appel GB. Renal pathology of human immunodeficiency virus infection.Semin Nephrol
20. Centers for Disease Control and Prevention. 1993 Revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults.MMWR
21. U.S. Renal Data System. USRDS 2002 Annual Data Report: Atlas of End-stage-renal Diseases in the United States.
Bethesda, MD: National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health; 2002.
22. Bruggeman LA, Dikman S, Meng C, Quaggin SE, Coffman TM, Klotman PE. Nephropathy in human immunodeficiency virus-1 transgenic mice is due to renal transgene expression.J Clin Invest
23. Kimmel PL. The nephropathies of HIV infection: pathogenesis and treatment.Curr Opin Nephrol Hypertens
24. Szczech LA, Gange SJ, van der HC, Bartlett JA, Young M, Cohen MH, et al
. Predictors of proteinuria and renal failure among women with HIV infection.Kidney Int
. 2002, 61
25. Freedman BI, Soucie JM, Stone SM, Pegram S. Familial clustering of end-stage renal disease in blacks with HIV-associated nephropathy.Am J Kidney Dis
26. Department of Health and Human Services and Henry J. Kaiser Family Foundation. Guidelines for the Use of Antiretroviral Agents in HIV-infected Adults and Adolescents.
Available at http://www.hivatis.org
. Last accessed December 2002.