Acute renal failure (ARF) is characterized by rapid deterioration of renal function and typically occurs in the setting of severe illness, dehydration, exposure to nephrotoxic drugs or all [1,2]. The incidence of ARF in HIV-infected patients has ranged from 2.7 to 5.9 per 100 person-years [3,4], with substantially higher rates in those with recently diagnosed HIV infection , and the associated early mortality is high (27–33%) [3,5]. Several HIV-associated parameters, including low CD4 cell count, AIDS, intravenous drug use (IVDU) and hepatitis C coinfection have been identified as risk factors for ARF [3–5]. However, the relationship between combination antiretroviral therapy (cART) and ARF remains unclear, with one study suggesting that receipt of antiretroviral therapy was a risk factor for ARF  and case reports highlighting the potential of indinavir, tenofovir and atazanavir to cause ARF [6–8].
Chronic kidney disease (CKD), defined as an estimated glomerular filtration rate (eGFR) less than 60 ml/min or the presence of proteinuria, is present in 15–20% of HIV-infected patients . HIV-associated nephropathy (HIVAN) is the predominant cause of CKD and end-stage renal failure (ESRF) in black patients [10,11]. Although cART may prevent the development of HIVAN  and delay its progression to ESRF [13–15], tenofovir, indinavir and atazanavir have been associated with CKD or progression of CKD [10,16,17]. Although CKD is a risk factor for ARF in general and HIV-infected populations [5,18,19], the contribution of impaired renal function to ARF in HIV-infected patients is unknown.
We performed a longitudinal, observational cohort study of HIV-associated ARF with the objective to define the incidence of ARF and the relative importance of time-updated CD4 cell count, HIV RNA level, cART status and eGFR as risk factors for ARF.
All HIV-positive patients aged at least 18 years who attended King's College Hospital (KCH) between January 1999 and December 2008 were identified. KCH cares for a multiethnic population in South London, UK. Clinical assessment and laboratory parameter testing are routinely performed every 3–6 months, and cART is provided free of charge. Serum creatinine values were obtained from the hospital electronic patient record system and converted into eGFR, using age, sex, ethnicity and the four-variable modification of diet in renal disease (MDRD) equation .
Episodes of ARF were ascertained by all of the following criteria: confirmed eGFR less than 60 ml/min; nadir eGFR more than 40% reduced from baseline; and duration of ARF less than 90 days. A 40% decline in eGFR equates to an approximate 55% increase in serum creatinine, and we have previously shown 95% concordance between ARF defined by more than 40% reduction in eGFR and ARF defined by more than 50% increase in serum creatinine . Patients with eGFR less than 60 ml/min at baseline who experienced more than 40% decline in renal function (relative to baseline or the level of renal function post-ARF), and patients who experienced more than 40% eGFR decline within 3 months without subsequent recovery of renal function were also included in the analyses. A maximum of one ARF episode for every 3 months of follow-up was included in the analyses.
Demographic, clinical and laboratory parameters were abstracted from the HIV clinical database and electronic patient records. Medical notes were reviewed for all patients who experienced ARF to obtain information on comorbidities [the presence of reduced renal blood flow (dehydration, hypotension, shock), infections, liver disease or cancer] and exposure to potential nephrotoxic medication at the time of ARF . The study was approved by the KCH Research Ethics Committee.
Data were analysed using STATA (version 11; Stata Corporation, College Station, Austin, Texas, USA). Person-time of follow-up time was calculated from the date of initiating HIV care at KCH (or date of entry into the cohort for those in care prior to 1 January 1998) to the date of death or last visit, censored at 31 December 2008. Follow-up time was divided into 1-month intervals; each interval was assigned the most recent CD4 cell count until a new measurement became available (strata: <50, 50–99, 100–199, 200–350, >350 cells/μl), eGFR (most recent value preceding ARF for cases; <60, 60–74, 75–89, ≥90 ml/min per 1.73 m2), HIV RNA (<400, ≥400 copies/ml) and cART exposure status (no cART, indinavir, tenofovir or atazanavir-containing cART, or cART not containing indinavir, tenofovir or atazanavir). Fewer than 10% of patients had CD4 or eGFR measurement gaps of more than 12 months – these patients were excluded from the analysis. The incidence rate of ARF within each CD4 cell count and eGFR stratum was calculated.
Generalized estimation equation (GEE) Poisson models with Huber–White sandwich (robust) estimator were used to estimate the crude and adjusted associations between baseline and time-updated (CD4 cell counts, eGFR, HIV RNA and cART) parameters and ARF incidence. In addition, the effects of indinavir, tenofovir and atazanavir use on ARF incidence were investigated. Unlike ordinary Poisson model, GEE Poisson with robust estimator allows variances to adjust for clustering of within-person time on individual patients. The results are presented as incidence rate ratios (IRRs) with 95% confidence intervals (CIs). All statistical tests are two-sided; associations with P value less than 0.05 in univariate analyses were considered to be statistically significant and taken forward into multivariate models.
During the study period, 2556 patients received HIV care at KCH, of whom 184 (7.2%) experienced 202 episodes of ARF. The overall ARF incidence rate was 2.8 (95% CI 2.41–3.24) episodes per 100 person-years. Patients who developed ARF were older and more likely to be men, of black ethnicity and to have acquired HIV through IVDU. They also had lower nadir CD4 T-cell counts and more often AIDS and hepatitis C coinfection (Table 1).
At the time of ARF, only 45% of patients were receiving cART. Opportunistic infections were the commonest underlying illness in patients with CD4 cell counts less than 50 μl, nonopportunistic infections were common throughout and liver disease, cancer and CKD were present in a significant minority of patients, whereas reduced renal blood flow and exposure to nephrotoxic medications was identified in 80 and 76%, respectively (Table 1). Death during follow-up was more common among patients who experienced ARF compared to those without ARF (32.1 vs. 3.7%, P < 0.001), and patients with ARF had shorter mean observation times [2.37 (SD 1.99) years vs. 2.81 (2.15) years, P = 0.001] and more serum creatinine measurements [mean 86.0 (69.9) vs. 36.4 (33.4), P < 0.001], though they were less likely to become lost to follow-up (4.8 vs. 8.6%, P = 0.13).
When follow-up time was stratified by CD4 cell count, we observed a stepwise increase in ARF incidence of 1.2 (0.89–1.59), 2.3 (1.7–3.1), 5.9 (4.2–8.2), 13.3 (8.8–20.0) and 26.4 (20.3–34.5) episodes per 100 person-years for person-time accrued with CD4 cell counts of more than 350, 200–350, 100–199, 50–99 and less than 50 cells/μl. Similarly, the incidence of ARF increased as the eGFR declined, with ARF incidence rates of 1.4 (1.1–1.8), 1.9 (1.3–2.8), 7.1 (4.9–10.3) and 56.1 (45.0–69.9) episodes per 100 person-years for person-time accrued with eGFR at least 90, 75–89, 60–74 and less than 60 ml/min per 1.73 m2. In univariate analysis, current CD4 cell count and current eGFR were strongly associated with ARF. In addition, older age, history of IVDU or AIDS and hepatitis B/C coinfection were associated with an increased incidence of ARF, whereas ethnicity, sex, use of cART and HIV RNA were not associated with ARF (Table 2). In multivariate analysis, lower current CD4 cell count and lower current eGFR remained significantly associated with ARF, whereas the association with IVDU and hepatitis B or C coinfection became nonsignificant. AIDS at or prior to ARF, although strongly associated with ARF in univariate analysis [IRR = 19.8 (95% CI 13.40–29.21), P < 0.001], was excluded from the multivariable analysis because of its strong association with CD4 cell count less than 50 cells/μl. Current CD4 cell count was found to be a better predictor of ARF than nadir CD4 cell count (data not shown). When assessed in multivariate models, exposure to any cART [IRR 1.40 (0.92–2.15)] or exposure to cART containing potentially nephrotoxic agents (indinavir, tenofovir or atazanavir) was not significantly associated with an increased incidence of ARF (Table 2).
This is the first study to comprehensively evaluate the effects of cART, CD4 cell count and HIV RNA level, coinfection with viral hepatitis and renal function as predictors of ARF. We found that lower current CD4 cell count and lower current eGFR were independent risk factors for developing ARF. Hepatitis B or C coinfection, HIV viraemia and exposure to cART with or without indinavir, tenofovir or atazanavir were not significantly associated with ARF.
CKD, as defined by a recent eGFR of less than 60 ml/min, was associated with an approximately 25-fold increased incidence of ARF. Impaired renal function has been identified as a risk factor for ARF in patients undergoing cardiac surgery [18,19] and in hospitalized HIV-positive patients . Our results showed that lesser degrees of renal impairment (eGFR 60–74 ml/min) may also be associated with an increased risk of ARF. We recently reported clinically significant atherosclerosis in 21% of HIV patients with CKD , whereas others have observed impaired renal function (mean eGFR 69 ml/min) in HIV patients with myocardial infarction . Reduced eGFR may, thus, be a manifestation of underlying cardiovascular disease, the presence of which could predispose to ARF by a reduced ability to maintain adequate renal blood flow during severe illness or by an increased susceptibility to drug-induced nephrotoxicity. Irrespective of renal function prior to ARF, reduced renal blood flow and exposure to potentially nephrotoxic drugs were common in patients who developed ARF.
In accordance with previous studies [3–5], the incidence of ARF was markedly increased in patients with severe immunodeficiency. A small but significantly increased risk of ARF was apparent in patients with CD4 cell counts of 200–350 cells/μl, which is consistent with the increased risk of AIDS-defining illnesses  and non-AIDS morbidity  in this stratum, and provides further support for current guidelines to initiate cART at CD4 cell counts at or above 350 cells/μl [24,25]. It is worth noting that, after adjustment for current CD4 cell count and other factors, cART use was not associated with ARF in our cohort. This may to some extent be due to the high proportion of ‘late presenters’ (new HIV diagnosis in patients with CD4 cell count <200 cells/μl) among our patients with ARF , low nephrotoxic potential of current cART regimens, minimal usage of indinavir  and avoidance of tenofovir in patients with reduced eGFR .
The present study has several limitations. In the absence of a generally accepted case definition, ARF episodes were identified on the basis of declines in eGFR , and clinical information was obtained through retrospective case note review. During most of the study period, viral hepatitis status was determined at baseline only and not routinely reevaluated annually, and prevalence of hepatitis B/C coinfection was low. As this was an observational study, the results may have been affected by bias, unmeasured confounding and loss to follow-up. Nonetheless, the systematic case ascertainment, availability of detailed clinical information, prolonged study period (10 years), provision of free healthcare, including cART to the study participants and the multiethnic population add to the strengths of the study and the more general applicability of the study findings.
In summary, we have shown that current CD4 cell count and renal function are the most important predictors of ARF in HIV-infected patients. After adjustment for level of immunodeficiency and renal dysfunction, we observed no association of ARF with cART in general, or with specific antiretrovirals with nephrotoxic potential. ARF is a complication of late HIV diagnosis  and, as for other forms of severe kidney disease [10,11,13], earlier HIV diagnosis and timely initiation of cART is likely to be the best way to reduce the burden of HIV-associated ARF.
F.I., C.S. and F.A.P. designed the study. C.N., E.C. and J.R. reviewed case notes and abstracted clinical information. F.I. performed the analyses, with assistance from L.J.C., L.B. and C.S. F.I., C.S., B.M.H. and F.P. assisted with the interpretation of the results. F.I. and F.A.P. wrote the manuscript with input from all authors. The final version of the manuscript was approved by all authors.
C.N. has received funding for conference attendance from Bristol-Myers Squibb and ViiV Healthcare, and J.R. from GlaxoSmithKline. B.M.H. has received honoraria for consultancy from Abbott Laboratories, AM Pharma and Gilead Sciences. C.S. has received funding for conference attendance and honoraria for membership of advisory boards, data safety and monitoring boards and consultancy from Gilead Sciences, Bristol-Myers Squibb and Janssen-Cilag. F.A.P. has received funding for conference attendance and honoraria for membership of advisory boards and consultancy from Gilead Sciences, Bristol-Myers Squibb, Janssen-Cilag, GlaxoSmithKline, ViiV healthcare, Merck and Roche, and research funding from Bristol-Myers Squibb, GlaxoSmithKline and ViiV healthcare. F.I., E.C., L.J.C. and L.B. have no conflict of interest to report.
Part of this paper was presented at the 17th Conference on Retroviruses and Opportunistic Infection (CROI) held on 16–19 February 2010 in San Francisco, California, USA, [abstract #734].
1. Lameire N, Van Biesen W, Vanholder R. The changing epidemiology of acute renal failure. Nat Clin Pract Nephrol 2006; 2:364–377.
2. Izzedine H, Baumelou A, Deray G. Acute renal failure in HIV patients. Nephrol Dial Transplant 2007; 22:2757–2762.
3. Roe J, Campbell LJ, Ibrahim F, Hendry BM, Post FA. HIV care and the incidence of acute renal failure. Clin Infect Dis 2008; 47:242–249.
4. Franceschini N, Napravnik S, Eron JJ Jr, Szczech LA, Finn WF. Incidence and etiology of acute renal failure among ambulatory HIV-infected patients. Kidney
Int 2005; 67:1526–1531.
5. Wyatt CM, Arons RR, Klotman PE, Klotman ME. Acute renal failure in hospitalized patients with HIV: risk factors and impact on in-hospital mortality. AIDS 2006; 20:561–565.
6. Zimmermann AE, Pizzoferrato T, Bedford J, Morris A, Hoffman R, Braden G. Tenofovir-associated acute and chronic kidney
disease: a case of multiple drug interactions. Clin Infect Dis 2006; 42:283–290.
7. Schmid S, Opravil M, Moddel M, Huber M, Pfammatter R, Keusch G, et al
. Acute interstitial nephritis of HIV-positive patients under atazanavir and tenofovir therapy in a retrospective analysis of kidney
biopsies. Virchows Arch 2007; 450:665–670.
8. Berns JS, Cohen RM, Silverman M, Turner J. Acute renal failure due to indinavir crystalluria and nephrolithiasis: report of two cases. Am J Kidney
Dis 1997; 30:558–560.
9. Post FA, Holt SG. Recent developments in HIV and the kidney
. Curr Opin Infect Dis 2009; 22:43–48.
10. Campbell LJ, Ibrahim F, Fisher M, Holt SG, Hendry BM, Post FA. Spectrum of chronic kidney
disease in HIV-infected patients. HIV Med 2009; 10:329–336.
11. Bansi L, Hughes A, Bhagani S, Mackie NE, Leen C, Levy J, et al
. Clinical epidemiology of HIV-associated end-stage renal failure in the UK. AIDS 2009; 23:2517–2521.
12. Lucas GM, Eustace JA, Sozio S, Mentari EK, Appiah KA, Moore RD. Highly active antiretroviral therapy and the incidence of HIV-1-associated nephropathy: a 12-year cohort study. AIDS 2004; 18:541–546.
13. Post FA, Campbell LJ, Hamzah L, Collins L, Jones R, Siwani R, et al
. Predictors of renal outcome in HIV-associated nephropathy. Clin Infect Dis 2008; 46:1282–1289.
14. Szczech LA, Gupta SK, Habash R, Guasch A, Kalayjian R, Appel R, et al
. The clinical epidemiology and course of the spectrum of renal diseases associated with HIV infection. Kidney
Int 2004; 66:1145–1152.
15. Atta MG, Gallant JE, Rahman MH, Nagajothi N, Racusen LC, Scheel PJ, et al
. Antiretroviral therapy in the treatment of HIV-associated nephropathy. Nephrol Dial Transplant 2006; 21:2809–2813.
16. Mocroft A, Kirk O, Gatell J, Reiss P, Gargalianos P, Zilmer K, et al
. Chronic renal failure among HIV-1-infected patients. AIDS 2007; 21:1119–1127.
17. Kirk O, Mocroft A, Reiss P, De Wit S, Sedlacek D, Beniowski M, et al.Chronic kidney disease and exposure to ART in a large cohort with long-term follow-up: the EuroSIDA study.17th Conference on Retroviruses and Opportunistic Infections
; 16–19 February 2010; San Francisco, California, USA; 2010.
18. Mangano CM, Diamondstone LS, Ramsay JG, Aggarwal A, Herskowitz A, Mangano DT. Renal dysfunction after myocardial revascularization: risk factors, adverse outcomes, and hospital resource utilization. The Multicenter Study of Perioperative Ischemia Research Group. Ann Intern Med 1998; 128:194–203.
19. Thakar CV, Arrigain S, Worley S, Yared JP, Paganini EP. A clinical score to predict acute renal failure after cardiac surgery. J Am Soc Nephrol 2005; 16:162–168.
20. Gupta SK, Eustace JA, Winston JA, Boydstun II, Ahuja TS, Rodriguez RA, et al
. Guidelines for the management of chronic kidney
disease in HIV-infected patients: recommendations of the HIV Medicine Association of the Infectious Diseases Society of America. Clin Infect Dis 2005; 40:1559–1585.
21. George E, Lucas GM, Nadkarni GN, Fine DM, Moore R, Atta MG. Kidney
function and the risk of cardiovascular events in HIV-1-infected patients. AIDS 2010; 24:387–394.
22. Podlekareva D, Mocroft A, Dragsted UB, Ledergerber B, Beniowski M, Lazzarin A, et al
. Factors associated with the development of opportunistic infections in HIV-1-infected adults with high CD4+
cell counts: a EuroSIDA study. J Infect Dis 2006; 194:633–641.
23. Phillips AN, Neaton J, Lundgren JD. The role of HIV in serious diseases other than AIDS. AIDS 2008; 22:2409–2418.
24. Gazzard BG. British HIV Association Guidelines for the treatment of HIV-1-infected adults with antiretroviral therapy 2008. HIV Med 2008; 9:563–608.
25. EACS. Prevention and management of noninfectious co-morbidities in HIV. http://www.europeanaidsclinicalsociety.org/guidelinespdf/2_Non_Infectious_Co_Morbidities_in_HIV.pdf
. [Accessed 15.04.2010].
26. Roling J, Schmid H, Fischereder M, Draenert R, Goebel FD. HIV-associated renal diseases and highly active antiretroviral therapy-induced nephropathy. Clin Infect Dis 2006; 42:1488–1495.