Antiretroviral therapy improves renal function among HIV-infected Ugandans
2008; Elsevier BV; Volume: 74; Issue: 7 Linguagem: Inglês
10.1038/ki.2008.305
ISSN1523-1755
AutoresPhilip J. Peters, David Moore, Jonathan Mermin, John T. Brooks, Robert Downing, Willy Were, Aminah Kigozi, Kate Buchacz, Paul J. Weidle,
Tópico(s)HIV/AIDS Research and Interventions
ResumoRenal dysfunction is a severe complication of advanced HIV disease. We evaluated the impact of highly active antiretroviral therapy (HAART) on renal function among HIV-infected Ugandans in the Home-Based AIDS Care clinical trial. The patients presented with symptomatic HIV disease or CD4 cell count ≤250 cells/mm3 and creatinine clearances above 25 ml/min determined by the Cockcroft-Gault equation. Of the 508 patients at baseline, 8% had a serum creatinine over 133 μmol/l and about 20% had reduced renal function evidenced by a creatinine clearance between 25 and 50 ml/min. After 2 years of HAART, the median serum creatinine was significantly decreased by 16% while the median creatinine clearance significantly increased 21%. The median creatinine clearance of patients with renal dysfunction at baseline, increased by 53% during 2 years of treatment. In multivariable analysis, a baseline creatinine above 133 μmol/l, a weight gain of more than 5 kg over the 2 years, female gender and a WHO stage 4 classification were all associated with greater improvements in creatinine clearance on HAART. Our study shows that renal dysfunction was common with advanced HIV disease in Uganda but this improved following 2 years of HAART. Renal dysfunction is a severe complication of advanced HIV disease. We evaluated the impact of highly active antiretroviral therapy (HAART) on renal function among HIV-infected Ugandans in the Home-Based AIDS Care clinical trial. The patients presented with symptomatic HIV disease or CD4 cell count ≤250 cells/mm3 and creatinine clearances above 25 ml/min determined by the Cockcroft-Gault equation. Of the 508 patients at baseline, 8% had a serum creatinine over 133 μmol/l and about 20% had reduced renal function evidenced by a creatinine clearance between 25 and 50 ml/min. After 2 years of HAART, the median serum creatinine was significantly decreased by 16% while the median creatinine clearance significantly increased 21%. The median creatinine clearance of patients with renal dysfunction at baseline, increased by 53% during 2 years of treatment. In multivariable analysis, a baseline creatinine above 133 μmol/l, a weight gain of more than 5 kg over the 2 years, female gender and a WHO stage 4 classification were all associated with greater improvements in creatinine clearance on HAART. Our study shows that renal dysfunction was common with advanced HIV disease in Uganda but this improved following 2 years of HAART. Renal dysfunction progressing to end-stage renal disease is a common complication of advanced human immunodeficiency virus (HIV) infection, particularly among African Americans.1.Gupta S.K. Eustace J.A. Winston J.A. 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: 1559-1585Crossref PubMed Scopus (501) Google Scholar,2.Winston J.A. Burns G.C. Klotman P.E. The human immunodeficiency virus (HIV) epidemic and HIV-associated nephropathy.Semin Nephrol. 1998; 18: 373-377PubMed Google Scholar,3.Choi A.I. Rodriguez R.A. Bacchetti P. et al.The impact of HIV on chronic kidney disease outcomes.Kidney Int. 2007; 72: 1380-1387Abstract Full Text Full Text PDF PubMed Scopus (53) Google Scholar,4.Wyatt C.M. Klotman P.E. HIV-1 and HIV-associated nephropathy 25 years later.Clin J Am Soc Nephrol. 2007; 2: S20-S24Crossref PubMed Scopus (32) Google Scholar Although HIV-associated renal disease has been documented in sub-Saharan Africa,5.Gerntholtz T.E. Goetsch S.J. Katz I. HIV-related nephropathy: a South African perspective.Kidney Int. 2006; 69: 1885-1891Abstract Full Text Full Text PDF PubMed Scopus (122) Google Scholar,6.Han T.M. Naicker S. Ramdial P.K. et al.A cross-sectional study of HIV-seropositive patients with varying degrees of proteinuria in South Africa.Kidney Int. 2006; 69: 2243-2250Abstract Full Text Full Text PDF PubMed Scopus (153) Google Scholar,7.Wools-Kaloustian K. Gupta S.K. Muloma E. et al.Renal disease in an antiretroviral-naive HIV-infected outpatient population in Western Kenya.Nephrol Dial Transplant. 2007; 22: 2208-2212Crossref PubMed Scopus (78) Google Scholar,8.Emem C.P. Arogundade F. Sanusi A. et al.Renal disease in HIV-seropositive patients in Nigeria: an assessment of prevalence, clinical features and risk factors.Nephrol Dial Transplant. 2008; 23: 741-746Crossref PubMed Scopus (73) Google Scholar,9.Kalayjian R.C. Franceschini N. Gupta S.K. et al.Suppression of HIV-1 replication by antiretroviral therapy improves renal function in persons with low CD4 cell counts and chronic kidney disease.AIDS. 2008; 22: 481-487Crossref PubMed Scopus (99) Google Scholar little is known about its prevalence or impact. Highly active antiretroviral therapy (HAART) may be protective against10.Lucas G.M. Eustace J.A. Sozio S. et al.Highly active antiretroviral therapy and the incidence of HIV-1-associated nephropathy: a 12-year cohort study.AIDS. 2004; 18: 541-546Crossref PubMed Scopus (212) Google Scholar and therapeutic for11.Cosgrove C.J. bu-Alfa A.K. Perazella M.A. Observations on HIV-associated renal disease in the era of highly active antiretroviral therapy.Am J Med Sci. 2002; 323: 102-106Abstract Full Text Full Text PDF PubMed Scopus (51) Google Scholar HIV-associated renal disease. However, renal dysfunction may complicate antiretroviral treatment because some medications require dose adjustments. Certain antiretrovirals, such as tenofovir12.Gallant J.E. Parish M.A. Keruly J.C. et al.Changes in renal function associated with tenofovir disoproxil fumarate treatment, compared with nucleoside reverse-transcriptase inhibitor treatment.Clin Infect Dis. 2005; 40: 1194-1198Crossref PubMed Scopus (262) Google Scholar,13.Zimmermann A.E. Pizzoferrato T. Bedford J. et al.Tenofovir-associated acute and chronic kidney disease: a case of multiple drug interactions.Clin Infect Dis. 2006; 42: 283-290Crossref PubMed Scopus (260) Google Scholar and indinavir,14.Brodie S.B. Keller M.J. Ewenstein B.M. et al.Variation in incidence of indinavir-associated nephrolithiasis among HIV-positive patients.AIDS. 1998; 12: 2433-2437Crossref PubMed Scopus (45) Google Scholar can also cause renal dysfunction. To better understand changes in renal function associated with HAART in resource-limited settings where laboratory-based monitoring may not be routinely available, we evaluated data from persons participating in the Home-Based AIDS Care (HBAC) project in rural Uganda. We evaluated 508 participants who had been followed for at least 24 months after HAART initiation and had creatinine results at baseline and at their 12-month and 24-month follow-up visits. Participants were initiated on stavudine plus lamivudine with either nevirapine (496 participants, 98%) or efavirenz (12 participants, 2%). The study population (n=508) included 301 (59%) women and had a median age of 39 years, CD4 cell count of 122 cells per mm3, HIV viral load of 244 500 copies per ml, weight of 54 kg, and body mass index (BMI) of 19.6 kg/m2 at baseline. The baseline median creatinine was 97 μmol/l (1.10 mg per 100 ml), and 43 (8%) of 508 participants had a creatinine of ≥133 μmol/l (1.50 mg per 100 ml). The baseline median creatinine clearance (CLcr) by the Cockcroft–Gault equation was 63 ml/min, and 101 (20%) of 505 participants had renal dysfunction (CLcr 25–50 ml/min). Baseline CLcr could not be calculated for three participants because their baseline weight was not recorded. Characteristics associated with renal dysfunction at baseline in multivariate analysis included female gender, age ≥35 years, World Health Organization (WHO) stage 4 disease, baseline weight ≤60 kg, and a baseline BMI <18 kg/m2 (P<0.05). Renal dysfunction at baseline, however, was not associated with the baseline viral load (P=0.12) or CD4 cell count (P=0.74). Among the 508 participants who survived and received 24 months of HAART, median creatinine decreased by 16 μmol/l (16% decline, P<0.0001) and median CLcr increased by 13 ml/min (21% increase, P<0.0001) (Table 1). The proportion of participants with renal dysfunction decreased from 20% at baseline to 6% by month 24 (P<0.0001). Similar improvements in renal function were observed when the simplified modification of diet in renal disease (MDRD) equation was used to estimate glomerular filtration rate (Table 1). Participants with renal dysfunction at baseline (n=101) had the most dramatic improvements in CLcr. Their median CLcr increased by 23 ml/min (53% increase, P 50 ml/min) over 24 months on HAART (Figure 1). Among HBAC participants with diarrheal disease (a surrogate for dehydration) at baseline (n=35), there was no difference in CLcr improvements on HAART compared with participants without diarrhea. Among participants who achieved viral load suppression (viral load of 50 ml/min at baseline, the CLcr of 19 (5%) participants declined to ≤50 ml/min after 24 months on HAART. These participants experienced a median CLcr decline of 13 ml/min, although 18 of 19 had achieved viral load suppression (<400 copies per ml) on HAART. At 24 months, no participant had a CLcr of less than 25 ml/min and the lowest CLcr was 37 ml/min.Table 1Renal function during 24 months on HAART (n=508)Month 0Month 12Month 24Median creatinine μmol/l978881†P<0.0001 compared with months 0 and 12, Wilcoxon signed-rank test.Interquartile range (IQR) μmol/l(80–107)(75–103)(72–95)Creatinine ≥133 μmol/l (1.5 mg per 100 ml), n (%)43 (8%)17 (3%)13 (3%)‡P=0.0002 compared with month 0 and P=0.4 compared with month 12, McNemar's test.Median CLcr (ml/min)aThree participants did not have baseline weights.637376§P<0.0001 compared with month 0 and P=0.002 compared with month 12, Wilcoxon signed-rank test.IQR (ml/min)(53–77)(61–87)(65–91)CLcr ≤50 ml/minaThree participants did not have baseline weights., n (%)101 (20%)38 (8%)32 (6%)P<0.0001 compared with month 0 and P=0.5 compared with month 12, McNemar's test.CLcr ≤60 ml/minaThree participants did not have baseline weights., n (%)212 (42%)116 (23%)91 (18%)¶P<0.0001 compared with month 0 and P=0.02 compared with month 12, McNemar's test.Median MDRD (ml/min per 1.73 m2)839096**P<0.0001 compared with months 0 and 12, Wilcoxon signed-rank test. The correlation coefficient for CLcr versus MDRD was 0.68 at month 0, 0.73 at month 12, and 0.61 at month 24.IQR (ml/min per 1.73 m2)(69–98)(76–108)(82–114)MDRD ≤60 ml/min per 1.73 m2, n (%)59 (12%)30 (6%)18 (4%)P<0.0001 compared with month 0 and P=0.045 compared with month 12, McNemar's test.Median weight (kg)aThree participants did not have baseline weights.545757¶P<0.0001 compared with month 0 and P=0.02 compared with month 12, McNemar's test.IQR (kg)(48–60)(52–63)(52–63)CLcr, creatinine clearance; HAART, highly active antiretroviral therapy; IQR, interquartile range; MDRD, modification of diet in renal disease.† P<0.0001 compared with months 0 and 12, Wilcoxon signed-rank test.‡ P=0.0002 compared with month 0 and P=0.4 compared with month 12, McNemar's test.§ P<0.0001 compared with month 0 and P=0.002 compared with month 12, Wilcoxon signed-rank test.∣∣ P<0.0001 compared with month 0 and P=0.5 compared with month 12, McNemar's test.¶ P<0.0001 compared with month 0 and P=0.02 compared with month 12, McNemar's test.** P<0.0001 compared with months 0 and 12, Wilcoxon signed-rank test. The correlation coefficient for CLcr versus MDRD was 0.68 at month 0, 0.73 at month 12, and 0.61 at month 24.†† P<0.0001 compared with month 0 and P=0.045 compared with month 12, McNemar's test.a Three participants did not have baseline weights. Open table in a new tab Table 2MultivariableaWe used a linear regression model and initially included all covariates that in univariate analyses had P 0.10. An F-test with 7 degrees of freedom using the Type III sum of squares of the five dropped variables from our saturated model was non-significant (P=0.84). The six dropped variables from the multivariable analysis were viral load suppression at one year, baseline viral load, weight, BMI, LFT results, and diarrheal disease. analysis of factors associated with improved CLcr over 24 months on HAART (N=501bSeven participants were excluded for incomplete data.)n (%)Adjusted mean change CLcr, ml/min (95% CI)Adjusted P-valueCreatinine at baseline ≥133 μmol/l (1.5 mg per 100 ml)43 (9%)30 (24–36)<0.0001 < 133 μmol/l (1.5 mg per 100 ml)458 (91%)11 (9–13)Weight gain on HAART ≥5 kg149 (30%)20 (17–24)<0.0001 <5 kg352 (70%)9 (7–12)Gender Female296 (59%)16 (13–18)<0.0001 Male205 (41%)8 (5–11)WHO stage at baseline Stage 441 (8%)20 (13–26)0.02 Stages 1–3460 (92%)12 (10–14)Age (years) <35156 (31%)15 (12–18)0.09 ≥35345 (69%)12 (9–14)CI, confidence interval; CLcr, creatinine clearance; HAART, highly active antiretroviral therapy; IQR, interquartile range; WHO, World Health Organization.a We used a linear regression model and initially included all covariates that in univariate analyses had P 0.10. An F-test with 7 degrees of freedom using the Type III sum of squares of the five dropped variables from our saturated model was non-significant (P=0.84). The six dropped variables from the multivariable analysis were viral load suppression at one year, baseline viral load, weight, BMI, LFT results, and diarrheal disease.b Seven participants were excluded for incomplete data. Open table in a new tab CLcr, creatinine clearance; HAART, highly active antiretroviral therapy; IQR, interquartile range; MDRD, modification of diet in renal disease. CI, confidence interval; CLcr, creatinine clearance; HAART, highly active antiretroviral therapy; IQR, interquartile range; WHO, World Health Organization. As a separate analysis, we compared the baseline renal function of the 72 HBAC participants who died within the first 12 months of starting HAART with the 508 renal analysis participants who received HAART and survived 24 months. No deaths in HBAC were known to be due to renal failure, although a definite cause of death could not always be determined. HBAC participants who died within the first 12 months of starting HAART had a baseline median CLcr of 63 ml/min (P=0.35 compared with renal analysis participants who also had a baseline median CLcr of 63 ml/min) and 19 (29%) deceased participants had baseline renal dysfunction (P=0.09 compared with renal analysis participants of whom 20% had baseline renal dysfunction). Renal dysfunction was common in this population of rural Ugandans with advanced HIV disease, but generally improved during 2 years on HAART. The greatest improvements occurred in patients with renal dysfunction at baseline. Our results suggest that the inability to monitor renal function should not be a contraindication to providing first-line HAART and support current WHO guidelines, which do not recommend monitoring renal function with first-line HAART in resource-limited settings. The pre-HAART prevalence of renal dysfunction in this setting was greater than rates observed in several US cohort studies15.Gardner L.I. Holmberg S.D. Williamson J.M. et al.Development of proteinuria or elevated serum creatinine and mortality in HIV-infected women.J Acquir Immune Defic Syndr. 2003; 32: 203-209Crossref PubMed Scopus (138) Google Scholar,16.Szczech L.A. Gange S.J. van der H.C. et al.Predictors of proteinuria and renal failure among women with HIV infection.Kidney Int. 2002; 61: 195-202Abstract Full Text Full Text PDF PubMed Scopus (178) Google Scholar and a cross-sectional study from Kenya.7.Wools-Kaloustian K. Gupta S.K. Muloma E. et al.Renal disease in an antiretroviral-naive HIV-infected outpatient population in Western Kenya.Nephrol Dial Transplant. 2007; 22: 2208-2212Crossref PubMed Scopus (78) Google Scholar Two factors likely explain this discrepancy. Renal disease, such as HIV-associated nephropathy (HIVAN), typically worsens with advanced HIV disease,17.Winston J.A. Klotman M.E. Klotman P.E. HIV-associated nephropathy is a late, not early, manifestation of HIV-1 infection.Kidney Int. 1999; 55: 1036-1040Abstract Full Text Full Text PDF PubMed Scopus (152) Google Scholar and the participants in this clinical trial uniformly had advanced HIV disease at baseline. Additionally, in the US conditions such as HIVAN occur primarily in individuals of African descent,18.Shahinian V. Rajaraman S. Borucki M. et al.Prevalence of HIV-associated nephropathy in autopsies of HIV-infected patients.Am J Kidney Dis. 2000; 35: 884-888Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar suggesting that Africans may also be at greater risk for this complication. In resource-limited settings, the ability to monitor renal function is often limited. If, as our data suggest, renal function will improve in most patients with advanced HIV disease who initiate HAART, a rational approach would be to use standard weight-based dosing without adjustment for renal function at the initiation of antiretroviral therapy, even for patients in whom renal dysfunction can be identified. Such a strategy avoids sub-optimal HAART dosing, which could result in virologic treatment failure. Where monitoring is available, patients with renal dysfunction at HAART initiation may warrant repeat renal function testing during the first few months of treatment. If renal function does not improve despite HAART, then antiretroviral dose adjustments based on renal function should be considered. As participants did not receive tenofovir or indinavir as part of their initial HAART regimens, this suggested strategy cannot be extrapolated to tenofovir- or indinavir-based regimens without further evaluation. Although renal function improved in the majority of participants in this study, 5% of participants had a significant decline in renal function despite achieving viral load suppression on HAART. This decline in renal function for a minority of participants may reflect HIV-associated renal disease that does not improve with HAART or other renal pathologies that would not be expected to respond to HAART.19.Szczech L.A. Gupta S.K. Habash R. et al.The clinical epidemiology and course of the spectrum of renal diseases associated with HIV infection.Kidney Int. 2004; 66: 1145-1152Abstract Full Text Full Text PDF PubMed Scopus (271) Google Scholar There were several limitations in this study. First, the underlying etiology of renal disease was unknown in study participants, and urine specimens were not available to evaluate for proteinuria. Rates of diabetes and hypertension, although believed to be low in rural Uganda, were also not defined in this study. Second, we do not have adequate data to assess how HAART improved renal function. Some improvement in renal function may have occurred from an overall improvement in general health. For instance, chronic HIV-associated diarrhea in persons with advanced disease can cause chronic dehydration, which could rapidly improve with HAART. Notably, however, in our study reported diarrhea (our best surrogate for dehydration with advanced HIV disease) did not correlate with improvements in renal function on HAART. HAART may also have had a more direct effect on renal function. Several studies have suggested that HAART can slow or reverse the progression of HIVAN and other HIV-related renal diseases,10.Lucas G.M. Eustace J.A. Sozio S. et al.Highly active antiretroviral therapy and the incidence of HIV-1-associated nephropathy: a 12-year cohort study.AIDS. 2004; 18: 541-546Crossref PubMed Scopus (212) Google Scholar,20.Atta M.G. Gallant J.E. Rahman M.H. et al.Antiretroviral therapy in the treatment of HIV-associated nephropathy.Nephrol Dial Transplant. 2006; 21: 2809-2813Crossref PubMed Scopus (117) Google Scholar,21.Kirchner J.T. Resolution of renal failure after initiation of HAART: 3 cases and a discussion of the literature.AIDS Read. 2002; 12: 103-1012PubMed Google Scholar,22.Wali R.K. Drachenberg C.I. Papadimitriou J.C. et al.HIV-1-associated nephropathy and response to highly-active antiretroviral therapy.Lancet. 1998; 352: 783-784Abstract Full Text Full Text PDF PubMed Scopus (158) Google Scholar,23.Reid A. Stohr W. Walker S. et al.Glomerular dysfunction and associated risk factors following initiation of ART in adults with HIV infection in Africa (abstract THAB0105).in: Program and Abstracts of the 16th International AIDS Conference (Toronto, Canada) International AIDS Society, Toronto, Canada2006Google Scholar especially in the context of advanced HIV infection and impaired baseline renal function.9.Kalayjian R.C. Franceschini N. Gupta S.K. et al.Suppression of HIV-1 replication by antiretroviral therapy improves renal function in persons with low CD4 cell counts and chronic kidney disease.AIDS. 2008; 22: 481-487Crossref PubMed Scopus (99) Google Scholar In our study, renal function continued to improve from month 12 to month 24 on HAART, suggesting the gradual improvement of a chronic renal condition. Third, the accuracy of the Cockcroft–Gault equation to estimate renal function is unknown in the context of advanced HIV disease in Africa, especially with AIDS-associated wasting and cachexia.24.Heald A. Yuen G. Mydlow P. et al.Optimal estimation of glomerular filtration rate in patients infected with human immunodeficiency virus (abstract no. 114).in: the 34th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC) Orlando: Florida1994Google Scholar,25.Kotler D.P. Glyptis A. Grunfeld C. et al.Systematic errors in estimating renal function by Cockcroft–Gault (CG) or modification of diet in renal disease (MDRD) equations (abstract 18).in: the 8th International Workshop on Adverse Drug Reactions and Lipodystrophy in HIV San Francisco, California2006Google Scholar In addition, lower rates of renal dysfunction were identified by the MDRD, which has also been demonstrated among HIV-infected patients in Western Kenya.7.Wools-Kaloustian K. Gupta S.K. Muloma E. et al.Renal disease in an antiretroviral-naive HIV-infected outpatient population in Western Kenya.Nephrol Dial Transplant. 2007; 22: 2208-2212Crossref PubMed Scopus (78) Google Scholar Data, however, indicate that the Cockcroft–Gault equation is more likely to overestimate renal function in adults with advanced HIV disease,24.Heald A. Yuen G. Mydlow P. et al.Optimal estimation of glomerular filtration rate in patients infected with human immunodeficiency virus (abstract no. 114).in: the 34th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC) Orlando: Florida1994Google Scholar,25.Kotler D.P. Glyptis A. Grunfeld C. et al.Systematic errors in estimating renal function by Cockcroft–Gault (CG) or modification of diet in renal disease (MDRD) equations (abstract 18).in: the 8th International Workshop on Adverse Drug Reactions and Lipodystrophy in HIV San Francisco, California2006Google Scholar which suggests that we may have underestimated the improvements in renal function. Fourth, our data are subject to survivor bias, as we excluded participants who died within the first 12 months of starting HAART and changes in renal function prior to death in these participants were unknown. Finally, participants with severe renal dysfunction (CLcr <25 ml/min) at baseline were also excluded from this study. Although less than 1% of participants (n=7) were excluded from HBAC for this indication, we cannot generalize our observations to HIV-infected persons with more severe renal dysfunction. Although renal dysfunction may be an important complication of advanced HIV infection in Africa, among adults with advanced HIV disease in rural Uganda renal function improved over 2 years on HAART. Our results suggest that monitoring renal function with first-line HAART may not be mandatory. Identifying people infected with HIV at earlier stages of disease and initiating HAART before severe complications of renal disease develop may reduce the impact of HIV-associated renal disease in Africa. HBAC is a randomized clinical trial (registered at http://www.clinicaltrials.gov; identifier: NCT00119093) of three different monitoring strategies for HIV-infected individuals receiving HAART in Tororo and Busia districts of rural Eastern Uganda as previously described.26.Weidle P.J. Wamai N. Solberg P. et al.Adherence to antiretroviral therapy in a home-based AIDS care programme in rural Uganda.Lancet. 2006; 368: 1587-1594Abstract Full Text Full Text PDF PubMed Scopus (181) Google Scholar HBAC is a collaboration of the Ugandan Ministry of Health, The AIDS Support Organization (TASO), the Tororo and Busia District Health Departments, and the US Centers for Disease Control and Prevention (CDC). Inclusion criteria required participants be 18 years or older and have either a CD4-cell count of ≤250 cells per mm3 or symptomatic HIV disease (WHO stage 3 or 4 disease or recurrent herpes zoster; isolated pulmonary tuberculosis was not an inclusion criterion). Participants with baseline CLcr <25 ml/min (severe renal dysfunction) were excluded from the study. Of 1171 HIV-infected adults eligible for HAART by clinical and CD4 cell count criteria, seven (0.6%) were excluded from the randomized trial because of a CLcr of <25 ml/min. All HBAC participants received standard weight-based dosages of HAART, but medications were not adjusted for renal function. Overall adherence to antiretrovirals was high and 96% of participants had a viral load of 250 cells per mm3 (n=19). As we wanted to evaluate changes in renal function on HAART, we limited our primary analysis to the 508 participants who had been followed for at least 24 months after HAART initiation and had creatinine results at baseline and at their 12- and 24-month follow-up visits. The 96 participants excluded from the renal analysis had either died during the first 12 months after starting HAART (n=72) or were lost to follow-up (n=24). Overall, the 508 renal study participants were significantly more likely to be male, older, have a higher BMI, and have a baseline LFT elevation compared with the entire HBAC cohort. There was no difference, however, in baseline CD4 cell count or viral load. All analyses in this manuscript were repeated after excluding the additional participants (LFT elevation at baseline or a baseline CD4 cell count of >250 cells per mm3) and we obtained similar results (data not shown). We elected, therefore, to include these additional participants in the primary renal analysis, as they had a similar proportion of participants with renal dysfunction (defined as a CLcr of 25–50 ml/min) at baseline as the entire HBAC cohort. As this selection process excludes any participant who has died in follow-up, in a separate analysis we also compared the baseline renal function of the 72 HBAC participants who died within the first 12 months of starting HAART with the 508 renal analysis participants who survived 24 months on HAART. CLcr, an estimate of the glomerular filtration rate, was calculated using the Cockcroft–Gault equation28.Cockcroft D.W. Gault M.H. Prediction of creatinine clearance from serum creatinine.Nephron. 1976; 16: 31-41Crossref PubMed Scopus (12500) Google Scholar at each time point. Glomerular filtration rate was also estimated using the simplified MDRD equation.29.Levey A.S. Coresh J. Greene T. et al.Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate.Ann Intern Med. 2006; 145: 247-254Crossref PubMed Scopus (3697) Google Scholar CLcr could not be calculated for three participants who lacked baseline weight measurements. Renal dysfunction was defined as a CLcr of 25–50 ml/min. We dichotomized renal function at 50 ml/min because HIV treatment guidelines recommend dosage modifications for several antiretrovirals with CLcr of ≤50 ml/min.1.Gupta S.K. Eustace J.A. Winston J.A. 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: 1559-1585Crossref PubMed Scopus (501) Google Scholar,30.Panel on Antiretroviral Guidelines for Adults and Adolescents Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents.in: Department of Health and Human Services. 29 January 2008: 1-128http://www.aidsinfo.nih.gov/ContentFiles/AdultandAdolescentGL.pdfGoogle Scholar We repeated the analysis with renal function dichotomized at 60 ml/min and obtained similar results (data not shown). Serum specimens were drawn at participants' homes by field phlebotomists and transported to the CDC laboratory at the Uganda Virus Research Institute in Entebbe for processing, testing, and storage. Serum concentrations of creatinine were assessed using the Jaffe reaction (Randox CREA CR510, Randox Laboratories Ltd, Crumlin, UK). The National Institutes of Health Division of AIDS (DAIDS) table for grading the severity of adult and pediatric adverse events classifies a creatinine greater than or equal to 1.4 times the upper limit of normal (ULN; in our lab ULN=97 μmol/l) as moderately abnormal,31.NIH Division of AIDS NIH Division of AIDS 2004. Available at http://rcc.tech-res.com/tox_tables.htm Accessed 20 September 2007Google Scholar which corresponded to a serum creatinine of ≥133 μmol/l (1.5 mg per 100 ml). Baseline diarrheal disease was used as a surrogate for baseline dehydration in this cohort. Differences in serum creatinine, CLcr, and change in CLcr over 24 months on HAART between comparison groups were tested by the Wilcoxon rank sum test (continuous variables), the Wilcoxon signed-rank test (continuous, matched variables), the χ2 test (dichotomous variables), and McNemar's test (matched, nominal variables). P-values <0.05 were considered statistically significant. We used a multivariable linear regression model to identify variables independently associated with changes in CLcr from baseline to 24 months. The model was created using general linear modeling (SAS Proc GLM) with calculation of the least-squares means (LSMEANS). LSMEANS were also weighted with the observed marginals option so that changes in coefficients were proportional to the input dataset. Dichotomous variables in univariate analysis that were associated (P 0.10 to create a parsimonious model and manually reviewed the results after each step to ensure that our parsimonious model did not significantly differ from our original saturated model. We also performed an F-test using the Type III sum of squares of the dropped variables from our saturated model to verify that the dropped variables were not important to the model. We also used logistic regression to identify variables independently associated with baseline renal dysfunction. All variables in univariate analysis that were associated (P<0.20) with renal dysfunction at baseline were included in that model. Analyses were performed using SAS version 9.1 (SAS Institute Inc., Cary, NC, USA). Written informed consent to participate in HBAC was obtained from each participant in English or one of the six local languages. The Uganda National Council of Science and Technology (FWA no. 00001293) and the Institutional Review Boards of the Uganda Virus Research Institute (IRB no. 00001693) and CDC (IRB no. 00000183–185) approved the main study (HBAC) and this renal analysis. We thank the volunteers, staff, and clients of HBAC and TASO; the staff of CDC-Uganda, the Ugandan Ministry of Health, the districts of Tororo and Busia, and the CDC Global AIDS Program Headquarters. Data presented in abstract form (no. N-236) at the 14th Conference on Retroviruses and Opportunistic Infections (Los Angeles, CA, USA), 25–28 February 2007. This study was supported by the US Centers for Disease Control and Prevention and the US Agency for International Development through the President's Emergency Plan for AIDS Relief.
Referência(s)