https://orcid.org/0000-0001-8901-7850
Abstract
It is unclear whether use of contemporary protease inhibitors pose a similar risk of chronic kidney disease (CKD) as use of older protease inhibitors.
Participants in the Data Collection on Adverse Events of Anti-HIV Drugs (D:A:D) study were followed up until the earliest occurrence of CKD, the last visit plus 6 months, or 1 February 2016. Adjusted Poisson regression was used to assess associations between CKD and the use of ritonavir-boosted atazanavir (ATV/r) or ritonavir-boosted darunavir (DRV/r).
The incidence of CKD (10.0/1000 person-years of follow-up; 95% confidence interval, 9.5–10.4/1000 person-years of follow-up) increased gradually with increasing exposure to ATV/r, but the relation was less clear for DRV/r. After adjustment, only exposure to ATV/r (adjusted incidence rate ratio, 1.4; 95% confidence interval, 1.2–1.6), but not exposure to DRV/r (1.0; .8–1.3), remained significantly associated with CKD.
While DRV/r use was not significantly associated with CKD an increasing incidence with longer ATV/r use was confirmed.
Prior studies, including analyses of the Data Collection on Adverse Events of Anti-HIV Drugs (D:A:D) study, have shown an association between incident chronic kidney disease (CKD) and longer cumulative exposure to several human immunodeficiency virus (HIV) protease inhibitors (PIs), including indinavir (IDV), ritonavir-boosted atazanavir (ATV/r), and ritonavir-boosted lopinavir (LPV/r) [1, 2]. The association between ritonavir-boosted PI use and CKD may be explained by the increased propensity of these drugs to cause crystalluria, urolithiasis, and interstitial nephritis [3–5].
Ritonavir-boosted darunavir (DRV/r) was widely implemented as part of routine clinical care for HIV in Europe from 2009 onward, with steadily increasing use owing to its efficacy and high genetic barrier to resistance. In contrast to several of the older PIs, only a very limited number of case reports have linked use of DRV/r with development of urolithiasis [6–8]. A recent switch study from the United Kingdom further suggested that DRV/r use may even exert a positive effect on estimated glomerular filtration rate (eGFR) trajectories compared with other PIs [7].
CKD is increasingly common among people living with HIV (PLWH), with a wide spectrum of potential underlying risk factors, and it is associated with considerable morbidity and mortality risks. Improved insights into both primary and secondary preventive measures are therefore urgently required [9]. The aim of the current analysis was to assess whether cumulative use of more contemporary PIs, including DRV/r and ATV/r, are associated with an increased incidence of CKD to a similar extent as some of the older PIs.
METHODS
The D:A:D study is a large cohort collaboration established in 1999 with >49 000 HIV-1–positive persons undergoing prospective follow-up in Europe, Australia, and the United States; details have been published elsewhere [10]. Data on demographics, CD4 cell count, HIV RNA and other laboratory measurements, antiretroviral therapy (ART), cardiovascular risk factors, and AIDS events are collected electronically at the time of enrollment and every 6 months thereafter. In addition, clinical events, including end-stage renal disease, myocardial infarction, stroke, invasive cardiovascular procedures, and death, are reported during routine clinical care, validated centrally, and regularly monitored. All participating cohorts followed local national guidelines and regulations regarding patient consent and/or ethical review.
CKD was defined as confirmed (measured >3 months apart) eGFR ≤60 mL/min/1.73 m2 [11, 12]. As in prior D:A:D renal analyses, the Cockcroft-Gault equation, standardized for body surface area, was used to estimate creatinine clearance and as a surrogate for eGFR [13]. Because several participating cohorts are prohibited by law from collecting information on ethnicity, the Cockcroft-Gault equation was used rather than an equation including ethnicity. This equation also has the advantage of weight adjustment, which is relevant in a population with lipohypertrophy and lipoatrophy.
Study participants with ≥3 eGFR measurements (1 at or before baseline and 2 after), ≥3 months of follow-up, baseline eGFR >60 mL/min/1.73 m2, and data on CD4 cell count and HIV viral load (VL) at baseline were included in the analysis. The study baseline was defined as 1 January 2009, reflecting the broader licensing of DRV/r in Europe. Participants were followed up until the earliest occurrence of CKD, the last visit plus 6 months, or 1 February 2016.
Poisson regression was used to model the association between CKD and cumulative use of the presently most commonly used PIs, DRV/r and ATV/r, while adjusting for demographic variables (ie, sex, age, and cohort), other ART that may affect renal function (ie, tenofovir disoproxil fumarate, TDF), traditional renal risk factors (ie, hypertension, diabetes, baseline eGFR, and cardiovascular disease) and HIV-related risk factors (ie, CD4 cell count, viral hepatitis coinfection, and prior AIDS).
Based on earlier D:A:D renal analyses, the association between CKD and longer ATV/r use is expected to be gradual and could therefore be reasonably fitted as a continuous variable. However, because such a relation may not exist for DRV/r in this analysis, ART exposure was fitted categorically [11]. Variables not changing over time were fitted as time-fixed (baseline) values, whereas variables that changed during follow-up (ie, use of ART and CD4 cell count) were fitted as time-updated values. A separate Poisson regression model assessed adjusted associations of switching away from DRV/r and ATV/r with declining eGFRs. All statistical analyses were carried out using SAS software, version 9.3.
RESULTS
Of the 36 283 persons in D:A:D with prospective follow-up after 1 January 2009, 8494 were excluded from the analysis because of a baseline eGFR ≤60 mL/min/1.73 m2, <2 eGFR measurements after baseline, or <3 months follow-up. Another 114 persons were excluded owing to missing baseline CD4 cell count and/or VL data. Compared with the 27 675 persons included in the analysis, those excluded were less likely to be on ART and have an undetectable VL and more likely to have lower CD4 cell counts and be older, of white origin, and hepatitis C virus positive.
The median age (interquartile range) at baseline was 44 (38–50) years; the median eGFR, 101 (87–117) mL/min/1.73 m2; and the median CD4 cell count, 510/μL (340–699/μL); 80.1% of participants had a VL <400 copies/mL, and 28.7%, 35.6%, and 35.7%, respectively, were at low, medium, and high 5-year CKD risk, as estimated by the D:A:D CKD risk score [12]. Most participants were male (73.8%), of white origin (45.8%), and men who have sex with men (47.0%). Of the total follow-up time (164 983 person-years of follow-up [PYFU]), 14.4% and 25.0% were accrued after DRV/r and ATV/r initiation, respectively (Supplementary Table 1).
CKD developed in a total of 1642 persons (5.9%) (incidence rate [IR], 10.0/1000 PYFU; 95% confidence interval [CI], 9.5–10.4/1000 PYFU) during a median follow-up (interquartile range) of 6.8 (5.4–7.1) years. The crude IR of CKD was 9.3/1000 PYFU (95% CI, 8.8–9.8/1000 PYFU) in persons unexposed to DRV/r and 8.7/1000 PYFU (8.1–9.2/1000 PYFU) in those unexposed to ATV/r. There was a consistently increasing IR of CKD with increasing exposure to ATV/r, and while there was some increase in CKD IR with increasing exposure to DRV/r, the IR for DRV/r was more variable (Figure 1).
Crude incidence rates of chronic kidney disease (CKD) per 1000 person-years of follow-up (PYFU), with 95% confidence intervals (CIs), stratified by cumulative use of ritonavir-boosted atazanavir (ATV/r) and ritonavir-boosted darunavir (DRV/r).
After adjustment for potential confounding factors, only cumulative exposure to ATV/r (adjusted IR ratio, 1.4; 95% CI, 1.2–1.6), but not exposure to DRV/r (1.0; 0.8–1.3) (both after >4 years of use vs no exposure) remained significantly associated with increased incidence of CKD (Figure 2). These associations remained similar when the analysis was restricted to individuals with a baseline eGFR >90 mL/min/1.73 m2 (data not shown).
Multivariate relationship between chronic kidney disease (CKD) and cumulative exposure to ritonavir-boosted atazanavir (ATV/r) and ritonavir-boosted darunavir (DRV/r). Multivariate models were adjusted for sex, race, human immunodeficiency virus (HIV) exposure group, enrollment cohort, prior cardiovascular disease (CVD), age, CD4 cell count nadir, baseline date and estimated glomerular filtration rate (all fixed at baseline), HIV load, current CD4 cell count, prior AIDS, hepatitis B virus, hepatitis C virus, diabetes, hypertension, dyslipidemia, smoking status, body mass index, family history of CVD, CVD, cancer, and cumulative exposure to tenofovir, atazanavir (unboosted), lopinavir, abacavir, tipranavir, and other ritonavir-boosted protease inhibitors (all time updated). Abbreviation: CI, confidence interval.
A total of 3580 persons discontinued ATV/r use during follow-up (IR, 183.6/1000 PYFU; 95% CI, 177.8–189.4/1000 PYFU). At eGFRs >90 mL/min/1.73 m2, discontinuation rates for ATV/r were 181.9/1000 PYFU (95% CI, 174.1–189.6/1000 PYFU), gradually increasing with declining eGFRs to 318.2/1000 PYFU (255.2–381.2/1000 PYFU) at eGFRs <60 mL/min/1.73 m2. The adjusted IR ratio for discontinuing ATV/r use during follow-up was 80% higher at a current eGFR ≤60 mL/min/1.73 m2 than at a current eGFR >90 mL/min/1.73 m2 (1.8; 95% CI, 1.4–2.1).
For DRV/r, 2084 persons discontinued use during follow-up (IR, 111.6/1000 PYFU; 95% CI, 106.8–116.4/1000 PYFU), with rates of 113.3/1000 PYFU (106.8–119.5/1000 PYFU) and 138.4/1000 PYFU (95.5–181.3/1000 PYFU) at eGFRs >90 and ≤60 mL/min/1.73 m2, respectively. In contrast to ATV/r, discontinuation of DRV/r use was largely unaffected by declining eGFRs (adjusted IR ratio for eGFRs ≤60 vs >90 mL/min/1.73 m2, 1.2; 95% CI, .9–1.7). The ATV/r discontinuations also increased in those at high estimated risk of CKD, from 58% of these ATV/r discontinuations in 2009 to 65% in 2015 (P = .0033).
DISCUSSION
The current study is the first to systematically investigate associations between incident CKD and longer cumulative use of the contemporary PIs DRV/r and ATV/r. Although prior studies have linked use of the older PIs IDV and LPV/r and the more contemporary ATV/r with CKD at increased rates between 11% and 20% per additional year of use, controversies have existed about DRV/r and about a possible PI class effect on CKD risk [7, 11, 12]. With a median follow-up of >6 years, we were unable to find a statistically significant, gradual or equally strong association, compared with several other PIs, between more extended use of DRV/r and CKD.
In contrast, the year-on-year risk previously observed between longer ATV/r use and CKD remained, with a 40% increased incidence of CKD after 4 years use compared with no use in fully adjusted analyses, including adjustment for concomitant use of TDF. The strength of the ATV/r-associated CKD risk has decreased over time (initially reported to be up to 20% per additional year of use), which is likely explained by the increased general awareness of the nephrotoxic potential of ATV/r use and the subsequent high rates of switching away from ATV/r in PLWH with high predicted CKD risks and/or declining eGFRs [1, 7, 12]. In contrast, rates of DRV/r discontinuations were unrelated to eGFR level, and the lack of an association with CKD is therefore unlikely to be explained by channeling [11].
The D:A:D study does not collect information on drug dosages, and we are therefore unable to address whether CKD risks may differ according to DRV/r dosage, although our findings were unchanged by adjustment for factors associated with increased dosing (ie, low CD4 cell count, viremia, and prior AIDS events), and the 2009 study baseline reflects a wider use of DRV/r with a mixture of dosing regimens. Although a cross-sectional study found that both contemporary PIs, ATV/r and DRV/r, may precipitate as crystals in urine in a small proportion of PLWH and therefore have a similar theoretical potential for inducing urolithiasis, both a UK and a Japanese study found that individuals taking ATV/r had significantly higher rates of urolithiasis than those taking DRV/r, with adjusted rates between 3.8 and 21.5, respectively [6, 8, 14].
Using CKD, a more rigorously defined clinical endpoint, for the first time and in a large-cohort setting, we report findings that support those of previous studies using eGFR slopes and rates of urolithiasis formation, suggesting there is no uniform or equally strong nephrotoxic effect of all contemporarily used PIs [7, 8]. Instead, there seems to be a gradually increasing risk of CKD related to use of specific PIs, such as ATV/r, even after adjustment for other potentially nephrotoxic antiretrovirals. This finding has direct clinical implications when considering ART drug choices for the increasing population of PLWH at increased risk of renal disease or with prevalent CKD.
As our analysis was limited by the 6.8 years median follow-up time, and as DRV/r may precipitate in urine, albeit relatively rarely, it is not possible to exclude the possibility that there is an association between DRV/r and CKD, but that this only emerges with very extended drug use. However, our data did not indicate a year-on-year increase in risk. Within the D:A:D study, there is only very limited follow-up among participants receiving cobicistat, so we could not analyze the impact of using an alternative PI boosting agent on CKD incidence, but this question is nevertheless relevant. Likewise, the D:A:D study does not systematically collect data on proteinuria or genetic predisposition to CKD, which may have modified the effects observed between PI use and CKD. The observations presented in this analysis are therefore conservative estimates.
In conclusion, in this large heterogeneous cohort of PLWH, more extended use of DRV/r was not significantly associated with a gradually increasing incidence of CKD, even after a median follow-up of >6 years. Discontinuation of DRV/r use was, in contrast to discontinuation of ATV/r, unrelated to declining eGFRs. A gradually increasing CKD risk with longer use of ATV/r was confirmed, with a 40% increased CKD incidence after 4 years of use, when compared with those never exposed to ATV/r.
Supplementary Data
Supplementary materials are available at The Journal of Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
References
Notes
D:A:D cohorts. The participating cohorts from the Data Collection on Adverse Events of Anti-HIV Drugs (D:A:D) study include the Australian HIV Observational Database (AHOD) (Australia), Aquitaine (France), AIDS Therapy Evaluation in the Netherlands (ATHENA) (the Netherlands), Barcelona Antiretroviral Surveillance Study (BASS) (Spain), Community Programs for Clinical Research on AIDS (CPCRA) (United States), EuroSIDA (multinational), HivBivus (Sweden), Italien Cohort of Antiretroviral Naive individuals (ICONA) (Italy), Nice (France), Swiss HIV Cohort Study (SHCS) (Switzerland), and St Pierre (Belgium). For a complete list of the members of the 11 participating cohorts, see Appendix 1 in the Supplementary Material.
Cohort principal investigators. W. E.-S. (member of D:A:D steering committee [SC]) (CPCRA), G. Calvo (SC) (BASS), F. Bonnet (SC) and F. Dabis (SC) (Aquitaine), O. K. (SC) and A. M. (SC) (EuroSIDA), M. L. (SC) (AHOD), A. d. M. (SC) (ICONA), L. Morfeldt (SC) (HivBivus), C. Pradier (SC) (Nice), P. R. (SC) (ATHENA), R. Weber (SC) (SHCS), S. D. W. (SC) (St Pierre).
Cohort coordinators and data managers. A. Lind-Thomsen (coordinator), R. Salbøl Brandt, M. Hillebreght, S. Zaheri, F. W. N. M. Wit (ATHENA), A. Scherrer, F. Schöni-Affolter, M. Rickenbach (SHCS), A. Tavelli, I. Fanti (ICONA), O. Leleux, J. Mourali, F. Le Marec, E. Boerg (Aquitaine), E. Thulin, A. Sundström (HivBivus), G. Bartsch, G. Thompsen (CPCRA), C. Necsoi, M. Delforge (St Pierre), E. F., C. Caissotti, K. Dollet (Nice), S. Mateu, F. Torres (BASS), K. Petoumenos, A. Blance, R. Huang, R. Puhr (AHOD), K. Grønborg Laut, D. Kristensen (EuroSIDA).
Statisticians. C. S. (SC), A. P. (SC), D. A. Kamara, C. J. Smith, A. M. (SC).
D:A:D coordinating office. C. I. H., L. R. (SC), A. Lind-Thomsen, R. S. Brandt, D. Raben, C. Matthews, A. Bojesen, A. L. Grevsen, J. D. L. (SC chair).
D:A:D oversight committee. B. Powderly (SC), N. Shortman (SC), C. Moecklinghoff (SC), G. Reilly (SC), X. Franquet (SC).
D:A:D working group experts. Kidney: L. R. (SC), A. M. (SC), O. K. (SC), P. R. (SC), C. Smit, M. R., C. A. F., P. M., E. F., D. A. Kamara, C. J. Smith, J. D. L. (SC chair). Mortality: C. J. Smith, L. R. (SC), C. I. H., A. P. (SC), R. Weber (SC), P. M., C. Pradier (SC), P. R. (SC), F. W. N. M. Wit, N. Friis-Møller, J. Kowalska, J. D. L. (SC chair). Cancer: C. S. (SC), L. R. (SC), C. I. H., M. L. (SC), A. d. M. (SC), F. Dabis (SC), F. Bonnet (SC), P. R. (SC), F. W. N. M. Wit, C. J. Smith, D. A. Kamara, J. Bohlius, M. Bower, G. Fätkenheuer, A. Grulich, J. D. L. (SC chair).
External end-point reviewers. A. Sjøl (cardiovascular disease), P. Meidahl (oncology), J. S. Iversen (nephrology).
Disclaimer. The content of this publication is solely the responsibility of the authors and does not necessarily represent the official views of any of the institutions mentioned above. No funding bodies had any role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Financial support. The D:A:D study was supported by the Highly Active Antiretroviral Therapy Oversight Committee, a collaborative committee with representation from academic institutions, the European Agency for the Evaluation of Medicinal Products, the US Food and Drug Administration, the patient community, and pharmaceutical companies with licensed anti-HIV drugs in the European Union: AbbVie, Bristol-Myers Squibb (BMS), Gilead Sciences, ViiV Healthcare, Merck, and Janssen Pharmaceuticals. It was also supported also by the Danish National Research Foundation (CHIP and PERSIMUNE) (grant DNRF126), the Dutch Ministry of Health, Welfare and Sport through the Center for Infectious Disease Control of the National Institute for Public Health and the Environment to Stiching HIV Monitoring (to ATHENA), and the Agence Nationale de Recherches sur le Sida et les Hépatites Virales (Action Coordonnée 7 [Cohortes] to the Aquitaine Cohort). AHOD is funded as part of the Asia Pacific HIV Observational Database, a program of The Foundation for AIDS Research, and is supported in part by the US National Institutes of Health’s National Institute of Allergy and Infectious Diseases (grant U01-AI069907) and by unconditional grants from Merck Sharp & Dohme, Gilead Sciences, BMS, Boehringer Ingelheim, Janssen-Cilag, and ViiV Healthcare. The Kirby Institute is funded by the Australian Government Department of Health and Ageing and is affiliated with the Faculty of Medicine, The University of New South Wales. BASS is supported by the Fondo de Investigación Sanitaria (grant FIS 99/0887) and Fundación Para la Investigación y la Prevención del SIDA en Espanã (grant FIPSE 3171/00). The CPCRA is supported by the National Institute of Allergy and Infectious Diseases, National Institutes of Health (grants 5U01AI042170-10 and 5U01AI046362-03). Primary funding is provided by the European Union’s Seventh Framework Programme for research, technological development, and demonstration (under EuroCoord grant 260694) and by unrestricted grants by BMS, Janssen Research & Development, Merck, Pfizer., and GlaxoSmithKline. The participation of centers from Switzerland is supported by the Swiss National Science Foundation (grant 108787 to the EuroSIDA study). The ICONA Foundation is supported by unrestricted educational grants from AbbVie, BMS, Gilead Sciences, GlaxoSmithKline, Pfizer, and Janssen Pharmaceuticals. The SHCS is supported by the Swiss National Science Foundation (grant 148522).
Potential conflicts of interest. P. R. has served as a scientific advisor to BMS, Gilead Sciences, Grupo Ferrer, GlaxoSmithKline, Janssen Pharmaceuticals, Merck, and ViiV Healthcare. He has served on data and safety monitoring boards and end-point adjudication committees for Janssen Pharmaceuticals and his institution has received honoraria for speaking engagements at scientific conferences from BMS, Gilead Sciences, GlaxoSmithKline. He has received research support from Gilead Sciences, ViiV Healthcare, Merck, Janssen Pharmaceuticals, BMS, Abbott, and Boehringer Ingelheim Pharmaceuticals. O. K. had prior/present board membership at ViiV Healthcare, Gilead Sciences and Merck, received payment for lectures and/or for development of educational presentations from Abbott, Gilead Sciences and Tibotec and had travel/accommodations/meeting expenses paid by Abbott, BMS, Gilead Sciences, Merck and ViiV Healthcare. M. L. has received research grants from Boehringer Ingelheim, BMS, Gilead, GlaxoSmithKline, Janssen-Cilag, Merck Sharp & Dohme, Pfizer and Roche. P. M. has received honorarium and support for travel to meeting from Gilead Sciences, ViiV Healthcare and Merck. C. A. F. is an advisory board member for Gilead Sciences and MSD, has pending grants from Gilead Sciences and Abbott and received payment for lectures by Gilead HIV and the body. A. d. M. has past board membership at AbbVie, BMS, Gilead Sciences, Janssen Pharmaceuticals and Merck. A. P. received personal fees from Gilead Sciences, AbbVie, GlaxoSmithKline Vaccines and grants from BMS. C. S. received personal fees from Gilead Sciences, BMS, Janssen Pharmaceuticals, Abbott Pharmaceuticals, and ViiV Healthcare. A. M. has received consultancy fees/honoraria/speaker fees from BMS, Pfizer, Merck, Boehringer Ingelheim, and Gilead Sciences. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
Presented in part: Conference on Retroviruses and Opportunistic Infections (CROI) Feb 2017, Seattle, Washington. Poster no. 653
Author notes
See Acknowledgments for D:A:D study group details.
