icon-folder.gif   Conference Reports for NATAP  
  HIV Glasgow
23-26 October 2016
Glasgow, UK
Back grey_arrow_rt.gif
Changes in kidney function in patients with suppressed HIV RNA who substitute their protease inhibitor with atazanavir/ritonavir
  Download the PDF here
AIDS 2015
Hamzah, Lisaa; Campbell, Lucya; Wandolo, Emilyb; Engler, Birgitb; Cheserem, Emilya; Naftalin, Clairea; Taylor, Chrisb; Post, Frank A.a,b aKing's College Hospital NHS Foundation Trust bKing's College London School of Medicine, London, UK.
Atazanavir (ATV) is a protease inhibitor licensed for the treatment of HIV infection, and one of the preferred agents for first-line therapy in the UK [1]. The use of ATV, however, has been associated with a number of kidney syndromes, including acute interstitial nephritis [2,3], chronic kidney disease (CKD) [4], rapid decline in estimated glomerular filtration rate (eGFR) [5], proximal kidney tubular dysfunction [6,7] and nephrolithiasis [8,9]. The risk factors for ATV-associated kidney dysfunction remain to be defined.
There is a paucity of data on renal biomarkers in patients switching from other ritonavir-boosted protease inhibitor (PI/r) to ATV/r. Most clinical trials focussed on lipid profiles and bilirubin and did not report data on renal function [10,11]. A change in the London HIV Consortium treatment guidelines in 2010 allowed us to study the effects of ATV/r on renal function of individuals with suppressed HIV viral loads switching protease inhibitor therapy. Patients deemed suitable for protease inhibitor switching were invited to consent to storage of fasted plasma, serum and urine samples at their routine clinic visits at baseline (preswitch), and 2-4 and 24-48 weeks postswitch. These samples were stored at -70•C until analysis for urinary total protein, albumin, retinol-binding protein (RBP) [12] and phosphate, and for serum cystatin C (CysC) concentrations. This observational study was approved by the BioBank Ethics Committee.
The current analyses focus on individuals who were on two nucleoside reverse transcriptase inhibitor (NRTI, with no subsequent change in NRTI) and a non-ATV ritonavir-boosted protease inhibitor for more than 6 months prior to switching protease inhibitor. Routinely collected serum creatinine data were used to calculate eGFR slopes (CKD-EPI-creatinine [13]) pre and postswitch (reflecting exposure to non-ATV PI/r and ATV/r, respectively). For those individuals with stored samples, estimated GFR (CKD-Epi-creatinine and CKD-EPI-CysC), tubular function [RBP expressed as a ratio with urinary creatinine (RBPCR)] and the ratio of the maximum rate of tubular phosphate reabsorption to eGFR (TmPO4/GFR) at 2-4 weeks and 24-48 weeks were compared with baseline. Preswitch and 2-4 and 24-48 weeks postswitch data were compared using the one-way repeated-measures analysis of variance, Friedman [14] or Skillings-Mack tests [15]. To determine whether switching to ATV/r had an effect on eGFR slope, pre and postswitch eGFR slopes were calculated and compared using a linear mixed effects model [16].
Overall, 159 individuals were identified as being on a non-ATV protease inhibitor containing regime at baseline. Twenty-eight (17.6%) were on lopinavir (LPV)/r monotherapy and were not switched. Of the remaining 131 individuals on triple therapy, 52 switched to ATV/r between November 2010 and April 2013. Reasons for not switching PI included current pregnancy, viral nonsuppression, age less than 18 years and patient unwillingness to change protease inhibitor regime. One patient also switched NRTI backbone at the time of protease inhibitor switch and three had HIV RNA levels more than 400 copies/ml; these were excluded from the analyses. Of the 48 remaining individuals, 47 had at least three creatinine measurements pre and postswitch on protease inhibitor therapy and were included in the eGFR slope analyses, and 30 provided samples for biomarker analysis. The mean age of the patients was 42.6 (SD 7.2) years, 44% were female and 73% were of black ethnicity. The median time since HIV diagnosis was 7.6 [interquartile range (IQR) 3.1-13.1] years, and the median CD4+ cell count was 594 (491-781) cells/μl. Thirty patients (63%) received tenofovir disoproxil fumarate (TDF) as part of their antiretroviral regimen; preswitch protease inhibitor exposure consisted of LPV/r (n = 35), fosamprenavir/r (n = 7), saquinavir/r (SQV/r, n = 4) and darunavir/r (DRV/r, n = 2). Patients included in the renal biomarker analyses had similar characteristics as those included in the eGFR slope analysis, with the exception that they were more likely to be men (56.7vs. 22.2%) who acquired HIV through sex with men (36.7 vs. 5.6%) and be on TDF vs. a non-TDF containing regime (80.0 vs. 33.3%, P < 0.01 for all).
The median (IQR) duration of follow up on PI/r-containing regimens was 3.8 (1.7-7.3) and 1.8 (1.5-2.1) years pre and postswitch, respectively, and the median number of eGFR measurements during this period was 33 (17-42) and 7.5 (6-9), respectively. The median (IQR) eGFR was 118 (100-126) at the start of PI/r therapy and 113 (90-126) ml/min/1.73 m2 at the time of ATV switch. The crude eGFR slopes [95% confidence interval (95% CI)] were -0.7 (-1.1 to -0.3) preswitch and -0.6 (-4.2 to 3.0) ml/min/1.73 m2/year postswitch. Using mixed effect models adjusted for age, ethnicity and CD4+ cell count, there was no significant difference between the pre and postswitch eGFR slope [1.1 (-1.5 to 3.6) ml/min/1.73 m2/year, P = 0.4].
In the subset of patients with stored samples, there was no statistically significant difference from baseline in eGFR (based on creatinine and cystatin C measurements), proteinuria (protein/creatinine ratio), albuminuria (albumin/creatinine ratio), RBPCR or TmPO4/GFR at either weeks 2-4 or weeks 24-48 postswitch (Table 1).
In this observational study of patients with well controlled HIV infection, switching from a non-ATV PI/r to ATV/r-containing antiretroviral regimen did not have a demonstrable effect on renal glomerular or tubular function. Our data provide reassurance regarding the use of ATV/r in protease inhibitor experienced patients with preserved renal function and are consistent with studies in treatment-naive patients in whom ATV/r was compared with LPV/r [17], SQV/r [18], DRV/r [19] or efavirenz [20]. By contrast, although large observational cohort studies such as EuroSIDA, Veterans Health Administration and D:A:D are able to provides risk estimates of rapid eGFR decline and/or kidney disease progression [4,5,21], these studies may be subject to channelling bias if ATV/r was preferentially prescribed to individuals at an increased risk of CKD.
Our study is limited by the relatively small sample size, duration of follow-up, lack of a control group and its observational design. It does, however, reflect real-life clinical practice. Our results endorse the London Consortium guidance to use ATV/r as the preferred protease inhibitor in people living with HIV who have preserved renal function. Our data argue against a widespread detrimental effect of ATV/r on the kidney but are unable to exclude the possibility of a direct nephrotoxic effect in a small subset of patients.