Early Safety of Tenofovir Alafenamide in Patients With a History of Tubulopathy on Tenofovir Disoproxil Fumarate

A Randomized Controlled Clinical Trial

L Hamzah; D Williams; AC Bailey; R Jones; F Ibrahim; CG Musso; K Burling; B Barbini; L Campbell; FA Post


HIV Medicine. 2020;21(3):198-203. 

In This Article


The Fanconi-Tenofovir Alafenamide study is a prospective, open-label, phase IV randomized controlled trial conducted at five HIV centres in the UK. HIV-1-positive patients aged > 16 years on stable ART for the past 6 months with viral load < 200 HIV-1 RNA copies/mL at their most recent assessment were eligible if they had a documented history of treatment-limiting PRT while receiving TDF, defined by at least two of the following: proteinuria (at least 1+ on urinary dipstick or urine protein:creatinine ratio of at least 30 mg/mmol); normoglycaemic glycosuria (at least 1+ on urinary dipstick); hypophosphatemia (serum phosphate < 0.64 mmol/L); rapid eGFR decline (> 5 mL/min/1.73m2/year with> 25% reduction from baseline);[5] or a renal biopsy showing acute tubular injury not explained by other causes.[7] Patients with diabetes mellitus, glycosuria (at least 1+ on dipstick), proteinuria (at least 2+ on dipstick) or urine protein:creatinine ratio (PCR) ≥ 100 mg/mmol or creatinine clearance < 30 mL/min (Cockcroft–Gault) at screening or baseline, and those on TDF- or TAF-containing ART were ineligible. All participants provided written informed consent. The trial was registered under European Union Drug Regulating Authorities Clinical Trials Database 2016–003345-29.

Patients were randomized to receive co-formulated emtricitabine (200 mg)/TAF [25 mg (10 mg with boosted regimens); F/TAF] (immediate arm) or to continue their current ART for a further 12 weeks (deferred arm); those randomized to F/TAF discontinued other nucleoside reverse transcriptase inhibitors (NRTIs) (e.g. lamivudine and abacavir) and any medications for hepatitis B at the time of randomization as clinically appropriate while the remaining ART continued unchanged. At 12 weeks, those randomized to remain on current ART were offered F/TAF, with further modification or simplification of ART regimens allowed at the discretion of the investigators. Plasma, serum and urine samples were obtained at baseline, week 4 and week 12 and stored at −70°C for analysis of renal and bone biomarkers in central laboratories.

The primary outcome was change in retinol-binding protein:creatinine ratio (RBPCR) from baseline to week 12. Secondary outcomes in this ongoing study are changes in other markers of kidney function [eGFR based on creatinine and cystatin C using the Chronic Kidney Disease-Epidemiology equation[16] albumin:creatinine ratio (ACR), PCR, ratio of maximum rate of renal tubular reabsorption of phosphate to glomerular filtration rate (TmPO4/GFR),[17] fractional excretion of urea (FE-urea), and (optional) early-morning urine osmolality], bone regulatory hormones [25-hydroxy vitamin D (25(OH)D) and parathyroid hormone (PTH)], markers of bone turnover [total procollagen type 1 N-terminal propeptide (P1NP), a marker of bone formation, and C-terminal telopeptide (CTx), a marker of bone resorption], bone mineral density and incidence of proximal tubulopathy up to 96 weeks. The current analyses focus on the primary endpoint and other biomarkers up to week 12.

Statistical Analysis

Based on our earlier study of TDF-associated PRT,[5] we estimated that we could enrol 40 individuals. Assuming a mean baseline RBPCR of 2.93 μg/mmol (the upper limit of the reference range),[18] a standard deviation (SD) of 1.1 μg/mmol and 20 participants in each arm, the trial had 87% power to detect a 50% difference in RBPCR from baseline to week 12 between the two study arms.

Baseline characteristics were summarized by randomization group as means and SDs (continuous normally distributed variables), medians and interquartile ranges (IQRs) (non-normally distributed variables), and frequencies and percentages (categorical variables).

For the primary outcome, the mean difference in RBPCR between the two study arms was analysed using linear regression adjusted for baseline RBPCR measurements, with robust standard errors. Assumptions were checked graphically. All participants had complete observations for the primary outcome. Secondary longitudinal outcomes, measured at baseline, week 4 and week 12, were analysed using repeated measures mixed effects models with an unstructured variance-covariance matrix to assess mean difference between arms. Non-normally distributed variables were transformed to approximate normality. Covariates such as age, ethnicity and sex were considered for inclusion in the model. All analyses were conducted using STATA (version 12; StataCorp LP, College Station, TX).