Estimated Glomerular Filtration Rate Slopes on Tenofovir Alafenamide

F Ibrahim; L Campbell; AC Bailey; S Stockwell; L Waters; C Orkin; M Johnson; M Gompels; A De Burgh-Thomas; R Jones; G Schembri; PW Mallon; FA Post

Disclosures

HIV Medicine. 2020;21(9):607-612.

Discussion

Our study demonstrates that eGFR slopes improved in individuals who switched from TDF- to TAF-containing ART, particularly in those who experienced rapid eGFR decline or CKD while receiving TDF. These data expand clinical trial experience which has shown that TAF is associated with stable renal function, even in those with mild/moderate CKD, and a more favourable renal biomarker profile than TDF.^[12,13] Moreover, TAF had been successfully used and was shown to have no significant effect on biomarkers of renal tubular function in individuals who experienced proximal tubulopathy/Fanconi syndrome while receiving TDF,^[16–20] suggesting that the renal safety of TAF extends to a broad population of people with HIV infection.

In this observational study, eGFR decline during TDF use improved, and in some cases resolved, with switching from TDF to TAF, suggesting that in many cases the observed eGFR decline may have been caused, or at least contributed to, by TDF. The somewhat greater eGFR decline during TDF exposure in those who discontinued TDF prior to TAF initiation as compared to those who directly switched from TDF to TAF probably reflects a greater proportion of renal discontinuations and/or more pronounced adverse effects of tenofovir on the kidney in the former group. In settings such as the UK, where TDF remains the tenofovir formulation of choice for people without established CKD or bone fragility, safety monitoring should include review of trends in renal function (eGFR and proteinuria).^[21] Individuals who experience progressive eGFR decline or worsening proteinuria while receiving TDF should discontinue this drug to avoid further worsening of kidney function and potential kidney damage.^[7] Our study suggests that TAF may be an appropriate option for such patients.

These data should be viewed with several limitations in mind. Identification of early TAF recipients was retrospective, and the median follow-up was relatively short. In addition, data on albuminuria and renal tubular function (low-molecular-weight proteinuria, glycosuria and phosphate wasting) were not (or only sparsely) available. Finally, the health care systems differed between the two participating countries, with limited access to TAF in the UK and unrestricted access in Ireland.

In conclusion, we observed stable or improved eGFR in the vast majority of early TAF recipients, some of whom had renal impairment and/or experienced renal dysfunction while receiving TDF. Our data suggest that rapid decline in eGFR seen in some TDF-treated individuals may be abrogated by switching from TDF to TAF. As eGFR decline may be an early manifestation of proximal renal tubulopathy,^[6] switching patients who experience eGFR decline from TDF to TAF may be an attractive strategy to avoid renal tubular injury.

1 2 3 4

Next Section

Acknowledgements
The authors would like to thank Melissa Tan who assisted with data collection at the North Bristol NHS Trust, Deborah Williams who oversaw participation at Brighton and Sussex NHS Trust, and members of the University College Dublin Infectious Diseases Cohort Study Team (Aoife Cotter, Gerard Sheehan, Jack Lambert, Eavan Muldoon, Tara McGinty, Willard Tinago, Padraig McGettrick, Stefano Savinelli, Cathal O'Brien, Alan Macken, Elena Alvarez Barco, Alejandro Abner Garcia Leon and Christine Kelly) who provided clinical care to and/or assisted with data provision for the Irish participants.

Financial disclosure
Funding was provided by Gilead Sciences for this investigator-initiated study (IN-UK-311-4340). The funder was not involved in study design, data analysis or the decision to publish our findings. We notified the funder of our intention to submit this manuscript for publication, and the funder was provided with an opportunity to review and comment on the content of the manuscript.

Table 1. Patient characteristics at initiation of tenofovir alafenamide (TAF)-containing antiretroviral therapy
		Prior TDF exposure		Total (n = 357)
		Yes (n = 309)	No (n = 48)	Total (n = 357)
Age (years)	Median (IQR)	50 (42, 59)	49 (34, 58)	50 (42, 59)
Gender (male)	n (%)	250 (81)	36 (75)	286 (80)
Ethnicity
White*/other	n (%)	256 (83)	37 (77)	293 (82)
Black (incl. mixed black)	n (%)	53 (17)	11(23)	64 (18)
HIV transmission risk
Sex between men	n (%)	162 (52)	20 (42)	182 (51)
Heterosexual	n (%)	89 (29)	16 (33)	105 (29)
IDU	n (%)	31 (10)	4 (8)	35 (10)
Other/unknown	n (%)	27 (9)	8 (17)	35 (10)
Time since HIV diagnosis (years)	Median (IQR)	11 (5, 17)	7 (1,14)	10 (4, 17)
Nadir CD4 count (cells/μL)	Median (IQR)	212 (93, 355)	246 (128, 387)	216 (102, 359)
Prior AIDS diagnosis	n/total (%)	69/249 (28)	9/28 (32)	78/277 (28)
Current CD4 count (cells/μL)	Median (IQR)	569 (412, 787)	462 (314, 619)	557 (399, 769)
HIV viral load < 50 copies/mL	n (%)	277 (90)	29 (60)	306 (86)
Hepatitis B (surface antigen positive)	n/total (%)	22/284 (8)	3/44(7)	25/328 (8)
Hepatitis C (antibody positive)	n/total (%)	19/224 (8)	2/29 (7)	21/253 (8)
Hypertension	n/total (%)	41/303 (14)	8/46(17)	49/349 (14)
Diabetes	n/n (%)	28/309 (10)	6/48 (13)	34/357 (10)
eGFR (mL/min/1.73m²)	Median (IQR)	78.6 (63.3, 96.4)	94.1 (54.4, 108.5)	79.5 (62.9, 98.3)
Duration of exposure to TDF	Median (IQR)	4.8 (1.9, 7.9)	-	4.8 (1.9, 7.9)
Number of eGFR measures on TDF	Median (IQR)	10 (5, 15)	-	10 (5, 15)
Time between TDF discontinuation and TAF initiation (months)	Median (IQR)	0 (0, 5.5)	-	0 (0, 5.5)
Duration of exposure to TAF (years)	Median (IQR)	2.0 (1.6, 2.3)	2.1 (1.8, 2.4)	2.0 (1.6, 2.3)
Number of eGFR measures on TAF	Median (IQR)	6 (4, 7)	6 (5, 7)	6 (4, 7)

TDF, tenofovir disoproxil fumarate; IDU, injecting drug use; IQR, interquartile range; eGFR, estimated glomerular filtration rate.
*n = 269 of white ethnicity.

Table 2. Estimated glomerular filtration rate (eGFR) slopes in people with HIV infection who initiated tenofovir alafenamide (TAF)-containing antiretroviral therapy prior to October 2016
	TDF	TAF	P value
Participants with prior TDF exposure (n = 309)
Crude	−2.20 (−2.36, −2.03)	1.16 (0.55, 1.28)	< 0.001
Adjusted	−2.08 (−2.24, −1.92)	1.18 (0.20, 1.52)	< 0.001
Prior TDF exposure; immediate switch from TDF to TAF (n = 91)
Crude	−1.76 (−2.09, −1.42)	0.80 (−0.47,2.06)	< 0.001
Adjusted	−1.67 (−2.00, −1.34)	0.81 (−0.44,2.06)	0.004
Prior TDF exposure; not on TDF at TAF initiation (n = 218)
Crude	−2.38 (−2.57, −2.19)	0.89 (0.27, 1.51)	< 0.001
Adjusted	−2.27 (−2.46, −2.09)	0.90 (0.28, 1.52)	< 0.001
No prior TDF exposure (n = 48)
Crude	-	0.30 (−0.70, 1.30)	-
Adjusted	-	0.18 (−0.91, 1.27)	-
Rapid eGFR decline (> 3 mL/min/1.73 m²/year) on TDF (n = 126)
Crude	−4.20 (−4.44, −3.97)	2.68 (0.84,2.92)	< 0.001
Adjusted	−4.12 (−4.35, −3.89)	2.72 (0.89,3.95)	< 0.001
Rapid eGFR decline (> 5 mL/min/1.73 m²/year) on TDF (n = 70)
Crude	−6.31 (−6.71, −5.92)	2.55 (0.62, 3.06)	< 0.001
Adjusted	−6.20 (−6.59, −5.80)	2.78 (0.95, 3.15)	< 0.001
CKD (eGFR < 60 mL/min/1.73m²) on TDF (n = 99)
Crude	−1.98 (−2.19, −1.76)	0.11 (−1.19, 1.42)	< 0.001
Adjusted	−1.96 (−2.17, −1.74)	0.24 (−1.07, 1.55)	0.001

Slopes are expressed as mean (95% confidence interval) in mL/min/1.73 m² per year; adjusted slopes were adjusted for age at switch, gender, ethnicity, eGFR at TDF/TAF initiation, and time-updated CD4 cell count, HIV RNA [undetectable (< 50 copies/mL) or detectable] and cumulative TDF use (in those with prior TDF exposure).
CKD, chronic kidney disease; TDF, tenofovir disoproxil fumarate.

Table S1. Sensitivity analysis which excludes eGFR measurements during the first month of TDF or TAF exposure (to minimise the effects of co-administered agents on tubular creatinine secretion)
	TDF	TAF	P value
Participants with prior TDF exposure (n=309)
Crude	-2.18 (-2.35, -2.02)	1.20 (0.44, 1.31)	<0.001
Adjusted	-2.07 (-2.23, -1.91)	1.25 (0.21, 1.54)	<0.001
Prior TDF exposure; immediate switch from TDF to TAF (n=91)
Crude	-1.74 (-2.08, -1.41)	0.88 (0.25,1.51)	<0.001
Adjusted	-1.65 (-1.98, -1.32)	0.89 (0.26,1.52)	<0.001
Prior TDF exposure; not on TDF at TAF initiation (n=218)
Crude	-2.37 (-2.56, -2.18)	0.93 (-0.37, 2.24)	<0.001
Adjusted	-2.26 (-2.45, -2.07)	0.94 (-0.35, 2.22)	<0.001
No prior TDF exposure (n=48)
Crude	-	0.39 (-1.44, 0.65)	-
Adjusted	-	0.28 (-1.43, 0.87)	-
Rapid eGFR decline (>3 mL/min/1.73m²/year) on TDF (n=126)
Crude	-4.56 (-4.82, -4.30)	2.70 (0.90, 2.99)	<0.001
Adjusted	-4.47 (-4.73, -4.21)	2.81 (0.92,3.50)	<0.001
Rapid eGFR decline (>5 mL/min/1.73m²/year) on TDF (n=70)
Crude	-6.34 (-6.77, -5.93)	2.59 (0.63, 3.19)	<0.001
Adjusted	-6.22 (-6.64, -5.81)	2.81 (0.96, 3.29)	<0.001
CKD (eGFR<60 mL/min/1.73m²) on TDF (n=99)
Crude	-1.95 (-2.16, -1.73)	0.39 (-0.73, 1.51)	<0.001
Adjusted	-1.92 (-2.14, -1.70)	0.37 (-0.74, 1.48)	<0.001