Associations between tenofovir use and renal complications in VA cohort
Simon Collins, HIV i-Base
Tenofovir is one of the most widely used antiretrovirals and the association with a generally low risk of renal complications has been widely reported. However, there has been conflicting data on potential for renal complications with cumulative use or in patients with normal renal function.
An analysis from the US Veterans Association (VA) cohort published in the 24 April 2012 edition of AIDS reported that cumulative use of tenofovir was associated with renal complications and that this might not be reversible. 
From 1997-2007, more than 19,700 treatment naïve patients were reported as starting ART in the VA cohort. Discounting those without at least one of the key parameters: CD4, viral load, out-patient visit, renal markers or with renal failure however, reduced this study to 10,841 patients, 4,303 of whom used tenofovir. An era-of-use analysis adjusted for tenofovir not being approved until 2001: 85% of patients used tenfovir in the period 2005-07; 54% in 2003-05 and 17% prior to 2003).
Changes in renal function were determined by one of three criteria.
- Proteinuria (urine dipstick >=30 mg/dL on two consecutive tests).
- Rapid decline in kidney function (>3 mL/min/1.73m2 annual decline for at least two years), and
- Chronic kidney disease (CKD) (eGFR < 60 mL/min/1.73m2 on two occasions at least 3 months apart).
Additional sensitivity analyses were also performed for time to events and for more extreme renal dysfunction. Patients with proteinuria or CKD at baseline were excluded from those analyses. Hazard ratios (HR) were calculated adjusting for demographic, time dependent and marginal structural models.
Median age was 46 years (IQR 40-52), and approximately 98% of participants were men. Ethnicity included approximately 50% black, 40% white and 10% other. Median CD4 count and viral load before treatment were approximately 200 (IQR 50-400) cells/mm3 and 60,000 (IQR 15,000-220,000) copies/mL.
Prevalence of comorbid conditions at baseline (in the TDF vs no-TDF groups) included hypertension (38% vs 39%), diabetes (6.8 vs 7.9%), HCV (14 vs 17%), smoking (18% vs 19%) and dyslipidaemia (15% both groups). Renal disease at baseline included approximate median eGFR 96 (IQR 82-114) mL/min per 1.73 m2, with 4.7% vs 7.3% with eGRF <60 mL/min/1.73m2 and 19% vs 21% with proteinuria (>30 mg/dL).
Median follow-up per individual ranged from 3.9 years (for proteinuria) to 5.5 years (for CKD), during which there were 3,400 cases of proteinuria (>38,000 patient years), 3078 of rapid kidney decline (>51,500 PY) and 533 CKD events (>56,400 PY). However, participants using tenofovir only had a median of 1.0 year exposure (IQR 0.5-1.9). Therefore 25% of people providing data used tenofovir for less than 6 months, 50% for less than 12 months and 75% less than 2 years. Maximum tenofovir use was 6.3 years. The summary of events shown in Table 1, published as supplementary information, is important to estimate rates in the tenofovir vs no-tenofovir groups, given that the results are in the main paper are based on hazard ratios.
|Tenofovir ever used||2,646||32,421||81.6|
|Tenofovir ever used||754||5,711||132.0|
|Tenofovir ever used||2,349||43,693||53.8|
|Tenofovir ever used||729||7,896||92.3|
|Chronic Kidney Disease|
|Tenofovir ever used||352||46,724||7.5|
|Tenofovir ever used||181||9,692||18.7|
All ARVs were included in the analysis, but only tenofovir had an increased association with all three renal markers, in all three adjusted analyses, see Table 2. In the time dependent analysis (adjusting for ARV use in addition to baseline demographics), each year of exposure to tenofovir was associated with 34% increased risk of proteinuria (95%CI 25-45%, p < 0.0001), 11% increased risk of rapid decline (3-18%, p = 0.0033), and 33% increased risk of CKD (18-51%; p < 0.0001). Controlling for slightly more frequent monitoring in tenofovir users did not affect the results. Pre-existing renal risk factors did not appear to worsen the effects of tenofovir.
|Demographic adjusted model. Hazard ratio (95% CI)||Time dependent Cox model. Hazard ratio (95% CI)||Marginal structural model. Hazard ratio (95% CI)|
|Cumulative exposure to tenofovir (per year)|
|Proteinuria||1.30 * (1.22 – 1.37)||1.34 * (1.25 – 1.45)||1.24 * (1.17 – 1.32)|
|Rapid decline||1.1 * (1.11 – 1.24)||1.11 ** (1.03 – 1.18)||1.16 * (1.09 – 1.23)|
|CKD||1.44 * (1.30 – 1.60)||1.33 * (1.18 – 1.51)||1.36 * (1.22 – 1.51)|
|Ever exposure to tenofovir|
|Proteinuria||1.70 * (1.57 – 1.85)||1.68 * (1.52 – 1.85)||1.51 * (1.36 – 1.66)|
|Rapid decline||1.51 * (1.39 – 1.64)||1.36 * (1.23 – 1.50)||1.50 * (1.36 – 1.67)|
|CKD||2.11 * (1.76 – 2.54)||1.71 * (1.38 – 2.12)||1.88 * (1.50 – 2.36)|
* all P <0.0001 except ** p=0.0033
Other ARVs showed weaker or inconsistent associations with kidney disease events, notably with ritonavir and lopinavir/r associated with proteinuria, atazanavir with rapid decline and indinavir with CKD. see Table 3.
|ARV||% pts with exposure||proteinuria||p||rapid decline||p||CKD||p|
|Tenofovir||39.7||1.34 (1.25 – 1.45)||<0.0001||1.11 (1.03 – 1.18)||0.0033||1.33 (1.18 – 1.51)||<0.0001|
|AZT||68.3||0.98 (0.93-1.03)||0.42||0.98 (0.93-1.02)||0.29||0.89 (0.81-0.98)||0.020|
|Efavirenz||49.0||0.94 (0.90-0.99)||0.026||1.01 (0.97-1.05)||0.64||0.88 (0.79-0.98)||0.018|
|Ritonavir||35.7||1.18 (1.09-1.27)||<0.0001||0.96 (0.89-1.04)||0.34||0.97 (0.84-1.14)||0.74|
|Indinavir||24.6||1.04 (0.99 – 1.09)||0.15||0.99 (0.95-1.04)||0.67||1.16 (1.06-1.27)||0.0019|
|Atazanavir||17.1||0.93 (0.79-1.08)||0.34||1.22 (1.07-1.40)||0.0035||0.96 (0.77-1.18)||0.69|
|Lopinavir/r||15.3||0.77 (0.68-0.86)||<0.0001||1.05 (0.94-1.17)||0.39||1.21 (0.91-1.60)||0.18|
|Saquinavir||10.7||0.91 (0.83-0.99)||0.035||1.00 (0.92-1.08)||0.97||0.89 (0.72-1.09)||0.24|
The association with tenofovir exposure was consistent across sub groups by age, race, all baseline comorbidities except diabetes and CKD, viral load, CD4 and BMI.
Among those who discontinued tenofovir use, risk of kidney disease events did not appear to increase or decrease during median follow-up of 1.2 years. Previous use of tenofovir was associated with a higher risk of all complications compared to never-use.
This study was widely reported based on statistically significantly increases of 34% (proteinuria), 11% (rapid decline) and 33% (CKD) per year of exposure to tenofovir, after adjusting for traditional risks for renal complications, with increases from ever-use of 68%, 36% and 71% respectively.
As will all medical reports, relative rates (in this case, hazard ratios) have to also be interpreted together with data that supports the absolute risks associated with both tenofovir and non-tenofovir use.
Even given the generally low duration of use with tenofovir and limited follow-up after discontinuation, and that was a male study, these results are clearly important, especially when supported by other studies, such as the D:A:D analysis presented at CROI (see earlier in this issue of HTB).
Although the optimal way to define a rapid decline in kidney function is unclear, these seem like reasonable markers to have selected even though other groups (including D:A:D) use different criteria.
While it is unclear whether any minimum number of eGRF measures were needed when calculating the rate of decline within a year, as three or more would more accurately reflect a true decline rather than annual fluctuation but it is good that they excluded assessments of renal function during in-patient episodes as many other studies have not been able to do this.
With limited follow-up, it is difficult to separate the effect of ‘ever exposure’ from ‘cumulative exposure’ (and even with longer follow-up, this isn’t always straightforward), but this will only become clear in future analyses.
As renal function was not always assessed during early the HAART period, patients with these early data may have had other renal complications requiring monitoring. However, the results did not change significantly when patients prior to 2001 were excluded.
Although the method of fitting the marginal structural models may be unclear, it is somewhat reassuring that similar results were found for all three models, suggesting that the results are robust to the choice of methodological approach.
The perceived risk of tenofovir and renal complications clearly affected the choice of early switching and explains the lack of associations with more advanced stages of CKD.
Scherzer R et al. Association of tenofovir exposure with kidney disease risk in HIV infection. AIDS. 26(7):867-875, April 24, 2012. doi: 10.1097/QAD.0b013e328351f68f