CROI: Is tenofovir causing kidney problems?

This article is more than 15 years old. Click here for more recent articles on this topic

Tenofovir (Viread) has been associated with kidney (renal) toxicity in certain cases, although this has proven to be nearly always mild, rare, and reversible. Researchers continue to monitor the long-term safety of this drug for any warning signals, and several poster presentations at the Fourteenth Conference on Retroviruses and Opportunistic Infections, Los Angeles, investigated the likelihood of, and risk factors for, kidney toxicity among people taking tenofovir.

Presentations showed that tenofovir treatment may be associated with modest declines in kidney function, especially in treatment-experienced patients, and that exposure to low-dose ritonavir may increase the risk of impaired kidney function in tenofovir-treated individuals.

Johns Hopkins cohort

Researchers at Johns Hopkins University in Baltimore took data from the Johns Hopkins HIV Clinical Cohort, a prospective observational cohort study which has enrolled over 5,500 patients since 1990. In this analysis, they looked at all patients with an estimated GFR of greater than 50 ml/minute/1.73 m2(calculated by the MDRD equation), who had started NRTI-containing combination therapy after January 1st, 2001. Of this group, 627 (565 treatment-experienced, 62 treatment-naïve) were taking tenofovir (TDF) and 311 (211 experienced, 100 naïve) were taking alternate NRTIs. There was no difference in baseline GFR (115 vs. 114 ml/minute/1.73 m2, p = 0.27 or how long they had been on therapy (median 506 days TDF vs. 519 days NRTI, p = 0.25).

The team found a small decline in GFR in ART-experienced patients (but not ART-naïve patients) on tenofovir. There were greater GFR declines in ART-experienced patients on tenofovir than in those on other NRTIs (15 vs. 11 ml/min/1.73 m2, p < 0.01; 14% vs. 9% decline p < 0.001), with the greatest decline occurring around the 20-week point. No significant differences were seen in ART-naïve patients (15 vs. 14 ml/minute/1.73 m2, p < 0.10). Other risk factors for GFR decline were CD4 cell counts < 50 cells/mm3, diabetes, and age greater than 45 years. Tenofovir was still associated with greater GFR declines (in treatment-experienced people only) after adjusting for these and other variables.



In HIV, an individual who is ‘treatment naïve’ has never taken anti-HIV treatment before.




Relating to the kidneys.


Breakdown product of creatine phosphate in muscle, usually produced at a fairly constant rate by the body (depending on muscle mass). As a blood test, it is an important indicator of the health of the kidneys because it is an easily measured by-product of muscle metabolism that is excreted unchanged by the kidneys.

boosting agent

Booster drugs are used to ‘boost’ the effects of protease inhibitors and some other antiretrovirals. Adding a small dose of a booster drug to an antiretroviral makes the liver break down the primary drug more slowly, which means that it stays in the body for longer times or at higher levels. Without the boosting agent, the prescribed dose of the primary drug would be ineffective.

The Johns Hopkins researchers concluded that tenofovir “was associated with a 4% greater decline in GFR compared with alternative NRTIs in ARV-experienced patients… Although statistically significant, the GFR decline may not be of clinical significance. The decline … was small, most commonly occurred within the first 6 months of treatment and was not progressive.”

Measures of kidney toxicity

The simplest “flag” of possible kidney problems is elevated blood levels of creatinine. The standard indicator of kidney function (and hence of kidney toxicity) is glomerular filtration rate (GFR). GFR is related to the creatinine clearance rate, and can be calculated from creatinine levels and other factors (including sex and age). Various equations (principally Cockcroft-Gault and MDRD) can be used to calculate estimates of kidney toxicity, which are therefore not always directly comparable (and may sometimes be expressed in differing units).

Swiss HIV Cohort Study

In another poster presentation, researchers in Switzerland drew on the Swiss HIV Cohort Study, a prospective cohort study of HIV-positive adults. People who were either treatment-naïve or had interrupted treatment for at least twelve months, and had started combination therapy either including tenofovir (n=213; 100 naïve, 113 restart) or not (n=494; 357 naïve, 137 restart). Calculated GFRs were retrospectively analysed for people thus extracted from the larger prospective cohort study.

In this analysis, the Cockcroft-Gault (CG) equation was used to estimate GFR with the modification of diet in renal disease (MDRD) as a second, confirmatory calculation, the results of the two were “highly consistent”. The endpoint was a GFR decrease of at least 10 mL/minute by CG calculation, sustained over at least a month.

The time to sustained 10 and 20 ml/min reductions of GFR were significantly shorter for patients taking tenofovir than for those who were not (p < 0.001). Over two years, the risk of a sustained 10 ml/minute decrease in calculated GFR was 52% in those on tenofovir and 21% in others. Boosted protease inhibitor use was identified as a risk factor (hazard ratio=1.63 by CG calculation, p=0.01, HR =2.21 by MDRD, p < 0.0001). Other risk factors (female gender, prior AIDS diagnosis, diabetes) were identified by some analysis methods but not by others. Altogether, 46 (2%) of 2592 tenofovir-treated patients had to stop TDF because of drug-related kidney toxicity after a mean of 442 days.

The Swiss HIV Cohort Study team concluded that beginning combination therapy of any kind was associated with “a substantial decrease in cGFR [(calculated GFR)] in all subpopulations studied. This decrease was significantly more frequent in patients treated with TDF. HIV-related patient characteristics had little impact on cGFR dynamics.”

Two Californian studies

Another study team in California confirmed the increased risk of kidney toxicity resulting from tenofovir combined with boosted protease inhibitors (PIs). This team used stored blood samples from CCTG 578, a completed therapeutic drug monitoring study. Participants were included if they had received tenofovir, a ritonavir-boosted PI, or an NNRTI for at least 48 weeks. This resulted in a total of 147 participants: 51 on TDF plus a boosted PI, 29 on TDF plus an NNRTI, and 67 on combinations not including tenofovir. Baseline mean age (40 years), CD4 cell count (197 cells/mm3) and GFR (C-G: 107 ml/minute by CG, 108 ml/min/1.73 m2 by MDRD) were similar between the three groups.

After 48 weeks on treatment, the estimated GFR rate declined more quickly in those taking tenofovir and boosted PIs compared to tenofovir and NNRTIs: (by CG: 13.9 vs. 6.2 ml/minute/year, p = 0.033; by MDRD: 15 vs. 4 ml/minute/1.73 m2/year, p = 0.018). Treatment history (naïve vs. experienced), baseline CD4 counts and viral load did not appear as risk factors in this study; older age did (GFR decline of 1 ml/minute/year of age, p = 0.001). The researchers in this study concluded that: “therapy with TDF+PI/r was strongly associated with greater renal function decline over 48 weeks compared to TDF+NNRTI-based regimens. Differences in week 2 TDF plasma exposure did not explain this observation.”

Another retrospective cohort study from the University of California San Diego also found low-dose ritonavir to be a risk factor for kidney toxicity with tenofovir-containing regimens. In this group of 635 adults without pre-existing kidney disease (creatinine clearance equal or above 75 ml/minute) and on stable combination ART from 2000 to 2005, three study groups were compared: 184 on combinations including tenofovir and ritonavir (RTV), 71 taking TDF but not RTV, and 380 not taking TDF.

Various kidney function endpoint measurements were used – elevated serum creatinine, and decreased creatinine clearance of 30 ml/minute or by 25% (by CG and MDRD calculations). Although rates varied by endpoint used, 33% of the 635 met at least one of the toxicity endpoints. After adjusting for body mass index, baseline CD4 count, viral load, and MDRD clearance, being on tenofovir plus ritonavir “remained a significant predictor of MDRD ¯ 30 cc/minutes (adjusted HR 1.9, p = 0.001), and MDRD ¯ 25% (adjusted HR 1.8, p < 0.0001), whereas TDF without RTV was not associated with increased risk.” The UCSD team concluded that “low-dose RTV in TDF-containing regimens was associated with nephrotoxicity … in patients without pre-existing renal disease, while TDF alone posed no additional risk.”


Moore R et al. Tenofovir and renal dysfunction in clinical practice. Fourteenth Conference on Retroviruses and Opportunistic Infections, Los Angeles, abstract 832, 2007.

Fux C et al. Tenofovir treatment is associated with a decrease in calculated glomerular filtration rates in a large observational cohort. Fourteenth Conference on Retroviruses and Opportunistic Infections, Los Angeles, abstract 834, 2007.

Goicoechea M et al.Increased renal impairment in patients receiving TDF+PI vs. TDF+NNRTI. Fourteenth Conference on Retroviruses and Opportunistic Infections, Los Angeles, abstract 835, 2007.

Wai H et al. Risk factors for tenofovir-associated nephrotoxicity identified in an HIV clinic cohort. Fourteenth Conference on Retroviruses and Opportunistic Infections, Los Angeles, abstract 833, 2007.