Kidney toxicity is most likely to occur in patients taking tenofovir if they are controlling high blood pressure with potentially kidney-toxic drugs, and also taking protease inhibitors, delegates at the XVII International AIDS Conference heard on Wednesday.
Tenofovir (Viread) is a widely-used antiretroviral and a preferred component of first-line regimens whose safety has been generally well-established.
However, it is known to cause relatively rare but potentially serious kidney toxicity, including acute renal failure and Fanconi's syndrome, in some patients. Previous studies have investigated some risks for developing toxicity, including high blood pressure, diabetes, and ritonavir use.
During an oral session on Wednesday morning, Chelsea Castellano of Duke University Medical Center presented findings from a carefully designed, NIH-funded case-control study which aimed to estimate the frequency of kidney toxicity in HIV-positive patients taking tenofovir, and to find factors which predicted its occurrence. There was no pharmaceutical sponsorship.
Nephrotoxicity (kidney toxicity) was defined as: a decrease in glomerular filtration rate (GFR) of more than 50% (as per "RIFLE" criteria, a standard accepted definition), or an absolute decrease in creatinine clearance (CrCl) of ≥ 25 ml/minute. This additional component of the definition was added, Castellano stated, to capture kidney function declines in people with good creatinine clearance at baseline. (Questioned about this later, she added that most of the identified cases of toxicity – 29 out of 35 – fit the standard definition. An additional six were identified purely due to the absolute CrCl decrease of ≥ 25 ml/minute.)
The study group in this investigation was drawn from 1574 patients seen at the HIV clinic of North Carolina's Duke University Medical Center between October 2001 (when tenofovir was approved by the US FDA) and August 2007.
From this group, records were selected for all patients who had received tenofovir for at least three months, and for whom serum creatinine levels had been recorded one year prior to starting the drug. This resulted in a study group of 744 patients. From the remaining 830 clinic patients, 191 (who had been on a non-tenofovir-containing regimen for ≥ one year) were randomly selected as controls for comparison.
Of these 744 patients, 56 cases of tenofovir-associated nephrotoxicity were identified. These were compared to 74 controls, randomly selected from tenofovir recipients who had had no significant change in serum creatinine levels. Demographically, the two groups were both similar to the overall clinic population, which was: 71% male, 55% African-American, 37% Caucasian, 37% on private insurance, and 35% on Medicaid or Medicare. Duration of HIV infection was not presented, but was "not different between the groups." Twenty percent of the control group had a history of opportunistic infections (OIs) (a slightly lower incidence than the overall clinic population).
Information was also available on a wide range of other clinical factors, co-morbidities including high blood pressure, hyperlipidaemia, diabetes, smoking, hepatitis B or hepatitis C infection, and concomitant and potentially kidney-toxic medications including NSAIDs, foscarnet, amphotericin, ACE inhibitors and others.
What predicted kidney toxicity?
Kidney toxicity occurred at similar rates in patients on tenofovir and in the control group: 35/744 (4.7%) versus 8/191 (4.2%).
Note: the authors of this study originally stated that kidney toxicity occurred nearly twice as frequently in patients on tenofovir than in the control group: 35/744 (7.5%) versus 8/191 (4.2%), with a borderline statistical significance (p=.052), but subsequently issued a correction.
Of the 35 cases of nephrotoxicity, 20 (57%) discontinued tenofovir, and 16 (80%) of these saw significant renal improvement after discontinuation. Of the 15 who continued on tenofovir, 10 (67%) continued to have abnormal renal function.
Significant predictors of developing nephrotoxicity included:
- concurrent nephrotoxic medications (odds ratio [OR] 6.36; 95% confidence interval [CI], 2.21 to 18.4; p<.001)
- other medical co-morbidities (OR 5.43; 95% CI, 2.22 to 13.25; p<.001)
- high blood pressure (OR 4.79; 95% CI, 2.0 to 11.45; p<.001)
- chronic pain (a marker of potential NSAID use) (OR 4.58; 95% CI, 1.59 to 13.21; p=.003)
- concurrent protease inhibitor use (OR 3.79; 95% CI, 1.47 to 9.75; p=.004)
- previous protease inhibitor use (OR 2.80; 95% CI, 1.18 to 6.63; p=.02)
- history of opportunistic infections (OR 2.40; 95% CI, 1.03 to 5.60; p=.04)
Being on Medicaid or Medicare was also a risk factor for nephrotoxicity (p<.001).
The risk of nephrotoxicity was diminished to nearly a third when tenofovir was used with concurrent NNRTIs (OR 0.36; 95% CI, 0.16 to 0.83; p=.01).
Nephrotoxicity was seen in all patients (25/25) for whom both high blood pressure and nephrotoxic drug use were present. Only one patient in 39 (3%) experienced nephrotoxicity when both were absent.
Castellano concluded that nephrotoxicity was more common among HIV-positive clinic patients who were taking tenofovir, but that "clinical risk factors for significant tenofovir-associated nephrotoxicity were readily identifiable." Most cases of toxicity occurred in patients with high blood pressure requiring other nephrotoxic drugs, especially if given with a protease inhibitor. Toxicity was generally reversible on discontinuation of tenofovir, and persistent if the drug was continued. Antiretroviral regimens using tenofovir in combination with NNRTIs were "rarely associated" with renal toxicity.