Tenofovir-related nephrotoxicity in HIV-infected patients

Tenofovir disoproxil fumarate (TDF) is the first nucleotide analogue approved for the treatment of HIV infection. Its potential for mitochondrial dysfunction is lower than when using nucleoside analogues [^1,2], but other nucleotide analogues, such as cidofovir and adefovir, may induce kidney damage as a result of mitochondrial DNA depletion in kidney tubular cells [³]. As the structure of TDF is closely related to those molecules, concern has been raised about its potential nephrotoxicity. Cell culture studies, however, have demonstrated that TDF is only minimally toxic for renal tubular cells [⁴], even though the use of supratherapeutic doses of TDF in animals may cause renal toxicity [⁵]. Although human trials with TDF have not shown clinically relevant renal dysfunction [^6,7], up to 22 cases of TDF-related nephrotoxicity have been reported up to August 2003 (Table 1) [^8-17]. We review these cases here, and describe a new patient with acute renal failure associated with TDF.

Table 1

A 46-year-old man with HIV infection and a previous history of arterial hypertension, hypercholesterolemia and high alcohol intake, was seen at our outpatient clinic in March 2003. He CD4 cell count was 819 cells/μl and his plasma HIV-RNA level was less than 50 copies/ml. His weight was 81 kg and he had no previous history of renal dysfunction. He had initiated treatment with stavudine, didanosine, and efavirenz 3 years earlier, switching to didanosine, lamivudine and nelfinavir one year later because of liver toxicity. Nineteen months later he was invited to switch to a simplified regimen based on didanosine, TDF and efavirenz. Concomitant medications were enalapril and pravastatin.

One month after initiating this antiretroviral regimen, the patient complained of cough, rhinorrhea and vomiting, and metoclopramide was prescribed. The next day he returned to the hospital complaining of abnormal perioral sensations. Renal dysfunction was diagnosed, and he was admitted to hospital. His plasma creatinine level was 159 μmol/l (normal < 115 μmol/l). Sodium, potassium, amylase and uric acid levels were within normal levels. Urinoanalysis showed three to 10 white cells per high-power field, glucosuria, hyaline and granulose casts, and mild proteinuria. Urine culture was sterile. Creatinine clearance was slightly reduced (77 ml/min).

Antiretroviral drugs were discontinued, clinical symptoms improved and creatinine levels returned to normal values within 48 h. One month after discharge, the patient resumed treatment with didanosine, lamivudine and efavirenz. He has remained asymptomatic, and blood biochemistry and urinoanalysis were normal for the following 4 months, except for mild proteinuria (480 mg/day), which disappeared thereafter.

The incidence of TDF-induced nephrotoxicity is not well known. Significant creatinine elevations have been reported in 5-7% of patients taking TDF for at least 6 months [^6,15]. However, many cases may have been misdiagnosed, as renal damage may occur despite normal creatinine levels. Only 1-4% of patients discontinue TDF because of nephrotoxicity during the first year of treatment [^10,13,15].

TDF-related kidney toxicity seems to be more frequent in patients with CD4 cell counts of less than 150 cells/μl. Therefore, immunosuppression might increase the risk of nephrotoxicity in patients exposed to TDF, as occurs with many other adverse events associated with antiretroviral drugs [¹⁸]. In addition, many patients with TDF nephrotoxicity were underweight and received low doses of ritonavir [^10-12]. In four cases, renal insufficiency was present before TDF was initiated. Therefore, factors leading to a greater exposure to TDF might also facilitate kidney toxicity. This hypothesis is supported by animal studies, in which nephrotoxicity was the main TDF dose-limiting adverse event [⁵]. With respect to the co-administration of ritonavir, it may increase TDF plasma levels by 30% [⁵], which might be clinically relevant [¹⁸].

Several facts support the involvement of TDF in the development of renal dysfunction in our patient. First, acute renal failure occurred soon after TDF was initiated. Second, the patient had no previous renal dysfunction and TDF was the only new agent added before renal abnormalities appeared. Third, blood and urine analyses were similar to those reported in other cases from the literature. The most frequent manifestation of TDF renal toxicity is proximal tubulopathy, occasionally leading to Fanconi's syndrome (proximal tubular acidosis, hypophosphoremia, hypouricemia, glucosuria and proteinuria). As expected, the discontinuation of TDF led to the normalization of renal function.

Our case differs from others in several features. The patient had high CD4 cell counts and was neither underweight nor receiving interacting drugs. TDF is eliminated by the kidney by both filtration and mainly active tubular secretion via human organic anion transporter 1, as are cidofovir and adefovir. Once the drug is accumulated within the tubular cells, its high intracellular concentration may interfere with cell function. This susceptibility of the kidney to the cytotoxic effect of nucleotides is caused by the high expression of human organic anion transporter 1 in kidney cells [¹⁹]. Hypothetically, TDF nephrotoxicity in patients without any other predisposing factors might be genetically determined by a higher expression of this transporter.

In conclusion, physicians should be aware of the potential nephrotoxicity of TDF, even in patients lacking any predisposing factor. In all patients receiving TDF, routine biochemical monitoring should include urinoanalysis and plasma phosphorus, lactate and uric acid levels. If any sign of tubulopathy develops, TDF should be stopped as soon as possible. The contribution of weight, concomitant drugs, CD4 cell counts, and genetic predisposition requires further investigation.