• highly active antiretroviral therapy;
  • osteomalacia;
  • renal disease;
  • tenofovir;
  • toxicity


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References

Objectives The aims of the study were to describe the clinical presentation and renal and bone abnormalities in a case series of HIV-infected patients receiving treatment with tenofovir (TDF), and to recommend appropriate screening for toxicity related to TDF.

Methods Patients were identified from referrals to a specialist HIV renal clinic. Patients were included if treatment with TDF was assessed as the primary cause of the renal function impairment and clinical data were available prior to and following discontinuation of TDF treatment. Data were collected from case note review and clinic databases.

Results Twenty-two patients (1.6% of all those who received TDF) were identified with TDF-associated renal toxicity. All had normal serum creatinine prior to TDF therapy. All presented with proteinuria. On stopping TDF, renal function improved. Eight patients had confirmed Fanconi syndrome. Twelve patients presented with bone pain and osteomalacia was confirmed on an isotope bone scan in seven of these patients. The findings (in those patients tested) of tubular proteinuria, reduced tubular transport maximum of phosphate (TmP), and glycosuria were all consistent with the proximal tubule being the site of toxicity.

Conclusion Renal toxicity remains a concern in patients treated with TDF. Clinical presentation may be with renal dysfunction, Fanconi syndrome or osteomalacia. Our investigations suggest proximal tubular toxicity as a common pathogenic mechanism.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References

Tenofovir (TDF) has been approved for the treatment of HIV infection in highly active antiretroviral therapy (HAART) combinations since October 2001. In randomized controlled trials of TDF, renal toxicity is uncommon, with grade 3 and 4 elevations in serum creatinine reported in <1.0% of patients over 96 weeks [1,2], although studies consistently report up to a 10% decrease in estimated glomerular filtration rate (GFR) over time in patients treated with TDF compared with controls [1,3,4]. Trials of drug toxicity that rely on serum creatinine, along with urine dipstick testing for albuminuria, to detect renal involvement are essentially screening for markers of glomerular disease, and may not detect abnormalities in renal tubular function.

The renal proximal tubule (PT) is a major site for the excretion of xenobiotics, including TDF. PT dysfunction leads to the wasting of substances in the urine that are normally freely filtered by the glomerulus and then reabsorbed by the PT, such as low-molecular-weight proteins (LMWPs), phosphate or glucose. The outcome will be determined by the severity of the insult. Mild PT dysfunction may be detected by the asymptomatic presence of LMWPs in the urine, such as retinol-binding protein (RBP), whereas severe disease results in the classical renal Fanconi syndrome (FS), which is characterized by osteomalacia, metabolic acidosis and glycosuria (in the absence of hyperglycaemia).

Renal tubular toxicity occurs frequently in animals treated with TDF but at drug concentrations higher than used in humans [5]. Multiple case reports have been published describing renal toxicity including FS and osteomalacia in patients treated with TDF [6–8]. The mechanism for TDF-associated renal tubular toxicity remains unclear.

We describe the clinical presentation and renal abnormalities in a case series of HIV-infected patients who developed renal toxicity while receiving treatment with TDF as part of their antiretroviral therapy (ART). We propose that the site of TDF toxicity is the PT, and that screening for renal toxicity using serum creatinine alone is ineffective. We recommend that screening for renal toxicity in patients receiving TDF should include estimation of the urine protein/creatinine ratio (Up/c) and monitoring of serum phosphate. TDF-associated renal toxicity appears to be reversible on stopping therapy, underlining the value of effective screening.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References

Patients were identified from referrals to a joint specialist HIV renal clinic from two London-based HIV clinical centres. Referral to the specialist clinic was at the judgement of the referring clinician and largely determined by concerns over impairment of renal function, which included graded increases in serum creatinine, progressive decline in estimated GFR (eGFR) and/or development of clinical proteinuria. Patients were included in the study if treatment with TDF was assessed as the primary cause of the renal function impairment and clinical data were available prior to and following discontinuation of TDF treatment. Assessment was based on the finding of development of renal impairment while on TDF, diagnosis of proximal tubular dysfunction, the improvement of renal function following the discontinuation of TDF and an absence of an alternative primary cause. Data were collected from case note review and clinic databases. Data collected included (where possible): patient demographics, HIV disease stage, current and previous ART, concomitant medications, significant past medical history, serum biochemistry, Up/c, urine RBP/creatinine ratio (Ur/c), urine albumin/creatinine ratio (Ua/c), renal biopsy and bone imaging.

eGFR was calculated from serum creatinine using the modification of diet in renal disease (MDRD) formula [9]. Ua/c and Ur/c were used as markers of glomerular and PT disease, respectively. PT function was further assessed where possible by dipsticking urine for glycosuria, and calculating tubular transport maximum of phosphate (TmP). TmP reflects the capacity of the PT to reabsorb urinary phosphate. It is calculated from the formula: [phosphate]plasma−([phosphate]urine × [creatinine]plasma/[creatinine]urine), and is adjusted for GFR.

The Wilcoxon matched pairs signed rank test was used to compare changes in measured variables.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References


The two clinical centres provided HIV care for 5687 patients between 2004 and 2006. Twenty-two patients were identified with TDF-associated renal toxicity. Of these patients, 21 were male and one female, with a median age of 45.1 years (range 27–59 years), and 21 of the 22 patients were of white ethnicity. The median time from HIV diagnosis was 9.1 years (range 1.4–20.8 years), with 13 of the 22 patients having a previous AIDS-defining disease. Four patients had hypertension (on therapy); two patients were coinfected with the hepatitis C virus; none was diabetic. Ten of the 22 patients were receiving treatment with other potentially nephrotoxic drugs at the time of presentation [aciclovir, n=5; cotrimoxazole, n=4; Non-Steroidal Anti -Infammatory Drugs (NSAIDS), n=4]. In only one of these 10 cases was the drug discontinued at the same time as TDF.

HIV status and ART

The median length of time on ART was 7.0 years (range 1.2–17.8 years) and the median number of antiretroviral drugs ever received was 3.7 (range 0–10). Six of the 22 patients (27.3%) were on first-line ART. The median time on TDF was 2.5 years (range 0.2–4.2 years). Five of the 22 patients (22.7%) were concurrently receiving didanosine. Twenty of the 22 patients (90.9%) were taking TDF in combination with a ritonavir-boosted protease inhibitor (PI). At the time of presentation, 19 of the 22 patients (86.3%) had a plasma HIV RNA level <50 HIV-1 RNA copies/mL (one patient had a missing value, one had 50–400 copies/mL, and one had >400 copies/mL) and their median CD4 count was 521 cells/μL (range 130–1116 cells/μL).

Serum creatinine and estimated GFR

Prior to starting treatment with TDF, all patients with available data (n=19) had a serum creatinine concentration within the normal range (median 84 μmol/L; range 68–111 μmol/L). On TDF, all of these patients experienced a rise in serum creatinine (median rise 88 μmol/L; range 5–659 μmol/L), 15 to above the normal range. Overall, 18 of the 22 patients presented with a raised serum creatinine level. Following discontinuation of TDF, serum creatinine decreased in 95.5% of patients (21 of 22), in 59% (13 of 22) to within the normal range (Table 1).

Table 1.   Median (range) pre, post and on treatment with tenofovir (TDF) of selected laboratory tests
 Pre-TDFOn TDFPost TDFTime post TDF (days)P-value (on vs. post TDF)
  1. Creatinine, serum creatinine; eGFR, estimated glomerular filtration rate (mL/min/1.73 m2); Ur P/Cr, urinary protein:creatinine ratio; Phosphate, serum inorganic phosphate; ALP, alkaline phosphatase; Time post TDF, time in days between discontinuation of TDF and observed value; On TDF, observed value at time of diagnosis of renal toxicity.

Creatinine (mmol/mL)84165110230<0.001
eGFR (mL/min/1.73 m2)935369230<0.001
Ur P/Cr (mg/mmol)n/a16636183<0.001
Phosphate (mmol/L)1.120.591258<0.001
ALP (iu/L)83176116233<0.001

Prior to starting TDF, 19 of 22 patients (three patients had data missing) had an eGFR >60 (Fig. 1a). On TDF, GFR decreased in all of these patients (median decrease 41.3 mL/min/1.73 m2; range 8–91 mL/min/1.73 m2), with 14 of the 22 patients (63.6%) developing an eGFR of <60 mL/min/1.73 m2. On discontinuation of TDF, although the eGFR increased in all patients, it was normal (>90 mL/min/1.73 m2) in only 9% of patients (two of 22) (Fig. 1a), between 60 and 90 mL/min/1.73 m2 in 63.6% of patients (14 of 22) and <60 mL/min/1.73 m2 in 27.3% of patients (six of 22). In 10 patients with >6 months of post-TDF follow-up, three (30%) continued to have an eGFR of <60 mL/min/1.73 m2.


Figure 1.  Laboratory markers of renal function in patients developing tenofovir (TDF)-associated renal toxicity. (a) Estimated glomerular filtration rate (eGFR); (b) urinary protein:creatinine ratio (Pr/Cr); (c) urinary retinol-binding protein:creatinine ratio (RBP/Cr); (d) plasma inorganic phosphate (Phos).

Download figure to PowerPoint


All patients were found to have an elevated Up/c ratio at presentation, which decreased significantly on discontinuation of TDF (Fig. 1b, Table 1). In the 10 patients with >6 months of follow-up after stopping TDF, none had significant proteinuria (Up/c>45 mg/mmol). Six patients had Ua/c measured. The median was 6.3 mg/mmol (range 1.18–32.7 mg/mmol), which did not account for the elevated total urinary protein in these patients and does not support significant glomerular injury.

Twelve of the 22 patients had Ur/c measured, and all were found to have increased excretion, consistent with PT dysfunction (Fig. 1c).

Phosphate excretion and bone metabolism

Of 16 patients with available data, 12 had a plasma phosphate level in the normal range prior to starting TDF. In all these patients, the plasma phosphate level decreased on treatment with TDF. Nineteen of the 22 patients (86.4%) were found to have hypophosphataemia at diagnosis of renal toxicity, and in 18, phosphate levels returned to normal after discontinuation of TDF (Fig. 1d, Table 1). Thirteen of 14 patients (92.9%) were found to have reduced TmP, suggesting that PT dysfunction was at least in part responsible for the low level of serum phosphate.

Of the 19 patients with available data prior to starting TDF, all had a normal alkaline phosphatase (ALP) level. In 17 of these 19 patients (89.5%), the ALP level increased on TDF treatment, with 12 of the 19 patients (63.2%) having ALP rising to greater than the upper limit of the normal range (Table 1). Fourteen of the 22 patients (63.6%) were found to have an elevated ALP, above the normal range, at diagnosis of renal toxicity. Seventeen of 20 patients (85%) (data missing for two patients) had an ALT in the normal range at diagnosis of renal toxicity. In 20 of the 22 patients (90.1%), the ALP level decreased on stopping TDF (Table 1). Twelve patients (54.5%) presented with bone pain and osteomalacia was confirmed on an isotope bone scan in seven patients, all of whom had an elevated ALP at presentation.

Parathyroid hormone (PTH) levels were within normal limits in 13 of 17 (76%) of patients tested, with a median of 4.7 pmol/L (range 1.8–22.9 pmol/L). Four patients had elevated PTH levels. 25 OH vitamin D levels were above the lower limit of the referenced normal range for the local laboratory (15–120 nmol/L) in all 15 patients tested, with a median of 73.4 nmol/L (range 19.2–183.3 nmol/L). 25 OH vitamin D results were available in six of the seven patients with confirmed osteomalacia, with a median of 65.8 nmol/L (range 19.2–121.8 nmol/L); of these, two had levels below 35 nmol/L.


In eight of nine patients tested, glycosuria was present (in the absence of diabetes), in addition to phosphate wasting and tubular proteinuria, confirming a diagnosis of FS.

Renal biopsy

Renal biopsies were performed in three patients. The major feature on light microscopy was acute and chronic tubular damage with flattening of the epithelium and interstitial oedema but with no associated inflammatory infiltrate. On electron microscopy, degenerative tubules had evidence of misshapen and occasional giant mitochondria. Two patients had additional evidence of glomerular immunoglobulin A (IgA) deposition in mesangial areas, suggesting co-existing IgA nephropathy. In both these patients, Up/c and eGFR improved substantially on stopping TDF. Neither of these patients had detectable proteinuria on recent follow-up. Nevertheless, it is possible that pre-existing glomerular disease was a factor in the development of toxicity in these patients. (Both had serum creatinine just at the upper end of normal prior to starting TDF.)


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References

We identified 22 patients who we assessed as having developed TDF renal toxicity. Presentation was with elevated serum creatinine, isolated proteinuria, FS or osteomalacia. The majority had an elevated creatinine (82%) and had experienced a fall in eGFR to <60 mL/min/1.73 m2 (64%) at presentation. A minority presented with isolated proteinuria with neither an elevated creatinine nor a fall in eGFR to <60 mL/min/1.73 m2. All patients had proteinuria at presentation. The findings (in those patients tested) of tubular proteinuria, reduced TmP, and glycosuria were all consistent with the PT being the site of toxicity. The diagnosis of TDF renal toxicity is supported by the improvement of renal function abnormalities following discontinuation of the drug. All patients experienced substantial falls in Up/c and all but three significant increases in eGFR, suggesting that TDF renal toxicity is largely reversible. Although 19% of patients continued to have clinical proteinuria (>45 mg/mmol) and 28% an eGFR of <60 mL/min/1.73 m2 at their last follow-up visit, renal function may or may not have continued to improve on further follow-up. We are unable to estimate the level of permanent renal impairment as the follow-up time varied among patients.

Up/c is a convenient screening test for renal disease. It only requires a spot sample of urine, strongly correlates with 24-hour collections for proteinuria [10,11], and is a reliable predictor of progression of renal disease [12]. All of our patients had an elevated Up/c at presentation, suggesting that it is a sensitive screening test for TDF-associated renal toxicity. Although a useful screening test, Up/c does not distinguish between proteinuria of glomerular or tubular origin. In all patients tested, Ur/c was raised, consistent with tubular proteinuria. Testing for renal tubular dysfunction has not been reported in large randomized control trials of TDF, although smaller studies suggest that abnormalities of renal tubular function may be common [13].

The incidence of clinically significant renal toxicity in randomized control trials of TDF is very low [1,14] and appears no different from that for the comparator ART regimen. In addition, phase III trials of TDF in ART-naïve patients have reported no difference in the incidence of hypophosphataemia or of proteinuria (as measured by urine dipstick testing) out to 144 weeks of follow-up [14]. However, urine dipstick testing mainly detects albuminuria and not tubular proteinuria. Although we did observe increases in serum creatinine in the majority of patients (with corresponding decreases in GFR), Up/c was raised in all patients, suggesting that this is a better screening test for renal toxicity. Small rises in serum creatinine can be caused by tubular toxicity, as a result of impaired secretion of creatinine by the PT.

A striking feature of many cases was the severity of bone disease. In these cases, osteomalacia was detected by isotope bone scan and was the major clinical feature. Patients with HIV infection have a higher incidence of osteopaenia and osteoporosis [15]. In this series, seven patients developed severe metabolic bone disease accompanied by hypophosphataemia and increased urinary wasting of phosphate. Hypophosphataemia is common in patients with HIV infection and is thought to be multifactorial, although interestingly Badiou et al. [16] have reported significant urinary phosphate wasting in patients with HIV infection. The PT has been suggested to be a reservoir for viral infection and this, coupled with drug-induced tubular toxicity, may explain phosphate wasting. Alternatively, HIV infection and TDF itself may have direct effects on bone metabolism or influence phosphate metabolism through circulating factors such as phosphatonins. This potential link between the increased risk of osteopaenia in patients with HIV infection and renal regulation of phosphate balance deserves further study. Classical FS is generally considered to be a rare disease; however, milder forms of PT disease are likely to be far more prevalent, particularly in the setting of drug toxicity, as the PT is a major site for the excretion of xenobiotics from the body. It has been suggested that, in milder cases of PT dysfunction, LMWPs that are normally reabsorbed in the PT may be presented to more distal parts of the tubule, where they may have toxic effects [17]. HIV itself can cause elevated levels of LMWPs in the urine [18]; however, in our patients the clear temporal relationship between Up/c and TDF usage, combined with the generally low levels of viral load, suggests that the drug, rather than the virus, was responsible for the PT toxicity.

The mechanism of TDF-associated renal toxicity is currently unknown. Other nucleoside reverse transcriptase inhibitors (NRTIs), such as didanosine, are known to be toxic to mitochondria; however, both in vivo and in vitro tests have suggested that TDF is not [19,20]. Of the three patients of ours who underwent renal biopsy, two revealed abnormalities in mitochondrial appearance on electron microscopy, but no formal measures of mitochondrial function were performed. The majority of our patients were being treated with a ritonavir-boosted PI regimen, which increases the potential for a drug–drug interaction, perhaps increasing the risk of renal tubular toxicity. Treatment with a boosted PI may be a risk factor for TDF-associated renal toxicity. In case reports of renal FS associated with TDF, treatment with a PI is commonly reported. Although lopinavir/ritonavir has been shown to increase plasma concentrations of TDF [21], the mechanism for a potential increased risk of renal toxicity with TDF when combined with PIs is uncertain. Active renal secretion of TDF across the PT is mediated by the human organic anion transporters 1 and 3, together with efflux by the multidrug resistance protein 4 [22]. Recent in vitro data suggest that PIs at plasma concentrations achieved in vivo have limited potential to interfere with these tubular transport mechanisms [23]. Further data are needed to explore the interaction with PIs and TDF. However, it should be emphasized that TDF-associated renal toxicity may occur in the absence of concurrent treatment with a boosted PI, and monitoring for renal toxicity should occur in all patients treated with TDF.

Renal toxicity remains a concern in some patients treated with TDF and may have variable clinical presentation. Our investigations support PT toxicity as the common pathogenic mechanism. Although TDF-associated renal toxicity appears largely reversible, a small number of patients continued to have impaired renal function in terms of GFR more than 6 months after TDF discontinuation. Renal follow-up of these patients is important. We recommend that all patients on TDF should be regularly screened for renal toxicity with Up/c, and if detected further tests for PT dysfunction should be considered. Reliance on measurement on eGFR alone is not recommended as some patients may present with isolated proteinuria and the accuracy of eGFR above 60 mL/min is low. More research is required to identify: firstly, those patients who are most at risk of TDF renal toxicity; secondly, the incidence of proteinuria in patients treated with TDF and other antiretroviral drugs; and finally, the underlying cellular mechanisms of TDF toxicity. When screening any drug for renal toxicity, it is important that appropriate tests are carried out, to exclude both tubular and glomerular involvement.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. References
  • 1
    Pozniak AL, Gallant JE, Dejesus E et al. Tenofovir disoproxil fumarate, emtricitabine, and efavirenz vs. fixed-dose zidovudine/lamivudine and efavirenz in antiretroviral-naïve patients: virologic, immunologic, and morphologic changes – a 96-week analysis. J Acquir Immune Defic Syndr 2006; 43: 535540.
  • 2
    Izzedine H, Hulot JS, Vittecoq D et al. Long-term renal safety of tenofovir disoproxil fumarate in antiretroviral-naïve HIV-1-infected patients. Data from a double-blind randomized active-controlled multicentre study. Nephrol Dial Transplant 2005; 20: 743746.
  • 3
    Gallant JE, Parish MA, Keruly JC, Moore RD. Changes in renal function associated with tenofovir disoproxil fumarate treatment, compared with nucleoside reverse-transcriptase inhibitor treatment. Clin Infect Dis 2005; 41: 570571.
  • 4
    Winston A, Amin J, Mallon PWG et al, Minor changes in calculated creatinine clearance and anion-gap are associated with tenofovir disoproxil fumarate-containing highly active antiretroviral therapy. HIV Med 2006; 7: 105111.
  • 5
    Van Rompay KK, Brignolo LL, Meyer DJ et al. Biological effects of short-term or prolonged administration of 9-[2-(phosphonomethoxy)propyl]adenine (tenofovir) to newborn and infant rhesus macaques. Antimicrob Agents Chemother 2004; 48: 14691487.
  • 6
    Peyriere H, Reynes J, Rouanet I et al. Renal tubular dysfunction associated with tenofovir therapy: report of 7 cases. J Acquir Immune Defic Syndr 2004; 35: 269273.
  • 7
    Izzedine H, Isnard-Bagnis C, Hulot JS et al. Renal safety of tenofovir in HIV treatment-experienced patients. AIDS 2004; 18: 10741076.
  • 8
    Parsonage MJ, Wilkins EGL, Snowden N, Issa BG, Savage MW. The development of hypophosphataemic osteomalacia with myopathy in two patients with HIV infection receiving tenofovir therapy. HIV Med 2005; 6: 341346.
  • 9
    Levey AS, Greene T, Kusek JW, Beck GJ, Group MS. A simplified equation to predict glomerular filtration rate from serum creatinine. J Am Soc Nephrol 2000; 11: 155A.
  • 10
    Ginsberg JM, Chang BS, Matarese RA, Garella S. Use of single voided urine samples to estimate quantitative proteinuria. N Engl J Med 1983; 309: 15431546.
  • 11
    Lane C, Brown M, Dunsmuir W, Kelly J, Mangos G. Can spot urine protein/creatinine ratio replace 24hr urine protein in usual clinical nephrology? Nephrology (Carlton) 2006; 11: 245249.
  • 12
    Ruggenenti P, Gaspari F, Perna A, Remuzzi G. Cross sectional longitudinal study of spot morning urine protein: creatinine ratio, 24 hour urine protein excretion rate, glomerular filtration rate, and end stage renal failure in chronic renal disease in patients without diabetes. BMJ 1998; 316: 504509.
  • 13
    Mauss S, Berger F, Schmutz G. Antiretroviral therapy with tenofovir is associated with mild renal dysfunction. AIDS 2005; 19: 9395.
  • 14
    Gallant JE, Staszewski S, Pozniak AL et al. & 903 Study Group. Efficacy and safety of tenofovir DF vs. stavudine in combination therapy in antiretroviral-naïve patients: a 3-year randomized trial. JAMA 2004; 292: 191201.
  • 15
    Brown TT, Quarqish RB. Antiretroviral therapy and the prevalence of osteopenia and osteoporosis: a meta-analytic review. AIDS 2006; 20: 21652174.
  • 16
    Badiou S, De Boever CM, Terrier N, Baillat V, Cristol JP, Reynes J. Is tenofovir involved in hypophosphatemia and decrease of tubular phosphate reabsorption in HIV-positive adults? J Infect 2006; 52: 335338.
  • 17
    Norden AG, Lapsley M, Lee PJ et al. Glomerular protein sieving and implications for renal failure in Fanconi Syndrome. Kidney Int 2001; 60: 18851892.
  • 18
    Kabanda A, Vandercam B, Lauwerys R, Van Ypersele de Strihou C. Low molecular weight proteinuria in human immunodeficiency virus-infected patients. Am J Kidney Dis 1996; 27: 803808.
  • 19
    Biesecker G, Karimi S, Desjardins J et al. Evaluation of mitochondrial DNA content and enzyme levels in tenofovir DF-treated rats, rhesus monkeys and woodchucks. Antiviral Res 2003; 58: 217225.
  • 20
    Birkus G, Hitchcock MJ, Cihlar T. Assessment of mitochondrial toxicity in human cells treated with tenofovir: comparison with other nucleoside reverse transcriptase inhibitors. Antimicrob Agents Chemother 2002; 46: 716723.
  • 21
    Kearney BP, Mathias A, Mittan A, Sayre J, Ebrahimi R, Cheng AK. Pharmacokinetics and safety of tenofovir disoproxil fumarate on coadministration with lopinavir/ritonavir. J Acquir Immune Defic Syndr 2006; 43: 278283.
  • 22
    Ray AS, Cihlar T, Robinson KL et al. Mechanism of active renal tubular efflux of tenofovir. Antimicrob Agents Chemother 2006; 50: 32973304.
  • 23
    Cihlar T, Ray AS, Laflamme G et al. Molecular assessment of the potential for renal drug interactions between tenofovir and HIV protease inhibitors. Antivir Ther 2007; 12: 267272.