icon-folder.gif   Conference Reports for NATAP  
  12th Conference on Retroviruses and Opportunistic Infections (CROI)
Feb 22-25, 2005
Boston, MA
Back greyarrowrt.gif
Tenofovir and Kidney Toxicity: What do the studies show?

Judith Aberg, MD, Bellevue Hospital, New York University Medical Center

"……Renal impairment appears to be underestimated among the HIV infected population…… HIV and kidney disease has not been well studied. Members of the AIDS Clinical Trails Group Renal (kidney) Subcommittee recently developed "Guidelines on the Management of Chronic Kidney Disease in HIV-Infected Patients" for the Infectious Disease Society of America HIV Medicine Association and are expected to be published in the next few months….."

Acute kidney toxicity has been described with certain antiviral agents such as adefovir and cidofovir. Both these drugs can cause toxicity to the proximal renal tubule and patients may develop Fanconi Syndrome characterized by presence of protein and glucose in the urine plus a loss of phosphate, potassium and bicarbonate in the blood which may lead to acidosis and a reduction in the glomerular function and an increase in creatinine. This means the kidney cannot work as well as it should to clear waste and remove fluids. I recommend patients go to the National Kidney Foundation web site and the site’s kidney learning system (http://www.kidney.org/kls/public/howkidneyswrk.cfm)

Previous HIV treatment studies showed that approximately one third of patients taking adefovir at 120 mg daily developed some kidney dysfunction. Fortunately for most, the damage was mild and reversible. Because of the kidney toxicity, adefovir was not approved for use as a HIV medicine (antiretroviral therapy). Adefovir is now currently approved at very low doses (10 mg daily) for the treatment of Hepatitis B. Tenofovir is a similar antiviral agent that is approved for treatment of HIV and there have been case reports of kidney damage in HIV infected patients taking tenofovir. There were 3 abstracts presented at CROI 2005 which explored the association of kidney toxicity and tenofovir. Jules Levin has previously reviewed these abstracts and I refer you to the NATAP web site for full details of these studies. I will list the 3 abstracts and then briefly discuss some of the problems in sorting out what is and isn’t related to tenofovir.

Beyond Serum Creatinine: Identification of Renal Insufficiency Using GFR: Implications for Clinical Research and Care


Improving the Clinical Assessment of Renal Function in HIV


Decline in Renal Function Associated with Tenofovir DF (TDF) Compared to Nucleoside Reverse Transcriptase Inhibitor (NRTI) Treatment


Tenofovir has not been associated with higher rates of kidney failure in the initial treatment studies compared with the more commonly prescribed antiretroviral agents, though study subjects with any pre-existing kidney dysfunction were not allowed to be enrolled in these studies. However, there are several case reports and case series of patients developing Fanconi Syndrome after taking tenofovir-containing regimens. These cases report that the syndrome resolved or improved after tenofovir was stopped. These reports suggest that kidney toxicity is more common with those that already have some kidney dysfunction, are on other kidney toxic medications and/or have advanced HV disease. The time to development of kidney dysfunction associated with tenofovir-containing regimens varies from weeks to more than one year; therefore it is difficult to say whether the kidney dysfunction truly is related to the tenofovir. This also makes it difficult to predict when/if tenofovir toxicity will develop and how best to screen for tenofovir associated kidney toxicity.



Poster 818: Chronic Kidney Disease and the Use of HAART presented by Ron Reisler and colleagues.


Background: HIV-infected individuals with worsening renal function have an increasing risk of progression to AIDS and death. The prevalence and predictors of chronic kidney disease (CKD) in HIV-infected individuals have not been well defined. Methods: We examined the prevalence and predictors of CKD in a cross sectional analysis of 1470 HIV-infected and uninfected men enrolled in the Multicenter AIDS Cohort Study, using data collected from men who had serum creatinine measurements obtained at follow-up visit 40 (October 2003 to March 2004). Subjects’ glomerular filtration rates (GFR) were estimated using the simplified modification of diet in renal disease (MDRD) equation, which incorporates age and race; this then allowed CKD staging using the accepted NIH/NIDDK Kidney Disease Outcomes Quality Initiative standards. We performed multivariate logistic regression analyses to assess relationships between HIV status, HAART use (while controlling for pre-HAART HIV-1 RNA level, pre-HAART CD4 cell count, hypertension, diabetes mellitus, smoking, and body mass index), and the presence of more advanced CKD stage. Associations between the recent use of tenofovir and CKD stage were also investigated. Results: HAART-using infected men were more likely to have GFR < 60 mL/min/1.73 m2 (stages 3 to 5 CKD) compared to uninfected men (adjusted odds ratio (AOR) = 2.5, 95% CI = 1.4 to 4.5) but did not differ for having 60 < GFR < 89 mL/min/1.73 m2 (stage 2 CKD). More HIV-infected HAART users had GFR < 90 mL/min/1.73 m2 than HIV-infected HAART-naïve men (AOR = 1.7, 95% CI = 1.1 to 2.7). Pre-HAART HIV-1 RNA level and pre-HAART CD4 cell count were not associated with GFR. However, restricting the analysis to HAART users, use of tenofovir was associated with lower GFR than in non-users, yielding an OR = 1.7 (95% CI = 1.1 to 2.5) for stage 2 CKD and an OR = 2.0 (95% CI = 0.8 to 4.9) for stages 3-5 CKD. Length of time since initiation of HAART did not diminish these associations.Conclusions: Advanced CKD is significantly more common among HIV-infected men on HAART as compared to HIV-uninfected men. Decreased GFR in HAART users was not associated with pre-HAART HIV-1 RNA level or pre-HAART CD4 cell count. The association seen between tenofovir and lower GFR warrants additional investigation.

Poster 819

Beyond Serum Creatinine: Identification of Renal Insufficiency Using Glomerular Filtration: Implications for Clinical Research and Care presented by Stephen Becker and colleagues. Updated results since abstract submission available on Jules Levin review.

Background: Recent reports of renal insufficiency associated with the nucleoside reverse transcriptase inhibitors (NRTI) have included elevations in serum creatinine, proximal tubular dysfunction, nephrogenic diabetes insipidus, renal failure, Fanconi-like syndrome, and death. Appropriate methods for identification of patients at risk for severe renal injury are needed for monitoring and clinical dosage modification.Methods: Using an antiretroviral known to cause renal impairment (tenofovir disproxil fumarate [TDF]), we evaluated the rate of grade 1 and grade 4 renal insufficiency using serum creatinine and a method to more precisely evaluate glomerular filtration. Literature suggests that even minor changes in serum creatinine result in substantial decreases in glomerular filtration. Moreover, serum creatinine is influenced by other factors including age, race, sex, and body mass. According to the National Kidney Foundation, glomerular filtration is the "best overall indicator of the level of kidney function." This is especially true in patients with co-morbid conditions. Multivariable Cox models were fit to consider age, race, study site, sex, route of HIV infection, AIDS, baseline viral load and CD4, baseline renal function, concurrent use of other nephrotoxic drugs, history of renal disease, hypertension, and diabetes as predictors of renal insufficiency in patients treated with TDF. Results: A total of 1298 patients were initiated first TDF on or after consent to CHORUS. Using a combined endpoint of serum creatinine and clinical events, 22 patients (1.7%) experienced a grade-1 event, 1 patient (< 0.1%) experienced a grade 4 event. Using a combined endpoint of glomerular filtration and clinical events, 128 patients (9.9%) experienced a grade-3 event, 7 patients (< 0.1%) experienced a grade-4 event. Multivariable analysis identified past history of renal disease (hazard ratio = 4.9; 95% CI = 2.7 to 8.8), abnormal baseline glomerular filtration (18.9; 9.2 to 39.1), hypertension (1.6; 1.1 to 2.3), and concurrent use of other nephrotoxic medications (2.7; 1.9 to 3.9) as important predictors of renal insufficiency. Conclusions: The elevations in serum creatinine observed in CHORUS are consistent with those seen in other populations, observational or otherwise. The proportion of patients with extreme grade 4+ glomerular filtration values is comparable to results seen for grade 4+ serum creatinine in other studies. However, nearly 10% of our patients have renal dysfunction using the grade 3+ glomerular filtration criterion. The National Kidney Foundation defines this level as chronic kidney disease. Glomerular filtration should be considered for the evaluation of renal insufficiency in HIV.

Poster 820

Decline In renal Failure Associated with Tenofovir DF Compared with Nucleoside Reverse Transcriptase Inhibitor Treatment By Joel Gallant, et al and subsequently published in Clinical Infectious Diseases, April 15 2005;40:1194-1198.

Background: Tenofovir DF (TDF) is a nucleotide analog reverse transcriptase inhibitor (NRTI) used for the treatment of HIV disease. Despite demonstrated renal safety in several clinical trials, case reports, and observational data suggest that it may be associated with nephrotoxicity. Methods: We analyzed data from a large prospective observational cohort, comparing all patients who initiated TDF (n = 344) or an alternative NRTI (n = 314) after January 1, 2001 and who had at least 2 blood samples drawn to test for serum creatinine (Cr) within 90 days of initiation. We assessed change and percentage change in calculated creatinine clearance (CLCr) over the study period for each subject. Baseline CLCr was calculated using the Cockcroft-Gault equation using the average of the 2 serum Cr measurements obtained closest to the start of therapy. Results: TDF recipients had greater increases in Cr and greater absolute and percentage declines in CLCr than those in the NRTI group, although there was no difference in rates of discontinuation coincident with maximum decline in CLCr. Median baseline Cr and CLCr were 0.8 mg/dL and 117.5 mL/min, respectively, with no differences between groups. Median Cr change was +0.15 and +0.10 in the TDF and NRTI groups, respectively (p = 0.01). Median CLCr change was —13.35 and —7.5 mL/min (p = 0.005). The changes were apparent by 90 days of therapy and persisted over the entire year. The percentage change in CLCr was also associated with longer duration of treatment with TDF or the NRTI, diabetes, higher baseline Cr, and CD4 count <50 cells/mm3 (p <0.05). In a multivariate analysis only TDF use and lower CD4 count were associated with CLCr decline (p < 0.05), with a trend toward association with lower baseline CLCr and diabetes. In this analysis, hypertension, use of ritonavir-boosted lopinavir or any other concomitant ART, viral load, previous use of adefovir, age, sex, race, and HIV transmission risk group were not associated with CLCr decline. Conclusions: Use of TDF is associated with a modest decline in CLCr, especially in patients with baseline renal insufficiency, diabetes, or low CD4 count. While the differences were statistically significant, they were small in magnitude and of unclear clinical significance. Patients at risk for renal dysfunction should be carefully monitored when taking TDF, and the dosing interval should be adjusted when indicated by a reduced Cr clearance.



First in order to interpret the above studies one needs to be familiar with kidney function, measurement of GFR and methodologies used for these studies. What is GFR? The glomerular filtration rate (GFR) is the best measurement of kidney function. GFR cannot be measured directly. The urinary clearance of an ideal filtration marker, such as inulin, iothalamate or iohexol, is the gold standard for the measurement of GFR. However this is cumbersome to do in clinical practice. Therefore serum levels of endogenous filtration markers, such as serum creatinine, have traditionally been used to estimate GFR. (National Kidney Foundation) The calculated creatinine clearance is a kidney test used to estimate GFR. Exercise may cause increased creatinine clearance. The glomerular filtration rate is substantially increased in pregnancy. Ascorbic acid, ketone bodies (acetoacetate), hydantoin, numerous cephalosporins and glucose may influence creatinine determinations. Trimethoprim, cimetidine, quinine, quinidine, procainamide reduce creatinine excretion. Icteric samples, lipemia, and hemolysis may interfere with determination of creatinine. (Clinician's Guide to Laboratory Medicine, Lexi-Comp On- Line) Since tubular secretion of creatinine is fractionally more important in progressing renal failure, the creatinine clearance overestimates GFR with high serum creatinine levels. While ingestion of meats may cause some increase in creatinine excretion, in practice this seems to make little difference. Intraindividual variation in creatinine clearance is about 15%. Males excrete more creatinine and have slightly higher clearance than females. Because of the exponential rise in serum creatinine concentration with decline of GFR, a 25% increase in serum creatinine actually represents a substantial diminution of GFR. When muscle mass and kidney function diminish in parallel with advancing age, an elderly woman with perceived normal creatinine concentration may have a GFR only 30% that of a young adult. Normal Adult male creatinine clearance is 85-125 mL/minute/1.73 m2 (SI: 1.42-2.08 mL/s/1.73 m2) and for females is 75-115 mL/minute/1.73 m2 (SI: 1.25-1.92 mL/s/1.73 m2). For each age decade after 40, creatinine clearance decreases 6-7 mL/minute/1.73 m2. (Clinician's Guide to Laboratory Medicine, Lexi-Comp On-Line)

Let’s go back to the 3 abstract above. Two of the abstracts (818 and 819) used a formula called the Modification of Diet in Renal Disease (MDRD) and abstract 820 used the Cockcroft-Gault equation. The MDRD equation uses serum creatinine in combination with age, sex and race to estimate GFR and therefore improves upon several of the limitations with the use of serum creatinine only. The MDRD Study equation has not been validated in children (age <18 years), pregnant women, the elderly (age >70 years), racial or ethnic subgroups other than Caucasians and African Americans, in individuals with normal kidney function who are at increased risk for CKD, or in normal individuals. This equation was initially validated in whites with known kidney disease but without diabetes. Later studies validated its use among African Americans with kidney disease excluding diabetes. The MDRD formula is adjusted for age, race and gender using the serum creatinine. There is an extended formula which utilizes blood urea nitrogen and albumin laboratory values. The Cockcroft-Gault equation is most frequently used by clinicians in determining adjustments to doses of medications based upon the patient’s body mass index, age and gender. This formula was derived based upon 24 hour urine creatinine clearances collected on 270 hospitalized white men. The adjustment recommended for women was based upon a best guess estimate accounting for the change in muscle mass between genders. So, this formula has its problems as the MDRD does. It is recommended to use a 24 hour urine collection in age extremes, severe malnutrition or obesity, prior to dosing kidney toxic drugs, pregnancy and other disease states. No studies have been performed in a HIV-infected population to validate any calculation. Other problems with these calculations include the variability of creatinine based upon state of hydration (creatinine is increased when dehydrated) and for the C-G equation, body weight differences can significantly alter the results.

I decided to use the formulas available on the National Kidney Foundation (http://www.kidney.org/kls/patients/gfr_calculator.cfm) and the Nephron Information Center (http://nephron.com/ ) sites . I used normal laboratory values from my hospital and my age/BMI for Patient X. According to the MDRD formula, Patient X had mild kidney dysfunction. The MDRD calculated to be 63 both by the regular and extended. I then used the C-G equation which calculated it to be anywhere from 66-78 depending on what weight I chose to use. Some recommend that one should use ideal body weight while others suggest using actual body weight. The baseline creatinine clearance was 73 (normal for the age). I chose one to be 15 pounds less than usual weight (not uncommon with untreated HIV) and the creatinine clearance was 66. I then chose to add 10 lbs which is not uncommon after HIV infected persons begin therapy. Patient X’s creatinine clearance increased to 78. My point is that these calculations can be quite variable depending on BMI and creatinine which can vary over time depending on nutritional status, medications, hydration status and other factors.

Another difficulty with these studies is that the change in creatinine clearance was statistically significant although Dr. Gallant does note that this decrease may not be clinically significant. Of note, in his study, there was an association of decreased creatinine clearance in subjects with lower CD4+ T-cell counts. This would not be surprising in that advanced disease is associated with lower serum creatinine as seen in wasted states. Other HIV treatment studies have demonstrated that individuals typically gain weight after starting HAART therefore one would expect and improvement (increase) in serum creatinine so therefore the calculated creatinine clearance should decrease. Unfortunately, we do not have the data comparing the change in body mass index within the group or between the two groups.

Another important variable we do not have is what the baseline kidney function is prior to HIV infection. Sharon Riddler has previously reported the change in lipids from HIV infection itself in a cohort of men who had lipid levels prior to acquisition of HIV. She then reported the changes after HAART. (Riddler SA, Smit E, Cole SR, et al. Impact of HIV infection and HAART on serum lipids in men. JAMA 2003; 289:2978-82.) It would be interesting to evaluate the kidney function of a cohort of similar individuals.

Do I think tenofovir is associated with kidney toxicity? Based upon these 3 abstracts and other reports, I agree with the conclusions that tenofovir appears to contribute to kidney dysfunction in the setting of pre-existing kidney impairment whether it be secondary to advancing age, advanced HIV disease, other disease states and/or use of concomitant nephrotoxic medications. However, one should not over react to these findings either. Tenofovir is a potent antiretroviral agent that is generally well tolerated. At this juncture, one cannot predict per se who will develop kidney toxicity but it appears unlikely in individuals who do have normal kidney function and do not have other risk factors for kidney disease. In addition, it does appear that this decline in kidney function is reversible when one withdraws tenofovir or other contributing factors. Also, some of this decline may be artifactual as it may be secondary to improvement in body weight. It is also important to remember that it was determined that the dose of tenofovir needs to be adjusted for kidney impairment. These studies included subjects who were given full dose of tenofovir as that information was not available at the time they initiated therapy with tenofovir. I also agree that "time will tell". Clearly in the studies with adefovir, we learned that it was not only dose dependent but also duration on drug so that it was overall cumulative exposure to adefovir that was associated with kidney toxicity. This may account for why the case reports of tenofovir associated kidney toxicity have been variable as to the time before development of toxicity.

These abstracts all share a similar theme and one that cannot be ignored. Renal impairment appears to be underestimated among the HIV infected population especially when clinicians rely upon serum creatinine only as the marker of kidney function. Studies are needed to validate these equations or determine modifications that accurately reflect the glomerular filtration rate in the HIV infected population. HIV and kidney disease has not been well studied. Members of the AIDS Clinical Trails Group Renal (kidney) Subcommittee recently developed "Guidelines on the Management of Chronic Kidney Disease in HIV-Infected Patients" for the Infectious Disease Society of America HIV Medicine Association and are expected to be published in the next few months. These guidelines should prove useful for assisting clinicians in the diagnosis and management as well as serving to remind us of the importance of screening and early detection of kidney disease.