Impairment of Kidney Function among US Hispanics on Tenofovir

Tenofovir disoproxil fumarate (TDF), a nucleotide analogue reverse-transcriptase inhibitor is one of the most widely used antiretroviral medications and was approved by the US Food and Drug Administration for pre-exposure prophylaxis of primary human immunodeficiency virus (HIV) infection [1]. Although there is evidence that TDF impairs the renal function [2,3], previous studies have not reported the concurrent assessment of kidney function among TDF use in Hispanics. Hispanics comprise the largest ethnic minority group in the US and experience disproportionate burden of diabetes and kidney disease. For this reason, as part of a Phase IV safety study on the use of TDF among HIV-infected Hispanic men who have sex with men, we conducted a pharmacovigilance retrospective study in El Paso, Texas, a predominantly Hispanic community. The specific objective of the study was to assess the effect that TDF has on estimated glomerular filtration rates among US Hispanics.


Introduction
Tenofovir disoproxil fumarate (TDF), a nucleotide analogue reverse-transcriptase inhibitor is one of the most widely used antiretroviral medications and was approved by the US Food and Drug Administration for pre-exposure prophylaxis of primary human immunodeficiency virus (HIV) infection [1]. Although there is evidence that TDF impairs the renal function [2,3], previous studies have not reported the concurrent assessment of kidney function among TDF use in Hispanics. Hispanics comprise the largest ethnic minority group in the US and experience disproportionate burden of diabetes and kidney disease. For this reason, as part of a Phase IV safety study on the use of TDF among HIV-infected Hispanic men who have sex with men, we conducted a pharmacovigilance retrospective study in El Paso, Texas, a predominantly Hispanic community. The specific objective of the study was to assess the effect that TDF has on estimated glomerular filtration rates among US Hispanics.

Population and Methods
This study was approved by the University of Texas at El Paso Institutional Review Board (80204-1); only de-identified data was made available to the investigators.
The study population consisted of HIV patients who took part in the above-mentioned Phase IV study and received care at La Fe Care, a Federally Qualified Health Center, which provides medical services to underserved communities throughout the region of El Paso, Texas. Study personnel abstracted data on demographics, laboratory assessments and start dates for TDF medications. In total, 117 male patients on TDF were evaluated; 11of these were excluded due to having less than 6 months of TDF therapy. The analytic study was comprised of 510 observations from106 individuals receiving TDF from 2004 to 2006.
We estimated the glomerular filtration rate (eGFR) at each visit using a reduced version of the Modification of Diet in Renal Disease Study (MDRD) equation:eGFR = 175×(serum creatinine) -1.154 ×(age in years) -0.203 [4]. We also used the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) [5] formula to estimate eGFR.
The authors used repeated measures mixed modeling to estimate the changes in eGFR for each patient and across the cohort. These models adjusted for age at baseline, and reported diabetes, as well as the presence of hepatitis B surface antigen (HBsAg). Statistical significance was defined as two-sided P-values<0.05. We conducted all statistical analyses in SAS version 9.3 (SAS Institute, Cary, NC).

Results
A total of 106 patients were followed anywhere from 6 months to 36 months (typically for 16 months, interquartile range=11-25). Almost forty-percent of participants were followed for at least 2 years. Participants were typically middle aged (median=42 years, interquartile range=37-47). Ten percent had been diagnosed with diabetes mellitus, while 8.5% were chronic carriers of HBsAg (Table 1). None of the participants was concomitantly receiving other nephrotoxic drugs commonly prescribed to HIV patients such as amphotericin B, cidofivir, foscarnet, pentamidine [5], and only one patient received acyclovir during the study period.
At baseline, the typical participant of this study population had an eGFR MDRD of 91.4 (SD=22.5), and based of CKDEPI a mean eGFR of 97.3 (SD=18.6). Depending on the equation used, there were 8 subjects (7.6%) using CKDEPI or 9 subjects (8.5%) using MDRD with eGFR below 60 mL/min, one cut-off value for screening of chronic kidney disease. Table 2

Discussion
In our cohort of Hispanic male HIV patients on TDF, a clinically and statistically significant reduction in renal function as measured by eGFR was observed. For perspective of these findings, Table 2 also shows that the effect of having a medical diagnosis of diabetes mellitus in our study population amounted to a 10-13 mL/min difference, a magnitude of decrease that can occur in two years on TDF if the observed relation were proven to be causal. To our knowledge, no previous published studies have evaluated the effects of TDF on renal function among Hispanic patients. In a retrospective evaluation, Rawlings et al. [6] reported the impact of TDF and comorbidities in the development of renal impairment in 323 patients, including 96 African Americans and 19 Hispanics. Such a small sample of Hispanics makes it difficult to draw any meaningful conclusions. Therefore; our study is the first to specifically describe the change in levels of eGFR among Hispanics on TDF.
A number of comprehensive reviews of published studies have given us detailed information regarding patterns and mechanisms of TDFinduced renal dysfunction [2,3]. In 2010, Cooper et al. [2] conducted a systematic review and meta-analysis of the effects of TDF on renal function. Based on their criteria they included 17 eligible studies, but only six studies reported changes in MDRD-eGFR (including 2,334 participants). The authors found a mean difference of 2.6 mL/min/1.73 m 2 , although this difference was not statistically significant, and there was a high degree of heterogeneity across the studies. Demographic data for the six studies included in that meta-analysis appears to show samples that consisted only of White and African-American participants.
A more recent review by Hall et al. [3] points to the occurrence of TDF-associated nephrotoxicity as a common reason for HIV-related referral to renal services. Osteomalacia and hypophosphatemia are conditions considered to be associated with TDF use. They propose that mitochondrial toxicity may be an important mechanism underlying the TDF-associated kidney impairment and endorse the Infectious Disease Society of America guidelines for monitoring renal toxicity in persons on TDF: requiring measurements of eGFR, serum phosphate, and urine analysis at least every 6 months in patients with eGFR<90 ml/min/1.73 m 2 , the use of other drugs eliminated through renal secretions, other co-morbid diseases such as diabetes and hypertension, and use of alternative protease inhibitor such as a ritonavir-boosted regimen [7]. Our findings suggest that these guidelines would be applicable to Hispanic patients on TDF.
Both reviews concluded that although the clinical significance was modest, TDF-containing ART regimens were associated with a significant loss of kidney function. One more recent observational study described changes of eGFR by each year on TDF and found changes consistent with our estimates: the mean losses observed during the first year were -3.0 and -4.0 ml/min/1.73 m 2 [8]. A separate study reports that direct measurements of TDF in plasma are correlated with decreased levels of eGFR [9]. The authors highlight the need for close monitoring of kidney function for patients on TDF. More compelling evidence from yet another observational study among persons without previous impairment of kidney function showed that improvement of kidney function followed the discontinuation of TDF and ritonavir [10]. Finally, the largest observational study found that anti-retroviral therapies that include TDF had a higher risk of chronic kidney disease [11]. Yet recent studies point out that the observed effect on eGFR may reflect the effect of TDF on tubular rather than glomerular dysfunction [12].

Limitations
One needs to consider at least three major limitations to this study when interpreting its findings: this was an observational, retrospective chart review cohort study, lacking a concurrent comparison group, and therefore potentially affected by unmeasured confounders. Data on non-nephrotoxic drugs and dietary components that have been identified as inhibitors of organic anion transporters was not collected; markers of tubule dysfunction such as glucose and phosphate in urine were not collected, nor were data on proteinuria. Also the history of blood pressure was sparse. In addition, the study sample size was relatively modest, and although observations in this study total over 500, they are based on only 106 study participants. In addition, the follow-up period was relatively short in comparison to other studies. Although none of our participants was taking any other potentially nephrotoxic drugs [6], one cannot rule out rule out the possibility that other concomitant medications may have played a role.

Conclusions
In our retrospective review of HIV infected US Hispanic men on TDF, we found a modest but statistically significant reduction of eGFR (based on MDRD a decrease of 0.5 mL/min/1.73 m 2 per month or 0.4 mL/min per month using the CKDEPI). The effect of TDF use among Hispanics is similar to that reported among other ethnic groups. Our data support the notion that guidelines for routine monitoring of renal function should be applied to Hispanics patients on TDF therapy. In our opinion, the clinical significance of this reduction in eGFR may vary from patient to patient, but should be considered in Hispanic patients receiving TDF therapy.