Plasma Tenofovir Concentrations and Proximal Tubular Dysfunction in HIV-Infected Adults Receiving Tenofovir in Thailand

Background: HIV-infected Asians may be at risk for tenofovir toxicity due to their relatively low body weight (BW). We assessed the prevalence of proximal tubular dysfunction (PRTD) and risk associated with PRTD in HIV-infected adults receiving tenofovir in Thailand. Methods: A cross-sectional study in HIV-infected adults (≥ 18years) treated with tenofovir for >1 year. Twenty-four- hour urine samples were collected to assess PRTD. PRTD was defined as the presence of >2 of the following criteria: Mid-dose tenofovir plasma concentrations were determined and concentrations >160 ng/mL were used as a cut-off for assessing risk of PRTD. Results: 351 subjects (52% males) with median age of 40.2 years, BW of 58.9 kg, and duration of tenofovir treatment 4.7 years were included. 93% had a HIV-1 RNA <50 copies/mL and 7% were co-infected with HCV. Fifty-four (15.4%) patients were diagnosed with PRTD. In a multivariate analysis, only a mid-dose tenofovir concentration >160ng/mL was associated with PRTD [odds ratio: OR 2.02 (95% CI 1.13-3.66)]; 32 of 54 (59.2%) with PRTD had a tenofovir concentration >160 ng/ml; versus 124 of 297 patients (41.7%) without PRTD (p = 0.02). Predictors of a tenofovir concentration >160ng/mL were BW<55 kg [OR2.32 (95% CI 1.45-3.68)], chronic HCV [OR 2.64 (95% CI 1.13-6.17)], eGFR<90 mL/min/1.73m 2 [OR 2.66 (95% CI 1.64-4.31)], lopinavir/ritonavir [OR 2.47 (95% CI 1.32-4.6)] and PRTD [OR 2.08(95% CI 1.10-3.92)]. Conclusion: Mid-dose tenofovir concentrations >160 ng/mL were independently associated with PRTD. Tubular function should be closely monitored in patients using tenofovir with BW<55 kg, lopinavir/ritonavir use, chronic hepatitis C, or low eGFR.


Introduction
Tenofovir disoproxil fumarate (TDF) is a nucleotide reverse transcriptase inhibitor (N (t) RTI) that has activity against both HIV and hepatitis B virus (HBV). Because of its high efficacy, once-daily dose, and availability within several fixed-dose combination tablets, TDF is one of the most widely used ARV drugs as part of combination antiretroviral treatment (cART). TDF is a prodrug of tenofovir (TFV) that is rapidly hydrolyzed to tenofovir monophosphate in the blood, and then converted to its active tenofovir diphosphate form intracellularly. TFV is primarily eliminated unchanged by a combination of glomerular filtration and proximal tubular secretion [1]. Tenofovir clearance is impaired in patients with abnormal renal function and its accumulation inside the tubular cells may result in further renal damage [2]. Tenofovir is generally well-tolerated but several degrees of renal toxicity, including acute tubular necrosis, nephritis, proximal tubulopathy, proteinuria, diabetes insipidus, hypokalemia and subclinical proximal renal tubular dysfunction (PRTD) has been reported [3][4][5][6][7][8][9][10][11][12][13][14]. In clinical trials, the risk of clinically significant TDF related renal toxicity is relatively low (0.3-2% Fanconi syndrome) [15][16][17]. However, trial participants may not represent HIV patients at large as they are generally in good health and have fulfilled strict inclusion criteria.
Several studies conducted in HIV-infected patients found that older age, presence of metabolic disease and concomitant use of TDF and protease inhibitor are associated with an increased risk of tubular dysfunction [18][19][20][21]. Recently, several pharmacogenomics studies have reported that PRTD in HIV-infected patients receiving tenofovir were associated with certain genetic polymorphisms within genes of several efflux transporters (ABCC2 or ABCC4 or ABCC10) [22][23][24]. Higher plasma concentration of tenofovir has also been recently associated with proximal tubular dysfunction [10,25]. Even though the isolated presentation of tubular dysfunction such as hypophosphatemia, tubular proteinuria, normoglycemic glycosuria, and a decrease in phosphate reabsorption has been associated with tubular cell damage but the clinical significance of isolated or subclinical tubular dysfunction in the short-and long-term remains largely unclear whether these particular patients are at increased risk for developing Fanconi syndrome, progressive alteration of eGFR, osteomalacia or reduced bone mass density. However, PRTD may indeed contribute to the persistent loss of the phosphate and subsequent prematuration of osteopenia/ osteoporosis. Clinical research on tubular function is challenging because of the complexity of kidney exploration. Few studies have assessed prospectively the tubular function of HIV infected patients, and have shown that approximately 6.5-22% of HIV-infected patients may have subclinical PRTD [10,26]. There are limited data on the prevalence of PRTD of HIV infected patients from Asia whom generally have lower body weight than those in resource-rich countries.
In South East Asia, TDF is now widely used for HIV treatment with increased usage of TDF, many patients with low BW and/or pre-existing renal disease may exhibit clinical symptoms of TDF nephrotoxicity later in life. The studies from Thailand, Vietnam, and Japan in HIV-infected patients with median BW of 55-63 kg showed that the high prevalence of TDF-related renal dysfunction (14.4%-19.6%) was strongly associated with low BW [27][28][29][30]. Since TDF-related renal toxicity is linked to an increased plasma concentration, possibly in a dose-dependent manner, Therefore, the Asian population with small body weight is expected to be at higher risk of developing TDF-related renal toxicity [27,28,30,31]. Thus, it is important to understand the relationship between plasma tenofovir concentrations and factors associated with tubular dysfunction among Asian HIV-infected patients. It is also critical to identify subpopulations at risk of developing long-term TDF nephrotoxicity. As a result of this, we assessed the prevalence of PRTD and risk factors associated with proximal tubular dysfunction among HIV-infected patients receiving TDF containing antiretroviral therapy within a large HIV cohort in Thailand.

Study population
HIV infected patients participating in a cohort study (clinicaltrials. gov identifier: NCT 00411983) at the HIV Netherlands Australia Thailand Research collaboration (HIV-NAT), Bangkok, Thailand, were included in this analysis. This sub study was reviewed and registered by the Institutional Review Boards of the Faculty of Medicine, Chulalongkorn University (NCT01138241). All patients provided a written consent. A total of 351 patients treated with TDF-containing cART with a normal estimated glomerular filtration rate (eGFR > 90 ml/min/1.73m 2 ) underwent a 24-hour urinary analysis for assessment of proximal tubular dysfunction. Plasma samples were collected "mid-dose (11-13 hours post dose)" to measure the tenofovir plasma concentration. Patients were excluded if they were currently on nephrotoxic agents such as pentamidine, ganciclovir, amphotericin B, adefovir, cidofovir or indinavir. Patient characteristics were extracted from the HIV-NAT 006 cohort database including sex, age, route of HIV transmission, HBV and/or HCV co-infection, hypertension, diabetes, clinically defined lipodystrophy, ARVs, duration of ARVs and concomitant drugs. Chronic hepatitis B (Hep B) and hepatitis C (Hep C) infection were defined as having a two positive tests (i.e., Hep B S antigen (HBsAg or anti-HCV antibody, respectively) more than 6 months apart.

Assessment of proximal tubular function
Blood and 24-hour urine samples were used to assess the kidney glomerular and tubular parameters. All patients were required to fast for at least 8 hours before blood samples were collected. Written and verbal instructions for urine collection were provided along with the container. Urine was collected over a 24-hour period, which also included the samples from the morning of the blood draw. Various multiparameter blood chemistry and hematology profiles were assessed, such as glucose, creatinine, sodium, potassium, chloride, bicarbonate, phosphorus, total calcium, uric acid, albumin, and total protein. Urine pH, glucose, creatinine, phosphorus, uric acid, albumin, proteins, β2 microglobulin, white and red blood cells in urine was assessed. All parameters for renal proximal tubular function were performed at the laboratory of the Division of Nephrology, Chulalongkorn University. 4. β2 microglobulin more than 1 mg/day or β2 microglobulin/ urinary creatinine more than 0.3 mg/l. 5. Non-diabetic glucosuria (urine glucose > 300 mg/day or positive for urine glucose) with normal glycemic levels (plasma glucose < 100 mg/dl).
The eGFR was calculated using the Modification of Diet in Renal Disease (MDRD) equation [32].

Assessment of tenofovir plasma concentrations
Tenofovir plasma concentrations were measured using a validated reversed-phase high-performance liquid chromatography (HPLC) method with fluorescent detection. This method was validated using the AIDS Clinical Trials Group (ACTG) method validation guidelines over the concentration range of 15 to 1,500 ng/mL [33]. Average accuracy for tenofovir was 99-102% and precision (inter-and intra-assay) was <5% of the coefficient of variation (CV). Samples were assayed at the PHPT-IRD laboratory at Chiang Mai University which participates in the ACTG Pharmacology Quality Control program, which performs standardized inter-laboratory testing twice a year [34].

Statistical analysis
Continuous values are presented as median (interquartile range: IQR) and categorical data as percentages. A multivariable analysis using logistic regression models evaluated factors associated with PRTD. A mid-dose plasma tenofovir concentration cut-off of 160 ng/ mL was used as a threshold to test for an association with PRTD [25]. Variables obtained from the univariate analysis with a p-value < 0.10 were included in the full model. A stepwise selection procedure was used to assess the relative role of the risk factors. P-values of <0.05 were considered significant. All statistical analyses were carried out using STATA version 12.0 (StataCorp, College Station, TX, USA)

Risk factors associated with kidney tubular dysfunction
In a univariate analysis, previous IDV exposure, current ATV exposure, hypertension and a TDF plasma concentration >160 ng/ mL were associated with PRTD (Table 2). However, in the multivariate analysis, only a tenofovir plasma concentration >160 ng/mL remained significantly associated with PRTD (OR 2.02, 95% CI 1.13-3.66, p = 0.02). Low body weight, chronic HCV, lopinavir/r use and duration of TDF were not associated with PRTD.

Discussion
In this study, 15.4% of 351 TDF-treated patients had subclinical PRTD. We found that: 1) patients with PRTD had higher tenofovir plasma concentrations compared to those without (168 ng/mL versus 149 ng/mL); 2) tenofovir plasma concentrations of >160 ng/mL was associated with PRTD; 3) BW< 55 kg, eGFR<90 ml/min/1.73m 2 , chronic HCV, PRTD, and use of LPV/r were associated with high tenofovir plasma concentrations. The overall prevalence of PRTD in TDF-treated patients (15.4%) in this study is relatively lower when compared to the study from Spain (22%) and higher than study from Italy (Aquitaine cohort) (6.5%) and the United States (7.6%) [10,26,35]. Our findings are in agreement with prior studies performed in Europe which reported that high tenofovir plasma concentrations were associated with PRTD [25,31]. In the current study, tenofovir plasma concentrations were 2.5-fold higher in patients with eGFR < 90 ml/min/1.73m 2 and 2-fold higher in patients with subclinical PRTD. Likewise, patients with plasma tenofovir concentrations >160ng/mL were associated with a 2-fold higher risk of developing PRTD, consistent with previous findings showing tenofovir plasma concentrations was independently associated with tubular dysfunction [25,31,36]. It has been theorized that high intracellular concentrations of tenofovir can disrupt the mitochondrial function and damage the proximal tubular cells [37]. The higher the tenofovir plasma concentrations were, the greater the accumulation of tenofovir in the proximal renal tubular cells. This can subsequently result in having a higher risk for tubular cell dysfunction. In this study, we observed that patient with low eGFR had significantly higher plasma tenofovir concentrations and patients with PRTD had lower eGFR than patients without PRTD. Because tenofovir is secreted via glomerular filtration and proximal tubular secretion, thus when the  patients have tubular dysfunction/low eGFR, this will possibly cause higher plasma tenofovir concentrations [1,2]. Therefore, it is possible that PRTD could be caused by cumulative exposure to tenofovir. On the other words, PRTD with low renal function can possibly cause high plasma tenofovir concentrations. Our results support data that the effect of tenofovir on PRTD is dose-dependent. However, in this study we could not confirm whether high plasma concentration is mainly caused by PRTD or it is directly caused by low renal clearance.
Three studies from Asia evaluated the risk of developing TDF nephrotoxicity in HIV infected patients with median BW of 55-63 kg. They found that lower body weight was significantly associated with TDF related nephrotoxicity [28][29][30]. This is consistent with our findings that plasma tenofovir concentrations were 2.2-fold higher among patients with BW < 55 kg. Higher tenofovir plasma concentrations in patients with low BW < 55 kg would lead to further renal damage. In addition, this study found that BW/serum creatinine (Scr) ratio was negatively correlated with tenofovir plasma concentrations (Correlation Coefficient = -0.28; P < 0.001) which was consistent with the findings reported by Rodriguez-Novoa et al. and Calcagno et al. in which tenofovir plasma concentrations decreased as the BW/SCr ratio increased [25,38].
However it is unclear why HCV co-infection was strongly correlated with higher tenofovir plasma concentrations in our study. Patients co-infected with HIV and HCV had higher tenofovir plasma concentrations [median190 (IQR146-223) ng/mL] compared to those with HIV mono-infection [median 146 (IQR 120-197) ng/mL] whereas only 2 out of 54 (4%) patients co-infected with HIV/HCV had PRTD. In addition, HCV was strongly associated with higher tenofovir plasma concentration (3.6-fold). Currently, pharmacokinetic data of TDF in HIV and HCV co-infection is largely unknown. Since TDF is not a substrate of CYP450 enzyme, HCV co-infection is not expected to influence the plasma concentrations of tenofovir. There is a higher risk of renal impairment in HIV and hepatitis B co-infection with advanced liver fibrosis (fibrosis score >F3). However, in our study, 20% of the patients co-infected with HIV and HCV had advanced liver fibrosis (fibrosis score > 9.5kPa) based on transient elastography scores. Median plasma tenofovir concentrations did not differ between those with or without advanced liver fibrosis (191 ng/mL versus 136 ng/mL; p = 0.15). In addition, advanced liver fibrosis was not significantly associated with lower eGFR and higher tenofovir plasma concentrations. Liver damage can lead to impaired blood flow and production of toxins that can have a detrimental effect on kidney function. However, this study did not show any correlation between liver fibrosis and eGFR which may be due to its relatively small sample size. Thus, we cannot ascertain what factors are causing the patients co-infected with HCV to have high TDF plasma concentrations.
Although almost half of the study population (48%) was female, we did not find any association between female sex and high tenofovir plasma concentrations. This finding disagrees with a previous Spanish pharmacokinetic study conducted in 2009 which reported that the female sex was associated with high tenofovir plasma concentrations [25]. Other studies also failed to find an association of sex on tenofovir plasma concentrations. Recently, Gervasoni et al evaluated tenofovir plasma concentrations among 22 post-menopausal and 28 pre-menopausal women from Italy and found that tenofovir plasma concentrations were not difference between pre-and post-menopausal women [39].
Several studies conducted in Caucasians found an association between high plasma tenofovir concentrations with unboosted or boosted PI and a higher risk of developing renal toxicity when compared to EFV, NVP or RAL [21,[39][40][41][42][43]. Our results did not observe a correlation of boosted PI and tubular dysfunction, but LPV/r use was associated with higher tenofovir plasma concentrations (2.5 fold). No correlation between ATV/r use and plasma tenofovir concentrations was observed, unlike those patients on unboosted ATV from Calcagno's study which found an association with the high tenofovir plasma concentration [38]. As our ATV/r patients were using ATV/ 200/100 mg once daily, we speculate that receiving a lower dose may lower the  impact on inhibition of multiple resistance protein (MRP) transporters at the apical side of renal tubular cells, and consequently less tenofovir intra-tubular accumulation.
Our study has several limitations. Firstly, due to the nature of the cross-sectional design, we could not foresee whether high plasma tenofovir concentrations and PRTD will predict future renal impairment. Secondly, only patients with eGFR > 60 ml/min/1.73m 2 were enrolled and approximately 50% of patients using tenofovir at our research center were included in this analysis so it is possible that patients with more serious renal toxicity were excluded. Thirdly, most of the patients on boosted ATV were on ATV/r 200/100 mg once daily instead of the standard dose of ATV/r 300/100 mg once daily. Hence we cannot draw strong conclusion whether boosted ATV/r use is associated with high tenofovir plasma concentration. Lastly, this study had only a single tenofovir plasma measurement at 12-hours post-drug administration. Oral apparent clearance of TDF was not determined. High intraindividual variation of tenofovir clearance have been reported in Caucasians , but little is known of tenofovir clearance in Asians [44].
In conclusion, mid-dose tenofovir concentrations >160 ng/ mL was independently associated with PRTD. Low BW, chronic HCV, low eGFR, LPV/r use and PRTD were associated with higher tenofovir plasma concentration and tubular function should be closely monitored in these patients. In addition, the association of higher plasma tenofovir concentrations and PRTD need to be further explored in a large prospective cohort to determine whether this association is linked to rapid renal deterioration in the future. Furthermore, lower dosing strategies of TDF in patients at highest risk of TDF associated renal toxicity should be further explored.