Pharmacogenetics of tenofovir and emtricitabine penetration into cerebrospinal fluid

Background Blood-cerebrospinal fluid (CSF) barrier transporters affect the influx and efflux of drugs. The antiretrovirals tenofovir and emtricitabine may be substrates of blood-brain barrier (BBB) and blood-CSF barrier transporters, but data are limited regarding the pharmacogenetics and pharmacokinetics of their central nervous system (CNS) penetration. Objectives We investigated genetic polymorphisms associated with CSF disposition of tenofovir and emtricitabine. Method We collected paired plasma and CSF samples from 47 HIV-positive black South African adults who were virologically suppressed on efavirenz, tenofovir and emtricitabine. We considered 1846 single-nucleotide polymorphisms from seven relevant transporter genes (ABCC5, ABCG2, ABCB1, SLCO2B1, SCLO1A2, SLCO1B1 and ABCC4) and 782 met a linkage disequilibrium (LD)-pruning threshold. Results The geometric mean (95% confidence interval [CI]) values for tenofovir and emtricitabine CSF-to-plasma concentration ratios were 0.023 (0.021–0.026) and 0.528 (0.460–0.605), respectively. In linear regression models, the lowest p-value for association with the tenofovir CSF-to-plasma ratio was ABCB1 rs1989830 (p = 1.2 × 10−3) and for emtricitabine, it was ABCC5 rs11921035 (p = 1.4 × 10−3). None withstood correction for multiple testing. Conclusion No genetic polymorphisms were associated with plasma, CSF concentrations or CSF-to-plasma ratios for either tenofovir or emtricitabine.


Introduction
Tenofovir and emtricitabine are part of the current first-line antiretroviral therapy (ART) regimens for HIV-positive adults in resource-limited settings and both are widely used in high-income countries. 1 Infection of the central nervous system (CNS) by HIV-1 occurs early in infection and its clearance is reliant on adequate CNS antiretroviral concentrations. 2 However, there are limited data regarding determinants of cerebrospinal fluid (CSF) penetration by tenofovir and emtricitabine. Data from small cohorts indicate that CSF concentrations of tenofovir and emtricitabine are 5% and 50% of plasma concentrations, respectively. 3,4,5 However, higher CSF tenofovir concentrations and lower emtricitabine concentrations have been reported, which may be explained by polymorphisms in drug transporters or altered blood-brain barrier (BBB) permeability. 4,5 Transporters in the BBB and blood-CSF barrier (BCB) affect the influx and efflux of drugs, including tenofovir and emtricitabine. 3,6,7 Multidrug resistance protein-5 (MRP-5, encoded by ABCC5) is ubiquitous and mediates the efflux of nucleoside reverse transcriptase inhibitors. 8 Lower CSF emtricitabine exposure in females compared to males is hypothesised to reflect differential expression of MRP transporters at the BBB and BCB. 3 In vitro, tenofovir is a substrate of the breast cancer resistance protein (BCRP, encoded by ABCG2), MRP-4 (encoded by ABCC4) and P-glycoprotein (encoded by ABCB1). 9,10,11 A polymorphism in ABCG2 rs2231142 has been associated with 1.5-fold increased plasma tenofovir exposure and Thai patients carrying ABCC4 3463 AG or GG (rs1751034) had an 11% greater tenofovir clearance compared with AA. 12,13 Loss-of-function ABCC4 polymorphisms have been associated with reduced clearance of tenofovir. 11,14 In genome-wide analyses, SLC17A1 rs12662869 was associated with an increase in tenofovir clearance. 15 It is possible that genetic polymorphisms that affect transporter function will affect tenofovir or emtricitabine CSF penetration. The pharmacogenetics of CSF penetration of tenofovir and emtricitabine have not been described.
Africans are the most genetically diverse population worldwide. 16 South Africa has the world's largest ART programme, with most patients currently receiving efavirenz-based regimens that include the nucleos(t)ides tenofovir and emtricitabine. 17 We previously reported on the pharmacogenetics of CSF penetration of efavirenz in black South Africans. 18 Here, we characterise the associations between transporter gene polymorphisms and CSF penetration of tenofovir and emtricitabine in the same cohort.

Participants
Adults (≥ 18 and ≤ 70 years of age) from a randomised control trial (PACTR201310000635418) that investigated lithium for HIV-associated neurocognitive impairment were invited to participate in the present study. 19 We also invited participants who were screened for that trial but were excluded based on cognitive impairment criteria. All participants provided written informed consent. This study was approved by the University of Cape Town Human Research Ethics Committee (HREC 071/2013).

Pharmacokinetic sampling
We collected paired plasma and CSF samples for tenofovir and emtricitabine assays. Participants recorded the dosing time the night before and were admitted in the morning for pharmacokinetic sampling. Whole blood was collected within 45 min of CSF sampling and centrifuged within 1 h of collection. Plasma and CSF aliquots were stored at −80 °C until analysis.

Tenofovir and emtricitabine measurement
The analytical laboratory in the Division of Clinical Pharmacology at the University of Cape Town quantified total tenofovir and emtricitabine in plasma and CSF using validated liquid chromatography tandem mass spectrometry assays.
The lower limits of quantification (LLQs) for plasma tenofovir and emtricitabine were 10.0 ng/mL and 37.5 ng/mL, respectively. For CSF, the LLQs for total tenofovir and emtricitabine were 0.5 ng/mL. Concentrations below the limits of quantification were treated as missing data.

Characterisation of genetic polymorphisms
We extracted DNA from the buffy coat using the QIAsymphony kit. Genotyping was performed using the Infinium® Expanded Multi-Ethnic Genotyping Array (MEGA EX ; Illumina, San Diego, CA, USA).
All genotyping was performed at Vanderbilt Technologies for Advanced Genomics (VANTAGE), by laboratory personnel with no knowledge of clinical data. All samples were genotyped in duplicate. The final dataset included 1846 polymorphisms from 47 participants.

Genetic polymorphisms
Amongst the 47 participants, 1846 polymorphisms were successfully genotyped. Only SLCO1B1 rs4149056 was monomorphic (i.e. no minor alleles). The remaining 1845 polymorphisms were in Hardy-Weinberg equilibrium based on a Bonferroni-adjusted p-value threshold of 6.4 × 10 -5 ; 56 had unadjusted Bonferroni p-values of < 0.05. Minor allele frequencies for all polymorphisms are provided in Appendix Table 1.

Discussion
We characterised the associations between 782 genetic polymorphisms and CSF disposition of tenofovir and emtricitabine in black South African adults. The lowest  p-value for tenofovir CSF-to-plasma ratio was ABCB1 rs1989830 (p = 1.2 × 10 -3 ), and for emtricitabine was ABCC5 rs11921035 (p = 1.4 × 10 -3 ). None were significant after correcting for multiple testing. In addition, we found no significant associations with absolute CSF or plasma concentration after correcting for multiple testing.
Associations with tenofovir pharmacokinetics and genetic polymorphisms were found in other populations. An increase in tenofovir plasma concentrations were independently associated with ABCC4 4131T→G (genotype TG or GG) in 150 Thai HIV-infected adults. 14 ABCB1 rs3213619 was associated with increased tenofovir bioavailability in a predominantly African-American patient population (n = 45) and thought to be a result of decreased P-glycoprotein function. 21 A genomewide and candidate gene association analyses with tenofovir pharmacokinetics showed that SLC17A1 rs12662869 was associated with an increase in tenofovir clearance (p = 7.1 × 10 -9 ) but failed to show significant associations in candidate genes (including ABCC4, ABCC10, ABCB1, ABCC2, SLC22A11, AK2 and AK3) after correction for multiple comparisons. 15 Our study has limitations. With our sample size, we were underpowered to detect associations with small effect sizes. We could only detect associations with relatively frequent polymorphisms and with large effect sizes. Therefore, these data should be regarded as exploratory. Polymorphisms not genotyped in our study may be associated with tenofovir or emtricitabine disposition into CSF. Whilst we did not adjust for creatinine clearance, this should not be a confounder that affects drug disposition into CSF. We included 33 (70%) participants with mild to moderate neurocognitive impairment, as previously reported. 18 We may therefore have introduced a selection bias.

Conclusion
In conclusion, we found no significant associations between any of the 782 polymorphisms and plasma concentrations, CSF concentrations or CSF-to-plasma ratios for either tenofovir or emtricitabine in univariate linear regression models after correcting for multiple testing.
E.H.D. was responsible for the study concept and design, data acquisition, data analysis and interpretation of data, drafting and revising the manuscript for content, study supervision and obtaining funding. P.Z.S. contributed to data analysis and interpretation of data and drafted and revised the manuscript for content. L.W. conducted the sample analysis and revised the manuscript for content. J.A.J. contributed to the study concept and design, revised the manuscript for content and supervised the study. D.W.H. conducted the sample analysis and data analysis, interpreted the data and revised the manuscript for content. G.M. contributed to the study concept and design and revised the manuscript for content.

Data availability
The data that support the findings of this study are available from the corresponding author, E.H.D., upon reasonable request.