EVG/COBI/FTC/TAF Bioequivalence Comparing Whole Tablets with Tablets Dissolved in Tap Water

Abstract

Medication adherence can be challenging for persons with difficulty swallowing tablets. We investigated the bioequivalence of a dissolved tablet when compared with that of a whole tablet of the fixed-dose combination elvitegravir (EVG)/cobicistat (COBI)/emtricitabine (FTC)/tenofovir (TFV) alafenamide fumarate (TAF). A within-subject fixed-order two-period open-label study was conducted in 12 HIV-negative research participants after obtaining informed consent. Participants took a single dose each of the whole tablet and dissolved tablet under direct observation, separated by a 14-day washout period. The dissolved tablet was prepared by adding one whole EVG/COBI/FTC/TAF tablet to 120 mL tap water and stirring. Both dosage types were taken with a standardized meal. Plasma samples were obtained for 72 h postdose. Plasma EVG, FTC, TAF, and TFV were analyzed with liquid chromatographic-tandem mass spectrometric methods. Peak plasma concentration (C_max) and the area under the concentration-time curve extrapolated to infinity (AUC_0−∞) were estimated using WinNonlin software (v.8.3). The primary outcome was bioequivalence consistent with FDA guidance using the 90% confidence interval or the geometric mean ratio. Of 12 participants, 7 were black (58%) and 5 were white (42%), 4 were women (33%), 8 were men (67%), and the mean age was 43.6 years (23–54). There were no complaints about taste with the dissolved tablet. Bioequivalence was established only for FTC. EVG C_max and AUC_0–∞ were higher by 18% and 12%, respectively, when taking the dissolved compared with the whole tablet. TAF AUC_0–∞ and C_max were both 8% lower, whereas TFV C_max and AUC_0–∞ were 8% and 5% lower, respectively, when taken after dissolution. EVG/COBI/FTC/TAF dissolved rapidly in water and had no unpleasant taste. Increases in EVG and decreases in TAF and TFV concentrations were observed when taking dissolved EVG/COBI/FTC/TAF. These changes were judged to be clinically insignificant. Dissolving EVG/COBI/FTC/TAF in water may be suitable for those with pill swallowing challenges. The trial was registered on (//clinicaltrials.gov NCT03717129).

Introduction

Poor medication adherence is a common problem for many patients with chronic illnesses, including HIV, especially those patients with complex regimens.^1,2 There are numerous patient, clinician, and health care system causes of poor adherence.^3–6 A high pill burden is associated with lower adherence, higher hospitalization rates, increased risk of drug resistance, and treatment failure.^7,8 Numerous fixed-dose once-daily antiretroviral (ARV) drug combinations simplify regimens and encourage adherence.^2,8–14

However, such combinations typically create larger pills, posing new adherence challenges for some due to multifactorial causes of pill aversion or trouble swallowing.^15,16 To solve this problem, many people prefer to dissolve or crush their pills instead of swallowing them whole. Even though this is a widely used practice, crushing or dissolving tablets may alter drug bioavailability, with potential for secondary effects on efficacy or toxicity.^17–19

Elvitegravir/cobicistat/emtricitabine/tenofovir alafenamide (EVG/COBI/FTC/TAF; Genvoya^®) is a fixed-dose combination that includes an integrase strand transfer inhibitor (INSTI), (EVG) 150 mg, a CYP3A4 inhibitor and INSTI pharmaco-enhancer, (COBI) 150 mg, and 2 nucleoside reverse transcriptase inhibitors, (FTC) 300 mg, and (TAF) 10 mg, a prodrug of tenofovir (TFV), which is metabolized intracellularly to TFV diphosphate (TFV-DP), the active metabolite.^20,21 EVG/COBI/FTC/TAF is a once-daily one-tablet regimen proven effective for the treatment of HIV-1 The current guidelines do not recommend crushing or chewing (EVG/COBI/FTC/TAF) tablets, since no data exist on the bioequivalence of this modification.²⁰

The primary objective of this study was to investigate EVG/COBI/FTC/TAF bioequivalence administered as single dissolved tablet compared with that of a whole tablet in HIV-negative participants to inform counseling of persons who experience dosing challenges with this fixed-dose combination. We evaluated plasma area under the concentration-time curve extrapolated to infinity (AUC_0–∞) and peak plasma concentration (C_max) of EVG, FTC, TAF, and TFV, after single-dose administration of each formulation, both whole and dissolved.

Methods

Study design

This was an open label two-period fixed-drug sequence bioequivalence study that assessed the pharmacokinetics (PK) of EVG, FTC, TAF, and TFV components of the EVG/COBI/FTC/TAF tablet administered as a whole or dissolved tablet. The study was conducted at the Johns Hopkins Hospital Drug Development Unit (DDU). The protocol was reviewed and approved by the institutional review board (IRB) at Johns Hopkins Medicine and conducted in accordance with good clinical practice. The subjects provided their written consent before enrolling in the study.

The study included a screening visit, at which the eligibility of the participants was determined after obtaining their informed consent. After an overnight fast, eligible participants received a series of two single-dose administrations of EVG/COBI/FTC/TAF (Gilead Sciences, Inc., Foster City, CA, USA) after a standardized meal (400 Kcal, 20% fat). Plasma PK sampling followed over 3 days with at least a 14-day washout between the two dosing periods. The first dosing sequence was the whole tablet, a safety visit on day 8, the 14-day washout period, a similar sequence with the dissolved tablet.

To ensure complete dissolution, tablets were dissolved for at least 5 min in 60 mL room temperature tap water with gentle stirring. After swallowing the EVG/COBI/FTC/TAF dose dissolved in 60 mL tap water, the same dosing cup was rinsed with a second 60 mL aliquot of tap water and then swallowed. All study treatments were administrated under direct observation by the study team. Each study treatment was taken at least 30 min after starting the meal and before the passage of 10 min after finishing the meal.

Study subjects

Eligible research participants were ≥18 years old, HIV-negative, HBsAg-negative, with negative point of care testing urine pregnancy test for cisgender women, and with no active attempt to conceive. Any medication as well as any herbal, dietary, nutritional supplements, or alternative medicine compounds was prohibited for 14 days prior, and 3 days after treatment administration, unless medically needed and approved by the study investigator.

Specimen collection and bioanalytical testing

Blood was collected once before administration of the drug and 0.25, 0.5, 0.75, 1, 2, 3, 4, 5, 6, 8, 10, 14, 24, 48, and 72 h after the observed study drug dose. Samples were processed and plasma was analyzed through liquid chromatographic-tandem mass spectrometric (LC-MS/MS) analysis by the Clinical Pharmacology Analytical Laboratory (CPAL) within the Johns Hopkins University School of Medicine. LC-MS/MS methods for FTC, TAF, and TFV quantification have been previously described.^22,23 For EVG quantification, two calibration curves were generated (low assay: 0.05–100 ng/mL; high assay: 100–5,000 ng/mL); 0.025 and 0.100 mL of plasma was used for the low and high EVG assay, respectively.

EVG was extracted through protein precipitation, and 0.010 mL of extracted material was introduced into the mass analyzer. EVG was monitored in positive ionization and selective reaction monitoring modes using an API5000 mass analyzer (SCIEX, Redwood City, CA, USA) interfaced with a Waters Acquity LC system (Waters Corporation, Milford, MA, USA). Intra- and interassay coefficients of variation (%CVs) for EVG low-, mid-, and high-level quality controls ranged from 1.65% to 7.50% and 2.70% to 6.41% (low assay) and from 0.77% to 4.56% and 2.51% to 4.35% (high assay), respectively. In aggregate, assays lower limits of quantitation for EVG, FTC, TAF, and TFV were 0.05, 5.00, 0.03, and 1.00 ng/mL, respectively.

Assays were validated in accordance with FDA bioanalytical recommendations. PK profiles for plasma EVG, FTC, TAF, and TFV were determined over 72 h postdose that provided sampling over five half-lives for each of the administered drugs, except for TFV (only —three to four half-lives).

Safety assessment

Safety was assessed with physical examination, vital signs, and documentation of laboratory and clinical adverse events. A physical examination was performed during the screening visit, with vital signs and adverse events documented at every visit. Safety laboratory tests, including hematology and chemistry panels, were collected at the two safety visits, which occurred 7 days after each dose.

Data analysis

PK parameters were calculated with noncompartmental analysis using Phoenix WinNonlin version 8.3 (Certara L.P., St. Louis, MO, USA). The main PK parameters calculated were area under the concentration-time curve extrapolated to infinity (AUC_0–∞), and peak plasma concentration (C_max). We followed the FDA guidance on bioequivalence²⁴ in calculating boundaries of 80%–125% used for the 90% confidence interval (CI) of the geometric mean ratio (GMR) for the PK parameters (C_max and AUC_0–∞) of the test formulation (dissolved tablet) compared with the reference formulation (whole tablet).

The sample size of 12 was based on a coefficient of variation for C_max and AUC_0–∞ of <25% for all analytes, and the hypothesized test/reference GMR of 1, to assess bioequivalence with 5% alpha and 80% power. We used nonparametric Wilcoxon signed rank test with exact significance to assess differences for PK parameters beyond the formal bioequivalence (BE) definition, namely, time to C_max (T_max) and concentration 24 h after dosing (C24).

Results

Twelve research participants enrolled in the study: eight male (67%) and four female (33%) sex assigned at birth, seven black (58%), and five white (42%), age (mean ± standard deviation) 43.3 ± 10.4 years, weight 95.8 ± 19.2 kg, and body mass index 30.7 ± 5.5 kg/m². No research participants complained of unpleasant taste after administration of the dissolved tablets. No adverse events were attributed to study drug, and no clinically relevant changes were noted in laboratory parameters, vital signs, or physical examination. Although the protocol specified a 5-min period of stirring for tablet dissolution, tablets were largely dissolved without any stirring within 5 min; after another 5 min of stirring, the tablets were completely dissolved.

Plasma concentration–time profiles for whole and dissolved tablets were largely overlapping within each analyte, with only EVG showing a visibly noticeable difference between the two curves (Fig. 1). Mean EVG C_max and AUC_0–∞ were 18% and 12% higher when the tablet was dissolved than when taken whole, respectively (Table 1), although only C_max failed to meet bioequivalence criteria. Mean FTC C_max and AUC_0–∞ were 7% and 5% lower, respectively, when dissolved, although both parameters met the bioequivalence criteria.

FIG. 1.

Plasma drug concentration versus time plots for each of four analytes related to three of four drugs in the fixed-dose combination (excluding COBI) compared with whole tablet (closed circles) and dissolved tablet (open circles) administration. Mean + SD concentrations summarize concentrations for all 12 research participants. COBI, cobicistat; SD, standard deviation.

Table 1.

Summary of Pharmacokinetics Parameters and Bioequivalence Criteria of Antiretroviral Constituents of EVG/COBI/FTC/TAF Tablets Administrated As a Whole or Dissolved

PK parameter	Whole tablet	Dissolved tablet	GMRa	Bioequivalence criteria met
Units	Mean (CV%)	Mean (CV%)	90% CI	Yes/nob
EVG
Tmax (h)	3.7 (36.7%)	4.6 (37.9%)
Cmax (ng/mL)	1,650 (22.5%)	1,946 (28%)	1.18 (1.03–1.35)	No
C24 (ng/mL)	259 (70%)	297 (65.5%)
AUC0–24 (ng.h/mL)	20,702 (20.9%)	23,021 (18.2%)	1.12 (1.04–1.22)
AUCinfc (ng.h/mL)	24,219 (25%)	26,948 (22.1%)	1.12 (1.04–1.22)	Yes
FTC
Tmax (h)	1.25 (56.8%)	1.15 (60%)
Cmax (ng/mL)	2,095 (21.2%)	1,968 (24.4%)	0.93 (0.81–1.09)	Yes
C24 (ng/mL)	47 (32.5%)	43 (21.7%)
AUC0–24 (ng.h/mL)	10,446 (20.3%)	9,907 (16.8%)	0.95 (0.90–1.01)
AUCinfc (ng.h/mL)	11,603 (19.8%)	10,969 (16.6%)	0.95 (0.90–1.00)	Yes
TAF
Tmax (h)	0.7 (65.2%)	0.4 (53.7%)
Cmax (ng/mL)	186 (43%)	171 (42.7%)	0.92 (0.68–1.23)	No
C24 (ng/mL)
AUC0–24 (ng.h/mL)	145 (43.9%)	130 (35.9%)	0.94 (0.82–1.07)
AUCinfc (ng.h/mL)	145 (44.1%)	128 (37.5%)	0.92 (0.79–1.08)	No
TFV
Tmax (h)	1.1 (73.1%)	0.8 (55%)
Cmax (ng/mL)	11 (37.3%)	9.5 (24.2%)	0.92 (0.77–1.10)	No
C24 (ng/mL)	2.60 (20.4%)	2.5 (24.4%)
AUC0–24 (ng.h/mL)	89 (18.1%)	86 (17.1%)	0.98 (0.90–1.06)
AUCinfc (ng.h/mL)	253 (18.6%)	241 (19.1%)	0.95 (0.87–1.04)	Yes

^a GMR is for dissolved/whole tablets.

^b Yes indicates 90% CI of GMR falls within the 80%–125%. No indicates GMR 90% CI falls outside those bounds.

^c Percent extrapolated for AUC_inf was low (all <4%) except for TFV (range 19%–49%). Because of these high values for TFV, we also evaluated AUC_0–24, even though this was not conducted at steady state.

AUC, area under curve; C_max, peak plasma concentration; CI, confidence interval; CV%, percent coefficient of variation; EVG, elvitegravir; FTC, emtricitabine; GMR, geometric mean ratio; PK, pharmacokinetics; T_max, time to C_max; TAF, tenofovir alafenamide; TFV, tenofovir.

Mean prodrug TAF C_max and AUC_0–∞ were both 8% lower when taken dissolved, with both lower 90% CI falling below the lower BE boundary. Mean TFV C_max and AUC_0–∞ were 8% and 5% lower, respectively, with dissolved tablets, with the C_max 90% CI falling outside the lower BE boundary. TFV T_max and C24 were not significantly different between the whole and dissolved tablets (all p > .12). (Table 1)

Discussion

In this bioequivalence study comparing dissolved with whole (reference) EVG/COBI/FTC/TAF tablets, FTC fully met the BE criteria for both C_max and AUC_0–∞. However, EVG either met BE criteria (AUC_0–∞) or rose above the BE boundary (C_max). The TAF and TFV PK parameters fell slightly below the lower BE bound.

Although we did not assess the clinical or virological impact of these PK differences, we believe the observed BE differences are not clinically significant. INSTI are well tolerated and more potent than the other ARVs included in this fixed dose combination; therefore, modest increases in EVG concentrations do not raise concerns for toxicity and provide reassurance that the overall antiviral effect is maintained despite TAF falling below the lower BE boundary.^25,26 The lower bound of TFV C_max after TAF dosing fell below the BE boundary; however, plasma TFV entering CD4+ cells makes a small contribution to formation of intracellular TFV-DP when compared with plasma TAF entering cells. Furthermore, the GMR of TFV AUC_0-24 was within the established BE bounds. Lastly, the other nucleoside reverse transcriptase inhibitor (NRTI) backbone drug, FTC, was fully within BE boundaries.

Accordingly, the reduction in TAF concentration with tablet dissolution has the greatest potential for negative clinical impact. The lower 90% CI for TAF AUC_0–∞ and C_max was only 1% and 12% below the lower BE boundary, respectively. Plasma trough concentrations (Ctau), arguably a more important PK parameter for NRTI antiviral effect, were not directly assessed in this single-dose study; therefore, the relative plasma concentrations 24 h after a single dose (C24), consistent with time of the next daily EVG/COBI/FTC/TAF dose, though not at steady-state, provide a reasonable surrogate of relative dissolved versus whole tablet TAF and TFV-DP concentrations at steady-state Ctau. We noted no statistically significant differences in C24 between dissolved and whole tablets for any analyte. Another limitation is that the active intracellular anabolite of TAF, TFV-DP, was not evaluated.

Previous reports note that TAF 10 mg achieves more than twice the PBMC TFV-DP achieved with TDF 300 mg—itself a highly effective dose in combination treatment and prevention regimens—thus, mitigating concerns regarding the small plasma TAF decrease observed with pill dissolution (relative to the twofold TDF-DP increase associated with TAF compared with TDF).²⁷ Lastly, our study did not investigate COBI levels, though any concentration changes would be of less interest given the very small increases in EVG concentrations and the fact that COBI has no ARV activity.

The EVG/COBI/FTC/TAF package insert does not indicate whether it is acceptable to crush and dissolve the tablet,²⁰ and concerns exist regarding the low water solubility of EVG and COBI.²⁸ In this study, we demonstrated that the tablets were readily dissolvable in tap water with stirring and without any undesirable taste. This provides a simple dosing solution for those challenged with taking the tablets.

In conclusion, EVG/COBI/FTC/TAF is a highly effective HIV treatment with convenient daily dosing, enabling high levels of adherence.²⁹ We demonstrated that dissolving the tablet in tap water results in an easily palatable solution and minor changes in ARV PK that are not expected to be clinically significant. For patients prescribed EVG/COBI/FTC/TAF who struggle with swallowing the EVG/COBI/FTC/TAF tablet, tap water dissolution of the tablet provides a viable solution to maintain medication adherence.

Authors' Contributions

A.A. contributed to conceptualization, methodology, writing—review and editing, supervision, and funding acquisition. E.J.F. was involved in conceptualization, methodology, investigation, data curation, writing—review and editing, supervision, and project administration. M.A.M. carried out methodology, validation, formal analysis, investigation, data curation, and writing—review and editing. S.A.M. was in charge of formal analysis, data curation, writing—original draft, writing—review and editing, and visualization.

J.B. was in charge of methodology, investigation, data curation, writing—review and editing, and project administration. S.B. took care of methodology, investigation, and writing—review and editing. I.M. carried out methodology, resources, and writing—review and editing. C.W.H. was involved in conceptualization, methodology, formal analysis, investigation, data curation, writing—original draft, writing—review and editing, visualization, and supervision.

Disclaimer

The views expressed in this article do not necessarily represent those of the NIH.

Author Disclosure Statement

C.W.H. has served on a Scientific Advisory Board of Gilead Sciences. I.M. is an employee of Gilead Sciences. All other authors report no conflict of interest relevant to this publication.

Funding Information

The funding for this study and study drug was provided by Gilead Sciences, Inc., Partial support was also provided by the Johns Hopkins Center for AIDS Research (P30AI094189).