CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

Background: Patient non-adherence to therapy is an ongoing problem, which results in sub-therapeutic drug concentrations that increases the risk for emergence of drug/class resistant strains of HIV. Recently, long-acting injectable (LAI) formulations of antiretroviral drugs have been developed, which may overcome some of the problems with patient non-adherence to oral regimens. We report here a preclinical assessment of LAI solid drug nanoparticle (SDN) formulations of novel emtricitabine (FTC) prodrugs. Methods: To decrease aqueous solubility for compatibility with SDN formation, 3 varying chain length carbamate/carbonate prodrugs of FTC were produced to mask key hydrophilic groups (PD1, PD2 and PD3). SDN formulations were then generated using a proprietary emulsion-templated freeze-drying technology. An initial screen of 12 candidate formulations was conducted in Wistar rats. Rats were dosed with 10 mg/kg FTC equivalent via a single intramuscular injection in the musculus biceps femoris. Plasma samples were taken over 7 days and analysed via LC-MS/MS. For three promising lead formulations PD2SDN9; PD3SDN10; PD3SDN12), a longer-term in vivo study was conducted. For this, Wistar rats were dosed with two IM injections (20 mg/kg) and plasma drug concentrations were monitored over 28 days. Results: In the initial screen, FTC concentrations in candidates SDN1-8 and SDN11 fell below the limit of detection (2 ng/mL) within 4 days. PD2SDN9, PD3SDN10 and PD3SDN12 resulted in detectable plasma FTC concentrations for 7 days (FTC > 16 ng/mL). SDN formulations 9, 10 and 12 were therefore progressed to a longer-term study (Figure 1). Cmax was reached at 6 hours for all formulations (SDN9 127 ± 52.9 ng/mL, SDN10 95 ± 39.4 ng/mL and SDN12 119 ±20.9 ng/mL). FTC plasma concentrations were detectable for all three formulations until 14 days (SDN9 21 ± 16.2 ng/mL, SDN10 31 ± 7.1 ng/mL and SDN12 28 ± 22.7 ng/mL). FTC concentrations in all 3 formulations were below the limit of detection (2 ng/mL) by 21 days. Conclusion: The in vivo data presented here demonstrate that the combined prodrug/SDN approach can provide plasma exposure in rats for 14 days. Species differences in renal clearance of FTC mean that exposures longer than 14 days are likely to be achievable in humans. Further studies will aim to optimise formulations to produce exposure beyond 14 days in rats and to elucidate the biological mechanisms involved in LA release of SDN prodrugs. 485 PBPK MODELING OF MICROARRAY PATCHES FOR LONG-ACTING INTRADERMAL DRUG DELIVERY Rajith Kumar Reddy Rajoli 1 , Charles W. Flexner 2 , Andrew Owen 1 , Ryan Donnelly 3 , Marco Siccardi 1 1 University of Liverpool, Liverpool, UK, 2 Johns Hopkins Hospital, Baltimore, MD, USA, 3 Queen’s University Belfast, Belfast, UK Background: Intradermal delivery using microarray patches (MAPs) represent a blood-free, minimally invasive and patient friendly route of administration. MAPs can deliver long-acting (LA) nanoformulations (NFs), providing systemic exposure for an extended time period. The aim of this study was to design and evaluate intradermal physiologically-based pharmacokinetic (PBPK) models to simulate the pharmacokinetics (PK) resulting from the administration of rilpivirine or cabotegravir through MAPs. Methods: Simulations were conducted in 100 healthy individuals using published anthropometric equations. A novel PBPK model for transdermal delivery was constructed using Simbiology v. 4.3.1 (MATLAB 2013b). The model was qualified against observed data in rats for intramuscular (IM) and intradermal rilpivirine, and for clinical data in healthy adults for both oral and IM LA NFs of cabotegravir and rilpivirine. The PBPK models were assumed to be qualified if the simulated values were ± 50% from the mean clinical values. Pharmacokinetics for various four-weekly intradermal loading dose (LD) and maintenance doses (MDs) were predicted following four weeks of oral administration. C trough values obtained from steady-state oral administration were considered as the target C trough values. Optimal drug release rates from the MAPs were identified for both drugs such that the minimal dose and target C trough values at the end of the dosing interval were achieved.

Results: The PBPK models were successfully qualified resulting in a difference of less than ± 50% from the observed pharmacokinetic values. For cabotegravir and rilpivirine, LDs ranging from 180 – 540 mg and 720 – 1080 mg respectively and MDs from 180 – 360 mg and 360 – 720 mg respectively were assessed across six different release rates between 0.0005 – 0.003 h -1 (Figure). An LD of 360 mg and MD of 180 mg between 0.001 – 0.0015 h -1 for cabotegravir, and LD of 1080 mg and MD of 720 mg between 0.001 – 0.002 h -1 for rilpivirine, were identified as the optimal settings for NF and MAP design. Conclusion: MAPs could be an effective tool for delivery of LA NFs for chronic antiretroviral therapy. PBPK models identified optimal dose and release characteristics for cabotegravir and rilpivirine, supporting rational development of future formulations. This approach can identify potential drug candidates for LA therapy delivered using MAPs.

Poster Abstracts

486 PHARMACOKINETICS OF TENOFOVIR ALAFENAMIDE BY SUBCUTANEOUS IMPLANT FOR HIV PrEP Greg Gatto 1 , Natalie Girouard 1 , Rhonda M. Brand 2 , Leah Johnson 1 , Mark A. Marzinke 3 , Sudie Rowshan 1 , Jarret C. Engstrom 2 , Ian McGowan 2 , Zach Demkovich 4 , Ellen Luecke 4 , Ariane van der Straten 4 1 RTI International, Research Triangle Park, NC, USA, 2 University of Pittsburgh, Pittsburgh, PA, USA, 3 Johns Hopkins Hospital, Baltimore, MD, USA, 4 RTI International, San Francisco, CA, USA Background: We evaluated the pharmacokinetics of tenofovir alafenamide (TAF), an antiretroviral prodrug with high potency and potential for HIV Pre- exposure prophylaxis (PrEP), released from polycaprolactone (PCL) implants at the midway point of a 3-month study in rabbits. Methods: PCL tubes (40 mm x 2.5 mm x 70 µmwall thickness) were fabricated by extrusion and loaded with a slurry of TAF and Castor Oil (2:1 ratio by mass). Devices were sterilized by gamma-irradiation post-fabrication. Ten female New Zealand white rabbits were surgically implanted with active (TAF; n=6) or placebo devices (n=4). Blood samples were taken 6 h post-surgery on day 0 and on days 1, 4, 7 and weekly thereafter, for TAF and tenofovir (TFV) drug quantification (lower limits of quantification: TAF, 0.03 ng/mL; TFV, 0.31 ng/mL). Peripheral blood mononuclear cells (PBMCs) were isolated from blood samples collected on days 7 and weekly thereafter, for TFV-diphosphate (TFV-DP) determination (as fmol/10 6 cells). At day 49, some devices were recovered for residual drug analysis. Similar TAF devices were concurrently evaluated in-vitro in PBS at 37°C. In-vitro sampling occurred every other day and TAF release was measured by UV spectrophotometer.

CROI 2018 174