A novel formulation enabled transformation of 3 HIV drugs Tenofovir-Lamivudine-Dolutegravir (TLD) from Short-Acting to Long-Acting All-in-One Injectable

Abstract

Objective:

To develop an injectable dosage form of the daily oral HIV drugs, tenofovir (T), lamivudine (L), and dolutegravir (D), creating a single, complete, all-in-one TLD 3-drug-combination that demonstrates long-acting pharmacokinetics.

Design:

Using drug-combination-nanoparticle (DcNP) technology to stabilize multiple HIV drugs, the 3 HIV drugs TLD, with disparate physical-chemical properties, are stabilized and assembled with lipid-excipients to form TLD-in-DcNP. TLD-in-DcNP is verified to be stable and suitable for subcutaneous administration. To characterize the plasma time-courses and PBMC concentrations for all 3 drugs, single subcutaneous injections of TLD-in-DcNP were given to nonhuman primates (NHP, M. nemestrina).

Results:

Following single-dose TLD-in-DcNP, all drugs exhibited long-acting profiles in NHP plasma with levels that persisted for 4 weeks above predicted viral-effective concentrations for TLD in combination. Times-to-peak were within 24 hr in all NHP for all drugs. Compared to a free-soluble TLD, TLD-in-DcNP provided exposure enhancement and extended duration 7.0-, 2.1-, and 20-fold as AUC boost and 10-, 8.3-, and 5.9-fold as half-life extension. Additionally, DcNP may provide more drug exposure in cells than plasma with PBMC-to-plasma drug ratios exceeding one, suggesting cell-targeted drug-combination delivery.

Conclusions:

This study confirms that TLD with disparate properties can be made stable by DcNP to enable TLD concentrations of 4 weeks in NHP. Study results highlighted the potential of TLD-in-DcNP as a convenient all-in-one, complete HIV long-acting product for clinical development.

INTRODUCTION

The 3-HIV oral drugs–tenofovir (TFV: T), lamivudine (3TC: L), and dolutegravir (DTG: D)–TLD is the WHO’s recommended first-line treatment for people living with HIV. Daily oral TLD is effective for people living with HIV to achieve undetectable plasma virus levels only when taken consistently. Pill fatigue and discontinuation of daily TLD tablets increase the risk of viral rebound and progression to AIDS [1].

Cabenuva is an approved injectable containing two separate long-acting products, cabotegravir and rilpivirine. Initiation of cabenuva typically includes one-month daily tablet intake, with verified viral suppression before every 1-or-2-month intramuscular(IM)-assisted injection [2]. Its implementation has proven challenging [3,4]. Cabenuva lacks hepatitis B virus (HBV) coverage and is not recommended for people living with HIV and HBV [5].

Building on success in formulating water insoluble-and-soluble HIV drugs with biocompatible lipid-excipient to form drug-combination-nanoparticles–referred to as DcNP platform [6–8] –we investigated whether short-acting TLD can be transformed into a single long-acting formulation (with TFV providing HBV coverage). A TLD drug-combination was stabilized by two-lipid-excipients in suspension as an injectable, referred to as TLD-in-DcNP. This product-candidate was characterized by long-acting pharmacokinetics in NHP. We found that LA-TLD can be formulated in the DcNP dosage form, and a single subcutaneous injection in NHP yields sustained concentrations for all 3-HIV drugs for 4 weeks.

MATERIALS AND METHODS

Chemicals

Tenofovir/TFV PMPA (((1-(6-Amino-9H-purin-9-yl)propan-2-yl)oxy)methyl) phosphonic acid, Laurus, India), lamivudine/3TC and dolutegravir sodium/DTG (Cipla, India) were cGMP-grade. Lipoid GmbH (Germany) provided cGMP DSPC (1,2-Distearoyl-sn-glycero-3-phosphocholine) and mPEG₂₀₀₀-DSPE (N-(carbonylmethoxypolyethyleneglycol-2000)-1,2-distearoyl-sn-glycero-3-phosphoethanolamine sodium salt).

Methods

TLD-in-DcNP injectable suspension was prepared following a process previously detailed [6,9], which was adjusted here. The adjustments were that 5.66 mmol DTG; 5.97mmol HCL; 40.27mmol DSPC and 4.49mmol mPEG₂₀₀₀-DSPE were dissolved in 472mL ethanol at 70°C; 28mL of 200mM NaHCO₃ buffer containing 5.85mmol TFV and 5.85mmol 3TC was added; solution was spray-dried (4M8Trix Unit; ProCepT, Belgium) under controlled-solvent-removal process to get TLD-in-DcNP powder. TLD-in-DcNP powder in 0.45%w/v NaCl-20mM NaHCO₃ buffer suspension was held at 75°C; homogenized (Emulsiflex-C5; Avestin, Canada) to achieve stable particles (50–70nm). The suspension was cooled to 25°C and stored at 4°C. The TLD % associated with DcNP, measured by equilibrium dialysis under 200-time-excess-volume, was 21.5%-T, 15%-L, and 95%-D. Finally, TLD-in-DcNP (52±5nm) sterile suspension was NHP-ready.

Pharmacokinetic Study

Seven NHPs (Macaca nemestrina, 7–12kg) were enrolled under approved Institutional Animal Care and Use Committee protocol. 5 NHP were dosed 6.2, 5.1, 10 mg/kg with TLD-in-DcNP, while 2 NHPs were dosed with the free-mixture TLD, both by subcutaneous route in the back mid-scapular region. Blood samples for the TLD-in-DcNP were collected at 0, 0.25, 0.5, 1, 3, 5, 8, 24, 48, 120, 168, 192, 336, 50, 672 hr (4 weeks) for drug analysis–the free TLD samples were collected with same schedule, but until 120 hr. Furthermore, in 4 NHP dosed with TLD-in-DcNP, in the 48 and 168 hr blood samples, peripheral blood mononuclear cells (PBMC) were isolated, frozen immediately to avoid drug loss, and analyzed for drug content. For TFV and 3TC, we could only measure the total content (parent+metabolites), without discerning triphosphates due to in-column conversion to parent TFV or 3TC. All 3-drugs in cell and plasma were analyzed using a published-validated LC-MS/MS assay [10], adapted for TLD with specific mass (m/z). Based on 5µL injection, LOD/LOQ were recorded at 0.25/0.05ng/mL for DTG, 0.59/0.2ng/mL for TFV, and 0.49/0.16ng/mL for 3TC.

Noncompartmental pharmacokinetic analysis was carried out using MATLAB2023a. Times-to-peak, peak concentrations (C_max), concentrations at 4 weeks (C_4w), apparent terminal half-lives, and areas under the curves until 4 weeks (AUC) were calculated (directly from integrating time-concentration curve based on trapezoidal rules) for each animal and reported as mean ± SD. Based on sample quality, DTG assay for NHP#5’s PBMC and free DTG in NHP#7’s plasma did not meet quality standard, and therefore they were not included in the final analysis.

For antiviral activity, an in-vitro HIV-1 inhibition assay was performed to estimate the 90% inhibitory drug concentration (IC₉₀), as previously described [6], in presence of a serum protein. The IC₉₀ for TLD-in-DcNP and free mixture TLD (fixed-dose-ratio) were established as 0.3, 0.2, and 0.25 ng/mL for TFV, 3TC and DTG, respectively, and similar between free and DcNP combinations.

RESULTS

After systematic studies with DSPC and mPEG₂₀₀₀-DSPE lipid excipients, TLD 3-HIV drugs with disparate water solubilities (TFV, 3TC, DTG: Log-P −3.4, −1.3, 2.4) were successfully stabilized in a DcNP nanosuspension. The final TLD-in-DcNP suspension contained 6.2, 5.1, 10 mg/kg TLD, and 115.1, 45.8 DSPC and mPEG₂₀₀₀-DSPE (mg/mL). No NHP safety issues, including injection site reactions, were observed with both TLD-in-DcNP and free. Drug concentrations in plasma and PBMC were determined following a subcutaneous dose: [TFV], [3TC] and [DTG] were detectable in plasma throughout a 4-week NHP study (Figure 1A). In comparison, plasma drug levels in the NHP receiving TLD-free dosage fell below detectable levels within 3 days.

Figure 1. Effects of DcNP dosage form on the time-course of plasma tenofovir (TFV or T), lamivudine (3TC or L), and dolutegravir (DTG or D) in non-human primates (NHPs). Panel A: plasma drug concentrations over time are presented after a single subcutaneous dose of the TLD-in-DcNP or free-soluble mixture given to NHP. Panel B: noncompartmental pharmacokinetic (PK) parameters results.

Panel A. All drugs in NHP receiving TLD-in-DcNP show extended time as prolonged TFV, 3TC, and DTG presence in the plasma as opposed to free-soluble drugs control at equivalent subcutaneous dose of 6.2 mg/kg TFV, 5.1 mg/kg 3TC, and 10 mg/kg DTG. Filled symbols represent plasma drug concentrations of TLD drugs in DcNP for 5 NHP, as mean ± SD. Empty symbols represent TLD given as free-soluble TLD combination in 2 NHP as mean (only one NHP with free DTG was available). Horizontal dashes lines denote the in vitro IC₉₀ values in serum (protein) as a fixed-ratio presented in the TLD-in-DcNP or fee-soluble drug-in-combination. In each graph, there is an inset representing a zoomed-in view of the first 72 hours after injection. Panel B. Noncompartmental PK parameter estimates are based on the concentration-time course data presented in Panel A. Each PK parameter is presented as mean ± SD. Time-to-peak is the time needed for the drug to rise to the first identifiable concentration peak collected at concentration C_max. C_4w is the drug concentration at 4 weeks; AUC is the area under the curve based on integration of plasma time-courses using classical trapezoidal rules for the 4-weeks study; AUC ratio as the fold enhancement of TLD formulated in DcNP after a single dose; t_1/2,z as the apparent terminal elimination half-life.

The plasma concentration peaks for TLD-in-DcNP were all recorded within the first day (Figure 1B); comparable to the equivalent free-mixture. TLD-in-DcNP’s AUC were 7.0-, 2.1-, and 20-fold higher, for TFV, 3TC, and DTG, respectively (Figure 1B). Effects of DcNP on TLD half-life extension were recorded as 10-, 8.3-, and 5.9-fold over the free-soluble drug-combination equivalent. All drug concentrations were above the estimated IC₉₀ value for the 4-week study.

Effects of DcNP on intracellular TLD enhancement were notable as the average cell-to-plasma ratio were greater than one, for all 3 drugs (Table 1). Intracellular TFV concentration in PBMC was 2.2- and 3.1-fold higher than plasma, at 48 and 168 hr, respectively. Similarly, [total] 3TC in PBMC was 3- and 15-fold higher than in plasma. Finally, DTG concentration in PBMC was 29- and 4.1-fold higher than in plasma (Table 1).

Table 1.

Effects of DcNP formulated 3-HIV drugs, TLD on enhanced cellular concentrations as measured in the Peripheral Blood Mononuclear Cells from NHP.^*

Sample Type	NHPs dose with TLD in DcNP dosage form
	Tenofovir (T)		Lamivudine (L, 3TC)		Dolutegravir (D)
	48 h	168 h	48 h	168 h	48 h	168 h
[Cell]	1,631.8 ± 114.7	989.0± 136.9	50.6 ± 4.5	13.4 ± 10.5	1303.8 ± 107.6	278.0 ± 74.9
[Plasma]	731.3 ± 190.8	320.5 ± 89.5	16.8 ± 5.8	0.9 ± 0.3	45.1 ± 18.0	68.6 ± 22.6
Cell/Plasma	2.2	3.1	3.0	15	29	4.1

^*NHP, non-human primates M. Nemestrina were given a single subcutaneous injection of TLD-in-DcNP dosage form as described in Materials and Methods. The peripheral blood mononuclear cells [PBMC or Cell] were immediately isolated from blood samples and analyzed for drug concentrations in the PBMCs. Data as mean±SD. PBMC-to-plasma noted as cell/plasma drug ratio are based on the mean values.

Collectively, these data indicate that water-soluble TFV, 3TC, and water-insoluble DTG combined into the single TLD-in-DcNP injectable provided with 4 weeks plasma exposure in NHP above detectable, measurable, and IC₉₀-predicted levels.

DISCUSSION

A stable-scalable 3-in-1 HIV drug combination, based on a novel injectable drug combination DcNP platform was successfully made. This novel TLD-in-DcNP formulation in NHP sustained plasma levels for 4 weeks despite having disparate physiochemical properties of TFV, 3TC, and DTG. The higher intracellular PBMC drug levels relative to plasma for all three drugs suggest that DcNP deliver these antivirals into cells with efficiency. Given oral TLD is a current WHO-recommended “one-pill-a-day” HIV treatment, having a long-acting TLD based on DcNP dosage form may overcome daily pill fatigue. The subcutaneous LA-TLD dosage-forms may provide a self-administration option as opposed to assisted IM-injection. While the 4-week duration of the 3 drugs in TLD-in-DcNP may be shorter than the eventual every-two-month dosing with LA-CAB and LA-RPV [11–13], the LA TLD-in-DcNP reaches C_max without delay and within 1 day (Figure 1B). As sustained/extended-release dosage forms, LA-CAB and LA-RPV deposited in the muscular space exhibit delayed peak plasma concentrations which are typically reached by 7–10 days [12]. To ensure tolerability and address delays in reaching therapeutic levels, an oral lead-in phase is advisable for cabenuva treatment. However, for TLD-in-DcNP, an oral lead-in may not be necessary due to shorter times-to-peak, and because of the well-documented safety profile of TLD in humans. This could facilitate clinical implementation.

Previously, the DcNP technology was reported to transform short-acting drug combinations into long-acting combinations [7,8,14–16]. We have leveraged the physical and chemical interactions of water-soluble and insoluble drug substances with amphipathic lipid-excipients at varying temperatures. As a result, DcNP technology is able to stabilize physically disparate water-soluble and -insoluble HIV and cancer drugs, such as lopinavir, ritonavir, efavirenz, dolutegravir or paclitaxel with water-soluble lamivudine, tenofovir, or gemcitabine [7,8,14–16]. The capacity of formulating multiple agents as one injectable suspension is notable in the HIV long-acting landscape as DcNP can transform water-soluble drugs tenofovir and lamivudine into a long-acting product along with water-insoluble DTG [17].

The 3-HIV drugs containing a potent integrase inhibitor in one complete HIV-treatment may also provide sustained viral suppression due to higher and more consistent intracellular drug levels. By week 1 in NHP, cell-to-plasma ratios for all 3-TLD drugs exhibited 3 to 15-fold higher than what found in plasma with potential synchronized TLD enhancement. As TFV is the current active drug substance of interest (formulated as prodrug TAF or TDF) to treat people with chronic HBV, TLD-in-DcNP could provide antiviral coverage for people with HBV and HIV-HBV.

To inhibit viral replication, TFV and 3TC are phosphorylated intracellularly to the active metabolites (TFV-dP and 3TC-tP). Previously, we reported with DcNP, TFV and 3TC, PBMCs isolated from NHP exhibited 60–70% of intracellular total-TFV concentrations as TFV-dP [14], and 40% of 3TC as 3TC-tP [8]. These data were similar to that reported for oral TDF and 3TC [18–20]. While remaining to be directly measured, it is likely that DcNP formulated 3TC and TFV in the PBMCs are biologically active with similar fractions of active metabolites to that with oral dosages.

The predicted target concentrations of the 90% antiviral activity as TLD-combination in serum, IC₉₀ were noted to be similar between DcNP vs free-soluble formulation. The IC₉₀ estimated with serum protein, were noted to be 0.2 to 0.3 ng/mL (Figure 1A). The IC₉₀ values for TLD are based on the fixed-dose combination, which may be significantly lower than the reported protein-adjusted IC₉₀ for each drug. For instance, the IC₉₀ for the single-agent DTG is 64 ng/mL [19], whereas in the TLD formulation, we found it as 0.25 ng/mL. Moreover, comparing the oral dose-normalized exposure of DTG (AUC/dose) between NHPs and humans [21,22], the elevated UGT metabolic clearance of DTG in NHPs could potentially result in a 24-fold increase in DTG exposure in humans [23]. Thus, given a 10mg/kg dose in NHP, the levels observed could be upscaled to therapeutic ranges for humans.

Previously, we reported that drugs in DcNP formulations do not form a local depot [24]. Instead, DcNP are rapidly and completely absorbed by the lymphatic, as opposed to blood capillary system. The DcNP enable preferential drug penetration into lymphatic, instead of less permeable blood, vessels as a first-passage after subcutaneous dosing; there was no significant retention at the injection site [25]. It is likely that all-drugs in TLD-in-DcNP remain associated with the DcNP particles, contributing to in-vivo stability leading to long-acting pharmacokinetics for all drugs. Earlier studies with TFV-in-DcNP containing lopinavir/ritonavir introduced into systemic circulation by IV dosing indicates that greater than 90% of TFV remained associated with DcNP in blood circulation [24]. Therefore, the observed peaks followed by sustained levels in the time-course are likely due to a combination of lymphatic “first pass” absorption, cellular uptake and retention, and in-vivo formulation stability [24,26,27].

Collectively, the successful transformation of short-acting to long-acting TLD-in-DcNP may provide an all-in-one long-acting first-line, complete HIV treatment. This innovative approach utilizes lipid-stabilized drug-combination in suspensions, resulting in prolonged pharmacokinetics, enhanced AUC, and increased concentrations in PBMC. With dose adjustments, this novel all-in-one 3 HIV drug formulation opens new possibilities for long-acting HIV therapy, potentially transforming the lives of people living with HIV and HIV-HBV.