Ultra-Long-Term Delivery of Hydrophilic Drugs Using Injectable In Situ Cross-Linked Depots

Abstract

Achieving ultra-long-term release of hydrophilic drugs over several months remains a significant challenge for existing long-acting injectables (LAIs). Existing platforms, such as in situ forming implants (ISFI), exhibit high burst release due to solvent efflux and microsphere-based approaches lead to rapid drug diffusion due to significant water exchange and large pores. Addressing these challenges, we have developed an injectable platform that, for the first time, achieves ultra-long-term release of hydrophilic drugs for over six months. This system employs a methacrylated ultra-low molecular weight pre-polymer (polycaprolactone) to create in situ cross-linked depots (ISCD). The ISCD's solvent-free design and dense mesh network, both attributed to the ultra-low molecular weight of the pre-polymer, effectively minimizes burst release and water influx/efflux. In vivo studies in rats demonstrate that ISCD outperforms ISFI by achieving lower burst release and prolonged drug release. We demonstrated the versatility of ISCD by showcasing ultra-long-term delivery of several hydrophilic drugs, including antiretrovirals (tenofovir alafenamide, emtricitabine, abacavir, and lamivudine), antibiotics (vancomycin and amoxicillin) and an opioid antagonist naltrexone. Additionally, ISCD achieved ultra-long-term release of the hydrophobic drug tacrolimus and enabled co-delivery of hydrophilic drug combinations encapsulated in a single depot. We also identified design parameters to tailor the polymer network, tuning drug release kinetics and ISCD degradation. Pharmacokinetic modeling predicted over six months of drug release in humans, significantly surpassing the one-month standard achievable for hydrophilic drugs with existing LAIs. The platform's biodegradability, retrievability, and biocompatibility further underscore its potential for improving treatment adherence in chronic conditions.

Figure 1.. Injectable in situ crosslinked depot (ISCD) platform for sustained release of hydrophilic therapeutics.

A. The main component of ISCD is low molecular weight liquid methacrylated PCL, for example, PCLDMA. The liquid pre-polymer can suspend or dissolve both hydrophilic and hydrophobic drugs and can be easily injected through a standard 18–23 gauge needle. Upon adding an initiator (BPO) and accelerator (DMT) to PCLDMA, the pre-polymer mixture undergoes radical polymerization transitioning from a free-flowing liquid solution to a solid monolithic depot, resulting in physical encapsulation of the drug. B. ISCD has two key features enabling ultra-long-term release of hydrophilic drugs: a solvent-free design and a dense mesh network, both attributed to the ultra-low-molecular weight of the pre-polymer, PCLDMA. The liquid state of the pre-polymer obviates the need for a solvent, minimizing burst release. Cross-linking of the ultra-small chains of the pre-polymer results in a dense network (as shown in the SEM image of an ISCD depot formed in vitro) that limits water influx and efflux, minimizing the drug release rate. C. Design parameters to tailor the ISCD network to tune the drug release kinetics. Modulating the intrinsic factors, including decreasing the concentrations of BPO and DMT, or using higher molecular weight PCLDMA increases drug release. Additionally, adding external polymer additives including polyethylene glycol (PEG) and PCL-diol alongside PCLDMA can enhance the depot’s hydrophilicity, increasing drug release. External additives with different degrees of methacrylation (mono or di) can further tune the drug release. The cumulative release profiles of TAF from two different ISCD depots injected subcutaneously into rats are shown as examples of tailored drug release with varying release rates. Lowering the crosslinking density and increasing the hydrophilicity of the polymer chains achieve faster drug release.

Figure 2.. Synthesis and physiochemical characterization of the ISCD platform.

A. The cross-linking time for PCLDMA at different BPO/DMT concentrations is shown. Cross-linking time was measured as the point at which the viscosity of the pre-polymer mixture, monitored with a rheometer, begins to increase rapidly, as shown in the inset. B. Exothermic heat released during the cross-linking of ISCD with varying concentrations of BPO/DMT, measured using DSC. C. Injection parameters for the Hagen-Poiseuille equation used to calculate the injection force for PCLDMA, with or without TAF, using a 23-gauge needle. D. Viscosities of PCLDMA, with or without TAF, measured using a rheometer. E. Injection force calculated for PCLDMA, with or without TAF. The maximum acceptable injection force is 80 N. F. In vitro release profile of TAF in PBS (37°C) from ISCD comprising either PCLDMA or PCLTMA. G. In vitro release profile of TAF in PBS (37°C) from ISCD loaded with concentrations of TAF. H. Compressive stress-strain curves for ISCD with or without TAF, measured using a mechanical tester. I. Elastic moduli and yield stress for ISCD with or without TAF. J. ISCD can be polymerized ex vivo into various shapes such as cylinders, pipes, or disks that can be used as ultra-long-acting implants. Data in A, B, E-G, and I are presented as mean ± standard deviation (n=3, replicates performed at least twice). Data in D and H are representative of a single experiment (repeated three times).

Figure 3.. Tailoring the drug release kinetics and degradation of ISCDs in vitro and in vivo.

A. In vitro release profiles of TAF in PBS (37°C) from ISCDs prepared with PCLDMA of different molecular weights (630 Da and 2100 Da). B. In vitro release profiles of TAF in PBS (37°C) from ISCDs prepared with varying BPO/DMT concentrations. C. Crosslinking density of unmodified ISCD prepared with different concentrations of BMP/DMT and ISCD containing different external polymer additives (25 wt%) (*P<0.05 and **P<0.01). D. In vitro release profiles of TAF in PBS (37°C) from unmodified ISCD (prepared using PCLDMA only) or ISCD containing 25 wt% of an external polymer additive (PEG, PCL, or PDMS) alongside PCLDMA. E. In vitro release profiles of TAF and F. and percentage depot degradation in PBS (37°C) for unmodified ISCD or ISCD containing 25 wt% of PEG with varying degrees of methacrylation. G. SEM images of unmodified ISCD or ISCD containing different external polymer additives (25 wt%) showing the cross-section of depot structure at week 1 post-incubation in PBS (37°C). H. Plasma level of TFV in rats injected with 500 μl of TAF-loaded ISFI (control) or TAF-loaded unmodified ISCDs of ISCD containing 25 wt% PEGMMA. All depots were loaded with 90 mg/mL of TAF. The inset shows plasma levels up to day 30. (*P<0.05 for the overall comparison of plasma levels of the two ISCDs over the entire study duration). I. In vivo daily release rate and J. cumulative release of TAF from unmodified ISCD or ISCD containing 25 wt% of PEGMMA, as determined by PK modeling. K. Camera images of TAF-loaded ISFI or TAF-loaded unmodified ISCD or ISCD containing 25 wt% of PEGMMA, retrieved from rats at month 7 post-injection, and L. Remaining TAF amount in the depots (*P<0.05 and ***P<0.001) and M. Remaining mass of the depot (*P<0.05). Due to the disintegration of ISFI within the animal, the remaining mass of the ISFI depots could not be measured. N. MRI images of subcutaneously injected unmodified ISCD or ISCD containing 25 wt% PEGMMA at different time points. Data in A, B, C, E, F, and G are presented as mean ± standard deviation (n=3, experiments performed at least twice). Data in H, L, and M are presented as mean ± standard deviation of technical repeats (n=3, experiment performed twice). Data in I and J present predictions from PK modeling of the average plasma levels of TFV obtained experimentally. P-value in H was determined using two-way ANOVA with Bonferroni correction, with time and different ISCD formulations as the two variables. The P-value in C and L was determined using one-way ANOVA with Tukey’s post hoc analysis. The P-value in M was determined using Student’s t-test.

Figure 4.. In vivo biocompatibility and safety of ISCD.

A. Representative image (10X magnification) of an H&E stained section of local tissue, explanted with the ISCD depot one week after subcutaneous injection of 500 μl PCLDMA-based ISCD in rats. B. Left side shows high magnification (20X) representative images of H&E-stained sections of local tissue, explanted with the ISCD depot at different time points. Yellow arrows show inflammatory cells. The right side shows representative immunofluorescence images of local tissue sections explanted at different time points and stained against CD3 (green) and CD68 (red) markers to visualize T cells and macrophages, respectively. Yellow arrows show cells positive for CD3 or CD68. C. Fluorescence intensity quantified for CD3 and CD68 immunofluorescence. (*P<0.05, **P<0.01). D. Camera image taken during the procedure of retrieving ISCD from a rat, showing safe retrievability via a small incision. E. Plasma levels of TFV following ISCD removal. Data in C and E are presented as mean ± standard deviation of technical repeats (n=3). The P-value in C was determined using one-way ANOVA with Tukey’s post hoc analysis.

Figure 5.. The versatility of the ISCD platform and human PK prediction.

A. In vitro release profile of different drugs with varying water solubilities encapsulated into the ISCD platform. The release was studied in PBS (37°C). B. Correlation of cumulative release at day 1 with different drugs with varying water. C. In vitro release profile of TAF and FTC, when loaded into the ISCD platform individually versus in combination. The release was studied in PBS (37°C). D. In vitro release profile of ABC and LAM, when loaded into the ISCD platform individually versus in combination. The release was studied in PBS (37°C). E. Plasma concentration of NAL in rats subcutaneously injected with 500 μl of NAL-loaded ISCD (45 mg/ml NAL). F. Whole blood concentration of TAC in rats subcutaneously injected with 500 μl of TAC-loaded ISCD (28 mg/ml TAC). In vivo daily release rate and cumulative release profile of G. NAL and H. TAC, as predicted by PK modeling of systemic drug levels in rats, following subcutaneous injection of 500 μl of NAL- or TAC-loaded ISCD. I. Convolution analysis-based prediction of human PK of a single subcutaneous dose of NAL-loaded ISCD in comparison to clinically established PK profile of once-daily oral dose of NAL (green region), and once monthly intra-muscular injection – Vivitrol^® (purple lines). J. Convolution analysis-based prediction of human PK of a single subcutaneous dose of TAC-loaded ISCD (at different dosages) in comparison to clinically established PK profile of twice-daily oral doses of TAC (green region). Data in A-D are presented as mean ± standard deviation (n=3, experiments performed at least twice). Data in E and F are presented as mean ± standard deviation of technical repeats (n=3). Data in G and H present predictions from PK modeling of the average plasma level of TAC and NAL obtained experimentally. Data in I and J present human PK prediction based on convolution analysis of experimentally obtained PK data of TAC and NAL in rats.