Self-aggregating long-acting injectable microcrystals

Abstract

Injectable drug depots have transformed our capacity to enhance medication adherence through dose simplification. Central to patient adoption of injectables is the acceptability of needle injections, with needle gauge as a key factor informing patient discomfort. Maximizing drug loading in injectables supports longer drug release while reducing injection volume and discomfort. Here, to address these requirements, we developed self-aggregating long-acting injectable microcrystals (SLIM), an injectable formulation containing drug microcrystals that self-aggregate in the subcutaneous space to form a monolithic implant with a low ratio of polymer excipient to drug (0.0625:1 w/w). By minimizing polymer content, SLIM supports injection through low-profile needles (<25 G) with high drug loading (293 mg ml^-1). We demonstrate in vitro and in vivo that self-aggregation is driven by solvent exchange at the injection site and that slower-exchanging solvents result in increased microcrystal compaction and reduced implant porosity. We further show that self-aggregation enhances long-term drug release in rodents. We anticipate that SLIM could enable low-cost interventions for contraceptives.

Keywords: Drug delivery; Materials for devices.

Fig. 1. Overview of SLIM.

a, Schematic of the self-injection procedure. b, An image of different needle gauges ranging from 18 G to 28 G, compared against a grain of rice. c, Schematic of subcutaneous environment highlighting the solvent-exchange-driven self-aggregation of microcrystals into a compacted implant. Illustrations in a and c by Virginia E. Fulford, Alar Illustration.

Fig. 2. In situ self-aggregation of drug crystals via solvent exchange.

a,b, A visual representation of the effect of three different solvents with different miscibility with the aqueous exchange medium: (i) NMP in PBS (perfectly miscible), (ii) BA (partially miscible) in PBS, (iii) BB in PBS with 10% SDS (partially miscible) and (iv) BB in PBS (immiscible). The figure illustrates the effect of the three solvents on the self-aggregation dynamics (a) and on the mechanical integrity of the equilibrated depots (b). Scale bars, 1 mm. c, Self-supported concave structure observed for 50 mg ml⁻¹ of LNG in BA. Scale bar, 300 μm. d, Enhanced self-supported concave structure observed for 50 mg ml⁻¹ of LNG in BA with small addition of PCL. Scale bar, 1 mm. e, No concave structure observed for slower-exchanging system of 50 mg ml⁻¹ LNG in BB with PBS/5 wt.% SDS. Scale bar, 500 μm. f,g, Droplets of 50 mg ml⁻¹ LNG in BB shrank in CCA mode over 4 days in PBS with 5% SDS (f), with an approximately constant contact angle and a continually shrinking contact radius (g). Scale bars, 1 mm. h, Schematics illustrating the compaction of LNG crystals into a dense sphere as a result of droplet shrinking with an unpinned contact line. Source data

Fig. 3. Assessment of solvent exchange in vivo.

a–c, Ultrasound images depicting the clearance of BB within subcutaneous tissue over 34 days (a), PBS clearance over 8 days (b) and NMP clearance over 20 days (c). Yellow dashed circles indicate the contour of the injected solvent. Scale bars, 5 mm.

Fig. 4. Mechanical and rheological characterization of SLIM.

a–d, Changes in storage modulus (a) and tan delta (b) of LNG–BB and storage modulus (c) and tan delta (d) of LNG–BB + PCL in a medium of 5% SDS in PBS over 7 days. e, Axial force as a function of compressive displacement on depots self-aggregated from BB formulations. Increasing PCL concentration increases the mechanical integrity of the depots. f, Injection force at 6 ml min⁻¹ as a function of needle size for LNG–BB (293 mg ml⁻¹, light blue) and LNG–BB (293 mg ml⁻¹) + PCL (1.62 wt.%, dark blue) formulations, highlighting the self-injectability threshold (red). Source data

Fig. 5. In vivo demonstration of SLIM.

a, In vivo release of LNG–PBS, LNG–BB and LNG–BB + PCL over 97 days. b, Magnified version for the initial time points up to day 12. c–e, µCT images of LNG–PBS (c), LNG–BB (d) and LNG–BB + PCL (e) depots excised from rats after 97 days, highlighting the higher packing efficiency achieved from formulations with BB compared with PBS. The region of interest (dashed red) is 200 × 200 pixels. Scale bars, 2 mm. n = 3 per group. Data represent mean ± s.d. Source data