(M1330-07-38) Proof of Concept In Situ Salt Formation during Amorphous Solid Dispersion Fabrication to Improve Physical Stability and Drug Release Performance

Purpose: Amorphous solid dispersions (ASD) represent a promising strategy to enhance the solubility and dissolution rate of poorly water-soluble compounds. However, the risk of drug crystallization remains a significant challenge for ASDs, potentially compromising their physical stability and release performance. Salt formation is a common technique used to modify the physicochemical properties of ionizable drugs. Notably, although not extensively studied, salt formation can alter the crystallization tendency of a drug. The strong electrostatic interactions between the drug and counterions likely play a role in stabilizing the amorphous form of the drug. Given that salt formation can affect crystallization tendencies, combining salt formation with ASDs presents a potential strategy to inhibit the crystallization of rapidly crystallizing drugs while maintaining the benefits of an amorphous formulation. This approach is particularly relevant for weakly basic compounds with a high propensity for crystallization, such as delamanid (DLM), where ASDs containing the drug salt are of significant interest. The objective of the current study was to enhance the physical stability and release performance of DLM in ASDs containing the drug salt. We compared ASDs fabricated in situ via a single step with systems prepared from DLM salts. We hypothesized that DLM salts with hydrochloric acid (HCl) would exhibit greater resistance to solid-state crystallization than the corresponding amorphous free base. Additionally, maintaining the amorphous nature of the drug salt in the ASDs was expected to result in improved release performance.

Methods: Crystalline and amorphous salt form of DLM free base and HCl (1:1 molar ratio) were prepared by antisolvent and solvent evaporation methods, respectively. Subsequently, ASDs containing the drug salt were fabricated by dissolving the crystalline or amorphous DLM chloride salt and hydroxypropyl methylcellulose phthalate (HPMCP) in methanol:DCM (1:1 v/v) mixture. Alternatively, DLM free base and HCl (1:1 molar ratio) as well as HPMCP were simultaneously dissolved in the same solvent mixture. In the three ASD systems, the drug loading was kept at 25% w/w. For comparison purpose, an additional ASD containing DLM free base was prepared with the same drug loading of 25%. The solvents were removed by rotary evaporation. ASD samples were kept under a vacuum, cryo-milled, and sieved to obtain the particle size fraction of 106–250 μm. The ASDs were characterized by X-ray diffraction (XRD), differential scanning calorimetry (DSC), Fourier transform infrared (FTIR) and ¹H nuclear magnetic resonance (NMR) spectroscopy. Release performance of the ASDs were evaluated using a single-stage or two-stage dissolution test consisting of an initial low pH gastric stage followed by transfer to a higher pH medium corresponding to intestinal pH conditions.

Results: XRD, DSC, ¹H NMR and FTIR data all point to the successful generation of DLM chloride salt. Firstly, X-ray diffractograms exhibited distinct patterns of crystalline DLM base or salt versus their amorphous counterparts. Secondly, DSC thermograms showed a dramatic jump in the glass transition temperature (T_g) from 43^oC for the amorphous free base to 108^oC for the amorphous salt. Thirdly, the solution-state ¹H NMR revealed that the peaks assigned to protons of the piperidine ring of the free base significantly shifted downfield in the salt. Fourthly, protonation of a nitrogen group is further confirmed by FTIR spectra, with the appearance of a very broad and intense peak within the range 2380-2770 cm⁻¹ in the salt, which was assigned to the characteristic tertiary N⁺-H stretching vibration (Figure 1). For the ASDs, X-ray diffractograms showed that all dispersions of DLM chloride salt were amorphous whereas the ASD containing 25% base remained partially crystalline (Figure 2A). Consequently, three ASD samples of DLM salts exhibited similar DSC thermograms with a single T_g at around 112^oC without recrystallization upon further heating. The ASD of the free base showed a much lower T_g at 68^oC, followed by a recrystallization exotherm starting from 85^oC prior to fusion at 179^oC (Figure 2B). The ASDs of DLM salts had similar FTIR spectra. The XRD, DSC and FTIR data are in good agreement with each other and indicate that the three ASD systems (prepared from the crystalline salt, amorphous salt or in situ using the free base and HCl) are chemically and physically equivalent. The dissolution performance of ASD samples was displayed in Figure 3. Clearly, the three ASDs of DLM salts showed similar release patterns and were significantly superior to DLM base in term of drug release in both single-stage and pH-shift dissolution.

Conclusion: Crystallization represents a primary failure mechanism for ASDs. Herein, we have demonstrated that the tendency for drug crystallization can be modulated through salt formation. By employing in situ salt formation during single-step ASD manufacture, drug crystallization was effectively inhibited, and the release profiles were enhanced. Thus, the combination of salt formation with ASD formulation offers a dual strategy to address the dissolution and solubility challenges associated with poorly soluble compounds that are difficult to formulate using a single enabling approach.

Acknowledgements: The authors gratefully acknowledge the Bill and Melinda Gates Foundation, Seattle, WA for funding this study through award number OPP1159809.

Figure 1. FTIR spectra of (A) crystalline DLM base and salt, and (B) amorphous DLM base and salt.

Figure 2. (A) X-ray diffractograms and (B) DSC thermograms of ASD samples.

Figure 3. Drug release profiles of ASD samples containing 25% drug loading in (A) pH 6.5 phosphate buffer, and (B) pH-shift experiment with dashed line indicates the pH change from pH 1.6 to pH 6.5.