Analytical considerations for measuring the globule size distribution of cyclosporine ophthalmic emulsions.

Measurement of particle size and size distribution of complex drug products exhibiting complex rheological behaviors can be challenging as these properties may be beyond the theoretical assumptions of the measurement technique. Herein cyclosporine (CsA) ophthalmic emulsion was selected as a model complex system, and an in-depth assessment of particle size was performed using five fundamentally different particle sizing techniques, including dynamic light scattering (DLS), laser diffraction (LD), nanoparticle tracking analysis (NTA), cryogenic transmission electron microscopy (Cryo-TEM) and 2-dimensional diffusion ordered spectroscopy nuclear magnetic resonance (2D DOSY-NMR). The effect of various viscosity modifying and stabilizing excipients in the emulsions was assessed using four types of CsA formulations, i.e., 1) no viscosity modifying excipients, 2) carbomer copolymer type A (CCA), 3) Carbopol 1342, or 4) hydroxypropyl methyl cellulose (HMPC). In general, the variability of reported particle size increased, and is not as accurate, for emulsions dispersed in a non-Newtonian fluid and at higher emulsion concentrations. This effect was reduced in part by diluting the samples to lower volume fraction and a more Newtonian regime. To address the concern that sample dilution prior to measurement may induce physical instability in the emulsions, NTA was used to monitor average size at dilutions of up to 1:50,000. The size was found to remain constant and independent of the presence or type of stabilizer used. Cryo-TEM further confirmed that dilution did not alter particle size or morphology. Of the five evaluated techniques, Cryo-TEM and 2D DOSY NMR did not require dilution for measurement. The overestimate in DLS size measurements for certain CsA formulations was attributed to complex dispersant rheological behavior, particle-particle interactions, multiple light scattering events, and/or scattering interference from the polymers, which can be overcome by either testing under dilutions or by selecting one of the techniques less impacted by the interference of polymer.