Comparison of two otitis media models for the study of middle ear antimicrobial pharmacokinetics

Abstract

We compared two models of acute otitis media that estimate middle ear antimicrobial pharmacokinetics. Using a crossover study design, we compared a systemic drug administration model with a diffusion model we devised that measures the disappearance of antimicrobials from the middle ear. We induced acute otitis media in 14 chinchillas by inoculating S. pneumoniae into the middle ear, then administered 3 antimicrobials: amoxicillin, trimethoprim, and sulfamethoxazole. Next we collected middle ear fluid samples to analyze drug concentrations and compare rate constants for the systemic and diffusion models by analysis of variance. We found that amoxicillin K values were not affected by model testing sequence (p = 0.827) or model type (systemic versus diffusion, p = 0.310), nor were sulfamethoxazole K values: model testing sequence (p = 0.917), model type (p = 0.963). Trimethoprim K values were also not affected by model testing sequence (p = 0/760), but were by model type (p = 0.0001). Trimethoprim elimination from the diffusion model was faster (K = 0.33 +/- 0.17 versus 0.57 +/- 0.09 hr-1) than from the systemic model, although it appears this was caused by sampling before drug distribution into the middle ear was complete. In conclusion, it appears K values derived from either systemic antimicrobial administration or direct middle ear instillation are similar for assessing middle ear antimicrobial pharmacokinetics, and these models can be used interchangeably to study factors affecting otitis media treatment response.