OBJECTIVES: To evaluate a noninvasive method of bacterial identification via fluorescence spectroscopy in the setting of acute otitis media in a chinchilla model. STUDY DESIGN: Prospective. METHODS: For each chinchilla, transbullar inoculation with Streptococcus pneumoniae, Haemophilus influenzae, or Staphylococcus aureus was performed bilaterally and clinical infection was determined by otoscopy. An optical fiber coupled to a spectrofluorometer allowed for the delivery of an excitation wavelength to the inflamed tympanic membrane and the acquisition of the resulting emission signal. Sequential emission spectra obtained over a range of excitation wavelengths were assembled by a computer algorithm, and a single, three-dimensional plot was created for each test ear. Similarly, plots from the healthy external auditory canal (EAC) were also recorded. Twelve animals were used to establish a library of four reference plots representing the three bacteria and the EAC. Of the 24 ears available for study, 10 were excluded from analysis because of lack of clinical infection or presence of tympanic membrane perforation with purulent drainage. From four additional animals, four samples from ears infected with the above bacteria and three samples of the EAC served as unknowns. The unknown plots were analyzed by an investigator blinded to their identity. RESULTS: Using a multiple correlation of the unknown to the reference plots, seven of seven samples were correctly identified. CONCLUSIONS: We were able to establish a prototype method for the noninvasive identification of a limited library of pathogens in a chinchilla model of acute otitis media.
OBJECTIVES: To evaluate a noninvasive method of bacterial identification via fluorescence spectroscopy in the setting of acute otitis media in a chinchilla model. STUDY DESIGN: Prospective. METHODS: For each chinchilla, transbullar inoculation with Streptococcus pneumoniae, Haemophilus influenzae, or Staphylococcus aureus was performed bilaterally and clinical infection was determined by otoscopy. An optical fiber coupled to a spectrofluorometer allowed for the delivery of an excitation wavelength to the inflamed tympanic membrane and the acquisition of the resulting emission signal. Sequential emission spectra obtained over a range of excitation wavelengths were assembled by a computer algorithm, and a single, three-dimensional plot was created for each test ear. Similarly, plots from the healthy external auditory canal (EAC) were also recorded. Twelve animals were used to establish a library of four reference plots representing the three bacteria and the EAC. Of the 24 ears available for study, 10 were excluded from analysis because of lack of clinical infection or presence of tympanic membrane perforation with purulent drainage. From four additional animals, four samples from ears infected with the above bacteria and three samples of the EAC served as unknowns. The unknown plots were analyzed by an investigator blinded to their identity. RESULTS: Using a multiple correlation of the unknown to the reference plots, seven of seven samples were correctly identified. CONCLUSIONS: We were able to establish a prototype method for the noninvasive identification of a limited library of pathogens in a chinchilla model of acute otitis media.
Authors: Gislaine Vieira-Damiani; Marna Elise Ericson; Marilene Neves da Silva; Kalpna Gupta; Tânia Benetti Soares; Amanda Roberta de Almeida; Vitor Bianchin Pelegati; Mariana Ozello Baratti; Carlos Lenz Cesar; Maria Letícia Cintra; Paulo Eduardo Neves Ferreira Velho Journal: J Trop Dis Public Health Date: 2016-02-20