Evaluation of the use of paired modified Wright's and periodic acid Schiff stains to identify microbial aggregates on cytological smears of dogs with microbial otitis externa and suspected biofilm.

Abstract

Micro-organisms associated with canine otitis externa (OE) may cause biofilm-associated infections (BAI). A key component of biofilm is microbial aggregate and extracellular polymeric substance (EPS). Periodic acid Schiff (PAS) can stain polysaccharide EPS in human otitis media with effusion, but this has not been tested in canine OE. There is no cytological definition for microbial aggregate, and definitive methods for identifying BAI in a clinical setting in canine OE have not been defined. To establish whether PAS stain can identify polysaccharide matrix on cytological smears; and to determine the reproducibility of identification of microbial aggregates within a discrete area of stained matrix, using paired modified Wright's and PAS-stained smears. Forty privately-owned dogs presenting to a dermatological referral practice. In this prospective, cross-sectional study, three investigators independently and blindly classified 40 paired modified Wright's-PAS slide sets into groups: aggregate-associated infection (AAI) and non-AAI (n=27); and control (n=13). Agreement between investigators for presence of AAI was measured using Fleiss' kappa statistic (FK). Agreement between investigators and dermatologists for presence of AAI upon cytological evaluation, and suspected BAI based on clinical examination, was measured using Cohen's kappa statistic. The matrix was confirmed to stain PAS-positive. Interinvestigator agreement for AAI was very good using PAS (0.82 FK) and fair using modified-Wright's (MW) (0.33 FK). Reproducible cytological features associated with AAI were the presence of: three or more distinct aggregates (0.76 FK); discrete areas of PAS-positive matrix (0.70 FK); and the presence of high-density material (0.70 FK) using PAS stain. PAS can stain the extracellular matrix on otic smears, and a novel protocol for reproducible identification of cytological features such as microbial aggregates has been established.