“Checkerboard” DNA‐Probe Analysis and Anaerobic Culture of Initial Periodontal Lesions

In this study, we compared a new, rapid DNA-probe assay with anaerobic culture for characterizing the microbiology of initial periodontal lesions in adults. We used the microbiological data from both assays to identify the microbiota associated with the conversion from gingival health to periodontal disease. In previous studies of the microbiota of periodontal infections, investigators have mainly examined the microbiota of periodontal pockets in patients with established disease and have used anaerobic cultural methods. Little information is available about the microbiota associated with the initiation of periodontitis [1]. An understanding of the initial stages of periodontal disease (i.e., when loss of attachment first occurs around a previously healthy tooth) will be important for risk assessment and prevention of these infections. By using cultural methods, in which the predominant species are isolated in pure culture and identified by phenotypic tests, all isolates can be characterized, but these methods are labor intensive and time-consuming and therefore limit the number of samples that can be analyzed. To facilitate large-scale clinical studies, which are necessary to assess microbiological risk markers and risk factors for periodontal disease, more-rapid methods for assaying multiple bacterial species in plaque samples will be required. We assessed one such method, the "checkerboard" DNA-probe assay [2], which enables simultaneous analysis of 28 plaque samples with - 1.5-mm increase in periodontal attachment loss during monitoring. Subgingival plaque from active and inactive sites was sampled for anaerobic culture. Fifty isolates from each sample were cultured and identified on the basis of gram staining, growth in air, fluorogenic substrate tests [3], and SDS-PAGE of whole cell proteins [4]. The cultural data were expressed as the percentage of total cultivable microbiota represented by each species identified in each sample. We calculated the absolute level (count of cfu per