Identification of Actinobacillus actinomycetemcomitans in subgingival plaque by PCR

The purpose of this study was to assess the sensitivity and specificity of the PCR in detecting Actinobacillus actinomycetemcomitans. The PCR's detection capability was compared with those of three other methods: culture-enhanced PCR (CE-PCR), colony hybridization (CH), and conventional culture with presumptive biochemical identification. A 285-bp stretch of the leukotoxin gene lktA of A. actinomycetemcomitans was amplified by PCR with primers TT-15 and TT-16. For CH, the PCR product was labeled with digoxigenin and used as a hybridization probe. Nucleotide sequence analysis of the PCR product of A. actinomycetemcomitans 1D4 and 1664 and three clinical isolates revealed complete homology among the tested strains, with only one base substitution (at position 1344) in comparison with the published sequence. With artificially infected subgingival plaque, the detection limit of PCR for A. actinomycetemcomitans was 10(3) CFU/ml of plaque suspension. Culturing subgingival plaque on tryptic soy-serum-bacitracin-vancomycin agar prior to PCR (CE-PCR) improved the limit of detection to 10(2) CFU/ml. Analysis of subgingival plaque samples from 35 patients with periodontal disease and 10 periodontally healthy subjects revealed that CE-PCR and CH had the highest overall rate of A. actinomycetemcomitans detection (both 58%), followed by PCR and culture (both 42%). With CH as the "gold standard", the sensitivities of CE-PCR, PCR, and culture were 88, 65, and 58%, respectively; the specificities were 84, 89, and 79%, respectively. The CE-PCR provided acceptable positive and negative predictive values (> or = 70%) when the prevalence of A. actinomycetemcomitans varied between 30 and 70%. PCR alone provided comparable predictive values over a narrower range of prevalence rates (30 to 50%), while culture did not afford acceptable predictive values at any prevalence rate. PCR and CE-PCR were found to be superior to culture with presumptive biochemical identification and should be the preferred methods for the detection of A. actinomycetemcomitans in subgingival plaque.