The Diagnostic Accuracy of Autofluorescence Microscopy for Oral Candidiasis

Objectives: The purpose of this study was to evaluate the sensitivity and specificity of screening Papanicolaou-stained oral cytology specimens using UV LED autofluorescent microscopy to diagnose subclinical oral candidal infections. Methods: Oral mucosal specimens were sampled from patients presenting at the dental clinics of the University of the Western Cape, in Cape Town, South Africa (n=80). The dorsal tongue, buccal mucosae, and labial vestibulae were sampled using brush cytology. Two parallel exfoliative smears were prepared from each brushing sample. Each slide pair was stained, one with Periodic Acid-Schiff (PAS) and the corresponding slide with Papanicolaou (Pap) stain. The slides were then screened by medical cytologists of the NHLS at Tygerberg Hospital. The baseline screening results formed the control group for this study. The slides were randomly numbered and then viewed under LED UV microscopy by an observer, A. Slides were then randomly renumbered and viewed again, by observer B. Results were then correlated against the control group to determine sensitivity, specificity, and diagnostic accuracy. Results: Pap stained exfoliated oral squamous cells and oral microflora autofluoresce under LED UV light. Candidal hyphae contrast distinctly bright green in colour. Results indicate a sensitivity of 39.4% and specificity of 83% for observer A. Sensitivity and specificity for observer B were 30.3% and 95.7% . Diagnostic accuracy for observer A was 63.8%, 67.5% for observer B. Conclusion: Orientation of pathogens and squamous cells affect screening results. Pathogens otherwise missed in regular microscopy viewing may be identified under UV microscopy, and vice versa. After adequate calibration autofluorescence screening for Candida spp. could be a good complimentary examination method to normal light microscopy.