Accuracy of special histochemical staining methods in diagnosis of oral pathology: A systematic review and meta‑analysis

Maryam Ghelichli, Abdolhalim Rajabi, Niyayesh Mirsaeedi, Masoud Mohammadi, Sina Sohrabi, Niloufar Darabi

Abstract


Clinical decision‑making and biomedical research heavily rely on imaging techniques to visualize
tissue morphology. To examine tissues in detail, it is necessary to use special histochemical stains
to enhance contrast. This meta‑analysis aimed to assess the sensitivity and specificity of these
stains in diagnosing oral pathologic specimens. We conducted a search in 8 databases, including
EMBASE, PubMed, Web of Science, Scopus, ProQuest, Ovid, Cinahl, and Cochrane, up to June 2022.
Of 87,393 studies, 41 articles were selected for inclusion in our study. The results revealed that the
specificity and sensitivity of the special histochemical stains were 86% with confidence interval (CI)
95%: 80%–90% and 83% with CI 95%: 75%–89%, respectively. Among the stains evaluated, toluidine
blue, Papanicolaou, silver stain, Giemsa, Gram, feulgen, and periodic acid–Schiff (PAS) were the most
frequently used for the detection of malignancy, premalignant lesions, dysplasia, and candidiasis. The
specificity and sensitivity of each stain were analyzed individually, considering the type of specimen.
Toluidine blue was the most commonly utilized special histochemical stain, particularly effective, for
detecting malignancy, with a specificity of 97% with CI 95%: 88%–99% and sensitivity of 76% with
CI 95%: 56%–89%. In conclusion, special histochemical stains are effective in diagnosing oral lesions,
exhibiting reasonable specificity and sensitivity, especially in cases of premalignant and malignant
lesions. Based on the reviewed articles in our study, the silver stain was identified as highly sensitive,
while Giemsa and Papanicolaou stain exhibited the highest specificity.
Key Words: Accuracy, oral pathology, sensitivity, special histochemical stain, specificity

 

 

Highlight

Niyayesh Mirsaeedi: Pubmed,Google Scholar


Full Text:

PDF xml

Refbacks

  • There are currently no refbacks.