C.albicans is the most prevalent of the Candidal species where > 80% of isolates are of this type (20). They can present in health and disease as yeast, pseudohyphae or hyphae(21, 22). The progression from yeast to hyphae by a phenomenon called “switching” is considered critical in the expression of their virulence, adhesion, tissue penetration and secretion of proteases(12). Under favorable condition the hyphae withstand adverse oral environment with the propensity for host tissue invasion by circumventing marcrophageal phagocytosis (23–28). Both PAS and CFW can detect fungal elements under direct microscopic examination.
To the best of our knowledge, only three studies had reported the diagnostic parameters comparing CFW and PAS in the identification of C.albican in OSCC (15, 29). Of them, only two had used PAS as reference standard where it is reported that CFW had a sensitivity and specificity of 4.9% and 4.3% more to that of PAS. Despite this, limited information is available as to how the intermediate test results (both false positive and false negative) were assessed in the final analysis particularly when the reference standards in itself in itself are not an error free gold standard.
Our results identified a sensitivity of 75.8% and specificity of 10.0% with a degree of concordance of 60.4% for C.albicans. Seventeen of the 43 tests results between the observers showed disagreement and had to be misclassified as either false positive or false negative. Our study is free of verification bias, the sensitivity and specificity values is in line with previous diagnostic accuracy study where CFW was compared to PAS in archival records of OSCC(15). In general, tests that have high sensitivity are often accompanied with low values for specificity. The high sensitivity is also accompanied with high rate of false positive that leads to disease free individuals being subjected to more invasive procedures. Mammogram is an excellent example of this.
There are many reasons why intermediate test results can arise, imperfect reference standard, technical reasons in classification due to retrospective analysis of archival data can pose a challenge while assessing the performance of diagnostic tests. Frequencies of up to 40% had been reported in literature for intermediate test outcomes that varies from one test to the other(30). Such findings can lead to erroneous outcome. Disease prevalence and sample size also play crucial role in diagnostic accuracy studies. Although, our study includes a small number of archival records yet it is not uncommon in diagnostic accuracy studies to have small sample numbers. As such, any generalization or blanked extrapolation of our findings will need to be interpreted with caution.
Accurate diagnosis of candida infection is important as some of the untreated CHC may endow dysplastic characteristics and a potential risk for the development of OSCC. But the critical question, was there candidal infection prior to the initiation of OSCC or the association was more of casual rather causal? is yet to be explored and longitudinal studies are needed to address this.