Oral Leukoplakia Microbiome Predicts the Degree of Dysplasia and is Shaped by Smoking and Tooth Loss

ABSTRACT Objective This study aimed to determine if the oral potentially malignant disorder, oral leukoplakia (OLK), exhibited microbiome changes that predict the degree of dysplasia and the risk of malignant progression. Results We examined the microbiome in 216 swabs of OLK from 177 patients. Compared to healthy controls ( n = 120 swabs from 61 patients), who were less likely to smoke and had better oral health, OLK patients exhibited an increased abundance of Rothia mucilaginosa , Streptococcus parasanguinis and S. salivarius , resembling acetaldehyde generating communities described previously. Compared to the patients' healthy contralateral normal (CLN) mucosa ( n = 202), which acts as a matched control for oral health parameters, OLK exhibited increased S. infantis , Leptotrichia spp., Bergeyella spp., Porphyromonas spp. and F. nucleatum . Machine learning with clinical and microbiome data could discriminate high‐risk dysplasia (moderate to severe) from low‐risk dysplasia (none or mild) (sensitivity 87.4%; specificity 76.5%). Follow‐up swabs were recovered from 58 patients, eight of whom progressed to a higher grade of dysplasia or OSCC and these eight patients exhibited a higher abundance of Fusobacterium species at their initial presentation. Conclusions Our study suggests that the OLK microbiome has potential to be an aid to the prediction of dysplasia grade and the risk of malignant transformation.