Grading systems of oral cavity pre-malignancy: a systematic review and meta-analysis

Machine learning methods in predicting the risk of malignant transformation of oral potentially malignant disorders: A systematic review

BACKGROUND

Oral Potentially Malignant Disorders (OPMDs) refer to a heterogenous group of clinical presentations with heightened rate of malignant transformation. Identification of risk levels in OPMDs is crucial to determine the need for active intervention in high-risk patients and routine follow-up in low-risk ones. Machine learning models has shown tremendous potential in several areas of dentistry that strongly suggest its application to estimate rate of malignant transformation of precancerous lesions.

METHODS

A comprehensive literature search was performed on Pubmed/MEDLINE, Web of Science, Scopus, Embase, Cochrane Library database to identify articles including machine learning models and algorithms to predict malignant transformation in OPMDs. Relevant bibliographic data, study characteristics, and outcomes were extracted for eligible studies. Quality of the included studies was assessed through the IJMEDI checklist.

RESULTS

Fifteen articles were found suitable for the review as per the PECOS criteria. Amongst all studies, highest sensitivity (100%) was recorded for U-net architecture, Peaks Random forest model, and Partial least squares discriminant analysis (PLSDA). Highest specificity (100%) was noted for PLSDA. Range of overall accuracy in risk prediction was between 95.4% and 74%.

CONCLUSION

Machine learning proved to be a viable tool in risk prediction, demonstrating heightened sensitivity, automation, and improved accuracy for predicting transformation of OPMDs. It presents an effective approach for incorporating multiple variables to monitor the progression of OPMDs and predict their malignant potential. However, its sensitivity to dataset characteristics necessitates the optimization of input parameters to maximize the efficiency of the classifiers.

Calibration improves the agreement in grading oral epithelial dysplasia-Findings from a National Workshop in Malaysia

INTRODUCTION

A major pitfall of many of the established oral epithelial dysplasia (OED) grading criteria is their lack of reproducibility and accuracy to predict malignant transformation. The main objective of this study was to determine whether calibration of practicing oral pathologists on OED grading could improve the reproducibility of the WHO 2017 and the binary OED grading systems.

METHODS

A nationwide online exercise was carried out to determine the influence of calibration on the reproducibility of the WHO 2017 and the binary OED grading systems.

RESULTS

A significant improvement was observed in the inter-observer agreement for the WHO 2017 OED grading system (K 0.196 vs. 0.448; Kw 0.357 vs. 0.562) after the calibration exercise. The significant difference (p = 0.027) in the level of agreement between those with five or more years and less than 5 years of experience was no more observed (p = 0.426) after the calibration exercise. The percent agreement for binary grading was significantly higher (91.8%) for buccal mucosal lesions as compared to lesions on the tongue after the calibration exercise.

CONCLUSION

This study validates the significance of calibration in improving the reproducibility of OED grading. The nationwide exercise resulted in a statistically significant improvement in the inter-observer agreement for the WHO 2017 OED grading system among a large number of oral pathologists. It is highly recommended that similar exercises should be organized periodically by professional bodies responsible for continuing education among oral pathologists to improve the reliability of OED grading for optimal treatment of oral potentially malignant disorders.

Deep learning of cell spatial organizations identifies clinically relevant insights in tissue images

Recent advancements in tissue imaging techniques have facilitated the visualization and identification of various cell types within physiological and pathological contexts. Despite the emergence of cell-cell interaction studies, there is a lack of methods for evaluating individual spatial interactions. In this study, we introduce Ceograph, a cell spatial organization-based graph convolutional network designed to analyze cell spatial organization (for example,. the cell spatial distribution, morphology, proximity, and interactions) derived from pathology images. Ceograph identifies key cell spatial organization features by accurately predicting their influence on patient clinical outcomes. In patients with oral potentially malignant disorders, our model highlights reduced structural concordance and increased closeness in epithelial substrata as driving features for an elevated risk of malignant transformation. In lung cancer patients, Ceograph detects elongated tumor nuclei and diminished stroma-stroma closeness as biomarkers for insensitivity to EGFR tyrosine kinase inhibitors. With its potential to predict various clinical outcomes, Ceograph offers a deeper understanding of biological processes and supports the development of personalized therapeutic strategies.

Deep learning-based pathology image analysis predicts cancer progression risk in patients with oral leukoplakia

BACKGROUND

Oral leukoplakia (OL) is associated with an increased risk for oral cancer (OC) development. Prediction of OL cancer progression may contribute to decreased OC morbidity and mortality by favoring early intervention. Current OL progression risk assessment approaches face large interobserver variability and is weakly prognostic. We hypothesized that convolutional neural networks (CNN)-based histology image analyses could accelerate the discovery of better OC progression risk models.

METHODS

Our CNN-based oral mucosa risk stratification model (OMRS) was trained to classify a set of nondysplastic oral mucosa (OM) and a set of OC H&E slides. As a result, the OMRS model could identify abnormal morphological features of the oral epithelium. By applying this model to OL slides, we hypothesized that the extent of OC-like features identified in the OL epithelium would correlate with its progression risk. The OMRS model scored and categorized the OL cohort (n = 62) into high- and low-risk groups.

RESULTS

OL patients classified as high-risk (n = 31) were 3.98 (95% CI 1.36-11.7) times more likely to develop OC than low-risk ones (n = 31). Time-to-progression significantly differed between high- and low-risk groups (p = 0.003). The 5-year OC development probability was 21.3% for low-risk and 52.5% for high-risk patients. The predictive power of the OMRS model was sustained even after adjustment for age, OL site, and OL dysplasia grading (HR = 4.52, 1.5-13.7).

CONCLUSION

The ORMS model successfully identified OL patients with a high risk of OC development and can potentially benefit OC early diagnosis and prevention policies.

Association between socioeconomic status and severity of oral epithelial dysplasia using a Taiwanese Nationwide Oral Mucosal Screening Program: a retrospective analysis

BACKGROUND

The study aimed to investigate the association between socioeconomic status and severity of oral epithelial dysplasia (OED) using current data from the Taiwanese Nationwide Oral Mucosal Screening Program (TNOMSP).

METHODS

This retrospective analysis was conducted in the Department of Oral and Maxillofacial Surgery at a general hospital in Taipei, Taiwan. A total of 134 participants were analysed from a previous study database of 150 patients. The inclusion criteria included age > 20 years and a history of either tobacco or betel nut use. Background information, including para-habits such as betel and tobacco use, was analysed using the Pearson chi-square (χ2) test; furthermore, the correlation of background information with OED severity was investigated using logistic regression (mild or moderate/severe).

RESULTS

High school education level (P < 0.001), poor self-awareness (P = 0.002), current betel use (P < 0.001), and tobacco use (P = 0.003) were highly correlated with moderate- and severe OED (P < 0.05). The odds ratio (OR) of education status above senior high school was 0.03 (95% confidence interval [CI] 0.01-0.15, P < 0.001), while that of junior high school was 1. Current betel chewing (OR 6.57 [95% CI 1.17-37.0], P = 0.033) was significantly associated with OED severity compared with never or ex-use of betel.

CONCLUSIONS

We found a strong correlation between the severity of OED and current betel use and low education status. The current study revealed that the socioeconomic status, poor self-awareness, and para-habit history of the patients with OED should be evaluated to identify high-risk individuals using TNOMSP.

Oral Epithelial Dysplasia Grading Systems: A Systematic Review & Meta-Analysis

This systematic review and meta-analysis aims to provide a robust qualitative and quantitative analysis of the different systems used to assess the grade of oral epithelial dysplasia (OED). This study was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-analyzes (PRISMA) statement. Six electronic databases were searched for primary research published over the past four decades. Overall quality and level of evidence were based on the Johns Hopkins Research Evidence Appraisal Tool, while evidence of heterogeneity was determined by the Q-statistic and I^{^2} statistic. Evidence of publication bias was determined using Egger's regression and the Rank correlation tests. A total of 170 records were identified. Only 9 primary research articles were included in the qualitative systematic review. Four studies (4/9) were included in the final quantitative meta-analysis. The grading systems analyzed included the WHO, binary, Ljubljana, Smith and Pindborg, Brothwell, and the oral intraepithelial neoplasia. The results demonstrate the binary system to be superior to the WHO system in grading OED, by providing better inter-observer agreement, however, the substantial error among the inter-observer κ values analyzed indicates the significance of this finding to be of minimal impact. Lack of reliable reproducibility of the grading systems and lack of common effect size (heterogeneity analysis) were noted.

The clinical utility of contemporary oral epithelial dysplasia grading systems

INTRODUCTION

Clinical management of oral potentially malignant disorders relies on accurate histopathological assessment of the presence and grade of oral epithelial dysplasia. Whilst adjunctive laboratory tests have provided useful prognostic information, none are in widespread clinical use. This study was performed to assess the clinical utility of two contemporary oral epithelial dysplasia grading systems.

METHODS

Patients were identified from a clinical database. Oral epithelial dysplasia grading was performed by three oral and maxillofacial pathologists blinded to clinical outcome using the WHO 2017 system and a binary classification. The primary outcome measure was the development of oral squamous cell carcinoma, termed 'malignant transformation'.

RESULTS

131 cases satisfied the inclusion criteria, of which 23 underwent malignant transformation. There was substantial inter-rater agreement between the study pathologists for both grading systems, measured using kappa statistics (κ = 0.753-0.784). However, there was only moderate agreement between the consensus WHO 2017 dysplasia grade for the study against the original grade assigned by a pool of six pathologists in the context of the clinical service (κ = 0.491). Higher grade categories correlated with an increased risk of developing cancer using both grading systems.

CONCLUSION

This study demonstrates that the WHO 2017 and binary grading systems are reproducible between calibrated pathologists and that consensus reporting is likely to improve the consistency of grading. The WHO and binary systems were prognostically comparable. We recommend that institutions implement consensus oral epithelial dysplasia grading and prospectively audit the effectiveness of risk stratifying their patients with oral potentially malignant disorders. (249 words).

Identification of specific clinical risk factors associated with the malignant transformation of oral epithelial dysplasia

BACKGROUND

Factors that increase the risk of malignant transformation of oral epithelial dysplasia (OED) are not completely elucidated.

METHODS

A retrospective chart review was performed assessing risk factors for transformation of OED, and cancer staging for transformed cases at Sunnybrook Health Sciences Centre.

RESULTS

Two-hundred four patients were diagnosed with OED, and 16.7% (34) underwent malignant transformation. Risk factors associated with transformation included: heavy tobacco smoking, excessive EtOH consumption, non-homogenous leukoplakia, size >200 mm² , moderate dysplasia or greater than moderate, progression of dysplasia grades, and immunosuppression. Transformed cases followed for a dysplastic lesion were associated with a stage-I cancer diagnosis, and cancer cases with no prior biopsy were associated with a stage-IV diagnosis.

CONCLUSIONS

In addition to commonly cited risk factors, immunosuppression was associated with malignant transformation, including the use of topical steroids. Analyzing risk factors can help clinicians define risk of progression in patients with OED.

Oral epithelial dysplasia: Recognition, grading and clinical significance

Histopathological grading of epithelial dysplasia remains the principal laboratory method for assessing the risk of malignant transformation in oral potentially malignant disorders (OPMDs). Current views on the molecular pathogenesis and histological interpretation of the features of epithelial dysplasia are described, and the use of grading systems for epithelial dysplasia is discussed. Changes to the current 2017 WHO criteria for diagnosis are proposed with emphasis on the architectural features of epithelial dysplasia. The predictive values of three-grade and binary systems are summarised, and categories of epithelial dysplasia are reviewed, including lichenoid and verrucous lesions, keratosis of unknown significance, HPV-associated dysplasia, differentiated and basaloid epithelial dysplasia. The implications of finding epithelial dysplasia in an oral biopsy for clinical management are discussed from the pathologists' viewpoint.

Recurrence and malignant transformation rates of high grade oral epithelial dysplasia over a 10 year follow up period and the influence of surgical intervention, size of excision biopsy and marginal clearance in a UK regional maxillofacial surgery unit

OBJECTIVES

To determine the overall recurrence rate (RR) and malignant transformation rate (MTR) of high grade oral mucosal epithelial dysplasias (HGOED).

MATERIALS & METHODS

A clinicopathological review of records of patients diagnosed with a unifocal HGOED between 2004 and 2016 on incisional biopsy who then underwent excision. The mean duration of follow-up was 47.7 months (±29.9 SD).

RESULTS

Full demographic, historical and histopathological data were available for 120 patients. Six were lost to follow-up after excisional biopsy. Invasive squamous cell carcinoma (SCC) was present in 19 (18.3%) excisions. HGOED affected the lateral and ventral tongue in 58% of patients. Fourteen (11.7%) were not treated surgically but kept under surveillance. The overall RR was 34.7% (33 patients) and MTR 17.8% (17 patients). Four of the 14 (28.6%) patients who had not had the HGOED excised developed SCC, by contrast to the 13 of the 106 (12.3%) who had been treated. RR was significantly associated with positive excision margins (p = 0.007; OR = 3.6) and a clinical presentation of erythroplakia (p = 0.023; OR = 1.5). MTR was significantly associated with age (p = 0.034), clinical appearance (p = 0.030), site (p = 0.007), treatment received (p = 0.012) and positive excision margins (p = 0.007). The mean time for recurrence to develop was 62 months (±31.5 SD) (range 22-144 months), that for malignant transformation was 50 months (±32.5 SD) (range 8-97 months).

CONCLUSION

Patients with HGOED require follow-up for at least 10 years after treatment. Younger age, homogeneous clinical appearance, complete excision, a larger excision specimen and clear margins all improve prognosis.

Binary and WHO dysplasia grading systems for the prediction of malignant transformation of oral leukoplakia and erythroplakia: a systematic review and meta-analysis

OBJECTIVES

The aim of this systematic review was to examine the evidence of the binary histologic grading system capacity for predicting malignant transformation and to compare it with that of the WHO systems.

MATERIAL AND METHODS

A systematic review was conducted, using PubMed, EMBASE, LILACS, Web of Science, Scopus, and LIVIVO databases without any language or timeframe restrictions. Studies were included if they compared the binary and the WHO histologic grading systems in the prediction of malignant transformation of oral epithelial dysplasia (OED).

RESULTS

The capacity of the WHO and binary grading systems to predict malignant transformation ranged from 16 to 80% and from 5 to 80%, respectively. The pooled malignant transformation rate of lesions classified as severe dysplasia or carcinoma in situ by the WHO grading was 40% (95% confidence interval (CI), 0.02-0.87; I² = 92%; P = 0.00), while the corresponding value for lesions classified as high-risk by the binary grading system was 31% (95% CI, 0.00-0.84; I² = 97%; P = 0.00). Overall, there was no significant difference in prognostication accuracy between the WHO and the binary systems (odds ratio = 2.02; 95% CI, 0.88-4.64).

CONCLUSIONS

Although some studies suggest that the binary system is associated with lower inter-rater variability when grading OED, the evidence remains inconclusive on whether this system is superior to that of the WHO at predicting malignant transformation.

CLINICAL RELEVANCE

The reproducibility of the binary system has the potential to be better for prognostic purposes. However, there is no high-quality evidence to confirm if this advantage may assist clinicians in decision-making.