Deep Learning Predicts the Malignant-Transformation-Free Survival of Oral Potentially Malignant Disorders

Machine-intelligence platforms for the prediction of the probability of malignant transformation of oral potentially malignant disorders are required as adjunctive decision-making platforms in contemporary clinical practice. This study utilized time-to-event learning models to predict malignant transformation in oral leukoplakia and oral lichenoid lesions. A total of 1098 patients with oral white lesions from two institutions were included in this study. In all, 26 features available from electronic health records were used to train four learning algorithms—Cox-Time, DeepHit, DeepSurv, random survival forest (RSF)—and one standard statistical method—Cox proportional hazards model. Discriminatory performance, calibration of survival estimates, and model stability were assessed using a concordance index (c-index), integrated Brier score (IBS), and standard deviation of the averaged c-index and IBS following training cross-validation. This study found that DeepSurv (c-index: 0.95, IBS: 0.04) and RSF (c-index: 0.91, IBS: 0.03) were the two outperforming models based on discrimination and calibration following internal validation. However, DeepSurv was more stable than RSF upon cross-validation. External validation confirmed the utility of DeepSurv for discrimination (c-index—0.82 vs. 0.73) and RSF for individual survival estimates (0.18 vs. 0.03). We deployed the DeepSurv model to encourage incipient application in clinical practice. Overall, time-to-event models are successful in predicting the malignant transformation of oral leukoplakia and oral lichenoid lesions.

Comparative Analyses of Clinical Features, Histopathology, and CD123 Immunohistochemistry of Oral Lupus Erythematosus, Lichen Planus, and Other Lichenoid Lesions

<b><i>Background:</i></b> Oral lupus erythematosus (OLE) and oral lichen planus (OLP) are among the common causes of oral lichenoid lesions (OLLs). The differential diagnosis among causes of OLLs, particularly between OLE and OLP, is challenging as they have significant clinical and histopathological overlap. <b><i>Objectives:</i></b> To compare and summarize the clinical, histopathological, and direct immunofluorescence (DIF) findings between OLE, OLP, and other OLLs and to explore the diagnostic value of CD123 immunohistochemistry. <b><i>Methods:</i></b> A retrospective study on patients with OLE, OLP, and other OLLs was performed between January 2014 and December 2019. The baseline characteristics, the clinical, histopathological, and DIF features, as well as CD123 immunohistochemistry for plasmacytoid dendritic cells (PDCs) were statistically analyzed and compared between groups. <b><i>Results:</i></b> Of 70 patients, 12 had OLE, 39 had OLP, and 19 had other OLLs. Oral erosions/ulcers were the most common findings in all three groups. Red macules, telangiectases, and discoid plaques were more common in OLE patients, while OLP cases were typified by reticulated patches (<i>p</i> &#x3c; 0.05). Additionally, white patches were found more often in other OLLs than in both OLE and OLP (<i>p</i> = 0.002). Histologically, mucosal atrophy, basal vacuolization, and perivascular infiltrate were observed in OLE, whereas OLP specimens possessed mucosal hyperplasia, hypergranulosis, and compact orthokeratosis (<i>p</i> &#x3c; 0.05). Mucosal spongiosis was a histologic feature that favored other OLLs over OLE and OLP (<i>p</i> &#x3c; 0.001). Data on DIF were nonspecific for all three conditions. For immunohistochemical staining, the median number of total CD123+ PDCs was observed to be higher in OLE than OLP in the mucosal-submucosal junction (MSJ) (<i>p</i> = 0.021), the superficial perivascular area (<i>p</i> = 0.026), and the superficial and deep perivascular areas (<i>p</i> = 0.001). Likewise, PDCs in clusters ≥2+ were seen in significantly higher numbers on OLE than OLP along the MSJ (<i>p</i> = 0.002), the superficial perivascular area (<i>p</i> &#x3c; 0.001), as well as the superficial and deep perivascular areas (<i>p</i> = 0.011). CD123+ PDCs were found to be significantly more numerous in both OLE and OLP than other OLLs in all of the abovementioned areas (all <i>p</i> &#x3c; 0.05). <b><i>Conclusion:</i></b> While there are some differences in the clinicopathological features between OLE, OLP, as well as other OLLs, a significant overlap remains. The quantity and distribution pattern of CD123 immunohistochemical staining has a diagnostic implication in differentiating OLE from OLP and other OLLs.

Malignant transformation of oral lichen planus: a retrospective study of 565 Japanese patients

Background Oral lichen planus (OLP) is a chronic inflammatory oral mucosa disease that is recognized as an oral potentially malignant disorder. However, the potentially malignant nature of OLP remains unclear. Methods We designed this study to examine the demographic and clinical characteristics of patients with OLP and evaluate the associated malignant transformation rate. A total of 565 patients with a clinical and histopathological diagnosis of OLP who presented at our department between 2001 and 2017 were retrospectively studied. Patients who had clinical and histopathological features of oral lichenoid lesions (OLLs) classified as oral lichenoid contact lesions, oral lichenoid drug reactions and oral lichenoid lesions of graft-versus-host disease were excluded. Results The study population included 123 men and 442 women aged 21–93 years (mean ± standard deviation, 60.5 ± 11.8). The 565 patients were followed up for a duration of 55.9 ± 45.3 months, during which 4 (0.7%) patients developed squamous cell carcinoma (SCC). In three of these 4 patients who developed SCC, the clinical type of OLP was the red type. Conclusions Our results suggested that OLP was associated with a low risk of malignant transformation. We recommend regular follow-up for OLP patients and clear differentiation of oral epithelial dysplasia and OLLs to enable early detection of malignant transformation. Further investigation of the clinical risk factors associated with malignant transformation is necessary.

Inter-Observer and Intra-Observer Variations in the Assessment of Epithelial Dysplasia in Oral Lichenoid Diseases

Oral lichen planus (OLP) and oral lichenoid lesions (OLL) can both present with histological dysplasia. Despite the presence of WHO-defined criteria for the evaluation of epithelial dysplasia, its assessment is frequently subjective (inter-observer variability). The lack of reproducibility in the evaluation of dysplasia is even more complex in the presence of a lichenoid inflammation. We evaluated dysplasia in 112 oral biopsies with lichenoid inflammation in order to study the inter-observer and the intra-observer variability.