A digital score of peri-epithelial lymphocytic activity predicts malignant transformation in oral epithelial dysplasia.

Oral squamous cell carcinoma (OSCC) is amongst the most common cancers, with more than 377,000 new cases worldwide each year. OSCC prognosis remains poor, related to cancer presentation at a late stage, indicating the need for early detection to improve patient prognosis. OSCC is often preceded by a premalignant state known as oral epithelial dysplasia (OED), which is diagnosed and graded using subjective histological criteria leading to variability and prognostic unreliability. In this work, we propose a deep learning approach for the development of prognostic models for malignant transformation and their association with clinical outcomes in histology whole slide images (WSIs) of OED tissue sections. We train a weakly supervised method on OED cases (n = 137) with malignant transformation (n = 50) and mean malignant transformation time of 6.51 years (±5.35 SD). Stratified five-fold cross-validation achieved an average area under the receiver-operator characteristic curve (AUROC) of 0.78 for predicting malignant transformation in OED. Hotspot analysis revealed various features of nuclei in the epithelium and peri-epithelial tissue to be significant prognostic factors for malignant transformation, including the count of peri-epithelial lymphocytes (PELs) (p < 0.05), epithelial layer nuclei count (NC) (p < 0.05), and basal layer NC (p < 0.05). Progression-free survival (PFS) using the epithelial layer NC (p < 0.05, C-index = 0.73), basal layer NC (p < 0.05, C-index = 0.70), and PELs count (p < 0.05, C-index = 0.73) all showed association of these features with a high risk of malignant transformation in our univariate analysis. Our work shows the application of deep learning for the prognostication and prediction of PFS of OED for the first time and offers potential to aid patient management. Further evaluation and testing on multi-centre data is required for validation and translation to clinical practice. © 2023 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.

ChatGPT and dentistry: a step toward the future.

This article aims to explore the integration of ChatGPT, an advanced conversational artificial intelligence model, in the field of dentistry. The review primarily consists of information related to the capabilities and functionalities of ChatGPT and how these abilities can aid dental professionals. This study includes data from research papers, case studies, and relevant literature on language models, as well as papers on dentistry, patient communication, dental education, and clinical decision-making. A systematic approach was used to select relevant studies and literature. The selection criteria focused on papers that specifically discussed the integration of language models, ChatGPT in particular, in dentistry and their applications. The study findings revealed that ChatGPT has significant potential to revolutionize dentistry by offering various applications and benefits. It can enhance patient engagement and understanding through personalized oral health information and guidance. In dental education, ChatGPT can provide interactive learning, case studies, and virtual patient simulations. ChatGPT can also assist researchers in analyzing dental literature, identifying patterns, and generating insights. Moreover, it supports dentists with evidence-based recommendations, treatment options, and diagnostic support. Integrating ChatGPT in dentistry can be highly beneficial, but it is crucial to address ethical considerations, accuracy, and privacy concerns. Responsible implementation and continuous improvement of its functionalities are necessary to ensure that patient care and outcomes are improved.

Development and validation of a multivariable model for prediction of malignant transformation and recurrence of oral epithelial dysplasia.

BACKGROUND: Oral epithelial dysplasia (OED) is the precursor to oral squamous cell carcinoma which is amongst the top ten cancers worldwide. Prognostic significance of conventional histological features in OED is not well established. Many additional histological abnormalities are seen in OED, but are insufficiently investigated, and have not been correlated to clinical outcomes. METHODS: A digital quantitative analysis of epithelial cellularity, nuclear geometry, cytoplasm staining intensity and epithelial architecture/thickness is conducted on 75 OED whole-slide images (252 regions of interest) with feature-specific comparisons between grades and against non-dysplastic/control cases. Multivariable models were developed to evaluate prediction of OED recurrence and malignant transformation. The best performing models were externally validated on unseen cases pooled from four different centres (n = 121), of which 32% progressed to cancer, with an average transformation time of 45 months. RESULTS: Grade-based differences were seen for cytoplasmic eosin, nuclear eccentricity, and circularity in basal epithelial cells of OED (p < 0.05). Nucleus circularity was associated with OED recurrence (p = 0.018) and epithelial perimeter associated with malignant transformation (p = 0.03). The developed model demonstrated superior predictive potential for malignant transformation (AUROC 0.77) and OED recurrence (AUROC 0.74) as compared with conventional WHO grading (AUROC 0.68 and 0.71, respectively). External validation supported the prognostic strength of this model. CONCLUSIONS: This study supports a novel prognostic model which outperforms existing grading systems. Further studies are warranted to evaluate its significance for OED prognostication.

A Novel Deep Learning Algorithm for Human Papillomavirus Infection Prediction in Head and Neck Cancers Using Routine Histology Images.

The etiology of head and neck squamous cell carcinoma (HNSCC) involves multiple carcinogens, such as alcohol, tobacco, and infection with human papillomavirus (HPV). Because HPV infection influences the prognosis, treatment, and survival of patients with HNSCC, it is important to determine the HPV status of these tumors. In this article, we propose a novel deep learning pipeline for HPV infection status prediction with state-of-the-art performance in HPV detection using only whole-slide images of routine hematoxylin and eosin-stained HNSCC sections. We show that our Digital-HPV score generated from hematoxylin and eosin slides produces statistically significant patient stratifications in terms of overall and disease-specific survival. In addition, quantitative profiling of the spatial tumor microenvironment and analysis of the immune profiles show relatively high levels of lymphocytic infiltration in tumor and tumor-associated stroma. High levels of B cells and T cells and low macrophage levels were also identified in HPV-positive patients compared to HPV-negative patients, confirming different immune response patterns elicited by HPV infection in patients with HNSCC.

A digital score of tumour-associated stroma infiltrating lymphocytes predicts survival in head and neck squamous cell carcinoma.

The infiltration of T-lymphocytes in the stroma and tumour is an indication of an effective immune response against the tumour, resulting in better survival. In this study, our aim was to explore the prognostic significance of tumour-associated stroma infiltrating lymphocytes (TASILs) in head and neck squamous cell carcinoma (HNSCC) through an AI-based automated method. A deep learning-based automated method was employed to segment tumour, tumour-associated stroma, and lymphocytes in digitally scanned whole slide images of HNSCC tissue slides. The spatial patterns of lymphocytes and tumour-associated stroma were digitally quantified to compute the tumour-associated stroma infiltrating lymphocytes score (TASIL-score). Finally, the prognostic significance of the TASIL-score for disease-specific and disease-free survival was investigated using the Cox proportional hazard analysis. Three different cohorts of haematoxylin and eosin (H&E)-stained tissue slides of HNSCC cases (n = 537 in total) were studied, including publicly available TCGA head and neck cancer cases. The TASIL-score carries prognostic significance (p = 0.002) for disease-specific survival of HNSCC patients. The TASIL-score also shows a better separation between low- and high-risk patients compared with the manual tumour-infiltrating lymphocytes (TILs) scoring by pathologists for both disease-specific and disease-free survival. A positive correlation of TASIL-score with molecular estimates of CD8+ T cells was also found, which is in line with existing findings. To the best of our knowledge, this is the first study to automate the quantification of TASILs from routine H&E slides of head and neck cancer. Our TASIL-score-based findings are aligned with the clinical knowledge, with the added advantages of objectivity, reproducibility, and strong prognostic value. Although we validated our method on three different cohorts (n = 537 cases in total), a comprehensive evaluation on large multicentric cohorts is required before the proposed digital score can be adopted in clinical practice. © 2021 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.

Machine learning for detection and classification of oral potentially malignant disorders: A conceptual review.

Oral potentially malignant disorders represent precursor lesions that may undergo malignant transformation to oral cancer. There are many known risk factors associated with the development of oral potentially malignant disorders, and contribute to the risk of malignant transformation. Although many advances have been reported to understand the biological behavior of oral potentially malignant disorders, their clinical features that indicate the characteristics of malignant transformation are not well established. Early diagnosis of malignancy is the most important factor to improve patients' prognosis. The integration of machine learning into routine diagnosis has recently emerged as an adjunct to aid clinical examination. Increased performances of artificial intelligence AI-assisted medical devices are claimed to exceed the human capability in the clinical detection of early cancer. Therefore, the aim of this narrative review is to introduce artificial intelligence terminology, concepts, and models currently used in oncology to familiarize oral medicine scientists with the language skills, best research practices, and knowledge for developing machine learning models applied to the clinical detection of oral potentially malignant disorders.

Machine learning concepts applied to oral pathology and oral medicine: A convolutional neural networks' approach.

INTRODUCTION: Artificial intelligence models and networks can learn and process dense information in a short time, leading to an efficient, objective, and accurate clinical and histopathological analysis, which can be useful to improve treatment modalities and prognostic outcomes. This paper targets oral pathologists, oral medicinists, and head and neck surgeons to provide them with a theoretical and conceptual foundation of artificial intelligence-based diagnostic approaches, with a special focus on convolutional neural networks, the state-of-the-art in artificial intelligence and deep learning. METHODS: The authors conducted a literature review, and the convolutional neural network's conceptual foundations and functionality were illustrated based on a unique interdisciplinary point of view. CONCLUSION: The development of artificial intelligence-based models and computer vision methods for pattern recognition in clinical and histopathological image analysis of head and neck cancer has the potential to aid diagnosis and prognostic prediction.

The use of deep learning state-of-the-art architectures for oral epithelial dysplasia grading: A comparative appraisal.

BACKGROUND: Dysplasia grading systems for oral epithelial dysplasia are a source of disagreement among pathologists. Therefore, machine learning approaches are being developed to mitigate this issue. METHODS: This cross-sectional study included a cohort of 82 patients with oral potentially malignant disorders and correspondent 98 hematoxylin and eosin-stained whole slide images with biopsied-proven dysplasia. All whole-slide images were manually annotated based on the binary system for oral epithelial dysplasia. The annotated regions of interest were segmented and fragmented into small patches and non-randomly sampled into training/validation and test subsets. The training/validation data were color augmented, resulting in a total of 81,786 patches for training. The held-out independent test set enrolled a total of 4,486 patches. Seven state-of-the-art convolutional neural networks were trained, validated, and tested with the same dataset. RESULTS: The models presented a high learning rate, yet very low generalization potential. At the model development, VGG16 performed the best, but with massive overfitting. In the test set, VGG16 presented the best accuracy, sensitivity, specificity, and area under the curve (62%, 62%, 66%, and 65%, respectively), associated with the higher loss among all Convolutional Neural Networks (CNNs) tested. EfficientB0 has comparable metrics and the lowest loss among all convolutional neural networks, being a great candidate for further studies. CONCLUSION: The models were not able to generalize enough to be applied in real-life datasets due to an overlapping of features between the two classes (i.e., high risk and low risk of malignization).

Prediction of malignant transformation and recurrence of oral epithelial dysplasia using architectural and cytological feature specific prognostic models.

Oral epithelial dysplasia (OED) is a precursor state usually preceding oral squamous cell carcinoma (OSCC). Histological grading is the current gold standard for OED prognostication but is subjective and variable with unreliable outcome prediction. We explore if individual OED histological features can be used to develop and evaluate prognostic models for malignant transformation and recurrence prediction. Digitised tissue slides for a cohort of 109 OED cases were reviewed by three expert pathologists, where the prevalence and agreement of architectural and cytological histological features was assessed and association with clinical outcomes analysed using Cox proportional hazards regression and Kaplan-Meier curves. Within the cohort, the most prevalent features were basal cell hyperplasia (72%) and irregular surface keratin (60%), and least common were verrucous surface (26%), loss of epithelial cohesion (30%), lymphocytic band and dyskeratosis (34%). Several features were significant for transformation (p < 0.036) and recurrence (p < 0.015) including bulbous rete pegs, hyperchromatism, loss of epithelial cohesion, loss of stratification, suprabasal mitoses and nuclear pleomorphism. This led us to propose two prognostic scoring systems including a '6-point model' using the six features showing a greater statistical association with transformation and recurrence (bulbous rete pegs, hyperchromatism, loss of epithelial cohesion, loss of stratification, suprabasal mitoses, nuclear pleomorphism) and a 'two-point model' using the two features with highest inter-pathologist agreement (loss of epithelial cohesion and bulbous rete pegs). Both the 'six point' and 'two point' models showed good predictive ability (AUROC ≥ 0.774 for transformation and 0.726 for recurrence) with further improvement when age, gender and histological grade were added. These results demonstrate a correlation between individual OED histological features and prognosis for the first time. The proposed models have the potential to simplify OED grading and aid patient management. Validation on larger multicentre cohorts with prospective analysis is needed to establish their usefulness in clinical practice.

Observer agreement in the diagnosis of oral lichen planus using the proposed criteria of the American Academy of Oral and Maxillofacial Pathology.

BACKGROUND: Oral lichen planus (OLP) is a common chronic inflammatory condition with an undefined malignant transformation potential. There have been many attempts at providing a specific definition of OLP without conclusive outcomes. A new set of diagnostic criteria was proposed in 2016 by the American Academy of Oral and Maxillofacial Pathology (AAOMP) in an endeavour to resolve this issue, and this has not yet been evaluated. This study aimed to assess the utility of AAOMP proposed criteria for the diagnosis of OLP. METHODS: Five pathologists blindly assessed a cohort of 215 digital whole slide images (WSI) obtained from haematoxylin and eosin-stained microscopic slides. Forty-six WSI were included twice to assess the intra-observer agreement. Included cases were diagnosed clinically as either OLP or oral lichenoid reaction. Each pathologist was asked to utilize the AAOMP histopathological criteria while assessing slides. The variations in diagnoses were assessed by unweighted kappa statistics. RESULTS: The level of intra-observer agreement was very good (0.801 to 0.899). The level of inter-observer agreement among the observers varied from good (0.658) to very good (0.842) when the responses were categorized as evident/compatible OLP versus no OLP and was good (0.62 to 0.725) when the responses were categorized as evident OLP, versus compatible OLP, versus no OLP. The clinico-pathological correlation was 87.6%. CONCLUSION: A reliable level of agreement can be achieved by pathologists for the diagnosis of OLP using the AAOMP criteria for differentiation between lichenoid and other conditions. There are still limitations in discriminating OLP from oral lichenoid lesions microscopically.

Clinicopathological analysis and survival outcomes of primary salivary gland tumors in pediatric patients: A systematic review.

BACKGROUND: Salivary gland tumors are a diverse group of uncommon neoplasms that are rare in pediatric patients. The aim of this study was to evaluate the clinicopathological profile and survival outcomes of pediatric patients affected by salivary gland tumors. MATERIALS AND METHODS: An extensive search was carried out using the MEDLINE/PubMed, EMBASE, Scopus databases, and grey literature. The risk of bias was available in all papers included. RESULTS: A total of 2,830 articles were initially retrieved with 54 remaining for data extraction, resulting in 2,937 cases. This comprised forty-five case series' and nine cohort studies. These tumors were slightly more prevalent in females (57.4%). The patients' age ranged from 0.3 to 19 years old, with a mean age of 13.3 years. Parotid was the most affected site (81.9%), and 99.2% of cases clinically exhibited a swelling. Presence of pain/tenderness was reported in 13.5% of the cases, with an average duration of 12.6 months for the appearance of symptoms. Most of the reported cases were malignant tumors (75.4%), with mucoepidermoid carcinoma the most common tumor of all tumors (44.8%), followed by pleomorphic adenoma (24.1%). Surgery alone was the leading treatment choice in 74.9% cases, and the 5-year overall survival rate of patients was 93.1%. Patients with symptoms (P = .001), local recurrence (P < .001), metastasis (P < .001), and those not undergoing surgery or surgery combined with radiotherapy (P < .001) showed lower survival rates. CONCLUSION: The pediatric patients present a high frequency of malignant salivary neoplasms and a high overall survival rate.

Burkitt lymphoma of the head and neck: An international collaborative study.

BACKGROUND: Burkitt lymphoma (BL) is an aggressive B-cell lymphoma with three variants (endemic, sporadic, and immunodeficiency-associated), presenting with specific epidemiological and clinical features. Burkitt lymphoma affects the head and neck region (BLHN) in approximately 10% of cases. The aim of this study was to undertake a comparative analysis of the clinicopathologic and immunohistochemical (IHC) features of BLHN diagnosed in patients from Africa, Guatemala, and Brazil. METHODS: Cases diagnosed as BLHN were collected from the files of six oral pathology laboratory services (Brazil, South Africa, and Guatemala) and one Brazilian pediatric oncology hospital from 1986 to 2020. Clinicopathological and IHC data, and Epstein-Barr virus (EBV) status by in situ hybridization data for each case were reviewed and described. RESULTS: Of the 52 cases, BLHN was predominant in pediatric patients [43 (82.69%)] and males [43 (82.69%)], with a mean age of 11.26 ± 9.68 years (range, 1-39 years). Neck and cervical lymph nodes [14 (26.92%)], and involvement of both maxilla and mandible [8 (15.38%)], were the most common anatomical sites. Clinically, tumor/swelling [40 (31.25%)], cervical lymphadenopathy [14 (10.94%)], pain [12 (9.38%)], and bone destruction [12 (9.38%)] were frequent findings. All cases showed typical morphological characteristics of BL. IHC profiles included positivity for CD20 [52 (100%)], CD10 [38 (79.17%)], Bcl6 [29 (87.88%)], and c-Myc protein [18 (81.82%)]. EBV was positive in 18 cases (62.07%). The Ki-67 index ranged from 90 to 100%. CONCLUSION: The clinicopathological and EBV profile of BLHN in South African, Guatemalan, and Brazilian patients is similar.

Human papillomavirus (HPV) can establish productive infection in dysplastic oral mucosa, but HPV status is poorly predicted by histological features and p16 expression.

AIMS: Previous studies have reported the presence of high-risk human papillomavirus (HR-HPV) in a subset of dysplastic oral epithelial lesions. Many cases show a histological spectrum of atypia similar to that seen in non-human papillomavirus (HPV) severe epithelial dysplasia, but some studies have suggested that HPV status can be inferred on the basis of histological features. We aimed to assess the utility of such histological features and p16 as surrogate markers of HPV infection in a retrospective cohort of 33 cases of severe epithelial dysplasia, with matched clinicopathological data and histological features. METHODS AND RESULTS: Tissue sections were assessed for the expression of p16, minichromosome maintenance 2, HPV E4 and HPV L1 by the use of immunohistochemistry. HPV16/18 E6 and E7 expression was assessed by the use of RNA in-situ hybridisation (RNAScope). In the cohort, 18.2% of cases (6/33) were HR-HPV-positive, with no age/gender differences between the HPV-positive and HPV-negative groups. HPV E4 and HPV L1 were expressed in surface keratinocytes in four of six (66%) HPV-positive cases, indicative of productive HPV infection. Lack of p16 expression was predictive of HPV-negative status, but sensitivity and specificity varied according to the cut-off. Histologically, the presence of karyorrhectic nuclei and abnormal mitotic figures was higher in HPV-positive lesions (P < 0.05), but the predictive specificity and sensitivity were suboptimal (sensitivity, 0.75; specificity, 0.52). CONCLUSIONS: This study demonstrates, for the first time, that a minority of severely dysplastic oral lesions harbour productive, biologically relevant HPV infection. Consideration should be given to the specific assessment of HPV status in severe epithelial dysplasia cases, as both p16 status and the presence of karyorrhectic cells are poor predictive markers of HPV status.

The chemokine lymphotactin and its recombinant variants in oral cancer cell regulation.

BACKGROUND: The expression of XCR1 receptor and its metamorphic ligand lymphotactin (hLtn) has been shown in cancers but their precise role in tumorigenesis is poorly understood including the significance of the physiologically existing hLtn monomeric (CC3) and dimeric (W55D) confirmations where the latter thought to function as the receptor antagonist. The aim of this study was to explore the functional role of bioengineered hLtn variants and the role of fibroblasts in XCR1/hLtn expression regulation in oral cancer cells (OCCL). MATERIAL AND METHODS: qRT-PCR and flow cytometry were performed to evaluate mRNA and protein expression of XCR1 and hLtn. Recombinant hLtn variants (wild-type, CC3 and W55D mutant) were designed, expressed, purified and evaluated using proliferation, adhesion and chemotaxis assays. XCR1 and hLtn expression regulation by fibroblasts was determined using indirect co-culture. XCR1 and hLtn expression in primary and metastatic OSCC tissue was assessed using immunohistochemistry. RESULTS: hLtn caused a significant decrease in OCCL XCR1 surface protein expression. hLtn CC3 mutant was highly functional facilitating proliferation and migration. Conditioned media from primary cancer-associated and senescent fibroblasts significantly upregulated XCR1 and hLtn mRNA expression in OCCL. Immunohistochemistry revealed higher XCR1 and hLtn expression in metastatic tumour deposits and surrounding stroma compared to primary OSCC tissue. CONCLUSIONS: The development of hLtn biological mutants, regulation of XCR1 expression by its ligand hLtn and crosstalk with fibroblasts are novel findings suggesting an important role for the XCR1/hLtn axis within the OSCC tumour microenvironment. These discoveries build upon previous studies and suggest that the hLtn/XCR1 axis has a significant role in stromal crosstalk and OSCC progression.

Demystifying oral epithelial dysplasia: a histological guide.

Oral epithelial dysplasia is a histologically diagnosed potentially premalignant disorder of the oral mucosa, which carries a risk of malignant transformation to squamous cell carcinoma. The diagnosis and grading of oral epithelial dysplasia is challenging, with cases often referred to specialist oral and maxillofacial pathology centres for second opinion. Even still there is poor inter-examiner and intra-examiner agreement in a diagnosis. There are a total of 28 features of oral epithelial dysplasia listed in the 5th edition of World Health Organization classification of tumours of the head and neck. Each of these features is poorly defined and subjective in its interpretation. Moreover, how these features contribute to dysplasia grading and risk stratification is even less well defined. This article discusses each of the features of oral epithelial dysplasia with examples and provides an overview of the common mimics, including the normal histological features of the oral mucosa which may mimic atypia. This article also highlights the paucity of evidence defining these features while offering suggested definitions. Ideally, these definitions will be refined, and the most important features identified to simplify the diagnosis of oral epithelial dysplasia. Digital whole slide images of the figures in this paper can be found at: https://www.pathogenesis.co.uk/r/demystifying-dysplasia-histology-dataset.

The Value of Labial Gland Biopsies as a Diagnostic Test for Sjögren's Syndrome.

PURPOSE: There are a number of diagnostic criteria that can be used to support a diagnosis of Sjögren's syndrome (SS), a chronic autoimmune condition often characterised by xerostomia and xerophthalmia. Of the available investigations, the most invasive is the labial gland biopsy (LGB) for histopathology, which is associated with a risk of long-term altered sensation to the lip. A positive histological diagnosis is currently considered to be one of the most objective criteria, however there is debate about the interobserver agreement between pathologists, as well as the sensitivity and specificity of this test. We aim to determine if the diagnostic value of the LGB is significant enough to warrant the surgical procedure and its associated risks. METHODS: This study involved assessing the degree of agreement between members of a pathology team for a cohort of 50 LGBs taken for the purpose of confirming or excluding SS. The Tarpley system was used, which involves the allocation of a 'focus score'. Additionally, the histological diagnoses were compared to the relevant serological findings where available. RESULTS: All cases within the cohort had adequate tissue for assessment. 84% agreement (Cohen's Kappa = 0.585) was seen between the current team's consensus and the original reporting pathologist on whether the appearance was supportive of SS. However, only 58% agreement was seen for focus scores (Weighted Kappa = 0.496). The agreement between the serology result and whether the histology was supportive of SS was 79% (Cohen's Kappa = 0.493). CONCLUSION: The findings raise the possibility that undue emphasis is placed on the value of a histological SS diagnosis. The current system for assessing and grading these biopsies is ambiguous in nature, with a low threshold considered indicative of SS. Due to the risk of complications associated with a LGB, alternative minimally invasive investigations should always be considered. The histological findings in isolation, particularly when a low focus score is seen, may not be predictive of a diagnosis of SS.

Prognostic importance of mitosis quantification and PHH3 expression in oral epithelial dysplasia.

Oral epithelial dysplasia (OED) is diagnosed and graded using a range of histological features, making grading subjective and challenging. Mitotic counting and phosphohistone-H3 (PHH3) staining have been used for the prognostication of various malignancies; however, their importance in OED remains unexplored. This study conducts a quantitative analysis of mitotic activity in OED using both haematoxylin and eosin (H&E)-stained slides and immunohistochemical (IHC) staining for PHH3. Specifically, the diagnostic and prognostic importance of mitotic number, mitotic type and intra-epithelial location is evaluated. Whole slide images (WSI) of OED (n = 60) and non-dysplastic tissue (n = 8) were prepared for analysis. Five-year follow-up data was collected. The total number of mitosis (TNOM), mitosis type and intra-epithelial location was manually evaluated on H&E images and a digital mitotic count performed on PHH3-stained WSI. Statistical associations between these features and OED grade, malignant transformation and OED recurrence were determined. Mitosis count increased with grade severity (H&E: p < 0.005; IHC: p < 0.05), and grade-based differences were seen for mitosis type and location (p < 0.05). The ratio of normal-to-abnormal mitoses was higher in OED (1.61) than control (1.25) and reduced with grade severity. TNOM, type and location were better predictors when combined with histological grading, with the most prognostic models demonstrating an AUROC of 0.81 for transformation and 0.78 for recurrence, exceeding conventional grading. Mitosis quantification and PHH3 staining can be an adjunct to conventional H&E assessment and grading for the prediction of OED prognosis. Validation on larger multicentre cohorts is needed to establish these findings.

Spatial transcriptomic interrogation of the tumour-stroma boundary in a 3D engineered model of ameloblastoma.

Stromal cells are key components of the tumour microenvironment (TME) and their incorporation into 3D engineered tumour-stroma models is essential for tumour mimicry. By engineering tumouroids with distinct tumour and stromal compartments, it has been possible to identify how gene expression of tumour cells is altered and influenced by the presence of different stromal cells. Ameloblastoma is a benign epithelial tumour of the jawbone. In engineered, multi-compartment tumouroids spatial transcriptomics revealed an upregulation of oncogenes in the ameloblastoma transcriptome where osteoblasts were present in the stromal compartment (bone stroma). Where a gingival fibroblast stroma was engineered, the ameloblastoma tumour transcriptome revealed increased matrix remodelling genes. This study provides evidence to show the stromal-specific effect on tumour behaviour and illustrates the importance of engineering biologically relevant stroma for engineered tumour models. Our novel results show that an engineered fibroblast stroma causes the upregulation of matrix remodelling genes in ameloblastoma which directly correlates to measured invasion in the model. In contrast the presence of a bone stroma increases the expression of oncogenes by ameloblastoma cells.

A fully automated and explainable algorithm for predicting malignant transformation in oral epithelial dysplasia.

Oral epithelial dysplasia (OED) is a premalignant histopathological diagnosis given to lesions of the oral cavity. Its grading suffers from significant inter-/intra-observer variability, and does not reliably predict malignancy progression, potentially leading to suboptimal treatment decisions. To address this, we developed an artificial intelligence (AI) algorithm, that assigns an Oral Malignant Transformation (OMT) risk score based on the Haematoxylin and Eosin (H&E) stained whole slide images (WSIs). Our AI pipeline leverages an in-house segmentation model to detect and segment both nuclei and epithelium. Subsequently, a shallow neural network utilises interpretable morphological and spatial features, emulating histological markers, to predict progression. We conducted internal cross-validation on our development cohort (Sheffield; n = 193 cases) and independent validation on two external cohorts (Birmingham and Belfast; n = 89 cases). On external validation, the proposed OMTscore achieved an AUROC = 0.75 (Recall = 0.92) in predicting OED progression, outperforming other grading systems (Binary: AUROC = 0.72, Recall = 0.85). Survival analyses showed the prognostic value of our OMTscore (C-index = 0.60, p = 0.02), compared to WHO (C-index = 0.64, p = 0.003) and binary grades (C-index = 0.65, p < 0.001). Nuclear analyses elucidated the presence of peri-epithelial and intra-epithelial lymphocytes in highly predictive patches of transforming cases (p < 0.001). This is the first study to propose a completely automated, explainable, and externally validated algorithm for predicting OED transformation. Our algorithm shows comparable-to-human-level performance, offering a promising solution to the challenges of grading OED in routine clinical practice.

LYSTO: The Lymphocyte Assessment Hackathon and Benchmark Dataset.

We introduce LYSTO, the Lymphocyte Assessment Hackathon, which was held in conjunction with the MICCAI 2019 Conference in Shenzhen (China). The competition required participants to automatically assess the number of lymphocytes, in particular T-cells, in images of colon, breast, and prostate cancer stained with CD3 and CD8 immunohistochemistry. Differently from other challenges setup in medical image analysis, LYSTO participants were solely given a few hours to address this problem. In this paper, we describe the goal and the multi-phase organization of the hackathon; we describe the proposed methods and the on-site results. Additionally, we present post-competition results where we show how the presented methods perform on an independent set of lung cancer slides, which was not part of the initial competition, as well as a comparison on lymphocyte assessment between presented methods and a panel of pathologists. We show that some of the participants were capable to achieve pathologist-level performance at lymphocyte assessment. After the hackathon, LYSTO was left as a lightweight plug-and-play benchmark dataset on grand-challenge website, together with an automatic evaluation platform.