Intelligent deep learning supports biomedical image detection and classification of oral cancer.

BACKGROUND: Oral cancer is a malignant tumor that usually occurs within the tissues of the mouth. This type of cancer mainly includes tumors in the lining of the mouth, tongue, lips, buccal mucosa and gums. Oral cancer is on the rise globally, especially in some specific risk groups. The early stage of oral cancer is usually asymptomatic, while the late stage may present with ulcers, lumps, bleeding, etc. OBJECTIVE: The objective of this paper is to propose an effective and accurate method for the identification and classification of oral cancer. METHODS: We applied two deep learning methods, CNN and Transformers. First, we propose a new CANet classification model for oral cancer, which uses attention mechanisms combined with neglected location information to explore the complex combination of attention mechanisms and deep networks, and fully tap the potential of attention mechanisms. Secondly, we design a classification model based on Swim transform. The image is segmented into a series of two-dimensional image blocks, which are then processed by multiple layers of conversion blocks. RESULTS: The proposed classification model was trained and predicted on Kaggle Oral Cancer Images Dataset, and satisfactory results were obtained. The average accuracy, sensitivity, specificity and F1-Socre of Swin transformer architecture are 94.95%, 95.37%, 95.52% and 94.66%, respectively. The average accuracy, sensitivity, specificity and F1-Score of CANet model were 97.00%, 97.82%, 97.82% and 96.61%, respectively. CONCLUSIONS: We studied different deep learning algorithms for oral cancer classification, including convolutional neural networks, converters, etc. Our Attention module in CANet leverages the benefits of channel attention to model the relationships between channels while encoding precise location information that captures the long-term dependencies of the network. The model achieves a high classification effect with an accuracy of 97.00%, which can be used in the automatic recognition and classification of oral cancer.

Comparison of the 7th and revised 8th UICC editions (2020) for oral squamous cell carcinoma: How does the reclassification impact staging and survival?

Since its introduction in 1968, the TNM (tumor, node, metastasis) classification established by the International Union Against Cancer has provided a consistent framework for staging of oral squamous cell carcinoma (OSCC). The introduction of the 8th edition in 2017 brought about significant modifications, encompassing the integration of depth of invasion (DOI) and extranodal extension (ENE) into the T and N classifications. Further, the UICC the criteria for the T3 and T4a categories were amended in 2020. This study aimed to evaluate the impact of reclassification on staging and, subsequently, the survival of patients with OSCC. Primary OSCCs from 391 patients were classified according to the 7th and revised 8th UICC editions (2020). Stage migration was assessed, and stage-specific progression-free survival (PFS) and overall survival (OS) were evaluated using the Kaplan-Meier method. The log-rank test was used to compare the different stages. Cox-proportional hazard modeling was used to compare the two editions. Incorporating the DOI into the T classification resulted in an upstaging of 77 patients, constituting 19.69% of the cohort. In addition, 49 (12.53%) patients experienced an upstaging when considering ENE in the N classification. Consequently, 103 patients underwent upstaging in UICC staging, accounting for 21.74% of cases. Upstaging mainly occurred from stage III to IVA (26.92%) and from stage IVA to IVB (31.78%). Upon comparing the categories in survival analysis, significant differences in OS and PFS were especially observed between stage IVB and lower stages. When examining the hazard ratios, it became evident that UICC 8 stage IVB is burdened by a 5.59-fold greater risk of disease progression than stage I. Furthermore, UICC 8 stage IVB exhibits a 3.83 times higher likelihood of death than stage I disease. We demonstrated significant stage migration from the 7th to the revised 8th UICC edition. Overall, incorporating DOI and ENE into the T and N classifications represents a substantial clinical advancement, leading to a more accurate staging of OSCC patients. Both staging systems exhibited statistically significant discrimination between stages; however, the 8th UICC edition allowed for a more precise categorization of patients based on their prognosis and led to enhanced hazard discrimination, particularly within higher stages.

Update from the 5th Edition of the World Health Organization Classification of Head and Neck Tumors: Tumours of the Oral Cavity and Mobile Tongue.

The fifth chapter of the upcoming fifth edition of the 2022 World Health Organization Classification of Tumours of the Head and Neck titled Tumours of the oral cavity and mobile tongue, has had some modifications from the 2017 fourth edition. A new section "Non-neoplastic Lesions", introduces two new entries: necrotizing sialometaplasia and melanoacanthoma. The combined Oral potentially malignant disorders and Oral epithelial dysplasia section in the 2015 WHO has now been separated and submucous fibrosis and HPV-associated dysplasia are also discussed in separate sections. Carcinoma cuniculatum and verrucous carcinoma are described in dedicated sections, reflecting that the oral cavity is the most common location in the head and neck for both these entities which have distinct clinical and histologic features from conventional squamous cell carcinoma. This review summarizes the changes in Chapter 5 with special reference to new additions, deletions, and sections that reflect current clinical, histological, and molecular advances.

Study of morphological and textural features for classification of oral squamous cell carcinoma by traditional machine learning techniques.

BACKGROUND: Oral squamous cell carcinoma (OSCC) is the most prevalent form of oral cancer. Very few researches have been carried out for the automatic diagnosis of OSCC using artificial intelligence techniques. Though biopsy is the ultimate test for cancer diagnosis, analyzing a biopsy report is a very much challenging task. To develop computer-assisted software that will diagnose cancerous cells automatically is very important and also a major need of the hour. AIM: To identify OSCC based on morphological and textural features of hand-cropped cell nuclei by traditional machine learning methods. METHODS: In this study, a structure for semi-automated detection and classification of oral cancer from microscopic biopsy images of OSCC, using clinically significant and biologically interpretable morphological and textural features, are examined and proposed. Forty biopsy slides were used for the study from which a total of 452 hand-cropped cell nuclei has been considered for morphological and textural feature extraction and further analysis. After making a comparative analysis of commonly used methods in the segmentation technique, a combined technique is proposed. Our proposed methodology achieves the best segmentation of the nuclei. Henceforth the features extracted were fed into five classifiers, support vector machine, logistic regression, linear discriminant, k-nearest neighbors and decision tree classifier. Classifiers were also analyzed by training time. Another contribution of the study is a large indigenous cell level dataset of OSCC biopsy images. RESULTS: We achieved 99.78% accuracy applying decision tree classifier in classifying OSCC using morphological and textural features. CONCLUSION: It is found that both morphological and textural features play a very important role in OSCC diagnosis. It is hoped that this type of framework will help the clinicians/pathologists in OSCC diagnosis.

The 8th TNM classification for oral squamous cell carcinoma: What is gained, what is lost, and what is missing.

OBJECTIVES: The 8th TNM edition remarkably changed the classification of T and N categories for oral squamous cell carcinoma (OSCC). The present study aims at evaluating the improvement in prognostic power compared to the 7th edition, pros and cons of the modifications, and parameters deserving consideration for further implementations. MATERIALS AND METHODS: All OSCCs treated with upfront surgery at our institution between 2002 and 2017 were included. Demographics, clinical-pathological and treatment variables were retrieved. All tumors were classified according to both the 7th and 8th TNM edition, and patients were grouped according to the shift in T category and stage. Survivals were calculated with the Kaplan-Meier method. Univariate and multivariate analysis were carried out. Receiver Operating Characteristics (ROC) curve analyses were performed to find the best cut-off of DOI (in patients with DOI > 10 mm) and number of involved nodes (in positive neck patients). RESULTS: 244 patients were included. T, N categories, and stage changed in 59.2%, 20.5%, and 49.1% patients, respectively; 41.5% of patients were upstaged. The new T classification well depicted prognosis according to OS. Five-year overall (OS), disease-specific, recurrence-free (RFS) survivals were 60.5%, 70.9%, 59.8%, respectively. According to ROC curves, DOI > 20 mm and 4 positive nodes were the best cutoffs for OS and RFS. CONCLUSION: The novelties introduced in 8th TNM edition were positive. DOI > 20 mm for T4 definition and number of positive nodes (0, <4, 4 or more) for N classification emerged as the most urgent factors to be implemented.

Prognosis of oral squamous cell carcinoma with perineural invasion: A comparative study of classification types.

OBJECTIVE: To investigate the histological location and extent of perineural invasion (PNI) as prognostic factors. DESIGN: Retrospective review of medical records and histological analysis of 116 patients with oral squamous cell carcinoma (OSCC). SETTING: Two major public tertiary hospitals treating head and neck cancer, Royal Adelaide Hospital and Flinders Medical Centre, in South Australia. PARTICIPANTS: Patients diagnosed with OSCC who underwent primary surgical treatment with curative intent at these two centres from January 1, 2005 through December 31, 2015. MAIN OUTCOME MEASURES: The primary end points were disease-free survival (DFS) and disease-specific survival (DSS). RESULTS: The presence of PNI as a binary factor alone did not significantly influence the clinical outcomes. Extratumoural (ET) PNI as measured from the tumour edge was associated with worse DFS on multivariate analyses. Multifocal PNI was associated with worse DFS and DSS. DFS in multifocal PNI was worse irrespective of whether adjuvant therapy was administered. CONCLUSIONS: The presence of multifocal and ET PNI in OSCC is associated with poor clinical outcomes. Patients with multifocal PNI were associated with worse DFS even with adjuvant therapy.

Fog Computing Employed Computer Aided Cancer Classification System Using Deep Neural Network in Internet of Things Based Healthcare System.

Computer assisted automatic smart pattern analysis of cancer affected pixel structure takes critical role in pre-interventional decision making for oral cancer treatment. Internet of Things (IoT) in healthcare systems is now emerging solution for modern e-healthcare system to provide high quality medical care. In this research work, we proposed a novel method which utilizes a modified vesselness measurement and a Deep Convolutional Neural Network (DCNN) to identify the oral cancer region structure in IoT based smart healthcare system. The robust vesselness filtering scheme handles noise while reserving small structures, while the CNN framework considerably improves classification accuracy by deblurring focused region of interest (ROI) through integrating with multi-dimensional information from feature vector selection step. The marked feature vector points are extracted from each connected component in the region and used as input for training the CNN. During classification, each connected part is individually analysed using the trained DCNN by considering the feature vector values that belong to its region. For a training of 1500 image dataset, an accuracy of 96.8% and sensitivity of 92% is obtained. Hence, the results of this work validate that the proposed algorithm is effective and accurate in terms of classification of oral cancer region in accurate decision making. The developed system can be used in IoT based diagnosis in health care systems, where accuracy and real time diagnosis are essential.

Adenocarcinoma de células basales / Basal cell adenocarcinoma

El adenocarcinoma de células basales, también conocido como carcinoma salival basaloide, adenoma maligno de células basales, es una neoplasia epitelial de bajo grado, infiltrante, localmente destructivo y con tendencia a ser recidivante. Su aparición es entre la 5ª y 6ª década de vida, sin predilección por sexo. Clínicamente se manifiesta con un edema o un aumento repentino de tamaño en la zona, de consistencia firme, crecimiento lento e indoloro. El diagnóstico de certeza es a través de la histopatología; su tratamiento quirúrgico, y tiene buen pronóstico en sus estadios iniciales (AU)

Basal cells adenocarcinoma also known as salivary basaloide carcinoma basal cells malignant adenoma is a low degree, infiltrating, locally destructive and prone to be relapsing, epithelial neoplasia. It occurs between the 5th and 6th decade of life, with no predilection for sex. Clinically it manifests with an edema or sudden increased size in the area, of firm consistency, slow growth and pain-less. Its treatment is surgical and the diagnosis of certainty is histopathological with a good prognosis. The purpose of this presentation is to show the case of a 57- years-old male patient with clinical and anatomopathological diagnosis of adenocarcinoma of basal cells located in the yugal mucosa (AU)

A Novel Surface Enhanced Raman Catheter for Rapid Detection, Classification, and Grading of Oral Cancer.

Fabrication and testing of a novel nanostructured surface-enhanced Raman catheter device is reported for rapid detection, classification, and grading of normal, premalignant, and malignant tissues with high sensitivity and accuracy. The sensor part of catheter is formed by a surface-enhanced Raman scattering (SERS) substrate made up of leaf-like TiO2 nanostructures decorated with 30 nm sized Ag nanoparticles. The device is tested using a total of 37 patient samples wherein SERS signatures of oral tissues consisting of malignant oral squamous cell carcinoma (OSCC), verrucous carcinoma, premalignant leukoplakia, and disease-free conditions are detected and classified with an accuracy of 97.24% within a short detection-cum-processing time of nearly 25-30 min per patient. Neoplastic grade changes detected using this device correlate strongly with conventional pathological data, enabling correct classification of tumors into three grades with an accuracy of 97.84% in OSCC. Thus, the potential of a SERS catheter device as a point-of-care pathological tool is shown for the rapid and accurate detection, classification, and grading of solid tumors.

Revisão dos aspectos morfológicos e imunoistoquímicos do linfoma difuso de grandes células B (LDGCB) com ênfase em boca / Review of the morphological and immunohistochemical aspects of diffuse large B cell lymphoma (DLBCL) with emphasis on the mouth

Os linfomas compreendem um grupo diverso de neoplasias malignas, provenientes de células do sistema imunológico em diferentes estágios de diferenciação. Objetivo: o propósito deste artigo é facilitar o diagnóstico do linfoma difuso de grandes células B (LDGCB) por meio de seus aspectos clínicos, morfológicos e imunoistoquímicos, além de suas peculiaridades como manifestação primária em boca. Revisão de literatura: foi realizada uma revisão narrativa da literatura por intermédio de artigos selecionados nas bases de dados PubMed, Medline, SciELO e Lilacs, pela busca por palavras-chave. Aspectos relacionados a classificação e manifestações clínicas também foram considerados, a fim de facilitar o entendimento da lesão e de suas particularidades em boca. Verificou-se que o LDGCB representa a variante mais comum em boca. Os sinais e sintomas clínicos relacionados a essa condição podem ser: aumento de volume, dor, ulceração, alteração de cor da mucosa ou até mesmo parestesia. Morfologicamente, os LDGCBs apresentam células grandes, com padrão de crescimento difuso, citoplasma escasso, nucléolos evidentes e mitoses. Na imunoistoquímica, os LDGCBs são geralmente positivos para CD20 e outros marcadores da linhagem B (CD19, CD79a, PAX5 e CD138), dependendo do estágio de maturação em que se encontram as células B. Considerações finais: o diagnóstico do LDGCB em boca representa um desafio contínuo para os patologistas, em função da heterogeneidade de suas características morfológicas e imunofenotípicas.(AU)

Lymphomas comprise a diverse group of malignant neoplasias from cells of the immune system at different stages of differentiation. Objective: this article aimed to facilitate the diagnosis of diffuse large B-cell lymphoma (DLBCL) through its clinical, morphological, and immunohistochemical aspects, as well as its particularities as a primary manifestation in the mouth. Literature Review: hence, a narrative review of the literature was performed using articles selected in the PubMed, Medline, SciELO, and Lilacs databases through keyword search. Aspects related to classification and clinical manifestations were also considered to facilitate the understanding of the lesion and its particularities in the mouth. It was verified that the diffuse large B-cell lymphoma (DLBCL) represents the most common variant in the mouth. The clinical signs and symptoms related to this condition may be increased volume, pain, ulceration, changed mucosal color, or even paresthesia. Morphologically, DLBCL presents large cells with diffuse growth pattern, scarce cytoplasm, evident nucleoli, and mitoses. In immunohistochemistry, DLBCL is usually positive for CD20 and other markers of lineage B (CD19, CD79a, PAX5, and CD138) depending on the maturation stage in which B cells are found. Final considerations: the diagnosis of oral DLBCL represents a continuous challenge for pathologists due to the heterogeneity of its morphological and immunophenotypic characteristics.(AU)

Addition of tumour infiltration depth and extranodal extension improves the prognostic value of the pathological TNM classification for early-stage oral squamous cell carcinoma.

AIMS: In the 8th edition of the American Joint Committee on Cancer TNM staging manual, tumour infiltration depth and extranodal extension are added to the pathological classification for oral squamous cell carcinoma. The currently available 8th TNM validation studies lack patients with conservative neck treatment, and changes in the classification especially affect patients with small tumours. The aim of this study was to determine the potential impact of the changes in the 8th edition pTNM classification on the prognosis and treatment strategy for oral squamous cell carcinoma in a well-defined series of pT1-T2 patients with long-term follow-up. METHODS AND RESULTS: Two hundred and eleven first primary pT1-T2 oral squamous cell carcinoma patients, with surgical resection as primary treatment, were analysed retrospectively. One hundred and seventy-three patients underwent a neck dissection, and 38 patients had frequent clinical neck assessments. Long-term follow-up (median 64 months) and reassessed tumour infiltration depth were available. Classification according to the 8th edition criteria resulted in 36% total upstaging with the T classification and 16% total upstaging with the N classification. T3-restaged patients (n = 30, 14%) had lower 5-year disease-specific survival rates than T2-staged patients (81% versus 67%, P = 0.042). Postoperative (chemo)radiotherapy could have been considered in another seven (3%) patients on the basis of the 8th edition criteria. CONCLUSIONS: Addition of tumour infiltration depth and extranodal extension in the 8th TNM classification leads to the identification of oral squamous cell carcinoma patients with a worse prognosis who might benefit from an improved postoperative treatment strategy.

Increased Presence of Perineural Invasion in the Tongue and Floor of the Mouth: Could It Represent a More Aggressive Oral Squamous Cell Carcinoma, or Do Larger Aggressive Tumors Cause Perineural Invasion?

PURPOSE: Despite data showing worse outcomes and aggressive disease behavior, perineural invasion (PNI) has not been well characterized in terms of tumor location, histopathologic features, or cervical lymph node status. The specific aims of this study were to measure correlations between PNI, tumor location, and other known histopathologic characteristics used to define aggressive disease. MATERIALS AND METHODS: This was a retrospective cohort study of adult patients with primary squamous cell carcinoma of the oral cavity who underwent neck dissection. We excluded patients whose neck was previously treated with surgery or radiation therapy. Demographic and histopathologic variables of interest were obtained from patient charts. The primary outcome of interest was PNI, and the predictors of interest included tumor location, histopathologic tumor characteristics, and cervical lymph node status. For continuous variables, mean differences were compared by t tests. For categorical variables, the differences in the distribution of the proportions were analyzed with the χ2 test. All variables were entered simultaneously into a multivariate logistic regression model to control for possible confounding. Statistical significance for the study was set at P < .05. RESULTS: Three hundred sixty-eight patients met the study criteria. PNI showed statistically significant correlations with lymph node status, tumor depth, and specific primary tumor location. PNI was more likely to be seen in tumors located in the tongue or floor of the mouth. Tumors with PNI had a deeper depth of invasion: 15.9 ± 10.9 mm versus 10.2 ± 10.0 mm (P < .001). PNI tumors had a higher mean total number of positive nodes: 2.85 ± 5.23 versus 0.83 ± 1.80 (P < .001). CONCLUSIONS: PNI is statistically correlated with tongue and floor-of-the-mouth subsites within the oral cavity, as well as larger tumors, deeper tumors, and disease that has progressed to the lymph nodes. Whether this correlation represents causation in either direction remains unknown.

Historical Evolution of the Polymorphous Adenocarcinoma.

The 2017 World Health Organization Classification of Head and Neck Tumors introduced for the first time the diagnostic terminology "cribriform variant of polymorphous adenocarcinoma". This nomenclature attempts to reconciliate the ongoing taxonomical controversy related to cribriform adenocarcinoma of tongue. In order to better understand this classification conundrum, it is imperative for pathologist to comprehend the historical evolution of polymorphous adenocarcinoma formerly known as polymorphous "low grade" adenocarcinoma. This review highlights our understanding of these tumors since their origins.

Comparison of the Seventh and Eighth Editions of the American Joint Committee on Cancer pT and pN Classifications as Predictors of Survival in Patients With Oral Squamous Cell Carcinoma.

Objectives: We retrospectively compared the seventh and eighth editions of American Joint Committee on Cancer (AJCC) TNM frameworks as disease-free survival (DFS) and overall survival (OS) predictors in oral squamous carcinomas (OSCCs). Methods: We restaged the 342 patients with the revised pT and pN criteria and performed survival analyses. Results: The 3-year DFS (mean follow-up, 364 days; recurrences, 99) was 50%, and the 5-year OS (mean follow-up, 615 days; deaths, 69) was 42%. The eight edition pN classification was an independent multivariate survival predictor. The revised TNM criteria upstaged pT, pN, and stage groupings in 99 (38.8%), 58 (37.3%), and 101 (29.5%) patients. The latter two groups revealed significantly worse DFS and OS compared with those whose categorizations had remained unaltered. In addition, their classification/staging criteria demonstrated superior discrimination, monotonicity, and accuracy for survival estimations. Conclusions: Of the competing AJCC staging criteria, the revised pN classification was the most powerful system to predict OSCC survivals.

What does the 4th edition of the World Health Organization Classification of Head and Neck Tumors (2017) bring new about mucosal melanomas?

The recently published 4th Edition of the World Health Organization Classification of Head and Neck Tumors addresses the most relevant and updated aspects of tumor biology, including clinical presentation, histopathology, immunohistochemistry, and prognosis of head and neck tumors. The objective of the present study is to compare these updates to the 3rd edition of that book with regard to mucosal melanomas and to highlight the potential factors that differ those tumors from cutaneous melanomas. We observed progress in the understanding of oral and sinonasal mucosal melanomas, which also present themselves, in the molecular scope, differently form cutaneous melanomas.

Tumor classification with MALDI-MSI data of tissue microarrays: A case study.

With mass spectrometry imaging (MSI) on tissue microarrays (TMAs) a large number of biomolecules can be studied for many patients at the same time, making it an attractive tool for biomarker discovery. Here we investigate whether lymph node metastasis can be predicted from MALDI-MSI data. Measurements are performed on TMAs and then filtered based on spectral intensity and the percentage of tumor cells, after which the resulting data for 122 patients is further preprocessed. We assume differences between patients with and without metastasis are expressed in a limited number of features. Two univariate feature selection methods are applied to reduce the dimensionality of the MALDI-MSI data. The selected features are then used in combination with three classifiers. The best classification scores are obtained with a decision tree classifier, which classifies about 72% of patients correctly. Almost all the predictive power comes from a single peak (m/z 718.4). The sensitivity of our classification approach, which can be generically used to search for biomarkers, is investigated using artificially modified data.

Is the 8th edition of the Union for International Cancer Control staging of oral cancer good enough?

Depth of invasion is an important predictor of survival. A study by the International Consortium (ICOR) for Outcome Research proposed incorporation of it (together with the greatest surface dimension, or the anatomical criteria, or both) into the T stage. This has been adopted in part by the 8th edition of the Union for International Cancer Control (UICC) TNM 8 classification of malignant tumours for oral squamous cell carcinoma (SCC). Our aim was to verify depth of invasion as an independent prognostic factor, and to validate the staging by comparing it with that specified in the 7th edition (TNM 7) and the T-staging model proposed by the International Consortium. We retrospectively studied 449 patients who had had operations for a previously untreated primary oral cancer between 2006 and 2014 at a single centre, and analysed the independent predictive value of depth of invasion for both disease-specific and overall survival. It was an independent predictor of disease-specific survival as were sex, perineural invasion, and N stage. It was also an independent predictor of overall survival together with sex and N status. Staging in TNM 8 gave a better balance of distribution than that in TNM 7, but did not discriminate between prognosis in patients with T3 and T4 disease. The proposed International Consortium rules for T-staging gave an improved balance in distribution and hazard discrimination. The incorporation of depth of invasion into the T-staging rules for oral SCC improved prognostic accuracy and is likely to influence the selection of patients for adjuvant treatment. Our findings suggest that the TNM 8 staging lacks hazard discrimination in patients with locally-advanced disease because its T4 staging is restricted to anatomical criteria.

What does the 4th edition of the World Health Organization Classification of Head and Neck Tumors (2017) bring new about mucosal melanomas?

Abstract: The recently published 4th Edition of the World Health Organization Classification of Head and Neck Tumors addresses the most relevant and updated aspects of tumor biology, including clinical presentation, histopathology, immunohistochemistry, and prognosis of head and neck tumors. The objective of the present study is to compare these updates to the 3rd edition of that book with regard to mucosal melanomas and to highlight the potential factors that differ those tumors from cutaneous melanomas. We observed progress in the understanding of oral and sinonasal mucosal melanomas, which also present themselves, in the molecular scope, differently form cutaneous melanomas.

Trend of oral and pharyngeal cancer mortality in Brazil in the period of 2002 to 2013.

OBJECTIVE: To analyze the trend of oral and pharyngeal cancer mortality rates in the period of 2002 to 2013 in Brazil according to sex, anatomical site, and macroregion of the country. METHODS: The mortality data were obtained from the Mortality Information System and the population data were obtained from the Brazilian Institute of Geography and Statistics. The trend of the rates standardized by sex and age was calculated using the Prais-Winsten estimation, and we obtained the annual percentage change and the respective 95% confidence intervals, analyzed according to sex, macroregion, and anatomical site. RESULTS: The average coefficient of oral cancer mortality was 1.87 per 100,000 inhabitants and it remained stable during the study period. The coefficient of pharyngeal cancer mortality was 2.04 per 100,000 inhabitants and it presented an annual percentage change of -2.6%. Approximately eight in every 10 deaths occurred among men. There was an increase in the rates of oral cancer in the Northeast region (annual percentage change of 6.9%) and a decrease in the Southeast region (annual percentage change of -2.9%). Pharyngeal cancer mortality decreased in the Southeast and South regions with annual percentage change of -4.8% and -5.1% respectively. Cancer mortality for tonsil, other major salivary glands, hypopharynx, and other and unspecified parts of mouth and pharynx showed a decreasing trend while the other sites presented stability. CONCLUSIONS: Pharyngeal cancer mortality decreased in the period of 2002 to 2013. Oral cancer increased only in the Northeast region. Mortality for tonsil cancer, other major salivary glands, hypopharynx, and other and ill-defined sites in the lip, oral cavity, and pharynx decreased.