Contexte : Peu d’études ont recherché le rôle des facteurs de risque professionnels dans la survenue des cancers des voies aérodigestives supérieures (VADS). Ces études ont été conduites principalement chez des hommes.Objectif : L’objectif de cette thèse était d’étudier les associations entre les cancers des VADS et les expositions professionnelles chez les femmes.Méthodes : Icare est une étude cas-témoins en population générale incluant 296 cas féminins de cancers épidermoïdes des VADS et 775 femmes témoins. Les historiques de carrières recueillis ont été codés et croisés avec les matrices emplois-expositions du programme Matgéné. Outre les intitulés d’emplois, les expositions à 5 solvants chlorés (chloroforme, chlorure de méthylène, perchloréthylène, trichloréthylène, tétrachlorure de carbone), 5 solvants oxygénés (éthylène glycol, tétrahydrofurane, éther éthylique, cétones, alcools), 5 solvants pétroliers (essences carburants, essences spéciales, gazole, benzène, white-spirit) et à 7 poussières et fibres (amiante, farine, cuir, fibres céramiques réfractaires, ciment, laines minérales, silice) ont été étudiées. Les odds-ratios et leurs intervalles de confiance à 95% ajustés sur l’âge, le département, les consommations de tabac et d’alcool ont été estimés par régressions logistiques non conditionnelles.Résultats :Plusieurs professions et secteurs d’activité associés à un risque élevé de cancer des VADS ont été identifiés. Certaines professions (ouvrières de l’alimentation et des boissons, monteuses en appareillage électrique ou électronique, soudeuses) peuvent être à l’origine d’expositions professionnelles aux solvants, aux métaux, aux fumées de soudage et à diverses poussières. Les analyses par nuisance ont mis en évidence des associations significatives entre le risque de cancer des VADS et l’exposition au perchloréthylène et au au trichloréthylène. Aucune association claire n’est observée avec les solvants pétroliers et oxygénés, certains largement utilisés par les femmes L’exposition aux poussières de farine augmente significativement le risque de cancer des VADS. Une exposition probable à l’amiante est associée à une augmentation modérée et non significative du risque. Les analyses par localisation de cancer (cavité orale, pharynx, larynx), limitées par des effectifs faibles, ne mettent pas en évidence d’association spécifique.Conclusion : Nos résultats suggèrent un rôle des expositions professionnelles au trichloréthylène, au perchloréthylène et aux poussières de farine dans la survenue des cancers des VADS chez les femmes. / Background : Few occupational studies have addressed head and neck cancer, and these studies have been predominantly conducted in men. Objective : Our objective was to investigate the associations between head and neck cancer and occupational exposures in women Population and methods : ICARE, a French population-based case–control study, included 296 squamous cell carcinomas of the head and neck (HNSCC) in women and 775 female controls. Lifelong occupational history was collected. Job-exposure matrices were used to assess exposure to five chlorinated solvents (carbon tetrachloride; chloroform; methylene chloride; perchloroethylene; trichloroethylene), 5 petroleum solvents (benzene; special petroleum product; gasoline; white-spirits and other light aromatic mixtures; diesel, fuels and kerosene), 5 oxygenated solvents (alcohols; ketones and esters; ethylene glycol; diethyl ether; tetrahydrofuran) and 7 fibers and dusts (asbestos, flour dust, leather dust, refractory ceramic fibers, cement dust, mineral wools and silica) . An analysis by job title was conducted, and then associations with specific occupational exposures were investigated.Odds ratio (ORs) and 95% confidence intervals (CI), adjusted for smoking, alcohol drinking, age and residence area, were estimated with logistic models. Results : Significantly increased HNSCC risks were found for several jobs and industries. Some of these occupations (food and beverage processors, electrical and electronic equipment assemblers, welders and flame cutters) may entail exposure to agents such as solvents, metals, welding fumes and various dusts. Analyses for specific occupational exposures showed a significantly elevated risk of HNSCC associated with exposure to trichloroethylene and perchloroethylene. There is no clear evidence that petroleum or oxygenated solvents, some of them commonly used by women, are risk factors for HNSCC. Exposure to flour dust increased significantly HNSCC risk. Probable exposure to asbestos was associated with a moderate, non-significant elevation in risk. Analyses by cancer site (oral cavity, pharynx, larynx) were hampered by small numbers and did to reveal any specific association.Conclusion : These findings suggest that occupational exposure to perchloroethylene, trichloroethylene and flour dust may increase the risk of HNSCC in women.

Os objetivos deste estudo foram identificar os fatores de risco para complicações pulmonares pós-operatórias em pacientes laringectomizados por carcinoma espinocelular de laringe, assim como os fatores de risco para apresentação de metástases pulmonares e os fatores de risco para apresentação de segundo tumores primários em pulmão destes pacientes no período de 1985 a 1996. O estudo consistiu em uma análise retrospectiva de 291 pacientes admitidos no Hospital do Câncer A. C. Camargo, no período de 1985 a 1996, portadores de carcinoma espinocelular de laringe, submetidos a tratamento cirúrgico com intenção curativa seguido ou não de radioterapia. Foram analisados as variáveis demográficas, as comorbidades, a localização do sítio primário, o estadiamento clínico, o tratamento do tumor primário e do pescoço, o tratamento radioterápico, a diferenciação celular, as margens cirúrgicas, as recidivas locais, as recidivas regionais, a presença de complicações maiores, de complicações menores, as metástases pulmonares e a presença de múltiplos tumores primários pulmonares. O teste de associação do qui-quadrado foi utilizado para análise univariada descritiva das diversas variáveis comparando-se os grupos com complicação pulmonar e sem complicação pulmonar. A análise multivariada através da regressão logística foi utilizada na determinação dos fatores de risco para apresentação de metástases pulmonares e múltiplos tumores primário pulmonares. A incidência de complicações pulmonares foi de 31,3%, foram identificados como fatores de risco para complicações pulmonares somente a epiglote (p=0,004; RR 2,1), tendo a variável gênero associação marginal (p=0,081; RR 2,8). As metástases pulmonares tiveram incidência de 7,2%, na análise univariada foram identificados como fatores de risco o estadiamento N (p=0,032), diferenciação histológica (p=0,004), margens cirúrgicas (p=0,017) e recidivas locoregionais (0,002). Os múltiplos tumores primários pulmonares apresentaram incidência de 3,1% e na análise univariada foram identificados como fatores de risco o estadiamento N (p=0,048) e sítio aritenóide (p=0,001). Na análise multivariada foram significativos somente a diferenciação histológica: moderamente diferenciado (p=0,007; RR 2,9) e pouco diferenciado (p=0,032; RR 4,0); e as margens cirúrgicas: exíguas (p=0,003; RR 6,4) para apresentação de metástases pulmonares e múltiplos tumores pulmonares. Este estudo demonstra a importância do estadiamento clínico como fator de risco para complicações pulmonares, metástases à distância e múltiplos tumores primário em pulmão. Os fatores de risco determinantes para aparecimento de metástases pulmonares e múltiplos tumores primários pulmonares foram a diferenciação histológica e as margens cirúrgicas / Objective: To identify the risk factors to postoperative pulmonary complications in laryngeal cancer patients submitted to surgical treatment, the risk factors to development of lung metastasis and second lung primary tumor. Patients and Methods: Retrospective study of a cohort of 291 patients admitted at Hospital do Câncer A.C.Camargo from January, 1985 to December 1996. All patients were submitted to some kind of laryngectomy with curative intent as part of treatment of a proven laryngeal cancer, followed or not by radiotherapy. The following variables were analized: demographic, comorbidities, primary site, clinical stage, primary and neck surgical treatment, histopathologic differentiation grade, surgical margins, recurrences, postoperative pulmonary complications, lung metastasis and second lung primary tumor. The univariate and multivariate analysis were utilized to built the model to predict the risks factors and the factors of prognostic significance. Results: The overall pulmonary complications incidence were 31,3%, epiglottis were identified as significant single risk factor to pulmonary complications (p=0.004; RR 2,1). Lung metastasis had 7,2% incidence to this, the risk factors were N stage (p=0.032 ), histopathologic differentiation grade (p=0.004), surgical margins (p=0.017) and locoregional recurrence (p=0.002). The second lung primary tumor incidence were 3,1%, univariate analysis showed N stage (p=0,048) and arithenoid site (p=0,001) as significant risk factors. The multivariate analysis showed the histopathologic differentiation: moderate grade (p=0.007 RR 2,9) and poor grade (p=0.032 RR 4,0) and surgical margins: close (p=0.003 RR 6,4) as prognostic factors to deveopment of lung metastasis and second lung primary tumor. Conclusions: This study showed the clinical stage importance as risk factor to development of postoperative pulmonary complication, lung metastasis and second lung primary tumor in laryngectomy cancer patients. The prognostic factors associated with lung metastasis and second lung primary tumor were the histopathologic differentiation and the surgical margins

Os objetivos deste estudo foram identificar os fatores de risco para complicações pulmonares pós-operatórias em pacientes laringectomizados por carcinoma espinocelular de laringe, assim como os fatores de risco para apresentação de metástases pulmonares e os fatores de risco para apresentação de segundo tumores primários em pulmão destes pacientes no período de 1985 a 1996. O estudo consistiu em uma análise retrospectiva de 291 pacientes admitidos no Hospital do Câncer A. C. Camargo, no período de 1985 a 1996, portadores de carcinoma espinocelular de laringe, submetidos a tratamento cirúrgico com intenção curativa seguido ou não de radioterapia. Foram analisados as variáveis demográficas, as comorbidades, a localização do sítio primário, o estadiamento clínico, o tratamento do tumor primário e do pescoço, o tratamento radioterápico, a diferenciação celular, as margens cirúrgicas, as recidivas locais, as recidivas regionais, a presença de complicações maiores, de complicações menores, as metástases pulmonares e a presença de múltiplos tumores primários pulmonares. O teste de associação do qui-quadrado foi utilizado para análise univariada descritiva das diversas variáveis comparando-se os grupos com complicação pulmonar e sem complicação pulmonar. A análise multivariada através da regressão logística foi utilizada na determinação dos fatores de risco para apresentação de metástases pulmonares e múltiplos tumores primário pulmonares. A incidência de complicações pulmonares foi de 31,3%, foram identificados como fatores de risco para complicações pulmonares somente a epiglote (p=0,004; RR 2,1), tendo a variável gênero associação marginal (p=0,081; RR 2,8). As metástases pulmonares tiveram incidência de 7,2%, na análise univariada foram identificados como fatores de risco o estadiamento N (p=0,032), diferenciação histológica (p=0,004), margens cirúrgicas (p=0,017) e recidivas locoregionais (0,002). Os múltiplos tumores primários pulmonares apresentaram incidência de 3,1% e na análise univariada foram identificados como fatores de risco o estadiamento N (p=0,048) e sítio aritenóide (p=0,001). Na análise multivariada foram significativos somente a diferenciação histológica: moderamente diferenciado (p=0,007; RR 2,9) e pouco diferenciado (p=0,032; RR 4,0); e as margens cirúrgicas: exíguas (p=0,003; RR 6,4) para apresentação de metástases pulmonares e múltiplos tumores pulmonares. Este estudo demonstra a importância do estadiamento clínico como fator de risco para complicações pulmonares, metástases à distância e múltiplos tumores primário em pulmão. Os fatores de risco determinantes para aparecimento de metástases pulmonares e múltiplos tumores primários pulmonares foram a diferenciação histológica e as margens cirúrgicas / Objective: To identify the risk factors to postoperative pulmonary complications in laryngeal cancer patients submitted to surgical treatment, the risk factors to development of lung metastasis and second lung primary tumor. Patients and Methods: Retrospective study of a cohort of 291 patients admitted at Hospital do Câncer A.C.Camargo from January, 1985 to December 1996. All patients were submitted to some kind of laryngectomy with curative intent as part of treatment of a proven laryngeal cancer, followed or not by radiotherapy. The following variables were analized: demographic, comorbidities, primary site, clinical stage, primary and neck surgical treatment, histopathologic differentiation grade, surgical margins, recurrences, postoperative pulmonary complications, lung metastasis and second lung primary tumor. The univariate and multivariate analysis were utilized to built the model to predict the risks factors and the factors of prognostic significance. Results: The overall pulmonary complications incidence were 31,3%, epiglottis were identified as significant single risk factor to pulmonary complications (p=0.004; RR 2,1). Lung metastasis had 7,2% incidence to this, the risk factors were N stage (p=0.032 ), histopathologic differentiation grade (p=0.004), surgical margins (p=0.017) and locoregional recurrence (p=0.002). The second lung primary tumor incidence were 3,1%, univariate analysis showed N stage (p=0,048) and arithenoid site (p=0,001) as significant risk factors. The multivariate analysis showed the histopathologic differentiation: moderate grade (p=0.007 RR 2,9) and poor grade (p=0.032 RR 4,0) and surgical margins: close (p=0.003 RR 6,4) as prognostic factors to deveopment of lung metastasis and second lung primary tumor. Conclusions: This study showed the clinical stage importance as risk factor to development of postoperative pulmonary complication, lung metastasis and second lung primary tumor in laryngectomy cancer patients. The prognostic factors associated with lung metastasis and second lung primary tumor were the histopathologic differentiation and the surgical margins

As patologias que afetam a laringe estão aumentando consideravelmente nos últimos anos devido à condição da sociedade atual onde há hábitos não saudáveis como fumo, álcool e tabaco e um abuso vocal cada vez maior, talvez por conta do aumento da poluição sonora, principalmente nos grandes centros urbanos. Atualmente o exame utilizado pela endoscopia per-oral, direcionado a identificar patologias de laringe, são a videolaringoscopia e videoestroboscopia, ambos invasivos e por muitas vezes desconfortável ao paciente. Buscando melhorar o bem estar e minimizar o desconforto dos pacientes que necessitam submeter-se a estes procedimentos, este estudo tem como objetivo reconhecer padrões que possam ser aplicados à identificação de patologias de laringe de modo a auxiliar na criação de um novo método não invasivo em substituição ao método atual. Este trabalho utilizará várias configurações diferentes de redes neurais. A primeira rede neural foi gerada a partir de 524.287 resultados obtidos através das configurações k-k das 19 medidas acústicas disponíveis neste trabalho. Esta configuração atingiu uma acurácia de 99,5% (média de 96,99±2,08%) ao utilizar uma configuração com 11 e com 12 medidas acústicas dentre as 19 disponíveis. Utilizando-se 3 medidas rotacionadas (obtidas através do método de componentes principais), foi obtido uma acurácia de 93,98±0,24%. Com 6 medidas rotacionadas, o resultado obtido foi de acurácia foi de 94,07±0,29%. Para 6 medidas rotacionadas com entrada normalizada, a acurácia encontrada foi de 97,88±1,53%. A rede neural que fez 23 diferentes classificações, voz normal mais 22 patologias, mostrou que as melhores classificações, de acordo com a acurácia, são a da patologia hiperfunção com 58,23±18,98% e a voz normal com 52,15±18,31%. Já para a pior patologia a ser classificada, encontrou-se a fadiga vocal com 0,57±1,99%. Excluindo-se a voz normal, ou seja, utilizando uma rede neural composta somente por vozes patológicas, a hiperfunção continua sendo a mais facilmente identificável com uma acurácia de 57,3±19,55%, a segunda patologia mais facilmente identificável é a constrição ântero-posterior com 18,14±11,45%. Nesta configuração, a patologia mais difícil de se classificar continua sendo a fadiga vocal com 0,7±2,14%. A rede com re-amostragem obteve uma acurácia de 25,88±10,15% enquanto que a rede com re-amostragem e alteração de neurônios na camada intermediária obteve uma acurácia de 21,47±7,58% para 30 neurônios e uma acurácia de 18,44±6,57% para 40 neurônios. Por fim foi feita uma máquina de vetores suporte que encontrou um resultado de 67±6,2%. Assim, mostrou-se que as medidas acústicas precisam ser aprimoradas para a obtenção de melhores resultados de classificação dentre as patologias de laringe estudadas. Ainda assim, verificou-se que é possível discriminar locutores normais daqueles pacientes disfônicos. / Diseases that affect the larynx have been considerably increased in recent years due to the condition of nowadays society where there have been unhealthy habits like smoking, alcohol and tobacco and an increased vocal abuse, perhaps due to the increase in noise pollution, especially in large urban cities. Currently the exam performed by per-oral endoscopy (aimed to identify laryngeal pathologies) have been videolaryngoscopy and videostroboscopy, both invasive and often uncomfortable to the patient. Seeking to improve the comfort of the patients who need to undergo through these procedures, this study aims to identify acoustic patterns that can be applied to the identification of laryngeal pathologies in order to creating a new non-invasive larynx assessment method. Here two different configurations of neural networks were used. The first one was generated from 524.287 combinations of 19 acoustic measurements to classify voices into normal or from a diseased larynx, and achieved an max accuracy of 99.5% (96.99±2.08%). Using 3 and 6 rotated measurements (obtained from the principal components analysis method), the accuracy was 93.98±0.24% and 94.07±0.29%, respectively. With 6 rotated measurements from a previouly standardization of the 19 acoustic measurements, the accuracy was 97.88±1.53%. The second one, to classify 23 different voice types (including normal voices), showed better accuracy in identifying hiperfunctioned larynxes and normal voices, with 58.23±18.98% and 52.15±18.31%, respectively. The worst accuracy was obtained from vocal fatigues, with 0.57±1.99%. Excluding normal voices of the analysis, hyperfunctioned voices remained the most easily identifiable (with an accuracy of 57.3±19.55%) followed by anterior-posterior constriction (with 18.14±11.45%), and the most difficult condition to be identified remained vocal fatigue (with 0.7±2.14%). Re-sampling the neural networks input vectors, it was obtained accuracies of 25.88±10.15%, 21.47±7.58%, and 18.44±6.57% from such networks with 20, 30, and 40 hidden layer neurons, respectively. For comparison, classification using support vector machine produced an accuracy of 67±6.2%. Thus, it was shown that the acoustic measurements need to be improved to achieve better results of classification among the studied laryngeal pathologies. Even so, it was found that is possible to discriminate normal from dysphonic speakers.

As patologias que afetam a laringe estão aumentando consideravelmente nos últimos anos devido à condição da sociedade atual onde há hábitos não saudáveis como fumo, álcool e tabaco e um abuso vocal cada vez maior, talvez por conta do aumento da poluição sonora, principalmente nos grandes centros urbanos. Atualmente o exame utilizado pela endoscopia per-oral, direcionado a identificar patologias de laringe, são a videolaringoscopia e videoestroboscopia, ambos invasivos e por muitas vezes desconfortável ao paciente. Buscando melhorar o bem estar e minimizar o desconforto dos pacientes que necessitam submeter-se a estes procedimentos, este estudo tem como objetivo reconhecer padrões que possam ser aplicados à identificação de patologias de laringe de modo a auxiliar na criação de um novo método não invasivo em substituição ao método atual. Este trabalho utilizará várias configurações diferentes de redes neurais. A primeira rede neural foi gerada a partir de 524.287 resultados obtidos através das configurações k-k das 19 medidas acústicas disponíveis neste trabalho. Esta configuração atingiu uma acurácia de 99,5% (média de 96,99±2,08%) ao utilizar uma configuração com 11 e com 12 medidas acústicas dentre as 19 disponíveis. Utilizando-se 3 medidas rotacionadas (obtidas através do método de componentes principais), foi obtido uma acurácia de 93,98±0,24%. Com 6 medidas rotacionadas, o resultado obtido foi de acurácia foi de 94,07±0,29%. Para 6 medidas rotacionadas com entrada normalizada, a acurácia encontrada foi de 97,88±1,53%. A rede neural que fez 23 diferentes classificações, voz normal mais 22 patologias, mostrou que as melhores classificações, de acordo com a acurácia, são a da patologia hiperfunção com 58,23±18,98% e a voz normal com 52,15±18,31%. Já para a pior patologia a ser classificada, encontrou-se a fadiga vocal com 0,57±1,99%. Excluindo-se a voz normal, ou seja, utilizando uma rede neural composta somente por vozes patológicas, a hiperfunção continua sendo a mais facilmente identificável com uma acurácia de 57,3±19,55%, a segunda patologia mais facilmente identificável é a constrição ântero-posterior com 18,14±11,45%. Nesta configuração, a patologia mais difícil de se classificar continua sendo a fadiga vocal com 0,7±2,14%. A rede com re-amostragem obteve uma acurácia de 25,88±10,15% enquanto que a rede com re-amostragem e alteração de neurônios na camada intermediária obteve uma acurácia de 21,47±7,58% para 30 neurônios e uma acurácia de 18,44±6,57% para 40 neurônios. Por fim foi feita uma máquina de vetores suporte que encontrou um resultado de 67±6,2%. Assim, mostrou-se que as medidas acústicas precisam ser aprimoradas para a obtenção de melhores resultados de classificação dentre as patologias de laringe estudadas. Ainda assim, verificou-se que é possível discriminar locutores normais daqueles pacientes disfônicos. / Diseases that affect the larynx have been considerably increased in recent years due to the condition of nowadays society where there have been unhealthy habits like smoking, alcohol and tobacco and an increased vocal abuse, perhaps due to the increase in noise pollution, especially in large urban cities. Currently the exam performed by per-oral endoscopy (aimed to identify laryngeal pathologies) have been videolaryngoscopy and videostroboscopy, both invasive and often uncomfortable to the patient. Seeking to improve the comfort of the patients who need to undergo through these procedures, this study aims to identify acoustic patterns that can be applied to the identification of laryngeal pathologies in order to creating a new non-invasive larynx assessment method. Here two different configurations of neural networks were used. The first one was generated from 524.287 combinations of 19 acoustic measurements to classify voices into normal or from a diseased larynx, and achieved an max accuracy of 99.5% (96.99±2.08%). Using 3 and 6 rotated measurements (obtained from the principal components analysis method), the accuracy was 93.98±0.24% and 94.07±0.29%, respectively. With 6 rotated measurements from a previouly standardization of the 19 acoustic measurements, the accuracy was 97.88±1.53%. The second one, to classify 23 different voice types (including normal voices), showed better accuracy in identifying hiperfunctioned larynxes and normal voices, with 58.23±18.98% and 52.15±18.31%, respectively. The worst accuracy was obtained from vocal fatigues, with 0.57±1.99%. Excluding normal voices of the analysis, hyperfunctioned voices remained the most easily identifiable (with an accuracy of 57.3±19.55%) followed by anterior-posterior constriction (with 18.14±11.45%), and the most difficult condition to be identified remained vocal fatigue (with 0.7±2.14%). Re-sampling the neural networks input vectors, it was obtained accuracies of 25.88±10.15%, 21.47±7.58%, and 18.44±6.57% from such networks with 20, 30, and 40 hidden layer neurons, respectively. For comparison, classification using support vector machine produced an accuracy of 67±6.2%. Thus, it was shown that the acoustic measurements need to be improved to achieve better results of classification among the studied laryngeal pathologies. Even so, it was found that is possible to discriminate normal from dysphonic speakers.

Oral squamous cell carcinoma (OSCC) is the most common head and neck cancer, with a worldwide incidence of 275,000 new cases annually (Warnakulasuriya, 2009). Globally, the head and neck carcinoma represents a major cause of morbidity and mortality and is the sixth most commonly occurring cancer (Warnakulasuriya, 2009). A majority (>90%) of the head and neck cancers are squamous in origin and thus are linguistically referred to as head and neck squamous cell carcinoma (HNSCC) (Warnakulasuriya, 2009). HNSCC includes cancers of the oral cavity, larynx and pharynx; oral cancer being the most common (Warnakulasuriya, 2009). Although, HNSCC is the sixth most common cancer globally (Warnakulasuriya, 2009), the Indian scenario is graver. According to GLOBOCAN 2008 (http://globocan.iarc.fr), the worldwide age standardized incidence rate (ASR) for HNSCC (and thus OSCC) is 5.3 and 2.5 per 100,000 males and females respectively (Ferlay et al., 2010). In India, the ASR is 9.8 and 5.2 per 100,000 males and females respectively, clearly demonstrating a remarkably high incidence rate of OSCC (Ferlay et al., 2010; http://globocan.iarc.fr). OSCC is a peculiar cancer which is largely preventable and rarely presents as a familial disorder. The most common etiological factors associated with OSCC include tobacco and alcohol consumption (Johnson, 2001). Additionally, high risk human papillomaviruses (HPV strains 16 and 18) as well as genetic predispositions have been implicated. The treatment of OSCC mainly relies on surgical resection of the tumor. The site, size, depth of infiltration and proximity to the bone of the tumor determine whether a combination of surgery with radiation therapy or chemotherapy would be advised (Scully and Bagan, 2009). The concomitant chemo-radiation therapy is the most commonly used strategy in locally advanced cancer. Taxanes (e.g., paclitaxel and docetaxel) and platinum-based induction chemotherapy (e.g., cisplatin) are the options in the treatment of locally advanced cancer. Epidermal growth factor receptor (EGFR) targeted with cetuximab in combination with radiotherapy has been successfully tested in a large randomized trial and thus is currently a new option (Scully and Bagan, 2009). The success of cetuximab has paved the path for the development and implementation of molecules targeting various signaling pathways. Despite extensive research on oral squamous cell carcinoma (OSCC), the five-year survival rate has not changed in several decades with the exception of the targeted treatment strategies involving cetuximab as discussed above. The current chemotherapeutic approaches lack selectivity and are flagitious. Thus, effective treatment of OSCC requires the identification of molecular targets to design appropriate therapeutic strategies. To this end, the present study took three distinct approaches in order to validate the use of existing targets and to reveal novel prognostic biomarkers and therapeutic targets. 1) Targeting the PI3K-AKT-MTOR pathway in OSCC and identification of determinants of its sensitivity. 2) Gene expression analysis of ectopically overexpressed TSC2 to identify new therapeutic targets and prognostic biomarkers as well as to elucidate the genes regulated by it. 3) Expression profiling of CYP1B1 in order to validate the use of CYP1B1 based prodrug therapy in OSCC. Investigations pertaining to the changes in gene and protein expression profiles in malignant as well as pre-malignant lesions have documented the deregulation of the PI3K-AKT-MTOR (phosphoinositide 3-kinase-AKT-mechanistic target of rapamycin) and EGFR (epidermal growth factor receptor) pathways in OSCC which are being widely targeted in many therapeutic strategies (Molinolo et al., 2007; Chakraborty et al., 2008; Matta and Ralhan, 2009; Molinolo et al., 2009; Stransky et al., 2011). The PI3K-AKT-MTOR pathway is a central hub for controlling cellular proliferation and growth in response to various intracellular as well as extracellular stimuli. Crucial signaling cascades including WNT, RAS, HIF-1α and AMPK cross-talk with the PI3K-AKT-MTOR pathway at a variety of molecular junctions. Thus, making this pathway sensitive to perceiving various growth modulatory conditions, ranging from the presence of growth factors to hypoxia and nutrient deprivation (Sengupta et al., 2010; Yang and Guan, 2007). The aberrant expression of the PI3K-AKT-MTOR pathway in OSCC advocated the targeting of this coveted pathway (Chakraborty et al., 2008). In various cancers, the monotherapeutic treatments with inhibitors like LY294002 (PI3K inhibitor) and rapamycin (MTOR inhibitor) demonstrated reduced efficacies. Such reduced efficacies were attributed to the drug toxicity and non-specific action of LY294002 (Davies et al., 2000; Sun et al., 2005; Ikezoe et al., 2007; Wang et al., 2008; Liu et al., 2009), or the ablation of a feedback inhibition loop leading to the reactivation of the PI3K-AKT-MTOR pathway by rapamycin (O'Reilly et al., 2006; Carracedo et al., 2008). Thus, rapamycin or its analogues demonstrated mediocre efficacy due to cytostatic effects in clinical trials, primarily due to the paradoxical activation of major survival kinases namely MAPK and AKT (O'Reilly et al., 2006; Carracedo et al., 2008). The present study aimed at increasing the efficacy of these drugs by incorporating a combinatorial approach. The MTT assay demonstrated that prolonged monotherapeutic treatments with rapamycin led to a modest growth inhibition in three OSCC (KB, SCC131 and SCC084) and HeLa cell lines. Western blot analysis of the phosphorylation status of AKT and RPS6KB1 revealed that monotherapeutic treatments with rapamycin for 96 hr led to the reactivation of the PI3K-AKT-MTOR pathway. Thus, the modest growth inhibitory effect of rapamycin was attributed to the reactivation of the PI3K-AKT-MTOR pathway. A combinatorial treatment approach was hence believed to circumvent this problem in order to increase the efficacy of targeting the PI3K-AKT-MTOR pathway. The PI3K inhibitor LY294002 was used combinatorially with rapamycin. This prolonged dual combinatorial treatment regime was distinctly more efficacious than either of the drugs alone and led to a reduction in cellular viability accompanied by increased sub-G1 population, indicating marked cell death that was characterized as caspase-3 dependent apoptosis. The differential sensitivity of the cell lines towards this combinatorial treatment revealed a novel determinant of the sensitivity, the transactivation of EGFR. The cell lines (SCC131 and SCC084) that were capable of transactivating EGFR were relatively resistant to the dual targeting of PI3K and MTOR in comparison to cell lines that did not transactivate EGFR (HeLa and KB). Further, targeting PI3K, MTOR and EGFR simultaneously was more efficacious in the presence of EGFR transactivation than dually targeting PI3K and MTOR. The results conclusively proved that the combinatorial therapeutic approach dually targeting PI3K and MTOR is a promising treatment strategy as compared to a monotherapeutic treatment and a major factor determining the sensitivity towards this treatment is the status of autophosphorylation of EGFR (Tyr1173) which governs the potential for EGFR transactivation by the combinatorial treatment. Thus, this study demonstrated that the status of EGFR autophosphorylation (Tyr1173) can be used as a biomarker to predict the sensitivity towards the combinatorial targeting of PI3K and MTOR in OSCC. The PI3K-AKT-MTOR pathway is negatively regulated by TSC2 (tuberous sclerosis complex 2; tuberin) (Tee et al., 2002). The importance of the TSC2 gene in the regulation of cell growth and proliferation is irrefutable. TSC2 facilitates the crosstalk between a variety of cellular signals, making it a crucial hub where many cellular networks integrate like AKT, MAPK and AMPK (Clements et al., 2007; Rosner et al., 2007; Rosner et al., 2008). It is a tumor suppressor gene and is downregulated in many cancers including OSCC (Chakraborty et al., 2008). In order to identify the genes regulated by TSC2 in OSCC, we stably overexpressed TSC2 in KB cells and the changes in the gene expression profiles caused by this ectopic overexpression were observed using a whole genome expression microarray. The results showed differential regulation of 268 genes (107 genes were upregulated and 161 genes were downregulated, p<0.05, fold change ≥ 1.5). A majority of these genes were functionally associated with transcription, cell growth and proliferation, apoptosis, cell cycle and neurogenesis. Functional annotation and network analysis was performed by using the DAVID v6.7 and IPA version 8.7 softwares. The microarray data revealed a novel aspect in the crosstalk between WNT signaling and TSC2, namely the transcriptional regulation of WNT signaling by TSC2. Further, in the context of therapeutic applications, the microarray analysis revealed multiple genes that were functionally categorized to be involved in response to radiation, UV and drugs (e.g., SERPINB13 and IL1B). Future studies on the regulation of such genes that are involved in responses to drugs and radiation may give insights into the role of TSC2 in resistance or sensitivity towards chemotherapy and radiation therapy. Moreover, EREG, a member of the epidermal growth factor family, was found to be the most downregulated gene in the microarray analysis. Previous reports have documented elevated levels of EREG in tuberous sclerosis lesions and its association with poor clinical prognosis in OSCC patients (Li et al., 2008; Shigeishi et al., 2008), making its regulatory aspects intriguing. Additionally, published data on the transcriptional functions of TSC2 instigated us to analyze the role of TSC2 in the regulation of EREG. TSC2 has been shown to modulate the transcription mediated by members of the steroid receptor superfamily of genes (Henry et al., 1998) and was shown to bind specifically to ERα and inhibit estrogen induced proliferation (Finlay et al., 2004). Also, TSC2 has been shown to possess C-terminal transcriptional activation domains (Tsuchiya et al., 1996). We have therefore attempted to investigate the transcription related functional aspects of TSC2 by exploiting the observed transcriptional repression of EREG. The physiological roles of TSC1 and TSC2 that are independent of the PI3K-AKT-MTOR pathway have been termed as ‘non-canonical’ (Neuman and Henske, 2011). The repression of EREG by TSC2 was observed to be insensitive to rapamycin, suggesting that it was independent of MTORC1 and thus a non-canonical function of TSC2. To determine whether the repression in EREG was at the level of the promoter, we performed a dual luciferase reporter assay. The results showed that the EREG promoter was repressed by stable as well as transient overexpression of TSC2. In order to elucidate the mechanism of transcriptional regulation by TSC2, we performed the ChIP analysis to observe the in vivo binding of TSC2 to the EREG promoter. In the ChIP analysis with the anti-TSC2 antibody, we observed that TSC2 did not bind to the EREG promoter between the regions -857 bp to -302 bp or -325 bp to +165 bp. Further, in silico analysis revealed an interesting trend among the transcription factors that were differentially regulated by TSC2 and had putative binding sites on the EREG promoter. A majority of these transcription factors (17/21) were downregulated by the overexpression of TSC2. This observation suggested that the repression of EREG could be an indirect effect due to repression of transcription factors caused by overexpression of TSC2. On the whole, this study revealed novel functions of TSC2 in OSCC with implications in determining novel biomarkers and therapeutic targets. As discussed previously, OSCC has a very flagitious treatment regime. A prodrug approach is thought to aid in targeting chemotherapy (Rooseboom et al., 2004). CYP1B1, a member of the cytochrome P450 family, has been implicated in chemical carcinogenesis (Bandiera et al., 2005; Sliwinski et al., 2010). There exists a general accordance that this protein is overexpressed in a variety of cancers (e.g., colon, lung, renal, bladder, prostate, breast, endometrial and esophageal cancers), making it an ideal candidate for a prodrug therapy (McFadyen et al., 1999; Murray et al., 2001; McFadyen et al., 2004; Sissung et al., 2006; Wen and Walle, 2007; Sliwinski et al., 2010). The activation of the prodrug facilitated by CYP1B1 would enable the targeting of chemotherapy to tumor tissues in which CYP1B1 is specifically overexpressed as a result reducing the non-specific side effects that the current chemotherapy elicits (Rooseboom et al., 2004). This study was aimed at validating the use of CYP1B1 as a target for the prodrug therapy in OSCC. The expression profile of CYP1B1 was analysed in a panel of 51 OSCC tumors, their corresponding normal tissues, an epithelial dysplasia lesion and its matched normal tissue by qRT-PCR, Western blotting and Immunohistochemistry. Counterintuitively, CYP1B1 was found to be downregulated in 77.78% (28/36) tumor tissues in comparison to their corresponding normal tissues as well as in the epithelial dysplasia lesion compared to its matched normal tissue at the transcriptional level, and in 92.86% (26/28) of tumor tissues at the protein level. This clearly demonstrated the downregulation of CYP1B1 at the transcriptional and translational levels in tumor tissues in comparison to their corresponding normal tissues. These observations indicate that caution should be observed as this therapy may not be applicable universally to all cancers. Since CYP1B1 has been shown to be involved in the activation of pro-carcinogens (Murray et al., 2001; Bandiera et al., 2005; Sissung et al., 2006), its inhibition could facilitate the development of a prophylactic therapy for oral cancer. Overall, this study has identified the transactivation of EGFR as a determinant of sensitivity towards combinatorial targeting of PI3K and MTOR in OSCC and has demonstrated that the autophosphorylation of EGFR (Tyr1173) can be used as a marker to judge the sensitivity towards this treatment. In the clinical perspective, the identification of such markers would aid in predicting the efficacy of targeted therapies. Such investigations would enable the strategic treatment of OSCC patients, thus decreasing the time lost in trial and errors for determining the appropriate treatment. Additionally, this study elucidated a novel role of TSC2 in the transcriptional repression of EREG, a prognostic biomarker for OSCC. Further, the study revealed potential prognostic biomarkers as well as therapeutic targets that are regulated by TSC2 by using a whole genome expression microarray. Moreover, the counterintuitive downregulation of CYP1B1 in OSCC tumors suggested the possibility of a prophylactic therapy for oral cancer but also advised a precautionary note for the application of prodrug treatments based on CYP1B1 overexpression in OSCC.