Cheung Ka Wai. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2004. / Includes bibliographical references (leaves 83-93). / Abstracts in English and Chinese. / Acknowledgments --- p.i / Publications --- p.ii / Awards --- p.iii / List of abbreviations --- p.iv / List of figures --- p.vi / List of tables --- p.vii / Abstract in English --- p.viii / Abstract in Chinese --- p.ix / Table of Contents / Chapter Chapter 1 --- Introduction --- p.1 / Chapter Chapter 2 --- literature Review / Chapter 2.1 --- Anatomy and Physiology of Endometrium --- p.2 / Chapter 2.2 --- Endometrial cancer --- p.5 / Chapter 2.2.1 --- Epidemiology --- p.6 / Chapter 2.2.2 --- Etiologies and Risk Factors --- p.7 / Chapter 2.3 --- Pathology --- p.11 / Chapter 2.3.1 --- Grading of endometrial cancer --- p.14 / Chapter 2.3.2 --- Staging of endometrial cancer --- p.14 / Chapter 2.4 --- Prevention and Treatment --- p.16 / Chapter 2.5 --- Molecular alterations in endometrial cancer --- p.17 / Chapter 2.5.1 --- Genetic alterations in endometrial cancer --- p.18 / Chapter 2.5.1.1 --- Oncogene activation --- p.19 / Chapter 2.5.1.2 --- Tumor suppressor gene inactivation --- p.20 / Chapter 2.5.1.2.1 --- Mutation and loss of heterozygosity of tumor suppressor genes in endometrial cancer --- p.22 / Chapter 2.5.2 --- Epigenetic alterations --- p.27 / Chapter 2.5.2.1 --- CpG islands methylation --- p.29 / Chapter 2.5.2.2 --- de novo methylation --- p.29 / Chapter 2.5.2.3 --- Detection of gene promoter hypermethylation --- p.31 / Chapter 2.5.2.4 --- Epigenetic alteration in endometrial cancer --- p.31 / Chapter 2.5.2.5 --- Promoter hypermethylation of tumor suppressor genes in other cancers --- p.39 / Chapter 2.5.3 --- Microsatellite instability --- p.42 / Chapter Chapter 3 --- The objectives of study --- p.45 / Chapter Chapter 4 --- Materials and Methods --- p.46 / Chapter 4.1 --- Samples --- p.46 / Chapter 4.1.1 --- Formalin fixed paraffin embedded tissues --- p.46 / Chapter 4.1.2 --- Cell lines --- p.46 / Chapter 4.2 --- Histological grading and staging of samples --- p.47 / Chapter 4.3 --- Microdissection on tissue sections --- p.47 / Chapter 4.4 --- Extraction of nucleic acid / Chapter 4.4.1 --- Extraction of DNA from paraffin-embedded tissues --- p.48 / Chapter 4.4.2 --- Extraction of DNA from cell lines --- p.49 / Chapter 4.5 --- DNA methylation analysis --- p.49 / Chapter 4.5.1 --- Overview of Methylation-Specific PCR (MSP) --- p.49 / Chapter 4.5.2 --- Bisulfite modification of DNA --- p.50 / Chapter 4.5.3 --- Methylation specific PCR (MSP) --- p.51 / Chapter 4.6 --- Microsatellite Analysis --- p.53 / Chapter 4.7 --- Statistical analysis --- p.56 / Chapter Chapter 5 --- Results / Chapter 5.1 --- Clinical-pathological features of endometrioid adenocarcinoma --- p.57 / Chapter 5.2 --- Promoter hypermethylation in endometrial cancer --- p.57 / Chapter 5.3 --- Microsatellite status (MSI) analysis --- p.65 / Chapter Chapter 6 --- Discussion / Chapter 6.1 --- Promoter hypermethylation in endometrial cancer --- p.71 / Chapter 6.1.1 --- Concurrent hypermethylation of multiple genesin endometrioid adenocarcinoma and its precursor lesions --- p.72 / Chapter 6.1.1.1 --- Promoter hypermethylation of E-cad --- p.73 / Chapter 6.1.1.2 --- Promoter hypermethylation of APC --- p.73 / Chapter 6.1.1.3 --- Promoter hypermethylation of MGMT --- p.74 / Chapter 6.1.1.4 --- Promoter hypermethylation of RASSF1A --- p.75 / Chapter 6.1.1.5 --- Promoter hypermethylation of hMLH-1 --- p.76 / Chapter 6.1.1.6 --- Promoter hypermethylation in ECA coexisting with hyperplasia and not coexisting with hyperplasia --- p.77 / Chapter 6.1.2 --- Promoter hypermethylation in SCA --- p.77 / Chapter 6.2 --- Microsatellite status analysis --- p.78 / Chapter 6.2.1 --- MSI in endometrial cancer --- p.78 / Chapter 6.2.2 --- MSI and concurrent promoter hypermethylation --- p.79 / Chapter 6.2.3 --- MSI and promoter hypermethylation of hMLH-1 --- p.80 / Chapter Chapter 7 --- Conclusion --- p.81 / Further studies --- p.82 / References --- p.83

子宮内膜癌是最常见的妇科恶性肿瘤，其中有80%属于子宮内膜腺样癌，这一癌症发病的分子机制尚未清楚。研究表明，癌基因表达异常或者功能异常在肿瘤的发生发展过程中具有重要的作用。另一方面，抑癌基因在肿瘤细胞中特异性甲基化失活通常导致细胞恶性转变和肿瘤的发生。本实验将阐明癌基因阴阳1和抑癌基因PCDH10在人类子宮内膜腺样癌发病中的作用。 / 本实验第一部分研究多功能转录因子阴阳1（YY1）在子宮内膜腺样癌发病过程中的作用。首先本实验证实YY1在子宮内膜腺样癌临床标本和癌细胞系中均明显表达上调，并且上调的程度与肿瘤的FIGO分期相关。接着体外细胞培养和裸鼠荷瘤模型的实验均提示抑制YY1 表达可抑制癌细胞增殖和体外迁移，而过表达YY1则促进癌细胞增殖。这些结果表明YY1在子宮内膜腺样癌发病中具有促进作用。进一步全细胞基因组转录谱分析提示YY1 介入子宮内膜腺样发病的各个方面，并通过抑制抑癌基因APC的表达发挥发挥重要作用。深入的分子机制研究发现一个新的表观抑制作用模型：YY1可募集EZH2等多梳蛋白到APC启动子区并导致后者组蛋白3赖氨酸27上三甲基化，从而抑制APC基因转录。此外，本实验还发现YY1在子宮内膜腺样癌的表达增高是由于微小RNA，miR-193a-5p，在此癌中表达下降所导致的。所以，本实验第一部分的结果揭示了miR-193a-5p-YY1-APC这条全新的信号通路在子宮内膜腺样癌发病中发挥重要作用，并可作为潜在的治疗靶点。 / 本实验第二部分鉴定出PCDH10作为子宮内膜腺样癌一个新的抑癌基因。通过5-氮杂-2'-氧胞嘧啶处理和亚硫酸氢钠测序的方法，我们证实抑癌基因PCDH10在子宮内膜腺样癌中失活是由于其启动子区DNA甲基化所致，并且这种DNA甲基化介导的PCDH10表达沉默在子宮内膜腺样癌临床标本和癌细胞系中很常见，但不存在于正常子宮内膜组织。另外，在子宮内膜腺样癌细胞系体外实验中恢复PCDH10的表达可抑制细胞增殖、单细胞克隆形成，促进细胞凋亡。 同时在体实验荷瘤模型中恢复PCDH10的表达也可抑制肿瘤细胞增殖，这些结果与其肿瘤抑制功能相符。 / 总之，本实验结果阐明了YY1和PCDH10在子宮内膜腺样癌发病过程中新的作用，拓展了子宮内膜腺样癌发病分子机制的研究并为其药物治疗提供了潜在的靶点。 / Endometrial cancer is the most common gynecologic malignancy and about 80% of these cancers are endometrial Endometrioid carcinoma (EEC). The molecular mechanisms underlying EEC tumorigenesis are under-explored. Aberrant expression and function of oncogenes promote tumor development by modulating many aspects of tumor cell growth. On the other hand, tumor specific promoter methylation on tumor suppressor genes (TSG), which are generally unmethylated in normal cells, usually initiate and promote malignant transformation and cancer initiation. Our study aims to characterize the functions of an oncogenic transcription factor Yin Yang 1 (YY1) and a novel tumor suppressor gene PCDH10 in Human Endometrioid Endometrial Adenocarcinoma. / In the first part of our study, we investigated the function of a multifunctional TF, YY1 in EEC tumorigenesis. We demonstrated YY1 is up-regulated in EEC cell lines and primary tumors and its expression is associated with FIGO stages. Depletion of YY1 inhibits EEC cell proliferation and migration both in vitro and in vivo whereas over-expression of YY1 promotes EEC cell growth. These results suggest that YY1 functions as an onocogenic factor in EEC. Transcriptome analysis revealed a significant effect of YY1 on critical aspects of EEC tumorigenesis and its down-regulation of APC transcripts. Further mechanistic investigation uncovered a new epigenetic silencing mode of Adenomatosis Polyposis Coli (APC) by YY1 through recruitment of EZH2 and trimethylation of histone 3 lysine 27 in its promoter region. Additionally, YY1 over-expression was found to be a consequence of miR-193a-5p down-regulation through direct miR-193a-5p-YY1 interplay. Our results therefore established a novel miR-193a-5p-YY1-APC regulatory axis contributing to EEC development, which may serve as future intervention target. / In the second part of our study, we identified PCDH10 as a novel tumor suppressor gene in EEC. By using bisulfate genomic sequencing combined with pharmacologic demethylation drug treatment, we elucidated that PCDH10 inactivation in EEC is a consequence of DNA hypermethylation on its promoter region. Further study suggested that hypermethylation-mediated PCDH10 silencing was a common event in EEC cell lines and clinical samples, but not in normal endometrial tissues. Restoration of PCDH10 expression in EEC cells suppressed cell proliferation, inhibited single cell colony formation and induced cell apoptosis; moreover, overexpression of PCDH10 inhibited EEC xenograft tumor growth in vivo.These results suggest PCDH10 acts as a tumor suppressor. / Together, our results reveal the novel functions of YY1 and PCDH10 in EEC. These findings add novel insights into the molecular mechanisms of EEC development and progression, which may serve as potential therapeutic targets for this disease. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Yang, Yihua. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2012. / Includes bibliographical references (leaves 170-186). / Abstracts also in Chinese. / TITLE --- p.I / ABSTRACT --- p.III / ACKNOWLEDGEMENTS --- p.VII / PUBLICATION --- p.IX / ABBREVIATIONS --- p.X / LIST OF FIGURES --- p.XIII / LIST OF TABLES --- p.XVI / TABLES OF CONTENT --- p.XVII / Chapter Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Endometrioid Endometrial Adenocarcinoma (EEC) --- p.1 / Chapter 1.1.1 --- Epidemiology --- p.2 / Chapter 1.1.2 --- Etiology and risk factors --- p.3 / Chapter 1.1.3 --- Treatment and prognosis --- p.8 / Chapter 1.1.4 --- Molecular Mechanisms --- p.9 / Chapter 1.1.5 --- APC and Wnt/β-catenin signaling pathway --- p.12 / Chapter 1.1.6 --- Summary --- p.18 / Chapter 1.2 --- Epigenetic modifications and EEC --- p.20 / Chapter 1.2.1 --- Epigenetic modifications --- p.20 / Chapter 1.2.2 --- Epigenetic and cancer --- p.21 / Chapter 1.2.5 --- Summary --- p.30 / Chapter Chapter 2 --- Material and Method --- p.31 / Chapter 2.1 --- Tissue samples --- p.31 / Chapter 2.2 --- Cell culture --- p.32 / Chapter 2.3 --- Cell proliferation assays --- p.33 / Chapter 2.4 --- Cell migration assay --- p.34 / Chapter 2.5 --- 3-deazaneplanocin A (Dznep) or 5-aza-2'-deoxycytidine (5-aza) treatment --- p.34 / Chapter 2.6 --- Computational prediction --- p.35 / Chapter 2.7 --- Cell cycle assay --- p.35 / Chapter 2.8 --- Apoptosis assay --- p.36 / Chapter 2.9 --- Total RNAs, Total proteins and Genomic DNA extraction --- p.36 / Chapter 2.10 --- Bisulfite Genomic Sequencing --- p.38 / Chapter 2.11 --- Oligonucleotides --- p.39 / Chapter 2.12 --- RT-PCR, Semi-quantitative PCR and Real-time PCR --- p.41 / Chapter 2.13 --- microRNA validation --- p.43 / Chapter 2.14 --- Plasmid construction --- p.43 / Chapter 2.15 --- Transfection --- p.45 / Chapter 2.16 --- Luciferase reporter assay --- p.45 / Chapter 2.17 --- Western blotting --- p.46 / Chapter 2.18 --- Immunofluorescence ( IF ) --- p.48 / Chapter 2.19 --- Immunohistochemistry (IHC) --- p.50 / Chapter 2.20 --- ChIP assay --- p.53 / Chapter 2.21 --- Sequencing and base calling --- p.55 / Chapter 2.22 --- Read mapping to genome with splice-aware aligner sequenced --- p.55 / Chapter 2.23 --- Xenograft mouse model --- p.55 / Chapter 2.24 --- Statistical analysis --- p.57 / Chapter Chapter 3 --- Yin Yang 1 Plays an Oncogenic Role in Human Endometrioid Endometrial Adenocarcinoma --- p.58 / Chapter 3.1 --- YIN YANG 1(YY1) --- p.58 / Chapter 3.1.1 --- YY1 structure --- p.58 / Chapter 3.1.2 --- YY1 function --- p.59 / Chapter 3.1.3 --- YY1 and epigenetic --- p.61 / Chapter 3.1.4 --- YY1 and cancer --- p.62 / Chapter 3.1.5 --- Regulation of YY1 expression and activity --- p.66 / Chapter 3.2 --- Results --- p.68 / Chapter 3.2.1 --- YY1 is up-regulated in EEC lines and localizes in nuclei of EEC cells --- p.68 / Chapter 3.2.2 --- YY1 expression level is associated with EEC clinicopathological features --- p.72 / Chapter 3.2.3 --- Knock-down of YY1 by RNAi inhibits EEC cell proliferation and migration --- p.77 / Chapter 3.2.4 --- Ectopic expression of YY1 promotes EEC cell proliferation --- p.84 / Chapter 3.2.5 --- YY1 does not affect EEC cell cycle and cell apoptosis --- p.91 / Chapter 3.2.6 --- Genome-wide characterization of YY1-mediated transcriptome changes --- p.94 / Chapter 3.2.7 --- Gene Ontology analysis of YY1 targets on EEC tumorigenesis --- p.98 / Chapter 3.2.8 --- YY1 inhibits APC gene expression and functions --- p.101 / Chapter 3.2.9 --- YY1 inhibits APC expression through recruiting EZH2 and causing H3K27me3. --- p.105 / Chapter 3.2.10 --- Knock-down of YY1 does not change DNA methylation status of CpG island of APC gene --- p.117 / Chapter 3.2.11 --- SiYY1 oligo injection inhibits tumor grows in vivo --- p.119 / Chapter 3.2.12 --- miR-193a-5p is down-regulated in EEC cell lines and clinical samples --- p.126 / Chapter 3.2.13 --- miR-193a-5p targets YY1 3’UTR and inhibits YY1 expression --- p.128 / Chapter 3.2.14 --- miR-193a-5p inhibits tumor grow in vivo --- p.133 / Chapter 3.3 --- Discussion --- p.136 / Chapter 3.3.1 --- YY1 oncogenic functions in EEC --- p.136 / Chapter 3.3.2 --- YY1 epigenetically silences APC --- p.138 / Chapter 3.3.3 --- miR-193a-5p down-regulates YY1 in EEC --- p.139 / Chapter 3.4 --- Conclusion --- p.141 / Chapter Chapter 4 --- The tumor suppressive functions of PCDH10 in Human Endometrioid Endometrial Adenocarcinoma --- p.143 / Chapter 4.1 --- Introduction --- p.143 / Chapter 4.1.1 --- PCDH10 structure and function --- p.143 / Chapter 4.1.2 --- PCDH10 and tumor --- p.146 / Chapter 4.1 --- Results --- p.149 / Chapter 4.2.1 --- PCDH10 is down-regulated in EEC cell lines and clinical samples --- p.149 / Chapter 4.2.2 --- PCDH10 is hypermethylated in EEC cell lines and clinical samples --- p.150 / Chapter 4.2.3 --- Pharmacologic demethylation restores PCDH10 expression in EEC cell lines --- p.152 / Chapter 4.2.4 --- Ectopic over-expression of PCDH10 inhibits EEC cell proliferation --- p.154 / Chapter 4.2.5 --- PCDH10 over-expression induces EEC cell apoptosis --- p.161 / Chapter 4.2.6 --- PCDH10 over-expression inhibits tumor grows in vivo --- p.166 / Chapter 4.3 --- Discussion and future plan --- p.169 / REFERENCE --- p.170

INTRODUÇÃO: A ressecção do ramo ascendente da mandíbula foi, durante várias décadas, considerada o tratamento de eleição para os tumores da loja amigdalina e região retromolar, independente do grau de acometimento do osso mandibular, ocasionando um déficit funcional e estético considerável, muitas vezes, com prejuízos irreparáveis à qualidade de vida, às vezes desnecessário. Assim, a ressecção marginal do osso mandibular surgiu como uma alternativa de tratamento viável, uma vez que a manutenção de um segmento do ramo mandibular, em lesões sem comprometimento ósseo, não aumenta os índices de recidiva, tampouco compromete os princípios de radicalidade oncológica. OBJETIVO E MÉTODOS: Por meio de estudo retrospectivo de Outubro de 1994 a Dezembro de 2001, foram comparados 42 pacientes portadores de tumores avançados de região retromolar e loja amigdalina, sendo 20 deles submetidos a ressecção marginal do osso mandibular e 22 submetidos a ressecção do ramo ascendente da mandíbula, em relação a complicações, seqüela de procedimentos, recidiva locorregional e sobrevida. RESULTADOS: Dos 20 pacientes tratados com mandibulectomia marginal, avaliados por um período de 9 a 60 meses, sete (35%) pacientes morreram com doença, com sobrevida mínima de 09 meses, 3 por recidiva local, 3 por recidiva regional e 1 por recidiva locorregional. Um paciente morreu no pós-operatório imediato. Na avaliação da peça cirúrgica encontramos todas as margens livres, considerada exígua em profundidade em dois pacientes, um deles falecido por recidiva local. Houve disseminação linfonodal em 15 pacientes sendo com ruptura extracapsular em 4, encontrada em 2 pacientes com recidiva regional. O controle locorregional foi obtido em 63% dos pacientes. Dos 22 pacientes tratados com ressecção segmentar do osso mandibular, com intervalo de seguimento de 14 a 60 meses, 8 (36,4%) morreram pela doença, com sobrevida mínima de 9 meses, 5 por recidiva local e 3 por recidiva à distância. Um paciente morreu no pós-operatório imediato. As margens foram livres em 20 pacientes e, em 3 exíguas, um deles falecido por recidiva local. Houve disseminação linfonodal em 12 pacientes com ruptura extracapsular em 7 pacientes. O controle locorregional foi obtido em 61% dos pacientes. Na curva de análise de sobrevida, pelo método de Kaplan-Meier, o grupo tratado com mandibulectomia marginal apresentou uma taxa de 42%, com intervalo de 31 a 52 meses, erro padrão de 5 meses e intervalo de confiança de 95% e o grupo tratado com ressecção segmentar 38% com intervalo de 27 a 48 meses, erro padrão de 5 meses, um intervalo de confiança de 95%. A comparação pelo teste de Log Rank, não paramétrico apresentou p<0,8329 e pelo teste t-Student p< 0,621 ambos não significantes. As principais complicações foram a infecção local em 5 (11,9%) pacientes e a fístula orocutânea em 4 (9,5%). Houve uma fratura da placa de titânio, dois pacientes evoluíram com osteorradionecrose e nove com disfunção da articulação têmporo-mandibular. CONCLUSÕES: Não houve diferenças estatisticamente significantes entre os grupos de ressecção marginal e segmentar nos critérios analisados. Portanto, a conservação do ramo ascendente da mandíbula, em lesões que não apresentem envolvimento mandibular, mesmo avançadas, não aumenta o índice de recidiva. / INTRODUCTION: For several decades, resection of the ascending ramus of the mandible was considered to be mandatory for the treatment of tonsil and retromolar trigone tumours, independent from the damage degree of the mandibular bone, causing considerable functional and aesthetic deficit, many times with irreparable quality of life loss, sometimes unnecessary. Thus, marginal resection of the mandibular bone appeared as a feasible alternative treatment, since maintaining a segment of the mandibular ramus in lesions without bone involvement did not increase the recurrence indexes or compromise the principles of oncological radicalness. OBJECTIVES AND METHODS: Through a retrospective study from October 1994 to December 2001, 42 patients with advanced retromolar and tonsil tumors were compared, 20 undergoing marginal resection of the mandibular bone and 22 undergoing segmental resection of the ascending ramus of the mandible, with regard complications, injury originated from the procedure, locoregional recurrence and survival. RESULTS: From the 20 patients undergoing to marginal mandibulectomy, assessed for a period of 09 to 60 months, seven (35%) patients died of the disease, with a minimal survival of 9 months: 3 due to local recurrence, 3 due to regional recurrence and 1 due to local and regional recurrence. One patient died in the immediate postoperative period. When assessing the surgical part, all the margins were found to be free and in two patients, were considered to be of little depth, one of them being found in one of the patients that died from local recurrence. There was lymph nodal dissemination in 15 patients, with capsular rupture in 4, of which two presented regional recurrence. The locoregional control was obtained in 63% of the patients. From the 22 patients undergoing to segmental resection of the mandibular bone, with a follow-up varying from 14 to 60 months, 8 (36.4%) died of the disease, with a minimum survival of 9 months, 5 due to local recurrence and 3 due to distant recurrence. One patient died in the immediate postoperative period. The surgical margins were considered free from disease in 20 patients and, in three they were too small, one patient died of local recurrence. There was lymph nodal dissemination in 12 patients, and 7 presented capsular rupture. The locoregional control was obtained in 61% of the patients. In the survival analysis curve, by the Kaplan-Meier method, the group submitted to marginal mandibulectomy presented with a rate of 42%, with an interval of 31 to 52 months, standard error of 5 months and confidence interval of 95%, and the group submitted to segmentary resection, 38% with an interval of 27 to 48 months, a standard error of 5 months, and a confidence interval of 95%. The comparison using the non-parametric Log-Rank test presented p<0.8326 and the t-Student test p< 0.621, both not statistically significant. The main complications were local infection in 5 (11.9%) patients and oro-cutaneous fistula in 4 (9.5%). There was one titanium plate fracture, two patients developed osteoradionecrosis and nine temporo-mandibular dysfunction. CONCLUSIONS: There weren t statistically significant differences between the marginal and the segmental resection groups in relation to the analyzed criteria. So, the preservation of the ascending ramus of the mandible, in lesions which do not present mandible commitment, even if not advanced, doesn t increase the recurrence rate.

INTRODUÇÃO: A ressecção do ramo ascendente da mandíbula foi, durante várias décadas, considerada o tratamento de eleição para os tumores da loja amigdalina e região retromolar, independente do grau de acometimento do osso mandibular, ocasionando um déficit funcional e estético considerável, muitas vezes, com prejuízos irreparáveis à qualidade de vida, às vezes desnecessário. Assim, a ressecção marginal do osso mandibular surgiu como uma alternativa de tratamento viável, uma vez que a manutenção de um segmento do ramo mandibular, em lesões sem comprometimento ósseo, não aumenta os índices de recidiva, tampouco compromete os princípios de radicalidade oncológica. OBJETIVO E MÉTODOS: Por meio de estudo retrospectivo de Outubro de 1994 a Dezembro de 2001, foram comparados 42 pacientes portadores de tumores avançados de região retromolar e loja amigdalina, sendo 20 deles submetidos a ressecção marginal do osso mandibular e 22 submetidos a ressecção do ramo ascendente da mandíbula, em relação a complicações, seqüela de procedimentos, recidiva locorregional e sobrevida. RESULTADOS: Dos 20 pacientes tratados com mandibulectomia marginal, avaliados por um período de 9 a 60 meses, sete (35%) pacientes morreram com doença, com sobrevida mínima de 09 meses, 3 por recidiva local, 3 por recidiva regional e 1 por recidiva locorregional. Um paciente morreu no pós-operatório imediato. Na avaliação da peça cirúrgica encontramos todas as margens livres, considerada exígua em profundidade em dois pacientes, um deles falecido por recidiva local. Houve disseminação linfonodal em 15 pacientes sendo com ruptura extracapsular em 4, encontrada em 2 pacientes com recidiva regional. O controle locorregional foi obtido em 63% dos pacientes. Dos 22 pacientes tratados com ressecção segmentar do osso mandibular, com intervalo de seguimento de 14 a 60 meses, 8 (36,4%) morreram pela doença, com sobrevida mínima de 9 meses, 5 por recidiva local e 3 por recidiva à distância. Um paciente morreu no pós-operatório imediato. As margens foram livres em 20 pacientes e, em 3 exíguas, um deles falecido por recidiva local. Houve disseminação linfonodal em 12 pacientes com ruptura extracapsular em 7 pacientes. O controle locorregional foi obtido em 61% dos pacientes. Na curva de análise de sobrevida, pelo método de Kaplan-Meier, o grupo tratado com mandibulectomia marginal apresentou uma taxa de 42%, com intervalo de 31 a 52 meses, erro padrão de 5 meses e intervalo de confiança de 95% e o grupo tratado com ressecção segmentar 38% com intervalo de 27 a 48 meses, erro padrão de 5 meses, um intervalo de confiança de 95%. A comparação pelo teste de Log Rank, não paramétrico apresentou p<0,8329 e pelo teste t-Student p< 0,621 ambos não significantes. As principais complicações foram a infecção local em 5 (11,9%) pacientes e a fístula orocutânea em 4 (9,5%). Houve uma fratura da placa de titânio, dois pacientes evoluíram com osteorradionecrose e nove com disfunção da articulação têmporo-mandibular. CONCLUSÕES: Não houve diferenças estatisticamente significantes entre os grupos de ressecção marginal e segmentar nos critérios analisados. Portanto, a conservação do ramo ascendente da mandíbula, em lesões que não apresentem envolvimento mandibular, mesmo avançadas, não aumenta o índice de recidiva. / INTRODUCTION: For several decades, resection of the ascending ramus of the mandible was considered to be mandatory for the treatment of tonsil and retromolar trigone tumours, independent from the damage degree of the mandibular bone, causing considerable functional and aesthetic deficit, many times with irreparable quality of life loss, sometimes unnecessary. Thus, marginal resection of the mandibular bone appeared as a feasible alternative treatment, since maintaining a segment of the mandibular ramus in lesions without bone involvement did not increase the recurrence indexes or compromise the principles of oncological radicalness. OBJECTIVES AND METHODS: Through a retrospective study from October 1994 to December 2001, 42 patients with advanced retromolar and tonsil tumors were compared, 20 undergoing marginal resection of the mandibular bone and 22 undergoing segmental resection of the ascending ramus of the mandible, with regard complications, injury originated from the procedure, locoregional recurrence and survival. RESULTS: From the 20 patients undergoing to marginal mandibulectomy, assessed for a period of 09 to 60 months, seven (35%) patients died of the disease, with a minimal survival of 9 months: 3 due to local recurrence, 3 due to regional recurrence and 1 due to local and regional recurrence. One patient died in the immediate postoperative period. When assessing the surgical part, all the margins were found to be free and in two patients, were considered to be of little depth, one of them being found in one of the patients that died from local recurrence. There was lymph nodal dissemination in 15 patients, with capsular rupture in 4, of which two presented regional recurrence. The locoregional control was obtained in 63% of the patients. From the 22 patients undergoing to segmental resection of the mandibular bone, with a follow-up varying from 14 to 60 months, 8 (36.4%) died of the disease, with a minimum survival of 9 months, 5 due to local recurrence and 3 due to distant recurrence. One patient died in the immediate postoperative period. The surgical margins were considered free from disease in 20 patients and, in three they were too small, one patient died of local recurrence. There was lymph nodal dissemination in 12 patients, and 7 presented capsular rupture. The locoregional control was obtained in 61% of the patients. In the survival analysis curve, by the Kaplan-Meier method, the group submitted to marginal mandibulectomy presented with a rate of 42%, with an interval of 31 to 52 months, standard error of 5 months and confidence interval of 95%, and the group submitted to segmentary resection, 38% with an interval of 27 to 48 months, a standard error of 5 months, and a confidence interval of 95%. The comparison using the non-parametric Log-Rank test presented p<0.8326 and the t-Student test p< 0.621, both not statistically significant. The main complications were local infection in 5 (11.9%) patients and oro-cutaneous fistula in 4 (9.5%). There was one titanium plate fracture, two patients developed osteoradionecrosis and nine temporo-mandibular dysfunction. CONCLUSIONS: There weren t statistically significant differences between the marginal and the segmental resection groups in relation to the analyzed criteria. So, the preservation of the ascending ramus of the mandible, in lesions which do not present mandible commitment, even if not advanced, doesn t increase the recurrence rate.

O carcinoma basocelular (CBC) é tumor constituído por diferentes tipos histológicos, que demonstram diverso potencial de agressividade. Sabe-se que a correlação entre os tipos histológicos de CBC encontrados no material de biópsia pré-operatória e no material da peça cirúrgica excisional não é total. Na literatura esta correlação varia de 42,7 a 80,0% quando analisados os tipos histológicos predominantes (THP). No presente estudo foi feita análise retrospectiva de 70 casos de CBC primário submetidos a biópsia préoperatória e cirurgia excisional. A amostra foi analisada estatisticamente quanto ao gênero e idade dos doentes e localização anatômica dos CBC, demonstrando ser comparável aos demais estudos da literatura. Também foram avaliados o tamanho médio tumoral e o tipo de reconstrução utilizado. A média do maior eixo dos CBC foi de 20 mm e 54% dos casos necessitaram reconstrução complexa, como retalhos e enxertos, mostrando ser amostra representativa de tumores de médio a grande porte. A avaliação histológica foi feita de modo padronizado, determinando tanto o THP quanto os tipos histológicos acessórios (THA) encontrados no material das biópsias pré-operatórias e nas peças cirúrgicas excisionais. Houve 78,3% de correlação entre THP da biópsia e peça cirúrgica, 87,0% de correlação entre THP e/ou THA da biópsia e THP da peça cirúrgica e 92,7% de correlação entre tipos agressivos ou não agressivos. Conclui-se que a biópsia préoperatória é útil para predizer o THP de CBC da peça cirúrgica excisional na maioria dos casos. No entanto, é importante ressaltar que, quando descrito apenas o THP encontrado na biópsia, ocorre 21,7% de falha no diagnóstico. Quando descritos THP e THA encontrados na biópsia a falha diagnóstica cai para 13%. Quando a intenção da biópsia for a determinação da presença de tipos de CBC agressivos ou não, a falha no diagnóstico é de apenas 7,3% / Basal cell carcinoma (BCC) is a tumor presenting many histological types, each one possessing a specific aggressivity potential. It\'s known that correlation between histological types found in preoperative biopsy specimens and excisional surgery specimens is not total. When correlation between predominant histological types (PHT) is analyzed, concordance value varies from 42,7 to 80,0% in the literature. In the present study 70 primary BCC submitted to preoperative biopsy and excisional surgery were retrospectively analyzed. The sample was statistically analyzed in terms of patients\' gender and age and anatomical location of the tumour and was found to be similar to other reports in the literature. Average size of tumors and type of surgical reconstruction employed were also evaluated. Average size of the largest tumour axis was 20 mm and 54% of the cases needed complex reconstructions, such as flaps and grafts, demonstrating that the sample was represented by medium to large sized tumors. Histological evaluation was made in a patterned way, determining PHT and accessory histological types (AHT) in both preoperative biopsies and excisional surgery specimens. Results obtained were: 78.3% correlation between biopsy PHT and excisional surgery PHT, 87.0% correlation between biopsy PHT and/or AHT and excisional surgery PHT and 92.7% correlation when BCC were classified as \"aggressive\" or \"non aggressive\" . Conclusion: preoperative biopsy is useful to predict BCC\'s PHT of excisional surgery specimen in most cases. However, it\'s important to note that when biopsy findings are limited to the description of the PHT , there is a 21.7% diagnostic failure. When both PHT and AHT found in biopsy are described, diagnostic failure falls to 13%. When the intention is determining the presence of aggressive or non aggressive types of BCC, diagnostic failure is only 7.3%

O carcinoma basocelular (CBC) é tumor constituído por diferentes tipos histológicos, que demonstram diverso potencial de agressividade. Sabe-se que a correlação entre os tipos histológicos de CBC encontrados no material de biópsia pré-operatória e no material da peça cirúrgica excisional não é total. Na literatura esta correlação varia de 42,7 a 80,0% quando analisados os tipos histológicos predominantes (THP). No presente estudo foi feita análise retrospectiva de 70 casos de CBC primário submetidos a biópsia préoperatória e cirurgia excisional. A amostra foi analisada estatisticamente quanto ao gênero e idade dos doentes e localização anatômica dos CBC, demonstrando ser comparável aos demais estudos da literatura. Também foram avaliados o tamanho médio tumoral e o tipo de reconstrução utilizado. A média do maior eixo dos CBC foi de 20 mm e 54% dos casos necessitaram reconstrução complexa, como retalhos e enxertos, mostrando ser amostra representativa de tumores de médio a grande porte. A avaliação histológica foi feita de modo padronizado, determinando tanto o THP quanto os tipos histológicos acessórios (THA) encontrados no material das biópsias pré-operatórias e nas peças cirúrgicas excisionais. Houve 78,3% de correlação entre THP da biópsia e peça cirúrgica, 87,0% de correlação entre THP e/ou THA da biópsia e THP da peça cirúrgica e 92,7% de correlação entre tipos agressivos ou não agressivos. Conclui-se que a biópsia préoperatória é útil para predizer o THP de CBC da peça cirúrgica excisional na maioria dos casos. No entanto, é importante ressaltar que, quando descrito apenas o THP encontrado na biópsia, ocorre 21,7% de falha no diagnóstico. Quando descritos THP e THA encontrados na biópsia a falha diagnóstica cai para 13%. Quando a intenção da biópsia for a determinação da presença de tipos de CBC agressivos ou não, a falha no diagnóstico é de apenas 7,3% / Basal cell carcinoma (BCC) is a tumor presenting many histological types, each one possessing a specific aggressivity potential. It\'s known that correlation between histological types found in preoperative biopsy specimens and excisional surgery specimens is not total. When correlation between predominant histological types (PHT) is analyzed, concordance value varies from 42,7 to 80,0% in the literature. In the present study 70 primary BCC submitted to preoperative biopsy and excisional surgery were retrospectively analyzed. The sample was statistically analyzed in terms of patients\' gender and age and anatomical location of the tumour and was found to be similar to other reports in the literature. Average size of tumors and type of surgical reconstruction employed were also evaluated. Average size of the largest tumour axis was 20 mm and 54% of the cases needed complex reconstructions, such as flaps and grafts, demonstrating that the sample was represented by medium to large sized tumors. Histological evaluation was made in a patterned way, determining PHT and accessory histological types (AHT) in both preoperative biopsies and excisional surgery specimens. Results obtained were: 78.3% correlation between biopsy PHT and excisional surgery PHT, 87.0% correlation between biopsy PHT and/or AHT and excisional surgery PHT and 92.7% correlation when BCC were classified as \"aggressive\" or \"non aggressive\" . Conclusion: preoperative biopsy is useful to predict BCC\'s PHT of excisional surgery specimen in most cases. However, it\'s important to note that when biopsy findings are limited to the description of the PHT , there is a 21.7% diagnostic failure. When both PHT and AHT found in biopsy are described, diagnostic failure falls to 13%. When the intention is determining the presence of aggressive or non aggressive types of BCC, diagnostic failure is only 7.3%

INTRODUÇÃO: o padrão-ouro para o diagnóstico histológico dos tumores corticais adrenais (TCAs) e sua diferenciação entre adenomas e carcinomas é o sistema de Weiss, cuja aplicação é limitada pela baixa reprodutibilidade de alguns dos critérios que o compõe. Recentemente foi proposto e validado um algoritmo diagnóstico para os TCAs baseado na integridade do arcabouço de reticulina e da membrana basal. Os carcinomas adrenais são tumores raros e apresentam prognóstico reservado, mesmo nos pacientes com doença aparentemente localizada. Além do estadiamento e da extensão da ressecção cirúrgica, outros dados foram reportados na literatura como tendo importância prognóstica, tais como idade ao diagnóstico, padrão funcional, tamanho tumoral, extensão local do tumor primário e alguns achados histológicos e imuno-histoquímicos, com destaque à taxa mitótica e ao índice de Ki-67. O sistema de Weiss, embora permita o diagnóstico diferencial entre adenomas e carcinomas, não foi testado completamente como uma ferramenta para distinguir os carcinomas com boa evolução clínica daqueles com desfecho desfavorável. OBJETIVOS: o presente estudo teve como objetivo primário construir um nomograma para estimar o risco de metástases e recorrência local em portadores de carcinoma adrenal, a partir de dados clínico-patológicos. O objetivo secundário foi avaliar o desempenho do algoritmo da reticulina no diagnóstico diferencial entre adenomas e carcinomas do córtex adrenal. MÉTODOS: para a construção do nomograma, foram analisados dados clínico-patológicos de 129 portadores de carcinomas adrenais atendidos no Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo entre 1976 e 2010. A avaliação do desempenho do algoritmo da reticulina para o diagnóstico histológico dos TCAs foi feita a partir do exame de 89 lâminas (45 adenomas e 44 carcinomas adrenais), que foram classificadas de acordo com o sistema de Weiss e pelo algoritmo da reticulina, pelo mesmo patologista e de forma cega. RESULTADOS: utilizando a análise de regressão logística binária, foi proposto um escore prognóstico com cinco covariáveis: padrão funcional outro que não o hiperandrogenismo isolado, tamanho tumoral > 7,5 cm, tumor primário classificado como T3 ou T4, presença de invasão venosa microscópica e índice mitótico > 5/50 campos de grande aumento. O escore prognóstico estava calibrado de acordo com o teste de aderência de Hosmer-Lemeshow (P =0,9329) e apresentou excelente desempenho global (escore de Brier =0,0738). Finalmente, a capacidade discriminatória do modelo, determinada pela área sob a curva ROC (AROC), foi considerada quase perfeita (AROC, 0,9611; IC95%, 0,92676-0,99552). O modelo preditivo foi validado internamente com 200 reamostragens por bootstrap. A partir destes resultados, desenvolvemos um aplicativo para iOS7 para estimar o risco de metástases e recorrência local em portadores de carcinomas adrenais (ACCPS - Adrenal Cortical Carcinoma Prognostic Score). O aplicativo está disponível para download gratuito na App Store. Em relação à avaliação do desempenho do algoritmo da reticulina, o mesmo apresentou sensibilidade =84,1% e especificidade =95,6% para o diagnóstico de carcinoma adrenal. Não se observou diferença estatisticamente significante entre os diagnósticos obtidos pelo escore de Weiss e pelo algoritmo da reticulina (P =0,1797; Teste exato de McNemar). A concordância dos dois sistemas, avaliada pelo coeficiente k, foi considerada substancial (k =0,7975; IC95%, 0,6728-0,9221). CONCLUSÕES: propusemos e validamos internamente um escore prognóstico baseado em informações clínicopatológicas que estão facilmente disponíveis ao clínico. O modelo proposto está disponível para uso como um aplicativo gratuito para iOS7 e é capaz de estimar o risco de desfechos desfavoráveis nos pacientes com carcinoma cortical adrenal. O escore pode ter grande utilidade na determinação do tipo e frequência do acompanhamento clínico-radiológico e na decisão de um tratamento profilático, após a retirada cirúrgica completa do carcinoma adrenal. O algoritmo da reticulina teve desempenho equivalente ao sistema de Weiss para o diagnóstico histológico dos tumores corticais adrenais, mas oferece a vantagem de ter aplicação mais fácil / INTRODUCTION: The gold standard for the histological diagnosis of adrenal cortical tumors (ACTs) and for the differentiation between adenomas and carcinomas is the Weiss system, whose application is limited by poor reproducibility of some of its criteria. Recently, a diagnostic algorithm for ACT diagnosis based on the integrity of the reticulin network and the basal membrane has been proposed and validated. Adrenal carcinomas are rare tumors and have a poor prognosis, even in patients with apparently localized disease. Besides tumor staging and extent of surgical resection, other data have been reported in the literature as having prognostic importance, such as age at diagnosis, the functional pattern, tumor size, local extension of the primary tumor and some histological and immunohistochemical findings, such as the mitotic rate and the Ki-67 index. The Weiss system, while allowing the differential diagnosis between adrenal cortical adenomas and carcinomas, has not been fully tested as a tool for distinguishing carcinomas with favorable clinical outcome from those with unfavorable outcome. OBJECTIVES: The primary objective of this study was to construct a nomogram for estimating the risk of metastasis and local recurrence in patients with adrenal cortical carcinoma, based on clinical and pathological data. The secondary objective was to evaluate the performance of the reticulina algorithm in the differential diagnosis between adenomas and carcinomas of the adrenal cortex. METHODS: For the construction of the nomogram, clinical and pathological data from 129 patients with adrenal cortical carcinomas treated at the Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo between 1976 and 2010 were analyzed. The evaluation of the performance of the reticulin algorithm for the histological diagnosis of ACTs was made from the examination of 89 slides (45 adenomas and 44 adrenal carcinomas), which were classified according to the Weiss system and the reticulin algorithm by the same pathologist in a blinded fashion. RESULTS: Using binary logistic regression analysis, we proposed a prognostic score with five covariates: a functional pattern other than isolated hyperandrogenism, a tumor size > 7.5 cm, a primary tumor classified as T3 or T4, the presence of microscopic venous invasion and a mitotic index > 5/50 high-power fields. The prognostic score was calibrated according to the Hosmer-Lemeshow goodness-of-fit test (P =0.9329) and showed excellent overall performance (Brier score =0.0738). Finally, the discriminatory ability of the model, as determined by the area under the ROC curve (AROC), was considered near-perfect (AROC, 0.9611; 95%CI, 0.92676 to 0.99552). The predictive model was internally validated with 200 bootstrap resamples. Based on these results, we develop an app for iOS7 for estimating the risk of metastasis and local recurrence in patients with adrenal carcinomas (ACCPS - Adrenal Cortical Carcinoma Prognostic Score). The app is available for free download on the App Store. Regarding the assessment of the performance of the reticulin algorithm, it had a sensitivity of 84.1% and specificity of 95.6% for the diagnosis of adrenal cortical carcinoma. No statistically significant difference between the diagnoses obtained by Weiss score and the reticulin algorithm was observed (P =0.1797, McNemar\'s exact test). The concordance between the two systems, assessed by the k coefficient, was considered substantial (k =0.7975; 95%CI, 0.6728 to 0.9221). CONCLUSIONS: we proposed and internally validated a prognostic score based on clinical and pathological information that are readily available to the attending physician. The proposed model is available as a free app for iOS7 and can estimate the risk of adverse outcomes in patients with adrenal cortical carcinoma. The score may be useful in determining the type and frequency of clinical and radiological evaluation and the decision of adjuvant treatment after complete surgical removal of the adrenal carcinoma. The reticulin algorithm was equivalent to the Weiss system for the histological diagnosis of adrenal cortical tumors, but offers the advantage of an easier application

INTRODUÇÃO: o padrão-ouro para o diagnóstico histológico dos tumores corticais adrenais (TCAs) e sua diferenciação entre adenomas e carcinomas é o sistema de Weiss, cuja aplicação é limitada pela baixa reprodutibilidade de alguns dos critérios que o compõe. Recentemente foi proposto e validado um algoritmo diagnóstico para os TCAs baseado na integridade do arcabouço de reticulina e da membrana basal. Os carcinomas adrenais são tumores raros e apresentam prognóstico reservado, mesmo nos pacientes com doença aparentemente localizada. Além do estadiamento e da extensão da ressecção cirúrgica, outros dados foram reportados na literatura como tendo importância prognóstica, tais como idade ao diagnóstico, padrão funcional, tamanho tumoral, extensão local do tumor primário e alguns achados histológicos e imuno-histoquímicos, com destaque à taxa mitótica e ao índice de Ki-67. O sistema de Weiss, embora permita o diagnóstico diferencial entre adenomas e carcinomas, não foi testado completamente como uma ferramenta para distinguir os carcinomas com boa evolução clínica daqueles com desfecho desfavorável. OBJETIVOS: o presente estudo teve como objetivo primário construir um nomograma para estimar o risco de metástases e recorrência local em portadores de carcinoma adrenal, a partir de dados clínico-patológicos. O objetivo secundário foi avaliar o desempenho do algoritmo da reticulina no diagnóstico diferencial entre adenomas e carcinomas do córtex adrenal. MÉTODOS: para a construção do nomograma, foram analisados dados clínico-patológicos de 129 portadores de carcinomas adrenais atendidos no Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo entre 1976 e 2010. A avaliação do desempenho do algoritmo da reticulina para o diagnóstico histológico dos TCAs foi feita a partir do exame de 89 lâminas (45 adenomas e 44 carcinomas adrenais), que foram classificadas de acordo com o sistema de Weiss e pelo algoritmo da reticulina, pelo mesmo patologista e de forma cega. RESULTADOS: utilizando a análise de regressão logística binária, foi proposto um escore prognóstico com cinco covariáveis: padrão funcional outro que não o hiperandrogenismo isolado, tamanho tumoral > 7,5 cm, tumor primário classificado como T3 ou T4, presença de invasão venosa microscópica e índice mitótico > 5/50 campos de grande aumento. O escore prognóstico estava calibrado de acordo com o teste de aderência de Hosmer-Lemeshow (P =0,9329) e apresentou excelente desempenho global (escore de Brier =0,0738). Finalmente, a capacidade discriminatória do modelo, determinada pela área sob a curva ROC (AROC), foi considerada quase perfeita (AROC, 0,9611; IC95%, 0,92676-0,99552). O modelo preditivo foi validado internamente com 200 reamostragens por bootstrap. A partir destes resultados, desenvolvemos um aplicativo para iOS7 para estimar o risco de metástases e recorrência local em portadores de carcinomas adrenais (ACCPS - Adrenal Cortical Carcinoma Prognostic Score). O aplicativo está disponível para download gratuito na App Store. Em relação à avaliação do desempenho do algoritmo da reticulina, o mesmo apresentou sensibilidade =84,1% e especificidade =95,6% para o diagnóstico de carcinoma adrenal. Não se observou diferença estatisticamente significante entre os diagnósticos obtidos pelo escore de Weiss e pelo algoritmo da reticulina (P =0,1797; Teste exato de McNemar). A concordância dos dois sistemas, avaliada pelo coeficiente k, foi considerada substancial (k =0,7975; IC95%, 0,6728-0,9221). CONCLUSÕES: propusemos e validamos internamente um escore prognóstico baseado em informações clínicopatológicas que estão facilmente disponíveis ao clínico. O modelo proposto está disponível para uso como um aplicativo gratuito para iOS7 e é capaz de estimar o risco de desfechos desfavoráveis nos pacientes com carcinoma cortical adrenal. O escore pode ter grande utilidade na determinação do tipo e frequência do acompanhamento clínico-radiológico e na decisão de um tratamento profilático, após a retirada cirúrgica completa do carcinoma adrenal. O algoritmo da reticulina teve desempenho equivalente ao sistema de Weiss para o diagnóstico histológico dos tumores corticais adrenais, mas oferece a vantagem de ter aplicação mais fácil / INTRODUCTION: The gold standard for the histological diagnosis of adrenal cortical tumors (ACTs) and for the differentiation between adenomas and carcinomas is the Weiss system, whose application is limited by poor reproducibility of some of its criteria. Recently, a diagnostic algorithm for ACT diagnosis based on the integrity of the reticulin network and the basal membrane has been proposed and validated. Adrenal carcinomas are rare tumors and have a poor prognosis, even in patients with apparently localized disease. Besides tumor staging and extent of surgical resection, other data have been reported in the literature as having prognostic importance, such as age at diagnosis, the functional pattern, tumor size, local extension of the primary tumor and some histological and immunohistochemical findings, such as the mitotic rate and the Ki-67 index. The Weiss system, while allowing the differential diagnosis between adrenal cortical adenomas and carcinomas, has not been fully tested as a tool for distinguishing carcinomas with favorable clinical outcome from those with unfavorable outcome. OBJECTIVES: The primary objective of this study was to construct a nomogram for estimating the risk of metastasis and local recurrence in patients with adrenal cortical carcinoma, based on clinical and pathological data. The secondary objective was to evaluate the performance of the reticulina algorithm in the differential diagnosis between adenomas and carcinomas of the adrenal cortex. METHODS: For the construction of the nomogram, clinical and pathological data from 129 patients with adrenal cortical carcinomas treated at the Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo between 1976 and 2010 were analyzed. The evaluation of the performance of the reticulin algorithm for the histological diagnosis of ACTs was made from the examination of 89 slides (45 adenomas and 44 adrenal carcinomas), which were classified according to the Weiss system and the reticulin algorithm by the same pathologist in a blinded fashion. RESULTS: Using binary logistic regression analysis, we proposed a prognostic score with five covariates: a functional pattern other than isolated hyperandrogenism, a tumor size > 7.5 cm, a primary tumor classified as T3 or T4, the presence of microscopic venous invasion and a mitotic index > 5/50 high-power fields. The prognostic score was calibrated according to the Hosmer-Lemeshow goodness-of-fit test (P =0.9329) and showed excellent overall performance (Brier score =0.0738). Finally, the discriminatory ability of the model, as determined by the area under the ROC curve (AROC), was considered near-perfect (AROC, 0.9611; 95%CI, 0.92676 to 0.99552). The predictive model was internally validated with 200 bootstrap resamples. Based on these results, we develop an app for iOS7 for estimating the risk of metastasis and local recurrence in patients with adrenal carcinomas (ACCPS - Adrenal Cortical Carcinoma Prognostic Score). The app is available for free download on the App Store. Regarding the assessment of the performance of the reticulin algorithm, it had a sensitivity of 84.1% and specificity of 95.6% for the diagnosis of adrenal cortical carcinoma. No statistically significant difference between the diagnoses obtained by Weiss score and the reticulin algorithm was observed (P =0.1797, McNemar\'s exact test). The concordance between the two systems, assessed by the k coefficient, was considered substantial (k =0.7975; 95%CI, 0.6728 to 0.9221). CONCLUSIONS: we proposed and internally validated a prognostic score based on clinical and pathological information that are readily available to the attending physician. The proposed model is available as a free app for iOS7 and can estimate the risk of adverse outcomes in patients with adrenal cortical carcinoma. The score may be useful in determining the type and frequency of clinical and radiological evaluation and the decision of adjuvant treatment after complete surgical removal of the adrenal carcinoma. The reticulin algorithm was equivalent to the Weiss system for the histological diagnosis of adrenal cortical tumors, but offers the advantage of an easier application

Fibroblast growth factor receptors (FGFR) play key roles in proliferation, differentiation, and tumorigenesis. Many urothelial carcinomas contain activating point mutations or increased expression of FGFR3. However, little is known about the role of other FGFRs. We examined FGFR expression in telomerase-immortalized normal human urothelial cells, urothelial carcinoma cell lines, and tumor samples and showed that FGFR1 expression is increased in a high proportion of cell lines and tumors independent of stage and grade. To determine the role of FGFR1 in low-stage bladder cancer, we overexpressed FGFR1 in telomerase-immortalized normal human urothelial cells and examined changes in proliferation and cell survival in response to FGF2. FGFR1 stimulation increased proliferation and reduced apoptosis. To elucidate the mechanistic basis for these alterations, we examined the signaling cascades activated by FGFR1. FRS2alpha and PLCgamma were activated in response to FGF2, leading to activation of the mitogen-activated protein kinase pathway. The level of mitogen-activated protein kinase activation correlated with the level of cyclin D1, MCL1, and phospho-BAD, which also correlated with FGFR-induced proliferation and survival. Knockdown of FGFR1 in urothelial carcinoma cell lines revealed differential FGFR1 dependence. JMSU1 cells were dependent on FGFR1 expression for survival but three other cell lines were not. Two cell lines (JMSU1 and UMUC3) were dependent on FGFR1 for growth in soft agar. Only one of the cell lines tested (UMUC3) was frankly tumorigenic; here, FGFR1 knockdown inhibited tumor growth. Our results indicate that FGFR1 has significant effects on urothelial cell phenotype and may represent a useful therapeutic target in some cases of urothelial carcinoma.