INTRODUÇÂO: O termo lesão subepitelial se refere a qualquer protrusão ao lúmen do trato gastrointestinal recoberta por mucosa de aspecto normal. A realização de biópsias endoscópicas apresenta rendimento diagnóstico limitado. A ecoendoscopia é considerada o teste diagnóstico de escolha para avaliar diversas características da lesão subepitelial. OBJETIVO: Em relação às lesões subepiteliais gástricas, avaliar dados clínicos, topográficos e ecoendoscópicos como fatores preditores do diagnóstico histopatológico. MÉTODOS: selecionados 55 pacientes adultos atendidos no Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo, no período de outubro de 2003 a agosto de 2011 com diagnóstico de lesão subepitelial gástrica à endoscopia digestiva alta, submetidos à ecoendoscopia, com diagnóstico histopatológico, utilizando-se como procedimentos de investigação a ecoendoscopia e a análise histológica/imuno-histoquímica do material obtido por punção ecoguiada e/ou ressecção/biópsia endoscópica/cirúrgica. As seguintes variáveis foram incluídas para análise, relacionando-as ao diagnóstico histopatológico: tamanho; camada ecoendoscópica; aspecto ecográfico; limites; detecção de fluxo ao Doppler e distribuição topográfica gástrica. RESULTADO: Utilizando modelo logístico para variáveis associadas aos diagnósticos (p <,05), observamos que no caso do tumor gastrointestinal estromal (GIST) a probabilidade da lesão localizar-se na cárdia é baixa (4,5%); há maior risco de GIST em pacientes acima de 57 anos (RC=8,9, IC95%7.6,10.2), lesão >= 21mm (RC=7,15, IC95%5.88, 8.43), com fluxo ao Doppler (RC =9, IC95%6.6, 11.4), limite irregular (RC= 7,75, IC95%6, 9.4) e inserida na 4ª. camada parietal (RC=18,8 IC95% 16.7, 20.94); o leiomioma apresentou alta probabilidade (95%) para cárdia com RC = 390 (IC95% 387, 394); o modelo de regressão múltipla indicou as variáveis dimensão, distribuição topográfica gástrica e camada parietal como significativas para GIST, e distribuição topográfica gástrica (cárdia) para leiomioma. CONCLUSÃO: Existe associação entre a localização da lesão subepitelial gástrica na topografia da cárdia e os diagnósticos de Leiomioma e GIST, com comportamento inverso, sendo o leiomioma o diagnóstico mais provável nesta situação.O modelo logístico de regressão múltipla indica que as variáveis significativas para afastar o diagnóstico de GIST são localização na cárdia, fora da 4a. camada parietal ecoendoscópica e diâmetro da lesão de até 20mm / BACKGROUND: The term subepithelial mass (SEM) refers to any protrusion of the lumen of the gastrointestinal tract covered by a normal appearance mucosa. The performance of endoscopic biopsies has limited diagnostic yield. Endoscopic ultrasonography (EUS) is considered the diagnostic test of choice to assess various characteristics of SEM. AIM: to investigate the association between patients\' clinical characteristics, EUS features and gastric topography with the histopathological diagnosis of gastric SEM, using as diagnostic gold standard the histological and immunohistochemical analysis of the material obtained by fine-needle aspiration and/or surgical resection. METHODS: fifty-five patients selected at the Clinics Hospital - University of São Paulo, from October 2003 to August 2011 with a endoscopic diagnosis of gastric SEM, who underwent EUS, with histopathologic confirmed diagnosis. The following variables were included for analysis: size, echoendoscopic layer, sonographic appearance, echogenicity, irregular outer limits, Doppler flow signal and topographic distribution. RESULTS: Applying logistic regression for variables associated with the diagnoses (P < .05), we found that in the case of gastrointestinal stromal tumor (GIST) the probability of the lesion to be located in the cardia is low (4.5%); there is greater risk of GIST in patients older than 57 years (OR = 8.9, 95% CI 7.6,10.2), with lesion >= 21mm (OR = 7.15, 95% CI 5.88, 8:43), positive Doppler (OR = 9, 95% CI 6.6, 11.4), irregular outer limits (OR = 7.75, 95% CI 6, 9.4) and located at 4th. parietal layer (OR = 18.8 95% CI16.7, 20.94); if leiomyoma, the likelihood of this lesion in the cardia was high (greater than 95%) with odds ratio of 390; multiple regression model indicated the size, topographic distribution and gastric parietal layer as significant for GIST, and gastric topographic distribution (cardia) for leiomyoma. CONCLUSION: There is an association between the location of gastric subepithelial lesion in the topography of cardia and diagnostics of leiomyoma and GIST, with opposite behavior, being leiomyoma the most likely diagnosis. Multiple regression analysis indicates cardia location, lesion outside 4th. parietal layer and diameter of up to 20mm as significant variables to exclude GIST diagnosis

In der Diagnostik der gastrointestinalen Stromatumoren (GIST) spielt neben der Histologie die Immunhistochemie eine zentrale Rolle. Die vorliegende Arbeit befasst sich mit der Fragestellung, welche Wertigkeit der Mutationsanalyse im diagnostischen Kontext zukommt und wie stabil Immunphänotyp und Mutationsstatus im Verlauf der Erkrankung tatsächlich sind. In drei Fällen rezidivierter GIST war die Histomorphologie, die Immunhistochemie und der Mutationsstatus im Vergleich zum Primärtumor stabil. Bei den untersuchten synchron auftretenden Tumoren von drei Patienten waren in der Mutationsanalyse unterschiedliche Ergebnisse zu erheben. Bei zwei Patienten unterstützte das unterschiedliche Mutationsmuster das Vorliegen synchroner Tumoren, bei einem Patienten ist das Vorliegen eines Primärtumors und einer Metastase statt einem synchronen GIST wahrscheinlich. Die Untersuchung metastasierter GIST wurde an verschiedenen Tumoren von neun Patienten durchgeführt. Acht der neun Fälle zeigten sich bezüglich der Metastasen genotypisch stabil, einer der acht Fälle wies zusätzlich einen Zugewinn einer Punktmutation auf, die als Möglichkeit eines Tumormosaiks oder als neu erworbene zusätzliche Mutation zu werten sein könnte. Zudem wurden 28 Fälle unklarer spindelzelliger Tumoren mit uneinheitlichem immunhistochemischen Profil untersucht. In Zusammenschau mit der Mutationsanalyse war eine eindeutige Bestimmung der Tumorentität möglich. Abschließend zeigt sich die Kombination aus Histomorphologie, immunhistochemischer Untersuchung und Mutationsanalyse als gutes diagnostisches Mittel zur Sicherung der Tumorentität und Entdeckung eventuell neu aufgetretener prognostisch relevanter Mutationen mit therapeutischer Konsequenz.

INTRODUÇÂO: O termo lesão subepitelial se refere a qualquer protrusão ao lúmen do trato gastrointestinal recoberta por mucosa de aspecto normal. A realização de biópsias endoscópicas apresenta rendimento diagnóstico limitado. A ecoendoscopia é considerada o teste diagnóstico de escolha para avaliar diversas características da lesão subepitelial. OBJETIVO: Em relação às lesões subepiteliais gástricas, avaliar dados clínicos, topográficos e ecoendoscópicos como fatores preditores do diagnóstico histopatológico. MÉTODOS: selecionados 55 pacientes adultos atendidos no Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo, no período de outubro de 2003 a agosto de 2011 com diagnóstico de lesão subepitelial gástrica à endoscopia digestiva alta, submetidos à ecoendoscopia, com diagnóstico histopatológico, utilizando-se como procedimentos de investigação a ecoendoscopia e a análise histológica/imuno-histoquímica do material obtido por punção ecoguiada e/ou ressecção/biópsia endoscópica/cirúrgica. As seguintes variáveis foram incluídas para análise, relacionando-as ao diagnóstico histopatológico: tamanho; camada ecoendoscópica; aspecto ecográfico; limites; detecção de fluxo ao Doppler e distribuição topográfica gástrica. RESULTADO: Utilizando modelo logístico para variáveis associadas aos diagnósticos (p <,05), observamos que no caso do tumor gastrointestinal estromal (GIST) a probabilidade da lesão localizar-se na cárdia é baixa (4,5%); há maior risco de GIST em pacientes acima de 57 anos (RC=8,9, IC95%7.6,10.2), lesão >= 21mm (RC=7,15, IC95%5.88, 8.43), com fluxo ao Doppler (RC =9, IC95%6.6, 11.4), limite irregular (RC= 7,75, IC95%6, 9.4) e inserida na 4ª. camada parietal (RC=18,8 IC95% 16.7, 20.94); o leiomioma apresentou alta probabilidade (95%) para cárdia com RC = 390 (IC95% 387, 394); o modelo de regressão múltipla indicou as variáveis dimensão, distribuição topográfica gástrica e camada parietal como significativas para GIST, e distribuição topográfica gástrica (cárdia) para leiomioma. CONCLUSÃO: Existe associação entre a localização da lesão subepitelial gástrica na topografia da cárdia e os diagnósticos de Leiomioma e GIST, com comportamento inverso, sendo o leiomioma o diagnóstico mais provável nesta situação.O modelo logístico de regressão múltipla indica que as variáveis significativas para afastar o diagnóstico de GIST são localização na cárdia, fora da 4a. camada parietal ecoendoscópica e diâmetro da lesão de até 20mm / BACKGROUND: The term subepithelial mass (SEM) refers to any protrusion of the lumen of the gastrointestinal tract covered by a normal appearance mucosa. The performance of endoscopic biopsies has limited diagnostic yield. Endoscopic ultrasonography (EUS) is considered the diagnostic test of choice to assess various characteristics of SEM. AIM: to investigate the association between patients\' clinical characteristics, EUS features and gastric topography with the histopathological diagnosis of gastric SEM, using as diagnostic gold standard the histological and immunohistochemical analysis of the material obtained by fine-needle aspiration and/or surgical resection. METHODS: fifty-five patients selected at the Clinics Hospital - University of São Paulo, from October 2003 to August 2011 with a endoscopic diagnosis of gastric SEM, who underwent EUS, with histopathologic confirmed diagnosis. The following variables were included for analysis: size, echoendoscopic layer, sonographic appearance, echogenicity, irregular outer limits, Doppler flow signal and topographic distribution. RESULTS: Applying logistic regression for variables associated with the diagnoses (P < .05), we found that in the case of gastrointestinal stromal tumor (GIST) the probability of the lesion to be located in the cardia is low (4.5%); there is greater risk of GIST in patients older than 57 years (OR = 8.9, 95% CI 7.6,10.2), with lesion >= 21mm (OR = 7.15, 95% CI 5.88, 8:43), positive Doppler (OR = 9, 95% CI 6.6, 11.4), irregular outer limits (OR = 7.75, 95% CI 6, 9.4) and located at 4th. parietal layer (OR = 18.8 95% CI16.7, 20.94); if leiomyoma, the likelihood of this lesion in the cardia was high (greater than 95%) with odds ratio of 390; multiple regression model indicated the size, topographic distribution and gastric parietal layer as significant for GIST, and gastric topographic distribution (cardia) for leiomyoma. CONCLUSION: There is an association between the location of gastric subepithelial lesion in the topography of cardia and diagnostics of leiomyoma and GIST, with opposite behavior, being leiomyoma the most likely diagnosis. Multiple regression analysis indicates cardia location, lesion outside 4th. parietal layer and diameter of up to 20mm as significant variables to exclude GIST diagnosis

O presente estudo tem por objetivo avaliar as características clinicopatológicas, imunoistoquímicas e moleculares de 513 casos brasileiros de tumor estromal gastrintestinal, incluindo o estudo da expressão imunoistoquímica de CD117, CD34, proteína DOG1, actina de músculo liso, desmina, proteína S-100, antígeno Ki-67, proteína p53, molécula de adesão CD44v3, receptor para fator de crescimento epidérmico, proteína HER2 e receptor para fator de crescimento derivado de plaqueta alfa, além da análise molecular de mutações envolvendo os genes KIT e receptor para fator de crescimento derivado de plaqueta alfa e da pesquisa de amplificação dos genes receptor para fator de crescimento epidérmico e HER2 por hibridização \"in situ\" fluorescente. As características clínicas e morfológicas dos casos estudados não foram diferentes das referidas pela literatura. Houve expressão de CD117 (KIT) em 95,7% dos casos. A proteína DOG1 foi expressa em 85,9% dos tumores, sendo capaz de diagnosticar 20% dos casos com ausência de expressão de CD117 (KIT). O índice de proliferação celular determinado pelo antígeno Ki-67 foi superior nos casos classificados como de alto risco para comportamento biológico agressivo segundo critérios do \"National Institutes of Health\". A expressão da proteína p53 esteve restrita aos casos classificados como de alto risco. Não se observou expressão imunoistoquímica da molécula de adesão CD44v3. A proteína receptora para fator de crescimento epidérmico foi expressa em 84,4% dos casos, porém não se observou superexpressão da proteína HER2. A expressão imunoistoquímica da proteína receptora para fator de crescimento derivado de plaqueta alfa foi observada em 94,4% dos casos estudados, não apresentando relação com o tipo de mutação encontrado. As mutações do gene KIT foram as mais frequentemente observadas, principalmente do éxon 11. Mutações do gene receptor para fator de crescimento derivado de plaqueta alfa foram observadas em 8,1% dos casos, sendo que 54,5% dos casos com ausência de expressão imunoistoquímica de CD117 (KIT) apresentavam mutações nesse gene. A hibridização \"in situ\" fluorescente não demonstrou amplificação dos genes receptor para fator de crescimento epidérmico e HER2 nos tumores estromais gastrintestinais. / The present study presents the clinicopathologic, immunohistochemical and molecular genetic features of 513 Brazilian cases of gastrointestinal stromal tumor, including the immunohistochemical expression of CD117, CD34, DOG1 protein, smooth muscle actin, desmin, S-100 protein, Ki-67 antigen, p53 protein, CD44v3 adhesion molecule, epidermal growth factor receptor, HER2 protein and platelet derived growth factor receptor alpha, the mutation analysis of KIT and platelet-derived growth factor receptor alpha genes, and the investigation of amplification of HER2 and epidermal growth factor receptor genes by fluorescence \"in situ\" hybridization. The clinicopathologic features of Brazilian gastrointestinal stromal tumors do not differ from those published from other countries. CD117 (KIT) was expressed by immunohistochemistry in 95.7% of cases. DOG1 protein was expressed in 85.9% of tumors, being able to establish the diagnosis of GIST in 20% of those cases that were negative for CD117 (KIT). Ki-67 proliferation index was higher in those cases classified as high-risk of aggressive behavior by the National Institutes of Health consensus approach. The immunohistochemical expression of p53 protein was restricted to cases classified as high-risk of aggressive behavior. The adhesion molecule CD44v3 was not expressed in any of the cases. The epidermal growth factor receptor protein was overexpressed in 84.4% of cases, and HER2 protein was not expressed in all cases. The platelet-derived growth factor receptor alpha protein was detected by immunohistochemistry in 94.4% of tumors, and there was no relationship between its expression and the mutation status. KIT mutations were the most frequent, mainly of exon 11. Plateletderived growth factor receptor alpha mutations were found in 8.1% of the cases, and 54.5% of those cases that were negative for CD117 (KIT) had mutations in platelet-derived growth factor receptor alpha gene. Fluorescence \"in situ\" hybridization revealed no amplification of epidermal growth factor receptor and HER2 genes in gastrointestinal stromal tumors.

O presente estudo tem por objetivo avaliar as características clinicopatológicas, imunoistoquímicas e moleculares de 513 casos brasileiros de tumor estromal gastrintestinal, incluindo o estudo da expressão imunoistoquímica de CD117, CD34, proteína DOG1, actina de músculo liso, desmina, proteína S-100, antígeno Ki-67, proteína p53, molécula de adesão CD44v3, receptor para fator de crescimento epidérmico, proteína HER2 e receptor para fator de crescimento derivado de plaqueta alfa, além da análise molecular de mutações envolvendo os genes KIT e receptor para fator de crescimento derivado de plaqueta alfa e da pesquisa de amplificação dos genes receptor para fator de crescimento epidérmico e HER2 por hibridização \"in situ\" fluorescente. As características clínicas e morfológicas dos casos estudados não foram diferentes das referidas pela literatura. Houve expressão de CD117 (KIT) em 95,7% dos casos. A proteína DOG1 foi expressa em 85,9% dos tumores, sendo capaz de diagnosticar 20% dos casos com ausência de expressão de CD117 (KIT). O índice de proliferação celular determinado pelo antígeno Ki-67 foi superior nos casos classificados como de alto risco para comportamento biológico agressivo segundo critérios do \"National Institutes of Health\". A expressão da proteína p53 esteve restrita aos casos classificados como de alto risco. Não se observou expressão imunoistoquímica da molécula de adesão CD44v3. A proteína receptora para fator de crescimento epidérmico foi expressa em 84,4% dos casos, porém não se observou superexpressão da proteína HER2. A expressão imunoistoquímica da proteína receptora para fator de crescimento derivado de plaqueta alfa foi observada em 94,4% dos casos estudados, não apresentando relação com o tipo de mutação encontrado. As mutações do gene KIT foram as mais frequentemente observadas, principalmente do éxon 11. Mutações do gene receptor para fator de crescimento derivado de plaqueta alfa foram observadas em 8,1% dos casos, sendo que 54,5% dos casos com ausência de expressão imunoistoquímica de CD117 (KIT) apresentavam mutações nesse gene. A hibridização \"in situ\" fluorescente não demonstrou amplificação dos genes receptor para fator de crescimento epidérmico e HER2 nos tumores estromais gastrintestinais. / The present study presents the clinicopathologic, immunohistochemical and molecular genetic features of 513 Brazilian cases of gastrointestinal stromal tumor, including the immunohistochemical expression of CD117, CD34, DOG1 protein, smooth muscle actin, desmin, S-100 protein, Ki-67 antigen, p53 protein, CD44v3 adhesion molecule, epidermal growth factor receptor, HER2 protein and platelet derived growth factor receptor alpha, the mutation analysis of KIT and platelet-derived growth factor receptor alpha genes, and the investigation of amplification of HER2 and epidermal growth factor receptor genes by fluorescence \"in situ\" hybridization. The clinicopathologic features of Brazilian gastrointestinal stromal tumors do not differ from those published from other countries. CD117 (KIT) was expressed by immunohistochemistry in 95.7% of cases. DOG1 protein was expressed in 85.9% of tumors, being able to establish the diagnosis of GIST in 20% of those cases that were negative for CD117 (KIT). Ki-67 proliferation index was higher in those cases classified as high-risk of aggressive behavior by the National Institutes of Health consensus approach. The immunohistochemical expression of p53 protein was restricted to cases classified as high-risk of aggressive behavior. The adhesion molecule CD44v3 was not expressed in any of the cases. The epidermal growth factor receptor protein was overexpressed in 84.4% of cases, and HER2 protein was not expressed in all cases. The platelet-derived growth factor receptor alpha protein was detected by immunohistochemistry in 94.4% of tumors, and there was no relationship between its expression and the mutation status. KIT mutations were the most frequent, mainly of exon 11. Plateletderived growth factor receptor alpha mutations were found in 8.1% of the cases, and 54.5% of those cases that were negative for CD117 (KIT) had mutations in platelet-derived growth factor receptor alpha gene. Fluorescence \"in situ\" hybridization revealed no amplification of epidermal growth factor receptor and HER2 genes in gastrointestinal stromal tumors.

In der Diagnostik der gastrointestinalen Stromatumoren (GIST) spielt neben der Histologie die Immunhistochemie eine zentrale Rolle. Die vorliegende Arbeit befasst sich mit der Fragestellung, welche Wertigkeit der Mutationsanalyse im diagnostischen Kontext zukommt und wie stabil Immunphänotyp und Mutationsstatus im Verlauf der Erkrankung tatsächlich sind. In drei Fällen rezidivierter GIST war die Histomorphologie, die Immunhistochemie und der Mutationsstatus im Vergleich zum Primärtumor stabil. Bei den untersuchten synchron auftretenden Tumoren von drei Patienten waren in der Mutationsanalyse unterschiedliche Ergebnisse zu erheben. Bei zwei Patienten unterstützte das unterschiedliche Mutationsmuster das Vorliegen synchroner Tumoren, bei einem Patienten ist das Vorliegen eines Primärtumors und einer Metastase statt einem synchronen GIST wahrscheinlich. Die Untersuchung metastasierter GIST wurde an verschiedenen Tumoren von neun Patienten durchgeführt. Acht der neun Fälle zeigten sich bezüglich der Metastasen genotypisch stabil, einer der acht Fälle wies zusätzlich einen Zugewinn einer Punktmutation auf, die als Möglichkeit eines Tumormosaiks oder als neu erworbene zusätzliche Mutation zu werten sein könnte. Zudem wurden 28 Fälle unklarer spindelzelliger Tumoren mit uneinheitlichem immunhistochemischen Profil untersucht. In Zusammenschau mit der Mutationsanalyse war eine eindeutige Bestimmung der Tumorentität möglich. Abschließend zeigt sich die Kombination aus Histomorphologie, immunhistochemischer Untersuchung und Mutationsanalyse als gutes diagnostisches Mittel zur Sicherung der Tumorentität und Entdeckung eventuell neu aufgetretener prognostisch relevanter Mutationen mit therapeutischer Konsequenz.:Inhaltsverzeichnis Abkürzungsverzeichnis 1 1. Einleitung 2 1.1 Definition und Epidemiologie 2 1.1.1 Definition 2 1.1.2 Epidemiologie 3 1.2 Histologie 3 1.2.1 Spindelzellige GIST 4 1.2.2 Epitheloide GIST 5 1.2.3 Intermediäre GIST 6 1.2.4 Mitosen 7 1.3 Immunhistochemie 8 1.4 Molekulare Pathologie 9 1.5 Klinische Diagnostik 11 1.6 Krankheitsverlauf und Risikoabschätzung 11 1.7 Therapie 13 1.7.1 Patienten mit lokalem Tumorgeschehen 13 1.7.2 Patienten mit lokal fortgeschrittenem oder metastasiertem GIST 13 2. Zielsetzung 15 3. Material und Methoden 16 3.1 Untersuchungsgut 16 3.1.1 Rezidivierte gastrointestinale Stromatumoren 16 3.1.2 Synchrone gastrointestinale Stromatumoren 16 3.1.3 Metastasierte gastrointestinale Stromatumoren 16 3.1.4 Unklare spindelzellige Tumoren 17 3.2 Chemikalien 17 3.3 Verbrauchsmaterialien 17 3.4 Chemikalien-Zusammensetzungen 18 3.5 Geräte 19 3.6 Antikörper Immunhistochemie 20 3.7 Oligonukleotide 20 3.8 Bestimmung des Risikoprofils und TNM-Klassifikation 22 3.9 Immunhistochemie 24 3.9.1 Probenaufarbeitung 25 3.9.2 Durchführung und Färbung 25 3.9.3 Auswertung und Kontrolle 26 3.10 Mutationsanalyse 27 3.10.1 Probenaufbereitung 27 3.10.2 Entparaffinierung 27 3.10.3 DNA-Extraktion 27 3.10.4 Amplifikation / Polymerase-Kettenreaktion 29 3.10.5 Konzentrationsbestimmung des PCR-Produktes 30 3.10.6 Kapillargelelektrophorese 30 3.10.7 Aufreinigung der PCR-Produkte 32 3.10.8 Sanger-Sequenzierung 32 3.10.9 Sequenzauswertung 33 4. Ergebnisse 35 4.1 Patienten mit Lokalrezidiv eines GIST 37 4.1.1 Risiko der Krankheitsprogression nach Fletcher und Miettinen 37 4.1.2 Immunhistochemie 38 4.1.3 Mutationsanalyse 38 4.2 Patienten mit synchron aufgetretenen GIST 39 4.2.1 Risiko der Krankheitsprogression nach Fletcher und Miettinen 40 4.2.2 Immunhistochemie 41 4.2.3 Mutationsanalyse 41 4.3 Patienten mit metastasiertem GIST 42 4.3.1 Risiko der Krankheitsprogression nach Fletcher und Miettinen 44 4.3.2 Immunhistochemie 45 4.3.3 Mutationsanalyse 47 4.4 Patienten mit unklaren spindelzelligen Tumoren 49 4.4.1 Risiko der Krankheitsprogression nach Fletcher und Miettinen 51 4.4.2 Immunhistochemie 53 4.4.3 Mutationsanalyse 54 5. Diskussion 57 5.1 Patienten mit Lokalrezidiv eines GIST 57 5.2 Patienten mit synchronen GIST 59 5.3 Patienten mit metastasiertem GIST 61 5.4 Patienten mit unklaren spindelzelligen Tumoren 70 Zusammenfassung 73 Tabellenverzeichnis 76 Abbildungsverzeichnis 78 Literaturverzeichnis 79 Erklärung 84 Danksagung 85 Lebenslauf 86

No description available.

Diss. (sammanfattning) Uppsala : Uppsala universitet, 2007. / Härtill 4 uppsatser.

The present thesis project focused on the preclinical study of Neurotensin receptor 1 (Ntsr1), Endoglin/CD105 (Eng), Glypican 6 (Gpc6) and Sprouty homolog 4 (Spry4) as potential markers or molecular targets for future therapeutic interventions of gastrointestinal stromal tumors (GIST). Ntsr1 expression was characterized was reported in a paper that I co-authored entitled “Neurotensin receptor 1 is expressed in gastrointestinal stromal tumors but not in interstitial cells of Cajal.” by Gromova et al. PLOS ONE, 2009. As radio-labeled NTSR1 ligand analogues have already been reported for whole-body imaging and therapeutic interventions, prompting us to investigate NTSR1 as a target for in vivo imaging of GIST.Next, expression of Eng was characterized in the KitK641E murine GIST model, human GIST, GIST882 and BaF3 cells. This study has been reported in “Endoglin/CD105 is expressed in KIT positive cells in the gut and gastrointestinal stromal tumors” by Gromova et al. JCMM, 2011, a paper I co-authored. As result, an American start-up approached us to assess their proprietary compound targeting ENG on GIST882 cells. However, concerns were raised about possible non-selective action and the project was stalled by the company.Subsequently, GIST tissue microarrays were examined by immunohistochemistry using the sole commercially available GPC6 antibody. No statistical correlation could be found between GPC6-ir and GIST clinic-pathological features and concerns were raised about the reliability of the GPC6 antibody used. Ultimately, Spry4 was investigated in the last part of my thesis. In vitro, we have demonstrated that Spry4 is specifically upregulated by the ERK pathway in GIST882 cells. In vivo, Spry4 deficient mice showed an ICC hyperplasia in antrum and colon, using a new ICC quantification method developed in the lab, which was reminiscent of the oncogenic GIST murine model KitK641E. Similarities between Spry4 KO and KitK641E heterozygous animals were even further emphasized by functional studies, as both genotypes showed a delay in transit time. This study lead to the publication “Hyperplasia of interstitial cells of Cajal in Sprouty homolog 4 deficient mice” by Thys et al. 2015, PLOS ONE. / Doctorat en Sciences biomédicales et pharmaceutiques (Médecine) / info:eu-repo/semantics/nonPublished

No description available.