Mostrar registro simples

dc.contributor.advisorScola, Rosana Herminia, 1959-pt_BR
dc.contributor.otherTeixeira, Bernardo Corrêa de Almeida, 1986-pt_BR
dc.contributor.otherUniversidade Federal do Paraná. Setor de Ciências da Saúde. Programa de Pós-Graduação em Medicina Interna e Ciências da Saúdept_BR
dc.creatorSarolli, Bruna Maria Stofelapt_BR
dc.date.accessioned2024-11-08T19:07:29Z
dc.date.available2024-11-08T19:07:29Z
dc.date.issued2023pt_BR
dc.identifier.urihttps://hdl.handle.net/1884/92236
dc.descriptionOrientadora: Profª Drª Rosana Herminia Scolapt_BR
dc.descriptionCoorientador: Dr. Bernardo Corrêa de Almeida Teixeirapt_BR
dc.descriptionDissertação (mestrado) - Universidade Federal do Paraná, Setor de Ciências da Saúde, Programa de Pós-Graduação em Medicina Interna e Ciências da Saúde. Defesa : Curitiba, 20/10/2023pt_BR
dc.descriptionInclui referênciaspt_BR
dc.description.abstractResumo: As miopatias inflamatórias (MI) são um grupo heterogêneo de doenças idiopáticas que têm como principal característica a fraqueza muscular proximal progressiva. São subdivididas em cinco categorias (dermatomiosite, polimiosite, miosite por corpúsculos de inclusão, miopatia necrotizante e síndrome de sobreposição), todas com manifestações clínicas, radiológicas e histológicas específicas. Por se tratar de um conjunto de doenças complexas mas potencialmente tratáveis, a implementação de tecnologias que proporcionem uma avaliação rápida, não-invasiva e multiplanar desses pacientes tornou-se imprescindível. A ressonância magnética (RM) cumpre esse papel quando estabelece, de forma cada vez mais assetiva, padrões específicos de distribuição de danos estimando inclusive o potencial de gravidade dessa doença. O objetivo desse trabalho é validar os achados de imagem na RM capazes de aumentar a detecção diagnóstica das MI. Estabelecer através desse estudo quais são as características distintivas de imagem que, correlacionadas com o resultado das biópsias musculares, permitem estreitar os diagnósticos diferenciais, podendo inclusive determinar o subtipo da MI de cada paciente. Na metodologia desse estudo analítico, observacional, longitudinal e retrospectivo foram catalogados paciente com hipótese diagnóstica de MI registrados no banco de dados, que fizeram exames de imagem (RM) e biópsia muscular por histoquímica. Os pacientes selecionados com suspeita de MI constam no banco de dados do Serviço de Doenças Neuromusculares do Complexo Hospital de Clínicas/UFPR, registrados entre 2010 a 2021. Os exames de RM são do arquivo da Unidade de Diagnóstico por Imagem (UDIM) do CHC-UFPR, feitos de 2014 a 2021. Os resultados das biópsias musculares são do laboratório de Histoquímica do Serviço de Doenças Neuromusculares do Complexo Hospital de Clínicas/UFPR, feitos de 2010 a 2021. Dos 144 pacientes selecionados apenas 51 pacientes preencherem esses três requisitos principais: suspeita diagnóstica de MI, biópsia muscular e exame de RM. Cada um desses 51 pacientes tinha mais de uma RM, totalizando 168 exames que foram devidamente analisados por dois radiologistas experientes que classificaram os pacientes quanto a probabilidade de MI com base nos achados de imagem e cegos para os resultados das biópsias. Somente na segunda fase do estudo que o terceiro pesquisador revelou o resultado das biópsias para que fossem comparados com as análises das RM. O estudo estatístico comparando os resultados da RM com os resultados da bióspia muscular foi o parâmetro utilizado para avaliar e validar a performance diagnóstica desse método de imagem. O primeiro resultado foi em relação ao grau de concordância entre os radiologistas que analisaram as imagens de RM; segundo o teste Kappa (IC 95%) houve um grau de concordância quase perfeito para três critérios (0,81-1,00), não se observando nenhum ítem com grau de concordância baixo (menor que 0,57), o que nos permite afirmar que houve uma interpretação equivalente, não sendo esse um viés do estudo. Foram então correlacionados os achados de imagem com as biópsias musculares. A RM apresentou uma sensibilidade diagnóstica de 94% (IC 95%: 80,32% a 99,28%), especificidade de 47% (IC 95%: 22,98 a 72,19%), e uma acurácia de 78% (IC 95%: 64,68% a 88,71%). Com o Nomograma de Fagan observamos que a probabilidades pré-teste da doença no grupo amostral (51 pacientes) era de 67%, aumentando 11 pontos percentuais a probabilidade no pós-teste com um exame positivo (67% para 78%) e reduzindo 47 pontos percentuais no pósteste com exame negativo (67% para 20%), ou seja, um paciente com suspeita de MI e com um exame negativo tem cerca de 80% de chance de, de fato, não ter a doença. O poder discriminante desse teste diagnóstico (RM) foi bom/muito próximo de muito bom, com área sob a curva ROC de 0,77. A conclusão do trabalho foi de que a RM como ferramenta diagnóstica não-invasiva apresentou boa precisão, bom poder discriminante, alta acurácia, alta sensibilidade e baixa especificidade na investigação de pacientes com suspeita de MI. Os padrões de imagem bem definidos na RM são excelentes preditores de MI, aumentando a probabilidade da doença quando demonstra achados positivos e reduzindo consideravelmente essa chance com achados negativos. A RM além de estreitar os diagnósticos diferencias é bastante efetiva no controle evolutivo e de tratamento desses pacientes bem como na marcação e orientação das biópsias musculares. Contudo não se trata de um método infalível, sendo mandatório o estudo histopatológico, genético ou, mais recentemente, com anticorpos, nos paciente com alta suspeição clínica e sem achados de imagem característicos.pt_BR
dc.description.abstractAbstract: Inflammatory myopathies (MI) are a heterogeneous group of idiopathic diseases whose main characteristic is progressive proximal muscle weakness. They are subdivided into five categories (dermatomyositis, polymyositis, inclusion body myositis, necrotizing myopathy and overlap syndrome), all with specific clinical, radiological and histological manifestations. As this is a set of complex but potentially treatable diseases, the implementation of technologies that provide a rapid, non-invasive and multiplanar assessment of these patients has become necessary. Magnetic resonance imaging (MRI) fulfills this role when it establishes, in an increasingly assertive way, specific damage distribution patterns, including estimating the potential severity of this disease. The objective of this work is to validate MR imaging findings capable of increasing the diagnostic detection of MI. Establish through this study which are the distinctive image characteristics that, correlated with the results of muscle biopsies, allow narrowing the differential diagnoses, and can even determine the MI subtype of each patient. The methodology of this analytical, observational, longitudinal and retrospective study involved cataloging patients with diagnostic hypotheses of MI registered in the database, who underwent imaging tests (MRI) and muscle biopsy by histochemistry. The selected patients with suspected IM are included in the database of the Neuromuscular Diseases Service of the Complexo Hospital de Clínicas/UFPR, registered between 2010 and 2021. The MRI exams are from the archive of the Imaging Diagnostic Unit (UDIM) of the CHC- UFPR, carried out from 2014 to 2021. The results of the muscle biopsies are from the Histochemistry laboratory of the Neuromuscular Diseases Service of the Complexo Hospital de Clínicas/UFPR, carried out from 2010 to 2021. Of the 144 patients selected, only 51 patients meet these three main requirements: suspected diagnosis of MI, muscle biopsy and MRI examination. Each of these 51 patients had more than one MR, totaling 168 scans that were confirmed by two experienced radiologists who assessed patients for likelihood of MI based on imaging findings and blinded to biopsy results. It was only in the second phase of the study that the third researcher revealed the results of the biopsies so that they could be compared with the MRI analyses. The statistical study comparing MRI results with muscle biopsy results was the parameter used to evaluate and validate the diagnostic performance of this imaging method. The first result was in relation to the degree of agreement between the radiologists who analyzed the MR images; According to the Kappa test (95% CI), there was an almost perfect degree of agreement for three criteria (0.81-1.00), with no item being observed with a low degree of agreement (less than 0.57), which gives us allows us to state that there was an equivalent interpretation, which was not a view of the study. The imaging findings were then correlated with the muscle biopsies. MRI had a diagnostic sensitivity of 94% (95% CI: 80.32% to 99.28%), specificity of 47% (95% CI: 22.98 to 72.19%), and an accuracy of 78% (95% CI: 64.68% to 88.71%). Using the Fagan Nomogram, we observed that the pre-test probability of the disease in the sample group (51 patients) was 67%, increasing the post-test probability by 11 percentage points with a positive (67% to 78%) and negative test 47 percentage points in the post-test with a negative test (67% to 20%), that is, a patient with suspected IM and a negative test has around an 80% chance of, in fact, not having the disease. The discriminating power of this diagnostic test (MRI) was good/very close to very good, with an area under the ROC curve of 0.77. The conclusion of the study was that MRI as a non-invasive diagnostic tool presented good precision, good discriminating power, high accuracy, high sensitivity and low specificity in the investigation of patients with suspected MI. Well-defined image patterns on MRI are excellent predictors of MI, increasing the probability of the disease when they demonstrate positive results and considerably reducing this chance with negative findings. MRI, in addition to narrowing the differences between diagnoses, is very effective in monitoring the evolution and treatment of these patients, as well as marking and guiding muscle biopsies. However, it is not an infallible method, and histopathological, genetic or, more recently, antibody studies are mandatory in patients with high clinical suspicion and without specific imaging findings.pt_BR
dc.format.extent1 recurso online : PDF.pt_BR
dc.format.mimetypeapplication/pdfpt_BR
dc.languagePortuguêspt_BR
dc.subjectMiositept_BR
dc.subjectEspectroscopia de ressonância magnéticapt_BR
dc.subjectMúsculos - Doençaspt_BR
dc.subjectBiópsiapt_BR
dc.subjectClinica Médicapt_BR
dc.titleValidação dos achados de ressonância magnética no diagnóstico das miopatias inflamatóriaspt_BR
dc.typeDissertação Digitalpt_BR


Arquivos deste item

Thumbnail

Este item aparece na(s) seguinte(s) coleção(s)

Mostrar registro simples