Orientador: Kleber Gomes Franchini / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Ciencias Medicas / Made available in DSpace on 2018-08-07T12:07:47Z (GMT). No. of bitstreams: 1 Borges_MariaCandidaCalzada_D.pdf: 3690302 bytes, checksum: 0c73a392de92a2836881cbe3034c49f5 (MD5) Previous issue date: 2006 / Resumo: Estudos epidemiológicos têm estabelecido uma relação contínua entre a massa do ventrículo esquerdo (VE) e o risco cardiovascular na população geral e na população hipertensa. O pior prognóstico de pacientes hipertensos com aumento na massa do VE tem sido, em parte, atribuído à disfunção do miocárdio, mas permanece desconhecido se os pacientes hipertensos sem hipertrofia do VE (HVE) definida pelos critérios habituais apresentam alterações na função do miocárdio. Os índices ecocardiográficos derivados das dimensões do VE e das velocidades do fluxo mitral aferidas através do Doppler têm provado ser geralmente não específicos e insensíveis para a detecção de alterações discretas da dinâmica ventricular. O Doppler Tecidual (DT) tem demonstrado ser um método eficiente e rápido para avaliar a função do miocárdio. Nos pacientes hipertensos com hipertrofia do VE tem sido descrita redução das velocidades sistólica e diastólica inicial. Entretanto não tem sido relatada investigação da função sistólica do miocárdio do VE através do DT em pacientes hipertensos sem ou com discreta elevação da massa do VE. O presente estudo avalia se, através da velocidade sistólica e diastólica aferidas pelo DT, é possível identificar alterações na contração e relaxamento do miocárdio de indivíduos hipertensos com ou sem hipertrofia ventricular definida pelos valores de corte habituais e com fração de ejeção preservada. Para tanto indivíduos normotensos e hipertensos com ou sem hipertrofia do ventrículo esquerdo (índice de massa do VE [IMVE] > 51g/m2.7) foram avaliados através da ecocardiografia convencional e do DT em 5 segmentos do anel mitral. Os subgrupos incluíram indivíduos normotensos não obesos (n=16; idade 51 ± 9 anos; 11 feminino; pressão arterial sistólica [PAS] 109 ± 11 mmHg; índice de massa corpórea [IMC] 24 ± 2,7 Kg/m2 ; IMVE 32 ± 5.5g/m2.7), hipertensos não obesos sem HVE (n=16; idade 54 ± 12 anos ; 12 femininos; PAS 166 ± 15 mmHg; IMC 25 ± 2,7 g/m2; IMVE 42 ± 5,5 g/m2.7) e hipertensos não obesos com HVE (n=22; idade 56 ± 10 anos; 10 feminino; PAS 181 ± 19 mmHg; IMC 26 ± 2,3 g/m2; IMVE 69 ± 16 g/m2.7). A fração de ejeção foi comparável entre os subgrupos, mas a fração de encurtamento da parede média foi reduzida nos pacientes hipertensos com hipertrofia do VE (¿26%). O tempo de relaxamento isovolumétrico foi aumentado nos pacientes hipertensos com hipertrofia do VE, enquanto a velocidade A do fluxo mitral encontrou-se aumentada em indivíduos hipertensos com e sem hipertrofia do VE. A velocidade sistólica (SM) e diastólica inicial (EM) aferida através do DT ao nível do anel mitral encontraram-se significativamente reduzidas nos indivíduos hipertensos com e sem HVE quando comparadas com as dos indivíduos normotensos. Evidenciou-se correlação positiva entre SM e EM (r=0,68; p<0,0001) e correlação negativa entre essas duas variáveis e o IMVE (SM, r= -0,41; p=0,002; EM, r= -0,56; p<0,0001). Portanto a redução na velocidade sistólica e diastólica ao nível do anel mitral acompanha o aumento da massa do VE em indivíduos hipertensos, iniciando-se em níveis de IMVE que se encontram abaixo dos valores definidos clinicamente como normais / Abstract: Epidemiological studies have established a continuous relationship between the left ventricular (LV) mass and cardiovascular risk in the general and hypertensive population. The poorer prognosis of hypertensive subjects with major increases in LV mass has been, in part, attributed to myocardial dysfunction, but it remains unknown whether hypertensive subjects without clinically defined LV hypertrophy have subtle abnormalities of myocardial function. Echocardiographic indices derived from LV chamber dimensions and Doppler measurements of flow velocities have been proved to be generally nonspecific and insensitive for the detection of minor abnormalities of cardiac relaxation and contraction. Tissue Doppler imaging (TDI) has been advocated as a reliable, rapid, and efficient method to assess myocardial function. In hypertensive patients with LV hypertrophy, TDI mitral annulus systolic and diastolic velocities have been shown to be reduced. However, a comprehensive TDI approach has not been reported in hypertensive subjects without or with minor increases in LV mass. The present study examines whether, by using TDI early systolic and diastolic velocities, one might identify changes in LV myocardial contraction and relaxation in subsets of hypertensive with and without clear-cut, clinically defined LV hypertrophy and normal LV ejection fraction. Normotensive and hypertensive subjects with and without left ventricular (LV) hypertrophy (LV mass index [LVMI] =51 g/m2.7) were examined by conventional echocardiography and tissue Doppler imaging of mitral annulus motion. The subgroups included non obese normotensive subjects (n=16; age 51±9 years; 11 female; systolic blood pressure [SBP] 109±11 mm Hg, body mass index [BMI] 24±2.7 kg/m2; LVMI 32±5.5 g/m2.7), non obese hypertensive subjects without LV hypertrophy (n=16; age 54±12 years; 12 female; SBP 166±15 mm Hg; BMI 25±2.7 kg/m2; LVMI 42_5.5 g/m2.7), and hypertensive subjects with LV hypertrophy (n=22; age 56±10 years; 10 female; SBP 181±19 mm Hg; BMI 26±2.3 kg/m2; LVMI 69±16 g/m2.7). Ejection fraction was comparable among the subgroups, but midwall fractional shortening was reduced in hypertensive subjects with LV hypertrophy (¿26%). Isovolumic relaxation time was increased in subjects with LV hypertrophy, whereas mitral wave A velocity was increased in hypertensive subjects with and without LV hypertrophy. Tissue Doppler imaging mitral annulus systolic (SM) and diastolic (EM) velocities were reduced in subjects with and without LV hypertrophy compared with normotensive subjects. There was a positive correlation between SM and EM (r=0.68; P<0.0001) and negative correlations between these 2 variables and LVMI (SM, r= -0.41; P=0.002; EM, r=-0.56; P<0.0001). Thus, reductions in mitral annulus systolic and diastolic velocities parallel increases in LV mass in hypertensive subjects, beginning at LV mass within the clinically defined normal values / Doutorado / Medicina Experimental / Doutor em Fisiopatologia Medica

Idrottshjärta är ett kardiovaskulärt tillstånd som uppträder under längre perioder av intensiv träning som orsakar strukturella, funktionella och elektriska förändringar hos hjärtat och är en fysiologisk anpassning som svar på ett ökat hemodynamiskt behov under fysisk ansträngning. De fysiologiska anpassningarna har dock blivit ett diagnostiskt dilemma att urskilja från de patologiska förändringarna såsom hypertrofisk kardiomyopati. Det finns därför ett behov av standardisering av kardiovaskulär screening hos idrottare för att upptäcka underliggande eller dolda kardiomyopatier som kan leda till allvarliga konsekvenser under fysisk ansträngning. Studiens ändamål var att undersöka den systoliska och diastoliska vänsterkammarfunktionen under ansträngning hos idrottare och öka förståelsen om vad som händer med de olika variablerna under arbete. Nio friska idrottare genomförde stressekokardiografi där cardiac index, ejektionsfraktion, fyllnadstryck, mitralisklaffplanets longitudinella rörelse (MAPSE), mitralisinflöde, vävnadsdoppler (e´ och s´) och veninflöde undersöktes före, under och efter ett ansträngningstest på ergometercykel. Variablerna under och efter cykeltestet jämfördes sedan med värdena i vila. Resultaten visade en signifikant ökning av cardiac index, MAPSE och vävnadsdoppler under ansträngning. Sammanfattningsvis visade studien att flera av variablerna förbättrades under ansträngning och en del av de visade sig vara relativt okänsliga för störningar och artefakter vilket kan vara användbart för framtida studie protokoll som avser utföra en hjärtstudie under arbete. / Athlete’s heart is a cardiovascular condition that occurs during extended periods of intense exercise that causes structural, functional and electrical changes of the heart and is a physiological adaptation in response to increased hemodynamic needs during physical exertion. However, the physiological adaptations have become a diagnostic dilemma to distinguish from the pathological changes such as hypertrophic cardiomyopathy. Therefore, there is a need for standardization of cardiovascular screening in athletes to detect underlying or hidden cardiomyopathies that can lead to severe consequences during physical exercise. The aim of the present study was to investigate the systolic and diastolic left ventricular function during exercise in athletes and to increase the understanding of what happens to the various variables during exertion. Nine healthy athletes conducted stress echocardiography where cardiac index, ejection fraction, filling pressure, mitral annular plane systolic excursion (MAPSE), mitral inflow, tissue Doppler imaging (e 'and s') and pulmonary venous inflow were examined before, during and after a cycle ergometer test. The variables during and after the cycle test were then compared to baseline. The results showed a significant increase in cardiac index, MAPSE, and tissue Doppler imaging during exertion. In conclusion, the study showed that several of the variables improved during exertion and some of them proved to be quite insensitive to disturbances and artifacts, which may be useful in future study protocols that consider carrying out a cardiac study during work.

During the last two decades, since the heart has become a direct target for various surgical procedures for the correction of several abnormalities, interest in understanding its basic regulatory mechanisms and performance has increased considerably. Many experiments have been devised to study cardiac performance, under different circumstances, as an isolated preparation or as an integrated part of a functioning system. The results derived from these investigations not only proved to be important conceptual tools in making many observed phenomena more readily comprehensible, but also provided us with valuable basic principles to treat cardiac patients before and after open heart operations. The purpose of this paper is to investigate the control of the function of the heart in dogs with the chest and pericardium open, and under strictly controlled conditions. In spite of all the work j done in this field, there still exists some degree of confusion regarding the relative importance of autoregulatory mechanisms and extrinsic factors governing cardiac performance. Without trying to take sides as to the priority of either of these mechanisms, we have made an effort to confine our interest mainly to the intrinsic autoregulation of the heart.

Background: Chronic aortic regurgitation (AR) patients demonstrate left ventricular (LV) remodeling with increased LV mass and volume but may have a preserved LV ejection fraction (EF). We hypothesize that in chronic AR, global longitudinal systolic and diastolic function will be reduced despite a preserved LV EF. Methods: We studied with Doppler echocardiography 27 normal subjects, 87 patients with chronic AR with a LV EF > 50% (AR + PEF), 66 patients with an EF < 50% [AR + reduced LV ejection fraction (REF)] and 82 patients with hypertensive heart disease. LV volume, transmitral spectral and tissue Doppler were obtained. Myocardial velocities and their timing and longitudinal strain of the proximal and mid wall of each of the 3 apical views were obtained. Results: As compared to normals, global longitudinal strain was reduced in AR + PEF (13.8 ± 4.0%) and AR + REF (11.4 ±4.7%) vs. normals (18.4 ± 3.6%, both p < 0.001). As an additional comparison group for AR + PEF, global longitudinal strain was reduced as compared to patients with hypertensive heart disease (p = 0.032). The average peak diastolic annular velocity (e’) was decreased in AR + PEF (6.9 ± 3.3 cm/s vs. 13.4 ± 2.6 cm/s, p < 0.001) and AR + REF (4.8 ± 2.1 cm/s, p < 0.001). Peak rapid filling velocity/e’ (E/e’) was increased in both AR + PEF (14.4 ± 6.2 vs. 6.2 ± 1.3, p < 0.001) and AR + REF (18.8 ± 6.4, p <0.001 vs. normals). Independent correlates of global longitudinal strain (r = 0.6416, p < 0.001) included EF (p < 0.0001), E/e’ (p < 0.0001), and tricuspid regurgitation velocity (p = 0.0176). Conclusion: With chronic AR, there is impaired longitudinal function despite preserved EF. Moreover, global longitudinal strain was well correlated with noninvasive estimated LV filling pressures and pulmonary systolic arterial pressures.

Miniswine underwent procedures to evaluate treating chronic ischemia with the implantation of autologous satellite cells and laser transmyocardial revascularization (TMR). The objective was to combine two therapies to restore cardiac function. This experiment involved three surgical procedures: (1) placing a constrictor on the coronary artery; (2) producing channels and implanting cells; (3) obtaining samples. The swine were divided into groups: Group 1, Ischemia; Group 2, Ischemia + Laser TMR; Group 3, Ischemia + Laser TMR+ Cells; Group 4, Ischemia + Cells. Sonomicrometry and Millar pressure transducers were used to determine contractility, left ventricular pressure, and pressure-volume loops. There were no significant differences (p<0.05)among the hemodynamic data except for Group 4, which produced significantly lower output values. Morphological evaluation revealed a significantly reduced scar area in Group 3. Although there was a significant difference in scar area, the phenomena behind this improvement as compared to the unimproved hemodynamic function is not understood.

This project offers a multidisciplinary approach towards the acquisition, analysis and visualization of experimental data that pertain to cardiovascular applications. First and foremost, the capabilities of our Time-Resolved Digital Particle Image Velocimetry (TRDPIV) system were improved, allowing near-wall wall TRDPIV on compliant, dynamically moving boundaries. As a result, false flow-field vectors due to reflections from the boundary walls were eliminated, and allowing measurement of wall shear stress, wall shear rate, and oscillating shear index within as little as fifty microns of the boundary. Similar in-vitro measurements have not been reported to date by any other group. Second, an immersive, virtual environment (VE) was developed for the investigation and analysis of vortical, spatio-temporally developing flows with complex fluid-structure interactions. This VE was used to study flows in the cardiovascular system, particularly for flow through mechanical heart valves and inside the heart left ventricle (LV). The simulation provides three-dimensional (3-D) visualization of in-vitro heart flow mechanics, allowing global, volumetric flow analysis, and a useful environment for comparison with in-vivo MRI velocimetry data. 3-D glyphs (symbols representing informational parameters) are used to visually represent the flow parameters in the form of an ellipse attached to a cone, where the ellipse represents a second-order Reynolds stress tensor, and the cone represents the velocity magnitude and direction at a particular point in space, and the color corresponds to an out-of-plane vorticity. This new system has a major advantage over conventional 2-D systems in that it successfully doubles the number of visualized parameters, and allows for visualization of a time-dependent series of flow data in the Virginia Tech CAVETM immersive VE. The user controls his/her viewpoint, and can thus navigate through the simulation and view the flow field from any perspective in the immersive VE. Finally, an edge detection algorithm was developed to determine the inner and outer myocardial boundaries, and from this information calculate the local thickness distribution of the myocardium and a myocardial area approximation. This information is important in validating our in-vitro system, and is integral to the evaluation and diagnosis of congestive heart disease and its progression. / Master of Science

A rejeição é uma das principais complicações após transplante cardíaco (TC). A biópsia endomiocárdica do ventrículo direito (BEVD) continua a ser o padrão-ouro para seu diagnóstico. Há uma necessidade de alternativas não invasivas que permitam um diagnóstico seguro e precoce de rejeição cardíaca, antes do início dos sintomas clínicos. A ecocardiografia com speckle tracking (EST) permite a análise da dinâmica de contração ventricular, possibilitando a detecção precoce de disfunção miocárdica. Os objetivos deste estudo foram comparar os parâmetros da dinâmica ventricular obtidos com a EST em pacientes transplantados e indivíduos não transplantados cardíacos e avaliar o valor da EST e da dosagem sérica de troponina I e peptídeo cerebral natriurético (BNP), como ferramentas não invasivas para a detecção precoce de rejeição celular aguda significativa (RCA) após o TC. Entre janeiro de 2014 e novembro de 2017, foram estudados, prospectivamente, 49 pacientes transplantados cardíacos com função sistólica normal, tanto do ventrículo esquerdo (VE) como do direito (VD) submetidos à BEVD para vigilância ou por suspeita clínica de rejeição. A RCA foi definida como >= 2R pelos critérios da Sociedade Internacional para Transplante de Coração e Pulmão. Quarenta e nove indivíduos saudáveis pareados por idade e sexo constituíram o grupo controle. Todos os pacientes foram submetidos à ecocardiografia convencional e à EST. Os pacientes transplantados foram submetidos, horas antes da BEVD, à dosagem de troponina I e BNP séricos. Parâmetros ecocardiográficos clássicos, strain e strain rate longitudinal global, radial e circunferencial do VE, além de strain longitudinal da parede livre do VD (SLPL-VD) foram analisados. Os 49 pacientes transplantados cardíacos (média etária 45,2 ± 11,5 anos, 28 homens) foram submetidos a 66 biópsias entre 6-12 meses após o TC. A RCA foi detectada em 17 (26%) e ausência de RCA em 49 (74%) biópsias. Strain e strain rate longitudinal, circunferencial e radial globais do VE e SL-PLVD foram significantemente menores, em valores absolutos, nos grupos dos transplantados cardíacos do que no grupo controle. O SL-PLVD apresentou valor absoluto menor no grupo com RCA do que no sem RCA (-18,28 ± 4,8% versus -22,11 ± 2,9%; p < 0,001). Um valor de SL-PLVD < 17,57% (valor absoluto) teve sensibilidade de 71%, especificidade de 90%, valor preditivo positivo de 75%, acurácia de 84% e valor preditivo negativo de 88% para detecção de RCA. A dosagem da troponina I foi significantemente mais alta em pacientes com RCA (p < 0,01). A mediana para este grupo foi 0,19 [intervalo interquartil 0,09-1,31 ng/mL], versus 0,05 [intervalo interquartil, 0,01-0,18ng/mL] para o grupo sem RCA. Na análise multivariada, o SL-PLVD foi o único preditor independente de RCA com razão de chance de 1,70 (intervalo de confiança de 95% = 1,17- 2,48); p = 0,006. Concluiu-se que o SL-PLVD derivado da EST foi marcador sensível e específico de RCA. Estes dados sugerem que esta técnica é adequada para detectar alterações na deformação ventricular direita durante um episódio de RCA / Rejection is a major complication after heart transplantation (HT). Right ventricular endomyocardial biopsy (EMB) remains the gold standard for diagnosis of rejection. There is a need for non-invasive alternatives that allow for a safe and early diagnosis of cardiac rejection prior to the onset of clinical symptoms. Speckle tracking echocardiography (STE) enables the analysis of left and right ventricular contraction dynamics, thus allowing for the early detection of myocardial dysfunction. The aims of the present study were to compare ventricular dynamics parameters obtained by STE in heart transplanted patients and control individuals. We also aimed to assess the value of STE, troponin I, and brain natriuretic peptide (BNP) serum levels as noninvasive tools for the early detection of significant acute cellular rejection (ACR) after HT. From January 2014 to November 2017, we prospectively studied 49 transplanted patients with normal left and right ventricular systolic function, who had undergone EMB for either surveillance or clinical suspicion of rejection. ACR was defined as >= 2R graded according to the revised International Society for Heart and Lung Transplantation by EMB. A total of 49 age- and sex-matched healthy individuals formed the control group. All studied patients underwent conventional echocardiography with the analysis of STE. The transplanted group also had their serum troponin I and BNP levels measured hours before undergoing EMB. Classic echocardiographic parameters left ventricular global longitudinal strain (LV-GLS), radial and circumferential strain and strain rate, and right ventricular free wall longitudinal strain (RV-FWLS) were analyzed. The 49 heart transplanted patients (mean age 45.2 ± 11.5 years, 28 men) underwent 66 biopsies, 6-12 months after HT. ACR was detected in 17 (26%) and no ACR in 49 (74%) of biopsies. LV-GLS, circumferential and radial strain and strain rate and RVFWLS values were significantly lower in heart transplant group than in control group. RV-FWLS was lower (absolute value) in the group with ACR than in the group without ACR (-18.28 ± 4.8% versus -22.11 ± 2.9%; p < 0.001). A RVFWLS < 17.57% had 71% sensitivity, 90% specificity, 75% positive predictive value, 84% accuracy, and 88% negative predictive value for detection of ACR. The troponin I level was significantly higher in patients with ACR (p < 0.01). The median value for this group was 0.19 [interquartile interval 0.09-1.31 ng/mL] vs. 0.05 [interquartile interval 0.01-0.18 ng/mL] for the group without ACR. In the multivariate analysis, RV-FWLS was the only independent predictor of ACR, with odds ratio of 1.70 (CI 95% = 1.17-2.48); p=0.006. We concluded that STE-derived RV-FWLS was a sensitive and specific marker of ACR. Our data suggest that this is a suitable technique for noninvasive detection of changes in the right ventricular deformation during an episode of ACR

Para o diagnóstico de rejeição celular foi utilizado o Doppler tecidual para o estudo da movimentação miocárdica de 14 regiões de 54 pacientes transplantados que realizaram 129 biópsias endomiocárdicas do ventrículo direito. Foram comparadas as médias e desvios padrão das velocidades de movimentação entre os pacientes com e sem rejeição celular = 3A.Verificou-se que a velocidade de movimentação diastólica tardia do anel lateral do ventrículo esquerdo quando menor do que 5,6 cm/s, isoladamente, apresentou sensibilidade de 73,8% e na análise multivariada encontrou-se modelo matemático com um conjunto de seis parâmetros, com sensibilidade de 88,2% para o diagnóstico de rejeição celular = 3A / For the diagnosis of cellular acute cardiac allograft rejection pulsed tissue Doppler imaging was used for measuring myocardial wall motion of 14 sites of 54 transplant patients who underwent 129 endomyocardial biopsies.The mean and standard deviation of myocardial velocities among patients displaying or not acute rejection = 3A were compared.It was observed that the analysis of late diastolic mitral annular motion at lateral site on its own showed a sensibility of 73.8% (5.6 cm/s was the best cut off value) and in the multivariate analysis a mathematical model was found with a set of 6 parameters with a sensibility of 88.2% for the diagnosis of cellular rejection = 3A

Para o diagnóstico de rejeição celular foi utilizado o Doppler tecidual para o estudo da movimentação miocárdica de 14 regiões de 54 pacientes transplantados que realizaram 129 biópsias endomiocárdicas do ventrículo direito. Foram comparadas as médias e desvios padrão das velocidades de movimentação entre os pacientes com e sem rejeição celular = 3A.Verificou-se que a velocidade de movimentação diastólica tardia do anel lateral do ventrículo esquerdo quando menor do que 5,6 cm/s, isoladamente, apresentou sensibilidade de 73,8% e na análise multivariada encontrou-se modelo matemático com um conjunto de seis parâmetros, com sensibilidade de 88,2% para o diagnóstico de rejeição celular = 3A / For the diagnosis of cellular acute cardiac allograft rejection pulsed tissue Doppler imaging was used for measuring myocardial wall motion of 14 sites of 54 transplant patients who underwent 129 endomyocardial biopsies.The mean and standard deviation of myocardial velocities among patients displaying or not acute rejection = 3A were compared.It was observed that the analysis of late diastolic mitral annular motion at lateral site on its own showed a sensibility of 73.8% (5.6 cm/s was the best cut off value) and in the multivariate analysis a mathematical model was found with a set of 6 parameters with a sensibility of 88.2% for the diagnosis of cellular rejection = 3A

A gordura epicárdica é biologicamente ativa e sua espessura nos obesos é aumentada. A repercussão da gordura epicárdica sobre o remodelamento cardíaco ainda não está completamente elucidada. No presente estudo, foi avaliada a gordura epicárdica e sua influência no remodelamento cardíaco de obesos mórbidos, antes e após a cirurgia bariátrica. Métodos: No Hospital das Clínicas da Universidade de São Paulo, foram recrutados de forma prospectiva 20 obesos mórbidos sem outras comorbidades e 20 controles. Os participantes realizaram avaliação clínica e laboratorial, medida da duração da onda P no ECG e ecocardiograma transtorácico. O grupo de obesos repetiu essa avaliação 12 meses após a cirurgia bariátrica. A medida da gordura epicárdica foi feita pelo ecocardiograma. Para comparar as variáveis contínuas, foram utilizados os testes t de Student (não pareado e pareado), de Mann-Whitney ou de Wilcoxson. Para definir correlação entre as variáveis lineares, foi utilizado o coeficiente de correlação de Pearson. Para definir a associação entre variáveis categóricas, foi usado o teste exato de Fisher. Para avaliar a associação entre variáveis dependentes e independentes, foi realizada uma análise de regressão múltipla. Os dados foram examinados no software R. Um valor de p abaixo de 0,05 foi considerado significativo. Resultados: No préoperatório, foram observados níveis elevados de proteína C reativa, uma maior duração da onda P, da massa ventricular e do diâmetro do átrio esquerdo nos obesos em relação aos controles (p < 0,05). Uma menor fração de ejeção do ventrículo esquerdo foi observada no grupo de obesos (p < 0,05). A gordura epicárdica foi maior nos obesos (p < 0,01). Uma correlação positiva foi encontrada entre a gordura epicárdica, a duração da onda P (r=0,70; p < 0,01), o diâmetro do átrio esquerdo (r=0,67; p < 0,01) e a massa ventricular (r=0,58; p < 0,01). Uma correlação inversa foi observada entre a gordura epicárdica e a fração de ejeção do ventrículo esquerdo (r=- 0,52; p < 0,01). Na análise de regressão múltipla, a gordura epicárdica permaneceu correlacionada com a duração da onda P, o diâmetro do átrio esquerdo e a fração de ejeção do ventrículo esquerdo (p < 0,05). Em 60% dos obesos, foi identificada alguma alteração na geometria ventricular. Uma associação entre a espessura da gordura epicárdica maior ou igual a 3,7 mm e a presença de remodelamento ventricular geométrico foi demonstrada (p=0,03). No pós-operatório, observou-se uma redução do índice de massa corporal, da proteína C reativa e da gordura epicárdica (p < 0,01). Uma redução da duração da onda P e um aumento da fração de ejeção do ventrículo esquerdo também foram observados (p < 0,01). Na análise de regressão múltipla esses achados permaneceram correlacionados à redução da gordura epicárdica (p < 0,05), independente da variação do índice de massa corporal e da proteína C reativa. Conclusões: Em obesos mórbidos sem outras comorbidades, a gordura epicárdica foi associada a um aumento da duração da onda P, do diâmetro do átrio esquerdo e da massa ventricular, além de uma menor fração de ejeção do ventrículo esquerdo. A espessura da gordura epicárdica igual ou acima de 3,7 mm foi associada a alterações do remodelamento ventricular. A redução da gordura epicárdica após a cirurgia bariátrica foi associada com a redução da duração da onda P e o aumento da fração de ejeção do ventrículo esquerdo, independente da variação do índice de massa corporal e da proteína C reativa / Epicardial fat is biologically active and its thickness is increased in obese subjects. The effects of epicardial fat on cardiac remodeling are still not fully understood. In the present study we evaluated epicardial fat and its influence on cardiac remodeling of morbidly obese, before and after bariatric surgery. Methods: We prospectively recruited 20 morbid obese subjects without other comorbidities and 20 control subjects at Hospital das Clínicas, Universidade de São Paulo. Participants underwent clinical and laboratory assessment, measure of P-wave duration on ECG and transthoracic echocardiogram. The obese group repeated this evaluation 12 months after the bariatric surgery. To compare continuous variables, we used t Student test (paired and nonpaired), Mann-Whitney and Wilcoxson tests. To define the correlation between linear variables we used Pearson correlation coefficient. To define the association between categorical variables we used Fisher exact test. A multiple regression analysis was performed to assess the association between dependent and independent variables. Data were analyzed by software R. A p value below 0.05 was considered statistically significant. Results: Preoperatively, we observed high levels of C-reactive protein, longer P-wave duration, larger ventricular mass and left atrial diameter in obese subjects compared to the controls (p < 0.05). Lower left ventricle ejection fraction was observed in the obese group (p < 0.05). Epicardial fat was higher among obese subjects (p < 0.01). A positive correlation was found between epicardial fat and P-wave duration (r=0.70; p < 0.01), left atrial diameter (r=0.67; p < 0.01), and ventricular mass (r=0.58; p < 0.01). An inverse correlation was observed between epicardial fat and left ventricle ejection fraction (r=-0.52; p < 0.01). In the multiple regression analysis, epicardial fat remained correlated with P-wave duration, left atrial diameter and left ventricle ejection fraction (p < 0.05). In 60% of the obese subjects, there was some abnormality in ventricular geometry. We showed association between thickness of epicardial fat equal to or higher than 3.7 mm and presence of geometric ventricular remodeling (p=0.03). Postoperatively, we observed reduction in body mass index, C-reactive protein and epicardial fat (p < 0.01). Reduction in P-wave duration and an increase in left ventricle ejection fraction were also observed (p < 0.01). In the multiple regression analysis, these findings were correlated with reduction in epicardial fat (p < 0.05), regardless of the variation in body mass index and C-reactive protein. Conclusion: In morbid obese subjects without other comorbidities, epicardial fat was associated with increase in P-wave duration, left atrial diameter and ventricular mass, in addition to smaller left ventricle ejection fraction. Epicardial fat thickness equal to or greater than 3.7 mm was associated with abnormalities in ventricular remodeling. Reduction of epicardial fat after bariatric surgery was associated with reduction of P-wave duration and increase in left ventricle ejection fraction, regardless of the variation in body mass index and C-reactive protein