Thesis (MScMedSc)--University of Stellenbosch, 2011. / ENGLISH ABSTRACT: INTRODUCTION: Glycogen synthase kinase-3 (GSK-3) is a serine-threonine protein kinase that was first discovered as a regulator of glycogen synthase thus playing a role in glycogen synthesis (Embi et al. 1980). GSK-3 has also been shown to down regulate the expression of SERCA-2a (a calcium ATPase pump) thus playing a role in myocardial contractility (Michael et al. 2004). However, SERCA-2a activity is regulated by phospholamban (PLM) and sarcolipin (SLN) (Asahi et al. 2003). GSK-3 is constitutively active in cells and can be acutely inactivated by insulin through phosphorylation by PKB/Akt. However, GSK-3 is known to phosphorylate and inhibit IRS-1 protein, thus disrupting insulin signaling (Eldar-Finkelman et al. 1996). In addition, abnormally high activities of GSK-3 protein has been implicated in several pathological disorders which include type 2 diabetes, neuron degenerative and affective disorders (Eldar-Finkelman et al 2009). This led to the development of new generations of inhibitors with specific clinical implications to treat these diseases (Martinez 2008). GSK-3 inhibition has been shown to improve insulin and blood glucose levels and to be cardioprotective during ischemia/reperfusion (Nikoulina et al. 2002; Kumar et al. 2007). AIMS: To determine whether myocardial GSK-3 protein and its substrate proteins are dysregulated in obesity and insulin resistance, and whether a specific GSK-3 inhibitor can prevent or reverse the cardiovascular pathology found in obese and insulin resistant animals. OBJECTIVES: To correlate the alterations in expression and activation of GSK-3 protein in a well characterised rat model of obesity coupled to insulin resistance with: i) myocardial contractile dysfunction and an inability of hearts to withstand ischemia/reperfusion, ii) the activation and expression of phospholamban and SERCA-2a in the sarcoplasmic reticulum, iii) the activation of intermediates (IRS-1, IRS-2 and PKB/Akt) that lie upstream in the activation pathway of GSK-3 and iv) to determine the effects of inhibition of GSK-3 on the abovementioned parameters. METHODS: Age and weight matched male Wistar rats (controls and diet induced obese (DIO) animals) were used in the present study. Controls were fed normal rat chow, while DIOs were fed a rat chow diet supplemented with sucrose and condensed milk, for 8 or 16 weeks. Half of each group of animals were treated with the GSK-3 inhibitor for 4 weeks (from 12 to 16 weeks). After the feeding and treatment period, animals were weighed, sacrificed, hearts removed and freeze clamped immediately or perfused with Krebs-Henseleit buffer and subjected to low flow ischemia (25 min) followed by 30 min reperfusion. Biometric (body weight, intraperitoneal fat, ventricular weight and tibia length) and biochemical (fasting blood glucose and insulin levels) parameters were determined. Expression of GSK-3, PKB/Akt, IRS-1, IRS-2, SERCA-2a and Phospholamban were determined by Western blotting. Ca2+ ATPase activity was determined spectrophotometrically. RESULTS: At both 8 and 16 weeks DIO animals were significantly bigger than control animals and this was associated with increased intraperitoneal fat in DIOs. In DIO animals: IRS-1 was downregulated at 8 weeks and both IRS-1 and IRS-2 as well as PKB/Akt at 16 weeks. There was an increased tendency of GSK-3 expression at both 8 and 16 weeks in DIO animals while SERCA-2a was severely downregulated from 8 weeks onwards and associated with lower Ca2+-ATPase activity. PLM expression was upregulated but its phosphorylation was attenuated. At 16 weeks, baseline heart rate (225 vs 275 in control, P<0.0001, n=6) and rate pressure product (21000 vs 30000 in control, P=0.019, n=6) were significantly lower in hearts from DIO animals. Functional recovery was unchanged but the time to ischemic contracture development was increased (11.6±0.4 control vs 16.2±0.5 min DIO, P<0.01, n=6). Treatment had no effect on total GSK-3 expression. However, GSK-3 phosphorylation was significantly increased in treated controls, while there was no significant difference in DIO animals. However, there was a tendency for an increased GSK-3 phosphorylation in treated DIO animals. GSK-3 inhibitor, improved hypertrophy in DIO animals, while it led to its development in control animals. GSK-3 inhibitor improved IRS-2 expression in both control and DIO animals while it had no effect on IRS-1 and SERCA-2a expression and activity. However, GSK-3 inhibition increased PKB/Akt and phospholamban phosphorylation in DIO animals. CONCLUSION: These findings show that high calorie diet as well as imbalance between energy intake and expenditure lead to the development of obesity and insulin resistance in male Wistar rats. We showed that GSK-3 and its substrate proteins are dysregulated in obesity and insulin resistance. The reduced SERCA-2a expression at baseline may have a negative impact on cardiac function. By treating the animals with GSK-3 inhibitor, we showed that GSK-3 protein may not be responsible for changes seen at baseline. The decreased IRS-1 and SERCA-2a expression may have been caused by a different mechanism other than the actions of GSK-3. However, according to this study, GSK-3 may play a role in regulation of IRS-2 expression but not in IRS-1. Increased PKB/Akt phosphorylation may contribute to the GSK-3 inhibition. In addition, GSK-3 inhibition may reverse cardiac hypertrophy in DIO animals, thus acting as a negative regulator of hypertrophy. / AFRIKAANSE OPSOMMING: Inleiding: Glikogeen sintase kinase-3 (GSK-3), 'n serien/threonien proteïen kinase, is oorspronklik ontdek as 'n rolspeler in glikogeen sintese, aangesien dit 'n reguleerder van glikogeen sintase is (Embi et al.1980). Intussen is dit ook bevind dat GSK-3 die uitdrukking van SERCA-2a ('n kalsium ATPase pomp) kan afreguleer en dus sodoende 'n rol speel in miokardiale kontraktiliteit (Michael et al. 2004). Die aktiwiteit van SERCA-2a kan egter ook gereguleer word deur fosfolamban (PLM) en sarkolipin (Asahi et al. 2003). GSK-3 is deurgaans aktief, maar kan tydelik geïnaktiveer word onder kondisies van insulien stimulasie deur PKB/Akt gemedieerde fosforilering. Aan die ander kant is dit bekend dat GSK-3 die IRS-1 proteïen kan fosforileer om dus sodoende insulien sein-transduksie af te reguleer (Eldar-Finkelman et al. 1996). Daarmee saam is abnormaal hoë vlakke van GSK-3 aktiwiteit geassosieer met verskeie patologiese versteurings, insluitend tipe 2 diabetes, neuron degeneratiewe en affektiewe versteurings (Eldar-Finkelman et al. 2009). Daar is dus nuwe generasies GSK-3 inhibitore ontwikkel met die kliniese potensiaal om hierdie patologieë te behandel (Martinez 2008). Dit is al bevind dat GSK-3 inhibisie geassosieer kan word met beide die normalisering van plasma insulien- en glukose vlakke, asook kardiobeskerming in die konteks van iskemie/herperfusie (Nikoulina et al. 2002; Kumar et al. 2007). Doelwitte: Om te bepaal of GSK-3 proteïen en sy substraat proteïene gedisreguleer is onder kondisies van obesiteit en insulien weerstandigheid, asook om vas te stel of 'n spesifieke GSK-3 inhibitor die kardiovaskulêre patologie wat gevind word in obese en insulien weerstandige diere kan verhoed of omkeer. Mikpunte: Om veranderinge in uitdrukking en aktiwiteit van GSK-3 proteïen in 'n goed gekarakteriseerde rotmodel van obesiteit, gekoppel aan insulien weerstandigheid, te korreleer met die volgende: i) miokardiale kontraktiele disfunksie en onvermoë om kardiale iskemie/herperfusie besering te weerstaan, ii) aktivering en uitdrukking van PLM en SERCA-2a in die sarkoplasmiese retikulum, iii) die aktivering van intermediêres wat proksimaal geleë is in die insulienseintransduksiepad van GSK-3 (IRS-1, IRS-2 en PKB/Akt) en iv) om die effek van behandeling met 'n spesifieke inhibitor van GSK-3 op die bogenoemde punte te bepaal. Metodes: Ouderdoms- en gewigsgepaarde manlike Wistar rotte (kontrole en dieet geïnduseerde obees (DIO) diere) is in die studie gebruik. Kontrole diere was normale rotkos gevoer, terwyl die DIO diere op 'n dieet van rotkos aangevul met sukrose en kondensmelk geplaas is vir 'n periode van 8 of 16 weke. Helfte van die diere van elke groep is behandel met die GSK-3 inhibitor vir 4 weke (vanaf week 12 tot 16). Na afloop van die voer- en behandelingsperiode is die diere geweeg, doodgemaak en die harte verwyder om dan of onmiddelik gevriesklamp te word, of retrograad geperfuseer te word met Krebs-Hensleit buffer. Ex vivo geperfuseerde harte is dan blootgestel aan 25 minute lae vloei iskemie gevolg deur 30 minute herperfusie. Biometriese (liggaamsgewig, intraperitoneale vet, ventrikulêre gewig en tibia lengte) en biochemiese (vastende bloedglukose en -insulien vlakke) parameters is telkens bepaal. Western klad tegnieke is gebruik om die uitdrukking en fosforilering van GSK-3, PKB/Akt, IRS-1, IRS-2, SERCA-2a en PLM te bepaal. Ca2+-ATPase aktiwiteit is spektrofotometries bepaal. Resultate: Na beide 8 en 16 weke was die DIO diere beduidend swaarder as die kontrole diere. Hierdie gewigstoename was geassosieer met meer intraperitoneale vet in die DIO diere. Verder, in die DIO diere was IRS-1 afgereguleer na 8 weke, terwyl beide IRS-1 en IRS-2 asook PKB/Akt afgereguleer was na 16 weke. GSK-3 uitdrukking het 'n neiging getoon om toe te neem na beide 8 en 16 weke in die DIO diere, terwyl SERCA-2a beduidend afgereguleer was reeds vanaf 8 weke, geassosieer met laer Ca2+-ATPase aktiwiteit. PLM uitdrukking het toegeneem en die fosforilering daarvan was verlaag. Op 16 weke was die basale harttempo (225 vs 275 in die kontrole groep, P<0.0001, n=6) en tempo druk produk (21000 vs 30000 in die kontrole groep, P=0.019, n=6) betekenisvol laer in die DIO diere. Funksionele herstel het onveranderd gebly, alhoewel die tyd tot iskemiese kontraktuur toegeneem het in die DIO groep (kontrole: 11.6±0.4 min vs DIO: 16.2±0.5 min, P<0.01, n=6). Toediening van die inhibitor het geen effek op totale GSK-3 uitdrukking gehad nie. Fosforilering van GSK-3 was egter wel beduidend verhoog in die behandelde kontrole diere, terwyl daar geen verskille in die DIO groep was nie. Die fosforilering van GSK-3 het wel geneig na 'n toename in die behandelde DIO diere. Die GSK-3 inhibitor het kontrasterende effekte op hipertrofie gehad: dit het dit omgekeer in die DIO groep, maar veroorsaak in die kontrole diere. Daarmee saam het die inhibitor die uitdrukking van IRS-2 in beide DIO en kontrole diere gestimuleer, maar geen effek op IRS-1 en SERCA-2a uitdrukking en aktiwiteit gehad nie. GSK-3 inhibisie het wel PKB/Akt en PLM fosforilering in die DIO diere verhoog. Gevolgtrekking: Hierdie bevindinge toon dat 'n hoë kalorie dieet, tesame met 'n wanbalans tussen energie inname en verbruiking, lei tot die ontwikkeling van obesiteit en insulien weerstand in manlike Wistar rotte. Die studie het ook getoon dat GSK-3 en sy substraat proteïene wel gedisreguleer is in obesiteit en insulien weerstandigheid. Die verlaagde basale uitdrukking van SERCA-2a mag dalk 'n negatiewe impak hê op kardiale funksie. Behandeling van die diere met 'n GSK-3 inhibitor het getoon dat GSK-3 moontlik nie verantwoordelik is vir die basislyn veranderinge nie. Die afname in IRS-1 en SERCA-2a uitdrukking kan moontlik toegeskryf word aan ander meganismes buiten die effekte van GSK-3. Hierdie studie toon wel dat GSK-3 moontlik 'n rol speel in die regulering van die uitdrukking van IRS-2, maar nie IRS-1 nie. Verhoogde PKB/Akt fosforilering mag dalk bydra tot die inhibisie van GSK-3. Daarmee saam blyk dit dat GSK-3 inhibisie hipertrofie kan omkeer in DIO diere, om dan sodoende op te tree as 'n negatiewe reguleerder van hipertrofie, maar in normale kontrole diere, hipertrofie in die hand werk. / South African Medical Research Council / University of Stellenbosch, Dept. of medical Physiology

INTRODUÇÃO: A endomiocardiofibrose (EMF) é uma miocardiopatia de padrão restritivo de etiologia desconhecida, prevalente em regiões tropicais. Caracteriza-se por espessamento fibroso do endocárdio e miocárdio subjacente, comprometendo ponta e via de entrada de um ou de ambos os ventrículos. Sua etiopatogenia é pouco conhecida e muitos autores a associam à doença infecciosa cardíaca ou sistêmica, à eosinofilia prévia e/ou à carência nutricional. O prognóstico em geral é grave e a ressecção cirúrgica da lesão é indicada aos pacientes com insuficiência cardíaca refratária a tratamento clínico, em classe funcional III ou IV (NYHA). Os estudos anatomopatológicos até o momento foram realizados em material de autópsia ou de biópsia endomiocárdica. OBJETIVOS: Este estudo retrospectivo se propõe a (1) descrever os três aspectos morfológicos da lesão endocárdica (fibrose, infiltrado inflamatório e vasos) utilizando microscopia óptica comum e técnicas imunoistoquímicas, assim como correlacioná-los aos dados clínicos, laboratoriais e de imagem dos pacientes (2) comparar os aspectos morfológicos de espécimes de ressecção cirúrgica com os de autópsia a fim de verificar se os primeiros podem ser empregados para diagnóstico histológico da doença; e (3) discutir a patogenia da EMF e realizar pesquisa de agentes infecciosos cardiotrópicos em amostra endomiocárdica incluída em parafina por técnica de biologia molecular. MÉTODOS: Foram utilizadas amostras de ressecção cirúrgica endocárdica incluídas em parafina provenientes de 31 pacientes com diagnóstico clínico e cineangiocardiográfico de EMF, operados no InCor entre 1991 e 2005. As amostras foram coradas por técnicas convencionais (HE, tricômico de Masson, Verhoeff e reticulina) e submetidas a reações imunoistoquímicas para fibras colágenas tipo I, III e IV, para células inflamatórias (CD3, CD20, CD68) e para endotélio de linfáticos (D2-40). Amostras de nove corações de autópsia de pacientes com o mesmo diagnóstico serviram de controle positivo da doença. A pesquisa de agentes infecciosos foi feita por reação de cadeia de polimerase e reação de transcrição reversa de cadeia de polimerase (PCR e RT-PCR) em amostras endomiocárdicas incluídas em parafina, para T. gondii e para vírus cardiotrópicos (enterovirus, adenovirus, influenza A e B, citomegalovirus, parvovirus B19 e herpes simples). Para identificar alterações vasculares intramiocárdicas, procedeu-se à revisão de prontuários clínicos dos pacientes e de 16 cineangiocoronariografias. RESULTADOS: Foi observado intenso espessamento endocárdico ventricular à custa de fibrose hialina superficial escassamente celular, com fibras colágenas tipos I e III, predominando o tipo I sobre o III. O colágeno tipo IV foi identificado na membrana basal de vasos. Na porção profunda da lesão endocárdica observou-se escasso infiltrado inflamatório crônico com macrófagos, linfócitos T e B em menor número. O infiltrado estava distribuído ao redor de vasos proliferados com intensas alterações estruturais na parede e com participação de linfáticos. No miocárdio superficial identificou-se miocardite \"borderline\" (critério de \"Dallas\"). Foram obtidos ácidos nucléicos em quantidade suficiente para a reação de PCR/RT-PCR em 12/36 (33%) amostras. Genomas de agentes infecciosos foram identificados em 6/12 (50%) pacientes. Dois casos foram positivos para enterovirus (EV), dois para citomegalovirus (CMV), um para ambos (CMV e EV) e um para T. gondii. Não foram observadas diferenças histopatológicas entre as amostras cirúrgicas e as de autópsia. Foram detectadas alterações vasculares à cineangiocoronariografia em 9/16 (56%) pacientes. Não houve correlação anatomoclínica definida entre os múltiplos dados comparados. CONCLUSÕES: Os resultados indicam que na EMF ocorre processo inflamatório crônico mantido por rede vascular anômala rica em linfáticos localizada na profundidade da lesão endocárdica. A rede vascular provavelmente contribui para a manutenção da placa fibrótica e deve ser considerada como fator importante na patogenia da doença. O diagnóstico anatomopatológico pode ser feito com segurança em material de ressecção cirúrgica. A análise molecular do endomiocárdio possibilitou a detecção de alta incidência de genomas de agentes infecciosos cardiotrópicos. Seu significado, contudo, permanece controverso. / BACKGROUND: Endomyocardial fibrosis (EMF) is a restrictive cardiomyopathy of unknown etiology prevalent in tropical regions. The disease involves the inflow tract and apex of either one or both ventricles and is characterized by a fibrous thickening of the endocardium and the underlying myocardium. Although its etiology remains unknown, most authors believe it could be related to systemic or heart infection/parasitism, previous blood eosinophilia or malnutrition. Surgical resection of the thickened endocardium is recommended to patients with advanced heart failure of functional class III or IV, New York Heart Association (NYHA). The gross and histological features of the heart have been comprehensively studied in autopsies and endomyocardial biopsies. Studies in surgical samples, however, are still lacking. AIMS: This study was conducted to evaluated: (1) the histomorphological changes of EMF as seen in surgical specimens by means of routine histological and immunohistochemical methods in an attempt to correlate them with clinical symptoms and coronary angiographic features; (2) to compare histological data between surgical and autopsy samples, and (3) to discuss probable pathogenetic mechanisms of the disease, as well as to investigate cardiotropic infective agents by means of molecular analysis of endomyocardial surgical samples. METHODS: We collected all available clinical records and endomyocardial surgical samples from 31 patients with EMF who had been submitted to surgery between 1991 and 2005 at InCor. The diagnosis was based on clinical, hemodynamic and angiocardiographic findings. The surgical samples were fixed in 10% formalin, submitted to standard processing, and stained with H&E, Masson\'s trichrome, reticulin and elastic stains. Immunohistochemical methods were employed to detect collagen fibers type I, III, and IV, inflammatory cells (CD3, CD20, CD68) and lymphatic vessels\' endothelium (D2-40). Nine samples from autopsied hearts of EMF patients were used as a positive control group. Polymerase chain reaction (PCR) and reverse transcription-PCR were used retrospectively to search for genomes of T. gondii and cardiotropic viruses (enterovirus, adenovirus, influenza A e B, cytomegalovirus, parvovirus B19 and herpes simplex) in the surgical material. All clinical and surgical reports were reviewed, including follow-ups and 16 coronary cineangiocardiographies. RESULTS: Ventricular endocardium was thickened by superficial acellular hyaline collagen fibers type I and III. Type-IV collagen fibers were seen only around vessels. Focal chronic inflammatory infiltrate with T-lymphocytes, macrophages and a few B-lymphocytes was seen around blood vessels with a peculiar pattern of vascular changes and numerous lymphatics within the endocardium. The superficial myocardium showed borderline myocarditis (Dallas criteria). RNA and DNA were successfully extracted from 12/36 samples. Infective agents were detected in 6/12 (50%) patients; two of them were positive for cytomegalovirus (CMV), two for enterovirus (EV), one for both (CMV and EV) and one for T. gondii. No histopathological differences between surgical samples and autopsy fragments were observed. Vascular blush or neovascularity was detected in 9 of the 16 coronary cineangiocardiographies reviewed. Clinicopathologic characteristics are associated neither with infective genomes in the endocardium nor with vascular blush. CONCLUSIONS: Results indicate that there is an non especific chronic inflammatory process maintained by an anomalous vascular net rich in lymphatics situated deep within the endocardium. This angiolymphatic web probably contributes to the maintenance of the fibrotic plaque and might be considered an important pathological finding concerning in the pathogenesis of EMF. Histopathological changes as seen in surgical material are diagnostic of EMF. Molecular analysis of the endomyocardium revealed high incidence of cardiotropic infective agents, but their role in the pathogenesis of the disease is still controversial.

INTRODUÇÃO: A endomiocardiofibrose (EMF) é uma miocardiopatia de padrão restritivo de etiologia desconhecida, prevalente em regiões tropicais. Caracteriza-se por espessamento fibroso do endocárdio e miocárdio subjacente, comprometendo ponta e via de entrada de um ou de ambos os ventrículos. Sua etiopatogenia é pouco conhecida e muitos autores a associam à doença infecciosa cardíaca ou sistêmica, à eosinofilia prévia e/ou à carência nutricional. O prognóstico em geral é grave e a ressecção cirúrgica da lesão é indicada aos pacientes com insuficiência cardíaca refratária a tratamento clínico, em classe funcional III ou IV (NYHA). Os estudos anatomopatológicos até o momento foram realizados em material de autópsia ou de biópsia endomiocárdica. OBJETIVOS: Este estudo retrospectivo se propõe a (1) descrever os três aspectos morfológicos da lesão endocárdica (fibrose, infiltrado inflamatório e vasos) utilizando microscopia óptica comum e técnicas imunoistoquímicas, assim como correlacioná-los aos dados clínicos, laboratoriais e de imagem dos pacientes (2) comparar os aspectos morfológicos de espécimes de ressecção cirúrgica com os de autópsia a fim de verificar se os primeiros podem ser empregados para diagnóstico histológico da doença; e (3) discutir a patogenia da EMF e realizar pesquisa de agentes infecciosos cardiotrópicos em amostra endomiocárdica incluída em parafina por técnica de biologia molecular. MÉTODOS: Foram utilizadas amostras de ressecção cirúrgica endocárdica incluídas em parafina provenientes de 31 pacientes com diagnóstico clínico e cineangiocardiográfico de EMF, operados no InCor entre 1991 e 2005. As amostras foram coradas por técnicas convencionais (HE, tricômico de Masson, Verhoeff e reticulina) e submetidas a reações imunoistoquímicas para fibras colágenas tipo I, III e IV, para células inflamatórias (CD3, CD20, CD68) e para endotélio de linfáticos (D2-40). Amostras de nove corações de autópsia de pacientes com o mesmo diagnóstico serviram de controle positivo da doença. A pesquisa de agentes infecciosos foi feita por reação de cadeia de polimerase e reação de transcrição reversa de cadeia de polimerase (PCR e RT-PCR) em amostras endomiocárdicas incluídas em parafina, para T. gondii e para vírus cardiotrópicos (enterovirus, adenovirus, influenza A e B, citomegalovirus, parvovirus B19 e herpes simples). Para identificar alterações vasculares intramiocárdicas, procedeu-se à revisão de prontuários clínicos dos pacientes e de 16 cineangiocoronariografias. RESULTADOS: Foi observado intenso espessamento endocárdico ventricular à custa de fibrose hialina superficial escassamente celular, com fibras colágenas tipos I e III, predominando o tipo I sobre o III. O colágeno tipo IV foi identificado na membrana basal de vasos. Na porção profunda da lesão endocárdica observou-se escasso infiltrado inflamatório crônico com macrófagos, linfócitos T e B em menor número. O infiltrado estava distribuído ao redor de vasos proliferados com intensas alterações estruturais na parede e com participação de linfáticos. No miocárdio superficial identificou-se miocardite \"borderline\" (critério de \"Dallas\"). Foram obtidos ácidos nucléicos em quantidade suficiente para a reação de PCR/RT-PCR em 12/36 (33%) amostras. Genomas de agentes infecciosos foram identificados em 6/12 (50%) pacientes. Dois casos foram positivos para enterovirus (EV), dois para citomegalovirus (CMV), um para ambos (CMV e EV) e um para T. gondii. Não foram observadas diferenças histopatológicas entre as amostras cirúrgicas e as de autópsia. Foram detectadas alterações vasculares à cineangiocoronariografia em 9/16 (56%) pacientes. Não houve correlação anatomoclínica definida entre os múltiplos dados comparados. CONCLUSÕES: Os resultados indicam que na EMF ocorre processo inflamatório crônico mantido por rede vascular anômala rica em linfáticos localizada na profundidade da lesão endocárdica. A rede vascular provavelmente contribui para a manutenção da placa fibrótica e deve ser considerada como fator importante na patogenia da doença. O diagnóstico anatomopatológico pode ser feito com segurança em material de ressecção cirúrgica. A análise molecular do endomiocárdio possibilitou a detecção de alta incidência de genomas de agentes infecciosos cardiotrópicos. Seu significado, contudo, permanece controverso. / BACKGROUND: Endomyocardial fibrosis (EMF) is a restrictive cardiomyopathy of unknown etiology prevalent in tropical regions. The disease involves the inflow tract and apex of either one or both ventricles and is characterized by a fibrous thickening of the endocardium and the underlying myocardium. Although its etiology remains unknown, most authors believe it could be related to systemic or heart infection/parasitism, previous blood eosinophilia or malnutrition. Surgical resection of the thickened endocardium is recommended to patients with advanced heart failure of functional class III or IV, New York Heart Association (NYHA). The gross and histological features of the heart have been comprehensively studied in autopsies and endomyocardial biopsies. Studies in surgical samples, however, are still lacking. AIMS: This study was conducted to evaluated: (1) the histomorphological changes of EMF as seen in surgical specimens by means of routine histological and immunohistochemical methods in an attempt to correlate them with clinical symptoms and coronary angiographic features; (2) to compare histological data between surgical and autopsy samples, and (3) to discuss probable pathogenetic mechanisms of the disease, as well as to investigate cardiotropic infective agents by means of molecular analysis of endomyocardial surgical samples. METHODS: We collected all available clinical records and endomyocardial surgical samples from 31 patients with EMF who had been submitted to surgery between 1991 and 2005 at InCor. The diagnosis was based on clinical, hemodynamic and angiocardiographic findings. The surgical samples were fixed in 10% formalin, submitted to standard processing, and stained with H&E, Masson\'s trichrome, reticulin and elastic stains. Immunohistochemical methods were employed to detect collagen fibers type I, III, and IV, inflammatory cells (CD3, CD20, CD68) and lymphatic vessels\' endothelium (D2-40). Nine samples from autopsied hearts of EMF patients were used as a positive control group. Polymerase chain reaction (PCR) and reverse transcription-PCR were used retrospectively to search for genomes of T. gondii and cardiotropic viruses (enterovirus, adenovirus, influenza A e B, cytomegalovirus, parvovirus B19 and herpes simplex) in the surgical material. All clinical and surgical reports were reviewed, including follow-ups and 16 coronary cineangiocardiographies. RESULTS: Ventricular endocardium was thickened by superficial acellular hyaline collagen fibers type I and III. Type-IV collagen fibers were seen only around vessels. Focal chronic inflammatory infiltrate with T-lymphocytes, macrophages and a few B-lymphocytes was seen around blood vessels with a peculiar pattern of vascular changes and numerous lymphatics within the endocardium. The superficial myocardium showed borderline myocarditis (Dallas criteria). RNA and DNA were successfully extracted from 12/36 samples. Infective agents were detected in 6/12 (50%) patients; two of them were positive for cytomegalovirus (CMV), two for enterovirus (EV), one for both (CMV and EV) and one for T. gondii. No histopathological differences between surgical samples and autopsy fragments were observed. Vascular blush or neovascularity was detected in 9 of the 16 coronary cineangiocardiographies reviewed. Clinicopathologic characteristics are associated neither with infective genomes in the endocardium nor with vascular blush. CONCLUSIONS: Results indicate that there is an non especific chronic inflammatory process maintained by an anomalous vascular net rich in lymphatics situated deep within the endocardium. This angiolymphatic web probably contributes to the maintenance of the fibrotic plaque and might be considered an important pathological finding concerning in the pathogenesis of EMF. Histopathological changes as seen in surgical material are diagnostic of EMF. Molecular analysis of the endomyocardium revealed high incidence of cardiotropic infective agents, but their role in the pathogenesis of the disease is still controversial.