ABSTRACT BACKGROUND: Serological tests are practical, with low cost, but no noninvasive tests are available for diagnosing Helicobacter pylori (H. pylori) infection in Brazil. The aim here was to develop and validate enzyme-linked immunosorbent assay (ELISA) serological tests to detect anti-H. pylori immunoglobulin G antibodies, based on cultured strains from Brazilian patients. DESIGN AND SETTING: Cross-sectional, diagnostic accuracy study comparing a locally developed and validated ELISA and invasive tests among dyspeptic patients at two public hospitals in São Paulo, Brazil. METHODS: An ELISA test was prepared using whole-cell antigen from 56 strains. After genotypic characterization, it was standardized and optical density (OD) cutoffs were determined based on the serum antibody response of 100 H. pylori-negative samples, compared with 82 H. pylori-positive samples. Validation was performed on 174 symptomatic patients. RESULTS: The optimal OD cutoffs established (for monoclonal and polyclonal tests, respectively) were 0.167 and 0.164; overall ELISA sensitivity: 84.3%, 78.9%; specificity: 88.6%, 90.6%; positive predictive value (PPV): 75.4%, 80%; negative predictive value (NPV): 93.1%, 81.8%; accuracy: 87.3%, 86.2%; child and adolescent ELISA sensitivity: 74.2%, 81.8%; specificity: 90.8%, 86.7%; PPV: 66.6%, 84.3%; NPV: 95.8%, 84.8%; accuracy: 88.5%, 84.6; adult ELISA sensitivity: 84.4%, 75%; specificity: 86.9%, 93%; PPV: 81.8%, 78.3%; NPV: 88.9%, 91.8%; accuracy: 85.9%, 88.5%. CONCLUSION: The polyclonal serological test developed using local strains presented better diagnostic performance among children and adolescents, while the monoclonal test was better among adults. The results from both tests suggest that these in-house serological tests could be used to detect anti-H. pylori antibodies in our population, for screening purposes.

CONTEXTO: A ressecção endoscópica mucosa é uma modalidade estabelecida para a excisão de lesões sésseis no trato gastrointestinal. A injeção de fluídos na submucosa cria uma coxim que pode prevenir lesão térmica e perfuração. OBJETIVO: Este estudo cego investiga o desempenho de três diferentes soluções para criar um coxim fluído submucoso no estômago suíno. MÉTODOS: Três soluções foram injetadas no estômago de nove porcos BR1: soro fisiológico, carboximetilcelulose 0.5% e hidroxipropil metilcelulose 0.25%. Em cada porco, injeções submucosas com 6 mL por solução-teste foram realizadas. Uma gota de azul de metileno foi adicionada a cada injeção para melhor visualização. O tempo de desaparecimento de cada coxim foi registrado. RESULTADOS: O tempo mediano total do coxim submucoso visível foi de 37 minutos (faixa 12-60 min) para hidroxipropil metilcelulose, 31 minutos para carboximetilcelulose (faixa 10-43 min) e 19 minutos para soro fisiológico (faixa 8-37 min). Não houve significância estatística entre soro fisiológico e carboximetilcelulose (P = 0.146), assim como entre carboximetilcelulose e hidroxipropil metilcelulose (P = 0.119), mas a duração mediana de hidroxipropil metilcelulose foi significativamente maior que a do soro fisiológico (P = 0.039). CONCLUSÃO: A duração do coxim submucoso com hidroxipropil metilcelulose é maior em comparação com o do soro fisiológico. O tempo mediano da carboximetilcelulose não foi maior que do soro fisiológico. A hidroxipropil metilcelulose, na concentração de 0.25%, pode ser uma alternativa durável para injeção submucosa.

CONTEXT: Endoscopic mucosal resection is an established modality for excision of sessile lesions in the gastrointestinal tract. Submucosal fluid injection creates a cushion and may prevent thermal injury and perforation. OBJECTIVES: This blind study investigated the performance of three different solutions to create submucosal fluid cushions in porcine stomach. METHODS: Three solutions were injected in the stomach of nine pigs BR1: normal saline solution, carboxymethylcellulose 0.5% and hydroxypropyl methylcellulose 0.25%. In each pig, submucosal injections with 6 mL per test-solution were performed. One drop of methylene blue was added to all injections for better visualization. The time for the bleb to disappear was recorded. RESULTS: The overall median time of visible submucosal cushion was 37 minutes (range 12-60 min) for hydroxypropyl methylcellulose, 31 minutes for carboxymethylcellulose (range 10-43 min) and 19 minutes for normal saline solution (range 8-37 min). There was no statistically significant difference neither between normal saline solution and carboxymethylcellulose (P = 0.146) nor carboxymethylcellulose and hydroxypropyl methylcellulose (P = 0.119) but the median duration of hydroxypropyl methylcellulose was significantly longer than normal saline solution (P = 0.039). CONCLUSIONS: The length of hydroxypropyl methylcellulose submucosal fluid cushion is longer in comparison with normal saline solution. The median time for carboxymethylcellulose was not longer than normal saline solution. Hydroxypropyl methylcellulose, in the concentration of 0.25%, may be a durable alternative for submucosal injection.