Análise do Escore Poligênico de Risco como preditor de resposta à risperidona: aplicação em uma coorte de primeiro episódio psicótico
Arquivos
Data
2022-08-05
Autores
Luiz, Matheus Rodrigues [UNIFESP]
Orientadores
Belangero, Sintia Iole Nogueira [UNIFESP]
Tipo
Trabalho de conclusão de curso de graduação
Título da Revista
ISSN da Revista
Título de Volume
Resumo
Objetivo: analisar o escore poligênico de risco para esquizofrenia como possível variável preditora da resposta ao tratamento com risperidona, baseando-se em dois diferentes critérios de resposta.
Método: Foram selecionados 99 pacientes em primeiro episódio psicótico, previamente virgens de tratamento com antipsicótico que foram avaliados por meio da escala PANSS (Escala de Sintomas Positivos e Negativos em Esquizofrenia) em dois momentos, no início do tratamento com antipsicóticos e após dois meses de acompanhamento. A partir desse acompanhamento foi possível calcular a taxa de resposta ao tratamento e classificar os pacientes em respondedor e não respondedor baseado em dois critérios de resposta (redução de 30% e 50% da escala PANSS). Os pacientes tiveram seu sangue coletado para realizar a genotipagem, utilizada para calcular o escore poligênico de risco para esquizofrenia que foi analisado como variável preditora do desfecho resposta, junto a outras covariáveis, por meio de uma regressão logística.
Resultado: os modelos de predição mostraram que pacientes com maior escore poligênico de risco para esquizofrenia tem maior chance de não responder ao tratamento, para os dois critérios de resposta. Os melhores modelos de predição, para os dois critérios de resposta, envolveram a covariável tempo de psicose não tratada, mas em ambos os critérios, nenhuma variável foi suficiente para predizer a resposta ao tratamento sozinha e o melhor modelo em ambos os critérios envolvia todas as variáveis. Houve uma tendência à significância estatística em muitos achados, evidenciando a necessidade de se replicar este estudo em um número amostral maior.
Conclusão: pacientes com maior escore poligênico para esquizofrenia demonstraram maior risco para não responderem ao tratamento com antipsicóticos, e o escore poligênico de risco se mostrou um bom preditor de resposta quando atrelado a variáveis clínicas como tempo de psicose não tratada, histórico familiar, idade e sexo biológico.
Objective: Analyze the polygenic risk score for schizophrenia as a possible predictor variable for treatment response with risperidone, based in two different response criteria Methods: We recruited 99 patients in First-Episode of Psychosis, previously antipsychotic-naive, who were evaluated by the PANSS scale in two different moments: in the beginning of the treatment with antipsychotics (baseline) and a follow-up after two months in treatment. This follow-up made possible to calculate de response rate and to classify patients in treatment responders and non-responders (based in two different criteria – 30% and 50% reduction of PANSS score). Patients had their blood collected for genotyping and the genotype date was used to generate the polygenic risk score for schizophrenia, which was analyzed as a predictor for the response outcome, along with the covariables, in a logistic regression. Results: The prediction models show that patients with higher PRS for schizophrenia have higher chance of not responding to treatment, in both response criteria. The best prediction models had the covariable Duration of Untreated Psychosis but, in both criteria, a single variable, alone, was not efficient in predicting the response outcome. Therefore, the best prediction model in both criteria was the model with all the independent variables. Results show a tendency to statistical significance, indicating the need to replicate this study in a bigger sample size. Conclusion: Patients with higher PRS for schizophrenia showed higher risk for not responding to the treatment with antipsychotics. The PRS was a good predictor for response in a model with another clinical variables as predictors, such as duration of untreated psychosis, familiar history, age and biological sex.
Objective: Analyze the polygenic risk score for schizophrenia as a possible predictor variable for treatment response with risperidone, based in two different response criteria Methods: We recruited 99 patients in First-Episode of Psychosis, previously antipsychotic-naive, who were evaluated by the PANSS scale in two different moments: in the beginning of the treatment with antipsychotics (baseline) and a follow-up after two months in treatment. This follow-up made possible to calculate de response rate and to classify patients in treatment responders and non-responders (based in two different criteria – 30% and 50% reduction of PANSS score). Patients had their blood collected for genotyping and the genotype date was used to generate the polygenic risk score for schizophrenia, which was analyzed as a predictor for the response outcome, along with the covariables, in a logistic regression. Results: The prediction models show that patients with higher PRS for schizophrenia have higher chance of not responding to treatment, in both response criteria. The best prediction models had the covariable Duration of Untreated Psychosis but, in both criteria, a single variable, alone, was not efficient in predicting the response outcome. Therefore, the best prediction model in both criteria was the model with all the independent variables. Results show a tendency to statistical significance, indicating the need to replicate this study in a bigger sample size. Conclusion: Patients with higher PRS for schizophrenia showed higher risk for not responding to the treatment with antipsychotics. The PRS was a good predictor for response in a model with another clinical variables as predictors, such as duration of untreated psychosis, familiar history, age and biological sex.