Luiz Eduardo Wildemberg1,2, Aline Helen da Silva Camacho3, Renan Lyra Miranda3, Paula C L Elias4, Nina R de Castro Musolino5, Debora Nazato6, Raquel Jallad7,8, Martha K P Huayllas9, Jose Italo S Mota10, Tobias Almeida11, Evandro Portes12, Antonio Ribeiro-Oliveira13, Lucio Vilar14, Cesar Luiz Boguszewski15, Ana Beatriz Winter Tavares16, Vania S Nunes-Nogueira17, Tânia Longo Mazzuco18, Carolina Garcia Soares Leães Rech19, Nelma Veronica Marques1, Leila Chimelli3, Mauro Czepielewski11, Marcello D Bronstein7,8, Julio Abucham6, Margaret de Castro4, Leandro Kasuki1,2, Mônica Gadelha1,2,3. 1. Endocrine Unit and Neuroendocrinology Research Center, Medical School and Hospital Universitário Clementino Fraga Filho-Universidade Federal do Rio de Janeiro, Rio de Janeiro, RJ, Brazil. 2. Neuroendocrine Unit-Instituto Estadual do Cérebro Paulo Niemeyer, Secretaria Estadual de Saúde , Rio de Janeiro, RJ, Brazil. 3. Neuropathology and Molecular Genetics Laboratory, Instituto Estadual do Cérebro Paulo Niemeyer, Secretaria Estadual de Saúde, Rio de Janeiro, RJ, Brazil. 4. Division of Endocrinology-Department of Internal Medicine, Ribeirao Preto Medical School-University of Sao Paulo, São Paulo, SP, Brazil. 5. Neuroendocrine Unit, Division of Functional Neurosurgery, Hospital das Clinicas, Faculdade de Medicina, Universidade de São Paulo, São Paulo, SP, Brazil. 6. Neuroendocrine Unit-Division of Endocrinology and Metabolism-Escola Paulista de Medicina-Universidade Federal de São Paulo (Unifesp), São Paulo, SP, Brazil. 7. Neuroendocrine Unit, Division of Endocrinology and Metabolism, Hospital das Clínicas, University of São Paulo Medical School, São Paulo, SP, Brazil. 8. Cellular and Molecular Endocrinology Laboratory/LIM25, Discipline of Endocrinology, Hospital das Clinicas HCFMUSP, Faculty of Medicine, University of Sao Paulo, São Paulo, SP, Brazil. 9. Neuroendocrinology and Neurosurgery unit Hospital Brigadeiro, São Paulo, SP, Brazil. 10. Endocrinology and Metabolism Unit, Hospital Geral de Fortaleza, Secretaria Estadual de Saúde, Fortaleza, CE, Brazil. 11. Division of Endocrinology, Hospital de Clinicas de Porto Alegre (UFRGS), Porto Alegre, RS, Brazil. 12. Institute of Medical Assistance to the State Public Hospital, São Paulo, SP, Brazil. 13. Faculdade de Medicina, Universidade Federal de Minas Gerais, Belo Horizonte, MG, Brazil. 14. Neuroendocrine Unit, Division of Endocrinology and Metabolism, Hospital das Clínicas, Federal University of Pernambuco Medical School, Recife, PE, Brazil. 15. Endocrine Division (SEMPR), Department of Internal Medicine, Universidade Federal do Parana, Curitiba, PR, Brazil. 16. Endocrine Unit-Department of Internal Medicine, Faculty of Medical Sciences, Universidade do Estado do Rio de Janeiro, RJ, Brazil. 17. Department of Internal Medicine, São Paulo State University/UNESP, Medical School, Botucatu, SP, Brazil. 18. Division of Endocrinology of Medical Clinical Department, Universidade Estadual de Londrina (UEL), Londrina, PR, Brazil. 19. Santa Casa de Porto Alegre, Porto Alegre, RS, Brazil.
Abstract
CONTEXT: Artificial intelligence (AI), in particular machine learning (ML), may be used to deeply analyze biomarkers of response to first-generation somatostatin receptor ligands (fg-SRLs) in the treatment of acromegaly. OBJECTIVE: To develop a prediction model of therapeutic response of acromegaly to fg-SRL. METHODS: Patients with acromegaly not cured by primary surgical treatment and who had adjuvant therapy with fg-SRL for at least 6 months after surgery were included. Patients were considered controlled if they presented growth hormone (GH) <1.0 ng/mL and normal age-adjusted insulin-like growth factor (IGF)-I levels. Six AI models were evaluated: logistic regression, k-nearest neighbor classifier, support vector machine, gradient-boosted classifier, random forest, and multilayer perceptron. The features included in the analysis were age at diagnosis, sex, GH, and IGF-I levels at diagnosis and at pretreatment, somatostatin receptor subtype 2 and 5 (SST2 and SST5) protein expression and cytokeratin granulation pattern (GP). RESULTS: A total of 153 patients were analyzed. Controlled patients were older (P = .002), had lower GH at diagnosis (P = .01), had lower pretreatment GH and IGF-I (P < .001), and more frequently harbored tumors that were densely granulated (P = .014) or highly expressed SST2 (P < .001). The model that performed best was the support vector machine with the features SST2, SST5, GP, sex, age, and pretreatment GH and IGF-I levels. It had an accuracy of 86.3%, positive predictive value of 83.3% and negative predictive value of 87.5%. CONCLUSION: We developed a ML-based prediction model with high accuracy that has the potential to improve medical management of acromegaly, optimize biochemical control, decrease long-term morbidities and mortality, and reduce health services costs.
CONTEXT: Artificial intelligence (AI), in particular machine learning (ML), may be used to deeply analyze biomarkers of response to first-generation somatostatin receptor ligands (fg-SRLs) in the treatment of acromegaly. OBJECTIVE: To develop a prediction model of therapeutic response of acromegaly to fg-SRL. METHODS:Patients with acromegaly not cured by primary surgical treatment and who had adjuvant therapy with fg-SRL for at least 6 months after surgery were included. Patients were considered controlled if they presented growth hormone (GH) <1.0 ng/mL and normal age-adjusted insulin-like growth factor (IGF)-I levels. Six AI models were evaluated: logistic regression, k-nearest neighbor classifier, support vector machine, gradient-boosted classifier, random forest, and multilayer perceptron. The features included in the analysis were age at diagnosis, sex, GH, and IGF-I levels at diagnosis and at pretreatment, somatostatin receptor subtype 2 and 5 (SST2 and SST5) protein expression and cytokeratin granulation pattern (GP). RESULTS: A total of 153 patients were analyzed. Controlled patients were older (P = .002), had lower GH at diagnosis (P = .01), had lower pretreatment GH and IGF-I (P < .001), and more frequently harbored tumors that were densely granulated (P = .014) or highly expressed SST2 (P < .001). The model that performed best was the support vector machine with the features SST2, SST5, GP, sex, age, and pretreatment GH and IGF-I levels. It had an accuracy of 86.3%, positive predictive value of 83.3% and negative predictive value of 87.5%. CONCLUSION: We developed a ML-based prediction model with high accuracy that has the potential to improve medical management of acromegaly, optimize biochemical control, decrease long-term morbidities and mortality, and reduce health services costs.
Authors: Daniel G Henriques; Elisa B Lamback; Romulo S Dezonne; Leandro Kasuki; Monica R Gadelha Journal: Int J Mol Sci Date: 2022-08-04 Impact factor: 6.208