Fatores clínicos associados ao óbito e à recidiva em pacientes com carcinoma mamário invasivo: análise com regressão logística penalizada em coorte hospitalar

Juarez Antônio  de Sousa; Waldemar  Naves Amaral; André  Maroccolo de Sousa; Débora  Nascimento Dias Neves; Laura  Sousa e Sousa; Luiza  Sousa e Sousa; Letícia  Caiado Madi

doi:10.63162/v67n70e26720

Authors

Juarez Antônio de Sousa Doutor, Professor do Departamento de Ginecologia e Obstetrícia da Faculdade de Medicina da Universidade Federal de Goiás (UFG), Goiânia, GO, Brasil. Autor https://orcid.org/0000-0001-5986-7926
Waldemar Naves Amaral Doutor, Professor do Departamento de Ginecologia e Obstetrícia da Faculdade de Medicina da Universidade Federal de Goiás (UFG), Goiânia, GO, Brasil. Autor https://orcid.org/0000-0002-0824-1138
André Maroccolo de Sousa Médico Residente de Cirurgia Geral, Universidade de São Paulo (USP), São Paulo, SP, Brasil. Autor https://orcid.org/0009-0004-5613-1410
Débora Nascimento Dias Neves Médica, Universidade Federal de Goiás (UFG), Goiânia, GO, Brasil. Autor https://orcid.org/0000-0003-1280-5415
Laura Sousa e Sousa Acadêmica de Medicina, Universidade UniEvangélica, Anápolis, GO, Brasil. Autor https://orcid.org/0009-0004-9019-8846
Luiza Sousa e Sousa Acadêmica de Medicina, Universidade UniEvangélica, Anápolis, GO, Brasil. Autor https://orcid.org/0009-0007-3855-8682
Letícia Caiado Madi Médica Residente de Ginecologia e Obstetrícia, Maternidade Aristina Cândida, Goiânia, GO, Brasil. Autor https://orcid.org/0000-0002-2038-1379

DOI:

https://doi.org/10.63162/v67n70e26720

Keywords:

Breast neoplasms, Prognosis, Logistic regression, Survival analysis, Neoplasm recurrence local

Abstract

Introduction: Identifying clinical predictors of death and recurrence in breast cancer is essential for prognostic stratification and individualized treatment. In cohorts with few events, conventional statistical methods are inadequate, requiring penalized approaches such as Firth’s regression. Objective: To identify clinical factors associated with death and tumor recurrence in patients with invasive breast carcinoma in a hospital cohort, using Firth’s penalized logistic regression. Methods: Retrospective cohort study with 97 patients with confirmed invasive breast carcinoma, excluding ductal carcinoma in situ. Univariate analysis used Fisher’s exact test. Firth’s penalized logistic regression models were fitted for death and recurrence, expressed as odds ratios (OR) with 95% CI. Event concentration by subgroup was presented in a staging × T-component cross-table. Results: Six deaths (6.2%) and 9 recurrences (9.3%) were recorded. Advanced staging (III–IV) was stronglyassociated with death (OR = 57.67; 95%CI: 9.39–643.91; p < 0.001) and recurrence (OR = 73.29; 95%CI: 14.05–544.21; p < 0.001). T3–T4 component showed associations of similar magnitude. In a multivariable model for recurrence, staging III–IV maintained independent association (adjusted OR = 39.99; 95%CI: 7.41–290.74; p < 0.001) after adjustment for axillary approach and HER2. Patients with stage III–IV and T3–T4 tumors had death rates of 50% and recurrence rates of 70%. The adjusted model ROC curve reached AUC = 0.942. Conclusion: Advanced clinical staging and T3–T4 component are fundamental determinants of adverse prognosis in invasive breast carcinoma. Firth’s penalized logistic regression proved appropriate for analyzing rare outcomes in small cohorts.

References

Instituto Nacional de Câncer José Alencar Gomes da Silva (INCA). Estimativa 2023: Incidência de Câncer no Brasil. Rio de Janeiro: INCA; 2022.

Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71(3):209-49.

Harbeck N, Penault-Llorca F, Cortes J, Gnant M, Houssami N, Poortmans P, et al. Breast cancer. Nat Rev Dis Primers. 2019;5(1):66.

Firth D. Bias reduction of maximum likelihood estimates. Biometrika. 1993;80(1):27-38.

Heinze G, Schemper M. A solution to the problem of separation in logistic regression. Stat Med. 2002;21(16):2409-19.

Heinze G, Ploner M. Fixing the nonconvergence bug in logistic regression with SPLUS and SAS. Comput Methods Programs Biomed. 2003;71(2):181-7.

Rakha EA, El-Sayed ME, Lee AH, Elston CW, Grainge MJ, Hodi Z, et al. Prognostic significance of Nottingham histologic grade in invasive breast carcinoma. J Clin Oncol. 2008;26(19):3153-8.

Lehmann BD, Bauer JA, Chen X, Sanders ME, Chakravarthy AB, Shyr Y, et al. Identification of human triple-negative breast cancer subtypes and preclinical models for selection of targeted therapies. J Clin Invest. 2011;121(7):2750-67.

Gradishar WJ, Moran MS, Abraham J, Aft R, Agnese D, Allison KH, et al. NCCN Guidelines® Insights: Breast Cancer, Version 4.2023. J Natl Compr Canc Netw. 2023;21(6):594-608.

R Core Team. R: A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing; 2023. Disponível em: https://www.R-project.org/.