Combination of metabolomic and lipidomic analyses for the differentiation of high-grade precancerous lesions from cervical cancer

Luana Quadros de Souza Leão1, Giovanna Melo Marques1, Adriano de Souza Carolino1, Ana Carolina Rosa Silva2, Sofia Angiole Cavalcante3, Elzalina Ribeiro Soares1, Jonas Balan de Padua4, Anderson Ferreira Gonçalves4, Marina Amaral Alves2, Rafael Garrett 2, Priscila Ferreira de Aquino3

1. UFAM, Federal University of Amazonas ; Av. General Rodrigo Octavio Jordão Ramos, 1200 - Coroado I, Manaus - AM, 69067-005
2. UFRJ, Federal University of Rio de Janeiro; Antonio Barros de Castro Street, 119 - University City, Rio de Janeiro, Rio de Janeiro, Brazil
3. ILMD/FIOCRUZ Amazônia, Leonidas and Maria Deane Institute/Oswaldo Cruz Foundation ; Terezina Street, 476 - Adrianopolis, Manaus, Amazonas, Brazil
4. FCECON, Amazonas State Oncology Control Center Foundation ; Francisco Orellana Street, 215 - Planalto, Manaus, Amazonas, Brazil

The magnitude of the public health problem posed by cervical cancer (CC) in Brazil is reflected primarily by its persistently high incidence and mortality rates. Specifically, the situation is critical in Amazonas, which ranks first in incidence and mortality from this neoplasm. Among the measures adopted to prevent the progression of this disease, screening and treatment of CIN 2 and CIN 3 cervical intraepithelial lesions stand out. Although they present different probabilities of regression and risks of progression to CC, these neoplasms are grouped as high-grade cervical lesions in screening and, consequently, are treated similarly through the conization procedure, which can carries risks associated with reproductive complications. In this context, this study aimed to identify metabolic and lipid indicators present in cervical tissue samples capable of discriminating between CIN 2 and CIN 3 lesions. For this, cervical tissue from lesions and adjacent margins was collected from women diagnosed with CIN 2 and CIN 3 (CAAE 71342417.4.3002.502). The tissues underwent alternating freeze-thaw maceration cycles for metabolomic and lipidomic analysis, followed by metabolite/lipid extraction using the Matyash method. Subsequently, the samples were analyzed by LC-HRMS. Multivariate statistical analyses and correlation were performed to interpret the data obtained. Additionally, metabolite and lipid identification were performed using MS-DIAL software. As a result, a separation was observed between the CIN 2 (lesion and margin) and CIN 3 (lesion margin) categories. Furthermore, a positive correlation was observed between metabolites such as taurine and succinic acid and triglyceride lipids, which may be associated with the mechanism of suppression and progression of these neoplasms. Finally, six metabolites and three lipids were identified as potential markers to differentiate CIN 2 and CIN 3 lesions. Therefore, it was possible to demonstrate the potential of using metabolomics and lipidomics approaches to discriminate patients with high-grade precancerous lesions from cervical cancer, which may contribute in the future to the emergence of complementary tools for screening these lesions, enabling more individualized treatment.

Agradecimentos: CAPES - Finance Code 001, FAPEAM (PPSUS Program, PAIC Program and POSGRAD Program), FCECON, UFAM, LADETEC-LabMETA/UFRJ and PROEP/ILMD-FIOCRUZ AMAZÔNIA – LDMAIS.