2022-LBA-1382-ESGO Genomic signatures for the prediction of recurrence and metastasis in Endometrial Cancer. (27th October 2022)

• Record Type: Journal Article
• Authors: Farres, Alba; Teixeira, Natalia; Espanol, Pia; Magret, Eva; Luna, Rocio; Soler, Cristina; Martin-Malpartida, Pau; Macías, Maria Jesús; Céspedes, Maria Virtudes; Rovira, Ramon
• Is Part Of: International journal of gynecological cancer. Volume 32(2022)Supplement 2
• Page Start: A471
• Page End: A471
• Publication Date: 2022-10-27
• Subjects: Generative organs, Female -- Cancer -- Periodicals; 616.99465
• DOI: 10.1136/ijgc-2022-ESGO.1016
• Languages: English
• ISSNs: 1048-891X

Abstract

Abstract : Introduction: Lymphadenectomy is normally indicated in Endometrial Cancer (EC) staging. Only 20–25% of patients are diagnosed with nodal involvement, therefore 70% of patients will have lymphadenectomy-related risk without therapeutic benefit. Moreover, one fourth of node-negative EC recur, suggesting that some high-risk tumors are not detected using lymphadenectomy. We hypothesize that genetic/transcriptomic and clinical differences among patients might help predict tumor recurrence risk. This study aims to develop a diagnostic tool to determine the probability of relapse/metastasis using genetic indicators. Methods: Tumor samples from a cohort of patients surgically staged for EC were collected. Total RNA was obtained, then, gene and microRNA expression arrays were performed. Clinical and pathological data were extracted from patient files. Artificial intelligence using the machine learning protocol was first trained with clinical and genetic data downloaded from public databases. Re-training of the system will be performed with the newly acquired data from the current study. Results: Genetic material has been extracted from 120 of the 150 samples considered optimal. Transcriptomic profiles have been quantified and obtained from 90 of these samples. The patient samples studied were classified into low-risk (groups A and B) and high-risk tumors (remaining), and each of the groups has been subdivided according to surgical treatment (with or without lymphadenectomy). Each subgroup has been classified into reurrence (yes/no), resulting in groups of patients between A-H . With the available transcriptomic data, we can identify potential patients that although classified as low risk, would suffer recurrence. So far, these are preliminary data obtained from a pilot study. Additional samples are currently being analysed to increase the statistical value of the observations. Conclusions: With the transcriptomic data obtained to date, we can identify patients with a higher risk of recurrence despite being classified as low risk.