Personalised medicine in endometrial cancer through the use of liquid biopsy

• Autores: Carlos Casas Arozamena
• Directores de la Tesis: Laura Muinelo Romay (dir. tes.), Miguel Abal Posada (dir. tes.), Rafael López López (tut. tes.)
• Lectura: En la Universidade de Santiago de Compostela ( España ) en 2024
• Idioma: inglés
• Tribunal Calificador de la Tesis: Camilla Krakstad (presid.), Mónica Martínez Fernández (secret.), Frédéric Amant (voc.)
• Programa de doctorado: Programa de Doctorado en Investigación Clínica en Medicina por la Universidad de Santiago de Compostela
• Materias:
  • Ciencias médicas
    • Ciencias clínicas
      • Oncología
      • Ginecología
    • Patología
      • Oncología
• Enlaces
  • Tesis en acceso abierto en: MINERVA
• Resumen

  • Endometrial cancer, the most common gynaecological malignancy in developed countries, has seen rising incidence and mortality rates. Despite generally favourable outcomes when detected early, 10-20% of patients experience tumour relapse post-surgery, leading to poor clinical outcomes. Molecular classification, though promising, has yet to fully inform treatment decisions, with chemotherapy remaining the mainstay for recurrent or metastatic cases although immunotherapy is gaining relevance as a second or a first line of treatment. Precision medicine holds promise in tailoring treatments, but lacks accurate predictive markers and relevant preclinical models to further test new alternatives. Liquid biopsy-based approaches offer minimally invasive disease characterization and circulating biomarkers for improved patient stratification and longitudinal monitoring. The current doctoral thesis addresses these challenges, focusing on non-invasive diagnostics, accurate risk assessment, personalized treatment selection, and the development of relevant preclinical models. In Chapter I, the efficacy of two targeted-NGS approaches in minimally invasive samples from endometrial cancer patients was assessed. The uterine aspirate showed prevalent mutations in key genes like PTEN, PIK3CA, and TP53, with TP53 alterations indicating poor prognosis. Moreover, mutations in RB1, MDM2, and ARID1A were linked to disease progression, suggesting potential therapeutic targets. The analysis also identified patients with intratumoral heterogeneity, crucial for treatment response variations. The cfDNA analysis in a smaller cohort revealed pathogenic mutations, predominantly in TP53, PIK3CA, and/or GNAS, with ctDNA dynamics correlating with relapse, indicating evolving mutational landscapes. In Chapter II, the methylome od UAs from endometrial cancer patients was analysed, revealing a hypomethylation profile associated with the aggressive serous histology. Key pathways like GABAergic and calcium signalling were implicated, suggesting potential therapeutic targets. The study supported the clinical relevance of uterine aspirate-based methylome analysis to better understand the biology of the most aggressive endometrial tumours. Chapter III evaluated the performance of a highly sensitive ddPCR assay to characterize microsatellite instability. The analysis of UAs from endometrial cancer patients showed high concordance with the standard method, highlighting its clinical utility. Besides, the assay showed value to detect the presence of ctDNA in cfDNA from high-risk tumours and monitor the disease evolution. In Chapter IV we analysed pre-surgery and longitudinal plasma cfDNA and ctDNA from endometrial cancer patients, correlating high pre-surgical levels of both markers with poor prognosis. A combined analysis of both markers improved risk stratification, identifying patients with a fast and poor disease evolution. Longitudinal cfDNA/ctDNA analysis proved effective in predicting relapse and monitoring the therapy response, enhancing endometrial cancer management. In Chapter V, circulating tumour cells were evaluated in localized endometrial cancer patients, with higher levels associated with aggressive phenotypes and shorter survival times. CTC presence correlated with poor prognosis but lacked independent prognostic significance. Chapter VI explored the clinical utility of preclinical models generated from minimally invasive samples. Patient-derived spheroids and organoids proved reliable in recapitulating the tissue of origin and identifying effective treatments, showcasing their potential for personalized medicine in endometrial cancer management. The liquid biopsy-based pipeline described in the thesis demonstrates significant clinical utility, as evidenced by its implementation at the clinical hospital of Santiago de Compostela, resulting in early relapse detection, treatment initiation, and the generation of reliable preclinical models. Overall, the thesis aimed to improve patient management in both localized and advanced endometrial cancer through the clinical application of liquid biopsy techniques.