Periodic Reporting for period 3 - AVATAR (Integrating Genomics and Avatar Mouse Models to Personalize Pancreatic Cancer Treatment)
Periodo di rendicontazione: 2018-11-01 al 2020-04-30
Pancreatic Ductal Adenocarcinoma (PDAC) is one of the most lethal cancers in humans. It is expected that pancreatic cancer will be the second cause of cancer related mortality after lung cancer by 2030. Patients with this disease are treated with chemotherapy with limited results. So far, newer approaches to cancer treated such as precision medicine or immune-oncology have not resulted in meaningful improvements. Our goal was to demonstrate that the survival of patients with this disease could be improved by a comprehensive approach to precision medicine. Our central idea, which has been proven correct in other cancers, in that the response of a tumour to treatment drugs is related to specific molecular and biological features in cancer cells and that this information can be used for individualized patient treatment. To achieve these goals, we have performed a randomized clinical trial in patients with metastatic PDAC to compare the outcome of patients treated with conventional therapies with those treated with a personalized approach. This personalized approach included the generation of an Avatar mouse model (PDX) and of Patient Derived Organoid (PDO) from each patient, as well as a full genome analysis (WES) of the patients own tumour as well as tumour DNA obtained from blood. The information obtained from these different studies is then integrated to derive a personalize treatment plan for the patient. The overall goal is to demonstrate that this approach increases patient survival.
Eligible patients were randomized to either conventional treatment or to a personalize medicine approach in a 1:2 ratio. Patients randomized to the conventional arm were treated with standard of care therapy. Patients randomized to the experimental arm underwent a tumor biopsy that was used to determine Whole Exome Sequencing (WES) (NextSeq, Illumina) and to generate an Avatar model. Once the patient derived xenograft (PDX) had been established, part of this sample was used to establish a patient derived organoid (PDO) model. Treatment was based on the targets identified by genetic and bioinformatic analysis, as well as by phenotypic screening of 40 anticancer agents (High Throughput Screening, HTS) in the patient derived models.
Until January 2020, 128 of the 146 patients have been enrolled. Of the 128 patients enrolled in the study, 83 were randomized to the experimental arm and 45 to the conventional one. A WES and PDX has been obtained from 60 % and 65 % of patients respectively. All PDX models can be converted in a PDO. Patients are currently in follow up to determine outcome.
Another goal for this project was to evaluate the feasibility of identifying druggable somatic mutations on PDACs from patients’ plasma performing WES of circulating free DNA (cfDNA). Although relatively low levels of cfDNA have been observed in PDACs on different studies, yielding poor and/or inconsistent concordant results between liquid biopsy and tumour-based assays in this type of cancer, we have shown that the WES-cfDNA is feasible and enables the identification of somatic mutations of potential clinical interest in some cases with PDACs when the exome sequencing of the biopsy came as non-informative. Using this innovative technology, we identified mutations (like ROS1 and ARID1A) that were confirmed by WES performed in patients’ tumours and/or PDX. In other patients we have been able to monitor the evolution of some genes that have led to drug resistance through treatment.
Until January 2020, 128 of the 146 patients have been enrolled. Of the 128 patients enrolled in the study, 83 were randomized to the experimental arm and 45 to the conventional one. A WES and PDX has been obtained from 60 % and 65 % of patients respectively. All PDX models can be converted in a PDO. Patients are currently in follow up to determine outcome.
Another goal for this project was to evaluate the feasibility of identifying druggable somatic mutations on PDACs from patients’ plasma performing WES of circulating free DNA (cfDNA). Although relatively low levels of cfDNA have been observed in PDACs on different studies, yielding poor and/or inconsistent concordant results between liquid biopsy and tumour-based assays in this type of cancer, we have shown that the WES-cfDNA is feasible and enables the identification of somatic mutations of potential clinical interest in some cases with PDACs when the exome sequencing of the biopsy came as non-informative. Using this innovative technology, we identified mutations (like ROS1 and ARID1A) that were confirmed by WES performed in patients’ tumours and/or PDX. In other patients we have been able to monitor the evolution of some genes that have led to drug resistance through treatment.
This work is still ongoing and at this time we cannot confirm whether this approach will lead to changes in the management of patients with PDAC. We have learned that Avatar models alone, because of the time it takes to be generated and tested, cannot be used in all patients for therapy selection. Thus, we are incorporating new approaches such as more extensive and deep sequencing (HTS) to explore targetable mutations as well as phenotypic screening using PDOs. We expect that these interventions will allow the development of a truly personalized medicine approach that leads to a significant change in the management of PDAC.