Longitudinal Molecular Characterization of Lethal Prostate Cancer: A Major Step Towards Precision Oncology

Project goals:

  1. Identify drivers of lethality by multi-omic sequencing (genome, methylome, transcriptome, proteome) of primary tumours and benign prostate tissues from 300 radical prostatectomy (RP) patients: 100 with lethal prostate cancer (PCa), 100 presumably cured (no recurrence at >10 years post-RP) and 100 with recurrence and stable for >5 years under androgen-deprivation therapy (ADT).
  2. Trace most promising molecules and molecular changes longitudinally in banked serial liquid biopsies from these patients. This will include genes representing prostate cell subtypes and encoding drug targets.
  3. Profile PCa cell subtypes in tumours by spatial imaging and CTCs in serially collected liquid biopsies by single cell transcriptomic sequencing.
  4. Re-enroll patients experiencing late recurrence and/or failing treatment for closer blood draws to test gene panels defined in (2) as a function of further therapies through in-house or pharma-sponsored clinical trials.
  5. Integrate findings by merging data from omics with patient clinical and sociodemographic data and response/resistance to drugs or new therapeutics.
  6. Develop machine learning and artificial intelligence approaches to identify clinically applicable parameters to distinguish specific categories of patients and have an impact on PCa death.



PCa is the most prevalent malignancy in Canadian men, accounting for 20% of all cancers. The disease is highly heterogeneous and clinically challenging due to its unpredictable biologic nature. While being curable by surgery and/or radiation therapy, 25-35% of patients experience a recurrence and are managed by successive non-curative therapies. They eventually develop metastases and progress from being androgen-sensitive/responsive to castration-resistant/androgen-independent when reaching the end stage. PCa ranks 3rd as a cause of cancer death in Canadian men. Amongst other cancers, lethal prostate cancer displays an unusually long trajectory, often beyond 10 years, making it difficult to study longitudinal biologic changes related with cancer progression, resistance to therapy, and death. Research in individual cancer profiling has dramatically improved due to more precise molecular techniques. However, findings have not yet led to significant changes in the clinic, being limited by lack of follow up information and low number of lethal cases in cohorts. To overcome these limitations, we have access to prostate tissues and longitudinally collected blood samples, coupled with detailed clinical information from patients of the prospective PROCURE Biobank. The cohort comprises over 2 000 Quebec participants enrolled between 2006 and 2013 and still followed in 2021. This long follow-up period has allowed us to identify most severe and lethal cases to conduct a comprehensive profiling study from diagnosis, through various treatment steps, and ultimately death, something never done before.  We anticipate identifying key molecular and genetic changes from the outset that drive lethality, or acquired throughout the course of disease. This comparison of original cancer characteristics at diagnosis with the genetic changes in periodic serial blood tests throughout a patient’s trajectory makes this a very unique and valuable combination for achieving true precision medicine.