Transcript variation in multifocal prostate cancer

Prostate cancer is the most common type of cancer in Norway, where about 5000 men get this diagnosis every year. Although only a minority of the patients have the most aggressive type, there is a lack of tools to predict who is in need of the most radical treatment, and about 1000 Norwegian men suffer cancer-related death annually. Thus, it is challenging to customize how the patients should be treated and followed-up. Despite the fact that the majority of prostate cancer patients have multiple tumours when they are diagnosed, previous research on prostate cancer tissue have almost exclusively been performed on only a single tumour sample per patient. This may be why no good prognostic biomarker has been developed for this disease. In the TRANSPROCAN project, the research is based on tissue and blood samples alongside with clinical and pathological data from more than 500 patients who were operated for prostate cancer at the Oslo University Hospital-Radiumhospitalet between 2010 and 2012. Multiple tissue samples from multiple tumours were frozen from each patient to enable research taking the multi-focal nature of prostate cancer into account. By use of state-of-the-art genome technologies, DNA and RNA changes in cancer cells are used to investigate how often the so-called multi-focal prostate cancers originate from the same cancer cell, or whether they are independent tumours. The project found that the set of mutations from different tumour foci within the same prostate is virtually allways completely different (Løvf et al., European Urology 2019). Further, the project found that classification of the disease based on particular RNA expression produces different results when the RNA is sampled from different areas in the prostate (Carm et al., Scientific Reports 2019). Therefore, to meet the next goal of the project, to identify molecules which can distinguish aggressive from mild prostate cancer, this heterogeneity, or differences, between tumours of the same patient are taken into account. Despite extensive spatial heterogeneity, the researchers found that the amount of the long noncoding RNA SCHLAP1 has prognostic value (Kidd et al., Neoplasia 2022). The same was the case for the gene ERG (Kidd et al., Molecular Oncology 2022). Analyses of blood samples taken at the time the patients experienced relapse has made it possible to trace which of the original tumours that were responsible for the metastatic and/or relapsed disease. The resulting mutational patterns were quite complex, and it was evident that the relationship between the samples can only rarely be read directly from the mutational patterns. There were only few shared mutations both between different tumors within the same prostate, and there were only few shared mutations between the original primary tumors and the metastatic tumours. The project concluded that for the treatment of prostate cancer to be based on gene based tests, the analyses must be performed on multiple tumor samples. Further, it became clear that for metastatic and/or recurrent disease, a relatively newly taken sample must be analysed.

Prosjektet har ikke ved prosjektslutt gitt noen direkte endringer i oppfølging av pasienter eller implementerte genbaserte tester. Likefullt er det resultater fra prosjektet som endrer kunnskapsgrunnlaget ganske radikalt. Derfor vil videre utvikling av genbasert testing i prostatakreft måtte ta hensyn til resultater fra prosjektet. Konklusjonene fra prosjektet er tydelige på at for å basere behandling av prostatakreft på genbaserte tester, noe som er vanlig for stadig flere former for kreft, så må det analyseres flere prøver fra den enkelte pasient. Videre må det ved sykdom som allerede har spredd seg, eller ved tilbakefall etter operasjon, gjøres analyser på en nokså nylig tatt prøve.

Prostate cancer is a major health burden, and improved biomarkers are needed for both detection and prognostication of the disease. However, the mutation spectrum and expressed RNA-variants are complex, and different driver genes are commonly active within different tumour foci of the same prostate. We have competence in identification and characterization of cancer-specific transcripts, and in particular in fusion gene detection. There is still a need to study the heterogeneity of different tumor foci in sufficiently large clinical cohorts of prostate cancer. Genome technologies will enable us to detect both novel genetic elements and unravel how existing genes and biomarkers are distributed across multiple tissue cores in this heterogeneous disease. Further, we will use various genome sequencing and informatics technologies to explore on the sources of transcriptome instability in cancer. This phenotype, originally described by us, denotes cancers with particularly high fraction of miss-spliced transcripts. However, sequence-level and genome-scale characterization is lacking, and no information exists in a multi-focality/heterogeneity context.