Abstract
Men with low-risk prostate cancer (PCa) (Gleason Score 6) have several treatment options including active surveillance (AS) and radical prostatectomy (RP). At present, there are no genomic copy number means of identifying which of those men treated by either strategy will go on to develop a more aggressive form of PCa. There is, therefore, an urgent need to identify the patterns able to discriminate indolent and aggressive phenotypes of low-risk PCa, to inform patient stratification and to tailor management.I therefore aim to use somatic copy number aberrations (SCNAs) to determine which men who present as Gleason score 6 will upgrade to a higher Gleason score on subsequent sampling if enrolled onto AS. In addition, a proportion of men post RP will develop new local or distal disease detectable as a biochemical recurrence. I aim to use genetic predisposition markers and somatic copy number aberrations to identify which of these men will undergo biochemical recurrence post RP.
In order to identify which men will develop a biochemical recurrence I first conducted a systematic review of 698 GWAS studies to collate potential risk alleles associated with progression of the disease. Fifty-three unique SNPs residing in 29 genomic regions, including 8q24, 10q11 and 19q13, were recurrently cited to be associated with aggressive PCa. By corroborating these findings with functional, and copy number studies, 21/54 SNPs were reported as modulating genes in pathways related to disease progression in PCa, thereby strengthening their tumorigenic potential towards aggressive PCa forms.
There is no publicly available cohort to date which contains both genomic and upgrading information. In order to assess which SCNAs may be associated with men with Gleason 6 disease who upgraded whilst on AS, SCNAs from 288 low-grade PCas (Gleason 6, Gleason 7 (4+3) and (3+4)s) in the TCGA cohort were pooled. Unsupervised hierarchical clustering of the SCNAs of these low-grade PCas identified a subset of Gleason 6s as those which would potentially upgrade as they clustered with higher-GS PCas and were termed putative upgrading GS6s. The cluster containing a majority of GS6 PCas contained no men who underwent biochemical recurrence and were termed putative stable GS6s. As there were no means of discerning definitively whether or not this method would identify those men who would potentially upgrade, orthogonal validation was conducted. The gene expression was investigated between the dichotomised Gleason 6s revealing genes in the putative upgrading group, previously associated with aggressive PCa progression (e.g. HOXC6). Using elastic net regularisation, 93 SCNAs were found to be associated with GS6s in the putative upgrading group and 11 were associated with the putative stable group. Those SCNAs associated with GS6s in the putative upgrading group were termed potential driver SCNAs. Of the SCNAs associated with this putative upgrading group, 6q13 was a predominantly subclonal aberration in the GS7(3+4) and was significantly more clonal in the GS7(4+3)s (p=0.03) indicating potential selection for progression. Two losses 8p23.1 and 8p11.23 which increase in frequency with GS were also associated with the putative upgrading GS6s, however these aberrations remained clonal across GS meaning they are more likely to be associated with tumour initiation rather than progression.
A local cohort of 13 men on AS with available initial and follow-up biopsies were used to validate these potential driver SCNAs. They were sequenced to such a low-depth that a validation of focal aberrations could not take place. However, ten potential driver SCNAs, present in the initial and repeat biopsies of more than one man who upgraded, were in the same location or neighbouring (6q13) as those identified by elastic net regularisation in the putative upgrading GS6s. Secondary observations from this cohort of men on AS was made using spatial information of the initial and repeat biopsies from each of the 13 patients. More similar somatic copy number aberrations were seen between the tumours taken from initial and repeat biopsies from men whose cancer had upgraded rather than those men who remained stable despite being physically far apart.
Somatic copy number aberrations associated with and without biochemical recurrence were assessed in a combined data-set of 45 Gleason score 6 cases from The Cancer Genome Atlas (TCGA) and 29 Gleason score 6 cases from (ICGC) where a total of 10 men had undergone biochemical recurrence and 64 had not. SCNAs associated with biochemical recurrence were identified through elastic net regression. A high Shannon-diversity score capturing the amount of intra-tumoural heterogeneity from SNParray data, was significantly associated with biochemical recurrence in patients with low-risk PCas in the complete low-grade TCGA cohort.
These extensively collected genomic and genetic features have been found to be associated with Gleason Score 6 tumours from men whose disease has progressed.
| Date of Award | 1 Aug 2019 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Anita Grigoriadis (Supervisor) & Mieke Van Hemelrijck (Supervisor) |