Renal cell carcinoma (RCC) represents the 6th most commonly diagnosed cancer in males and the 10th most commonly diagnosed cancer in females, worldwide, with incidence rates rising. It has been estimated that 3-5% of RCC is inherited.

The aim of this thesis was to improve the understanding of the genetic basis of predisposition to RCC by identifying the genetic architecture of inherited RCC, potential novel genetic causes and mechanisms of tumourigenesis. Additionally, evidence-based recommendations for the molecular investigation, germline variant interpretation and clinical management of patients with syndromic and non-syndromic inherited RCC are suggested.

A variety of investigative approaches were utilised for this thesis. Firstly, a service evaluation was undertaken which showed a low diagnostic yield (2-3%) for a current clinical RCC panel consisting of 6 genes (BAP1, FH, FLCN, MET, SDHB, VHL). Following this I explored whether a larger panel consisting of 121 cancer susceptibility genes (CSGs) would increase the diagnostic yield and identify novel CSG and RCC phenotype associations in participants recruited with renal cancer (N=1,336) in the 100 000 Genomes Project (100KGP) which was unselected for inherited RCC risk factors. Whole genome sequencing (WGS) analysis identified 6.4% (95%CI [5.1, 7.8]) pathogenic/likely pathogenic variants (85 SNVs/indels and 3 CNVs) in a wider range of CSGs than previously recognised and mean RCC onset was 58.6 years. 24 ‘hot’ VUSs were also identified potentially increasing the diagnostic yield to 8%. Burden test analyses demonstrated excess of CHEK2 variants in European RCC cases compared to controls (P=0.0019) and a trend of association with double strand homologous (DS HR) and nonhomologous repair (DS NHR) pathways.

Candidate novel RCC genes were sought by familial studies on 13 trios recruited to the 100KGP and the Human Molecular Pathology of Genetic Disease (HumGenDis) study with WGS/WES data available. Trio analysis identified a novel de novo pathogenic variant in elongin C (ELOC), NM_005648.4:c.236A>G[p.Tyr79Cys] in a proband with a von Hippel Lindau phenotype. This variant occurs as a somatic hotspot mutation in RCC and disrupts pVHL-ELOC interaction mimicking the effects of pVHL inactivation on expression of hypoxia response elements. ELOC functions as a tumour suppressor gene and chromosome 8 deletion was demonstrated in the RCC tissue from the proband and other sporadic ELOC-mutated RCC tumours in the 100KGP and the literature. Five additional novel candidate genes (CTTN, DMXL1, MUC17, GANAB and DET1) were identified in other probands, all with missense de novo variants, which require validation in larger cohorts and/or additional functional studies to definitively determine their pathogenicity and clinical significance.

To enhance clinical management of families with suspected inherited RCC, I explored the sensitivity, specificity and positive predictive value of immunohistochemical (IHC) staining for the fumarate hydratase protein expression and the antibody detection of S-(2-succino)-cysteine (2-SC) in determining FH variant pathogenicity. A clinical pathway of investigation was proposed for individuals presenting with leiomyomatosis. IHC for 2SC and FH used together had a 100% sensitivity, specificity and PPV in identifying P/LP FH variants. In addition, the frequency of candidate P/LP FH variants in the 100KGP cohort was explored and age-related RCC risks were calculated for participants with fumarase deficiency (FMRD) related variants (4%; 95%CI:0.88-1 by age 75 years). Overall frequency of candidate P/LP FH variants in the 100KGP cohort was 1 in 200; 1 in 5,000 for HLRCC-related variants and 1 in 250 for FMRD related FH variants suggesting that HLRCC is underdiagnosed and/or less penetrant than previously thought.

In conclusion, about 6% of unselected RCC cases carry a P/LP variant showing that inherited RCC is more common than previously thought, thus I suggest expanding the existing clinical RCC panel and raise the age threshold for testing. ELOC is a novel candidate gene for VHL-disease related phenotypes and testing should be offered to VHL patients with no pathogenic VHL variant identified. I propose a management pathway for the investigation of suspected HLRCC in individuals with multiple leiomyomas and I showed that candidate P/LP FH variants are more common than previously thought with implications for RCC screening. Further work could involve validating findings in additional cohorts and with additional technologies like long read sequencing.