Austin Health, Monash Health, Peter McCallum Cancer Centre
Introduction: Familial prostate cancer accounts for around 10% of prostate cancer diagnoses. In Australia, this represents a cohort of around 2000 men each year. Strong evidence exists that familial prostate cancers are more aggressive than sporadic cancers, best exemplified by the BRCA2 cohort. BRCA gene identification and clinical correlation with aggressive disease has allowed for modified treatment and surveillance protocols, delivering improved outcomes. Despite advances, relatively few non-BRCA germline mutations have been identified in association with prostate cancer. Lack of evidence has resulted in limited testing in a clinical setting. In this study, we aim to identify novel mutations associated with prostate cancer in a cohort of men with familial disease and correlate these mutations with clinical features of disease.
Methods: 94 families with at least two cases of verified prostate cancer were identified from within the KConFab (familial cancer) cohort. Two-thirds of families had initially been recruited due to a strong presence of breast, ovarian and colorectal cancers with limited genetic testing identifying no pathogenic mutations. Remaining families had been recruited due to multi-case prostate cancer and had not previously undergone testing. 84-cancer gene panel testing was conducted for index cases from each family, and their relatives with verified prostate cancer if the index case was found to a mutation. Clinical data including age at diagnosis, PSA, pathology stage and grade were collected for mutation carriers and non-carriers. Descriptive statistics were used to explore clinical features of each mutation.
Results: 9 clinically notifiable, non-BRCA, pathogenic mutations were identified (ATM n=4, HOXB13 n=3, CHEK2 n=2). ATM carriers were the largest carrier group, with nine carriers with prostate cancer confirmed across four novel ATM mutations. Mean age at diagnosis was 65 years. Six ATM carriers were stratified to the D’Amico high-risk group on the basis of Grade Group 5 pathology (n=3) or advanced stage at diagnosis (n=3). Remaining carriers were stratified into intermediate (n=2) or low risk (n=1) groups. Despite this classification, age at diagnosis for the patient in the low-risk group was just 42 years. Specialist uropathology review identified high-risk architecture (IDCP, cribiform) in four available histopathology samples. Additionally, primary lung, bladder, renal, skin (melanoma) and blood (myelodysplastic syndrome) cancers were identified among this group.
Conclusions: Men with germline ATM mutations in this cohort developed high-risk prostate cancer with concerning pathological architecture. This is consistent with previous evidence associating familial prostate cancers with high-risk disease, even without germline mutations identified. Associating novel mutations with clinical disease is an important step forward in identifying those at risk of developing clinically significant cancers. Although small, in combination with further research, this data set contributes significantly to evidence for ATM gene testing in the clinical setting.
Source of Funding: kConFab is supported by a grant from the National Breast Cancer Foundation and Prostate Cancer Foundation Australia.