Search

Objective: To determine the rates of germline BRCA1 and BRCA2 mutations in Scottish patients with ovarian cancer, before and after a change in testing policy.Design: Retrospective cohort study.Setting: Four cancer/genetics centres in Scotland.Population: Patients with ovarian cancer undergoing germline BRCA1 and BRCA2 (gBRCA1/2) sequencing before 2013 (under the 'old criteria', with selection based solely on family history), after 2013 (under the 'new criteria', with sequencing offered to newly presenting patients with non-mucinous ovarian cancer), and in the 'prevalent population' (who presented before 2013, but were not eligible for sequencing under the old criteria but were sequenced under the new criteria).Methods: Clinicopathological and sequence data were collected before and for 18 months after this change in selection criteria.Main Outcome Measures: Frequency of germline BRCA1, BRCA2, RAD51C, and RAD51D mutations.Results: Of 599 patients sequenced, 205, 236, and 158 were in the 'old criteria', 'new criteria', and 'prevalent' populations, respectively. The frequency of gBRCA1/2 mutations was 30.7, 13.1, and 12.7%, respectively. The annual rate of gBRCA1/2 mutation detection was 4.2 before and 20.7 after the policy change. A total of 48% (15/31) 'new criteria' patients with gBRCA1/2 mutations had a Manchester score of <15 and would not have been offered sequencing based on family history criteria. In addition, 20 patients with gBRCA1/2 were identified in the prevalent population. The prevalence of gBRCA1/2 mutations in patients aged >70 years was 8.2%.Conclusions: Sequencing all patients with non-mucinous ovarian cancer gives a much higher annual gBRCA1/2 mutation detection rate, with the frequency of positive tests still exceeding the 10% threshold upon which many family history-based models operate.Tweetable Abstract: BRCA sequencing all non-mucinous cancer patients increases mutation detection five fold. [ABSTRACT FROM AUTHOR]

We explore the prediction of individuals' phenotypes for complex traits using genomic data. We compare several widely used prediction models, including Ridge Regression, LASSO and Elastic Nets estimated from cohort data, and polygenic risk scores constructed using published summary statistics from genome-wide association meta-analyses (GWAMA). We evaluate the interplay between relatedness, trait architecture and optimal marker density, by predicting height, body mass index (BMI) and high-density lipoprotein level (HDL) in two data cohorts, originating from Croatia and Scotland. We empirically demonstrate that dense models are better when all genetic effects are small (height and BMI) and target individuals are related to the training samples, while sparse models predict better in unrelated individuals and when some effects have moderate size (HDL). For HDL sparse models achieved good across-cohort prediction, performing similarly to the GWAMA risk score and to models trained within the same cohort, which indicates that, for predicting traits with moderately sized effects, large sample sizes and familial structure become less important, though still potentially useful. Finally, we propose a novel ensemble of whole-genome predictors with GWAMA risk scores and demonstrate that the resulting meta-model achieves higher prediction accuracy than either model on its own. We conclude that although current genomic predictors are not accurate enough for diagnostic purposes, performance can be improved without requiring access to large-scale individual-level data. Our methodologically simple meta-model is a means of performing predictive meta-analysis for optimizing genomic predictions and can be easily extended to incorporate multiple population-level summary statistics or other domain knowledge., (© The Author 2015. Published by Oxford University Press.)