Penetrance Estimation in Breast Cancer Patients via Modified Kin-Cohort Method

Introduction: The diagnosis of symptoms and the estimation of cancer risk in individuals are of great interest for research teams. Several methods developed for estimation of the risk of cancer in the individuals exposed to risk. Because of its capability to illustrate the correlation between genotype and phenotype, the Kin-Cohort Method has become a method of choice for many researchers. This study aims to employ a modified Piecewise Weibull Model for estimation of penterance in patients with breast cancer characterized by BRCA1 and BRCA2 gene mutations carriers.

Materials & Methods: The research follows a descriptive and analytical method. A set of data was simulated to resemble the true values. The simulated data set was analyzed using both Piecewise Exponential and Piecewise Weibull models. EM algorithm and bootstrap sampling were employed for maximization. Akaike’s criterion was used to compare the two methods, and graphs of the Penetrance values were plotted to show the differences. To estimate the penetrance of BRCA1/2 gene mutations, data was collected from the patients voluntarily referred to the Department of Medical Genetics and the Department of Cancer and Cytological Genetics of Medical School of Tehran University of Medical Sciences.

Results: Results of the simulated data showed that the estimates of cancer penetrance in the carriers and non-carriers in all age groups were closer to the default values in the Weibull model in comparison with the exponential model. The differences in Akaike’s criterion, 311 and 2753 for the non-carriers and carriers respectively in both models, showed a significant difference between the Exponential and Weibull models. Estimated penetrance for the age groups below and over 50 among BRCA1/2 carriers for breast cancer was 31.9% and 46.2% respectively.

Conclusion: The knowledge of penetrance is important in genetic counseling. Therefore methods capable of generating most accurate estimations are preferred. Results of the simulation revealed that the piecewise Weibull model is preferred for the estimation of Penetrance in Cancer Patients. The low value of the estimated penetrance in this study can be attributed to the rare mutation in Iranian population. Establishment and use of a Kin-Cohort gene databank is proposed as a solution for preparation of screening programs and estimation of the penetrance to help reduce the risk of cancer.