Abstract 2434: Characterizing age-of-onset of multiple primary cancers and specific cancer types in families with Li-Fraumeni syndrome

Abstract

Abstract Purpose: Li-Fraumeni syndrome (LFS) is a rare autosomal dominant disorder associated with a germline TP53 mutation that increases an individual’s risk for a spectrum of cancers as well as multiple primary diagnoses. Penetrance estimations of multiple primary cancer (MPC) risk and cancer specific (CS) risk associated with TP53 germline mutations remains a challenge because of limited data available for statistically meaningful estimation and the difficulty of accounting for variables like previous primary cancer effect or ascertainment bias inevitable in rare diseases. Accurate penetrance estimations are crucial to improve the clinical characterization and management of high-risk individuals of LFS. Modeling MPC and CS risks are also relevant questions for the general cancer population. Methods: Through a novel non-homogenous Poisson process to model the non-recurrent event approach, we estimated the MPC penetrance from a large pediatric sarcoma cohort from MD Anderson Cancer Center (MDACC). We used data from all individuals, with or without TP53 genotype information, and specifically considered age as well as ages of first primary cancer diagnosis for model parameter estimation. Next, we developed a model under the competing risk framework that takes into account the pedigree structure into the penetrance estimation and corrects for ascertainment bias to address the TP53 penetrance for three cancer subtypes (breast, BR/sarcoma, SA/others, OT) on the same pediatric cohort. Results: We then validated the risk prediction performance of the two sets of penetrance estimates via independent cohort datasets: a clinically ascertained dataset collected at MDACC (total number of individuals=774; SPC=264; MPC=252) for MPC penetrance validation and a LFS dataset collected at NCI combined with a dataset from the International Sarcoma Kindred Study (total number of individuals=1,083, BR=85, SA=540, OT=158) for CS outcome validation. Among carriers, the MPC penetrance estimates, which is dependent on TP53 genotype, gender, current age and age of first primary diagnosis are significantly different from those of a single primary cancer (SPC). We achieved an AUC=0.73 for predicting individual outcomes of multiple primaries vs. a single primary cancer. For the CS penetrance of TP53 mutation carriers for breast cancer, sarcoma, and other cancers related to LFS, we obtained AUCs of 0.92 (BR), 0.75 (SA), and 0.81 (OT). Conclusion: We have obtained two new sets of penetrance to accurately characterize the age-of-onset of multiple primary cancers and specific cancer types like breast, sarcoma or others in families with LFS. Incorporating associated risk prediction models into LFS counseling can potentially facilitate clinical decision-making to both patients and their physicians. Future work will be focused on integrative analysis of multi-cohort datasets. Citation Format: Seung Jun Shin, Elissa Dodd, Gang Peng, Jasmina Bojadzieva, Jingxiao Chen, Jing Ning, Phuong L. Mai, Sharon A. Savage, Mandy L. Ballinger, David M. Thomas, Ying Yuan, Louise C. Strong, Wenyi Wang. Characterizing age-of-onset of multiple primary cancers and specific cancer types in families with Li-Fraumeni syndrome [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2434.