5-10% of differentiated thyroid cancers cluster in families (familial non-medullary thyroid cancer, FNMTC). Few (<5%) families have thyroid cancer syndromically (e.g. Cowden’s disease); little is known about non-syndromic FNMTC. Recently, a germline mutation of HABP2 was identified through whole exome sequencing (WES) in a family with three affected members; but the generalizability of this finding to other FNMTC families has been strongly challenged. We aimed to identify whether FNMTC kindreds carry mutations in genes associated with known syndromic FNMTC (DICER1, APC, PRKAR1A, PTEN, WRN) or in HAPB2 or harbour novel germline mutations.
With ethics approval, a family with eight individuals with differentiated thyroid cancer was recruited from Royal Brisbane and Women’s Hospital Thyroid Cancer Clinic. DNA was extracted from saliva of living individuals and tumour blocks of deceased individuals. Exome capture and massively parallel sequencing was performed on three individuals (grandmother and two grandchildren from separate families). Good quality variants with minor allele frequency <5% were retained. The data were analysed for novel or rare (MAF<0.001) heterozygous variants predicted to be deleterious/damaging by at least one protein predication algorithm (SIFT, Mutation Taster, Polyphen). Five genes contained rare variants fulfilling these criteria; of these, three were novel (S100A3, RASGRF2, TLN1). The RASGRF2 variant was of particular interest,given the known role of somatic BRAF mutations in thyroid cancer and the interaction of BRAF and RAS pathways. Additionally, somatic mutations in RASGRF2 have been identified in human and rodent tumours (pancreatic, mammary, colon and lung).
We believe we may have identified a novel cause of FNMTC. Segregation of all five variants is currently being assessed. A priori the chance of sharing a variant amongst the three sequenced individuals was 1/16; the probability of segregation with all affected members is 1/512 (0.00195).