TSC2 rare germline variants in non-tuberous sclerosis patients with neuroendocrine neoplasias

Neuroendocrine neoplasias (NENs) represent a heterogeneous group of diseases with a wide spectrum of morbidity and lethality. These tumors can arise in almost any organ of the body, with the pancreas and the gastrointestinal tract being the most common primary sites (Yao et al. 2017). NENs are mainly considered sporadic diseases, although 5–20% of all cases occur in the context of genetic syndromes such as multiple endocrine neoplasia (MEN) type 1, MEN type 2, von Hippel–Lindau (VHL) disease, neurofibromatosis syndrome and tuberous sclerosis complex (TSC), which are caused by the presence of germline mutations in the MEN1, RET, VHL, NF1 and TSC1/TSC2 genes, respectively (Kim & Hong 2016). The incidence of NEN among patients with these syndromes varies considerably. Indeed, the incidence is very high in patients with MEN type 1 and is represented by isolated reports describing a small number of patients with TSC (Larson et al. 2012, Koc et al. 2017). Notably, epidemiological studies have demonstrated that non-syndromic individuals with an affected relative are at an increased risk of developing gastroenteropancreatic (GEP) NENs. In individuals with affected siblings, the relative risk of GEP NEN increases by 13.4-fold, and an increased risk (2.3-fold) is present even in third-degree relatives (Neklason et al. 2016). Collectively, these epidemiological studies suggest the existence of additional genetic determinants leading to the development of GEP NENs in non-syndromic patients. Here, we screened non-syndromic patients with GEP NEN for the presence of rare germline variants in genes previously associated with NEN-predisposing syndromes and searched for variants that could increase the risk of developing NENs without being associated with the corresponding syndromes. Ninety-three patients with any grade GEP NENs and with no clinical or molecular diagnosis of MEN type 1 or 2, von-Hippel–Lindau disease, neurofibromatosis syndrome or TSC who were being treated at Hospital Sírio-Libanês or at Instituto do Câncer do Estado de São Paulo were enrolled in this study. All patients provided written informed consent, and the study was approved by the ethics committees of both institutions (HSL2015-15 and NP762/15). Germline DNA was obtained from peripheral leucocytes. DNA libraries covering the coding regions of the MEN1, RET, VHL, NF1, TSC1 and TSC2 genes were prepared using TruSeq Custom Amplicon (Illumina) and sequenced on MiSeq (Illumina). Sequences were aligned to the human genome reference (hg19) using BWA, and nucleotide variants (SNVs and indels) were called using GATK. Only variants with predicted damaging impact on protein function (determined by SIFT or PolyPhen-2) and minor allele frequency (MAF) reported by ExAC (Exome Aggregation Consortium) lower than 0.1% were kept for further analysis. We identified 8 patients with missense rare germline variants in TSC2 (8.6%) and one in RET (1.1%). The presence of these variants was further validated by Sanger sequencing. There were no duplicated variants and no patient had more than one alteration. All missense rare germline variants identified in the TSC2 and RET genes have a predicted damaging effect on protein function, but none is considered diagnostic for TSC or MEN2 syndromes by consensus criteria (Northrup & Krueger 2013) and, as of this writing, their clinical significance has not been established. Four of these variants have been described in LOVD database (TSC2_03292, TSC2_02023, TSC2_01014 and TSC2_0008), but none of them has been considered causative of the disease (Table 1). 2 25