Introduction: The identification of actionable molecular targets has been recently adopted to guide therapeutic decisions in diverse tumor types. The identification of the target genes and the respective drugs is of remarkable importance for patients diagnosed with rare tumors, where the therapeutic arsenal is more limited. This is the case of salivary gland tumors, a group of neoplastic lesions with a frequency of ~1/100,000, that has the mucoepidermoid carcinoma (MEC) as its most common form. The high-grade MEC patients have a poor prognosis with a 5-years survival rate of only about 30%. Complete resection of primary tumor is the standard treatment for all grades of MEC and high-grade tumors are generally treated with surgical excision followed by postoperative radiotherapy. Currently, there is no prognostically useful regimen of chemotherapy or biotherapy. In this sense, a view of druggable target options for these tumors - where targetable variants are matched with available molecularly targeted therapies - is deeply needed to guide therapeutics of MECs.
Methods: We performed whole exome sequencing (WES) in 22 samples derived from 15 MEC-patients - including 8 independent cases as well as 7 normal/tumor primary salivary MEC pairs – to investigate the presence of mutations affecting actionable variants. Samples were collected during surgery, before chemotherapy and/or radiotherapy. WES was performed using the Ion Torrent TargetSeq kit. Single Nucleotide Variants (SNVs) as well as insertions or deletions (InDels) were detected by GATK (variant database; dbSNP and 1000G) after mapping to UCSC Hg19. Mutations were investigated in a list of the top-50 most explored genes for targeted therapies.
Results: Our WES allowed at least 80% of the exome to be covered >50x. From the 50 druggable target genes investigated here, we observed mutations in 3/15 patients (20%). Most of these were missense (66%), followed by splice-site mutations (34%). Patients and mutations found are as described: #1 - CTNNB1 and ERBB4 missense mutations; #2 CDKN2A mutation in a splicing acceptor site; #3 EGFR and KDR missense mutations and FGFR1 splicing site. These mutations were found only in patients with worse outcome, i.e. patients that died due to this cancer.
Conclusions: Precision therapy of patients with salivary MEC remains to be achieved. Whereas most of the genes with mutations found here have had treatment options tested on salivary MEC, few objective responses to these agents were achieved in phase II trials. Remarkably selection for these trials was based in the overexpression of these markers instead of identification of mutations in the target genes. Since no molecular signatures appeared to be more prevalent in MEC in this limited cohort, our data suggests that molecular-guided Basket Clinical Trials could help identifying those most likely to benefit from available drugs. Additional studies are required to verify the mutational frequency of these genes in a larger patient cohort. Research supported by FAPESP (Grants: 14/06186-1 & 14/07249-7).