Rare tumor with matched germline whole exome sequencing to identify somatic and inherited variants of clinical significance.

Abstract

1523 Background: Rare cancer incidence is defined as <6/100,000 cases. These cases are challenging due to delays in diagnosis, late stage at diagnosis, lack of standard of care and poor outcomes. We present molecular profiling of 28 rare tumor cases analyzed by whole exome sequencing (WES) seen in the Early Phase Therapeutics Program. Methods: We performed WES on 15 rare tumors with matched germline DNA, as well as tumor only on 2 additional cases. We sequenced a panel of 500 candidate cancer genes in 6 tumor/normal and 6 tumor only cases. Copy number alterations (CNAs) were assessed for the tumor-normal WES cases with SNVs and small indels for all cases. Caris (AZ) (tumor only) and liquid biopsy (Gaurdant360, CA) were also conducted in these patients for comparison. Results: Rare tumors affected 18 different tissues of the body with 1 case affecting tissue typical of common cancers. The rare tumor group contained 25% pathogenic or likely pathogenic germline variants compared to 7.1% for common tumors. This increased rate of inherited pathogenic variants met statistical significance: Fisher exact test (p = 0.0118). A total of 69 sequence variants and 8 CNAs were identified, 37 were actionable. For all but 3 patients, there was at least one variant associated with an actionable outcome given off-label use or clinical trial participation. There was excellent concordance with results from commercial sequencing; however, our tumor/germline sequencing identified additional germline variants of clinical significance, all were loss-of-function variants in tumor suppressors. Tumor vs. normal analysis showed that the majority of the commercially reported VUS were in fact germline VUS. In one case, a reported pathogenic variant we found to be an inherited pathogenic germline variant. 15 patients received at least 1 line of therapy indicated by genomic sequencing. For these patients the PFS ratio when compared to the most recent line of therapy prior to targeted therapy was 2.02, with 10/15 (67%) patients having a PFS ratio of >1.0, and 6/15 (40%) of patients having PFS ratio >1.3. Conclusions: We demonstrate the importance of tumor-normal analysis, especially in the context of rare tumors. Rare tumor patients may disproportionately benefit from tumor vs. normal WES at diagnosis to improve PFS within targeted therapy trials, guided by genomics.