Max D.M. Leiserson

Background: Thymoma and thymic carcinoma are the most common malignancies of the anterior mediastinum. Additionally, thymoma has a unique association with autoimmune disorders, notably myasthenia gravis (MG). Histologic classification of thymic epithelial tumors (TETs) has been largely based on the gross description of the epithelial cell appearance and the relative abundance of associated lymphocytes. A comprehensive molecular analysis of TETs has not heretofore been conducted. Methods: The TCGA Research Network conducted multi-platform analyses of 117 TETs (thymoma =105; thymic carcinoma =10 and micronodular thymoma =2), which included whole-exome, transcriptome, methylome and targeted proteome analysis. Patient characteristics: median age =60 years (range, 17–84 years); M:F (%) =52:48; Masaoka stage {I [36], IIA [39], IIB [19]; III [15]; IVA [1]; IVB [5]}; MG was present in 32 patients. No patient had prior therapy for metastatic disease, but 14 had prior chemotherapy and 39 had prior radiation therapy in the adjuvant setting. WHO histologic classification (blinded review) revealed A =10; AB =48, B1 =12, B2 =25, B3 =10, micronodular thymoma =2 and TC =10. Results: Thymoma has the lowest tumor mutation burden among adult malignancies in the TCGA. A unique transcription factor, GTF2I, was the most commonly observed mutation in WHO Types A and A/B. All GTF2I mutations were exclusively at the amino acid 424 locus. This is the only tumor with this specific mutation within the entire TCGA database. Differential expression of the RNA and protein data revealed dysregulation of several oncogenic pathways in GTF2I mutants vs. wild-type. Oncogenic HRAS, NRAS and TP53 mutations were also observed, but at a lower frequency among all TETs. We further describe an MSI-unstable thymic carcinoma that was hyper-mutated. Using multi-platform analyses, four distinct molecular-driven subtypes of TETs were identified that strongly correlated with the current WHO histologic classification and were associated with survival. Genomic hallmarks of these subtypes were identified to aid pathologic diagnosis. Lastly, when comparing MG-positive vs. -negative thymomas, we observed increased aneuploidy and overexpression of muscle auto-antigens in MG-positive tumors, providing a pathophysiologic link between thymoma and MG. Conclusions: Based on molecular analysis, four clusters were identified that correlated strongly with the current WHO Histologic Classification. Also identified was a unique mutation in GTF2I, which was associated with WHO Type A and A/B thymoma. Lastly, a molecular link between MG and thymoma characterized by increased aneuploidy and tumoral over-expression of muscle auto-antigens was observed. This international effort represents the largest and most comprehensive molecular analysis of TETs conducted to date is expected to have important clinical and translational implications for this rare disease.

Recent genomic analyses of pathologically-defined tumor types identify “within-a-tissue” disease subtypes. However, the extent to which genomic signatures are shared across tissues is still unclear. We performed an integrative analysis using five genome-wide platforms and one proteomic platform on 3,527 specimens from 12 cancer types, revealing a unified classification into 11 major subtypes. Five subtypes were nearly identical to their tissue-of-origin counterparts, but several distinct cancer types were found to converge into common subtypes. Lung squamous, head & neck, and a subset of bladder cancers coalesced into one subtype typified by TP53 alterations, TP63 amplifications, and high expression of immune and proliferation pathway genes. Of note, bladder cancers split into three pan-cancer subtypes. The multi-platform classification, while correlated with tissue-of-origin, provides independent information for predicting clinical outcomes. All datasets are available for data-mining from a unified resource to support further biological discoveries and insights into novel therapeutic strategies.