Paraneoplastic autoimmune diseases associate occasionally with small cell lung cancers and gynecologic tumors. However, myasthenia gravis (MG) occurs in at least 30% of all patients with thymomas (usually present at MG diagnosis). These epithelial neoplasms almost always have numerous admixed maturing polyclonal T cells (thymocytes). This thymopoiesis-and export of mature CD4(+)T cells-particularly associates with MG, though there are rare/puzzling exceptions in apparently pure epithelial WHO type A thymomas. Other features potentially leading to inefficient self-tolerance induction include defective epithelial expression of the autoimmune regulator (AIRE) gene and/or of major histocompatibility complex class II molecules in thymomas, absence of myoid cells, failure to generate FOXP3(+) regulatory T cells, and genetic polymorphisms affecting T-cell signaling. However, the strong focus on MG/neuromuscular targets remains unexplained and suggests some biased autoantigen expression, T-cell selection, or autoimmunization within thymomas. There must be further clues in the intriguing serological and cellular parallels in some patients with late-onset MG but without thymomas-and in others with AIRE mutations-and in the contrasts with early-onset MG, as discussed here.

Thymic neuroendocrine tumors (TNET) are rare primary epithelial neoplasms of the thymus. This study aimed to determine clinically relevant parameters for their classification and for therapeutic decisions. We performed a comprehensive histological, clinical, and genetic study of 73 TNET cases (13 thymic typical carcinoids [TTC], 40 thymic atypical carcinoids [TAC], and 20 high-grade neuroendocrine carcinomas [HGNEC] of the thymus), contributed by multiple institutions. The mean number of chromosomal imbalances per tumor was 0.8 in TTC (31% aberrant cases) versus 1.1 in TAC (44% aberrant cases) versus 4.7 in HGNEC (75% aberrant cases). Gains of 8q24 (MYC gene locus) were the most frequent alteration and one of the overlapping features between carcinoids and HGNEC. The 5-year survival rates for TTC, TAC, and HGNEC were 100, 60, and 30%. The 10-year survival rates for TTC and TAC were 50 and 30% (P = 0.002). Predictive mitotic cut-off values for TTC versus TAC were 2.5 per 10 high-power fields (HPF; indicating a higher death rate, P = 0.062) and 15 per 10 HPF (indicating higher risk of recurrence, P = 0.036) for separating HGNEC from TAC. We conclude that the current histopathologic classifications of TNET reflect tumor biology and provide important information for therapeutic management.

USCNB should be preferred to USFNA when a definite diagnosis of a parotid solid mass is needed.