OBJECTIVE: To determine whether atrial myxomas express antigens suggesting a neural origin. DESIGN: A retrospective analysis based on immunohistochemical examination of myxoma tissue. SETTING: Atrial myxomas excised by two tertiary referral cardiothoracic surgical units. SUBJECTS: 24 excised atrial myxomas. Three were from known cases of familial myxoma syndrome. METHODS: Immunohistochemical identifications of three neuroendocrine markers (protein gene product (PGP) 9.5, neurone specific enolase (NSE), synaptophysin) and S100 antigen; CD34 and von Willebrand factor; and chi smooth muscle actin to identify possible Schwann cell differentiation, endothelial cells, and smooth muscle cells respectively. RESULTS: The myxoma cells were PGP 9.5 positive in 18, S100 positive in 16, and NSE positive in 12. Of the 12 NSE positive myxomas seven were synaptophysin positive. All tumours that were NSE positive were also S100 and PGP 9.5 positive. The tumour surface was partially covered by myxoma cells, partly by endothelial cells. CONCLUSION: The histological appearances of myxomas with stellate cells embedded within a loose connective tissue stroma, abundant basophil cell infiltration, and the presence of pericellular type IV collagen are similar to nerve sheath tumours (neurofibromas) at other sites. A significant proportion of myxomas also express Schwann cell and neuroendocrine differentiation markers. These features cannot prove the origin of myxomas because tumours may develop aberrant phenotype expression but they do accord with the view that myxomas originate from endocardial sensory nerve tissue.
OBJECTIVE: To determine whether atrial myxomas express antigens suggesting a neural origin. DESIGN: A retrospective analysis based on immunohistochemical examination of myxoma tissue. SETTING:Atrial myxomas excised by two tertiary referral cardiothoracic surgical units. SUBJECTS: 24 excised atrial myxomas. Three were from known cases of familial myxoma syndrome. METHODS: Immunohistochemical identifications of three neuroendocrine markers (protein gene product (PGP) 9.5, neurone specific enolase (NSE), synaptophysin) and S100 antigen; CD34 and von Willebrand factor; and chi smooth muscle actin to identify possible Schwann cell differentiation, endothelial cells, and smooth muscle cells respectively. RESULTS: The myxoma cells were PGP 9.5 positive in 18, S100 positive in 16, and NSE positive in 12. Of the 12 NSE positive myxomas seven were synaptophysin positive. All tumours that were NSE positive were also S100 and PGP 9.5 positive. The tumour surface was partially covered by myxoma cells, partly by endothelial cells. CONCLUSION: The histological appearances of myxomas with stellate cells embedded within a loose connective tissue stroma, abundant basophil cell infiltration, and the presence of pericellular type IV collagen are similar to nerve sheath tumours (neurofibromas) at other sites. A significant proportion of myxomas also express Schwann cell and neuroendocrine differentiation markers. These features cannot prove the origin of myxomas because tumours may develop aberrant phenotype expression but they do accord with the view that myxomas originate from endocardial sensory nerve tissue.