OBJECTIVE: To specify uncharacterized amyloid deposits in gastrointestinal vessels of the elderly. MATERIALS AND METHODS: The gastrointestinal tracts from 110 consecutive autopsies of individuals aged 85 years and older were examined for amyloid using Congo red staining. Immunohistochemical classification of the amyloid deposits was conducted using antisera directed against amyloid A, apolipoprotein A-I, apolipoprotein A-II, apolipoprotein B, apolipoprotein C-I, lysozyme, lambda and kappa light chain amyloid fibril proteins, transthyretin, beta2-microglobulin, and amyloid P component. Electron microscopic examination assessed the ultrastructural features. RESULTS: Thirty-eight (35%) of the 110 cases had gastrointestinal amyloid deposits. In 17 cases the amyloid fibril proteins were defined immunohistochemically. In five cases (5%) the amyloid could not be classified because amyloid deposits were not present in the deeper serial sections used for immunohistochemistry. In 13 cases (11%) the vascular amyloid deposits could not be characterized because they did not demonstrate immunoreactivity with any of a panel of antibodies specific for the fibril proteins of all major extracerebral amyloids. In three individual cases, the vascular amyloid deposits showed variable immunoreactivity, with deposits being negative in some vessels. The immunohistochemically nonreactive vascular amyloid in these 16 cases had several consistent features: it affected only vessels of the small and large intestine, it was limited to mesenteric veins, it consisted of small dot- or comma-like deposits located in close proximity to fragmented elastic fibers, and it demonstrated inconsistent immunostaining for amyloid P component. CONCLUSIONS: The similar morphologic characteristics of nonreactive gastrointestinal amyloid deposits, which we have designated "portal amyloid," suggest a common origin. Determination of whether portal amyloid represents a new type of amyloid will require chemical analysis.
OBJECTIVE: To specify uncharacterized amyloid deposits in gastrointestinal vessels of the elderly. MATERIALS AND METHODS: The gastrointestinal tracts from 110 consecutive autopsies of individuals aged 85 years and older were examined for amyloid using Congo red staining. Immunohistochemical classification of the amyloid deposits was conducted using antisera directed against amyloid A, apolipoprotein A-I, apolipoprotein A-II, apolipoprotein B, apolipoprotein C-I, lysozyme, lambda and kappa light chain amyloid fibril proteins, transthyretin, beta2-microglobulin, and amyloid P component. Electron microscopic examination assessed the ultrastructural features. RESULTS: Thirty-eight (35%) of the 110 cases had gastrointestinal amyloid deposits. In 17 cases the amyloid fibril proteins were defined immunohistochemically. In five cases (5%) the amyloid could not be classified because amyloid deposits were not present in the deeper serial sections used for immunohistochemistry. In 13 cases (11%) the vascular amyloid deposits could not be characterized because they did not demonstrate immunoreactivity with any of a panel of antibodies specific for the fibril proteins of all major extracerebral amyloids. In three individual cases, the vascular amyloid deposits showed variable immunoreactivity, with deposits being negative in some vessels. The immunohistochemically nonreactive vascular amyloid in these 16 cases had several consistent features: it affected only vessels of the small and large intestine, it was limited to mesenteric veins, it consisted of small dot- or comma-like deposits located in close proximity to fragmented elastic fibers, and it demonstrated inconsistent immunostaining for amyloid P component. CONCLUSIONS: The similar morphologic characteristics of nonreactive gastrointestinal amyloid deposits, which we have designated "portal amyloid," suggest a common origin. Determination of whether portal amyloid represents a new type of amyloid will require chemical analysis.