Background-The term primary lymph node gastrinoma was first used to describ
e a group of patients with gastrin-producing tumors present in lymph nodes
located in a well-defined anatomic region. The patients had no known primar
y tumors in the pancreas or gastrointestinal tract and had disease-free sur
vival for up to 18 years. The anatomic region in question has a triangular
shape that extends from the cystic and common bile ducts to the second and
third portion of the duodenum and the neck and body of the pancreas. The te
rm gastrinoma triangle was coined to identify the area; in addition, it was
postulated that lymph nodes located in the gastrinoma triangle normally co
ntained neuroendocrine cells capable of secreting gastrin and other neurope
ptides. From its inception, the postulate became the subject of controversy
.
Design.-To extend previous observations, we examined the lymph nodes locate
d in the gastrinoma triangle of 20 autopsy cases for the presence of neuroe
ndocrine cells, as determined by immunohistochemistry, using antibodies to
a panneuroendocrine substance (eg, synaptophysin) and a specific neuropepti
de (eg, gastrin). Scanning for positive cells was performed by 2 observers
(M.E.H. and M.C.C.). We compared the findings in these lymph nodes with lym
ph nodes obtained from axillary and inguinal dissections during surgical pr
ocedures.
Results.-In all, 417 lymph nodes were studied. Five of the 20 gastrinoma tr
iangle cases contained synaptophysin reactive cells, whereas 3 had gastrin
reactive cells. None of the axillary and inguinal lymph nodes contained neu
roendocrine cells.
Conclusion.-Our findings support the hypothesis of entrapment of neuroendoc
rine cells during development and the presence of primary nodal gastrinomas
.