BACKGROUND: Pancreaticoduodenectomy is an accepted surgical option for
certain benign conditions and biopsy proven cancer. Whether this proc
edure should be performed when malignancy of the pancreas and periampu
llary region is suspected but not confirmed represents a fairly common
intraoperative dilemma. PATIENTS AND METHODS: Sixty-seven patients wh
o had undergone pancreaticoduodenectomy during a 15-year period were e
valuated retrospectively. RESULTS:The indications for resection were s
ymptomatic benign conditions (n = 10, 15%), proven pancreatic or peria
mpullary cancer (n = 37, 55%), and suspected but unproven malignancy (
n = 20, 30%). The patients with suspected malignancy ranged in age fro
m 27 to 73 years. Common findings in this group were abdominal pain (7
5%), jaundice (70%), weight loss (65%), and alcohol use (45%). There w
ere 14 pancreatic and 6 ampullary masses. Biopsies obtained preoperati
vely (n = 15) and intraoperatively (n = 11) were nonconfirmatory. Post
operatively 9 patients (45%) were found to have tumors, including 6 pa
ncreatic adenocarcinoma, 2 duodenal adenocarcinoma, and 1 islet cell t
umor. Six of the 8 adenocarcinomas (75%) were stage I. Seven patients
were alive 11 to 108 months later. The most common benign diagnosis wa
s pancreatitis. There were 8 complications and 1 death. CONCLUSIONS: P
ancreaticoduodenectomy performed based on suspicion alone frequently r
eveals malignancy. Immediate and long-term outcomes are acceptable. Th
ese findings justify a continued aggressive approach to suspected panc
reatic and periampullary malignancy.