Abstracts
1999 Digestive Disease Week

# 3471 IS CT-ANGIOGRAPHY SUFFICIENT FOR PREDICTION OF RESECTABILITY OF PANCREATICO-DUODENAL NEOPLASMS?
P F Saldinger, R Richards, J Goldberg, V Raptopoulos, M L Steer, Jeffrey B Matthews, Beth Israel Deaconess Med Ctr, Boston, MA

Purpose: The optimal preoperative evaluation of pancreatico-duodenal neoplasms remains controversial. It is unclear whether a single radiologic modality is sufficient to determine resectability. The value of additional modalities such as laparoscopy ± ultrasound in reducing the failure rate is still being evaluated. Methods: Sixty-two patients with pancreatico-duodenal neoplasms were explored with intent to perform a pancreaticoduodenectomy between 3/96 and 11/98. All of these patients were prospectively staged by helical CT and CT angiography with three-dimensional reconstruction. Vascular involvement was graded from 0 - 4 (Raptopoulos et al AJR:168,1997). No grade 4 patients were explored. Results: Overall resectability was 74% (46/62). Resectability in grade 0 and 1 (n=44) was 96%; 2 patients had unsuspected hepatic disease (4%). Resectability in grade 2 and 3 (n=18) was 22%, unsuspected hepatic disease accounting for only one patient (6%). The rest of the patients had unresectable disease due to local vascular invasion (72%). Conclusion: The high resectability rate in grade 0 and 1 patients suggests that other staging modalities are superfluous. In this series unresectability in grade 2 and 3 patients was largely dictated by the already-known vascular involvement suggesting that laparoscopy alone would be unlikely to improve the assessment of resectability.

Grade	Finding	n
0	Normal, preserved fat plane around the portal vein/SMV	40
1	Loss of fat plane, smooth displacement of the vessel	4
2	Flattening or irregularity on one side of the vessel	7
3	Encased narrowed vessel with tumor > 50% of the circumference	11
4	Occluded vessel	N/A

Grading was compared to operative resectability.

Copyright 1996 - 1999, SSAT, Inc.