Abstracts
2002 Digestive Disease Week

# 101736 Abstract ID: 101736 Intraductal Papillary Mucinous Tumors (IPMT) of the Pancreas: Mucus Is a Predictor of Better Survival and Benign Disease
Yuichi Kitagawa, Trisha A Unger, Richard A Kozarek, William Traverso, Seattle, WA

Aim: To examine a large number of resected IPMT cases for predictors of malignancy and survival. Methods: Between 1989 and 2000 at one institution, 63 resected patients were pathologically diagnosed as having IPMT using the WHO definition. Each patient?s demographic data, symptoms, imaging studies, and pathological findings were recorded. Variables analyzed were alcohol abuse, history of pancreatitis, abdominal pain, jaundice, weight loss, increased LFTs or CA19-9, mass on CT or US, gross mucus present in the ductal system, location (head, body/tail or diffuse), location in the ducts: side-branch type, main-duct (MPD) type, type of resection (PPWhipple, distal, or total pancreatectomy). Factors found predictive on univariate analysis were then analyzed with multivariate analysis. Results: The tumors were benign (n=30, 48%), invasive carcinoma (n=28, 44%) and carcinoma in-situ (CIS, n=5, 8%). Resections involved PPW in 34, total pancreatectomy in 8, distal pancreatectomy in 15, and other types or resection in 6. Average follow-up time was 43 ? 36 mo. After univariate analysis predictors of malignancy were jaundice (p=0.0006, relative risk (RR) = 10.32), CA19-9 (p=0.0004, RR=15.0), increased LFTs (p=0.0003, RR=7.69), and MPD-type (p=0.02, RR=3.49) while a predictor of benign disease was grossly observed mucus (p=0.006, RR 4.34). After the above were subjected to multivariate analysis then predictors of malignancy were LFTs (RR=5.09) and MPD type (RR=2.72) and a predictor of benign disease was still gross mucus (RR=5.9). Actuarial 3 and 5 yr survival was 95% and 83% for benign cases and 52% and 44% for malignant cases (both p<0.01). Among 31 patients with invasive or CIS disease (follow-up of 34 ? 32 mo) there was one significant predictor of survival after univariate and multivariate analysis - the presence of gross mucus (RR=4.34 and 4.55, respectively. Alcohol abuse was the only predictor of not surviving (R=3.41 and 3.60 respectively). Conclusion: After almost 4 years of follow-up it was not surprising that patients with benign IPMT had a better survival. However, the presence of gross mucus indicated that the IPMT was likely to be benign and, if malignant, the presence of mucus indicated better survival. Any process that resulted in LFT changes predicted malignancy. The role of alcohol abuse toward a lower likelihood of survival is speculative.