Long-term survival in pancreatic cancer: Pylorus-preserving versus Whipple pancreatoduodenectomy France Mosca, MD, FACS, Pier C. Giulianotti, MD, Tommaso Balestracci, MD, Giulio Di Candio, MD, Andrea Pietrabissa, MD, Fabio Sbrana, MD, and Giuseppe Rossi, MSc, Piss, Itab Background. This study compared long-term survival in pancreatic or periampullary cancer treated with Whipple pancreatoduodenectomy (PO) and pylorus$reserving pancreatoduodenectomy (PPPD). Methods. Two hundred twenty-one patients with pancreatic head or periampullary cancer were treated. Prognostic variables included age, gendq type and period of operation, and tumor stage. In the ductal adenocarcinomas variables also included tumor and node status, type of lymphadenectomy, pathologic grade, and presence of microscopic residual tumor The end point was death as a result of neoplastic recurrence. Survival curves were estimated by using the Kaplan-Me&r method, and multifactorial analysis was also performed on the data from the ductal adenocarcinoma group. Results. The mortality rate was 8.2 % in the PD group versus 7.0% in the PPPD group. Morbidity rates were 34.4 % for PD and 45.8% for PPPD. Five-year survival was 9.6% in the ductal adenocar- cinoma and 63.8 % in the periampullary carcinoma groups. Univariate analysis failed to show statisti- cally significant diff erences in survival curves between the two treatments in either patient group. Correcting for multiple variables in the ductal adenocarcinoma group did not reveal any significant dif- ferences in survival rates between the two treatments. Conclusions. PPPD was as successful as classic PD in the treatment of ductal adenocarcinoma and periampullary cancer of the pancreas. Long-term survival was not influenced by the type of resection. (Surgery 1997;122:553-66.) From the Istituto di Chirurgia Generale e Sperimentale and Unit& di Epiclemiologia e Biostatistica Istituto di Fisiologia Clinica CN& Uniuevsitci di Piss, Piss, Italy FOR MANY YEARS the Whipple pancreatoduodenec- tomy (PD) has been the standard treatment of pancreatic and periampullary neop1asms.l In spite of continuous technical improvements that have decreased operative morbidity and mortality rates worldwide, long-term survival of patients with pan- creatic cancers remained unsatisfactory. Even the most aggressive procedures appeared to offer only a few months of palliation for most patients.2>3 As a consequence, some surgeons explored the potential benefits of more radical procedures, such as regional pancreatectomy and extended lymphadenectomy,4,5 and others abandoned resec- tion in favor of palliative bypass.6 Although it appears that surgery alone will never be the sole treatment option for pancreatic cancer, the low Accepted for publication March 5, 1997. Reprint requests: Pier C. Giulianotti, MD, Istituto di Chirurgia Generale e Sperimentale, U.O. di Chirurgia Generale e Vascolare, Universiti di Pisa, Via Paradisa Z/a, 56124 Pisa, Italy. Copyright 0 1997 by Mosby-Year Book, Inc. 0039.6060/97/$5.00+0 11/56/81643 mortality rate currently associated with resection in many centers justifies the use of surgical proce- dures that remain the only hope of curing this dis- ease. The pylorus-preserving pancreatoduodenecto- my (PPPD) , introduced by Traverso and Longmire in 1978, offers theoretical advantages over PD in both ease of performance and postoperative func- tion.@ For these reasons PPPD has become the treatment of choice for chronic pancreatitis and periampullary cancer. However, many surgeons remain reluctant to use this procedure, which is supposed to be less radical than Whipple, for patients with pancreatic cancer.g,10 This retrospective analysis compares the results of PD and PPPD in the management of ductal adenocarcinoma and periampullary carcinoma of the pancreas. MATERIAL Between January 1, 1980, and December 31, 1994, 513 patients underwent surgery at the Istituto di Chirurgia Generale e Sperimentale for SURGERY 553