1 2 4 3 5 8 Since randomized trials on the value of neoadjuvant therapy have not yet been published, it also remains unclear whether—particularly in younger patients—neoadjuvant treatment (or reoperation after initially palliative chemotherapy) may be beneficial, at least in some patients, and may thus lead to resectability of an initially unresectable tumor. 9 18 9 10 18 In this study, we retrospectively analyzed our single-center experience with a group of patients who had initially been diagnosed as unresectable and who were subsequently referred to us for re-exploration. In addition, a literature review of the available (retrospective) studies was performed to more precisely define the value of re-exploration. PATIENTS AND METHODS From our database we identified 33 patients with ductal adenocarcinoma of the pancreas who had initially been diagnosed as unresectable (at other institutions) but were re-explored in the course of their disease. Patient characteristics, initial procedures, subsequent palliative or neoadjuvant therapies, operative details of the re-exploration, morbidity, and mortality (death within 30 days after surgery) were prospectively recorded. Patients were followed up until February 2007. At the time of analysis, two out of 14 patients in the group of unresectable patients and 11 out of 19 patients in the resection group were alive. 1 19 For the preoperative imaging of operable patients, computed tomography of the abdomen or magnetic resonance imaging/magnetic resonance cholangiopancreaticography (MRI/MRCP) was the standard. Statistical Analysis P Literature Review 20 20 RESULTS 1 2 Table 1. Patient characteristics  Unresectable patients ( n  = 15) Resectable patients ( n  = 18) Age 60 (41–68) 61 (33–71) Female 3 4 Male 12 14 Initial operation   Explorative laparotomy 4 8   Gastroenterostomy 5 0   Double bypass 3 5   Lymph node biopsy 1 0   Resection of abdominal wall tumor 1 0   Laparoscopy 0 2   Hepaticojejunostomy 1 3 Interval between surgeries (median) 88 days 101 days Preoperative tumor markers   CA19-9 (median) 608 U/l 117 U/l   CEA (median) 39143 U/l 136 U/l Diabetes mellitus 3 6 Neoadjuvant therapy 12 13 Table 2. Criteria for initial inoperability  Unresectable patients ( n  = 15) Resectable patients ( n  = 18) Peritoneal metastasis 4 0 Vascular infiltration 6 9 Liver metastasis 3 1 Duodenal infiltration 1 1 Lymph node infiltration 1 4 Other/unknown 0 3 4 3 2 1 Table 3. Surgical procedures at second operation, morbidity and mortality  Unresectable patients ( n  = 15) Resectable patients ( n  = 18) Exploration with biopsy 7 – Double bypass 3 – Gastroenterostomy 3 – Hepaticojejunostomy 2 – Pancreaticoduodenectomy – 5 Pylorus-preserving pancreaticoduodenectomy – 10 Total pancreatectomy – 2 Left resection – 1 Duration of operation (min; median) 105 440 Blood loss (ml; median) 100 500 Morbidity 2 3 Delayed gastric emptying 1 0 Bilioma 1 0 Wound dehiscence 0 1 Cholangitis 0 1 Lymph fistula 0 1 Mortality 0 1 Length of stay (days; median) 11 12 Table 4. Criteria for unresectability at reoperation  Unresectable patients ( n  = 15) Peritoneal metastasis 6 Liver metastasis 5 Omental metastasis 1 Peritoneal and liver metastasis 2 Retroperitoneal infiltration 1 Table 5. Review of studies on re-exploration for pancreatic cancer Author Year Institution Patients Resection rate Moosa 1979 University of Chicago 24 17/24 (71%) Jones 1985 University of Toronto 50 N/A Hashimi 1989 Bradford Royal Infirmary 26 11/26 (42%) McGuire 1991 Johns Hopkins Medical Institutions 55 33/55 (60%) Tyler 1994 M.D. Anderson Cancer Center 19 14/19 (74%) Robinson 1996 M.D. Anderson Cancer Center 29 29/29 (100%) Johnstone 1996 Naval Medical Center, San Diego 29 16/29 (55%) Sohn 1999 Johns Hopkins Medical Institutions 78 52/78 (67%) Chao 2000 Fox Chase Cancer Center 40 22/40 (55%) Shukla 2005 Tata Memorial Hospital, India 15 15/15 (100%) This series 2007 University of Heidelberg 33 18/33 (55%) 3 3 3 Pathology reporting revealed that a T3 tumor was present in most of the cases. Only one patient was found with a T2 tumor. Nine pancreatic specimens (50%) were node positive, with an average of 21.5 harvested lymph nodes. A grade 1 tumor was found in only two patients, whereas a grade 2 tumor was present in nine patients and a grade 3 pancreatic cancer was present in seven patients. P P 1 FIG. 1. . n n P  P 2 P 3 FIG 2. n P  FIG. 3. n n P  7 Review of Retrospective Studies 9 18 9 11 10 12 16 P 13 P 14 15 18 17 DISCUSSION This single-center experience shows that re-exploration of patients initially deemed unresectable may be an option for the treatment of pancreatic cancer. In concordance with previously published studies (which were all retrospectively conducted), re-exploration can yield resectability rates greater than 50% and can be performed with low morbidity and mortality. However, whether reoperation confers a survival benefit is still a matter of discussion. 21 22 3 7 8 12 13 15 17 23 In conclusion, reoperation for pancreatic cancer after initial classification of unresectability revealed resectability in half of patients. Since the majority had received chemotherapy/chemoradiation after the initial operation, the concept of selecting patients by neoadjuvant therapy may be supported. Furthermore, a subgroup analysis revealed that a large number of patients who were initially deemed unresectable at smaller hospitals would have been resectable at a large tertiary referral center, which again promotes the concept of patient centralization in pancreatic surgery. A survival benefit for the resected patients underlines the efficacy of a surgical resection even in a situation in which the initial findings preclude a potentially curative approach.