- Email: [email protected]
923 Islet Cell Autotransplantation and Morbidity After Operations for Chronic Pancreatitis
has not been reached. Algenpantucel-L was well tolerated with no grade 4 or 5 adverse events. There were nine grade 3 adverse events directly or possibly due to the immunotherapy. The most common adverse events were injection site pain and induration. Conclusion: Addition of algenpantucel-L to standard adjuvant therapy for resected pancreatic cancer may improve survival. A multi-institutional, phase 3 study began patient enrollment in May 2010. 921 Hereditary Pancreatitis: Endoscopic and Surgical Management Eugene P. Ceppa, Henry A. Pitt, JoAnna Hunter, Charles Leys, Nicholas J. Zyromski, Frederick J. Rescorla, Kumar Sandrasegaran, Evan L. Fogel, Lee McHenry, James L. Watkins, Stuart Sherman, Glen A. Lehman Introduction: Hereditary pancreatitis is a very rare cause of chronic pancreatic inflammation. In recent years the genetic mutations leading to hereditary pancreatitis have been characterized. Patients with hereditary pancreatitis present in childhood and, as adults, are reported to have an extremely high risk of pancreatic cancer. However, the rarity of this disorder has resulted in a gap in clinical knowledge. Therefore, the aims of this analysis are to characterize a large series of patients with hereditary pancreatitis and to establish clinical guidelines. Methods: Pediatric and adult endoscopic, surgical, radiologic and genetic databases were searched from 1994-2011. Patients with chronic pancreatitis and genetic mutations for PRSS-1 or SPINK-1 as well as those with a significant family history of chronic pancreatitis were included. Patients with chronic pancreatitis due to other causes, idiopathic pancreatitis without a positive family history, and familial pancreatic cancer were excluded. Data were gathered on genetic testing, endoscopic and surgical management as well as the occurrence of pancreatic cancer. Results: Eighty-six patients were identified, and the mean age at presentation was 17 years. Forty-four (51%) were female. Genetic testing confirmed the diagnosis in 38 families (44%) while 48 patients (56%) had a significant family history. In recent years adult patients were counseled to avoid tobacco and alcohol. Eighty patients (93%) with a median age of 15 years were managed endoscopically with sphincterotomy (69%), stone removal (31%) and/or stenting of pancreatic duct strictures (85%). In recent years endoscopic ultrasound was performed in 13 patients (15%) to screen for cancer. Twenty-nine patients (34%) have undergone 35 operations at a mean age of 19 years. Surgery included 15 drainage procedures (lateral pancreaticojejunostomy-9, cystenterostomy-3, Duval-2, sphincteroplasty-1) and 20 resections (proximal-10, distal-9, total-1). Islet transplantation was performed in the one patient who underwent a total pancreatectomy. While 22 patients (26%) were older than 40 years, only one 67 year old man developed pancreatic cancer and died. The remaining 85 patients are all alive and cancer free. Conclusions: Many children and young adults with hereditary pancreatitis can be managed initially with endoscopic therapy. When surgery is undertaken, the procedure should be tailored to the pancreatic anatomy and the cancer risk. With aggressive endoscopic and surgical management as well as avoidance of cocarcinogens the incidence of pancreatic cancer is surprisingly low in patients with hereditary pancreatitis.
Figure 2. * p < 0.05 vs pre-values 822a Metachronous Colorectal Cancer Risk Following Surgery for First Rectal Cancer in Mismatch Repair Gene Mutation Carriers Aung Ko Win, Susan Parry, Bryan Parry, Matthew F. Kalady, Finlay A. Macrae, Noralane M. Lindor, Robert W. Haile, Polly A. Newcomb, Loïc Le Marchand, Steven Gallinger, John Hopper, Mark A. Jenkins Background: Metachronous colorectal cancer risk for Lynch syndrome patients with primary colon cancer is high and total colectomy is the preferred option. However if the index primary cancer is in the rectum, management advice is complicated by considerations of worsening bowel function or stoma formation. To aid surgical decision-making, we estimated the risk of metachronous colon cancer for Lynch syndrome patients who underwent either anterior resection or abdominoperineal resection for primary rectal cancer. Methods: This retrospective cohort study comprised 79 MMR gene mutation carriers (18 MLH1, 55 MSH2, 4 MSH6 and 2 PMS2) from the Colon Cancer Family Registry who had a surgical resection for their first primary rectal cancer. Age-dependent cumulative risks of metachronous colon cancer were calculated using the Kaplan-Meier method. Risk factors for metachronous colon cancer were assessed using a Cox proportional hazards regression. Results: During 866 person-years of observation (median 9 years; range 1-32 years) since diagnosis of first rectal cancer, a total of 21 (27%) carriers were diagnosed with metachronous colon cancer (incidence 24.2; 95% CI 15.8-37.2 per 1000 person-years). Incidence for carriers who had an anterior resection (26.8; 95% CI 15.5-46.1 per 1000 person-years) was not different from that for carriers who had an abdominoperineal resection (21.0; 95% CI 10.5-42.1 per 1000 person-years) (P=0.1). Cumulative risk of metachronous colon cancer was 19% (95% CI 9-31%) at 10 years, 47% (95% CI 31-68%) at 20 years and 69% (95% CI 45-89%) at 30 years after surgical resection. There was no difference in the frequency of surveillance colonoscopy between the two types of surgery (one colonoscopy per 1.1 (95% CI 0.9-1.2) years after anterior resection vs. one colonoscopy per 1.4 (95% CI 1.0-1.8) years after abdominoperineal resection). Conclusions: For carriers of MMR gene mutations diagnosed with rectal cancer, the metachronous colon cancer risk is substantial and mirrors that seen for carriers who have undergone segmental resection for primary colon cancer, despite the majority continuing to receive frequent surveillance colonoscopy. Whereas total colectomy for primary colon cancer in mutation carriers is appropriate, for primary rectal cases this strategy has major implications for continence and need for stoma. Nevertheless, given the high metachronous risk, this procedure needs serious consideration especially for younger patients.
Islet Cell Autotransplantation and Morbidity After Operations for Chronic Pancreatitis John C. McAuliffe, Sandre F. McNeal, Manasi S. Kakade, Brandon A. Singletary, John D. Christein Background: Quality of life studies after pancreatic resection and islet cell autotransplantation have shown improvement and already been published. Mortality rates have improved, but morbidity remains high after pancreatic operations, in particular total pancreatectomy (TP) and pancreaticoduodenectomy (PD). Few studies have evaluated outcomes after pancreatic operations specifically for chronic pancreatitis, with or without islet cell autotransplantation (IAT), and compared these to operations for pancreatic cancer. Methods: A retrospective review for patients undergoing operation for chronic pancreatitis from 2005-2011 by a single surgeon at an academic center. Morbidity was evaluated to 90 days according to the Clavien Classification (CC). Patients undergoing pancreatic resection with IAT were evaluated as a subgroup. Both groups were compared to those undergoing similar operations for pancreatic cancer. Statistical analysis was applied. Results: Of the 200 patients (55% men, mean age 49 years), ninety-eight underwent resection alone (65 PD, 27 distal (DP) and 6 TP)), 67 underwent resection with IAT (47 TP, 18 PD, 2 DP), and 22 underwent drainage with lateral pancreaticojejunostomy (LPJ). There was no mortality; however, the overall morbidity rate was 55% (CC 1 - 5) and 29% of these experienced a more severe complication requiring intervention (CC 3-4). Severe complications (CC3 - 4) occurred more commonly after TP (29%) than DP (28%), LPJ (10%), or PD (10%) (p < 0.01). Resections with IAT did not have a higher overall (66% v. 53%) (p > 0.05) nor severe (20% v. 16%) complication rate than those without IAT (p > 0.05). Specifically looking at PD with and without IAT, length of stay (14 v. 10) and complication rate (72% v. 46%) appeared to be higher, but neither reached statistical significance (both p > 0.05). There was no difference in complication rate between TP-IAT and PD-IAT (67% v. 72%) (p > 0.05). Overall (CC1-5) and severe (CC34) complication rate was similar when all pancreatic resections with IAT (65% and 20%) and those without IAT (53% and 16%) were compared to those undergoing PD for pancreatic cancer (n = 133, 65% and 20%) (all p > 0.05). Reoperation for bleeding after IAT was not different than after PD for pancreatic cancer (p >0.05). Partial portal vein thrombosis (4%)after IAT and had no long term sequelae. Conclusions: Operations for chronic pancreatitis are well established and pose no greater risk than resections, specifically PD, for malignancy. Complication rates remain formidable and mortality rates are low. Improvements to quality of life after IAT have been documented; furthermore, the addition of IAT to resections for chronic pancreatitis adds no risk when compared to those for malignancy. At institutions with capability, IAT should be offered to patients during resection for chronic pancreatitis.
920 Addition of Algenpantucel-L Immunotherapy to Standard Adjuvant Therapy for Pancreatic Cancer: A Phase 2 Study Jeffrey M. Hardacre, Mary Mulcahy, William Small, Mark Talamonti, Jennifer Obel, Caio S. Rocha-Lima, Howard Safran, Heinz-Joseph Lenz, Elena G. Chiorean Background: Pancreatic cancer portends a poor prognosis with ~4% long-term survival. Among the estimated 20% of patients who have resectable disease, the 1/3/5-year survival rates approximate only 70%/30%/18%, even with adjuvant therapy. Better treatment options are needed, and addition of algenpantucel-L immunotherapy to standard adjuvant therapy is proposed to improve prospects for survival. Algenpantucel-L is composed of irradiated, live, allogeneic human pancreatic cancer cells expressing the enzyme α-1,3 galactosyl transferase (α-GT), which is the major barrier to xenotransplantation from lower mammals to humans (e.g., hyperacute rejection). Up to 2% of circulating human antibodies are directed against the α-GT epitope of algenpantucel-L and are the proposed mechanism of initiating the anti-tumor immune response. Methods: Open-label, dose-finding, multi-institutional Phase 2 study evaluating algenpantucel-L (100 or 300 million cells per dose) + standard adjuvant therapy (RTOG-9704, JAMA, 2008: gemcitabine + 5-FU-XRT) for pancreatic cancer patients undergoing R0/R1 resection. Disease-free survival (DFS) was the primary endpoint with overall survival (OS) and toxicity being secondary endpoints. Results: 70 patients with a 21-month median follow-up received gemcitabine + 5-FU-XRT + algenpantucel-L (mean 12 doses, range 1-14). Demographics and prognostic factors: median age 62 years, 47% female, 81% lymph node positive, median tumor size 3.2 cm (range 2-15 cm; 25% > 4cm), and 17% post-operative CA 19-9 ≥ 180. The primary endpoints of median and 12-month DFS were 14.3 months and 63%, respectively, for the entire cohort. These compare favorably to rates of 11.4 months and < 50% in historical controls treated with standard adjuvant therapy alone. Subgroup analysis showed that patients receiving 300 million cells per dose had a longer 12-month DFS compared to those receiving 100 million cells per dose, 81% vs. 52% (p = 0.02). Overall survival at one year in the entire cohort was 86% and compares favorably to 69% in historical controls. Subgroup analysis showed that patients receiving 300 million cells per dose tended toward a longer OS compared to those receiving 100 million cells per dose, 96% vs. 80% (p = 0.053). As of this analysis, median overall survival
S-1045
SSAT Abstracts
SSAT Abstracts
923
924
vascular resection had longer survival than non-NA patients who required vascular resection (23.6 months vs. 14.4 months, p<0.05). Conclusion: Many patients with locally advanced pancreatic cancer can be brought to resection through NA therapy and vascular resection with acceptable morbidity and mortality. These patients have significantly improved survival over patients deemed locally inoperable by traditional criteria.
Venous Involvement During Pancreaticoduodenectomy: Is There a Need for Redefinition of "Borderline Resectable Disease"? Kaitlyn J. Kelly, David Kooby, Alex Parikh, Clifford S. Cho, Emily Winslow, Charles R. Scoggins, Syed A. Ahmad, Robert C. Martin, Shishir K. Maithel, Hong Jin Kim, Nipun Merchant, Sharon M. Weber
SSAT Abstracts
Introduction: The consensus definition of borderline resectable pancreas cancer includes patients with any venous (SMV-PV) or limited arterial (SMA or GDA/CHA) involvement. Recent recommendations suggest that patients with borderline resectable pancreatic adenocarcinoma should receive neoadjuvant therapy to increase the likelihood of achieving R0 resection. It is established that a subset of patients with limited venous involvement can achieve R0 resection by utilizing vein resection. This study compares outcome of patients who underwent pancreaticoduodenectomy with (VR-PD) or without (PD) vein resection, and is unique because none received neoadjuvant therapy. Methods: A large, multi-institutional database of patients who underwent PD without neoadjuvant therapy was reviewed. Patients who required vein resection due to SMV-PV involvement by tumor were compared to those who underwent PD without vein resection. Results: Of 492 total patients, 70 (14%) underwent VR-PD and 422 (86%) underwent PD. There was no difference in R0 resection (66% VRPD vs. 75% PD, p=0.11) or local recurrence rate (18% VR-PD vs. 14% PD, p=0.33), at a median follow up of 16 months (range 3.0 - 129.7). There was no difference in median DFS (10.1 months VR-PD vs. 15.2 months PD, p =NS, HR 1.24 (0.94 - 1.64)). Positive margin, increased EBL, advanced tumor grade, and lymph node involvement, but not vein involvement, were independent predictors of DFS. These same factors, as well as use of adjuvant therapy, predicted OS (Table). Conclusion: This is the largest modern series examining patients with borderline resectable pancreas cancer due to SMV-PV involvement, none of whom received neoadjuvant therapy. This cohort of patients with vein involvement selected for up-front surgical resection demonstrates that oncologic outcomes, including R0 rate, local recurrence rate, and DFS, were not compromised when vein resection was required. These data suggest that up-front surgical resection is an appropriate option for patients with isolated vein involvement and calls into question the inclusion of all SMV-PV involvement within the definition of “borderline resectable disease.” Multivariate analysis of demographic, pathologic, and operative factors for patients undergoing PD for adenocarcinoma that may influence disease free (DFS) and overall survival (OS)
Neoadjuvant (NA) therapy and vascular resection can bring select patients with initially locally unresectable pancreatic cancer to pancreaticoduodenectomy with a significant survival benefit. In this series, median survival of resected NA patients was similar to resected nonNA patients (24.9 months vs. 19.3 months, p>0.05) and longer than non-NA patients aborted for locally advanced disease (24.9 months vs. 8.9 months, p<0.05). Median survival of NA patients aborted for locally advanced disease was 13.3 months. 1004 Greater Complexity of Liver Surgery is Not Associated With an Increased Incidence of Liver Related Complications Except for Bile Leak: an Experience With 2628 Consecutive Resections Robert E. Roses, Giuseppe Zimmitti, Thomas Aloia, Andreas Andreou, Steven Curley, Jean-Nicolas Vauthey Background: Advances in technique, technology and perioperative care have allowed for the more frequent performance of complex and extended hepatic resections. The purpose of this study was to determine if this increasing complexity has been accompanied by a rise in liver related complications. Methods: A large prospectively maintained single institution database of patients who underwent hepatic resection was utilized to identify the incidence of liver related complications. Patients were divided into two groups of equal size: an earlier group and a later group (surgery performed before or after 5/18/2006). Patient characteristics and perioperative factors were compared between the two groups. Abdominal collection was defined as a non-bilious fluid collection requiring drainage. Bile leak was defined as bilious drainage from the postoperative drain or incision for 3 days or more, or a bilious collection requiring drainage or reoperation. Hemorrhage was defined as bleeding requiring reoperation. Hepatic insufficiency was defined as a peak serum bilirubin level of >7mg/dL at any time postoperatively. Results: 2628 hepatic resections were performed between 1997 and 2011 with a 90 day morbidity of 37% and mortality of 2%. Comparison of case type between the later and earlier cohorts revealed an increase in complexity over time as evidenced by the greater number of re-resections (11.2% vs 4.3%; p<0.001), second stage resection(4.0% vs 0.9%; p<0.001) and extended right hepatectomies (18.4% vs 14.9%; p= .017) and increasing use of preoperative portal vein embolization (9.1% vs 5.9%; p=0.002) in the later group. Despite this, the incidence of abdominal collection (2.1% vs 3.4%; p= .031) and hemorrhage (0.3% vs 0.9%; p=.045) decreased and the incidence of hepatic insufficiency (2.6% vs 3.1%; p=0.41) remained stable. In contrast, the rate of bile leak increased (5.9% vs 3.7%; p=0.011). For the entire cohort, bile leak was associated with increased 90 day mortality (11.1% vs 1.8%; p<0.001) and increased length of stay (mean 13 vs 8 days; p<0.001). Independent predictors of bile leak included bile duct resection (OR 3.9; p<0.001), resection of >3 segments (OR 3.1; p<0.001), second stage resection (OR 2.5; p=0.019) and intraoperative blood loss> 1 liter (OR 1.9; p=0.019). Conclusions: Despite an increasing complexity of liver surgery, the rates of liver related complications have remained stable or decreased. An important exception to this is bile leak which has increased in incidence and is now a major cause of surgical morbidity. Given the strong association between bile leak and other poor outcomes, the development of novel technical strategies to reduce bile leak is indicated.
925 Neoadjuvant Therapy and Vascular Resection During Pancreaticoduodenectomy: Shifting the Survival Curve for Patients With Locally Advanced Pancreatic Cancer Joseph DiNorcia, Megan Winner, Minna K. Lee, Irene Epelboym, James A. Lee, Beth Schrope, John A. Chabot, John D. Allendorf Background: Neoadjuvant chemoradiation therapy and more aggressive surgery with vascular resection are two strategies to bring patients with locally advanced pancreatic cancer to the operating room for potential cure. Methods: We reviewed the medical records of all patients with pancreatic ductal adenocarcinoma who underwent pancreaticoduodenectomy (PD) at our institution between March 1992 and March 2011. We identified patients who received neoadjuvant (NA) therapy or required major vascular resection and evaluated demographics, operative characteristics, morbidity, mortality, and survival. Student's t- or Mann-Whitney U tests and Chi-squared or Fisher's exact tests were used to compare continuous and categorical variables, respectively. Kaplan-Meier curves and Cox proportional hazards models were used to compare survival. Results: A total of 600 patients were brought to the operating room for attempted resection. One hundred fifty-four (25.6%) had received NA therapy for initially unresectable disease whereas 446 (74.4%) were explored at presentation. One hundred twenty-four (80.5%) NA patients underwent successful PD compared to 340 (76.5%) non-NA patients. The NA patients were younger (62.7 years vs. 67.5 years, p<0.001), more often had prior resection attempts (26.6% vs. 0.9%, p<0.001), had longer median operative times (390 min vs. 328 min, p<0.001), and had higher median estimated blood loss (1500mL vs. 1000mL, p<0.001) than non-NA patients. There were no statistically significant differences in R0 resection rate (82.3% vs. 78.2%, p=0.34), median length of stay (9 days vs. 10 days, p=0.16), morbidity (50% vs. 49.4%, p=0.91), or mortality (6.5% vs. 2.7%, p=0.09) between the two groups. Sixty-one percent of NA patients versus 18.8% non-NA patients required vascular resection (p<0.001), therefore a subset analysis was performed. PD with vascular resection resulted in increased morbidity (54.9% vs. 33.7%, p<0.001) and mortality (1.8% vs. 6.3%, p=0.008), but similar R0 resection rates (74% vs. 81%, p=0.07) and lengths of stay (10 days vs. 9 days, p=0.07) compared to standard PD. Median survival of resected NA patients was similar to resected non-NA patients (24.9 months vs. 19.3 months, p>0.05) and significantly longer than non-NA patients aborted for locally advanced disease (24.9 months vs. 8.9 months, p<0.05). NA patients who required
SSAT Abstracts
Complexity of Liver Surgery
S-1046
Figure 2. * p < 0.05 vs pre-values 822a Metachronous Colorectal Cancer Risk Following Surgery for First Rectal Cancer in Mismatch Repair Gene Mutation Carriers Aung Ko Win, Susan Parry, Bryan Parry, Matthew F. Kalady, Finlay A. Macrae, Noralane M. Lindor, Robert W. Haile, Polly A. Newcomb, Loïc Le Marchand, Steven Gallinger, John Hopper, Mark A. Jenkins Background: Metachronous colorectal cancer risk for Lynch syndrome patients with primary colon cancer is high and total colectomy is the preferred option. However if the index primary cancer is in the rectum, management advice is complicated by considerations of worsening bowel function or stoma formation. To aid surgical decision-making, we estimated the risk of metachronous colon cancer for Lynch syndrome patients who underwent either anterior resection or abdominoperineal resection for primary rectal cancer. Methods: This retrospective cohort study comprised 79 MMR gene mutation carriers (18 MLH1, 55 MSH2, 4 MSH6 and 2 PMS2) from the Colon Cancer Family Registry who had a surgical resection for their first primary rectal cancer. Age-dependent cumulative risks of metachronous colon cancer were calculated using the Kaplan-Meier method. Risk factors for metachronous colon cancer were assessed using a Cox proportional hazards regression. Results: During 866 person-years of observation (median 9 years; range 1-32 years) since diagnosis of first rectal cancer, a total of 21 (27%) carriers were diagnosed with metachronous colon cancer (incidence 24.2; 95% CI 15.8-37.2 per 1000 person-years). Incidence for carriers who had an anterior resection (26.8; 95% CI 15.5-46.1 per 1000 person-years) was not different from that for carriers who had an abdominoperineal resection (21.0; 95% CI 10.5-42.1 per 1000 person-years) (P=0.1). Cumulative risk of metachronous colon cancer was 19% (95% CI 9-31%) at 10 years, 47% (95% CI 31-68%) at 20 years and 69% (95% CI 45-89%) at 30 years after surgical resection. There was no difference in the frequency of surveillance colonoscopy between the two types of surgery (one colonoscopy per 1.1 (95% CI 0.9-1.2) years after anterior resection vs. one colonoscopy per 1.4 (95% CI 1.0-1.8) years after abdominoperineal resection). Conclusions: For carriers of MMR gene mutations diagnosed with rectal cancer, the metachronous colon cancer risk is substantial and mirrors that seen for carriers who have undergone segmental resection for primary colon cancer, despite the majority continuing to receive frequent surveillance colonoscopy. Whereas total colectomy for primary colon cancer in mutation carriers is appropriate, for primary rectal cases this strategy has major implications for continence and need for stoma. Nevertheless, given the high metachronous risk, this procedure needs serious consideration especially for younger patients.
Islet Cell Autotransplantation and Morbidity After Operations for Chronic Pancreatitis John C. McAuliffe, Sandre F. McNeal, Manasi S. Kakade, Brandon A. Singletary, John D. Christein Background: Quality of life studies after pancreatic resection and islet cell autotransplantation have shown improvement and already been published. Mortality rates have improved, but morbidity remains high after pancreatic operations, in particular total pancreatectomy (TP) and pancreaticoduodenectomy (PD). Few studies have evaluated outcomes after pancreatic operations specifically for chronic pancreatitis, with or without islet cell autotransplantation (IAT), and compared these to operations for pancreatic cancer. Methods: A retrospective review for patients undergoing operation for chronic pancreatitis from 2005-2011 by a single surgeon at an academic center. Morbidity was evaluated to 90 days according to the Clavien Classification (CC). Patients undergoing pancreatic resection with IAT were evaluated as a subgroup. Both groups were compared to those undergoing similar operations for pancreatic cancer. Statistical analysis was applied. Results: Of the 200 patients (55% men, mean age 49 years), ninety-eight underwent resection alone (65 PD, 27 distal (DP) and 6 TP)), 67 underwent resection with IAT (47 TP, 18 PD, 2 DP), and 22 underwent drainage with lateral pancreaticojejunostomy (LPJ). There was no mortality; however, the overall morbidity rate was 55% (CC 1 - 5) and 29% of these experienced a more severe complication requiring intervention (CC 3-4). Severe complications (CC3 - 4) occurred more commonly after TP (29%) than DP (28%), LPJ (10%), or PD (10%) (p < 0.01). Resections with IAT did not have a higher overall (66% v. 53%) (p > 0.05) nor severe (20% v. 16%) complication rate than those without IAT (p > 0.05). Specifically looking at PD with and without IAT, length of stay (14 v. 10) and complication rate (72% v. 46%) appeared to be higher, but neither reached statistical significance (both p > 0.05). There was no difference in complication rate between TP-IAT and PD-IAT (67% v. 72%) (p > 0.05). Overall (CC1-5) and severe (CC34) complication rate was similar when all pancreatic resections with IAT (65% and 20%) and those without IAT (53% and 16%) were compared to those undergoing PD for pancreatic cancer (n = 133, 65% and 20%) (all p > 0.05). Reoperation for bleeding after IAT was not different than after PD for pancreatic cancer (p >0.05). Partial portal vein thrombosis (4%)after IAT and had no long term sequelae. Conclusions: Operations for chronic pancreatitis are well established and pose no greater risk than resections, specifically PD, for malignancy. Complication rates remain formidable and mortality rates are low. Improvements to quality of life after IAT have been documented; furthermore, the addition of IAT to resections for chronic pancreatitis adds no risk when compared to those for malignancy. At institutions with capability, IAT should be offered to patients during resection for chronic pancreatitis.
920 Addition of Algenpantucel-L Immunotherapy to Standard Adjuvant Therapy for Pancreatic Cancer: A Phase 2 Study Jeffrey M. Hardacre, Mary Mulcahy, William Small, Mark Talamonti, Jennifer Obel, Caio S. Rocha-Lima, Howard Safran, Heinz-Joseph Lenz, Elena G. Chiorean Background: Pancreatic cancer portends a poor prognosis with ~4% long-term survival. Among the estimated 20% of patients who have resectable disease, the 1/3/5-year survival rates approximate only 70%/30%/18%, even with adjuvant therapy. Better treatment options are needed, and addition of algenpantucel-L immunotherapy to standard adjuvant therapy is proposed to improve prospects for survival. Algenpantucel-L is composed of irradiated, live, allogeneic human pancreatic cancer cells expressing the enzyme α-1,3 galactosyl transferase (α-GT), which is the major barrier to xenotransplantation from lower mammals to humans (e.g., hyperacute rejection). Up to 2% of circulating human antibodies are directed against the α-GT epitope of algenpantucel-L and are the proposed mechanism of initiating the anti-tumor immune response. Methods: Open-label, dose-finding, multi-institutional Phase 2 study evaluating algenpantucel-L (100 or 300 million cells per dose) + standard adjuvant therapy (RTOG-9704, JAMA, 2008: gemcitabine + 5-FU-XRT) for pancreatic cancer patients undergoing R0/R1 resection. Disease-free survival (DFS) was the primary endpoint with overall survival (OS) and toxicity being secondary endpoints. Results: 70 patients with a 21-month median follow-up received gemcitabine + 5-FU-XRT + algenpantucel-L (mean 12 doses, range 1-14). Demographics and prognostic factors: median age 62 years, 47% female, 81% lymph node positive, median tumor size 3.2 cm (range 2-15 cm; 25% > 4cm), and 17% post-operative CA 19-9 ≥ 180. The primary endpoints of median and 12-month DFS were 14.3 months and 63%, respectively, for the entire cohort. These compare favorably to rates of 11.4 months and < 50% in historical controls treated with standard adjuvant therapy alone. Subgroup analysis showed that patients receiving 300 million cells per dose had a longer 12-month DFS compared to those receiving 100 million cells per dose, 81% vs. 52% (p = 0.02). Overall survival at one year in the entire cohort was 86% and compares favorably to 69% in historical controls. Subgroup analysis showed that patients receiving 300 million cells per dose tended toward a longer OS compared to those receiving 100 million cells per dose, 96% vs. 80% (p = 0.053). As of this analysis, median overall survival
S-1045
SSAT Abstracts
SSAT Abstracts
923
924
vascular resection had longer survival than non-NA patients who required vascular resection (23.6 months vs. 14.4 months, p<0.05). Conclusion: Many patients with locally advanced pancreatic cancer can be brought to resection through NA therapy and vascular resection with acceptable morbidity and mortality. These patients have significantly improved survival over patients deemed locally inoperable by traditional criteria.
Venous Involvement During Pancreaticoduodenectomy: Is There a Need for Redefinition of "Borderline Resectable Disease"? Kaitlyn J. Kelly, David Kooby, Alex Parikh, Clifford S. Cho, Emily Winslow, Charles R. Scoggins, Syed A. Ahmad, Robert C. Martin, Shishir K. Maithel, Hong Jin Kim, Nipun Merchant, Sharon M. Weber
SSAT Abstracts
Introduction: The consensus definition of borderline resectable pancreas cancer includes patients with any venous (SMV-PV) or limited arterial (SMA or GDA/CHA) involvement. Recent recommendations suggest that patients with borderline resectable pancreatic adenocarcinoma should receive neoadjuvant therapy to increase the likelihood of achieving R0 resection. It is established that a subset of patients with limited venous involvement can achieve R0 resection by utilizing vein resection. This study compares outcome of patients who underwent pancreaticoduodenectomy with (VR-PD) or without (PD) vein resection, and is unique because none received neoadjuvant therapy. Methods: A large, multi-institutional database of patients who underwent PD without neoadjuvant therapy was reviewed. Patients who required vein resection due to SMV-PV involvement by tumor were compared to those who underwent PD without vein resection. Results: Of 492 total patients, 70 (14%) underwent VR-PD and 422 (86%) underwent PD. There was no difference in R0 resection (66% VRPD vs. 75% PD, p=0.11) or local recurrence rate (18% VR-PD vs. 14% PD, p=0.33), at a median follow up of 16 months (range 3.0 - 129.7). There was no difference in median DFS (10.1 months VR-PD vs. 15.2 months PD, p =NS, HR 1.24 (0.94 - 1.64)). Positive margin, increased EBL, advanced tumor grade, and lymph node involvement, but not vein involvement, were independent predictors of DFS. These same factors, as well as use of adjuvant therapy, predicted OS (Table). Conclusion: This is the largest modern series examining patients with borderline resectable pancreas cancer due to SMV-PV involvement, none of whom received neoadjuvant therapy. This cohort of patients with vein involvement selected for up-front surgical resection demonstrates that oncologic outcomes, including R0 rate, local recurrence rate, and DFS, were not compromised when vein resection was required. These data suggest that up-front surgical resection is an appropriate option for patients with isolated vein involvement and calls into question the inclusion of all SMV-PV involvement within the definition of “borderline resectable disease.” Multivariate analysis of demographic, pathologic, and operative factors for patients undergoing PD for adenocarcinoma that may influence disease free (DFS) and overall survival (OS)
Neoadjuvant (NA) therapy and vascular resection can bring select patients with initially locally unresectable pancreatic cancer to pancreaticoduodenectomy with a significant survival benefit. In this series, median survival of resected NA patients was similar to resected nonNA patients (24.9 months vs. 19.3 months, p>0.05) and longer than non-NA patients aborted for locally advanced disease (24.9 months vs. 8.9 months, p<0.05). Median survival of NA patients aborted for locally advanced disease was 13.3 months. 1004 Greater Complexity of Liver Surgery is Not Associated With an Increased Incidence of Liver Related Complications Except for Bile Leak: an Experience With 2628 Consecutive Resections Robert E. Roses, Giuseppe Zimmitti, Thomas Aloia, Andreas Andreou, Steven Curley, Jean-Nicolas Vauthey Background: Advances in technique, technology and perioperative care have allowed for the more frequent performance of complex and extended hepatic resections. The purpose of this study was to determine if this increasing complexity has been accompanied by a rise in liver related complications. Methods: A large prospectively maintained single institution database of patients who underwent hepatic resection was utilized to identify the incidence of liver related complications. Patients were divided into two groups of equal size: an earlier group and a later group (surgery performed before or after 5/18/2006). Patient characteristics and perioperative factors were compared between the two groups. Abdominal collection was defined as a non-bilious fluid collection requiring drainage. Bile leak was defined as bilious drainage from the postoperative drain or incision for 3 days or more, or a bilious collection requiring drainage or reoperation. Hemorrhage was defined as bleeding requiring reoperation. Hepatic insufficiency was defined as a peak serum bilirubin level of >7mg/dL at any time postoperatively. Results: 2628 hepatic resections were performed between 1997 and 2011 with a 90 day morbidity of 37% and mortality of 2%. Comparison of case type between the later and earlier cohorts revealed an increase in complexity over time as evidenced by the greater number of re-resections (11.2% vs 4.3%; p<0.001), second stage resection(4.0% vs 0.9%; p<0.001) and extended right hepatectomies (18.4% vs 14.9%; p= .017) and increasing use of preoperative portal vein embolization (9.1% vs 5.9%; p=0.002) in the later group. Despite this, the incidence of abdominal collection (2.1% vs 3.4%; p= .031) and hemorrhage (0.3% vs 0.9%; p=.045) decreased and the incidence of hepatic insufficiency (2.6% vs 3.1%; p=0.41) remained stable. In contrast, the rate of bile leak increased (5.9% vs 3.7%; p=0.011). For the entire cohort, bile leak was associated with increased 90 day mortality (11.1% vs 1.8%; p<0.001) and increased length of stay (mean 13 vs 8 days; p<0.001). Independent predictors of bile leak included bile duct resection (OR 3.9; p<0.001), resection of >3 segments (OR 3.1; p<0.001), second stage resection (OR 2.5; p=0.019) and intraoperative blood loss> 1 liter (OR 1.9; p=0.019). Conclusions: Despite an increasing complexity of liver surgery, the rates of liver related complications have remained stable or decreased. An important exception to this is bile leak which has increased in incidence and is now a major cause of surgical morbidity. Given the strong association between bile leak and other poor outcomes, the development of novel technical strategies to reduce bile leak is indicated.
925 Neoadjuvant Therapy and Vascular Resection During Pancreaticoduodenectomy: Shifting the Survival Curve for Patients With Locally Advanced Pancreatic Cancer Joseph DiNorcia, Megan Winner, Minna K. Lee, Irene Epelboym, James A. Lee, Beth Schrope, John A. Chabot, John D. Allendorf Background: Neoadjuvant chemoradiation therapy and more aggressive surgery with vascular resection are two strategies to bring patients with locally advanced pancreatic cancer to the operating room for potential cure. Methods: We reviewed the medical records of all patients with pancreatic ductal adenocarcinoma who underwent pancreaticoduodenectomy (PD) at our institution between March 1992 and March 2011. We identified patients who received neoadjuvant (NA) therapy or required major vascular resection and evaluated demographics, operative characteristics, morbidity, mortality, and survival. Student's t- or Mann-Whitney U tests and Chi-squared or Fisher's exact tests were used to compare continuous and categorical variables, respectively. Kaplan-Meier curves and Cox proportional hazards models were used to compare survival. Results: A total of 600 patients were brought to the operating room for attempted resection. One hundred fifty-four (25.6%) had received NA therapy for initially unresectable disease whereas 446 (74.4%) were explored at presentation. One hundred twenty-four (80.5%) NA patients underwent successful PD compared to 340 (76.5%) non-NA patients. The NA patients were younger (62.7 years vs. 67.5 years, p<0.001), more often had prior resection attempts (26.6% vs. 0.9%, p<0.001), had longer median operative times (390 min vs. 328 min, p<0.001), and had higher median estimated blood loss (1500mL vs. 1000mL, p<0.001) than non-NA patients. There were no statistically significant differences in R0 resection rate (82.3% vs. 78.2%, p=0.34), median length of stay (9 days vs. 10 days, p=0.16), morbidity (50% vs. 49.4%, p=0.91), or mortality (6.5% vs. 2.7%, p=0.09) between the two groups. Sixty-one percent of NA patients versus 18.8% non-NA patients required vascular resection (p<0.001), therefore a subset analysis was performed. PD with vascular resection resulted in increased morbidity (54.9% vs. 33.7%, p<0.001) and mortality (1.8% vs. 6.3%, p=0.008), but similar R0 resection rates (74% vs. 81%, p=0.07) and lengths of stay (10 days vs. 9 days, p=0.07) compared to standard PD. Median survival of resected NA patients was similar to resected non-NA patients (24.9 months vs. 19.3 months, p>0.05) and significantly longer than non-NA patients aborted for locally advanced disease (24.9 months vs. 8.9 months, p<0.05). NA patients who required
SSAT Abstracts
Complexity of Liver Surgery
S-1046