- Email: [email protected]
518 Preoperative 18-FDG PET Predicts Survival in Resectable Pancreatic Cancer
516
518
The Incidence of Pancreatogenic Diabetes After Major Partial Pancreatic Resection May Be Greater Than You Think Richard A. Burkhart, Susan M. Gerber, Renee Tholey, Kathleen Lamb, Anitha Somasundaram, Caitlin McIntyre, Eliza Fradkin, Annie Ashok, Robert Felte, Jaya Mehta, Ernest L. Rosato, Harish Lavu, Serge A. Jabbour, Charles J. Yeo, Jordan M. Winter
Preoperative 18-FDG PET Predicts Survival in Resectable Pancreatic Cancer Mario Gruppo, Valentina Beltrame, Enrico Dalla Bona, Sergio Bissoli, Sergio Pedrazzoli, Stefano Merigliano, Cosimo Sperti Context The identification of prognostic factors useful in selecting patients with pancreatic cancer who will benefit from surgery or other treatments, is still debating. Preliminary reports showed that 18-Fluorodeoxyglucose Positron Emission Tomography (18-FDG PET) was predictive of prognosis in patients resected on for pancreatic adenocarcinoma (Sperti et al , J.Gastrointest Surg., 2003). Objective The aim of this study is to evaluate the role of 18 FDG PET as a prognostic factor for patients who underwent resection for pancreatic cancer. Methods From January 2004 to July 2011, a total of 110 patients who underwent resection for pancreatic cancer, were examined with 18-FDG PET (with-CT acquisition) in their preoperative work-up. The standardized uptake value (SUV) of 18 FDG was calculated and the patients were divided into high (> 4.0) and low ( ≤4.0) SUV groups. They were also evaluated according to the tumor node status, stage (TNM), tumor grade, radicality of resection, and serum CA 19-9 levels . Results Fifty-four cancers showed low and 56 high SUVs. Disease free survival (DFS) was significantly influenced by tumor stage (p=0.003), node status (p=0.003), radicality of resection (p=0.004), SUV (p=0.0001), CA 19-9 (p= 0.001) and grading (p=0.04). Multivariate analysis showed that only stage (p=0.02) and CA 19-9 (p=0.008) were independent predictors of DFS. Overall survival was significantly influenced by node status (p=0.001), radicality of resection (p=0.001), stage (p=0.0001), SUV (p=0.001), CA 19-9 (p=0.0001) and grading (p=0.008). Multivariate analysis showed that stage (p=0.004), CA 19-9 (p=0.001), SUV (p=0.004) and radicality (p=0.02) were independent predictors of survival. When patients analyzed for SUV were stratified according to stage, FDG uptake was related to DFS (p=0.01) and overall survival (p=0.001) in tumor's stage I-II. Stratification for CA 19-9 levels, showed that SUV was related to DFS (P=0.01) and overall survival (p=0.002) in patients with high tumor marker levels. In these patients, multivariate analysis confirmed that SUV was independent predictor of survival. Conclusion This study confirm that tumor stage and CA 19-9 serum levels are the strongest independent factors influencing disease-free and overall survival after resection for pancreatic cancer. SUV, calculated with 18 FDG PET, is an independent predictor of overall survival, and an important prognostic parameter in stage I-II pancreatic carcinoma and may be useful in selecting patients for neo-adjuvant therapy.
Background: The number of pancreatic resections performed each year in the United States according to the Nationwide Inpatient Sample is growing (currently around 5000). Nearly half of the patients undergoing resection (many with benign disease) will experience prolonged survival (>2 years), yet the long-term risk of pancreatogenic (type III) diabetes remains unknown. Previous estimates from small studies outside of the U.S. suggest a risk between 10-20%. Methods: After IRB approval, 1107 patients were identified who underwent pancreatectomy at a single institution from 2005 to 2012. Living patients were contacted by telephone, and pre- and postoperative diagnoses of diabetes mellitus (DM) were confirmed using a verbal questionnaire. Only individuals who completed the survey were included in the study. Results: Calls were made to all 691 living patients who underwent partial pancreatectomy. DM-specific information was successfully obtained for 257 patients (23% of the total cohort), including 179 pancreaticoduodenectomies (PD) and 78 distal pancreatectomies (DP). In the PD group: 44 (25%) patients reported having DM prior to resection (median onset 7 years prior, range 0.1 to 30), with 10 (6%) reporting onset within 1 year of resection. Of the group carrying a pre-operative diagnosis of DM, 3 (7%) had improved glucose control after resection (dose reduction in DM medicines), while 21 (48%) required escalated doses/medicines. Of 135 patients without preoperative DM, 24 (18%) were newly diagnosed with DM postoperatively (median onset 7.5 months postoperatively, range 1-64). In the DP group: 23 patients (29%) had DM preoperatively, with just 4 having onset within 1 year of resection (5%). No patients had improved glucose control after resection, while 6 (26%) patients carrying a preoperative diagnosis of DM had worse control after resection. Out of 55 patients without preoperative DM, 17 (31%) developed new onset DM after resection (median 6 months, range 0-60). In the total cohort of patients who developed DM after resection (n=41), the most common pathologic diagnoses at the time of resection were ductal adenocarcinoma (36%), IPMN adenoma (28%), and chronic pancreatitis (8%). Out of 190 patients without a diagnosis of preoperative DM, a high preoperative HgbA1C level (> 6%) was associated with an increased risk of developing new-onset DM postresection (38% of patients in the >6% group vs. 11% in the ≤6% group, p<0.02). Conclusions: Twenty-six percent of patients undergoing partial pancreatectomy experience new onset DM (22%) and/or worsening DM after resection (40%), a higher percentage of patients than previously reported. With a median follow-up time of 2.1 years after resection in this patient cohort, these figures may still underestimate the incidence of pancreatogenic diabetes. A high preoperative HgbA1C >6% identifies patients at highest risk. Incidence of perioperative DM with partial pancreatectomy
519
Background: Laparoscopic low anterior resection (LLAR) is considered an acceptable approach to rectal cancer. Due to limited oncologic and patient outcomes data, however, robotic low anterior resection (RLAR) remains more experimental. This study compared 30day outcomes and pathologic surrogates of oncologic results between rectal cancer patients undergoing either RLAR or LLAR. Methods: All rectal cancer patients in the National Cancer Database (NCDB) undergoing RLAR or LLAR from 2010-2011 were stratified by operative approach. The NCDB collects data from more than 1,500 cancer centers across the United States and Puerto Rico, and is estimated to capture approximately 70 percent of all newly diagnosed cases of colorectal cancer. Predictors of RLAR were modeled with multivariable logistic regression. Groups were matched on propensity to undergo RLAR using a 2:1 nearestneighbor matching algorithm. Primary oncologic endpoints included lymph node (LN) retrieval and margin status, while secondary 30-day outcomes were mortality, hospital length of stay, and unplanned readmission rates. Group characteristics were compared pre- and post-matching using Pearson's chi-square and one-way ANOVA. Results: 6,403 patients met inclusion criteria, of which 956 (14.9%) underwent RLAR. Baseline characteristics between groups were highly similar, except that RLAR patients were more likely to be treated at academic centers, to receive neoadjuvant therapy, had higher T-stage, and longer time from diagnosis to surgery (all p values <.001). Neoadjuvant therapy and treatment at an academic/ research center remained the only significant predictor of robotic use following multivariable adjustment. Following propensity adjustment, RLAR was associated with reduced risk of conversion to open (9.5 vs. 16.4%, p<.001). There were no significant differences in LN retrieval, margin status, 30-day mortality, unplanned readmission, or hospital length of stay (Table). Conclusions: In the largest series to date using a nationwide clinical database, we demonstrated equivalent perioperative safety and patient outcomes for robotic compared to laparoscopic LAR in the setting of rectal cancer. While the robotic approach required significantly fewer conversions to open, surrogates for proper oncologic surgery were nearly identical between the two approaches, suggesting that a robotic approach to colorectal cancer may be a suitable alternative. Further studies comparing long-term cancer recurrence and survival should be performed.
DM=diabetes mellitus, PD=pancreaticoduodenectomy, DP=distal pancreatectomy 517 Timing of Surgical Intervention After Percutaneous Catheter Drainage (PCD) in Step Up Approach of Severe Acute Pancreatitis Sunil D. Shenvi, Rajesh Gupta, Rajinder Singh, Madhu Khullar, Mandeep Kang, Surinder S. Rana, Deepak K. Bhasin Aims and Objectives -To determine appropriate timing of surgical intervention after PCD in infected pancreatic necrosis (IPN), as a part of step up approach and to see change in morbidity and mortality after changing the interval of surgery after PCD. Materials and methods - Randomized controlled trial was tried to find out the optimal timings of surgery following PCD in a patients with IPN who are not improving significantly within one week of PCD. The trail has to be stopped prematurely because of difficulty in recruitment of patients and apparent increase in the mortality in patients undergoing surgery early following PCD compared to patients who were managed with continued treatment of PCD. Following this rest 32 patients were managed by extended treatment with PCD in prospective manner and weekly inflammatory and nutritional markers were monitored in these patients. Surgery was performed in these patients when required. Out of 40 patients managed prospectively, 36 patients who underwent continued treatment with PCD were analyzed. Results - Duration between first PCD and surgery ranged from 10- 58 days with a mean of 41±16.15days and median of 43 days. The efficacy of Extended treatment PCD alone in achieving complete recovery in patients with infected pancreatic necrosis was 68.5%.Overall mortality in the present study was 15.1%.Disease specific morality in PCD alone group was 5.5%. Disease specific mortality in surgery group was 42.8%.ICU stay and number of Extrapancreatic necrosis in posterior pararenal space were predicting need of surgery on univariate and multivariate analysis. Need of mechanical ventilation and organ failure at admission had high odd's ratio when factors affecting mortality were compared in patients managed by PCD alone and patients managed by surgery after PCD. Serial estimation of inflammatory and nutritional markers does not help in predicting the timing of surgery. Conclusion- Early surgery following PCD results in high mortality in our randomized controlled study. Timing of surgery cannot be fixed and needs to be tailored depending on patient's response to extended treatment of PCD.
S-1019
SSAT Abstracts
SSAT Abstracts
Robotic Low Anterior Resection for Rectal Cancer and Short-Term Oncologic Outcomes Paul J. Speicher, Brian R. Englum, Asvin M. Ganapathi, Christopher R. Mantyh, John Migaly
518
The Incidence of Pancreatogenic Diabetes After Major Partial Pancreatic Resection May Be Greater Than You Think Richard A. Burkhart, Susan M. Gerber, Renee Tholey, Kathleen Lamb, Anitha Somasundaram, Caitlin McIntyre, Eliza Fradkin, Annie Ashok, Robert Felte, Jaya Mehta, Ernest L. Rosato, Harish Lavu, Serge A. Jabbour, Charles J. Yeo, Jordan M. Winter
Preoperative 18-FDG PET Predicts Survival in Resectable Pancreatic Cancer Mario Gruppo, Valentina Beltrame, Enrico Dalla Bona, Sergio Bissoli, Sergio Pedrazzoli, Stefano Merigliano, Cosimo Sperti Context The identification of prognostic factors useful in selecting patients with pancreatic cancer who will benefit from surgery or other treatments, is still debating. Preliminary reports showed that 18-Fluorodeoxyglucose Positron Emission Tomography (18-FDG PET) was predictive of prognosis in patients resected on for pancreatic adenocarcinoma (Sperti et al , J.Gastrointest Surg., 2003). Objective The aim of this study is to evaluate the role of 18 FDG PET as a prognostic factor for patients who underwent resection for pancreatic cancer. Methods From January 2004 to July 2011, a total of 110 patients who underwent resection for pancreatic cancer, were examined with 18-FDG PET (with-CT acquisition) in their preoperative work-up. The standardized uptake value (SUV) of 18 FDG was calculated and the patients were divided into high (> 4.0) and low ( ≤4.0) SUV groups. They were also evaluated according to the tumor node status, stage (TNM), tumor grade, radicality of resection, and serum CA 19-9 levels . Results Fifty-four cancers showed low and 56 high SUVs. Disease free survival (DFS) was significantly influenced by tumor stage (p=0.003), node status (p=0.003), radicality of resection (p=0.004), SUV (p=0.0001), CA 19-9 (p= 0.001) and grading (p=0.04). Multivariate analysis showed that only stage (p=0.02) and CA 19-9 (p=0.008) were independent predictors of DFS. Overall survival was significantly influenced by node status (p=0.001), radicality of resection (p=0.001), stage (p=0.0001), SUV (p=0.001), CA 19-9 (p=0.0001) and grading (p=0.008). Multivariate analysis showed that stage (p=0.004), CA 19-9 (p=0.001), SUV (p=0.004) and radicality (p=0.02) were independent predictors of survival. When patients analyzed for SUV were stratified according to stage, FDG uptake was related to DFS (p=0.01) and overall survival (p=0.001) in tumor's stage I-II. Stratification for CA 19-9 levels, showed that SUV was related to DFS (P=0.01) and overall survival (p=0.002) in patients with high tumor marker levels. In these patients, multivariate analysis confirmed that SUV was independent predictor of survival. Conclusion This study confirm that tumor stage and CA 19-9 serum levels are the strongest independent factors influencing disease-free and overall survival after resection for pancreatic cancer. SUV, calculated with 18 FDG PET, is an independent predictor of overall survival, and an important prognostic parameter in stage I-II pancreatic carcinoma and may be useful in selecting patients for neo-adjuvant therapy.
Background: The number of pancreatic resections performed each year in the United States according to the Nationwide Inpatient Sample is growing (currently around 5000). Nearly half of the patients undergoing resection (many with benign disease) will experience prolonged survival (>2 years), yet the long-term risk of pancreatogenic (type III) diabetes remains unknown. Previous estimates from small studies outside of the U.S. suggest a risk between 10-20%. Methods: After IRB approval, 1107 patients were identified who underwent pancreatectomy at a single institution from 2005 to 2012. Living patients were contacted by telephone, and pre- and postoperative diagnoses of diabetes mellitus (DM) were confirmed using a verbal questionnaire. Only individuals who completed the survey were included in the study. Results: Calls were made to all 691 living patients who underwent partial pancreatectomy. DM-specific information was successfully obtained for 257 patients (23% of the total cohort), including 179 pancreaticoduodenectomies (PD) and 78 distal pancreatectomies (DP). In the PD group: 44 (25%) patients reported having DM prior to resection (median onset 7 years prior, range 0.1 to 30), with 10 (6%) reporting onset within 1 year of resection. Of the group carrying a pre-operative diagnosis of DM, 3 (7%) had improved glucose control after resection (dose reduction in DM medicines), while 21 (48%) required escalated doses/medicines. Of 135 patients without preoperative DM, 24 (18%) were newly diagnosed with DM postoperatively (median onset 7.5 months postoperatively, range 1-64). In the DP group: 23 patients (29%) had DM preoperatively, with just 4 having onset within 1 year of resection (5%). No patients had improved glucose control after resection, while 6 (26%) patients carrying a preoperative diagnosis of DM had worse control after resection. Out of 55 patients without preoperative DM, 17 (31%) developed new onset DM after resection (median 6 months, range 0-60). In the total cohort of patients who developed DM after resection (n=41), the most common pathologic diagnoses at the time of resection were ductal adenocarcinoma (36%), IPMN adenoma (28%), and chronic pancreatitis (8%). Out of 190 patients without a diagnosis of preoperative DM, a high preoperative HgbA1C level (> 6%) was associated with an increased risk of developing new-onset DM postresection (38% of patients in the >6% group vs. 11% in the ≤6% group, p<0.02). Conclusions: Twenty-six percent of patients undergoing partial pancreatectomy experience new onset DM (22%) and/or worsening DM after resection (40%), a higher percentage of patients than previously reported. With a median follow-up time of 2.1 years after resection in this patient cohort, these figures may still underestimate the incidence of pancreatogenic diabetes. A high preoperative HgbA1C >6% identifies patients at highest risk. Incidence of perioperative DM with partial pancreatectomy
519
Background: Laparoscopic low anterior resection (LLAR) is considered an acceptable approach to rectal cancer. Due to limited oncologic and patient outcomes data, however, robotic low anterior resection (RLAR) remains more experimental. This study compared 30day outcomes and pathologic surrogates of oncologic results between rectal cancer patients undergoing either RLAR or LLAR. Methods: All rectal cancer patients in the National Cancer Database (NCDB) undergoing RLAR or LLAR from 2010-2011 were stratified by operative approach. The NCDB collects data from more than 1,500 cancer centers across the United States and Puerto Rico, and is estimated to capture approximately 70 percent of all newly diagnosed cases of colorectal cancer. Predictors of RLAR were modeled with multivariable logistic regression. Groups were matched on propensity to undergo RLAR using a 2:1 nearestneighbor matching algorithm. Primary oncologic endpoints included lymph node (LN) retrieval and margin status, while secondary 30-day outcomes were mortality, hospital length of stay, and unplanned readmission rates. Group characteristics were compared pre- and post-matching using Pearson's chi-square and one-way ANOVA. Results: 6,403 patients met inclusion criteria, of which 956 (14.9%) underwent RLAR. Baseline characteristics between groups were highly similar, except that RLAR patients were more likely to be treated at academic centers, to receive neoadjuvant therapy, had higher T-stage, and longer time from diagnosis to surgery (all p values <.001). Neoadjuvant therapy and treatment at an academic/ research center remained the only significant predictor of robotic use following multivariable adjustment. Following propensity adjustment, RLAR was associated with reduced risk of conversion to open (9.5 vs. 16.4%, p<.001). There were no significant differences in LN retrieval, margin status, 30-day mortality, unplanned readmission, or hospital length of stay (Table). Conclusions: In the largest series to date using a nationwide clinical database, we demonstrated equivalent perioperative safety and patient outcomes for robotic compared to laparoscopic LAR in the setting of rectal cancer. While the robotic approach required significantly fewer conversions to open, surrogates for proper oncologic surgery were nearly identical between the two approaches, suggesting that a robotic approach to colorectal cancer may be a suitable alternative. Further studies comparing long-term cancer recurrence and survival should be performed.
DM=diabetes mellitus, PD=pancreaticoduodenectomy, DP=distal pancreatectomy 517 Timing of Surgical Intervention After Percutaneous Catheter Drainage (PCD) in Step Up Approach of Severe Acute Pancreatitis Sunil D. Shenvi, Rajesh Gupta, Rajinder Singh, Madhu Khullar, Mandeep Kang, Surinder S. Rana, Deepak K. Bhasin Aims and Objectives -To determine appropriate timing of surgical intervention after PCD in infected pancreatic necrosis (IPN), as a part of step up approach and to see change in morbidity and mortality after changing the interval of surgery after PCD. Materials and methods - Randomized controlled trial was tried to find out the optimal timings of surgery following PCD in a patients with IPN who are not improving significantly within one week of PCD. The trail has to be stopped prematurely because of difficulty in recruitment of patients and apparent increase in the mortality in patients undergoing surgery early following PCD compared to patients who were managed with continued treatment of PCD. Following this rest 32 patients were managed by extended treatment with PCD in prospective manner and weekly inflammatory and nutritional markers were monitored in these patients. Surgery was performed in these patients when required. Out of 40 patients managed prospectively, 36 patients who underwent continued treatment with PCD were analyzed. Results - Duration between first PCD and surgery ranged from 10- 58 days with a mean of 41±16.15days and median of 43 days. The efficacy of Extended treatment PCD alone in achieving complete recovery in patients with infected pancreatic necrosis was 68.5%.Overall mortality in the present study was 15.1%.Disease specific morality in PCD alone group was 5.5%. Disease specific mortality in surgery group was 42.8%.ICU stay and number of Extrapancreatic necrosis in posterior pararenal space were predicting need of surgery on univariate and multivariate analysis. Need of mechanical ventilation and organ failure at admission had high odd's ratio when factors affecting mortality were compared in patients managed by PCD alone and patients managed by surgery after PCD. Serial estimation of inflammatory and nutritional markers does not help in predicting the timing of surgery. Conclusion- Early surgery following PCD results in high mortality in our randomized controlled study. Timing of surgery cannot be fixed and needs to be tailored depending on patient's response to extended treatment of PCD.
S-1019
SSAT Abstracts
SSAT Abstracts
Robotic Low Anterior Resection for Rectal Cancer and Short-Term Oncologic Outcomes Paul J. Speicher, Brian R. Englum, Asvin M. Ganapathi, Christopher R. Mantyh, John Migaly