Azadeh Elmi1, Janet Murphy2, Sandeep Hedgire3, Shaunagh McDermott1, Seyed-Mahdi Abtahi1, Elkan Halpern4, Carlos Fernandez-Del Castillo5, Mukesh Harisinghani1. 1. Division of Abdominal Imaging and Interventional Radiology, Massachusetts General Hospital, 55 Fruit St., White 270, Boston, MA, 02114, USA. 2. Division of Hematology and Oncology, Massachusetts General Hospital, 55 Fruit St., Boston, MA, 02114, USA. 3. Division of Abdominal Imaging and Interventional Radiology, Massachusetts General Hospital, 55 Fruit St., White 270, Boston, MA, 02114, USA. Hedgire.Sandeep@mgh.harvard.edu. 4. Institute for Technology Assessment, Radiology - Data Group, Massachusetts General Hospital, Harvard Medical School, Boston, MA, 02114, USA. 5. General and Gastrointestinal Surgery Department, Massachusetts General Hospital, 55 Fruit St., Boston, MA, 02114, USA.
Abstract
PURPOSE: To compare the survival outcomes in patients with pancreatic ductal adenocarcinoma (PDAC) who had regular imaging surveillance with those who had clinical follow-up after Whipple. METHOD: We identified 229 patients, who underwent Whipple for resection of PDAC from 2005 to 2011, and had regular postoperative clinical follow-up at our hospital. Patients were retrospectively selected for two follow-up groups: imaging surveillance (IS) defined as routine imaging at scheduled intervals, vs. clinical (C) defined as imaging triggered by either change in clinical status or change in CA19-9. Follow-up was obtained through the hospital and Cancer Data Registry records. Survival was calculated from the date of surgery to death or last follow-up, with data censored as of March 13, 2013. Kaplan-Meier survival curves were compared using the log-rank test, and Cox regression models were used for multivariate analysis. RESULTS: Patients were followed for a mean period of 24.35 ± 2.56 months. IS-group underwent significantly more imaging (4.41 vs. 2.08 scans/year). The most frequent imaging was CT of chest and abdomen at 3-4 month interval. Univariate associations with overall survival were detected with post-Whipple ECOG status, T-stage, N-stage, tumor grade, surgical margin, recurrence, and IS. In multivariate analysis, grade, ECOG status, and recurrence were independent predictors of survival. Also, our predictor of interest, IS, was highly associated with longer survival in multivariate modeling (median overall survival, 30.4±3.85 (IS-group) vs. 17.1 ± 2.42 (C-groups) month, log-rank p = 0.002). CONCLUSION: Routine imaging surveillance was associated with prolonged overall survival post-Whipple in a multivariate model. This is a hypothesis-generating finding that should be studied prospectively and could ultimately impact surveillance guidelines.
PURPOSE: To compare the survival outcomes in patients with pancreatic ductal adenocarcinoma (PDAC) who had regular imaging surveillance with those who had clinical follow-up after Whipple. METHOD: We identified 229 patients, who underwent Whipple for resection of PDAC from 2005 to 2011, and had regular postoperative clinical follow-up at our hospital. Patients were retrospectively selected for two follow-up groups: imaging surveillance (IS) defined as routine imaging at scheduled intervals, vs. clinical (C) defined as imaging triggered by either change in clinical status or change in CA19-9. Follow-up was obtained through the hospital and Cancer Data Registry records. Survival was calculated from the date of surgery to death or last follow-up, with data censored as of March 13, 2013. Kaplan-Meier survival curves were compared using the log-rank test, and Cox regression models were used for multivariate analysis. RESULTS:Patients were followed for a mean period of 24.35 ± 2.56 months. IS-group underwent significantly more imaging (4.41 vs. 2.08 scans/year). The most frequent imaging was CT of chest and abdomen at 3-4 month interval. Univariate associations with overall survival were detected with post-Whipple ECOG status, T-stage, N-stage, tumor grade, surgical margin, recurrence, and IS. In multivariate analysis, grade, ECOG status, and recurrence were independent predictors of survival. Also, our predictor of interest, IS, was highly associated with longer survival in multivariate modeling (median overall survival, 30.4±3.85 (IS-group) vs. 17.1 ± 2.42 (C-groups) month, log-rank p = 0.002). CONCLUSION: Routine imaging surveillance was associated with prolonged overall survival post-Whipple in a multivariate model. This is a hypothesis-generating finding that should be studied prospectively and could ultimately impact surveillance guidelines.