Chromogranin A levels to predict survival in resectable pancreatic neuroendocrine tumors.

Abstract

e15775 Background: To determine whether Chromogranin A (CGA) predicts outcome after Cytoreductive Surgery (CRS) for Pancreatic Neuroendocrine Tumors (PNETs). CGA is used by clinicians to monitor progression of disease and response to treatment, but has not been objectively evaluated. Methods: A retrospective study of patients with PNETs seen at the Georgia Cancer Center, formerly the Medical College of Georgia, from 1980 to 2011 was performed. All patients who had CGA concentrations measured were evaluated (n = 126). Data relating to diagnosis, therapies, and survival outcome were noted. Patients with resectable and unresectable disease were included. Those with unresectable disease where evaluated for receiving Octreotide treatment. Results: The initial patient pool yielded a group (n = 118) with 72% (n = 85) having undergone Cytoreductive Surgery (CRS). Within the CRS group decreased survival was noted in patients with an elevated CGA level 4 months post-operatively (p = 0.002) and an elevated median CGA level of > 224ng/ml correlated with a shorter survival time (p = < 0.005). Increased survival was seen in CRS patients with median CGA levels of < 200ng/ml between 4-12 months post-operatively with a 4 year survival of 91.08%. The median time to death was 80.27 months. Elevated CGA levels in CRS patients correlated with a survival probability of 63.25% at 4 years with a median death of 48.25 months. Overall 28% of the total patient group had unresectable disease (n = 33): 61% of this subgroup (n = 20) received Octreotide therapy and exhibited significant decrease in CGA level at a median of 3.6 months (p = < 0.024); no correlation existed between CGA levels and survival. Conclusions: PNETs amenable to resection by Cytoreductive Surgery (CRS) and had a median Chromogranin A (CGA) level < 200 ng/ml measured post-operatively after 4 months showed increased survival. No correlation existed between CGA levels and survival in patients with unresected disease despite Octreotide therapy lowering CGA levels. As such, CGA is predictive of survival in patients with resectable PNETs and worthy of prospective study.