Introduction: Falsely elevated serum chromogranin A (CgA) is associated with the use of proton pump inhibitors, the presence of renal impairment and systemic inflammation. We aimed to investigate which laboratory parameters are independently associated with increased CgA and to develop a nomogram, in order to improve the diagnostic accuracy of CgA in detecting patients with neuroendocrine tumors (NET).
Methods: Our retrospective study included 155 subjects (controls) and 55 treatment naïve patients with NET, with available data on CgA, other laboratory tests, medical history and antropomethric parameters. Nomogram was developed in a form of scoring system, based on z-score obtained from receiver operating curve analysis for each parameter that was independently associated with CgA.
Results: CgA was positively associated with erythrocyte sedimentation rate, red cell distribution width, serum creatinin, glucose, urine leukocyte casts and the use of proton pump inhibitors. The combination of all these parameters was associated with increased CgA with an area under the curve of 0.771 (P<0.001). Overall, CgA level of 189 had a sensitivity of 56.4% (42.369.7) and a specificity of 76.8% (69.383.2) in detecting patients with NET (area under the curve 0.656, P<0.001). In subjects with a score of <6, CgA level of 150 ng/ml had a sensitivity of 68.2% (45.186.1) and specificity of 89.4% (80.895.0), (AUC 0.767, P<0.001). In subjects with a score≥6, AUC decreased to 0.534 (P=0.538).
Conclusion: CgA should not be used as a biomarker for NET in patients with laboratory signs of inflammation and renal impairment. Our study suggests the possibility to adjust CgA in these patients in order to increase its diagnostic accuracy.
20 - 23 May 2017
European Society of Endocrinology