The value of serum chromogranin A as a predictor of tumor burden, therapeutic response, and nomogram-based survival in well-moderate nonfunctional pancreatic neuroendocrine tumors with liver metastases.

OBJECTIVE

To investigate the usefulness of serum chromogranin A (CgA) for the prediction of tumor burden, therapeutic response, and nomogram-based survival in well-moderate nonfunctional pancreatic neuroendocrine tumors (NF-PNETs) with liver metastases (LMs).

MATERIALS AND METHODS

This prospective study included 51 NF-PNETs of LMs patients, 134 other neuroendocrine tumors, and 125 controls. Serum CgA levels were determined by enzyme-linked immunosorbent assay at baseline and after treatment. LMs tumor burden was computed simultaneously from computed tomography/MRI scan with thin slices using a semiquantitative three-dimensional reconstruction approach. Predictive CgA for therapeutic response was assessed using the response evaluation criteria in solid tumors criteria. A nomogram to predict the prognostic value of CgA with variables selected in the multivariate Cox proportional hazards model was constructed; the accuracy of the nomogram was quantified by the (concordance index) C-index and a calibration plot.

RESULTS

Considering NF-PNETs, CgA correlated with the tumor grade and differentiation (P<0.05). There was a nonlinear exponential regression between LMs tumor burden and CgA levels (P<0.001). The alteration in CgA correlated with therapeutic response (P<0.001). Increased CgA presented significantly lower progression-free survival than the stable/decreased CgA subgroup (P<0.001). For overall survival, a baseline CgA increase greater than 2.5 upper limit of normal level was predictive of a poor prognosis (P<0.001). Baseline CgA level, LMs tumor burden, and Ki-67 were selected as independent factors for the nomogram to predict overall survival; the nomogram showed fitting calibration with a C-index of 0.87 (95% confidence interval, 0.82-0.92).

CONCLUSION

Serum CgA could be used to reflect tumor burden, evaluate the therapeutic response, and predict the survival outcomes for NF-PNETs with LMs. An effective nomogram including CgA was proposed for prediction.