Prognostic Threshold for Circulating Tumor Cells in Patients With Pancreatic and Midgut Neuroendocrine Tumors

Background: Circulating tumor cells (CTCs) are detectable in patients with neuroendocrine tumors (NETs) and are accurate prognostic markers although the optimum threshold has not been defined. Objective: This work aims to define optimal prognostic CTC thresholds in PanNET and midgut NETs. Patients and Methods: CellSearch was used to enumerate CTCs in 199 patients with metastatic pancreatic (PanNET) (90) or midgut NETs (109). Patients were followed for progression-free survival (PFS) and overall survival (OS) for a minimum of 3 years or until death. Results: The area under the receiver operating characteristic curve (AUROC) for progression at 12 months in PanNETs and midgut NETs identified the optimal CTC threshold as 1 or greater and 2 or greater, respectively. In multivariate logistic regression analysis, these thresholds were predictive for 12-month progression with an odds ratio (OR) of 6.69 (P<.01) for PanNETs and 5.88 (P<.003) for midgut NETs. The same thresholds were found to be optimal for predicting death at 36 months, with an OR of 2.87 (P<.03) and 5.09 (P<.005) for PanNETs and midgut NETs, respectively. In multivariate Cox hazard regression analysis for PFS in PanNETs, 1 or greater CTC had a hazard ratio (HR) of 2.6 (P<.01), whereas 2 or greater CTCs had an HR of 2.25 (P<.01) in midgut NETs. In multivariate analysis OS in PanNETs, 1 or greater CTCs had an HR of 3.16 (P<.01) and in midgut NETs, 2 or greater CTCs had an HR of 1.73 (P<.06). Conclusions: The optimal CTC threshold to predict PFS and OS in metastatic PanNETs and midgut NETs is 1 and 2, respectively. These thresholds can be used to stratify patients in clinical practice and clinical trials.

Automated summary

The study identified the optimal CTC thresholds of 1 and 2 for predicting survival and progression-free survival in metastatic PanNET and midgut NETs, respectively. These thresholds can assist in patient stratification and clinical trial design.