Detection rate of unknown primary tumour by using somatostatin receptor PET/CT in patients with metastatic neuroendocrine tumours: a meta-analysis.

Abstract

The high diagnostic performance of somatostatin receptor positron emission tomography with computed tomography (PET/CT) in neuroendocrine tumours (NETs) was demonstrated by several articles. However, only some studies evaluated the detection rate (DR) of this imaging method in patients with metastatic NETs and unknown primary tumours (CUP-NETs). Therefore, we aimed to perform a meta-analysis to add evidence-based data in this setting. A comprehensive computer literature search of studies listed in PubMed/MEDLINE, EMBASE, and Cochrane library databases through December 2018 and regarding the use of somatostatin receptor PET/CT in patients with CUP-NETs was carried out. Pooled DR of CUP-NETs by using somatostatin receptor PET/CT was calculated. A pooled analysis evaluating the percentage of change of management by using somatostatin receptor PET/CT in these patients was also performed. Twelve studies on the use of somatostatin receptor PET/CT in detecting CUP-NETs in 383 metastatic patients were included. The meta-analysis of all these studies provided the following DR on a per patient-based analysis: 56% (95% confidence interval (95% CI): 48-63%). Moderate heterogeneity among the selected studies was found (I2 = 50%), whereas a significant publication bias was excluded by Egger's test (p = 0.45). The most common primary tumour sites were the bowel and the pancreas. A change of management by using somatostatin receptor PET/CT was demonstrated in 20% (95% CI: 10-33%) of patients with CUP-NET. Somatostatin receptor PET/CT is very useful in detecting CUP-NETs in patients with metastatic disease. More studies on the change of management by using this imaging method in this setting are needed.