Endocrine Abstracts (2019) 65 P280 | DOI: 10.1530/endoabs.65.P280

Prognostic indicators of metastatic neuroendocrine tumour of unknown primary site: a single centre retrospective study

Katherine English1, Ultan Healy1, Michael Tadman1, Rajesh V Thakker2, Denis Talbot3, Dayakshi Abeyaratne1, Michael Silva1, Zahir Soonawalla1, Eve Fryer1, Brian Shine1, Ashley B Grossman1, Aparna Pal1, Christine May1, Andrew Weaver1 & Bahram Jafar-Mohammadi1


1Oxford University Hospital, Oxford, UK; 2Academic Endocrine Unit, Radcliffe Department of Medicine, University of Oxford, Oxford, UK; 3Department of Oncology, Radcliffe Department of Medicine, University of Oxford, Oxford, UK


Neuroendocrine tumours (NETs) presenting as metastatic cancer of unknown primary site (CUP) are suspected to confer poorer prognosis compared to metastatic NETs of known primary site. We performed a retrospective, single centre study to determine the prognostic indicators in CUP-NETs compared to metastatic small intestinal NET (SiNET), before and after adjusting for factors known to affect overall survival. Subjects were selected from a departmental database of 1050 NET patients discussed by the Oxford neuroendocrine service between 2011 and 2019. Inclusion criteria were histologically proven NET with radiological evidence of metastatic disease at diagnosis. Survival time began from the date of histological diagnosis until the last known follow-up. The primary end-point was death. Patients were divided into 3 cohorts: 1) CUP-NET, no primary identified; 2) likely SiNET, radiological evidence of mesenteric/SiNET, and 3) histologically confirmed SiNET. Cox proportional hazards models were constructed to compare unadjusted and adjusted hazard ratios (HR) for age, tumour differentiation [well vs. poor], grade and primary tumour resection [likely SiNET and SiNET] between each group: 233/1050 (21%) patients with metastatic NET (median follow-up of 22.5 months [95% CI 16−29, IQR 9−55]) were identified including 52/1050 (5%) CUP-NET, 66/1050 (6%) likely SiNET and 105/1050 (10%) SiNET. Unadjusted mean (standard error) overall survival data were 19 (2.4), 64 (5.8) and 102 (5.6) months respectively (P<0.02). Table 1 reports the unadjusted and adjusted HR for overall survival between each group. In conclusion, CUP-NET confers a poorer prognosis compared to metastatic SiNET. However, this difference is largely driven by patient age, tumour grade and differentiation.

Table 1 Unadjusted and adjusted hazard ratios (HR) for all cause mortality
GroupsUnadjusted HR (95% CI)Adjusted HR (95% CI)
CUP-NET to likely SiNET3.1 (1.5−6.3)0.8 (0.3−2.1)
CUP-NET to SiNET6.5 (3.3−13.0)1.8 (0.6−5.6)
Likely SiNET to SiNET2.5 (1.4−5.7)1.6 (0.2−13.8)