The utility of blood neuroendocrine gene transcript measurement in the diagnosis of bronchopulmonary neuroendocrine tumours and as a tool to evaluate surgical resection and disease progression.

Abstract

The management of bronchopulmonary neuroendocrine tumours (BPNETs) is difficult, since imaging, histology and biomarkers have a limited value in diagnosis, predicting outcome and defining therapeutic efficacy. We evaluated a NET multigene blood test (NETest) to diagnose BPNETs, assess disease status and evaluate surgical resection. (i) Diagnostic cohort: BP carcinoids (n = 118)-typical carcinoid, n = 67 and atypical carcinoid, n = 51; other lung NEN (large-cell neuroendocrine carcinoma and small-cell lung carcinoma, n = 13); adenocarcinoma, (n = 26); squamous cell carcinoma (n = 23); controls (n = 90) and chronic obstructive pulmonary disease (n = 18). (ii) Surgical cohort, n = 28: BP carcinoids (n = 16: typical carcinoid 12; atypical carcinoid 4); large-cell neuroendocrine carcinoma, n = 3; lung adenocarcinoma, n = 8 and squamous cell carcinoma, n = 1. Blood sampling was performed presurgery and 30 days post-surgery. Transcript levels measured by quantitative polymerase chain reaction were calculated as activity scores (0-100% scale: normal < 14%) and compared with chromogranin A (enzyme-linked immunosorbent assay; normal <109 ng/ml). NETest was significantly elevated in carcinoids (48.7 ± 27%) versus controls (6 ± 6%, P < 0.001) with metrics: sensitivity 93%, specificity 89%, positive predictive value 92% and negative predictive value 91%. NETest differentiated progressive disease (73 ± 22%) from stable disease (36 ± 19%, P < 0.001) and R0 resections (10 ± 5%, P < 0.001, area under the curve: 0.98). Levels in chronic obstructive pulmonary disease and lung cancers were 18-24% while elevated in small-cell lung carcinoma/large-cell neuroendocrine carcinoma (59 ± 10%). In BPNETs on postoperative Day 30, NETest decreased by 60% (P < 0.001). Chromogranin A was elevated in only 40% of carcinoids and not altered by surgery. Blood NET gene levels accurately identified BPNETs (100%) and differentiated these from controls, benign and malignant lung disease. Progressive disease could be identified and surgical resection verified. Chromogranin A had no clinical utility. Monitoring NET transcript levels in blood will facilitate management by detecting residual tumour and identifying progressive disease.