Continuous glucose monitoring improves the diagnosis of insulinoma and surveillance after endoscopic ablation therapy

Abstract Background Insulinomas are the most common functioning pancreatic neuroendocrine tumours (PNET) and are classically diagnosed according to Whipple’s triad. The gold-standard diagnostic test (72 hour fasting glucose/C-peptide) is difficult for patients, and risks mis-diagnosing patients who only experience post-prandial symptoms. Moreover, the advent of endoscopic ablative techniques means that confirming successful treatment can be challenging when compared with surgical resection. This study aimed to investigate the value of continuous glucose monitoring (CGM) to: a) confirm the diagnosis of insulinoma; b) guide symptom management whilst awaiting definitive treatment; and c) as a surveillance tool in patients electing for non-surgical interventions. Methods A 61-year-old healthy male presented with post-prandial sweating, vacant episodes and reduced exercise tolerance which resolved upon consuming glucose. Computed Tomography demonstrated a 2cm exophytic hypervascular lesion in the pancreatic body, consistent with a NET, and correlated with an Octreotide avid lesion upon SPECT/CT. Overnight fasting glucose/c-peptide was not confirmatory. The target glucose range was defined as 4.0-9.9 mmol/L and critical hypoglycaemia as ≤2.9mmol/L. Data from CGM were analysed (Time in range and critical hypoglycaemia [Chi2]; glycaemic control and glycaemic variability [T-test]) during diagnosis, in the first week post-ablation, and the recovery phase beyond 3 weeks. Results Data from 22,008 glucose levels were analysed over a 90 day period. Time in range during diagnosis was 67.25%, 82.8% (p<0.0001) post-ablation, and 98.9% (P<0.0001) during recovery. Critical hypoglycaemia occurred in 5.2% of measurements during diagnosis, 1.9% (p<0.0001) post-ablation, and 0.1% (p<0.0001) during recovery. Glucose levels (mean [SD] mmol/L) were 4.4 (0.47) during diagnosis, 4.9 (0.29) (p=0.0001) post-ablation, and 5.7 (0.42) (p<0.0001) during recovery. Glycaemic variability (mean [SD] %CV) was 19.4 (5.09) during diagnosis, 20.17 (5.60) post-ablation (p=0.630), and 11.3 (5.53) (p<0.0001) during recovery. Temporal changes identifying post-prandial hypoglycaemia during diagnosis, and resolution during recovery, are illustrated (Figure 1). Conclusions CGM has utility to accurately diagnose insulinoma when standard tests are negative, and to improve quality of life by alerting patients to life-threatening critical hypoglycaemia. As parenchymal preserving non-operative therapies increase in popularity, especially for patients with small PNETs, sensitive, reliable, and acceptable surveillance techniques are required in order to confirm treatment success and monitor for recurrence. Proactive engagement with commissioning and regulatory bodies is necessary in order to harness the potential from emerging technologies, so that PNET patients can benefit in a similar manner as those with type 1 Diabetes Mellitus.