Endocrine Abstracts

Diagnosis and localisation of insulinoma: the value of modern MRI in conjunction with calcium stimulation catheterisation

Vasantha M Muthuppalaniappan¹, Maralyn R Druce¹, Benjamin O’Leary¹, Shern L Chew¹, William M Drake¹, John P Monson¹, Scott A Akker¹, Michael Besser¹, Anju Sahdev², Andrea Rockall², Soumil Vyas⁴, Matthew Matson², Daniel Berney³, Satya Bhattacharya⁴ & Ashley B Grossman¹

¹Department of Endocrinology, Barts and the London Medical School, St Bartholomew’s Hospital, London, UK; ²Department of Radiology, Barts and the London Medical School, St Bartholomew’s Hospital, London, UK; ³Department of Histopathology, Barts and the London Medical School, St Bartholomew’s Hospital, London, UK; ⁴Department of Surgery, Barts and the London Medical School, St Bartholomew’s Hospital, London, UK.

---

Objective: To review the diagnostic features and localization accuracy of different investigations for insulinomas diagnosed 1990–2009 at a single tertiary referral centre.

Design: A cross-sectional, restrospective analysis, including sporadic tumours and those in multiple endocrine neoplasia syndromes.

Methods: Case notes and investigation results were reviewed from patients with biochemically or histologically-proven insulinoma. The accuracies of tumour localization by computed tomography (CT), magnetic resonance imaging (MRI), octreotide scanning, endoscopic ultrasound (EuS) and calcium stimulation test (CaStim) were compared to histologically-confirmed location following surgical excision.

Results: Thirty-seven cases were identified, in 36 patients: 32 sporadic cases (including one recurrence) and 5 in MEN-1 syndrome. Data from 34 prolonged supervised fasts revealed one negative result, while the majority were terminated (glucose <2.2 mmol/l) within 24 h.

Of 30 patients managed surgically, 23 (76.6%) had CT while 27 (90%) had MRI and 20 (66.7%) patients underwent both types of imaging. 15 (65.2%) lesions were correctly localized by CT and 20 (74.1%) by MRI. When considered together, such cross-sectional imaging correctly localised 80% of lesions. Inclusion of medically managed patients in the localisation analysis (excluding those with obvious metastatic disease for whom localisation was evident) results in an amended localisation rate for MRI of 20 out of 31 patients (65%).

Radiolabelled octreotide scanning was used in 20 cases and was positive in 11 cases (55%) although the technique was useful for localisation in just over half of these. Data were available for EuS in 21 patients; 12 lesions (57%) were correctly identified. Twenty-seven patients had CaStim. Six (22%) did not localise, 17 (63%) correctly localised and 4 (15%) gave discordant or confusing results.

Conclusion: MRI, combined pragmatically with CT, is valuable for insulinoma localization. CaStim remains valuable for functional confirmation and endoscopic imaging and octreotide scanning may be useful in selected cases.

---

Endocrine Abstracts (2010) 21 OC2.8