Aim: Pre-operative localisation of primary hyperparathyroidism (PHPT) has been based on ultrasound (US) of neck and MIBI scans with an expectation that solitary parathyroid adenomas (PA) would be localised with reasonable accuracy and sensitivity, to enable mini-parathyroidectomy. The aim of our audit was to assess the utility of both these imaging in patients with parathyroid adenoma.
Method: Retrospective analysis was performed on all patients who had surgery for PHPT. A list of patients who had PA proven on post-operative histology was collated. Pre-op localisation modalities on these patients were analysed.
Results: Baseline: Data on 220 surgeries were analysed. Mean age 58 years (1884); 77% were females. Post-op histology showed: Parathyroid adenoma 155 (70.5%), hyperplasia 10 (4.5%), diffuse hyperplasia 20 (9.1%), parathyroid carcinoma 3 (1.4%), normal 17 (7.7%), and others 12(6.8%). US was done in 164 (75%), MIBI in (81%), and both 149 (68%).
PA overall: Of the 168/220 who had a PA localized on US or MIBI, 130 of these were confirmed accurate after histology (77.4%); 20 had hyperplasia (11.9%); nine were normal (5.4%); six were others (3.6%); and parathyroid cancer 3 (1.7%). Sensitivity, specificity, and positive predictive value (PPV) to identify adenoma for US was 80, 36, and 78% and for MIBI 83, 40, and 80% respectively.
US and MIBI preformed: 149 patients had both US+MIBI (112 of these were PA on histology). 78/112 (70%) were concordantly identified. US identified a further 14 (12.5%) and MIBI a further 13 (11.6%). So the sensitivity of US or MIBI on their own was 82%. The sensitivity and PPV of US and/or MIBI localising an PA were 94 and 78% respectively. On 16/149 (11%) patients both US+MIBI showed false localisation when histology was not consistent with PA (five were normal).
US and MIBI negative: 14/149 (9.4%) had both scans negative: Histology showed: PA 7(50%), normal 4 (29%), and hyperplasia in three.
Conclusion: i) Sensitivity of US or MIBI is around 80%; combining the two scans enhances the chance of localisation. ii) Pre-op localisation may not be possible in 5% of patients despite the pathology being an adenoma. iii) False localisation can be an issue in about 4% of patients, even allowing for hyperplasia as an accepted confounder.