Magnetic resonance imaging in primary hyperparathyroidism

Aim: The primary aim of this study is to assess the efficiency of magnetic resonance imaging in pathologic evaluation of parathyroid lesions in patients with primary hyperparathyroidism. Materials and Methods: Retrospective observation study performed by selecting subjects among patients referred to Adnan Menderes University Hospital department of radiology with the diagnosis of primary hyperparathyroidism between September 2014 and October 2016. Among 116 patients (89 female, 27 male) whom received imaging including neck MRI, USG and parathyroid scintigraphy to evaluate for etiology and localization of the lesion, 48 patients had surgical intervention with histopathological analysis of the lesions. Also 2 patients had surgical intervention due to tertiary hyperparathyroidism. In total, 50 patients (37 female, 13 male) included in the study. All 50 patients had preoperative MRI, USG and scintigraphy evaluation. Demographic information such as age and sex, laboratory information such as serum PTH and calcium levels as well as histopathologic features such as size and weight of lesions recorded. The volume of lesions calculated using the standard formula. In order to detect the success rate in identification of lesion, size and localization, imaging results from MRI, USG and Tc-99m 2-methoxyisobutylisonitrile (Tc-99 MIBI) were compared to histopathology report. For MRI imaging T1 signal, T2 signal with fat suppression, contrast enhancement pattern, lesion Wall features, volume (cm3) and length (mm) were recorded. In diffusion weighted imaging signal features were identified and ADC measurements were recorded. In ultrasonography images size, wall features, and echogenicity of the lesions were recorded as well as vascular features using Doppler Ultrasonography. The location and number of lesions were obtained using surgical records. For analysis purposes, each patient assumed to have four parathyroid glands (integrative method). For all three methods sensitivity, specificity, positive predictive and negative predictive values calculated. Statistical analysis performed using IBM SPSS Statistics program (Statistical package for Social Sciences, version 22,0, IBM American multinational technology and consulting corporation, USA). Intergroup analysis of nonparametric data performed usingMann-Whitney U test (25-75 percentile). Standard deviation of the mean was assessed for regression analysis of parametric data. Spearman correlation was performed for analysis of ordinal data. Chi square test used to assess the correlation of categorical variables. Alfa value set as 0.05 for statistical analysis( p<0.05). Results: From 50 patients who had surgical excision of parathyroid glands, 53 lesions were obtained. Among these lesions 48 were parathyroid adenoma and 5 parathyroid hyperplasia. Measurements of excised parathyroid tissues were as following: average length 17,40 ± 8,26 mm (minimum 8, maximum 47mm), volume 1,87 ± ,53 cm3 (minimum 0,12, maximum 21 cm3), weight 5,14 ±14,38 gram (minimum 0,20, maximum 92 gram). Localization of the lesions were as following: right upper 5,6%, right lower 41,5%, left upper 5,6%, left lower 39,6%. Additionally, one lesion (1,8%) located intra-parenchymal embedded in thyroid gland and 3 lesions (5,6%) in ectopic locations. For ultrasonography evaluation, sensitivity of 86,7%, positive predictive value of 92% and negative predictive value of 95,3% was calculated. With integrative method calculated sensitivity, specificity, positive predictive and negative predictive values were %86,7, %92,0, %92,0 and %95,3 respectively. For Tc99MIBI scintigraphy, sensitivity of 67,9%, positive predictive value of 100% was calculated. With integrative method calculated sensitivity, specificity, positive predictive and negative predictive values were%67,9, %100, %100 and %89,6 respectively. ROC analysis of Tc-99m MIBI method, at cut off point of 0,4 cm3sensitivity of 75%, specificity of 76,9% and at 1,24 gram cut-off point, sensitivity of 47,2% and specificity of 92,3% was calculated. Contrast-enhanced MRI had a sensitivity of 98,2% and positive predictive value of 94,5 % whereas non-contrast MRI had a sensitivity of 88,7% and positive predictive value of 94%. With integrative method calculated sensitivity, specificity, positive predictive and negative predictive values for contrast-enhanced MRI were %97,9, %98,1, %95,5 and %99,3,and for non-contrast MRI %88,6, %97,9, %94,0 and %96,0 respectively. In fat-supression T2 weighted series all lesions demonstrated hyper-intense signal while only 4 lesions demonstrated iso-intense signal pattern. In diffusion weighted images, median of mean ADC values for parathyroid adenomas was 1,37 (1,27-1,50) (10-3 mm2/s) whereas for parathyroid hyperplasias it was 0,857 (0,768-0,926) (10-3 mm2/s), and there was a statistically significant difference between these to values. Conclusion: According to the findings in this study, it is possible to conclude that MRI imaging is highly sensitive and specific in identifying parathyroid pathologies. Ultrasonography may provide an additional value in identifying lesions such as lymph nodes and thyroid nodules which can have similar signal patterns as parathyroid adenomas and may decrease the rate of false positive diagnosis which can provide benefit in presurgical evaluation of parathyroid pathologies. Therefore, in presurgical evaluation of parathyroid pathologies if there is no contraindications, MRI with contrast can be used to detect parathyroid adenoma vs hyperplasia and additional USG evaluation might decrease the false positive results which can be beneficial in surgical planning. Also in patients who have contrast allergy, non-contrast MRI would still be valuable. In this study, a significant difference between parathyroid adenoma and hyperplasia mean ADC values was demonstrated. Further studies are required to validate these findings by increasing the number of subjects as well as evaluation of images by different radiologists.