Presurgical localization of abnormal parathyroid glands using a single injection of technetium-99m methoxyisobutylisonitrile: comparison of different techniques including factor analysis of dynamic structures.

Abstract

Single-tracer methoxyisobutylisonitrile (MIBI) imaging is considered to be a sensitive method for the localization of abnormal parathyroid glands. The aims of this study were to determine which of the analytical techniques described for this method - visual comparison of early (15-min) and late (120-min) images, use of time-activity curves (TACs) generated on regions of interest and factor analysis of dynamic structures (FADS) - corresponds best with surgical findings, and to ascertain the potential overall contribution of presurgical scintigraphy. Fifty-five patients were studied, 34 of whom presented with primary hyperparathyroidism (HPT) and 21 with secondary HPT. After a 925 MBq injection of technetium-99m MIBI, a 40-min dynamic acquisition was performed and static images were acquired at 5, 20, 40 and 120 min using a gamma camera equipped with a pinhole collimator. The dynamic series were submitted to FADS, an attractive non-operator-dependent technique, and TACs were generated on regions of interest after the visual comparison of early and 120-minute images (15'-120'). The presumed localizations of abnormal glands were compared with a sketch drawn by the surgeon. Sensitivity was defined as the percentage of true-positive localizations and was 84.4%, 74% and 65% in adenoma and 76%, 66.6% and 45% in hyperplasia for 15'-120', FADS and TACs, respectively. Surgical accuracy, i.e. the percentage of patients accurately and completely described, was 72%, 56% and 59% in adenoma and 53%, 30% and 22% in hyperplasia for 15'-120', FADS and TACs, respectively. The visual comparison method scored best in all cases. FADS was found to be sensitive in cases of adenoma but was handicapped by more false-positive localizations. TACs were particular inefficient in hyperplasia. With respect to the detection of adenomas, we found a relationship between the gland weight and scintigraphic positivity. This dependence on gland weight was not found in hyperplasia. The poorer results obtained with all techniques for surgical accuracy can be explained by the need for a complete scintigraphic description of all pathological glands found by the surgeon in a patient. This study demonstrates that the 15'-120' visual comparison method is more efficient and less cumbersome than TAC or the attractive FADS technique. However, it was less efficient than neck exploration by an experienced surgeon. Therefore, in our institution, scintigraphic studies are now only requested in selected cases of HPT, usually primary HPT and cases undergoing re-operation.