Prospective evaluation of selective venous sampling for parathyroid hormone concentration in patients undergoing reoperations for primary hyperparathyroidism.

BACKGROUND

The utility of standard radiologic imaging studies in guiding reoperative parathyroid surgery for primary hyperparathyroidism is widely known and accepted. The additional information gained by selective venous sampling in that patient population has not been well defined. We report the results of our experience with this method.

METHODS

Between 1982 and 1992, 223 consecutive patients underwent reoperations for persistent or recurrent primary hyperparathyroidism after a prospectively determined series of imaging studies. Patients underwent noninvasive testing consisting of ultrasonography, computed tomography, technetium thallium scanning, and magnetic resonance imaging. Patients with negative, equivocal, or discordant results on the noninvasive studies proceeded to angiography. If angiography was negative, selective venous sampling was performed.

RESULTS

Eighty-six patients (39%) with negative or equivocal noninvasive test and angiogram results underwent selective venous sampling. Seventy-six patients (88%) had a significant gradient in levels of parathyroid hormone from veins draining the left side of the neck (n = 25), the right side of the neck (n = 33), both sides of the neck (n = 7), and the thymus (n = 11). Correlation of these findings with operative findings revealed a sensitivity of 88% and a specificity of 86%. In the subgroup of patients who underwent venous sampling and had completely negative results of standard radiologic studies (35 of 86; 40%), 28 patients (80%) had venous gradients and seven patients (20%) had no gradient. Of those 28 patients in whom the venous sampling gradients were the only positive localization study, the venous samplings were helpful in 23 patients (true positive gradients), and the operative success rate was 93%. In the seven patients with no positive preoperative localizing studies including venous sampling, there were two operative failures (operative success, 71%).

CONCLUSIONS

Our results show that selective venous sampling is a highly sensitive and specific method to regionally localize abnormal parathyroid glands not imaged by standard noninvasive and invasive radiologic techniques. Venous sampling is the study of choice in guiding reoperative procedures for occult abnormal parathyroid glands that are undetected despite the use of all available imaging studies.