Objective: To retest the hypothesis that imprint cytology may be used to reliably diagnose parathyroid tissue and, if so, to ascertain whether accuracy in this technique may be easily attained.
Study design: Imprint preparations from 15 parathyroid, 10 thyroid, 8 lymphoreticular and 2 adipose tissue specimens were assessed blindly by two pathologists, one of whom (pathologist B) had only limited experience with endocrine tissue imprint cytology.
Results: Both assessors consistently distinguished parathyroid and thyroid preparations from lymphoreticular and adipose tissue preparations. While there was occasional difficulty in distinguishing between parathyroid and thyroid preparations, this was usually attributable to the scanty nature of the preparations. No single cytologic feature allowed a distinction between parathyroid and thyroid tissue. However, by considering several relatively diagnostic features collectively, pathologist B showed an increase in specificity and sensitivity rates for distinguishing parathyroid from thyroid imprints from 82% to 100% and 57% to 83%, respectively.
Conclusion: The high accuracy rates and rapid [table: see text] learning curve shown by imprint cytology in distinguishing between different neck or mediastinal tissue types, together with its time- and cost-cutting potential, support a role for the technique in the intraoperative diagnosis of parathyroid tissue.