Intraoperative Differentiation of Normal Pituitary Gland from Adenoma Using Frozen Section and Touch Preparation Techniques

Introduction: Pituitary adenomas comprise approximately 90% of sellar neoplasms, and transphenoidal surgery has become a mainstay in the treatment of sellar lesions. Failure to achieve total resection of a pituitary adenoma results in higher recurrence rates with associated morbidity and mortality, and may necessitate further surgery or irradiation. Postoperative recurrence primarily occurs at tumor margins, which may be difficult to distinguish from normal gland. Thus, increasing precision of tumor margin identification intraoperatively may improve remission rates while preserving normal gland and minimizing the risk of hypopituitarism.

Accuracy rates of frozen section diagnosis in pituitary surgery are relatively low, with rates reported at approximately 80%, as opposed to those achieved at other sites of greater than 90%. Diagnosis of pituitary adenomas using frozen sections is rendered difficult by small sample sizes and significant artifact. Methods that have been cited in the literature to improve the accuracy of intraoperative diagnosis of pituitary adenomas include rapid modified reticulin staining, multiphoton microscopy, and confocal scanning microscopy, as well as ACTH-antibody-guided assays in patients with Cushing disease.

At our institution, we supplement frozen sections with intraoperative touch preparation technique to help guide pituitary adenoma resection and distinguish tumor from gland, with high cellularity as the key identifying feature of adenomas. This relatively simple and readily available method can be efficiently and effectively utilized at most institutions. Here, we compare our intraoperative results of tumor versus gland using these techniques with the final permanent pathology findings.

Methods: A single-institution, retrospective study examining the accuracy of intraoperative tumor versus gland results using both frozen section and touch preparation techniques. Combined frozen and touch preparation pathology results were compared with permanent sections to determine the accuracy, sensitivity, and specificity of our intraoperative results.

Results: A total of 381 patients who underwent endoscopic transphenoidal resections between October 2013 and May 2020 was evaluated in the study. Fifty-nine of these patients were diagnosed with a pituitary adenoma and intraoperative pathology results were used to identify tumor versus gland and guide surgical resection. Sixty-six specimens were sent for tumor versus gland analysis, of which 51 contained benign tissue and 15 consisted of tumor. Of these, 64 samples (97%) were correctly identified as tumor or gland, and 2 (3%) were incorrectly identified, one incorrectly as tumor and the other incorrectly as gland. Thus, we were able to identify tumor versus gland intraoperatively with 97% accuracy, 93.33% sensitivity, and 98% specificity.

Conclusion: Intraoperative frozen section and touch preparation are widely available and relatively simple techniques that can together be utilized to accurately distinguish between tumor and gland, and thus guide resection of pituitary adenomas to maximize tumor resection and gland preservation.

Fig. 1

Publication History

Article published online:
12 February 2021