An Algorithmic Approach to Sellar Region Masses

Abstract

Context Most sellar region masses (85%–90%) are pituitary adenomas; however, other neoplasms or even inflammatory or cystic nonneoplastic lesions may occasionally be encountered in this location. A practical, non–electron-microscopically based approach is essential for the daily practice of diagnosing and subclassifying adenomatous and nonadenomatous sellar region lesions. Objective To provide an algorithmic approach to sellar region masses for the pathologist and to formulate a cost-effective, limited panel of stains and immunostains that can be used in daily practice at most small to medium-sized centers. Design Pool collective experience of 3 neuropathologists practicing at academic medical centers with expertise in diagnosis and treatment of sellar region masses to craft a single-page algorithmic diagram and to liberally illustrate the range of lesions present in the sellar region. Results After formulating a differential diagnosis, the general pathologist can generate a confident final diagnosis of adenoma using 1 histochemical (reticulin) and 1 immunohistochemical (synaptophysin) stain, supplemented by 5 immunohistochemical stains (CAM5.2, follicle-stimulating hormone, growth hormone, prolactin, and adrenocorticotropic hormone), which provide subtyping of the adenoma in the overwhelming majority of examples. CAM5.2 and clinical information further help identify clinically aggressive variants such as sparsely granulated growth hormone adenomas and silent adrenocorticotropic hormone adenomas, respectively. MIB-1, thyroid transcription factor 1, and S-100 protein can be of further assistance in select cases where increased mitotic activity or possible nonadenomatous spindle cell lesions are suspected. Conclusions Adenomas, normal anterior or posterior gland, and nonadenomatous masses can be easily diagnosed in a nontertiary pathology laboratory setting.