Pituitary adenomas may remain intrasellar or infiltrate dura and bone. Invasive adenomas are not considered to be malignant; in biological behavior they are between non-infiltrative adenomas and pituitary carcinomas. The latter are defined as tumors with subarachnoid, brain, or systemic metastasis. Invasion may be defined radiologically, operatively, or histologically. On the basis of operatively assessed tumor size and gross invasion of dura and bone as well as immunocytochemical and ultrastructural analysis of 365 pituitary adenomas, the following data were obtained. There were 23 growth hormone (GH)-cell adenomas: 14% microadenomas and 86% macroadenomas; their overall frequency of invasion was 50%. There were 24 prolactin (PRL)-cell adenomas: 33% microadenomas and 67% macroadenomas, with an overall frequency of invasion of 52%. Mixed GH-cell and PRL-cell adenomas were found in 35 cases; 26% were microadenomas and 74% were macroadenomas, and the overall frequency of invasion was 31%. Sixty patients had adrenocorticotropic hormone (ACTH)-cell adenomas (Cushing's disease): 87% microadenomas and 13% macroadenomas; the overall frequency of invasion was 25% (in 8% of microadenomas and 62% of macroadenomas). Twenty patients had ACTH-cell adenomas (Nelson's syndrome): 30% microadenomas and 70% macroadenomas; the overall frequency of invasion in these cases was 50% (in 17% of microadenomas and 64% of macroadenomas). Silent ACTH-cell adenomas, 100% macroadenomas, were found in 11 patients, with an 82% frequency of invasion. There were 32 follicle-stimulating and luteinizing hormone adenomas, all macroadenomas, with a frequency of invasion of 21%. Four patients had thyroid-stimulating hormone adenomas, all macroadenomas, with a 75% frequency of invasion. Null-cell adenomas were found in 93 cases: 2% microadenomas and 98% macroadenomas, with a frequency of invasion of 42%. There were 63 plurihormonal adenomas (GH, PRL, glycoprotein): 25% microadenomas and 75% macroadenomas, with a 50% overall frequency of invasion. Based on this study, and on their usual frequency of occurrence, the estimated rate of gross invasion by pituitary adenomas of all types is approximately 35%. It is concluded that immunocytochemical and ultrastructural characteristics of pituitary adenomas reflect the tendency of these tumors to infiltrate and hence may be of prognostic significance.

Little has been written about the long-term results of transsphenoidal treatment for clinically nonfunctioning pituitary adenomas. The records of 100 patients who had undergone a transsphenoidal procedure for excision of such tumors were reviewed. Immunocytology for pituitary hormones was performed in all cases. The group consisted predominantly of null-cell adenomas, although a small number of prolactinomas and gonadotropic tumors were found. The mean diameter of the tumors at the time of detection was slightly more than 2 cm. In most cases, the presenting symptoms were due to the mass effect of the tumor (that is, visual symptoms in 72 patients, hypopituitarism in 61, headache in 36, and cranial nerve disturbance other than visual loss in 10). Radiation therapy was recommended for patients in whom subtotal removal of the adenoma was expected. Six patients developed symptomatic tumor recurrence, and 10 patients demonstrated radiographic recurrence during the 48 to 100 months (mean 73.4 months) of follow-up observation. Only three of 10 deaths during the follow-up period were due to pituitary disease or treatment.

Twenty patients with a novel, frequently aggressive type of pituitary adenoma, termed silent subtype 3 adenoma on the basis of fine structural criteria, are reported. The surgically removed tumors were studied by morphological techniques, and the findings were correlated with clinical and biochemical data. All tumors were macroadenomas, often with parasellar extension. The histologically diffuse tumors frequently exhibited focal immunopositivity for one or more adenohypophysial hormones, although the majority of adenoma cells were negative. The tumors had characteristic electron microscopic features, assuring specific diagnosis and delineating this tumor type as a distinct ultrastructural entity. The tumors were removed from 9 women and 11 men (median ages, 27 and 41 yr, respectively). In all women, mild to moderate hyperprolactinemia and its sequelae were present from the early phase of the disease, leading to the erroneous diagnosis of prolactinoma. Bromocriptine therapy (3 patients) reduced serum PRL levels to normal, but failed to halt tumor growth. In men, most adenomas were nonfunctioning; 4 men had mild to moderate hyperprolactinemia. Three men had elevated serum GH levels and acromegaly, suggestive of multidirectional differentiation. Although the putative cell type giving rise to silent subtype 3 adenomas is not known, the tumor should be recognized to avoid erroneous diagnosis and inappropriate treatment.