The long-term significance of microscopic dural invasion in 354 patients with pituitary adenomas treated with transsphenoidal surgery

Object. Pituitary adenomas are considered benign tumors; however, they may infiltrate surrounding tissues including the dura mater. In this paper the authors analyze the clinical significance of microscopically confirmed dural invasion by comparing a range of variables (age and sex of patients, adenoma type, adenoma size on magnetic resonance [MR] images, remission, residual pituitary disease, recurrence, survival, and disease-free interval after surgery) between patients with noninvasive adenomas and those with invasive ones. Methods. Between 1992 and 1997 dural specimens were obtained in 354 patients with pituitary adenomas who underwent transsphenoidal surgery performed by the senior author (E.R.L.). Dural specimens were examined using routine histological methods and assessed for invasion by pituitary adenoma tissue. The dura was invaded by the pituitary adenoma in 161 patients (45.5%), and in 192 patients (54.5%) no evidence of dural invasion was found. Dural invasion was present significantly more frequently in the repeated surgery group (69%, 55 patients) than in the primary transsphenoidal surgery group (41%, 291 patients). The mean age of patients undergoing primary transsphenoidal surgery was significantly older in cases of invasive adenomas (50 years) compared with cases of noninvasive adenomas (43 years), and these age differences also correlated with adenoma size. Women tend to develop clinically evident, smaller adenomas at a younger age than men. Of the patients with pituitary adenomas that were 20 mm or smaller, 117 (76%) of 154 were women, whereas of the patients with adenomas that were larger than 20 mm, 74 (54%) of 137 were men. The frequency of dural invasion increased with increasing size of the pituitary adenoma as measured on MR images. In 291 patients who underwent primary pituitary surgery, the frequency of dural invasion according to adenoma size was 24% (≤ 10 mm), 35% (> 10 to ≤ 20 mm), 55% (> 20 to ≤ 40 mm), and 70% (> 40 mm). In patients who underwent primary transsphenoidal surgery, dural invasion was present in more than 50% of those with nonfunctioning adenomas and in 30 to 35% of patients with endocrinologically active adenomas. The mean diameter of the gonadotrophic adenomas and null-cell adenomas was significantly larger than that of each of the endocrinologically active adenomas. In 58 (20%) of 291 patients who underwent primary pituitary surgery there was residual pituitary disease postsurgery, and 20% of this subset of patients showed clinical improvement to such an extent that no further management was recommended. After pituitary surgery, residual tumor tissue was demonstrable significantly more frequently in patients with invasive adenomas than in those with noninvasive adenomas. Recurrences after initial remission (cure) of pituitary disease occurred in 18 (8.8%) of 205 patients between 2 and 79 months after primary pituitary surgery (median 25 months). The recurrence rate was not related to dural invasion in a consistent or significant fashion. Seven patients died between 14 and 79 months after pituitary surgery and all had invasive adenomas identified on gross observation at surgery and on microscopy. The survival rate was slightly but significantly decreased for patients with invasive adenomas (91%) compared with patients with noninvasive adenomas (100%) at 6 years postsurgery. Conclusions. The principal significance of dural invasion by pituitary adenoma is the persistence of tumor tissue after transsphenoidal surgery (incomplete adenomectomy; 20% in primary pituitary tumor resections). The increase in adenoma size with time and the concurrent development of dural invasion are the major factors that determine an incomplete adenomectomy. When the adenoma remains restricted to the sellar compartment or shows only moderate suprasellar extension, dural invasion may not yet have developed and conditions for complete selective adenomectomy are improved.