Stereotactic Radiosurgery for Pituitary Adenomas

Introduction

This chapter discusses the use of stereotactic radiosurgery (SRS) in the treatment of secretory and nonfunctional pituitary adenomas (NFPAs). SRS refers to several modalities that deliver precisely targeted, high-energy photons or protons to target tissue. This energy is delivered as conformal radiation therapy in one or very few fractions with strict control over dosing of adjacent tissue. The main technologies used include Gamma Knife (GK), CyberKnife (CK), linear accelerators (LINACs), and proton beams.

SRS for pituitary adenomas is most often used as an adjuvant to surgical resection and medical management to induce biochemical remission of endocrinologically active adenomas or to halt tumor progression in NFPAs ( Figs. 91.1 and 91.2 ). Other modalities of irradiation, such as fractionated radiosurgery (FSR), have also been used and continue to play specific roles in certain clinical scenarios. In many situations, however, the precise outcomes of SRS are still incompletely understood, mostly owing to the lack of data with long-term follow-up. Therefore the care of patients with pituitary adenomas necessitates an interdisciplinary approach with collaboration between endocrinologists, neurosurgeons, radiation oncologists, and ophthalmologists. The approaches, outcomes, and complications of SRS in secretory and nonfunctional adenomas of the pituitary are discussed in the following paragraphs.

Adrenocorticotropic Hormone–Producing Tumors (Cushing Disease)

Cushing disease, a cause of Cushing syndrome due to adrenocorticotropic (ACTH)-producing pituitary adenomas, is associated with significant morbidity and premature death. Up to 30% of patients may experience persistent or recurrent disease after transsphenoidal surgery. Most studies define remission as normalization of 24-hour urinary free cortisol (UFC) in the absence of medical suppression. Remission rates after radiosurgery reported in the literature vary widely, from 17% to 83%, because of the varying indications for radiosurgery, dosing, definitions of remission, and follow-up length. The time to remission also varies widely, from 2 months to 8 years, although the majority occur within 2 years of radiosurgery.

Sheehan et al. reviewed the GK experience in 96 patients with Cushing disease treated between 1990 and 2012, all of whom had at least 12 months of clinical and radiographic follow-up. The indication for GK in all patients was persistent elevation of 24-hour UFC following surgical resection. Patients presenting with Nelson syndrome, characterized by aggressive growth of the residual ACTH-secreting adenoma due to reduced feedback inhibition by cortisol typically seen after bilateral adrenalectomy, were not included in this study. The mean follow-up post-GK was 48 months (range 12 to 209.8 months), the mean marginal dose was 22 Gy, and the mean maximal dose was 47.2 Gy. Remission was achieved in 70% of patients at an average time of 16.6 months (range 1 to 165.7 months). The authors also analyzed the impact of ketoconazole administration at the time of GK on GK outcomes in the subset of patients taking the drug. Patients receiving GK after 2010 discontinued ketoconazole 2 to 3 weeks before radiosurgery and then resumed taking it 2 to 6 weeks postoperatively to control hypercortisolism. ^, This contrasted with patients receiving GK before 2010, who continued to receive ketoconazole before, during, and after radiosurgery. Interestingly, patients who temporarily stopped ketoconazole had a significantly shorter median time to remission than those who did not stop taking the drug (12.6 months vs. 21.8 months, respectively). Among patients with tumors evident on magnetic resonance imaging (MRI), tumor volume decreased in 70%, remained constant in 28%, and increased in 2% of cases. Of the patients who experienced biochemical remission, 15.6% suffered relapse at a mean time of 38 months (range 5 to 120.7 months) after remission. In patients without an apparent adenoma on MRI, the entire sella was irradiated and there was no difference in remission outcomes between these patients and those with discrete adenomas. Radiosurgical complications included hormone deficiencies in 36% of patients and cranial nerve (CN) deficits in 5.2%, with CN II and CN III most commonly affected. The risk of hypopituitarism was found to be significantly higher in cases where the entire sella was targeted. However, the authors of that study believe that targeting the entire sella is still preferable to a bilateral adrenalectomy to minimize risk of corticotroph tumor growth syndrome. ^, Of note, this preference is not shared by all centers.

In the aforementioned study, no correlation between either maximum dose, marginal dose, or treatment volume and biochemical remission or tumor control was found, consistent with findings from Castinetti et al. In two other studies, ^, however, the authors found the maximum dose and prescription isodose volume to be significantly correlated with hormone response. It is worth noting that when hormone response is analyzed, these two studies lumped all functioning adenomas together; therefore the specific applicability of the conclusions to Cushing disease may be brought into question. Indeed, as mentioned previously, functioning adenomas show differential responses to irradiation as reflected by their differential response rates. A recent study by Mehta et al. found maximum treatment doses (mean 45.3 Gy) and reduced margins to be significant predictors of lower recurrence rates on univariate analysis, although this was not significant in multivariate analysis. To maximize the likelihood of treatment success, some authors ^, advocate the use of more than 20 Gy at the 50% prescription isodose line when it is safe to do so.

A recent systematic review and meta-analysis by Dabrh et al. examined rates and predictors of remission as well as recurrence following SRS in Cushing disease. The authors included 24 studies with 571 patients, of whom 383 had prior transsphenoidal surgery. The overall biochemical remission rate at least 6 months after radiosurgery was 68%, and the overall recurrence rate was 32% at latest follow-up. Factors increasing remission rate included shorter follow-up (higher at >12-month follow-up vs. <12-month follow-up as well as >24-month vs. <24-month follow-up), higher radiation dose, age (higher in children vs. adults, based on a single study), different/multiple versus same/single surgeons, and negative history of transsphenoidal surgery. Factors that increased recurrence included female sex and lower radiation dose (≤45 Gy vs. >45 Gy). There was insufficient evidence to determine whether other predictors—including dural invasion, center experience, and history of MRI use to visualize the adenoma—affected remission or recurrence. The authors noted overall that SRS was associated with reasonable remission rates after a follow-up of at least 6 months, although it also had clinically significant recurrence rates (with no specified follow-up period) that should be considered in individualized treatment planning and prognostication. They noted that the overall quality of the evidence was low owing to the lack of controlled studies, heterogeneity between studies, and imprecision in meta-analysis.

Relapse of hypercortisolism after initial remission is an important problem in Cushing disease; in some series ^{, ,} it occurs in approximately 20% of patients who had apparently been cured by radiosurgery (based on the studies’ criteria). In a study by Jagannathan et al., recurrence occurred at a mean time of 27 months after radiosurgery. Of the seven patients who experienced recurrences, three had their disease brought into remission again after a second radiosurgery treatment. Castinetti et al. observed a 50% cure rate in 18 patients with Cushing disease and recurrence in 2 patients (22%) at 6 and 8 years after initial treatment. Therefore long-term endocrinologic follow-up is recommended.

Patients undergoing bilateral adrenalectomy due to persistent Cushing disease risk the development of corticotroph adenoma growth, often referred to as Nelson syndrome. Periodic imaging and monitoring of plasma ACTH levels have enabled the early detection and treatment of these tumors. Clinically significant growth is rare in irradiated patients but may occur in up to 50% of nonirradiated patients within 3 years of adrenalectomy. However, few studies have been conducted to specifically investigate the safety and efficacy of SRS for Nelson syndrome.

Growth Hormone–Producing Tumors

Acromegaly is a hormonal disorder resulting from excess growth hormone (GH), typically due to a GH-producing pituitary adenoma. Biochemical remission for GH–producing pituitary adenomas is commonly defined as normalization of age-adjusted serum level of insulin-like growth factor-1 (IGF-1) in the absence of medical therapy as well as a basal GH level of less than 2.5 μg/L or less than 0.4 μg/L after glucose challenge (see Figs. 91.1A and B ). ^{, ,} In general, GH-secreting adenomas are particularly resistant to the effects of radiation. In a retrospective study of 136 patients with GH-secreting pituitary adenomas who received SRS as treatment for residual or recurrent disease following transsphenoidal surgery, biochemical remission was achieved in 67% after a median time of 29 months postradiosurgery. The median marginal dose at the 50% prescription isodose line was 25.0 Gy, the median maximal dose was 50.0 Gy, and patients who went into remission after therapy were significantly more likely to have received higher marginal and maximal doses based on univariate analysis. Based on multivariate analysis, patients who went into remission were also more likely to have lower pre-SRS serum IGF-1 levels than those who did not achieve remission (531 vs. 673 ng/dL, P = .002). After SRS, tumor volume decreased in 47% of patients, remained constant in 51.5%, and grew in 1.5%. This reflected an overall tumor control rate of 98.5% and mean follow-up of 61.5 months. New-onset hormonal deficiencies occurred in 31.6% of patients after SRS, with median time to development of 50.5 months after SRS. The most common deficiencies were hypothyroidism and hypogonadism. Marginal doses greater than 25 Gy and greater tumor volumes were significantly associated with post-SRS hormonal deficiencies. Of the 136 patients, 4 developed deficits in their visual fields after SRS, but all 4 had extension of tumor into suprasellar or cavernous sinus regions.

Some studies have shown that lower basal GH levels are associated with an increased rate of cure after radiosurgery. ^, Losa et al., using multivariate analysis, found an increased likelihood of remission when basal GH levels were below 7.0 μg/L and serum IGF-1 levels were below 1.83 times of upper normal limit, with hazard ratios of 2.7 (95% confidence interval [CI] = 1.4 to 5.3) and 2.6 (95% CI = 1.3 to 5.1), respectively. This finding has since been replicated by Castinetti et al. in a series of 43 acromegaly patients with a mean follow-up of 102 months (42% remission). However, a large study by Lee et al. did not observe outcome differences based on initial GH levels.

A recent systematic review and meta-analysis by Dabrh et al., comparing SRS with conventional fractionated radiotherapy in treating acromegaly, analyzed trends in remission, IGF-1, GH, and complication rates between the two modalities. The authors included 30 studies with 2464 total patients. They found that overall, SRS was associated with a nonsignificant trend of increased remission rate at latest follow-up versus conventional radiation therapy (52% versus 36%). SRS was also associated with significantly lower follow-up IGF-1 levels but not GH levels versus conventional radiotherapy. There was no difference between all-cause mortality, hypothyroidism, or hypoadrenalism between SRS and conventional radiation. The authors noted that the overall quality of this evidence was low owing to the lack of controlled studies, heterogeneity between studies, and imprecision in meta-analysis.

Another recent review by Gheorghiu examined the use of SRS for acromegaly in 35 studies (26 with GK, 4 with LINAC, 3 with CK, and 2 with proton SRS) in 1868 patients. Overall tumor control was 93% to 100% in 33 studies, and weighted mean tumor control was 98% at a median follow-up of 59 months. The overall biochemical remission rate was 44% to 52% at 5 years with a median dose of 23.5 Gy; the weighted mean biochemical control rate was 44.3% at median follow-up of 59 months. There was significant variability in normalization rate of GH and IGF-1 serum levels after 5 or more years of median follow-up, ranging from 12% to 68% of patients. This normalization rate increased up to 47% to 86% 10 years after radiosurgery. The author notes that this significant heterogeneity may have occurred owing to different definitions of hormonal normalization, different follow-up durations, different baseline hormone levels, differences in tumor size, and differences in overall treatment regimens between studies. Prognostic factors increasing hormonal remission after SRS included higher marginal radiation dose, higher maximum radiation dose, and lower initial GH and IGF-1 levels. ^, The review also summarized SRS outcomes by treatment modality. The author concluded, based on current data, that all photon SRS techniques achieve similar results. Proton SRS was noted to achieve results similar to those of photon SRS.

There is emerging evidence that somatostatin analogs such as octreotide or lanreotide, when present at the time of radiosurgery, have a detrimental effect on achieving biochemical remission. In the initial study by Landolt et al., 9 patients treated with octreotide at the time of radiosurgery were compared with 22 patients who did not receive octreotide but had similar baseline characteristics with respect to age, sex, GH and IGF-1 levels, treatment volume, and radiation dose. Although patients who did not receive octreotide achieved a remission rate of 60%, patients who did receive octreotide achieved a remission rate of only 20% at 3-year follow-up. The radioprotective effect of octreotide has since been observed in other studies as well, ^, albeit inconsistently. ^, It is hypothesized that antisecretory agents such as octreotide may exert a radioprotective effect by decreasing cell cycling, thereby rendering cells less vulnerable to radiation-induced DNA damage. Based on these data, it is recommended that patients undergoing SRS for GH-releasing pituitary adenomas should discontinue long-acting somatostatin analogs at least a few months prior to radiosurgery and not resume them until 6 weeks afterward. It is important to note, however, that patients in these studies were not randomized and therefore selection biases, such as patients on octreotide being likely to have more severe and intractable disease, cannot be excluded.