Tumors of the sellar region are relatively common maladies of the neuroendocrine system that require long-term interdisciplinary management. Pituitary adenomas have been discovered in up to 23% of humans across autopsy and radiographic studies.1,2 Treatment paradigms depend on tumor physiology and size. Successful strategies aim to control tumor growth, decompress critical structures from mass effect, and minimize the burden of neuroendocrine impairment. Stereotactic radiosurgery (SRS) is an adjuvant therapy for pituitary adenoma patients with residual tumor burden (e.g., complete resection precluded in the setting of tumor invasion of the cavernous sinus or dura), as well as a primary treatment for those with unjustifiably high surgical morbidities.3–11 In general, SRS achieves tumor volume control rates of approximately 90%,9,12,13 with published actuarial tumor control rates of 98%, 95%, 91%, and 85% at 3, 5, 8, and 10 years of follow-up, respectively.10,14–16 New or delayed endocrinopathy remains the primary complication of SRS, occurring in 25%–40% of patients treated with older radiosurgery paradigms and closer to 20% in more modern multicenter investigations. Functioning adenomas often require higher radiosurgical doses than their hormonally quiescent counterparts, and this can alter the risk-benefit profile of long-term endocrine remission.10
Conventional fractionated radiotherapy includes a risk profile of endocrinopathy, radiation-induced tumors, internal carotid artery stenosis with or without ischemic changes, and cognitive dysfunction.17–19 The etiology of delayed endocrine dysfunction after fractionated radiotherapy remains unclear.17,18 SRS can offset some of these risks through higher precision of targeted treatment, which more effectively shields neighboring healthy brain tissue from long-term damage. The advent of more predictable differential radiation doses inside and outside of planned target volumes has resulted in greater tumor control and less hypopituitarism.11,20,21
The mechanism of delayed endocrine deficits from targeted radiation to the hypothalamic-pituitary axis (HPA) remains unclear. Radiation to normal neuroendocrine structures likely contributes to delayed hypopituitarism after SRS. One of the index investigations of point doses along the HPA by Feigl and colleagues implied that mean point doses to the pituitary stalk and gland could predict endocrine insufficiency, and suggested a dose cutoff of 5.5 Gy to the pituitary stalk to ameliorate the risk of post-SRS hypopituitarism.14,22 A prior single-institution study by the current authors demonstrated that higher doses to the pituitary stalk relative to normal gland are associated with a greater risk of endocrinopathy after SRS.23 The aim of this multicenter study by the International Radiosurgery Research Foundation (IRRF) was to evaluate the radiation tolerance of neuroanatomical structures surrounding pituitary adenomas and identify predictors of delayed hypopituitarism following SRS for these tumors. A recent systematic review of 2671 patients on outcomes and toxicities of SRS for nonfunctioning adenomas demonstrated post-SRS hypopituitarism with a random effects estimate of 21.0%.24 Secondary hypopituitarism is particularly concerning for patients susceptible to nonfunctioning pathology with a long natural history. While the report included multiple factors and treatment thresholds associated with post-SRS hypopituitarism, there is still an unmet clinical opportunity to define adaptive dosimetry parameters that are patient- and pathology-specific. This study, which to our knowledge represents the largest international cohort with multivariate dosimetry analysis and endocrine function after radiosurgery, was performed to meet this need.
Methods
Data Collection
Sixteen medical centers participating in the IRRF received approval from their respective institutional review boards to collect and submit analyses of outcomes for patients with pituitary macroadenomas who underwent SRS. The following centers contributed data to the final analysis: University of Virginia (n = 207), Taipei Veterans General Hospital (n = 75), Na Homolce Hospital (n = 63), University of Pennsylvania (n = 45), Hospital Ruber Internacional (n = 29), Humanitas Research Hospital (n = 23), NYU Langone Medical Center (n = 20), Centre Hospitalier Universitaire de Sherbrooke (n = 18), University of Pittsburgh Medical Center (n = 12), Penn State Health–Hershey Medical Center (n = 11), West Virginia University Medical Center (n = 11), and Jewish Hospital, Mayfield Clinic (n = 7).
Participating centers reviewed medical records and entered data into a spreadsheet with predefined variables created by investigators at the University of Virginia. Pooled and deidentified data were submitted for independent review by a third party before transmission to the institution of the first and senior authors for analysis. All patients with pituitary adenomas who were treated with SRS in a single session between 1997 and 2019 were evaluated for eligibility. In total, 674 patients underwent SRS for pituitary tumors. Patients with less than 1 year of follow-up, incomplete point dosimetric information, and/or CT-based SRS treatment plans were excluded. Data included patient demographics, baseline symptomatology, imaging and reports before and after resection, surgical and procedural dictations and treatment data, histopathological reports, and follow-up clinic notes.
Patient Evaluation
All patients underwent evaluation and treatment consistent with institutional protocols. Pre-SRS imaging included MRI and/or CT. Clinical information included subspecialty endocrine and neurological assessment. Pituitary function before and after SRS was ascertained via serum laboratory values, including thyroid (free thyroxine, thyroid-stimulating hormone), testes and ovaries (luteinizing hormone, follicle-stimulating hormone, testosterone), mammary (prolactin [PRL]), adrenal cortex (cortisol, 24-hour urine free cortisol for patients with Cushing’s disease), and body (growth hormone [GH], insulin-like growth factor 1 [IGF-1] levels). Additional glucose tolerance testing was performed for patients with signs and symptoms consistent with acromegaly. Formal neuro-ophthalmological assessment was performed for patients with subjective visual deficits or objective evidence of chiasmal compression on imaging. Nonfunctioning macroadenoma was defined by 1) sellar mass > 1 cm in dimension; 2) absence of significantly elevated levels of PRL (< 200 ng/ml), adrenocorticotropic hormone (ACTH), GH, IGF-1, or 24-hour urine free cortisol; and 3) absence of clinical and/or biochemical features of Cushing’s disease or acromegaly. Patients with depressed serum cortisol or thyroxine were treated with supplemental oral medications. Clinical and radiographic information from the day of the SRS procedure served as index data for relative changes during follow-up.
Clinical and Radiographic Follow-Up
Patients who underwent resection before SRS were assessed for new or worsening endocrine and neurological deficits at 2 months after surgery but prior to SRS. After SRS, all patients were followed with routine clinical assessment and imaging of the sella (typically MRI) at 6-month intervals for 1 year and then annually thereafter for 5 years. Patients continued follow-up at the discretion of their primary endocrinologist, typically on an annual or biannual basis. Endocrine assessment and neurological examination were performed concurrently with interval neuroimaging. Endocrine assays included hormones that were abnormal before SRS in addition to other endocrine axes at the discretion of the treating neuroendocrine team. Serial neurological and ophthalmological assessments were performed to measure deficits over time. Treatment response (of the tumor) was assessed by serial neuroimaging and endocrine workup.
Criteria for new-onset hypopituitarism have been well documented. For the purposes of this study, hypopituitarism was defined as deficiency of a pituitary hormone below normal limits requiring medical replacement and consistent with prior reviews of the institutional database.25,26 Normal ranges for morning and evening serum cortisol levels were 7–25 and 2–14 mg/dl, respectively.27 All imaging studies were independently reviewed by a team of neurosurgeons and neuroradiologists at the treating institution. Tumor dimension measurements were performed in the axial, sagittal, and coronal planes. The “dimensional indices” of the tumors were measured and recorded in three orthogonal planes: transverse (TR), anteroposterior (AP), and craniocaudal (CC). The volumes of the tumors were estimated by using the following formula: V = [π × (TR × AP × CC)]/6.28 Tumor control was defined by any volumetric reduction or volumetric enlargement ≤ 15% compared to baseline volume.29
SRS Technique
All SRS procedures were performed in accordance with each respective treating institution’s protocols, and all treatments were delivered in a single fraction using the Gamma Knife. The Gamma Knife model varied depending on the time and institution of treatment. Treatment planning was performed using KULA (earlier cases) and GammaPlan (more recent cases) software packages (Elekta). In general (not all institutions), on the day of the procedure, patients underwent temporary placement of a stereotactic Leksell G frame under local anesthesia with or without conscious sedation. After frame placement, high-resolution stereotactic MRI using pre- and postcontrast thin-slice (approximately 1 mm) images through the sellar region were obtained for dose and treatment planning. Fat suppression MRI and other imaging sequences were performed at the discretion of the treating clinicians. SRS was subsequently performed under the interdisciplinary care of a neurosurgeon, radiation oncologist, and medical physicist.10 Dose selection and constraints to critical structures were set by the treating teams, but they generally included a maximum dose of 8–12 Gy to the optic apparatus. Shielding techniques were employed, as needed, to limit radiation of the optic apparatus.
Point Dose Selection and Dosimetry
In accordance with prior investigations, 14 distinct points along the HPA were selected on the stereotactic planning MRI and analyzed using the GammaPlan software, as follows: left (n = 3) and right (n = 3) hypothalamus, pituitary gland (n = 3), and pituitary stalk (n = 5; Fig. 1).23 Descriptions and illustrations of these points were provided to each participating center so that data could be obtained in a consistent fashion. Each unique point consisted of zero volume and had no volume constraints. Patients with expansile tumor or persistent treatment changes that precluded clear identification of individual structures were excluded from analysis. Specific doses, along with descriptive statistics of the doses, were collected. Point dose measurement data remained blinded to outcome, including tumor control and hypopituitarism.
Point dosimetry example of treatment measures along the hypothalamus, pituitary stalk, and pituitary gland. Fourteen distinct points along the HPA were selected: three specific doses were made to both the left and right hypothalamus, three doses to the pituitary gland, and five doses to the pituitary stalk. The green crosses correspond to the points along the HP axis that were measured (e.g., the point doses). There are fewer crosses than the 14 points because only the first and last pituitary stalk (PS1, PS5) and not the others along the stalk are shown; 50 = 50% prescription isodose curve (yellow circle). LH = left hypothalamus; NP = normal pituitary gland; PS = pituitary stalk; RH = right hypothalamus. Figure is available in color online only.
Statistical Analysis
Data were grouped as either continuous (presented as median and interquartile range) or categorical (presented as frequency and percentages) variables. Time-to-event variables (e.g., development of specific hormone deficiencies) were based on the interval between SRS and the first report of corresponding pituitary dysfunction. Binary logistic regression models were used to determine predictors of new-onset hypopituitarism. Evaluated factors included patient demographics (age reported at time of SRS), tumor characteristics, margin dose, treatment volume, and the point doses along the hypothalamus, normal pituitary gland, and pituitary stalk, as well as various ratios of these doses. Statistical analyses were performed using R version 3.5.3 (2019). All statistical tests were two-sided, and a p value ≤ 0.05 was considered significant.
Results
Patient Demographics and Presenting Symptoms
After applying the exclusion criteria, 521 patients comprising the study cohort were included in the final analysis. There were approximately equal proportions of males (51.4%) and females (48.6%) with a median age of 54.0 years at the time of SRS. The median clinical and radiographic follow-up durations were 60.1 and 58.6 months, respectively. Most patients (94.4%) underwent prior transsphenoidal resection and some were treated with prior radiation (3.1%). More nonfunctioning (77.5%) than functioning (22.5%) adenomas were included in the analysis (Table 1).
Patient demographics
| Factor | Value |
|---|---|
| Male | 268 (51.4) |
| Female | 253 (48.6) |
| Age, yrs | 54.0 ± 20.0 (16.0–88.0) |
| Follow-up, mos | |
| Clinical | 60.1 (12.0–240.8) |
| Imaging | 58.6 (12.0–184.2) |
| Histological subtype | |
| Nonfunctioning | 404 (77.5) |
| Functioning | 117 (22.5) |
| Prior TSR | 492 (94.4) |
| Prior craniotomy | 39 (7.5) |
| Prior radiation | 16 (3.1) |
TSR = transsphenoidal resection.
Values are presented as number (%) of patients, median ± IQR (range), or median (range).
SRS Parameters
The median margin dose and treatment volume were 15.0 Gy and 3.2 cm3, respectively. The median margin doses were 14.5 and 25.0 Gy for nonfunctioning and functioning adenomas, respectively. The medians of three point doses to the left and right hypothalami were each 1.4 Gy. The doses to ventral structures were higher, with median doses of 7.2 Gy to five points along the pituitary stalk and 11.3 Gy to three points along the normal pituitary gland (Table 2).
SRS treatment parameters
| Factor | Median ± IQR | Range |
|---|---|---|
| Margin dose, Gy | 15.0 ± 7.0 | 10.0–25.0 |
| Isodose, % | 50.0 ± 0.0 | 32.0–75.0 |
| Treatment vol, cm3 | 3.2 ± 3.3 | 0.3–32.0 |
| Lt hypothalamus, Gy | 1.4 ± 1.9 | 0.1–13.1 |
| Rt hypothalamus, Gy | 1.4 ± 2.0 | 0.1–12.9 |
| Normal pituitary, Gy | 11.3 ± 9.2 | 0.2–48.9 |
| Pituitary stalk, Gy | 7.2 ± 5.5 | 0.7–40.3 |
Clinical and Radiographic Outcomes
Before SRS, 363 (69.7%) of 521 patients demonstrated pituitary hormone insufficiency. After SRS, 124 (23.8%) patients developed new hypopituitarism due to either SRS (22.5%) or tumor progression (1.3%). The rate of post-SRS neurological deficits was low: 12 patients (2.3%) developed new or worsening visual deficits, including vision loss and/or cranial neuropathy, and 1 patient (0.2%) developed diabetes insipidus. The tumor control rate was 93.9% (Table 3).
Outcomes
| Factor | No. of Pts | Percentage |
|---|---|---|
| Pre-SRS endocrinopathy | 363 | 69.7 |
| New endocrinopathy after radiosurgery | 124 | 23.8 |
| Endocrinopathy from tumor growth | 7 | 1.3 |
| Endocrinopathy from SRS | 117 | 22.5 |
| Tumor control | 489 | 93.9 |
| Post-SRS visual deficits | 75 | 14.4 |
| Post-SRS diabetes insipidus | 30 | 5.8 |
| SRS-induced visual deficits | 12 | 2.3 |
| SRS-induced diabetes insipidus | 1 | 0.2 |
Pt = patient.
The number and types of endocrinopathy after SRS were classified by endocrine axis. Of 124 (23.8%) patients with new endocrinopathy after SRS, the number of affected endocrine axes was 1, 2, and 3 in 86 (16.5%), 24 (4.6%), and 14 (2.7%), respectively. The most common endocrinopathy involved the thyroid axis (9.6%), followed by ACTH ± cortisol (9.4%), gonadotropin (7.9%), and GH ± IGF-1 (5.6%) deficits (Table 4).
Types of endocrinopathy after SRS
| Factor | No. of Pts | Percentage |
|---|---|---|
| New endocrinopathy after radiosurgery | 124 | 23.8 |
| Cases from tumor growth | 7 | 1.3 |
| Cases from SRS | 117 | 22.5 |
| Endocrine axes affected | ||
| 1 | 86 | 16.5 |
| 2 | 24 | 4.6 |
| ≥3 | 14 | 2.7 |
| New endocrinopathy by type* | ||
| ACTH ± cortisol | 49 | 9.4 |
| GH ± IGF-1 | 29 | 5.6 |
| Thyroid | 50 | 9.6 |
| GN | 41 | 7.9 |
| PRL | 4 | 0.8 |
GN = gonadotropin.
The sum of individual endocrinopathy types is greater than the total number of patients due to multiple new endocrinopathies in 38 patients.
Dosimetric Analysis of the HPA
Factors associated with new pituitary insufficiency after SRS were identified using logistic regression analysis. A separate analysis was performed for the number of affected endocrine axes. In the univariate logistic regression analysis, nonfunctioning adenoma (p = 0.0005), younger age (p = 0.0004), higher margin dose (p = 0.008), higher dose to the normal pituitary gland (p = 0.0001), and higher dose to the pituitary stalk (p = 0.0009) were significantly associated with post-SRS hypopituitarism (Table 5).
Univariate logistic regression analysis for new or worsening endocrinopathy after SRS
| Variable | p Value | OR | 95% CI |
|---|---|---|---|
| Sex | 0.63 | 0.90 | 0.60–1.35 |
| NFA vs other subtypes | 0.0005 | 0.45 | 0.29–0.71 |
| Age | 0.0004 | 0.97 | 0.96–0.98 |
| Margin dose | 0.008 | 1.06 | 1.01–1.11 |
| Volume | 0.95 | 1.0 | 0.94–1.05 |
| Max hypothalamus | 0.08 | 1.08 | 0.98–1.18 |
| Mean hypothalamus | 0.28 | 1.08 | 0.93–1.25 |
| Median hypothalamus | 0.57 | 1.04 | 0.89–1.21 |
| Max normal pituitary gland | 0.0001 | 1.04 | 1.01–1.06 |
| Mean normal pituitary gland | 0.00007 | 1.04 | 1.02–1.06 |
| Median normal pituitary gland | 0.00006 | 1.04 | 1.02–1.06 |
| Max pituitary stalk | 0.0009 | 1.04 | 1.01–1.07 |
| Mean pituitary stalk | 0.003 | 1.07 | 1.02–1.13 |
| Median pituitary stalk | 0.01 | 1.06 | 1.00–1.11 |
| Ratio of dose | |||
| Max stalk/max pituitary gland | 0.11 | 0.57 | 0.28–1.08 |
| Mean stalk/mean pituitary gland | 0.19 | 1.47 | 0.83–2.70 |
| Max hypothalamus/max pituitary gland | 0.26 | 0.57 | 0.19–1.32 |
| Mean stalk/mean hypothalamus | 0.12 | 0.98 | 0.95–1.005 |
| Max hypothalamus/max stalk | 0.24 | 1.74 | 0.7–4.57 |
| 10.7-Gy threshold of median dose to pituitary stalk | 0.006 | 1.77 | 1.17–2.68 |
NFA = nonfunctioning adenoma.
New or worsening post-SRS endocrinopathy was further analyzed by specific endocrine axis in a multivariate logistic regression analysis (Table 6). Higher maximum dose to the normal pituitary gland was a predictor of dysfunction of every endocrine axis ACTH (p = 0.049), GH (p = 0.0072), thyroid (p = 0.00003), and gonadotropin (p = 0.045), except for PRL. Higher maximum dose to the pituitary stalk was a predictor of thyroid hormone deficit (p = 0.00006). A threshold of median dose to the pituitary stalk for new endocrinopathy was 10.7 Gy (p = 0.006, OR 1.77).
Multivariate logistic regression analysis for new or worsening endocrinopathy, by endocrine axis, after SRS
| Variable | ACTH | GH | TH | GN | PRL |
|---|---|---|---|---|---|
| Max hypothalamus | 0.39 | 0.76 | 0.16 | 0.24 | 0.17 |
| Max pituitary stalk | 0.12 | 0.16 | 0.00006 | 0.39 | 0.98 |
| Max normal pituitary gland | 0.049 | 0.0072 | 0.00003 | 0.045 | 0.63 |
| Max hypothalamus/max normal pituitary gland | 0.79 | 0.14 | 0.39 | 0.86 | 0.096 |
| Max pituitary stalk/max normal pituitary gland | 0.78 | 0.07 | 0.46 | 0.052 | 0.5 |
TH = thyroid hormone.
In terms of analyzing thresholds separately between patients with pre-SRS endocrinopathy (n = 363, 69.7%) versus without pre-SRS endocrinopathy (n = 158, 30.3%), there were no significant differences in development of new endocrinopathy after radiosurgery (p = 0.91). The only significant differences between groups over the studied interval were post-SRS diabetes insipidus (7.2% in patients with pre-SRS endocrinopathy versus 2.5% without pre-SRS endocrinopathy, p = 0.04) and new gonadotropin deficit (6.3% in patients with pre-SRS endocrinopathy versus 11.4% without pre-SRS endocrinopathy, p = 0.03).
Discussion
Tumors of the pituitary gland can wield crippling local and systemic effects secondary to involvement of nearby central nervous system structures and the HPA.30,31 SRS is a safe and effective paradigm for controlling residual or progressive pituitary disease that cannot be cured with surgery alone.27,32–34 In rare cases, SRS can be employed as a primary treatment for pituitary adenomas in patients who are medically unfit to undergo surgery or who refuse resection.35 Despite continued therapeutic innovation, many patients suffer from an extensive cumulative lifetime risk of neuroendocrine dysfunction.19
The benefits afforded by the relative safety and precision of SRS are not novel.10,16,26,27 However, the mechanism of delayed endocrine deficits based on targeted radiation to the HPA remains unclear. The idea of dynamic gradients of radiation to normal neuroendocrine structures and their role in delayed hypopituitarism has been briefly explored. Feigl and colleagues reported a direct association between total radiation dose and pituitary hormone insufficiency in 43.5% of 108 SRS-treated pituitary adenoma patients over a median follow-up period of 6.6 years. Higher radiation point doses to the pituitary stalk and gland correlated with new hypopituitarism after treatment.14 Prior investigation by the current authors at the University of Virginia analyzed point dosimetry in 236 patients with a post-SRS hypopituitarism rate of 18.6% over a median interval of 21 months.23 In contrast to the findings by Feigl et al., radiation dose to the hypothalamus did not portend endocrine dysfunction, which may have been related to the relatively low doses delivered to that structure using contemporary multiisocentric dose planning. Interestingly, point dosimetry demonstrated the utility of dose ratios along the axis. A ratio of pituitary stalk to gland radiation dose of at least 0.8 significantly increased risk of endocrinopathy. Sicignano and colleagues reported that a threshold of 15.7 Gy to the gland and 7.3 Gy to the stalk should be followed in order to prevent new endocrinopathy.36
The current multicenter study draws from a much larger cohort of nonfunctioning and functioning pituitary adenomas that reflects more generalizable SRS treatment patterns throughout the world. The radiological tumor control rate was high (93.9%), but 22.5% of patients developed at least one new endocrinopathy due to SRS. An additional 1.3% experienced new post-SRS endocrinopathy due to tumor progression. Doses to the normal pituitary gland and stalk were independent predictors of new or worsening hypopituitarism in the multivariate analysis. Sex, tumor volume, and dose to the hypothalamus were not associated with hypopituitarism following SRS.
The relatively low median doses to the right and left hypothalami of 1.4 Gy each are consistent with prior recommendations to limit irradiation of the hypothalamus. While Feigl and colleagues did not find a significant difference between radiation doses to the hypothalamus in patients with versus those without new endocrine dysfunction after SRS, the mean hypothalamic doses associated with the presence versus absence of new hypopituitarism were 1.3 versus 0.8 Gy, respectively.14 This lack of prognostication was contextualized in the setting of lower doses that are typically delivered to the hypothalamus.37,38 Other studies of SRS for acromegaly and prolactinomas endorsed the prognostic role of radiation doses to the median eminence, though with a higher mean dose of 5.5 Gy.39 In terms of relevant pathophysiology, Chieng and colleagues reasoned that if blood flow to the hypothalamus was not disturbed in patients who developed endocrinopathy after conventional radiotherapy for nasopharyngeal carcinoma, then radiation-induced neuronal injury and not vascular injury was more likely the culprit portending negative outcomes.40 Other groups have argued that radiosensitivity may be variable among hypothalamic nuclei and that this nonuniform tolerance may underlay observations that certain hormonal axes characteristically demonstrate insufficiency before others (e.g., growth hormone) or in certain treatment contexts (e.g., hyperprolactinemia as a hallmark feature of damage to the hypothalamus after conventional cranial radiotherapy).41–43 The current study demonstrates that not all endocrine axes are equally affected in terms of delayed treatment effects, with the caveat that radiation doses delivered to the hypothalamus were not themselves predictors of new or worsening endocrinopathy.
The present study demonstrated that the maximum dose to the normal pituitary gland and stalk increased the likelihood of long-term endocrinopathy after SRS. This finding highlights the principle of shielding healthy vital structures from differential thresholds of radiation. Our results support the concept of the pituitary gland and stalk as discrete anatomical entities, and we found that a median dose threshold of 10.7 Gy delivered to the pituitary stalk protects against post-SRS endocrine dysfunction (p = 0.006, OR 1.77). Our analysis suggests that various endocrine axes have differential sensitivities to radiation. Specifically, maximum dose to the pituitary gland predicted dysfunction of every axis except for prolactin, whereas maximum dose to the pituitary stalk was only predictive of thyroid hormone deficiency. Of 124 patients with new post-SRS endocrinopathy, a steep falloff occurred from patients with 1 affected endocrine axis (16.5%) to those with 2 (4.6%) or ≥ 3 affected axes (2.6%). This implies that a treatment gradient exists with respect to both neuroanatomical terrain and specific hormone physiology. Tumor control typically requires a modest dose for most pituitary adenomas. The challenge is achieving endocrine remission in functioning adenomas, as this often requires a higher dose. In general, the primary goals of tumor control (for functioning and nonfunctioning adenomas) and endocrine remission (for functioning adenomas) should not be compromised in exchange for a lower risk of endocrinopathy. However, based on some of the findings in the current study and using appropriate dose planning techniques, it may be feasible to refine the SRS delivery to accomplish tumor control and, where necessary, endocrine remission while still offering a better change of normal endocrine function preservation.
Study Limitations
We acknowledge that this study has several limitations. Our results are subject to the systematic bias of dosimetry analysis and idiosyncratic bias of multiple treating institutions. In order to analyze unique point doses of radiation along the HPA, there is bias of subjectively determining precise treatment volumes in patients with less definitive boundaries between tumor and normal structures. The study design was retrospective, and therefore, our findings were influenced by referral biases of the participating centers and their respective clinicians. Additionally, the selection and treatment biases of contributing centers may have affected our analyses. The intervals among pituitary adenoma diagnosis, resection, and SRS were likely inconsistent across different centers. Finally, the rate of hypopituitarism may have been underreported due to inclusion of patients with less than 3 years of follow-up after SRS.44
The primary endpoint of hypopituitarism was defined relative to baseline endocrine status prior to SRS. Since the majority of patients suffered from hypopituitarism before SRS (70%), not all endocrine axes were susceptible to the same degree of risk for each patient. The differential radiosensitivity profiles of various endocrine axes warrant further investigation.25 Further analysis requires investigating dose thresholds and dose ratios in patients with versus those without preexisting endocrinopathies as well as with and without prior radiation, which may be confounding effects in terms of understanding the drivers of post-SRS endocrinopathy. However, those patients who did receive prior radiation therapy provide a real-world experience to the current study. Commonly, some patients receive radiation therapy in the community setting, demonstrate tumor and/or symptom progression, and are subsequently referred to an academic setting for SRS. In the current study, 30.3% of patients did not have pre-SRS endocrinopathy and 96.9% of patients did not have prior radiation. The development of prospective SRS registries may help to clarify long-term endocrine and radiological outcomes based on differential radiation patterns in patients with pituitary adenomas.45
Conclusions
SRS for the treatment of pituitary adenomas affords a high tumor control rate with a low risk of new or worsening endocrinopathy. Our evaluation of point dosimetry to adjacent neuroanatomical structures revealed that doses to the pituitary stalk, with a threshold of 10.7 Gy, and normal gland significantly increased the risk of post-SRS hypopituitarism. In patients with preserved pre-SRS neuroendocrine function, limiting the dose to the pituitary stalk and gland while still delivering an optimal dose to the tumor appears prudent.
Acknowledgments
We sincerely thank Lisa Baxendall, Clinical Research Manager at the University of Pittsburgh Medical Center, for her continued dedication and hard work in coordinating all data collection and participating author and center information.
Disclosures
Roman Liscak reports a consulting relationship with Elekta AB. Roberto Martinez-Alvarez reports a consulting relationship with Elekta AB. L. Dade Lunsford reports direct stock ownership in Elekta AB and consulting relationships with the Insightec DSMB. Brad E. Zacharia reports consulting relationship with Medtronic and being on the Speakers Bureau of NICO Corp.
Author Contributions
Conception and design: Pomeraniec, Xu, JP Sheehan. Acquisition of data: all authors. Analysis and interpretation of data: all authors. Drafting the article: Pomeraniec, JP Sheehan. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Pomeraniec. Statistical analysis: Pomeraniec, Xu, JP Sheehan. Administrative/technical/material support: Pomeraniec, JP Sheehan. Study supervision: Pomeraniec, JP Sheehan.
References
- 1↑
Ezzat S, Asa SL, Couldwell WT, et al. The prevalence of pituitary adenomas: a systematic review. Cancer. 2004;101(3):613–619.
- 2↑
Teramoto A, Hirakawa K, Sanno N, Osamura Y. Incidental pituitary lesions in 1,000 unselected autopsy specimens. Radiology. 1994;193(1):161–164.
- 3
Iwata H, Sato K, Tatewaki K, et al. Hypofractionated stereotactic radiotherapy with CyberKnife for nonfunctioning pituitary adenoma: high local control with low toxicity. Neuro Oncol. 2011;13(8):916–922.
- 4
Liscák R, Vladyka V, Marek J, et al. Gamma knife radiosurgery for endocrine-inactive pituitary adenomas. Acta Neurochir (Wien). 2007;149(10):999–1006.
- 5
Mingione V, Yen CP, Vance ML, et al. Gamma surgery in the treatment of nonsecretory pituitary macroadenoma. J Neurosurg. 2006;104(6):876–883.
- 6
Picozzi P, Losa M, Mortini P, et al. Radiosurgery and the prevention of regrowth of incompletely removed nonfunctioning pituitary adenomas. J Neurosurg. 2005;102(suppl):71–74.
- 7
Pollock BE, Cochran J, Natt N, et al. Gamma knife radiosurgery for patients with nonfunctioning pituitary adenomas: results from a 15-year experience. Int J Radiat Oncol Biol Phys. 2008;70(5):1325–1329.
- 8
Sheehan JP, Kondziolka D, Flickinger J, Lunsford LD. Radiosurgery for residual or recurrent nonfunctioning pituitary adenoma. J Neurosurg. 2002;97(5)(suppl):408–414.
- 9↑
Sheehan JP, Niranjan A, Sheehan JM, et al. Stereotactic radiosurgery for pituitary adenomas: an intermediate review of its safety, efficacy, and role in the neurosurgical treatment armamentarium. J Neurosurg. 2005;102(4):678–691.
- 10↑
Sheehan JP, Starke RM, Mathieu D, et al. Gamma Knife radiosurgery for the management of nonfunctioning pituitary adenomas: a multicenter study. J Neurosurg. 2013;119(2):446–456.
- 11↑
Voges J, Kocher M, Runge M, et al. Linear accelerator radiosurgery for pituitary macroadenomas: a 7-year follow-up study. Cancer. 2006;107(6):1355–1364.
- 12↑
Park KJ, Kano H, Parry PV, et al. Long-term outcomes after Gamma Knife stereotactic radiosurgery for nonfunctional pituitary adenomas. Neurosurgery. 2011;69(6):1188–1199.
- 13↑
Sheehan JP, Kondziolka D, Flickinger J, Lunsford LD. Radiosurgery for nonfunctioning pituitary adenoma. Neurosurg Focus. 2003;14(5):e9.
- 14↑
Feigl GC, Pistracher K, Berghold A, Mokry M. Pituitary insufficiency as a side effect after radiosurgery for pituitary adenomas: the role of the hypothalamus. J Neurosurg. 2010;113(suppl):153–159.
- 15
Gopalan R, Schlesinger D, Vance ML, et al. Long-term outcomes after Gamma Knife radiosurgery for patients with a nonfunctioning pituitary adenoma. Neurosurgery. 2011;69(2):284–293.
- 16↑
Pomeraniec IJ, Dallapiazza RF, Xu Z, et al. Early versus late Gamma Knife radiosurgery following transsphenoidal resection for nonfunctioning pituitary macroadenomas: a matched cohort study. J Neurosurg. 2016;125(1):202–212.
- 17↑
Brada M, Rajan B, Traish D, et al. The long-term efficacy of conservative surgery and radiotherapy in the control of pituitary adenomas. Clin Endocrinol (Oxf). 1993;38(6):571–578.
- 18↑
Minniti G, Jaffrain-Rea ML, Osti M, et al. Radiotherapy for nonfunctioning pituitary adenomas: from conventional to modern stereotactic radiation techniques. Neurosurg Rev. 2007;30(3):167–176.
- 19↑
Starke RM, Williams BJ, Jane JA Jr, Sheehan JP. Gamma Knife surgery for patients with nonfunctioning pituitary macroadenomas: predictors of tumor control, neurological deficits, and hypopituitarism. J Neurosurg. 2012;117(1):129–135.
- 20↑
Hayashi M, Izawa M, Hiyama H, et al. Gamma Knife radiosurgery for pituitary adenomas. Stereotact Funct Neurosurg. 1999;72(suppl 1):111–118.
- 21↑
Landolt AM, Lomax N. Gamma knife radiosurgery for prolactinomas. J Neurosurg. 2000;93(suppl 3):14–18.
- 22↑
Feigl GC, Bonelli CM, Berghold A, Mokry M. Effects of gamma knife radiosurgery of pituitary adenomas on pituitary function. J Neurosurg. 2002;97(5)(suppl):415–421.
- 23↑
Pomeraniec IJ, Taylor DG, Cohen-Inbar O, et al. Radiation dose to neuroanatomical structures of pituitary adenomas and the effect of Gamma Knife radiosurgery on pituitary function. J Neurosurg. 2019;132(5):1499–1506.
- 24↑
Kotecha R, Sahgal A, Rubens M, et al. Stereotactic radiosurgery for non-functioning pituitary adenomas: meta-analysis and International Stereotactic Radiosurgery Society practice opinion. Neuro Oncol. 2020;22(3):318–332.
- 25↑
Trifiletti DM, Xu Z, Dutta SW, et al. Endocrine remission after pituitary stereotactic radiosurgery: differences in rates of response for matched cohorts of Cushing disease and acromegaly patients. Int J Radiat Oncol Biol Phys. 2018;101(3):610–617.
- 26↑
Xu Z, Lee Vance M, Schlesinger D, Sheehan JP. Hypopituitarism after stereotactic radiosurgery for pituitary adenomas. Neurosurgery. 2013;72(4):630–637.
- 27↑
Cohen-Inbar O, Ramesh A, Xu Z, et al. Gamma knife radiosurgery in patients with persistent acromegaly or Cushing’s disease: long-term risk of hypopituitarism. Clin Endocrinol (Oxf). 2016;84(4):524–531.
- 28↑
Attanasio R, Epaminonda P, Motti E, et al. Gamma-knife radiosurgery in acromegaly: a 4-year follow-up study. J Clin Endocrinol Metab. 2003;88(7):3105–3112.
- 29↑
Snell JW, Sheehan J, Stroila M, Steiner L. Assessment of imaging studies used with radiosurgery: a volumetric algorithm and an estimation of its error. Technical note. J Neurosurg. 2006;104(1):157–162.
- 30↑
Dekkers OM, Pereira AM, Romijn JA. Treatment and follow-up of clinically nonfunctioning pituitary macroadenomas. J Clin Endocrinol Metab. 2008;93(10):3717–3726.
- 32
Cohen-Inbar O, Xu Z, Lee CC, et al. Prognostic significance of corticotroph staining in radiosurgery for non-functioning pituitary adenomas: a multicenter study. J Neurooncol. 2017;135(1):67–74.
- 33
Sokolowski JD, Cohen-Inbar O, Sheehan JP. Radiosurgery for infundibulum adenoma: stalk dose implications. Acta Neurochir (Wien). 2016;158(9):1697–1700.
- 34
Cohen-Inbar O, Xu Z, Schlesinger D, et al. Gamma Knife radiosurgery for medically and surgically refractory prolactinomas: long-term results. Pituitary. 2015;18(6):820–830.
- 35↑
Mohammed N, Ding D, Hung YC, et al. Primary versus postoperative stereotactic radiosurgery for acromegaly: a multicenter matched cohort study. J Neurosurg. 2019;132(5):1507–1516.
- 36↑
Sicignano G, Losa M, del Vecchio A, et al. Dosimetric factors associated with pituitary function after Gamma Knife Surgery (GKS) of pituitary adenomas. Radiother Oncol. 2012;104(1):119–124.
- 37↑
Lam KS, Wang C, Yeung RT, et al. Hypothalamic hypopituitarism following cranial irradiation for nasopharyngeal carcinoma. Clin Endocrinol (Oxf). 1986;24(6):643–651.
- 38↑
Samaan NA, Vieto R, Schultz PN, et al. Hypothalamic, pituitary and thyroid dysfunction after radiotherapy to the head and neck. Int J Radiat Oncol Biol Phys. 1982;8(11):1857–1867.
- 39↑
Landolt AM, Haller D, Lomax N, et al. Stereotactic radiosurgery for recurrent surgically treated acromegaly: comparison with fractionated radiotherapy. J Neurosurg. 1998;88(6):1002–1008.
- 40↑
Chieng PU, Huang TS, Chang CC, et al. Reduced hypothalamic blood flow after radiation treatment of nasopharyngeal cancer: SPECT studies in 34 patients. AJNR Am J Neuroradiol. 1991;12(4):661–665.
- 41
Ciccarelli E, Corsello SM, Plowman PN, et al. Long-term effects of radiotherapy for acromegaly on circulating prolactin. Acta Endocrinol (Copenh). 1989;121(6):827–832.
- 42
O’Halloran DJ, Shalet SM. Radiotherapy for pituitary adenomas: an endocrinologist’s perspective. Clin Oncol (R Coll Radiol). 1996;8(2):79–84.
- 43
Degerblad M, Brismar K, Rähn T, Thorén M. The hypothalamus-pituitary function after pituitary stereotactic radiosurgery: evaluation of growth hormone deficiency. J Intern Med. 2003;253(4):454–462.
- 44↑
Cordeiro D, Xu Z, Mehta GU, et al. Hypopituitarism after Gamma Knife radiosurgery for pituitary adenomas: a multicenter, international study. J Neurosurg. 2019;131(4):1188–1196.
- 45↑
Sheehan JP, Kavanagh BD, Asher A, Harbaugh RE. Inception of a national multidisciplinary registry for stereotactic radiosurgery. J Neurosurg. 2016;124(1):155–162.
