Browse the corpus

Long-Term Outcomes After CyberKnife Radiosurgery for Nonfunctioning Pituitary Adenomas. BACKGROUND AND OBJECTIVES: Stereotactic radiosurgery (SRS) has been widely adopted as an important adjunctive treatment modality for managing nonfunctioning pituitary adenomas (NFPAs). However, current studies on the long-term effects of SRS on pituitary adenomas have been largely limited by small sample sizes and short follow-up periods. The aim of this study was to evaluate the long-term outcomes of SRS for NFPAs. METHODS: We conducted a retrospective review of 178 patients with NFPAs who received CyberKnife radiosurgery at a single institution between February 2008 and July 2021. Long-term outcomes of tumor control, new-onset hypopituitarism, and new visual disorders were assessed. RESULTS: During a median radiological follow-up of 49.7 months (range, 2.5-158.1 months), only 11 (7.0%) patients experienced tumor progression. The progression-free survival at 3, 5, and 10 years was 97.47%, 95.57%, and 93.04%, respectively. New-onset hypopituitarism was diagnosed in 27 (16.9%) patients with a median clinical follow-up duration of 71.2 months (range, 11.5-175.4 months). The median time from SRS to new-onset hypopituitarism was 28.3 months (range, 2.8-101.7 months). The cumulative incidence of new-onset hypopituitarism at 3, 5, and 10 years was 8.47%, 12.43%, and 15.25%, respectively. Biological effective dose >140 Gy and single fraction equivalent dose >16.0 Gy were significant risk factors for new-onset hypopituitarism ( P = .046). Other adverse events were experienced by 15 (8.4%) patients, 9 (5.1%) of whom presented with new visual disorders. Development of new visual disorders was associated with a pretreatment tumor volume of >2.5 mL ( P = .044). CONCLUSION: SRS is an effective and relatively safe means of managing both primary and residual/recurrent NFPAs.

We conducted a retrospective review of all patients with NFPAs who received SRS at National Taiwan University Hospital between February 2008 and July 2021. All clinical information was collected within the guidelines approved by the institutional review board (National Taiwan University Hospital, 202307214RINB). Informed consent was waived by the institutional review board. Patients with pathologies other than pituitary adenomas who received SRS and patients who did not complete the entire SRS course were excluded from the study.

s approved by the institutional review board (National Taiwan University Hospital, 202307214RINB). Informed consent was waived by the institutional review board. Patients with pathologies other than pituitary adenomas who received SRS and patients who did not complete the entire SRS course were excluded from the study. A total of 178 patients, 82 (46.1%) of whom were male and 96 (53.9%) of whom were female, were treated with SRS at our institution and were selected for inclusion. The median age of the patients at the time of SRS was 54 years (range, 18-87 years). Most patients received previous surgery and therefore had histological confirmation of the tumor; 85.4% of patients underwent previous transsphenoidal adenomectomy, and 6.2% of patients underwent craniotomy before SRS. Of the 152 patients who underwent transsphenoidal adenomectomy, 98 were operated by the microscopic approach, 33 were operated by the endoscopic approach, and 21 by an unspecified approach at another hospital. These surgeries were performed by multiple neurosurgeons. Usually, the first follow-up MRI was performed 3 to 6 months after surgery. Thus, if the patient received SRS within one year of surgery, the tumor was considered residual; if SRS was arranged more than one year after surgery, the tumor was considered recurrent. Patients with sizeable tumors near the optic apparatus were not offered a repeat resection because of technical challenges encountered during previous operations. All 15 patients who did not undergo surgery also did not receive any previous conventional fractionated radiotherapy or SRS. Approximately one third of patients (32.6%) were on hormone supplementation before SRS. Baseline demographics are shown in Table 1.

of technical challenges encountered during previous operations. All 15 patients who did not undergo surgery also did not receive any previous conventional fractionated radiotherapy or SRS. Approximately one third of patients (32.6%) were on hormone supplementation before SRS. Baseline demographics are shown in Table 1. Baseline Patient Demographics (n = 178) SRS, stereotactic radiosurgery; TSA, transsphenoidal adenomectomy.

of technical challenges encountered during previous operations. All 15 patients who did not undergo surgery also did not receive any previous conventional fractionated radiotherapy or SRS. Approximately one third of patients (32.6%) were on hormone supplementation before SRS. Baseline demographics are shown in Table 1. Baseline Patient Demographics (n = 178) SRS, stereotactic radiosurgery; TSA, transsphenoidal adenomectomy. Generally, neuroimaging and clinical follow-up were performed at baseline, 6 months after SRS for the first year, and then every 1 to 2 years thereafter. If the patient was already on hormone supplementation or diagnosed with new-onset hypopituitarism, clinical follow-up would be more frequent at the discretion of the physician. The primary outcome was the local control rate after SRS. Tumor control was defined as stationary or smaller tumor size on follow-up MRI. Progression-free survival (PFS) was defined as the time between SRS and first progression or death. The secondary outcomes assessed were new-onset hypopituitarism and other complications after radiosurgery. Duration of radiological follow-up was defined as the time between SRS and the last follow-up MRI. Duration of clinical follow-up was defined as the time between SRS and the last outpatient department visit at our institution. Twenty patients were not followed up for imaging study at our institution and were hence excluded from statistical analysis of tumor control rates. New-onset hypopituitarism was defined as hypopituitarism requiring new hormone supplementation for more than one year after SRS or hypopituitarism requiring new hormone supplementation that was prescribed at least one year after SRS. Eighteen patients were lost to clinical follow-up before the 1-year mark and were hence excluded from statistical analysis of new-onset hypopituitarism. Visual complications that were caused by tumor enlargement, which was confirmed on imaging, were not counted as new visual disorders.

ibed at least one year after SRS. Eighteen patients were lost to clinical follow-up before the 1-year mark and were hence excluded from statistical analysis of new-onset hypopituitarism. Visual complications that were caused by tumor enlargement, which was confirmed on imaging, were not counted as new visual disorders. Radiosurgery was performed using the CyberKnife (Accuray, Inc.) model G4 before 2020 and model M6 since 2021. Dose planning was performed using stereotactic MRI to achieve optimal tumor irradiation while minimizing exposure to surrounding structures. The optic apparatus can generally tolerate maximum radiation doses of up to 8 to 12 Gy in a single fraction.18,19 If the radiation dose to the optic apparatus exceeded the maximum tolerable single fraction dose or if the tumor was adjacent to the optic apparatus, dose fractionation would be used to reduce radiation injury to surrounding critical structures.20-22 When the dose was fractionated, it would be divided into equal fractions and given on consecutive days. Biological effective dose (BED) was calculated using an α/β ratio of 2.1.23 Single fraction equivalent dose (SFED) was calculated based on values reported by Puataweepong et al.24 After the tumor and critical structures were defined, treatment planning was performed using the MultiPlan/Precision software (Accuray, Inc.) using inverse planning. The final treatment plan was reviewed and approved by the neurosurgeon and radiation oncologist before treatment delivery. Multiple neurosurgeons and radiation oncologists were involved in the radiation planning of all patients.

as performed using the MultiPlan/Precision software (Accuray, Inc.) using inverse planning. The final treatment plan was reviewed and approved by the neurosurgeon and radiation oncologist before treatment delivery. Multiple neurosurgeons and radiation oncologists were involved in the radiation planning of all patients. Data are presented as the median (range), mean (SD), or number (%). Continuous variables were divided into 2 groups using the median. The nonparametric Mann-Whitney U test was used for continuous variables. Fisher's exact test was used for categorical variables. Continuous variables divided into 3 groups or more were compared using the Kruskal–Wallis test. Post hoc analysis was performed using Dunn's test. The log-rank test and Fisher's exact test were used to evaluate prognostic factors for tumor control, new-onset hypopituitarism, and new visual disorders. If the log-rank test resulted in a significant P value, the Cox model was used to estimate the hazard ratio (HR) and confidence interval. The Hodges–Lehmann estimator (HLΔ) was used to find the confidence interval for the median of all possible differences between 2 groups. The 3-year, 5-year, and 10-year rates of local control and new-onset hypopituitarism were calculated using the Kaplan-Meier method. Statistical analysis was performed using Stata/SE 15.0 (StataCorp). All statistics were 2-tailed, and P < .05 was considered statistically significant.

Generally, neuroimaging and clinical follow-up were performed at baseline, 6 months after SRS for the first year, and then every 1 to 2 years thereafter. If the patient was already on hormone supplementation or diagnosed with new-onset hypopituitarism, clinical follow-up would be more frequent at the discretion of the physician. The primary outcome was the local control rate after SRS. Tumor control was defined as stationary or smaller tumor size on follow-up MRI. Progression-free survival (PFS) was defined as the time between SRS and first progression or death. The secondary outcomes assessed were new-onset hypopituitarism and other complications after radiosurgery. Duration of radiological follow-up was defined as the time between SRS and the last follow-up MRI. Duration of clinical follow-up was defined as the time between SRS and the last outpatient department visit at our institution. Twenty patients were not followed up for imaging study at our institution and were hence excluded from statistical analysis of tumor control rates. New-onset hypopituitarism was defined as hypopituitarism requiring new hormone supplementation for more than one year after SRS or hypopituitarism requiring new hormone supplementation that was prescribed at least one year after SRS. Eighteen patients were lost to clinical follow-up before the 1-year mark and were hence excluded from statistical analysis of new-onset hypopituitarism. Visual complications that were caused by tumor enlargement, which was confirmed on imaging, were not counted as new visual disorders.

Radiosurgery was performed using the CyberKnife (Accuray, Inc.) model G4 before 2020 and model M6 since 2021. Dose planning was performed using stereotactic MRI to achieve optimal tumor irradiation while minimizing exposure to surrounding structures. The optic apparatus can generally tolerate maximum radiation doses of up to 8 to 12 Gy in a single fraction.18,19 If the radiation dose to the optic apparatus exceeded the maximum tolerable single fraction dose or if the tumor was adjacent to the optic apparatus, dose fractionation would be used to reduce radiation injury to surrounding critical structures.20-22 When the dose was fractionated, it would be divided into equal fractions and given on consecutive days. Biological effective dose (BED) was calculated using an α/β ratio of 2.1.23 Single fraction equivalent dose (SFED) was calculated based on values reported by Puataweepong et al.24 After the tumor and critical structures were defined, treatment planning was performed using the MultiPlan/Precision software (Accuray, Inc.) using inverse planning. The final treatment plan was reviewed and approved by the neurosurgeon and radiation oncologist before treatment delivery. Multiple neurosurgeons and radiation oncologists were involved in the radiation planning of all patients.

Data are presented as the median (range), mean (SD), or number (%). Continuous variables were divided into 2 groups using the median. The nonparametric Mann-Whitney U test was used for continuous variables. Fisher's exact test was used for categorical variables. Continuous variables divided into 3 groups or more were compared using the Kruskal–Wallis test. Post hoc analysis was performed using Dunn's test. The log-rank test and Fisher's exact test were used to evaluate prognostic factors for tumor control, new-onset hypopituitarism, and new visual disorders. If the log-rank test resulted in a significant P value, the Cox model was used to estimate the hazard ratio (HR) and confidence interval. The Hodges–Lehmann estimator (HLΔ) was used to find the confidence interval for the median of all possible differences between 2 groups. The 3-year, 5-year, and 10-year rates of local control and new-onset hypopituitarism were calculated using the Kaplan-Meier method. Statistical analysis was performed using Stata/SE 15.0 (StataCorp). All statistics were 2-tailed, and P < .05 was considered statistically significant.

The median tumor volume was 2.5 mL (range, 0.3-40.1 mL), and the median distance to the optic apparatus was 1.0 mm (range, 0.0-10.2 mm). The median tumor prescription dose, BED, and SFED were 18.0 Gy (range, 12.0-33.0 Gy), 137.90 Gy (range, 68.09-252.47 Gy), and 16.0 Gy (range, 10.6-22.0 Gy), respectively. Although analyses of median tumor volume and median distance from the optic apparatus between previously resected and treatment-naïve tumors did not reach statistical significance, our results in Table 2 show that treatment-naïve tumors tended to be smaller and closer to the optic apparatus. Compared with treatment-naïve tumors, we found that previously resected tumors were treated with a higher median BED (P = .034, 93.5 vs 137.9 Gy, HLΔ [95% CI] = −21.57 Gy [−46.29 to 0.00 Gy]) and SFED (P = .031, 13.0 vs 16.0 Gy, HLΔ [95% CI] = −1.90 Gy [−3.00 to 0.00 Gy]). In addition, there was a higher incidence of cavernous sinus invasion in previously resected tumors (46.6%) compared with treatment-naïve tumors (20.0%). Tumors that were treated in a single fraction were significantly farther away from the optic apparatus than those that were treated in multiple fractions (P < .001, 2.0 vs 0.0 mm, HLΔ [95% CI] = 2.00 mm [1.83-2.00 mm]). Details on SRS treatment parameters are summarized in Table 2. Compared with those who underwent craniotomy or endoscopic resection, those who underwent microscopic resection were significantly farther away from the optic apparatus (P = .004, .0 vs 1.4 mm, HLΔ [95% CI] = −1.00 mm [−1.00 to 0.00 mm]), were smaller in size in median tumor volume (P = .016, 3.5 vs 2.2 mL, HLΔ [95% CI] = 1.29 mL [0.35-2.45 mL]), and were treated with a significantly higher median BED (P = .004, 137.9 vs 107.5 Gy, HLΔ [95% CI] = 24.43 Gy [0.57-38.40 Gy]) and SFED (P = .003, 16.0 vs 13.8 Gy, HLΔ [95% CI] = 1.80 Gy [0.50-2.80 Gy]). The prescribed dose, BED, and SFED for all dose schedules in our study are summarized in Supplemental Digital Content 1 (http://links.lww.com/NEU/E465). The most common radiation schedules were 16.0 and 18.0 Gy in 1 fraction and 18.0 and 21.0 Gy in 3 fractions.

, 16.0 vs 13.8 Gy, HLΔ [95% CI] = 1.80 Gy [0.50-2.80 Gy]). The prescribed dose, BED, and SFED for all dose schedules in our study are summarized in Supplemental Digital Content 1 (http://links.lww.com/NEU/E465). The most common radiation schedules were 16.0 and 18.0 Gy in 1 fraction and 18.0 and 21.0 Gy in 3 fractions. Stereotactic Radiosurgery Treatment Parameters BED, biological effective dose; SFED, single fraction equivalent dose. Data are presented as number (%) or median (range) and mean ± SD. Bold values are statistically significant, P < .05. Post-SRS MRI follow-up was available for 158 patients. Outcomes regarding tumor control are reported in Supplemental Digital Content 2 (http://links.lww.com/NEU/E466). Most of the NFPAs remained stable or regressed, and only 11 (7.0%) patients experienced radiological progression at a median interval from SRS to progression of 44.3 months (range, 2.5-115.7 months). With a median radiological follow-up duration of 49.7 months (range, 2.5-158.1 months), the PFS at 3, 5, and 10 years was 97.47%, 95.57%, and 93.04%, respectively (Figure 1). Statistical analysis did not reveal any significant predictors of tumor control by the log-rank and Fisher's exact test (Supplemental Digital Content 3 [http://links.lww.com/NEU/E467] and Supplemental Digital Content 4 [http://links.lww.com/NEU/E468]). Kaplan–Meier plot demonstrating overall progression-free survival rate.

Post-SRS MRI follow-up was available for 158 patients. Outcomes regarding tumor control are reported in Supplemental Digital Content 2 (http://links.lww.com/NEU/E466). Most of the NFPAs remained stable or regressed, and only 11 (7.0%) patients experienced radiological progression at a median interval from SRS to progression of 44.3 months (range, 2.5-115.7 months). With a median radiological follow-up duration of 49.7 months (range, 2.5-158.1 months), the PFS at 3, 5, and 10 years was 97.47%, 95.57%, and 93.04%, respectively (Figure 1). Statistical analysis did not reveal any significant predictors of tumor control by the log-rank and Fisher's exact test (Supplemental Digital Content 3 [http://links.lww.com/NEU/E467] and Supplemental Digital Content 4 [http://links.lww.com/NEU/E468]). Kaplan–Meier plot demonstrating overall progression-free survival rate. A total of 160 patients received clinical follow-up for at least one year. Outcomes regarding new-onset hypopituitarism are reported in Supplemental Digital Content 5 (http://links.lww.com/NEU/E469). New-onset hypopituitarism was diagnosed in 27 (16.9%) patients at a median interval from SRS to hypopituitarism of 28.3 months (range, 2.8-101.7 months). With a median clinical follow-up duration of 71.2 months (range, 11.5-175.4 months), the cumulative incidence of new-onset hypopituitarism at 3, 5, and 10 years was 8.47%, 12.43%, and 15.25%, respectively (Figure 2). In the univariate analysis (Supplemental Digital Content 6 [http://links.lww.com/NEU/E470] and Supplemental Digital Content 7 [http://links.lww.com/NEU/E471]), BED >140 Gy and SFED >16.0 Gy were significant risk factors for new-onset hypopituitarism (log-rank, P = .046, HR [95% CI] = 2.13 [1.00-4.56], Figure 3).

re 2). In the univariate analysis (Supplemental Digital Content 6 [http://links.lww.com/NEU/E470] and Supplemental Digital Content 7 [http://links.lww.com/NEU/E471]), BED >140 Gy and SFED >16.0 Gy were significant risk factors for new-onset hypopituitarism (log-rank, P = .046, HR [95% CI] = 2.13 [1.00-4.56], Figure 3). Kaplan–Meier plot demonstrating overall cumulative incidence of new-onset hypopituitarism. Kaplan–Meier survival curves demonstrating the effect of BED and SFED on new-onset hypopituitarism. Patients treated with a BED of >140 Gy and a SFED of >16.0 Gy were more likely to develop new-onset hypopituitarism (log-rank, P = .046, hazard ratio [95% CI] = 2.13 [1.00-4.56]). BED, biological effective dose; SFED, single fraction equivalent dose.

emonstrating the effect of BED and SFED on new-onset hypopituitarism. Patients treated with a BED of >140 Gy and a SFED of >16.0 Gy were more likely to develop new-onset hypopituitarism (log-rank, P = .046, hazard ratio [95% CI] = 2.13 [1.00-4.56]). BED, biological effective dose; SFED, single fraction equivalent dose. A total of 15 (8.4%) patients developed other complications during the follow-up period after SRS. Nine (5.1%) patients were complicated with new visual disorders, 6 of whom developed decreased visual acuity, 2 of whom developed new visual field defects, and one of whom developed ptosis which was later diagnosed as oculomotor nerve palsy. The median time to the development of new visual disorders was 42.1 months (range, 1.6-86.4 months). In the univariate analysis (Supplemental Digital Content 8 [http://links.lww.com/NEU/E472] and Supplemental Digital Content 9 [http://links.lww.com/NEU/E473]), pretreatment tumor volume >2.5 mL (log-rank, P = .044, HR [95% CI] = 4.41 [1.01-19.26], Figure 4) increased the risk of developing new visual disorders. In addition, 6 other major complications were noted during the follow-up period: 3 (1.7%) patients developed hydrocephalus, 2 of whom were treated with ventriculoperitoneal shunts and one of whom preferred conservative treatment; 2 (1.1%) patients developed cerebrospinal fluid (CSF) leakage, one of whom underwent microscopic resection and the other craniotomy; and one (0.6%) patient, who experienced a transient ischemic attack (TIA), was diagnosed with internal carotid artery 60% stenosis. No radiation necrosis or secondary neoplasm was observed during the follow-up period. Other complications of SRS are summarized in Table 3.

microscopic resection and the other craniotomy; and one (0.6%) patient, who experienced a transient ischemic attack (TIA), was diagnosed with internal carotid artery 60% stenosis. No radiation necrosis or secondary neoplasm was observed during the follow-up period. Other complications of SRS are summarized in Table 3. Kaplan–Meier survival curves demonstrating the effect of pretreatment tumor volume on new visual disorders. Patients with a pretreatment tumor volume of >2.5 mL were more likely to develop new visual disorders (log-rank, P = .044, hazard ratio [95% CI] = 4.41 [1.01-19.26]). Complications CSF, cerebrospinal fluid; ICA, internal carotid artery.

, 16.0 vs 13.8 Gy, HLΔ [95% CI] = 1.80 Gy [0.50-2.80 Gy]). The prescribed dose, BED, and SFED for all dose schedules in our study are summarized in Supplemental Digital Content 1 (http://links.lww.com/NEU/E465). The most common radiation schedules were 16.0 and 18.0 Gy in 1 fraction and 18.0 and 21.0 Gy in 3 fractions. Stereotactic Radiosurgery Treatment Parameters BED, biological effective dose; SFED, single fraction equivalent dose. Data are presented as number (%) or median (range) and mean ± SD. Bold values are statistically significant, P < .05.

Post-SRS MRI follow-up was available for 158 patients. Outcomes regarding tumor control are reported in Supplemental Digital Content 2 (http://links.lww.com/NEU/E466). Most of the NFPAs remained stable or regressed, and only 11 (7.0%) patients experienced radiological progression at a median interval from SRS to progression of 44.3 months (range, 2.5-115.7 months). With a median radiological follow-up duration of 49.7 months (range, 2.5-158.1 months), the PFS at 3, 5, and 10 years was 97.47%, 95.57%, and 93.04%, respectively (Figure 1). Statistical analysis did not reveal any significant predictors of tumor control by the log-rank and Fisher's exact test (Supplemental Digital Content 3 [http://links.lww.com/NEU/E467] and Supplemental Digital Content 4 [http://links.lww.com/NEU/E468]). Kaplan–Meier plot demonstrating overall progression-free survival rate.

A total of 160 patients received clinical follow-up for at least one year. Outcomes regarding new-onset hypopituitarism are reported in Supplemental Digital Content 5 (http://links.lww.com/NEU/E469). New-onset hypopituitarism was diagnosed in 27 (16.9%) patients at a median interval from SRS to hypopituitarism of 28.3 months (range, 2.8-101.7 months). With a median clinical follow-up duration of 71.2 months (range, 11.5-175.4 months), the cumulative incidence of new-onset hypopituitarism at 3, 5, and 10 years was 8.47%, 12.43%, and 15.25%, respectively (Figure 2). In the univariate analysis (Supplemental Digital Content 6 [http://links.lww.com/NEU/E470] and Supplemental Digital Content 7 [http://links.lww.com/NEU/E471]), BED >140 Gy and SFED >16.0 Gy were significant risk factors for new-onset hypopituitarism (log-rank, P = .046, HR [95% CI] = 2.13 [1.00-4.56], Figure 3). Kaplan–Meier plot demonstrating overall cumulative incidence of new-onset hypopituitarism. Kaplan–Meier survival curves demonstrating the effect of BED and SFED on new-onset hypopituitarism. Patients treated with a BED of >140 Gy and a SFED of >16.0 Gy were more likely to develop new-onset hypopituitarism (log-rank, P = .046, hazard ratio [95% CI] = 2.13 [1.00-4.56]). BED, biological effective dose; SFED, single fraction equivalent dose.

A total of 15 (8.4%) patients developed other complications during the follow-up period after SRS. Nine (5.1%) patients were complicated with new visual disorders, 6 of whom developed decreased visual acuity, 2 of whom developed new visual field defects, and one of whom developed ptosis which was later diagnosed as oculomotor nerve palsy. The median time to the development of new visual disorders was 42.1 months (range, 1.6-86.4 months). In the univariate analysis (Supplemental Digital Content 8 [http://links.lww.com/NEU/E472] and Supplemental Digital Content 9 [http://links.lww.com/NEU/E473]), pretreatment tumor volume >2.5 mL (log-rank, P = .044, HR [95% CI] = 4.41 [1.01-19.26], Figure 4) increased the risk of developing new visual disorders. In addition, 6 other major complications were noted during the follow-up period: 3 (1.7%) patients developed hydrocephalus, 2 of whom were treated with ventriculoperitoneal shunts and one of whom preferred conservative treatment; 2 (1.1%) patients developed cerebrospinal fluid (CSF) leakage, one of whom underwent microscopic resection and the other craniotomy; and one (0.6%) patient, who experienced a transient ischemic attack (TIA), was diagnosed with internal carotid artery 60% stenosis. No radiation necrosis or secondary neoplasm was observed during the follow-up period. Other complications of SRS are summarized in Table 3.

To the best of our knowledge, this is currently the largest and most comprehensive single-center CyberKnife study on NFPAs. Current available CyberKnife studies on NFPAs show promising results with low risks of complications (Table 4). Summary of Current Available CyberKnife Studies on Nonfunctioning Pituitary Adenomas BED, biological effective dose; F, functioning pituitary adenoma; Fr, fraction; NF, nonfunctioning pituitary adenoma.

To the best of our knowledge, this is currently the largest and most comprehensive single-center CyberKnife study on NFPAs. Current available CyberKnife studies on NFPAs show promising results with low risks of complications (Table 4). Summary of Current Available CyberKnife Studies on Nonfunctioning Pituitary Adenomas BED, biological effective dose; F, functioning pituitary adenoma; Fr, fraction; NF, nonfunctioning pituitary adenoma. Even with longer follow-up periods (median radiological follow-up duration of 49.7 months) and larger sample size, our tumor control rate (93.0%) compares favorably with previous CyberKnife studies that report tumor control rates ranging from 92.3% to 100%.11,12,14-16,24,25 In addition, we report rates of PFS at 3, 5, and 10 years to be 97.47%, 95.57%, and 93.04%, respectively. These rates are consistent with those reported by Iwata et al,14 who reported rates of PFS at 3, 4, and 5 years to be 96%, 93%, and 93%, respectively. In our study, a tumor control rate of 92.3% was achieved at last MRI follow-up in patients with advanced age or significant surgical risks who underwent SRS as the primary treatment (Supplemental Digital Content 2, http://links.lww.com/NEU/E466). Thus, we believe that SRS may be a viable therapeutic option in selected patients with primary NFPAs who carry a higher risk for resection because of concomitant medical illness or advanced age. Similarly, SRS may be considered for patients with functioning pituitary adenomas who are deemed medically unfit for surgical resection.5-7 Previous studies on SRS have reported large tumor volumes,26,27 suprasellar extension,28 and a history of multiple surgical procedures29 to be risk factors for tumor recurrence in NFPAs. However, we did not find any significant predictors of tumor control in our patients.

emed medically unfit for surgical resection.5-7 Previous studies on SRS have reported large tumor volumes,26,27 suprasellar extension,28 and a history of multiple surgical procedures29 to be risk factors for tumor recurrence in NFPAs. However, we did not find any significant predictors of tumor control in our patients. Similarly, with longer follow-up periods (median clinical follow-up duration of 71.2 months) and larger sample size, our rate of new-onset hypopituitarism (16.9%) compares favorably with previous CyberKnife studies that report new-onset hypopituitarism rates ranging from 0% to 15.8%.11,12,14-16,24,25 In addition, we are the first to report the cumulative incidences of new-onset hypopituitarism for CyberKnife at 3, 5, and 10 years to be 8.47%, 12.43%, and 15.25%, respectively, which are lower than those reported in previous Gamma Knife radiosurgery (GKRS) studies.26,30,31 In a single-center retrospective study of patients with NFPA treated with GKRS, Deng et al31 reported the cumulative rates of developing new hypopituitarism at 3, 5, and 10 years to be 21%, 30%, and 57%, respectively. In another multicenter, international cohort of patients with both functioning pituitary adenomas and NFPA treated with GKRS, Cordeiro et al30 reported the actuarial 3-year, 5-year, and 10-year rates of hypopituitarism to be 16.2%, 22.4%, and 31.3%, respectively. In our study, patients who received a BED of >140 Gy or an SFED of >16.0 Gy were significantly more likely to develop new-onset hypopituitarism. These findings are consistent with those reported by Leenstra et al32 who found that new endocrine deficits after pituitary adenoma radiosurgery correlated with increasing radiation dose to the pituitary gland, with reports of safe mean radiation doses ranging from 9.5 to 18 Gy.33-36

develop new-onset hypopituitarism. These findings are consistent with those reported by Leenstra et al32 who found that new endocrine deficits after pituitary adenoma radiosurgery correlated with increasing radiation dose to the pituitary gland, with reports of safe mean radiation doses ranging from 9.5 to 18 Gy.33-36 All 178 patients in this study were followed up for complications. Our new visual disorder rate (5.1%) compares favorably with previous CyberKnife studies that report visual complication rates ranging from 0% to 13.6%.11,12,14-16,24,25 Previous studies have shown that multisession SRS allows for high rates of tumor control while preserving visual function in patients with perioptic lesions.20-22 We arranged dose fractionation if the radiation dose to the optic apparatus exceeded the maximum tolerable single fraction dose or if the tumor was adjacent to the optic apparatus. As a result, tumors that were treated in a single fraction were significantly farther away from the optic apparatus than those that were treated in multiple fractions. In addition, we found that the development of new visual disorders was significantly associated with a pretreatment tumor volume of >2.5 mL. Similarly, in a retrospective study conducted by Sun et al,37 tumor volume ≥5 mL was found to be a significant risk factor of visual dysfunction. In particular, we observed the development of oculomotor nerve palsy in a patient with subclinical apoplexy who was treated with a prescription dose of 18.0 Gy in 3 fractions. The initial follow-up MRI of this patient revealed evidence of enlarging hematoma. However, decreased tumor size was noted in subsequent MRI scans. Thus, we believe that this patient's oculomotor palsy was caused by pituitary apoplexy after SRS. Asymptomatic subclinical apoplexy after SRS based on radiological findings or autopsy is not that uncommon, with incidence rates of up to 25%.38 However, symptomatic pituitary apoplexy after SRS is rare and there have been only 2 reports of SRS-associated clinical pituitary apoplexy in the literature.38,39

y after SRS. Asymptomatic subclinical apoplexy after SRS based on radiological findings or autopsy is not that uncommon, with incidence rates of up to 25%.38 However, symptomatic pituitary apoplexy after SRS is rare and there have been only 2 reports of SRS-associated clinical pituitary apoplexy in the literature.38,39 We observed other major adverse events of hydrocephalus, delayed CSF leakage, and TIA, which are less commonly reported in the literature. Communicating hydrocephalus requiring CSF diversion is an unusual complication of SRS in patients with pituitary adenomas. Narayan et al27 reported that the complication rate of hydrocephalus after GKRS for pituitary adenomas was 2.7%. It has been proposed that radiotherapy-induced fibrosis of arachnoid granulations and radiotherapy-induced tumor necrosis with subsequent CSF protein elevation may lead to CSF malabsorption.40,41 Although there are no studies in the current literature regarding the risk factors of communicating hydrocephalus after SRS for pituitary adenomas, previous studies on vestibular schwannomas have reported large tumor volume to be a significant risk factor.42-44 Nevertheless, one of the 3 patients in our study, who were complicated with hydrocephalus, had a tumor volume of 0.9 mL. Thus, we cannot determine whether there is a correlation between tumor volume and post-SRS hydrocephalus for pituitary adenomas.

have reported large tumor volume to be a significant risk factor.42-44 Nevertheless, one of the 3 patients in our study, who were complicated with hydrocephalus, had a tumor volume of 0.9 mL. Thus, we cannot determine whether there is a correlation between tumor volume and post-SRS hydrocephalus for pituitary adenomas. Delayed CSF leak and ischemic stroke after pituitary radiosurgery are even less commonly reported in the literature.28,45-48 In our study, 1 patient experienced a TIA 81 months after SRS was diagnosed with a moderate degree of right internal carotid artery stenosis. This patient was a long-time smoker and had multiple risk factors for stroke including hypertension, dyslipidemia, and a family history of stroke. Thus, we believe that the development of TIA in this patient was multifactorial, which warrants further research to evaluate the contribution of radiation in the development of ischemic stroke.

Our study has some limitations. The retrospective study design of a single institution may present with some bias. Furthermore, most of the decreased visual acuity and new visual field defect complications were based on subjective data rather than examination results. Visual acuity and Humphrey visual field tests should be conducted in future studies to provide more objective data on the visual complications of SRS. Finally, our study did not collect data on the radiation doses to surrounding critical structures such as the optic pathways and pituitary stalk and cavernous sinuses, which would provide further insight into the optimal radiosurgical dose for NFPAs.

In this clinical experience with longer follow-up periods, we found that SRS is an effective and relatively safe means of managing both primary and residual/recurrent NFPAs. To the best of our knowledge, this is the largest single-center study on CyberKnife for NFPAs to date. We believe that our results provide additional long-term evidence that SRS is a safe and effective treatment option capable of improving outcomes for NFPAs. Multicenter, international studies are needed to confirm our results and develop radiosurgery guidelines for these types of tumors.