Multisession CyberKnife radiosurgery for post-surgical residual and recurrent pituitary adenoma: preliminary result from one center

Abstract

Objective

To evaluate the efficacy of CyberKnife (Accuray, Inc., Sunnyvale, CA) stereotactic radiosurgery (SRS) for patients with pituitary adenoma.

Methods

We conducted a retrospective review of all patients treated by image-guided radiosurgery at our institution between August 2007 and June 2009. Twentytwo patients with pituitary adenoma were identified. The median follow-up period from date of treatment was 30.8 + 7.7months. The median patient age was 56 years. The mean tumor volume was 4.1 + 2.8 mL (range, 0.8-16.0 mL), and the prescribed dose was 25 Gy (5 Gy x 5).

Results

Tumors were treated with a mean coverage of 95.1 ± 1.3%(range 90.2-99.1%), a mean conformality index of 1.4 ± 0.4 (range, 1.2-1.8), and a mean treatment isodose line of 76.4 ± 3.0%(range70-83%). The primary endpoints were radiographic and endocrinological tumor control. The local radiographic control rate in this series was 100% (22/22). Radiographic evidence of tumor necrosis developed in 3 patients (13.7%).

Conclusion

CyberKnife multisession radiosurgery of 25 Gy (5 Gy x 5) provided excellent local control with acceptable adverse effect in patients with pituitary adenoma post-transsphenoidal resection. This preliminary study provides data that supports the belief that the optic nerve and chiasm tolerate the above dosing regimens in perichiasmatic pituitary adenomas, and this dosing regimen may achieve satisfactory radiographic and endocrinological tumor control for post-transsphenoidal resection.

INTRODUCTION

Pituitary adenomas are benign intracranial tumors and constitute 10-20% of central nervous system tumors¹. Pituitary adenomas are traditionally classified as functional and nonfunctioning pituitary adenomas (NFPAs) according to the secretory character of tumor; patients with secretory adenomas most frequently present with endocrinopathies. Although pituitary adenomas are histologically benign, the neurological and physiological consequences can be devastating, particularly if these tumors are left untreated. In NFPAs, compression of the optic apparatus (nerves, chiasm, tracts) may lead to blindness or loss of peripheral vision. Invasion into the cavernous sinus can result in diplopia and/or facial paresthesias. Further expansion of the tumor laterally into the temporal lobes or posteriorly into the hypothalamus can lead to significant cognitive problems. On the other hand, the patients with functional adenomas typically have abnormal neuroendocrine secretion with resultant health consequences. Excessive growth hormone (GH) production associated with acromegaly can lead to life-threatening cardiovascular and respiratory conditions, diabetes mellitus, and possibly an increased risk of colon cancer. Prolonged hypersecretion of ACTH in Cushing disease can lead to severe problems with hypertension and osteoporosis. In patients with prolactinomas, galactorrhea and infertility may occur.^{2, 3}

The goal of therapy in functional adenoma is normalizing the endocrinopathy and eliminating further lesion growth. First-line therapy for prolactin-secreting tumors, however, is medical, using dopamine-agonist drugs such as bromocriptine or cabergoline; while surgery is reserved for adrenocorticotropic and GH-secreting pituitary adenomas, or in combination with medical therapy in lesions not controlled by either therapy alone.⁴ For NFPAs, surgery is the treatment of choice, but patients remain at risk for tumor recurrence for several years afterward. Therefore, the surgical intervention such as trans-sphenoidal approach is still the main curative method for this disease.

Technologic advancements in endoscopy and increased experience in pituitary surgery have reduced surgical mortality.⁵ Introduction of the endoscope have revolutionized pituitary surgery, with the remission results for patients with nonfunctioning and functional adenomas reported to be 83% and 76.3% respectively.⁶ However, the post-operative complications, residual tumor and recurrence of the tumor are still the primary obstacle for the clinician in treating theses patients. Post-operative complications rate were present in 13.9% to 42.1% of cases. The most frequent complications included temporary and permanent diabetes insipidus, syndrome of inappropriate antidiuretic hormone secretion and CSF leaks.^{6, 7} Recurrence rates of NFPAs are about 33.5%⁸. There was a higher proportion of cavernous sinus invasion in recurrent patients (47.2%), which has been reported to be the most important unfavorable factor influencing the surgical outcome after initial surgery.⁹ Furthermore, even when performed by experienced pituitary surgeons, repeat surgery still created significant post-operative morbidity with complication rates around 20%¹⁰.

Stereotactic radiosurgery is now been recognized as a useful method to treat some cases of invasive pituitary adenoma, especially those extends into the suprasellar and cavernous sinus regions. Long-term data confirm the antisecretory efficacy of the procedure (about 50% remission in hypersecreting tumors). The antitumoral efficacy of Gamma Knife treatment against nonsecreting tumors is observed in about 90% of cases. The time to remission is estimated to range from 12 to 60 months. Although antisecretory drugs, particularly for acromegaly and prolactinomas, have been increasingly used more effectively, tumor recurrences or new growth from residual tissue of such cases may occur (8-40% over 10 years).¹¹ For those residual or recurrent tumors, the radiosurgery treatment had also the potential to reduce some of the preoperative morbidity associated with traditional open surgery.

In this study, we show good local radiographic control rates (100%) for patients who received the CyberKnife therapy in treating the residual or recurrent pituitary adenoma after transsphenoid surgery.

Patients and Methods

Patients: Data was collected retrospectively from 22 consecutive patients who underwent multisession radiosurgery for primary, recurrent or residual pituitary adenomas with the CyberKnife at the Tri-Service General Hospital between August 2007 and June 2011. All clinical information was obtained within guidelines approved by the Institutional Review Board at Tri-Service General Hospital. At the time of treatment, the patients’ median age was 58.0 ± 15.9 years (range, 27.5-83.8 years). There were 14 female patients and 8 male patients.

All patients had undergone previous operations except one patient received the SRS for his pituitary tumor based on radiographic appearance alone: 17 patients had single transsphenoidal resections, 1 patient had one previous transsphenoidal resection as well as previous craniotomy; 2 patients had 2 previous transsphenoidal resections then followed with craniotomy approach to remove residual tumor, 1 had 2 previous transsphenoidal resection and received conventional radiotherapy. Patients underwent CyberKnife radiosurgery for residual tumor within 6 months of their last operation (mean, 3.5 ± 1.1 months; range, 1.7-5.6 months).

Patient demographic data is presented in Table 1.

Table 1.

Summary of patient demographics at time of treatment.

Patient No.	Age at treatment/sex	adenoma type	previous operation	CyberKnife dosing (Gy)
1	25.5/F	Acro	TS	5x5
2	67.9/F	NF	TS	5x5
3	28.1/M	NF	TS	5x5
4	66.1/M	NF	Nil	5x5
5	66.9/F	NF	TS	5x5
6	63.0/M	NF	TS	5x5
7	59.8/F	NF	TS	5x5
8	56.1/F	PRL	TS	5x5
9	59.8/M	NF	TS	5x5
10	53.5/M	NF	TS	5x5
11	55.8/F	NF	TS	5x5
12	33.8/M	PRL	TS	5x5
13	40.1/F	PRL	TS	5x5
14	81.8/F	NF	TS	5x5
15	35.2/F	NF	TS	5x5
16	68.1/F	NF	TS	5x5
17	76.2/F	NF	TSx2, 3D-CRT and X-knife	5x5
18	39.2/F	NF	TSx2, Crani	5x5
19	76.1/M	NF	TS	5x5
20	49.4/F	NF	TS	5x5
21	46.7/M	PRL	TS, Crani	5x5
22	50.3/F	NF	TS	5x5

NF, nonfunctioning; TS, transsphenoidal; Crani, craniotomy; Acro, acromegalic; PRL, prolactinoma.

Tumors: The tumors enrolled in this study included 18 postsurgical remnants in which 14 had cavernous sinus involvement (11 had unilateral cavernous sinus involvement, 3 had bilateral cavernous sinus extension); 3 with parasellar invasion and 1 had suprasellar extension. The remained 3 patients had recurrent tumors. The tumor types in this study are seventeen nonfunctional adenomas, four prolactinomas and 1 growth hormone secreting tumor. Tumor characteristics are presented in Table 2.

Table 2.

Patient and tumor characteristics.

Characteristics	Number of patients
Total	22
Age (years), median (range)	56 (27.3-83.5)
Gender, male/female	M15/F7
KPS 100/90/80/70	21/1/0/0
After surgery	21
Refusal of surgery or medical inoperability	1
Tumor volume (cc), median (range)	4.1 (0.8-16.0)
Pre-SRT visual disorders	17
Tumor Types	NF17/PR4/AC1
Interval between final operation and SRT (months)Median (range)	58 (25.2~76)

SRT: Stereotactic radiotherapy; NF: nonfunctional adenoma; PR: prolactinoma; AC: acromegaly

Radiosurgery: All patients were treated using the CyberKnife system. Thin-slice, high-resolution, computer tomographic scans with 1.25-mm slice intervals and magnetic resonance imaging (MRI) scans with 2- mm slice intervals, obtained after intravenous administration of gadolinium, were used for treatment planning. The Accuray MultiPlan treatment software was used to fuse these image sets to better visualize tumor and neural structures (Fig.1). The treating surgeon manually outlined the lesion as well as critical structures, including the optic nerves, the chiasm, the brainstem, and the eyes. An inverse planning method was used to deliver a tumor marginal dose of 5 Gy per session. Tumor margins were treated with a 6-MV X-band photon source to a mean isodose line of 76.4 ± 3.0% (range, 70-83%) with an average of 191 ± 57 beams (range, 99-283 beams) with a single collimator in all but 2 cases, in which a second collimator was used. All patients were treated on consecutive days until their radiosurgery was completed.

Figure 1.

Representative CyberKnife treatment plans showing solid isodose lines outlining the tumors and dotted outlines of critical structures, based on computed tomographic scans (A) and cumulative-dose volume histogram (DVH) and maximal dose at critical organ were tabulated (B).

Clinical Evaluation

The pretreatment evaluation included formal visual field testing with an ophthalmologist, gadolinium-enhanced MRI scanning, and an evaluation by an endocrinologist. Our postradiosurgery protocol included clinical evaluation every 3 months and visual field checks and gadolinium-enhanced MRI scanning at 3 months and 1 year post treatment and annually thereafter. Endocrine follow-up was at the discretion of the patient’s endocrinologist and varied on the basis of pretreatment endocrine status and the adenoma’s functional status.

RESULT

CyberKnife Treatments

All patients received the full 25Gy in 5 fractions in 5 consecutive daily sessions. The mean tumor coverage with the full prescribed dose was 95.1 ± 1.3% (range, 90.2-99.1%), with an average conformality index of 1.4 ± 0.4 (range, 1.2-1.8) to an average tumor volume of 4.1 + 2.8 mL (range, 0.8-16.0 mL). The mean maximal dose to the tumor per session was 6.5 ± 0. 3 Gy (range, 6.0-7.1 Gy) and 2.4 ± 0.31 Gy (range, 1.5-3.0 Gy) to the optic chiasm, with an average minimal dose of 3.8 ± 0.8 Gy per session (range, 2.3-4.9 Gy per session). Treatment data is summarized in Table 2.

Tumor response

In our study, of the 21 patients that had received the surgery removal of the tumor, 3 patients had radiosurgery after demonstrating postresection radiographic tumor progression, on average 58.1 ± 28.6 months (range, 25.2-76 months) after surgery. The remaining 18 patients were referred for treatment with a mean interval of 6.7 ± 6.9 months (range, 0.7-25.2 months) after surgery because of residual tumor. Patient demographics are presented in tables 1 & 2.

Radiographic follow-up with gadolinium-enhanced MRI was available for all patients. All imaging studies were reviewed by radiologists who were blinded to the clinical information. According to tumor response to the CyberKnife treatment, the outcome of patient could be divided into three groups: those tumors which decreased after treatment indicated as tumor-response (TR) group, those tumors which did not change size or shape indicated as stable-disease (SD) group, and those tumors growing after treatment indicated as disease-progression (PD) (Fig.2). Tumor reduction was seen in 9 tumors (39.1%, representative case Fig.4), and stable disease in 13 patients (56.5 %). Figure 3 shows that the cumulative tumor response rate during our follow up period. Tumors which decreased in size typically showed reduction at 3~6 months post treatment. Tumor control defined as tumor reduction and stable disease, was 95.6%. Radiographic evidence of radiation necrosis developed in 3 patients (13%).

Figure 2.

The outcome of the patient was classified based on tumor response to the CyberKnife treatment: Tumor reduction (TR) was seen in 9 tumors (39.1%), the remained (60.9%) are stable disease (SD) , and none showed disease progression (PD) during our following period.

Figure 4.

The representative case of tumor response (No. 3 patient, a 28-y-o male with non-functional adenoma (A, F, M) after transsphenoidal approach with tumor residual (B, I, N)). The tumor started to shrink 3 months later(C, J, O), 6 months (D, K, P) and 1.5 years later (E, L, Q); and response continuously to near completely remission.

Figure 3.

The Kaplan-Meier curve of cumulative tumor response rate during following up period and most of the tumor start to shrink from 3 months after the CyberKnife treatment.

Vision Outcome

Follow-up data was available for 21of 22 patients for visual fields and acuity, with a mean follow-up period of 26.6 ± 10.5 months (range, 10.5-41 months). Before treatment, 5 patients had intact vision, and 17 had impaired vision. At the last follow-up examination, 5 of 5 patients with intact vision prior to treatment still had intact vision at last follow up (Table 5); among the 17 patients with impaired vision, most were stable (11 of 17 patients) or improved (6 of 17 patients)

Table 5.

CyberKnife treatment of perioptic tumor: review of literature.

Author (Year)	No. of Patients	Mean follow up (mos)	Marginal dose (Gy)	Tumor response (%)	Tumor Control (%)	New Hormone deficit (%)	New Visual deficit (%)
Pham et al., (2004)	34	29	20	ND	94	ND	8.8
Kajiwara et al., (2005)	21	32	12.6	7.1	92.9	7.1	4.8
Adler et al., (2006)	49	49	20.3	64	94%	ND	6
Roberts et al., (2007)	9	25.4	21	44.4%	100	33	0
Killory et al., (2009)	20	29.3	25/5fr	ND	100	ND	0
Chen et al., (2011)	22	30.7	25/5fr	13.6	100	0	0

Hormone level

In our series, there were 4 prolactinomas and 1 GH secreting tumor; the other 17 patients were nonfunctional adenoma. Of the 5 secreting tumors followed post treatment (mean, 33 months), only 1 patient with a prolactinoma continued to have abnormal hormone levels. This was subsequently treated with bromocriptine. There was no hormone deficit noted post treatment in the patients with nonfunctional tumor.

Table 4.

The outcome of visual field after CyberKnife treatment.

Outcome of Visual field	Defect	Intact
Pre-CyberKnife treatment	18	5
Post-CyberKnife unchanged	11	5
Post-CyberKnife improve	6	0

Complication

Treatment-related morbidity in our study revealed transient and fleeting headaches and an occasional complaint of transient diplopia lasting for less than 6 weeks in three patients, all of whom responded to a short course of dexamethasone. There was no other morbidity observed during this study. The only significant long-term morbidity related to visual loss in the single patient with tumor progression as described above.

DISCUSSION

The treatment of pituitary tumor is based on the functional type (secretory pattern) of the adenoma.¹² For GH-secreting adenomas, trans-sphenoidal surgery is the first-line therapy while postoperative radiation therapy (fractionated, or radiosurgery) is performed for partially resected tumors or when GH levels remain elevated after operation. Somatostatin analogs, are proposed when surgery is contra-indicated, or has failed to normalize GH levels, or in waiting for the delayed effects of radiation therapy. For prolactinoma, dopamine-agonists are now considered as primary treatment because the effects on visual disturbances and tumoral shrinkage are usually significant and surgery in this group is reserved for patients who do not tolerate the DA agonists. For ACTH-secreting adenomas, primary therapy is generally trans-sphenoidal surgery and radiotherapy, with adrenal steroidogenesis inhibitors reserved for patients who are subtotally resected or remain hyper-secretory after surgery. For NFPAs, trans-sphenoidal surgery with or without postoperative radiation therapy is performed for almost all patients particularly if they have visual symptoms related to their tumor.¹³

Although usually very effective, pituitary surgery does not provide cure in all patients. Recurrent rates are significant in secretory pituitary adenomas (recurrence is highest in patients with a prolactinoma).¹⁴ The highest incidence of tumor recurrence is between 1 and 5 years after surgery.

The recurrent rate of NFPAs is about 33.5%.⁸ Even after complete or near complete surgical resection, NFMAs regrow in 12-58% of patients within 5 years.^15-17 Size of the post-operative tumor remnant and length of follow-up are the two major determinants of recurrence/regrowth.¹⁸ The presence of a tumor with an extrasellar remnant was associated with the highest risk of recurrence (odds ratio 3.73 [CI: 1.97-7.09]).⁸ Tumor recurrences or new growth from residual tissue of such cases may occur (8-40% over 10 years).¹¹ For invasive pituitary adenomas, stereotactic radiosurgery is now been recognized as a useful method of treatment, especially in those tumors extending into the suprasellar and cavernous sinus regions. More than 95% of pituitary adenoma patients have either tumor shrinkage or stabilization and biochemical remission is possible in approximately 80% of properly selected patients with hormone-producing pituitary adenomas after radiosurgery.¹² The antitumoral efficacy for nonsecreting tumors is observed in about 90% of cases. The time to remission is estimated to range from 12 to 60 months.¹⁹ The majority of patients in our study had nonfunctional adenomas (17/22) with different degree visual field deficit due to optic chiasm compression. Following surgical interventions, residual tumors were treated with CyberKnife radiosurgery and most of patients were either stabilize or regress with minimal side effects during the study period.

The visual field defects were found 17 out of 22 patients in our series and there was no newly defect after the CyberKnife treatment within these patients. One patient (who received could not tolerate radiosurgery and therefore had an incomplete treatment) was not enrolled to this series and she had worsening new visual defect after the radiosurgery because tumor progression. Visual loss after radiosurgery is rare if the maximum radiation dose to the optic apparatus is kept below 12 Gy delivered in a single session.¹² The patients with post- radiosurgery exacerbation of visual defect within the studies (²⁰~²¹) and in our series all appeared to have visual worsening related to uncontrolled tumor growth or cystic enlargement instead of direct radiological toxicity. In our study, the local control rate approximated 95.5% after a mean follow up period of 34 months. Compared to the reports from other centers(²⁰~²¹), our therapy regiment of 5 Gy x5 time could also provide good local control rate. (Table 5)

The tolerance of radiation of the adjacent critical organ (ex. optic chiasm) is important. As a general rule, every effort was made not to exceed a maximum of 8 Gy per session to any portion of the anterior visual pathway, when this structure was displaced and could not be delineated separately from tumor, it was generally impossible to meet this objective. The maximal number of sessions could be used to the cases with the longest involvement of the optic apparatus and where the nerve or chiasm was most displaced and as a result, could not be clearly distinguished (contoured) on imaging studies.²⁰ So it was generally possible to keep the single session dose to the visible portions of the visual pathways to less than 5 Gy in this situation. In our study, there was no newly visual field deficit formation after CyberKnife therapy. The dosing schema of 5 consecutive daily sessions of 5 Gy that we used is similar to Killory et al²² and this dosing is biologically equivalent to a single- fraction dose of 13.2 Gy when α/β coefficient was set of 3 respectively by using the linear quadratic formula.²³

Only one patient in our series had the transient elevated prolactin level after Cyberknife treatment, and no new hormonal deficits occurred in nonfunctional tumor patients during our following period. CK radiosurgery resulted in complete biochemical remission of prolactinoma in 4/4 (100%) subjects, and in biochemical control with the concomitant use of a somatostatin analog in an additional subject. Delayed anterior pituitary deficits occur in 20-50% of patients depending on the length and quality of the endocrine follow-up.¹² In contrast to the optic apparatus, the pituitary gland usually embed within or compressed by the tumor, which may be vulnerable to the radiation and result in newly post-therapeutic hormone disturbance.

In addition to antisecretory drugs, particularly for acromegaly and prolactinomas, stereotactic radiosurgery plays an increasing role in the treatment of these tumors. The overall disease control rate was approximately 48% without suppressive medications after radiosurgery for acromegaly.²⁴ Long-term data confirms the antisecretory efficacy of stereotactic radiosurgery with the Gamma Knife in secreting and nonsecreting pituitary adenomas (about 50% remission in hypersecreting tumors) but also a previously unknown low risk of recurrence (2-10% of cases).

Compared with conventional radiotherapy, the incidence of hypopituitarism appears lower with radiosurgery. For residual or recurrent pituitary adenomas after surgery, traditional radiotherapy remains an important treatment option after failure of surgery. Although it is effective in the long-term tumor- and hormone hypersecretion control of ACTH-secreting pituitary adenomas, high prevalence of hypopituitarism developed in 62% and in 76% of patients at 5 and 10 years after conventional radiotherapy.²⁵ Hypopituitarism following radiosurgery may correlate with the radiation dose to the pituitary stalk,²⁶ and certain normal adenohypophysis cell types are more susceptible to radiation than others.²⁷ Well-respected groups have reported a low incidence (0-36%) of pituitary dysfunction following radiosurgery.^{4, 28-30}

The radiosurgical dose for treatment of a pituitary adenoma is still a controversial issue and current recommendations are for single- fraction stereotactic radiosurgery doses of 15 to 18 Gy for nonfunctioning adenomas and of as much as 30 Gy or more for functioning adenomas.^31-33 The marginal and maximal dose and the number of sessions were determined by both attending neurosurgeon and radiation oncologist while this decision was influenced by a multitude of factors including tumor volume, proximity and extent of irradiated optic nerve, as well as a previous history of radiation therapy.²⁰ We choose the dosage 5 Gy x 5 fractions with considering the following reasons: 1) the tumor location and tumor shape, 2) safety of margin and diminished the side effect, and 3) the tumor control rate. Although functional tumor probably need higher dose of radiosurgery, but in our series, the functional tumor response well by using 5 X 5 Gy protocol and local control rate is high so far. It is too early to conclude that our protocol is totally suitable for functional tumor.

Figure 5.

The representative case (Case No.14, 39-y-o female with medical uncontrolled prolactinoma) of stable disease .The large post-transsphenoidal surgery remnant was noted because right cavernous sinus invasion and carotid artery encasement (A). The tumor did not change size and shape during following up post-treatment 3 months(B,F,J); 6months later(C,G,K); 12 months (D,H,L)and; 24 months later(E,I,M).

Previous reports indicated that radiosurgical indices may be affected by target volume.³⁴ To investigate the relationship between tumor volume and radiosurgical parameter, we calculated the homogeneity index (HI), percentage of target coverage, conformity index (CI) and new conformity index (nCI) in each therapeutic planning (Fig.6). The simple linear regressions on the tumor volume for each of four indices were performed. Pearson’s correlation coefficients of all indices were also calculated by using linear regression analysis. The estimated slopes for HI and percentage target coverage are near zero (Fig. 6). Pearson correlation coefficients of HI and percentage target coverage were less than ± 0.4 suggesting a poor correlation between these examined variables. Therefore, tumor volume does not appear to markedly affect HI and Percentage target coverage when using the CyberKnife radiosurgical treatment planning system in our patient population. But the correlation coefficients of CI and nCI were more than ± 0.4 suggesting a correlation between these examined variables. Our data revealed that HI and the percentage of tumor coverage are unrelated to the tumor volume, but not the CI and nCI That is, in our study, as tumor volume increases, a more concise conformity for radiation dosage delivery is achieved.

Figure 6.

The relationship between target volume and radiosurgical indeces, using scatter plots constructed from data obtained from CyberKnife planning system. There is no significant relationship between HI, percentage target coverage and target volume (A and B); but CI (C, r=0.43, p<0.05) and nCI(D, R=0.50, p<0.05) negatively correlate to target volume.

In comparing our clinical result and treatment related-morbidity with other reports (Table 3), our data compatible with the conclusion made by Kajiwara³⁵ which indicated that image-guided stereotactic radiosurgery/radiotherapy with the gamma knife, CyberKnife, or LINAC system is effective and safe against pituitary adenomas.

Table 3.

Summary of radiosurgical treatment.

Patient No.	Cyberkinfe dosing (Gy)	Treatment IDL(%)	Gross tumor volume (cm3)	% tumor covered with treatment IDL	Conformity index	Tumor dose/fraction (Gy)		Max chiasm dose/ fraction (Gy)
						Max	Min
1	5x5	73%	7.7	95.0%	1.3	6.9	2.3	2.3
2	5x5	73%	2.9	95.2%	1.5	6.9	3.4	2.6
3	5x5	80%	16.0	94.2%	1.3	6.3	3.3	2.6
4	5x5	80%	6.0	90.2%	1.2	6.3	2.6	2.9
5	5x5	78%	2.9	95.5%	1.3	6.4	3.5	2.7
6	5x5	78%	6.1	95.2%	1.4	6.4	3.8	2.5
7	5x5	83%	7.0	93.0%	1.3	6.0	4.1	2.4
8	5x5	70%	1.5	93.4%	1.8	7.1	4.0	2.2
9	5x5	75%	3.3	93.1%	1.5	6.7	3.3	2.7
10	5x5	70%	1.0	95.8%	1.4	7.1	4.6	2.5
11	5x5	80%	4.5	95.6%	1.4	6.3	4.1	2.2
12	5x5	75%	3.9	95.3%	1.4	6.6	4.3	2.6
13	5x5	75%	8.0	96.0%	1.3	6.7	3.2	2.5
14	5x5	80%	6.6	95.8%	1.4	6.3	3.1	2.4
15	5x5	74%	2.2	95.3%	1.3	6.8	4.3	2.5
16	5x5	75%	3.9	95.4%	1.5	6.6	4.0	3.4
17	5x5	76%	4.7	96.2%	1.4	6.6	4.4	1.5
18	5x5	76%	4.4	96.3%	1.5	6.6	4.2	2.1
19	5x5	78%	1.7	95.5%	1.4	6.4	4.4	2.3
20	5x5	80%	1.6	98.2%	1.4	6.2	4.8	2.5
21	5x5	71%	10.2	92.4%	1.3	7.0	2.4	2.6
22	5x5	80%	0.8	99.1%	1.4	6.3	4.9	2.1

IDL, isodose line; max, maximum; min, minimum.

CONCLUSION

Stereotactic radiosurgery performed with the CyberKnife is characterized by a high precision and a sharp dose fall-off at the edge of the target, together with excellent target coverage and dose conformity to the target volume. The CyberKnife has emerged as an adjuvant radiation-based therapy for pituitary adenomas and should be considered as a therapeutic option for recurrent or residual adenoma. Our result suggests that multisession radiosurgery of 25Gy in 5 fractions could achieve satisfactory radiographic and endocrinological tumor control.