CLINICAL STUDY

The feasibility of frameless stereotactic radiosurgeryin the management of pediatric central nervous system tumors

Ronica Nanda • Anees Dhabbaan • Anna Janss •

Hui-Kuo Shu • Natia Esiashvili

Received: 15 August 2013 / Accepted: 26 January 2014

� Springer Science+Business Media New York 2014

Abstract Recurrent malignant primary and metastatic

central nervous system (CNS) tumors in pediatric patients

are devastating, and efforts to improve outcomes for these

patients have been disappointing. Conventional re-irradia-

tion in these patients increases the risk of significant tox-

icity. We therefore evaluated feasibility and outcomes

using frameless radiosurgery (FRS) in children with

recurrent primary and metastatic brain tumors. We

reviewed five cases of recurrent primary and metastatic

brain tumors treated with frameless radiosurgery between

2008 and 2013. We analyzed safety and feasibility, dosi-

metric data, local control, and adverse effects. Five patients

were treated with frameless radiosurgery for palliation.

Fifteen target volumes were treated using our institutional

FRS system. The volumes of targets ranged from 0.08 to

51.67 cm3 with doses ranging from 15 to 21 Gy. Radio-

surgery was well tolerated, decreased the need for large-

volume CNS irradiation, and allowed for effective pallia-

tion in this small cohort. Frameless radiosurgery is feasible

in this patient population. Frameless radiosurgery should

be considered in management of select patients with

recurrent primary or metastatic brain tumors.

Keywords Frameless � Radiosurgery � CNS � Pediatric �Image-guided � Palliation � Brain metastases

Introduction

Pediatric central nervous system (CNS) tumors, repre-

senting 20 % of all pediatric tumors, are difficult to treat at

initial presentation. Maximally safe surgery, often limited

by adjacent critical structures, and adjuvant chemoradia-

tion are standard of care at initial presentation of most

high-grade malignant tumors [1]. However, they represent

a significant treatment dilemma when they recur due to

structural changes from prior surgeries and radiation [2, 3].

Radiation therapy, like surgery, presents treatment diffi-

culties both in the upfront and the relapse setting, due to

normal tissue dose constraints. In children, the risks of

radiation to the CNS and head and neck are amplified in

their yet developing vital organs. Moreover, in the pediatric

population, additional considerations, such as the admin-

istration of daily anesthesia for radiation treatments, pres-

ent their own challenges.

Stereotactic radiosurgery (SRS) is playing an increas-

ingly frequent role, and in pediatrics can help alleviate the

need for daily, long-term, anesthesia. This modality is

gaining recognition in adult patients as an alternative to

whole brain irradiation with metastatic brain tumors [4]

and in some cases for relapsed primary CNS tumors. In

children, SRS can be an attractive option in selected cases

to avoid or delay craniospinal irradiation (CSI) or whole

brain radiation therapy (WBRT) and resultantly minimize

associated toxicities. However, SRS also has its disad-

vantages, including the need for a fixed head frame and

scheduled operating room time, requiring additional

resources. Conventional hypofractionated RT on the other

hand requires larger margins due to reduced precision

compared with SRS, with consequent increased normal

tissue irradiation. Frameless radiosurgery, which addresses

the above concerns, therefore represents an attractive

R. Nanda � A. Dhabbaan � H.-K. Shu � N. Esiashvili (&)

Departments of Radiation Oncology, Winship Cancer Institute at

Emory University, 1365 Clifton Road, Atlanta, GA 30322, USA

e-mail: nesiash@emory.edu

A. Janss

Pediatric Oncology, Children’s Healthcare of Atlanta, AFLAC

Cancer and Blood Disorder Center, Atlanta, GA, USA

123

J Neurooncol

DOI 10.1007/s11060-014-1392-7

option in treating children. The frameless option has pre-

cision on the order of what is achieved with conventional,

framed SRS, while potentially permitting hypofractionated

treatments [5]. In this report, we reviewed five pediatric

patients at our institution that received frameless single

fraction or hypofractionated FRS for recurrences of pri-

mary intracranial tumors and for metastatic brain disease.

The techniques used, dosimetric data, and treatment out-

comes will be discussed.

Methods

Five consecutively pediatric patients treated in our insti-

tution to date with frameless radiosurgery for intracranial

tumors were included in this retrospective review. All

patients had treatment to the primary disease site prior to

relapsing distantly.

The frameless radiosurgery technique was chosen to

avoid placement of a frame, and to facilitate hypofrac-

tionation when needed. It also allowed more rapid initia-

tion of therapy and a shorter duration of anesthesia

administration. Patients in this series were followed

through June 1, 2013 or date of death, and no patients were

lost to follow-up.

Frameless technique

In this study, one of two frameless systems was used to

position pediatric patients. Prior to 2010, an optically guided

frameless positioning system and in-room kV imaging sys-

tem for cone beam computer tomography (CBCT) was used

for patient positioning. The patients were initially positioned

for treatment delivery with a custom molded bite block

using the frameless array of optical markers and optical

guidance camera system. The bite block was designed to

allow airway access, including the use of an laryngeal mask

airway or endotracheal tube as needed. As an image verifi-

cation check, in-room CBCT was acquired and rigidly reg-

istered to the planning computer tomography (PCT) using

the Varian 3D review system. Patient positioning setups

were compared for the optical guidance versus CBCT and

the attending radiation oncologist would make the final

decision regarding positioning based on this information.

In 2010, an in-house 6 degrees of freedom (6DOF)

radiosurgery frameless system was developed. This system

does not use the bite block, which makes it more suitable

for pediatric patients. Using this system, patient motion

during setup and treatment is restricted by a custom ther-

moplastic mask. Accurate positioning is achieved by

matching an in-room CBCT to the PCT using a 6DOF rigid

registration method customized to use mutual information

metric in Mattes’ formulation [5] (Fig. 1). This registration

system calculates not only three translational shifts but also

three angles of rotations. Couch translations and rotation

are applied using the treatment console while pitch and tilt

are applied using a customized couch mount. The posi-

tioning accuracy of this system was evaluated and found to

be comparable to that of the frame-based stereotactic sys-

tem [5]. For this frameless setup, patients are first aligned

using the laser marks. Next, AP and lateral kV images are

taken, registered to their corresponding DRRs, and used to

calculate translational couch shifts based on the kV-DRR

match. Once the patient’s position is approximated, an

initial CBCT is then obtained and registered to the PCT.

Based on this registration, rotational adjustments (tilt and

spin) using the customized couch mount are made. Finally,

a second CBCT is obtained and registered to the PCT. This

registration permits the final shifts and couch rotations to

be applied. After final position verification by the attending

radiation oncologist, treatment delivery is initiated.

Patients were planned and treated either with dynamic

conformal arcs (DCA) or with an IMRT technique. A

planning CT with 0.625 mm slice thickness and a fine res-

olution magnetic resonance imaging (MRI) study was

obtained for each patient. The sequence that allowed the best

visualization of the tumor extent was used, which in most

cases was a fine resolution T1 post-contrast sequence. Other

sequences were used as indicated to verify our target vol-

umes. The planning CT and the selected MRI sequence were

registered to allow for MRI-based target delineation. The

attending radiation oncologist and neurosurgeon outlined

target volumes and critical structures during the planning

process. A 1-mm margin was added to the tumor to create a

planning target volume. In a well-defined metastatic tumor

on imaging, there is considered to be no subclinical spread

and therefore no clinical target volume is created. When

DCA were used, the multileaf collimator (MLC) automati-

cally conforms to the target volume outlines over the arc

path via software optimization. A typical DCA plan would

average 100 degrees per arc over four non-coplanar arcs. For

cases treated with IMRS, 12 static fields with modulation

with a sliding window technique were used to achieve good

conformity (Fig. 2). The plans were evaluated using

benchmarks such as isodoses and conformity indices (CIs) to

quantify the target coverage and normal tissue sparing

(Table 2). For this study, CIs were defined as volume

receiving the prescription dose divided by the PTV volume.

Results

Patient and tumor characteristics

Tumor types included posterior fossa ependymoma (two

patients), posterior fossa atypical teratoid rhabdoid tumor

J Neurooncol

123

(one patient), medulloblastoma (one patient), and meta-

static Ewing sarcoma (one patient). All patients were ini-

tially treated with multimodality therapy and by current

Children’s Oncology Group (COG) protocols respective to

their diagnoses. Patients with non-disseminated primary

CNS malignancies (ependymoma, ATRT) initially under-

went maximally safe resection followed by chemotherapy

and intensity-modulated radiotherapy (IMRT) to

54–59.4 Gy to the primary tumor site. The patient with

medulloblastoma underwent gross total resection followed

by chemotherapy alone given young age at presentation

(1 year). Craniospinal irradiation was delayed, despite M1

disease at presentation until the patient reached age 3, in

order to avert major toxicities and after discussion with the

patient’s caregivers. Finally, the Ewing sarcoma patient

presented with primary tumor of the femur with metastases

to the bilateral hips and was treated with neo-adjuvant and

adjuvant chemotherapy and had the primary site addressed

by radical resection with prosthesis placement. This was

eventually followed with definitive radiation therapy to a

total dose 55.8 Gy to the metastatic lesions.

At time of intracranial recurrence or metastases, patients

were typically treated with multimodality therapy, includ-

ing surgical resection, radiation therapy, and chemother-

apy. Table 1 summarizes patient and disease characteristics

as well as treatment and outcome data. Patients received

frameless radiosurgery to up to eight lesions over time.

There were three in-field recurrences after radiosurgery (on

the older, optically guided system), but these were suc-

cessfully salvaged with additional radiosurgery or surgical

Fig. 1 Set up of patient using former system with bite block and cone beam registration to ensure set up accuracy

Fig. 2 Occipital lesions before FSR (a), FSR dose plan (b) and 1 year post-FSR (c)

J Neurooncol

123

resection. Surgical salvage was offered in a few cases

depending on the size of the lesion and overall prognosis of

the patient. Radiosurgery was offered in the adjuvant set-

ting in some cases (such as in subtotal or near total

resections) to improve local control if patients had

acceptable post-surgical morbidity.

Frameless SRS treatment details

Patients with primary brain tumors, without metastatic

disease at presentation, developed intracranial recurrences

distant from initial site of disease 33-37 months after initial

surgery and chemo-radiation. The patient with metastatic

Ewing sarcoma developed a brain metastasis 11.5 months

after initial diagnosis.

General anesthesia for FRS was used in four of the five

cases due to the young age of patients at the time of

treatment. In total, 15 volumes were treated using our

institutional frameless SRS system. All treatments were

performed on the Novalis Tx treatment unit (Varian, Palo

Alto, CA, USA) utilizing a high definition multileaf colli-

mator (leaf width at isocenter is 2.5 mm). Doses used

ranged from 15–20 Gy for single fraction treatments and

18–21 Gy for hypofractionated treatments. Dose and frac-

tionation decisions were, similar to adult radiosurgery

treatments, largely based on patient age, tumor volume

(favoring fractionation or lower doses for larger tumor

volumes), proximity to normal organs at risk, and risk to

previously irradiated tissue. Radiation specific treatment

details are included on Table 2.

Dose to normal structures

Maximum dose to critical structures was consistently less than

3–5 Gy per treatment, unless they were part of the PTV, and in

most cases was less than 1 Gy. These structures include the

cochlea, eyes, lenses, hypothalamus/pituitary, hippocampi,

brainstem, optic nerves/chiasm, and spinal cord, when rele-

vant. The brainstem and optic nerves/chiasm are especially

important to keep within tolerance. In one exceptional case,

maximum dose to the brainstem was 21.4 Gy in a single

treatment (mean and median doses of 4.3 and 3.2 Gy,

respectively), resulting in a sum dose of 32.5 Gy, but the

volume of this high-dose region was kept within standard

tolerances. No clinical signs of toxicities to these structures

were noted following radiosurgery. The maximum sum doses

to critical normal tissues are summarized on Table 3. These

were determined by registration of consecutive FRS plans to

determine cumulative doses to normal structures.

Outcome

Of the fifteen volumes treated, there were three in-field

failures after radiosurgery to a given lesion, treated using

the optically guided system, but these were successfully

salvaged with FRS (in two cases) or surgery (in one case).

Both patients with ependymoma did require multiple FRS

treatments, but to distinct regions every time. There were

no other in-field failures in any patients; however, one

patient required craniospinal irradiation 25 months after

initial FRS due to disease progression (ependymoma), and

one patient required whole brain radiation 6 months after

FRS due to leptomeningeal disease (Ewing sarcoma case).

He passed away shortly thereafter. At time of last follow-

up, two patients were alive with disease. The two other

patients with primary CNS tumors passed away from

causes unrelated to treatment and had no known residual

disease at time of death. Disease status and patient outcome

details are defined in Table 1.

Adverse effects

No patient experienced new major deficits or toxicities that

could be directly attributable to radiosurgery. Acute

Table 1 Disease characterisitcs and treatment outcome

Pt. Age at

diagnosis

Tumor

type

Time to first relapse/

metastasis

Time to first

radiosurgery

Duration of local

control

Time to CSI/

WBRT

Survival Length of

follow-up

#1 23 mo EP 21 mo 33 mo 25 mo N/A 32 mo 32 mo

(deceased)

#2 16 mo EP 33 mo 33 mo 29 mo 25 mo 45 mo 45 mo (alive)

#3 11 mo ATRT 34 mo 34 mo 24 mo N/A 26 mo 26 mo

(deceased)

#4 23 mo MB N/A 9 mo 1 mo N/A 1 mo 1 mo (alive)

#5 11 yr EWS 4 mo 4 mo 6 mo 6 mo 10 mo 10 mo

(deceased)

Local control, time to CSI/WBRT, survival, and length of follow up determined from time of first frameless radiosurgery treatment. Time to first

relapse/metastasis: from completion of primary treatment

Number of lesions total # of lesions treated with FRS, Mo months, Yr years, EP ependymomas, ATRT atypical teratoid rhabdoid teratoma

J Neurooncol

123

toxicities were minimal (grade 1) and self-limited. One

patient, treated for an ependymoma, developed radio-

graphic changes consistent with radiotherapy effects

(‘‘radionecrosis’’) in the lateral ventricles but did not

manifest any clinical symptoms and did not require inter-

vention. No other patients were noted to develop radio-

graphic sequelae from radiosurgery. The patient treated for

metastatic Ewing sarcoma developed persistent mild to

moderate headaches months after FRS, resulting in steroid

dependence. The etiology of this remains unclear given the

lack of radiographic correlates to suggest radiation necro-

sis. He did shortly thereafter develop leptomeningeal dis-

ease and it is possible that his headaches were associated

with subclinical disseminated disease. No acute or late

Table 2 Treatment overview

Patient Initial

tumor

Recurrence

no.

Treatment Lesion

previously

treated?

PTV

volume

(cc)

Conformity

index

No of

fields

Dose (Gy) Fractions Technique

1 PF EP Surgery, RT,

chemo

No N/A N/A N/A 59.8 33 IMRT

1 Surgery No N/A N/A N/A N/A N/A N/A

2 FRS No 0.22 2 6 20 1 DCA

2 FRS No 2.69 1.02 4 20 1 DCA

3 Fractionated RT

(spine)

No N/A N/A N/A 45 25 IMRT

3 FRS No 2.29 1.61 4 20 1 DCA

3 FRS No 2.52 1.96 4 20 1 DCA

3 FRS No 1.56 1.96 4 20 1 DCA

4 Chemo No N/A N/A N/A N/A N/A N/A

2 PF EP Surgery, RT,

Chemo

No N/A N/A N/A 59.8 33 IMRT

1 Surgery, FRS No 1.28 1.84 4 15 1 DCA

2 CT, Surgery,

FRS

Yes 1.09 2.04 4 15 1 DCA

3 FRS No 0.33 1.53 4 18 1 DCA

3 FRS No 0.15 1.15 12 18 1 IMRS

4 CSI ? 4

fractionated

RT boost

Yes N/A N/A N/A 54 30 IMRT

5 Fractionated

FRS

No 1.24 0.49 12 18 3 IMRS

6 FRS No 0.15 1.15 12 15 1 IMRS

7 FRS No 0.39 2.23 12 15 1 IMRS

8 FRS No 0.01 1.69 12 15 1 IMRS

3 PF

ATRT

Surgery, RT,

Chemo

No N/A N/A N/A 54 30 IMRT

1 FRS No 5.82 1.7 4 18 1 DCA

2 Surgery Yes N/A N/A N/A N/A N/A N/A

4 Pelvic

ES

Surgery, RT,

Chemo

No N/A N/A N/A 55.8 31 IMRT

1 CT, surgery,

FRS

No 1.6 1.52 12 21 3 IMRS

2 Whole brain RT No N/A N/A N/A N/A N/A N/A

5 PF MB Surgery, RT,

Chemo

No N/A N/A N/A 36 (CSI)

54

(Boost)

30 CSI?IMRT

1 FRS No 51.67 1.16 12 21 3 IMRS

EP ependymomas, ATRT atypical teratoid rhabdoid teratoma, MB medulloblastoma, ES Ewing’s sarcoma, RT radiation therapy, FSR fractionated

radiosurgery, CSI cranio-spinal irradiation, PTV planning target volume, DCA dynamic conformal arcs, IMRS intensity-modulated radiosurgery,

IMRT intensity-modulated radiation therapy

J Neurooncol

123

grade 3 or 4 toxicities were noted. In addition, no com-

plications associated with general anesthesia were noted.

Discussion

Typical management of most malignant CNS tumors

involves maximally safe resection followed by external

beam radiation therapy. However, recurrences of these

tumors still remain an important problem. Repeated sur-

geries carry the risks of anesthesia, surgical complications,

and prolonged recovery. In patients with recurrent disease,

prognosis and survival is often limited and therefore

treatment options need to be focused not only on local

control and survival, but also on quality of life for these

patients. Likewise, most available chemotherapies have

minimal effect on CNS tumors, in addition to the incon-

venience of repeated administrations and side effects such

as nausea and vomiting, among others. Although re-irra-

diation to the CNS does raise concern for significant tox-

icity, it can be considered for palliation in select cases.

Given toxicity, wide-field salvage irradiation, such as

whole brain or craniospinal irradiation, may not always be

the best choice, particularly for very young children, and

should be chosen carefully after consideration of potential

adverse outcomes.

Depending on the tumor histology or clinical situation,

focal radiation therapy may be a good option for salvage.

In a series by Merchant et al. [2], focal fractionated

radiotherapy had promising preliminary results with good

disease control and overall survival (*67 % at 5 years).

However, delivery of regular daily fractionated radiother-

apy still requires an extended period of treatment and may

be challenging, especially given the need for daily sedation

and the relatively limited life expectancy for these patients

at the time of recurrence. Alternatively, conventional

frame-based SRS has been used in children with relapsed

brain tumors. Good local control rates in pediatric patients

with a range of brain tumor histologies treated with this

modality have been reported [6]. Thus, SRS may be a

useful means of treating selected patients with focal ther-

apy. However, the need for a fixed frame with associated

trauma to the cranium, the potential need for general

anesthesia for an extended period of time for treatment

planning, increased resource utilization, and strict target

volume constraints for single fraction treatments limits the

use of frame-based radiosurgery. These limitations are

addressed with frameless radiosurgery.

Hypofractionated radiation for recurrences is also being

used in order to minimize the number of treatments com-

pared with conventional fractionation, especially when

single-fraction radiosurgery may not be feasible due to

tumor volume or dose constraints. Hypofractionation, as

opposed to conventionally fractionated palliation regimens,

not only adds convenience for the family, but also reduces

the need for daily anesthesia. Several series [5, 7] have

suggested that hypofractionation can offer effective palli-

ation and short-term local control, without significant

toxicities. Improving precision of hypofractionated radio-

therapy and reduction of margins around the target can

further improve the outcome, and frameless, rather than

frame-based, radiosurgery can offer the benefits of hypo-

fractionation as well.

Frameless radiosurgery presents an alternative to frame-

based radiosurgery in select cases for treating children with

recurrent intracranial tumors. In addition to the aforemen-

tioned limitations of the frame itself, it may also obstruct

proper positioning of the head, depending on the location

of the tumor [8]. Preliminary results of feasibility and

accuracy have been promising, with outcomes comparable

to those of frame-based SRS [7]. Studies from our own

institution regarding accuracy are described in detail by

Dhabaan [5] and indicate that while both optical-guided

systems and the image-guided systems provide sub-milli-

meter accuracy, the latter is preferred as it is less invasive.

Kamath et al. [9] also found this technique to be accurate

and reproducible with sub-millimeter accuracy in a group

of adolescents and adults. Use of frameless radiosurgery in

patients with brain metastases appears to be safe [10], and

may help defer the need for whole brain or craniospinal

irradiation while maintaining a reasonable quality of life in

this patient population. This study is one of only two to

examine the role of frameless radiosurgery specifically in

pediatric patients. Our institutional data to date suggests

that frameless stereotactic radiosurgery is a feasible alter-

native to whole brain radiation or frame-based radiosurgery

as part of a strategy to provide local control of intracranial

masses in this population. We found no grade 3 or higher

toxicities in any patient treated with frameless radiosur-

gery, and patients demonstrated excellent tolerance to this

technique. Although there were three in-field failures for

the duration of follow-up, these were done using older

Table 3 Sum doses to critical structures (Gy)

Patient Brainstem L. optic

N

R. optic

N

L.

eye

R.

eye

Chiasm

#1 2.30 0.25 0 0 0.40 0.35

#2 98 56.5 56 57 50 60

#3 0.30 0.10 0 0 0.10 0.10

#4 0.09 0.02 0.02 0.01 0.01 0.02

#5 32.50 12.50 13 12 10.50 13.00

Sum doses determined by registration of successive radiosurgery

plans and analysis of cumulative doses to critical structures

J Neurooncol

123

systems and were successfully salvaged with additional

therapy.

Unfortunately, in-field control of the treated lesions did

not translate to adequate overall CNS control, with all

patients eventually succumbing to disease, or continuing to

progress, similar to findings by Kelly et al. [10]. It is

unlikely however that upfront whole brain radiation ther-

apy would have averted this eventual outcome in these

patients given its palliative nature. Treatment with frame-

less radiosurgery did help minimize the volume of normal

brain tissue that was irradiated upfront by delaying whole

brain radiation therapy or CSI until progression to multi-

focal disease, or possibly by averting the need for it alto-

gether. FRS allowed for multiple courses of CNS

reirradiation to focal lesions, with low sum doses delivered

to critical structures (Table 3). Although the data is limited

in this small series and definitive conclusions cannot be

made, this approach may have helped decrease potential

neurotoxicity and therefore maintained or improved quality

of life in this population. Moreover, by minimizing overall

treatment time compared with WBRT or CSI, FRS

potentially allows for greater quality of life in patients with

limited life expectancy.

Overall, intracranial tumors represent difficult to control

disease, and even more so with recurrent primary brain

tumors or metastases from extracranial sites. Prognosis

remains poor, with high rates of morbidity and mortality

despite aggressive therapy. Therapy to improve both local

and distant control remains a priority of further research

investigations. Frameless radiosurgery and hypofractiona-

tion methods are feasible in pediatric patients and represent

reasonable options for palliation in this population. Patients

with primary brain tumors should be carefully selected for

frameless radiosurgery on an individual basis with con-

sideration of patterns of disease failure and progression, as

well as toxicity risks and quality of life factors. Further

investigation in a rigorous, controlled setting may be

warranted to help further define the role of this technique in

the pediatric population.

Disclosures The authors have no conflicts of interest or funding to

disclose.

References

1. Albright AL, Pollack IF, Adelson PD (2008) Principles and

practice of pediatric neurosurgery, vol xxiii, 2nd edn. Thieme,

New York, p 1286

2. Merchant TE et al (2008) A retrospective study of surgery and

reirradiation for recurrent ependymoma. Int J Radiat Oncol Biol

Phys 71(1):87–97

3. Sangra M et al (2009) Management strategies for recurrent

ependymoma in the paediatric population. Child’s Nerv Syst

25(10):1283–1291

4. Halasz LM et al (2012) Use of stereotactic radiosurgery for brain

metastases from non-small cell lung cancer in the United States.

Int J Radiat Oncol Biol Phys 85(2):e109–e116

5. Dhabaan A (2012) Six degrees of freedom CBCT-based posi-

tioning for intracranial targets treated with frameless stereotactic

radiosurgery. J Appl Clin Med Phys 13(6):3916

6. Bauman G et al (2004) Radiotherapy for pediatric central nervous

system tumors: a regional cancer centre experience. J Neurooncol

68(3):285–294

7. Breneman JC et al (2009) Frameless image-guided intracranial

stereotactic radiosurgery: clinical outcomes for brain metastases.

Int J Radiat Oncol Biol Phys 74(3):702–706

8. Pollack IF et al (1995) The role of frameless stereotactic tech-

niques in the treatment of pediatric CNS lesions. Pediatr Neurol

13(3):209–216

9. Kamath R et al (2005) Initial clinical experience with frameless

radiosurgery for patients with intracranial metastases. Int J Radiat

Oncol Biol Phys 61(5):1467–1472

10. Kelly PJ et al (2012) Stereotactic irradiation of the postoperative

resection cavity for brain metastasis: a frameless linear acceler-

ator-based case series and review of the technique. Int J Radiat

Oncol Biol Phys 82(1):95–101

J Neurooncol

123