- Email: [email protected]
Tumor volume threshold for achieving improved conformity in VMAT and Gamma Knife stereotactic radiosurgery for vestibular schwannoma
Radiotherapy and Oncology xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Radiotherapy and Oncology journal homepage: www.thegreenjournal.com
Original article
Tumor volume threshold for achieving improved conformity in VMAT and Gamma Knife stereotactic radiosurgery for vestibular schwannoma Hyun Kim a, Peter Potrebko a, Amanda Rivera a, Haisong Liu b, Harriet B. Eldredge-Hindy a, Vickie Gunn b, Maria Werner-Wasik a, David W. Andrews b, James J. Evans b, Christopher J. Farrell b, Kevin Judy b, Wenyin Shi a,⇑ a
Department of Radiation Oncology; and b Department of Neurological Surgery, Sidney Kimmel Medical College at Thomas Jefferson University, Philadelphia, United States
a r t i c l e
i n f o
Article history: Received 12 January 2015 Received in revised form 12 March 2015 Accepted 27 March 2015 Available online xxxx Keywords: Radiosurgery Vestibular schwannoma Gamma knife VMAT
a b s t r a c t Background and purpose: Recent advances in multileaf collimator field shaping technology and inverse planning software have resulted in highly conformal LINAC based stereotactic radiosurgery (SRS) plans with minimal dose to critical structures. This modeling study compares Gamma Knife (GK) and LINAC SRS for vestibular schwannoma (VS). Materials and methods: 76 treatment plans from nineteen patients with VS were planned using GK forward planning and volumetric arc therapy (VMAT) inverse planning software. VMAT plans were generated with 1 coplanar, 3 and 5 non-coplanar arcs. Dose to normal structures and beam-on time (dose rate 600 MU/min) were compared using Kruskal–Wallis and Dunn’s post hoc test. Results: Median tumor volume was 1.2 cm3 (range 0.1–4.8 cm3). A peripheral tumor dose of 12 Gy was prescribed. Tumor coverage was >99.8%. VMAT plans had lower target D2% and mean dose, as well as decreased beam-on time, compared to GK plans (p < 0.0001). Paddick conformity index in VMAT 5 arc plans was superior to that of GK plans for targets >0.5 cm3 (p = 0.002). Similar dose to cochlea, normal brain tissue and brainstem was observed. Conclusion: VMAT should be considered as a safe, alternative modality to GK for VS SRS treatment, especially for tumors larger than 0.5 cm3. Ó 2015 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology xxx (2015) xxx–xxx
Stereotactic radiosurgery (SRS) is a common treatment option for vestibular schwannoma (VS, acoustic neuroma). SRS provides a non-surgical approach for excellent tumor control in a single treatment session and achieves a steep dose gradient that minimizes normal tissue dose. Gamma Knife (GK) (Elekta AB, Stockholm, Sweden), the traditional technique for VS SRS, is a frame-based treatment modality allowing cranial sites to be treated using cobalt-60 sources embedded within a shielded radiation unit. The newest generation model, the Gamma Knife Perfexion, uses 192 cobalt-60 sources arranged in eight sectors. The delivered dose is shaped to the target by a combination of shots generated by different collimator sizes and plugging of sectors. In GK treatment planning, treatment volumes are usually encompassed by the 50% isodose line (IDL) with a penumbral width of 1 mm, resulting in minimal dose to normal brain and cochlea [1]. Treatment time
⇑ Corresponding author at: Department of Radiation Oncology, Bodine Center, Suite G-301, 111 S. 11th Street, Philadelphia, PA 19107, United States. E-mail address: [email protected] (W. Shi).
increases with source age, with the need to replace the sources every 5 years for approximately 1/5 of the unit purchase price [2]. With the advent of new technology and planning software, gantry-mounted linear accelerator (LINAC) based SRS is an increasingly implemented alternative to GK. Traditionally, LINAC machines were not viewed as suitable for SRS due to mechanical instability and lack of precision. Current stabilization sub-devices and high precision bearings for couch and gantry now allow a radiation beam precision of up to 0.2 ± 0.1 mm [3]. LINAC SRS beam conformity is achieved through multileaf collimation (MLC), intensity-modulation and multiple non-coplanar arcs [4]. Targets are usually encompassed by the 80% IDL with a 2 mm penumbral width [1]. Appealing aspects of LINAC SRS include avoiding invasive frame-based immobilization, inverse-planning dose limits to critical structures, high dose rate settings, a wider availability of LINACs and potentially lower cost [2]. Comparisons of SRS in brain metastases and benign lesions (VS, meningiomas, arteriovenous malformations) have shown similar clinical outcome between GK and LINAC [3,5]. Long term LINAC SRS outcome data for benign lesions are lacking. Much of the long-term outcome data for VS SRS is from GK treated lesions, with
http://dx.doi.org/10.1016/j.radonc.2015.03.031 0167-8140/Ó 2015 Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Kim H et al. Tumor volume threshold for achieving improved conformity in VMAT and Gamma Knife stereotactic radiosurgery for vestibular schwannoma. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.03.031
2
VMAT vs GK SRS for vestibular schwannoma
local control of 92% at >10 years and hearing preservation in 58% (3 years) and 46% (5 years) of patients [6,7]. Increasing evidence indicates that hearing preservation can be improved by limiting dose to the cochlea and increasing the plan conformity index [8,9]. With the development of intensity modulated radiation therapy (IMRT) and volumetric modulated arc therapy (VMAT) planning, plans with improved conformity and decreased treatment times compared to conformal radiation therapy are possible [10]. Previous studies have compared various LINAC SRS techniques for the treatment of benign intracranial lesions. Authors have shown similar tumor coverage and normal tissue sparing with VMAT compared to dynamic conformal arcs [11], static beam IMRT [12], helical tomotherapy [13,14] and CyberKnife [12]. Studies comparing GK to VMAT treatment planning for VS are limited, and the two available retrospective studies are limited to 6 patients or less [15]. This study is, to our knowledge, the largest patient comparison of GK to LINAC VMAT inverse-planning with multiple non-coplanar arcs. By utilizing the tools available with LINAC treatment inverse-planning, it was possible to improve the conformity index compared to GK Perfexion plans. We also present data to suggest tumor size criteria for which VMAT SRS may result in improved dosimetric outcomes. Materials and methods Patients and treatment volumes CT and MRI data of 19 patients with VS who were treated with Novalis Tx or GK from 1994–2001, at the Jefferson Hospital for Neuroscience, were selected for the generation of VMAT and GK plans. CT simulation was obtained using 2.5 mm slice images, with patients in supine position and mask immobilization (BrainLAB). Gadolinium-enhanced 1 mm slice MRI T1 and T2 sequences were used by radiation oncologists and neurosurgeons to contour the VS gross tumor volume (GTV) and critical normal structures (brainstem, cochlea) in the GammaPlan treatment planning system (v 10, Elekta AB, Stockholm, Sweden). No margin was added to the GTV to define the planning target volume (PTV). Structures were exported via DICOM to Eclipse for VMAT planning. VMAT planning The Eclipse treatment planning system (v11.0.42, Varian Medical Systems, Palo Alto, CA) was used to generate VMAT (RapidArc) treatment plans using the progressive-resolution optimizer algorithm for a TrueBeam (Varian Medical Systems, Palo Alto, CA), 6 MV photon beam model with 120 leaf Millennium MLC (5 mm leaf width at isocenter). The final dose calculation was performed using the Anisotropic Analytical Algorithm using a 1 mm grid size and with the heterogeneity correction turned off. VMAT plans with one coplanar full arc, three non-coplanar partial arcs, and five non-coplanar partial arcs were investigated. The gantry and couch angles for the non-coplanar VMAT plans are presented in Table 1. In general, the treatment arc orientations were chosen to be incident on the side of the patient containing the target and to avoid entering through the optic structures. Furthermore, these arc orientations were developed at our institution from extensive treatment planning experience and were found to provide excellent dose conformity. Dose constraints for the tumor, cochlea, brainstem, and surrounding normal tissue (conformity ring structure) were defined for VMAT inverse planning (Table 2). Per our institutional practice, we created a 0.3 cm thick ring planning structure around the tumor with a 0.15 cm gap, and assigned dose objectives to this structure to produce excellent dose conformity and a steep dose gradient. The clinical VMAT dose
Table 1 VMAT arc orientations. 3 non-coplanar arcs
5 non-coplanar arcs
Gantry
Couch
Gantry
Couch
Right sided target
181–359 210–320 210–330
0 330 300
181–359 210–320 210–330 30–150 50–150
0 330 300 90 60
Left sided target
0–179 40–140 50–150
0 30 60
0–179 40–140 50–150 30–150 210–330
0 30 60 90 300
Table 2 VMAT dose constraints. Structure
Objective type
Volume (%)
Dose (cGy)
Priority
Tumor Ring Ring Ring Cochlea Brainstem
Lower Upper Upper Upper Upper Upper
100 30 10 0 0 0
1200 550 780 1200 750 950
100 100 50 50 50 50
constraints were based on the goals of achieving 100% target dose coverage and keeping the cochlea and brainstem maximum doses below 12 Gy and 14 Gy, respectively. GK planning The GammaPlan treatment planning system was used to generate GK Perfexion treatment plans using the Tissue-Maximum-Ratio 10 dose algorithm using a 1 mm dose grid size and skull measurements [16]. For each plan, the inverse planning algorithm was initially used to optimize target coverage, selectivity, and gradient index [17] followed by manual fine-tuning. It should be noted that the current version of the GK Perfexion inverse planning algorithm does not include the ability to define dose constraints for specific structures. Instead, the inverse planning algorithm is based on optimizing the three previously mentioned metrics for the target volume. Plan optimization and dose prescription After initial structure contouring, each plan (VMAT or GK) required approximately 30 min of manipulation, not including plan optimization by software. This time included generating the conformity ring structure, arc orientations, inverse planning constraints, and plan renormalization (VMAT) and manual adjustment of radiation shot isocenters (GK). Dose was prescribed to the 65% isodose line in VMAT plans and 50% isodose line for GK plans. The plan normalization mode in Eclipse was then used to renormalize the VMAT plans to achieve the clinical goal of 100% dose coverage of the PTV by the prescription dose of 12 Gy. Typical plan normalization values were on the order of 90%. Therefore, the normalization metrics for all the GK and VMAT plans were 99.8% to 100% target dose coverage by 12 Gy, limited by dose constraints of 12 and 14 Gy maximum to the cochlea and brainstem, respectively. Dose matrices with a 1 mm grid resolution were DICOM exported from GammaPlan to Eclipse for each structure to maintain a consistent DVH comparison between GK and VMAT plans. The dose matrix for the entire ‘skull’ was exported with a 1 mm resolution so that a DVH for the entire brain dose could be computed. Both GK and LINAC plans were reviewed by an attending radiation oncologist and clinical physicist.
Please cite this article in press as: Kim H et al. Tumor volume threshold for achieving improved conformity in VMAT and Gamma Knife stereotactic radiosurgery for vestibular schwannoma. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.03.031
3
H. Kim et al. / Radiotherapy and Oncology xxx (2015) xxx–xxx Table 3 VMAT treatment planning results in significantly shorter treatment times with similar dosimetric outcomes compared to GK.
GK VMAT 1 CA VMAT 3 NCA VMAT 5 NCA
Brain (cm3)
600 MU/min (min)
Target (Gy) D2%
Mean
D2%
Mean
D2%
Mean
V6
V8
V12
20.9 ± 7.4 8.5 ± 1.7* 8.6 ± 1.3* 9.6 ± 1.9*
22.9 ± 0.3 19.8 ± 2.5* 19.6 ± 2.1* 19.7 ± 2.2*
17.2 ± 0.5 16.1 ± 1.0* 15.9 ± 0.8* 16.0 ± 0.9*
5.1 ± 3.4 6.5 ± 2.9 5.1 ± 2.9 4.6 ± 2.6
1.4 ± 1.0 1.5 ± 0.9 1.2 ± 0.8 1.2 ± 0.7
9.3 ± 2.6 9.8 ± 1.4 8.8 ± 1.7 8.4 ± 1.8
6.3 ± 1.9 7.9 ± 1.1* 6.4 ± 1.2 6.0 ± 1.4
8.4 ± 7.4 10.4 ± 8.2 8.3 ± 6.3 7.6 ± 5.7
5.4 ± 4.8 5.9 ± 4.7 5.2 ± 4.1 4.8 ± 3.7
2.5 ± 2.3 2.5 ± 2.1 2.3 ± 2.0 2.2 ± 1.8
Brainstem (Gy)
Cochlea (Gy)
Abbreviations: GK, gamma knife; VMAT, volumetric modulated arc therapy; CA, coplanar arc; NCA, non-coplanar arcs; PCI, Paddick conformity index. * Statistically significant difference compared to GK.
Dose–volume histogram and analysis Composite dose–volume histograms (DVH) were generated for tumor, body, brainstem and cochlea. The DVH was compiled by selecting representative dose and volume points in each structure and graphing the average value with the standard error of the mean. The Paddick Conformity Index (PCI), or the product of a treatment plan’s ‘‘undertreatment’’ and ‘‘overtreatment’’ ratios, was calculated as:
PCI ¼
ðTV PIV Þ2 ðTV PIVÞ
where TVPIV is the volume of the target covered by the prescription isodose line, TV is the total target volume, and PIV is the volume encompassed by the prescription isodose line. Perfect conformity
results in a PCI of unity or 1, with potential values ranging from 0 to 1 [18]. The gradient index (GI), which provides a measure of the spread of low dose outside the target volume into the surrounding healthy tissue, was calculated for all plans:
GI ¼
PIV 1=2 PIV
where PIV1/2 is the volume encompassed by half the prescription isodose line. Keeping a low value of GI indicates low dose to the surrounding tissue [19]. PCI and GI were calculated for the entire data set and then stratified by greater or less than 0.5 cm3 in volume. GK beam-on time was calculated assuming a new Co-60 source while LINAC beam-on time was estimated using a nominal maximum dose rate of 600 MU/min. The treatment time reported here was not intended to represent time included for patient setup and frame/mask removal.
Fig. 1. The average DVH ± SEM for tumor (A), cochlea (B), brainstem (C) and body (D) shows no significant difference between VMAT 1, 3 and 5 arc plans compared to GK. DVH, dose volume histogram; VMAT, volumetric modulated arc therapy; GK, gamma knife.
Please cite this article in press as: Kim H et al. Tumor volume threshold for achieving improved conformity in VMAT and Gamma Knife stereotactic radiosurgery for vestibular schwannoma. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.03.031
4
VMAT vs GK SRS for vestibular schwannoma
Statistics Dosimetric outcomes for GK and VMAT plans were compared using a Kruskal–Wallis test and Dunn’s post hoc test with correction for multiple comparisons (GraphPad Prism v5.02). Comparison was done for all patients and stratified by tumor size (tumor volume greater and less than 0.5 cm3). P value less than 0.05 was considered statistically significant.
GI for 1 arc VMAT plans was significantly inferior to GK and VMAT 5 arc plans for the entire cohort (p < 0.0001). When stratified by tumor size, GI in GK was superior to VMAT 1 arc plans for tumors <0.5 cm3 (p = 0.02). GI for VMAT 1 arc plans were inferior to GK, VMAT 3 and 5 arc plans for tumors greater than 0.5 cm3; there was no significant difference between the GK and VMAT 3 and 5 non-coplanar arc plans (Table 4).
Results
Discussion
GK plans and VMAT 1 coplanar, and 3 and 5 non-coplanar arc plans were generated for each of the 19 patients treated for VS. A total of 76 VMAT and GK plans were included in the analysis (3 sets of LINAC and 1 GK plan per patient). In the 19 patient cohort, there were 6 patients with tumor size <0.5 cm3 (range 0.1–0.4 cm3), 6 patients with tumor size >0.5 cm3 but <2 cm3 (0.6–1.8 cm3) and 7 patients with tumor size P2 cm3 (2–4.7 cm3). Median tumor volume was 1.2 cm3. The resultant treatment times and dose metrics for the 19 GK and 57 VMAT plans are summarized in Table 1. GK beam-on time was significantly longer (20.9 ± 7.4 min) than VMAT 600 MU/min beam-on times for 1 co-planar (8.5 ± 1.7 min), 3 non-coplanar (8.6 ± 1.3 min) and 5 non-coplanar (9.6 ± 1.9 min) arc plans (p < 0.0001). The dose received by 2% of the structure volume (D2%) for tumors in GK plans was 22.9 ± 0.3 Gy. The D2% for VMAT 1, 3 and 5 arc plans were significantly lower at 19.8 ± 2.5, 19.6 ± 2.1 and 19.7 ± 2.2 Gy, respectively (p < 0.0001; Table 3). The mean dose for the target was decreased with VMAT 1 coplanar (16.1 ± 1.0 Gy), 3 non-coplanar (15.9 ± 0.8 Gy) and 5 non-coplanar (16.0 ± 0.9 Gy) arcs, compared to GK plans (17.2 ± 0.5 Gy) (p < 0.0001). Fig. 1A illustrates the reduced tumor dose achieved with VMAT by comparing the DVH of VMAT and GK plans. Normal tissue sparing was similar with GK and VMAT 3 and 5 non-coplanar arc plans, as was reflected in no statistically significant difference between the dose to normal brain tissue, brainstem (p = 0.2) and cochlea (p = 0.06) (Table 3, Fig. 1B–D). However, GK plans demonstrated improved mean cochlear dose when compared to 1 coplanar arc VMAT plans (p = 0.002; Table 3). VMAT and GK plans displayed the same degree of conformity as indicated by similar PCI when evaluated for all tumor sizes (p = 0.3). For tumors larger than 0.5 cm3, the 5 non-coplanar VMAT plans had a PCI of 0.77 ± 0.07. This was significantly improved compared to GK and VMAT 1 arc plans with PCI values of 0.69 ± 0.04 and 0.69 ± 0.05, respectively (p = 0.002; Table 4). For tumors smaller than 0.5 cm3, the GK plan PCI was significantly improved compared to the PCI values of VMAT 1 (0.50 ± 0.11), 3 (0.49 ± 0.05) and 5 (0.47 ± 0.04) arc plans (p = 0.004; Table 4). There was no difference in dose to normal structures when the data were stratified by tumor size (Figs. 2A and B).
Low dose spill to critical structures or normal brain tissue and the potential for radiation associated malignancies is of special concern in the treatment of benign brain lesions where patient life expectancy is high [20,21]. Doses as low as 1 Gy have been implicated in radiation associated malignancies and long term data are relatively unavailable, given the time necessary for detection of secondary tumors [11]. Recent data have shown that frame-based and frameless immobilizations have comparable three dimensional magnitudes of errors of 1.19 ± 0.45 mm and 0.76 ± 0.46 mm, respectively [22]. Thus comparing dosimetric outcomes of frameless, LINAC SRS plans and frame-based, GK plans may provide guidance in making treatment decisions. In our study, we present the largest series, to our knowledge, of VS SRS comparison of GK and VMAT plans. Our data indicate that VMAT plans for treatment of VS offer similar low dose to cochlea, brainstem and normal brain tissue without compromise in tumor coverage. An important observation is that VMAT plans with 5 non-coplanar arcs allowed increased conformity in comparison to GK plans for tumor volumes greater than 0.5 cm3. Conversely, GK plans for tumors smaller than 0.5 cm3, had increased conformity compared to VMAT plans. This may be due to the size of the 5 mm MLC leaves in our current study, with the Perfexion offering the slightly smaller 4 mm collimator size. The actual dosimetric significance of this size disparity is unknown. Smaller MLC size may result in a more tightly shaped beam profile and provide similar or improved conformity compared to GK for small tumor sizes [23]. Further studies comparing GK plans to VMAT plans using micro or high-definition MLC leaves are warranted. The existing literature comparing VMAT to GK for SRS treatment of VS is limited. One previous study published by Abacioglu et al. compared VMAT to GK plans for VS. The authors concluded that target coverage, homogeneity, and OAR sparing were similar between the two modalities with improvement in treatment time [15]. Their effort is commended for being the first study to compare the two modalities in VS. However, the data were limited to 6 patients and it is unclear what statistical methods were used to find statistically significant differences between the two treatment modalities in this sample size. The current study corroborates many of their findings based on a larger cohort of 19 patients with 3 distinct LINAC plans (1, 3 and 5 arcs) compared to one GK plan per patient. Further, Abacioglu et al. grouped VMAT treatments with a range of non-coplanar arcs in their analysis. Our data suggest that the dosimetric outcomes may significantly be affected by the number of non-coplanar arcs in a treatment plan. Finally, this study reports treatment time based on treatment without flattening filter free (FFF) beams. While FFF treatment is becoming more commonly available, it is important to demonstrate the generalizability of these data to institutions that may not have FFF treatment as an option. At our institution, GK treatment requires patients to be evaluated by two medical subspecialties (radiation oncology and neurosurgery), undergo invasive frame placement and wait on-site during treatment planning. The entire treatment procedure with no delays and new Co-60 sources can be approximately 4 h, although there are multiple factors that may lengthen the
Table 4 Paddick conformity index and gradient index for GK and VMAT plans differ according to tumor volume.
GK VMAT 1 CA VMAT 3 NCA VMAT 5 NCA
PCI Tumor <0.5 cm3
PCI Tumor >0.5 cm3
GI Tumor <0.5 cm3
GI Tumor >0.5 cm3
0.67 ± 0.03 0.50 ± 0.11* 0.49 ± 0.05* 0.47 ± 0.04*
0.69 ± 0.04# 0.69 ± 0.05# 0.74 ± 0.06 0.77 ± 0.07
3.73 ± 1.12 6.04 ± 1.14* 4.85 ± 0.89 4.53 ± 0.9
3.36 ± 0.48à 4.26 ± 0.42 3.72 ± 0.55à 3.48 ± 0.40à
Abbreviations: GK, gamma knife; VMAT, volumetric modulated arc therapy; CA, coplanar arc; NCA, non-coplanar arcs; PCI, Paddick conformity index; GI, gradient index. Statistically significant difference compared to GK (*), VMAT 5 NCA (#) and VMAT 1 CA (à).
Please cite this article in press as: Kim H et al. Tumor volume threshold for achieving improved conformity in VMAT and Gamma Knife stereotactic radiosurgery for vestibular schwannoma. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.03.031
H. Kim et al. / Radiotherapy and Oncology xxx (2015) xxx–xxx
5
Fig. 2. Representative VMAT 1, 3 and 5 arc plans for tumor size <0.5 cm3 (A) and >0.5 cm3 (B) show similar dose conformity for each respective size category.
treatment time such as the strength of the cobalt sources and the availability of the treating physicians. For LINAC SRS, patient CT simulation, setup and treatment occur on two separate days, but the total time of the patient in clinic is limited to 4 h. Thus, LINAC SRS allows the physician and patient flexibility in the schedule as well as the potential for increased patient throughput. As radiation oncologists are required to be present for SRS delivery, shortened patient stay and treatment times are pertinent for clinic efficiency. With the recent cuts on health care reimbursement and increased pressure to improve clinic efficiency, it is important to pursue more cost and time efficient treatment modalities without
compromising on treatment plans. Other authors have reported the cost disparity between LINAC and GK based radiosurgical treatments, with GK being 6% more expensive per patient than a dedicated SRS LINAC and 18% more expensive per patient compared to a LINAC used for both SRS and standard fractionated treatment of extracranial sites [2,11]. It is difficult to compare the costs incurred from changing out cobalt-60 sources every 5–10 years versus the required maintenance for LINAC and eventual replacement (which occurs approximately every 10 years). If GK treatment actually requires more hospital resources, then there is a financial disincentive to offer this treatment modality for the same reimbursement fee as LINAC SRS [2].
Please cite this article in press as: Kim H et al. Tumor volume threshold for achieving improved conformity in VMAT and Gamma Knife stereotactic radiosurgery for vestibular schwannoma. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.03.031
6
VMAT vs GK SRS for vestibular schwannoma
The authors recognize the limitations of this study as it is retrospective in nature and generated from a single institution where treatment preference or experience may lead to biased results. However, we propose that the data presented were generated in an objective manner consistent with our regular treatment procedures. It is also noted that our study does not address the affect of tumor shape irregularity on conformity index. It seems logical that MLCs would allow improved PCI with more irregularly shaped tumors due to the leaf configuration to match any shape. However, as there is no standard mechanism for quantifying tumor irregularity this assessment would have been highly subjective and subject to criticism. Finally, although the study is limited to VS in 19 patients, it still remains the largest series comparing GK and VMAT SRS for VS to date and raises interesting points of discussion that require further investigation (e.g. tumor volume relationship to CI in SRS). Thus we propose that these data indicate that strong consideration should be given for treatment of VS with VMAT 5 (or more) non-coplanar arcs for tumor volumes greater than 0.5 cm3. Until long term follow up for LINAC-based SRS for VS is available, it remains a cost effective and dosimetrically comparable modality to GK. Conflicts of interest James Evans is a consultant for Stryker and develops surgical instruments with Mizuho. No other conflicts of interest. References [1] Gerbi BJ, Higgins PD, Cho KH, Hall WA. Linac-based stereotactic radiosurgery for treatment of trigeminal neuralgia. J Appl Clin Med Phys Am Coll Med Phys 2004;5:80–92. [2] Griffiths A, Marinovich L, Barton MB, Lord SJ. Cost analysis of Gamma Knife stereotactic radiosurgery. Int J Technol Assess Health Care 2007;23:488–94. http://dx.doi.org/10.1017/S0266462307070584. [3] Deinsberger R, Tidstrand J. Linac radiosurgery as a tool in neurosurgery. Neurosurg Rev 2005;28:79–88. http://dx.doi.org/10.1007/s10143-005-0376-7 (discussion 89–90). [4] Clark GM, Popple RA, Young PE, Fiveash JB. Feasibility of single-isocenter volumetric modulated arc radiosurgery for treatment of multiple brain metastases. Int J Radiat Oncol Biol Phys 2010;76:296–302. http://dx.doi.org/ 10.1016/j.ijrobp.2009.05.029. [5] Andrews DW, Scott CB, Sperduto PW, Flanders AE, Gaspar LE, Schell MC, et al. Whole brain radiation therapy with or without stereotactic radiosurgery boost for patients with one to three brain metastases: phase III results of the RTOG 9508 randomised trial. Lancet 2004;363:1665–72. http://dx.doi.org/10.1016/ S0140-6736(04)16250-8. [6] Hasegawa T, Kida Y, Kato T, Iizuka H, Kuramitsu S, Yamamoto T. Long-term safety and efficacy of stereotactic radiosurgery for vestibular schwannomas: evaluation of 440 patients more than 10 years after treatment with Gamma Knife surgery. J Neurosurg 2013;118:557–65. http://dx.doi.org/10.3171/ 2012.10.JNS12523.
[7] Yang I, Aranda D, Han SJ, Chennupati S, Sughrue ME, Cheung SW, et al. Hearing preservation after stereotactic radiosurgery for vestibular schwannoma: a systematic review. J Clin Neurosci 2009;16:742–7. http://dx.doi.org/10.1016/ j.jocn.2008.09.023. [8] Lasak JM, Klish D, Kryzer TC, Hearn C, Gorecki JP, Rine GP. Gamma knife radiosurgery for vestibular schwannoma: early hearing outcomes and evaluation of the cochlear dose. Otol Neurotol Off Publ Am Otol Soc Am Neurotol Soc Eur Acad Otol Neurotol 2008;29:1179–86. http://dx.doi.org/ 10.1097/MAO.0b013e31818b6639. [9] Brown M, Ruckenstein M, Bigelow D, Judy K, Wilson V, Alonso-Basanta M, et al. Predictors of hearing loss after gamma knife radiosurgery for vestibular schwannomas: age, cochlear dose, and tumor coverage. Neurosurgery 2011;69:605–14. http://dx.doi.org/10.1227/NEU.0b013e31821a42f3. [10] Wolff HA, Wagner DM, Christiansen H, Hess CF, Vorwerk H. Single fraction radiosurgery using Rapid Arc for treatment of intracranial targets. Radiat Oncol 2010;5:77. http://dx.doi.org/10.1186/1748-717X-5-77. [11] Lagerwaard FJ, Meijer OWM, van der Hoorn EAP, Verbakel WFAR, Slotman BJ, Senan S. Volumetric modulated arc radiotherapy for vestibular schwannomas. Int J Radiat Oncol Biol Phys 2009;74:610–5. http://dx.doi.org/10.1016/ j.ijrobp.2008.12.076. [12] Mayo CS, Ding L, Addesa A, Kadish S, Fitzgerald TJ, Moser R. Initial experience with volumetric IMRT (RapidArc) for intracranial stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 2010;78:1457–66. http://dx.doi.org/10.1016/ j.ijrobp.2009.10.005. [13] Fogliata A, Clivio A, Nicolini G, Vanetti E, Cozzi L. Intensity modulation with photons for benign intracranial tumours: a planning comparison of volumetric single arc, helical arc and fixed gantry techniques. Radiother Oncol J Eur Soc Ther Radiol Oncol 2008;89:254–62. http://dx.doi.org/10.1016/ j.radonc.2008.07.021. [14] Lee T-F, Fang F-M, Chao P-J, Su T-J, Wang LK, Leung SW. Dosimetric comparisons of helical tomotherapy and step-and-shoot intensity-modulated radiotherapy in nasopharyngeal carcinoma. Radiother Oncol 2008;89:89–96. http://dx.doi.org/10.1016/j.radonc.2008.05.010. [15] Abacioglu U, Ozen Z, Yilmaz M, Arifoglu A, Gunhan B, Kayalilar N, et al. Critical appraisal of RapidArc radiosurgery with flattening filter free photon beams for benign brain lesions in comparison to GammaKnife: a treatment planning study. Radiat Oncol 2014;9:119. http://dx.doi.org/10.1186/1748-717X-9-119. [16] White Paper: A new TMR dose algorithm in Leksell GammaPlanÒ n.d. [17] Inverse Planning in Leksell GammaPlanÒ 10 - MedicalPhysicsWeb n.d. http:// medicalphysicsweb.org/cws/article/whitepapers/52773 [accessed August 27, 2014]. [18] Paddick I. A simple scoring ratio to index the conformity of radiosurgical treatment plans. J Neurosurg 2000;93:219–22. http://dx.doi.org/10.3171/ jns.2000.93.supplement. [19] Paddick I, Lippitz B. A simple dose gradient measurement tool to complement the conformity index. J Neurosurg 2006;105:194–201. http://dx.doi.org/ 10.3171/sup.2006.105.7.194. [20] Evans DGR, Birch JM, Ramsden RT, Sharif S, Baser ME. Malignant transformation and new primary tumours after therapeutic radiation for benign disease: substantial risks in certain tumour prone syndromes. J Med Genet 2006;43:289–94. http://dx.doi.org/10.1136/jmg.2005.036319. [21] Loeffler JS, Niemierko A, Chapman PH. Second tumors after radiosurgery: tip of the iceberg or a bump in the road? Neurosurgery 2003;52:1436–40 (discussion 1440–2). [22] Gevaert T, Verellen D, Tournel K, Linthout N, Bral S, Engels B, et al. Setup accuracy of the Novalis ExacTrac 6DOF system for frameless radiosurgery. Int J Radiat Oncol 2012;82:1627–35. http://dx.doi.org/10.1016/ j.ijrobp.2011.01.052. [23] Shiu AS, Kooy HM, Ewton JR, Tung SS, Wong J, Antes K, et al. Comparison of miniature multileaf collimation (MMLC) with circular collimation for stereotactic treatment. Int J Radiat Oncol Biol Phys 1997;37:679–88.
Please cite this article in press as: Kim H et al. Tumor volume threshold for achieving improved conformity in VMAT and Gamma Knife stereotactic radiosurgery for vestibular schwannoma. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.03.031
Contents lists available at ScienceDirect
Radiotherapy and Oncology journal homepage: www.thegreenjournal.com
Original article
Tumor volume threshold for achieving improved conformity in VMAT and Gamma Knife stereotactic radiosurgery for vestibular schwannoma Hyun Kim a, Peter Potrebko a, Amanda Rivera a, Haisong Liu b, Harriet B. Eldredge-Hindy a, Vickie Gunn b, Maria Werner-Wasik a, David W. Andrews b, James J. Evans b, Christopher J. Farrell b, Kevin Judy b, Wenyin Shi a,⇑ a
Department of Radiation Oncology; and b Department of Neurological Surgery, Sidney Kimmel Medical College at Thomas Jefferson University, Philadelphia, United States
a r t i c l e
i n f o
Article history: Received 12 January 2015 Received in revised form 12 March 2015 Accepted 27 March 2015 Available online xxxx Keywords: Radiosurgery Vestibular schwannoma Gamma knife VMAT
a b s t r a c t Background and purpose: Recent advances in multileaf collimator field shaping technology and inverse planning software have resulted in highly conformal LINAC based stereotactic radiosurgery (SRS) plans with minimal dose to critical structures. This modeling study compares Gamma Knife (GK) and LINAC SRS for vestibular schwannoma (VS). Materials and methods: 76 treatment plans from nineteen patients with VS were planned using GK forward planning and volumetric arc therapy (VMAT) inverse planning software. VMAT plans were generated with 1 coplanar, 3 and 5 non-coplanar arcs. Dose to normal structures and beam-on time (dose rate 600 MU/min) were compared using Kruskal–Wallis and Dunn’s post hoc test. Results: Median tumor volume was 1.2 cm3 (range 0.1–4.8 cm3). A peripheral tumor dose of 12 Gy was prescribed. Tumor coverage was >99.8%. VMAT plans had lower target D2% and mean dose, as well as decreased beam-on time, compared to GK plans (p < 0.0001). Paddick conformity index in VMAT 5 arc plans was superior to that of GK plans for targets >0.5 cm3 (p = 0.002). Similar dose to cochlea, normal brain tissue and brainstem was observed. Conclusion: VMAT should be considered as a safe, alternative modality to GK for VS SRS treatment, especially for tumors larger than 0.5 cm3. Ó 2015 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology xxx (2015) xxx–xxx
Stereotactic radiosurgery (SRS) is a common treatment option for vestibular schwannoma (VS, acoustic neuroma). SRS provides a non-surgical approach for excellent tumor control in a single treatment session and achieves a steep dose gradient that minimizes normal tissue dose. Gamma Knife (GK) (Elekta AB, Stockholm, Sweden), the traditional technique for VS SRS, is a frame-based treatment modality allowing cranial sites to be treated using cobalt-60 sources embedded within a shielded radiation unit. The newest generation model, the Gamma Knife Perfexion, uses 192 cobalt-60 sources arranged in eight sectors. The delivered dose is shaped to the target by a combination of shots generated by different collimator sizes and plugging of sectors. In GK treatment planning, treatment volumes are usually encompassed by the 50% isodose line (IDL) with a penumbral width of 1 mm, resulting in minimal dose to normal brain and cochlea [1]. Treatment time
⇑ Corresponding author at: Department of Radiation Oncology, Bodine Center, Suite G-301, 111 S. 11th Street, Philadelphia, PA 19107, United States. E-mail address: [email protected] (W. Shi).
increases with source age, with the need to replace the sources every 5 years for approximately 1/5 of the unit purchase price [2]. With the advent of new technology and planning software, gantry-mounted linear accelerator (LINAC) based SRS is an increasingly implemented alternative to GK. Traditionally, LINAC machines were not viewed as suitable for SRS due to mechanical instability and lack of precision. Current stabilization sub-devices and high precision bearings for couch and gantry now allow a radiation beam precision of up to 0.2 ± 0.1 mm [3]. LINAC SRS beam conformity is achieved through multileaf collimation (MLC), intensity-modulation and multiple non-coplanar arcs [4]. Targets are usually encompassed by the 80% IDL with a 2 mm penumbral width [1]. Appealing aspects of LINAC SRS include avoiding invasive frame-based immobilization, inverse-planning dose limits to critical structures, high dose rate settings, a wider availability of LINACs and potentially lower cost [2]. Comparisons of SRS in brain metastases and benign lesions (VS, meningiomas, arteriovenous malformations) have shown similar clinical outcome between GK and LINAC [3,5]. Long term LINAC SRS outcome data for benign lesions are lacking. Much of the long-term outcome data for VS SRS is from GK treated lesions, with
http://dx.doi.org/10.1016/j.radonc.2015.03.031 0167-8140/Ó 2015 Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Kim H et al. Tumor volume threshold for achieving improved conformity in VMAT and Gamma Knife stereotactic radiosurgery for vestibular schwannoma. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.03.031
2
VMAT vs GK SRS for vestibular schwannoma
local control of 92% at >10 years and hearing preservation in 58% (3 years) and 46% (5 years) of patients [6,7]. Increasing evidence indicates that hearing preservation can be improved by limiting dose to the cochlea and increasing the plan conformity index [8,9]. With the development of intensity modulated radiation therapy (IMRT) and volumetric modulated arc therapy (VMAT) planning, plans with improved conformity and decreased treatment times compared to conformal radiation therapy are possible [10]. Previous studies have compared various LINAC SRS techniques for the treatment of benign intracranial lesions. Authors have shown similar tumor coverage and normal tissue sparing with VMAT compared to dynamic conformal arcs [11], static beam IMRT [12], helical tomotherapy [13,14] and CyberKnife [12]. Studies comparing GK to VMAT treatment planning for VS are limited, and the two available retrospective studies are limited to 6 patients or less [15]. This study is, to our knowledge, the largest patient comparison of GK to LINAC VMAT inverse-planning with multiple non-coplanar arcs. By utilizing the tools available with LINAC treatment inverse-planning, it was possible to improve the conformity index compared to GK Perfexion plans. We also present data to suggest tumor size criteria for which VMAT SRS may result in improved dosimetric outcomes. Materials and methods Patients and treatment volumes CT and MRI data of 19 patients with VS who were treated with Novalis Tx or GK from 1994–2001, at the Jefferson Hospital for Neuroscience, were selected for the generation of VMAT and GK plans. CT simulation was obtained using 2.5 mm slice images, with patients in supine position and mask immobilization (BrainLAB). Gadolinium-enhanced 1 mm slice MRI T1 and T2 sequences were used by radiation oncologists and neurosurgeons to contour the VS gross tumor volume (GTV) and critical normal structures (brainstem, cochlea) in the GammaPlan treatment planning system (v 10, Elekta AB, Stockholm, Sweden). No margin was added to the GTV to define the planning target volume (PTV). Structures were exported via DICOM to Eclipse for VMAT planning. VMAT planning The Eclipse treatment planning system (v11.0.42, Varian Medical Systems, Palo Alto, CA) was used to generate VMAT (RapidArc) treatment plans using the progressive-resolution optimizer algorithm for a TrueBeam (Varian Medical Systems, Palo Alto, CA), 6 MV photon beam model with 120 leaf Millennium MLC (5 mm leaf width at isocenter). The final dose calculation was performed using the Anisotropic Analytical Algorithm using a 1 mm grid size and with the heterogeneity correction turned off. VMAT plans with one coplanar full arc, three non-coplanar partial arcs, and five non-coplanar partial arcs were investigated. The gantry and couch angles for the non-coplanar VMAT plans are presented in Table 1. In general, the treatment arc orientations were chosen to be incident on the side of the patient containing the target and to avoid entering through the optic structures. Furthermore, these arc orientations were developed at our institution from extensive treatment planning experience and were found to provide excellent dose conformity. Dose constraints for the tumor, cochlea, brainstem, and surrounding normal tissue (conformity ring structure) were defined for VMAT inverse planning (Table 2). Per our institutional practice, we created a 0.3 cm thick ring planning structure around the tumor with a 0.15 cm gap, and assigned dose objectives to this structure to produce excellent dose conformity and a steep dose gradient. The clinical VMAT dose
Table 1 VMAT arc orientations. 3 non-coplanar arcs
5 non-coplanar arcs
Gantry
Couch
Gantry
Couch
Right sided target
181–359 210–320 210–330
0 330 300
181–359 210–320 210–330 30–150 50–150
0 330 300 90 60
Left sided target
0–179 40–140 50–150
0 30 60
0–179 40–140 50–150 30–150 210–330
0 30 60 90 300
Table 2 VMAT dose constraints. Structure
Objective type
Volume (%)
Dose (cGy)
Priority
Tumor Ring Ring Ring Cochlea Brainstem
Lower Upper Upper Upper Upper Upper
100 30 10 0 0 0
1200 550 780 1200 750 950
100 100 50 50 50 50
constraints were based on the goals of achieving 100% target dose coverage and keeping the cochlea and brainstem maximum doses below 12 Gy and 14 Gy, respectively. GK planning The GammaPlan treatment planning system was used to generate GK Perfexion treatment plans using the Tissue-Maximum-Ratio 10 dose algorithm using a 1 mm dose grid size and skull measurements [16]. For each plan, the inverse planning algorithm was initially used to optimize target coverage, selectivity, and gradient index [17] followed by manual fine-tuning. It should be noted that the current version of the GK Perfexion inverse planning algorithm does not include the ability to define dose constraints for specific structures. Instead, the inverse planning algorithm is based on optimizing the three previously mentioned metrics for the target volume. Plan optimization and dose prescription After initial structure contouring, each plan (VMAT or GK) required approximately 30 min of manipulation, not including plan optimization by software. This time included generating the conformity ring structure, arc orientations, inverse planning constraints, and plan renormalization (VMAT) and manual adjustment of radiation shot isocenters (GK). Dose was prescribed to the 65% isodose line in VMAT plans and 50% isodose line for GK plans. The plan normalization mode in Eclipse was then used to renormalize the VMAT plans to achieve the clinical goal of 100% dose coverage of the PTV by the prescription dose of 12 Gy. Typical plan normalization values were on the order of 90%. Therefore, the normalization metrics for all the GK and VMAT plans were 99.8% to 100% target dose coverage by 12 Gy, limited by dose constraints of 12 and 14 Gy maximum to the cochlea and brainstem, respectively. Dose matrices with a 1 mm grid resolution were DICOM exported from GammaPlan to Eclipse for each structure to maintain a consistent DVH comparison between GK and VMAT plans. The dose matrix for the entire ‘skull’ was exported with a 1 mm resolution so that a DVH for the entire brain dose could be computed. Both GK and LINAC plans were reviewed by an attending radiation oncologist and clinical physicist.
Please cite this article in press as: Kim H et al. Tumor volume threshold for achieving improved conformity in VMAT and Gamma Knife stereotactic radiosurgery for vestibular schwannoma. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.03.031
3
H. Kim et al. / Radiotherapy and Oncology xxx (2015) xxx–xxx Table 3 VMAT treatment planning results in significantly shorter treatment times with similar dosimetric outcomes compared to GK.
GK VMAT 1 CA VMAT 3 NCA VMAT 5 NCA
Brain (cm3)
600 MU/min (min)
Target (Gy) D2%
Mean
D2%
Mean
D2%
Mean
V6
V8
V12
20.9 ± 7.4 8.5 ± 1.7* 8.6 ± 1.3* 9.6 ± 1.9*
22.9 ± 0.3 19.8 ± 2.5* 19.6 ± 2.1* 19.7 ± 2.2*
17.2 ± 0.5 16.1 ± 1.0* 15.9 ± 0.8* 16.0 ± 0.9*
5.1 ± 3.4 6.5 ± 2.9 5.1 ± 2.9 4.6 ± 2.6
1.4 ± 1.0 1.5 ± 0.9 1.2 ± 0.8 1.2 ± 0.7
9.3 ± 2.6 9.8 ± 1.4 8.8 ± 1.7 8.4 ± 1.8
6.3 ± 1.9 7.9 ± 1.1* 6.4 ± 1.2 6.0 ± 1.4
8.4 ± 7.4 10.4 ± 8.2 8.3 ± 6.3 7.6 ± 5.7
5.4 ± 4.8 5.9 ± 4.7 5.2 ± 4.1 4.8 ± 3.7
2.5 ± 2.3 2.5 ± 2.1 2.3 ± 2.0 2.2 ± 1.8
Brainstem (Gy)
Cochlea (Gy)
Abbreviations: GK, gamma knife; VMAT, volumetric modulated arc therapy; CA, coplanar arc; NCA, non-coplanar arcs; PCI, Paddick conformity index. * Statistically significant difference compared to GK.
Dose–volume histogram and analysis Composite dose–volume histograms (DVH) were generated for tumor, body, brainstem and cochlea. The DVH was compiled by selecting representative dose and volume points in each structure and graphing the average value with the standard error of the mean. The Paddick Conformity Index (PCI), or the product of a treatment plan’s ‘‘undertreatment’’ and ‘‘overtreatment’’ ratios, was calculated as:
PCI ¼
ðTV PIV Þ2 ðTV PIVÞ
where TVPIV is the volume of the target covered by the prescription isodose line, TV is the total target volume, and PIV is the volume encompassed by the prescription isodose line. Perfect conformity
results in a PCI of unity or 1, with potential values ranging from 0 to 1 [18]. The gradient index (GI), which provides a measure of the spread of low dose outside the target volume into the surrounding healthy tissue, was calculated for all plans:
GI ¼
PIV 1=2 PIV
where PIV1/2 is the volume encompassed by half the prescription isodose line. Keeping a low value of GI indicates low dose to the surrounding tissue [19]. PCI and GI were calculated for the entire data set and then stratified by greater or less than 0.5 cm3 in volume. GK beam-on time was calculated assuming a new Co-60 source while LINAC beam-on time was estimated using a nominal maximum dose rate of 600 MU/min. The treatment time reported here was not intended to represent time included for patient setup and frame/mask removal.
Fig. 1. The average DVH ± SEM for tumor (A), cochlea (B), brainstem (C) and body (D) shows no significant difference between VMAT 1, 3 and 5 arc plans compared to GK. DVH, dose volume histogram; VMAT, volumetric modulated arc therapy; GK, gamma knife.
Please cite this article in press as: Kim H et al. Tumor volume threshold for achieving improved conformity in VMAT and Gamma Knife stereotactic radiosurgery for vestibular schwannoma. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.03.031
4
VMAT vs GK SRS for vestibular schwannoma
Statistics Dosimetric outcomes for GK and VMAT plans were compared using a Kruskal–Wallis test and Dunn’s post hoc test with correction for multiple comparisons (GraphPad Prism v5.02). Comparison was done for all patients and stratified by tumor size (tumor volume greater and less than 0.5 cm3). P value less than 0.05 was considered statistically significant.
GI for 1 arc VMAT plans was significantly inferior to GK and VMAT 5 arc plans for the entire cohort (p < 0.0001). When stratified by tumor size, GI in GK was superior to VMAT 1 arc plans for tumors <0.5 cm3 (p = 0.02). GI for VMAT 1 arc plans were inferior to GK, VMAT 3 and 5 arc plans for tumors greater than 0.5 cm3; there was no significant difference between the GK and VMAT 3 and 5 non-coplanar arc plans (Table 4).
Results
Discussion
GK plans and VMAT 1 coplanar, and 3 and 5 non-coplanar arc plans were generated for each of the 19 patients treated for VS. A total of 76 VMAT and GK plans were included in the analysis (3 sets of LINAC and 1 GK plan per patient). In the 19 patient cohort, there were 6 patients with tumor size <0.5 cm3 (range 0.1–0.4 cm3), 6 patients with tumor size >0.5 cm3 but <2 cm3 (0.6–1.8 cm3) and 7 patients with tumor size P2 cm3 (2–4.7 cm3). Median tumor volume was 1.2 cm3. The resultant treatment times and dose metrics for the 19 GK and 57 VMAT plans are summarized in Table 1. GK beam-on time was significantly longer (20.9 ± 7.4 min) than VMAT 600 MU/min beam-on times for 1 co-planar (8.5 ± 1.7 min), 3 non-coplanar (8.6 ± 1.3 min) and 5 non-coplanar (9.6 ± 1.9 min) arc plans (p < 0.0001). The dose received by 2% of the structure volume (D2%) for tumors in GK plans was 22.9 ± 0.3 Gy. The D2% for VMAT 1, 3 and 5 arc plans were significantly lower at 19.8 ± 2.5, 19.6 ± 2.1 and 19.7 ± 2.2 Gy, respectively (p < 0.0001; Table 3). The mean dose for the target was decreased with VMAT 1 coplanar (16.1 ± 1.0 Gy), 3 non-coplanar (15.9 ± 0.8 Gy) and 5 non-coplanar (16.0 ± 0.9 Gy) arcs, compared to GK plans (17.2 ± 0.5 Gy) (p < 0.0001). Fig. 1A illustrates the reduced tumor dose achieved with VMAT by comparing the DVH of VMAT and GK plans. Normal tissue sparing was similar with GK and VMAT 3 and 5 non-coplanar arc plans, as was reflected in no statistically significant difference between the dose to normal brain tissue, brainstem (p = 0.2) and cochlea (p = 0.06) (Table 3, Fig. 1B–D). However, GK plans demonstrated improved mean cochlear dose when compared to 1 coplanar arc VMAT plans (p = 0.002; Table 3). VMAT and GK plans displayed the same degree of conformity as indicated by similar PCI when evaluated for all tumor sizes (p = 0.3). For tumors larger than 0.5 cm3, the 5 non-coplanar VMAT plans had a PCI of 0.77 ± 0.07. This was significantly improved compared to GK and VMAT 1 arc plans with PCI values of 0.69 ± 0.04 and 0.69 ± 0.05, respectively (p = 0.002; Table 4). For tumors smaller than 0.5 cm3, the GK plan PCI was significantly improved compared to the PCI values of VMAT 1 (0.50 ± 0.11), 3 (0.49 ± 0.05) and 5 (0.47 ± 0.04) arc plans (p = 0.004; Table 4). There was no difference in dose to normal structures when the data were stratified by tumor size (Figs. 2A and B).
Low dose spill to critical structures or normal brain tissue and the potential for radiation associated malignancies is of special concern in the treatment of benign brain lesions where patient life expectancy is high [20,21]. Doses as low as 1 Gy have been implicated in radiation associated malignancies and long term data are relatively unavailable, given the time necessary for detection of secondary tumors [11]. Recent data have shown that frame-based and frameless immobilizations have comparable three dimensional magnitudes of errors of 1.19 ± 0.45 mm and 0.76 ± 0.46 mm, respectively [22]. Thus comparing dosimetric outcomes of frameless, LINAC SRS plans and frame-based, GK plans may provide guidance in making treatment decisions. In our study, we present the largest series, to our knowledge, of VS SRS comparison of GK and VMAT plans. Our data indicate that VMAT plans for treatment of VS offer similar low dose to cochlea, brainstem and normal brain tissue without compromise in tumor coverage. An important observation is that VMAT plans with 5 non-coplanar arcs allowed increased conformity in comparison to GK plans for tumor volumes greater than 0.5 cm3. Conversely, GK plans for tumors smaller than 0.5 cm3, had increased conformity compared to VMAT plans. This may be due to the size of the 5 mm MLC leaves in our current study, with the Perfexion offering the slightly smaller 4 mm collimator size. The actual dosimetric significance of this size disparity is unknown. Smaller MLC size may result in a more tightly shaped beam profile and provide similar or improved conformity compared to GK for small tumor sizes [23]. Further studies comparing GK plans to VMAT plans using micro or high-definition MLC leaves are warranted. The existing literature comparing VMAT to GK for SRS treatment of VS is limited. One previous study published by Abacioglu et al. compared VMAT to GK plans for VS. The authors concluded that target coverage, homogeneity, and OAR sparing were similar between the two modalities with improvement in treatment time [15]. Their effort is commended for being the first study to compare the two modalities in VS. However, the data were limited to 6 patients and it is unclear what statistical methods were used to find statistically significant differences between the two treatment modalities in this sample size. The current study corroborates many of their findings based on a larger cohort of 19 patients with 3 distinct LINAC plans (1, 3 and 5 arcs) compared to one GK plan per patient. Further, Abacioglu et al. grouped VMAT treatments with a range of non-coplanar arcs in their analysis. Our data suggest that the dosimetric outcomes may significantly be affected by the number of non-coplanar arcs in a treatment plan. Finally, this study reports treatment time based on treatment without flattening filter free (FFF) beams. While FFF treatment is becoming more commonly available, it is important to demonstrate the generalizability of these data to institutions that may not have FFF treatment as an option. At our institution, GK treatment requires patients to be evaluated by two medical subspecialties (radiation oncology and neurosurgery), undergo invasive frame placement and wait on-site during treatment planning. The entire treatment procedure with no delays and new Co-60 sources can be approximately 4 h, although there are multiple factors that may lengthen the
Table 4 Paddick conformity index and gradient index for GK and VMAT plans differ according to tumor volume.
GK VMAT 1 CA VMAT 3 NCA VMAT 5 NCA
PCI Tumor <0.5 cm3
PCI Tumor >0.5 cm3
GI Tumor <0.5 cm3
GI Tumor >0.5 cm3
0.67 ± 0.03 0.50 ± 0.11* 0.49 ± 0.05* 0.47 ± 0.04*
0.69 ± 0.04# 0.69 ± 0.05# 0.74 ± 0.06 0.77 ± 0.07
3.73 ± 1.12 6.04 ± 1.14* 4.85 ± 0.89 4.53 ± 0.9
3.36 ± 0.48à 4.26 ± 0.42 3.72 ± 0.55à 3.48 ± 0.40à
Abbreviations: GK, gamma knife; VMAT, volumetric modulated arc therapy; CA, coplanar arc; NCA, non-coplanar arcs; PCI, Paddick conformity index; GI, gradient index. Statistically significant difference compared to GK (*), VMAT 5 NCA (#) and VMAT 1 CA (à).
Please cite this article in press as: Kim H et al. Tumor volume threshold for achieving improved conformity in VMAT and Gamma Knife stereotactic radiosurgery for vestibular schwannoma. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.03.031
H. Kim et al. / Radiotherapy and Oncology xxx (2015) xxx–xxx
5
Fig. 2. Representative VMAT 1, 3 and 5 arc plans for tumor size <0.5 cm3 (A) and >0.5 cm3 (B) show similar dose conformity for each respective size category.
treatment time such as the strength of the cobalt sources and the availability of the treating physicians. For LINAC SRS, patient CT simulation, setup and treatment occur on two separate days, but the total time of the patient in clinic is limited to 4 h. Thus, LINAC SRS allows the physician and patient flexibility in the schedule as well as the potential for increased patient throughput. As radiation oncologists are required to be present for SRS delivery, shortened patient stay and treatment times are pertinent for clinic efficiency. With the recent cuts on health care reimbursement and increased pressure to improve clinic efficiency, it is important to pursue more cost and time efficient treatment modalities without
compromising on treatment plans. Other authors have reported the cost disparity between LINAC and GK based radiosurgical treatments, with GK being 6% more expensive per patient than a dedicated SRS LINAC and 18% more expensive per patient compared to a LINAC used for both SRS and standard fractionated treatment of extracranial sites [2,11]. It is difficult to compare the costs incurred from changing out cobalt-60 sources every 5–10 years versus the required maintenance for LINAC and eventual replacement (which occurs approximately every 10 years). If GK treatment actually requires more hospital resources, then there is a financial disincentive to offer this treatment modality for the same reimbursement fee as LINAC SRS [2].
Please cite this article in press as: Kim H et al. Tumor volume threshold for achieving improved conformity in VMAT and Gamma Knife stereotactic radiosurgery for vestibular schwannoma. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.03.031
6
VMAT vs GK SRS for vestibular schwannoma
The authors recognize the limitations of this study as it is retrospective in nature and generated from a single institution where treatment preference or experience may lead to biased results. However, we propose that the data presented were generated in an objective manner consistent with our regular treatment procedures. It is also noted that our study does not address the affect of tumor shape irregularity on conformity index. It seems logical that MLCs would allow improved PCI with more irregularly shaped tumors due to the leaf configuration to match any shape. However, as there is no standard mechanism for quantifying tumor irregularity this assessment would have been highly subjective and subject to criticism. Finally, although the study is limited to VS in 19 patients, it still remains the largest series comparing GK and VMAT SRS for VS to date and raises interesting points of discussion that require further investigation (e.g. tumor volume relationship to CI in SRS). Thus we propose that these data indicate that strong consideration should be given for treatment of VS with VMAT 5 (or more) non-coplanar arcs for tumor volumes greater than 0.5 cm3. Until long term follow up for LINAC-based SRS for VS is available, it remains a cost effective and dosimetrically comparable modality to GK. Conflicts of interest James Evans is a consultant for Stryker and develops surgical instruments with Mizuho. No other conflicts of interest. References [1] Gerbi BJ, Higgins PD, Cho KH, Hall WA. Linac-based stereotactic radiosurgery for treatment of trigeminal neuralgia. J Appl Clin Med Phys Am Coll Med Phys 2004;5:80–92. [2] Griffiths A, Marinovich L, Barton MB, Lord SJ. Cost analysis of Gamma Knife stereotactic radiosurgery. Int J Technol Assess Health Care 2007;23:488–94. http://dx.doi.org/10.1017/S0266462307070584. [3] Deinsberger R, Tidstrand J. Linac radiosurgery as a tool in neurosurgery. Neurosurg Rev 2005;28:79–88. http://dx.doi.org/10.1007/s10143-005-0376-7 (discussion 89–90). [4] Clark GM, Popple RA, Young PE, Fiveash JB. Feasibility of single-isocenter volumetric modulated arc radiosurgery for treatment of multiple brain metastases. Int J Radiat Oncol Biol Phys 2010;76:296–302. http://dx.doi.org/ 10.1016/j.ijrobp.2009.05.029. [5] Andrews DW, Scott CB, Sperduto PW, Flanders AE, Gaspar LE, Schell MC, et al. Whole brain radiation therapy with or without stereotactic radiosurgery boost for patients with one to three brain metastases: phase III results of the RTOG 9508 randomised trial. Lancet 2004;363:1665–72. http://dx.doi.org/10.1016/ S0140-6736(04)16250-8. [6] Hasegawa T, Kida Y, Kato T, Iizuka H, Kuramitsu S, Yamamoto T. Long-term safety and efficacy of stereotactic radiosurgery for vestibular schwannomas: evaluation of 440 patients more than 10 years after treatment with Gamma Knife surgery. J Neurosurg 2013;118:557–65. http://dx.doi.org/10.3171/ 2012.10.JNS12523.
[7] Yang I, Aranda D, Han SJ, Chennupati S, Sughrue ME, Cheung SW, et al. Hearing preservation after stereotactic radiosurgery for vestibular schwannoma: a systematic review. J Clin Neurosci 2009;16:742–7. http://dx.doi.org/10.1016/ j.jocn.2008.09.023. [8] Lasak JM, Klish D, Kryzer TC, Hearn C, Gorecki JP, Rine GP. Gamma knife radiosurgery for vestibular schwannoma: early hearing outcomes and evaluation of the cochlear dose. Otol Neurotol Off Publ Am Otol Soc Am Neurotol Soc Eur Acad Otol Neurotol 2008;29:1179–86. http://dx.doi.org/ 10.1097/MAO.0b013e31818b6639. [9] Brown M, Ruckenstein M, Bigelow D, Judy K, Wilson V, Alonso-Basanta M, et al. Predictors of hearing loss after gamma knife radiosurgery for vestibular schwannomas: age, cochlear dose, and tumor coverage. Neurosurgery 2011;69:605–14. http://dx.doi.org/10.1227/NEU.0b013e31821a42f3. [10] Wolff HA, Wagner DM, Christiansen H, Hess CF, Vorwerk H. Single fraction radiosurgery using Rapid Arc for treatment of intracranial targets. Radiat Oncol 2010;5:77. http://dx.doi.org/10.1186/1748-717X-5-77. [11] Lagerwaard FJ, Meijer OWM, van der Hoorn EAP, Verbakel WFAR, Slotman BJ, Senan S. Volumetric modulated arc radiotherapy for vestibular schwannomas. Int J Radiat Oncol Biol Phys 2009;74:610–5. http://dx.doi.org/10.1016/ j.ijrobp.2008.12.076. [12] Mayo CS, Ding L, Addesa A, Kadish S, Fitzgerald TJ, Moser R. Initial experience with volumetric IMRT (RapidArc) for intracranial stereotactic radiosurgery. Int J Radiat Oncol Biol Phys 2010;78:1457–66. http://dx.doi.org/10.1016/ j.ijrobp.2009.10.005. [13] Fogliata A, Clivio A, Nicolini G, Vanetti E, Cozzi L. Intensity modulation with photons for benign intracranial tumours: a planning comparison of volumetric single arc, helical arc and fixed gantry techniques. Radiother Oncol J Eur Soc Ther Radiol Oncol 2008;89:254–62. http://dx.doi.org/10.1016/ j.radonc.2008.07.021. [14] Lee T-F, Fang F-M, Chao P-J, Su T-J, Wang LK, Leung SW. Dosimetric comparisons of helical tomotherapy and step-and-shoot intensity-modulated radiotherapy in nasopharyngeal carcinoma. Radiother Oncol 2008;89:89–96. http://dx.doi.org/10.1016/j.radonc.2008.05.010. [15] Abacioglu U, Ozen Z, Yilmaz M, Arifoglu A, Gunhan B, Kayalilar N, et al. Critical appraisal of RapidArc radiosurgery with flattening filter free photon beams for benign brain lesions in comparison to GammaKnife: a treatment planning study. Radiat Oncol 2014;9:119. http://dx.doi.org/10.1186/1748-717X-9-119. [16] White Paper: A new TMR dose algorithm in Leksell GammaPlanÒ n.d. [17] Inverse Planning in Leksell GammaPlanÒ 10 - MedicalPhysicsWeb n.d. http:// medicalphysicsweb.org/cws/article/whitepapers/52773 [accessed August 27, 2014]. [18] Paddick I. A simple scoring ratio to index the conformity of radiosurgical treatment plans. J Neurosurg 2000;93:219–22. http://dx.doi.org/10.3171/ jns.2000.93.supplement. [19] Paddick I, Lippitz B. A simple dose gradient measurement tool to complement the conformity index. J Neurosurg 2006;105:194–201. http://dx.doi.org/ 10.3171/sup.2006.105.7.194. [20] Evans DGR, Birch JM, Ramsden RT, Sharif S, Baser ME. Malignant transformation and new primary tumours after therapeutic radiation for benign disease: substantial risks in certain tumour prone syndromes. J Med Genet 2006;43:289–94. http://dx.doi.org/10.1136/jmg.2005.036319. [21] Loeffler JS, Niemierko A, Chapman PH. Second tumors after radiosurgery: tip of the iceberg or a bump in the road? Neurosurgery 2003;52:1436–40 (discussion 1440–2). [22] Gevaert T, Verellen D, Tournel K, Linthout N, Bral S, Engels B, et al. Setup accuracy of the Novalis ExacTrac 6DOF system for frameless radiosurgery. Int J Radiat Oncol 2012;82:1627–35. http://dx.doi.org/10.1016/ j.ijrobp.2011.01.052. [23] Shiu AS, Kooy HM, Ewton JR, Tung SS, Wong J, Antes K, et al. Comparison of miniature multileaf collimation (MMLC) with circular collimation for stereotactic treatment. Int J Radiat Oncol Biol Phys 1997;37:679–88.
Please cite this article in press as: Kim H et al. Tumor volume threshold for achieving improved conformity in VMAT and Gamma Knife stereotactic radiosurgery for vestibular schwannoma. Radiother Oncol (2015), http://dx.doi.org/10.1016/j.radonc.2015.03.031