Practical Radiation Oncology (2011) 1, 126–134
www.practicalradonc.org
Original Report
A novel modified dynamic conformal arc technique for treatment of peripheral lung tumors using stereotactic body radiation therapy Christopher C. Ross MD a , John J. Kim CMD a , Zhe J. Chen PhD a , David J. Grew MPH b , Bryan W. Chang MD a , Roy H. Decker MD, PhD a,⁎ a
Department of Therapeutic Radiology, Yale University School of Medicine, New Haven, Connecticut Tulane University School of Medicine, New Orleans, Louisiana
b
Received 28 September 2010; revised 12 November 2010; accepted 13 November 2010
Abstract Purpose: To describe and compare a novel, modified dynamic conformal arc (MDCA) technique for lung stereotactic body radiation therapy with the standard noncoplanar beam (NCB) technique based on stereotactic body radiation therapy (SBRT) coverage, dose conformality, normal tissue constraints, and treatment time. Materials and Methods: Twenty consecutive medically inoperable patients with early stage, peripheral, non–small cell lung cancer treated with SBRT using an NCB technique were replanned with a novel MDCA technique. Treatment plans were compared based on Radiation Therapy Oncology Group (RTOG) 0236 criteria for planning treatment volume (PTV) coverage and normal tissue dose constraints, as well as high- and moderate-dose conformality. Treatment times necessary to deliver the NCB plans were compared with the times of a separate group of 12 consecutive patients treated with the MDCA technique at our institution. Results: The MDCA technique resulted in improved coverage of the cranial and caudal regions of the PTV and generated plans that were significantly more conformal by all high-dose criteria proposed by the RTOG protocol. In terms of moderate-dose criteria, MDCA plans had a significantly lower maximum dose (2 cm from the PTV), whereas the ratio of the 50% dose volume to the volume of the PTV was equivalent between the 2 techniques. All normal tissue dose constraints proposed in the RTOG 0236 protocol were met by each plan, although the median lung V20 and mean lung dose were slightly higher in the MDCA plans, whereas the chest wall dose was slightly lower. A 42% reduction in treatment time was observed when patients treated with the NCB technique were compared with a separate cohort of 12 patients treated with the MDCA technique (P b .0001). Conclusions: The new MDCA technique described in this study resulted in enhanced PTV coverage, improved high- and moderate-dose conformality, simplified treatment planning, and reduced treatment time compared with results using the standard NCB technique. © 2011 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved.
Conflicts of interest: None. ⁎ Corresponding author. Department of Therapeutic Radiology, Yale University School of Medicine, PO Box 208040, New Haven, CT 06520-8040. E-mail address: [email protected] (R.H. Decker). 1879-8500/$ – see front matter © 2011 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.prro.2010.11.002
Practical Radiation Oncology: April-June 2011
Introduction Surgical resection remains the standard of care for the management of early stage non–small cell lung cancer (NSCLC), with local control rates of approximately 94%1 and 5-year overall survival rates of approximately 70%.1 Unfortunately, many patients are deemed medically inoperable because of comorbid conditions that preclude safe resection and thus require a nonsurgical approach. Stereotactic body radiation therapy (SBRT) has demonstrated promise as an alternative to conventional radiation therapy because it offers improved disease control. In SBRT, the treatment volume is limited by utilizing precise targeting and conformal treatment delivery to increase the dose of radiation that is delivered to the tumor and to minimize radiation to the adjacent normal tissues. This type of therapy allows a much higher biologically equivalent dose to be delivered to the tumor, usually during the course of 1 to 5 fractions. Several retrospective studies, as well as prospective trials, have demonstrated excellent 2- and 3year local control rates of 80%-95%,2–11 and phase III studies are currently under way to compare primary resection with SBRT in medically operable patients. Although many institutions have published their results using SBRT for lung tumors, relatively little data exist on optimal beam arrangements. The protocol, used in the recently published 3-year results9 from Radiation Therapy Oncology Group (RTOG) 0236, recommended using an arc rotational technique or a coplanar or noncoplanar static beam (NCB) arrangement. Seven to 10 static, nonopposing, noncoplanar beams with roughly equal weighting were preferred, as this typically results in more symmetric dose fall-off in all directions, including the cranial and caudal directions, and better restriction of low- and intermediatedose volumes. Although SBRT treatments using static noncoplanar beams typically provide excellent dose distributions, the beam arrangement must be optimized for each patient, and this procedure is associated with lengthy delivery times because of couch rotations between beams. We have developed a novel technique using 6 partially expanded, dynamic, conformal, coplanar arc segments, which significantly simplifies SBRT planning and reduces the time required for treatment delivery. The present study was undertaken to compare and evaluate treatment plans generated using the standard NCB technique with plans generated using this novel arc technique, according to the normal tissue dose constraints and conformality criteria provided by the RTOG 0236 protocol.
Materials and methods Twenty consecutive patients with peripheral lung tumors treated at our institution with SBRT using an NCB technique consisting of 54 Gy in 3 fractions were
Modified dynamic conformal arc technique
127
chosen for comparison. A second plan was generated for each patient using 6 partially expanded, dynamic, conformal coplanar arc segments. Treatment plans were calculated using the analytic anisotropic algorithm available in the Eclipse treatment planning system (Varian Medical Systems, Palo Alto, CA) for tissue heterogeneity correction and were compared using the RTOG 0236 criteria.
Simulation All patients were immobilized using a full-length vacuum cushion, and 4-dimensional computed tomographic simulation with 2.5-mm slice thickness was performed during free breathing to capture tumor motion throughout all phases of the respiratory cycle. The internal target volume was contoured on the average intensity projection using the Advantage Workstation (GE Healthcare, Waukesha, WI) and was modified to account for tumor motion throughout all respiratory phases. An isovolumetric expansion of 7 mm was added to the internal target volume edge to create the planning treatment volume (PTV) consistent with our institutional experience based on our own internal data on patient intrafraction motion. The heart, lungs, esophagus, trachea and ipsilateral bronchus, spinal cord, and brachial plexus were contoured according to the guidelines provided by the RTOG 0236 protocol.
Noncoplanar static beam technique At the time of the computed tomographic simulation, the isocenter was placed within the tumor volume. A treatment plan was then generated using 7-13 static, nonopposing, noncoplanar 6 mV photon beams that were conformed to the PTV using a multileaf collimator. A total dose of 54 Gy in 3 fractions was prescribed to the edge of the PTV. The weighting and arrangement of the beams was optimized based on tumor location and surrounding normal structures. Plans were normalized such that 95% of the PTV was covered by 100% of the prescription dose, and 99% of the PTV was covered by at least 90% of the prescription dose, with an expected maximum heterogeneity of 111%-143% within the tumor (corresponding to 70%-90% of the maximum dose at the edge of the PTV). Plans were also optimized to meet all normal tissue dose constraints proposed by the RTOG 0236 protocol.
Modified dynamic conformal arc technique The modified dynamic conformal arc (MDCA) technique requires identification of an isocenter, which allows dynamic arc rotation 360° around the patient. Because of the frequent peripheral location of lung tumors, placing the
128
C.C. Ross et al
isocenter within the tumor itself would result in a gantry collision with the patient or treatment table, or both. Therefore, the isocenter was placed at the lateral midpoint of the table, and the vertical midpoint of the patient and immobilization device, which was approximately 14.5 cm above the tabletop. Expanded PTVs were then created for treatment planning purposes. A 1-slice expanded PTV (PTV-1slc) was created in which the second-to-last superior-most and inferior-most PTV slices were copied and pasted over the last slice, and 1 slice further in the superior or inferior direction (each slice corresponds to 2.5 mm). A 3-slice expanded PTV (PTV-3slc) was also created by pasting the superior-most and inferior-most slices, 3-slices beyond the PTV, and pasting the second-tolast superior- and inferior-most slices over the last slice and intervening 2 slices beyond the PTV (Fig 1). A treatment plan was then generated using 6 coplanar arc segments of 60° each. Alternating arc segments were then conformed to the expanded PTVs using a multileaf collimator (MLC), such that 3 arc segments conformed to the PTV-1slc and 3 segments conformed to the PTV3slc (Fig 2). Determining which arc segments should be conformed to which expanded PTV is dependent on which lung is involved. For left-sided lung lesions, it is beneficial to have more segments conformed to the larger PTV-3slc entering the patient from the left side, as these beams will pass through less normal tissue before reaching the target. Therefore, for left-sided lesions, the left anterior oblique, left posterior oblique, and right lateral arc segments are conformed to PTV-3slc, and the right anterior oblique, right posterior oblique, and left lateral segments are conformed to the PTV-1slc. Conversely, for right-sided lesions, the right anterior oblique, right posterior oblique, and left lateral arc segments are conformed to the PTV3slc; and the left anterior oblique, left posterior oblique, and right lateral segments are conformed to the PTV-1slc (see Fig 2). A total dose of 54 Gy in 3 fractions was prescribed to the edge of the PTV. Plans were normalized
Practical Radiation Oncology: April-June 2011
Figure 2 Example showing the 6-arc segment technique. Alternating arc segments are conformed to the 1-slice expanded planning target volume (PTV), right lateral (RL), left anterior oblique (LAO), and left posterior oblique (LPO) and the 3-slice expanded PTV, right posterior oblique (RPO), right anterior oblique (RAO), and left lateral (LL). The isocenter is placed at the lateral midpoint of the table, and vertical midpoint of the patient and immobilization device, approximately 14.5 cm above the table top to allow the gantry full 360° rotation.
such that 95% of the PTV was covered by 100% of the prescription dose, and 99% of the PTV was covered by at least 90% of the prescription dose. If necessary, beam weighting was modified to meet all normal tissue dose constraints.
Plan comparison Plans were compared based on PTV coverage, normal tissue dose constraints, high- and moderate-dose conformality parameters, and treatment delivery times. Deviations were defined for the various parameters used in the RTOG
Figure 1 Representative coronal (left) and sagital (right) planes through the planning target volume (PTV), demonstrating PTV expansions. The contoured structures are the PTV (innermost), 1-slice expanded PTV (middle), and 3-slice expanded (outermost).
Practical Radiation Oncology: April-June 2011
0236 protocol according to the guidelines provided by the RTOG, including both minor and major deviations.
PTV coverage Plans were compared using the RTOG guidelines for PTV coverage, in which 95% of the PTV was covered by 100% of the prescription dose, and 99% of the PTV was covered by 90% of the prescription dose.
Normal tissue dose comparisons Plans were compared according to their ability to respect the normal tissue dose constraints recommended by the RTOG protocol to the lungs, heart, esophagus, trachea and ipsilateral bronchus, spinal cord, and ipsilateral brachial plexus. The volume of chest wall receiving ≥ 30 Gy (V30) and ≥ 60 Gy (V60) were also compared for each plan.
High-dose conformality parameters V105/ptv Volume of tissue outside the PTV receiving a dose greater than 105% (56.7 Gy) must be less than 15% of the PTV. Target conformality index The target conformality index (CI) is the ratio of the volume receiving the prescription dose to the volume of the PTV and must be ≤ 1.2. Paddick conformality index TVPIV2/(TV × PIV) (where TV = PTV volume; PIV = prescription isodose volume (54 Gy); and TVPIV = volume of PTV covered by the prescription isodose volume. This CI was proposed by Paddick12 as a more informative highdose CI. Table 1
Modified dynamic conformal arc technique
129
Moderate-dose conformality parameters D2cm The maximum dose of 2 cm from the PTV in any direction was defined for each plan according to RTOG criteria (Table 1). R50% The ratio of the 50% (27 Gy) prescription isodose volume to the volume of the PTV was defined for each plan according to the RTOG criteria (see Table 1).
Treatment delivery time For patients treated using the NCB technique, the treatment delivery time from the start of the first beam to the start of the last beam was recorded for each fraction. These treatment times were compared with those from a different group of 12 patients treated using the MDCA technique at our institution. Treatment time was measured from the start of the first arc segment to the start of the last arc segment for each fraction.
Statistical analysis Comparisons for normal tissue dose constraints and high- and moderate-dose conformality criteria between NCB and MDCA plans were performed using the Wilcoxon signed rank test. Comparisons for treatment delivery times between the 2 groups of patients were performed using the Mann-Whitney U test.
Results Patient characteristics Twenty consecutive patients with peripheral lung tumors treated at our institution with SBRT using a NCB
Conformality parameters based on PTV size according to the RTOG 0236 protocol
Maximum PTV size
Maximum dose 2 cm from the PTV edge in any direction, D2cm (Gy) a
Ratio of the 50% isodose volume to the volume of the PTV, R50%
Dimension (cm)
Volume (cc)
No Deviation
Minor Deviation b
No Deviation
Minor Deviation b
≤ 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0
≤13.2 21.9 33.8 49.6 69.9 95.1 125.8 162.6
b25.29 b27.36 b29.43 b31.59 b33.66 b35.73 b37.80 b39.87
25.29–27.09 27.36–29.16 29.43–31.23 31.59–33.39 33.66–37.53 35.73–37.53 37.80–39.60 39.87–41.67
b3.9 b3.8 b3.7 b3.6 b3.5 b3.3 b3.1 b2.9
3.9–4.1 3.8–4.0 3.7–3.9 3.6–3.8 3.5–3.7 3.3–3.5 3.1–3.3 2.9–3.1
PTV, planning treatment volume; RTOG, Radiation Therapy Oncology Group. a Adjusted from the RTOG 0236 protocol based on a prescription dose of 54 Gy, as opposed to 60 Gy. The doses listed remain consistent with the RTOG protocol as a percentage of the prescribed dose. b Deviations greater than those listed were considered major deviations.
130 Table 2
C.C. Ross et al
Practical Radiation Oncology: April-June 2011
Comparison of techniques
Comparison parameter
Treatment technique
P value
NCB a MDCA b Chest wall dose Volume receiving N 30 Gy (cc) Volume receiving N 60 Gy (cc) Lung dose Volume receiving N 20 Gy (%) Mean lung dose (Gy) V105/PTV a Target conformality index Paddick conformality index R50% b D2cm (Gy) c
25.3 1.0
25.0 0.35
.17 .054
2.5% 2.3 6.9% 1.14 0.74 4.12 33.7
2.7% 2.6 1.4% 1.06 0.78 4.12 28.1
.01 .002 .002 .004 .0004 .80 .006
MDCA, modified dynamic conformal arc; NCB, noncoplanar beam. a Defined by the 105% dose volume outside of the planning treatment volume/PTV (expressed as a percentage). b Ratio of the 50% dose volume to the PTV. c Maximum dose 2 cm from the PTV in any direction.
technique, consisting of 54 Gy in 3 fractions, were chosen for comparison. The median age of the patients was 73 years (range, 60-84 years). There were 11 right lung tumors, and 9 left lung tumors treated. Nineteen were T1 lesions, and 1 was a T2 lesion. Volume of PTVs ranged from 4.79 cc-256.86 cc with a median of 30.16 cc.
Dosimetric plan comparison All plans were normalized to meet RTOG 0236 protocol criteria for PTV coverage (95% of PTV is covered by 100% of prescription dose, and 99% of PTV is covered by 90% of prescription dose). Dose-volume histograms (DVH) for each plan were compared based on PTV coverage, and a representative DVH is shown in Figure 3. Overall, PTV coverage demonstrated by the DVHs was very similar between the 2 techniques. All plans met RTOG criteria, and the median maximum inhomogeneity within the PTV for the MDCA and NCB plans were similar (131.2% and 132.1%, respectively). However, when NCB and MDCA plans were compared side by side, the superior-most and inferior-most portions of the PTV were noted to be more fully covered by the 95% isodose line using the MDCA technique. A representative dose distribution is shown in Figure 4. Delivering beams off-axis to a large degree can be technically difficult because of concern about gantry collision with the patient. Therefore, even with a noncoplanar beam arrangement, dose fall-off in the superior and inferior directions is rapid, and thus the 50% isodose coverage for these regions of the PTV is very close to the PTV edge when using the traditional NCB technique. The MDCA technique, however,
Figure 3 Representative dose-volume histogram showing the planning target volume (gray) and lung (black) dose for the noncoplanar beam (triangle) and modified dynamic conformal arc (square) plans.
provides more generous 50% coverage superiorly and inferiorly (Fig 5). All normal tissue dose constraints proposed by the RTOG protocol (for heart, lungs, esophagus, trachea and ipsilateral bronchus, spinal cord, and ipsilateral brachial plexus) were respected in each of the NCB and MDCA plans. The median lung V20 was higher for the MDCA plans than for the NCB plans (2.7% and 2.5%, respectively; P = .01); however, no plans generated by using either technique exceeded the 10% maximum allowed by the RTOG protocol (range, 1.4%-7.6% and 1.3%-7.8% for MDCA and NCB plans, respectively). The mean lung dose was also slightly higher for MDCA plans than that for NCB plans (median, 2.6 Gy and 2.3 Gy, respectively; P = .002). The chest wall V30 and V60 were not included in the RTOG protocol, but they were used in this study to evaluate dose gradient and were not significantly different for the 2 techniques. The median chest wall V30 for MDCA and NCB plans were 25.0 cc and 25.3 cc, respectively (P = .17), with an average V60 of 0.35 cc and 1.0 cc, respectively (P = .054). High-dose conformality was evaluated using the V105/PTV, and the target CI was proposed by the RTOG, as well as the Paddick CI. Moderate-dose conformality was evaluated using the D2cm and R50% described by the RTOG protocol. It appeared that the MDCA plans generated were generally more conformal by all measured parameters, aside from the R50% (Fig 6). All 20 of the MDCA plans met the V105/PTV cutoff (of b15% proposed by the RTOG) compared with 17 of 20 NCB plans (Fig 7). The MDCA plans had a median V105/PTV of 1.4% compared with an average of 6.9% with NCB plans (P = .002). All 20 MDCA plans also met the RTOG criteria for target CI (of ≤ 1.2) compared with only 13 of 20 NCB plans (see Fig 7). The MDCA plans had a median target CI of 1.06 compared with a median of 1.14 for NCB
Practical Radiation Oncology: April-June 2011
Modified dynamic conformal arc technique
131
Figure 4 Representative dose distribution demonstrating 95% dose coverage of the planning target volume for the modified dynamic conformal arc (top) and noncoplanar beam plans (bottom) in both the coronal (left) and sagittal (right) views.
plans (P = .0004). By the Paddick CI, 18 of 20 (90%) patients had a more conformal plan using the MDCA technique (see Fig 6), with a median Paddick CI of 0.78 compared with 0.74 for NCB plans (P = .0004). Comparing plans with the moderate-dose conformality criteria revealed that plans generated using the MDCA technique had a lower median dose 2 cm from the PTV (D2cm) than plans generated using the NCB technique (28.1 Gy and 33.7 Gy, respectively; P = .006) and had fewer deviations from the RTOG guidelines (see Fig 7). The ratio of the 50% dose volume to the volume of the PTV (R50%), however, was better for 12 of 20 (60%) patients using the NCB technique, although there was no difference in median R50% between the 2 techniques (4.12 for each; P = .80). We then compared treatment delivery times of these 20 patients with a separate cohort of 12 patients treated with the MDCA technique. Treatment delivery times for each fraction measured from initiation of the first beam or arc segment to initiation of the last beam or arc segment were significantly reduced with the MDCA technique. Treatment times per fraction using the NCB technique ranged from 9.4 to 34.3 minutes with a median time of 12.2 minutes. After the MDCA technique was adopted for routine use in treating peripheral tumors, the treatment time necessary for a separate group of 12 consecutive
patients treated with the technique to the same total dose of 54 Gy in 3 fractions was recorded for each fraction. Treatment times per fraction ranged from 3.8 to 26.7 minutes, with a median of 7.0 minutes. This corresponds to a 42% reduction in treatment time (P b .0001).
Discussion The SBRT has been adopted as a standard treatment option for patients with medically inoperable, peripheral, early stage NSCLC. Many series have been published demonstrating excellent 2- and 3-year local control rates with SBRT using several different treatment techniques, including multiple coplanar beams, multiple noncoplanar beams, and various arc techniques.2-11 To date, however, few studies have evaluated which treatment technique is preferable. A study at the University of Maryland evaluated several static coplanar, static noncoplanar, and rotational arc therapy plans for 37 different medically inoperable lung tumors on the basis of the RTOG 0236 criteria.13 They found that nearly all treatment techniques satisfied the high-dose conformality parameters and normal tissue dose constraints; however, 81% of cases had improved moderate-dose conformality when using an
132
C.C. Ross et al
Practical Radiation Oncology: April-June 2011
Figure 5 Representative dose distribution demonstrating 50% dose coverage of the planning target volume for the modified dynamic conformal arc (top) and noncoplanar beam plans (bottom) in both the coronal (left) and sagittal (right) views.
NCB technique. Conversely, a recent study from William Beaumont Hospital generated volumetric modulated arc therapy (VMAT) plans for 21 patients with stage IA NSCLC previously treated stereotactically with 7-10
Figure 6 Comparison showing the percentage of patients who had a more conformal plan generated with either the noncoplanar beam (NCB) or modified dynamic conformal arc (MDCA) technique by high- and moderate-dose conformality criteria. NT, normal tissue; PTV, planning target volume.
nonoppposing, noncoplanar static beams.14 They found that the 2 techniques generated plans that had equivalent mean maximum doses to the PTV and similar 95% dose conformality. The VMAT plans, however, were more conformal at the 80% and 50% isodose lines. The dose to normal tissues was also equivalent between the 2 techniques, with the exception of normal lung dose, which saw a reduction in all parameters (V20/12.5/10/5) with the VMAT technique. In addition, the VMAT technique reduced treatment times by 37%-63%. In this study, we describe a novel MDCA technique, which not only simplifies treatment planning and shortens overall treatment delivery time, but also generates more conformal treatment plans and provides better PTV coverage for the superior-most and inferior-most portions of the PTV. During normal respiration, lung tumors often experience significant motion in the superior and inferior directions. This makes coverage of the PTV in these areas critically important. In this regard, a noncoplanar technique generally provides a better dosimetric plan than a coplanar technique, as the noncoplanar beams are not tangential to the superior and inferior edges of the PTV. In particular, using a coplanar arc technique often generates a disc-shaped dose distribution with a very flat region of rapid-dose fall-off immediately superior and inferior to the
Practical Radiation Oncology: April-June 2011
Figure 7 Comparison showing percentage of plans generated with the noncoplanar beam (NCB) and modified dynamic conformal arc (MDCA) techniques that meet Radiation Therapy Oncology Group (RTOG) 0236 protocol parameters regarding high- and moderate-dose conformality factors. NT, normal tissue; PTV, planning target volume.
PTV. In this case, the dose distribution will be increased in the radial directions proportionally more than in the superior and inferior directions when normalizing the plan to meet PTV coverage parameters in the superior and inferior portions of the PTV. The technique described in this study makes use of arc segments expanded in the superior and inferior directions and allows for improved coverage of these regions without the need to expand the dose distribution radially to the same degree as is necessary with a traditional arc technique. Ultimately, this expansion results in a more spherical dose distribution. Even when compared with plans generated using a traditional, nonexpanded NCB technique, it provides improved coverage of the superior- and inferior-most regions of the PTV. It should be noted, however, that this improved dose coverage in the superior and inferior regions of the PTV can likely be attributed solely to the PTV expansion used in the MDCA treatment planning process. It can be assumed that a similar optimization PTV created for the NCB plans would result in similar coverage. In this study, the NCB plans were developed according to the guidelines set out in the RTOG 0236 protocol, and thus the PTV was not expanded for planning purposes. Previous studies evaluating optimal block margins around the PTV have suggested that plans with greater conformality and less normal lung dose are generated when using no additional margin.15 However, the NCB plans used in this study were re-planned with a 5-mm (2 computed tomographic sliced) expansion superiorly and inferiorly to approximate the average expansion used in the MDCA technique planning process. As a result, the coverage in these regions improved and was similar to the MDCA plans (data not shown).
Modified dynamic conformal arc technique
133
This more spherical dose distribution produced by the MDCA plans also resulted in improved target conformality. The MDCA technique produced plans that were significantly more conformal by all high-dose conformality criteria that were examined, including V105/PTV, target CI, and the Paddick CI, than plans generated using a NCB technique. Even when comparing moderate-dose conformality criteria, the MDCA technique produced plans with significantly lower maximum doses 2 cm from the PTV (D2cm), and the ratio of the 50% isodose volume to the volume of the PTV was equivalent between the 2 techniques. Again, it should be noted that at least part of the improved conformality with the MDCA technique is a result of the PTV expansion. When the PTVs for the NCB plans were expanded, the conformality of the resultant plans was much closer to the MDCA plans by the Paddick CI (median, 0.76 and 0.78 for the expanded NCB and MDCA plans, respectively; P = .004). It is expected that an arc technique would result in a greater volume of normal tissue being irradiated, as is evidenced by the higher median lung V20 and mean lung dose in this study. However, when examining the absolute magnitude of these values (V20 of 2.7% and 2.5% for MDCA and NCB plans, respectively; and the mean lung dose of 2.6 Gy and 2.3 Gy for MDCA and NCB plans, respectively), there is little difference, and all plans generated with both techniques were well within the lung dose constraints suggested in the RTOG 0236 protocol. Furthermore, the median volume of chest wall receiving N 30 Gy (V30) and N 60 Gy (V60) was slightly higher in the NCB plans than in the MDCA plans. Neither was statistically significant, although the increase in median V60 did approach significance (P = 0.054). The blocking along the chest wall could be optimized to reduce the chest wall dose for the NCB plans; however, it is our institutional practice not to limit the chest wall dose at the cost of PTV coverage. In addition, the absolute magnitude of the chest wall dose for each plan was not excessive in the cases evaluated. With regard to the remaining normal tissue constraints set forth in the RTOG protocol, all parameters were respected using both techniques, compared with a 73% normal tissue dose constraint compliance rate in the published results of RTOG 0236.9 Additional benefits of the MDCA technique include a simplified SBRT treatment planning process and shorter treatment delivery times. The simplification of the planning process stems from the ability to use a template as a starting point, in contrast to the need to identify 7-10 or more optimal beam angles for each patient when using the NCB method. By using a template, which automatically loads 6 arc segments of 60° each, the user needs only to define an isocenter and conform each equally weighted arc segment to the appropriate expanded PTV. In rare cases, the weighting of the arc segments must be modified to respect normal tissue dose constraints, but for the majority of peripheral lesions this is not necessary.
134
C.C. Ross et al
In addition to a more efficient treatment planning process, we also observed a significant reduction in treatment delivery time. The arc segments generated using the MDCA technique rotate around a single isocenter. The isocenter is placed in such a location as to allow full gantry rotation around the patient without obstruction, thus obviating the need to enter the treatment room between each beam to make adjustments in couch position. When we compared the treatment times for patients at our institution who were treated using the NCB technique with a group of 12 patients treated with the MDCA technique on the same equipment, we observed a 42% reduction in treatment time using the MDCA technique. In patients with medically inoperable lung tumors, a reduction in treatment time is especially important, because many patients in this population have significant difficulty remaining motionless during a lengthy treatment. This could potentially reduce error associated with patient movement, as well as improve clinical flow. As a result of the enhanced PTV coverage, improved high- and moderate-dose conformality, simplified treatment planning process, and reduced treatment time afforded by the MDCA technique, we have adopted it as the primary SBRT method for treating peripheral lung tumors at our institution. In addition, we are evaluating its utility for the treatment of both more centrally located early stage NSCLCs and liver SBRT.
References 1. Ginsberg RJ, Rubinstein LV. Randomized trial of lobectomy versus limited resection for T1N0 non–small cell lung cancer. Ann Thorac Surg. 1995;60:615-623. 2. Fakiris AJ, McGarry RC, Yiannoutsos CT, et al. Stereotactic body radiation therapy for early-stage non–small-cell lung carcinoma: four-year results of a prospective phase II study. Int J Radiat Oncol Biol Phys. 2009;75:677-682.
Practical Radiation Oncology: April-June 2011 3. Baumann P, Nyman J, Lax I, et al. Factors important for efficacy of stereotactic body radiotherapy of medically inoperable stage I lung cancer. A retrospective analysis of patients treated in the Nordic countries. Acta Oncol. 2006;45:787-795. 4. Baumann P, Nyman J, Hoyer M, et al. Outcome in a prospective phase II trial of medically inoperable stage I non–small-cell lung cancer patients treated with stereotactic body radiotherapy. J Clin Oncol. 2009;27:3290-3296. 5. Nagata Y, Negoro Y, Aoki T, et al. Clinical outcomes of 3D conformal hypofractionated single high-dose radiotherapy for one or two lung tumors using a stereotactic body frame. Int J Radiat Oncol Biol Phys. 2002;52:1041-1046. 6. Nagata Y, Takayama K, Matsuo Y, et al. Clinical outcomes of a phase I/II study of 48 Gy of stereotactic body radiotherapy in 4 fractions for primary lung cancer using a stereotactic body frame. Int J Radiat Oncol Biol Phys. 2005;63:1427-1431. 7. Hara R, Itami J, Kondo T, et al. Clinical outcomes of single-fraction stereotactic radiation therapy of lung tumors. Cancer. 2006;106: 1347-1352. 8. Onishi H, Shirato H, Nagata Y, et al. Hypofractionated stereotactic radiotherapy (HypoFXSRT) for stage I non–small cell lung cancer: updated results of 257 patients in a Japanese multi-institutional study. J Thorac Oncol. 2007;2:S94-S100. 9. Timmerman R, Paulus R, Galvin J, et al. Stereotactic body radiation therapy for inoperable early stage lung cancer. JAMA. 2010;303(11): 1070-1076. 10. Xia T, Li H, Sung Q, et al. Promising clinical outcome of stereotactic body radiation therapy for patients with inoperable stage I/II non–small cell lung cancer. Int J Radiat Oncol Biol Phys. 2006;66:117-125. 11. Timmerman R, McGarry R, Yiannoutsos C, et al. Excessive toxicity when treating central tumors in a phase II study of stereotactic body radiation therapy for medically inoperable early-stage lung cancer. J Clin Oncol. 2006;24:4833-4839. 12. Paddick I. A simple scoring ratio to index the conformity of radiosurgical treatment plans. J Neurosurg. 2000;93(Suppl 3): 219-222. 13. Lim DH, Yi BY, Mirmiran A, et al. Optimal beam arrangement for sterotactic body radiation therapy delilvery in lung tumors. Acta Oncol. 2010;49:219-224. 14. McGrath SD, Matuszak MM, Yan D, et al. Volumetric modulated arc therapy for delivery of hypofractionated stereotactic lung radiotherapy: a dosimetric and treatment efficiency analysis. Radiother Oncol. 2010;95:153-157. 15. Cardinale RM, Wu Q, Benedict SH, et al. Determining the optimal block margin on the planning target volume for extracranial stereotactic radiotherapy. Int J Radiat Oncol Biol Phys. 1999;45:515-520.
www.practicalradonc.org
Original Report
A novel modified dynamic conformal arc technique for treatment of peripheral lung tumors using stereotactic body radiation therapy Christopher C. Ross MD a , John J. Kim CMD a , Zhe J. Chen PhD a , David J. Grew MPH b , Bryan W. Chang MD a , Roy H. Decker MD, PhD a,⁎ a
Department of Therapeutic Radiology, Yale University School of Medicine, New Haven, Connecticut Tulane University School of Medicine, New Orleans, Louisiana
b
Received 28 September 2010; revised 12 November 2010; accepted 13 November 2010
Abstract Purpose: To describe and compare a novel, modified dynamic conformal arc (MDCA) technique for lung stereotactic body radiation therapy with the standard noncoplanar beam (NCB) technique based on stereotactic body radiation therapy (SBRT) coverage, dose conformality, normal tissue constraints, and treatment time. Materials and Methods: Twenty consecutive medically inoperable patients with early stage, peripheral, non–small cell lung cancer treated with SBRT using an NCB technique were replanned with a novel MDCA technique. Treatment plans were compared based on Radiation Therapy Oncology Group (RTOG) 0236 criteria for planning treatment volume (PTV) coverage and normal tissue dose constraints, as well as high- and moderate-dose conformality. Treatment times necessary to deliver the NCB plans were compared with the times of a separate group of 12 consecutive patients treated with the MDCA technique at our institution. Results: The MDCA technique resulted in improved coverage of the cranial and caudal regions of the PTV and generated plans that were significantly more conformal by all high-dose criteria proposed by the RTOG protocol. In terms of moderate-dose criteria, MDCA plans had a significantly lower maximum dose (2 cm from the PTV), whereas the ratio of the 50% dose volume to the volume of the PTV was equivalent between the 2 techniques. All normal tissue dose constraints proposed in the RTOG 0236 protocol were met by each plan, although the median lung V20 and mean lung dose were slightly higher in the MDCA plans, whereas the chest wall dose was slightly lower. A 42% reduction in treatment time was observed when patients treated with the NCB technique were compared with a separate cohort of 12 patients treated with the MDCA technique (P b .0001). Conclusions: The new MDCA technique described in this study resulted in enhanced PTV coverage, improved high- and moderate-dose conformality, simplified treatment planning, and reduced treatment time compared with results using the standard NCB technique. © 2011 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved.
Conflicts of interest: None. ⁎ Corresponding author. Department of Therapeutic Radiology, Yale University School of Medicine, PO Box 208040, New Haven, CT 06520-8040. E-mail address: [email protected] (R.H. Decker). 1879-8500/$ – see front matter © 2011 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.prro.2010.11.002
Practical Radiation Oncology: April-June 2011
Introduction Surgical resection remains the standard of care for the management of early stage non–small cell lung cancer (NSCLC), with local control rates of approximately 94%1 and 5-year overall survival rates of approximately 70%.1 Unfortunately, many patients are deemed medically inoperable because of comorbid conditions that preclude safe resection and thus require a nonsurgical approach. Stereotactic body radiation therapy (SBRT) has demonstrated promise as an alternative to conventional radiation therapy because it offers improved disease control. In SBRT, the treatment volume is limited by utilizing precise targeting and conformal treatment delivery to increase the dose of radiation that is delivered to the tumor and to minimize radiation to the adjacent normal tissues. This type of therapy allows a much higher biologically equivalent dose to be delivered to the tumor, usually during the course of 1 to 5 fractions. Several retrospective studies, as well as prospective trials, have demonstrated excellent 2- and 3year local control rates of 80%-95%,2–11 and phase III studies are currently under way to compare primary resection with SBRT in medically operable patients. Although many institutions have published their results using SBRT for lung tumors, relatively little data exist on optimal beam arrangements. The protocol, used in the recently published 3-year results9 from Radiation Therapy Oncology Group (RTOG) 0236, recommended using an arc rotational technique or a coplanar or noncoplanar static beam (NCB) arrangement. Seven to 10 static, nonopposing, noncoplanar beams with roughly equal weighting were preferred, as this typically results in more symmetric dose fall-off in all directions, including the cranial and caudal directions, and better restriction of low- and intermediatedose volumes. Although SBRT treatments using static noncoplanar beams typically provide excellent dose distributions, the beam arrangement must be optimized for each patient, and this procedure is associated with lengthy delivery times because of couch rotations between beams. We have developed a novel technique using 6 partially expanded, dynamic, conformal, coplanar arc segments, which significantly simplifies SBRT planning and reduces the time required for treatment delivery. The present study was undertaken to compare and evaluate treatment plans generated using the standard NCB technique with plans generated using this novel arc technique, according to the normal tissue dose constraints and conformality criteria provided by the RTOG 0236 protocol.
Materials and methods Twenty consecutive patients with peripheral lung tumors treated at our institution with SBRT using an NCB technique consisting of 54 Gy in 3 fractions were
Modified dynamic conformal arc technique
127
chosen for comparison. A second plan was generated for each patient using 6 partially expanded, dynamic, conformal coplanar arc segments. Treatment plans were calculated using the analytic anisotropic algorithm available in the Eclipse treatment planning system (Varian Medical Systems, Palo Alto, CA) for tissue heterogeneity correction and were compared using the RTOG 0236 criteria.
Simulation All patients were immobilized using a full-length vacuum cushion, and 4-dimensional computed tomographic simulation with 2.5-mm slice thickness was performed during free breathing to capture tumor motion throughout all phases of the respiratory cycle. The internal target volume was contoured on the average intensity projection using the Advantage Workstation (GE Healthcare, Waukesha, WI) and was modified to account for tumor motion throughout all respiratory phases. An isovolumetric expansion of 7 mm was added to the internal target volume edge to create the planning treatment volume (PTV) consistent with our institutional experience based on our own internal data on patient intrafraction motion. The heart, lungs, esophagus, trachea and ipsilateral bronchus, spinal cord, and brachial plexus were contoured according to the guidelines provided by the RTOG 0236 protocol.
Noncoplanar static beam technique At the time of the computed tomographic simulation, the isocenter was placed within the tumor volume. A treatment plan was then generated using 7-13 static, nonopposing, noncoplanar 6 mV photon beams that were conformed to the PTV using a multileaf collimator. A total dose of 54 Gy in 3 fractions was prescribed to the edge of the PTV. The weighting and arrangement of the beams was optimized based on tumor location and surrounding normal structures. Plans were normalized such that 95% of the PTV was covered by 100% of the prescription dose, and 99% of the PTV was covered by at least 90% of the prescription dose, with an expected maximum heterogeneity of 111%-143% within the tumor (corresponding to 70%-90% of the maximum dose at the edge of the PTV). Plans were also optimized to meet all normal tissue dose constraints proposed by the RTOG 0236 protocol.
Modified dynamic conformal arc technique The modified dynamic conformal arc (MDCA) technique requires identification of an isocenter, which allows dynamic arc rotation 360° around the patient. Because of the frequent peripheral location of lung tumors, placing the
128
C.C. Ross et al
isocenter within the tumor itself would result in a gantry collision with the patient or treatment table, or both. Therefore, the isocenter was placed at the lateral midpoint of the table, and the vertical midpoint of the patient and immobilization device, which was approximately 14.5 cm above the tabletop. Expanded PTVs were then created for treatment planning purposes. A 1-slice expanded PTV (PTV-1slc) was created in which the second-to-last superior-most and inferior-most PTV slices were copied and pasted over the last slice, and 1 slice further in the superior or inferior direction (each slice corresponds to 2.5 mm). A 3-slice expanded PTV (PTV-3slc) was also created by pasting the superior-most and inferior-most slices, 3-slices beyond the PTV, and pasting the second-tolast superior- and inferior-most slices over the last slice and intervening 2 slices beyond the PTV (Fig 1). A treatment plan was then generated using 6 coplanar arc segments of 60° each. Alternating arc segments were then conformed to the expanded PTVs using a multileaf collimator (MLC), such that 3 arc segments conformed to the PTV-1slc and 3 segments conformed to the PTV3slc (Fig 2). Determining which arc segments should be conformed to which expanded PTV is dependent on which lung is involved. For left-sided lung lesions, it is beneficial to have more segments conformed to the larger PTV-3slc entering the patient from the left side, as these beams will pass through less normal tissue before reaching the target. Therefore, for left-sided lesions, the left anterior oblique, left posterior oblique, and right lateral arc segments are conformed to PTV-3slc, and the right anterior oblique, right posterior oblique, and left lateral segments are conformed to the PTV-1slc. Conversely, for right-sided lesions, the right anterior oblique, right posterior oblique, and left lateral arc segments are conformed to the PTV3slc; and the left anterior oblique, left posterior oblique, and right lateral segments are conformed to the PTV-1slc (see Fig 2). A total dose of 54 Gy in 3 fractions was prescribed to the edge of the PTV. Plans were normalized
Practical Radiation Oncology: April-June 2011
Figure 2 Example showing the 6-arc segment technique. Alternating arc segments are conformed to the 1-slice expanded planning target volume (PTV), right lateral (RL), left anterior oblique (LAO), and left posterior oblique (LPO) and the 3-slice expanded PTV, right posterior oblique (RPO), right anterior oblique (RAO), and left lateral (LL). The isocenter is placed at the lateral midpoint of the table, and vertical midpoint of the patient and immobilization device, approximately 14.5 cm above the table top to allow the gantry full 360° rotation.
such that 95% of the PTV was covered by 100% of the prescription dose, and 99% of the PTV was covered by at least 90% of the prescription dose. If necessary, beam weighting was modified to meet all normal tissue dose constraints.
Plan comparison Plans were compared based on PTV coverage, normal tissue dose constraints, high- and moderate-dose conformality parameters, and treatment delivery times. Deviations were defined for the various parameters used in the RTOG
Figure 1 Representative coronal (left) and sagital (right) planes through the planning target volume (PTV), demonstrating PTV expansions. The contoured structures are the PTV (innermost), 1-slice expanded PTV (middle), and 3-slice expanded (outermost).
Practical Radiation Oncology: April-June 2011
0236 protocol according to the guidelines provided by the RTOG, including both minor and major deviations.
PTV coverage Plans were compared using the RTOG guidelines for PTV coverage, in which 95% of the PTV was covered by 100% of the prescription dose, and 99% of the PTV was covered by 90% of the prescription dose.
Normal tissue dose comparisons Plans were compared according to their ability to respect the normal tissue dose constraints recommended by the RTOG protocol to the lungs, heart, esophagus, trachea and ipsilateral bronchus, spinal cord, and ipsilateral brachial plexus. The volume of chest wall receiving ≥ 30 Gy (V30) and ≥ 60 Gy (V60) were also compared for each plan.
High-dose conformality parameters V105/ptv Volume of tissue outside the PTV receiving a dose greater than 105% (56.7 Gy) must be less than 15% of the PTV. Target conformality index The target conformality index (CI) is the ratio of the volume receiving the prescription dose to the volume of the PTV and must be ≤ 1.2. Paddick conformality index TVPIV2/(TV × PIV) (where TV = PTV volume; PIV = prescription isodose volume (54 Gy); and TVPIV = volume of PTV covered by the prescription isodose volume. This CI was proposed by Paddick12 as a more informative highdose CI. Table 1
Modified dynamic conformal arc technique
129
Moderate-dose conformality parameters D2cm The maximum dose of 2 cm from the PTV in any direction was defined for each plan according to RTOG criteria (Table 1). R50% The ratio of the 50% (27 Gy) prescription isodose volume to the volume of the PTV was defined for each plan according to the RTOG criteria (see Table 1).
Treatment delivery time For patients treated using the NCB technique, the treatment delivery time from the start of the first beam to the start of the last beam was recorded for each fraction. These treatment times were compared with those from a different group of 12 patients treated using the MDCA technique at our institution. Treatment time was measured from the start of the first arc segment to the start of the last arc segment for each fraction.
Statistical analysis Comparisons for normal tissue dose constraints and high- and moderate-dose conformality criteria between NCB and MDCA plans were performed using the Wilcoxon signed rank test. Comparisons for treatment delivery times between the 2 groups of patients were performed using the Mann-Whitney U test.
Results Patient characteristics Twenty consecutive patients with peripheral lung tumors treated at our institution with SBRT using a NCB
Conformality parameters based on PTV size according to the RTOG 0236 protocol
Maximum PTV size
Maximum dose 2 cm from the PTV edge in any direction, D2cm (Gy) a
Ratio of the 50% isodose volume to the volume of the PTV, R50%
Dimension (cm)
Volume (cc)
No Deviation
Minor Deviation b
No Deviation
Minor Deviation b
≤ 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0
≤13.2 21.9 33.8 49.6 69.9 95.1 125.8 162.6
b25.29 b27.36 b29.43 b31.59 b33.66 b35.73 b37.80 b39.87
25.29–27.09 27.36–29.16 29.43–31.23 31.59–33.39 33.66–37.53 35.73–37.53 37.80–39.60 39.87–41.67
b3.9 b3.8 b3.7 b3.6 b3.5 b3.3 b3.1 b2.9
3.9–4.1 3.8–4.0 3.7–3.9 3.6–3.8 3.5–3.7 3.3–3.5 3.1–3.3 2.9–3.1
PTV, planning treatment volume; RTOG, Radiation Therapy Oncology Group. a Adjusted from the RTOG 0236 protocol based on a prescription dose of 54 Gy, as opposed to 60 Gy. The doses listed remain consistent with the RTOG protocol as a percentage of the prescribed dose. b Deviations greater than those listed were considered major deviations.
130 Table 2
C.C. Ross et al
Practical Radiation Oncology: April-June 2011
Comparison of techniques
Comparison parameter
Treatment technique
P value
NCB a MDCA b Chest wall dose Volume receiving N 30 Gy (cc) Volume receiving N 60 Gy (cc) Lung dose Volume receiving N 20 Gy (%) Mean lung dose (Gy) V105/PTV a Target conformality index Paddick conformality index R50% b D2cm (Gy) c
25.3 1.0
25.0 0.35
.17 .054
2.5% 2.3 6.9% 1.14 0.74 4.12 33.7
2.7% 2.6 1.4% 1.06 0.78 4.12 28.1
.01 .002 .002 .004 .0004 .80 .006
MDCA, modified dynamic conformal arc; NCB, noncoplanar beam. a Defined by the 105% dose volume outside of the planning treatment volume/PTV (expressed as a percentage). b Ratio of the 50% dose volume to the PTV. c Maximum dose 2 cm from the PTV in any direction.
technique, consisting of 54 Gy in 3 fractions, were chosen for comparison. The median age of the patients was 73 years (range, 60-84 years). There were 11 right lung tumors, and 9 left lung tumors treated. Nineteen were T1 lesions, and 1 was a T2 lesion. Volume of PTVs ranged from 4.79 cc-256.86 cc with a median of 30.16 cc.
Dosimetric plan comparison All plans were normalized to meet RTOG 0236 protocol criteria for PTV coverage (95% of PTV is covered by 100% of prescription dose, and 99% of PTV is covered by 90% of prescription dose). Dose-volume histograms (DVH) for each plan were compared based on PTV coverage, and a representative DVH is shown in Figure 3. Overall, PTV coverage demonstrated by the DVHs was very similar between the 2 techniques. All plans met RTOG criteria, and the median maximum inhomogeneity within the PTV for the MDCA and NCB plans were similar (131.2% and 132.1%, respectively). However, when NCB and MDCA plans were compared side by side, the superior-most and inferior-most portions of the PTV were noted to be more fully covered by the 95% isodose line using the MDCA technique. A representative dose distribution is shown in Figure 4. Delivering beams off-axis to a large degree can be technically difficult because of concern about gantry collision with the patient. Therefore, even with a noncoplanar beam arrangement, dose fall-off in the superior and inferior directions is rapid, and thus the 50% isodose coverage for these regions of the PTV is very close to the PTV edge when using the traditional NCB technique. The MDCA technique, however,
Figure 3 Representative dose-volume histogram showing the planning target volume (gray) and lung (black) dose for the noncoplanar beam (triangle) and modified dynamic conformal arc (square) plans.
provides more generous 50% coverage superiorly and inferiorly (Fig 5). All normal tissue dose constraints proposed by the RTOG protocol (for heart, lungs, esophagus, trachea and ipsilateral bronchus, spinal cord, and ipsilateral brachial plexus) were respected in each of the NCB and MDCA plans. The median lung V20 was higher for the MDCA plans than for the NCB plans (2.7% and 2.5%, respectively; P = .01); however, no plans generated by using either technique exceeded the 10% maximum allowed by the RTOG protocol (range, 1.4%-7.6% and 1.3%-7.8% for MDCA and NCB plans, respectively). The mean lung dose was also slightly higher for MDCA plans than that for NCB plans (median, 2.6 Gy and 2.3 Gy, respectively; P = .002). The chest wall V30 and V60 were not included in the RTOG protocol, but they were used in this study to evaluate dose gradient and were not significantly different for the 2 techniques. The median chest wall V30 for MDCA and NCB plans were 25.0 cc and 25.3 cc, respectively (P = .17), with an average V60 of 0.35 cc and 1.0 cc, respectively (P = .054). High-dose conformality was evaluated using the V105/PTV, and the target CI was proposed by the RTOG, as well as the Paddick CI. Moderate-dose conformality was evaluated using the D2cm and R50% described by the RTOG protocol. It appeared that the MDCA plans generated were generally more conformal by all measured parameters, aside from the R50% (Fig 6). All 20 of the MDCA plans met the V105/PTV cutoff (of b15% proposed by the RTOG) compared with 17 of 20 NCB plans (Fig 7). The MDCA plans had a median V105/PTV of 1.4% compared with an average of 6.9% with NCB plans (P = .002). All 20 MDCA plans also met the RTOG criteria for target CI (of ≤ 1.2) compared with only 13 of 20 NCB plans (see Fig 7). The MDCA plans had a median target CI of 1.06 compared with a median of 1.14 for NCB
Practical Radiation Oncology: April-June 2011
Modified dynamic conformal arc technique
131
Figure 4 Representative dose distribution demonstrating 95% dose coverage of the planning target volume for the modified dynamic conformal arc (top) and noncoplanar beam plans (bottom) in both the coronal (left) and sagittal (right) views.
plans (P = .0004). By the Paddick CI, 18 of 20 (90%) patients had a more conformal plan using the MDCA technique (see Fig 6), with a median Paddick CI of 0.78 compared with 0.74 for NCB plans (P = .0004). Comparing plans with the moderate-dose conformality criteria revealed that plans generated using the MDCA technique had a lower median dose 2 cm from the PTV (D2cm) than plans generated using the NCB technique (28.1 Gy and 33.7 Gy, respectively; P = .006) and had fewer deviations from the RTOG guidelines (see Fig 7). The ratio of the 50% dose volume to the volume of the PTV (R50%), however, was better for 12 of 20 (60%) patients using the NCB technique, although there was no difference in median R50% between the 2 techniques (4.12 for each; P = .80). We then compared treatment delivery times of these 20 patients with a separate cohort of 12 patients treated with the MDCA technique. Treatment delivery times for each fraction measured from initiation of the first beam or arc segment to initiation of the last beam or arc segment were significantly reduced with the MDCA technique. Treatment times per fraction using the NCB technique ranged from 9.4 to 34.3 minutes with a median time of 12.2 minutes. After the MDCA technique was adopted for routine use in treating peripheral tumors, the treatment time necessary for a separate group of 12 consecutive
patients treated with the technique to the same total dose of 54 Gy in 3 fractions was recorded for each fraction. Treatment times per fraction ranged from 3.8 to 26.7 minutes, with a median of 7.0 minutes. This corresponds to a 42% reduction in treatment time (P b .0001).
Discussion The SBRT has been adopted as a standard treatment option for patients with medically inoperable, peripheral, early stage NSCLC. Many series have been published demonstrating excellent 2- and 3-year local control rates with SBRT using several different treatment techniques, including multiple coplanar beams, multiple noncoplanar beams, and various arc techniques.2-11 To date, however, few studies have evaluated which treatment technique is preferable. A study at the University of Maryland evaluated several static coplanar, static noncoplanar, and rotational arc therapy plans for 37 different medically inoperable lung tumors on the basis of the RTOG 0236 criteria.13 They found that nearly all treatment techniques satisfied the high-dose conformality parameters and normal tissue dose constraints; however, 81% of cases had improved moderate-dose conformality when using an
132
C.C. Ross et al
Practical Radiation Oncology: April-June 2011
Figure 5 Representative dose distribution demonstrating 50% dose coverage of the planning target volume for the modified dynamic conformal arc (top) and noncoplanar beam plans (bottom) in both the coronal (left) and sagittal (right) views.
NCB technique. Conversely, a recent study from William Beaumont Hospital generated volumetric modulated arc therapy (VMAT) plans for 21 patients with stage IA NSCLC previously treated stereotactically with 7-10
Figure 6 Comparison showing the percentage of patients who had a more conformal plan generated with either the noncoplanar beam (NCB) or modified dynamic conformal arc (MDCA) technique by high- and moderate-dose conformality criteria. NT, normal tissue; PTV, planning target volume.
nonoppposing, noncoplanar static beams.14 They found that the 2 techniques generated plans that had equivalent mean maximum doses to the PTV and similar 95% dose conformality. The VMAT plans, however, were more conformal at the 80% and 50% isodose lines. The dose to normal tissues was also equivalent between the 2 techniques, with the exception of normal lung dose, which saw a reduction in all parameters (V20/12.5/10/5) with the VMAT technique. In addition, the VMAT technique reduced treatment times by 37%-63%. In this study, we describe a novel MDCA technique, which not only simplifies treatment planning and shortens overall treatment delivery time, but also generates more conformal treatment plans and provides better PTV coverage for the superior-most and inferior-most portions of the PTV. During normal respiration, lung tumors often experience significant motion in the superior and inferior directions. This makes coverage of the PTV in these areas critically important. In this regard, a noncoplanar technique generally provides a better dosimetric plan than a coplanar technique, as the noncoplanar beams are not tangential to the superior and inferior edges of the PTV. In particular, using a coplanar arc technique often generates a disc-shaped dose distribution with a very flat region of rapid-dose fall-off immediately superior and inferior to the
Practical Radiation Oncology: April-June 2011
Figure 7 Comparison showing percentage of plans generated with the noncoplanar beam (NCB) and modified dynamic conformal arc (MDCA) techniques that meet Radiation Therapy Oncology Group (RTOG) 0236 protocol parameters regarding high- and moderate-dose conformality factors. NT, normal tissue; PTV, planning target volume.
PTV. In this case, the dose distribution will be increased in the radial directions proportionally more than in the superior and inferior directions when normalizing the plan to meet PTV coverage parameters in the superior and inferior portions of the PTV. The technique described in this study makes use of arc segments expanded in the superior and inferior directions and allows for improved coverage of these regions without the need to expand the dose distribution radially to the same degree as is necessary with a traditional arc technique. Ultimately, this expansion results in a more spherical dose distribution. Even when compared with plans generated using a traditional, nonexpanded NCB technique, it provides improved coverage of the superior- and inferior-most regions of the PTV. It should be noted, however, that this improved dose coverage in the superior and inferior regions of the PTV can likely be attributed solely to the PTV expansion used in the MDCA treatment planning process. It can be assumed that a similar optimization PTV created for the NCB plans would result in similar coverage. In this study, the NCB plans were developed according to the guidelines set out in the RTOG 0236 protocol, and thus the PTV was not expanded for planning purposes. Previous studies evaluating optimal block margins around the PTV have suggested that plans with greater conformality and less normal lung dose are generated when using no additional margin.15 However, the NCB plans used in this study were re-planned with a 5-mm (2 computed tomographic sliced) expansion superiorly and inferiorly to approximate the average expansion used in the MDCA technique planning process. As a result, the coverage in these regions improved and was similar to the MDCA plans (data not shown).
Modified dynamic conformal arc technique
133
This more spherical dose distribution produced by the MDCA plans also resulted in improved target conformality. The MDCA technique produced plans that were significantly more conformal by all high-dose conformality criteria that were examined, including V105/PTV, target CI, and the Paddick CI, than plans generated using a NCB technique. Even when comparing moderate-dose conformality criteria, the MDCA technique produced plans with significantly lower maximum doses 2 cm from the PTV (D2cm), and the ratio of the 50% isodose volume to the volume of the PTV was equivalent between the 2 techniques. Again, it should be noted that at least part of the improved conformality with the MDCA technique is a result of the PTV expansion. When the PTVs for the NCB plans were expanded, the conformality of the resultant plans was much closer to the MDCA plans by the Paddick CI (median, 0.76 and 0.78 for the expanded NCB and MDCA plans, respectively; P = .004). It is expected that an arc technique would result in a greater volume of normal tissue being irradiated, as is evidenced by the higher median lung V20 and mean lung dose in this study. However, when examining the absolute magnitude of these values (V20 of 2.7% and 2.5% for MDCA and NCB plans, respectively; and the mean lung dose of 2.6 Gy and 2.3 Gy for MDCA and NCB plans, respectively), there is little difference, and all plans generated with both techniques were well within the lung dose constraints suggested in the RTOG 0236 protocol. Furthermore, the median volume of chest wall receiving N 30 Gy (V30) and N 60 Gy (V60) was slightly higher in the NCB plans than in the MDCA plans. Neither was statistically significant, although the increase in median V60 did approach significance (P = 0.054). The blocking along the chest wall could be optimized to reduce the chest wall dose for the NCB plans; however, it is our institutional practice not to limit the chest wall dose at the cost of PTV coverage. In addition, the absolute magnitude of the chest wall dose for each plan was not excessive in the cases evaluated. With regard to the remaining normal tissue constraints set forth in the RTOG protocol, all parameters were respected using both techniques, compared with a 73% normal tissue dose constraint compliance rate in the published results of RTOG 0236.9 Additional benefits of the MDCA technique include a simplified SBRT treatment planning process and shorter treatment delivery times. The simplification of the planning process stems from the ability to use a template as a starting point, in contrast to the need to identify 7-10 or more optimal beam angles for each patient when using the NCB method. By using a template, which automatically loads 6 arc segments of 60° each, the user needs only to define an isocenter and conform each equally weighted arc segment to the appropriate expanded PTV. In rare cases, the weighting of the arc segments must be modified to respect normal tissue dose constraints, but for the majority of peripheral lesions this is not necessary.
134
C.C. Ross et al
In addition to a more efficient treatment planning process, we also observed a significant reduction in treatment delivery time. The arc segments generated using the MDCA technique rotate around a single isocenter. The isocenter is placed in such a location as to allow full gantry rotation around the patient without obstruction, thus obviating the need to enter the treatment room between each beam to make adjustments in couch position. When we compared the treatment times for patients at our institution who were treated using the NCB technique with a group of 12 patients treated with the MDCA technique on the same equipment, we observed a 42% reduction in treatment time using the MDCA technique. In patients with medically inoperable lung tumors, a reduction in treatment time is especially important, because many patients in this population have significant difficulty remaining motionless during a lengthy treatment. This could potentially reduce error associated with patient movement, as well as improve clinical flow. As a result of the enhanced PTV coverage, improved high- and moderate-dose conformality, simplified treatment planning process, and reduced treatment time afforded by the MDCA technique, we have adopted it as the primary SBRT method for treating peripheral lung tumors at our institution. In addition, we are evaluating its utility for the treatment of both more centrally located early stage NSCLCs and liver SBRT.
References 1. Ginsberg RJ, Rubinstein LV. Randomized trial of lobectomy versus limited resection for T1N0 non–small cell lung cancer. Ann Thorac Surg. 1995;60:615-623. 2. Fakiris AJ, McGarry RC, Yiannoutsos CT, et al. Stereotactic body radiation therapy for early-stage non–small-cell lung carcinoma: four-year results of a prospective phase II study. Int J Radiat Oncol Biol Phys. 2009;75:677-682.
Practical Radiation Oncology: April-June 2011 3. Baumann P, Nyman J, Lax I, et al. Factors important for efficacy of stereotactic body radiotherapy of medically inoperable stage I lung cancer. A retrospective analysis of patients treated in the Nordic countries. Acta Oncol. 2006;45:787-795. 4. Baumann P, Nyman J, Hoyer M, et al. Outcome in a prospective phase II trial of medically inoperable stage I non–small-cell lung cancer patients treated with stereotactic body radiotherapy. J Clin Oncol. 2009;27:3290-3296. 5. Nagata Y, Negoro Y, Aoki T, et al. Clinical outcomes of 3D conformal hypofractionated single high-dose radiotherapy for one or two lung tumors using a stereotactic body frame. Int J Radiat Oncol Biol Phys. 2002;52:1041-1046. 6. Nagata Y, Takayama K, Matsuo Y, et al. Clinical outcomes of a phase I/II study of 48 Gy of stereotactic body radiotherapy in 4 fractions for primary lung cancer using a stereotactic body frame. Int J Radiat Oncol Biol Phys. 2005;63:1427-1431. 7. Hara R, Itami J, Kondo T, et al. Clinical outcomes of single-fraction stereotactic radiation therapy of lung tumors. Cancer. 2006;106: 1347-1352. 8. Onishi H, Shirato H, Nagata Y, et al. Hypofractionated stereotactic radiotherapy (HypoFXSRT) for stage I non–small cell lung cancer: updated results of 257 patients in a Japanese multi-institutional study. J Thorac Oncol. 2007;2:S94-S100. 9. Timmerman R, Paulus R, Galvin J, et al. Stereotactic body radiation therapy for inoperable early stage lung cancer. JAMA. 2010;303(11): 1070-1076. 10. Xia T, Li H, Sung Q, et al. Promising clinical outcome of stereotactic body radiation therapy for patients with inoperable stage I/II non–small cell lung cancer. Int J Radiat Oncol Biol Phys. 2006;66:117-125. 11. Timmerman R, McGarry R, Yiannoutsos C, et al. Excessive toxicity when treating central tumors in a phase II study of stereotactic body radiation therapy for medically inoperable early-stage lung cancer. J Clin Oncol. 2006;24:4833-4839. 12. Paddick I. A simple scoring ratio to index the conformity of radiosurgical treatment plans. J Neurosurg. 2000;93(Suppl 3): 219-222. 13. Lim DH, Yi BY, Mirmiran A, et al. Optimal beam arrangement for sterotactic body radiation therapy delilvery in lung tumors. Acta Oncol. 2010;49:219-224. 14. McGrath SD, Matuszak MM, Yan D, et al. Volumetric modulated arc therapy for delivery of hypofractionated stereotactic lung radiotherapy: a dosimetric and treatment efficiency analysis. Radiother Oncol. 2010;95:153-157. 15. Cardinale RM, Wu Q, Benedict SH, et al. Determining the optimal block margin on the planning target volume for extracranial stereotactic radiotherapy. Int J Radiat Oncol Biol Phys. 1999;45:515-520.
