Evaluation of fixed‐jaw IMRT and tangential partial‐VMAT radiotherapy plans for synchronous bilateral breast cancer irradiation based on a dosimetric study

Abstract Purpose To investigate the fixed‐jaw intensity‐modulated radiotherapy (F‐IMRT) and tangential partial volumetric modulated arc therapy (tP‐VMAT) treatment plans for synchronous bilateral breast cancer (SBBC). Materials and method Twelve SBBC patients with pTis‐2N0M0 stages who underwent whole‐breast irradiation after breast‐conserving surgery were planned with F‐IMRT and tP‐VMAT techniques prescribing 42.56 Gy (2.66 Gy*16f) to the breast. The F‐IMRT used 8‐12 jaw‐fixed tangential fields with single (sF‐IMRT) or two (F‐IMRT) isocenters located under the sternum or in the center of the left and right planning target volumes (PTVs), and tP‐VMAT used 4 tangential partial arcs with two isocenters located in the center of the left and right PTVs. Plan evaluation was based on dose‐volume histogram (DVH) analysis. Dosimetric parameters were calculated to evaluate plan quality; total monitor units (MUs), and the gamma analysis for patient‐specific quality assurance (QA) were also evaluated. Results For PTVs, the three plans had similar Dmean and conformity index (CI) values. F‐IMRT showed a slightly better target coverage according to the V100% values and demonstrated an obvious reduction in V105% and Dmax compared with the values observed for sF‐IMRT and tP‐VMAT. Compared with tP‐VMAT, sF‐IMRT was slightly better in terms of V100%, V105% and Dmax. In addition, F‐IMRT achieved the best homogeneity index (HI) values for PTVs. Concerning healthy tissue, tP‐VMAT had an advantage in minimizing the high dose volume. The MUs of the tP‐VMAT plan were decreased approximately 1.45 and 1 times compared with the sF‐IMRT and F‐IMRT plans, respectively, and all plans passed QA. For the lungs, heart and liver, F‐IMRT achieved the smallest values in terms of Dmean and showed a significant difference compared with tP‐VMAT. Simultaneously, sF‐IMRT was also superior to tP‐VMAT. For the coronary artery, tP‐VMAT achieved the lowest Dmean, while the value for F‐IMRT was 2.24% lower compared with sF‐IMRT. For all organs at risk (OARs), tP‐VMAT was superior at the high dose level. In contrast, sF‐IMRT and F‐IMRT were obviously superior at the low dose level. The sF‐IMRT and F‐IMRT plans showed consistent trends. Conclusion All treatment plans for the provided techniques were of high quality and feasible for SBBC patients. However, we recommend F‐IMRT with a single isocenter as a priority technique because of the tremendous advantage of local hot spot control in PTVs and the reduced dose to OARs at low dose levels. When the irradiated dose to the lungs and heart exceed the clinical restriction, two isocenter F‐IMRT can be used to maximize OAR sparing. Additionally, tP‐VMAT can be adopted for improving cold spots in PTVs or high‐dose exposure to normal tissue when the interval between PTVs is narrow.

F-IMRT were obviously superior at the low dose level. The sF-IMRT and F-IMRT plans showed consistent trends.
Conclusion: All treatment plans for the provided techniques were of high quality and feasible for SBBC patients. However, we recommend F-IMRT with a single isocenter as a priority technique because of the tremendous advantage of local hot spot control in PTVs and the reduced dose to OARs at low dose levels. When the irradiated dose to the lungs and heart exceed the clinical restriction, two isocenter F-IMRT can be used to maximize OAR sparing. Additionally, tP-VMAT can be adopted for improving cold spots in PTVs or high-dose exposure to normal tissue when the interval between PTVs is narrow. dence showing that BCS plus radiotherapy is at least equivalent to mastectomy with respect to long-term survival. [1][2][3][4][5][6][7] Additionally, considering cosmetic and breast cancer-specific survival probabilities, breast-conservation treatment would be suggested, and standard treatment procedures include BCS followed by whole-breast irradiation (WBI). However, no definite radiation therapy technique has yet been reported or established for SBBC. Compared with UBC irradiation, SBBC irradiation is more complex, with the concomitant involvement of both the lungs and heart and a wider distribution of treatment volume.
Three-dimensional conformal radiotherapy (3D-CRT) represents the most common approach for WBI. 3D-CRT is generally delivered by two tangential fields for each breast and usually causes over/underdosage at field junctions and increased dose heterogeneity over the whole breast, especially in large-breasted patients. Additionally, organs at risk (OARs) that lie in the same direction as the target cannot be fully protected. 8,9 In recent years, highly conformal radiation therapy techniques, such as intensity-modulated radiation therapy (IMRT) and volumetric modulated arc therapy (VMAT), have been proposed to achieve the required target dose coverage while ensuring adequate normal tissue sparing. A few dosimetric studies have been conducted on IMRT, VMAT or helical tomotherapy for SBBC. [10][11][12][13][14] Most of those studies employed a single isocenter located under the sternum. This approach can compress treatment time but is limited in clinical use, especially for obese patients. Moreover, in the aforementioned studies, how the treatment plans were executed was not specified, and more specific information regarding the treatment planning for bilateral breast cases was not provided.
In the present investigation, we designed fixed-field IMRT (F-IMRT) and tangential partial VMAT (tP-VMAT) treatment plans for SBBC to identify an efficient method that can solve outstanding dose distribution problems and be applicable to various patients in the clinic.

2.A | Patient selection and planning objectives
From February 2017 to March 2018, 12 SBBC patients with medical histories of ductal or lobular carcinoma were referred to our institution to receive adjuvant radiotherapy. All patients were diagnosed with stage pTis-2N0M0 and underwent BCS; the median age was 45 (range, 31-64). CT scans were acquired with a thickness of 5 mm in free breathing mode; and the position was head-first supine as the treatment position, with arms elevated. The scan range was from the sixth cervical to the second lumbar vertebra and included the entire lung volume and liver.
All targets and structures were contoured by the same oncology physician according to ESTRO guidelines. The clinical target volumes (CTVs) were those encompassing the entire breast. The planning target volumes (PTVs) were obtained with an expansion of 8 mm in all directions from the CTVs and restricted to the skin cropping at 5 mm from the surface, excluding the ribs (Fig. 1) Gy and V 30Gy < 10%; for the coronary artery area, mean dose < 25 Gy and V 40Gy < 33%; and for the liver, V 30Gy < 30%.

2.B | Planning techniques
The plans were generated using the Varian Eclipse treatment plan- Analytical Algorithm (AAA) algorithm was used for dose calculation, and calculation grid was set to 2.5 mm. The photon energy used was 6 MV, and a maximum dose rate of 600 MU/min for VMAT and a fixed dose rate of 600 MU/min for IMRT were applied.

2.B.1 | sF-IMRT
The IMRT technique with multiple treatment fields has been reported to increase the low dose volumes in the ipsilateral and contralateral lung and heart. 10,13 In this study, we adopted the F-IMRT technique to decrease heart and lungs irradiation without increasing the low dose volume.
Initially, plans were optimized for a single isocenter approach, which was located medially under the sternum. Approximately 8-12 fixed-jaw beams were used, and two or three pairs were similarly tangential beams for each target to avoid anterior and posterior entrances [ Fig. 1(a)]. In the first step, we identified the inner tangent field in which the beam's eye view (BEV) of the target had the minimum projection. Then, taking this field as the starting point, the remaining one or two fields were identified in the clockwise direction (the counterclockwise direction was used for the right target), and each field was spaced 6-10 degrees. Finally, similar outer tangential fields were identified. All beams were coplanar beams, and the collimator angle was set parallel to the long axis of the focus targets. Fixed jaws have been used to maximize the OARs sparing in the outer tangential field. As shown in Fig. 2(a), the fixed X1 jaw can reduce lung exposure, and the shield target can be complemented by the opposite tangent fields. Similarly, the fixed X2 jaw in Fig. 2 can reduce heart irradiation.

2.B.2 | F-IMRT
Different patients exhibit significant differences in breast shape and width in different parts of the breast. When the patient is obese, or the volume of the lungs is relatively small, high doses to the lungs and heart can result. When the irradiated dose to the lungs and heart exceeds the clinical restrictions, two isocenter F-IMRT is used.
In order to compare the sF-IMRT approach, we used F-IMRT technique for all 12 patients in this paper.
There were two isocenters, each of which was located in the center of targets, one for the left PTV and one for the right PTV F I G . 1. Beam arrangements and isocenter positions for (a) single fixed-jaw intensity modulated radiotherapy (sF-IMRT), (b) F-IMRT, and (c) tangential partial volumetric modulated arc therapy. For F-IMRT, six fixed-jaw fields aimed to geometrically cover the left breast (red lines) and the other six (yellow lines) focused on the right breast. The interval angle for the ipsilateral fields was 6-10 degrees. For tangential partial VMAT, two partial arcs, rotating in opposite directions, aimed to geometrically cover primarily either the left (red line) or right (yellow line) breast. Dashed sectors were avoidance areas where the accelerator beam was off while the gantry was rotating. Usually, two to three times optimizations can produce satisfactory results. With additional optimizations, the MU will be increased, which can be improved with a high smooth factor during the optimization phase. Considering the intrafraction motion of organs and interfraction setup errors, the tangential breast field in 3D-CRT plans consistently and extensively irradiates the region outside the skin surface, which was defined by typically adding 2 cm to the anterior edge of the field. However, most commercially available TPSs assign a zero-dose region outside the skin, which leads to the MLC close to the surface in IMRT plans. In this study, we used a skin flash tool to extend the fluence outside the surface and selected the appropriate cut range parameter and brush ceiling in BEV to extend the fluence uniformly by 2 cm outside the skin.

2.B.3 | tP-VMAT
VMAT is a relatively new technique based on the simultaneous optimization of the MLC, gantry, and dose rate. Several studies have applied VMAT to UBC with varying results but with the same conclusions. Compared with tangential fields, VMAT achieves greater target coverage and homogeneity and reduces ipsilateral lung and heart doses in high dose volumes at the expense of increased lowdose volumes. [19][20][21] In previous dosimetric studies, full or partial arcs have been used for SBBC, [11][12][13][14] analogous to the use of VMAT for UBC, and the use of full or partial arcs consequently increased the irradiated volume. We introduced a restricted tangential partial VMAT technique to increase dose homogeneity and target coverage; furthermore, the radiation dose to the heart and lung can be decreased without increasing the low dose volume.
Example of fixed jaw use to reduce exposure of (a) right lung and (b) heart.
Example of fixed X2 jaw to reduce lung exposure. The X2 jaw can only exceed the central axis by 2 cm, and the white dotted line is where the X2 jaw is expected to be located.
In the initial attempt, we planned the treatment using a single isocenter. However, the results did not reach our goal described in Section 2.A. Therefore, two isocenters were used in this method and were located in the same manner as in the two isocenter F-IMRT plans. Four coplanar partial arcs were used in the present study: two for the left breast and two for the right breast, as shown in Fig. 1(c).
For the left breast, those two arcs started from 150 to 300 degrees (one was clockwise, the other was counterclockwise), and 90 to 0 degrees was the avoidance sector; therefore, the accelerator only beamed on the 120 degree sector (150-90 degrees, 0-300 degrees).
For the right breast, arcs started from 60 to 210 degrees (one was clockwise, the other was counterclockwise), and 350 to 280 degrees was the avoidance sector; the accelerator beamed on the 60-350 degree sector and the 280-210 degree sector. With respect to the heart, the avoidance sector for the left PTV was larger than that for the right PTV. To minimize the contribution of the tongue and groove effect during the gantry rotation, the TPS suggested a collimator angle fixed to 10-30 degrees instead of zero. 22 In this study, the collimator angle was set parallel to the long axis of the focus targets.
Because no fluence map is generated in VMAT plans, the skin flash tool cannot be used to expand the fluence outside the skin. A 2-cm-thick bolus was added in the region of PTV outside the skin during optimization, but deliverable beams and dose calculations were performed without a bolus, as suggested in ICRU 62. 23 In previous methods, all fields were simultaneously optimized to generate the desired dose distribution. The gamma passing rate should be ≥95%.

2.C | Evaluation tools
An independent sample t-test following a normality test (Shapiro-Wilk) was used to compare the results, and the threshold for significance was P < 0.05.

| RESULTS
A case of SBBC isodose distributions with sF-IMRT, F-IMRT, and tP-VMAT techniques is shown in Fig. 4 for the transverse, coronal, and sagittal planes. Figure 5 shows the mean DVHs for the lungs, heart, coronary artery area, and liver. The numerical DVH findings are F I G . 4. Case of isodose distribution on transverse, coronal, and sagittal views of (a) single fixed-jaw intensity modulated radiotherapy (sF-IMRT), (b) F-IMRT, and (c) tangential partial volumetric modulated arc therapy for synchronous bilateral breast cancer.
summarized in Tables 1 and 2. All data in the tables are normally distributed, and the P-values are shown.

3.A | PTV dose distribution
The comparison data in Table 1  Simultaneously, the MU ratio between sF-IMRT and tP-VMAT was 2.44. All QA results were above the criterion, which indicated those plans were clinically acceptable.
The data for the 12 patients showed that the three planning methods were eligible. Comparatively, the F-IMRT plan yielded the best results. During the planning process, F-IMRT reached the stated goals faster and more consistently.   Table 2. The mean values for the lungs, heart, and liver indicated in Table 2 were significantly different between the F-IMRT and tP-VMAT; however, the other two groups showed no significant differences. The DVHs in Fig. 5 indicate that the values for tP-VMAT were distinctly lower than those for sF-IMRT and F-IMRT at high dose levels and were higher at low dose levels. Obviously, F-IMRT had similar DVHs with sF-IMRT.

3.B | OAR dose distribution
Bilateral lung protection was our concern. All mean values for both lungs under the plans were managed under 9 Gy, and F-IMRT for right lung; in terms of the heart, the mean value was 8.18 ± 3.06 Gy. These data are higher than those presented in Table 2. As we known, 3D-CRT is typically affected by the risk of over/underdosage at the field junctions, increased dose heterogeneity over the targets and large portions of heart and lungs that cannot be dosimetrically spared. The purpose of this study was to provide an alternative protocol for SBBC irradiation treatment with highly conformal RT while assuring adequate normal tissue sparing.
Although several studies on SBBC have been conducted, the planning details have not been sufficient. The sF-IMRT, F-IMRT, and tP-VMAT methods were suggested in this study to improve target coverage and subsequent disease control while sparing normal tissue and reducing patient toxicity.
In present study, the patients were treated in the supine position. Several studies have shown that the ipsilateral lung dose is F I G . 5. Mean dose-volume histograms (averaged over the 12 patients) of the left (a) and right (b) lungs, heart (c), coronary artery area (d), and liver (e).
T A B L E 1 Comparison of PTVs coverage for single fixed-jaw intensity modulated radiotherapy (sF-IMRT), tangential partial volumetric modulated arc therapy (tP-VMAT), and F-IMRT.  11 In another study, two half-arcs (180 degrees) consisted of a rotating beam on each breast. 13 Those continued arcs increased the low-dose volume to the lungs and heart compared with the volume observed in the tangential field technique. Qiu et al. used a modified partial arc in which a portion of the arc was blocked to minimize the normal structure dose for partial breast irradiation. 30 Another study used two tangential dual arcs of 50-60°to decrease the low-dose volume for the ipsilateral lung and heart for left WBI. 31 In the present study, four tangential partial arcs were used: two for the left breast and two for the right breast. The first step was to determine the partial arc range based on the individual PTV location. Then, a shield area was selected within the arc to avoid angles directed toward the heart and lungs.
The avoidance sectors were defined before the optimization procedure. Usually, the avoidance sectors in the left breast are larger than those in the right breast, owing to the heart and coronary artery area.
The three techniques selected for this investigation were meticulously designed with tangential fields to improve dose coverage and decrease the radiation exposure of normal tissue. The data presented in the tables suggest that three techniques can obtain satisfactory results as described in section 2.A for the treatment of SBBC. As long-term breast pain, cosmesis, and quality of life highly correlate with moist desquamation and severe acute pain during radiation treatment, it is justified to optimize dose homogeneity to prevent acute side effects, particularly moist desquamation. 32 The 2018 ASTRO Guideline recommendation for dose homogeneity for hypofractionated whole-breast irradiation (HF-WBI) is that the volume of breast tissue receiving 105% of the prescribed dose should be minimized, regardless of dose fractionation. 33 According to the V 105% values in Table 1, F-IMRT showed the smallest dose distribution, with only approximately 1% of both PTVs, which is a great advantage compared with sF-IMRT and tP-VMAT. Additionally, the F-IMRT plan was superior to the other plans with regard to HI and CI.
However, hot spots out of target and cold spots in PTVs were challenges for sF-IMRT and F-IMRT because the number of tangential fields was limited, and the dose intensity modulation was limited from the tangential directions. Thus, hot spots out of target could often be found, especially above the sternum when the interval between the left and right PTVs was narrow. Based on this study, we found that when the interval distance was smaller than 2.5 cm, hot spots could only be eliminated at the expense of PTV coverage in sF-IMRT and F-IMRT plans. Simultaneously, cold spots in PTVs adjacent to the sternum were also a distinct disadvantage due to tight high-dose control in PTVs and strict normal tissue avoidance, especially when the tumor bed happened to be near the inner tangential field in the PTV area. Additionally, among BRCA carrier patients treated for early breast cancer, prophylactic irradiation to the contralateral breast to reduce the risk of subsequent contralateral breast cancer is gradually gaining wide acceptance. 34 Because there is no accurate tumor bed in the contralateral breast, the cold spots that appear in the breast adjacent to the sternum may influence this prophylactic irradiation effect. tP-VMAT further reduced the presence of hot spots out of target and improved cold spots in targets with multiple tangential partial arcs and simultaneous modulation of the dose rate. Therefore, in this situation, the dose coverage of tP-VMAT is superior to that of sF-IMRT and F-IMRT.
Regarding OARs and normal tissue, the reduction of the highdose region decreases the incidence of acute and late toxicities, such as acute radiodermatitis, symptomatic radiation pneumonitis and skin fibrosis. Moreover, the low-dose region within OARs may decrease the incidence of second primary malignant tumors induced by radiotherapy. 35 The tP-VMAT method can increase the volume of tissue exposed to low doses of radiation and thus increase the risk for radiation-induced carcinogenesis. The technique has also been shown to decrease the volume of tissue exposed to high doses of radiation, which could exert a beneficial impact on the risk for radiationinduced sarcoma (RIS). However, the overall risk for RIS is small compared with the potential benefit of radiotherapy. 36 To verify the efficiency of each technique, we compared MUs.
The MUs optimized by tP-VMAT were far less than those of sF-IMRT and F-IMRT. Because of the reduced number of MUs in tP-VMAT, the delivery time was reduced, which should not be neglected in clinical treatment.