Dosimetric comparison of four different external beam partial breast irradiation techniques: Three-dimensional conformal radiotherapy, intensity-modulated radiotherapy, helical tomotherapy, and proton beam therapy

Radiotherapy and Oncology 90 (2009) 66–73

Contents lists available at ScienceDirect

Radiotherapy and Oncology journal homepage: www.thegreenjournal.com

Partial breast irradiation

Dosimetric comparison of four different external beam partial breast irradiation techniques: Three-dimensional conformal radiotherapy, intensity-modulated radiotherapy, helical tomotherapy, and proton beam therapy Sung Ho Moon a, Kyung Hwan Shin a,b,*, Tae Hyun Kim a, Myonggeun Yoon a, Soah Park a, Doo-Hyun Lee a, Jong Won Kim a, Dae Woong Kim a, Sung Yong Park a, Kwan Ho Cho a a b

Proton Therapy Center, National Cancer Center, Gyeonggi-do, Republic of Korea Center for Breast Cancer, National Cancer Center, Gyeonggi-do, Republic of Korea

a r t i c l e

i n f o

Article history: Received 29 May 2008 Received in revised form 11 September 2008 Accepted 13 September 2008 Available online 5 November 2008 Keywords: Breast cancer Accelerated partial breast irradiation Protons Helical tomotherapy Intensity-modulated radiotherapy 3D-conformal Comparison

a b s t r a c t Background and purpose: As an alternative to whole breast irradiation in early breast cancer, a variety of accelerated partial breast irradiation (APBI) techniques have been investigated. The purpose of our study is to compare the dosimetry of four different external beam APBI (EB-APBI) plans: three-dimensional conformal radiation therapy (3D-CRT), intensity-modulated radiation therapy (IMRT), helical tomotherapy (TOMO), and proton beam therapy (PBT). Methods and materials: Thirty patients were included in the study, and plans for four techniques were developed for each patient. A total dose of 30 Gy in 6 Gy fractions once daily was prescribed in all treatment plans. Results: In the analysis of the non-PTV breast volume that was delivered 50% of the prescribed dose (PD), PBT (mean: 16.5%) was superior to TOMO (mean: 22.8%), IMRT (mean: 33.3%), and 3D-CRT (mean: 40.9%) (p < 0.001). The average ipsilateral lung volume percentage receiving 20% of the PD was significantly lower in PBT (0.4%) and IMRT (2.3%) compared with 3D-CRT (6.0%) and TOMO (14.2%) (p < 0.001). The average heart volume percentage receiving 20% and 10% of the PD in left-sided breast cancer (N = 19) was significantly larger with TOMO (8.0%, 19.4%) compared to 3D-CRT (1.5%, 3.1%), IMRT (1.2%, 4.0%), and PBT (0%, 0%) (p < 0.001). Conclusions: All four EB-APBI techniques showed acceptable coverage of the PTV. However, effective nonPTV breast sparing was achieved at the cost of considerable dose exposure to the lung and heart in TOMO. Ó 2008 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology 90 (2009) 66–73

Conservative surgery followed by whole breast irradiation (WBRT) has been established as an effective alternative to mastectomy in stages I–II breast cancer [1–4]. Recently, accelerated partial breast irradiation (APBI) has been investigated as an alternative to time-consuming WBRT in early stage breast cancer, as most local recurrences develop at, or in proximity to, the primary tumor bed [5–7]. A variety of APBI techniques are currently available, including interstitial brachytherapy, intracavitary brachytherapy, intraoperative radiation therapy (RT), and external beam RT. Despite the recent technical advancement of APBI using brachytherapy [8,9], external beam APBI (EB-APBI) has several advantages over brachytherapy-based APBI, including non-invasiveness and treatment initiation based on the final pathology [10]. In addition, EB-APBI offers a more homogenous dose distribution than

* Corresponding author. Address: Proton Therapy Center and Center for Breast Cancer, National Cancer Center, 809 Madu-1-dong, Ilsandong-gu, Goyang-si, Gyeonggi-do 411-769, Republic of Korea. E-mail address: [email protected] (K.H. Shin). 0167-8140/$34.00 Ó 2008 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.radonc.2008.09.027

brachytherapy-based APBI does. However, in a comparative study of APBI techniques, it was pointed out that the better target coverage of three-dimensional conformal RT (3D-CRT) over brachytherapy could be obtained with higher integral dose to the lung, heart, or to the remaining normal breast [11]. As external beam RT (EBRT) evolved from conventional RT into 3D-CRT and intensity-modulated RT (IMRT), treatment planning and delivery improved for various malignancies. Most planning studies have shown that IMRT is superior to 3D-CRT in treating the target volume adequately while sparing healthy organs and tissues [12–15]. Regarding the outcome of breast cosmesis at a short-term follow-up in a prospective phase II trial, IMRT was compared favorably with a previous series that used 3D-CRT [13]. Helical tomotherapy (TOMO) is another novel approach that enables effective intensity modulation of radiation delivery, and the feasibility of TOMO in APBI is currently being tested [14,16,17]. Proton beams, unlike X-ray beams, have a low entrance dose, followed by a region of uniform high dose (the spread-out Bragg peak) at the target, then a steep fall-off to zero dose. These charac-

S.H. Moon et al. / Radiotherapy and Oncology 90 (2009) 66–73

teristics minimize the dose delivered to normal tissues while maximizing the dose delivered to the tumor. Accelerated partial breast irradiation using proton beam therapy (PBT) has been reported to achieve excellent PTV coverage and dose homogeneity while significantly reducing the volume of irradiated non-target breast tissue by an average of 36%, compared to the 3D-CRT-based APBI [18–20]. Recently, a treatment planning study which compared several EB-APBI techniques was published [14], but the implication of the results was somewhat limited in that this study was based on a small number of patient population (N = 15), and the potential of PBT could not be evaluable compared with the other sophisticated techniques of EB-APBI, including 3D-CRT, IMRT, and TOMO, despite that the use of PBT in APBI is increasing. Our aim is to complete a comprehensive analysis of 3D CRT, IMRT, TOMO, and PBI, which includes a relatively sufficient number of patients from the standpoint of planning study, by comparing all techniques on the basis of several dosimetric criteria which are representative of homogeneity, coverage, conformity, and the ability to spare normal tissues. Methods and materials Study population The study population consisted of a subset of patients (N = 30) enrolled in an ongoing institutional trial, ‘A phase II study of proton beam accelerated partial breast irradiation after lumpectomy in patients with early breast cancer’ (NCCCTS-07-248). Eligible patients were characterized as having fully excised breast carcinoma that was not more than 3 cm in diameter, with negative margins of at least 2 mm and negative axillary lymph nodes. The lumpectomy cavity had to be marked with surgical clips, which were placed at the superior, inferior, medial, lateral, and the deep margins of the cavity. In accordance with the ongoing trial (NCCCTS-07-248), proton beam APBI was scheduled within 4 weeks after breast-conserving surgery to minimize migration of the surgical clips. The study was performed in accordance with the guidelines of our institutional review board, and an informed consent was obtained from each patient before the initiation of treatment. Treatment planning A treatment planning CT scan (Philips Ultra Z, Philips Medical System, Andover, MA, USA) was acquired with the patient in the supine position. The CT started at or above the mandible and extended several centimeters below the inframammary fold, at 3 mm intervals. The following structures were contoured in the planning stage: planning target volume (PTV), skin, ipsilateral and contralateral breasts, thyroid, ipsilateral and contralateral

67

lungs, and heart. Ipsilateral breast volume was defined as the soft tissue volume, excluding lung and bony structures, that would be irradiated in WBRT using parallel opposite tangential beams and composed of PTV and non-PTV breast volume. To define the PTV, unequal margins were added according to the directional safety margin status of each lumpectomy cavity (i.e., medial, lateral, superior, inferior, and deep margins measured from the tumor front after the examination of the surgical specimen: 2, 1.5, and 1 cm for resection margins <1 cm, 1–2 cm, and >2 cm, respectively). The PTV was then modified so that it was no closer than 3 mm to the skin surface and was no deeper than the lung–chest interface. Although one may freely choose beam arrangement and beam number as long as the necessary dose–volume constraint is met, we chose a four-field non-coplanar beam arrangement for 3DCRT and IMRT, in order to compare these two techniques in similar conditions. No beams that are directed toward critical normal structures, such as heart, lung, or contralateral breast, are permitted in 3D-CRT or IMRT plan. In principle, we used the techniques of Baglan et al. for planning of 3D-CRT and IMRT [21]. Wedges were used in all 3D-CRT fields and their usual orientation was heel anterior. Gantry, couch, wedge angles, and beam weights were chosen by the planner. For 3D-CRT and IMRT plans, a Pinnacle3 treatment planning system (Version 7.6, Philips Electronics, Eindhoven, Netherlands) was used. The TomoTherapy treatment planning system (TomoTherapy Incorporated, Madison, WI, USA) was used for TOMO planning. In general, parameters specified as part of the optimization process are a field width, pitch, and a modulation factor. These factors are similar to those in conventional diagnostic CT scanning with the exception of modulation factor. The user can choose 1.0, 2.5, and 5.0 cm field width, which is defined by the longitudinal extent of treatment field at machine isocenter. Typically, the field width will be 2.5 cm with 0.25 of pitch value, which is recommended by TomoTherapy Incorporated as the initial values. Modulation factor in TOMO determines the range of intensity levels that can be achieved. This factor was varied from its highest possible value of 5 to its lowest possible value of 1. At a modulation factor of 1, the only modulation that the leaves can make is to open at the same amount of time as all the other open leaves or the leaves which are closed. With a higher modulation factor (greater than 2.0), the plan can include a greater range of intensity values. For optimum dose distribution, small field width and pitch and high modulation factors should be employed at the expense, however, of longer treatment times. Although one must work hard to find the optimal modulation factor for each case, a modulation factor between 2.0 and 3.0 might be a good starting value for moderate modulation. In this study, we used a field width of 2.5 cm, a pitch of 0.3, and a modulation factor of 3.0. By increasing the field width, the pitch and by reducing the modulation factor, one can distinctly reduce the treatment time. For example, the total

Fig. 1. Beam arrangements for (a) three-dimensional conformal radiotherapy, (b) intensity-modulated radiotherapy, and (c) proton beam therapy.

68

Dosimetric comparison of four EB-APBI techniques

treatment time can be decreased from 10 to 8 min by decreasing the modulation factor to 80%. However, the resultant dose–volume histogram (DVH) should be evaluated to the original DVH to check whether it is at least compatible to the original plan. Our results show that the selected values for a field width, pitch, and a modulation factor worked well for APBI. Contralateral breast and lung were completely blocked, and the stepwise directional blocking technique and hundreds of iterations in each case of TOMO planning were done, as was reported in a previous study, to reduce the dose to the lung and heart using [16]. For proton treatment planning, the Eclipse proton planning system (Version 8.03, Varian Medical System, Inc., Palo Alto, CA, USA) was used. The proximal, distal, and transverse margins were 2 mm, 2 mm, and 10 mm, respectively, and the border smoothing and smearing margins were set to 0 mm and 5 mm, respectively. During PBT, there may be a misalignment of the range compensator and the patient, resulting in the potential dosimetric errors. The misalignment has been generally compensated for the margin smearing technique, formulated by Urie et al. [22]. Basically, this algorithm compares each pixel of the compensator and sets the pixel value of the thickness to the minimum of the neighboring pixel. A smearing margin of 3 mm, which we applied during planning, was also considered the margin for setup uncertainty only in aligning the compensator to the patient since the smearing margin for other uncertainties such as target motion had already been included in the PTV. Two or three fields were used to avoid the unnecessary entrance skin dose produced by direct spread-out Bragg peak when a single proton beam is used. In the case of PTV close to overlying skin, every effort was made to reduce the area of excessive skin dose. Beam arrangements for 3D-CRT, IMRT, and PBT are illustrated in Fig. 1, and the axial dose distributions for each of the four EBAPBI modalities are depicted in Fig. 2. Treatment plans for 3D-CRT, IMRT, and PBT were provided by two dosimetrists. Another senior dosimetrist made TOMO plans, and had a primary review of 3D-CRT, IMRT, and PBT plans. A phys-

icist and two radiation oncologists finally reviewed and confirmed those plans. Dosimetric parameters for plan comparison A total dose of 30 Gy in 6 Gy fractions once daily was prescribed in all treatment plans. Two guidelines were followed during APBI planning. These guidelines have been slightly modified from those suggested in the 3D-CRT guidelines of RTOG 0319 and NSABP B-39/ RTOG 0413 [23,24]. First, the 95% isodose surface should cover the 100% of the PTV (total coverage). Second, the maximum dose should not exceed 110% of the prescribed dose (PD). Marginal coverage (down to 95% of PTV covered by the 95% isodose surface) is accepted when total coverage is not possible without exceeding 110% of the PD. Dose limitations for normal tissues were also evaluated according to the RTOG guidelines mentioned above [23,24] (Table 1). When PTV coverage and normal tissue restriction criteria were contradictory, the former took priority. Although most of the clinical range-modulated proton beams are assumed to have a fixed overall relative biological effectiveness (RBE) of 1.1 [20], it

Table 1 Dose limitations for normal tissues. Uninvolved normal breast

Contralateral breast Ipsilateral lung Contralateral lung Heart (right-sided lesion) Thyroid

<60% whole reference breast volume should receive P50% of the PD <35% of whole reference breast volume should receive the PD <3% of the PD to any point <15% of the lung can receive 30% of the PD <15% of the lung can receive 5% of the PD <5% of the heart receiving 5% of the PD should be <40% Maximum point dose of 3% of the PD

Abbreviation: PD, prescribed dose.

Fig. 2. Dose distribution of (a) three-dimensional conformal radiotherapy, (b) intensity-modulated radiotherapy, (c) helical tomotherapy, and (d) proton beam therapy in the axial plane. Lumpectomy cavity (red), PTV (pink), and isodose lines of 103% (green), 100% (red), 90% (blue), 70% (yellow), 50% (cyan), and 30% (orange) are depicted.

S.H. Moon et al. / Radiotherapy and Oncology 90 (2009) 66–73

69

Table 2 Summary of dosimetric parameters.

Statistical analysis

Homogeneity index Conformity index Coverage index D2_PTV (Gy) D98_PTV (Gy) Dprescription_PTV (Gy) V100_NORM (mL)

One-way analysis of variance (ANOVA) was used to compare dosimetric differences among plans using the four EB-APBI techniques. For pairwise comparisons among the four EB-APBI techniques, a Bonferroni adjusted p value of 60.008 was considered to be significant to meet the overall significance level of <0.05. All statistical tests were two-sided, and were performed using SPSS software (release 12.0.1, SPSS Inc., Chicago, IL, USA).

V100_PTV (mL) VPTV (mL)

[(D2_PTV

D98_PTV)/Dprescription_PTV]  100

1 + (V100_NORM/V100_PTV) V100_PTV/VPTV Dose covering 2% volume of PTV Dose covering 98% volume of PTV Prescribed dose to PTV Volume of normal tissue receiving at least 100% of prescribed dose PTV receiving at least 100% of prescribed dose Actual volume of PTV

Abbreviation: PTV, planning target volume.

is well known that the RBE varies depending on various factors, such as beam energy and depth in the spread-out Bragg peak (SOBP). At our institution, we have adopted the use of a fixed 1.1 proton RBE until solid agreement is reached in the particle therapy community. Although we used 1.1 as a RBE for PBT, PD in PBT planning system is the same as the dose in photon treatment. Only output values or monitor units (MU) were adjusted based on the RBE values which provided the same PD. This means that the PD used in the planning stage is the same regardless of modality. Once planning is done, the output or MU can be decided based on the effective dose using RBE. Therefore, proton doses in this study represent Cobalt Gray Equivalent, and are presented as Gy. Definitions of the dosimetric parameters to evaluate and compare EB-APBI planning are summarized in Table 2. Homogeneity index (HI) was used to provide an adequate assessment of homogeneity in the PTV [25]. Dose covering 2% and 98% of the PTV (D2_PTV and D98_PTV) served as the maximum and the minimum doses, respectively. Therefore, a lower HI is indicative of a more homogeneous dose distribution across the PTV. The conformity index (CI) measures the degree of isodose conformity to the PTV, and it is crucial in the treatment efficiency [26–30]. Several definitions of CI are now available; by our definition, a plan with a lower CI value was more conformal. A DVH can also be used to assess a PTV coverage with a coverage index (CovI), defined as the fraction of the PTV receiving the PD. Ideally, 100% of the PTV will receive exactly the PD, but PTV coverage is by no means 100%, due to dose constraints of organ at risk. Another parameter V50_IB-NPTV was defined as the percentage of ipsilateral non-PTV breast volume receiving more than 50% of the PD. Unintended dose delivery to the lung and heart was evaluated using the DVH parameters, Vx_IL Vx_H and Dy%_IL, Dy%_H, Dy%_CL, Dy%_CB, defined as the percentage of ipsilateral lung and heart volume receiving more than x% of the PD, and the percentage of PD dose delivered to y% of the ipsilateral lung, heart, contralateral breast, and contralateral lung, respectively.

Results Eleven patients had right breast carcinoma, and 19 patients had left breast carcinoma. The majority of the tumors (63%) were located in the upper outer quadrant of the breast. Tumor size ranged between 0.1 and 2.9 cm in diameter (median: 1.2 cm). PTV ranged from 39.0 to 237.9 mL (mean: 83.8 mL, median: 94.6 mL). Mean and median ipsilateral breast volume (VIB) were 588.5 and 485.6 mL (range: 214.2–1712.3 mL), respectively. Mean and median PTV/VIB ratios were 17.0% and 16.7% (range: 9–30%), respectively. HI, CI, and CovI results for the four EB-APBI techniques are listed in Table 3. In one-way ANOVA, significant differences for mean value of HI, CI, and CovI were found among the planning techniques (p < 0.001). IMRT (mean: 5.48) provided the best HI, and no significant difference in HI was found among 3D-CRT (mean: 8.10), TOMO (mean: 8.34), and PBT (mean: 7.37). The no-variation conditions of dose homogeneity recommended by RTOG are (i) no more than 20% of any PTV will receive >110% of its prescribed dose, (ii) the prescription dose is the isodose that encompasses at least 95% of the PTV [25]. This suggests that for an acceptable plan, we can safely consider 95% and 110% as minimum (D98_PTV) and maximum (D2_PTV) doses, respectively. In this case, HI is equal to 15. Therefore, HI should be less than 15 for acceptable plans, whose condition was met in our data for all four modalities. Whereas the CI of TOMO (mean: 1.15) was significantly better than the others, 3D-CRT (mean: 2.89) was observed to have a significantly poorer CI. There was no difference between IMRT (mean: 1.94) and PBT (mean: 1.81) in terms of CI. CovI was acceptable in all EB-APBI techniques, with a mean value P0.95. Though significant differences in CovI existed across EB-APBI techniques, absolute differences were small and were not statistically significant in post hoc analysis. All EB-APBI plans met the primary planning goal at the level of marginal coverage (V95_PTV > 95%). Although the total coverage was not attainable without exceeding the maximum dose limitation of 110% of the PD in some proportion of patients, the minimum and the average value of V95_PTV was more than 97% and 99%, which is close to 100% (total coverage) in all EB-APBI techniques. There were significant differences for mean value of all parameters of PTV coverage, ipsilateral breast, and non-PTV breast among EBAPBI methods (p < 0.001). The average percent volume of the

Table 3 Comparison of homogeneity index (HI), conformity index (CI), and coverage index (CovI) for three-dimensional radiotherapy (3D-CRT), intensity-modulated radiotherapy (IMRT), helical tomotherapy (TOMO), and proton beam therapy (PBT). pa

Range (mean) 3D-CRT

IMRT

TOMO

HI

4.7–13.1 (8.10)

1.80–13.50 (5.48)

6.81–10.9 (8.34)

5.0–12.1 (7.37)

CI

1.9–8.9 (3.04)

1.42–2.85 (1.99)

1.10–1.44 (1.21)

1.66–2.20 (1.95)

3D-CRT vs. IMRT (<0.001), IMRT vs. PBT (1.000), PBT vs. TOMO (<0.001)

CovI

0.87–0.99 (0.95)

0.94–1.00 (0.97)

0.95–0.96 (0.95)

0.95–0.98 (0.96)

3D-CRT vs. TOMO (1.000), TOMO vs. PBT (0.331), PBT vs. IMRT (0.165)

a

PBT

By Bonferroni post hoc analysis with statistical significance defined as p < 0.008.

TOMO vs. 3D-CRT (1.000), 3D-CRT vs. PBT (1.000), PBT vs. IMRT (0.005)

70

Dosimetric comparison of four EB-APBI techniques

Table 4 Comparison of PTV and breast dosimetry for three-dimensional conformal radiotherapy (3D-CRT), intensity-modulated radiotherapy (IMRT), helical tomotherapy (TOMO), and proton beam therapy (PBT). pa

Range (mean) 3D-CRT PTV coverage V100_PTV V95_PTV V90_PTV

(%) 87.3–98.8 (94.8) 99.3–100 (99.9) 100 (100)

Ipsilateral breast (%) V100_IB 18.1–55.9 V75_IB 33.2–69.8 42.2–82.2 V50_IB 54.6–92.4 V25_IB

(32.8) (48.9) (57.6) (67.3)

Non-PTV breast (%) V50_IB-NPTV 25.0–64.7 (40.9)

IMRT

TOMO

PBT

94.4–99.9 (97.1) 97.3–100 (99.4) 98.7–100 (99.8)

94.5–95.8 (95.2) 98.8–99.9 (99.2) 99.7–99.9 (99.8)

95.0–97.9 (96.1) 98.6–100 (99.8) 99.5–100 (100)

3D-CRT vs. TOMO (1.000), TOMO vs. PBT (0.331), PBT vs. IMRT (0.165) TOMO vs. IMRT (0.731), IMRT vs. PBT(0.003), PBT vs. 3D-CRT (1.000) IMRT vs. TOMO (1.000), TOMO vs. PBT (0.010), PBT vs. 3D-CRT (1.000)

15.1–44.9 27.3–56.4 35.5–66.3 47.8–85.1

10.0–31.8 17.8–42.9 23.1–62.2 34.9–81.3

(18.7) (27.4) (39.8) (62.9)

9.9–29.0 (18.2) 16.2–41.5 (26.8) 21.0–50.1 (33.0) 25.4–55.2 (38.0)

3D-CRT 3D-CRT 3D-CRT 3D-CRT

10.6–32.4 (22.8)

10.3–23.2 (16.5)

3D-CRT vs. IMRT (<0.001), IMRT vs. TOMO (<0.001), TOMO vs. PBT (0.002)

(27.2) (41.4) (50.3) (60.5)

22.1–48.7 (33.3)

vs. vs. vs. vs.

IMRT (0.016), IMRT vs. TOMO (<0.001), TOMO vs. PBT (1.000) IMRT (0.002), IMRT vs. TOMO (<0.001), TOMO vs. PBT (1.000) IMRT (0.009), IMRT vs. TOMO (<0.001), TOMO vs. PBT (0.018) TOMO (0.503), TOMO vs. IMRT (1.000), IMRT vs. PBT (<0.001)

Abbreviations: PTV, planning target volume; Vx_IB, the percentage ipsilateral breast volume receiving more than x% of prescribed dose; V50_IB-NPTV, the percentage ipsilateral non-PTV breast volume receiving more than 50% of prescribed dose. a By Bonferroni post hoc analysis with statistical significance defined as p < 0.008.

ipsilateral breast receiving 100%, 75%, 50%, and 25% of the PD was the highest in 3D-CRT. Significantly less percent volume of the ipsilateral breast was exposed to 100%, 75%, and 50% of the PD in TOMO and PBT, compared with IMRT and 3D-CRT. Among all the EB-APBI techniques, PBT had the lowest volume of ipsilateral breast exposed to a lower dose level of V25_IB. On V50_IB-NPTV analysis, PBT was found to be the best technique for sparing ipsilateral normal breast. TOMO was better than IMRT on V50_IB-NPTV analysis, with 3D-CRT giving the poorest results on V50_IB-NPTV analysis. The results of PTV coverage and breast dosimetry are summarized in Table 4. The dose exposure for the ipsilateral lung was acceptable in almost all 3D-CRT, IMRT, and PBT plans, but we could not meet the normal tissue constraint guideline of V30_IL < 10% in four patients (13.3%) with TOMO. Furthermore, the dose to the heart could not be maintained below the constraint guideline of V5_H < 40% in four patients (21%) with left-sided breast cancer using the TOMO plan. Except for the contralateral lung (p = 0.396), significant differences were found among EB-APBI techniques for mean value of all dosimetric parameters of lung (p < 0.001), heart (p < 0.001), and contra-

lateral breast (p = 0.046). The average ipsilateral lung volume percentage receiving 20% of the PD was significantly lower in IMRT (2.3%) and PBT (0.4%), compared to 3D-CRT (6.0%) and TOMO (14.2%). The ipsilateral lung dose in PBT was significantly lower when compared to other techniques in terms of V5_IL. The findings observed in lung dosimetry were also noted in comparative analysis of the hearts of 19 patients with left-sided breast cancer. Concerning the average heart volume receiving 20% and 10% of the PD during the treatment of left breast cancer, TOMO (8.0% and 19.4%) was significantly larger than for 3D-CRT (1.5% and 3.1%), IMRT (1.2% and 4.0%), and PBT (0% and 0%). Contralateral lung and breast dose was minimal in all techniques (Table 5). For example, DVHs for the ipsilateral non-PTV breast volume and the ipsilateral lung volume of all EB-APBI techniques in the same patient are depicted in Fig. 3. Discussion The objective of this study was to compare the dosimetry of four EB-APBI techniques—3D-CRT, IMRT, TOMO, and PBT—and to

Table 5 Comparison of lung and heart dosimetry for three-dimensional conformal radiotherapy (3D-CRT), intensity-modulated radiotherapy (IMRT), helical tomotherapy (TOMO), and proton beam therapy (PBT). pa

Range (mean) 3D-CRT

IMRT

TOMO

PBT

0–14.3(2.3) 0–18.1 (4.7) 0–28.1 (8.2) 0–7.5 (1.8) 0–35.0 (4.8) 0–54.5(10.5)

3.9–29.1 (14.2) 13.2–57.7 (37.6) 23.1–78.3 (53.9) 6.8–24.5 (16.2) 11.3–34.8 (23.0) 16.4–46.3 (30.0)

0–1.7 0–2.6 0–3.6 0 0 0–1.0

Heart (%) (N = 19, left-sided) V20_H 0–9.0 (1.5) 0–13.3 (3.1) V10_H 0–22.5 (6.7) V5_H 0.7–5.5 (2.3) D20%_H D10%_H 1.1–24.0 (4.9)

0–7.9 (1.2) 0–40.2 (4.0) 0–50.4 (5.4) 0–15.3(1.6) 0–52.0 (7.1)

0–33.5 (8.0) 0–53.3 (19.4) 0.10–64.5 (25.7) 0.8–27.9 (8.8) 0.9–37.6 (14.6)

Contralateral lung (%) D5%_CL 0–2.4 (0.1)

0

Contralateral breast (%) D2.5%_CB 0–7.3 (0.7)

0–5.9 (0.2)

Ipsilateral lung (%) 1.8–18.9 V20_IL 3.7–23.1 V10_IL 7.4–29.8 V5_IL 1.5–16.2 D20%_IL D10%_IL 3.6–55.1 9.8–79.9 D5%_IL

(6.0) (9.6) (15.9) (4.0) (11.2) (26.1)

(0.4) (0.8) (1.2)

TOMO TOMO TOMO TOMO TOMO TOMO

vs.3D-CRT vs.3D-CRT vs.3D-CRT vs.3D-CRT vs.3D-CRT vs.3D-CRT

0–0.2 (0) 0–0.3 (0) 0–0.5 (0) 0 0

TOMO TOMO TOMO TOMO TOMO

vs. vs. vs. vs. vs.

0

0

3D-CRT vs. IMRT (0.960), IMRT vs. TOMO (1.000), TOMO vs. PBT (1.000)

0–0.02 (0)

0

3D-CRT vs. IMRT (0.458), IMRT vs. TOMO (1.000), TOMO vs. PBT (1.000)

(0.3)

(<0.001), 3D-CRT vs. IMRT (0.003), IMRT vs. PBT (0.398) (<0.001), 3D-CRT vs. IMRT (0.025), IMRT vs. PBT (0.118) (<0.001), 3D-CRT vs. IMRT (<0.001), IMRT vs. PBT (0.002) (<0.001), 3D-CRT vs. IMRT (0.014), IMRT vs. PBT (0.055) (<0.001), 3D-CRT vs. IMRT (0.001), IMRT vs. PBT (0.024) (0.840), 3D-CRT vs. IMRT (<0.001), IMRT vs. PBT (0.001)

3D-CRT (<0.001), 3D-CRT vs. IMRT (1.000), IMRT IMRT (<0.001), IMRT vs. 3D-CRT (1.000), 3D-CRT 3D-CRT (<0.001), 3D-CRT vs. IMRT (1.000), IMRT 3D-CRT (<0.001), 3D-CRT vs. IMRT (1.000), IMRT IMRT (<0.001), IMRT vs. 3D-CRT (1.000), 3D-CRT

vs. vs. vs. vs. vs.

PBT PBT PBT PBT PBT

(1.000) (1.000) (0.911) (1.000) (0.068)

Abbreviations: PTV, planning target volume; Vx_IL,Vx_H, the percentage ipsilateral lung and heart volume receiving more than x% of prescribed dose, respectively, Dy%_IL, Dy%_H, Dy%_CL, Dy%_CB, the percentage of prescribed dose delivered to y% of ipsilateral lung, heart, contralateral lung and contralateral breast volume, respectively. a By Bonferroni post hoc analysis with statistical significance defined as p < 0.008.

S.H. Moon et al. / Radiotherapy and Oncology 90 (2009) 66–73

71

Fig. 3. Dose–volume histogram (DVH) data for PTV (a), the ipsilateral non-PTV breast volume (b) and the ipsilateral lung volume (c).

evaluate the characteristics of each modality when applied to APBI. Reports on treatment toxicities related to APBI are available predominantly for brachytherapy experiences, and most of the reported toxicities have been confined to the breast or overlying skin [10]. These complications are likely to correlate with the total dose delivered, fraction size, level of inhomogeneity, and subsequent chemotherapy, finally resulting in esthetic problems [10]. However, toxicity profiles in EB-APBI are still largely unknown [21,31,32]. Therefore, it is important to maintain quality of treatment planning to reduce the potential toxicities. Determining if delineation of the lumpectomy cavity is adequate or not is a critical issue in EB-APBI, in that it determines oncologic outcomes; one of the important aims of APBI is to reduce the amount of unnecessary exposure to the non-targeted area. The issue of how to define the volume of the lumpectomy cavity remains controversial, especially in EB-APBI. Some investigators suggest using surgical clips, but others prefer to use contrast injection instead of seroma or ultrasonography [17,24]. In this study, we defined the volume of the lumpectomy cavity by the boundaries of the surgical clips. The mean and median volume measurements of the lumpectomy cavity were 17.8 and 18.4 mL (range: 7.6– 39.8 mL), respectively, values smaller than those of other studies [14,19]. This may be explained by a combination of factors, such as interobserver variation in target delineation, degree of seroma absorption, and interval between surgery and RT initiation. We thought that variability in the volume of the lumpectomy cavity itself definitely would not influence the interpretation of our results because we compared treatment plans for the four techniques in a single person. Our definition of PTV was somewhat different from that used in most of the APBI studies, in that we did not use the universal expansion of the lumpectomy or seroma cavity. Ideally, a full tumor excision by a skilled breast surgeon should leave minimal, but not close safety margins of equal distance in all directions. In general, actual safety margins reported from microscopic pathological examination of surgical specimens are not even. Therefore, it was our rationale to define the PTV by the lumpectomy cavity with different margins in each direction, according to the varying safety margin status of each lumpectomy cavity. In addition, universal expansion of the lumpectomy cavity sometimes results in a PTV too large to be accommodated in patients with small breasts. The average total ipsilateral breast volume in our study was smaller than those of other studies with the same breast definition [13,14,19,33]. This disparity may be attributable to racial differences between patients in Western and Eastern countries; patients in the current study were not selected based on breast size. While our average PTV was also smaller, and might be the result of the unique PTV definition used in our study, the average PTV to VIB ratio was comparable to that observed in most of the other studies [13–15,19,33].

According to NSABP B-39/RTOG 0413 guidelines, a three to five non-coplanar beam arrangement utilizing high-energy photons was recommended in 3D-CRT-based APBI [24,33]. In an analysis by Kozak et al. comparing 3D-CRT-based APBI techniques, both the multiple non-coplanar photon field technique and the threefield mixed-modality technique provided the effective normal tissue sparing and adequate PTV coverage [33]. Instead of using a mixed-modality technique, which is the combination of photon and electron fields, we used a four-field non-coplanar beam arrangement for 3D-CRT and IMRT. Above all, the four-field noncoplanar beam arrangement was convenient and efficient for effective dosimetric comparison of 3D-CRT and IMRT APBI plans under similar conditions. In the series of Kozak et al. comparing 3D-CRT and PBT APBI plans, a four-field non-coplanar beam arrangement was adopted in more than 80% of the patients with 3D-CRT plans [18]. In addition, the current study considered potential hot spots and the possibility that the electron beam would provide inadequate coverage in deep-seated lumpectomy cavities. The dosimetric results of 3D-CRT of our study were comparable to those of the previously published studies [21,33]. For example, V50_IB-NPTV of published data are 41% by Kozak et al. [20] and 43% by Baglan et al. [21], which is similar to 40.9% of our study. The ipsilateral lung dose in our study was also less than that of Baglan et al. [21]. The ipsilateral lung sparing in Kozak et al. [20] is superior to the current study, but this could be partly attributed to the mixed-modality approach in some patients. The results of IMRT dosimetry were excellent in our study. Intensity-modulated radiotherapy for APBI has recently been investigated [13,34]. Leonard et al. prospectively explored the feasibility of IMRT-based APBI, and reported that the IMRT dose delivery was acceptable [13]. After short-term follow-up of 10 months, clinical outcomes, including early cosmetic results, were excellent. Compared with historic controls for 3D-CRT, IMRT provided a similar dose delivery to the target while reducing the volume of the normal breast in the various levels of isodose lines. It usually takes less than one hour for 3D-CRT planning and less than three hours for IMRT, TOMO, and PBT planning. Although the beam-on time per patient is dependent on the dose rate, PD and the modality of the treatment planning, it was estimated less than 5 min in conventional cases for all modalities with the exception of TOMO. The beam-on time in TOMO seems to take up to 10–15 min depending on the cases. Like the beam-on time, the time for patient positioning varies depending on the modality of the treatment planning. In general, it takes 10–20 min in TOMO and PBT which is little longer than 3D-CRT or IMRT. Immobilization and geometric uncertainty are also important issues for APBI. It was reported that the average positional difference between normal inhalation and exhalation was 0.6–0.9 cm [21,35]. Based on the results of Baglan et al., a 5 mm CTV-to PTV margin was chosen to account for organ motion [21]. Instead of

72

Dosimetric comparison of four EB-APBI techniques

performing an individual analysis of organ motion study, we also thought that 5 mm margin for organ motion would be sufficient. Although our definition of PTV is different from other studies, the safety margin for organ motion and additional margin for set-up error are also considered in it. On the other hand, PBT is slightly more dependent upon organ motion because of the range uncertainty, which is different from 3D-CRT, IMRT, and TOMO. Similar to series of proton-based APBI [18–20], we do not use any kind of immobilization tool yet. A future study regarding range uncertainly of PBT for APBI is needed. Recently, several studies have been published related to APBI on TOMO [14,16,17]. Most of these reports have suggested that TOMO offers excellent results in terms of PTV coverage, conformity, and homogeneity, even with the normal tissue constraints for lung and heart. In this study, TOMO showed excellent dosimetric results for target coverage, though the dose exposure to the ipsilateral lung and heart was not ideal. The lower dose to the ipsilateral lung was considerable in our study, despite the efforts to maintain a minimal dose to normal tissues with stepwise optimization techniques from complete block to directional block, which were reported by Patel et al. [16]. In a recent comparative APBI planning study by Langen et al., prone position TOMO was found to effectively decrease the integral dose due to displacement of the target away from critical structures [17]. However, it required a specially constructed positioning system, limiting its widespread adoption; it is also questionable whether it can be effectively applied to patients with small breast volume or with challenging anatomy. Unfortunately, there is yet no clear evidence concerning the detrimental effects of higher integral doses to the lung in breast cancer patients treated with RT. Dosimetric comparison can be justified when it is based on the assumption that all plans are pursuing the same planning goal, while satisfying minimal requirements with adequate optimization. In this study, our planning goal of total PTV coverage corresponded well to this assumption. Therefore, the higher integral dose to the lung and heart in TOMO might have been reduced to some extent if the planning goal had been applied more flexibly. The Bragg peak, the unique dosimetric feature of PBT, made PBT one of the attractive non-invasive APBI techniques. Experiences with PBT for EB-APBI in patients with breast cancer are limited. The use of PBT was based on the first clinical report suggesting the superiority of PBT dose distribution over 3D-CRT using photons and electrons [18]. When compared with conventional 3D-CRT plans, PBT doses delivered to normal structures such as the ipsilateral and contralateral lungs, heart, and non-target breast tissues were significantly lower, yet did not compromise PTV coverage. In this study, PBT was found to be excellent in all comparative dosimetric parameters, as well as in factors related to sparing normal tissues. In our analysis, APBI using PBT was dosimetrically feasible to be applied to a prospective clinical trial, an observation consistent with the findings of Kozak et al. [18]. However, information on long-term oncologic and cosmetic outcomes is currently not available. In our study, treatment comparisons were made with three different treatment planning systems. Due to the complexity of radiation interactions with human tissues and due to the practical need for rapid calculation times, dose calculation algorithms selected by radiotherapy planning (RTP) systems have various inherent limitations with the corresponding uncertainties. In addition to this, dosimetric results may be dependent on the optimization algorithm, which may be different based on the RTP system. These facts suggest that it is possible that there may be slight differences in results due to the dose calculation and optimization algorithms used in each of the unique planning systems. In summary, all four EB-APBI techniques were found to have acceptable PTV coverage in our study. In this analysis, PBT had

the greatest capacity to spare normal breast tissue without increasing the dose delivered to the lung and heart. IMRT also showed a better ability to spare normal breast tissue compared with 3D-CRT, as was suggested by other investigators [13,15]. However, in TOMO, high conformity to the PTV and effective breast tissue sparing were achieved at the expense of considerable dose exposure to the lung and heart. Acknowledgement This study was supported by a research Grant (0710720) from the National Cancer Center, Korea. References [1] Fisher B, Dignam J, Wolmark N, et al. Lumpectomy and radiation therapy for the treatment of intraductal breast cancer: findings from National Surgical Adjuvant Breast and Bowel Project B-17. J Clin Oncol 1998;16:441–52. [2] Julien JP, Bijker N, Fentiman IS, et al. Radiotherapy in breast-conserving treatment for ductal carcinoma in situ: first results of the EORTC randomised phase III trial 10853. EORTC Breast Cancer Cooperative Group and EORTC Radiotherapy Group. Lancet 2000;355:528–33. [3] Fisher B, Jeong JH, Anderson S, Bryant J, Fisher ER, Wolmark N. Twenty-fiveyear follow-up of a randomized trial comparing radical mastectomy, total mastectomy, and total mastectomy followed by irradiation. N Engl J Med 2002;347:567–75. [4] Veronesi U, Cascinelli N, Mariani L, et al. Twenty-year follow-up of a randomized study comparing breast-conserving surgery with radical mastectomy for early breast cancer. N Engl J Med 2002;347:1227–32. [5] Clark RM, Whelan T, Levine M, et al. Randomized clinical trial of breast irradiation following lumpectomy and axillary dissection for node-negative breast cancer: an update. Ontario Clinical Oncology Group. J Natl Cancer Inst 1996;88:1659–64. [6] Liljegren G, Holmberg L, Bergh J, et al. 10-Year results after sector resection with or without postoperative radiotherapy for stage I breast cancer: a randomized trial. J Clin Oncol 1999;17:2326–33. [7] Veronesi U, Marubini E, Mariani L, et al. Radiotherapy after breast-conserving surgery in small breast carcinoma: long-term results of a randomized trial. Ann Oncol 2001;12:997–1003. [8] Aristei C, Tarducci R, Palumbo I, et al. Computed tomography for excision cavity localization and 3D-treatment planning in partial breast irradiation with high-dose-rate interstitial brachytherapy. Radiother Oncol 2009;90: 43–7. [9] Major T, Fröhlich G, Lövey K, Fodör J, Polgár C. Dosimetric experience with accelerated partial breast irradiation using image-guided interstitial brachytherapy. Radiother Oncol 2009;90:48–55. [10] Arthur DW, Vicini FA. Accelerated partial breast irradiation as a part of breast conservation therapy. J Clin Oncol 2005;23:1726–35. [11] Weed DW, Edmundson GK, Vicini FA, Chen PY, Martinez AA. Accelerated partial breast irradiation: a dosimetric comparison of three different techniques. Brachytherapy 2005;4:121–9. [12] Khan AJ, Kirk MC, Mehta PS, et al. A dosimetric comparison of threedimensional conformal, intensity-modulated radiation therapy, and MammoSite partial-breast irradiation. Brachytherapy 2006;5:183–8. [13] Leonard C, Carter D, Kercher J, et al. Prospective trial of accelerated partial breast intensity-modulated radiotherapy. Int J Radiat Oncol Biol Phys 2007;67:1291–8. [14] Oliver M, Chen J, Wong E, Van Dyk J, Perera F. A treatment planning study comparing whole breast radiation therapy against conformal, IMRT and tomotherapy for accelerated partial breast irradiation. Radiother Oncol 2007;82:317–23. [15] Rusthoven KE, Carter DL, Howell K, et al. Accelerated partial-breast intensitymodulated radiotherapy results in improved dose distribution when compared with three-dimensional treatment-planning techniques. Int J Radiat Oncol Biol Phys 2008;70:296–302. [16] Patel RR, Becker SJ, Das RK Mackie TR. A dosimetric comparison of accelerated partial breast irradiation techniques: multicatheter interstitial brachytherapy, three-dimensional conformal radiotherapy, and supine versus prone helical tomotherapy. Int J Radiat Oncol Biol Phys 2007;68:935–42. [17] Langen KM, Buchholz DJ, Burch DR, et al. Investigation of accelerated partial breast patient alignment and treatment with helical tomotherapy unit. Int J Radiat Oncol Biol Phys 2008;70:1272–80. [18] Kozak KR, Smith BL, Adams J, et al. Accelerated partial-breast irradiation using proton beams: initial clinical experience. Int J Radiat Oncol Biol Phys 2006;66:691–8. [19] Taghian AG, Kozak KR, Katz A, et al. Accelerated partial breast irradiation using proton beams: Initial dosimetric experience. Int J Radiat Oncol Biol Phys 2006;65:1404–10. [20] Kozak KR, Katz A, Adams J, et al. Dosimetric comparison of proton and photon three-dimensional, conformal, external beam accelerated partial breast irradiation techniques. Int J Radiat Oncol Biol Phys 2006;65:1572–8.

S.H. Moon et al. / Radiotherapy and Oncology 90 (2009) 66–73 [21] Baglan KL, Sharpe MB, Jaffray D, et al. Accelerated partial breast irradiation using 3D conformal radiation therapy (3D-CRT). Int J Radiat Oncol Biol Phys 2003;55:302–11. [22] Urie MM, Goitein M, Doppke K, et al. The role of uncertainty analysis in treatment planning. Int J Radiat Oncol Biol Phys 1991;21:91–107. [23] Radiation Therapy Oncology Group. NSABP B-39/RTOG 0413 protocol. Available from: www.RTOG.org. Accessed August 2006. [24] Vicini F, Winter K, Straube W, et al. A phase I/II trial to evaluate threedimensional conformal radiation therapy confined to the region of the lumpectomy cavity for stage I/II breast carcinoma: initial report of feasibility and reproducibility of Radiation Therapy Oncology Group (RTOG) Study 0319. Int J Radiat Oncol Biol Phys 2005;63:1531–7. [25] Wu Q, Mohan R, Morris M, Lauve A, Schmidt-Ullrich R. Simultaneous integrated boost intensity-modulated radiotherapy for locally advanced head-and-neck squamous cell carcinomas. I: dosimetric results. Int J Radiat Oncol Biol Phys 2003;56:573–85. [26] Leung LH, Chua DT, Wu PM. A new tool for dose conformity evaluation of radiosurgery treatment plans. Int J Radiat Oncol Biol Phys 1999;45: 233–41. [27] Lomax NJ, Scheib SG. Quantifying the degree of conformity in radiosurgery treatment planning. Int J Radiat Oncol Biol Phys 2003;55:1409–19. [28] Paddick I. A simple scoring ratio to index the conformity of radiosurgical treatment plans. Technical note.. J Neurosurg 2000;93(Suppl. 3):219–22.

73

[29] Shaw E, Kline R, Gillin M, et al. Radiation Therapy Oncology Group: radiosurgery quality assurance guidelines. Int J Radiat Oncol Biol Phys 1993;27:1231–9. [30] ICRU. Prescribing, recording, and reporting photon beam therapy (supplement to ICRU Report 50). Bethestda, Maryland: International Commission of Radiation Units and Measurements. 1999. [31] Formenti SC, Truong MT, Goldberg JD, et al. Prone accelerated partial breast irradiation after breast-conserving surgery: preliminary clinical results and dose–volume histogram analysis. Int J Radiat Oncol Biol Phys 2004;60:493–504. [32] Magee B, Swindell R, Harris M, Banerjee SS. Prognostic factors for breast recurrence after conservative breast surgery and radiotherapy: results from a randomised trial. Radiother Oncol 1996;39:223–7. [33] Kozak KR, Doppke KP, Katz A, Taghian AG. Dosimetric comparison of two different three-dimensional conformal external beam accelerated partial breast irradiation techniques. Int J Radiat Oncol Biol Phys 2006;65:340–6. [34] Coles CE, Moody AM, Wilson CB, Burnet NG. Reduction of radiotherapyinduced late complications in early breast cancer: the role of intensitymodulated radiation therapy and partial breast irradiation. Part II— radiotherapy strategies to reduce radiation-induced late effects. Clin Oncol (Royal College of Radiologists (Great Britain)) 2005;17:98–110. [35] Frazier RC, Vicini FA, Sharpe MB, et al. Impact of breathing motion on whole breast radiotherapy: a dosimetric analysis using active breathing control. Int J Radiat Oncol Biol Phys 2004;58:1041–7.