Abstract
Background/Aim: To compare heart, left ventricle (LV) and coronary artery dose-sparing with three-dimensional conformal radiotherapy (3D-CRT) vs. helical tomotherapy (HT) in left-sided breast cancer (BC). Patients and Methods: 3D-CRT and HT treatments were planned for 20 patients (pts). Computed tomography (CT) scans without and with intravenous contrast (ic) were performed and co-registered. Left breast and organs at risk (OARs) were contoured. Dose-volume histograms (DVHs) for 3D-CRT and HT treatment plans were evaluated in terms of planning target volume for evaluation (PTVeval) coverage and dose to the OARs. Results: HT provided the best target coverage and significantly reduced D2% and mean dose to the left anterior descending artery (LADA) and to the LADA-planning organ at risk volume (PRV), D2%, V5 and mean dose to the LV and D2% and V25 to the heart. As expected, due to the rotational delivery, the dose to all other coronary arteries and their PRV, contralateral breast and lungs was higher with HT. Conclusion: In left-sided BC, HT provided the best target coverage and significantly reduced LV and LADA doses. Moreover D2% and V25 to the heart were significantly reduced. Further studies are needed to correlate dosimetric findings with in-depth cardiac monitoring.
- Left-sided breast cancer
- three-dimensional conformal radiotherapy
- helical tomotherapy
- heart
- left ventricle
- coronary arteries
Radiation therapy (RT)-induced cardiac damage include pericarditis, myocarditis, arrhythmias, valvular dysfunction, congestive heart failure, ischemic heart disease, coronary artery stenosis, and arteriosclerosis (1, 2). In left breast cancer (BC) patients (pts), especially in those with unfavorable thoracic geometry, tangential field three-dimensional conformal RT (3DCRT), which is the standard technique for whole breast irradiation (WBI), risks delivering a very high dose to the left ventricle (LV) and left anterior descending artery (LADA) (2, 3). Strategies to minimize cardiac irradiation range from use of diverse positions, whether prone or lateral decubitus, to breath holding techniques (4, 5). Irradiation techniques include intensity modulated radiation therapy (IMRT), volumetric arc therapy (VMAT), helical tomotherapy (HT), proton beam RT and partial breast irradiation, which is a valid alternative to WBI in carefully selected pts (6, 7). IMRT, VMAT, and HT are associated with significant heart sparing and high dose homogeneity in the target, even in left-sided BC pts with unfavorable cardiac anatomy (8-14). In the past, at our Institution, breath holding (BH) techniques were not available, therefore we performed a dosimetric study to compare 3DCRT vs. HT treatment plans in left-sided BC RT, aiming to identify which technique most reduced the dose to the heart and its substructures.
Patients and Methods
Twenty pts with early stage left BC undergoing WBI were selected for this dosimetric study, which was conducted in accordance with the Helsinki Declaration of 1975, as revised in 2000. Written informed consent was obtained from all pts. The prescribed dose was 50 Gy in 25 fractions (5 fractions a week). Pts were positioned supine on a flat breast board immobilization device, with arms raised above the head and instructed to breathe gently during computed tomography (CT) scanning. Radiopaque wires identified palpable breast tissue. CT scans (Aquilion S16, Toshiba America Medical Systems, Inc., Tustin, CA, USA) with and without intravenous contrast (ic) were carried out with 3 mm slice thickness and step and acquired from the mandibular angle to the diaphragm. CT-images were transferred to the treatment planning system (TPS, Pinnacle3 Philips V 9.10, Philips Radiation Oncology Systems, Fitchburg, WI, USA) and co-registered. After co-registration, maximum shift of the heart (maxSHeart) as contoured on CT scans with (icHeart) and without ic (Heart) was measured in the anterior, posterior, left, right, cranial, and caudal directions for each patient. The left breast and organs at risk (OARs) were contoured by one radiation oncologist (RO) expert in BC RT and trained in coronary artery delineation. The clinical target volume (CTV) encompassed only the left breast and was contoured using the Italian guideline atlas (15). On CT images without ic, the CTV, lungs, right breast, and heart were contoured. The heart was contoured alone (Heart) and with the blood vessels (pulmonary artery, ascending aorta, superior and inferior vena cava-Heart+vessels). On CT images with ic the heart alone (ic Heart) and with the blood vessels (ic Heart+vessels), LV, left main coronary artery (LMCA), LADA, left circumflex (LCX) and right coronary artery (RCA) were contoured. The heart and its structures were delineated following the atlas proposed by Feng et al. (16). In accordance with our internal guidelines, the planning target volume (PTV) for 3DCRT planning was obtained by expanding the CTV 10 mm in all directions except the medial, where expansion was 5 mm. To obtain the PTV for HT planning, the CTV was expanded 5 mm in all directions. To create planning target volume for evaluations (PTVevals), all PTVs were drawn back to 5 mm under the skin surface. In accordance with literature data (17), a 5 mm expansion in all directions was added to LADA, LMCA, LCX, and RCA to define the corresponding planning organ at risk volume (PRV). In our series, 5 mm correspond to the average of the mean maxSHeart values in all directions calculated after co-registration of CT-images with and without ic (data not shown). CT without ic was used for dosimetric calculations. 3D-CRT treatments were planned on an isocentric technique. Beam energy was 6 MV. Two tangential, opposed, wedged half-beams were used. Gantry and collimator angles were adjusted, using the Pinnacle3 TPS beam eye view, minimizing heart and lung irradiation while maximizing target volume coverage. Dynamic wedges in the opposing tangential beams were used to improve target dose distribution. To avoid hot spots, dose distribution was optimized by adding extra tangent fields with a manually generated multi-leaf collimator configuration (field-in-field). In the PTVeval, 95-107% dose distribution was achieved, according to ICRU criteria. Dose volume specifications were: for the ipsilateral lung V30 Gy <20% and V20 Gy <30%, for the heart V25 Gy <10%. HT plans were generated and optimized using a Tomotherapy HD treatment planning station (version 5.1.0.4, Accuray, Sunnyvale, CA, USA). A field width of 2.5 cm, a pitch of 0.287, and a maximum modulation factor of 2.7 were used for all plans. To spare the heart, right breast, and lungs a L-shaped region of interest was created with directional blocking. Plans were optimized on PTVeval. Target and OARs constraints used for inverse planning optimization are reported in Table I. Dose-volume histograms (DVH) points and penalties were adjusted throughout the optimization process to best meet OARs dose constraints without penalizing PTVeval coverage. DVHs for 3DCRT and HT plans were evaluated in terms of PTVeval coverage and dose to the OARs. Dose volume specifications were: for the PTVeval D95%, D50%, D2% and mean dose (Dmean), for the heart D2%, V5 Gy, V25 Gy and Dmean, for the LV D2%, V5 Gy and Dmean, for all coronary arteries and their PRV and for contralateral breast D2% and Dmean, for the ipsilateral lung V20 Gy and V5 Gy and for the contralateral lung V15 Gy and V5 Gy. A power analysis using G-Power (18) determined a sample size of 20 was required. The α for Wilcoxon signed-rank test (matched pairs) was set at 0.05 to detect a significant difference of a large effect size (d=0.8) (19) with a power of 0.90. Mean, SD, median, and range of all dosimetric variables for 3DCRT and HT plans were calculated and compared. Tests were two tailed and p-values <0.05 were regarded as significant. All analyses were performed using IBM-SPSS® version 22.0 (IBM Corp., Armonk, NY, USA).
Target and OAR objectives used for inverse planning optimization.
Results
Left breast and OARs volumes are reported in Table II. Table III shows the dosimetric parameters for 3DCRT and HT plans. Target coverage resulted better with HT: PTVeval D95% and Dmean were significantly higher, D50% was higher, but this difference was not significant (p=0.912), D2% was slightly lower. Regarding the heart (contoured with or without vessels) HT, compared with 3DCRT, significantly decreased D2% (by about 60%) and reduced V25 to 0, while V5 resulted significantly increased; however, Dmean was not significantly different. HT, compared with 3DCRT, significantly lowered LV D2%, V5, and Dmean by about 78%, 43%, and 36%, respectively; LADA D2% and Dmean were significantly reduced (by about 85% and 83%, respectively), as well as PRV-LADA D2% and Dmean, which were lowered by about 71% and 75%, respectively. With 3DCRT, D2% and Dmean of MCA, LCX, RCA and their PRV were significantly lower; however, with HT plans Dmean of MCA, LCX, RCA and their PRV never exceeded 4 Gy. All dosimetric parameters for contralateral breast and lungs were significantly higher with HT planning, but all OAR constraints accepted for HT plans were respected. Figure 1 shows that HT planning resulted in a significant PRV-LADA dose reduction, compared with other coronary arteries.
Mean (±standard deviation) and median volumes (with range) in cm3 of left breast and organs at risk (OARs).
Dosimetric parameters (mean±SD, median, range) for three-dimensional conformal radiotherapy (3DCRT) and helical tomotherapy (HT) plans.
Average cumulative DVH for PRV-LADA (A), PRV-LCX (B), PRV-LCMA (C), PRV-RCA (D) comparing 3D-CRT and HT. DVH: Dose-volume histograms; PRV: planning organ at risk volume; LADA: left anterior descending artery; LCX: left circumflex artery; LMCA: left main coronary artery; RCA: right coronary artery.
Discussion
In our series HT planning yielded the best PTVeval coverage and a better dosimetric profile for the heart, LV and PRV-LADA. Doses to PRV-LMCA, PRV-LCX, and PRV-RCA, although significantly higher with HT than with 3DCRT, are probably not clinically relevant, since absolute values are quite small. As expected with rotational delivery, doses to contralateral breast and lungs were higher with HT. In our series, HT planning reduced LADA and PRV-LADA D2% and Dmean by more than 70%, LV V5 by about 40%, and all dose-volume dosimetric constraints for OARs were respected. Concurring with our results, which were obtained using conventional fractionation as in other studies, IMRT, VMAT, and HT resulted associated with relevant heart and heart substructures dose sparing in left-sided BC pts (8-12). To our knowledge, our dosimetric study is the first to compare 3DCRT and HT dose delivery to the heart, LV, and all coronary arteries in pts with left-sided BC undergoing WBI. Lohr et al. (8) showed that IMRT reduced the maximal dose to the LV in patients with unfavorable thoracic geometry. Fan et al. (11), comparing prone and supine positions in IMRT, evaluated dosimetric parameters for the heart, LV, and PRV-LADA and concluded that the prone position provided better OAR sparing. Unlike the present study, the PRV-LADA was defined as LADA contoured on CT coronary angiography images, with 1 cm margins in all directions. In pts with unfavorable cardiac anatomy, Coon et al. (9) showed that IMRT and HT significantly reduced doses to the heart and LV. Jin et al. (10) compared tangential wedge-based fields, field-in-field (FIF) technique, tangential inverse planning IMRT (T-IMRT), multi-field IMRT (M-IMRT) and VMAT in pts with small breast size, showing that T-IMRT reduced RT dose to OARs (including heart and coronary artery region) with reasonable target homogeneity. Haciislamoglu et al. (12) compared 3DCRT with forward-planned IMRT, inverse-planned IMRT, VMAT, and HT with HT providing, as in the present study, optimal target coverage, while delivering the lowest maximum doses to the heart and LADA (contoured on a CT scan without ic). As RT-related cardiotoxicity could potentially counteract the clinical benefit achieved with adjuvant RT, ROs should focus on cardiac dose sparing. Darby et al. showed that rates of major coronary events increased linearly with the heart Dmean by 7.4% per Gy, with no apparent threshold, whereas LADA Dmean did not improve prediction of major coronary events (2). However, in their study, information about irradiated structures within the heart for each patient were lacking, since a CT scan of a woman with typical anatomy was used to reconstruct each RT treatment. In our opinion, LV and LADA sparing is crucial, since LV-V5 Gy was highly associated with the excess risk of RT-induced acute coronary events in a large cohort of BC pts treated with 3DCRT (20) and the heart portion perfused by LADA is one of typical sites of ischemic disease in left-sided BC treated with tangential field 3DCRT (3). Furthermore, since excessive irradiation of any portion of coronary arteries should be avoided, as they have to be considered a “series subunit” organ (3), coronary artery contouring is a challenge. Indeed, despite the efforts to define them better (16, 21), difficulties in their definition remain. Therefore, in our opinion, the main strength of the present study is the accurate heart, LV, and coronary arteries delineation, obtained using CT scan with ic and performed by the same RO trained in heart substructures delineation. Regarding dosimetric results, major strengths of our study are the inclusion of non-selected left-sided BC pts, thus allowing data extrapolation for any pts (even without unfavourable anatomy), and the use of smaller PTV margins for HT planning compared to 3DCRT, thus allowing plan comparisons closer to the “real life”. Finally, regarding RT-related cardiotoxicity, diagnosis of early cardiac damage to ensure prompt therapy is crucial, particularly in pts with left-sided BC and pre-existing cardiac risk factors. These pts should be carefully monitored by means of cardiac biomarkers (22) and/or cardiac imaging, such as coronary CT angiography (23), cardiac magnetic resonance imaging (24), and echocardiography (23, 25).
In conclusion, in left-sided BC pts, HT provided dose sparing to the LV and LADA with an expected increase in low doses to all the other OARs. In our opinion, HT should be reserved, if BH techniques are not available, for selected subgroups of pts, such as left-sided BC pts with cardiovascular risk factors and/or a history of coronary artery disease (CAD) and myocardial infarction (MI), especially if they also receive anthracycline-based chemotherapy and anti-HER2 (2, 26). Prospective studies are particularly needed to correlate heart, LV, and coronary artery dosimetric findings with in-depth cardiac monitoring.
Footnotes
↵# Affiliation at the time of the study.
Authors’ Contributions
IP, MM, and CA conceptualized and designed this study. MS, VB, MDB, SN, CF, and EP collected and analyzed data. IP, MM, MS, and VB participated in the interpretation of results. IP and MM drafted the article, CA revised the article. All Authors read and approved the article.
Funding
This work was partially supported by the Italian Ministry of Education, University and Research (MIUR) within the Individual Annual Funding for Fundamental Research “FFABR” 2018 (I. Palumbo).
Conflicts of Interest
The Authors have no conflicts of interest to declare in relation to this study.
- Received July 7, 2023.
- Revision received August 21, 2023.
- Accepted August 25, 2023.
- Copyright © 2023, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).