Radiation dose distribution in functional heart regions from tangential breast cancer radiotherapy

Radiotherapy and Oncology xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Radiotherapy and Oncology journal homepage: www.thegreenjournal.com

Original article

Radiation dose distribution in functional heart regions from tangential breast cancer radiotherapy Daniel Wollschläger a,⇑, Heiko Karle b, Marcus Stockinger b, Detlef Bartkowiak c, Sandra Bührdel c, Hiltrud Merzenich a, Thomas Wiegel c, Maria Blettner a, Heinz Schmidberger b a Institute for Medical Biostatistics, Epidemiology and Informatics; b Department of Radiation Oncology, University Medical Center Mainz; and c Department of Radiation Oncology, University Hospital Ulm, Germany

a r t i c l e

i n f o

Article history: Received 2 October 2015 Received in revised form 8 January 2016 Accepted 12 January 2016 Available online xxxx Keywords: Breast cancer Heart dose Dosimetry Dose–volume histogram

a b s t r a c t Background and purpose: To analyze the distribution of individually-determined radiation dose to the heart and its functional sub-structures after radiotherapy in breast cancer patients treated in Germany during 1998–2008. Material and methods: We obtained electronic treatment planning records for 769 female breast cancer patients treated with megavoltage tangential field radiotherapy. All dose distributions were recalculated using Eclipse with the anisotropic analytical algorithm (AAA) for photon fields, and the electron Monte Carlo algorithm for electron boost fields. Based on individual dose volume histograms for the complete heart and several functional sub-structures, we estimated various dose measures in patient groups. Results: Mean heart dose spanned a range of 0.9–19.1 Gy for left-sided radiotherapy and 0.3–11.6 Gy for right-sided radiotherapy. Average (median) mean heart dose was 4.6 Gy (3.7 Gy) for left-sided radiotherapy, and 1.7 Gy (1.4 Gy) for right-sided RT. With left-sided radiotherapy, 66% of the patients had 2 cm3 of the complete heart exposed to at least 40 Gy. Younger age, higher body mass index, tumor location in a medial quadrant, and presence of a parasternal field were also associated with higher heart dose. Conclusion: Tumor location and treatment choices influence cardiac dose with complex interactions. There is considerable variability in heart dose, with dose metrics of different cardiac sub-structures showing different patterns in their dependency on external influences. Dose–response analysis of late cardiac effects after radiotherapy requires detailed individual dosimetry. Ó 2016 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology xxx (2016) xxx–xxx

For most women with breast cancer who undergo primary surgery, adjuvant radiotherapy (RT) helps reduce mortality and disease recurrence. Since the 1980 s, external beam RT began to standardize on tangential megavoltage photon fields and electron or photon boosts to the primary tumor region [1]. For breast cancer, this approach still dominates current RT among newer techniques like IMRT. Three-dimensional imaging methods and advanced treatment planning software algorithms helped reduce unwanted irradiation of non-tumor tissue compared to regimens used in the 1950s– 1990s [2–5]. However, the heart often remains partially exposed, mainly depending on tumor laterality and individual anatomy. For women treated in the 1950s–1990s, radiation exposure to the heart was shown to be associated with cardiac diseases and ⇑ Corresponding author at: Institute for Medical Biostatistics, Epidemiology and Informatics, University Medical Center Mainz, Obere Zahlbacher Str. 69, 55131 Mainz, Germany. E-mail address: [email protected] (D. Wollschläger).

excess mortality [6–8]. Most evidence comes from studies showing that women treated with left-sided RT have higher cardiac mortality than women with right-sided RT [8–10]. A recent study with quantitative information on heart dose found a linear dose– response relationship for the radiation-associated excess risk of major cardiac events [11]. As survival rates after breast cancer continue to improve [12– 14], long-term RT side effects become important for treatment planning. Detailed individual dosimetry information for the heart is required to better understand cardiac damage from radiation exposure, and to estimate the risk for late cardiac effects. Establishing a dose–response relationship from linking the dosimetric information to clinical endpoints should rely on a wide dose range of heart exposures. Since radiation exposure of different cardiac sub-structures might be associated with characteristic late effects, it is also desirable to characterize the received dose in different functional parts of the heart. The dose distribution in the heart after breast RT in the 1950s– 1990s has been carefully characterized [4,5,15], mostly using

http://dx.doi.org/10.1016/j.radonc.2016.01.020 0167-8140/Ó 2016 Elsevier Ireland Ltd. All rights reserved.

Please cite this article in press as: Wollschläger D et al. Radiation dose distribution in functional heart regions from tangential breast cancer radiotherapy. Radiother Oncol (2016), http://dx.doi.org/10.1016/j.radonc.2016.01.020

2

Heart dose in breast cancer radiotherapy

virtual simulation based on few average patient anatomies. So far, cardiac radiation exposure for current breast RT techniques based on individual anatomy and treatment planning has been analyzed for only a small sample of patients with predominantly left-sided tumors [16]. We here describe the cardiac dose distribution after tangential breast RT including regional lymph node irradiation and boost. The data are based on individual anatomy and treatment planning for a large sample of women who underwent breast RT from 1998–2008 in Germany. The primary goal is to use individual dose-volume histograms (DVHs) to investigate how anatomical characteristics and RT details affect cardiac dose. In addition to the complete heart, we aim to characterize the dose distribution in functional sub-structures as potential organs at risk for late cardiac effects. Materials and methods We randomly sampled 769 patients out of 12077 women included in the German PASSOS retrospective cohort study on late cardiac effects after treatment for breast cancer with good prognosis. Patients were first diagnosed with locoregional breast cancer from 1998–2008 at the university medical centers Mainz, Ulm, or one of 16 collaborating hospitals close to Ulm. PASSOS includes patients from the prospective BRENDA cohort [17]. Sampled patients completed their RT as prescribed either in Mainz from 2001 onward or in Ulm from 1998 onward. Henceforth, treatment centers will be labeled A (N = 395) and B (N = 374). Patients treated in the 16 collaborating hospitals were not included as their electronic treatment planning records were inaccessible. The sample was stratified for treatment center, age at diagnosis and tumor laterality. Patients with left-sided tumors (N = 486, 63%) were over-sampled in a 2:1 ratio to compensate their greater variability in dose metrics [18] (Table 1). Patients with bi-lateral tumors did not meet the inclusion criteria of the cohort. The mean (sd) age at diagnosis was 56.7 (12.7) years (A: 56.4 (12.9), B: 57.0 (12.5)).

Table 1 Frequency for radiotherapy details: Laterality, prescription dose to the planning target volume, boost to the primary tumor location, presence of at least one lymph node field.

Center A

Laterality Prescription dose

Boost

Parasternal field Center B

Laterality Prescription dose

Boost

Parasternal field

N

%

Left Right 46 Gy 50 Gy 50.4 Gy 54 Gy 56 Gy No Electron Photon No Yes

249 146 144 204 27 19 1 86 299 10 302 88

63 37 36 52 7 5 0.3 22 76 2 77 23

Left Right 50 Gy 50.4 Gy 56 Gy 60 Gy No Electron Photon Mixed No Yes

237 137 265 89 2 18 130 70 166 8 305 69

63 37 71 24 0.5 5 35 19 44 2 82 18

The study was approved by the ethics review board of the Rhineland-Palatinate chamber of physicians and the ethics review board of the university medical hospital Ulm. The requirement for individual informed consent was waived because of the anonymous analysis. Three-dimensional imaging for treatment planning came from computed tomography (CT) image slices of thickness 8 mm (center A) or 10 mm (center B) without contrast enhancement. Treatment planning software in center A was either Cadplan version 6.4.7 or Eclipse version 8 (Varian Medical Systems, Palo Alto, CA), in center B Helax TMS version 6.1b (MDS Nordion, Kanata, Canada). Original dose calculations were based on the pencil beam algorithm [19,20] in both centers. Standard planning comprised one pair of tangent fields of 6-MV photons with a total dose to the planning target volume (PTV) between 46 Gy and 60 Gy, applied in fractions of 1.8 Gy or 2 Gy (Table 1). 10-MV photons were used as supplementary field-in-field applications. Tangential field borders and angles were determined clinically by the attending physician to encompass the whole palpable breast. In mastectomy patients, the lateral field border would be the mid-axillary line and the medial field border would be the mid-line. Beam weights and wedge angles were optimized based on the dose distribution for the central axis plane. A boost to the primary tumor region was optional, as were supraclavicular and parasternal fields to reduce lymphatic dissemination. The parasternal field was sometimes omitted to reduce heart exposure. Patients were irradiated in a supine position while being immobilized on a breast board with both hands above their head. There was no use of breath gating, heart or lung shielding was not standard. Dosimetry was performed on the basis of individual electronic records from initial treatment planning with CT imaging data as well as details of the implemented RT fields. We included only patients with CT scans of sufficient quality showing the complete heart. Original treatment planning data were imported into Varian Eclipse version 8.65 or 11.031 before contouring. Data from Helax TMS were transferred via native DICOM format. Cadplan records were converted using VODCA [21]. For photon fields, dose recalculations were done with the anisotropic analytical algorithm (AAA) [22,23]. Electron boost fields were re-calculated using the electron Monte Carlo algorithm. We devised a contouring atlas with standard operating procedures for the complete heart and for six sub-structures as potential organs at risk. The goal was to robustly capture the substrate volume of a functional heart unit while ensuring reliable contouring even with non-contrast-enhanced CT scans. The sub-structures were therefore defined using a largely geometrical, rather than anatomical approach (Fig. 1):  The complete heart as in its RTOG definition for RT in breast cancer.  The aortic valve: A conical volume starting at the aortic root going 3 cm in the caudal direction.  The pulmonary valve: A spherical volume expanding uniformly around a point within the truncus pulmonalis at the level of the aortic root until it almost touches the aortic valve volume.  The heart wall including the pericardium, coronary arteries, and the myocardium without the ventricular septum: The outer layer of the complete heart below the aortic valve with a thickness of 1 cm.  The left anterior heart wall including the left main coronary artery and left anterior descending (LAD).  The right anterior heart wall including the right coronary artery and the sinoatrial node.

Please cite this article in press as: Wollschläger D et al. Radiation dose distribution in functional heart regions from tangential breast cancer radiotherapy. Radiother Oncol (2016), http://dx.doi.org/10.1016/j.radonc.2016.01.020

D. Wollschläger et al. / Radiotherapy and Oncology xxx (2016) xxx–xxx

3

Fig. 1. Simulation films of typical medio-lateral tangential beams for right-sided and left-sided postoperative whole breast irradiation. Superimposed contours of the complete heart (yellow) and functional sub-structures as defined by study protocol: Aortic valve (blue), Pulmonary valve (pink), AV node and bundle of His (green), right anterior heart wall (red).

 A central volume including the caudal aspect of the aortic valve and a plane between the inferior cava vein and the apex of the heart representing the atrioventricular node and bundle of His. For each patient and structure, a cumulative DVH was exported from Eclipse with a bin size of at most 5 cGy and imported into the statistical environment R [24,25]. Because of the heterogeneous dose distribution in the heart, we did not convert physical dose to 2 Gy fractions equivalent dose (EQD2). From the DVHs, we calculated three absorbed dose metrics each for all structures: DMEAN is the volume-weighted mean dose. D2CC is the dose received by the maximally exposed 2 cm3. V10Gy is the relative volume exposed to at least 10 Gy. Two groups are statistically compared with the Wilcoxon–Ma nn–Whitney test, more than two ordered groups with the Jonckheere–Terpstra test. Reported location differences between two groups are Hodges–Lehmann pseudo-medians. P-values are twotailed and not corrected for multiplicity. Confidence intervals (CIs) are for level 95%. Results Dose distribution Fig. 2 shows the dose distribution for each dose metric. The DMEAN distribution in the complete heart is strongly right skewed, especially for left-sided RT. For right-sided RT, the D2CC distribution has most mass around 3 Gy with a long tail. For left-sided RT, the D2CC distribution has most mass around 49 Gy and is strongly left-skewed. The V10Gy distribution for right-sided RT has almost all of its mass around 0.4 Gy while it is much more spread out and right-skewed for left-sided RT. Results for the mean, median, and standard deviation of each dose metric for the complete heart and its sub-structures structures are shown in Table 2. The average (median) DMEAN in the complete is 4.6 Gy (3.7 Gy) for left-sided RT, and 1.7 Gy (1.4 Gy) for right-sided RT. The variability in mean heart dose is considerable, both for SD (left-sided RT: 3.1 Gy, right-sided RT: 1.2 Gy), and for the range of values, especially for left-sided RT (0.9– 19.1 Gy). The effect for left- vs. right-sided RT is strongest in the left anterior myocardium where average (median) mean heart dose reaches

15.6 Gy (13.2 Gy) for left-sided RT. Likewise, RT laterality is important for the mean dose to the pulmonary valve (average (median): 6.1 Gy (3.3 Gy) left, 2.0 Gy (1.7 Gy) right) and the AV node (average (median): 2.9 Gy (2.2 Gy) left, 1.6 Gy (1.4 Gy) right). With leftsided RT, extreme exposure often approaches prescription dose. In the complete heart, 319 (66%) of patients with left-sided RT have a D2CC of at least 40 Gy. In contrast, only 8 (3%) of patients with right-sided RT have a D2CC of 40 Gy. For left-sided RT, the most extremely exposed part of the heart is the left anterior myocardium with very similar D2CC values as the complete heart. V10Gy values show that for left-sided RT, on average 8% of the complete heart volume is exposed to at least 10 Gy. For the left anterior myocardium, that proportion reaches 40%. In the pulmonary valve, 352 (72%) patients with left-sided RT have no exposure of at least 10 Gy, 48 (10%) have more than 50% of the pulmonary valve exposed to at least 10 Gy.

Systematic influences on the dose distribution Aside from tumor laterality, we assessed several potential influences on cardiac dose (Table 3). Results for three age groups with cut points at 50 and 70 years – roughly the 33.3% and 66.6% quantiles of the cohort, show a trend for reduced average as well as median mean heart dose with higher age at diagnosis for leftsided RT (p < 0.01). For 734 (95%) patients with information on height and weight, we calculated BMI and defined groups as the 33.3% (23.2) and 66.6% (27.0) quantiles. For left-sided RT, there is a significant trend for increased average as well as median mean heart dose with higher BMI (p < 0.001). For 444 (58%) patients, tumor location was known to be in a medial (ICD codes C50.2 and C50.3) or lateral (ICD codes C50.4, C50.5, and C50.6) quadrant [26]. Average and median mean heart doses are elevated for left-sided RT for tumors in a medial vs. lateral quadrant (p < 0.01, location difference 0.78 Gy, CI 0.26–1.31). DMEAN differed for patients treated in center A vs. B for leftsided RT (p < 0.0001, location difference 0.93 Gy, CI 0.57–1.29), and for right-sided RT (p < 0.001, location difference -0.25 Gy, CI 0.39 to 0.11). Treatment-defined patient sub-groups were compared separately for each treatment center (Table 4). In center A, patients

Please cite this article in press as: Wollschläger D et al. Radiation dose distribution in functional heart regions from tangential breast cancer radiotherapy. Radiother Oncol (2016), http://dx.doi.org/10.1016/j.radonc.2016.01.020

4

Heart dose in breast cancer radiotherapy

DMEAN

D2CC

V10GY_%

0.6 Side

left right

0.5

Density

0.4

0.3

0.2

0.1

0.0

0

5

10

15

20 0

20

40

60

0

20

40

60

Dose [Gy] / relative volume [%] Fig. 2. Histograms and nonparametric kernel density estimates for individual mean dose (DMEAN), dose in the most-exposed 2 cm3 (D2CC) and relative volume that received at least 10 Gy (V10Gy) in the complete heart for left-sided and for right-sided radiotherapy.

Complete heart

DMEAN [Gy] D2CC [Gy] V10Gy [%]

4.6 38.3 8.2

1.7 7.0 0.8

3.7 45.2 5.1

1.4 4.2 0

3.1 15.6 9.8

1.2 8.7 3.6

Aortic valve

DMEAN [Gy] D2CC [Gy] V10Gy [%]

2.1 2.5 0.8

2.5 3.1 1.5

1.7 2.0 0

1.8 2.1 0

1.8 2.2 7.2

2.4 3.9 8.4

with left-sided RT and electron boost had a significantly lower mean heart dose than those without boost (p < 0.0001, location difference 1.53 Gy, CI 0.90–2.17). In contrast, patients treated in center B with left-sided RT and electron or photon boost had slightly higher median heart dose than those without boost. A parasternal field increased mean heart dose, especially for left-sided RT (p < 0.0001 in both centers). DMEAN location difference with vs. without left-sided parasternal RT in center A was 2.10 Gy (CI 1.50–2.82), and 6.32 Gy in center B (CI 5.24–7.57). For right-sided RT, the difference was significant as well (p < 0.0001 in both centers, location difference A 0.61 (CI 0.43– 0.81), B 2.1 (CI 1.45–2.83)).

Pulmonary valve

DMEAN [Gy] D2CC [Gy] V10Gy [%]

6.1 7.3 11.5

2.0 2.0 0.5

3.3 3.5 0

1.7 1.7 0

7.4 10.0 27.3

1.3 1.6 5.8

Discussion

Left ant. myoc.

DMEAN [Gy] D2CC [Gy] V10Gy [%]

15.6 36.9 39.5

1.3 2.6 0.2

13.2 44.3 35.5

1.2 1.8 0

10.6 16.0 30.1

0.7 3.7 1.6

Right ant. myoc.

DMEAN [Gy] D2CC [Gy] V10Gy [%]

2.7 7.3 3.2

3.9 6.7 5.0

1.9 3.3 0

2.6 4.0 0

2.7 10.1 9.7

4.5 8.5 17.0

AV node

DMEAN [Gy] D2CC [Gy] V10Gy [%]

2.9 7.3 2.9

1.6 2.8 0.1

2.2 3.9 0.1

1.4 2.0 0

2.1 8.7 7.3

1.1 2.9 2.5

Table 2 Mean, median and standard deviations for three dose metrics calculated for six different heart structures for left-sided and right-sided irradiation. Structure

Metric

Median

SD

Left

M Right

left

Right

Left

Right

Based on individual anatomy and treatment planning, we analyzed radiation exposure of several heart structures in a sample of 769 women with breast cancer. Patients were treated 1998– 2008 in two German university medical centers with megavoltage external beam RT using tangential fields. Strengths and limitations Radiotherapy practice in large tertiary care centers may not be representative for the covered time period in general. The cohort is

Table 3 DMEAN average, median, standard deviations and number of observations for the complete heart in several patient sub-groups for left-sided and right-sided radiotherapy. M

Median

SD

N

Left

Right

Left

Right

Left

Right

Left

Right

Age group

650 50–70 70+

5.1 4.3 4.2

1.9 1.5 1.6

4.0 3.7 3.3

1.5 1.3 1.3

3.5 2.7 2.9

1.7 0.8 1.0

188 212 86

107 133 43

BMI group

623.3 23.3–27.0 27.0–50.4

3.9 4.6 5.1

1.5 1.8 2.0

3.4 3.5 4.2

1.3 1.4 1.5

2.5 3.2 3.3

1.1 1.1 1.6

148 160 157

98 85 86

Quadrant

Medial Lateral

5.3 3.9

1.6 1.7

4.1 3.5

1.5 1.4

3.5 1.9

0.5 1.0

85 192

41 126

Center

A B

4.9 4.3

1.4 2.0

4.3 3.1

1.2 1.5

2.7 3.4

0.6 1.6

249 237

146 137

Please cite this article in press as: Wollschläger D et al. Radiation dose distribution in functional heart regions from tangential breast cancer radiotherapy. Radiother Oncol (2016), http://dx.doi.org/10.1016/j.radonc.2016.01.020

D. Wollschläger et al. / Radiotherapy and Oncology xxx (2016) xxx–xxx Table 4 DMEAN average, median, standard deviations and number of observations for the complete heart in two treatment-defined patient sub-groups for left-sided and rightsided radiotherapy (RT). M

Center A

Boost

Parasternal field Center B

Boost

Parasternal field

Median

SD

Left

Right

Left

Right

Left

Right

No Electron Photon No Yes

6.2 4.5 4.2 4.3 6.5

1.5 1.4 1.9 1.3 2.0

5.5 3.9 3.9 3.9 6.1

1.5 1.4 1.9 1.1 1.8

3.1 2.4 1.3 2.4 2.8

0.7 0.6 1.1 0.5 0.7

No Electron Photon No Yes

4.7 4.2 4.1 3.1 9.7

1.9 1.6 2.1 1.5 3.5

2.6 3.4 3.3 2.7 9.2

1.1 1.5 1.8 1.4 3.5

4.3 3.2 2.6 1.6 4.0

2.3 0.8 1.3 0.6 2.7

limited to women with a good prognosis, e.g., there were no bi-lateral tumors or metastases. Within each treatment center, RT planning and implementation was largely consistent for all patients. Given the high compliance with RT treatment guidelines for breast cancer starting in the 1990s, and considering the standardization on using tangential megavoltage photon fields, treatment should also be broadly similar in other large centers with access to three-dimensional CT imaging and recent treatment planning software. Dose calculations are subject to several sources of uncertainty: The accuracy of the AAA algorithm for organs outside of the PTV on the fringe of the radiation field is lower than for the PTV itself. Other uncertainties arise from the reliability of the contouring procedure, day-to-day variations in patient positioning and immobilization, and organ movement. Taylor et al. [16] identified individual anatomy as the largest source of variability. Since our dose calculations were based on individual CT planning scans, errors should be small relative to mean doses.

Dose distribution The left anterior myocardium was impacted most by left-sided irradiation. The pulmonary valve and the AV node were also strongly affected by left-sided irradiation, while the aortic valve showed no dependency on tumor location. Each dose metric exhibited a characteristic distribution. Mean heart dose was unimodal and strongly right skewed. In contrast, D2CC for left-sided irradiation had a more bimodal shape, owing to the steep dose gradient at the field borders: The most exposed tip of the heart was either in the field, and thus received close to prescription dose, or was fully out. This characteristic may make D2CC more susceptible to uncertainty of patient positioning and organ movement relative to the planning CT scan. The V10Gy distribution showed much larger dispersion for left- than for rightsided RT, especially in the left anterior myocardium. Younger women, women with higher BMI, and with a tumor location in a medial quadrant had a higher mean heart dose. A large effect was apparent for being irradiated with a parasternal field. Despite following the same broad guidelines, one strong moderating factor for cardiac dose turned out to be the policy of the individual department. The likely cause for a lower average DMEAN with vs. without boost in center A is that patients with boost often had a prescription dose of 46 Gy via tangential fields plus electron boost with limited depth penetration. In contrast, patients without boost often had at least 50 Gy via tangential fields. Consistent with [16], boost RT in center B was associated with a slightly higher DMEAN, most pronounced in right-sided RT with photon boost.

5

Comparison to historical data Taylor et al. [10] provides an overview of heart exposure following breast RT from the 1950s to the 1990s. Our sample exhibits slightly lower heart exposure as reported for Swedish breast cancer patients treated in the 1990s with a prescription dose of about 54 Gy and an average mean heart dose of 5.8 Gy for left-sided RT [27]. Compared to an average mean heart dose of 2.3 Gy for UK patients treated with left-sided RT in 2006 with prescription doses of 40 Gy [16], our sample shows higher heart exposure.

Implications for the analysis of dose–response relationships For left-sided RT, there is high within-group variability in cardiac dose for the complete heart and its functional substructures. In dose–response analyses between absorbed dose and late cardiac effects [6,7,11], a left-RT vs. right-RT comparison ignores the large heterogeneity of left-RT patients. Therefore, individual dosimetry seems highly preferable. A feasible approach for the retrospective analysis of historic cases or large cohorts is matching individual patient and therapy data to standard anatomies and typical treatment styles for which heart doses are determined [5,11]. Our dose reconstruction for individual patients and functional cardiac substructures gives a better account of the true variability of exposure and may help improve dose–response modeling. The different distribution patterns of dose metrics in different cardiac sub-structures suggest that rather than relying solely to mean heart dose, one may try to link a particular dose metric in a particular organ at risk with a clinically connected cardiac event type. The influence of department policy on heart dose calls for a stratified analysis of late cardiac effects.

Conclusions With considerable inter-personal variability, systematic influences on cardiac dose included a complex interaction of tumor location, boost type, presence of a parasternal field, age, BMI, and department policy. Dose metrics and cardiac sub-structures showed different patterns in their distribution and dependency on external influences. The analysis of dose–response relationships for late cardiac effects should account for heterogeneous radiation exposure.

Conflict of interest statement All authors declare no conflict of interests.

Acknowledgments This work was supported by the German Federal Ministry of Education and Research (BMBF) with contract number 02NUK026B. The study sponsor had no involvement in the study design, in the collection, analysis and interpretation of data; in the writing of the manuscript; or in the decision to submit the manuscript for publication.

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.radonc.2016.01. 020.

Please cite this article in press as: Wollschläger D et al. Radiation dose distribution in functional heart regions from tangential breast cancer radiotherapy. Radiother Oncol (2016), http://dx.doi.org/10.1016/j.radonc.2016.01.020

6

Heart dose in breast cancer radiotherapy

References [1] Senkus E, Kyriakides S, Ohno S, et al. Primary breast cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol 2015;26:8–30. [2] Fuller SA, Haybittle JL, Smith REA, et al. Cardiac doses in post-operative breast irradiation. Radiother Oncol 1992;15:1348–53. [3] Gyenes G, Gagliardi G, Lax I, et al. Evaluation of irradiated heart volumes in stage I breast cancer patients treated with postoperative adjuvant radiotherapy. J Clin Oncol 1997;15:1348–53. [4] Taylor CW, McGale P, Darby SC. Cardiac risks of breast-cancer radiotherapy: a contemporary view. Clin Oncol (R Coll Radiol) 2006;18:236–46. [5] Taylor CW, Nisbet A, McGale P, et al. Cardiac doses from Swedish breast cancer radiotherapy since the 1950s. Radiother Oncol 2009;90:127–35. [6] Schulz-Hector S, Trott K-R. Radiation-induced cardiovascuar disease: is the epidemiologic evidence compatible with the radiobiologic data? Int J Radiat Oncol, Biol, Phys 2007;67:10–8. [7] Gagliardi G, Constine LS, Moiseenko V, et al. Radiation dose-volume effects in the heart. Int J Radiat Oncol, Biol, Phys 2010;76:77–85. [8] Early Breast Cancer Trialists’ Collaborative Group. Effect of radiotherapy after breast-conserving surgery on 10-year recurrence and 15-year breast cancer death: meta-analysis of individual patient data for 10 801 women in 17 randomised trials. Lancet 2011;378:1707–16. [9] Darby S, McGale P, Peto R, et al. Mortality from cardiovascular disease more than 10 years after radiotherapy for breast cancer: nationwide cohort study of 90,000 Swedish women. BMJ 2003;326:256–7. [10] Darby S, McGale P, Taylor CW, et al. Long-term mortality from heart disease and lung cancer after radiotherapy for early breast cancer: prospective cohort study of about 300 000 women in US SEER cancer registries. Lancet Oncol 2005;6:557–65. [11] Darby S, Ewertz M, McGale P, et al. Risk of ischemic heart disease in women after radiotherapy for breast cancer. N Engl J Med 2013;11:987–98. [12] de Angelis R, Sant M, Coleman MP, et al. Cancer survival in Europe 1999–2007 by country and age: Results of EUROCARE-5: A population-based study. Lancet Oncol 2014;15:23–4.

[13] Jansen L, Castro FA, Gondos A, et al. Recent cancer survival in Germany: an analysis of common and less common cancers. Int J Cancer 2015;136:2649–58. [14] Howlader N, Noone AM, Krapcho M, et al. SEER cancer statistics review, 1975– 2012. Bethesda, MD: National Cancer Institute; 2015. [15] Taylor CW, Nisbet A, McGale P, et al. Cardiac exposures in breast cancer radiotherapy: 1950s–1990s. Int J Radiat Oncol, Biol, Phys 2007;69:1484–95. [16] Taylor CW, Povall JM, McGale P, et al. Cardiac dose from tangential breast cancer radiotherapy in the year 2006. Int J Radiat Oncol, Biol, Phys 2008;72:501–7. [17] Schwentner L, Wöckel A, König J, et al. Adherence to treatment guidelines and survival in triple-negative breast cancer: a retrospective multi-center cohort study with 9156 patients. BMC Cancer 2013;13:487. [18] Neyman J. On the two different aspects of the representative method: the method of stratified sampling and the method of purposive selection. J R Stat Soc 1934;97:558–606. [19] Ahnesjö A, Saxner M, Trepp A. A pencil beam model for photon dose calculation. Med Phys 1992;19:263–73. [20] Ahnesjö A. Analytic modeling of photon scatter from flattening filters in photon therapy beams. Med Phys 1994;21:1227–35. [21] Medical Software Solutions. Visualisation and organisation of data for cancer analysis (VODCA). 2014. [22] Ulmer W, Harder D. A triple Gaussian pencil beam model for photon beam treatment planning. Z Med Phys 1995;5:25–30. [23] Sievinen J, Ulmer W, Kaissl W. AAA photon dose calculation model in Eclipse. Palo Alto, CA: Varian Medical Systems; 2005. 1–18. [24] R Development Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2015. [25] Wollschläger D, Karle H. DVHmetrics: Analyze dose-volume histograms and check constraints. 2015. http://CRAN.R-project.org/package=DVHmetrics. [26] Bouchardy C, Rapiti E, Usel M, et al. Excess of cardiovascular mortality among node-negative breast cancer patients irradiated for inner-quadrant tumors. Ann Oncol 2010;21:459–85. [27] Taylor CW, Brønnum D, Darby SC, et al. Cardiac dose estimates from Danish and Swedish breast cancer radiotherapy during 1977–2001. Radiother Oncol 2011;100:176–83.

Please cite this article in press as: Wollschläger D et al. Radiation dose distribution in functional heart regions from tangential breast cancer radiotherapy. Radiother Oncol (2016), http://dx.doi.org/10.1016/j.radonc.2016.01.020