Introduction
Worldwide, an estimated 1·7 million women were diagnosed with breast cancer in 2012.Reference Ferlay, Soerjomataram and Ervik1 For women in Malta, breast cancer is the most common female carcinoma in Malta, with 116·2 cases/100,000 population, accounting for an estimated 35% of all cancer incidence.2
Breast conservation surgery with adjuvant radiotherapy has become the standard treatment for women with early-stage breast cancer, providing between 85–95% equivalent survival and locoregional control when compared with mastectomy.Reference Keller, Cohen and Sopka3
National screening programmes using improved imaging modalities now mean that breast cancer is commonly detected at the earlier (0, I and II) stages.Reference Falahatpour, Aghamiri and Anbiaee4 Screening roughly halves the incidence of axillary node involvement.Reference Barco, Chabrera and Font5 Given the large number of women undergoing radiotherapy and the expected long-term survival,Reference Neal, Torr, Heyler and Yarnold6 it is imperative that, for these patients, irradiation of adjacent radiosensitive organs such as the lung is minimised.
Whole breast radiotherapy involves the use of a pair of wedged tangential fields with the patient lying in a supine position.Reference Ramella, Trodella and Ippolito7 The breast is treated isocentrically with the posterior edges of the beams parallel to each other to avoid divergence into the lung.Reference Kunkler8 However, the shape of the underlying chest cavity necessitates the inclusion of some lung within the fieldReference Minor, Yashar and Spanos9, 10 (Figure 1).
Wedged pair of tangential fields used in radiotherapy to the breast. Note: Image retrieved from www.intechopen.com on Saturday 9 May 2015.10
Late pulmonary toxicity arising from primary breast irradiation include radiation pneumonitis and fibrosis.Reference Bhatnagar, Heron, Deutsch, Brandner, Wu and Kalnicki11 Between 5–15% of patients develop symptomatic pneumonitis.Reference Jagannath, Lokesh and Thejaswini12 Although pneumonitis can often be managed, symptomatic pulmonary fibrosis is irreversible and should therefore be avoided.Reference Patterson and Stark13 These patients may also develop second malignancies, particularly lung cancer.Reference Hamilton, Tyldesley, Li, Olson and McBride14
As the lung functions as a parallel organ, the likelihood of radiation-induced toxicity is directly associated with both the dose and irradiated volume of lung.Reference Iyer and Libshitz15 Dose–volume parameters widely used as predictors of pulmonary complications are mean lung dose (MLD) and the volume of lung receiving more than a threshold dose, such as V20.Reference Seppenwoolde16 The summary three-dimensional (3D) parameter MLD (cGy) was found to be a significantly better predictor of radiation pneumonitis than V20 in a multi-institutional study of 540 patients by Kwa et al.Reference Kwa, Lebesque and Theuws17 More recently, it was demonstrated that MLD provided a strong correlation with fibrosis volume, and was deemed a convenient parameter to compare plans in terms of radiation-induced lung fibrosis.Reference Oh, Noh and Jang18 The role of low-dose parameters, such as V5, as predictors of radiation pneumonitis is equivocal.Reference Agrawal, Kuwar, Lawrence, Das and Kumar19, 20
Estimating the probability of developing radiation pneumonitis in patients receiving radiotherapy is important for devising a treatment plan that maximises tumour dose and minimises the normal tissue complication probability.Reference Seppenwoolde16 3D treatment planning systems have enabled quantitatively related dose/volume parameters to outcomes by calculating Dose–volume histograms (DVHs) of organs at risk (OARs).Reference Kwa, Lebesque and Theuws17 A DVH is a plot of percentage volume against indicated dose and is a tool available on a treatment planning system.Reference Halperin, Perez and Brady21 A number of mathematical models have therefore been designed to predict the normal tissue complication probability from a DVH.Reference Kwa, Lebesque and Theuws17
In patients with larger breast sizes, tangential radiotherapy may prove challenging owing to the breasts falling laterally, requiring beam angle manipulation to ensure complete breast tissue coverage posteriorly, at the expense of further increasing irradiated lung volumesReference Goldsmith, Haviland, Tsang, Syndenham and Yarnold22 (Figure 2). Therefore, MLD may be used to identify any association between breast size and lung dose, by correlation with breast size parameters.
Axial computed tomography images showing small breast (left) versus a larger breast (right), illustrating additional lung volume in the larger breast.
Malta has the second highest proportion of obese women in the European Union, with 21·1% of the population (aged 18 years and over) being obese.23 As a strong positive correlation is evident between body mass index and breast volumeReference Kunkler8, Reference Dorn, Corbin, Al-Hallaq, Hasan and Chmura24 and obesity is an aetiological factor for breast cancer in postmenopausal women,Reference Kunkler8 the number of large-breasted women in Malta requiring radiotherapy is expected to increase. This subgroup of patients may therefore require the introduction of different techniques, such as prone positioning,Reference Ramella, Trodella and Ippolito7 custom made microshellsReference Latimer, Beckham, West, Holloway and Delaney25 and intensity-modulated radiation therapy (IMRT)Reference Bhatnagar, Heron, Deutsch, Brandner, Wu and Kalnicki11 to provide optimised treatment delivery.
The aim of this study was to explore the impact of breast size on MLD for patients receiving standard tangential radiotherapy treatment to the whole breast for a Maltese cohort of patients.
Methods and Materials
This study was a quantitative, non-experimental retrospective plan review. All data were obtained from archived treatment plans of patients from a single radiotherapy centre within a public oncology hospital in Malta.
Sample and treatment characteristics
3D planning data were retrieved from a Monaco® 5.0 (Elekta-CMS, Crawley, UK) treatment planning system for female patients who received adjuvant radical radiotherapy to the whole breast from the period January 2012 to December 2014.
Eligibility criteria were adult female patients diagnosed with primary breast cancer in either breast who
o were treated with the standard whole breast protocol using tangential fields with a prescribed dose of 4,005 cGy;
o were not treated for regional lymph node involvement;
o were not prescribed a breast boost.
Data collection
For eligible patients, the following parameters were recorded:
∙ Breast size was determined in terms of breast volume and chest wall separation (CWS)
o Breast volume: there are two main techniques that can be used to define the whole breast volume, either through anatomical landmarks or by contouring the breast on each computed tomography (CT) image. Local departmental protocol makes use of the former technique and this was the one followed in this study. The whole breast tissue was based on the following anatomical landmarks:
▪ Superior border: 2 cm superior to palpable breast tissue (level with the superior sternal notch);
▪ Inferior border: 2 cm inferior to breast tissue;
▪ Medial border: to include the midline of the patient;
▪ Lateral border: 1·5 cm laterally from the lateral border of the breast;
▪ Anteriorly: 1·5 cm anterior to breast tissue.
The above data were then retrieved from the treatment planning station Monaco® 5.0 (Elekta-CMS). A treatment plan was created based on ICRU guidelines26 requiring that the breast tissue should receive at least 95% of the prescribed dose. Using the external outline structure, the volume of the tissue receiving 95% of the prescribed dose was recorded. By virtue of the whole breast being planned to receive this dose level, this is a reasonable proxy of breast volume (Figure 3).
o CWS, measured in cm, was obtained by activating the planning system ruler and measuring, on the central axial CT slice, the distance from the medial radioopaque marker/tattoo to the lateral radioopaque marker/tattoo (Figure 3).
∙ MLD is the parameter used at the local oncology hospital in Malta to define dose contraints and was obtained from the DVH of the patient plans, which summarised the maximum, minimum and mean doses to outlined structures. When the DVH of the contoured structure of interest (in this case the ipsilateral lung, as seen in Figures 4 and 5) is created, the software calculates, with the use of algorithms, the minimum, maximum and mean dose in cGy.
Axial computed tomography image shows 95% isodose line and separation.
Dose–volume histogram of a patient treated with tangential whole breast radiotherapy.
Illustrating a contoured ipsilateral lung as shown on the Monaco version 5.0 software.
Analysis of data
Breast size was classified objectively based on CWS and planned target volume (PTV) into small (CWS<25 cm and PTV<1500 cm3) and large (CWS>25 cm and PTV>1,500 cm3).Reference Hannan, Thompson and Chen27 This method of classification was selected based on the fact that other methods have utilised observers to classify breast size through photographic and clinical assessment,Reference Goldsmith, Haviland, Tsang, Syndenham and Yarnold22, Reference Latimer, Beckham, West, Holloway and Delaney25 which may have introduced a subjective element. In addition, combining two breast size parameters gives a better representation of the true size of the breast. Despite this method being based in the United States, it was the only objective analysis found in literature.
Pearson’s correlation was used to determine the relationship between the following variables: 95% isodose and CWS; and MLD with both 95% isodose and CWS.
The independent sample t-test was used to compare the mean of two independent samples from the same population. This was used to compare the mean values of 95% isodose and MLD between large and small breast volumes; 95% isodose, CWS and MLD between right and left breasts and 95% isodose, CWS and MLD between right small/large breasts and left small/large breasts.
Results
A total of 233 patients were treated in this department for breast cancer in the period January 2012 to December 2014. From this data 80 patients met the eligibility criteria, whereas 153 plans were not eligible for the study. The majority of plans failed to have both lungs contoured. Table 1 provides a summary of the sample analysed, whereas Table 2 provides a summary of the results obtained.
Sample overview
Summary of results
There was a strong positive correlation between volume of 95% isodose and CWS (r=0.74, p≤0·05, which was expected as both are considered as surrogates for breast size. However, the MLD was not statistically significantly affected by either the volume of 95% isodose (r=−0·08, p=0·49) or CWS (r=−0·13, p=0·24) (refer to Figures 6 and 7).
A scatter plot of volume of 95% isodose and mean lung dose.
A scatter plot of chest wall separation and mean lung dose.
There was a significant variation between large and small breasts in volume of 95% isodose (t=5·68, p≥0·05) and in CWS (t=8·24, p≤0·05). However, no significant variation could be noted in MLD between large and small breasts (t=−0·26, p=0·80) (Table 2).
There was no significant variation in 95% isodose volume (t=−1·08, p=0·28) and CWS (t=1·42, p=0·16) between left and right breasts. However, a significant variation in MLD between the left and right breast was noted (t=3·03, p=0·00), consistent with the fact that the left lung volume is smaller than the right.Reference Kong, Klein and Bradley28
Finally, for both subgroups (small/large left breast and small/large right breasts), significant variations in volume of 95% isodose and CWS were found, respectively (volume of 95% isodose: t=3·06, p≤0·05 and CWS: t=4·87, p≤0·05 and volume of 95% isodose: t=4·79, p≤0·5 and CWS: t=12·67, p≤0·05). However, no significant variation in MLD could be observed between large and small left-sided breasts (t=−0·02, p≥0·05 and) and between large and small right-sided breasts (t=0·56, p≥0·05).
Discussion
The study evaluated the impact of breast size in terms of 95% isodose on the MLD for patients receiving standard tangential radiotherapy treatment to the whole breast for a Maltese cohort of patients.
The current results show that the correlation between breast size in terms of both 95% isodose and CWS to MLD is not statistically significant. This finding was consistent with the study by Hannan et al.,Reference Hannan, Thompson and Chen27 which concluded that MLD of the ipsilateral lung was not affected by the size of the breast. Moreover, in the study by Bhatnagar et al.,Reference Bhatnagar, Heron, Deutsch, Brandner, Wu and Kalnicki11 it was also demonstrated that there was no correlation between breast volume and MLD, although it could be observed with the contralateral breast. However, both studies were based on IMRT treatment, which is less commonly used for breast radiotherapy in clinical practice, and thus limits comparison with the findings of this study, which uses conventional radiotherapy.
The aforementioned studies contrasted with the study by Kong et al.Reference Kong, Klein and Bradley28 where a weak correlation was found between breast volume and MLD. Moreover, in the aforementioned study, a positive correlation was found between lung dose and CWS.
A study by Heineman et al.Reference Heineman, Sabbas, Delamerced, Chiu, Smith and Parashar29 revealed a strong positive correlation between CWS and dose to the ipsilateral lung for V5, V10 and V20 with the greatest linearity observed at V20. The study by Heineman et al.,Reference Heineman, Sabbas, Delamerced, Chiu, Smith and Parashar29 however, was focussed on women with locally advanced breast cancer who also had nodal involvement, and thus that could be a potential reason for such results. Moreover, this study used V20 as a parameter as opposed to MLD used in the local study.
Based on these findings, it is reasonable to conclude that there is no clinically significant correlation between breast volume and MLD. A factor that could have influenced the findings could involve intended modifications of beam angles to purposely comply with dose constraints or to limit excess lung within the field. Inter-oberver variation in contouring of the lungs by the radiographers at the local department could have also played a part in the results of this study.
From the results of this study, it could be concluded that there is a statistically significant relationship between the volume of 95% isodose and CWS with breast size. The link between breast size, CWS and volume of the breast is very clear. Studies have since been using these two parameters as surrogates for breast size.Reference Neal, Torr, Heyler and Yarnold6, Reference Ramella, Trodella and Ippolito7, Reference Hannan, Thompson and Chen27 Large CWS and breast volumes directly correlated with a larger bust, but no definitive indication has ever been presented as to how the line between ‘small’ and ‘large’ was drawn using these two parameters. The fixed value by which breast was categorised in this study using CWS as a convenient 2D parameter, and the volume of 95% isodose as the 3D value (‘small’=CWS<25 cm and PTV<1,500 cm3; and ‘large’=CWS>25 cm and PTV>1,500 cm3) may not have been fully representative of the breast size of these patients, mainly pointing to the possibility of patients with smaller values than the large breast category still having considerably large busts. A range would perhaps have been more suitable to allow for a latitude of sizes in the two categories. Similar studiesReference Dorn, Corbin, Al-Hallaq, Hasan and Chmura24, Reference Moody, Mayles and Bliss30 made use of the clinical scenario to determine breast size based on observers classifying the breast during clinical assessment, yet for such classifications, an element of subjectivity could play a part in determining breast size.
The use of the 95% isodose as a volume, despite being justified as used by previous studies,Reference Neal, Torr, Heyler and Yarnold6, Reference Ramella, Trodella and Ippolito7, Reference Hannan, Thompson and Chen27 could also not have been the most ideal parameter of volume, as it may include volume of tissue that is not breast tissue and may also underestimate breast tissue volume at the build-up region. This was, however, used in this study as the PTV volume was not outlined on the vast majority of patient plans reviewed for this study. The result of this study, however, in comparison with studies using PTV volume,Reference Hannan, Thompson and Chen27 yielded a similar result.
Finally, a high variation in MLD for left and right breast was found, consistent in that it is known that left lung irradiated volume is smaller than the right lung irradiated volume during radiotherapy to the breast.Reference Kong, Klein and Bradley28 This is because anatomically, the left lung is smaller than the right, owing to the asymmetry of the position of the heart slightly shifted to the left.Reference Schunke, Schulte, Ross, Schumacher and Lamperti31 This shift of the heart to the left indicates, particularly with patients treated for radiotherapy to the left breast, that a larger area of the heart could be potentially included in the field, and thus more regard is to be given to the dose received by the heart, also defined as an OAR.Reference Nielsen, Berg and Pedersen32
Conclusion
As both the volume of 95% isodose and CWS are surrogates for breast size, it can be concluded that breast size does not affect the MLD in a clinically important way, in the cohort evaluated in this study. However, further studies on patients with larger bust sizes are still required, focussing on the impact of breast size in dosimetry, set-up reproducibility, cosmetic outcomes and other organs at risk, for example, the heart.
Despite the lack of correlation between breast size and MLD in this study, different techniques for different patient subgroups could be implemented to ensure the delivery of more optimised, patient-tailored treatment.
Acknowledgements
The authors would like to thank the management and staff of a public radiotherapy department in Malta for providing the required anonymous data for the study.
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Conflicts of Interest
None.
Ethical Standards
The authors assert that all procedures contributing to this work comply with the ethical standards of the Malta National Bioethics Consultative Committee on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008, and has been approved by the University of Malta Research and Ethics Committee (UREC reference number: 174/2014).
