|Year : 2013 | Volume
| Issue : 1 | Page : 154-160
Impact of a large breast separation on radiation dose delivery to the ipsilateral lung as result of respiratory motion quantified using free breathing and 4D CT-based planning in patients with locally advanced breast cancers: A potential for adverse clinical implications
Thomas E Heineman1, Albert Sabbas1, Marilynn Santos Delamerced1, Ya-lin Chiu2, Michael Smith1, Bhupesh Parashar1, A Gabriella Wernicke1
1 Department of Radiation Oncology, Stich Radiation Center, NewYork-Presbyterian Hospital, Weill Cornell Medical Center, New York, NY, USA
2 Department of Public Health, Division of Biostatistics, Weill Medical College of Cornell University, New York, NY, USA
|Date of Web Publication||10-Apr-2013|
A Gabriella Wernicke
Weill Medical College of Cornell University, Stich Radiation Oncology, 525 East 68th Street, New York, NY 10065
Source of Support: None, Conflict of Interest: None
Purpose: We examined the effects of large breast separation (BS) on dosimetric and positional differences of radiation treatment plans of locally advanced breast cancers during a free-breathing respiratory cycle.
Materials and Methods: Computed tomography (CT) datasets of 18 patients were acquired using 3D, 4D-0% (end-inspiration), and 4D-50% (end-exhalation). BS was examined in relation to the positional and dosimetric changes to organs-at-risk (OAR). Based on dosimetric analysis of all three plans, we compared 4D-0% and 4D-50% for V 5 , V 10 , and V 20 to 3D for heart and ipsilateral lung. Using 4D-0% and 4D-50% CTs, we recorded positional variations of the organs' centroid (centers of mass) and their effects on dosimetry.
Results: Median BS was 23.95 cm (range: 16.86-29.48 cm). There was a strong positive correlation between BS and dose to the ipsilateral lung for V 5 , V 10 , and V 20 , with the greatest linearity observed at V 20 (R 2 = 0.23). At BS ≥27 cm, the dose increased during inspiration at V 5 , V 10 , and V 20 (P < 0.05). When comparing 4D and 3D for the heart, V 5 and V 10, decreased by average of 0.94% and 0.96% (P < 0.008), respectively.
Conclusions: This study offers the first evidence of the impact of a large BS on radiation dose to the ipsilateral lung.
Keywords: Breast cancer, breast separation, four-dimensional computed tomography, ipsilateral lung, respiratory motion
|How to cite this article:|
Heineman TE, Sabbas A, Delamerced MS, Chiu Yl, Smith M, Parashar B, Wernicke A G. Impact of a large breast separation on radiation dose delivery to the ipsilateral lung as result of respiratory motion quantified using free breathing and 4D CT-based planning in patients with locally advanced breast cancers: A potential for adverse clinical implications. J Can Res Ther 2013;9:154-60
|How to cite this URL:|
Heineman TE, Sabbas A, Delamerced MS, Chiu Yl, Smith M, Parashar B, Wernicke A G. Impact of a large breast separation on radiation dose delivery to the ipsilateral lung as result of respiratory motion quantified using free breathing and 4D CT-based planning in patients with locally advanced breast cancers: A potential for adverse clinical implications. J Can Res Ther [serial online] 2013 [cited 2019 Nov 16];9:154-60. Available from: http://www.cancerjournal.net/text.asp?2013/9/1/154/110368
| > Introduction|| |
Radiotherapy to the cancer of the breast and its regional lymph nodes has historically struggled with means of delivering radiation more precisely to a moving tissue target. ,,,,,,,, Breath hold techniques, active breathing control, and respiratory gating have been promoted as a means to limit unwanted radiation to healthy tissues, such as heart and ipsilateral lung; however, there is a trade-off for greater treatment time and more complex treatment planning. ,,, With improved survival of patients with breast cancer, reducing ipsilateral lung dose during breast radiotherapy may diminish the risk of secondary lung cancers. 
There is a subset a patients receiving radiotherapy to the breast in which respiratory motion does not affect treatment and does not put healthy organ tissue at increased risk. Predominantly, these cases involve an early stage cancer in a patient with a large breast volume, allowing the patient to be treated in a prone position with wedges.  This treatment consistently spares the lung and has variable sparing results on the heart. , However, such sparing in patients with locally advanced breast cancers (LABC), who have no other alternative but to be treated in the supine position due to irradiation delivery to both breast tissue and nodal regions, is not always possible.
This study represents the first attempt to juxtapose 4D radiotherapy coverage with a 3D, free-breathing clinical plan in patients with LABC. We attempted to reflect a more realistic day-to-day picture based on the actual clinical 3D treatment plans and compared 3D with the 4D planning fused image over the entire breathing cycle, spanning from a full inspiration and a full expiration. The aim was to examine whether the 3D current treatment plan differed significantly from the experimental non-clinically applied 4D plan.
This study provides a first analysis of the relationship between breast separation (BS) measurements and radiation dose received by organs at-risk.
| > Materials and Methods|| |
After an institutional review board rendered approval for a radiation techniques study, we selected patients with LABC left-sided cancers chosen preferentially to right-sided patients so that the effects of radiation on the heart and lung could be studied. Between 2009 and 2010, a total of 14 patients with left-sided LABC American Joint Cancer Commission stages IIB-IIIB underwent two scans: a 3D-scan used for treatment and a 4D-scan used for research purposes. All 14 patients were simulated in the supine position. Patients with large breasts (>D cup) had a special plastic mesh-bra prepared to ensure that the breasts remained in the same position every treatment. Median bra size was 42 (range: 36-48), and a median bra cup was double D (range: B-J). Median number of positive axillary nodes (AN) was 8 (range: 4-12). Median age was 52 years (range: 44-60 years).
After the computed tomography (CT) scanner had been acquired and parsed, the data were imported into a commercially available planning system (Pinnacle, Version 8.0M, Philips Medical Systems, Milpitas, CA, USA). Using both 3D and 4D protocols, three datasets were transferred from the CT scanner to the Pinnacle planning system: 3D as well as 4D-0% (end of inspiration) and a 4D-50% (end of exhalation) scans. Reference plans were performed on 3D. The two 4D sets of images were part of a cine acquisition protocol that allowed for post-acquisition parsing of the images according to the breathing cycle. The patients were treated based on the 3D plan, with the 4D plans used only for research purpose. The heart, lungs breast, tumor bed, the supraclavicular (SCV) nodes, AN levels I-III, and internal mammary nodes (IMN) were contoured for all three datasets (3D, 4D-0%, and 4D-50%) for each patient. It is our standard practice to contour IMNs for patients treated for LABC. However, no patient underwent treatment planning to include IMNs as a target, as our institution does not perform elective IMNs irradiation. IMN node contouring was removed from the structures related to treatment planning.
This project required a large degree of consistency between structure contours at each phase to allow accurate comparisons to be drawn. To achieve this, all nodes and breasts were contoured according to the Radiation Therapy Oncology Group (RTOG) Breast Cancer Atlas for radiation therapy planning. Lung contours were generated using the auto-contour tool of the planning system in the lung window, whereas the heart was contoured in the thorax window. For an added degree of consistency, contours of the 4D research protocol were then corroborated with the actual 3D treatment contours. Additionally, all contours were performed by the same primary researcher in a methodical manner to reduce inter-physician/dosimetrist variability. [Figure 1] and [Figure 2] display all three CT datasets for the heart and the lung, as well as a fusion window for a given patient.
|Figure 1: Axial images of computed tomography plans of a patient simulated in 3D and 4D, demonstrating contours of the heart (left), lung (right), and their movements. All slices are aligned based on the left breast marker. (A) 3D; (B) 4D-0% (end of inspiration); (C) 4D-50% (end of exhalation)|
Click here to view
|Figure 2: Axial (A), sagittal (B), and coronal (C) images of computed tomography treatment planning scans of a patient simulated with a 3D superimposed on a 4D plan, demonstrating contours of the heart (left), lung (right), and their movements. 3D (lung and heart in black); 4D-0% (end of inspiration) (lung and heart in gray); 4D-50% (end of exhalation) (lung and heart in white)|
Click here to view
Once the contours were generated for all structures, the volumes were calculated within Pinnacle. An auto-placement tool was used to calculate the centroid of each structure to assess the position outpoint in cranial-caudal (CC), anterior-posterior (AP), and medial-lateral (ML) for each structure at each respiratory phase.
To assess dosimetry variations between phases, unique treatment plans using Pinnacle were generated for each dataset. The fields two tangential beams with wedges irradiating the whole breast and an anterior SCV field were set on the 3D scan and transposed onto the 4D scan without adjusting for a change in organ position due to respiration. The prescription dose was 50.4 Gy (1.8 Gy/day) to the entire breast and lumpectomy cavity, IMN (when clinically indicated), SCV, and AN using megavoltage photon beams, followed by a 10-Gy photon or electron boost to the tumor bed. A plan was considered acceptable if 95% of the planned treatment volume (PTV) received ≥95% of the prescription dose. The reference plan parameters (e.g., isocenter location, beam energy/angles, port shape, weighting, etc.) were copied to the 4D image sets at different phases to generate 3D dose distributions at these phases. In this work, two out of ten breathing phases of 0% (end of full inspiration) and 50% (end of full expiration) were selected to simulate the two extremes of the breathing cycle that would be encountered during the whole breast treatment. Dose-volume-histograms (DVHs) for each phase were generated.
The values for BS were generated by a standardized planning protocol defined by the distance between the entrance points of the medial and lateral tangential beams entering at the breast isocenter point plane. 
The position data for centroid of each region-of-interest (ROI) (nodes, heart, lung) was imported into Microsoft Excel. The displacement of the centroids between the 0% and 50% datasets were then calculated.
From the DVHs of the planning system, we extracted V 5 , V 10 , and V 20 for both the heart and ipsilateral lung. We also compared the mean dose to the above structures among the three CT datasets.
Differences in DVHs between respiratory phases were analyzed using a Wilcoxon signed-rank test to evaluate the differences between paired data. Statistical significance was assumed for P < 0.05. The protocol was designed to detect statistical differences in dose due to changes in organ volume or centroid position that altered the radiation therapy delivery to the heart and lungs. Using this data, it was possible to examine any statistical relationships between dose changes due to respiratory motion in relationship to BS. To generate linear regressions of BSs when compared with dose, a plot was generated in Excel (relative dose-volume vs. BS) and a trend-line was added with a best-fit equation and R2 -value output.
| > Results|| |
Breast separation measurements yielded a median BS of 23.95 cm (range: 16.86-29.48 cm). The heart was determined to have no linearity (R2 ≈ 0) with BS; however, the lung showed a strong relationship for a small sample size. The results of this analysis are given in [Figure 3]. The strongest relationship existed for V 20 lung, with an R2 value of 0.23.
|Figure 3: Linear regression analysis of the relationship between lung dose-volume histogram (DVH) values during the respiratory cycle with breast separation|
Click here to view
Using the equation generated by the linear regression (V20 = 0.0078x − 0.0613 where x = BS in cm), it is possible to set a threshold value of BS, by which a patient with a larger BS received a greater dose of radiation. The greatest increase in dose occurred between 25 and 27 cm BS. Also apparent on this graph was the different regression lines given by 0% and 50% [Figure 3]. As seen, there was nearly a constant increase between 4D-0% and 4D-50% at all three dose volumes (V 5 , V 10 , V 20 ) as BS increases. At BS >27 cm, the dose increased during inspiration at V 5 , V 10 , and V 20 (P < 0.05). Ipsilateral lung doses increased at all three dose volumes during inspiration.
Movement data of the ipsilateral lung and heart is given in [Table 1]a and b. The maximum centroid vector movement for ipsilateral lung and heart were 8.1 mm and 6.7 mm, respectively. The average movement for each direction for all patients for the lung was 0.4 mm, 1.3 mm, and 4.4 mm (ML, AP, and superior-inferior (SI), respectively). The average movement for each direction for all patients for heart was 2.1 mm, 3.9 mm, and 4.6 mm (ML, AP, and SI, respectively). The average 3D vector movement in the lung was 1.4 mm, whereas for the heart it was 3.8 mm. Individual patient data are shown in [Table 1] and [Table 2].
When performing dosimetry calculations, we evaluated 4D-0% with 3D as well as 4D-50% with 3D. The dose-volume histogram data for ipsilateral lung (V 5 , V 10 , V 20 ) as well as the heart (V 5 , V 10 , V 20 , V 30 ) at respiratory phases 0%, 50% and 3D-planning are provided in [Table 2]a and b. Overall, statistically significant differences were found for the heart at V 5 and V 10, decreasing by an average of 0.94% (P < 0.008) and 0.96% (P < 0.0088), respectively, when comparing 4D-0% and 3D. Only V 10 for the ipsilateral lung demonstrated a statistically significant difference between 3D and 4D-50% (P = 0.0245) [Table 3]. Both V 5 and V 10 were significant for the heart when 3D was compared with 4D-0% (P = 0.0081 and P = 0.0088, respectively).
To compare the decrease in mean dose across the entire range of the breathing cycle - from full inspiration to full expiration - the data from 4D-0% was compared with 4D-50%. The decrease in mean dose to the heart over this range was 6.3%, averaged over the 12 left-sided patients. The decrease in mean dose to the lung was much less significant, with an average decrease of 0.8% at full inspiration.
We measured changes for the ipsilateral lung dosage between 4D-0% and 4D-50%, they were 3.1%, 3.4%, and 3.0% for V 5 , V 10 , and V 20, respectively.
Similar percent changes were observed when the heart was examined: 3.2%, 1.8%, 1.6%, and 6.1%, respectively, for V 5 , V 10 , V 20 , and V 30 , respectively.
| > Discussion|| |
Advancements in the treatment of breast cancers have allowed patients to enjoy prolonged survival following initial radiation treatment. Inadvertent radiation dosage to healthy tissues - heart and lung - are presented with greater consequences because of direct risks of cardiac and pulmonary toxicity, including cancer.  Consequently, reducing ipsilateral lung dose during breast radiotherapy can reduce morbidity such as secondary lung cancers. A recent study of 300,000 women receiving breast radiotherapy demonstrated three times a greater risk of developing ipsilateral lung cancer when compared with the contralateral lung.  Ipsilateral lung cancers most often occur greater than 10 years post-treatment.
Consequently, with improvements in breast cancer treatment, larger proportions of women are surviving decades following RT treatment, making long-term morbidities relating to treatment evermore pertinent.
Several authors have used biologically effective uniform dose to quantify these risks. ,,,,,,, A number of methods have previously been used to examine the changes in dose received by target and non-target tissues during respiration. In one such study by George et al.  a simulation approach was used, whereas Yue et al. performed calculations on a 4D-CT dataset. 
One of the most pressing questions this study sought to answer was the dosimetry effect on the heart and ipsilateral lung, as well as any relationship to BS. Further, we attempted to elucidate whether there was an added benefit of 4D planning over a clinically utilized 3D plan and if so, in what category of patients?
It was hypothesized that greater the BS, greater was the effect of respiratory motion on inadvertent dose to the heart and lung. This question has not been examined in RT literature. The linear regression analysis of the relationship between the dose to these organs and the BS, for a cohort of 14 patients, represents substantial linearity. Our hypothesis with regard to effect of BS was not proven in the case of the heart but held true for the lung. Thus, the conservative threshold of BS ≥27 cm above which the dose of radiation is affecting a large volume of the lung may be the one to be considered for treatment with a 4D-based radiotherapy in a patient with LABC.
The average centroid movement of the heart was 3.6 mm (SD, 1.5 mm). The average centroid movement of the lung was 5.6 mm (SD, 1.6 mm). These were greater than the values reported by Li et al., which reported an average ipsilateral lung movement as 3.9 mm but similar to the heart movement value of 3.3 mm.  Our measured centroid movement values were smaller than that found in Weiss et al., which had an average 6.0 mm movement for the heart over 9 patients and 7.7 mm movement for the lung over 6 patients.  Additionally, we did observe that greater centroid movement of the lung correlated with greater centroid movement in the heart, as the patients with the three greatest changes in lung centroid position (patients 1, 10, and 12) had 3 of the 4 greatest heart centroid position changes. There was no correlation between the magnitude of the centroid position change and the anatomical size of the heart or lung.
We observed, as expected, the dose for the heart decreased with inspiration, which moved the chest wall and breast away from the heart. We found no correlation with the size of heart with the dose changes given to the heart.
The lack of BS correlation with higher doses to the heart was another interesting outcome of this study. Other studies have not shown statistical significance relating to decreases in mean heart dose with similar study sizes.  We postulate that lack of BS correlation with doses received by the heart is secondary to the fact that the cardiac contraction/relaxation cycle is independent of the breathing cycle. The BS is a relationship to this finding in that the larger the BS, the larger degree of breast area must be irradiated resulting in larger lung areas receiving radiation. With a small patient population it is hard to reach statistical significance but further study is needed to examine heart toxicity, especially relating to the concomitant use of many heart toxic chemotherapy regimens. Further study is needed to examine heart toxicity, especially relating to the concomitant use of many heart toxic chemotherapy regimens.
Additionally, there was much more variation in the dosages given to the lung than the heart. Our findings of dosage received by ipsilateral lung were supported by similar results of Qi et al. who reported values of 4.4%, 2.8%, and 1.9%.  This could be explained by the reality that the lung received a similar dose, but since the volume of lung decreased during exhalation, the dose per volume would increase. Again, our results were consistent with Li et al.  in V 10 and V 20 whose group reported values of 3.2% (V 10 ) and 1.2% (V 25.2 ) but a greater result for V 5 (8.6%).
Our results are in general agreement with other studies that have observed reductions in dose to organs at-risk by accounting for respiratory motion. For example, Korremann et al. did show that free breathing gating could reduce cardiac doses with significant lung sparing in a study of 17 patients. 
This study has limitations that deserve mention. The patients and plan generation originated from a single institution and the patients selected for the study were not randomized. A common theme between many of the observed volumetric and dosimetric changes to various target structures and at-risk structures was the anatomical size and positional variations between patients resulting in a large range for any calculations. This makes finding statistical significance difficult in small patient studies such as this.
| > Conclusion|| |
With ever-improving breast cancer screening and treatment, women are being treated earlier and are living longer. This further emphasizes the role in oncology research of reducing inadvertent side effects of treatment that may occur decades following treatment, which may be the synergistic effects of chemotherapy (including anthracyclines, taxanes, and others) with radiation.
This study proves a strong association between increasing BS and increases in dose to the ipsilateral lung. The heart is also at risk of an unwanted increase in dose over the breathing cycle but this finding is independent of BS. Additionally, it is clear that there is movement of target structures and organs between the free breathing clinical plans and the 4D CT plans, which will blur the dosage given even in circumstances of precise delivery and planning. Due to the increased planning and delivery time for 4D radiation therapy, this study offers the first argument for using a quantitative rationale for choosing 4D versus 3D planning based on BS. Limitations of the study include the small patient population which could result in some inter-patient variation in data. A larger study, which includes pulmonary function testing, might be warranted.
More specific planning based on BS could be applied to any organ-at-risk-reducing modality, one of the most exciting of which being image-guided radiotherapy (IGRT). IGRT is a logical and feasible extension of BS-based determination of women at risk for cardiac and pulmonary toxicity post-radiotherapy requiring treatment protocols that protect organs at-risk, still in its preliminary stages of implementation in the area of breast radiotherapy. Additionally, a prospective study comparing 3D and 4D treatment plans could generate a more precise algorithm when creating a 4D-treatment criterion.
| > References|| |
|1.||Qi XS, White J, Rabinovitch R, Merrell K, Sood A, Bauer A, et al. Respiratory organ motion and dosimetric impact on breast and nodal irradiation. Int J Radiat Oncol Biol Phys 2010;78:609-17. |
|2.||Stranzl H, Zurl B, Langsenlehner T, Kapp KS. Wide tangential fields including the internal mammary lymph nodes in patients with left-sided breast cancer. Influence of respiratory-controlled radiotherapy (4D-CT) on cardiac exposure. Strahlenther Onkol 2009;185:155-60. |
|3.||Yue NJ, Li X, Beriwal S, Heron DE, Sontag MR, Huq MS. The intrafraction motion induced dosimetric impacts in breast 3D radiation treatment: A 4D CT-based study. Med Phys 2007;34:2789-800. |
|4.||Li G, Citrin D, Camphausen K, Mueller B, Burman C, Mychalczak B, et al. Advances in 4D medical imaging and 4D radiation therapy. Technol Cancer Res Treat 2008;7:67-81. |
|5.||Frazier RC, Vicini FA, Sharpe MB, Yan D, Fayad J, Baglan KL, 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. |
|6.||George R, Keall PJ, Kini VR, Vedam SS, Siebers JV, Wu Q, et al. Quantifying the effect of intrafraction motion during breast IMRT planning and dose delivery. Med Phys 2003;30:552-62. |
|7.||Moeckly SR, Lamba M, Elson HR. Respiratory motion effects on whole breast helical tomotherapy. Med Phys 2008;35:1464-75. |
|8.||Su FC, Shi C, Mavroidis P, Goytia V, Crownover R, Rassiah-Szegedi P, et al. Assessing four-dimensional radiotherapy planning and respiratory motion-induced dose difference based on biologically effective uniform dose. Technol Cancer Res Treat 2009;8:187-200. |
|9.||Weiss E, Wijesooriya K, Dill SV, Keall PJ. Tumor and normal tissue motion in the thorax during respiration: Analysis of volumetric and positional variations using 4D CT. Int J Radiat Oncol Biol Phys 2007;67:296-307. |
|10.||Langen KM, Jones DT. Organ motion and its management. Int J Radiat Oncol Biol Phys 2001;50:265-78. |
|11.||Frazier RC, Vicini FA, Sharpe MB, Yan D, Fayad J, Baglan KL, 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. |
|12.||Remouchamps VM, Vicini FA, Sharpe MB, Kestin LL, Martinez AA, Wong JW. Significant reductions in heart and lung doses using deep inspiration breath hold with active breathing control and intensity-modulated radiation therapy for patients treated with locoregional breast irradiation. Int J Radiat Oncol Biol Phys 2003;55:392-406. |
|13.||Korreman SS, Pedersen AN, Nøttrup TJ, Specht L, Nyström H. Breathing adapted radiotherapy for breast cancer: Comparison of free breathing gating with the breath-hold technique. Radiother Oncol 2005;76:311-8. |
|14.||Darby SC, McGale P, Taylor CW, Peto R. 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. |
|15.||Varga Z, Hideghéty K, Mezo T, Nikolényi A, Thurzó L, Kahán Z. Individual positioning: A comparative study of adjuvant breast radiotherapy in the prone versus supine position. Int J Radiat Oncol Biol Phys 2009;75:94-100. |
|16.||De Wyngaert JK, Jozsef G, Mitchell J, Rosenstein B, Formenti SC. Accelerated intensity-modulated radiotherapy to breast in prone position: Dosimetric results. Int J Radiat Oncol Biol Phys 2007;68:1251-9. |
|17.||Griem KL, Fetherston P, Kuznetsova M, Foster GS, Shott S, Chu J. Three-dimensional photon dosimetry: A comparison of treatment of the intact breast in the supine and prone position. Int J Radiat Oncol Biol Phys 2003;57:891-9. |
|18.||Mihai A, Rakovitch E, Sixel K, Woo T, Cardoso M, Bell C, et al. Inverse vs. forward breast IMRT planning. Med Dosim 2005;30:149-54. |
|19.||Wu WC, Chan CL, Wong YW, Cuijpers JP. A study on the influence of breathing phases in intensity-modulated radiotherapy of lung tumours using four-dimensional CT. Br J Radiol 2010;83:252-6. |
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2], [Table 3]