- Email: [email protected]
Optimal Target Delineation and Treatment Techniques in the Era of Conformal Photon and Proton Breast and Regional Nodal Irradiation
Journal Pre-proof Optimal Target Delineation and Treatment Techniques in the Era of Conformal Photon and Proton Breast and Regional Nodal Irradiation Emily S. Kowalski, MD, Steven J. Feigenberg, MD, Justin Cohen, MD, Zachary Fellows, CMD, Patrick Vadnais, CMD, Stephanie Rice, MD, Mark V. Mishra, MD, Jason K. Molitoris, MD, PhD, Elizabeth M. Nichols, MD, James W. Snider, III, MD PII:
S1879-8500(19)30359-5
DOI:
https://doi.org/10.1016/j.prro.2019.11.010
Reference:
PRRO 1163
To appear in:
Practical Radiation Oncology
Received Date: 16 May 2019 Revised Date:
27 October 2019
Accepted Date: 14 November 2019
Please cite this article as: Kowalski ES, Feigenberg SJ, Cohen J, Fellows Z, Vadnais P, Rice S, Mishra MV, Molitoris JK, Nichols EM, Snider JW III, Optimal Target Delineation and Treatment Techniques in the Era of Conformal Photon and Proton Breast and Regional Nodal Irradiation, Practical Radiation Oncology (2019), doi: https://doi.org/10.1016/j.prro.2019.11.010. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Inc. on behalf of American Society for Radiation Oncology.
Title: Optimal Target Delineation and Treatment Techniques in the Era of Conformal Photon and Proton Breast and Regional Nodal Irradiation Short Title: Breast Nodal Optimal Target Delineation Authors: Emily S. Kowalski MD1, Steven J. Feigenberg MD3, Justin Cohen MD1, Zachary Fellows CMD1, Patrick Vadnais CMD1, Stephanie Rice MD1, Mark V. Mishra MD2, Jason K. Molitoris MD, PhD2, Elizabeth M. Nichols, MD2, James W. Snider III MD2
1
Department of Radiation Oncology, University of Maryland Medical Center, Baltimore, Maryland
2
Department of Radiation Oncology, University of Maryland School of Medicine, Baltimore, Maryland
3
Department of Radiation Oncology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania
Corresponding Author: James W. Snider III, MD Assistant Professor Department of Radiation Oncology University of Maryland School of Medicine Baltimore, Maryland 21201 Email: [email protected] Phone: 410-369-5239
Conflict of Interest: Mark Mishra MD has received honoraria and travel funding from Varian Medical Systems, which was not used for this project. James Snider III, MD has received honoraria and speakers bureau from Varian Medical Systems and a consulting award from Siemens Healthineers, neither of which were used for this project. Dr. Snider also has a patent pending for ProtonGRID Radiotherapy Technique, which is unrelated to this project.
Outside funding: No outside funding was used in the development of this manuscript
Title: Optimal Target Delineation and Treatment Techniques in the Era of Conformal Photon and Proton Breast and Regional Nodal Irradiation Short Title: Breast Nodal Optimal Target Delineation Abstract Purpose: Regional nodal irradiation (RNI) improves disease-free and distant disease–free survival in patients with high-risk breast cancer (BC). Trials demonstrating this utilized 2- or 3-dimensional conformal radiotherapy (2D or 3D CRT) fields based on bony anatomy. Modern volumetric-modulated arc therapy (VMAT) and pencil beam scanning proton therapy (PBSPT) may underdose regional nodes (RNs) not contoured but covered by 3D CRT. Multiple atlases guide modern treatment planning. This study addresses the risk of underdosing following published atlases using 3D CRT, VMAT and PBSPT. Methods and materials: Targets per the Radiation Therapy Oncology Group (RTOG), European Society for Radiotherapy and Oncology (ESTRO), and Radiotherapy Comparative Effectiveness Consortium (RADCOMP) atlases were contoured on a representative patient CT scan. 3D CRT plans based on anatomic borders and VMAT and PBSPT plans for each set of target volumes were generated. PET/CT scans were reviewed. CT-positive and 18F-FDG–avid RNs (n = 389) were mapped from 102 patients with locally advanced (n = 51; median 2; range, 1–8 nodes) and metastatic (n = 51; median 4; range, 1–19 nodes) BC: axillary (AX; n = 284), supraclavicular (SCV; n = 60), and internal mammary nodal (IMN; n = 45). 18F-FDG–avid RNs falling within the 95% isodose line were considered adequately covered. Results: 3D CRT plans provided excellent RN coverage. Low AX nodes were covered (≥99%) in all plans. Underdosing of 18F-FDG–avid RNs falling in the high AX (78%-92%), SCV (52%-75%), and IMN (84%-89%) volumes was observed following the RTOG and ESTRO atlases for VMAT and PBSPT plans. Use of the RADCOMP atlas provided coverage of these areas (89%-100%) with slightly increased heart and lung doses. Atlas guided VMAT/PBSPT plans provided cumulative nodal coverage as follows: ESTRO (89%/88%), RTOG (93%/91%), and RADCOMP (98%/96%).
Conclusions: VMAT and PBSPT for RNI in patients with high risk BC risks underdosage in the high AX, SCV, and IMN nodal regions unless comprehensive target delineation is performed.
Introduction In women with breast cancer (BC), the addition of regional nodal irradiation (RNI), in particular internal mammary nodal (IMN) radiation, has been controversial. However, the results of 2 recently reported randomized trials have reinvigorated the “believers” support of RNI in patients with <4 lymph nodes as well as high-risk node-negative women. Whelan et al. (MA.20) demonstrated improvement in diseasefree and distant disease–free survival with the addition of RNI in node-positive or high-risk node-negative women.1 Poortmans and co-investigators (EORTC 22922) echoed this benefit in node-positive patients and patients who were node-negative with centrally located tumors.2 For many radiation oncologists, the results from these trials lowered the threshold for considering RNI. Women registered in the MA.20 and EORTC 22922 trials were treated with traditional 2- or 3-dimensional conformal radiotherapy (2D or 3D CRT). These trials required quality assurance review, specified radiation fields, permissible delivery techniques, and acceptable IMN coverage, thereby assuring relative uniformity in planning and target coverage for patients across institutions.3 2D or 3D CRT plans treat larger volumes of normal tissue to full dose and often result in dose heterogeneity and hot spots that can have significant effects on critical structures as well as on cosmesis/breast pain.4 The Cambridge experience of >1,000 women treated employed field-in-field modulation versus standard wedged tangential fields and demonstrated that efforts to improve breast dose homogeneity improved 5-year cosmesis and while reducing late toxicity.4 In an attempt to improve conformality especially of the high dose regions, many radiation oncologists have opted for more sophisticated treatment techniques. Multifield, inverse-planned IMRT and VMAT often increase high isodose line conformity, at the expense of somewhat increased low dose volume. Perhaps, the most critical normal structure in this disease site is the heart, where radiation exposure can increase the risk of complication at even low to moderate doses. Darby et al. demonstrated a linear 7.4% relative increase in the rate of cardiac events per each mean Gray delivered to the heart.5 RT remains crucial to the cure of locally advanced BC; however, increased cardiovascular risks may negate some of these benefits if not properly mitigated.
Proton therapy is an attractive alternative method of delivering RNI. Protons deliver most of their energy within the target and substantially less dose beyond the target based on the Bragg peak energy deposition, often achieving mean heart doses ≤1 Gy.6 The clinical benefit of these dose reductions is being tested in the Radiotherapy Comparative Effectiveness Consortium (RADCOMP) trial. As the field moves toward treating BC with modern conformal techniques, the question remains as to whether nodal failure patterns once covered by the historical 2D fields are adequately covered with inverse-planned IMRT, VMAT and proton therapy. Conformal techniques require careful delineation of regions involved or at risk of harboring microscopic disease. Several contouring atlases have been developed by consensus groups to guide physicians. These are critical in deciding how to target these nodal regions; however, they differ in recommended volumes. We hypothesize that this variability in nodal coverage could lead to marginal failures as a result of underdosing lymph node regions when more conformal techniques are utilized. To test this hypothesis, we first identified nodal areas at risk using 18FFDG PET/CT, for which literature supports superior sensitivity for detecting extra-axillary lymph nodes and distant metastases compared with conventional imaging in the locally advanced setting.7-8
Methods and Materials Patient characteristics On an institutional review board–approved protocol, we retrospectively reviewed patients seen from 2009 to 2017 with an ICD 10 diagnosis code of BC, who had also undergone 18F-FDG PET/CT. Patients whose image sets were no longer available for review on our server were excluded. A total of 102 patient records were available for analysis. Patients with metastatic disease were included. Contouring per the atlas A representative patient who had undergone CT simulation with a diagnosis of left-sided BC was selected for contouring. The clinical target volumes (CTVs) were contoured by a single radiation oncologist according to 3 published BC atlases that have been central to current/previous large phase II/III studies: (1) the breast contouring atlas utilized in the RADCOMP trial9; (2) the European Society for Radiation and Oncology (ESTRO) consensus guidelines on target volume delineation for elective
radiation therapy of early-stage BC10; and (3) the Radiation Therapy Oncology Group (RTOG) atlas.11 A summary of key anatomic boundaries of lymph node volumes for each atlas is provided in the Supplement, Table I. The contouring physician was blinded to the locations of lymph nodes 18F-FDG avid on PET/CT to avoid contouring bias. Mapping PET/CT images and corresponding radiologists’ reports were reviewed by a second radiation oncologist. Lymph nodes were designated as positive if so indicated by the radiology report and/or clearly 18
F-FDG avid on PET/CT when the report was less specific. Nodes determined to be clinically positive
were mapped by the second radiation oncologist onto the representative patient utilizing anatomic landmarks in the treatment planning system. All lymph nodes were plotted on the left side of the representative patient to allow for uniform planning and analysis. A 5-mm diameter circle was placed at the epicenter of each lymph node–correlated location (Fig. 1). Planning Two dosimetrists utilized the RayStation (RaySearch Laboratories) planning system to create a 3D CRT plan based on anatomical landmarks as well as VMAT and pencil-beam scanning proton (PBSPT) plans for each atlas-based set of volumes contoured. The 3D CRT plan utilized 2 partially wide mixed energy 6 MV and 18 MV tangent beams to cover the breast and upper 3 intercostal spaces of the IMN chain matched to 2 parallel opposed 18 MV beams to treat the supraclavicular (SCV) fossa. Fields were weighted for homogeneity across the target. Contoured atlas-based volumes were utilized for inversely optimized VMAT plans, which consisted of 3 partial arcs. Planning target volumes (PTVs) for VMAT included a 5-mm expansion of the clinical target volume, removed 5mm from the skin’s surface. The proton plan included 2 anterior en face beams with a 35-degree hinge angle. A beam-specific pPTV of 5 mm was generated. Dosimetrists were instructed to achieve a conformal plan with near complete coverage of the target volume with at least 95% of the prescribed dose (Fig. 2). Dosimetrists were not informed of target coverage or organ at risk sparing achieved by competing plans, again to prevent bias.
Dosimetric analysis Plans were reviewed and approved as adequate by the first, contouring radiation oncologist. 18FFDG–avid lymph nodes were designated by the second, mapping radiation oncologist as adequately covered if the entire 5-mm circle fell within the 95% isodose line.
Results A total of 389 18F-FDG–avid lymph nodes were mapped from 102 patients with locally advanced (n = 51; median, 2; range, 1–8 nodes) and metastatic (n = 51; median, 4; range, 1–19 nodes) BC: axillary (AX; n = 284), SCV (n = 60), and internal mammary nodal (IMN; n = 45) (Table II). Patients with metastatic disease, on average, were found to have more evident lymph nodes per person than patients with locally advanced disease (Table I). 18F-FDG–avid lymph nodes were most frequently found in AX levels 1–3 (n = 284; 73%), followed by the SCV (n = 60; 15%), and the IMN (n = 45; 12%) regions. The 3D CRT plan provided excellent coverage (89%–100%) across all nodal levels (Table II). When analyzed by individual nodal regions, clear differences in coverage of 18F-FDG–avid lymph nodes were evident based on the contouring conventions. Level I–III axilla Level I and II AX lymph nodes were well covered in all plans (94%–99%; Table II). 18F-FDG– avid lymph nodes located in the junction between the SCV and level III volumes (the infraclavicular fossa) and those in the level III volume proper were frequently undercovered by both the VMAT and PBSPT plans using the ESTRO- and RTOG-based treatment volumes (Fig. 3A). Most of these undercovered nodes were found in the “posterior neck,” posterolateral to the traditional level III, diving deep toward the rotator cuff musculature and abutting the coracoid process and scapula (Fig. 3A). SVC fossa Mapped lymph nodes in the SCV fossa were also undercovered by the RTOG (75%, 72%)- and ESTRO (52%, 55%)- informed VMAT and PBSPT plans, respectively (Table II). First, multiple 18FFDG–avid lymph nodes were seen posterior to the sternocleidomastoid muscle, in the posterior neck, similar to region of level Vb as defined by head and neck RT volumes (Fig. 3B).12 Second, undercovered
18
F-FDG–avid nodes were found at the inferior edge of SCV volumes near the junction of the SCV fossa
with the upper mediastinum and IMNs. These nodes were clustered just superior to the attachment of the first rib to the sternum, posterior to the 95% isodose line around the common carotid artery, and at the separation of the brachiocephalic vein into the internal jugular and subclavian vein (Fig. 3C). Third, nodes were found at the level of the thyroid cartilage, superior to the delineation extent of all 3 atlases (Fig. 4B). In 2 patients, undercovered nodes were found in the substernal space and adjacent to the medial edge of the carotid. IMNs All 3 atlases denote the inferior extent of the third intercostal space as the border of the IMN volume. ESTRO guidelines indicate the fourth intercostal space may be included when primary tumors are located in the lower inner quadrant. In patients with metastatic disease, 18F-FDG–avid lymph nodes were found outside the 95% isodose line in the fourth (n=1) and fifth (n=1) intercostal spaces (Fig. 4A). Overall, the population contained a greater than expected number of IMNs (n = 45; 12%). The RADCOMP-based VMAT and PBSPT plans provided improved coverage of each of the aforementioned nodal volumes. 18F-FDG–avid nodes in levels 1–3 of the axilla were covered 100% in both the VMAT and PBSPT plans (Table II). 18F-FDG–avid nodes in the SCV fossa (95%/95%) and IMN (89%/89%) regions were also better covered by both plans (Table II). As with the ESTRO and RTOG atlases, the RADCOMP atlas does not extend superiorly to the thyroid cartilage, and undercoverage was encountered in this area in advanced cases.
Discussion The use of advanced RT techniques allows for conformal RT plans to complex targets, improved dose distributions in patients with challenging anatomy and lower doses to nearby organs at risk. As opposed to 2D and 3D techniques, which rely on bony landmarks for field design, VMAT and PBSPT treatment plans are highly dependent on target contouring. Multiple competing atlases guide radiation oncologists in contouring these volumes. We, along with previous investigators,16-20 have demonstrated that most of these atlases fail to incorporate several areas at risk for nodal metastases based on retrospective review of
PET/CT examinations in BC patients. This is, to our knowledge, the first effort to demonstrate the differential dosimetric impact of these flaws in target delineation based on treatment technique and is the first analysis of PBSPT in this setting. The inclusion of the RADCOMP atlas as a more comprehensive alternative is also unique. In our study, metabolically active lymph nodes undercovered by VMAT and PBSPT techniques were most frequently found in the following locations: cranial to the cricoid cartilage, in the posterolateral neck, at the junction of the IMN and SCV volumes, and in the fourth and fifth intercostal spaces. Our 3D CRT technique was unintentionally (not based on contouring) comprehensive in most of these areas of undercoverage. This highlights the caveat that comes with VMAT and PBSPT radiotherapy techniques. As dose distributions become more conformal to target volumes, radiation oncologists must be confident that areas at risk for harboring gross or microscopic disease are accurately delineated. On the other hand, overcontouring can also lead to increased organ-at-risk exposure. Therefore, a balance must be struck. Published reports have retrospectively evaluated the RTOG and ESTRO atlases for comprehensiveness of target volumes and found similar areas of undercoverage. Three modern efforts evaluated the delineation of SCV lymph node channels as provided by various atlases. Jing et al. mapped the location of 524 metastatic SCV lymph nodes and concluded that only 63% were covered by the RTOG atlas volumes.13 Pathologic nodes were found lateral to the sternocleidomastoid and scalene muscles in 80% of their population. In line with our observations, Jing et al also found a significant proportion (55.3%) of their patient population exhibited pathologic nodes above the base of the cricoid cartilage.13 Brown and colleagues reported that most patients with pathologic nodes in the cranial SCV volume, approaching the thyroid cartilage, had multiple additional SCV metastases, suggesting that this region is most at risk in patients with advanced nodal disease.y14 Also consistent with the current series, Jing et al. and Brown et al. found involved lymph nodes in the posterolateral neck outside of the RTOG volume. In a population of patients with recurrent BC, DeSelm et al. found that a majority of nodal
recurrences outside the RTOG and ESTRO CTVs within the SCV region were located in the fat spaces lateral to the sternocleinomastoid muscle and posterior to the transverse process.15 Most recently Borm and co-investigators reviewed 601 pathologic lymph nodes in 235 BC patients and evaluated the coverage provided by both the RTOG and ESTRO atlases.16 Patients with distant metastases were much more likely to have IMN and SCV fossa involvement. Overall, they found 17% and 21.2% of lymph node metastases were outside the RTOG and ESTRO guided boundaries, respectively. The volumes of covered lymph nodes were 69.8% ± 35.5% and 69.1% ± 36.3% for primary BC and 57.6% ± 38.9% and 51.1% ± 39.1% for recurrent BC using the RTOG and ESTRO guidelines, respectively. They identified many similar areas of undercoverage including the posterior SCV volume and the border between the SCV and IMN volumes.16 Jethwa et al. investigated RTOG atlas coverage of metastatic IMNs in 67 patients. Similar to our findings, they found that 78% of metastatic nodes fell within the first 3 intercostal spaces in a cranial caudal dimension; however, when also taking into consideration lateral expansions, 47% of metastatic nodes fell outside the RTOG atlas recommended volumes. Patients in this study found to have pathologic nodes caudal to the fourth rib (14%) typically exhibited additional pathologic nodes in the first 3 interspaces. Consistent with our findings, metastatic nodes (8%) were found cranial to the first intercostal space but caudal to the confluence of the IM, brachiocephalic, jugular, and subclavian veins.17 In our series, we utilized the RADCOMP atlas as a more comprehensive comparator. This atlas was informed by the above data and, therefore, exhibits more extensive volumes in the areas of concern. The RADCOMP atlas recommends a seamless junction between the SCV and IMN volumes and extends the IMN volume medially to include the fat space. It exhibits cranial SCV and inferior level I borders similar to those of the other 2 atlases but also extends the posterolateral border in the low neck near the junction of the infraclavicular and SCV volumes. Use of this atlas led to much better dosimetric coverage of the PET/CT positive lymph nodes in our series when utilizing VMAT and PBSPT techniques. The improved coverage provided by the RADCOMP atlas came at the expense of increased mean heart (950 cGy, VMAT) and ipsilateral lung V20 (16.6%, VMAT) exposure. Mean heart dose and
ipsilateral lung V20 with the RADCOMP atlas informed plans were only mildly increased over those of the RTOG and ESTRO atlases (Table II). When RADCOMP volumes were planned with PBSPT, these doses, as expected, decreased significantly (Table II). Of note, organs at risk were not prioritized by the dosimetrists during planning because this could have introduced confounders based on the representative patient’s anatomy or the planner’s aggressiveness in sparing these organs at risk. Poortmans et al.2 and Whelan et al.1 found no statistical difference in the rates of cardiac events and low rates of pneumonitis and pulmonary fibrosis when randomizing women to nodal irradiation versus breast/chest wall irradiation alone. Although these trials did not publish the organ at risk doses, the difference between heart and lung dose in patients treated with or without nodal irradiation is undoubtedly greater than the discrepancy encountered here based on one atlas-driven plan versus another. We present cardiac and lung doses only to highlight that the more comprehensive coverage provided by RADCOMP volumes does not come at a substantial cost to organs at risk versus the less comprehensive RTOG and ESTRO volumes. Therefore, we predict differential lung and heart toxicities in patients treated with RADCOMP-generated verses RTOG/ESTRO–generated plans to be minimal, if any. It is important to note that these plans were generated without dose-mitigating approaches such as deep inspiratory breath hold, which is regularly utilized in many photon/proton clinics to minimize cardiac dose when treating left-sided breast cancer. The planning directive to achieve 95% isodose coverage of the entire IMN volume likely increased the heart and lung doses above those that would be acceptable or normally encountered in the clinic. Many protocols allow for less (~80%) isodose coverage of this region due to the concern for increased cardiac toxicity, in particular. As a result of these directives and lack of breath-hold, the 3D CRT plans resulted in higher than clinically encountered doses to heart and lung; to avoid misrepresentation of clinically acceptable parameters, these were not presented. This study has several limitations. It is retrospective in nature and included patients with lymph nodes that were sufficiently large to be seen on 18F-FDG PET/CT. This resulted in a selection bias favoring patients with bulkier, locally advanced disease. The included patients with distant metastatic disease had twice as many 18F-FDG–avid lymph nodes as those with locally advanced disease, which led
to improved power of the analysis. However, it is likely that nodal patterns encountered in the patients in this study, while emblematic of potential regions of failure, are not representative of most early stage or minimally locally advanced patients. For earlier stage patients with nodal disease uncovered on sentinel lymph node biopsy, smaller treatment volumes may be adequate. This remains a question for prospective trials, but this study does demonstrate that modern techniques have become sufficiently conformal to “miss” areas not intentionally targeted that could be at risk. Additional limitations include that metabolically active nodes were presumed malignant without confirmatory biopsy. Radiologists reviewed the initial PET/CTs, and their reports were utilized to guide selection of clinically positive nodes. Transferring of the nodal positions onto the standard patient was performed manually by a radiation oncologist based on anatomic landmarks and represent approximations. Deformable registration was not utilized because of the significant differences in position (arm up versus arm down) for most scans. This effort is the first, to our knowledge, to demonstrate the dosimetric impact of various contouring conventions/atlases in the treatment of locally advanced BC. It highlights the critical importance of accurate delineation and contouring when shifting toward VMAT and PBSPT radiotherapy techniques. Several previous efforts have evaluated the RTOG and ESTRO atlases for under-contouring various regions of the AX, SCV, and IM chains. However, this is the first effort to test these formally against the more comprehensive RADCOMP atlas and to evaluate PBSPT as the most conformal modality currently being employed widely. Based on these results, we recommend utilization of the RADCOMP atlas over RTOG and ESTRO when using VMAT and PBSPT techniques to treat patients with locally advanced disease with a potential for high burden of involved nodes. This ensures coverage of at-risk nodal regions in a manner more consistent with 2D and 3D CRT techniques.
Conclusions The use of VMAT and PBSPT radiotherapy techniques in the treatment of locally advanced breast cancer should be accompanied by careful delineation of target volumes. A bias towards
more comprehensive nodal targets is recommended in more advanced cases. Without this, VMAT and PBSPT techniques may lead to underdosage of regions at risk that would have been covered with 2D or 3D CRT approaches.
Fig. 1. Digitally reconstructed radiograph. Lymph nodes 18F-FDG–avid on PET/CT in locally advanced (blue) and distant metastatic (red) patients. Fig. 2. Radiation plans. (A) 3D CRT; (B) VMAT; (C) PBSPT. VMAT and PBSPT plan utilized RADCOMP atlas volumes. Isodose lines: 105% = orange, 100% = red, 95% = pink, 70% = green, 50% = dark blue, 30% = light blue, 10% = lavender. Fig. 3. 18F-FDG–nodal coverage by atlas-driven VMAT and PBSPT plans. Level III axilla. (A). SCV (B). IMN SCV junction (C). Left to right: RTOG, ESTRO, and RADCOMP atlas guidance. Top row: VMAT; bottom row: PBSPT. Cyan = 95%, Green=50%, Lavender=30% isodose line. Fig. 4. 18F-FDG–avid nodes uncovered in all VMAT and PBSPT plans. IMNs in the fourth intercostal space (A.) and cranial SCV (B). Left to right: RTOG, ESTRO, and RADCOMP atlas guidance. Top row: VMAT; bottom row: PBSPT. Cyan = 95%, Green=50%, Lavender=30% isodose line.
References 1.
Whelan TJ, Olivotto I a., Parulekar WR, et al. Regional Nodal Irradiation in Early-Stage Breast Cancer. N Engl J Med. 2015;373(4):307-316. doi:10.1056/NEJMoa1415340
2.
Poortmans PM, Collette S, Kirkove C, et al. Internal Mammary and Medial Supraclavicular Irradiation in Breast Cancer. N Engl J Med. 2015;373(4):317-327. doi:10.1056/NEJMoa1415369
3.
Musat E, Poortmans P, den Bogaert W Van, et al. Quality assurance in breast cancer: EORTC experiences in the phase III trial on irradiation of the internal mammary nodes. Eur J Cancer. 2007;43(4):718-724. doi:10.1016/j.ejca.2006.11.015
4.
Mukesh MB, Barnett GC, Wilkinson JS, et al. Randomized controlled trial of intensity-modulated radiotherapy for early breast cancer: 5-year results confirm superior overall cosmesis. J Clin Oncol. 2013. doi:10.1200/JCO.2013.49.7842
5.
Darby SC, Ewertz M, McGale P, et al. Risk of Ischemic Heart Disease in Women after Radiotherapy for Breast Cancer. N Engl J Med. 2013;368(11):987-998. doi:10.1056/NEJMoa1209825
6.
MacDonald SM, Jimenez R, Paetzold P, et al. Proton radiotherapy for chest wall and regional lymphatic radiation; dose comparisons and treatment delivery. Radiat Oncol. 2013. doi:10.1186/1748-717X-8-71
7.
Groheux D, Giacchetti S, Delord M, et al. 18F-FDG PET/CT in staging patients with locally advanced or inflammatory breast cancer: comparison to conventional staging. J Nucl Med. 2013;54(1):5-11. doi:10.2967/jnumed.112.106864
8.
Ulaner GA, Castillo R, Goldman DA, et al. 18F-FDG-PET/CT for systemic staging of newly diagnosed triple-negative breast cancer. Eur J Nucl Med Mol Imaging. 2016;43(11):1937-1944. doi:10.1007/s00259-016-3402-9
9.
Macdonald, Shannon M; Cahlon O. Breast Atlas (RADCOMP).Pdf.; 2016. https://www.rtog.org/CoreLab/ContouringAtlases/RADCOMPBreastAtlas.aspx.
10.
Offersen B V., Boersma LJ, Kirkove C, et al. ESTRO consensus guideline on target volume delineation for elective radiation therapy of early stage breast cancer. Radiother Oncol J Eur Soc Ther Radiol Oncol. 2015;114(1):3-10. doi:10.1016/j.radonc.2014.11.030
11.
White JR, Tai A, Arthur D, et al. Breast Cancer Atlas for Radiation Therapy Planning: Consensus Definitions. https://www.rtog.org/LinkClick.aspx?fileticket=vzJFhPaBipE%3D&tabid=236.
12.
Grégoire V, Ang K, Budach W, et al. Delineation of the neck node levels for head and neck tumors: A 2013 update. DAHANCA, EORTC, HKNPCSG, NCIC CTG, NCRI, RTOG, TROG consensus guidelines. Radiother Oncol. 2014. doi:10.1016/j.radonc.2013.10.010
13.
Jing H, Wang SL, Li J, et al. Mapping patterns of ipsilateral supraclavicular nodal metastases in breast cancer: Rethinking the clinical target volume for high-risk patients. Int J Radiat Oncol Biol Phys. 2015. doi:10.1016/j.ijrobp.2015.08.022
14.
Brown LC, Diehn FE, Boughey JC, et al. Delineation of Supraclavicular Target Volumes in Breast Cancer Radiation Therapy. Radiat Oncol Biol. 2015. doi:10.1016/j.ijrobp.2015.02.022
15.
DeSelm C, Yang TJ, Cahlon O, et al. A 3-Dimensional Mapping Analysis of Regional Nodal
Recurrences in Breast Cancer. Int J Radiat Oncol Biol Phys. 2019;103(3):583-591. doi:10.1016/j.ijrobp.2018.10.021 16.
Borm KJ, Voppichler J, Düsberg M, et al. FDG / PET-CT based lymph node atlas in breast cancer patients. Int J Radiat Oncol. 2018;0(0). doi:10.1016/j.ijrobp.2018.07.2025
17.
Jethwa KR, Kahila MM, Hunt KN, et al. Delineation of Internal Mammary Nodal Target Volumes in Breast Cancer Radiation Therapy. Int J Radiat Oncol. 2017;97(4):762-769. doi:10.1016/J.IJROBP.2016.11.037
Table I. Patient population and nodal distribution. Locally advanced
Distant metastatic
Patients (no.)
50
51
Age (median)
57
55
FDG avid nodes
Median
Range
Total
Median
Range
Total
2
0-8
112
4
0-19
277
Axilla- level 1
1
0-5
54
3
0-9
108
Axilla- level 2
1
0-4
24
2
0-7
62
Axilla- level 3
1
0-2
10
1
0-5
26
Supraclavicular
1
0-3
10
2
0-6
50
Internal mammary
1
0-1
14
2
0-4
31
Total Regional distribution
Table II. Locoregional lymph nodes covered by the 95% isodose line, left lung V20 (%) and mean heart dose (cGy) for 3D-CRT, RTOG, ESTRO and RADCOMP VMAT and PBSPT plans.
Locoregional nodal coverage by 95% isodose line
Total nodes
All nodes n (%)
Level 1 Ax n (%)
Level 2 Ax n (%)
Level 3 Ax n (%)
SCV n (%)
IMN n (%)
389
162
86
36
60
45
Left lung V20 (%)
Heart mean (cGy)
Plan
Technique
3D-CRT
PWT/APPA SCV
379 (97%)
161 (99%)
86 (100%)
36 (100%)
56 (93%)
40 (89%)
RTOG
VMAT
362 (93%)
160 (99%)
86 (100%)
33 (92%)
45 (75%)
38 (84%)
16.7%
853
RTOG
PBSPT
354 (91%)
160 (99%)
81 (94%)
31 (86%)
43 (72%)
39 (87%)
10%
204
ESTRO
VMAT
347 (89%)
162 (100%)
86 (100%)
28 (78%)
31 (52%)
40 (89%)
15.4%
831
ESTRO
PBSPT
344 (88%)
162 (100%)
81 (94%)
28 (78%)
33 (55%)
40 (89%)
10.5%
206
RADCOMP
VMAT
381 (98%)
162 (100%)
86 (100%)
36 (100%)
57 (95%)
40 (89%)
16.6%
950
RADCOMP
PBSPT
381 (96%)
162 (100%)
86 (100%)
36 (100%)
57 (95%)
40 (89%)
12.1%
259
S1879-8500(19)30359-5
DOI:
https://doi.org/10.1016/j.prro.2019.11.010
Reference:
PRRO 1163
To appear in:
Practical Radiation Oncology
Received Date: 16 May 2019 Revised Date:
27 October 2019
Accepted Date: 14 November 2019
Please cite this article as: Kowalski ES, Feigenberg SJ, Cohen J, Fellows Z, Vadnais P, Rice S, Mishra MV, Molitoris JK, Nichols EM, Snider JW III, Optimal Target Delineation and Treatment Techniques in the Era of Conformal Photon and Proton Breast and Regional Nodal Irradiation, Practical Radiation Oncology (2019), doi: https://doi.org/10.1016/j.prro.2019.11.010. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Inc. on behalf of American Society for Radiation Oncology.
Title: Optimal Target Delineation and Treatment Techniques in the Era of Conformal Photon and Proton Breast and Regional Nodal Irradiation Short Title: Breast Nodal Optimal Target Delineation Authors: Emily S. Kowalski MD1, Steven J. Feigenberg MD3, Justin Cohen MD1, Zachary Fellows CMD1, Patrick Vadnais CMD1, Stephanie Rice MD1, Mark V. Mishra MD2, Jason K. Molitoris MD, PhD2, Elizabeth M. Nichols, MD2, James W. Snider III MD2
1
Department of Radiation Oncology, University of Maryland Medical Center, Baltimore, Maryland
2
Department of Radiation Oncology, University of Maryland School of Medicine, Baltimore, Maryland
3
Department of Radiation Oncology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania
Corresponding Author: James W. Snider III, MD Assistant Professor Department of Radiation Oncology University of Maryland School of Medicine Baltimore, Maryland 21201 Email: [email protected] Phone: 410-369-5239
Conflict of Interest: Mark Mishra MD has received honoraria and travel funding from Varian Medical Systems, which was not used for this project. James Snider III, MD has received honoraria and speakers bureau from Varian Medical Systems and a consulting award from Siemens Healthineers, neither of which were used for this project. Dr. Snider also has a patent pending for ProtonGRID Radiotherapy Technique, which is unrelated to this project.
Outside funding: No outside funding was used in the development of this manuscript
Title: Optimal Target Delineation and Treatment Techniques in the Era of Conformal Photon and Proton Breast and Regional Nodal Irradiation Short Title: Breast Nodal Optimal Target Delineation Abstract Purpose: Regional nodal irradiation (RNI) improves disease-free and distant disease–free survival in patients with high-risk breast cancer (BC). Trials demonstrating this utilized 2- or 3-dimensional conformal radiotherapy (2D or 3D CRT) fields based on bony anatomy. Modern volumetric-modulated arc therapy (VMAT) and pencil beam scanning proton therapy (PBSPT) may underdose regional nodes (RNs) not contoured but covered by 3D CRT. Multiple atlases guide modern treatment planning. This study addresses the risk of underdosing following published atlases using 3D CRT, VMAT and PBSPT. Methods and materials: Targets per the Radiation Therapy Oncology Group (RTOG), European Society for Radiotherapy and Oncology (ESTRO), and Radiotherapy Comparative Effectiveness Consortium (RADCOMP) atlases were contoured on a representative patient CT scan. 3D CRT plans based on anatomic borders and VMAT and PBSPT plans for each set of target volumes were generated. PET/CT scans were reviewed. CT-positive and 18F-FDG–avid RNs (n = 389) were mapped from 102 patients with locally advanced (n = 51; median 2; range, 1–8 nodes) and metastatic (n = 51; median 4; range, 1–19 nodes) BC: axillary (AX; n = 284), supraclavicular (SCV; n = 60), and internal mammary nodal (IMN; n = 45). 18F-FDG–avid RNs falling within the 95% isodose line were considered adequately covered. Results: 3D CRT plans provided excellent RN coverage. Low AX nodes were covered (≥99%) in all plans. Underdosing of 18F-FDG–avid RNs falling in the high AX (78%-92%), SCV (52%-75%), and IMN (84%-89%) volumes was observed following the RTOG and ESTRO atlases for VMAT and PBSPT plans. Use of the RADCOMP atlas provided coverage of these areas (89%-100%) with slightly increased heart and lung doses. Atlas guided VMAT/PBSPT plans provided cumulative nodal coverage as follows: ESTRO (89%/88%), RTOG (93%/91%), and RADCOMP (98%/96%).
Conclusions: VMAT and PBSPT for RNI in patients with high risk BC risks underdosage in the high AX, SCV, and IMN nodal regions unless comprehensive target delineation is performed.
Introduction In women with breast cancer (BC), the addition of regional nodal irradiation (RNI), in particular internal mammary nodal (IMN) radiation, has been controversial. However, the results of 2 recently reported randomized trials have reinvigorated the “believers” support of RNI in patients with <4 lymph nodes as well as high-risk node-negative women. Whelan et al. (MA.20) demonstrated improvement in diseasefree and distant disease–free survival with the addition of RNI in node-positive or high-risk node-negative women.1 Poortmans and co-investigators (EORTC 22922) echoed this benefit in node-positive patients and patients who were node-negative with centrally located tumors.2 For many radiation oncologists, the results from these trials lowered the threshold for considering RNI. Women registered in the MA.20 and EORTC 22922 trials were treated with traditional 2- or 3-dimensional conformal radiotherapy (2D or 3D CRT). These trials required quality assurance review, specified radiation fields, permissible delivery techniques, and acceptable IMN coverage, thereby assuring relative uniformity in planning and target coverage for patients across institutions.3 2D or 3D CRT plans treat larger volumes of normal tissue to full dose and often result in dose heterogeneity and hot spots that can have significant effects on critical structures as well as on cosmesis/breast pain.4 The Cambridge experience of >1,000 women treated employed field-in-field modulation versus standard wedged tangential fields and demonstrated that efforts to improve breast dose homogeneity improved 5-year cosmesis and while reducing late toxicity.4 In an attempt to improve conformality especially of the high dose regions, many radiation oncologists have opted for more sophisticated treatment techniques. Multifield, inverse-planned IMRT and VMAT often increase high isodose line conformity, at the expense of somewhat increased low dose volume. Perhaps, the most critical normal structure in this disease site is the heart, where radiation exposure can increase the risk of complication at even low to moderate doses. Darby et al. demonstrated a linear 7.4% relative increase in the rate of cardiac events per each mean Gray delivered to the heart.5 RT remains crucial to the cure of locally advanced BC; however, increased cardiovascular risks may negate some of these benefits if not properly mitigated.
Proton therapy is an attractive alternative method of delivering RNI. Protons deliver most of their energy within the target and substantially less dose beyond the target based on the Bragg peak energy deposition, often achieving mean heart doses ≤1 Gy.6 The clinical benefit of these dose reductions is being tested in the Radiotherapy Comparative Effectiveness Consortium (RADCOMP) trial. As the field moves toward treating BC with modern conformal techniques, the question remains as to whether nodal failure patterns once covered by the historical 2D fields are adequately covered with inverse-planned IMRT, VMAT and proton therapy. Conformal techniques require careful delineation of regions involved or at risk of harboring microscopic disease. Several contouring atlases have been developed by consensus groups to guide physicians. These are critical in deciding how to target these nodal regions; however, they differ in recommended volumes. We hypothesize that this variability in nodal coverage could lead to marginal failures as a result of underdosing lymph node regions when more conformal techniques are utilized. To test this hypothesis, we first identified nodal areas at risk using 18FFDG PET/CT, for which literature supports superior sensitivity for detecting extra-axillary lymph nodes and distant metastases compared with conventional imaging in the locally advanced setting.7-8
Methods and Materials Patient characteristics On an institutional review board–approved protocol, we retrospectively reviewed patients seen from 2009 to 2017 with an ICD 10 diagnosis code of BC, who had also undergone 18F-FDG PET/CT. Patients whose image sets were no longer available for review on our server were excluded. A total of 102 patient records were available for analysis. Patients with metastatic disease were included. Contouring per the atlas A representative patient who had undergone CT simulation with a diagnosis of left-sided BC was selected for contouring. The clinical target volumes (CTVs) were contoured by a single radiation oncologist according to 3 published BC atlases that have been central to current/previous large phase II/III studies: (1) the breast contouring atlas utilized in the RADCOMP trial9; (2) the European Society for Radiation and Oncology (ESTRO) consensus guidelines on target volume delineation for elective
radiation therapy of early-stage BC10; and (3) the Radiation Therapy Oncology Group (RTOG) atlas.11 A summary of key anatomic boundaries of lymph node volumes for each atlas is provided in the Supplement, Table I. The contouring physician was blinded to the locations of lymph nodes 18F-FDG avid on PET/CT to avoid contouring bias. Mapping PET/CT images and corresponding radiologists’ reports were reviewed by a second radiation oncologist. Lymph nodes were designated as positive if so indicated by the radiology report and/or clearly 18
F-FDG avid on PET/CT when the report was less specific. Nodes determined to be clinically positive
were mapped by the second radiation oncologist onto the representative patient utilizing anatomic landmarks in the treatment planning system. All lymph nodes were plotted on the left side of the representative patient to allow for uniform planning and analysis. A 5-mm diameter circle was placed at the epicenter of each lymph node–correlated location (Fig. 1). Planning Two dosimetrists utilized the RayStation (RaySearch Laboratories) planning system to create a 3D CRT plan based on anatomical landmarks as well as VMAT and pencil-beam scanning proton (PBSPT) plans for each atlas-based set of volumes contoured. The 3D CRT plan utilized 2 partially wide mixed energy 6 MV and 18 MV tangent beams to cover the breast and upper 3 intercostal spaces of the IMN chain matched to 2 parallel opposed 18 MV beams to treat the supraclavicular (SCV) fossa. Fields were weighted for homogeneity across the target. Contoured atlas-based volumes were utilized for inversely optimized VMAT plans, which consisted of 3 partial arcs. Planning target volumes (PTVs) for VMAT included a 5-mm expansion of the clinical target volume, removed 5mm from the skin’s surface. The proton plan included 2 anterior en face beams with a 35-degree hinge angle. A beam-specific pPTV of 5 mm was generated. Dosimetrists were instructed to achieve a conformal plan with near complete coverage of the target volume with at least 95% of the prescribed dose (Fig. 2). Dosimetrists were not informed of target coverage or organ at risk sparing achieved by competing plans, again to prevent bias.
Dosimetric analysis Plans were reviewed and approved as adequate by the first, contouring radiation oncologist. 18FFDG–avid lymph nodes were designated by the second, mapping radiation oncologist as adequately covered if the entire 5-mm circle fell within the 95% isodose line.
Results A total of 389 18F-FDG–avid lymph nodes were mapped from 102 patients with locally advanced (n = 51; median, 2; range, 1–8 nodes) and metastatic (n = 51; median, 4; range, 1–19 nodes) BC: axillary (AX; n = 284), SCV (n = 60), and internal mammary nodal (IMN; n = 45) (Table II). Patients with metastatic disease, on average, were found to have more evident lymph nodes per person than patients with locally advanced disease (Table I). 18F-FDG–avid lymph nodes were most frequently found in AX levels 1–3 (n = 284; 73%), followed by the SCV (n = 60; 15%), and the IMN (n = 45; 12%) regions. The 3D CRT plan provided excellent coverage (89%–100%) across all nodal levels (Table II). When analyzed by individual nodal regions, clear differences in coverage of 18F-FDG–avid lymph nodes were evident based on the contouring conventions. Level I–III axilla Level I and II AX lymph nodes were well covered in all plans (94%–99%; Table II). 18F-FDG– avid lymph nodes located in the junction between the SCV and level III volumes (the infraclavicular fossa) and those in the level III volume proper were frequently undercovered by both the VMAT and PBSPT plans using the ESTRO- and RTOG-based treatment volumes (Fig. 3A). Most of these undercovered nodes were found in the “posterior neck,” posterolateral to the traditional level III, diving deep toward the rotator cuff musculature and abutting the coracoid process and scapula (Fig. 3A). SVC fossa Mapped lymph nodes in the SCV fossa were also undercovered by the RTOG (75%, 72%)- and ESTRO (52%, 55%)- informed VMAT and PBSPT plans, respectively (Table II). First, multiple 18FFDG–avid lymph nodes were seen posterior to the sternocleidomastoid muscle, in the posterior neck, similar to region of level Vb as defined by head and neck RT volumes (Fig. 3B).12 Second, undercovered
18
F-FDG–avid nodes were found at the inferior edge of SCV volumes near the junction of the SCV fossa
with the upper mediastinum and IMNs. These nodes were clustered just superior to the attachment of the first rib to the sternum, posterior to the 95% isodose line around the common carotid artery, and at the separation of the brachiocephalic vein into the internal jugular and subclavian vein (Fig. 3C). Third, nodes were found at the level of the thyroid cartilage, superior to the delineation extent of all 3 atlases (Fig. 4B). In 2 patients, undercovered nodes were found in the substernal space and adjacent to the medial edge of the carotid. IMNs All 3 atlases denote the inferior extent of the third intercostal space as the border of the IMN volume. ESTRO guidelines indicate the fourth intercostal space may be included when primary tumors are located in the lower inner quadrant. In patients with metastatic disease, 18F-FDG–avid lymph nodes were found outside the 95% isodose line in the fourth (n=1) and fifth (n=1) intercostal spaces (Fig. 4A). Overall, the population contained a greater than expected number of IMNs (n = 45; 12%). The RADCOMP-based VMAT and PBSPT plans provided improved coverage of each of the aforementioned nodal volumes. 18F-FDG–avid nodes in levels 1–3 of the axilla were covered 100% in both the VMAT and PBSPT plans (Table II). 18F-FDG–avid nodes in the SCV fossa (95%/95%) and IMN (89%/89%) regions were also better covered by both plans (Table II). As with the ESTRO and RTOG atlases, the RADCOMP atlas does not extend superiorly to the thyroid cartilage, and undercoverage was encountered in this area in advanced cases.
Discussion The use of advanced RT techniques allows for conformal RT plans to complex targets, improved dose distributions in patients with challenging anatomy and lower doses to nearby organs at risk. As opposed to 2D and 3D techniques, which rely on bony landmarks for field design, VMAT and PBSPT treatment plans are highly dependent on target contouring. Multiple competing atlases guide radiation oncologists in contouring these volumes. We, along with previous investigators,16-20 have demonstrated that most of these atlases fail to incorporate several areas at risk for nodal metastases based on retrospective review of
PET/CT examinations in BC patients. This is, to our knowledge, the first effort to demonstrate the differential dosimetric impact of these flaws in target delineation based on treatment technique and is the first analysis of PBSPT in this setting. The inclusion of the RADCOMP atlas as a more comprehensive alternative is also unique. In our study, metabolically active lymph nodes undercovered by VMAT and PBSPT techniques were most frequently found in the following locations: cranial to the cricoid cartilage, in the posterolateral neck, at the junction of the IMN and SCV volumes, and in the fourth and fifth intercostal spaces. Our 3D CRT technique was unintentionally (not based on contouring) comprehensive in most of these areas of undercoverage. This highlights the caveat that comes with VMAT and PBSPT radiotherapy techniques. As dose distributions become more conformal to target volumes, radiation oncologists must be confident that areas at risk for harboring gross or microscopic disease are accurately delineated. On the other hand, overcontouring can also lead to increased organ-at-risk exposure. Therefore, a balance must be struck. Published reports have retrospectively evaluated the RTOG and ESTRO atlases for comprehensiveness of target volumes and found similar areas of undercoverage. Three modern efforts evaluated the delineation of SCV lymph node channels as provided by various atlases. Jing et al. mapped the location of 524 metastatic SCV lymph nodes and concluded that only 63% were covered by the RTOG atlas volumes.13 Pathologic nodes were found lateral to the sternocleidomastoid and scalene muscles in 80% of their population. In line with our observations, Jing et al also found a significant proportion (55.3%) of their patient population exhibited pathologic nodes above the base of the cricoid cartilage.13 Brown and colleagues reported that most patients with pathologic nodes in the cranial SCV volume, approaching the thyroid cartilage, had multiple additional SCV metastases, suggesting that this region is most at risk in patients with advanced nodal disease.y14 Also consistent with the current series, Jing et al. and Brown et al. found involved lymph nodes in the posterolateral neck outside of the RTOG volume. In a population of patients with recurrent BC, DeSelm et al. found that a majority of nodal
recurrences outside the RTOG and ESTRO CTVs within the SCV region were located in the fat spaces lateral to the sternocleinomastoid muscle and posterior to the transverse process.15 Most recently Borm and co-investigators reviewed 601 pathologic lymph nodes in 235 BC patients and evaluated the coverage provided by both the RTOG and ESTRO atlases.16 Patients with distant metastases were much more likely to have IMN and SCV fossa involvement. Overall, they found 17% and 21.2% of lymph node metastases were outside the RTOG and ESTRO guided boundaries, respectively. The volumes of covered lymph nodes were 69.8% ± 35.5% and 69.1% ± 36.3% for primary BC and 57.6% ± 38.9% and 51.1% ± 39.1% for recurrent BC using the RTOG and ESTRO guidelines, respectively. They identified many similar areas of undercoverage including the posterior SCV volume and the border between the SCV and IMN volumes.16 Jethwa et al. investigated RTOG atlas coverage of metastatic IMNs in 67 patients. Similar to our findings, they found that 78% of metastatic nodes fell within the first 3 intercostal spaces in a cranial caudal dimension; however, when also taking into consideration lateral expansions, 47% of metastatic nodes fell outside the RTOG atlas recommended volumes. Patients in this study found to have pathologic nodes caudal to the fourth rib (14%) typically exhibited additional pathologic nodes in the first 3 interspaces. Consistent with our findings, metastatic nodes (8%) were found cranial to the first intercostal space but caudal to the confluence of the IM, brachiocephalic, jugular, and subclavian veins.17 In our series, we utilized the RADCOMP atlas as a more comprehensive comparator. This atlas was informed by the above data and, therefore, exhibits more extensive volumes in the areas of concern. The RADCOMP atlas recommends a seamless junction between the SCV and IMN volumes and extends the IMN volume medially to include the fat space. It exhibits cranial SCV and inferior level I borders similar to those of the other 2 atlases but also extends the posterolateral border in the low neck near the junction of the infraclavicular and SCV volumes. Use of this atlas led to much better dosimetric coverage of the PET/CT positive lymph nodes in our series when utilizing VMAT and PBSPT techniques. The improved coverage provided by the RADCOMP atlas came at the expense of increased mean heart (950 cGy, VMAT) and ipsilateral lung V20 (16.6%, VMAT) exposure. Mean heart dose and
ipsilateral lung V20 with the RADCOMP atlas informed plans were only mildly increased over those of the RTOG and ESTRO atlases (Table II). When RADCOMP volumes were planned with PBSPT, these doses, as expected, decreased significantly (Table II). Of note, organs at risk were not prioritized by the dosimetrists during planning because this could have introduced confounders based on the representative patient’s anatomy or the planner’s aggressiveness in sparing these organs at risk. Poortmans et al.2 and Whelan et al.1 found no statistical difference in the rates of cardiac events and low rates of pneumonitis and pulmonary fibrosis when randomizing women to nodal irradiation versus breast/chest wall irradiation alone. Although these trials did not publish the organ at risk doses, the difference between heart and lung dose in patients treated with or without nodal irradiation is undoubtedly greater than the discrepancy encountered here based on one atlas-driven plan versus another. We present cardiac and lung doses only to highlight that the more comprehensive coverage provided by RADCOMP volumes does not come at a substantial cost to organs at risk versus the less comprehensive RTOG and ESTRO volumes. Therefore, we predict differential lung and heart toxicities in patients treated with RADCOMP-generated verses RTOG/ESTRO–generated plans to be minimal, if any. It is important to note that these plans were generated without dose-mitigating approaches such as deep inspiratory breath hold, which is regularly utilized in many photon/proton clinics to minimize cardiac dose when treating left-sided breast cancer. The planning directive to achieve 95% isodose coverage of the entire IMN volume likely increased the heart and lung doses above those that would be acceptable or normally encountered in the clinic. Many protocols allow for less (~80%) isodose coverage of this region due to the concern for increased cardiac toxicity, in particular. As a result of these directives and lack of breath-hold, the 3D CRT plans resulted in higher than clinically encountered doses to heart and lung; to avoid misrepresentation of clinically acceptable parameters, these were not presented. This study has several limitations. It is retrospective in nature and included patients with lymph nodes that were sufficiently large to be seen on 18F-FDG PET/CT. This resulted in a selection bias favoring patients with bulkier, locally advanced disease. The included patients with distant metastatic disease had twice as many 18F-FDG–avid lymph nodes as those with locally advanced disease, which led
to improved power of the analysis. However, it is likely that nodal patterns encountered in the patients in this study, while emblematic of potential regions of failure, are not representative of most early stage or minimally locally advanced patients. For earlier stage patients with nodal disease uncovered on sentinel lymph node biopsy, smaller treatment volumes may be adequate. This remains a question for prospective trials, but this study does demonstrate that modern techniques have become sufficiently conformal to “miss” areas not intentionally targeted that could be at risk. Additional limitations include that metabolically active nodes were presumed malignant without confirmatory biopsy. Radiologists reviewed the initial PET/CTs, and their reports were utilized to guide selection of clinically positive nodes. Transferring of the nodal positions onto the standard patient was performed manually by a radiation oncologist based on anatomic landmarks and represent approximations. Deformable registration was not utilized because of the significant differences in position (arm up versus arm down) for most scans. This effort is the first, to our knowledge, to demonstrate the dosimetric impact of various contouring conventions/atlases in the treatment of locally advanced BC. It highlights the critical importance of accurate delineation and contouring when shifting toward VMAT and PBSPT radiotherapy techniques. Several previous efforts have evaluated the RTOG and ESTRO atlases for under-contouring various regions of the AX, SCV, and IM chains. However, this is the first effort to test these formally against the more comprehensive RADCOMP atlas and to evaluate PBSPT as the most conformal modality currently being employed widely. Based on these results, we recommend utilization of the RADCOMP atlas over RTOG and ESTRO when using VMAT and PBSPT techniques to treat patients with locally advanced disease with a potential for high burden of involved nodes. This ensures coverage of at-risk nodal regions in a manner more consistent with 2D and 3D CRT techniques.
Conclusions The use of VMAT and PBSPT radiotherapy techniques in the treatment of locally advanced breast cancer should be accompanied by careful delineation of target volumes. A bias towards
more comprehensive nodal targets is recommended in more advanced cases. Without this, VMAT and PBSPT techniques may lead to underdosage of regions at risk that would have been covered with 2D or 3D CRT approaches.
Fig. 1. Digitally reconstructed radiograph. Lymph nodes 18F-FDG–avid on PET/CT in locally advanced (blue) and distant metastatic (red) patients. Fig. 2. Radiation plans. (A) 3D CRT; (B) VMAT; (C) PBSPT. VMAT and PBSPT plan utilized RADCOMP atlas volumes. Isodose lines: 105% = orange, 100% = red, 95% = pink, 70% = green, 50% = dark blue, 30% = light blue, 10% = lavender. Fig. 3. 18F-FDG–nodal coverage by atlas-driven VMAT and PBSPT plans. Level III axilla. (A). SCV (B). IMN SCV junction (C). Left to right: RTOG, ESTRO, and RADCOMP atlas guidance. Top row: VMAT; bottom row: PBSPT. Cyan = 95%, Green=50%, Lavender=30% isodose line. Fig. 4. 18F-FDG–avid nodes uncovered in all VMAT and PBSPT plans. IMNs in the fourth intercostal space (A.) and cranial SCV (B). Left to right: RTOG, ESTRO, and RADCOMP atlas guidance. Top row: VMAT; bottom row: PBSPT. Cyan = 95%, Green=50%, Lavender=30% isodose line.
References 1.
Whelan TJ, Olivotto I a., Parulekar WR, et al. Regional Nodal Irradiation in Early-Stage Breast Cancer. N Engl J Med. 2015;373(4):307-316. doi:10.1056/NEJMoa1415340
2.
Poortmans PM, Collette S, Kirkove C, et al. Internal Mammary and Medial Supraclavicular Irradiation in Breast Cancer. N Engl J Med. 2015;373(4):317-327. doi:10.1056/NEJMoa1415369
3.
Musat E, Poortmans P, den Bogaert W Van, et al. Quality assurance in breast cancer: EORTC experiences in the phase III trial on irradiation of the internal mammary nodes. Eur J Cancer. 2007;43(4):718-724. doi:10.1016/j.ejca.2006.11.015
4.
Mukesh MB, Barnett GC, Wilkinson JS, et al. Randomized controlled trial of intensity-modulated radiotherapy for early breast cancer: 5-year results confirm superior overall cosmesis. J Clin Oncol. 2013. doi:10.1200/JCO.2013.49.7842
5.
Darby SC, Ewertz M, McGale P, et al. Risk of Ischemic Heart Disease in Women after Radiotherapy for Breast Cancer. N Engl J Med. 2013;368(11):987-998. doi:10.1056/NEJMoa1209825
6.
MacDonald SM, Jimenez R, Paetzold P, et al. Proton radiotherapy for chest wall and regional lymphatic radiation; dose comparisons and treatment delivery. Radiat Oncol. 2013. doi:10.1186/1748-717X-8-71
7.
Groheux D, Giacchetti S, Delord M, et al. 18F-FDG PET/CT in staging patients with locally advanced or inflammatory breast cancer: comparison to conventional staging. J Nucl Med. 2013;54(1):5-11. doi:10.2967/jnumed.112.106864
8.
Ulaner GA, Castillo R, Goldman DA, et al. 18F-FDG-PET/CT for systemic staging of newly diagnosed triple-negative breast cancer. Eur J Nucl Med Mol Imaging. 2016;43(11):1937-1944. doi:10.1007/s00259-016-3402-9
9.
Macdonald, Shannon M; Cahlon O. Breast Atlas (RADCOMP).Pdf.; 2016. https://www.rtog.org/CoreLab/ContouringAtlases/RADCOMPBreastAtlas.aspx.
10.
Offersen B V., Boersma LJ, Kirkove C, et al. ESTRO consensus guideline on target volume delineation for elective radiation therapy of early stage breast cancer. Radiother Oncol J Eur Soc Ther Radiol Oncol. 2015;114(1):3-10. doi:10.1016/j.radonc.2014.11.030
11.
White JR, Tai A, Arthur D, et al. Breast Cancer Atlas for Radiation Therapy Planning: Consensus Definitions. https://www.rtog.org/LinkClick.aspx?fileticket=vzJFhPaBipE%3D&tabid=236.
12.
Grégoire V, Ang K, Budach W, et al. Delineation of the neck node levels for head and neck tumors: A 2013 update. DAHANCA, EORTC, HKNPCSG, NCIC CTG, NCRI, RTOG, TROG consensus guidelines. Radiother Oncol. 2014. doi:10.1016/j.radonc.2013.10.010
13.
Jing H, Wang SL, Li J, et al. Mapping patterns of ipsilateral supraclavicular nodal metastases in breast cancer: Rethinking the clinical target volume for high-risk patients. Int J Radiat Oncol Biol Phys. 2015. doi:10.1016/j.ijrobp.2015.08.022
14.
Brown LC, Diehn FE, Boughey JC, et al. Delineation of Supraclavicular Target Volumes in Breast Cancer Radiation Therapy. Radiat Oncol Biol. 2015. doi:10.1016/j.ijrobp.2015.02.022
15.
DeSelm C, Yang TJ, Cahlon O, et al. A 3-Dimensional Mapping Analysis of Regional Nodal
Recurrences in Breast Cancer. Int J Radiat Oncol Biol Phys. 2019;103(3):583-591. doi:10.1016/j.ijrobp.2018.10.021 16.
Borm KJ, Voppichler J, Düsberg M, et al. FDG / PET-CT based lymph node atlas in breast cancer patients. Int J Radiat Oncol. 2018;0(0). doi:10.1016/j.ijrobp.2018.07.2025
17.
Jethwa KR, Kahila MM, Hunt KN, et al. Delineation of Internal Mammary Nodal Target Volumes in Breast Cancer Radiation Therapy. Int J Radiat Oncol. 2017;97(4):762-769. doi:10.1016/J.IJROBP.2016.11.037
Table I. Patient population and nodal distribution. Locally advanced
Distant metastatic
Patients (no.)
50
51
Age (median)
57
55
FDG avid nodes
Median
Range
Total
Median
Range
Total
2
0-8
112
4
0-19
277
Axilla- level 1
1
0-5
54
3
0-9
108
Axilla- level 2
1
0-4
24
2
0-7
62
Axilla- level 3
1
0-2
10
1
0-5
26
Supraclavicular
1
0-3
10
2
0-6
50
Internal mammary
1
0-1
14
2
0-4
31
Total Regional distribution
Table II. Locoregional lymph nodes covered by the 95% isodose line, left lung V20 (%) and mean heart dose (cGy) for 3D-CRT, RTOG, ESTRO and RADCOMP VMAT and PBSPT plans.
Locoregional nodal coverage by 95% isodose line
Total nodes
All nodes n (%)
Level 1 Ax n (%)
Level 2 Ax n (%)
Level 3 Ax n (%)
SCV n (%)
IMN n (%)
389
162
86
36
60
45
Left lung V20 (%)
Heart mean (cGy)
Plan
Technique
3D-CRT
PWT/APPA SCV
379 (97%)
161 (99%)
86 (100%)
36 (100%)
56 (93%)
40 (89%)
RTOG
VMAT
362 (93%)
160 (99%)
86 (100%)
33 (92%)
45 (75%)
38 (84%)
16.7%
853
RTOG
PBSPT
354 (91%)
160 (99%)
81 (94%)
31 (86%)
43 (72%)
39 (87%)
10%
204
ESTRO
VMAT
347 (89%)
162 (100%)
86 (100%)
28 (78%)
31 (52%)
40 (89%)
15.4%
831
ESTRO
PBSPT
344 (88%)
162 (100%)
81 (94%)
28 (78%)
33 (55%)
40 (89%)
10.5%
206
RADCOMP
VMAT
381 (98%)
162 (100%)
86 (100%)
36 (100%)
57 (95%)
40 (89%)
16.6%
950
RADCOMP
PBSPT
381 (96%)
162 (100%)
86 (100%)
36 (100%)
57 (95%)
40 (89%)
12.1%
259