Abstract
Background/Aim: Accelerated partial breast irradiation (APBI) is an alternative to whole-breast irradiation in early-stage breast cancer. This study evaluated the setup accuracy of a workflow without skin markers that uses surface-guided radiotherapy (SGRT) in conjunction with clip-based alignment in patients undergoing APBI.
Patients and Methods: This study recruited 35 patients who underwent APBI after breast-conserving surgery. Treatment plans were generated with 30 Gy in five fractions. During the treatment period, patients were positioned using AlignRT, intentionally omitting the skin marks. After the initial setup, the position was verified via daily kV images or cone-beam computed tomography, with matching surgical clips serving as the basis. Translational and rotational shifts were recorded, along with the monitoring-on time by AlignRT.
Results: A total of 175 treatment fractions were analyzed. The mean±standard deviation (SD) residual setup error detected via image registration was 0.01±0.18, −0.09±0.22, and −0.04±0.19 cm in the vertical, longitudinal, and lateral axes, respectively. Setup accuracy within 5 mm in all axes was achieved in over 95% of the treatment fractions assessed. During the treatment period, 28% of patients (10 out of 35) maintained position deviations of less than 3 mm in the 3D vector direction. The mean±SD monitoring-on time was 603.3±214.1 s (range=349-1,353 s).
Conclusion: The integration of surface-guided radiation therapy and clip alignment effectively achieved accurate and efficient patient positioning; this can serve as an alternative to traditional skin markers in the external beam APBI workflow.
- Breast cancer
- accelerated partial breast irradiation
- intensity modulated radiation therapy
- surface-guided radiotherapy
Introduction
Accelerated partial breast irradiation (APBI) has emerged as a promising alternative to conventional whole-breast irradiation in select patients with early-stage breast cancer. The advantages of APBI include a shorter treatment duration, less toxicity to surrounding healthy tissues, and better cosmetic results (1-3). Moreover, the dosimetric advantages of APBI in combination with volumetric-modulated arc therapy (VMAT) have been demonstrated in recent clinical trials (4, 5).
The success of APBI is highly dependent on accurate radiation delivery, which requires an accurate and reproducible patient setup. However, conventional patient setup methods using skin marks or tattoos are subject to variability between patients, days, and therapists, thereby increasing patient burden in terms of time and radiation exposure. Recently, surface-guided radiation therapy (SGRT) has been increasingly adopted in clinical practice as an exposure-free, real-time patient positioning approach, since it can achieve submillimeter accuracy using 3D surface imaging. In addition, recent reports have discussed the usefulness of setups, which do not use skin marks or tattoos for various anatomical sites (6-8). However, despite the widespread use of SGRT, only a few studies have discussed its use in APBI (9-13); no reports have evaluated the accuracy of the SGRT setup in patients subjected to APBI without markings.
The residual setup error for SGRT and tattoos in patients with breast cancer ranges from 1.8 to 6 mm and 2.4 to 14 mm, respectively, with studies reporting improved accuracy in the former (14). However, SGRT depends on body surface reproducibility on a daily basis, and errors usually arise due to deformation, swelling, and respiratory movement during the treatment period. In clinical practice, patients with an obese physique or large breasts often cannot reproduce the position of the entire body, not just the body surface around the breast (15). Thus, in populations with various physical characteristics, the workflow and performance of the SGRT system may differ, which could affect the accuracy and reproducibility of the radiotherapy setup. However, SGRT may be a more powerful tool in patients with smaller body physiques and breast sizes who can more effectively benefit from its full potential.
Meanwhile, surgical clips inserted during breast-conserving surgery can serve as reliable internal landmarks as targets for radiation therapy. Previous reports have shown that fiducial marker-based imageguided radiotherapy (IGRT) improves interfractional setup accuracy (13, 16, 17). Furthermore, surgical clips improve delineation accuracy of the tumor bed since they are sutured directly around the surgical cavity. Weed et al. reported that surgical clips serve as reliable surrogates of the surgical cavity (18).
We hypothesized that SGRT could be integrated with clip-based alignment to achieve an efficient treatment workflow by combining the advantages of noninvasive surface imaging with the precision of in-body target localization, thereby eliminating the need for traditional skin marks. To test this hypothesis, this study evaluated the setup accuracy of a workflow that uses SGRT in conjunction with clip-based alignment instead of skin marks. The impact of this approach on treatment time and its effectiveness in clinical practice was also assessed.
Patients and Methods
Study population. This retrospective study analyzed 35 consecutive Japanese patients undergoing APBI who were treated for early-stage breast cancer (n=18, right breast; n=17, left breast) at our institution from July 2024 to December 2024. The study protocol was approved by our institutional review board (24-R060).
Delineation of target volumes and organs at risk. Computed tomography (CT) simulation was conducted in the supine position with the patient’s arms raised above the head through wing-board immobilization (CIVCO Radiotherapy, Orange City, IA, USA). Patients were instructed to breathe freely during the CT scan. CT images with a slice thickness of 2.0-mm were obtained using the SOMATOM Confidence RT Pro (Siemens Healthcare, Erlangen, Germany). Skin marks were not used on patients during the CT scans.
An experienced radiation oncologist (RO) determined the clinical target volume (CTV) using surgical clips as surrogates to include the pathological extent of the tumor plus 2 cm around it. The planning target volume (PTV) was defined by adding a 5-mm margin to the CTV. In compliance with American Society for Radiation Oncology (ASTRO) guidelines, a volume for dose evaluation of PTV (PTV_Eval) was defined as the PTV cropped 3 mm inside patient surface and limiting posteriorly by the pectoralis muscle (19). The organs at risk (OARs), including the heart, bilateral breasts, bilateral lungs, and thyroid, were also delineated by ROs and/or medical physicists (MP). These structure definitions were in accordance with the Radiation Therapy Oncology Group (RTOG) breast cancer atlas and the ASTRO clinical practice guidelines (19, 20). These processes were independently verified by RO and MP.
As a guide for target delineation and image registration, contours for the surgical clips with a 3-mm margin (clip+3 mm) were defined; this was used to account for clip migration due to body shape changes, artifacts resulting from respiratory motion in the CT simulation, and phase misalignment during kV imaging in the linear accelerator (21).
Treatment planning. All plans were generated by two partial arc VMAT plans using a radiation treatment planning system (RayStation version 10ASP1; RaySearch Laboratories, Stockholm, Sweden). The applied arc angle was approximately 200°, which was adjusted based on the patient’s body shape and the PTV position. The dose was calculated using a collapsed-cone convolution algorithm with heterogeneity correction and a 2-mm calculation grid size. The prescribed dose was 30 Gy in five fractions. The dose received by at least 50% of the PTV_Eval volume (D50%) was set to 30 Gy. A 6 MV flattening-filter-free beam (TrueBeam; Varian Medical Systems, Palo Alto, CA, USA) was employed for all the plans. All planning approaches were optimized based on the target and OAR dose constraints outlined in the ASTRO clinical practice guidelines (20).
Treatment process using SGRT. The workflow for patient setup during treatment, including the steps for alignment and verification, is illustrated in Figure 1. On the first day of treatment, the patient was aligned to the DICOM reference surface using AlignRT version 6.3 (VisionRT Ltd., London, UK) and shifted as necessary to ensure accurate positioning. At this time, two regions of interest (ROIs) were used: one for the entire affected breast and another for the trunk in an inverted T shape.
Workflow for patient setup and treatment using surface-guided radiotherapy plus clip-based alignment without skin markers. SGRT: Surface-guided radiotherapy; IGRT: image-guided radiotherapy; CBCT: cone-beam computed tomography; ROI: region of interest.
During setup, the patient’s posture was corrected by the therapist using either manual or couch guidance based on the difference calculated by AlignRT from the reference and live surfaces (Real-time delta; RTD). When using couch guidance, movement was allowed along any axis (6 degrees of freedom [6DoF]: 3 translational axes, 3 rotational axes). The arm and chin positions were corrected using the “posture video” function, which displays the patient’s body outline. This correction ensured that the position of the entire breast ROI only fluctuated within±3 mm/3° to account for respiratory-related RTD fluctuations.
Prior to treatment, x-ray images of the orthogonal kV beam were taken using the TrueBeam imaging system (Varian Medical Systems, Inc.), and these pretreatment kV images were visually compared with the digitally reconstructed radiographs of the treatment plan. After major alignment corrections of the bone landmarks, image registration was performed with reference to the clip or clip+3 mm ROI. On x-ray, if any shift was greater than 5 mm/2°, the patient was imaged again to confirm the treatment position after applying the shifts. Afterward, cone-beam CT (CBCT) imaging was conducted, and image registration was carried out using clips, breast tissue, and the body surface, similar to the orthogonal kV images. We also recorded the magnitude of movement along the 6DoF for the image registration of the orthogonal kV images and CBCT images. All images were checked by the ROs before treatment. In cases wherein the orthogonal kV image and CBCT were consistent, as determined by the ROs, CBCT was omitted from the second fraction onward.
Evaluation of accuracy and setup time. The setup accuracy of AlignRT was defined as the sum of the residual setup errors corrected using kV images and CBCT, which was subsequently used to calculate the 3D vector setup error. Using the log information from AlignRT, the monitoring-on time was calculated as the duration from the start of initial positioning until the completion of beam delivery in AlignRT. This excluded time it took for the patient to enter and leave the room. The analysis was based on all treatment fractions for all patients (i.e., 35 patients×5 fractions=175 fractions).
Results
The study cohort had a mean±standard deviation (SD) age of 55.6±7.3 years and body mass index (BMI) of 21.6± 3.3 kg/m2. The mean±SD volumes of CTV, PTV, PTV_Eval, and the ipsilateral breast were 41.5±15.7, 87.0±23.2, 66.9±23.6, and 450.3±245.5 cm3, respectively.
Figure 2 and Table I illustrate the distribution of the residual setup error in any axis detected by the kV images or CBCT after the SGRT setup. The mean±SD residual setup error was 0.01±0.18, −0.09±0.22, and −0.04±0.19 cm, respectively, in the vertical, longitudinal, and lateral axes, while in the rotational axes, this was −0.19°±0.88°, 0.04°±0.64°, and 0.14°±0.54°, respectively, for yaw, roll, and pitch. Almost all treatment fractions achieved setup accuracy in the vertical, longitudinal, and lateral axes within acceptable thresholds of 5 mm (98.2%, 96.6%, and 96.6%, respectively) and 3 mm (92.3%, 85.7%, and 92.0%, respectively). The cumulative probability for positioning a patient within a 3D vector of 0.5 cm from the isocenter was 90% (Figure 3), and in 50% of the setup cases, the 3D vector was within 0.25 cm. During the treatment period, 28% of the patients (10 out of 35) maintained position deviations of less than 3 mm in the 3D vector direction.
Distribution of the residual setup error in the translational and rotational axes detected by image registration after the setup using surface-guided radiotherapy without skin markers.
Residual setup error detected by image registration after the setup using surface-guided radiotherapy without skin markers.
Cumulative probability of the translational and 3D vector offset.
Lastly, Figure 4 shows the distribution of the monitoring-on time in this setup using SGRT without skin markers. The mean±SD monitoring-on time was 603.3±214.1 seconds (range=349-1,353 s), and the 90th percentile was 877.2 s.
Distribution of the monitoring-on time of the setup using surface-guided radiotherapy without skin markers. The monitoring-on time is defined as the time from the start of the initial positioning in AlignRT until the completion of beam delivery.
Discussion
This study evaluated residual setup errors using kV images or CBCT after a setup using SGRT without skin markers for patients with small physiques and breasts who were subjected to APBI. Confirming our hypothesis, the integration of SGRT and clip alignment effectively achieved accurate and efficient patient positioning without the need for traditional skin marks. The AlignRT system enabled precise initial positioning which minimizes the need for significant corrections during subsequent IGRT. Over 95% of the treatment fractions achieved setup accuracy within clinically acceptable thresholds of 5 mm in all axes, demonstrating the reliability of this approach. Additionally, 28% of the patients (10 out of 35) maintained position deviations of less than 3 mm in the 3D vector direction, demonstrating the high positional stability of this approach. Lastly, the average monitoring-on time was 10 min, which was sufficient for ensuring accurate positioning while maintaining an efficient clinical workflow.
To the best of our knowledge, this is the first report that evaluates the positioning accuracy of a setup using SGRT plus clip-based alignment without using skin markers in APBI. Previous studies on SGRT setup in radiotherapy for patients with breast cancer undergoing APBI reported a median vector offset of 0.45 cm (range=0.24-0.60 cm) (10, 14, 22) and 0.44 cm (range=0.40-0.49 cm) (10-13); our result of 0.3 cm is comparable or better. These favorable results could be attributed to patient characteristics, mechanical updates to the SGRT system, and technical knowledge of the SGRT setup.
In this study cohort, the mean BMI was 21.6 kg/m2 (range=17.0-30.6 kg/m2), with most patients considered to be underweight or have normal weight. Since BMI was one of the physique indexes considered in this study, we selected patients with a small physique relative to European or North American patients. Accordingly, the average breast volume in our cohort was 450.3±245.5 cm3, which is much smaller than the average of 857 cm3 reported in previous studies (11). Notably, setup reproducibility is often challenging in patients with a high BMI or large, pendulous breasts (15). Most patients in this study had small breasts, whereas in previous studies, their patients had larger breasts, which are more likely to have variations in shape, potentially reducing the setup accuracy of SGRT. Nevertheless, compared to the conventional 3-point marker setup, SGRT remains a suitable solution for improving the positioning reproducibility because it incorporates surface contour data.
The mechanical advancements in the SGRT systems may have also impacted the setup accuracy. For example, the frame rate of AlignRT has increased by 2-3 times throughout its version upgrades, and its overall system performance also improved due to advancements in camera technology and system processing. Furthermore, a new function that corrects the patient’s posture has also improved the accuracy of adjusting the patient’s arm position. Kugele et al. reported that a three-camera system, compared to a single system, enabled a wider surface area coverage of the treatment site, thereby increasing the accuracy of patient setup (22).
The publication of international guidelines for SGRT systems as well as numerous findings and publications on SGRT will greatly improve the setup workflow for SGRT users (23, 24). These guidelines provide recommendations for the treatment workflow and points to consider for major anatomical sites (e.g., breast), providing valuable assistance to SGRT users. The impact of ROI setting and quality control/quality assurance of SGRT systems are also important factors. The workflow at our institution was established by referring to these publications (25, 26).
Temporary or permanent skin markings are traditionally employed to ensure reproducibility of the radiotherapy setup. However, Yamauchi et al. reported that many female patients with breast cancer (84%, 90 out of 107) experienced psychological distress in various aspects of their daily life due to these skin markings (27). Thus, these patients preferred skin markerless radiotherapy despite the additional time and financial burdens (28). Our findings demonstrate that the SGRT system does not require the use of conventional practices such as skin markers or tattoos.
The average monitoring-on time per patient using the AlignRT was 10 min, with 90% of patients completing the process within 15 min, including the initial positioning, image matching, and beam-on time. The complete in-room time, including 3 min for entry and exit, was approximately 13 min, which is clinically acceptable. Thus, SGRT enables fast and accurate initial positioning; this may have helped reduce the subsequent image registration time. Furthermore, the surgical clips provided clear landmarks for the therapist, allowing for fast and accurate image registration independent of the operator. Moreover, additional time was saved since we did not need to use temporary skin markings, which had to be redrawn every few days.
The main limitation of this study is its small cohort of 35 patients who received APBI at a single medical institution. Because SGRT is affected by the patient’s body shape, breast size, and skin color, these results may not be directly applicable in other populations. Moreover, we did not compare the outcomes of the conventional three-point marker setup within the same facility. This is because our hospital had already introduced a markerless setup for whole-breast irradiation, and thus we simply followed the same procedure for APBI. Therefore, we needed to refer to previous studies to compare setup accuracy, and procedure improvements were not evaluated.
Conclusion
The study evaluated the residual setup errors detected via image registration after SGRT setup in patients undergoing APBI. This method demonstrated setup accuracy within the acceptable range of 5 mm along any axis for over 95% of the treatment fractions evaluated. This fully markerless SGRT workflow in conjunction with clip alignment is highly reliable, accurate, and efficient, making it a viable alternative to skin marking.
Acknowledgements
This work was supported by JSPS KAKENHI [grant number JP23K14850] and the Shintaro Akatsu Young Scientist Incentive Program of St. Luke’s Health Science Research Fund.
Footnotes
Authors’ Contributions
All Authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by RY and FT. The first draft of the article was written by RY, and all authors commented on previous versions of the article. All Authors read and approved the final article.
Conflicts of Interest
The Authors declare that there are no conflicts of interest in relation to this study.
- Received January 8, 2025.
- Revision received January 21, 2025.
- Accepted January 22, 2025.
- Copyright © 2025 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).










