Seed Placement in Permanent Breast Seed Implant Brachytherapy: Are Concerns Over Accuracy Valid?

International Journal of

Radiation Oncology biology

physics

www.redjournal.org

Physics Contribution

Seed Placement in Permanent Breast Seed Implant Brachytherapy: Are Concerns Over Accuracy Valid? Daniel Morton, MSc,*,y Michelle Hilts, PhD,*,y Deidre Batchelar, PhD,* and Juanita Crook, MDz *Department of Medical Physics, BC Cancer Agency, Centre for the Southern Interior, Kelowna, British Columbia, Canada; yDepartment of Physics and Astronomy, University of Victoria, Victoria, British Columbia, Canada; and zDepartment of Radiation Oncology, BC Cancer Agency, Centre for the Southern Interior, Kelowna, British Columbia, Canada Received Sep 22, 2015, and in revised form Jan 6, 2016. Accepted for publication Jan 27, 2016.

Summary Permanent breast seed implant brachytherapy is an attractive treatment option for early-stage breast cancer patients. To improve the accuracy and confidence in the procedure, uncertainties in the delivery must be identified. This study evaluates the accuracy of seed placement within the breast, to identify any systematic errors in the delivery and evaluate the impact of seed placement on the treatment.

Purpose: To evaluate seed placement accuracy in permanent breast seed implant brachytherapy (PBSI), to identify any systematic errors and evaluate their effect on dosimetry. Methods and Materials: Treatment plans and postimplant computed tomography scans for 20 PBSI patients were spatially registered and used to evaluate differences between planned and implanted seed positions, termed seed displacements. For each patient, the mean total and directional seed displacements were determined in both standard room coordinates and in needle coordinates relative to needle insertion angle. Seeds were labeled according to their proximity to the anatomy within the breast, to evaluate the influence of anatomic regions on seed placement. Dosimetry within an evaluative target volume (seroma þ 5 mm), skin, breast, and ribs was evaluated to determine the impact of seed placement on the treatment. Results: The overall mean (SD) difference between implanted and planned positions was 9  5 mm for the aggregate seed population. No significant systematic directional displacements were observed for this whole population. However, for individual patients, systematic displacements were observed, implying that intrapatient offsets occur during the procedure. Mean displacements for seeds in the different anatomic areas were not found to be significantly different from the mean for the entire seed population. However, small directional trends were observed within the anatomy, potentially indicating some bias in the delivery. Despite observed differences between the planned and implanted seed positions, the median (range) V90 for the 20 patients was 97% (66%-100%), and acceptable dosimetry was achieved for critical structures.

Reprint requests to: Daniel Morton, MSc, BC Cancer Agency, SAHCSI, 399 Royal Ave, Kelowna, BC V1Y 5L3, Canada. Tel: (250) 712-3966, ext. 686877; E-mail: [email protected] Int J Radiation Oncol Biol Phys, Vol. -, No. -, pp. 1e8, 2016 0360-3016/$ - see front matter Ó 2016 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ijrobp.2016.01.049

Conflict of interest: none.

2

International Journal of Radiation Oncology  Biology  Physics

Morton et al.

Conclusions: No significant trends or systematic errors were observed in the placement of seeds in PBSI, including seeds implanted directly into the seroma. Recorded seed displacements may be related to intrapatient setup adjustments. Despite observed seed displacements, acceptable postimplant dosimetry was achieved. Ó 2016 Elsevier Inc. All rights reserved.

Introduction Treatment of early-stage breast cancer typically involves breast-conserving surgery followed by adjuvant radiation therapy to the whole breast, which has disease-free survival rates equivalent to mastectomy and leads to improved cosmetic outcomes (1, 2). For early-stage breast cancer patients, accelerated partial breast irradiation (APBI), which irradiates only the lumpectomy cavity, is becoming more widely used as an alternative to whole-breast irradiation, primarily owing to decreased treatment times and excellent clinical outcomes (3, 4). A recent development in partial-breast irradiation is low-dose-rate permanent breast seed implant brachytherapy (PBSI) (5, 6). This procedure uses radioactive 103Pd (7) seeds implanted permanently around the excision cavity in the breast under template and ultrasound (US) guidance. Pignol and colleagues have published encouraging results from the procedure, achieving successful planning target volume (PTV) coverage, no significant risk of local recurrences over a median follow-up period of 63 months for 134 patients, minimal skin toxicity, and excellent patient satisfaction (6, 8, 9). Given these results and the convenience of a singleday treatment, PBSI is an attractive treatment option for early-stage breast cancer patients. However, concerns have been expressed regarding uncertainties in seed placement accuracy, as well as the possibility of seed migration given that the target is a fluid-filled excision cavity (10). In this study the former is addressed by identifying uncertainties in treatment delivery to improve implant accuracy. The accuracy of seed placement in the current PBSI procedure is quantified by assessing differences between planned and implanted seed positions, and the effects of such differences on initial dosimetry are evaluated.

Methods and Materials Patient eligibility Between May 2012 and October 2015, 30 women were treated with PBSI as part of an ongoing single-institute clinical trial. Patient eligibility criteria were age 60 years with ductal carcinoma in situ or low-risk invasive ductal carcinoma (clear margins, pT1pN0, estrogen receptor positive, and no evidence of metastatic disease). Technical eligibility criteria included seroma <3 cm in equivalent sphere diameter, PTV volume <120 cm3, seroma visible on US, and seroma position and breast size feasible for

implant. Seed placement was evaluated for 20 of 30 patients (13 left, 7 right); the others were excluded owing to ambiguity in image registration, intraoperative changes, or seed matching.

PBSI technique The PBSI procedure used was adapted from the technique developed by Pignol et al (5). The patient is positioned on a breast board and undergoes computed tomography (CT) simulation as for external beam radiation therapy. The seroma is contoured, and the PTV is defined as seroma plus a 1.25-cm margin, cropped to chest wall and 5 mm from skin surface. Using MIM Symphony (MIM Software, Cleveland, OH), needle insertion angle, depth, and template position are chosen by planning a fiducial needle to pass through the center of the seroma, tangential to chest wall. The plan is centered on the fiducial needle, and loaded needle positions are planned to achieve 100% prescription dose coverage (90 Gy) to the seroma, PTV coverage of V90 >98%, V100 >95%, V150 <70%, and V200 <25%, while limiting skin dose to less than 90% of the prescription over a 1-cm2 area (6). Treatments are planned using a 2.5-U seed activity (range, 2.4-2.7 U) and line-source geometries. Immediately before implant the patient undergoes a mark-up procedure in the CT simulator, designed to increase setup accuracy and reduce time spent in the operating room (11). After setup adjustments in the operating room to replicate the simulator positioning, the template (Breast Microseed, Mercer Island, WA) is positioned and the fiducial needle is inserted using freehand US guidance to ensure proper depth and positioning relative to seroma. Stranded 103Pd seeds (Theraseed-200; Theragenics, Buford, GA) are then implanted to their planned positions, again using freehand US for guidance. The implant takes between 1 and 2 hours. On the day of implant (day-0), postimplant CT is performed to evaluate seed positions and dosimetry. Patient positioning from the planning CT is replicated using the breast board and tattoos to achieve setup accuracy equivalent to EBRT. Deformable image registration and adaptive contouring (MIM Maestro, MIM Software) is used to define postimplant seroma from the preplan contour (12). Seeds are identified and dose-volume histograms calculated for the postimplant seroma, an evaluative target volume (ETV) defined as the seroma plus 0.5 cm, and surrounding normal tissues, using point-source calculations. The same procedure is repeated at a 30-day follow-up. Day-0 CT images are used here to assess initial seed placement and dosimetry.

Volume -  Number -  2016

Seed placement accuracy Image registration After clinical dosimetry a retrospective analysis was undertaken to determine the position of implanted seeds relative to their planned positions. Rigid registration (MIM Symphony), confirmed by visual inspection, was used to align a region of interest (seroma plus a 1-cm margin) and, thus, the treatment volumes. This process was validated by comparing the pre- and postimplant locations of surgical clips placed at the time of lumpectomy in 10 patients, and any noted differences were correlated with differences between planned and implanted seed positions. Seed matching In-house seed-matching software (SeedPreview) previously developed to quantify seed motion in prostate brachytherapy (13) was used to determine positional differences between planned and implanted seed positions (seed displacements). Coordinates for both planned and implanted seed positions, contours, and the transformation matrix derived from the registration of the 2 images sets were exported from MIM Symphony. The transformation matrix was used to superpose the 2 seed clouds in SeedPreview, and postimplant seeds for each patient were manually matched with their corresponding planned locations (Fig. 1). Because of the stranded nature of the seeds, the matching process was unambiguous for all cases included in the study. Seed placement analysis Using SeedPreview, the displacements of 1370 implanted seeds from 20 patients (median 67 seeds per patient; range, 52-94) were determined. Mean total displacements were

PBSI seed placement accuracy

3

calculated over all patients and within each individual patient. Directional displacements were evaluated using 2 frames of reference (Fig. 2a): standard room coordinates (medialelateral [ML], anterioreposterior [AP], superioreinferior [SI]) and needle coordinates, defined by the insertion direction. Positive values were defined to be in the lateral, anterior, and superior directions in room coordinates, and the shallow, up, and right directions in needle coordinates. The computed mean direction displacements imply global, systematic seed displacements. To approximate the contribution of random errors relative to these systematic displacements, for each patient mean directional seed displacements were subtracted from the directional displacement recorded for each seed and an average residual seed displacement computed. Implanted seeds were further classified into 4 anatomic regions (Fig. 2b): seeds in the plane closest to chest wall (Chest Wall), seeds in the plane closest to skin (Skin), seeds implanted through the seroma (Seroma), and seeds at the anterior boundary of the seroma (Edge of Seroma). Mean values for the total and directional displacements of the seeds, using needle coordinates, were calculated within these regions to determine the effect of anatomic location on placement accuracy.

Dosimetry Postimplant dosimetry was assessed on day-0 CT images, to evaluate the initial treatment quality relative to the calculated seed displacement. The volume of the ETV receiving at least 90% of the prescribed does (V90) was calculated, where a V90 of >90% is considered acceptable coverage for APBI (14). The D0.2cc in a 2-mm-thick skin layer (15), V150 and V200 within the ETV, and maximum point rib dose were also evaluated to further assess the treatment.

Results Image registration accuracy The median (range) time from surgery to planning CT was 60 (37-105) days, and from planning to the day-0 CT was 22 (15-43) days. The mean positional agreement between surgical clips within the region of interest (39 clips total: median 4 clips per patient; range, 1-7) from planning to implant was 1  1 mm and was independent of location within the breast. This is consistent with values recorded previously for clip migration (16) and validates the accuracy of the rigid registration process. No correlation was found between displacement of clips and seeds (rZ0.01). Fig. 1. Visualization of matched planned (blue) and implanted (pink) seed positions. Connecting lines (white) indicate matched seed pairs. The purple, blue, and white contours correspond to the skin surface, chest wall, and seroma, respectively. A color version of this figure is available at www.redjournal.org.

Seed placement accuracy Figure 3a shows the total and directional distributions of the aggregate seed population (all patients). Mean (SD) total displacement was 9  5 mm; there are no systematic

4

International Journal of Radiation Oncology  Biology  Physics

Morton et al.

Fig. 2. (a) Implant geometry and coordinate systems in standard room coordinates (white) and rotated needle coordinates (yellow). (b)Visualization orthogonal to the insertion direction demonstrating the anatomic regions relative to the skin (orange), seroma (blue), and chest wall (green). A color version of this figure is available at www.redjournal.org. intrapatient systematic offsets do occur during the procedure. However, the magnitude of shifts varied widely among patients (for example, 7 to 8 mm in the AP direction) and thus do not amount to any interpatient trend. Subtraction of mean displacements from the individual patient distributions gave an approximate mean random component to seed displacement of 7  4 mm. This signifies that a significant component of observed seed displacements is due to random variations in seed placement. Figure 4 shows seed displacement distributions from the entire seed population in the needle coordinate system. Similar to the room coordinate system (Fig. 3a), the

displacements in room coordinates, and all directional displacements appear normally distributed around zero; means are 0  7 mm, 0  7 mm, and 1  5 mm, in the ML, AP, and SI directions, respectively. For individual patients, mean total seed displacement ranged from 6  3 mm to 16  6 mm, with individual seeds placed between 1 and 31 mm from their planned positions. In contrast to the whole population, for individual patients systematic displacements are observed, as is shown in Figure 3b. Here, mean shifts in the directional displacement of the seeds are 1  5 mm, 4  4 mm, and 3  4 mm, in the ML, AP, SI directions, respectively. This implies that

1 2 Distance (cm)

0

3

4

160

200

AP 120 Seeds

40 2

100

0

1 0.5 Distance (cm)

0 –1 –0.5 0 0.5 1 Distance (cm)

1.5

4

SI

–1 0 1 Distance (cm)

2

0 –0.5

4 2

2 –2

1.5

6 AP

6

50 –1 0 1 Distance (cm)

0

8

SI

150

80

0 –2

0 0 2 Distance (cm)

6 3

2 –2

ML

9 Seeds

120 60

0

12

Magnitude

6 Seeds

Seeds

Seeds

80 40

Seeds

8

ML

180

120

0

b

240

Magnitude

Seeds

160

Seeds

a

0 0.5 1 Distance (cm)

1.5

0 –1

0 –0.5 Distance (cm)

0.5

Fig. 3. Histograms of the magnitude and directional (ML, AP, SI) seed displacements for the cohort seed population (a) and for a single patient (b). The directional distributions are normal and centered at zero for the cohort but not for the individual patient. Abbreviations: AP Z anterioreposterior; ML Z medialelateral; SI Z superioreinferior.

Volume -  Number -  2016

PBSI seed placement accuracy

160 160

Magnitude Seeds

Seeds

120 80 40 0

120 80 40

0

1 2 Distance (cm)

0

3

160

0 2 Distance (cm)

40

R/L

180 Seeds

80

0 –2

–2

240 Height

120 Seeds

Depth

120 60

–1 0 1 Distance (cm)

0

–2

–1 0 1 Distance (cm)

2

Fig. 4. Histograms of the magnitude and directional seed displacements in the needle coordinate system for the entire seed cohort. Abbreviation: R/L Z righteleft. directional distributions in Figure 4 are centered around zero in depth (0  8 mm), height (0  4 mm), and righteleft (R/L) (0  5 mm) of needle insertion.

Seed placement within anatomy Mean total displacements were 9  4 mm, 9  4 mm, 9  5 mm, and 10  5 mm for seeds implanted within the Seroma, Edge of Seroma, Chest Wall, and Skin regions, respectively, and are not significantly different from the mean displacements for all seeds (9  5 mm), regardless of anatomic position (PZ.34, .07, .82, and .68, respectively). Figure 5 shows the total and directional distributions, measured in needle coordinates, of seed displacements within anatomic regions. The directional distributions within the anatomy contain several small mean directional shifts, most notably in the height of insertion in Skin and Edge of Seroma seeds (1  5 mm and 1  4 mm, respectively). Figure 5 shows a large number of seeds displaced toward the negative height within these areas, corresponding to seeds implanted downward from their planned positions, away from skin. Conversely, seeds within the Chest Wall region were seen to on average be implanted upward from their planned positions (1  3 mm). Displacement in depth was seen to vary widely within the cohort, but no direction was favored, with a mean of 0  8 mm in all regions except for Edge of Seroma (0  7 mm).

Dosimetry The ETV V90 (Fig. 6) ranged from 66% to 100% (median 97%) across the 20 patients. A V90 >90% was achieved in 15 patients, and >95% in 13 patients. Figure 6 also shows how dosimetry may be affected by the distance of the seeds

5

from their planned positions. The maximum mean displacement observed that resulted in acceptable dosimetry was 11  4 mm. One patient with a mean seed displacement lower than this value (7  3 mm) had ETV coverage falling into the unacceptable range, but at 89% was only very slightly below the acceptable value (90%). Median ETV V150 and V200 were 75% (range, 43%-97%) and 50% (range, 23%-66%), respectively. Dose to normal tissues was also observed to be acceptable for the patients within this study. The median D0.2cc to the 2-mm-thick skin layer was 80% (range, 9%-202%) of the prescription dose, and the median maximum point dose to the ribs was 101% (range, 2%-437%). The immediate postimplant dosimetry achieved within this study was consistent with previously reported data (6).

Discussion Concern has been expressed regarding the uncertainty in seed placement and thus in the safety and efficacy of PBSI (10). This stems from the assertion that large seed placement errors may result in reduced coverage of the target volume and increased chance of recurrence, or increased radiation hot spots resulting in undesirable radiation side effects, neither of which has been observed in published experience to date (9). This study examined the validity of these concerns by evaluating the placement accuracy of seeds in the current PBSI procedure, to determine the magnitude of such placement errors and their effect on treatment dosimetry, as well as to expose any potential systematic errors affecting the treatment. The results presented in Figures 3 and 4 show that no significant interpatient systematic errors occur in the placement of seeds, in either coordinate system. Although systematic placement errors are absent, an overall mean displacement of 9  5 mm exists in the evaluated seeds. Despite this, acceptable implant dosimetry of ETV V90 exceeding 90% (14) is largely maintained. Random variations in seed placement, which were shown to form a large portion of the seed displacements, do not result in net displacement, and because good ETV coverage was achieved it is unlikely that they have a resulting effect on the dosimetry. The results of this study indicate that normal tissue tolerance and acceptable treatment dosimetry are achieved immediately after the implant within the limits of the seed placement accuracy of the current PBSI procedure. A learning curve in the PBSI implant technique has been observed (5, 6), so placement accuracy and target coverage could potentially improve beyond the initial patients included in this study. In contrast to the lack of systematic directional displacements observed across the entire patient cohort, mean directional displacements do occur within each individual patient. One possible explanation is that individual patient setup inconsistencies between planning and implant contribute to the seed displacements. Time lag between

International Journal of Radiation Oncology  Biology  Physics

Morton et al. 60

20

10

0 –2

3

–1

0 1 2 Distance (cm)

0.5 1 1.5 2 Distance (cm)

–1 0 1 Distance (cm)

d

10

1

Depth 30

0

0.5 1 1.5 2 Distance (cm)

0

2.5

1.5

–2

–1 0 1 Distance (cm)

40 20 0

2

60 Height

Seeds

0 –1 –0.5 0 0.5 1 Distance (cm)

20 10

60

20

1.5

40

20

0

2

10

0 –1.5 –1 –0.5 0 0.5 Distance (cm)

40

R/L Seeds

10

0 –1 –0.5 0 0.5 1 Distance (cm)

0 1 Distance (cm)

10 –1 0 1 Distance (cm)

20

Magnitude

30

20

–1

30

Height 30

R/L

10

Depth

40

40

20

0

2

20

0 –2

2.5

0 2 Distance (cm)

30

10

Seeds

Seeds

10

0

20

–2

–2

40

30

30

Magnitude

0

3

Height

40

0

1.5

20

2 1 Distance (cm)

0

40

Seeds

Seeds

Seeds

20

20 10

R/L

30

Seeds

0

3

60

40

20

Seeds

1 2 Distance (cm)

0 –1 –0.5 0 0.5 1 Distance (cm)

Seeds

Depth 30

Seeds

0

40

Height

0

40 Magnitude Seeds

Seeds

Seeds

20

60

c

40

Depth 30

40

0

b

60

Magnitude

Seeds

a

R/L Seeds

6

–1

0 1 Distance (cm)

40 20 0 –2

–1 0 1 Distance (cm)

2

Fig. 5. Histograms of the magnitude and directional seed displacements, measured in needle coordinates, in the 4 anatomic areas: (a) Chest Wall, (b) Skin, (c) Edge of Seroma, and (d) Seroma. Abbreviation: R/L Z righteleft. 100 95

ETV V90 (%)

90 85 80 75 70 65 0.4

0.6

0.8 1 1.2 1.4 Mean Displacement (cm)

1.6

Fig. 6. V90 of the evaluative target volume (ETV) relative to the mean seed displacement within each patient. Dashed lines represent the target V90 of 90% and the maximum displacement observed to give an acceptable dose (1.1 cm).

planning and implant could affect seroma size and shape and therefore how seeds are implanted relative to the plan, and breast density or tissue type (scar vs skin) may impact tissue deformation and needle bending during insertion. Another contributing factor may be the use of 2 different imaging techniques within the PBSI procedure: CT for treatment planning and US guidance for implant. Although US image guidance is used to match the patient treatment position to the planned position, the US-delineated seroma can vary in size, shape, and position compared with the seroma contoured on planning CT (17-20). Therefore, shifting the plan to align it to the US image may create a discrepancy between treatment delivery and planned location, thus introducing offsets to the seed placements. The results given in Figure 5 show that the anatomic position does not affect the magnitude of the displacement of the implanted seeds, thus refuting the hypothesis (10) that seeds within certain anatomic areas are prone to larger displacements immediately after their implant. A small mean vertical displacement of the seeds on the

Volume -  Number -  2016

anterior edge of the seroma may indicate the seeds settling toward the seroma, but seeds placed directly in the seroma did not experience any greater displacement compared with the seeds placed in the other anatomic areas. The only notable trend was the small vertical displacements of seeds away from the chest wall and skin. This difference may indicate an operator bias to intentionally direct needles to avoid penetrating the lungs or distal skin. Seed placement in prostate brachytherapy is estimated to be accurate to approximately 3 to 6 mm (21-24). Although relatively smaller than the PBSI displacement observed in this study, the prostate brachytherapy procedure has advantages such as the use of US for both planning and implant guidance, and the use of a US stabilizer and stepper to provide a fixed reference for implant coordinates. Thus greater accuracy is not unexpected. Additionally, the robust method used in this study of registering the CT data according to the anatomy and manually performing the seed matching is more rigorous than previously presented evaluations of prostate seed implant accuracy via orthogonal films or automatic registration of the seed clouds (22, 23). A favorable result of this study in comparison with prostate brachytherapy was the lack of any systematic displacements. Multiple prostate brachytherapy sector dosimetry studies have shown systematic under-dosing of the prostate anterior base (25-27). Seeds being pulled toward the apex during implant, in part owing to needle drag (25), did not have an analog in this study, and thus PBSI seed displacement seems to be more random and unique to each individual patient. The success of the procedure relies not only on the accurate placement of seeds, but whether the seeds remain in the desired location over the course of the delivery. Pignol and colleagues have observed an increase in the V100 and V200 within the PTV day 0 to day 60, likely due to the reduction of edema over time bringing the seeds closer together (6, 28). Later-date dosimetry will be influenced by seed motion, and work is ongoing to determine how seeds migrate owing to factors such as recovery or motion of the breast, and how migration relates to initial seed placement. Further, continued technique developments in PBSI may provide improvements in seed placement accuracy. As mentioned above, biases associated with attempting to center the CT-planned implant on the US seroma could potentially be reduced by implementing co-registered 3dimensional US images into the planning procedure, thus allowing patient setup to be performed using more analogous images. Improved image guidance during treatment delivery, similar to the well-established prostate brachytherapy techniques, may further reduce seed displacements by providing precise needle and plan localization at depth.

Conclusion Permanent breast seed implant brachytherapy seeds were found to be placed, on average, 9  5 mm from their

PBSI seed placement accuracy

7

planned location. No significant trends or systematic errors were observed in the placement of seeds, including seeds implanted directly into the seroma cavity. Recorded differences in seed position before and after implant may be related to necessary adjustments made during implant based on US guidance and are currently being investigated. Even with the observed discrepancies in seed placement, acceptable postimplant dosimetry was achieved.

References 1. Fisher B, Anderson S, Bryant J, et al. Twenty-year follow-up of a randomized trial comparing total mastectomy, lumpectomy, and lumpectomy plus irradiation for the treatment of invasive breast cancer. N Engl J Med 2002;347:1233-1241. 2. Veronesi U, Cascinelli N, Mariani L, et al. Twenty-year follow-up of a randomized study comparing breast-conserving surgery with radical mastectomy for early breast cancer. N Engl J Med 2002;347:12271232. 3. Shah C, Vicini F, Wazer DE, et al. The American Brachytherapy Society consensus statement for accelerated partial breast irradiation. Brachytherapy 2013;12:267-277. 4. Vicini FA, Kestin L, Chen P, et al. Limited-field radiation therapy in the management of early-stage breast cancer. J Natl Cancer Inst 2003; 95:1205-1210. 5. Pignol JP, Keller B, Rakovitch E, et al. First report of a permanent breast 103Pd seed implant as adjuvant radiation treatment for earlystage breast cancer. Int J Radiat Oncol Biol Phys 2006;64:176-181. 6. Keller BM, Ravi A, Sankreacha R, et al. Permanent breast seed implant dosimetry quality assurance. Int J Radiat Oncol Biol Phys 2012;83:84-92. 7. Keller B, Sankreacha R, Rakovitch E, et al. A permanent breast seed implant as partial breast radiation therapy for early-stage patients: A comparison of palladium-103 and iodine-125 isotopes based on radiation safety considerations. Int J Radiat Oncol Biol Phys 2005;62:358-365. 8. Pignol JP, Rakovitch E, Keller BM, et al. Tolerance and acceptance results of a palladium-103 permanent breast seed implant phase I/II study. Int J Radiat Oncol Biol Phys 2009;73:1482-1488. 9. Pignol JP, Caudrelier JM, Crook J, et al. Report on the clinical outcomes of permanent breast seed implant for early-stage breast cancers. Int J Radiat Oncol 2015;93:614-621. 10. Godinez J, Gombos EC. A permanent breast seed implant as partialbreast radiation therapy for early-stage patients: In regards to Keller et al. (Int J Radiat Oncol Biol Phys 2005;62:358-365). Int J Radiat Oncol Biol Phys 2006;64:1611. 11. Batchelar D, Hilts M, Rose T, et al. Simulation and intraoperaive checks for improved standardization and reproducibility of partial breast seed implant technique. Brachytherapy 2014;13(Suppl. 1):S84 [Abstr.]. 12. Hilts M, Batchelar D, Rose J, et al. Deformable image registration for defining the postimplant seroma in permanent breast seed implant brachytherapy. Brachytherapy 2015;14:409-418. 13. Bowes D, Gaztan˜aga M, Araujo C, et al. A randomized trial comparing seed displacement of coated seeds to regular loose seeds at 30 days postimplant. Brachytherapy 2013;12:362-367. 14. Vicini F, White J, Arthur D, et al. NSABP protocol B-39/RTOG protocol 0413: A randomized phase III study of conventional whole breast irradiation (WBI) versus partial breast irradiation for women with stage 0, I, or II breast cancer. Available at: www.rtog.org/ClinicalTrials/Protocol Table/StudyDetails.aspx?studyZ0413. Accessed October 15, 2014. 15. Hilts M, Halperin H, Morton D, et al. Skin dose in breast brachytherapy: Defining a robust metric. Brachytherapy 2015;14:970-978. 16. Topolnjak R, de Ruiter P, Remeijer P, et al. Image-guided radiotherapy for breast cancer patients: Surgical clips as surrogate for breast excision cavity. Int J Radiat Oncol Biol Phys 2011;81:e187-e195.

8

Morton et al.

17. Berrang TS, Truong PT, Popescu C, et al. 3D ultrasound can contribute to planning CT to define the target for partial breast radiotherapy. Int J Radiat Oncol Biol Phys 2009;73:375-383. 18. Wong P, Muanza T, Reynard E, et al. Use of three-dimensional ultrasound in the detection of breast tumor bed displacement during radiotherapy. Int J Radiat Oncol Biol Phys 2011;79:39-45. 19. Landry A, Berrang T, Gagne I, et al. Investigation of variability in image acquisition and contouring during 3D ultrasound guidance for partial breast irradiation. Radiat Oncol 2014;9:35. 20. Smitt MC, Birdwell RL, Goffinet DR. Breast electron boost planning: Comparison of CT and US. Radiology 2001;219:203-206. 21. Podder T, Beaulieu L, Caldwell B, et al. AAPM and ESTRO guidelines for image-guided robotic brachytherapy: Report of Task Group 192. Med Phys 2014;41:101501. 22. Taschereau R, Pouliot J, Roy J, et al. Seed misplacement and stabilizing needles in transperineal permanent prostate implants. Radiother Oncol 2000;55:59-63.

International Journal of Radiation Oncology  Biology  Physics 23. Fu L, Liu H, Brasacchio R, et al. Clinical observation and modeling of postimplant seed displacement for prostate brachytherapy. Int J Radiat Oncol Biol Phys 2005;63(Suppl. 1):S504-S505 [Abstr.]. 24. Kaplan I, Meskell P, Lieberfarb M, et al. A comparison of the precision of seeds deposited as loose seeds versus suture embedded seeds: A randomized trial. Brachytherapy 2004;3:7-9. 25. Sidhu S, Morris WJ, Spandinger I, et al. Prostate brachytherapy postimplant dosimetry: A comparison of prostate quadrants. Int J Radiat Oncol Biol Phys 2002;52:544-552. 26. Nasser NJ, Wang Y, Borg J, et al. Sector analysis of dosimetry of prostate cancer patients treated with low-dose-rate brachytherapy. Brachytherapy 2014;13:369-374. 27. Merrick GS, Butler WM, Grimm P, et al. Multisector prostate dosimetric quality: Analysis of a large community database. Brachytherapy 2014;13:146-151. 28. Wazer DE, Aurthur DW, Vicini FA, editors. Accelerated Partial Breast Irradiation. Heidelberg: Springer-Verlag; 2009.