TG 303 MRI in HDR Brachytherapy Public Review Draft

Page 1 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1 The American Association of Physicists in Medicine (AAPM) is a nonprofit professional

2 society whose primary purposes are to advance the science, education and professional
3 practice of medical physics. The AAPM has more than 8,000 members and is the principal
organization of medical physicists in the United States.
4
5 The AAPM periodically publishes task group reports to help advance the science of medical
6 physics and to improve the quality of service to patients.
7
Each AAPM task group report goes through a thorough review process at the
8
corresponding task group, working group, subcommittee, and parent committee (Imaging
9 Physics Committee or Therapy Physics Committee) and requires the approval of Science
10 Council.
11
Science Council has adopted a process where the attached report is posted for 2 weeks to
12
gather feedback from members of the AAPM prior to the final stages of review.
13
14
15 MRI Implementation in gynecologic and prostate HDR brachytherapy -
16 considerations from simulation to treatment
17 The Report of AAPM Task Group #303, Endorsed by the ABS
18
19 Joann Prisciandaro Ph.D. (Vice-Chair)
20 University of Michigan Hospital and Health Systems, Ann Arbor, MI
21 Jacqueline (Esthappan) Zoberi, Ph.D.
22 Washington University School of Medicine, St. Louis, MO
23 Gil'ad Cohen Ph.D.
24 Memorial Sloan-Kettering Cancer Center, New York, NY
25 Yusung Kim, Ph.D.
26 University of Iowa, Iowa City, Iowa
27 Perry Johnson, Ph.D.
28 University of Florida Health Proton Therapy Institute, Jacksonville, FL
29 Eric Paulson, Ph.D.
30 Medical College of Wisconsin, Milwaukee, WI
31 William Song, Ph.D.
32 Virginia Commonwealth University, Richmond, VA
33 Ken-Pin Hwang, Ph.D.
34 The University of Texas MD Anderson Cancer Center, Houston, TX
35

1
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 2 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

36 Beth Erickson, M.D.

37 Medical College of Wisconsin, Milwaukee, WI
38 Sushil Beriwal, M.D., MBA
39 University of Pittsburgh Medical Center Hillman Cancer center, Pittsburgh, PA
40 Christian Kirisits, Ph.D.
41 Medical University of Vienna, Vienna, Austria
42 Firas Mourtada, Ph.D. (Chair)
43 Helen F. Graham Cancer Center, Newark, DE
44
45 Disclosure Statement:
46 The Chair of this task group reviewed the required Conflict of Interest statement on file for each
47 member of AAPM Task Group # 303 and determined that disclosure of potential Conflicts of Interest is
48 an adequate management plan. Disclosures of potential Conflicts of Interest for each member of AAPM
49 Task Group # 303 are found at the close of this document.
50
51

2
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 3 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

52 Table of Contents
53 I. Introduction: Purpose, Goals, and Rationale for the Task Group 6
54 A. Brief description of MRI-based HDR brachytherapy requirements 6
55 B. Intracavitary and interstitial applications for gynecological and prostate BT 6
56 II. Nomenclature and abbreviations 8
57 III. Background: Review of literature on MRI utilization for HDR BT 10
58 A. Summary of various approaches 10
59 i. Open bore MRI with <1.5 T magnetic strength 11
60 ii. Closed bore MRI with 1.5 T or 3.0 T magnetic strength 11
61 iii. Multi-imaging modality approach: MRI/Ultrasound (US), MRI/CT 12
62 B. MRI applicators and needles 12
63 C. Review of current MRI BT clinical trials 13
64 i. Prostate BT 13
65 ii. Gynecologic BT 14
66 D. Review of existing Task Group reports on HDR and MRI 15
67 E. New recommendations beyond CT based practices 16
68 IV. Clinical implementation of MRI for HDR BT 16
69 A. General Workflows from Simulation to Delivery 16
70 B. Brachytherapy-specific MR imaging acquisition parameters 18
71 C. Applicator/needle commissioning and reconstruction 20
72 D. Patient-transfer systems 22
73 E. Target and OAR Delineation using MRI 23
74 i. Delineation of GYN structures 23
75 ii. Delineation of prostate structures 25
76 F. Heterogeneities for calculations 26
77 G. Safety considerations 26
78 i. Procedure room equipment 27
79 ii. Applicator, needles, patient positioning systems 27
80 iii. Patient immobilization and transfer considerations 28
81 iv. Setup verification prior to HDR treatments 29

3
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 4 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

82 H. Logistical and economic considerations 30

83 V. Risk based analysis 31
84 VI. Recommendations to the medical physicists 33
85 A. Clinical commissioning tasks 33
86 B. MRI safety essential considerations 35
87 C. Quality Assurance 36
88 i. Traditional HDR BT Program QA 36
89 ii. QA considerations for MRI integration with HDR BT 36
90 D. Logistical and economic considerations 37
91 VII. Potential future development in MRI-guided BT 38
92 VIII. Summary 38
93 IX. References 58
94
95
96

4
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 5 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

97 Abstract:

98 The Task Group (TG) on Magnetic Resonance Imaging (MRI) implementation in High Dose Rate
99 (HDR) Brachytherapy - Considerations from Simulation to Treatment, TG 303, was constituted by the
100 American Association of Physicists in Medicine’s (AAPM’s) Science Council under the direction of the
101 Therapy Physics Committee, the Brachytherapy Subcommittee, and the Working Group on Brachytherapy
102 Clinical Applications. The TG was charged with developing recommendations for commissioning, clinical
103 implementation, and on-going quality assurance (QA), to describe HDR brachytherapy (BT) workflows and
104 to evaluate practical consideration that arise when implementing MR imaging. For brevity, the report is
105 focused on the treatment of gynecologic and prostate cancer.
106 The TG report provides an introduction and rationale for MRI implementation in BT, a review of
107 previous publications on topics including available applicators, clinical trials, previously published BT
108 related TG reports, and new image guided recommendations beyond CT based practices. The report
109 describes MRI protocols and methodologies, including recommendations for the clinical implementation
110 and logical consideration for MR imaging for HDR BT. Given the evolution from prescriptive to risk-based
111 QA, an example of risk-based analysis using MRI-based, prostate HDR BT is presented.
112 In summary, the TG report is intended to provide clear and comprehensive guidelines and
113 recommendations for commissioning, clinical implementation, and QA for MRI-based HDR BT that may
114 be utilized by the medical physics community to streamline this process. This report is endorsed by the
115 American Brachytherapy Society (ABS).
116

5
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 6 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

117 I. Introduction: Purpose, Goals, and Rationale for the Task Group
118 A. Brief description of MRI-based HDR brachytherapy requirements
119 Brachytherapy (BT) has a long history in the treatment of cancer, with its initial applications
120 performed shortly after the discovery and isolation of radium from pitchblende by Pierre and Marie
121 Curie in 1898. Two dimensional (2D) imaging was the mainstay for image guidance for many years, as
122 radiographs provided sharp subject contrast and detail between objects with significantly different
123 attenuation properties. However, due to the limited differences in attenuation between different soft
124 tissue types, 2D kilovoltage (kV) imaging does pose limitations in developing optimal treatment plans,
125 for instance in the case of intracavitary applications. As such, BT treatment plans were traditionally
126 designed to deliver a desired dose to a reference point relative to the applicator geometry.
127 As computed tomography (CT) and magnetic resonance imaging (MRI) have become more widely
128 available in clinics and hospitals, BT imaging has transitioned from the use of planar to volumetric
129 imaging, in particular for high dose rate (HDR) and pulsed dose rate (PDR) approaches using
130 afterloading technology. Relative to planar radiographs, volumetric images can potentially provide
131 better visualization of tumors and adjacent normal tissues. Over the last two decades, there has been
132 increasing interest in MRI either in conjunction with CT or ultrasound (US) through a multi-imaging
133 modality approach, or alone for image guidance.1,2 Compared with CT, MR images have superior soft
134 tissue contrast resolution that has demonstrated clear advantages for clinical assessment and
135 radiotherapy treatment planning.3,4 MRI has also proven to be advantageous because it allows for
136 improved organ at risk (OAR) sparing and dose escalation to targets, adaptive imaging protocols, and
137 multi-planar scanning and reconstruction.3,5,6 Additionally, MRI is non-ionizing and does not require
138 the use of iodine-based contrast, which can cause severe reactions in patients with an allergy to
139 iodine.3
140 The basic requirements for the successful implementation of MRI for HDR BT include: 1) access to
141 an MRI scanner system, either a shared imager or dedicated radiation oncology MRI simulator,7
142 applicators or needles, stirrups, immobilization devices, and a transport table, all of which must be
143 MRI compatible. 2) Integrated MRI safety procedures with ongoing training of staff to ensure patient
144 and hardware safety. 3) An optimized workflow designed to efficiently incorporate MRI into the
145 clinical environment.
146 In general, BT workflows involve diagnosis/staging, placement of intracavitary applicator(s)
147 and/or interstitial needles, simulation imaging, treatment planning, pre-treatment implant
148 verification, treatment delivery, and review of dose delivery to ensure the treatment was delivered
149 as planned. MRI can be incorporated during most of these key steps, but for the purpose of this report,
150 recommendations focus on clinical workflows, MRI acquisition parameters, commissioning and on-
151 going quality assurance (QA), and practical considerations when implementing MRI in HDR
152 brachytherapy.
153 B. Intracavitary and interstitial applications for gynecological and prostate BT
154 Over the last decade, several advancements have been made in the application of MRI for
155 gynecological (GYN) and prostate BT. The July 2017 issue of the Brachytherapy journal focused on the
156 utilization of MRI from diagnosis to treatment assessment in LDR and HDR BT for prostate cancer. The

6
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 7 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

157 articles in this issue noted the evolution of MRI in the management of prostate cancer and its potential
158 by providing clinicians the tools to ensure high quality curative treatment for patients. Additionally,
159 gaps of knowledge, technical challenges, and barriers to MRI implementation were identified in each
160 step of the process to provide a roadmap for future MRI education, training, research, innovation,
161 and commercialization in prostate BT. For prostate cancers, MRI has been shown to be a superior
162 diagnostic tool for the evaluation of dominant intraprostatic lesions (DILs), extracapsular extension
163 (ECE), seminal vesicle invasion, and neurovascular bundle involvement.8 Also, compared with US, MRI
164 can reduce uncertainties associated with simulation, treatment planning, and treatment delivery.9
165 For GYN cancers, GEC-ESTRO (the Groupe Européen de Curiethérapie – European Society for
166 Radiotherapy & Oncology) recognized the significance of volumetric imaging in the movement toward
167 three-dimensional (3D) treatment planning, namely for cervical cancer, with the formation of the GYN
168 GEC-ESTRO work group.10-13 In the twenty years since its creation, the work group has published a
169 series of recommendations to assist in the standardization of image-based BT treatment planning.
170 This has included the definition of a common language and means of delineating the target volumes
171 (i.e., low risk clinical target volume (CTV), intermediate risk CTV, and high risk CTV for the definitive
172 treatment of cervix cancer), discussions on issues related to applicator reconstruction, and
173 suggestions on the appropriate MR imaging sequences for treatment planning. Although these
174 recommendations are helpful, their scope is limited to the experience of a few key institutions from
175 Europe, North American, and Asia, and for magnetic field strengths of ≤ 1.5 T. The published
176 International Commission on Radiation Units and Measurements (ICRU) Report 8914 provides a
177 description of current, volumetric imaging of the cervix with the addition of 4D adaptive target
178 concepts, and updated radiobiology and dose volume histogram (DVH) parameter reporting for
179 targets and OARs. It contains clinical examples of MRI-based treatment plans from European, North
180 American, and Asian centers.
181 The present report covers the assigned Task Group (TG) 303 charges focused specifically on GYN
182 and prostate, however, the principles may be applicable to other treatment sites using intracavitary
183 and/or interstitial applications. The Task Group was charged to:
184 1. Describe workflow processes when implementing MRI in HDR brachytherapy from
185 simulation to treatment for common sites such as gynecologic and prostate.
186 2. Develop recommendations for brachytherapy-specific MRI acquisition parameters.
187 3. Develop recommendations for the commissioning and on-going QA of applicators and/or
188 needles when MRI is utilized with HDR brachytherapy.
189 4. Evaluate practical considerations arising when implementing MRI in HDR brachytherapy
190 a. MR safety awareness for patients and staff when using HDR applicators
191 b. Logistical and economic considerations for initial program development and
192 maintenance
193 c. Geometric uncertainty, dosimetric uncertainty, and MRI-specific imaging artifacts
194 The recommendations presented reflect clinical practice standards for AAPM and American
195 Brachytherapy Society (ABS) members. Readers should be aware that as technologies and methods

7
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 8 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

196 evolve, the recommendations presented here may become obsolete, and this report represents best
197 practices based on the current technical knowledge of the authors.
198 The report is organized into sections and subsections based on topics of interest. Section I
199 provides a brief introduction and rationale for the utilization of MRI within HDR BT workflows. Section
200 II includes a list of nomenclature and abbreviations used throughout the report. Section III provides a
201 review of literature of various approaches to MRI guided BT, available applicators, a review of
202 gynecologic and prostate clinical trials, a review of BT related TG reports, and new image guided
203 recommendations beyond CT-based practices. Section IV describes MRI protocols and methodologies,
204 including recommendations for the clinical implementation and logical consideration for MRI
205 guidance. Section V presents an example of risk-based analysis using MRI-based, prostate HDR BT as
206 a use case. Section VI provides recommendations to medical physicists related to clinical
207 commissioning, MRI safety, process QA, and logistic and economic considerations. Section VII outlines
208 potential future developments in MRI guided BT. Lastly, Section VIII provides a brief summary of the
209 intent of the TG report.
210 Terminology used in this report is modeled after that used in other TG reports: shall or must are
211 used when the activity is required by various regulatory agencies, recommend is used when it is
212 expected that the procedure should normally be followed as described. However, there may be
213 instances where other issues, techniques or priorities could force modification of the recommended
214 practice. The term “should” is used when it is expected that local analysis of the situation may change
215 the way a particular activity is performed.
216 Specific commercial equipment, instruments, and materials are described in the current report to
217 illustrate the necessary clinical procedures. Such identification does not imply recommendation or
218 endorsement by the AAPM, nor does it imply that the identified device is necessarily the best available
219 for these procedures.

220 II. Nomenclature and abbreviations

221 • 1D – one dimensional
222 • 2D – two dimensional
223 • 3D – three dimensional
224 • ABS – American Brachytherapy Society
225 • ADC – apparent diffusion coefficient
226 • ANR – artifact-to-noise ratio
227 • AAPM – American Association of Physicists in Medicine
228 • ASTM – American Society for Testing and Materials
229 • BT – brachytherapy
230 • CT – computed tomography
231 • CTV – clinical target volume
232 • DCE – dynamic contrast enhanced
233 • DIL - dominant intraprostatic lesion
234 • DIR – deformable image registration

8
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 9 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

235 • DMBT – direction modulated brachytherapy

236 • DVH – dose volume histogram
237 • DWI – diffusion weighted imaging
238 • Dxcc – maximum dose to the most exposed xcc of a given organ at risk (e.g., 2cc volume)
239 • DXX or DXX% - minimum dose to XX or XX% of a target volume (e.g., 90%)
240 • EBRT – external beam radiotherapy
241 • EMBRACE - IntErnational study on MRI-guided BRachytherapy in locally Advanced CErvical
242 cancer
243 • ECE – extracapsular extension
244 • EPI – echo planar imaging
245 • EQD2 – equivalent dose delivered in 2 Gy fractions
246 • ERC – endorectal coil
247 • ETL – echo train length
248 • fMRI – functional magnetic resonance imaging
249 • FDG – fluorodeoxyglucose
250 • FIGO – International Federation of Gynecology and Obstetrics
251 • FMEA – failure modes and effects analysis
252 • FOV – field-of-view
253 • FSE – fast spin echo
254 • GEC-ESTRO – the Groupe Européen de Curiethérapie – European Society for Radiotherapy &
255 Oncology
256 • GNL – gradient nonlinearity
257 • GTV – gross tumor volume
258 • GU – genitourinary
259 • GYN – gynecological
260 • HDR – high dose rate
261 • ICRU – International Commission on Radiation Units and Measurements
262 • IGABT – image guided adaptive brachytherapy
263 • IMBT – intensity modulated brachytherapy
264 • IsoFSE – isotropic fast spin echo
265 • kV - kilovoltage
266 • LDR – low dose rate
267 • MBDCA – model-based dose calculation algorithm
268 • mp-MRI – multi-parametric magnetic resonance imaging
269 • MR – magnetic resonance
270 • MRI – magnetic resonance imaging
271 • NRG – This is a national clinical trials network group. The acronym NRG is derived from the
272 names of its parental groups, the National Surgical Adjuvant Breast and Bowel Project (NSABP),
273 the Radiation Therapy Oncology Group (RTOG), and the Gynecologic Oncology Group (GOG)15

9
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 10 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

274 • OAR – organ at risk

275 • MAR – metal-artifact reduction
276 • PDR – pulsed dose rate
277 • PDW – proton density weighted
278 • PET – positron emission tomography
279 • PIBS – posterior-inferior border of the pubic symphysis
280 • PI-RADS – prostate imaging-reporting and data system
281 • pMRI – parallel magnetic resonance imaging
282 • PSI – prostate seed implant
283 • PTV – planning target volume
284 • QA – quality assurance
285 • QC – quality control
286 • QM – quality management
287 • QMP – qualified medical physicist
288 • RAPS – real-time applicator position monitoring system
289 • rBW – readout bandwidth
290 • RF – radiofrequency
291 • RPN – risk priority number
292 • ROI – region of interest
293 • RTOG – Radiation Therapy Oncology Group
294 • SAR – specific absorption rate
295 • SE – spin echo
296 • SNR – signal-to-noise ratio
297 • SPACE – sampling perfection with application optimized contrasts using different flip angle
298 evolution
299 • T1W – T1 weighted
300 • T2W – T2 weighted
301 • TE – echo time
302 • TG – task group
303 • TPS – treatment planning system
304 • TR – repetition time
305 • TRUS – trans-rectal ultrasound
306 • VISTA – volume isotropic turbo spin echo acquisition
307 • US – ultrasound

308 III. Background: Review of literature on MRI utilization for HDR BT

309 A. Summary of various approaches
310 CT is the most commonly available imaging modality in radiation oncology, and is widely used for
311 BT planning. While CT is adequate for OAR delineation, GYN and prostate target contours have been

10
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 11 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

312 shown to be larger on CT as compared to MRI.16,17 For GYN HDR, MRI is considered to be the gold
313 standard for target volume delineation, facilitating dose optimization to the target and adaptive
314 planning. For prostate HDR, use of MRI is not yet considered a standard-of-care due to resource and
315 infrastructure needs, complexity of an MRI integrated workflow, and reimbursement.2,18,19 New
316 approaches for prostate HDR BT have incorporated MRI for its improved visualization which can be
317 used to boost the prostate or intra-prostatic volumes, as adjuvant therapy as well as monotherapy.4,20
318 When considering the integration of MRI within an HDR BT workflow, one should review the
319 intended utilization (e.g., diagnosis, tumor response, implant guidance, post implant assessment and
320 adjustment, and/or treatment planning), the available technology, as well as the time and staff
321 resources that may be available. There are different approaches to incorporating MRI into the BT
322 workflow as described in Section IV.A.
323 Regardless of the approach, it is important for users to be aware of the available MRI technology,
324 namely the geometry of the magnetic bore and the magnetic field strength as these properties can
325 affect the implementation of MRI-based/-guided HDR BT.17,21
326 i. Open bore MRI with <1.5 T magnetic strength
327 Open MRI scanners are designed to allow the patient to lie between the two magnetic poles.
328 Compared to traditional MRI scanners, they offer greater patient comfort (reduced claustrophobia),
329 greater accessibility to patients especially during interventional procedures, accommodation of larger
330 patients, and greater variability in patient positioning. In terms of BT, the open configuration may be
331 of particular interest for the guidance of interstitial implants, as it allows for direct access to the
332 patient. The use of an open MRI unit has been reported for MRI-guided interstitial GYN BT.22,23 As
333 described in these references, the open MRI unit is located in a surgical suite and has a 60 cm bore
334 with a 56 cm gap producing a field of 0.5 T. However, the use of open MRI systems is not widespread
335 and no longer commercially available due to their limited field strength (≤ 1.2 T), which results in lower
336 image quality (e.g., lower signal-to-noise ratio (SNR), lower contrast and contour sharpness, lower
337 resolution), large geometric distortions due to field inhomogeneity and gradient nonlinearity, and
338 greater motion artifacts due to the longer acquisition times compared to their closed bore
339 counterparts.4,17,21,24 For the duration of this report, the discussion will be limited to the use of closed
340 bore MRI scanners alone or in combination with other imaging modalities, which will be termed as a
341 multi-imaging modality approach.25
342 ii. Closed bore MRI with 1.5 T or 3.0 T magnetic strength
343 Closed-bore MRI scanners at 1.5 T or 3.0 T are now widely available. The scanners have a
344 cylindrical design with a standard central bore averaging 60 cm, and a wide bore averaging 70 cm in
345 diameter. Even with wide bore scanners, accommodating the lithotomy position for implantation is
346 difficult. To address this issue, several potential imaging workflows are presented in Section IV.A.
347 Currently, 1.5 T MRI systems are more commonly available than 3.0 T systems, and have been
348 used to provide images with sufficient quality for GYN and prostate BT. However, there is increasing
349 interest in 3.0 T systems as they provide approximately twice the SNR, increased spatial resolution,
350 shorter scan times, and improved contrast-to-noise ratio.26 Transitioning from a 1.5 T to 3.0 T system
351 should not be done without testing and optimization, due to the enhancement of magnetic

11
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 12 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

352 susceptibility artifacts, chemical shift artifacts, standing wave effects, and the heating potential
353 observed for 3.0 T systems (see Section IV.G).
354 The use of MR imaging from closed bore MRI-linacs with magnetic field strengths of < 1.5 T has
355 been investigated for cervical HDR BT.27 However, this implementation is still developing, and as such,
356 is outside the scope of this report.
357 iii. Multi-imaging modality approach: MRI/Ultrasound (US), MRI/CT
358 Few radiation oncology clinics have a dedicated MRI scanner. Others may have to rely on
359 diagnostic radiology departments to coordinate MR imaging for their BT patients. Even for those
360 departments with a dedicated MRI(s), not all BT planning simulations may be performed on the MRI
361 due to the designed procedural workflow, logistical, and reimbursement issues. In these scenarios, a
362 multi-imaging modality approach may be selected that utilizes multiple imaging modalities such as
363 MRI and US or MRI and CT for image guided brachytherapy. There are several permutations that may
364 be used to integrate MR imaging into the BT workflow (see Section IV.A), each exploiting the superior
365 soft tissue contrast of MRI compared to CT and US to better visualize the pelvic anatomy and identify
366 the target tissues. Examples of multi-modality approaches include MRI-informed BT and MRI-based
367 BT.17 During MRI-informed BT, diagnostic or pre-implant simulation MR images may be reviewed
368 during the implant procedure to help guide needle placement, and reviewed and/or fused with the
369 planning CT or US to assist with the delineation of the target. In the case of MRI-based BT, an MRI
370 and CT dataset may be acquired following applicator/needle placement and registered based on the
371 applicator/needles.
372 B. MRI applicators and needles
373 The earliest examples of the application of MRI-based GYN BT were from the GEC-ESTRO working
374 group which provided recommendations for intracavitary tandem and ovoids and tandem and ring BT
375 for the treatment of cervical cancer.10-13 Since that time, MRI-compatible hybrid applicators, i.e.,
376 combined intracavitary-interstitial applicators like the Vienna,28 Utrecht,29 and Venezia30 applicators
377 have become commercially available for GYN HDR BT. Although not as common, there are also
378 examples of MRI-based HDR BT for post-operative endometrial cancers using vaginal single or multi-
379 channel cylinders and for inoperable endometrial using Heyman capsules or tandem(s) and/or
380 cylinders.31-33 Examples of template-based interstitial needle implants using, for example, the Syed-
381 Neblett template, have been published for MRI-guided GYN BT.22,23 There are also examples of more
382 customized applicators, including a 3D-printed vaginal template or hybrid applicator,34 a modified
383 tandem and colpostat,35 and custom-built MRI-conditional metallic alloy tandem applicator36 for MRI-
384 based BT in cervical cancer. With regard to prostate cancer, an example of MRI-guided placement of
385 an interstitial implant using a custom perineal template has been published for the treatment with
386 HDR BT.37 For MRI-based interstitial GYN and prostate implants, needles made of plastic or non-
387 ferromagnetic metals (e.g., titanium) are typically used.

12
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 13 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

388 C. Review of current MRI BT clinical trials

389 i. Prostate BT
390 Clinical prostate cancer trials combining HDR and MRI are fairly limited; most trials are ongoing
391 or in their early stages. Based on a search of the [Link] website using the terms MRI, HDR,
392 and prostate cancer several studies have been identified.
393 “MRI-guided HDR BT for prostate cancer”38 is a pilot study testing the technical and clinical
394 performance of MRI-guided prostate HDR BT. In Arm 1 of the study, patients with locally recurrent
395 prostate cancer receive targeted salvage HDR BT after EBRT. In Arm 2, patients with locally advanced
396 prostate cancer receive an HDR BT boost during EBRT.
397 The “prospective Phase II trial of single fraction real-time High-Dose-Rate BT in patients with low
398 and intermediate risk prostate cancer”39 is evaluating the safety, tolerance, and impact on quality of
399 life of single fraction, 19 Gy prostate HDR BT in patients with low and intermediate risk prostate
400 cancer. The trial involves the acquisition of T2 axial, pre-implant MR images that are imported and
401 fused with planning transrectal US images.
402 “HDR brachytherapy used as monotherapy for low and intermediate risk prostate cancer”40 is a
403 Phase II randomized trial evaluating HDR BT in 1 versus 2 fractions as monotherapy for the treatment
404 of low and intermediate risk prostate cancer. In one arm of this study, patients receive a single fraction
405 of 19.5 Gy, while in the second arm, patients receive a total of 29 Gy delivered in 2 fractions separated
406 by 6 hours with a single implant. The implant is performed under US guidance, but the treatment plan
407 is optimized based on post-implant MR imaging.
408 The “MRI assisted focal boost integrated with HDR monotherapy study in low and intermediate
409 risk prostate cancer patients (MARS)”41 is a pilot study investigating the feasibility and toxicities of an
410 integrated focal boost using whole gland prostate HDR BT. Patient eligibility is determined by multi-
411 parametric MRI (mp-MRI) to identify the dominant intraprostatic lesion (DIL), which is fused with the
412 preplanning US dataset. The prescribed dose is 19 Gy to the whole gland and 22.5 Gy to the MRI visible
413 lesion delivered in a single HDR fraction.
414 The “dominant intraprostatic lesion boost with focused LDR BT integrated to whole prostate
415 HDR BT: Safety and feasibility analysis”42 is a dose finding Phase I/II study. The protocol combines a
416 focused LDR BT boost with whole gland, single fraction HDR for patients with low and intermediate
417 risk prostate cancer and DILs which are visible on mp-MRI (Prostate Imaging-Reporting and Data
418 System [PI-RADS] 4 or 5). Patients enrolled on this protocol receive 19 Gy to the whole gland with HDR
419 and a concurrent LDR BT boost to the DIL. The primary endpoint for this study is early toxicity.
420 “HDR brachytherapy plus stereotactic ablative prostate radiotherapy for patients with
421 intermediate and high-risk prostate cancer”43 is a Phase II study. The intent of the study is to access
422 the impact on quality of life and tolerability of combining HDR BT and stereotactic radiotherapy in the
423 treatment of prostate cancer. Patients on this study receive a single HDR fraction of 15 Gy to the
424 followed by 5 fractions of 5 Gy (total dose of 25 Gy) to the prostate with stereotactic radiotherapy 2
425 – 4 weeks following BT.
426 The “focal salvage HDR brachytherapy for the treatment of prostate cancer”44 is a pilot study
427 investigating the feasibility and toxicities (e.g., acute and late urinary and rectal, biochemical disease

13
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 14 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

428 free survival, quality of life) associated with focal HDR BT. In this single arm study, patients will receive
429 2 fractions of HDR BT to the prostate of 13.5 Gy each, spaced 1 week apart.
430 ii. Gynecologic BT
431 For GYN cancer, the ICRU has updated their classical 1985 report 3845 with ICRU report 89.14 The
432 updated report provides an excellent description on the use of volumetric imaging (MRI and CT) for
433 cervical cancer with the addition of 4D adaptive target concepts, updated radiobiologic dose
434 assessment, and recommended DVH parameters for targets and OARs.14 Several of the updated
435 guidelines were based on GEC-ESTRO recommendations,10-13 taking into account the ABS consensus
436 guidelines for locally advanced carcinoma of the cervix1,46 and the state of the art for clinical
437 implementation as well as research oriented approaches. MR imaging was introduced into clinical
438 practice to assist in identifying the optimal implant technique (e.g., intracavitary, interstitial, hybrid
439 intracavitary/interstitial) and allows for improved visualization of targets and OARs for treatment
440 planning. Several centers from different regions of the world reported excellent feasibility and
441 promising clinical results.47-49
442 The prospective EMBRACE I (An IntErnational study on MRI-guided BRachytherapy in locally
443 Advanced CErvical cancer) multicenter protocol, which recorded MRI-based treatment planning
444 parameters and outcomes, and the retrospective retroEMBRACE allowed a comprehensive analysis
445 of the advantages of Image Guided Adaptive Brachytherapy (IGABT).50 Outcome data with MRI-based
446 planning and optimized treatment strategies are excellent in limited and well responding tumors
447 demonstrating improved local control and decreased morbidities in comparison to historical 2D
448 planning methods. Key findings include an improvement in overall survival by 10% when compared to
449 traditional cohorts across all FIGO stages, as well as up to a 50% decrease in major morbidities.51 The
450 late rectal morbidity appears to be lower when D2cc ≤ 65 Gy versus ≥ 75 Gy EQD2, and the bladder
451 morbidity appears lower when D2cc ≤ 80 Gy versus ≥ 90 Gy EQD2, even though the high risk CTV is
452 dose escalated with IGABT.52 Based on the outcomes from the retroEMBRACE and EMBRACE I
453 protocols, the EMBRACE research group has initiated the EMBRACE II protocol with the intention of
454 using state of the art treatment techniques for external beam and BT to enhance local, nodal, and
455 systemic control while minimizing normal tissue toxicity.52 Different from the previous EMBRACE
456 trials, the EMBRACE II protocol hypothesizes further improvement of outcome and morbidities
457 through the use of clearly defined dose and volume constraints for MRI-based treatment plans.
458 Additionally, the EMBRACE II protocol includes vaginal reference points (i.e., the posterior-inferior
459 border of the pubic symphysis (PIBS) < 65 Gy EQD2).
460 The NRG GY006 clinical trial is a randomized phase II/III trial of "radiation therapy and cisplatin
461 alone or in combination with intravenous triapine in women with newly diagnosed bulky stage IB2, II,
462 IIIB, or IVA cancer of the uterine cervix or stage II-IVA vaginal cancer.”53,54 This trial represents the first
463 NRG trial which requires image based planning for GYN BT. For volume-based BT, a pelvic MRI (≤ 3
464 mm slice thickness) is required for either the first or second insertion with an MRI-compatible
465 applicator. Subsequent insertions may use CT or MRI for planning.
466 The NRG GY017 clinical trial is a phase I study whose primary objective is to determine whether
467 differences in sequencing of atezolizumab and chemoradiation result in differential immune

14
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 15 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

468 activation in cervical cancer patients with FIGO stages IB2/IIA with positive para-aortic nodes or stages
469 IIB/IIIB/IVA with positive pelvic or para-aortic lymph nodes.55,56 The radiation component of this trial
470 involves EBRT followed by LDR, PDR, or HDR BT. This trial requires volumetric imaging, and encourages
471 the use of volume-based treatment planning in accordance with GEC-ESTRO/EMBRACE II dose
472 constraints.57
473 D. Review of existing Task Group reports on HDR and MRI
474 The AAPM has historically provided guidelines for commissioning and QA related to HDR BT. TG
475 41 described remote afterloader technology;58 TG 56 introduced a code of practice for BT physics
476 including QA for HDR remote afterloading units and HDR treatment planning and evaluation;59 TG 59
477 provided recommendations on safe HDR BT delivery;60 TG 40 included comprehensive QA guidelines
478 of BT;61 and TG 53 presented comprehensive QA and commissioning for treatment planning systems
479 (TPS), including BT.62 Although these TG reports provide a conceptual framework for use of MR images
480 in BT, they focus on technical issues related to 2D image-based BT procedures. The significant
481 advancements in clinical use of MR images such as MRI-guided implants, MRI-based treatment
482 planning, delivery, and verification in HDR BT have eclipsed the previous TG reports as it was
483 financially prohibitive.
484 Task Group reports that address the use of MRI include MRI Subcommittee TG 1,63 TG 118,64 and
485 TG 132.65 MRI Subcommittee TG 163 presented recommendations on the general aspects of
486 performing acceptance tests, commissioning, routine QA, and phantoms for MRI scanners. The tests
487 recommended by TG 1 for acceptance, commissioning, and QA of MRI scanners should be performed
488 prior to implementing the recommendations of the current report. TG 1 also provides
489 recommendations for acceptable MRI phantom materials, designs, and analysis procedures, which
490 can be referenced when developing MRI BT QA phantoms66,67 for applicator commissioning to assess
491 their reconstruction accuracy. Task Group 11864 introduced the use of parallel imaging in MRI in terms
492 of its technology, applications, and quality control (QC). TG 11864 addressed the issue of delays to
493 acquire MRI data within a clinically reasonable timeframe, and the introduction of parallel MRI (pMRI)
494 allowing physicians to assess large volumes of MRI data in a time-efficient manner. The use of pMRI
495 is still in the pioneering stage in radiotherapy with few studies conducted on its application.68 As the
496 use of mp-MRI in BT (e.g., diffusion-weighted imaging [DWI]69,70- and dynamic contrast enhanced
497 [DCE]-MRI71) is investigated, the application and QC of pMRI may also be considered after
498 implementing MRI in an HDR BT program. Task Group 13265 described the use of image registration,
499 fusion algorithms and techniques in radiotherapy, including current approaches and
500 recommendations for QA and QC of rigid and deformable image registration (DIR) processes, in the
501 context of external beam radiotherapy.65 This TG65 is beneficial as a reference when primary MRI
502 datasets are registered to secondary 3D image dataset for treatment planning or HDR delivery
503 through rigid registration or DIR. In this scenario, registration could follow the geometry-based metric
504 formalism of TG 132, in which at least three points are defined along the in situ applicators in both
505 sets of images. Registration is then performed to minimize the sum of squared differences between
506 corresponding points to align the applicators in the images. However, in the context of MRI-based

15
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 16 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

507 HDR BT, commissioning and QA aspects of 3D image registration as a secondary dataset for treatment
508 planning or HDR delivery were not specifically addressed in TG 132.
509 The AAPM and GEC-ESTRO TG 192 Report72 briefly discussed aspects of MRI guidance for prostate
510 cancer but their discussions are limited to robotic prostate seed implant (PSI) LDR BT, and were
511 focused on the introduction of MRI-guided robotic BT systems for prostate cancer. This report did not
512 address commissioning, QA, and QC aspects of MRI-based HDR programs.
513 E. New recommendations beyond CT based practices
514 According to the 2014 ABS practice pattern survey,16 the utilization of MRI in GYN HDR BT has
515 increased in North America from 2% in 2007 to 34% in 2014. Additionally, within this timeframe, an
516 increase in volume-based target delineation (i.e., high risk CTV) from 14% to 52% was observed.
517 Recommendations for MRI and volume-based treatment planning have largely been based on the
518 guidance of GEC ESTRO,10-13 which are based on magnetic field strengths of 1.5 T or lower. However,
519 the use of higher magnetic fields, i.e., 3.0 T, has been reported in the literature.67,73,74 Further, given
520 the availability of universal health care in many European countries, reimbursement, which can play
521 a significant role in healthcare decisions in the United States, was not factored into the
522 recommendations of GEC ESTRO. As a result, there is a clear need for additional recommendations
523 beyond those applicable to standard CT-based HDR BT to address MRI integration in the HDR BT
524 workflow.

525 IV. Clinical implementation of MRI for HDR BT

526 A. General Workflows from Simulation to Delivery
527
528 The general workflow of prostate and GYN HDR BT procedure involves six steps, with various
529 permutations in the order of execution: (1) disease staging/patient selection, (2)
530 applicator(s)/needle(s) placement, (3) treatment simulation, (4) treatment planning, (5) treatment
531 delivery, and (6) post-treatment response assessment/surveillance. For steps #2-5, where physicists
532 have relatively heavier involvement, the traditional imaging modality has been CT and/or US, and is
533 reflected in many of the professional society guidance reports.59-62 MRI, on the other hand, had been
534 limited in its use but the evidence of its usefulness is accumulating, especially for target volume
535 delineation and for identification of DILs during steps #2-4. Various approaches to incorporating MRI
536 into the BT workflow have been proposed, and Wang et al.17 broadly categorized them as MRI-
537 informed, MRI-based, and MRI-guided. It should be noted that the implementation of MRI into HDR
538 BT will require expertise and efforts from multiple disciplines (e.g., anesthesia, radiology, radiation
539 oncology) before a safe clinical workflow may be established. Since open bore MRI scanners are no
540 longer commercially available, the following discussion will be limited to workflows utilizing closed
541 bore MRI scanners.
542 In MRI-informed BT, diagnostic MRI and/or pre-implant MRI simulation images may be used to
543 guide needle placement for interstitial implants and target delineation for treatment planning. This
544 may be achieved by reviewing MR images in the operating or BT suite during the implant procedure,
545 and/or by importing and registering the pre-implant MRI dataset with the planning CT or US images.

16
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 17 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

546 This approach has been described for both GYN21,75,76 and prostate HDR BT.17,77 However, care must
547 be taken when utilizing this approach due to anatomical variations and deformations introduced
548 when comparing datasets acquired at different time points with and without the presence of the
549 applicator(s)/needles. An additional margin of error should be considered to account for these
550 uncertainties.
551 For MRI-based BT, the most common approach in North America involves the acquisition of MRI
552 alone or with CT datasets following applicator/needle placement (often guided by US for GYN78 and
553 prostate4,19) and registration based on the applicator/needles. This multi-modality imaging approach
554 is often utilized when MRI access is limited. For example, in the case of GYN BT, the MRI is commonly
555 acquired early in the BT process (e.g., during the first treatment fraction), and registered to CT images
556 for subsequent treatment fractions.4,21,79-81 The MRI dataset is used for target delineation, while the
557 CT is typically used for applicator reconstruction and to define the critical structures (Figures 1 and 2).
558 However, care must be taken when using this approach due to potential variations in the position of
559 the applicator as well as the anatomy as a result of organ, patient, and applicator motion as the patient
560 is transferred between the two imagers. An additional margin of error should be considered to
561 account for these uncertainties. User should also be cautious of the differences in the presentation of
562 the applicator(s) between imaging modalities. As shown in Figure 1, a gap is visible at the tip of the
563 tandem on CT due to the presence of a cervical sleeve, but appears uniformly hypo-intense on the
564 corresponding MRI, highlighting a potential pitfall of using MRI-only for tandem localization with a
565 sleeve (challenges are detailed in section IV.B and C). Further, the applicator is easier to discern on CT
566 than MRI, especially for metallic applicators and at higher field strengths.
567 The ideal and recommended clinical approach for MRI-based BT would involve MRI only
568 simulations for each HDR treatment fraction. Unlike the multi-modality approach, patient transfers
569 are kept at a minimum using an MRI only approach, thereby minimizing uncertainties associated with
570 applicator and organ motion, as well as uncertainties related to image registration. Further, the
571 superior soft tissue resolution of MRI compared to CT allows for the clear delineation of the tumor
572 from neighboring pelvic tissue (e.g., cervix, uterus, parametrium, vaginal tissue).14 This approach is
573 routinely used in Europe and some academic centers in North America, however, due to resource and
574 financial constraints, has not gained wide acceptance in North America.4,80
575 Finally, in MRI-guided BT, the applicator and needle insertion are guided intraoperatively in the
576 MRI/operating room. The initial experience with using a 0.5 T MRI during prostate LDR BT was
577 reported by the Dana Farber Cancer Institute, who treated 318 men from 1997 through 2007 using
578 an intraoperative real-time MRI guided BT technique targeting only the peripheral zone of the gland.82
579 The rationale for this resource-intensive approach included better visualizations of the gross tumor
580 and prostate boundaries than TRUS where uncertainties remain at the anterior aspect of the gland
581 because of echogenic artifacts created by implanted catheters, sacrificing workflow efficiency in the
582 hopes of improving accuracy. Therefore, this approach is best suited for salvage BT or dose escalation
583 to DILs. This approach has been shown to be feasible at several specialty centers, with overall
584 procedure times of 4-5 hours.82,83 Ménard et al.37 and Murgic et al.84 reported MRI-guided needle
585 insertion approaches for patients treated with prostate HDR BT in which the patient is positioned in

17
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 18 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

586 the decubitus and frog-legged positions, respectively, then moved in the bore of the scanner for
587 imaging, and removed for needle insertion and/or adjustments. For GYN BT, the Medical University
588 of Vienna first reported using MRI-guided applicator and needle insertions for difficult cases involving
589 extensive disease where patients were pulled in and out of the magnet bore between insertions using
590 a 0.35T MRI.85 Several other workflows for BT with the patient in the bore of the scanner for imaging,
591 and then withdrawn from the bore for implantation have been published in literature.37,84,86 Kapur et
592 al.86 described an MRI-guided insertion workflow for interstitial gynecologic BT, in which the patient
593 is in the lithotomy position with their pelvis outside of the scanner bore for needle placement, and
594 then with their legs partially down for imaging. Needle adjustments were performed by the physician
595 reaching inside the scanner bore to access the patient. More recently, Anderson et al.87 described an
596 MRI-guided needle insertion technique using T2W MR images with a short scanning time in an
597 interventional radiation oncology MRI-BT suite. Such MRI-guided insertion techniques are resource-
598 intensive, requiring access to an MRI system with trained operators, MRI safe/compatible equipment
599 (e.g., needles, template, anesthesia equipment, stretchers, etc.) and an anesthesia team familiar with
600 MRI restrictions on anesthesia equipment.84
601 The incorporation of MRI into any of the six steps above, or in combination thereof, is feasible
602 and has been proposed in workflows in the past, irrespective of fractionation schedules. It is up to the
603 individual institution to assess their infrastructure and personnel to determine the best workflow to
604 implement. If MRI is available, but not housed within the department, an MRI-informed or MRI-based
605 approach of using CT-MRI fusion to guide the treatment planning becomes the best option for patient
606 care.
607 B. Brachytherapy-specific MR imaging acquisition parameters
608 The development of an imaging protocol consisting of MRI acquisition sequences is dependent on
609 several factors, including the technical specifications of the MRI unit, patient preparation, and MRI-
610 compatibility of the BT applicator(s); the last of which in turn affects the interaction between imaging
611 and the applicator(s) leading to heating effects, image distortions, and artifacts. A combination of
612 phased-array RF receive coils is often utilized. Typically, the patient is positioned above a spine array
613 coil, integrated into the scanner couch, which provides posterior signal. Prior to imaging, a surface
614 array coil is positioned over the patient and lightly secured with straps. In general, the highest density
615 coil combination is desired to maximize SNR and acceleration. Use of RF coil bridges is usually
616 unnecessary for MRI-based BT because deformation of the surface anatomy is not a concern.
617 Acronyms and definitions of pulsed sequences from the major MRI vendors are shown in Table 1.
618 These sequences permit the collection of high-resolution, T2 weighted (T2W) volumetric images with
619 isotropic voxels within clinically acceptable scan times. The one dimensional (1D) or 2D parallel
620 imaging and partial Fourier techniques are often used to reduce acquisition times. The 3D fast spin
621 echo (FSE) sequences utilize very long echo trains with non-selective refocusing pulses of (typically)
622 variable flip angles to reduce tissue heating.88 Compared to multi-slice 2D sequences, images acquired
623 with isotropic 3D sequences can be reformatted into any arbitrary plane with relatively little loss of
624 image quality, permitting BT-eye-view reconstructions from a single acquisition. However, the flip
625 angle schedules of the refocusing pulses used in 3D FSE are designed using spin models with assumed

18
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 19 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

626 relaxation times which may not be representative of the relaxation times of tissues in the pelvis.88
627 Consequently, the T2 contrast on 3D FSE images is often altered compared to conventional, multi-
628 slice 2D FSE images. Table 2 compares the advantages and disadvantages of 2D and 3D FSE sequences
629 for BT planning, and Table 3 provides a set of generalized 2D and 3D FSE scan parameters (e.g.,
630 repetition time [TR], echo time [TE], echo train length [ETL], and readout bandwidth [rBW]) for 1.5
631 and 3.0 T systems for GYN and prostate BT. Beyond establishing the desired field-of-view (FOV) and
632 spatial resolution, this includes increasing the rBWs to reduce applicator-induced distortions arising
633 from magnetic susceptibility differences and controlling for gradient nonlinearity-induced distortions.
634 Susceptibility causes distortions in the local magnetic field (B0), and the resulting shifts in spin
635 frequency causes shifts and distortions along the frequency encoding or readout direction. To control
636 for applicator or needle induced geometric distortions, it is advisable to set sequence rBWs to twice
637 the fat-water shift at a particular magnetic field strength (e.g., 440Hz at 1.5T and 880Hz at 3T).89 It
638 should be noted that such a high rBW will result in substantial loss of SNR (approximately 30% or
639 more). Therefore, increases in the number of signal averages, use of phase and slice oversampling, or
640 reduction in parallel imaging acceleration factors may be required to recover SNR.90,91 Each of these
641 changes will result in increases in scan times. To minimize unnecessary SNR loss and scan time
642 increases, the rBW should be tailored based on an estimated maximum susceptibility induced
643 distortion map. This can be obtained by acquiring a B0 map of the applicator or needles during MRI-
644 based BT commissioning. The rBW can then be determined using the following equation:
𝐻𝐻𝐻𝐻 𝑀𝑀𝑀𝑀𝑀𝑀 ∆𝐵𝐵0 (𝑝𝑝𝑝𝑝𝑝𝑝) ∗ 𝑓𝑓0 (𝑀𝑀𝑀𝑀𝑀𝑀) ∗ 𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅 𝐹𝐹𝐹𝐹𝐹𝐹 (𝑚𝑚𝑚𝑚)
645 𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅 𝐵𝐵𝐵𝐵 � �=
𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 𝑆𝑆ℎ𝑖𝑖𝑖𝑖𝑖𝑖 𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴 𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅 (𝑚𝑚𝑚𝑚) ∗ 𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅 𝑀𝑀𝑀𝑀𝑀𝑀𝑟𝑟𝑖𝑖𝑖𝑖 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆
646 where f0 is the system frequency, and Readout FOV and Readout Matrix Size are the desired FOV and
647 matrix sizes, respectively.
648 Gradient nonlinearity (GNL)-induced distortions are expected to be small for MRI-based BT
649 because of the close proximity of the target volumes to the MRI scanner isocenter. GNL distortions
650 can be further minimized by ensuring that the center of the prescribed imaging volume matches the
651 MRI scanner isocenter along the table direction. In addition, vendor-provided 3D gradient distortion
652 correction should be applied prior to using the images for treatment planning.92
653 Metal-artifact reduction (MAR) sequences, which are becoming more widely available, have been
654 successfully applied to MRI of titanium applicators.74 These sequences were found to improve
655 visualization of the tandem tip on both T2W- and proton density weighted (PDW)-MRI sequences,
656 with a significant reduction in the blooming artifact at the tip of the tandem in PDW MRI.74 In all of
657 these studies, PDW-MRI was recommended as an addition to, and not as a substitute for, standard
658 T2W-MRI to improve visualization of the applicator.
659 Other sequences have also been utilized for improved reconstruction of the applicator and/or
660 detection of seeds or markers. These include 3D T1 weighted (T1W) gradient echo sequences such as
661 VIBE or LAVA,66,91 or 3D phase cycled steady-state free precession sequences such as CISS, FIESTA-C,
662 or balanced FFE.91,93 These sequences are generally fast compared to 3D FSE sequences but produce
663 less desirable contrast of the anatomy and lesion unless used with contrast agents.

19
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 20 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

664 C. Applicator/needle commissioning and reconstruction

665 When considering the selection of BT applicators that will be used in a clinical workflow involving
666 MR imaging, the user must verify the device is safe in a high magnetic field environment. Details on
667 safety risks and testing are provided in Section IV.G. As a component of applicator commissioning, the
668 user should review and be familiar with the applicator’s instruction for use and the conditions in which
669 the applicator is safe for use within an MRI environment (see Section [Link]). Additionally, regardless
670 of whether an MRI and its associated protocols are fully optimized for diagnostic purposes or for
671 external beam to serve as a secondary image dataset, during clinical commissioning MR images and
672 protocols should be cautiously validated in terms of geometric accuracy for plastic and/or metal
673 applicator(s)/needles (as discussed in Section IV.B). The level of distortions and general image artifacts
674 such as truncation, aliasing, magnetic susceptibility,36 and chemical shift artifacts94 should be
675 validated for the specific array coils used in the presence and absence of applicator(s)/needles.
676 Commissioning should also include an evaluation of the reconstruction accuracy of the
677 applicator(s) intended to be used in an MRI workflow. Applicator reconstruction can be affected by
678 distortions and susceptibility-related artifacts. The first issue is a valid concern in both plastic and
679 titanium applicators/needles, as distortion levels of 1-2 mm around the active dwell positions of a
680 GYN intracavitary applicator could cause > 2 mm uncertainties in applicator reconstruction which is
681 larger than the ICRU report # 89 recommended accuracy.14 On the other hand, susceptibility-related
682 artifacts are an issue for non-ferromagnetic metallic applicators/needles, and increases as the
683 magnetic field strength increases.4 The evaluation of applicator/needle reconstruction accuracy on
684 MRI requires the use of a unique QA phantom since the applicator/needles need to be embedded in
685 tissue-equivalent materials (e.g., Agarose gel95) to be identified on an imaging dataset. The phantom
686 should be composed of an MRI safe material (as discussed in Section III.D) that can fixate the
687 applicator and includes a coordinate system independent of the applicator that is used to register the
688 MRI dataset(s) acquired with the applicator to the gold standard imaging dataset (e.g., CT).66,67 In-
689 house MRI-QA phantoms designed for intracavitary applicators66,96 or needles96 have been discussed
690 in literature, as well as techniques to evaluate the accuracy of applicator reconstruction using CT and
691 MR imaging. Details of applicator reconstruction commissioning for 3D image (e.g., CT or MR images)
692 based treatment planning are described in Hellebust et al.13 For 1.5 - 3.0 T MRI scanners, applicator
693 reconstruction uncertainty should be less than 2 mm.14,66,67 If the uncertainty exceeds 2 mm, efforts
694 should be taken to improve the image quality such as reducing the slice thickness, optimizing MRI
695 sequences to minimize artifacts or distortion, and/or using MRI markers (i.e., catheters containing
696 MRI contrast agents that may assist with applicator reconstruction). Example CT and MR images
697 acquired during the commissioning of a plastic and titanium ring and tandem applicator set are shown
698 in Figure 3.
699 The use of MRI contrast agents has been investigated to improve intracavitary applicator35,66,96 or
700 needle96 reconstruction for plastic35,66 and titanium96 devices. The following MRI marker contrast
701 materials have been investigated for MRI-based BT using plastic applicators: CuSO4 solution,66 saline,35
702 cobalt-chloride complex contrast (C4),96,97 Vitamin E,96 1% Agarose gel,96 Conray-60 (Mallinckrodt

20
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 21 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

703 Medical, Pointe-Claire, QC, Canada),96 and fish oil.96 These markers cannot be visualized in titanium
704 applicators due to magnetic susceptibility artifacts.66 The choice of an MRI-marker agent should be
705 determined based on the MRI sequence (e.g., T1W- or T2W-MRI) used for applicator reconstruction.
706 For instance, CuSO466,96 and C496 are recommended when T1W-MRI are used for applicator
707 reconstruction, while saline35,96 is recommended for T2W-MRI. Since CuSO4, C4, and saline are liquids,
708 care should be taken to examine the vessel holding the MRI agent as leakage or evaporation of the
709 agent will introduce additional uncertainty in applicator reconstruction (as shown in Figure 4). MRI-
710 marker catheters are commercially available for plastic intracavitary applicators which may be filled
711 with saline for T2W-MRI (Elekta, Inc., Stockholm, Sweden) or are filled with C4 for T1W- or T2W- MRI
712 (C4 Imaging LLC, Doylestown, PA, USA). However, caution should be used in the application of saline
713 and water base markers as signal may be too weak for visualization in small diameter lumens. Example
714 images of a commercially available marker in a plastic ring and tandem applicator are shown in Figure
715 5. Efforts have been made to explore the use of MRI-marker agents for interstitial HDR BT using plastic
716 needles and presented more than 100% hyper-signals with CuSO4 and saline on T1W- and T2W- MRI,
717 respectively.96
718 Currently, most vendors provide a library of 3D applicator models for rigid applicators that
719 includes the detailed geometry and material of each applicator. Applicator libraries have been shown
720 to be more accurate and efficient than manual applicator reconstruction on volumetric imaging,13 but
721 should be used only after they have been properly commissioned.14 An MRI-marker flange for
722 titanium intracavitary applicators was proposed for use with its 3D-modeled applicator library to
723 improve the applicator reconstruction accuracy.96
724 Plastic applicators/needles have also been reconstructed on MRI in the absence of markers. Care
725 must be taken when relying on MR images alone for applicator reconstruction should air pockets be
726 present in the anatomy at the tip of the applicators/needles or in locations where applicators or
727 needles cross.
728 For titanium-based applicators/needles, no MRI-marker agent is feasible due to susceptibility
729 artifacts.66,96 For these applicators, image quality can be improved by optimizing MRI sequences and
730 acquisition techniques to minimize artifacts or distortion. This is of particular importance for a 3.0 T
731 MRI scanner as the artifact-to-noise ratio (ANR) is proportional to the main magnetic field,98 thus the
732 ANR of a 3.0 T MRI is theoretically twice that of 1.5 T scanner.98 The ANR is also proportional to the
733 voxel size, so the clinically feasible smallest slice thickness should be considered. When isotropic
734 reconstruction is achievable in a 1.5 T or higher strength MRI, the slice thickness should be maintained
735 at < 2 mm, especially when MR images will be used for applicator reconstruction. If possible, a T2W
736 MRI protocol should be developed for use in contouring and applicator reconstruction with an
737 approximate 1-mm isotropic pixel size (1 mm x 1mm x 1 mm). For cervical cancer HDR BT, standard
738 T2W-MRI has acceptable tumor-tissue contrast, but has poor uterus-tandem contrast making tandem
739 localization on MR images somewhat challenging. One approach for BT imaging of titanium
740 applicators is to incorporate PDW MRI sequences, which compared to T2W MRI, uses a very short TE
741 and a very long TR yielding images with high SNR and less susceptibility artifact.99 The use of PDW
742 MRI, versus T2W MRI, for titanium tandems, has been shown to significantly improve uterus-tandem

21
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 22 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

743 contrast and reduce blooming of the tandem diameter.99 PDW sequences are comparable in length
744 to T2W MRI scans (i.e., on the order of 5 – 10 minutes). Due to the high signal in PDW-MR images, the
745 slice thickness for 2D image acquisition protocols can be reduced. Para-sagittal acquisitions of 2.5 mm
746 slice thickness for PDW versus 5 mm thickness for T2W-MRI, have been shown to yield higher quality
747 reconstructed views of the titanium tandem as well as the plastic ovoid caps.100
748 The impact on dosimetric indices and DVHs (e.g., D2cc and D90, for OARs and targets,
749 respectively) due to uncertainties in the applicator/needle reconstruction in 3D image guided cervical
750 BT has been well characterized by various investigators.101-105 In general, the overall conclusion is that
751 if the reconstruction error is within ±2 mm (as recommended by ICRU Report # 89),14 the major
752 dosimetric indices for both the target and OARs remain within 1-2%, on average, with variations of up
753 to σ=1-3% from the average for individual cases, with the magnitude being larger for OARs than
754 targets. This is due to the geometric location of the OARs that generally in higher dose gradient
755 regions. Individual findings include Schindel et al.101, who simulated the applicator shifts and
756 concluded that in order to avoid more than a 10% variation in the relevant dosimetric indices, the
757 reconstruction uncertainty should be managed to be less than 3 mm. Tanderup et al.102 also simulated
758 the shifts in the applicator position and found that the mean changes were less than 4% per mm for
759 both OARs and targets in all directions including rotations, except in the anterior-posterior direction
760 where a mean change of 5 ± 1% per mm, in terms of the D2cc to the bladder and rectum, were
761 observed. De Leeuw et al.103 studied applicator shifts for PDR 192Ir-based BT and found that the
762 magnitude of the dosimetric impact was similar to those reported by Tanderup et al.102 and Schindel
763 et al.101 and, confirmed the impact to be more severe on OARs than targets. Such finding was also
764 confirmed by Deufel et al.105 in a recent publication.
765 It is worth noting that dosimetric uncertainties related to other geometric uncertainties (e.g.,
766 afterloader source positioning uncertainty) have been estimated in some studies to be up to 4% for
767 OARs (D2cc) or the target (D90).104 The authors have further reported that the overall uncertainty for
768 a treatment course, accounting for uncertainties in the source calibration, dose and DVH calculation,
769 geometry, contouring, and intra- and interfraction variability, may be on the order of 12% for targets
770 and 21 – 26% for the OARs.
771 For interstitial prostate HDR BT, the reconstruction uncertainties due to slice thickness were
772 evaluated for CT-based treatment planning. Under a realistic needle reconstruction error scenario
773 where all catheters were randomly distributed within the space, dose uncertainties of 0.7 – 1.7% were
774 reported.106 Acquiring MR images in multiple planar orientations (e.g., axial, sagittal, and coronal) can
775 help reduce this specific source of uncertainty.
776 D. Patient-transfer systems

777 When a HDR patient-transfer system, such as the hover-based Zephyr (Diacor, Inc., Salt Lake City,
778 Utah) or SymphonyTM (Qfix Inc., Avondale, PA) systems, or a removal table top/trolley (Philips MR
779 Therapy, Vantaa, Finland), is used to transfer a patient from the procedure suite/operating room to
780 the MRI scanning room or vice versa, the degradation of the MR image quality due to the presence of
781 the transport system should be verified. The quality evaluation should also be performed using

22
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 23 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

782 practical patient separations from the coils. The receiver coils detect signals emitted by spins in the
783 patient’s anatomy; reducing the distance between the coils and the region of interest increases SNR,
784 improving the overall image quality. However, geometric uncertainties caused by general image
785 artifacts as described above are challenging for medical physicists to quantify even using phantoms.
786 Image quality in the presence of these devices has not yet been quantitatively assessed. As such, in
787 vivo validation may be required, namely by acquiring CT and MR images with the applicator in place
788 before transitioning to MRI-based BT treatment planning.
789 E. Target and OAR Delineation using MRI

790 The geometric accuracy of patient anatomy produced from MR images is critical when MRI is
791 intended to be used for HDR treatment planning. MRI scan protocols must be optimized in terms of
792 image quality for target and OAR delineation as well as applicator/needle reconstruction (as discussed
793 in Section IV.B). Distortions due to the inhomogeneity of the static magnetic field and/or
794 nonlinearities in the gradient coils should be evaluated, along with vendor-supplied correction
795 algorithms to minimize these distortions. System related distortions are known to increase with
796 increasing distance from the MRI isocenter,89,107,108 and with increasing field strength.109 Thus the
797 geometric accuracy should be maximized in the region of dosimetric interest by positioning this
798 anatomical region at the center of the magnet,92,110 while optimizing the scan protocol. On a modern
799 3T scanner, maximum residual mean distortions of 3.2 mm were found at a radial distance of 25 cm,108
800 which may be considered clinically acceptable in the setting of prostate or GYN BT, if the pelvis is
801 appropriately aligned in the MRI scanner. In contrast, geometric accuracies may be assessed using
802 commercially available phantoms such as the QUASARTM MRID3D (Modus Medical Devices Inc.
803 London, ON N6H 5L6, Canada) or Magphan RT (Image Owl, Inc., Greenwich, NY).
804 i. Delineation of GYN structures
805 Clinical contouring definitions and considerations of target volumes on MR images for treatment
806 planning of cervical cancer have been well documented.12,14 These standard definitions for a series of
807 different gross tumor volumes (GTVs), CTV, and planning target volumes (PTVs) can be translated,
808 merged, and further elaborated for adaptive BT in cervical cancer to account for changes in the target
809 volume and anatomical changes of the OARs over the course of radiotherapy.12,14,111 Additionally,
810 these concepts can be applied to other adaptive gynecological BT when MRI-based treatment
811 planning is intended. Clinicians must follow the different target volume definitions as reported when
812 the high tumor contrast resolution of MR images is intended to deliver adaptive BT. For instance, the
813 high risk CTV is defined at time of BT and should include the residual tumor, cervix, and surrounding
814 tissue with potential microscopic disease.14 MR imaging may be utilized for contouring and treatment
815 planning in a number of different clinical scenarios as described in Section IV.A. Whenever contouring
816 is performed on MR images with the intention of adaptive BT, the use of T2W MR images is
817 recommended.14,112
818 With regard to anatomic MR imaging for GYN intracavitary BT, GEC-ESTRO describes the use of
819 multi-planar T2W-MRI acquired with pelvic surface coils as the “gold standard” for tumor and critical
820 structure visualization and as having sufficient contrast for applicator definition based on their

23
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 24 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

821 experience with 0.2-T and 1.5-T MRI units, with the use of complementary MRI sequences (e.g.,
822 contrast-enhanced T1W or 3D isotropic scans) to be considered optional.11 Thus, T2W sequences are
823 considered by GEC-ESTRO to be a mandatory sequence for BT, with emphasis on spin echo (SE) and
824 fast spin echo (FSE) techniques to reduce image acquisition time and maintain image quality in the
825 presence of local field inhomogeneities (e.g., presence of the BT applicator(s)). General scan protocol
826 parameters are provided in Table 3. However, these scan protocols may require further optimization
827 depending on the receiver coil hardware available for a specific scanner. GEC-ESTRO recommends
828 obtaining images that are oriented orthogonal and parallel relative to the applicator axis (e.g, para-
829 axial, para-sagittal, para-coronal), with imaging extents to capture at least the entire cervix, uterine
830 corpus, vagina, and parametria, and OARs.11 It should be noted that the recommendations from GEC-
831 ESTRO Working Group IV11 are based on the experience of a limited number of centers and on the
832 evidence from a limited amount of clinical imaging data. For example, with multiplanar acquisitions,
833 the reformatting may also result in the loss of image quality.113 With 3D imaging, one has the ability
834 to reconstruct images in any plane. Although considered optional by GEC-ESTRO, 3D T2W imaging is
835 also recommended with the understanding that this will generally lead to increased scanning time,
836 which increases artifacts due to motion. Thus, the GEC-ESTRO recommendations may be used to
837 provide a good starting point for MRI sequence development but will probably require some fine-
838 tuning by the use of complementary sequences based on the particular applicator type and MRI
839 scanner available.
840 There are no standard recommendations with regard to functional MR imaging (fMRI) for GYN
841 intracavitary BT; however, the utility of fMRI has been explored by different centers. It has been
842 suggested that functional imaging may further improve the accuracy of target and sub-volume target
843 delineation.113 With DWI one can identify areas of restricted water diffusion as areas of high
844 cellularity, e.g., tumors versus normal tissues. The utility of DWI and the derived apparent diffusion
845 coefficient (ADC) map has been used for tumor visualization and therapy response in cervical
846 cancer.69,70 Target volume borders can appear with more distinct edges on ADC maps than on T2W-
847 MRI.114 For pre-therapy ADC maps, with no applicator in place, regions with more restricted water
848 diffusion have been shown to correspond to areas of increased metabolic activity on 18F-
849 fluorodeoxyglucose-positron emission tomography (18F-FDG-PET) images.70 However, DWI is typically
850 acquired with echo planar imaging (EPI) sequences, which are prone to geometric distortion and
851 severe susceptibility artifacts, especially in regions near the applicator. Thus, it is generally
852 recommended that for BT imaging, DWI should be used as a supplement to, but not in place of, T2W
853 imaging.4,100,114,115 A combined approach using anatomic and functional imaging has been reported
854 where T2W images are supplemented with information from ADC maps for GTV delineation for
855 adaptive planning.100 In another combined approach using multi-parametric imaging, anatomic T2W
856 MRI is supplemented with functional imaging (e.g., DWI, DCE-MRI, and 18F-FDG-PET) reducing inter-
857 observer variability in the delineation of the GTV.71 With DCE-MRI one can identify areas of
858 homogeneous enhancement as small cervical tumors or areas surrounded by an enhancing rim as
859 large necrotic tumors. In this multi-parametric approach, DCE-MRI was most often used to modify the
860 GTV contour, and was found to be particularly helpful in visualizing residual disease involving the

24
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 25 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

861 myometrium.71 Given the recent data on the importance of dose to the GTV for local control, it has
862 been suggested that the role of fMRI for tumor delineation may expand in the future.116,117
863 The use of surface array coils wrapped around the pelvis is considered standard for diagnostic
864 pelvic MR imaging, and is also recommended for BT imaging. Intracavitary (i.e., rectal or vaginal) coils,
865 although useful for the visualization of small tumors, are not recommended for BT imaging due to the
866 deformations in organ shape after MR imaging and coil removal which would lead to uncertainties
867 between the planned and delivered dose.
868 Additional items to consider are patient preparation prior to MR imaging, including the
869 administration of antispasmodic agents, e.g., glucagon, to reduce artifacts due to bowel motion11 and
870 the use of packing soaked or filled with an MRI compatible contrast agent such as diluted gadolinium,
871 US gel, or saline water.118,119 Gauze or balloon-based packing may be used to increase the separation
872 between the BT applicator and OARs, as well as to enhance the visualization of the vagina and
873 cervix.46,118
874 ii. Delineation of prostate structures

875 Contouring on MR images for treatment planning of the prostate has been well documented in
876 the literature.18,19,120 While US and CT are still the standard imaging modalities used in the United
877 States,121 MRI-based planning has recently gained interest due to the superior soft tissue contrast and
878 resolution of MRI versus CT and US. The improved resolution results in smaller variability of prostate
879 volumes on MR, especially when compared to CT-based prostate volumes which are larger than MRI
880 and US, given uncertainties in the prostate boundaries on CT images near the seminal vesicles and
881 apex of the prostate.122-125 Prostate HDR monotherapy,126-128 salvage,129-131 and DIL focal treatment132
882 planning can benefit from MR imaging. Utilizing MR imaging, intra-prostatic tumors, macroscopic ECE
883 involvement, or seminal vesicle invasion can be identified and delineated.8 Additionally, MRI allows
884 for the visualization of the prostate-bladder interface, prostate-rectal interface, neurovascular
885 bundles, and the genitourinary diaphragm, allowing for the avoidance of radiosensitive structures
886 surrounding the target. Hydrogel is being utilized at some centers to increase the distance between
887 the prostate and rectal wall, to minimize dose to the rectum.133,134 They present as a hyperintense
888 structure on T2W sequences.135 Multi-parametric MRI has recently been used along with US-based
889 HDR real-time planning for the treatment of intermediate- and high-risk patients.132,136

890 Although MR imaging can improve the delineation of structures, large interobserver contouring
891 variabilities have been reported using both CT and T2W MR images. This variability can result in large
892 standard deviations of the prostate D90 (17 – 23%).137 Although contouring variability may be due in
893 part to poor soft tissue visualization on CT, the interobserver differences on MRI suggest that
894 additional guidelines and training in contouring is necessary when integrating MRI into the clinical
895 workflow.
896 The role of MRI for HDR prostate BT is less standardized, yet evolving as an area of
897 investigation.4,20,37,138 A multi-parametric approach including both anatomic T1W and T2W MRI and
898 the use of functional imaging sequences, DWI, DCE, and less frequently MRI spectroscopic imaging, is

25
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 26 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

899 being increasingly recognized as a standard approach in a diagnostic setting.4,139 With regard to
900 anatomic imaging for prostate HDR BT, T2W-MRI sequences provide adequate visualization of the
901 prostate gland as well as discrimination between peripheral and central zones. An endorectal coil may
902 be used for staging of prostate lesions for better visualization of intraprostatic lesions, especially for
903 low tesla MRI scanners (e.g., ≤ 1.5 T).140 However, the use of an endorectal coil is not recommended
904 for HDR prostate BT as it would introduce uncertainties in planned versus delivered dose due to organ
905 deformation.141 As an alternative, supplementing T2W-MRI with fMRI (i.e., DWI, DCE) for prostate
906 HDR BT can improve the identification of intraprostatic lesions for sub-volume boosting or recurrent
907 lesions after external beam radiotherapy for salvage treatment.142-145 Additionally, glucagon can be
908 used to reduce bowel peristalsis and motion-related artifacts.8
909 F. Heterogeneities for calculations
910
911 The general commissioning considerations for model-based dose calculation algorithms
912 (MBDCAs) for BT are detailed in TG 186.146 At present, to perform a heterogeneity-corrected dose
913 calculation, a CT dataset is required to account for the composition of various tissues in the patient.
914 However, an MRI dataset can also be used with MBDCAs with the assignment of tissue material as
915 water and regions of interest (ROIs) of air, such as air in rectum, which is feasible on MRI since air
916 produces a hypointense (dark) signal.147 For high energy sources such as 192Ir, such simplified tissue
917 assignment is acceptable.146 It should be noted that the most critical factor is the patient skin interface
918 (backscatter contribution to scatter) and the applicator material which can be accurately modeled in
919 most commercial TPS’ using solid applicator models.148,149
920 For unshielded plastic GYN applicators, minimal dosimetric changes were found using MBDCA
921 relative to the standard TG-43 formalism (a reduction of 2.1 +/- 1.1% to Point A as defined in ICRU
922 38,45 2.6 +/- 0.9% to the target D90, and 2.1 +/- 0.3% to the OARs).150 For the prostate, dose predictions
923 using MBDCAs for 192Ir HDR sources have minimal impact on the dose distribution relative to those
924 predicted by TG-43 (in water).149
925 The use of MRI-conditional shielded applicators in MR images has been commercially available
926 owing to the advances in intracavitary applicators such as a Fletcher CT/MRI shielded applicator
927 (Elekta, Inc.).151
928 Heterogeneity corrected dose calculations on MR images should be cautiously performed by: 1)
929 only accounting for high density shielding material while assuming all water conditions,151,152 2)
930 assigning bulk densities to individual segmented structures,147 or 3)converting MR images into a
931 synthetic/substitute/pseudo CT using a voxel,153-155 atlas,156,157 or multi-imaging modality
932 approach.158,159
933 G. Safety considerations

934 The MRI safety guidelines presented here are not comprehensive but are intended to supplement
935 and emphasize relevant points in the ACR Guidance Document on MRI Safe Practices and standards
936 for medical device use in MRI.160,161

26
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 27 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

937 i. Procedure room equipment

938 MRI safety of ancillary equipment and devices used during the procedure should be considered.
939 Equipment used for patient monitoring, anesthesia, medication, etc. need to be validated for use in
940 MRI, and should be kept within vendor specifications. As well, changes to applicator imaging
941 characteristics and heating susceptibility may occur when attached to immobilization devices. This
942 topic will be addressed further in an upcoming Task Group report, TG 334 (a guidance document to
943 using radiotherapy immobilization devices and accessories in an MRI environment).

944 ii. Applicator, needles, patient positioning systems

945 The main safety risks from the Radio-Frequency (RF) used under MR are tissue heating and
946 burns.162,163 Shellock’s book162 provides an excellent review of RF heating under MR. Titanium-based
947 intracavitary applicators can be used for MRI-based and MRI-guided HDR BT,36,67,73,164 along with
948 plastic-based applicators. For commercially available non-ferromagnetic metallic applicators, it is
949 strongly recommended that users verify the device is labeled MRI-conditional, review the conditions
950 in which it was tested (e.g., field strength, spatial gradient, RF fields, maximum specific absorption
951 rate (SAR)),165 and that the applicator(s) is(are) used in accordance with the vendors instructions for
952 use. If the applicator has been deemed MRI conditional for a lower magnetic field than one intends
953 to use (e.g. 1.5-T versus 3.0-T MRI), its clinical use is not recommended.67 If its clinical use needs to be
954 pursued, the commissioning procedures should follow the recommended American Society for
955 Testing and Materials (ASTM) International guidelines166-168 for RF induced heating167 and applicator
956 displacement166 and torque168 under the MRI conditions intended for clinical use. Before an in-house
957 intracavitary applicator undergoes MR imaging and is used clinically, additional tests, regardless of
958 the material composition of the applicator(s), should be performed, including sterilization-capability
959 and biocompatibility. Users should be mindful when using metal-based needles such as titanium
960 needles for MRI-based BT, even if the needle is FDA-approved and labeled MRI-conditional. Again, the
961 MRI conditions in which the needles were tested should be verified. Additionally, care should be taken
962 when implanting metallic needles. Should the needle tips come in close proximity or touch one
963 another, unexpected RF induced heat induction may occur.169 The use of plastic needles is
964 recommended for MRI-based or MRI-guided HDR BT (Figure 2).
965 One should be cautious of the presence of metallic objects, even outside of the patient during
966 MRI scans. For instance, when a hover based HDR patient-transfer system (e.g., Zephyr, Diacor Inc.,
967 Salt Lake City, UT) is used, its metal side rails should not come into contact with the patient’s arms.
968 Kim et al.170 reported an incident in which metal side rails touching a patient’s arm during an MRI scan
969 induced third degree skin burns. As such, all metallic objects on the patient-transfer system should be
970 removed before an MRI is initiated.
971 There are also some safety concerns with regard to patient positioning that are particularly
972 important when working with large patients, or with patients whose limbs are not aligned straight
973 with the bore. It should be noted that RF energy radiates from the RF coil in the bore and thus thermal

27
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 28 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

974 effects are greatest at the bore wall. Although the gradient and RF coils may themselves grow warm
975 and contribute a thermal load with sustained use, RF deposition also increases with greater proximity
976 to the RF coil. Thus pads are typically available to prevent bore contact and to ensure some
977 displacement from the bore wall. Care must also be taken to prevent skin-to-skin contact of the hands,
978 feet, and/or limbs, since the limbs can form large conductive loops that may result in thermal
979 injuries.160,161
980 iii. Patient immobilization and transfer considerations

981 To perform MRI-guided implants or MRI-based treatment planning, a patient typically needs to
982 be transferred between the MRI and HDR therapy rooms, unless MRI simulation and HDR delivery can
983 be performed in a single room. Inherently, there are concerns related to potential applicator/needle
984 displacement. The immobilization approaches may differ depending on the applicator used, and fall
985 into two categories: patient immobilization and applicator fixation. Patient immobilization in an MRI
986 environment can be challenging due to the limited bore size, but can be achieved using small leg
987 ramps as a substitute for stirrups. Such ramps should be made of MRI compatible plastic material,
988 should have a low profile to minimize the distance between the patient and the coil and keep the
989 patient’s knees from touching the bore of the scanner. An example of a commercially available ramp
990 is the CT/MRI Slessinger board (Radiation Products Design, Inc., Albertville, MN). This approach may
991 be useful for interstitial prostate as well as GYN cases, for which an external applicator fixation device
992 is not available. For instance, for cervical BT, tandem and ovoids or tandem and ring applicators are
993 often initially secured with vaginal packing such as gauze or vaginal balloons.171 Additional applicator
994 immobilization may be provided by a portable applicator supporting device (see Figure 1B. in Bou-
995 Zeid et al.172) that is not secured to a table but to the patient’s body. An articulating applicator clamp
996 with baseboard that is positioned below the patient has also been used. However, additional
997 precautions may be required with such devices as the applicator can inadvertently move and harm
998 the patient should they move (e.g., in response to a cough/sneeze) or as a result of unintended
999 movements during a patient’s transfer. Intracavitary applicators can also be secured to a hover based
1000 device with similar patient safety concerns.

1001 The degree of applicator displacement due to patient transfers is still being investigated.172-175
1002 Most of the studies172-175 use surrogates for the cervix such as bony anatomy on x-rays172,174,175 or
1003 patient-infrared markers172,173 on the skin surface to measure an applicator displacement due to the
1004 patient transfer. Gerszten et al.175 reported up to a 12 mm applicator displacement even when using
1005 an applicator immobilization device. Bou-Zeid et al.172 suggested a maximum applicator displacement
1006 of 6.7 mm (2.0 ± 1.5 mm) based on a real-time applicator monitoring system when using an in-house
1007 applicator-immobilization system, including a portable tandem and ovoids securing device and a
1008 modified transfer board. Andrew et al.174 reported a 2.3 – 3.4 mm tandem and ovoid displacement on
1009 x-ray radiographs, and found that the use of a hover based patient transfer system resulted in a
1010 statistically significant reduction in the displacement of the applicators (p < 0.01). Although some
1011 simple applicators (e.g., vaginal cylinders) can be implanted on a hover-based transfer system, this

28
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 29 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1012 may be challenging with more complex applicators (e.g., tandem and ovoids, tandem and ring) that
1013 may require greater accessibility and visibility for proper applicator placement (e.g., placement of the
1014 intrauterine tandem). Alternatively, the applicator can be implanted on a conventional procedure
1015 table and afterward, the patient can be transferred. Without using a conventional procedure table,
1016 the applicators can be implanted using a modified hover-based table system. Thus, a patient can
1017 remain on the hover device for the duration of the procedure.
1018 The use of a hover mattress (e.g., HoverMatt®, Hovertech International, Allentown, PA), or a slide
1019 board is often employed as a means for efficient patient transport between simulation and treatment
1020 tables. Precautions should be taken when using such a transfer device on patients with interstitial
1021 implants to ensure the needles are not displaced during the transfer, or transferring patients whose
1022 applicator(s) is(are) immobilized with an external fixation device clamped to the transport device. An
1023 attractive approach to minimize applicator displacement employs a dockable MRI table (e.g., Tim
1024 dockable table,176 Siemens Healthineers, Henkestr. 127, 91052 Erlangen), or a removable table
1025 top/trolley (Philips MR Therapy, Vantaa, Finland) transport system. These systems allow for the
1026 transport of the patient in the simulation position without the need for moving the patient to a
1027 different table for treatment. However, these are vendor specific and may not be available on all
1028 scanner models. Please note, considerable applicator displacements can be introduced if a transfer
1029 device needs to be inserted under a patient after the implant procedure. These displacements may
1030 be corrected at time of simulation, therefore the insertion and removal of the transfer device after
1031 simulation is not recommended.
1032 The consideration of improved workflow using immobilization/transport devices should be
1033 weighed against possible degradation of MR image quality. Positioning devices should be tested for
1034 MRI safety before incorporation into an MRI workflow.
1035 iv. Setup verification prior to HDR treatments
1036 Treatment setup verification is an integral part of the procedure workflow. To ascertain the
1037 treatment geometry conforms to the planned geometry, pre-treatment imaging is recommended for
1038 each treatment fraction (Figure 6). This will also require the establishment of action thresholds and
1039 remediation procedures in case a discrepancy is observed. To do so efficiently, it is recommended that
1040 fiducial markers implanted in the target volume be used as references to assess applicator or catheter
1041 displacement. Bony references and other anatomical surrogates (such as Foley balloons) are not
1042 recommended as they do not provide information on the position of the applicator with respect to
1043 the target. Markers that have been evaluated for CBCT, CT, kV, MV, US, and MRI include gold markers,
1044 gold coils, and polymer markers.177,178 The appropriate fiducial marker will be dependent on the
1045 imaging strategy employed, which in turn will be largely determined by the availability of imaging
1046 devices. Setup verification is recommended when a patient transfer is required as part of the
1047 treatment workflow, and in anticipation of unplanned events such as accidental patient movement or
1048 disturbance of the applicator.
1049 Ideally, the treatment setup would be verified through the use of 3D imaging. A volumetric scan
1050 would be used to confirm the position of the applicator with respect to the patient’s targeted
1051 anatomy, OARs, and fiducial markers. While MRI re-imaging prior to each fraction may not be

29
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 30 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1052 logistically practical, and CT scans provide inferior target visualization compared to MRI, CT does
1053 provide superior applicator, fiducial marker, and OAR visualization. The planning and treatment image
1054 set would be registered, e.g., based on the fiducial markers, and the distance between
1055 applicator/needle tip(s) and the markers would be used to assess the implant displacement.
1056 If a diagnostic quality CT is unavailable, kV CBCT may be used to assess applicator to fiducial
1057 marker consistency. However due to the low soft tissue contrast of CBCT, re-planning in case a
1058 displacement is observed may require re-simulating the patient. MV-based treatment verification
1059 adds to this challenge due to poor of visualization of titanium or plastic MRI compatible applicators
1060 and similar low-Z objects.
1061 In some clinics, the HDR treatment room may be equipped with 2D imaging only. While
1062 fluoroscopy/x-ray imaging has been shown to be less reliable in assessing applicator displacement
1063 than 3D imaging, some clinics utilized this imaging modality as an alternative when 3D imaging is
1064 unavailable (see Figure 6E).
1065 H. Logistical and economic considerations

1066 Before integrating MRI within the HDR BT workflow, discussions should be held with the
1067 institution’s finance department and the radiation oncology department administrator to seek the
1068 appropriate funding for initial and ongoing program maintenance. It is advantageous to hold these
1069 meetings early in the planning process to determine the required initial capital investment, and
1070 ongoing expected costs (e.g., FTE of MRI technologists, vendor related contracts). The department
1071 should be familiar with the appropriate charge codes when MRI is added to the clinical workflow, and
1072 develop a financial model with a conservative workload of patients to predict revenue streaming.
1073 Anderson et al.87 recently described the introduction of MRI logistics and financial hurdles they
1074 encounter. They concluded the integration of MRI into the HDR workflow is feasible with a sustainable
1075 financial structure if basic elements are formally introduced in collaboration with other departments
1076 (e.g., infection control, anesthesia, radiology) for GYN tandem and ovoid and later for interstitial HDR.
1077 A novel billing structure was created for their clinical environment, and was only possible after several
1078 iterations.
1079 The expected timeline to develop a budget/business plan may stretch over a month, however,
1080 equipment purchase, and commissioning will take considerably longer. Prior to developing a business
1081 model, the Authorized Medical Physicist and Authorized User physician with other relevant BT staff
1082 members (e.g., nursing, therapists, dosimetrists, administration) should make an effort to meet with
1083 infection control, the MR imaging nurse manager, anesthesia, and clinical engineering to develop a
1084 workflow and define the workspace. Most departments perform the applicator insertion outside the
1085 MRI vault, hence, following the TG-100179 methodology, a fishbone diagram should be developed with
1086 inputs from all stakeholders. This is highly recommended to ensure that all key elements required for
1087 patient care are considered. If the procedure is performed within a hospital environment, the
1088 anesthesia department usually has MRI compatible monitors and equipment for other patient
1089 populations that require anesthesia while undergoing MR imaging, otherwise these costs will need to
1090 be included in the operating and maintenance budget.

30
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 31 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1091 Workflow can also impact the cost structure. Kim et al.180 has reported on the experience of a
1092 single center performing MRI-based HDR for cervical cancer patients. They divided the procedure into
1093 four key steps based on where the procedure was performed: (1) applicator insertion under sedation,
1094 (2) MR imaging, (3) planning, and (4) treatment delivery. The reported mean total procedure time was
1095 149.3 min (SD 17.9, ranges 112–178), and the mean procedure times and range for each section
1096 (expressed in minutes) were (1) 56.2 (28.0–103.0), (2) 31.0 (19.0–70.0), (3) 44.3 (21.0–104.0), and (4)
1097 17.8 (9.0–34.0), respectively. In this setting, the authors stated that the combined mean procedure
1098 time for MR imaging and planning was 63.2 min, compared to 137.7 min during their initial
1099 implementation/learning phase in 2007–2008 (p < 0.001). Of note, the authors stressed the
1100 importance of seeking input and working with a skilled and multidisciplinary team to achieve an
1101 efficient clinical workflow.180

1102 V. Risk based analysis

1103 The inclusion of MRI guidance requires several changes to the current BT paradigm. Such changes
1104 include modifications in workflow, equipment, training, documentation, policies and procedures, and
1105 potentially personal/staffing. In considering the effects to patient safety and QA, one way to examine
1106 the risk associated with a new or modified process is through failure mode and effects analysis
1107 (FMEA). An FMEA comprises three key steps including process mapping, the collection of failure
1108 modes, and the scoring/ranking of those failure modes. The output of an FMEA is a list of ways in
1109 which a process can fail, prioritized based on risk. This list can be used to better distribute QA
1110 resources, create new checklists, and/or educate staff. At a deeper level, the output of an FMEA can
1111 be used along with other tools such as fault tree analysis to potentially create new safeguards or re-
1112 design a process to improve quality and safety.179,181,182
1113 In order to demonstrate this technique for an MRI-based BT procedure, an FMEA was performed
1114 for a representative use case, MRI-based, high dose rate BT of the prostate (MRI-HDRProstate). Sample
1115 FMEA’s for GYN BT are available for review in several publications.183,184 The process map for this
1116 example was based on the procedures as performed by one of the TG members at their local
1117 institution. The chosen workflow is best categorized as single fraction prostate HDR after the reversal
1118 of anesthesia. This workflow, as opposed to one where the entire procedure is performed
1119 intraoperatively, requires several patient transfers over the course of many hours. As such, there is
1120 an increased need for verification imaging, which in this workflow includes a CT scan after the patient
1121 recovers from anesthesia followed immediately by fluoroscopy once the patient has been transferred
1122 to the treatment room. This scenario is likely applicable to a broader audience as many institutions
1123 do not have MRI available within the department to facilitate intraoperative treatment.
1124 The process map for the procedure can be found in Figure 7. There are eight steps which represent
1125 the high level sub-processes. These include patient assessment, implantation, MRI simulation,
1126 treatment planning, plan review, pre-treatment actions/QA, treatment, and post-treatment
1127 actions/QA. Within these sub-processes, several subsequent actions have been outlined including
1128 items such as catheter insertion, image sequence selection, and contouring. The process map was

31
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 32 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1129 used to guide the collection of failure modes, which were submitted by six TG members with relevant
1130 experience. Collectively, the team identified 63 failure modes (see Table 4).
1131 The collected failure modes represent instances in which requirements to the MRI-HDRProstate
1132 process are not being met. In effect, they are the anti-functions to the proper actions needed for a
1133 successful outcome. In each case, the reason the failure occurred and the potential impact to the
1134 patient, staff, and the process itself were also identified. The three variables scored within an FMEA,
1135 occurrence (O), severity (S), and detectability (D), are related to these cause and effect mechanisms.
1136 Accordingly, occurrence is defined as the likelihood a failure mode occurs given a specific cause.
1137 Detectability is defined as the likelihood of not detecting a failure mode after it has happened but
1138 before the effects take place. And severity is defined simply as the magnitude of the effect considering
1139 aspects such as injury, cost, delays, etc. In each case, a higher score indicates that a failure mode is
1140 more likely to occur, less likely to be detected, and have more severe consequences. For this study,
1141 these three items were scored individually by the six team members using guidelines defined by AAPM
1142 TG 100.179 The median scores were then used to determine the risk priority number (RPN) of each
1143 failure mode, i.e., RPN = O x S x D.
1144 Table 4 highlights the top ranked failure modes according to RPN. The first item on the list relates
1145 to target contouring which is a commonly identified error pathway in radiation oncology risk
1146 assessment.183,185,186 The elevated RPN is driven by high scores for detection and severity indicating
1147 that the failure mode is not easily identified and capable of causing serious harm to the patient. The
1148 difficulty in detection is related to the fact that target contouring is a high level, knowledge-based
1149 action where the information and experience involved in the decision making process are not widely
1150 distributed. While in this setting the qualified medical physicist (QMP) is not optimally suited to detect
1151 subtle errors, familiarity with the planning process should allow for the discovery of obvious mistakes
1152 if the QMP is actively looking for them. Additionally, institutional controls such as peer review should
1153 be encouraged, and it is always prudent to verify that contours provided by a trainee are critically
1154 reviewed by an attending physician.
1155 Three prominent failure modes more specific to the current process are “needles
1156 crossed/doubled/swapped”, “calcification/artifact mistaken as needle tip”, and “needle tip/path not
1157 correctly defined.” These errors dealing with needle misidentification are closely related where the
1158 first two errors could additionally function as a cause for the latter. In any case, the end result is an
1159 implant model that does not accurately reflect the implant geometry. The causes associated with
1160 these failure modes are many including a sub-optimal implant, poor image quality, lack of training,
1161 and anatomical masking whereby a calcification is mistaken for a needle tip. Using FMEA as a guide,
1162 quality controls are primarily based in prevention or detection (a third option would be mitigation).
1163 Thus, in terms of preventing needle misidentification, quality measures addressing causal factors
1164 include proper commissioning of imaging protocols as discussed in Section IV.A, the critical review of
1165 MRI scans immediately after acquisition, and error training that familiarizes users with challenging
1166 situations where misidentification is likely to occur. In terms of detection, an independent review by
1167 a QMP should be performed, preferably within the planning system itself, as opposed to a review that
1168 relies on documentation alone.

32
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 33 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1169 Another top failure mode specific to the chosen workflow is the movement of needles during
1170 transport. As detailed in Figure 7, the patient is transferred from surgery to MRI, from the MRI to
1171 holding, from holding to CT, and from CT to treatment. During this route, any deviation in needle
1172 position will introduce changes not reflected in the MRI-based treatment plan. In terms of prevention,
1173 immobilization is important as well as the use of a suitable transfer mechanism such as a hover device,
1174 removable couch top, or portable mattress that can be transferred with the patient. In terms of
1175 detection, imaging for quality control purposes is built into the workflow at two specific time points,
1176 at CT that takes place after recovery from anesthesia and at planar x-ray acquisition that takes place
1177 immediately prior to treatment within the HDR suite. It is important not only to image the patient but
1178 also to develop protocols for registration and review65. As noted previously, error training can help
1179 familiarize users as to how images appear when needle movement has occurred.
1180 Data import and selection represent another type of action associated with several highly ranked
1181 failure modes. Here, the error often takes the form of slips, lapses, or misinterpretation of
1182 information. Since import and selection typically occur at the beginning of a process, it can be difficult
1183 to detect an error if the system does not make the origin/nature of the data apparent at subsequent
1184 time points. In these situations, it is prudent to document whenever important information enters
1185 the system or is selected as a key item. An example of this practice is the labeling of the planning MRI
1186 with relevant identifiers such as scan type, date, and/or intended use at the time of import. It can also
1187 be helpful to screen capture the import window showing exactly which scan is being brought into the
1188 system. The image can then be added to the plan documentation as a way to make this key item
1189 visible to a reviewer.
1190 To summarize, this FMEA is put forth as an example of how to conduct a prospective risk
1191 assessment for an MRI-based BT procedure. It is important to keep in mind that there are many ways
1192 to perform both FMEA and MRI-based BT. Failure modes which may be relevant in the current
1193 workflow may not be applicable to every clinic and vice versa. It is therefore important for each clinic
1194 to map their own processes and assess risk within their own clinical setting. Users are encouraged to
1195 periodically review previously reported medical events released by the Nuclear Regulatory
1196 Commission,187 and summarized in publications,188,189 as part of continuous quality improvement of
1197 their BT program. As demonstrated in Table 4, a significant source of these events are due to human
1198 errors, especially in situations of time pressures.

1199 VI. Recommendations to the medical physicists

1200 A. Clinical commissioning tasks
1201 During the initial discussions of MRI integration into an HDR clinical workflow, medical physicists
1202 should assume a leading role and assist in the formation of a multidisciplinary team. Additionally,
1203 given the expertise in Radiology departments with MRI, either contacting and working with experts in
1204 Radiology and/or including them in the multidisciplinary team to discuss possible workflow options
1205 and safety processes, and develop scanning protocols may prove to be helpful. The formed team will
1206 determine the proper resources for the safe implementation of MRI, as well as the appropriate MRI
1207 workflow for this implementation (e.g., MRI-informed, MRI-based, or MRI-guided) based on access to

33
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 34 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1208 the MRI, resources, and staffing. Perhaps the easiest form of integration is an MRI-informed approach
1209 using diagnostic or pre-implant MR images to guide applicator needle placement and/or target
1210 segmentation. This may be accomplished by reviewing the MR images immediately preceding or
1211 during implant, or by reviewing the MRI dataset with the planning CT. However, at this time, the Task
1212 Group does not recommend fusing the MRI dataset in the absence of applicators and/or needles with
1213 the planning CT or US (i.e., in the case of prostate BT) due to the potential for large uncertainties
1214 between imaging datasets with and without applicators. Although an MRI only approach is preferred
1215 as discussed in Section IV.A, if it is not feasible, the next approach to MRI integration, in terms of
1216 increasing level of complexity, is the MRI-based approach. Ideally, this approach would involve an MRI
1217 acquisition for each treatment fraction. However, in some cases the MRI dataset is only acquired for
1218 one treatment fraction, and registered to all subsequent CTs based on the same applicator(s) to help
1219 delineate the target.80 In the United States, this approach is feasible in most clinics and is
1220 recommended if MRI-only workflows are not available. However, it should be noted that this
1221 approach of a single MRI for patients treated with GYN malignancies may not be well-suited to clinics
1222 that deliver BT early on in conjunction with external beam therapy, due to tumor shrinkage early in
1223 the course of irradiation, thus requiring MRI for multiple fractions to capture the changing target
1224 volume over the course of BT.100 The ideal and most complex MRI approach is MRI-guided BT. In this
1225 workflow, the applicator and/or needles are implanted using real-time MRI guidance. In the United
1226 States, only a few centers have currently adopted this approach due to its time and resource demands.
1227 As such, the Task Group does not address specific recommendation for this workflow.
1228 Commissioning and Workflow Recommendations:
1229
1230 • For MRI-based and MRI-guided approaches, geometric and dosimetric uncertainties should
1231 be evaluated during clinical commissioning. The geometric accuracy should be maximized in
1232 regions of greatest dosimetric interest. High resolution CT images should be used as the gold
1233 standard for comparison.
1234 • Image degradation due to patient immobilization or transfer devices should be evaluated
1235 both in phantom and in vivo.
1236 • Acquisition time should be optimized to reduce the risk of motion blurring (ideally, the
1237 acquisition time for each MRI sequence should be < 5 minutes).
1238 • Intracavitary (i.e., rectal or vaginal) coils are not recommended for BT imaging due to organ
1239 deformations introduced by the coil during MR imaging, which would lead to uncertainties
1240 between the planned and delivered dose.
1241 • Surface array coils positioned around the pelvis are recommended for MR imaging for GYN
1242 and prostate BT.
1243 • A rectal enema or antispasmodic agents (e.g., glucagon, buscopan) may be used prior to MR
1244 imaging to reduce peristalsis and motion induced artifacts.
1245 • Gauze or balloon-based packing soaked or filled with MRI compatible contrast (e.g.,
1246 gadolinium, US gel, saline, sterile water) is recommended to increase the separation between

34
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 35 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1247 the BT applicator and OARs, as well as enhance the delineation between structures (e.g., the
1248 vagina and cervix).
1249 • Bladder filling should be consistent between planning simulation and treatment, and should
1250 be well tolerated by the patient.
1251 • If images show significant motion artifacts, the sequence(s) should be repeated.
1252 • Vendor-provided gradient distortion correction and optimization of bandwidth should be
1253 utilized to ensure geometric fidelity of the MRI dataset.
1254 • Work with an MRI physicist and/or vendor to optimize pulse sequences; 3D sequences with
1255 isotropic resolution of ≤ 2 mm are recommended for target delineation. In the absence of
1256 such 3D isotropic sequences, 2D T2W FSE acquired at ≤ 5 mm slice thicknesses and with <
1257 1mm in-plane resolution are recommended if acquired using multiple scan orientations (e.g.,
1258 axial, sagittal, coronal).
1259 • Acquisition of both T2W-MRI and fMRI (i.e., DWI, DCE) for HDR prostate BT is recommended
1260 to improve the identification of intraprostatic lesions.
1261 • MRI-based applicator reconstruction commissioning and validation should follow the
1262 recommendations of Hellebust et al.13 and ICRU 89.14,112 Applicator reconstruction
1263 uncertainties should be ≤ 2 mm. If the uncertainties exceed 2 mm, efforts should be made to
1264 improve quality (e.g., reduce slice thickness, optimize MRI sequences to minimize artifacts
1265 and distortions, use MRI compatible markers). When available, and following commissioning,
1266 use of 3D applicator models is recommended.
1267 • If available, verification imaging is recommended when a patient transfer is required as part
1268 of the treatment workflow to account for potential patient movement or disturbance of the
1269 applicator(s). The dosimetric significance of this potential motion remains controversial, and
1270 is dependent on the type of implant (e.g., intracavitary versus interstitial), sedation, and the
1271 immobilization equipment utilized.190-193 For instance, motion can be minimized with tighter
1272 packing although this may require the administration of anesthesia.
1273 • Work with an MRI physicist and/or vendor to verify SAR not exceeded for the different clinical
1274 sequences that have been developed.
1275 • Work with an MRI physicist and /or vendor to develop a QA program for the MRI scanner
1276 (e.g., image quality, image geometry, image transfer integrity, orientation).
1277 B. MRI safety essential considerations
1278 Care should be taken when considering a transition to a higher field strength MRI scanner (e.g.,
1279 1.5 to 3 T) for BT patients. It is imperative that commissioning tests be performed with the relevant
1280 BT equipment and accessories, in phantom, to ensure patients may be safely imaged with the existing
1281 equipment at a higher magnetic strength. Commissioning will require a review and re-optimization of
1282 the imaging sequences since the level of susceptibility artifacts can become substantial for certain
1283 sequences, such as diffusion weighted imaging.194 Also, the presence of metal (e.g., titanium
1284 applicators) can enhance magnetic susceptibility artifacts and heating at higher magnetic field
1285 strengths; therefore, the use of plastic applicators is preferred. Lastly, the user should investigate

35
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 36 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1286 whether their applicators/needles are labeled MRI safe or MRI conditional at 3.0 T. Not all devices
1287 that have been tested at 1.5 T have been tested at higher magnetic fields,195 and a device deemed
1288 MRI conditional under one environment (e.g., field strength) may not be safe to scan in another.165 To
1289 expedite this process, it may prove helpful to consult with colleagues in Radiology.
1290 MRI Safety Recommendations:
1291 • Establish a checklist for screening patients and equipment prior to each MRI scan.
1292 • Transition from 1.5 T to 3.0 T should not be done without repeating safety and image
1293 optimization tests as described in Section IV.
1294 • Ensure checks are in place to ensure MRI safe or conditional applicators are used for BT
1295 implants, and that applicators are used in accordance with the vendor instructions for use.
1296 • Members of the care team requiring access to the MRI suite should receive MRI safety training
1297 and annual refreshers to ensure they can safely practice near and/or in the MRI vault.
1298 • A barrier (e.g., pads, towels, linens) should be used to prevent the patient from coming into
1299 direct contact with metallic objects, even outside of the bore, and the bore wall during MRI
1300 scans to reduce the risk that the patient may experience thermal induced effects.
1301 • Care should be taken to prevent skin-to-skin contact of limbs during MRI scans to prevent
1302 thermal injuries.
1303 C. Quality Assurance
1304 Traditionally, recommendations for QA have been issued in publications as a series of
1305 prescriptive guidelines, with suggested frequencies and tolerances. However, with the increasing
1306 complexity of radiotherapy planning and delivery techniques, alternative approaches to QA have
1307 been considered.179 As BT physicists new to MRI will seek guidance to its integration to HDR BT, the
1308 Task Group has provided recommendations for traditional QA, as well as the AAPM TG-100 Report
1309 approach.
1310 i. Traditional HDR BT Program QA

1311 The Task Group recommends that clinics continue to perform their standard imaging and HDR
1312 BT QA in accordance with federal/state regulations and professional society recommendations. In
1313 addition, several additional QA tests should be performed to ensure the safe integration of MRI into
1314 the BT workflow. Table 5 provides a list of suggested QA procedures and frequencies in which these
1315 procedures should be performed.

1316 ii. QA considerations for MRI integration with HDR BT

1317 The Task Group recommends that each clinic follow the AAPM TG-100 Report methodology for
1318 process mapping, failure mode collection, scoring, and analysis as demonstrated in Section V for an
1319 MRI-HDRprostate workflow. Safety barriers should be used to address high priority failure modes and
1320 include mechanisms for both prevention and detection. A number of barriers were identified during
1321 the current FMEA analysis in Section V. Several of these apply universally to other workflows involving
1322 HDR BT and/or MRI.

36
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 37 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1323 Recommendations on Safety Barriers:

1324 • Radiation oncologist contouring remains a critical and often overlooked aspect of treatment
1325 planning. The QMP should actively look for contouring errors and encourage robust peer
1326 review and trainee supervision.
1327 • Needle/catheter digitization is a crucial step in any HDR BT workflow. The following safety
1328 barriers are recommended:
1329 - If MRI markers are used for plastic applicators, the marker should be inspected prior to
1330 use to verify the integrity of the marker, and the absence of air bubbles in critical locations
1331 of the marker (e.g., marker tip).
1332 - Critical review of planning images by physician and physicist prior to removing patient
1333 from the MRI table. The QMP involved in the procedure should be familiar with the types
1334 of applicators used and how these applicators present in both normal and abnormal
1335 scenarios.
1336 - Secondary review of applicator digitization should be performed by an independent QMP
1337 or a certified medical dosimetrist prior to resuming planning.
1338 • Whenever implantation and delivery are performed in separate locations, applicator
1339 movement during transport is a major concern. Robust immobilization is recommended as
1340 well as the use of a dedicated transfer device such as a portable mattress, hover unit, or
1341 removable couch top. A pre-treatment imaging protocol should be implemented to define
1342 image acquisition sequences, registration, and validation of applicator(s) positioning.
1343 • All data imported and selected for use during treatment planning should be made visible to
1344 an independent reviewer through descriptive labeling, the use of screen captures, or other
1345 suitable mechanisms.
1346 • While not specifically highlighted among the top failure modes, other routine safety barriers
1347 specific to current FMEA example apply such as verification of the:
1348 - Treatment length;
1349 - Plan transfer from treatment planning system to treatment control station;
1350 - Source model (i.e., if multiple models are commissioned in the TPS);
1351 - Source strength;
1352 - Afterloader (i.e., if multiple afterloaders are available in a clinic);
1353 - Template alignment; and
1354 - Plan normalization and prescription.
1355 Additionally, an independent secondary dose calculation should be performed 60,61.
1356 D. Logistical and economic considerations

1357 Before integrating MRI into a BT practice, it is important for the BT team to meet with all
1358 relevant stakeholders both within and outside of the department.

1359 Recommendations on Logistics and Economics

37
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 38 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1360 • Brachytherapy team should meet with parties outside of the department (e.g., radiology,
1361 anesthesia, infection control, clinical engineering) to develop a workflow and identify
1362 resource needs and workspaces.

1363 • Discussions on finances should be initiated with department administrators early in the
1364 planning phase to ensure appropriate funds.

1365 • Team should review and be familiar with the relevant charge codes that should be captured
1366 with the use of MRI within the BT workflow.

1367 • Charges should be periodically reviewed and updates made, when necessary (e.g.,
1368 introduction of new bundle payments).

1369 VII. Potential future development in MRI-guided BT

1370 Interests in MRI-integrated BT suites, equipped with both RF and radiation shielding for
1371 simultaneous MR imaging and treatment delivery, are increasing as the field recognizes the value and
1372 convenience of having an MRI available at the time of BT implants.196 Figure 8 illustrates an example
1373 setup where a typical afterloader is placed just outside the 5 Gauss line to ensure proper functioning
1374 of electronics, especially what drives the 192Ir source. In addition, towards true integration, Beld et
1375 al.197 recently proposed a prototype of an MRI-conditional HDR afterloader design that properly
1376 shields the electronics from RF, and contains a plastic source cable. If this line of research is successful,
1377 one day, it may be possible to observe real-time MR imaging while the HDR treatment is delivered,
1378 much like an MRI-linac system.198 This may also allow real-time visualization of source positions with
1379 respect to patient anatomy.

1380 VIII. Summary

1381 The Task Group presented a series of recommendations to ensure the successful implementation
1382 of MRI for HDR BT. Recommendations have been provided on potential workflows, simulation
1383 imaging, treatment planning, and pre-treatment verification for those centers that are integrating
1384 MRI into their HDR program, and to support centers with existing programs. MRI safety considerations
1385 were discussed, and a risk based analysis was provided for an MRI-guided prostate HDR workflow.
1386 Although these recommendations have been limited to gynecologic and prostate BT, many may be
1387 applicable to other disease sites.
1388
1389 Disclosure Statement
1390 The members of AAPM Task Group # 303 listed below disclose the following potential Conflict(s) of
1391 Interest related to subject matter or materials presented in this document.
1392 K-P. Hwang receives research funding from General Electric Healthcare.
1393 J. Prisciandaro has received funding for and is currently involved in a funded non-clinical evaluation
1394 agreement with Varian Medical Systems, Inc. She was also involved in an unfunded, non-clinical
1395 evaluation agreement with C4 Imaging LLC.

38
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 39 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1396 Y. Kim receives funding and equipment support through an industrial research collaboration with Varian
1397 Medical Systems, Inc.
1398 W. Song receives research funding from Varian Medical Systems, Inc.
1399 The Department of Radiation Oncology at the Medical College of Wisconsin receives research funding
1400 from Elekta.
1401 The Department of Radiation Oncology at the Medical University of Vienna receives financial and/or
1402 equipment support for research and educational purposes from Elekta (Nucletron B.V.) and Varian
1403 Medical Systems, Inc. Dr. Kirisits has received support for educational events from Elekta (Nucletron B.V.).
1404 The Department of Radiation Oncology at Memorial Sloan-Kettering Cancer Center receives research
1405 funding for investigator initiated clinical trials from Elekta.
1406
1407 Acknowledgements:
1408 The task group members would like to thank Elena Nioutsikou (Siemens Healthineers), Cristina
1409 Cozzini (GE), and Mo Kadbi (Philips) for their contributions to our many phone and email discussions.

39
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 40 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1410 FIGURES:
1411

A B

1412 Figure 1. Sagittal view of a GYN patient with a titanium tandem and ring applicator set, including a rectal
1413 retractor on (A) CT (Philips Brilliance large bore CT) and (B) MRI (sagittal 2D T2W FSE sequence acquired
1414 on a 3T Siemens Skyra MRI scanner). The high risk CTV is depicted in the red contour.
1415
1416
1417
1418
1419
1420

40
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 41 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1421
1422
CT-MR 50% Blend
1423 A B C
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
D E
1434 Figure 2. Mid-gland axial and sagittal views of a patient with a prostate HDR implant on (A) & (D) MRI
1435 and (C) & (E) intraoperative CBCT, respectively. The registration of the two datasets, with a 50% MRI-CT
1436 blending in the axial view is depicted in (B). The MRI dataset (axial and sagittal T2 sequences obtained
1437 with the patient under general anesthesia, using a 3T Philips Ingenia MRI scanner) is used for target
1438 delineation, while the CT is typically used for applicator reconstruction.
1439

41
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 42 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

B
A C

3D T1W 2D T2W
CT
D E F

CT 3D T1W 2D T2W
1440 Figure 3. Sagittal view of a (A-C) titanium and (D-F) plastic ring and tandem applicator set in an in-house
1441 phantom designed for applicator reconstruction commissioning on (A&D) CT (Philips Brilliance large
1442 bore CT), (B&E) 3D T1W images (3T Siemens Skyra MRI), and (C&F) 2D T2W images (3T Siemens Skyra
1443 MRI).
1444
1445

42
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 43 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

A
B

1446 Figure 4. Sagittal MRI images on A) CT (Philips Brilliance large bore CT) and B) 3D T1W MPRAGE MRI (3T
1447 Siemens Skyra MRI) of a plastic ring and tandem applicator in vivo. To visualize the lumen of the ring and
1448 tandem applicators, an x-ray marker wire and an in-house fabricated marker filled with gadolinium-
1449 doped water was inserted in the applicators during CT and MRI acquisition, respectively. The absence of
1450 a hyperintense signal at the tip of the tandem on MRI versus CT, is an indication of the presence of air-
1451 bubbles in the MRI marker. (Please note, given the contrast agent used for the in-house marker, it was
1452 not visible on T2W images. If a marker with a T2W contrast is used, a similar void may be expected in
1453 the presence of an air-bubble.)
1454

43
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 44 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

A B C

1455 Figure 5. Sagittal images of a plastic ring and tandem applicator set in an in-house phantom using (A) CT
1456 (Philips Brilliance large bore CT), (B) 3D T1W, and (C) 2D T2W images (3T Siemens Skyra MRI) during the
1457 evaluation of a commercially available MRI marker (C4 Imaging LLC, Doylestown, PA, USA). The presence
1458 of the marker in the MR images allows the lumen of the ring and tandem applicators to appear
1459 hyperintense on the MR imaging sequences.
1460

44
TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 45 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

45
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 46 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1461
1462 Figure 6: Example images acquired during implant and pre-treatment for a multi-fraction, single implant
1463 workflow as described in the risk based analysis in Section V. An intraoperative planning (A & B) T2W
1464 MRI [(A) axial and (B) sagittal T2W MRI] was registered to (D) a pre-treatment, post recovery CT using
1465 implanted fiducial markers. The circles denote one of the fiducial markers and arrows point to needles.
1466 The needle tip positions were then confirmed to coincide between CT and MRI, as shown in (C) a 50:50
1467 blended view. Artifacts from the marker reconstruction on the sagittal image is seen on CT only. After
1468 the patient is transferred to the treatment vault, (E) a planar x-ray image is taken to verify the distance
1469 from the selected needle tips to the respective fiducial markers has not changed.
1470
1471
1472

46
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 47 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1473

1474
1475 Figure 7: Process map for an example MRI-based, prostate HDR BT detailed in Section V.
1476

47
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 48 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1477
1478 Figure 8. An MRI-HDR BT-integrated suite at the University of Medical Center Utrecht, Utrecht, the
1479 Netherlands. Note that MRI-compatible anesthesia cart is available in-room enabling a wide range of
1480 treatment procedures. Picture courtesy of Dr Rien Moerland.196
1481

48
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 49 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1482 TABLES:
1483 Table 1: MRI vendor-specific 3D fast spin echo sequence acronyms and descriptions
Vendor Acronym Definition

Siemens SPACE Sampling Perfection with Application optimized Contrasts using different flip
angle Evolution

GE CUBE (Not an acronym)

Philips VISTA Volume Isotropic Turbo spin echo Acquisition

Toshiba 3D MVOX 3D MultiVOXel

Hitachi isoFSE isotropic Fast Spin Echo

1484
1485

49
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 50 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1486 Table 2: Comparison between multi-slice 2D FSE and 3D FSE sequences for MRI-based BT.
Advantages Disadvantages

2D FSE • True T2 contrast • Larger slice thickness

• Oriented to applicator • Multiple acquisitions typically required,
geometry increasing scan time
• High in-plane resolution • May require reformatting onto Cartesian grid
prior to import into BT planning systems,
resulting in decreased image quality
• May require separate sequence for applicator
reconstruction
• Full 3D gradient nonlinearity (GNL) correction
may not be supported for 2D FSE sequences

3D FSE • Isotropic spatial resolution • T2 contrast often different from 2D FSE1

(1mm3) • Motion sensitivity due to longer scan times
• Permits easy reformatting
into BT eye views
• Permits easy applicator
reconstruction

1487 1
The user should be advised that T2 contrast on 3D FSE images may differ from that on 2D FSE images
1488 obtained diagnostically for detection/staging or obtained during MRI simulation for external beam
1489 target delineation. Though the 3D FSE images are optimal for applicator reconstruction, the altered 3D
1490 FSE contrast may challenge interpretation of treatment related changes resulting from external beam
1491 when used alone and may require re-learning for contouring. Alternatively, a mixed-mode approach
1492 could be utilized in which the 3D FSE images serve as the reference and additional multi-planar 2D FSE
1493 images oriented to the applicator are acquired and used for delineation or verification of target
1494 contours. This approach maximizes the advantages of both 2D and 3D FSE sequences at the cost of
1495 longer scan times.

50
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 51 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1496 Table 3: An example of generalized 2D/3D FSE scan parameters for GYN and prostate BT from two
1497 institutions.
1498 GYN BT (Institution 1)
Slice TE (ms) TR (ms) Voxel Size ETL Readout BW Scan Time
Prescription1 (mm) (Hz/pix)5 (min)
2D FSE PA, PS, PC
2
85 2500 0.9x0.9x3.0 16 440/880 3
3D FSE Ax 85 3
2500 1.0x1.0x1.5 300 4
440/880 126
1499
1500 GYN BT (Institution 2, based on a 3.0 T Philips Ingenia scanner)
Slice TE (ms) TR (ms) Voxel Size ETL Readout BW Scan Time
Prescription 1
(mm) (Hz/pix) (min)
2D FSE Ax, Sag
2
100 4471 0.45x0.45x3.0 30 244.1 5:13
1501
1502 Prostate BT (Institution 2, based on a 3.0 T Philips Ingenia scanner)
Slice TE (ms) TR (ms) Voxel Size ETL Readout BW Scan Time
Prescription 1
(mm) (Hz/pix) (min)
2D FSE Ax, Sag
2
100 5194 0.6x0.6x2.0 29 244.1 3
3D FSE Ax 245 1800 0.65x0.65x2.0 79 455.3 5:40
1503
1504 1
Ax = Axial, Sag = Sagittal, and PA = Para-Axial, PS = Para-Sagittal, PC = Para-Coronal to Applicator for
1505 GYN
1506 2
Full 3D gradient non-linearity (GNL) correction may not be supported for 2D sequences.
1507 3
Effective TE reported for 3D FSE
1508 4
Echo train duration reported for 3D FSE
1509 5
Readout bandwidth reported for 1.5T/3.0T; Additional optimization to recover SNR may be required.
1510 6
Longer scan times may benefit from administration of antispasmodic agents to reduce motion.

51
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 52 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1511 Table 4. Failure modes identified by task group members for an example MRI-based, high dose rate BT of the prostate detailed in Section V.

1512
52
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 53 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1513
1514

53
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 54 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1515
1516 Table 5. Recommended QA procedures, tolerances (where applicable), and frequency with which to
1517 perform these tasks when MRI is integrated into the HDR clinical workflow. Please note, in accordance
1518 with TG 100, users should develop a process map based on their individual workflow and perform a risk
1519 based analysis to determine if additional checks should be performed. (TPS – Treatment Planning
1520 System)
Frequency Procedure Tolerances (if applicable)
Initially Review instructions for use for proposed N/A
HDR applicators/needles (e.g., MRI safety,
sterilization process and maximum number
of cycles)
Review imaging, HDR, and third party N/A
vendor websites to verify whether
customer bulletins have been released
regarding applicators, software, or
hardware
Verify MRI safety of ancillary equipment N/A
(e.g., used for patient monitoring,
anesthesia, medication delivery, patient
transport/transfer, applicator
immobilization devices), and write
standard operating procedures for their
use
Optimize MRI pulse sequences in phantom N/A
and in vivo, and evaluate image artifacts
and distortion and save sequence to MRI
console, limiting write access
Ensure BT staff receive MRI safety and N/A
appropriate procedure specific training199

Verify a pre-MRI screening questionnaire

has been developed200
Commission relevant applicator library N/A
models14 in the treatment planning system,
if applicable
Perform applicator reconstruction ≤ 2 mm14,66,67 and < 8 –
commissioning and evaluate geometric
13
10%,101,102 respectively
and dosimetric accuracy for relevant HDR
applicators compared with gold standard
imaging (e.g., CT)

54
TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 55 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

Evaluate MRI markers if planned for clinical ≤ 2 mm

use with plastic applicators
Perform end-to-end testing to evaluate N/A
workflow
Develop policies and procedures for N/A
appropriate use of applicators in an MRI
environment, imaging sequences, and QA
prior to clinical implementation of MRI
within the BT workflow
Document clinical workflow and update N/A
checklists based on changes introduced
with MRI integration
Each treatment day Perform HDR periodic QA per guidelines See references 60,61,201
(e.g., TG40,61 TG5960 and 10 CFR 35.643201)
Perform imaging QA per guidelines (e.g., See references 63,202-204
CT,202 MRI,63 kV/MV images,203 kV or MV-
CBCT203,204)
Perform pre-implant check to ensure the N/A
appropriate applicator/needles was
selected and that applicator/needles is/are
MR safe or conditional
Confirm pre-MRI screening questionnaire N/A
completed before start of procedure
Perform physical screening before patient
enters zone 4
Prior to MR imaging, prevent skin-to-skin N/A
contact of the hands, feet, and/or limb, and
bore contact
If a non-ferromatic metal N/A
applicator(s)/needles implanted, verify
applicator does not come in contact with
patients skin and applicator tips do not
cross
If applicable, verify the integrity of the MRI N/A
markers prior to MR imaging
Monthly Perform imaging QA per guidelines (e.g., See references 63,202-204
CT,202 MRI,63 kV/MV images,203 kV or MV-
CBCT203,204)

55
TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 56 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

Each source exchange or Perform HDR source exchange QA per See references 60,61,205
quarterly guidelines (e.g., TG40,61 TG5960 and 10 CFR
35.633205)
Review imaging and HDR vendor websites N/A
to verify whether new customer bulletins
have been released regarding applicators,
software, or hardware
Annual Perform HDR annual QA per professional See references 60,61
society recommendations (e.g., TG40,61
TG5960)
Verify staff has completed annual MRI and N/A
radiation safety refreshers, as well as HDR
emergency training
Review policies and procedures, and if N/A
necessary updated
Review clinical workflows and checklists, N/A
update as needed
Software/hardware Review software/hardware releases to N/A
upgrades determine which software/hardware
features should be tested post-upgrade
With updates to MRI software/hardware, ≤ 2 mm
verify clinical pulse sequences in phantom
and evaluate image artifacts and distortion
to determine if adjustments are needed in
imaging parameters
Verify whether upgrade has affected N/A
relevant applicator models,14 if so,
applicator models may need to be
recommissioned
Perform end-to-end testing to evaluate N/A
workflow
Purchase of new model Review instructions for use for relevant N/A
of applicator/needles HDR applicators; if third party applicator or
accessories, ensure compatibility with
existing devices
Verify MRI safety of relevant HDR N/A
applicators

56
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 57 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

Optimize MRI pulse sequences in phantom ≤ 2 mm

and in vivo, and evaluate image artifacts
and distortion
Commission applicator models,14 if N/A
applicable
Perform applicator reconstruction ≤ 2 mm14,66,67
commissioning13 and evaluate accuracy for
relevant HDR applicators
Evaluate MRI markers if planned for clinical ≤ 2 mm
use with plastic applicators
1521
1522

57
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 58 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1523 IX. References

1524 1. Viswanathan AN, Beriwal S, De Los Santos J, et al. The American Brachytherapy
1525 Society Treatment Recommendations for Locally Advanced Carcinoma of the Cervix
1526 Part II: High Dose-Rate Brachytherapy. Brachytherapy. Jan-Feb 2012;11(1):47-52.
1527 2. Pugh TJ, Pokharel SS. Magnetic resonance imaging in prostate brachytherapy:
1528 Evidence, clinical end points to data, and direction forward. Brachytherapy. 2017/07/01/
1529 2017;16(4):659-664.
1530 3. Dirix P, Haustermans K, Vandecaveye V. The Value of Magnetic Resonance Imaging for
1531 Radiotherapy Planning. Seminars in Radiation Oncology. 2014/07/01/ 2014;24(3):151-
1532 159.
1533 4. Tanderup K, Viswanathan AN, Kirisits C, et al. Magnetic resonance image guided
1534 brachytherapy. Semin Radiat Oncol. Jul 2014;24(3):181-191.
1535 5. Gay S, Chen N, Burch J, et al. Multiplanar Reconstruction in Magnetic Resonance
1536 Evaluation of the Knee: Comparison with Film Magnetic Resonance Interpretation.
1537 Investigative Radiology. 1993;28(2):142-145.
1538 6. Bruno F, Arrigoni F, Mariani S, et al. Advanced magnetic resonance imaging (MRI) of
1539 soft tissue tumors: techniques and applications. La radiologia medica. April 01
1540 2019;124(4):243-252.
1541 7. Glide-Hurst C, E. Paulson, McGee, K., Tyagi, N., Hu, Y., Balter, J., and Bayouth, J.
1542 Magnetic Resonance Imaging - Simulation in Radiotherapy: Considerations for clinical
1543 implementation, optimization, and quality assurance: Report of AAPM Task Group No.
1544 284. In progress.
1545 8. Venkatesan AM, Stafford RJ, Duran C, et al. Prostate magnetic resonance imaging for
1546 brachytherapists: Anatomy and technique. Brachytherapy. 2017;16(4):679-687.
1547 9. Blanchard P, Pugh TJ, Mahmood U, et al. MRI Simulation for LDR prostate
1548 brachytherapy: Can we replace ultrasound with MRI for treatment planning? Comparison
1549 of pre-planning, day 0, and day 30 MR dosimetry Brachytherapy. 2016;15(3):S57.
1550 10. Potter R, Haie-Meder C, Van Limbergen E, et al. Recommendations from gynaecological
1551 (GYN) GEC ESTRO working group (II): concepts and terms in 3D image-based
1552 treatment planning in cervix cancer brachytherapy-3D dose volume parameters and
1553 aspects of 3D image-based anatomy, radiation physics, radiobiology. Radiother Oncol.
1554 Jan 2006;78(1):67-77.
1555 11. Dimopoulos JC, Petrow P, Tanderup K, et al. Recommendations from Gynaecological
1556 (GYN) GEC-ESTRO Working Group (IV): Basic principles and parameters for MR
1557 imaging within the frame of image based adaptive cervix cancer brachytherapy.
1558 Radiother Oncol. Apr 2012;103(1):113-122.
1559 12. Haie-Meder C, Potter R, Van Limbergen E, et al. Recommendations from
1560 Gynaecological (GYN) GEC-ESTRO Working Group (I): concepts and terms in 3D
1561 image based 3D treatment planning in cervix cancer brachytherapy with emphasis on
1562 MRI assessment of GTV and CTV. Radiother Oncol. Mar 2005;74(3):235-245.

58
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 59 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1563 13. Hellebust TP, Kirisits C, Berger D, et al. Recommendations from Gynaecological (GYN)
1564 GEC-ESTRO Working Group: considerations and pitfalls in commissioning and
1565 applicator reconstruction in 3D image-based treatment planning of cervix cancer
1566 brachytherapy. Radiother Oncol. Aug 2010;96(2):153-160.
1567 14. ICRU Report No. 89. Prescribing, Recording, and Reporting Brachytherapy for Cancer of
1568 the Cervix. Journal of the ICRU. 2013;13(1).
1569 15. Curran WJ, DiSaia PJ, Wolmark N. NRG Oncology Research Opportunities within the
1570 New National Clinical Trials Network. Seminars in oncology. 09/06 2014;41(5):553-555.
1571 16. Grover S, Harkenrider MM, Cho LP, et al. Image Guided Cervical Brachytherapy: 2014
1572 Survey of the American Brachytherapy Society. Int J Radiat Oncol Biol Phys. Mar 01
1573 2016;94(3):598-604.
1574 17. Wang J, Tanderup K, Cunha A, et al. Magnetic resonance imaging basics for the
1575 prostate brachytherapist. Brachytherapy. Jul - Aug 2017;16(4):715-727.
1576 18. Blanchard P, Menard C, Frank SJ. Clinical use of magnetic resonance imaging across
1577 the prostate brachytherapy workflow. Brachytherapy. 2017;16:734 - 742.
1578 19. Buus S, Rylander S, Hokland S, et al. Learning curve of MRI-based planning for high-
1579 dose-rate brachytherapy for prostate cancer. Brachytherapy. Jul-Aug 2016;15(4):426-
1580 434.
1581 20. Hoskin PJ, Rojas AM, Ostler PJ, et al. Dosimetric predictors of biochemical control of
1582 prostate cancer in patients randomised to external beam radiotherapy with a boost of
1583 high dose rate brachytherapy. Radiother Oncol. Jan 2013;110(1):110-113.
1584 21. Damato AL, Viswanathan AN. Magnetic Resonance-Guided Gynecologic Brachytherapy.
1585 Magn Reson Imaging Clin N Am. Nov 2015;23(4):633-642.
1586 22. Viswanathan AN, Cormack R, Holloway CL, et al. Magnetic resonance–guided interstitial
1587 therapy for vaginal recurrence of endometrial cancer. International Journal of Radiation
1588 Oncology*Biology*Physics. 2006/09/01/ 2006;66(1):91-99.
1589 23. Viswanathan AN, Szymonifka J, Tempany-Afdhal CM, et al. A prospective trial of real-
1590 time magnetic resonance–guided catheter placement in interstitial gynecologic
1591 brachytherapy. Brachytherapy. 2013/05/01/ 2013;12(3):240-247.
1592 24. Enders J, Rief M, Zimmermann E, et al. High-Field Open versus Short-Bore Magnetic
1593 Resonance Imaging of the Spine: A Randomized Controlled Comparison of Image
1594 Quality. PLoS ONE. 2013;8(12):e83427.
1595 25. Beriwal S, Kannan N, Kim H, et al. Three-dimensional High Dose Rate Intracavitary
1596 Image-guided Brachytherapy for the Treatment of Cervical Cancer Using a Hybrid
1597 Magnetic Resonance Imaging/Computed Tomography Approach: Feasibility and Early
1598 Results. Clinical Oncology. 2011/12/01/ 2011;23(10):685-690.
1599 26. Wood R, Bassett, K., Foerster, V., Spry, C., and Tong, L., 1.5 Tesla Magnetic
1600 Resonance Imaging Scanners Compared with 3.0 Tesla Magnetic Resonance Imaging
1601 Scanners: Systematic Review of Clinical Effectiveness, 2001, Available at
1602 [Link] Accessed February 3, 2018.

59
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 60 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1603 27. Ko HC, Huang JY, Miller JR, et al. Novel use of ViewRay MRI guidance for high-dose-
1604 rate brachytherapy in the treatment of cervical cancer. Brachytherapy. 2018;17(4):680-
1605 688.
1606 28. Kirisits C, Lang S, Dimopoulos J, et al. The Vienna applicator for combined intracavitary
1607 and interstitial brachytherapy of cervical cancer: Design, application, treatment planning,
1608 and dosimetric results. International Journal of Radiation Oncology*Biology*Physics.
1609 2006/06/01/ 2006;65(2):624-630.
1610 29. Nomden CN, de Leeuw AAC, Moerland MA, et al. Clinical Use of the Utrecht Applicator
1611 for Combined Intracavitary/Interstitial Brachytherapy Treatment in Locally Advanced
1612 Cervical Cancer. International Journal of Radiation Oncology • Biology • Physics.
1613 2012;82(4):1424-1430.
1614 30. Walter F, Maihofer C, Schuttrumpf L, et al. Combined intracavitary and interstitial
1615 brachytherapy of cervical cancer using the novel hybrid applicator Venezia: Clinical
1616 feasibility and initial results. Brachytherapy. Jun 22 2018.
1617 31. Schwarz JK, Beriwal S, Esthappan J, et al. Consensus statement for brachytherapy for
1618 the treatment of medically inoperable endometrial cancer. Brachytherapy. Sep-Oct
1619 2015;14(5):587-599.
1620 32. Owrangi AM, Jolly S, Balter JM, et al. Clinical implementation of MR-guided vaginal
1621 cylinder brachytherapy. J Appl Clin Med Phys. Nov 8 2015;16(6):490-500.
1622 33. Chapman CH, Prisciandaro JI, Maturen KE, et al. MRI-Based Evaluation of the Vaginal
1623 Cuff in Brachytherapy Planning: Are We Missing the Target? Int J Radiat Oncol Biol
1624 Phys. Jun 1 2016;95(2):743-750.
1625 34. Lindegaard JC, Madsen ML, Traberg A, et al. Individualised 3D printed vaginal template
1626 for MRI guided brachytherapy in locally advanced cervical cancer. Radiotherapy and
1627 Oncology. 2016/01/01/ 2016;118(1):173-175.
1628 35. Perez-Calatayud J, Kuipers F, Ballester F, et al. Exclusive MRI-based tandem and
1629 colpostats reconstruction in gynaecological brachytherapy treatment planning.
1630 Radiotherapy and Oncology. 2009/05/01/ 2009;91(2):181-186.
1631 36. Soliman AS, Owrangi A, Ravi A, et al. Metal artifacts in MRI-guided brachytherapy of
1632 cervical cancer. J Contemp Brachytherapy. Aug 2016;8(4):363-369.
1633 37. Menard C, Susil RC, Choyke P, et al. MRI-guided HDR prostate brachytherapy in
1634 standard 1.5T scanner. Int J Radiat Oncol Biol Phys. Aug 1 2004;59(5):1414-1423.
1635 38. NCT00913939. MRI-Guided HDR Brachytherapy for Prostate Cancer. 2009;
1636 [Link] Accessed December 1, 2019.
1637 39. NCT02342054. Prospective Phase II Trial of Single Fraction Real-time High-Dose-Rate
1638 Brachytherapy in Patients With Low and Intermediate Risk Prostate Cancer. 2014;
1639 [Link] Accessed December 1, 2019.
1640 40. NCT03424694. HDR Brachytherapy Used as Monotherapy for Low and Intermediate
1641 Risk Prostate Cancer: a Phase II Randomized Trial. 2015;
1642 [Link] Accessed December 1, 2019.

60
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 61 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1643 41. NCT02623933. MRI Assisted Focal Boost Integrated With HDR Monotherapy Study in
1644 Low and Intermediate Risk Prostate Cancer Patients (MARS). 2015;
1645 [Link] Accessed December 1, 2019.
1646 42. NCT03323879. Dominant Intraprostatic Lesion Boost With Focused LDR Brachytherapy
1647 (BT) Integrated to Whole Prostate HDR BT: Safety and Feasibility Analysis. 2017;
1648 [Link] Accessed December 1, 2019.
1649 43. NCT04523896. HDR Brachytherapy Plus Stereotactic Ablative Prostate Radiotherapy for
1650 Patients With Intermediate and High-risk Prostate Cancer. 2020;
1651 [Link] Accessed March 10, 2021.
1652 44. NCT01583920. Focal Salvage HDR Brachytherapy for the Treatment of Prostate
1653 Cancer. 2020; [Link] Accessed March 10,
1654 2021.
1655 45. Chassagne D, Dutreix A, Almond P, et al. Dose and Volume Specification for Reporting
1656 Intracavitary Therapy in Gynecology. Journal of the International Commission on
1657 Radiation Units and Measurements. 1985;ICRU 38:1-23.
1658 46. Viswanathan AN, Thomadsen B. American Brachytherapy Society consensus guidelines
1659 for locally advanced carcinoma of the cervix. Part I: General principles. Brachytherapy.
1660 2012;11:33-46.
1661 47. Potter R, Georg P, Dimopoulos JC, et al. Clinical outcome of protocol based image
1662 (MRI) guided adaptive brachytherapy combined with 3D conformal radiotherapy with or
1663 without chemotherapy in patients with locally advanced cervical cancer. Radiother
1664 Oncol. Jul 2011;100(1):116-123.
1665 48. Lindegaard JC, Fokdal LU, Nielsen SK, et al. MRI-guided adaptive radiotherapy in
1666 locally advanced cervical cancer from a Nordic perspective. Acta Oncol. Oct
1667 2013;52(7):1510-1519.
1668 49. Mahantshetty U, Krishnatry R, Hande V, et al. Magnetic Resonance Image Guided
1669 Adaptive Brachytherapy in Locally Advanced Cervical Cancer: An Experience From a
1670 Tertiary Cancer Center in a Low and Middle Income Countries Setting. Int J Radiat
1671 Oncol Biol Phys. Nov 1 2017;99(3):608-617.
1672 50. Pötter R, Tanderup K, Kirisits C, et al. The EMBRACE II study: The outcome and
1673 prospect of two decades of evolution within the GEC-ESTRO GYN working group and
1674 the EMBRACE studies. Clinical and Translational Radiation Oncology. 2018/02/01/
1675 2018;9:48-60.
1676 51. Fokdal L, Sturdza A, Mazeron R, et al. Image guided adaptive brachytherapy with
1677 combined intracavitary and interstitial technique improves the therapeutic ratio in locally
1678 advanced cervical cancer: Analysis from the retroEMBRACE study. Radiother Oncol.
1679 Sep 2016;120(3):434-440.
1680 52. Mazeron R, Fokdal LU, Kirchheiner K, et al. Dose-volume effect relationships for late
1681 rectal morbidity in patients treated with chemoradiation and MRI-guided adaptive
1682 brachytherapy for locally advanced cervical cancer: Results from the prospective
1683 multicenter EMBRACE study. Radiother Oncol. Sep 2016;120(3):412-419.

61
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 62 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1684 53. NRG Oncology. NRG-GY006: A Randomized Phase II Trial of Radiation Therapy and
1685 Cisplatin Alone or in Combination with Intravenous Triapine in Women with Newly
1686 Diagnosed Bulky Stage IB2, Stage II, IIIB, or IVA Cancer of the Uterine Cervix or Stage
1687 II-IVA Vaginal Cancer [Link]
1688 54. [Link]. A Randomized Phase III Trial of Radiation Therapy and Cisplatin Alone
1689 or in Combination With Intravenous Triapine in Women With Newly Diagnosed Bulky
1690 Stage IB2, Stage II, IIIB, or IVA Cancer of the Uterine Cervix or Stage II-IVA Vaginal
1691 Cancer.
1692 [Link]
1693 &rank=2. Accessed October 11, 2019.
1694 55. NRG Oncology. NRG-GY017: Anti PD-L1 (Atezolizumab) as an Immune Primer and
1695 Concurrently with Extended Field Chemoradiotherapy for Node Positive Locally
1696 Advanced Cervical Cancer [Link]
1697 56. [Link]. Atezolizumab Before and/or With Chemoradiotherapy in Immune
1698 System Activation in Patients With Node Positive Stage IB2, II, IIIB, or IVA Cervical
1699 Cancer.
1700 [Link]
1701 +With+Chemoradiotherapy+in+Immune+System+Activation+in+Patients+With+Node+Po
1702 sitive+Stage+IB2%2C+II%2C+IIIB%2C+or+IVA+Cervical+Cancer&draw=2&rank=1.
1703 Accessed October 28, 2019.
1704 57. GEC-ESTRO, Image guided intensity modulated External beam radiochemotherapy and
1705 MRI based adaptive BRAchytherapy in locally advanced CErvical cancer (EMBRACE-II),
1706 2016, Available at [Link] Accessed November 11, 2013.
1707 58. Glasgow GP, Bourland DJ, Grigsby PW, et al., Remote Afterloading Technology, 1993,
1708 Available at Accessed
1709 59. Nath R, Anderson LL, Meli JA, et al. Code of practice for brachytherapy physics: report
1710 of the AAPM Radiation Therapy Committee Task Group No. 56. American Association of
1711 Physicists in Medicine. Med Phys. Oct 1997;24(10):1557-1598.
1712 60. Kubo HD, Glasgow GP, Pethel TD, et al. High dose-rate brachytherapy treatment
1713 delivery: Report of the AAPM Radiation Therapy Committee Task Group No. 59. Med.
1714 Phys. 1998;25(4):375-403.
1715 61. Kutcher GJ, Coia L, Gillin M, et al. Comprehensive QA for radiation oncology: Report of
1716 AAPM Radiation Therapy Committee Task Group 40. Med. Phys. 1994;21(4):581-618.
1717 62. Fraass B, Doppke K, Hunt M, et al. American Association of Physicists in Medicine
1718 Radiation Therapy Committee Task Group 53: Quality assurance for clinical
1719 radiotherapy treatment planning. Med Phys. Oct 1998;25(10):1773-1829.
1720 63. Jackson EF, Bronskill MJ, Drost DJ, et al., Acceptance Testing and Quality Assurance
1721 Procedures for Magnetic Resonance Imaging Facilities: Report of MR Subcommittee
1722 Task Group I, 2010, Available at Accessed
1723 64. Yanasak N, Clarke G, Stafford RJ, et al., Parallel Imaging in MRI: Technology,
1724 Applications, and Quality Control: The Report of AAPM Task Group 118, 2015, Available
1725 at Accessed

62
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 63 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1726 65. Brock KK, Mutic S, McNutt TR, et al. Use of image registration and fusion algorithms and
1727 techniques in radiotherapy: Report of the AAPM Radiation Therapy Committee Task
1728 Group No. 132. Medical Physics. 2017;44(7):e43-e76.
1729 66. Haack S, Nielsen SK, Lindegaard JC, et al. Applicator reconstruction in MRI 3D image-
1730 based dose planning of brachytherapy for cervical cancer. Radiother Oncol. May
1731 2009;91(2):187-193.
1732 67. Kim Y, Muruganandham M, Modrick JM, et al. Evaluation of artifacts and distortions of
1733 titanium applicators on 3.0-Tesla MRI: feasibility of titanium applicators in MRI-guided
1734 brachytherapy for gynecological cancer. Int J Radiat Oncol Biol Phys. Jul 1
1735 2011;80(3):947-955.
1736 68. Miquel ME, Blackall JM, Uribe S, et al. Patient-specific respiratory models using dynamic
1737 3D MRI: preliminary volunteer results. Phys Med. Mar 2013;29(2):214-220.
1738 69. Harry VN, Semple SI, Gilbert FJ, et al. Diffusion-weighted magnetic resonance imaging
1739 in the early detection of response to chemoradiation in cervical cancer. Gynecol Oncol.
1740 Nov 2008;111(2):213-220.
1741 70. Olsen JR, Esthappan J, DeWees T, et al. Tumor volume and subvolume concordance
1742 between FDG-PET/CT and diffusion-weighted MRI for squamous cell carcinoma of the
1743 cervix. Journal of Magnetic Resonance Imaging. 2013;37(2):431-434.
1744 71. Han K, Croke J, Foltz W, et al. A prospective study of DWI, DCE-MRI and FDG PET
1745 imaging for target delineation in brachytherapy for cervical cancer. Radiotherapy and
1746 Oncology. 2016/09/01/ 2016;120(3):519-525.
1747 72. Podder TK, Beaulieu L, Caldwell B, et al. AAPM and GEC-ESTRO guidelines for image-
1748 guided robotic brachytherapy: report of Task Group 192. Med Phys. Oct
1749 2014;41(10):101501.
1750 73. Sun W, Bhatia SK, Jacobson GM, et al. Target volume changes through high-dose-rate
1751 brachytherapy for cervical cancer when evaluated on high resolution (3.0 Tesla)
1752 magnetic resonance imaging. Practical Radiation Oncology. 2012;2(4):e101-e106.
1753 74. Rao YJ, Zoberi JE, Kadbi M, et al. Metal artifact reduction in MRI-based cervical cancer
1754 intracavitary brachytherapy. Phys Med Biol. 2017;62:3011-3024.
1755 75. Trifiletti DM, Libby B, Feuerlein S, et al. Implementing MRI-based target delineation for
1756 cervical cancer treatment within a rapid workflow environment for image-guided
1757 brachytherapy: A practical approach for centers without in-room MRI.
1758 Brachytherapy.14(6):905-909.
1759 76. Pötter R, Federico M, Sturdza A, et al. Value of Magnetic Resonance Imaging Without or
1760 With Applicator in Place for Target Definition in Cervix Cancer Brachytherapy.
1761 International Journal of Radiation Oncology • Biology • Physics. 2016;94(3):588-597.
1762 77. Maenhout M, Peters M, van Vulpen M, et al. Focal MRI-Guided Salvage High-Dose-Rate
1763 Brachytherapy in Patients With Radiorecurrent Prostate Cancer. Technology in Cancer
1764 Research & Treatment. 2017;16(6):1194-1201.
1765 78. Davidson MT, Yuen J, D'Souza DP, et al. Optimization of high-dose-rate cervix
1766 brachytherapy applicator placement: the benefits of intraoperative ultrasound guidance.
1767 Brachytherapy. Jul-Sep 2008;7(3):248-253.

63
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 64 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1768 79. GEC-ESTRO, A European study on MRI-guided brachytherapy in locally advanced

1769 cervical cancer (EMBRACE), 2008, Available at [Link]
1770 Accessed November 11, 2013.
1771 80. Nesvacil N, Potter R, Sturdza A, et al. Adaptive image guided brachytherapy for cervical
1772 cancer: a combined MRI-/CT-planning technique with MRI only at first fraction. Radiother
1773 Oncol. Apr 2013;107(1):75-81.
1774 81. Harkenrider MM, Shea SM, Wood AM, et al. How one institution overcame the
1775 challenges to start an MRI-based brachytherapy program for cervical cancer. Journal of
1776 Contemporary Brachytherapy. 2017;9(2):177-186.
1777 82. D'Amico AV, Tempany CM, Schultz D, et al. Comparing PSA outcome after radical
1778 prostatectomy or magnetic resonance imaging-guided partial prostatic irradiation in
1779 select patients with clinically localized adenocarcinoma of the prostate. Urology.
1780 2003/12/01/ 2003;62(6):1063-1067.
1781 83. Menard C, Pambrun JF, Kadoury S. The utilization of magnetic resonance imaging in the
1782 operating room. Brachytherapy. Jul - Aug 2017;16(4):754-760.
1783 84. Murgic J, Chung P, Berlin A, et al. Lessons learned using an MRI-only workflow during
1784 high-dose-rate brachytherapy for prostate cancer. Brachytherapy. Mar-Apr
1785 2016;15(2):147-155.
1786 85. Kirisits C, Schmid MP, Nesvacil N, et al. Chapter 19: Medical University of Vienna,
1787 Vienna, Austria. In: Song WY, Tanderup K, Pieters BR, eds. Emerging technologies in
1788 brachytherapy. Boca Raton, FL: CRC Press, Taylor & Francis Group; 2017.
1789 86. Kapur T, Egger J, Damato A, et al. 3-T MR-guided brachytherapy for gynecologic
1790 malignancies. Magn Reson Imaging. Nov 2012;30(9):1279-1290.
1791 87. Anderson R, Armour E, Beeckler C, et al. Interventional Radiation Oncology (IRO):
1792 Transition of a magnetic resonance simulator to a brachytherapy suite. Brachytherapy.
1793 2018;17(3):587-596.
1794 88. Busse RF, Hariharan H, Vu A, et al. Fast spin echo sequences with very long echo
1795 trains: design of variable refocusing flip angle schedules and generation of clinical T2
1796 contrast. Magn Reson Med. May 2006;55(5):1030-1037.
1797 89. Liney GP, Moerland MA. Magnetic resonance imaging acquisition techniques for
1798 radiotherapy planning. Semin Radiat Oncol. Jul 2014;24(3):160-168.
1799 90. Bernstein M, K. King, and X. Zhou. Handbook of MRI Pulse Sequences. 1st edition ed.
1800 Burlington, MA: Elsevier Academic Press; 2004.
1801 91. Ma J, Moerland MA, Venkatesan AM, et al. Pulse sequence considerations for
1802 simulation and postimplant dosimetry of prostate brachytherapy. Brachytherapy. Jul -
1803 Aug 2017;16(4):743-753.
1804 92. Paulson ES, Erickson B, Schultz C, et al. Comprehensive MRI simulation methodology
1805 using a dedicated MRI scanner in radiation oncology for external beam radiation
1806 treatment planning. Med Phys. Jan 2015;42(1):28-39.
1807 93. Jan JWL, Bas WR, Cornelis ATVdB, et al. MR guidance in radiotherapy. Physics in
1808 Medicine & Biology. 2014;59(21):R349.

64
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 65 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1809 94. Krupa K, Bekiesinska-Figatowska M. Artifacts in magnetic resonance imaging. Pol J

1810 Radiol. 2015;80:93-106.
1811 95. Mitchell MD, Kundel HL, Axel L, et al. Agarose as a tissue equivalent phantom material
1812 for NMR imaging. Magn Reson Imaging. 1986;4(3):263-266.
1813 96. Schindel J, Muruganandham M, Pigge FC, et al. Magnetic resonance imaging (MRI)
1814 markers for MRI-guided high-dose-rate brachytherapy: novel marker-flange for cervical
1815 cancer and marker catheters for prostate cancer. Int J Radiat Oncol Biol Phys. Jun 1
1816 2013;86(2):387-393.
1817 97. Frank SJ, Stafford RJ, Bankson JA, et al. A novel MRI marker for prostate
1818 brachytherapy. Int J Radiat Oncol Biol Phys. May 1 2008;71(1):5-8.
1819 98. Merkle EM, Dale BM. Abdominal MRI at 3.0 T: the basics revisited. AJR Am J
1820 Roentgenol. 2006;186:1524-1532.
1821 99. Hu Y, Esthappan J, Mutic S, et al. Improve definition of titanium tandems in MR-guided
1822 high dose rate brachytherapy for cervical cancer using proton density weighted MRI.
1823 Radiation Oncology. January 17 2013;8(1):16.
1824 100. Zoberi JE, Garcia-Ramirez J, Hu Y, et al. Clinical implementation of multisequence MRI-
1825 based adaptive intracavitary brachytherapy for cervix cancer. Journal of Applied Clinical
1826 Medical Physics. 2016;17(1):121-131.
1827 101. Schindel J, Zhang W, Bhatia SK, et al. Dosimetric impacts of applicator displacements
1828 and applicator reconstruction-uncertainties on 3D image-guided brachytherapy for
1829 cervical cancer. J Contemp Brachytherapy. 2013;5:250-257.
1830 102. Tanderup K, Hellebust TP, Lang S, et al. Consequences of random and systematic
1831 reconstruction uncertainties in 3D image based brachytherapy in cervical cancer.
1832 Radiotherapy and Oncology. 2008/11/01/ 2008;89(2):156-163.
1833 103. De Leeuw AAC, Moerland MA, Nomden C, et al. Applicator reconstruction and applicator
1834 shifts in 3D MR-based PDR brachytherapy of cervical cancer. Radiotherapy and
1835 Oncology. 2009/11/01/ 2009;93(2):341-346.
1836 104. Tanderup K, Nesvacil N, Potter R, et al. Uncertainties in image guided adaptive cervix
1837 cancer brachytherapy: impact on planning and prescription. Radiother Oncol. Apr
1838 2013;107(1):1-5.
1839 105. Deufel CL, Tian S, Yan BB, et al. Automated applicator digitization for high-dose-rate
1840 cervix brachytherapy using image thresholding and density-based clustering.
1841 Brachytherapy. Jan - Feb 2020;19(1):111-118.
1842 106. Kim Y, Hsu IC, Lessard E, et al. Dose uncertainty due to computed tomography (CT)
1843 slice thickness in CT-based high dose rate brachytherapy of the prostate cancer. Med
1844 Phys. 2004;31:2543-2548.
1845 107. Walker A, Metcalfe P, Liney G, et al. MRI geometric distortion: Impact on tangential
1846 whole-breast IMRT. Journal of applied clinical medical physics. 2016;17(5):7-19.
1847 108. Adjeiwaah M, Bylund M, Lundman JA, et al. Quantifying the Effect of 3T Magnetic
1848 Resonance Imaging Residual System Distortions and Patient-Induced Susceptibility
1849 Distortions on Radiation Therapy Treatment Planning for Prostate Cancer. International
1850 Journal of Radiation Oncology*Biology*Physics. 2018/02/01/ 2018;100(2):317-324.

65
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 66 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1851 109. Pappas EP, Alshanqity M, Moutsatsos A, et al. MRI-Related Geometric Distortions in
1852 Stereotactic Radiotherapy Treatment Planning: Evaluation and Dosimetric Impact.
1853 Technology in cancer research & treatment. 2017;16(6):1120-1129.
1854 110. Brunt JNH. Computed Tomography–Magnetic Resonance Image Registration in
1855 Radiotherapy Treatment Planning. Clinical Oncology. 2010/10/01/ 2010;22(8):688-697.
1856 111. Tan LT, Tanderup K, Kirisits C, et al. Image-guided Adaptive Radiotherapy in Cervical
1857 Cancer. Seminars in Radiation Oncology. 2019/07/01/ 2019;29(3):284-298.
1858 112. Kim Y, Kim Y, Todor D, et al. Recommendations on 3D image-based treatment planning,
1859 dosimetry and quality management for HDR intracavitary brachytherapy: Report of
1860 AAPM Task Group No. 236; Part II: Intracavitary Gynecological Brachytherapy. In
1861 progress.
1862 113. Petric P, Hudej R, Rogelj P, et al. Comparison of 3D MRI with high sampling efficiency
1863 and 2D multiplanar MRI for contouring in cervix cancer brachytherapy. Radiology and
1864 Oncology. 2012;46(3):242-251.
1865 114. Esthappan J, Ma DJ, Narra VR, et al. Comparison of apparent diffusion coefficient maps
1866 to T2-weighted images for target delineation in cervix cancer brachytherapy. J Contemp
1867 Brachytherapy. Dec 2011;3(4):193-198.
1868 115. Haack S, Pedersen EM, Jespersen SN, et al. Apparent diffusion coefficients in GEC
1869 ESTRO target volumes for image guided adaptive brachytherapy of locally advanced
1870 cervical cancer. Acta Oncol. Oct 2010;49(7):978-983.
1871 116. Dyk P, Jiang N, Sun B, et al. Cervical Gross Tumor Volume Dose Predicts Local Control
1872 Using Magnetic Resonance Imaging/Diffusion-Weighted Imaging—Guided High-Dose-
1873 Rate and Positron Emission Tomography/Computed Tomography—Guided Intensity
1874 Modulated Radiation Therapy. International Journal of Radiation
1875 Oncology*Biology*Physics. 2014/11/15/ 2014;90(4):794-801.
1876 117. Tanderup K, Fokdal LU, Sturdza A, et al. Effect of tumor dose, volume and overall
1877 treatment time on local control after radiochemotherapy including MRI guided
1878 brachytherapy of locally advanced cervical cancer. Radiother Oncol. Sep
1879 2016;120(3):441-446.
1880 118. Tanderup K, R. Pötter, J. Lindegaard, C. Kirisits, I. Juergenliemk-Schulz, A. de Leeuw, I.
1881 Fortin, K. Kirchheiner, D. Georg, R. Nout, Y. Seppenwoolde, W. Dörr, T. Liederer, and L.
1882 Tee Tan, Image guided intensity modulated External beam radiochemotherapy and MRI
1883 based adaptive BRAchytherapy in locally advanced CErvical cancer - EMBRACE-II,
1884 2016, Available at [Link] Accessed March 13, 2018.
1885 119. Swanick CW, Castle KO, Rechner LA, et al. Optimizing packing contrast for MRI-based
1886 intracavitary brachytherapy planning for cervical cancer. Brachytherapy. 2015;14(3):385-
1887 389.
1888 120. Citrin D, Ning H, Guion P, et al. Inverse treatment planning based on MRI for HDR
1889 prostate brachytherapy. Int J Radiat Oncol Biol Phys. Mar 15 2005;61(4):1267-1275.
1890 121. Yamada Y, Rogers L, Demanes DJ, et al. American Brachytherapy Society consensus
1891 guidelines for high-dose-rate prostate brachytherapy. Brachytherapy. Jan-Feb
1892 2012;11(1):20-32.

66
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 67 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1893 122. Debois M, Oyen R, Maes F, et al. The contribution of magnetic resonance imaging to the
1894 three-dimensional treatment planning of localized prostate cancer. Int J Radiat Oncol
1895 Biol Phys. Nov 1 1999;45(4):857-865.
1896 123. Rasch C, Barillot I, Remeijer P, et al. Definition of the prostate in CT and MRI: a multi-
1897 observer study. Int J Radiat Oncol Biol Phys. Jan 1 1999;43(1):57-66.
1898 124. Smith WL, Lewis C, Bauman G, et al. Prostate volume contouring: a 3D analysis of
1899 segmentation using 3DTRUS, CT, and MR. Int J Radiat Oncol Biol Phys. Mar 15
1900 2007;67(4):1238-1247.
1901 125. Christie DRH, Sharpley CF. How Accurately Can Prostate Gland Imaging Measure the
1902 Prostate Gland Volume? Results of a Systematic Review. Prostate Cancer.
1903 2019;2019:6932572.
1904 126. Demanes DJ, Martinez AA, Ghilezan M, et al. High-Dose-Rate Monotherapy: Safe and
1905 Effective Brachytherapy for Patients With Localized Prostate Cancer. International
1906 Journal of Radiation Oncology*Biology*Physics. 2011/12/01/ 2011;81(5):1286-1292.
1907 127. Zamboglou N, Tselis N, Baltas D, et al. High-Dose-Rate Interstitial Brachytherapy as
1908 Monotherapy for Clinically Localized Prostate Cancer: Treatment Evolution and Mature
1909 Results. International Journal of Radiation Oncology*Biology*Physics. 2013/03/01/
1910 2013;85(3):672-678.
1911 128. Demanes DJ, Ghilezan MI. High-dose-rate brachytherapy as monotherapy for prostate
1912 cancer. Brachytherapy. 2014/11/01/ 2014;13(6):529-541.
1913 129. Tisseverasinghe SA, Crook JM. The role of salvage brachytherapy for local relapse after
1914 external beam radiotherapy for prostate cancer. Transl Androl Urol. 2018;7(3):414-435.
1915 130. Murgic J, Morton G, Loblaw A, et al. Focal Salvage High Dose-Rate Brachytherapy for
1916 Locally Recurrent Prostate Cancer After Primary Radiation Therapy Failure: Results
1917 From a Prospective Clinical Trial. Int J Radiat Oncol Biol Phys. Nov 1 2018;102(3):561-
1918 567.
1919 131. Tharmalingam H, Hamada M, Tsang YM, et al. Salvage High-Dose Rate (HDR)
1920 Brachytherapy as a Treatment for Locally Recurrent Prostate Cancer after Primary
1921 Radiation Therapy. International Journal of Radiation Oncology • Biology • Physics.
1922 2018;102(3):e118-e119.
1923 132. Crook J, Ots A, Gaztanaga M, et al. Ultrasound-planned high-dose-rate prostate
1924 brachytherapy: dose painting to the dominant intraprostatic lesion. Brachytherapy. Sep-
1925 Oct 2014;13(5):433-441.
1926 133. Strom TJ, Wilder RB, Fernandez DC, et al. A dosimetric study of polyethylene glycol
1927 hydrogel in 200 prostate cancer patients treated with high-dose rate
1928 brachytherapy±intensity modulated radiation therapy. Radiotherapy and Oncology.
1929 2014/04/01/ 2014;111(1):126-131.
1930 134. Yeh J, Lehrich B, Tran C, et al. Polyethylene glycol hydrogel rectal spacer implantation
1931 in patients with prostate cancer undergoing combination high-dose-rate brachytherapy
1932 and external beam radiotherapy. Brachytherapy. May-Jun 2016;15(3):283-287.
1933 135. Sheridan AD, Nath SK, Huber S, et al. Role of MRI in the Use of an Absorbable
1934 Hydrogel Spacer in Men Undergoing Radiation Therapy for Prostate Cancer: What the

67
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 68 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1935 Radiologist Needs to Know. American Journal of Roentgenology. 2017/10/01

1936 2017;209(4):797-799.
1937 136. Gomez-Iturriaga A, Casquero F, Urresola A, et al. Dose escalation to dominant
1938 intraprostatic lesions with MRI-transrectal ultrasound fusion High-Dose-Rate prostate
1939 brachytherapy. Prospective phase II trial. Radiother Oncol. Apr 2016;119(1):91-96.
1940 137. De Brabandere M, Hoskin P, Haustermans K, et al. Prostate post-implant dosimetry:
1941 interobserver variability in seed localisation, contouring and fusion. Radiother Oncol. Aug
1942 2012;104(2):192-198.
1943 138. Frank SJ, Mourtada F, Crook J, et al. Use of magnetic resonance imaging in low-dose-
1944 rate and high-dose-rate prostate brachytherapy from diagnosis to treatment assessment:
1945 Defining the knowledge gaps, technical challenges, and barriers to implementation.
1946 Brachytherapy. 2017;16(4):672-678.
1947 139. Barentsz JO, Richenberg J, Clements R, et al. ESUR prostate MR guidelines 2012. Eur
1948 Radiol. Apr 2012;22(4):746-757.
1949 140. Pugh TJ, Frank SJ, Achim M, et al. Endorectal magnetic resonance imaging for
1950 predicting pathologic T3 disease in Gleason score 7 prostate cancer: implications for
1951 prostate brachytherapy. Brachytherapy. May-Jun 2013;12(3):204-209.
1952 141. Albert JM, Swanson DA, Pugh TJ, et al. Magnetic resonance imaging-based treatment
1953 planning for prostate brachytherapy. Brachytherapy. Jan-Feb 2013;12(1):30-37.
1954 142. Haider MA, Chung P, Sweet J, et al. Dynamic contrast-enhanced magnetic resonance
1955 imaging for localization of recurrent prostate cancer after external beam radiotherapy. Int
1956 J Radiat Oncol Biol Phys. Feb 1 2008;70(2):425-430.
1957 143. Bauman G, Haider M, Van der Heide UA, et al. Boosting imaging defined dominant
1958 prostatic tumors: a systematic review. Radiother Oncol. Jun 2013;107(3):274-281.
1959 144. Groenendaal G, Borren A, Moman MR, et al. Pathologic validation of a model based on
1960 diffusion-weighted imaging and dynamic contrast-enhanced magnetic resonance
1961 imaging for tumor delineation in the prostate peripheral zone. Int J Radiat Oncol Biol
1962 Phys. Mar 1 2012;82(3):e537-544.
1963 145. Menard C, Iupati D, Publicover J, et al. MR-guided prostate biopsy for planning of focal
1964 salvage after radiation therapy. Radiology. Jan 2015;274(1):181-191.
1965 146. Beaulieu L, Carlsson Tedgren A, Carrier JF, et al. Report of the Task Group 186 on
1966 model-based dose calculation methods in brachytherapy beyond the TG-43 formalism:
1967 current status and recommendations for clinical implementation. Med Phys. Oct
1968 2012;39(10):6208-6236.
1969 147. Lambert J, Greer PB, Menk F, et al. MRI-guided prostate radiation therapy planning:
1970 Investigation of dosimetric accuracy of MRI-based dose planning. Radiother Oncol. Mar
1971 2011;98(3):330-334.
1972 148. Mikell JK, Klopp AH, Gonzalez GM, et al. Impact of heterogeneity-based dose
1973 calculation using a deterministic grid-based Boltzmann equation solver for intracavitary
1974 brachytherapy. Int J Radiat Oncol Biol Phys. Jul 1 2012;83(3):e417-422.
1975 149. Rivard MJ, Venselaar JL, Beaulieu L. The evolution of brachytherapy treatment
1976 planning. Med Phys. Jun 2009;36(6):2136-2153.

68
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 69 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

1977 150. Jacob D, Lamberto M, DeSouza Lawrence L, et al. Clinical transition to model-based
1978 dose calculation algorithm: A retrospective analysis of high-dose-rate tandem and ring
1979 brachytherapy of the cervix. Brachytherapy. May - Jun 2017;16(3):624-629.
1980 151. Price MJ, Jackson EF, Gifford KA, et al. Development of prototype shielded cervical
1981 intracavitary brachytherapy applicators compatible with CT and MR imaging. Medical
1982 Physics. 2009;36(12):5515-5524.
1983 152. Rivard MJ, Melhus CS, Granero D, et al. An approach to using conventional
1984 brachytherapy software for clinical treatment planning of complex, Monte Carlo-based
1985 brachytherapy dose distributionsa). Medical Physics. 2009;36(6Part1):1968-1975.
1986 153. Johansson A, Karlsson M, Nyholm T. CT substitute derived from MRI sequences with
1987 ultrashort echo time. Med Phys. May 2011;38(5):2708-2714.
1988 154. Korhonen J, Kapanen M, Keyrilainen J, et al. A dual model HU conversion from MRI
1989 intensity values within and outside of bone segment for MRI-based radiotherapy
1990 treatment planning of prostate cancer. Med Phys. Jan 2014;41(1):011704.
1991 155. Edmund JM, Kjer HM, Van Leemput K, et al. A voxel-based investigation for MRI-only
1992 radiotherapy of the brain using ultra short echo times. Phys Med Biol. Dec 7
1993 2014;59(23):7501-7519.
1994 156. Sjolund J, Forsberg D, Andersson M, et al. Generating patient specific pseudo-CT of the
1995 head from MR using atlas-based regression. Phys Med Biol. Jan 21 2015;60(2):825-839.
1996 157. Dowling JA, Lambert J, Parker J, et al. An atlas-based electron density mapping method
1997 for magnetic resonance imaging (MRI)-alone treatment planning and adaptive MRI-
1998 based prostate radiation therapy. Int J Radiat Oncol Biol Phys. May 1 2012;83(1):e5-11.
1999 158. Gudur MS, Hara W, Le QT, et al. A unifying probabilistic Bayesian approach to derive
2000 electron density from MRI for radiation therapy treatment planning. Phys Med Biol. Nov 7
2001 2014;59(21):6595-6606.
2002 159. Siversson C, Nordstrom F, Nilsson T, et al. Technical Note: MRI only prostate
2003 radiotherapy planning using the statistical decomposition algorithm. Med Phys. Oct
2004 2015;42(10):6090-6097.
2005 160. Kanal E, Barkovich AJ, Bell C, et al. ACR guidance document on MR safe practices:
2006 2013. J Magn Reson Imaging. Mar 2013;37(3):501-530.
2007 161. Greenberg TD, Hoff MN, Gilk TB, et al. ACR guidance document on MR safe practices:
2008 Updates and critical information 2019. J Magn Reson Imaging. Feb 2020;51(2):331-338.
2009 162. Shellock FG. Reference Manual for Magnetic Resonance Safety, Implants, and Devices:
2010 Edition 2017. Los Angeles, CA: Biomedical Research Publishing Group; 2017.
2011 163. Hartwig V, Giovannetti G, Vanello N, et al. Biological effects and safety in magnetic
2012 resonance imaging: a review. Int J Environ Res Public Health. 2009;6:1778-1798.
2013 164. Nixon E, Kim Y, Kearney WR, et al. HDR brachytherapy tandem and ovoid titanium
2014 applicator safety assessment in 3T MRI. Brachytherapy. 2008;7(2):135-136.
2015 165. Woods TO. Standards for Medical Devices in MRI: Present and Future. J Magn Reson
2016 Imaging. 2007;26:1186 -1189.

69
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 70 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

2017 166. ASTM F2503-05 Standard Practice for Marking Medical Devices and Other Items for
2018 Safety in the Magnetic Resonance Environment, ASTM International. West
2019 Conshohocken, PA,2005, [Link].
2020 167. ASTM F2052-06 Standard Test Method for Measurement of Magnetically Induced
2021 Displacement Force on Medical Devices in the Magnetic Resonance Environment,
2022 ASTM International. West conshohocken, PA,2006, [Link].
2023 168. ASTM F2213-06 Standard Test Method for Measurement of Magnetically Induced
2024 Torque on Medical Devices in the Magnetic Resonance Environment, ASTM
2025 International. West Conshohocken, PA,2006, [Link].
2026 169. Murbach M, Zastrow E, Neufeld E, et al. Heating and Safety Concerns of the Radio-
2027 Frequency Field in MRI. Current Radiology Reports. October 28 2015;3(12):45.
2028 170. Kim Y, Chesnut D, Wagner BS, et al. Ferromagnetic Metal Side-Rails on Air-Hover HDR
2029 Patient Transport Table Can Cause Severe Skin Burns to Patients During MR
2030 Simulation for Brachytherapy. International Journal of Radiation
2031 Oncology*Biology*Physics. 2017/10/01/ 2017;99(2, Supplement):E556-E557.
2032 171. Rockey WR, Bhatia SK, Jacobson GM, et al. The dosimetric impact of vaginal balloon-
2033 packing on intracavitary high-dose-rate brachytherapy for gynecological cancer. J
2034 Contemp Brachytherapy. 2013;5(1):17-22.
2035 172. Bou-Zeid W, Bauer C, Kim Y, et al. Clinical validation of a real-time applicator position
2036 monitoring system for gynecologic intracavitary brachytherapy. Biomed Phys Eng
2037 Express. 2016;2:045008.
2038 173. Xia J, Waldron T, Kim Y. A Real-Time Applicator Position Monitoring System (RAPS) for
2039 High-Dose-Rate Intracavitary Brachytherapy. Med Phys. 2013;40:465.
2040 174. Andrew M, Kim Y, Ginader T, et al. Reduction of applicator displacement in MR/CT-
2041 guided cervical cancer HDR brachytherapy by the use of patient hover transport system.
2042 J Contemp Brachytherapy. 2018;10(1):85-90.
2043 175. Gerszten K, Faul C, Kounelis S, et al. The Impact of Adjuvant Radiotherapy on
2044 Carcinosarcoma of the Uterus. Gynecologic Oncology. 1998/01/01/ 1998;68(1):8-13.
2045 176. Siemens Healthineers. Magnetom Skyra - Transforming 3T productivity. 2010;
2046 [Link]
2047 root/wcm/idc/groups/public/@us/@imaging/@mri/documents/download/mdaw/ndq3/~edi
2048 sp/[Link].
2049 177. Chan MF, Cohen GaN, Deasy JO. Qualitative evaluation of fiducial markers for
2050 radiotherapy imaging. Technology in cancer research & treatment. 2015;14(3):298-304.
2051 178. Osman SOS, Russell E, King RB, et al. Fiducial markers visibility and artefacts in
2052 prostate cancer radiotherapy multi-modality imaging. Radiation Oncology. 2019/12/26
2053 2019;14(1):237.
2054 179. Huq MS, Fraass BA, Dunscombe PB, et al. The report of Task Group 100 of the AAPM:
2055 Application of risk analysis methods to radiation therapy quality management. Medical
2056 Physics. 2016;43(7):4209-4262.

70
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 71 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

2057 180. Kim H, Houser CJ, Kalash R, et al. Workflow and efficiency in MRI-based high-dose-rate
2058 brachytherapy for cervical cancer in a high-volume brachytherapy center.
2059 Brachytherapy. 2018;17(5):753-760.
2060 181. Ford EC, Gaudette R, Myers L, et al. Evaluation of safety in a radiation oncology setting
2061 using failure mode and effects analysis. Int J Radiat Oncol Biol Phys. Jul 1
2062 2009;74(3):852-858.
2063 182. Rath F. Tools for developing a quality management program: proactive tools (process
2064 mapping, value stream mapping, fault tree analysis, and failure mode and effects
2065 analysis). Int J Radiat Oncol Biol Phys. 2008;71(1 Suppl):S187-190.
2066 183. Mayadev J, Dieterich S, Harse R, et al. A failure modes and effects analysis study for
2067 gynecologic high-dose-rate brachytherapy. Brachytherapy. 2015;14(6):866-875.
2068 184. Richardson S, D. Scanderbeg, and J. Swamidas. FMEA for brachytherapy. In: Song WY,
2069 K. Tanderup, and B.R. Pieters, ed. Emerging technologies in brachytherapy: CRC Press,
2070 Taylor & Francis Group; 2017.
2071 185. Poder J, Brown R, Howie A, et al. A risk-based approach to development of ultrasound-
2072 based high-dose-rate prostate brachytherapy quality management. Brachytherapy. Sep -
2073 Oct 2018;17(5):788-793.
2074 186. TG275. Strategies for effective physics plan and chart review in radaition therapy In
2075 progress.
2076 187. Nuclear Regulatory Commission. Event Notification Reports.
2077 [Link]
2078 188. Richardson S. A 2-year review of recent Nuclear Regulatory Commission events: What
2079 errors occur in the modern brachytherapy era? Practical Radiation Oncology.
2080 2012/07/01/ 2012;2(3):157-163.
2081 189. Thomadsen BR, Erickson BA, Eifel PJ, et al. A review of safety, quality management,
2082 and practice guidelines for high-dose-rate brachytherapy: Executive summary. Practical
2083 Radiation Oncology. 2014/03/01/ 2014;4(2):65-70.
2084 190. Nesvacil N, Tanderup K, Hellebust TP, et al. A multicentre comparison of the dosimetric
2085 impact of inter- and intra-fractional anatomical variations in fractionated cervix cancer
2086 brachytherapy. Radiotherapy and Oncology. 2013/04/01/ 2013;107(1):20-25.
2087 191. Mazeron R, Champoudry J, Gilmore J, et al. Intrafractional organs movement in three-
2088 dimensional image-guided adaptive pulsed-dose-rate cervical cancer brachytherapy:
2089 Assessment and dosimetric impact. Brachytherapy. 2015;14(2):260-266.
2090 192. Meerschaert R, Nalichowski A, Burmeister J, et al. A comprehensive evaluation of
2091 adaptive daily planning for cervical cancer HDR brachytherapy. J Appl Clin Med Phys.
2092 Nov 8 2016;17(6):323-333.
2093 193. Kandasamy S, Reddy KS, Nagarajan V, et al. Inter-fraction variation in interstitial high-
2094 dose-rate brachytherapy. Journal of Radiotherapy in Practice. 2015;14(2):143-151.
2095 194. Haack S, Kallehauge JF, Jespersen SN, et al. Correction of diffusion-weighted magnetic
2096 resonance imaging for brachytherapy of locally advanced cervical cancer. Acta Oncol.
2097 Aug 2014;53(8):1073-1078.

71
 TG 303 MRI in HDR Brachytherapy Public Review Draft
Page 72 of 72 March 2021

DRAFT – NOT FOR PUBLICATION, QUOTATION, OR CITATION

2098 195. Wood R BK, Foerster V, Spry C, and Tong L., 1.5 Tesla Magnetic Resonance Imaging
2099 Scanners Compared with 3.0 Tesla Magnetic Resonance Imaging Scanners: Systematic
2100 Review of Clinical Effectiveness: Pilot Project 2011, Available at
2101 [Link]
2102 Accessed February 11, 2018.
2103 196. Song WY, Tanderup K, Pieters BR. Section III: Brachytherapy Suites. Emerging
2104 technologies in brachytherapy: CRC Press: Taylor & Francis Group; 2017.
2105 197. Beld E, Seevinck PR, Schuurman J, et al. Development and Testing of a Magnetic
2106 Resonance (MR) Conditional Afterloader for Source Tracking in Magnetic Resonance
2107 Imaging-Guided High-Dose-Rate (HDR) Brachytherapy. Int J Radiat Oncol Biol Phys.
2108 Nov 15 2018;102(4):960-968.
2109 198. Mutic S, Dempsey JF. The ViewRay System: Magnetic Resonance–Guided and
2110 Controlled Radiotherapy. Seminars in Radiation Oncology. 2014/07/01/ 2014;24(3):196-
2111 199.
2112 199. Prisciandaro J, Hadley S, Jolly S, et al. Development of a brachytherapy audit checklist
2113 tool. Brachytherapy. Nov-Dec 2015;14(6):963-969.
2114 200. [Link]. Screening Form. [Link]
2115 Accessed January 13, 2020.
2116 201. Nuclear Regulatory Commission. Title 10 of the Code of Federal Regulations, 35.643.
2117 [Link]
2118 Accessed January 13, 2020.
2119 202. Pei-Jan Paul Lin TJB, Caridad Borras, Gerald Cohen, Robert A. Jucius, Robert J. Kriz,
2120 Edward L. Nickoloff, Lawrence N. Rothenberg, Keith J. Strauss, Theodore Villafana,
2121 Robert K. Cacak, Joel E. Gray, Thomas N. Hangartner, R. Edward Hendrick, Raymond
2122 P. Rossi, AAPM Report No. 39: Specification and acceptance testing of computed
2123 tomography scanners, 1993, Available at
2124 [Link] Accessed
2125 203. Klein EE, Hanley J, Bayouth J, et al. Task Group 142 report: Quality assurance of
2126 medical acceleratorsa). Medical Physics. 2009;36(9Part1):4197-4212.
2127 204. Bissonnette J-P, Balter PA, Dong L, et al. Quality assurance for image-guided radiation
2128 therapy utilizing CT-based technologies: A report of the AAPM TG-179. Medical Physics.
2129 2012;39(4):1946-1963.
2130 205. Nuclear Regulatory Commission. Title 10 of the Code of Federal Regulations, 35.633.
2131 [Link]
2132 Accessed January 13, 2020.
2133

72