Clinical implementation, logistics and workflow guide for MRI image based interstitial HDR brachytherapy for gynecological cancers

Abstract Interstitial brachytherapy (IBT) is often utilized to treat women with bulky endometrial or cervical cancers not amendable to intracavitary treatments. A modern trend in IBT is the utilization of magnetic resonance imaging (MRI) with a high dose rate (HDR) afterloader for conformal 3D image‐based treatments. The challenging part of this procedure is to properly complete many sequenced and co‐related physics preparations. We presented the physics preparations and clinical workflow required for implementing MRI‐based HDR IBT (MRI‐HDR‐IBT) of gynecologic cancer patients in a high‐volume brachytherapy center. The present document is designed to focus on the clinical steps required from a physicist’s standpoint. Those steps include: (a) testing IBT equipment with MRI scanner, (b) preparation of templates and catheters, (c) preparation of MRI line markers, (d) acquisition, importation and registration of MRI images, (e) development of treatment plans and (f) treatment evaluation and documentation. The checklists of imaging acquisition, registration and plan development are also presented. Based on the TG‐100 recommendations, a workflow chart, a fault tree analysis and an error‐solution table listing the speculated errors and solutions of each step are provided. Our workflow and practice indicated the MRI‐HDR‐IBT is achievable in most radiation oncology clinics if the following equipment is available: MRI scanner, CT (computed tomography) scanner, MRI/CT compatible templates and applicators, MRI line markers, HDR afterloader and a brachytherapy treatment planning system capable of utilizing MRI images. The OR/procedure room availability and anesthesiology support are also important. The techniques and approaches adopted from the GEC‐ESTRO (Groupe Européen de Curiethérapie ‐ European Society for Therapeutic Radiology and Oncology) recommendations and other publications are proven to be feasible. The MRI‐HDR‐IBT program can be developed over time and progressively validated through clinical experience, this document is expected to serve as a reference workflow guideline for implementing and performing the procedure.


| INTRODUCTION
Cervical and endometrial cancer are the most common gynecologic cancer worldwide. 1 Because of the anatomic advantage of being able to place catheters directly within the tumor, a high dose can be given locally while still limiting the dose to neighboring organs.
Due to the global effort to provide improved access to radiation therapy and the importance of brachytherapy in the management of gynecologic malignancies in recent decades, there has been an increase in the number of high-dose rate (HDR) gynecologic (GYN) brachytherapy programs world-wide 2 . Brachytherapy has become an integral treatment for locally advanced cervical cancer and other gynecologic malignancies. 3 In cervical cancer management, the use of brachytherapy is found to have significantly associated with higher cause-specific survival and overall survival and is therefore considered as a vital component of therapy. 4 For bulky cervical or endometrial cancer, the interstitial brachytherapy (IBT) is often utilized to treat in situations that are not amendable to intracavitary treatments.
Although CT (computed tomography) images can be used for HDR IBT, magnetic resonance imaging (MRI) has been shown to be superior to CT in soft tissue delineation, which is helpful in accurate identification of the precise extent of the tumor. [5][6][7] However, because most radiation oncology departments do not have an MRI scanner within the department, the proper handling of MRI images in brachytherapy planning can be challenging to both physicists and physicians. Considering the substantial technical barriers to implementing and commissioning the procedure, a document providing a practical and easy-to-follow logistics, workflow and guideline can be helpful for a beginner to start a new MRI based HDR IBT (MRI-HDR-IBT) program, and for an existing program to find the pros and cons of peer groups.
Aimed at accomplishing that goal, this document presents the categorized physics preparation steps, a sequenced workflow and verified implementation checklists following the AAPM medical physics practice guideline. 8 A tabulated dosimetric metrics table was also presented. While the current document examines GYN IBT in detail, the basic techniques and preparation steps are also relevant for other anatomic sites. In this document we have chosen to focus our efforts on MRI-HDR-IBT of GYN origin, because the IBT of these cancer types has limited workflow guidelines in the literature compared to other types of brachytherapy procedures. 9 The primary recommendations are based on GEC-ESTRO publications, 10,11 but many of the specific techniques, checklists and work-flow recommendations are derived from our institution's experience.
Recently the TG-100, an AAPM task group focused on the application of risk analysis methods to radiation therapy quality management, has published the recommendations for establishing radiotherapy quality assurance (QA) programs and clinical procedures with the cross verified process-trees using a failure modes effects analysis (FMEA) or fault tree analysis (FTA). 12 Since the MRI -HDR-IBT involves many sub-steps and procedures, this document has generated a workflow chart and FTA chart following the TG-100 recommendations, and the potential errors and solutions as well. Some radiation oncologists prefer to use metal catheters (i, e., needles) for IBT, because metal catheters are rigid and their insertion is felt to be easier than more flexible plastic catheters. 13 However, the titanium catheters that were once classified as MRI compatible previously have been removed from the CT/MRI compatible equipment list by NIH owing to the distortions of surrounding tissue caused by the titanium needle artefact. 14 It has also been reported that precise localization of the titanium needle tip is often difficult, 15 and tumor delineation may be obscured by MRI signal cancellation of titanium material. 16  The consensus index length of catheters for Oncentra TPS (Oncentra, version 3.5, Elekta) and MicroSelectron TM HDR afterloader (Nucletron, Elekta AB, Stockholm, Sweden) is determined by the Elekta source position simulator (SPS) and verified by experimental film tests (In our system, the consensus index length is 1251 mm. Reader needs to get their own). For each new batch of catheters, a physicist is required to verify the length from a randomly picked catheter, making sure the purchase order was accurately carried out and manufacturing uncertainty is within expectations (±2 mm).
The graphical offset of first dwell point relative to the catheter tip is an important parameter for calculating accurate dose distributions of the targets and organs at risk (OARs). By taping a catheter onto Gafchromic film (Ashland, Inc., Covington, KY), programming a dwell point at the most distal source position of the catheter (e.g., the consensus index length), and delivering radiation over a duration of 0.5 s, the graphical offset from the first physical dwell point to the catheter tip was determined to be 6 mm (±1 mm). Based on the average value of measurements, 6 mm is utilized as the consensus offset in CT-based plan for our HDR afterloader. For the MRI-based plan, the catheter has a different visual structure. The catheter tube itself has a 2−3 mm solid end at the tip, and the tip of the MRI line marker tube has a 2-3 mm solid end as well. These solid ends are not visible on MRI images when the catheters are tracked, thus the 6 mm offset seen from the film and CT is compensated by the solid ends in the MRI image, so the graphical offset of the catheter in the MRI-based plan is approximated as 0.  17 and CuSo4 solution. 18 We tested a list of T2 contrast agents such as the radiology lab T2 agents (copper oxide and gadolinium solutions), medical saline and two types of nylon fishing lines in a water tank. The gel might be a better medium than water for testing the devices used in MRI imaging, but our department did not have gel for testing, so we did not use it. Of note, some difficulties were encountered during tests/dry runs, the problems were progressively resolved. The T2 MRI contrast agents recommended by MRI researchers at our institution were found to provide good visualization of plastic catheters when the surrounding medium was water. However, when these same agents were used in real cases of gynecology cancer patients, it was found that the catheters could not be accurately visualized given the gynecological cancer lesions were dark in the T2 MRI images like the catheters visualized by the tested agents. We performed additional dry runs in which we hoped the catheters could give a white hyperintense appearance. Two types of nylon fishing lines and a plastic tube filled with medical saline were tested in three real cases. It was found the plastic tubes (at 1 mm in diameter) filled with medical saline provided the best visualization of the catheters on T2-weighted MRI images.

2.A.3 | MRI compatible syed templates
There are a variety of MRI compatible templates available for IBT treatments. The template most commonly used at our institution is acquired through Best Medical International Incorporation and is F I G . 1. Two plastic interstitial catheters and a metal obturator. This picture shows an interstitial catheter with the metal obturator and an interstitial catheter with the protective cap and the obturator removed.  The patient will stay on the MRI couch for about 45 min to complete all above 10 sequences. The sequence of AX 3D T2 TSE is the primary scan utilized for brachytherapy planning, the scan slice size is 1 mm for a better spatial resolution, and it takes 6-9 min. Other listed sequences are undertaken for additional assistance in tumor verification and radiologic interpretation, readers may not need them.
A successful treatment plan relies on the quality of MRI images.
Given this, a rule was set at our institution that a medical physicist who is familiar with the MRI imaging protocol be physically present during MRI image acquisition.

MR imaging checklist:
1. Assure that the patient is transported between the cart and the MRI scanner couch in a fashion that minimizes catheter disruption to avoid altering patient position or moving the implant. ____   The most important step is the primary image must be loaded into the TPS first. The primary image is defined as the image to be used to track applicators or catheters for making the treatment plan. If the catheters can be tracked in the scanned T2 MRI image, then the T2 MRI image will be used as the primary image, the reconstruction of catheters, delineations of target volumes and OARs and development of treatment plans will be completed within it, the CT scan will not be needed.
In the case when the CT image had to be used as the primary image, the MRI image would become secondary image. All the OARs will be contoured from the CT image in addition to tracking all catheters. The target volumes will be delineated from the co-regis-

3.A | General strategies of planning
The optional planning strategies developed at our institution include: a The case label should be written with the modality and fraction number, such as MRI plan Fx1, or CT plan Fx2. Thus, it can be clearly identified in the treatment control system (TCS) to avoid loading the wrong plan. This is particularly helpful if a modification of plan using the same set of images is necessary during the current IBT course. For example, when a catheter obstruction is discovered after first fraction of treatment, a new plan will be made based on the available catheters with the same images, the fraction number (such as MR plan Fx2) will help identify the new plan in the treatment consol.  d Create a list of points onto the surface of the HR-CTV, then allow the TPS to normalize the plan with those points to make an initial basic plan. Starting from the initial basic plan, the MD will use the graphical optimization method to tune the plan for a desired coverage.

3.B | Recommended planning and treatment workflow
In order for readers to easily manage their own workflow of planning and treating the patient, a planning checklist adopted at our institution is listed here for reference.   | 43 axial view, the whole catheter pattern will be shown. (Fig. 5).

3.C | Treatment documentation
For a brachytherapy plan or course, the plan evaluation and dosimetric documentation are very important, because it will indicate if the delivered dose can achieve an expected tumor control (through the F I G . 7. An example workflow for MRI-HDR-IBT following the AAPM TG-100 guidelines. 12 summed dose to the target volume) and if the OARs will exceed the expected toxicity level (through the OAR dosimetric metrics).
The following dosimetric documents are implemented in our institution: a At the first fraction, MD will review and approve the dosimetric metrics (the DVH tables and iso-coverage graphs), b A physics consult document will be created at the first fraction and signed by MDs, The physics consult document will reflect the techniques used, image modalities utilized and physics equipment involved.
c If one or more channel(s) could not go through the obstruction test and had to be skipped during the treatments, an updated plan and DVH table reflecting the actual dose coverage must be provided. MD will review and sign the new DVH tables. e After the last treatment, all doses to the targets and OARs from all fractions will be summed, the physicist will provide the dose summary document. Table 1  | 45 by all team members. An example of MRI-HDR-IBT workflow is shown in Fig. 7.
It is noted that the workflow details will vary across institutions depending upon available resources and their unique situations.
Per TG-100, after the workflow is well understood by the involved personnel, failure mode and effect analysis (FMEA) should be performed by the team to identify which processes cause high risk among the possible failure modes. 12 As seen in Fig. 8 images.
An implant diagram reflecting the locations and numbers of catheters inserted during the OR procedure will be drawn by MD on a schematic Syed template graphic as a reference.
T A B L E 3 The times needed for each sub-parts of the procedure for delivering first fraction of treatment.

4.C | Times spent in different sub-procedures
Although complicated, the entire procedure takes approximately 9 h in total and can be divided into 8 major parts. The Table 3 presents the average time of each part. The first work is the MRI line marker preparation which happened one day before the treatment day.
Other sub-procedures happened on the treatment day. In brachytherapy planning room two hours of time is budgeted for MD's target delineation, this includes teaching residents, verifying resident's work and tuning the initial treatment plan by attending MD. If teaching is not involved, the target delineation could be done in one hour in our institution. Figure 9 is the pie map of the time budget for completing the procedure. It indicated that the insertion of catheters in OR and delineation of targets and OARs in brachytherapy planning room by MDs account for half of the procedure time (Fig. 9).

CONF LICT OF I NTEREST
The work was not supported, funded, or sponsored by any extra-institutional source, nor are there any actual or potential conflicts of interest with the production or publication of this work. No author has any direct or indirect commercial financial incentive associated with publishing this article.