High‐dose rate intracavitary brachytherapy pretreatment dwell position verification using a transparent applicator

Abstract Purpose The major errors in HDR brachytherapy are related to treatment distance, almost all of which are caused by incorrect applicator information. The aim of this study is to propose a quick pretreatment verification method to evaluate channel length and dwell position with a transparent applicator, which, in addition, is suitable as an education tool to assist in the understanding of the applicator structure. Methods A transparent applicator model was fabricated using a three‐dimensional printer and transparent resin. Its aim is to be a replica of a real gynecological applicator. The pretreatment verification is performed by observing the planned dwell positions of a check cable inside a transparent applicator. A digital camera acquired images and the dwell positions of the radioactive source and check cable were evaluated by comparing them with respect to the theoretical dwell positions marked by the proper x‐ray marker. The potential effectiveness of verification using a transparent applicator was also evaluated using brachytherapy events reported in the literature. Results The transparent applicator closely resembles the real applicator in shape and had an error of less than 0.2 mm. The average dwell position displacement between the radioactive source and check cable was 0.4 mm. The analysis of brachytherapy events showed that channel‐length, dwell‐position, and step‐size errors made up 50% of all events, but affected 64% of all patients. Conclusions The transparent applicator model enables a noninvasive, repeatable verification of the channel length and dwell positions to be performed before treatment. This verification has the potential to help prevent common errors in treatment delivery. In addition, the transparent applicator model can be used as a teaching tool to help clinicians understand the operation of the applicator, lowering the risk of events.


| INTRODUCTION
In high-dose rate (HDR) brachytherapy for cervical cancer, an applicator is placed in the vagina and penetrates the cervical canal. 1 One of several types of applicators is chosen based on the expansion of the cancer and patient's status. 2 In treatment planning, it is essential to input the correct applicator information (i.e., the offset value and channel length). However, the correct relationship between the first dwell position, offset value, and channel length can be hard to understand because each value depends not only on applicator type but also on the applicator reconstruction method in treatment planning. 3 Because applicators are opaque, users cannot directly visualize the source path inside it with the naked eye, which is one reason incorrect applicator information may be considered. Consequently, incorrect applicator information may result in irradiation at a position that differs from the treatment plan, which might cause unexpected adverse events and the expected therapeutic effects may not be obtained.
Several radiation misdeliveries caused by incorrect applicator information have been reported, all attributable to a lack of understanding of the applicator structure. [4][5][6][7][8][9][10][11][12][13] To prevent accidents, independent pretreatment verification is important. However, the common source position check ruler can be limited because it is generally restricted for use with a specific transfer tube type. 14 If the planned dwell position of the source inside an applicator can be observed, it can be compared with the dwell position displayed on the digitally reconstructed applicator in the treatment plan. An unintended dwell position will be instantly noticed based on an intuitive estimation. Therefore, this study aims to propose a quick pretreatment verification method to evaluate channel length and source dwell position with a transparent applicator and, in addition, suggest its use as an educational instrument.

2.A | Transparent applicator fabrication
To fabricate a transparent applicator, we chose to prepare a Fletcher CT/MR applicator (Nucletron, Elekta AB, Stockholm, Sweden), comprising a tandem and ovoids, which is used in intracavitary brachytherapy (ICBT) for cervical cancer. A commercially available Fletcher-type applicator was scanned using computed tomography (CT; Aquilion LB, Toshiba medical systems, Tochigi, Japan), and three-dimensional (3D) data were acquired. The CT data were acquired in the axial cine mode. The slice thickness was 0.5 mm, which was reconstructed in a field of view of 240 mm on a 512 × 512 grid. All CT slices were transferred to a Digital Imaging and Communication in Medicine (DICOM) viewer (OsiriX Lite version 7.0.3, Bernex, Switzerland) to perform surface rendering. The file format was then changed from DICOM to standard triangulation language (STL). The junction of the applicator and the transfer tube could not be accurately acquired from the CT data because the connector region structure was rather thin. The vendor (Nucletron, Elekta AB, Stockholm, Sweden) hence provided the design of the applicator after we signed a nondisclosure agreement. Based on this design, the 3D data of the connector region were modified using computer graphics software (Shade 3D version 15, Shade3D, Tokyo, Japan). The transparent applicators were fabricated using a 3D printer (ProX800; SOLIZE Products, Kanagawa, Japan) with a minimum layer height of 50 μm and accuracy of 0.05 mm using epoxy polypropylene-based resin as the printing material (Fig. 1). After sculpting a transparent applicator using the 3D printer, the manufacturing accuracy was confirmed by autoradiography (Fig. 2).

2.B | Evaluation of the practicality and accuracy of verification using a transparent applicator
In this study, the verification was performed with a microSelectron-HDR V2 (version 1.51) and Oncentra (version 10; Nucletron, Elekta AB, Stockholm, Sweden). First, an x-ray marker was inserted into the transparent applicators (right ovoid and tandem), and a picture was taken (Fig. 3). The position recognition scale marks were provided digitally based on the proper x-ray marker at 10-mm intervals from    The initial setting of each check cable dwell time is 1 s, but this time can be changed as required.

2.D | Previous HDR event collection
To assess the potential effectiveness of verification using a transparent applicator, we searched the literature for previous studies reporting HDR events and analyzed them. We used information con-  Table 1 shows 54 events that were categorized as dwell positionrelated, dwell time-related, and dwell-position/time-unrelated events.

3.B | Previous HDR event collection
Of these, the wrong channel length (22%) was the most common error, followed by the wrong dwell position (20%). For three subcategories (i.e., wrong channel length, wrong dwell position, and wrong step size), one event involved several patients, implying that these events were not noticed for several treatments. Moreover, the same error was repeated in multiple patients. Consequently, these three subcategories made up 50% of all events, but affected 64% of all patients.
Furthermore, we analyzed the theoretical detectability of the events using our approach. There are only three subcategories of events (i.e., wrong channel length, wrong dwell position, and wrong step size) that could be detected using a transparent applicator; however, these types of errors frequently occur. Table 2 shows the nine events reported from microSelectron-HDR UCS in Japan. While one event was for interstitial brachytherapy (inverse catheter direction), the others were ICBT errors. The wrong channel length and dwell position events made up 40% of all events. Notably, there were two events caused by time lag. One event was caused by an error in the computer system time: the treatment console system (TCS) time was 2 h earlier than the actual time. In the other one, the irradiation time was reduced to one eighth the intended time because of compatibility problems between the TCS and treatment planning system.

| DISCUSSION
In this study, we proposed a novel, simple, quick, and noninvasive pretreatment verification method that uses the check cable and a transparent applicator. Although we used a camera for quantitative observed at the curved portion of the ovoid applicator (Fig. 5), which is larger than the required 1 mm source position accuracy. 16 The cause of that the pathways are different between the x-ray marker and wire (radioactive source and check cable). 17 In this study, the transparent applicator was fabricated using a 3D printer to demonstrate its concept and practicality. The intent is not to recommend the creation of applicators using 3D printers. For clinical use worldwide, it will be necessary to create applicators using molds, and the vendor should produce a transparent applicator as an official product. Table 1 shows that events were frequently caused by the wrong channel length, wrong dwell position, and wrong step size.
These events have been reported to occur in several patients per event. In other words, the detection of these three events is essential for avoiding large-scale accidents. In December 2013, a systematic misdelivery involving as many as 100 patients was disclosed in Japan, all of which occurred because of a lack of independent verification methods before treatment delivery. 12 VariSource.
An institution that has a C-arm may monitor the check cable in the indwelling applicator in the patient using a fluoroscope. 18 Nose et al.
reported a real-time verification method of radioactive source position using modified C-arm fluoroscopy. 15 Although it is ideal to monitor the radioactive source during treatment, radiation regulations may restrict the use of this method. In some countries, regulations limit the concurrent use of two independent radiation sources in a single suite. 19,20 Researchers have summarized the errors and variations in brachytherapy along with their potential for detection by each verification method. 21 One event overlapping with dwell position related and dwell time related.
training are key for avoiding accidents. Training using transparent applicators will make the relationship between the x-ray marker position, radioactive source position, channel length, and offset easy to understand. Wang et al. provided a good tutorial on applicator reconstruction in the treatment planning process. 3 The use of transparent applicators will enable more effective training and assist understanding of the applicator structure.
The current transparent applicator was only designed for pretreatment verification, and the purpose of this verification is to catch gross channel-length, dwell-position, and step-size errors. Note that this verification is not meant to replace the commissioning and periodic QA of real applicators, which require millimeter accuracy because the transparent applicator is a replica and not available for use on patients yet. To use it with patients, it will be necessary to find a biocompatible material that meets the durability and sterilization requirements of applicators. If this problem can be solved, all applicators used for the patient can be transparent, which may be advantageous for QA. The evaluation of the source position is a crucial QA practice in HDR brachytherapy, and these procedures are laborious. A transparent applicator would make it possible to evaluate the source position visually.

| CONCLUSIONS
A transparent applicator will be a good educational tool for improving the understanding of the applicator structure. This implementation of the pretreatment verification method is a simple, quick, and noninvasive, and it enables instantaneous detection of dwell position error that can be estimated visually. This method not only detects dwell position errors, but also errors in the input of the applicator's geometrical parameters and step size. Any significant difference between the actual and intended dwell position of the check cable can be easily identified during verification, thus lowering the risk of fatal errors.