A review of nonstandardized applicators digitization in Nucletron™ HDR procedures

Abstract The major errors in HDR procedures were failures to enter the correct treatment distance, which could be caused by either entering wrong transmission lengths or imprecisely digitizing the dwelling positions. Most of those errors were not easily avoidable by enhancing the HDR management level because they were caused by implementations of nonstandardized applicators utilizing transmission tubes of different lengths in standard HDR procedures. We performed this comprehensive study to include all possible situations with different nonstandardized applicators that frequently occurred in HDR procedures, provide corresponding situations with standard applicator as comparisons, list all possible errors and in planning, clarify the confusions in offsets setting, and provide mathematical and quantitative solutions for each given scenarios. Training on HDR procedures with nonstandardized applicators are normally not included in most residential program for medical physics, thus this study could be meaningful in both clinical and educational purpose. At precision of 1 mm, our study could be used as the essential and practical reference for finding the correct treatment length as well as locating the accurate dwelling positions in any HDR procedure with nonstandardized applicators.

multiple clinical uses. 13 After years of clinical implementation, however, those early ideas were gradually adopted by brachytherapy vendors that started to market new, more powerful, and standardized applicators to meet the increasing use of HDR brachytherapy for different anatomical sites. Thus, contemporary HDR applicators tend to be standardized in several essential aspects. First, many applicators are compatible with radiograph or computed tomography (CT); i.e., when metallic markers are inserted into the applicators, they are visible in radiographs as well as in CT slices. Although applicators compatible to both CT and magnetic resonance (MR) have been developed for image-guided HDR brachytherapy, as recommended by GEC-ESTRO, 14 HDR planning is more frequently CT-based than MRbased. Second, in rigid applicators for example, gynecology (GYN) Tandem & Ovoid applicators by Nucletron TM (an Elekta Company), transfer tubes are applicator-specific and not exchangeable, and the first dwell position (FDP, also referred to as reference distance) of the radioactive source is standardized. Third, when flexible applicators (such as MammoSite of Hologic, Inc.) are implanted, total channel lengths (TCLs) are typically shorter than those of the rigid applicators and must be measured for each applicator or channel.
Nevertheless, these nonstandardized HDR applicators are still being used in many cancer centers nationwide. The word "nonstandardized" hereafter refers to applicators that are either non-CT compatible, or that do not have applicator-specific transfer tubes, or both.
Compared to the new and sophisticated HDR applicators, those "oldfashioned" ones were mostly customized for treatment sites with less concern for user convenience, and hence bring about drawbacks in multiple aspects. The greatest advantage of old-fashioned applicators, however, is that for the most part they were designed to be compatible with standardized transfer tubes. Therefore, there is no need to purchase extra transfer tubes particularly designed for them.
It is required that full calibration of an HDR unit should include determination of source positioning accuracy to within AE1 mm. 15 The typical verification or QA methods for HDR brachytherapy source dwell positions were proposed so that radiographic films could be used in direct contact with applicators. 16 According to the recommendations of the AAPM TG-59 report, 17 in recent years, revised radiographic methods for dwell position measurement were reported for specific applicators 17 ; other novel methods, such as using fluorescent screens were also proposed for dwell position verification. 18 With the recent development of image-guided radiotherapy, new instrumentation as well as planning tips were proposed for precise digitization of applicators in rapid HDR procedure workflows. 19,20 All those endeavors were essentially based on regular applicators implemented in HDR brachytherapy procedures. Difficulties or issues in digitizing nonstandardized applicators in prevalent procedures are always major causes for treatment errors.
Nonstandardized HDR applicators are neither CT nor radiograph compatible, and therefore markers are invisible in the images even if they were inserted to the applicators before the CT scan or radiography. Though some nonstandardized applicators meet the second or third aspect of standardization mentioned above, there are always many exceptions. For instance, when an older Miami applicator 13 is implanted for an HDR brachytherapy on GYN cervix cases, applicator-specific transfer tubes for vaginal cylinder applicators are used for all five channels (one channel for the tandem and four for the ovoids respectively). TCLs of the ovoids are much shorter than that of the tandem, one obvious drawback. In Table 1, scenarios for   implementing nonstandardized applicators are listed, compared to implementing standardized applicators.
As the major advantage for nonstandardized applicator designs, compatibility to existing applicator-specific transfer tubes is also a costly tradeoff because these can cause errors in digitization. Once an applicator is commissioned, it could be used properly for a long time regardless of standardization. However, using nonstandardized HDR applicators has been challenging to medical physicists, making it necessary and crucial to have comprehensive preparations prior to the procedure. According to our experience, in most CAMPEP residency programs, medical physicist residents are formally trained to perform HDR procedures only under standardized conditions. Experienced medical physicists who rarely have exposure to customized applicators would also expect an inevitable learning curve in their practice. Here, we address the major issues in commissioning nonstandardized HDR applicators, and provide a comprehensive guideline on digitization for all conditions with nonstandardized applicators. For convenience and consistency, standardized Nucletron applicators and transfer tubes are used as references. In typical HDR procedures, either tip-end or connector-end digitization could be performed, depending on user's preference. The most updated treatment planning system (TPS) uses tip-end digitization by default, and therefore here we will concentrate on that method. In tip-end digitization, distance is measured from the remote afterloader indexer, and thus all lengths and dwell positions have positive values.
For the purpose of applicator digitization, it is often important to measure the distance from the applicator tip to the FDP, which is thereafter termed the offset; the offset should have a negative value because the positive direction points toward the indexer.
The HDR vendor provides a simulator wire with built-in x-ray markers. The end portion of the wire consists of three metallic beads, shown as three consecutive bright spots on CT slices. The central spot is defined as the tip-end marker position. This is not the physical end of the applicator channel, even if the front marker bead touches the inner end. The end-to-end length of the three metallic beads is exactly 8 mm. If the front marker bead touches the channel end (the zero-gap scenario), there will be a 4 mm distance between the tip-end marker position and the channel end, and in that case the tip-end marker position is digitized as the FDP of the 192 Ir source. In this scenario, to maintain a necessary safety margin, there will be a 2 mm gap between the applicator channel end and the tip of the active source when located at the FDP.
Nonstandardized applicators, however, are manufactured such that they have slightly longer TCLs than standard applicators; these usually satisfy the zero-gap scenario. However, if the x-ray markers are extended in the nonstandardized applicators by the same distance as in standardized applicators, the front marker will see a distance of up to a few millimeters between front marker and the channel end (nonzero-gap scenario). This common feature in nonstandardized applicators makes their digitization different (at least in principle) from that of standardized applicators.

2.B | Use of rulers supplied by vendors
Even if an HDR catheter and the x-ray markers can be digitized in CT image or radiograph, the user must perform measurements to determine or verify the lengths defined in the previous section. In general, the reference distance of an applicator might not be exactly 4 mm from the channel end (nonzero-gap scenario). To take into account all possible scenarios, Fig. 3 shows the general situation of digitizing an HDR channel, but using a standardized applicator as an example. The offset of the FDP consists of the tip-end wall thickness Λ, the gap x between inner end of applicator channel and the front marker, and half-length (4 mm) of the x-ray markers. The parameters are correlated with the follow equation: The negative sign of offset means TCL is longer than the reference length, as required by the TPS. It should be noted that if x = 0, the equation reduces to the zero-gap scenario, as shown later in Fig. 4. In fact, if the FDP is 1500 mm, it is rare that x is exactly zero.
T A B L E 1 Comparisons of standardized to nonstandardized applicators in combination with applicator-specific or general transfer tubes, based on Nucletron TM products. 1350 1340 The position indicator (yellow piece) of a SPS, with its center located at 1340 mm. If the dummy wire of the simulator reaches the end of applicator, the indicator forehead shows the TCL (1342 mm), whereas its tail-end indicates the reference distance or FDP (1338 mm).
Furthermore, zero-gap scenarios occur more commonly in the channels of flexible applicators where FDP is less than 1500 mm.

2.C | Digitization of an HDR applicator
This subsection presents the principles of HDR applicator digitization that are common to both standardized and nonstandardized applicators (implementation of these principles is addressed in the next section). The first step of applicator digitization is to determine its total length, usually required by the TPS. The second step is to determine the tip-end marker position or the FDP. Next, the user digitizes the remaining dwell positions towards the connector end to allow for sufficient treatment length. For nonstandardized HDR applicators, special care should be always taken in the first two steps regardless of the transfer tubes being used. It is critical to note that each step could present challenges due to nonstandardization. The third step is not generally applicable for nonstandardized applicators, since many of them are not radiograph or CT compatible, and thus reasonable estimations are typically introduced.
The TCL and FDP can always be measured using SPS and x-ray markers respectively. The accuracy of SPS measurement can be veri-

Source position & offset
Offset 1500 X 4 F I G . 3. The relative positions of x-ray markers, SPS, and 192 Ir source in an applicator that has a reference distance of 1500 mm. Λ is the tip-end thickness, and x is the gap between the applicator end and the front marker. Please note that markers might be invisible in a nonstandardized applicator. (c) place a scaled film with the applicator channel and deliver the plan. The length between the center of the exposure spot and the outer end of applicator is the applicator offset in eq. (1), noting that offset is a negative number, thus

Source position & offset
3.B | Digitization of flexible applicators with nonapplicator-specific transfer tubes  Given that in such an applicator the FDP is shorter than 1500 mm, the dwell positions of the radioactive source will be accordingly shorter. In the digitization, we assume the zero-gap scenario for the dummy source when the SPS is used to measure the TCL and the FDP. To measure the tip-end thickness Λ, the same procedures addressed in Section 3.A may be followed. The formula for determining the FDP and the offset is: For negligible applicator thickness (Λ = 0), the offset can be set to 4 mm. In most cases, the typical offsets are 4-6 mm. The digitization procedure is also similar to that shown in Fig. 4, except that the metallic markers are invisible in these non-CT-compatible applicators.

3.D | Common errors in digitizing non-CTcompatible applicators
Two types of errors frequently occur in digitizing nonstandardized applicators with applicator-specific transfer tubes; i.e., applicators are not CT compatible, with FDP presumably to be 1500 mm.
The first type of error (Type-I) in digitization (also the most common one) is to directly digitize the physical tip of the catheter from the CT or the radiograph as the tip-end marker position, using 1500 mm as the FDP as well as zero offset. Thus the inner gap x, It is important to point out that neither Type-I nor Type-II errors will be identified by the remote afterloader system, because the source always travels as far as the FDP, and this is true even if the digitized spot is at a distance beyond it. This is also the reason that those errors could not be easily discovered, because delivery may go through successfully. If a positive offset rather than a negative offset is typed into the TPS, however, this error can be rejected by the delivery system, because in this situation the dummy wire will attempt to travel beyond the physical end of the catheter channel.
Another common mistake (Type-III) is to digitize the FDP by reading the central position of the SPS position indicator, and this mistake occurs frequently when flexible applicators are implemented.
Because the SPS indicator is 4 mm long, a Type-III mistake will cause the end of the active source to touch the inner end of the catheter channel. In most cases this will cause no issues from the TPS. Still, if a treatment plan based on the wrong measurement is delivered, the IDLs will obviously be shifted upstream by 2 mm (panel C of Fig. 7) with respect to the intended dose distribution (panel A of Fig. 7). We should also stress that standardized, flexible applicators of the same model may slightly differ in total channel length. It is therefore extremely important to commission each flexible applicator (or each channel of a multichannel applicator) before any patient treatment.

| DISCUSSION
The major difference between the TPS of Nucletron and Vari-Source is that the Nucletron TPS simplifies a dwelling position into a point (or a spot), whereas the VariSource TPS displays the physical length of the HDR source. In the two systems, the dwell position definitions are remarkably different: the dwell position in a Nucletron system is the center of the source, but in a VariSource system the dwell position is located at the tip of the active wire. Nevertheless, the same principles presented in this study could be implemented in a VariSource system even though they were discussed here as applied to a Nucletron system.

| CONCLUSION S
We performed a comprehensive review and study on nonstandardized applicators for typical HDR procedures using Elekta Nucletron TM system, with a view to recommending strategies that overcome the dominant errors or uncertainties caused by incorrect digitization of the channel length and/or dwell positions. We considered situations that are likely to occur in HDR procedures, listed possible errors in each of them, and provided corresponding solutions.

ACKNOWLEDG MENTS
We thank Dr. Adrian Koesters, Research Editor at UNMC, for her excellent contribution in professional English on revising our manuscript.

CONFLI CT OF INTEREST
The authors declare no conflict of interest.