Dosimetric assessment of an air‐filled balloon applicator in HDR vaginal cuff brachytherapy using the Monte Carlo method

Abstract Purpose As an alternative to cylindrical applicators, air‐inflated balloon applicators have been introduced into high‐dose‐rate (HDR) vaginal cuff brachytherapy to achieve sufficient dose to the vagina mucosa as well as to spare organs at risk, mainly the rectum and bladder. Commercial treatment planning systems which employ formulae in the AAPM Task Group No. 43 (TG 43) report do not take into account tissue inhomogeneity. Consequently, the low‐density air in a balloon applicator induces different doses delivered to the mucosa from planned by these planning systems. In this study, we investigated the dosimetric effects of the air in a balloon applicator using the Monte Carlo (MC) method. Methods The thirteen‐catheter Capri™ applicator by Varian™ for vaginal cuff brachytherapy was modeled together with the Ir‐192 radioactive source for the microSelectron™ Digital (HDR‐V3) afterloader by Elekta™ using the MCNP MC code. The validity of charged particle equilibrium (CPE) with an air balloon present was evaluated by comparing the kerma and the absorbed dose at various distances from the applicator surface. By comparing MC results with and without air cavity present, dosimetric effects of the air cavity were studied. Clinical patient cases with optimized multiple Ir‐192 source dwell positions were also explored. Four treatment plans by the Oncentra Brachy™ treatment planning system were re‐calculated with MCNP. Results CPE fails in the vicinity of the air‐water interface. One millimeter beyond the air‐water boundary the kerma and the absorbed dose are equal (0.2% difference), regardless of air cavity dimensions or iridium source locations in the balloon. The air cavity results in dose increase, due to less photon absorption in the air than in water or solid materials. The extent of the increase depends on the diameter of the air balloon. The average increment is 3.8%, 4.5% and 5.3% for 3.0, 3.5, and 4.0 cm applicators, respectively. In patient cases, the dose to the mucosa is also increased with the air cavity present. The point dose difference between Oncentra Brachy and MC at 5 mm prescription depth is 8% at most and 5% on average. Conclusions Except in the vicinity of the air‐mucosa interface, the dosimetric difference is not significant enough to mandate tissue inhomogeneity correction in HDR treatment planning.

the dose distributions. Via dwelling the radioactive source in peripheral channels, a non-circular and more conformal dose distribution to the mucosa may be achievable. A second major issue with solid applicators (single and multiple channel ones) has to do with the potential presence of air gaps between the applicator and the mucosa after applicator insertion. To deliver a sufficient dose to the vaginal mucosa, a direct contact of the applicator outer surface with the vaginal mucosa is imperative,as reported by the ABS. 2 Even a small gap of 1-2 mms can result in a significant dose reduction to the mucosa. As noticed by our and other institutions, 3,4 it is not uncommon to notice from patient CT scans the presence of air gaps at the apex of or along a solid applicator, as long as the vaginal cavity is not in a perfect cylindrical shape after applicator insertion.
The introduction of multi-catheter air balloon applicators is intended to resolve both issues associated with solid applicators.
With multiple catheters, the rectum and bladder are spared without coverage sacrifice thanks to source dwelling in lateral catheters.
Meanwhile, if the outer layer of the balloon applicator presents sufficient elasticity, the shape of the vaginal cavity would not pose a major issue. Pressure from the injected air or saline will push the balloon surface to firmly contact the vaginal wall. Our patient study has confirmed the efficacy of the balloon style Capri vaginal applicator in removing the air gaps around the applicator surface. 5 We randomly selected 25 patients treated with the Capri applicator and another 25 patients treated with the single-catheter cylinder applicators.
Among the patients treated with the cylinder applicators, ring-type air gaps (due to a smaller size of cylinder in the vagina) appeared in 12 patient CT scans. As a comparison, ring-type air gaps appeared in none of the patients treated with the Capri applicator.
One concern regarding air-inflated applicators is the accuracy of dosimetry. We use the program of Oncentra Brachy TM (version 4. Although in HDR brachytherapy, the distance accuracy plays a more dominant role than the tissue composition, it has been a consistent concern of our physicians that a clinically significant dose variation might be present with balloon applicators. Research on the SAVI TM partial breast air balloon applicator by Cianna Medical TM showed a dosimetric effect in the range of 3-9%. 7 The primary goal of this study is to evaluate the dosimetric effect of the air cavity in the Capri balloon applicator using the Monte Carlo (MC) method with the MCNP code.

2.A. | The Multi-catheter Capri balloon applicator
The Capri applicator was invented by Dr. Gail Lebovic, and was the most recent addition to the Varian's applicator inventory. The applicator is intended for brachytherapy treatment of vaginal and rectal diseases. The inflatable design makes the applicator flexible, accommodative, and comfortable to the potential patients. The applicator contains 13 lumens arranged in two concentric rings (six lumens each ring) surrounding a central lumen, as shown in Fig. 1. Markers of different length are attached to the lumens numbered 2, 4, and 6 to allow for easy identification of them in CT images. According to the manufacturer specifications, the deflated insertion diameter is 2.9 cm. Allowance of variable fill volume makes Capri a one-size-fitsall applicator. The balloon diameter after inflation was in the range of 3.5-4.1 cm for about 30 patients treated in our institution so far.
The balloon wall is made of soft silicone, and its thickness is about 0.5 mm when deflated. The balloon can be inflated with either saline or air. Saline was tried in our clinic for the first several cases to achieve water equivalent. However, our physician realized that it was challenging to fill up the applicator interior thoroughly, and for physicists it was difficult to recognize the catheters in CT images in saline-filled regions during treatment planning. Later on, a decision was made to adopt air instead.

2.B. | Ir-192 source modeling
The Nucletron microSelectron HDR Iridium-192 (Ir-192) source was modeled using the MC N-Particle transport code MCNP version 6.0 by Los Alamos National Laboratory. 8 The geometry of the Ir-192 source was illustrated and described in great details by others. 9 The energy distribution of Ir-192 photons has also been well summarized. 10 Photons of Bremsstrahlung, electron capture, and bdecay were all included in our MC simulations. To benchmark the correctness of our iridium source modeling, calculations of the TG 43 parameters adopted into Oncentra Brachy 9 were repeated, and excellent agreements were achieved. The discrepancy of dose rate per unit air-kerma strength (cGyh À1 U À1 in table III in reference 9 ) was generally within 2%. At the longest distances from the source (lowest dose regions), the relative difference might be over 2%, but the absolute difference was insignificant (<0.005 cGyh À1 U À1 ).
2.C. | Validity of charged particle equilibrium Comparison of the absorbed dose to the kerma indicated CPE validity at these preselected locations.

2.D | Dosimetric effects of the air cavity
By comparison of doses with or without the air cavity under the same iridium source condition, the impact of the air cavity on HDR dosimetry can be evaluated. An additional set of MC calculations were performed with the same models as in Section 2C, except that the air cavity was replaced with water.

2.E | Patient study
About 30 patients who suffered intravaginal diseases had been treated with Capri in our institution. The procedure of treatment planning was the following. After a Capri applicator was inserted and inflated with air, the patient was scanned with 2 mm slice thickness on a Philips TM Brilliance Big Bore TM 16-slice CT scanner. The CT images were imported into Oncentra Brachy, and organs at risk (the bladder, the rectum, the sigmoid, and the small bowel) were contoured by the physicians, following the RTOG (Radiation Therapy Oncology Group) guidelines (available at https://www.rtog.org/core lab/ContouringAtlases.aspx). The radiation dose was prescribed at

3.A | CPE
As plotted in Fig. 2, within 1 mm from the air-water interface, the absorbed dose can be significantly higher than the kerma. For

3.C | Patient study results
All the dose points in the patient study were placed at 5 mm depth from the applicator surface, and therefore no electrons were trans- In external beam radiotherapy with X rays, 5% of dose uncertainty is clinically meaningful. However, in brachytherapy with radioactive sources, the distance from the sources plays a much more dominant role, and the dose gradient is usually much higher than that in external beam radiotherapy. Five percent of dose discrepancy can be equivalently introduced by a positioning uncertainty of <1 mm, which is within the acceptable tolerance in brachytherapy. As for at the very vicinity of balloon surface (<1 mm), although not studied in patient cases, significantly higher doses than Oncentra Brachy could be expected.

| DISCUSSIONS
With the application of an air balloon applicator, CPE fails at the very proximity of air/tissue interface (<1 mm). The dose to the mucosa could be notably higher than reported by treatment planning programs. Within this 1 mm region, due to the steepness of the dose gradient, it might be challenging to accurately estimate the dose. Beyond 1 mm, the dose to the mucosa is only slightly elevated (3.8-5.3% on average for the case of a single source located at the center of the applicator), which is attributed to slightly less photon attenuation in the air than in an solid applicator. Our patient study did not discover significant dosimetric differences (8% at maximum and 5% on average) at 5 mm prescription depth between air-filled and water-equivalent applicators, either. Treatment planning with Oncentra Brachy or other TG-43 style planning systems is clinically acceptable for the air balloon applicators such as the Capri.
In HDR vaginal cuff brachytherapy, two methods of prescribing radiation dose was suggested by the ABS 11 : dose is prescribed to either the vaginal surface or at 5 mm depth. With balloon applicators, prescribing dose to the applicator surface may not be applicable anymore, because doses at the applicator surface in treatment plans might not be realistic. On the other hand, the dose at 5 mm depth is insignificantly affected by air balloon, and as a result the use of 5 mm as the prescription depth is still feasible.
T A B L E 1 Comparison between doses (cGy) calculated with Oncentra (air in balloon treated as water) and Monte Carlo (air in balloon modeled as air) for two patient cases of 35 mm Capri applicator.