Use of a PLA sleeve to remove electron enhancement in superficial X-ray therapy

A newly installed superficial X-ray unit was found to produce enhanced electron dose at the skin surface. The ACPSEM kilovoltage dosimetry recommendations suggest using nail varnish within the treatment cones as a method to reduce this dose. In this study, a 3D PLA sleeve was produced and used as an alternative to the nail varnish for energies between 55 and 100 kV. Further, plastic wrap was also investigated as an alternative method to reduce dose. It was found that a 1 mm printed sleeve, inserted into the treatment cone sufficiently reduced the enhanced dose as measured with a thin-window Exradin chamber to within 3.3% of the dose measured with a Farmer-type ionisation chamber. The use of plastic wrap also reduced the enhanced dose, but impracticalities in its use make it non-viable for routine clinical use.


Introduction
The incidence of Non-Melanoma Skin Cancer (NMSC) in the Australian population is estimated to be around 2% and it continues to be an increasing problem due to the nature of the climate and lifestyle [1]. Whilst the use of Mohs Micrographic Surgery is considered the gold standard for treatment of skin cancers, the use of superficial X-rays for the treatment of skin cancers has proven to be the preferred treatment for many patients [2]. The major recipients of superficial X-ray treatment (SXRT) will be patients typically greater than 70 years who may present with co-morbidities. Further, many of the sites that may be difficult to treat with surgical means, are ideal candidates for radiotherapy. The dosimetry of kilovoltage beams is well established, but with any new treatment machine design, issues can present [3,4]. This work is related to a newly released machine and the issue of electron contamination and subsequent surface dose enhancement.
Treatment with SXRT uses cones (also known as applicators) that are inserted into the treatment head and direct the collimated treatment beam to the targeted area. The cone design are typically stainless steel in construction through the main body. The cones that are supplied with the treatment machines vary in field size, SSD, shape, and whether they are open-ended or close-ended using a window of Perspex or similar material. There are also older models of cones which were made with lead glass but are less common in modern machines. Dose enhancement due to electron contamination from the walls of the cone is a known consequence of using open-ended cones [5]. In that work, the production of scattered dose from the metal applicator walls contributes dose to measurements of backscatter factors and dose outputs when measured with a thin window low-energy X-ray ionisation chamber. Without the addition of an absorbing material the dose measured at the end of the cone and delivered to the patient can be enhanced significantly due to contaminant electrons depositing in the first 1 3 few micrometres of skin [6]. Klevenhagen also showed the addition of 30 µm of polyester significantly reduced the dose measured. In comparison, close-ended cones provide sufficient shielding to remove that dose enhancement due to contaminant electrons.
One method suggested in the literature to remove the electron contamination dose is to coat the inner 10 cm of the cone with nail varnish [7]. The acrylic of the nail varnish sufficiently absorbs the electrons produced. This does provide an inexpensive method to alleviate the issue but is not clear in how many coats are sufficient.
Nail varnish is most commonly a nitrocellulose dissolved in a solvent like ethyl acetate. As the solvent evaporates, the resin in the nitrocellulose hardens forming a coating on the surface. Application of the nail varnish to the inner surface of the cones can be both time consuming and may give a non-uniform thickness throughout the cone. Further, the nail varnish will degrade over time, requiring it to be reapplied and potentially being a hygiene hazard when the cones are positioned directly in skin contact.
Hang et al. [8] found that a piece of paper placed at the end of the treatment cone would sufficiently absorb the electron enhanced dose. Similarly, the use of plastic wrap has been proposed to reduce the enhanced dose at the surface [6]. Plastic wrap is commonly used in the food industry and home applications. It is generally a form of PolyVinyl Chloride but depending upon manufacturer could also be a lowdensity polyethylene. It is generally found in thicknesses in the range of 7.6-10 μm [9].
An alternative proposed here is to use 3D printed resin to absorb the contaminant electrons. 3D printing is available within many clinical departments with the application including bolus for megavoltage treatment and to provide customised shielding in SXRT applications [10]. Online CAD software and desktop 3D printers enable the design of, and printing for a range of applications and uses.
The work here proposes to examine the reduction in electron enhancement dose at the central axis when using a 3D printed Poly-Lactic Acid (PLA) sleeve inserted into the treatment cone. In addition, the use of plastic wrap was investigated as another useful method.

Materials and methods
All tests were conducted on a WOmed T105 (Wolf Medical, Germany) SXRT machine. The machine has four available treatment energies listed in the table below. The Half-Value Layers (HVLs) listed in Table 1 were measured using the methodology recommended in AAPM TG61 protocol and as per ACPSEM kilovoltage dosimetry recommendations using high purity aluminium sheets (Gammex, USA) for all energies [7,11]. This was performed with the FC-65P ionisation chamber and the CNMC Model 10 electrometer (CNMC, USA).
The circular cones provided with the treatment unit were 1 cm, 2 cm, 3 cm, and 4 cm diameter open-ended stainlesssteel cones at 15 cm SSD. In addition, there were two circular stainless steel close-ended cones (Perspex); 5 cm at 15 cm SSD and 10 cm at 30 cm SSD. The cones in the range of 2-4 cm are primarily used in the routine clinical practice.
Polylactic Acid (PLA) is the most used material in desktop extrusion printers as it requires a low heat to print, does not require a heated bed, and is relatively inexpensive. It has a physical density of 1.210-1.430 g/cm 3 , and when printed in a thin layer is still relatively robust [12].
PLA has been assessed in its use for external beam radiotherapy [10, 13] and found to be a safe and effective material to use for electron and photon modalities for bolus. Other investigators have examined the use of PLA in 3D printing and the uncertainty and reproducibility in manufacture [14]. PLA manufacture can be inconsistent in measured densities, which would be problematic where the radiation beam will be passing through the material. In this application, any minor variation in production is not anticipated to significantly affect the dose measured at the centre of the treatment field but may have consequences for the field edge.
A 1 mm thick sleeve was designed in TinkerCad (Autodesk Inc, San Rafael, USA) software and printed at a commercial print company using the Original Prusa i3 MK3S 3D printer (Prusa Research a.s., Czech Republic). The wall thickness of 1 mm was the thinnest that could be reliably printed and still provide a robust structure. The sleeve was printed using the highest quality setting of 0.1 mm available to ensure a smooth curve, and 100% infill.
Plastic wrap is commercially available and used in many clinics to wrap lead used for shielding. Wrapping the lead aims to prevent toxicity to staff handling the shields, or skin contact for patients, as well as to prevent the contaminant electron dose at the field edge [6].
Multiple dose output readings in this work were measured using the AAPM TG61 protocol in-air methodology [11]. All measurements taken were delivered using 30 s readings, via the timer setting, and corrected for temperature and pressure. The 5 cm diameter close-ended cone is the reference cone for this machine. Measurements were limited in this work to the 4 cm diameter cone for proof of concept. Smaller cones were not measured due to the relatively large size of the measurement chambers.
In determining absolute dose for kV beams, reference protocols such as AAPM TG61 ( [11] recommend the use of chambers that will sufficiently absorb any contaminant electrons. For energies < = 100 kV, the wall thickness of the chamber should be at least 13.4 mg/cm 2 to absorb any electron contamination. The wall thickness for the Farmertype chamber FC-65P (IBA Dosimetry GmBH, Germany), is 57 mg/cm 2 , and is therefore a suitable chamber for absolute dosimetry for SXRT.
Hill et al. [7] recommend the use of a thin window parallel-plate chamber to check for electron contamination. The available low-energy thin window chamber was an Exradin A11TW chamber (RPD Inc, USA) which has a wall thickness of 3.86 mg/cm 2 . Therefore, measurements conducted for the same cones, using the Exradin chamber will include the electron contamination where present.

Electron dose contamination contribution
To estimate the electron dose contamination contribution, measurements were performed in two stages. In the first stage, the output for the selected cones was measured in-air using the FC-65P ionisation chamber. It is noted that filter 1 is below the recommended energy use for the FC-65P chamber but was included for comparison. In the second stage, the Exradin thin window chamber was used to measure the output dose once again for all treatment energies.
A ratio of these measured values was then used to calculate an Electron Dose Enhancement Factor (EDEF) as a measure of the electron enhanced dose. This ratio will also include other scatter and output components but this work will focus on the portion due to electron contribution.

Dose reduction using plastic wrap
The plastic wrap used was Glad® Wrap commonly used domestically. A series of measurements were performed for one energy (Filter 2/70 kV) with increasing layers of plastic wrap at the end of the cone to establish the thickness required to observe a significant dose reduction. This was then repeated at a single thickness for all available energies.

Dose Reduction with and without Sleeve
The initial printed sleeve was taped in place within the cone, shown in Fig. 1, and the output was measured and compared with the output without the sleeve in situ. A ratio of these values was calculated for the EDEF (with sleeve) and compared to the initial EDEF (without sleeve) with the Exradin Chamber.

Effect of sleeve on measured energy/HVL
To establish the effect of the sleeve on the dose measured, the HVLs were measured with and without the sleeve in place for all available treatment energies. The HVLs were measured with the FC-65P chamber, according to the method recommended in AAPM TG61 [11].

Electron dose contamination contribution
The dose at the end of the 4 cm and 5 cm cones was measured for all available clinical energies using the FC-65P and Exradin chambers. The Electron Dose Enhancement Factors (EDEFs) were calculated, and the difference between the two chambers has been attributed to electron enhanced dose, see Table 2. The increase in dose measured by the Exradin chamber varied between 9.9 and 14.4%.

Dose reduction using plastic wrap
The reduction in dose with added layers of plastic wrap is shown in Table 3 and was up to 3% for the maximum thickness measured.

Dose reduction with and without PLA sleeve
Measurements were made with the Exradin chamber and the reduction in dose was up to 15.2% as shown in Table 5. Table 6 compares the measured EDEFs with the sleeve in place as measured with the Exradin chamber, and the FC-65P measured EDEFs (no sleeve). The difference in the measured EDEFs is up to a maximum of 3.3%.

Effect of sleeve on measured energy/HVL
The graphs below show the results with and without the sleeve in place.  The reduction in dose output at the end of the cone was measured for all other available energies using eight layers of plastic wrap for comparison. The reduction in dose is shown in Table 4.

Discussion
Initial measurements undertaken in Table 2 show the ratio of output between the 4 cm and 5 cm cone between two different chambers to estimate the electron enhancement for the 4 cm cone. The 5 cm cone is close-ended, and thus absorbs the electron enhancement contribution. In addition, the measurements performed using the FC-65P reference chamber do not show the enhanced dose due to the wall thickness of the FC-65P. The wall thickness for the FC-65P chamber exceeds the recommended minimum thickness as listed Table 1 of AAPM TG61 [11]. Therefore, measured doses using the FC-65P chamber are only due to the kilovoltage photons. Measurements made with a low-energy chamber, such as the Exradin, will include the contaminant electron dose due to the thin front window. The EDEFs shown in Table 2 estimate the electron dose contaminant contribution to be within 9.9-14.4% for the range of energies measured. This measurement agrees with other researchers estimates of up to 14% [8].
The use of plastic wrap showed that there could be an appreciable dose reduction by up to 9.3% for the 4 cm cone up to 100 kV (filters 3 and 4). Plastic wrap could provide a reasonable solution as it is readily available. In this series of measurements, eight layers of wrap were used, but this does not remove all enhanced dose and further layers would be needed to eradicate the effect. In layering plastic wrap, it is difficult for the layers to be made uniformly and without wrinkles. The additional time taken in a busy clinic would be non-trivial to prepare this for all patients, as well as the need of consideration of re-using for the multiple fractions, or waste by making new for each fraction for each patient. Other commercially available thicker plastic sheeting (200 m) such as "Builders Film" may be reasonable to use, and reuse, but have not been measured in this study.
The use of the internal PLA sleeve showed a reduction in the dose delivered by up to 15.2%. This would indicate complete eradication of the contaminant dose to the measurement point. Table 6 compares the EDEFs measured with the FC-65P chamber (without sleeve), and the Exradin chamber (with sleeve). The sleeved measurements are within 3.3% of the FC-65P chamber measurements. It is noted that for filter 1, the EDEF is within 1.1% of the FC-65P chamber value, but the FC-65P chamber is not recommended for this energy range.
Measurements were made of the HVL, as an indicator of any change in energy. The measurements showed that the presence of the internal sleeve did not affect the HVL. The graphs shown in Fig. 2(a-d) show no change in the beam attenuation with added filtration for measurements with and without the sleeve. Only the first HVL was measured in this study to ascertain its relevance to clinical application.
One factor that has not been measured in this study was the reduction in field size by the inclusion of the sleeve. It is anticipated that film measurements would demonstrate some field size reduction, and penumbral broadening. Given these considerations, the clinic has decided to not include a sleeve for the smallest cone available. The 1 cm cone is typically used to treat inner canthus lesions. A 1 mm thick wall for the inner sleeve could potentially shrink the treatment size for full dose by up to 20%. For the small number of fields using this cone, the plastic wrap would provide a preferable option.
In addition, taping the sleeve in place, as performed here, is non-ideal. After these measurements were completed, showing proof of concept, the design was refined to ensure a snug fit of the sleeve in place. The sleeve can be removed for cleaning as required or replaced should it degrade. The sleeve is being used clinically and treatment staff have found the sleeve preferable to layering sheets of plastic wrap for its time efficiency.

Conclusion
This work examined the use of a simple 3D printing option to remove the dose enhancement due to electron contamination observed at the end of open-ended stainless-steel treatment cones. The sleeves provided a reduction in measured dose by up to 15.2%, aligning the measured EDEF closely to that from a Farmer-type chamber. Plastic food wrap is also able to remove some of the dose but with handling impracticalities is less desirable in a busy clinical department. The sleeve used is a robust addition into the clinic and is well received by staff.