Aims

The primary aim of this study is to investigate the benefits and limitations of supine and prone treatment positions for APBI planned for the MRL. This will involve a comparison of the location of the tumour bed and its proximity to OAR in each position, an assessment of the difference in geometric image distortion close to the target, and measurement of the of the patient’s breathing motion.

We hypothesise that prone positioning for APBI will result in less geometric distortion and less breathing (intrafraction) motion when compared to supine positioning. It will also increase the relative distance between the target volume and organs at risk (heart, lung(s), and chest wall) resulting in lower doses to these structures.

Primary outcomes:

1.

Quantification of geometric image distortion on planning MRI scans in both the supine and prone positions to determine whether there is a position specific difference.

2.

Measurement of relative distance from the Planning Target Volume (PTV) to the OAR in the supine and prone positions.

3.

Assessment of breathing motion for each patient in each position to determine whether motion is larger in the supine or prone position.

Secondary aims of this study include:

1.

To assess the differences in dose to the target, organs at risk and normal tissue for each patient between the supine and prone positions on planning MRI scans (simple synthetic computed tomography scans (CTs) generated from anatomical contours with bulk density overrides).

2.

To explore the patient experience of having an MRI simulation scan and whether there is a preference for the supine or prone position.

Secondary outcomes:

1.

Assess differences in dosimetry (dose in Gray) to the target, OAR and normal tissue for each patient between the supine and prone positions on synthetic CTs generated from the MRI scans.

2.

Analyse survey responses (content analysis) to ascertain whether there is a patient preference for the supine or prone position.

Study design

This is a prospective, exploratory single-centre pilot study of patients undergoing APBI for early-breast cancer. Patients meeting the American Society for Radiation Oncology (ASTRO) favourable and certain cautionary criteria for APBI will be invited to participate (Table 1) Patients with known contraindications to radiotherapy such as ataxia telangiectasia and systemic sclerosis will be excluded [13]. Participants will be asked to provide written informed consent. Participation will not alter the management of the patients but will inform the decision-making process for the choice of treatment position for future patients treated with ABPI, based on the study outcomes.

Table 1 ASTRO APBI Favourable and cautionary suitability criteria [13]

Participants will be scanned on a dedicated Philips Ingenia Ambition MR-RT 1.5T MRI simulator (Koninlijke Philips N.V., Amsterdam, Netherlands) in both the supine and prone positions for APBI as per current departmental standard of care. Assessment of positioning, geometric distortion, patient and breathing motion, and dosimetry will be completed using the de-identified data from these scans. In addition, immediately following the MRI simulation session, participants will be invited to complete a short anonymous electronic survey (using closed and open-ended questions) about their experience of the planning MRI scans (MR specific sensations, information provided about the procedure and positional preferences). CT simulation and treatment will be completed as per standard protocol with appropriate immobilisation equipment in the preferred position (supine or prone as a clinical decision after MR image review) with no further study-based requirements.

Recruitment

Participants will be invited to participate and provided with the relevant study information to make an informed decision about their participation by their Radiation Oncologist (RO) in their initial consultation if they are eligible for treatment with APBI and meet the study inclusion criteria.

Positioning assessment

Participants will attend for MRI simulation prior to CT simulation in alignment with standard departmental protocol. For the supine MRI scan, patients will be positioned with both arms up supported by a custom vacuum bag. For the prone MRI scan, patients will be positioned on the Low Procline™ (CDR Systems Inc. Calgary, Canada), which was developed in collaboration with one of our investigators. To make the positional, distortion and breathing motion assessment the standard breast MRI sequences (T1, T2 and Cine imaging) will be acquired in each position, optimised for each laterality. The minimum distance will be calculated between the Target volume and heart, lung(s), skin rind (5mm inside the patient contour), and chest wall.

Participants will be assigned to have either the supine or prone MRI scan first (alternating between the positions) as they are recruited, to reduce bias towards the first or second scan. The order will also be recorded and considered in the analysis of results.

Geometric distortion assessment

Geometric distortion in MRI is a known artifact arising from inherent magnetic field inhomogeneities that becomes larger further from the imaging isocentre [14]. Geometric distortion will be assessed in MIM Maestro® version 7.1 or later (MIM) utilising the dPhantom (3DOne Australia Pty Ltd, QLD, Australia), a specific MRI distortion phantom containing a 2 cm MRI visible grid. Two MR images will be taken of the phantom; a reference that is centred in the bore, and a second MR will be taken with the phantom shifted a known distance to cover the region of interest, e.g., Breast. Patient images in the supine and prone positions will be contoured, and structures transferred to the shifted MR. The shift from the ground truth 2 cm grid will be calculated within each contour. The mean and maximum shifts will then be calculated as a measure of geometric distortion for each contour to support comparison of distortion between the two positions.

Breathing motion assessment

Cine imaging sequences in sagittal-coronal directions acquired during the MRI simulation in both the supine and prone positions will be used to assess the range of breathing (intrafraction) motion. Images will be collected for 1-min intervals at multiple timepoints within the session in each position. Imaging and treatment will be performed with the patient free breathing. Motion will be measured across a minimum of five breathing cycles using tumour or surrogate edge displacement in Anterior/Posterior, Superior/Inferior and Left/Right directions. The exhale baseline drift will be assessed across the breathing phases and the mid-point between maximum and minimum exhale position will be used as the reference point. The maximum and average extents of the motion will be measured in both the supine and prone positions for each patient.

Dosimetric assessment

Patients will have all subsequent planning and treatment performed as per the standard departmental protocol. There is no change to usual care for radiotherapy planning procedures.

For study purposes, each patient’s supine and prone MR scans will be contoured (target volumes, appropriate OAR and structures requiring a specific electron density override) using MIM then used to make a pseudo-CT; average densities will be acquired from the collective supine and prone CT scans (acquired as part of the patient’s standard of care) and applied as average population bulk densities to structures as appropriate in the Elekta Monaco™ version 5.51.11 or later (Elekta solutions AB, Stockholm, Sweden) radiotherapy treatment planning system.

In this study protocol, a MR-only workflow will not be used. Every patient will receive a CT scan following their MR sim appointment in the treatment position which will provide patient specific electron densities as per current clinical practice in our department. Contours will be applied to the MRI and bulk electron densities will be assigned to the contours in the planning system as per the current MRL workflow to create the synthetic CT. Quality assurance of the average population electron density will be performed in line with the departmental protocol for online adaptive treatment, recalculating the treatment plan on the CT and comparing the dosimetry from the planning CT and the synthetic CT.

Target volumes will be contoured by RO study investigators and all other structures by RTs. The primary target volume will consist of visible seroma and the surrounding post operative tumour bed. Standard departmental margins will be applied to generate the Clinical Target Volume (CTV) and planning target volume (PTV). APBI plans for 30 Gray (Gy) in 5 fractions with the Elekta Unity™ (Elekta solutions AB, Stockholm, Sweden) MR Linac beam model will be completed for comparison on the pseudo-CT scans in both the supine and prone positions. Beam angles will be optimised to the ipsilateral side and adjusted appropriately for tumour location, patient position, treatment laterality and to avoid entry through the cryostat pipe. Planning considerations specific to the Unity; couch top, MR radiofrequency coil and structure/density layering will also be applied. Calculation and sequencing parameters will adhere to standard departmental protocol for APBI. Doses to the targets and OAR (Heart, Lung(s), ipsilateral chest wall, ipsilateral (breast–PTV), contralateral breast and skin rind (5mm)) will be collected to ascertain whether a particular position provides a benefit over the other dosimetrically.

Patient experience survey

Immediately following the simulation MR scans, participants will be asked to complete a simple, short anonymous survey to assess their experience of each position (see Appendix 1 “Patient Experience Survey”). The survey will ask patients to rate their experience of each of the positions and other common experiences of MRI scans on a Likert scale with an opportunity to explain their rating as free text comments.

This survey has been adapted and developed specifically for this study from a similar tool used by Barnes et al. [15] to investigate the patient experience of treatment on the MRL, as no validated tools were identified in the literature. The survey for this study was critically reviewed by four consumers with radiotherapy treatment experience prior to protocol submission for ethical approval. The survey will be made available to participants on an electronic device at the time of their simulation appointment.

Sample size estimation

Recruitment will cease at 12 months or when 30 patients have been recruited. This pragmatic sample size is based on the previous 12-month period of simulation appointments for this diagnosis.

Where available, existing MRI and CT datasets collected as standard of care for APBI patients will be included in the geometric distortion, positioning and dosimetry analysis aspects of the study. It is anticipated that up to an additional 30 datasets will be included in this analysis, with this de-identified data accessed under a Human Research Ethics Committee (HREC) approved waiver of consent.

Statistical analysisPositioning assessment

To compare the supine and prone positioning, categorical data will be presented as counts and frequencies, with means and standard deviations (for normally distributed data) or median and inter-quartile ranges (for non-normally distributed data) used to present continuous variables. Where appropriate, paired parametric or non-parametric tests will be applied to test for differences or equivalence between the supine and prone positions.

Patient experience

Quantitative survey results will be interpreted by comparison of percentage frequencies across the Likert scales. Chi Squared and Fishers Exact tests may be used to test for differences between categories, with McNemar’s test used to compare paired responses between prone and supine positioning.

Qualitative data from the open-ended survey questions will be analysed using thematic content analysis [16].

Data will be collected in a Research Electronic Data Capture (REDCap) database and/or Microsoft Excel with statistical analysis planned to be conducted using Stata version 18.0 (StataCorp, College Station, Texas, USA), with a p-value of < 0.05 used to indicate statistical significance.