Radiotherapy quality assurance

As stated in the Methods (see Chapter 2), all participating centres in the Fast-Forward Trial underwent a comprehensive RT QA programme.

Background

The complex nature of modern RT carries inherent problems both in ensuring reproducibility and accuracy within a RT unit and, more particularly, when carried out on a multicentre basis. Specific issues in the treatment of the breast/chest wall and nodal region arise from the geometry and proximity of the associated treatment volumes, with important radiation-sensitive structures underlying the breast, chest wall and nodal region including the lung, heart and brachial plexus.

Careful localisation, computerised planning, accurate verification of beam position and meticulous attention to alignment and matching during treatment are essential.

A QA programme is ‘a mandatory prerequisite when aiming at high dose, high precision radiotherapy’ and is an integral component of any RT trial as defined by the EORTC guidelines for trial protocols in RT.52,53

In this multicentre randomised trial, the RT QA programme enabled confirmation that technical guidelines within the protocol were understood and implemented correctly by participants and that the dose prescription was delivered according to protocol, together with appropriate documentation of technique and patient-related data. This ensured that clinical observations in terms of tumour control and normal tissue damage reflect differences in the randomised schedules rather than departures from trial protocol. Techniques used have been documented and these data are available should differences in observed outcomes emerge.

In this way the definition of RT QA as ‘all those planned and systematic actions necessary to provide adequate confidence that a product will satisfy given requirements of quality’54 can be satisfied and the scientific worth of the parent trial be validated.

The RT QA programme for the Fast-Forward Trial was built on the foundations of the START and IMPORT trials. This has provided an element of consensus in RT techniques among RT centres. FAST-Forward helped with the implementation of new technology in some centres where the use of intensity-modulated radiotherapy (IMRT) or image-guided radiotherapy (IGRT) had not been used previously.

Plan of investigation

The RT QA programme followed the guidelines set out by the EORTC53 and the RT QA component of the trial was coordinated by the National RTTQA Group based at Mount Vernon Cancer Centre.55,56 A total of 47 RT centres took part in the UK over a 3-year period 2011–2014. Protocol Versions 1.0–2.2 of the RT QA programme applied to whole breast/chest wall RT and, from version 3.0 onwards, centres had to gain additional QA approval for nodal RT.

The RT QA programme included both pre-trial and on-trial components. It was completed by all participating centres, unless streamlined through the IMPORT trial, and is detailed below:

Pre-trial QA

1. A facility questionnaire and process document collating details of techniques, equipment and processes to be used within the centre for FAST-Forward patients.
2. A dummy-run (centre-selected sample patient) plan containing target and organ-at-risk (OAR) structures outlined by the principal investigator and used to create a clinical plan, was assessed to ensure adherence to the trial protocol.
3. Once the nodal study was introduced through Protocol v3.0, each centre was required to undergo credentialing for lymphatic nodal outlining and planning through completion of an outlining and a planning benchmark case (RTTQA Group provided cases). Centres were also asked to submit an amended process document detailing additional information in relation to treatment of patients randomised to the nodal study.

A visit by the RTTQA Group was also performed prior to centre participation to validate, independently, the technique used against the information given in the process document. In particular, the following parameters were assessed:

1. Target volume and treatment technique used.
2. Confirmation of plan modulation – implementation, for example IMRT forward planning.
3. Planning of dose distribution across the treatment volume for homogeneity and prescription points.
4. Assessment of routine quality checks (QC) performed by the centre compared with current IPEM guidelines.57
5. Dose measurements across the treatment volume within a purpose-made phantom, if not performed for the same technique within the last 3 years.
6. Assessment of imaging verification protocol and technique.

On-trial QA

1. A prospective review of the outlines and RT plan was performed for the first breast and first chest wall patients recruited from each centre prior to the patient starting treatment. On introduction of the nodal study a prospective review of the nodal CTV outlines and RT plan was conducted for the first patient recruited at each centre. Feedback was provided to the investigator site within 48 hours of receipt of data. If any major protocol variations were identified, additional reviews were requested.
2. All patients’ RT planning CT, plan file, dose file, structure set and plan assessment forms were anonymised by the centre and submitted electronically to the RTTQA Group.
3. Retrospective outlining and planning reviews were performed to monitor continued protocol compliance. Any protocol variations were reported and discussed with the Chief Investigator (CI) and ICR-CTSU.

Quality control by department for IMRT

Where a centre had an established IMRT programme that had been previously credentialled by the RTTQA Group for an equivalent trial, some aspects of the FAST-Forward RT QA programme were streamlined. Where an established IMRT programme was not set up, additional QA was required such as verification of fluence maps for each field.

Analysis of the RT QA programme

The data from the RT QA programme have been analysed separately from the Main Trial. Any major variations from trial protocol would have been notified to the ICR-CTSU, the CI and participating centres. These included:

1. Discrepancies in documentation, dose prescription and dose recording.
2. Failure to meet upper and lower dose limits for treatment volumes.
3. Systematic errors of technique in any stage of treatment from planning through to implementation.

Detailed analysis of the RT QA data has produced quality information covering the following areas:

1. variations in breast RT practice in participating centres
2. a comparison of methods used for IMRT (multiple static fields, dynamic fields)
3. an assessment of the emerging technologies and their quality control
4. quantification of dose uniformity during the treatment period.

Results

A number of studies were carried out by the RTTQA Group evaluating the data collected within FAST-Forward. These are summarised below:

1. Field-based and volume-based PTVs as plan evaluation structures in the UK FAST-Forward breast trial.58

Centres were initially given the option of reporting either a field-based or volume-based breast PTV. Analysis of the initial 338 plans from eight centres showed that the majority of centres were using a field-based technique (seven centres). This study also showed that in some patients there were significant areas of high dose (105% of prescription) outside of the reported volume-based PTV. In order to ensure consistency in reporting across all patients and to ensure that all regions of high dose were captured within the reported PTV, field-based PTV reporting was used for all patients within the trial. Field-based PTVs were added by the RTTQA Group for any patients that were submitted with volume-based PTV.

1. Interim analysis of treatment plans in the FAST-Forward hypofractionation breast RT trial.59

This study analysed the RT treatment plans collected and evaluated their compliance with the trial protocol. An interim analysis was performed when the first 504 plans were available. This demonstrated that 95.6% of these plans complied with the dose requirements of the FAST-Forward protocol, but highlighted differences in OAR doses and PTV coverage related to the treatment technique used. Centres using field-based planning (the majority) included less lung and heart in the field compared to volume-based planning.

1. Results from the RT QA programme for the FAST-Forward breast trial.60

Additional work was carried out to analyse the RT treatment plans collected in the FAST-Forward Trial to monitor consistency of treatment across all centres and compliance with trial protocol.

The review performed on 3600 plans submitted as part of the FAST-Forward Trial demonstrated that, overall, the quality of the RT treatment plans had been maintained across centres. The majority of plans (97.3%) complied with the dose objectives specified in the protocol, with only 2.7% of all reviewed data not achieving one or more of the dose objectives. On further examination more than half of these variations were either unavoidable or not significant. Unavoidable variations (1.5%) included complex cases where a clinical compromise was deemed necessary, for example eight plans exceeded the V105% and V107% tolerances with higher doses to a non-clinically significant volume. Avoidable variations (0.8% of plans) included misinterpretation of the DVH information at the investigator site and/or inaccurate contouring of evaluation structures.

1. Variability of contouring using the European Society for Radiotherapy and Oncology (ESTRO) guideline for elective breast cancer RT.61

As part of the FAST-Forward Nodal Sub-Study, interobserver variability of nodal contouring was evaluated against the ESTRO consensus guideline for elective breast cancer RT.

It was concluded that contouring of the regional lymph nodes using the ESTRO consensus guidelines reduced interobserver variability in volumes contoured.62 Results suggested that experience and/or training is associated with less interobserver variability, promoting the role of RT QA when adopting new outlining guidelines as part of a multicentre trial. In addition, the methodology developed helped to inform evaluation of nodal contouring within the trial as part of prospective and retrospective analyses.

Discussion

Radiotherapy quality assurance plays an important role in ensuring trial protocol compliance and consistency in delivery of RT across multiple centres; establishing best practice and improving RT standards. The benefits of RT QA have been shown by numerous clinical trials in ensuring that clinical outcomes reflect differences in randomisation schedules rather than departures from trial protocol.63 It is particularly important when introducing novel techniques and frequently centres use clinical trial participation to implement new techniques supported by trial protocol and comprehensive RT QA programmes. The pre-trial and on-trial RT QA programmes implemented for the FAST-Forward Trial monitored the RT planning and delivery with the aim to deliver precise radiation dose to the target, maximising the dose to the tumour and minimising the dose to surrounding organs at risk, in order to improve the therapeutic index.64 A particular benefit of the detailed QA programme for this trial was that it gave confidence in adopting IGRT in breast cancer RT – which was not common, and often regarded as a low priority. This has benefits for patients both within and outside of the trial.

Findings from the RT QA studies provided reassurance that all FAST-Forward participating centres could comply with the standards required by the trial protocol. The first study examined the use of field-based and volume-based PTV to assess the suitability as breast plan evaluation structures. This informed centres on the difference between the two techniques and enabled centres to make a more informed choice.58–61 Further studies looked to monitor the consistency of RT treatment across all centres and compliance with the trial protocol. The majority of the plans were protocol compliant and any deviations were explored for future guidance.60 The variability of contouring using the ESTRO consensus guidelines was reviewed as the FAST-Forward Nodal Sub-Study required delineation of axillary nodal CTV comprising of levels 1–461 and nodal outlining was not routine practice across UK centres. The study helped to identify systematic regions of variance and informed the continued development of the RT planning and delivery guidelines in order to improve conformity. These studies demonstrate the importance of a robust RT QA process, particularly when introducing new techniques, either in clinical studies or in routine care.