Advanced Radiotherapy Technologies Network In The UK (ART-NET) - A focus on lung cancer

Sean Brown1,2, Marcel van Herk1, Robert Chuter2, Sally Falk2, Karen Kirkby1, Ranald Mackay1,2, Kevin Harrington3, Vivian Cosgrove4, Alistair Gray5, Emma Hall3, Maria Hawkins6, Dave Hawkes7,8, Ann Henry9, Tim Maughan5, Chris Nutting10, Uwe Oelfke3, Gary Royle7,8, David Sebag-Montefiore9, Ricky Sharma7,8, Frank Van den Heuvel5, Corinne Faivre-Finn1,2

University of Manchester1, The Christie NHS Foundation Trust2, Institute of Cancer Research3, Leeds Teaching Hospitals NHS Trust4, University of Oxford5, Cancer Research UK/MRC Oxford Institute for Radiation Oncology6, University College London Hospitals NHS Foundation Trust7, University College London8, University of Leeds9, Royal Marsden NHS Foundation Trust10

Background: • Lung cancer is the most frequently diagnosed cancer worldwide. • Survival remains poor for those with inoperable disease (median 28

months). • Whilst the role of curative-intent radiotherapy (RT) is widely accepted,

the thorax remains a challenging anatomical site to irradiate. • New advanced radiotherapy technologies; stereotactic ablative

radiotherapy (SABR), magnetic resonance image-guided radiotherapy (MR-Linac) and proton-beam therapy (PBT) have potential to yield substantial clinical benefits.

• However there are significant challenges for development, assessment and rational implementation within the NHS.

• ART-NET is a CRUK funded initiative comprising 7 workstreams across 5 UK centres (The Institute of Cancer Research/Royal Marsden Hospital (ICR), Leeds, Manchester Cancer Research Centre (MCRC), Oxford and University College London (UCL).

ART-NET Goals: • Accelerate clinical translation of SABR, MR-Linac and PBT around

the theme of hypofractionated RT to provide an evidence based rationale across a range of tumour types.

• Develop/disseminate expertise in medical physics to lead development of planning/image guided solutions for these technologies.

• Implement methodologies for reliable treatment plan optimization on MR-images, acquired directly prior to or during treatment, enabling the clinical benefit of MR-guided SABR and MR-Linac to be fully exploited.

• Co-ordinate methodological developments in design and conduct of clinical trials to streamline assessments of new health technologies in specific tumours.

Conclusion: • A key aim of ART-NET is to generate national

treatment protocols and disseminate them across the UK, improving and harmonising practice.2

• Additionally, involvement of medical physicists, research radiographers and clinical research fellows in ART-NET will develop a skilled workforce that will be a huge asset for future radiotherapy research in the UK.

• Ultimately this will improve patient access to state-of-the-art radiotherapy and increase recruitment into clinical trials.

WORKSTREAMS

Traffic light protocol: • This is a project being undertaken within the

motion management workstream 3 and is based on the traffic light system developed at the NKI

(see fig 1).1

• The aim is to develop a protocol to facilitate online radiographer review of cone beam computerised tomography (CBCTs) in patients undergoing lung SABR to assess organ motion and anatomical changes.

• This will assess the impact of these changes on PTV coverage and advise when to seek physics/clinical review.

• This will inform the need for plan adaptation nationally.

2 Fast/adaptive re-planning

• Provide QA guidance for MR and image registration software to enable the use of MR in SABR.

• Standardise algorithms that generate synthetic CT images from treatment-

integrated MR images for all tumour sites and applied imaging protocols.

• Implement image processing tools to allow verification of the geometrical accuracy of the acquired MR images.

• Share tools to generate images suitable for dose calculations from online images (CBCT, MR).

• Implement fast segmentation tools across the network.

• Disseminate ultra-fast Monte-Carlo dose algorithms for photon and proton beams.

• Multi-site implementation of a high-performance inverse planning tool exploiting modern computational CPU and GPU hardware architectures.

• Image library-based (3-D, 4-D) quantification of organ motion and anatomical changes in tumour and adjacent critical structures.

• Evaluation of novel 4-D acquisition, reconstruction, and motion modelling

approaches and determine their suitability for clinical use. • Quantification of treatment platform-independent uncertainties (e.g.

delineation variability and short-term organ motion) and their effect on predictions of tumour/normal organs motion.

• Evaluation of the accuracy of image-guidance methodologies for each treatment platform.

6 Health economics

• Development of MRI-based early predictors of radiotherapy response. • Design of one prospective validation study to demonstrate that predictors

of response have utility in a multi-site and multi-vendor trial.

• Recommendations for development pathways for new RT technologies into practice-changing clinical trials.

• Assess feasibility/delivery of PBT and MR-Linac clinical trials including assessment of logistics and equipoise.

• Development of core outcome set; treatment, dosimetry, safety, efficacy,

effectiveness, including electronic platforms for patient-reported outcomes. • Support for clinical trial grant applications in core TSGs.

• Protocol for correction of range differences at all PBT centres.

• In silico validation of protocol to assess proton range

using motion management system.

• Evaluate proton range verification techniques and identify margins for complex treatments.

• Development of health economic methodology. • Comparative assessment of SABR, PBT and MR-Linac for

prostate cancer. • Comparative assessment of SABR, PBT and MR-Linac for second

site (e.g. lung, oesophagus).

Figure 1 Examples of the Traffic Light Protocol developed by Kwint et al. 1

Left: planning CT. Right: cone beam CT Purple: planning target volume. Red: gross tumour volume. First row: level red change; T4N2M0 NSCLC; CBCT of week 3, the atelectasis resolved and this resulted in a tumour shift outside PTV. Second row: level orange change; T4N2M0 NSCLC; week 1 CBCT showed atelectasis without a tumour shift. Third row: level yellow change; T4N2M0 NSCLC; CBCT shows regression in week 3 of the treatment. Fourth row: level green; CBCT shows no change at fraction 15.

References: 1. Kwint M, Conijn S, Schaake E, et al. Intra thoracic anatomical changes in lung cancer patients during the course of radiotherapy. Radiother Oncol. 2014;113(3):392-397.doi:10.1016/j.radonc.2014.10.009.

2. Harrington K, Hall E, Hawkins M, et al. Introducing the Cancer Research UK Advanced Radiotherapy Technologies Network (ART-NET). Clin Oncol. 2017. doi:10.1016/j.clon.2017.07.016.