Journal of Medical Imaging and Radiation Oncology 59 (2015) 363–370
RADIATION O N C O LO GY —O R I G I N A L A RTICLE
bs_bs_banner
Trans Tasman Radiation Oncology Group: Development of the Assessment of New Radiation Oncology Technology and Treatments (ANROTAT) Framework Gillian M Duchesne,1,2,3 Mel Grand,4 Tomas Kron,1,3,5 Annette Haworth,1,2 June Corry,1,2 Michael Jackson,6 Michael Ng,7 Deidre Besuijen,4 Hannah E Carter,8 Andrew Martin,8 Deborah Schofield,8,9 Val Gebski,8 Joan Torony,4 Olga Kovacev,4 Rowena Amin10 and Bryan Burmeister11 1 2 3 4 5 6 7 8 9 10 11
Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia University of Melbourne, Melbourne, Victoria, Australia Monash University, Melbourne, Victoria, Australia Trans Tasman Radiation Oncology Group, Newcastle, New South Wales, Australia RMIT University, Melbourne, Victoria, Australia University of New South Wales, Sydney, New South Wales, Australia Radiation Oncology Victoria, Melbourne, Victoria, Australia NHMRC Clinical Trials Centre, University of Sydney, Sydney, New South Wales, Australia School of Public Health, University of Sydney, Sydney, New South Wales, Australia Health and Training Institute, Sydney, New South Wales, Australia Princess Alexandra Hospital, University of Queensland, Brisbane, Queensland, Australia
GM Duchesne MD; M Grand MHlthSc (MRS); T Kron PhD; A Haworth PhD; J Corry MD; M Jackson FRANZCR; M Ng FRANZCR; D Besuijen DPH; HE Carter BEcon; A Martin PhD; D Schofield PhD; V Gebski PhD; J Torony Grad Dip Clin Trials Management; O Kovacev BSc; R Amin M Mgt; B Burmeister MD. Correspondence Professor Gillian M Duchesne, Peter MacCallum Cancer Centre, Locked Bag 1, A’Beckett Street, Melbourne, Vic. 8006, Australia. Email: [email protected] Conflict of interest: None identified. Submitted 1 May 2014; accepted 16 September 2014. doi:10.1111/1754-9485.12255
Abstract Introduction: The study aim was to develop a generic framework to derive the parameters to populate health-economic models for the rapid evaluation of new techniques and technologies in radiation oncology. Methods: A draft framework was developed through horizon scanning for relevant technologies, literature review to identify framework models, and a workshop program with radiation oncology professionals, biostatisticians, health economists and consumers to establish the Framework’s structure. It was tested using four clinical protocols, comparing intensity modulated with 3D conformal therapy (post-prostatectomy, anal canal and nasopharynx) and image-guided radiation therapy techniques with off-line review of portal imaging (in the intact prostate). Results: The draft generic research framework consisted of five sequential stages, each with a number of components, and was assessed as to its suitability for deriving the evidence needed to populate the decision-analytic models required for the health-economic evaluations. A final Framework was established from this experience for use by future researchers to provide evidence of clinical efficacy and cost-utility for other novel techniques. The four clinical treatment sites tested during the project were considered suitable to use in future evaluations. Conclusions: Development of a generic research framework to predict early and long-term clinical outcomes, combined with health-economic data, produced a generally applicable method for the rapid evaluation of new techniques and technologies in radiation oncology. Its application to further health technology assessments in the radiation oncology sector will allow further refinement and support its generalisability. Key words: framework; health-economic evaluation; health technology assessment; radiation oncology techniques; radiation oncology technologies.
© 2014 The Royal Australian and New Zealand College of Radiologists
363
GM Duchesne et al.
Introduction Radiation oncology is a specialty in which technologies evolve rapidly but incrementally, with innovations that increasingly spare normal-tissue dose and reduce morbidity, as well as having the potential to increase tumour dose and hence local tumour control. Traditionally, practitioners in Australia have had to adopt new techniques before their public financial reimbursement has been approved by Medicare, because of the long lead times required to show benefit in terms of tumour control and reduction in morbidity. This gap between the desire for wide adoption of new technologies and approval for public funding can delay the uptake of the technology or leave patients or departments with large uncovered costs after introduction. Although new technologies may require, for example, resources that are not accounted for in the prevailing reimbursement models, in the long term they may result in lower disease and complication burdens and reduced overall costs. However, the financial support of new approaches that might significantly improve long-term health and health-economic outcomes may be delayed for many years until evidence of their benefit becomes available. The need for more rapid technological evaluation is well recognised and was the stimulus for the Australian Government Department of Health and Ageing (DoHA) to contract with the Trans Tasman Radiation Oncology Group (TROG) in 2009 to develop an Assessment of New Radiation Oncology Technology and Treatments (ANROTAT) Framework in a prospective and evidence-based project. Its development complements other Australian Government initiatives arising from the Review of Health Technology Assessment in Australia published in 2009.1 The gold standard for providing evidence of clinical outcomes is the randomised controlled trial, but there are certain limitations with this approach in this context. Firstly, mature clinical trial results for new technologies may take many years to produce. Secondly, very few of the clinical trials designed 10 or more years ago that are now providing clinical outcome evidence included parameters for health-economic evaluation, which is essential for effective health technology assessment (HTA) and required by government for public funding. Additionally, because developments in the sector are evolutionary rather than revolutionary, with clinical improvements arising step-wise one after another, it is rare to find randomised trials that test the possible benefits. Hence, a framework for evaluating the clinical measures of tumour control and toxicity together with costeffectiveness must include the necessary tools and steps to support timely and robust comparisons of new interventions and incumbent practice. This framework publication describes the development and evaluation of the ANROTAT Framework, together with its final structure and its potential future use in the sector.
364
Methods Personnel Members of the three core radiation oncology professions, namely radiation oncologists, radiation therapists and radiation oncology medical physicists, working throughout the TROG network of radiation oncology centres across Australia were invited to participate in the project. Twenty radiation oncology centres from all states, with a mix of metropolitan and regional centres and public and private practice, contributed clinical data. Health economics and biostatistical expertise was contracted from the National Health and Medical Research Council Clinical Trials Centre based at the University of Sydney. Consumers were included in the expert working groups and clinical site subgroups and in the various workshops that supported the Framework’s development. Representatives from industry and other expert groups were consulted as required. A full list of the contributors is presented in Appendix C of the Final Report.2
Framework development Development of the Framework was an iterative process. A literature review was undertaken to identify generic evaluation frameworks that might be adapted to suit the purpose for radiation oncology, and the findings were collated at an initial workshop. The resulting draft framework structure was then discussed and refined at a second workshop including representatives of all groups described in ‘Personnel’ above, and prepared for testing. Part of the contractual agreement with DoHA was to test the draft framework by preparing the evidence required for submission to the Government’s HTA process through the Medical Services Advisory Committee (MSAC) for two techniques that have entered into common practice around the world over the last decade, namely intensity-modulated radiation therapy (IMRT) and online image-guided radiation therapy (IGRT), with their comparators being three-dimensional conformal radiation therapy (3DCRT) and non-IGRT (offline) verification methods, respectively. Four clinical research protocols were developed to incorporate the clinical outcome and cost parameters required to support a health-economic evaluation of the new techniques. For IMRT and its comparator, these were treatment of the nasopharynx, anal canal and prostate bed after prostatectomy, and for IGRT and its comparator, it was the intact prostate. These sites were selected to represent a wide range of different treatment requirements and outcomes, enabling both intra-and inter-technique comparisons so that the resulting framework could then be adapted to the evaluation of any future techniques and tumour sites. All protocols received human research ethics committee approval at each centre. Clinical data were recorded and reported
© 2014 The Royal Australian and New Zealand College of Radiologists
The TROG ANROTAT framework
on case report forms as per standard clinical trial processes. The reports of these trial protocols are available through the TROG website (http://trog.com.au/ ANROTAT), together with the other reports from the project. Prospective clinical data provided by participating sites included dose–volume histogram (DVH) parameters (derived from radiation therapy planning studies), acute toxicity and quality of life (QoL), and resource usage evidence for patients undergoing treatment. Posttreatment data for QoL and patient costs at a number of time points were also captured. Costings in staff time, capital and resource usage were included. These prospective data were combined with evidence from other sources, including systematic literature review and expert opinion, to derive estimates of longer-term outcomes and to test and support the function of the Framework. A rigorous quality assurance program was applied to all aspects of the project through centre credentialing, data collection and data management processes, and review of radiation therapy planning and protocol compliance, according to the Good Clinical Practice guidelines for conduct of clinical research (http:// www.australianclinicaltrials.gov.au/researchers).
Development of Framework components After review of available framework literature, the initial planning workshop confirmed five requirements or stages for an effective framework structure, including preparation, assessment methodology, project management, analysis of the evidence, and ongoing framework evaluation, review and modification. Each stage comprised a series of steps and components leading to a defined outcome. For example, during the preparation stage, a number of resources and guidelines were listed to assist in the successful derivation of a clear project question, obviously of paramount importance. As the project proceeded, the supporting components were reviewed and modified as needed. The identification of robust and widely applicable predictive indicators through testing of the Framework was also a key component of the work. TROG also developed and evaluated the feasibility of running a longitudinal register (the ANROTAT Radiation Oncology Register Pilot, or ARORP; report also available on the website) of radiotherapy patient treatments, dosimetry and outcomes as a source of long-term prospective data to both feed into future applications of the Framework and use to validate the assumptions made in the current study.
relevant report.2 It defines each of the five stages that require consideration when developing and submitting an application to support the use of and reimbursement for a new technology or technique in the Australian health care system. Not all components in each stage might be required for a particular investigation, and the Framework is designed to be sufficiently flexible to allow future researchers to modify, select and include what is relevant to their particular purpose. The Framework is supported by a user guide3 containing templates, information and instructions on how to operate and populate the Framework template. Each stage and component included in Table 1 is fully discussed in the guide.
Using the Framework to support health technology assessment Predictors of clinical outcomes A set of predictive factors was developed from each of the four clinical protocols to be used as surrogate measures for modelling the long-term outcomes of clinical efficacy and treatment-related morbidity. As illustrated in Table 2, they were derived from the prospective DVH parameters of the interventional plans and those of their comparators. We selected for evaluation those indicators that could be derived at the time of treatment and were expected to have the greatest impact on quality of life (and longevity) up to a decade or more in the future. They reflected, firstly, the probability of the achievable tumour dose and dose distribution producing tumour control (tumour control probability, TCP) and cure, and, secondly, the early and late toxicity and the probability of complications from therapy as determined by quantitative analysis of normal-tissue effects in the clinic (QUANTEC) methodology,4 which may be used to estimate normal-tissue complication probability (NTCP). Different indicators were considered according to each clinical protocol, and those discriminating between the intervention and comparator were identified for use in future projects. The indicators cover a range of normal tissues and tumours in order to be readily applicable to the evaluation of future changes in technology and techniques, as future treatment evolutions are considered likely to be based on improving dose distributions and increasing the differential between tumour and normal-tissue doses. Their values are designed to be slotted into the decisionanalytic models in the Framework template to derive comparisons of costs and utilities between baseline technologies or techniques and new approaches.
Results Decision-analytic modelling
Framework structure The framework structure developed through this study is illustrated in Table 1 and may also be found in the
© 2014 The Royal Australian and New Zealand College of Radiologists
The development of decision-analytic models comparing the costs and clinical components of the treatment approaches and producing cost–utility endpoints to
365
GM Duchesne et al.
Table 1. The framework Stage 1. Preparation
Step Define question: horizon scan, professional expertise
Define technology Define evidence sources
2. Methodology
Clinical requirements
Economic analysis
Data protocols
3. Project management
Timelines Resources Process
Governance Communication Activation
Data management
4. Analysis
Modelling, statistics Sensitivity analysis
Interpretation
5. Evaluation
Review framework structure and function Follow-up
Components
Outcome
Patient population Intervention Comparator Outcome(s) of interest Setting Perspective Include processes Systematic review Prospective data Retrospective data Expert opinion Register (longitudinal data) Disease sites Disease endpoints Toxicity end points Quality of life Dosimetric/dose–volume histogram parameters Costings Utilities Decision-analytic model structure Statistics Prospective and/or retrospective data collection Other protocols including quality assurance Use of register data Early discussion/agreement Staff, hardware, software, budget, other Collaborations Define responsibilities Communication Site selection/recruitment Ethics Credentialling Collection (IT infrastructure) Centralisation Quality assurance Values of parameters Effects of varying assumptions
Context, perspectives Applicability Justification Report Does it need adjusting? ‘Phase 4’ – real-life application Value of register
Clarity of question to be answered Scope, goal posts and limitations
Agreed parameters, indicators What data are required?
Robust and transferable clinical indicators Rapid (surrogate) indicators
Valid and transferable model
Population of generic template
Agreed and realistic
Efficient project conduct Effective communication and buy-in Range of centres Technical validity Data validity
Informs cost-utility equation Clinical efficacy Technical efficacy Societal efficacy Demonstration of benefit or otherwise; relative societal value in the health-care context
Ongoing usefulness Confirmation of utility Ongoing informing of health technology assessments
The five stages are detailed in the user guide to the framework.
support decision-making for the allocation of the scarce health dollar, was a critical requirement of this process. Evidence was gathered from a number of sources to populate the decision-analytic models for each clinical protocol. The data sources included systematic reviews of the literature to identify any randomised trial data (Level II) or Level III or IV evidence5 available for cancer
366
control, early and late morbidity, QoL and health state utility, as well as expert opinion and health-care costs. A series of component studies within the ANROTAT project conducted to provide additional data for modelling included a dosimetric evaluation, a resource usage evaluation, and a Quality of Life (QoL) and toxicity evaluation. The construction, analysis and outcomes of the
© 2014 The Royal Australian and New Zealand College of Radiologists
The TROG ANROTAT framework
Table 2. Discriminatory predictive clinical indicators derived from the clinical protocols Protocol Intensity-modulated radiotherapy Post-prostatectomy 64 Gy 70 Gy
Anal canal carcinoma
Nasopharyngeal cancer
Image-guided radiation therapy Intact prostate
Indicator
Clinical outcome
Rectal V50, V60† Femoral head dose Rectal V50, V60† Rectal normal-tissue complication probability Femoral head dose Uncomplicated tumour control Bladder V35, V40, V50† Small bowel dose Dose to skin/external genitalia Femoral head dose Gross and planning target volume D95‡ Dose parameters for mandible, parotid, cochlea, and temporal lobe§ Rectal D60
Risk of acute and late rectal toxicity, such as bleeding Risk of femoral fracture Risk of acute and late rectal toxicity, such as bleeding Risk of femoral fracture Therapeutic ratio (cure vs toxicity) Acute and late bladder toxicity risk Small bowel toxicity risk Indirect indicator of risk of treatment break and decreased tumour control probability Risk of femoral fracture Tumour control probability Acute and late toxicity risks
Risk of late rectal toxicity
†The volume of the target organ receiving 35% (V35), 40%, 50% or 60% of the prescribed dose. ‡The dose in Gy covering 95% of the target or organ volume. §Only the organs with discriminatory dose parameter differences are included.
Markov modelling for one of the four clinical protocols, ‘A decision model to estimate the cost-effectiveness of intensity modulated radiation therapy (IMRT) compared to three dimensional conformal radiation therapy (3DCRT) in patients receiving radiotherapy to the prostate bed’, have been published online by Radiotherapy and Oncology.6 The article illustrates how the Framework stages were followed to produce the diverse pieces of evidence successfully, supporting the health-economic assessment of a radiation oncology technique. Review of the results across the four clinical protocols demonstrated consistency from one to another, supporting the robustness of both the process of the Framework and the indicators chosen as predictors. Use of four different clinical scenarios allowed a judgment as to the general applicability of the findings to other cancer sites. The Framework was revisited to ensure that the process, sequence and content were optimised for future application.
Discussion The use of evidence from a number of sources for the clinical outcomes, combined with the cost-analytic modelling in each clinical scenario, resulted in the derivation of the measures of comparative cost-effectiveness required to support application for reimbursement. These outcomes demonstrated that the Framework functioned effectively to support the health-economic evaluation of two common radiation oncology techniques. The work has demonstrated that it is feasible to use a generic framework for the rapid health-economic evaluation of radiation oncology innovations. The issue of how to
© 2014 The Royal Australian and New Zealand College of Radiologists
incorporate advances in complex and high-cost technology into routine clinical practice through demonstration of clinical benefit is highly topical, with discussions and debate occurring world-wide and not just in the Australian health-care environment.7 In April 2012, the European Society for Therapeutic Radiology and Oncology launched the HERO (Health Economics in Radiation Oncology) project8 with the aim of developing a knowledge base and a model for health-economic evaluation of radiation treatments. In the United States, the tight fiscal environment has led to specific targeting of medical interventions perceived to be costly, and radiation therapy has not been immune.9 Others in North America and Europe recognise that while the gold standard for collecting clinical evidence remains the randomised clinical trial, particularly for innovations in techniques such as altered fractionation schedules, other more timely and diverse methodologies are required to ensure rapid evaluation of the appropriateness of new technologies to demonstrate their comparative value.10,11 Health-economic evaluations using a framework such as ours will provide a rapid means of deciding which new technologies should be implemented and where; in the case of results that are not clear and conclusive, introduction should be delayed until more detailed or robust data become available. In addition, decision-modelling techniques may assist in informing the size and design of a randomised trial if one is still thought necessary. The value of having an established framework and models in facilitating future evaluations cannot be overstated. The requirements needed to make the Framework functional also informed the selection of the register data
367
GM Duchesne et al.
components, and details of the register’s piloting will be presented separately. Concise datasets reflective of dayto-day practice in every department will be very powerful in documenting trends in practice and the resulting patient outcomes on a population basis. They can be used to validate modelling assumptions and to provide a ready source of comparative evidence for implementation or otherwise of new approaches. We regard the ongoing development and maintenance of the register as a key component to support use of the Framework. One of the key aims of the project was to identify a means of supporting early health-economic evaluation of new approaches without having to wait many years for classical clinical trial results. As such, we sought to derive a number of factors predictive of the long-term outcomes of cancer control, morbidity and QoL. We elected to examine the dosimetric parameters of the treatment plans generated for the test interventions and the standard comparators, on the basis that technical developments in radiation oncology would be aimed at improving target coverage while limiting normal dose. In addition, dosimetric parameters also have the potential of being collected automatically from radiotherapy treatment planning and information systems. The quality assurance processes supporting this research ensured that all centres were credentialed before allowing patient recruitment and ensured compliance with protocol contouring, planning and dosimetry requirements to allow confidence in the identification of the parameters. It is recognised that a wealth of other factors, such as the use of combined modality therapy or individual predisposition, might alter the outcomes in individual patients, but what the Framework is designed to do is to compare potential outcomes of a standard therapy with an innovation, with these other influences remaining constant. We also recognise that these predictive indicators do not rigorously fulfil the Prentice criteria of surrogacy,12 which require the following four conditions to be satisfied: (i) that the treatment process determines the true endpoint; (ii) that the treatment process determines the surrogate endpoint; (iii) that the surrogate is prognostic for the true endpoint; and (iv) that the surrogate mediates the effect of treatment on the true endpoint. However, it is known that achieving all four criteria is difficult, and the factors identified relate to important clinical outcomes. The identification of similar predictive factors across the four protocols also suggests that these measures are likely to be applicable in other scenarios in the future in predicting the long-term outcomes of clinical advances. It is notable that the factors predictive of acute toxicity are those thought to be predictive of late morbidity outcomes, as reported, for example, by Heemsbergen et al.13 Although it is generally considered that the two outcomes are not directly linked because of different underlying pathological mechanisms, this finding may add support to their choice as predictive factors. One of the chief limitations of the project has been that we did
368
not have time to obtain prospective long-term follow-up data for all the patients enrolled on the clinical protocols (with the exception of a cohort of patients on the nasopharyngeal protocol), deriving such evidence instead from a number of sources, including systematic literature reviews, in order to populate the decisionanalytic models. However, this parallels the very nature of the project scope, which was to identify a means of early evaluation of innovative approaches without longterm follow up. A key step within the Framework is the ongoing review of its structure and function when it is used in other evaluations, also enabling it to be amended as required to improve its utility. Use of register-derived evidence will also determine the validity of the chosen predictive factors and will support modification of the Framework and models if required. There are also some limitations to be recognised in the use of NTCP parameters (and, to a lesser extent, TCP) as predictive indicators.14,15 Although the QUANTEC system4 is a significant step in the development of a comprehensive understanding of the relationships between dosimetric parameters and normal-tissue outcomes, the modelling is only as good as the data used to inform it. Key issues were the lack of uniformity in the earlier published literature regarding grading of side effects, incomplete reporting of results and the use of incompatible or ambiguous endpoints. Nonetheless, TCP and NTCP are internationally recognised parameters and provide a useful foundation for these analyses. The international community also recognises the value of radiation oncology registers in providing prospective and longitudinal evidence of health and economic outcomes that can support the modelled results obtained from use of the Framework, particularly where data are collected from multiple centres or a population base, representing real-life practice. Open-source-based infrastructure to facilitate data sharing between multiple centres is becoming available.16 A similar initiative is underway in North America under the auspices of the American Society for Therapeutic Radiology and Oncology to develop a National Radiation Oncology Registry17 with the goals of providing real-time comparative effectiveness data and circumventing the financial perils of technological introduction without thorough assessment.18 Another approach to gathering evidence sees the interim funding of new medical technologies for treating patients on clinical trials to generate the additional benefit and safety information needed to make an informed general-coverage decision. This approach is active in the United States, where Centers for Medicare and Medicaid and other funders treat patients under the Coverage with Evidence Development scheme. Although the ANROTAT Framework has its limitations, it has significant benefits in terms of providing timely data for decisions about public funding of new technologies, which are currently widely adopted without the required clinical evidence to endorse funding. It was successfully
© 2014 The Royal Australian and New Zealand College of Radiologists
The TROG ANROTAT framework
used to collate the necessary evidence for IMRT and IGRT for consideration by the Australian Government MSAC and could be adapted for similar health-care systems in other countries. Review of its use in future applications evaluating other innovative approaches is required to complete validation of the methodology. The use of this approach should seriously be considered as an alternative or adjunct to the randomised clinical trial pathway of obtaining the relevant evidence because of its rapidity and relatively low cost.
Conclusion Timely evaluation of the relative costs and benefits of introducing new approaches in radiation oncology may be facilitated by the use of the generic framework developed through the ANROTAT project. Use of early predictors of the late outcomes that are the ultimate determinants of cancer care success should permit appropriate reimbursement and support patient access to beneficial technologies and techniques.
Acknowledgements We acknowledge the participation of the Framework Workshops attendees: Eric Bone, Jeremy Booth, Celia Burrell, Haylea Cleaver, Peter Greer, Sally Hodgkinson, Robert Lin, John Stubbs, Chris Wratten, Phil Vial, Amanda Schell, Abel MacDonald, Scott Babington, Elizabeth Brown, Adrian Gibbs, Charles Lin, Sheryn Campbell, Sarat Chander, Alison Cray, Farshad Foroudi, Colin Hornby, Aldo Rolfo, Brindha Subramanian, Bronwyn Hilder, Craig Everitt, Maree Wood, Val Gebski, Merel Kimman, Andrew Martin, Deborah Schofield, Hannah Verry, Adrienne Kirby, Wal Crellin, Justin Dixon, Martin Ebert, Jim Frantzis, Chris Harper, Brendan Healy, Rebecca Montgomery, Clare Poprawski, Mark Sidhom, Natalia Vukolova and all staff members at the TROG Central Office, as well as all contributors to data collection in the TROG radiation oncology network. The Assessment of New Radiation Oncology Technology and Treatments project was funded by the Australian Government Department of Health and Ageing (DoHA). DoHA had oversight of the progress of the project but did not influence design, data handling or analysis and reporting.
References 1. Commonwealth of Australia. Review of health technology assessment in Australia. Canberra: Australian Government Department of Health, 2009. 2. Trans Tasman Radiation Oncology Group. The Assessment of New Radiation Oncology Technologies and Treatments (ANROTAT) Project. Final report. Newcastle, NSW: Trans Tasman Radiation Oncology Group, 2012. [Cited 1 Jul 2014.] Available from URL: http://www.TROG.com.au/ANROTAT
© 2014 The Royal Australian and New Zealand College of Radiologists
3. Trans Tasman Radiation Oncology Group. The Assessment of New Radiation Oncology Technologies and Treatments (ANROTAT). User guide to the Framework. Newcastle, NSW: Trans Tasman Radiation Oncology Group, 2012. 4. Marks LB, Yorke ED, Jackson A, Ten Haken RK, Constine LS, Eisbruch A. Use of normal tissue complication probability models in the clinic. Int J Radiat Oncol Biol Phys 2010; 76 (3 Suppl. ): S10–19. 5. Commonwealth of Australia. A guide to the development, implementation and evaluation of clinical practice guidelines. Canberra: Australian Government Department of Health, 1999. 6. Carter HE, Martin A, Schofield D, Duchesne G, Haworth A, Hornby C, Sidhom M, Jackson M. A decision model to estimate the cost-effectiveness of intensity modulated radiation therapy (IMRT) compared to three dimensional conformal radiation therapy (3DCRT) in patients receiving radiotherapy to the prostate bed. Radiother Oncol 2014. doi: 10.1016/j.radonc.2014.03.020. S0167-8140(14) 00163-7 [pii]. 7. Fraass BA, Moran JM. Quality, technology and outcomes: evolution and evaluation of new treatments and/or new technology. Semin Radiat Oncol 2012; 22: 3–10. 8. Lievens Y, Grau C. Health economics in radiation oncology: introducing the ESTRO HERO project. Radiother Oncol 2012; 103: 109–12. 9. Furlow BS. US Medicare to cut radiotherapy times and payments. Lancet Oncol 2012; 13: e334. 10. Zietman A, Ibbott G. A clinical approach to technology assessment: how do we and how should we choose the right treatment? Semin Radiat Oncol 2012; 22: 11–17. 11. van Loon J, Grutters J, Macbeth F. Evaluation of novel radiotherapy technologies: what evidence is needed to assess their clinical and cost effectiveness, and how should we get it? Lancet Oncol 2012; 13: e169–77. 12. Prentice RL. Surrogate endpoints in clinical trials: definition and operational criteria. Stat Med 1989; 8: 431–40. 13. Heemsbergen WD, Peeters ST, Koper PC, Hoogeman MS, Lebesque JB. Acute and late gastro-intestinal toxicity after radiotherapy in prostate cancer patients: consequential late damage. Int J Radiat Oncol Biol Phys 2006; 66: 3–10. 14. Jackson A, Marks LB, Bentzen SM et al. The lessons of QUANTEC: recommendations for reporting and gathering data on dose-volume dependencies of treatment outcome. Int J Radiat Oncol Biol Phys 2010; 76 (3 Suppl. ): S155–60. 15. Bentzen SM, Constine LS, Deasy JO et al. Quantitative analyses of normal tissue effects in the clinic (QUANTEC): an introduction to the scientific issues. Int J Radiat Oncol Biol Phys 2010; 76 (3 Suppl. ): S3–9.
369
GM Duchesne et al.
16. Roelofs E, Dekker A, Meldolesi E, van Stiphout RG, Valentini V, Lambin P. International data-sharing for radiotherapy research: an open-source based infrastructure for multicentric clinical data mining. Radiother Oncol 2014; 110: 370–4. 17. Palta JR, Efstathiou JA, Bekelman JE et al. Developing a national radiation oncology registry: from acorns to oaks. Pract Radiat Oncol 2012; 2: 10–17.
370
18. Nguyen PL, Gu X, Lipsitz SR et al. Cost implications of the rapid adoption of newer technologies for treating prostate cancer. J Clin Oncol 2011; 29: 1517–24.
© 2014 The Royal Australian and New Zealand College of Radiologists
RADIATION O N C O LO GY —O R I G I N A L A RTICLE
bs_bs_banner
Trans Tasman Radiation Oncology Group: Development of the Assessment of New Radiation Oncology Technology and Treatments (ANROTAT) Framework Gillian M Duchesne,1,2,3 Mel Grand,4 Tomas Kron,1,3,5 Annette Haworth,1,2 June Corry,1,2 Michael Jackson,6 Michael Ng,7 Deidre Besuijen,4 Hannah E Carter,8 Andrew Martin,8 Deborah Schofield,8,9 Val Gebski,8 Joan Torony,4 Olga Kovacev,4 Rowena Amin10 and Bryan Burmeister11 1 2 3 4 5 6 7 8 9 10 11
Peter MacCallum Cancer Centre, Melbourne, Victoria, Australia University of Melbourne, Melbourne, Victoria, Australia Monash University, Melbourne, Victoria, Australia Trans Tasman Radiation Oncology Group, Newcastle, New South Wales, Australia RMIT University, Melbourne, Victoria, Australia University of New South Wales, Sydney, New South Wales, Australia Radiation Oncology Victoria, Melbourne, Victoria, Australia NHMRC Clinical Trials Centre, University of Sydney, Sydney, New South Wales, Australia School of Public Health, University of Sydney, Sydney, New South Wales, Australia Health and Training Institute, Sydney, New South Wales, Australia Princess Alexandra Hospital, University of Queensland, Brisbane, Queensland, Australia
GM Duchesne MD; M Grand MHlthSc (MRS); T Kron PhD; A Haworth PhD; J Corry MD; M Jackson FRANZCR; M Ng FRANZCR; D Besuijen DPH; HE Carter BEcon; A Martin PhD; D Schofield PhD; V Gebski PhD; J Torony Grad Dip Clin Trials Management; O Kovacev BSc; R Amin M Mgt; B Burmeister MD. Correspondence Professor Gillian M Duchesne, Peter MacCallum Cancer Centre, Locked Bag 1, A’Beckett Street, Melbourne, Vic. 8006, Australia. Email: [email protected] Conflict of interest: None identified. Submitted 1 May 2014; accepted 16 September 2014. doi:10.1111/1754-9485.12255
Abstract Introduction: The study aim was to develop a generic framework to derive the parameters to populate health-economic models for the rapid evaluation of new techniques and technologies in radiation oncology. Methods: A draft framework was developed through horizon scanning for relevant technologies, literature review to identify framework models, and a workshop program with radiation oncology professionals, biostatisticians, health economists and consumers to establish the Framework’s structure. It was tested using four clinical protocols, comparing intensity modulated with 3D conformal therapy (post-prostatectomy, anal canal and nasopharynx) and image-guided radiation therapy techniques with off-line review of portal imaging (in the intact prostate). Results: The draft generic research framework consisted of five sequential stages, each with a number of components, and was assessed as to its suitability for deriving the evidence needed to populate the decision-analytic models required for the health-economic evaluations. A final Framework was established from this experience for use by future researchers to provide evidence of clinical efficacy and cost-utility for other novel techniques. The four clinical treatment sites tested during the project were considered suitable to use in future evaluations. Conclusions: Development of a generic research framework to predict early and long-term clinical outcomes, combined with health-economic data, produced a generally applicable method for the rapid evaluation of new techniques and technologies in radiation oncology. Its application to further health technology assessments in the radiation oncology sector will allow further refinement and support its generalisability. Key words: framework; health-economic evaluation; health technology assessment; radiation oncology techniques; radiation oncology technologies.
© 2014 The Royal Australian and New Zealand College of Radiologists
363
GM Duchesne et al.
Introduction Radiation oncology is a specialty in which technologies evolve rapidly but incrementally, with innovations that increasingly spare normal-tissue dose and reduce morbidity, as well as having the potential to increase tumour dose and hence local tumour control. Traditionally, practitioners in Australia have had to adopt new techniques before their public financial reimbursement has been approved by Medicare, because of the long lead times required to show benefit in terms of tumour control and reduction in morbidity. This gap between the desire for wide adoption of new technologies and approval for public funding can delay the uptake of the technology or leave patients or departments with large uncovered costs after introduction. Although new technologies may require, for example, resources that are not accounted for in the prevailing reimbursement models, in the long term they may result in lower disease and complication burdens and reduced overall costs. However, the financial support of new approaches that might significantly improve long-term health and health-economic outcomes may be delayed for many years until evidence of their benefit becomes available. The need for more rapid technological evaluation is well recognised and was the stimulus for the Australian Government Department of Health and Ageing (DoHA) to contract with the Trans Tasman Radiation Oncology Group (TROG) in 2009 to develop an Assessment of New Radiation Oncology Technology and Treatments (ANROTAT) Framework in a prospective and evidence-based project. Its development complements other Australian Government initiatives arising from the Review of Health Technology Assessment in Australia published in 2009.1 The gold standard for providing evidence of clinical outcomes is the randomised controlled trial, but there are certain limitations with this approach in this context. Firstly, mature clinical trial results for new technologies may take many years to produce. Secondly, very few of the clinical trials designed 10 or more years ago that are now providing clinical outcome evidence included parameters for health-economic evaluation, which is essential for effective health technology assessment (HTA) and required by government for public funding. Additionally, because developments in the sector are evolutionary rather than revolutionary, with clinical improvements arising step-wise one after another, it is rare to find randomised trials that test the possible benefits. Hence, a framework for evaluating the clinical measures of tumour control and toxicity together with costeffectiveness must include the necessary tools and steps to support timely and robust comparisons of new interventions and incumbent practice. This framework publication describes the development and evaluation of the ANROTAT Framework, together with its final structure and its potential future use in the sector.
364
Methods Personnel Members of the three core radiation oncology professions, namely radiation oncologists, radiation therapists and radiation oncology medical physicists, working throughout the TROG network of radiation oncology centres across Australia were invited to participate in the project. Twenty radiation oncology centres from all states, with a mix of metropolitan and regional centres and public and private practice, contributed clinical data. Health economics and biostatistical expertise was contracted from the National Health and Medical Research Council Clinical Trials Centre based at the University of Sydney. Consumers were included in the expert working groups and clinical site subgroups and in the various workshops that supported the Framework’s development. Representatives from industry and other expert groups were consulted as required. A full list of the contributors is presented in Appendix C of the Final Report.2
Framework development Development of the Framework was an iterative process. A literature review was undertaken to identify generic evaluation frameworks that might be adapted to suit the purpose for radiation oncology, and the findings were collated at an initial workshop. The resulting draft framework structure was then discussed and refined at a second workshop including representatives of all groups described in ‘Personnel’ above, and prepared for testing. Part of the contractual agreement with DoHA was to test the draft framework by preparing the evidence required for submission to the Government’s HTA process through the Medical Services Advisory Committee (MSAC) for two techniques that have entered into common practice around the world over the last decade, namely intensity-modulated radiation therapy (IMRT) and online image-guided radiation therapy (IGRT), with their comparators being three-dimensional conformal radiation therapy (3DCRT) and non-IGRT (offline) verification methods, respectively. Four clinical research protocols were developed to incorporate the clinical outcome and cost parameters required to support a health-economic evaluation of the new techniques. For IMRT and its comparator, these were treatment of the nasopharynx, anal canal and prostate bed after prostatectomy, and for IGRT and its comparator, it was the intact prostate. These sites were selected to represent a wide range of different treatment requirements and outcomes, enabling both intra-and inter-technique comparisons so that the resulting framework could then be adapted to the evaluation of any future techniques and tumour sites. All protocols received human research ethics committee approval at each centre. Clinical data were recorded and reported
© 2014 The Royal Australian and New Zealand College of Radiologists
The TROG ANROTAT framework
on case report forms as per standard clinical trial processes. The reports of these trial protocols are available through the TROG website (http://trog.com.au/ ANROTAT), together with the other reports from the project. Prospective clinical data provided by participating sites included dose–volume histogram (DVH) parameters (derived from radiation therapy planning studies), acute toxicity and quality of life (QoL), and resource usage evidence for patients undergoing treatment. Posttreatment data for QoL and patient costs at a number of time points were also captured. Costings in staff time, capital and resource usage were included. These prospective data were combined with evidence from other sources, including systematic literature review and expert opinion, to derive estimates of longer-term outcomes and to test and support the function of the Framework. A rigorous quality assurance program was applied to all aspects of the project through centre credentialing, data collection and data management processes, and review of radiation therapy planning and protocol compliance, according to the Good Clinical Practice guidelines for conduct of clinical research (http:// www.australianclinicaltrials.gov.au/researchers).
Development of Framework components After review of available framework literature, the initial planning workshop confirmed five requirements or stages for an effective framework structure, including preparation, assessment methodology, project management, analysis of the evidence, and ongoing framework evaluation, review and modification. Each stage comprised a series of steps and components leading to a defined outcome. For example, during the preparation stage, a number of resources and guidelines were listed to assist in the successful derivation of a clear project question, obviously of paramount importance. As the project proceeded, the supporting components were reviewed and modified as needed. The identification of robust and widely applicable predictive indicators through testing of the Framework was also a key component of the work. TROG also developed and evaluated the feasibility of running a longitudinal register (the ANROTAT Radiation Oncology Register Pilot, or ARORP; report also available on the website) of radiotherapy patient treatments, dosimetry and outcomes as a source of long-term prospective data to both feed into future applications of the Framework and use to validate the assumptions made in the current study.
relevant report.2 It defines each of the five stages that require consideration when developing and submitting an application to support the use of and reimbursement for a new technology or technique in the Australian health care system. Not all components in each stage might be required for a particular investigation, and the Framework is designed to be sufficiently flexible to allow future researchers to modify, select and include what is relevant to their particular purpose. The Framework is supported by a user guide3 containing templates, information and instructions on how to operate and populate the Framework template. Each stage and component included in Table 1 is fully discussed in the guide.
Using the Framework to support health technology assessment Predictors of clinical outcomes A set of predictive factors was developed from each of the four clinical protocols to be used as surrogate measures for modelling the long-term outcomes of clinical efficacy and treatment-related morbidity. As illustrated in Table 2, they were derived from the prospective DVH parameters of the interventional plans and those of their comparators. We selected for evaluation those indicators that could be derived at the time of treatment and were expected to have the greatest impact on quality of life (and longevity) up to a decade or more in the future. They reflected, firstly, the probability of the achievable tumour dose and dose distribution producing tumour control (tumour control probability, TCP) and cure, and, secondly, the early and late toxicity and the probability of complications from therapy as determined by quantitative analysis of normal-tissue effects in the clinic (QUANTEC) methodology,4 which may be used to estimate normal-tissue complication probability (NTCP). Different indicators were considered according to each clinical protocol, and those discriminating between the intervention and comparator were identified for use in future projects. The indicators cover a range of normal tissues and tumours in order to be readily applicable to the evaluation of future changes in technology and techniques, as future treatment evolutions are considered likely to be based on improving dose distributions and increasing the differential between tumour and normal-tissue doses. Their values are designed to be slotted into the decisionanalytic models in the Framework template to derive comparisons of costs and utilities between baseline technologies or techniques and new approaches.
Results Decision-analytic modelling
Framework structure The framework structure developed through this study is illustrated in Table 1 and may also be found in the
© 2014 The Royal Australian and New Zealand College of Radiologists
The development of decision-analytic models comparing the costs and clinical components of the treatment approaches and producing cost–utility endpoints to
365
GM Duchesne et al.
Table 1. The framework Stage 1. Preparation
Step Define question: horizon scan, professional expertise
Define technology Define evidence sources
2. Methodology
Clinical requirements
Economic analysis
Data protocols
3. Project management
Timelines Resources Process
Governance Communication Activation
Data management
4. Analysis
Modelling, statistics Sensitivity analysis
Interpretation
5. Evaluation
Review framework structure and function Follow-up
Components
Outcome
Patient population Intervention Comparator Outcome(s) of interest Setting Perspective Include processes Systematic review Prospective data Retrospective data Expert opinion Register (longitudinal data) Disease sites Disease endpoints Toxicity end points Quality of life Dosimetric/dose–volume histogram parameters Costings Utilities Decision-analytic model structure Statistics Prospective and/or retrospective data collection Other protocols including quality assurance Use of register data Early discussion/agreement Staff, hardware, software, budget, other Collaborations Define responsibilities Communication Site selection/recruitment Ethics Credentialling Collection (IT infrastructure) Centralisation Quality assurance Values of parameters Effects of varying assumptions
Context, perspectives Applicability Justification Report Does it need adjusting? ‘Phase 4’ – real-life application Value of register
Clarity of question to be answered Scope, goal posts and limitations
Agreed parameters, indicators What data are required?
Robust and transferable clinical indicators Rapid (surrogate) indicators
Valid and transferable model
Population of generic template
Agreed and realistic
Efficient project conduct Effective communication and buy-in Range of centres Technical validity Data validity
Informs cost-utility equation Clinical efficacy Technical efficacy Societal efficacy Demonstration of benefit or otherwise; relative societal value in the health-care context
Ongoing usefulness Confirmation of utility Ongoing informing of health technology assessments
The five stages are detailed in the user guide to the framework.
support decision-making for the allocation of the scarce health dollar, was a critical requirement of this process. Evidence was gathered from a number of sources to populate the decision-analytic models for each clinical protocol. The data sources included systematic reviews of the literature to identify any randomised trial data (Level II) or Level III or IV evidence5 available for cancer
366
control, early and late morbidity, QoL and health state utility, as well as expert opinion and health-care costs. A series of component studies within the ANROTAT project conducted to provide additional data for modelling included a dosimetric evaluation, a resource usage evaluation, and a Quality of Life (QoL) and toxicity evaluation. The construction, analysis and outcomes of the
© 2014 The Royal Australian and New Zealand College of Radiologists
The TROG ANROTAT framework
Table 2. Discriminatory predictive clinical indicators derived from the clinical protocols Protocol Intensity-modulated radiotherapy Post-prostatectomy 64 Gy 70 Gy
Anal canal carcinoma
Nasopharyngeal cancer
Image-guided radiation therapy Intact prostate
Indicator
Clinical outcome
Rectal V50, V60† Femoral head dose Rectal V50, V60† Rectal normal-tissue complication probability Femoral head dose Uncomplicated tumour control Bladder V35, V40, V50† Small bowel dose Dose to skin/external genitalia Femoral head dose Gross and planning target volume D95‡ Dose parameters for mandible, parotid, cochlea, and temporal lobe§ Rectal D60
Risk of acute and late rectal toxicity, such as bleeding Risk of femoral fracture Risk of acute and late rectal toxicity, such as bleeding Risk of femoral fracture Therapeutic ratio (cure vs toxicity) Acute and late bladder toxicity risk Small bowel toxicity risk Indirect indicator of risk of treatment break and decreased tumour control probability Risk of femoral fracture Tumour control probability Acute and late toxicity risks
Risk of late rectal toxicity
†The volume of the target organ receiving 35% (V35), 40%, 50% or 60% of the prescribed dose. ‡The dose in Gy covering 95% of the target or organ volume. §Only the organs with discriminatory dose parameter differences are included.
Markov modelling for one of the four clinical protocols, ‘A decision model to estimate the cost-effectiveness of intensity modulated radiation therapy (IMRT) compared to three dimensional conformal radiation therapy (3DCRT) in patients receiving radiotherapy to the prostate bed’, have been published online by Radiotherapy and Oncology.6 The article illustrates how the Framework stages were followed to produce the diverse pieces of evidence successfully, supporting the health-economic assessment of a radiation oncology technique. Review of the results across the four clinical protocols demonstrated consistency from one to another, supporting the robustness of both the process of the Framework and the indicators chosen as predictors. Use of four different clinical scenarios allowed a judgment as to the general applicability of the findings to other cancer sites. The Framework was revisited to ensure that the process, sequence and content were optimised for future application.
Discussion The use of evidence from a number of sources for the clinical outcomes, combined with the cost-analytic modelling in each clinical scenario, resulted in the derivation of the measures of comparative cost-effectiveness required to support application for reimbursement. These outcomes demonstrated that the Framework functioned effectively to support the health-economic evaluation of two common radiation oncology techniques. The work has demonstrated that it is feasible to use a generic framework for the rapid health-economic evaluation of radiation oncology innovations. The issue of how to
© 2014 The Royal Australian and New Zealand College of Radiologists
incorporate advances in complex and high-cost technology into routine clinical practice through demonstration of clinical benefit is highly topical, with discussions and debate occurring world-wide and not just in the Australian health-care environment.7 In April 2012, the European Society for Therapeutic Radiology and Oncology launched the HERO (Health Economics in Radiation Oncology) project8 with the aim of developing a knowledge base and a model for health-economic evaluation of radiation treatments. In the United States, the tight fiscal environment has led to specific targeting of medical interventions perceived to be costly, and radiation therapy has not been immune.9 Others in North America and Europe recognise that while the gold standard for collecting clinical evidence remains the randomised clinical trial, particularly for innovations in techniques such as altered fractionation schedules, other more timely and diverse methodologies are required to ensure rapid evaluation of the appropriateness of new technologies to demonstrate their comparative value.10,11 Health-economic evaluations using a framework such as ours will provide a rapid means of deciding which new technologies should be implemented and where; in the case of results that are not clear and conclusive, introduction should be delayed until more detailed or robust data become available. In addition, decision-modelling techniques may assist in informing the size and design of a randomised trial if one is still thought necessary. The value of having an established framework and models in facilitating future evaluations cannot be overstated. The requirements needed to make the Framework functional also informed the selection of the register data
367
GM Duchesne et al.
components, and details of the register’s piloting will be presented separately. Concise datasets reflective of dayto-day practice in every department will be very powerful in documenting trends in practice and the resulting patient outcomes on a population basis. They can be used to validate modelling assumptions and to provide a ready source of comparative evidence for implementation or otherwise of new approaches. We regard the ongoing development and maintenance of the register as a key component to support use of the Framework. One of the key aims of the project was to identify a means of supporting early health-economic evaluation of new approaches without having to wait many years for classical clinical trial results. As such, we sought to derive a number of factors predictive of the long-term outcomes of cancer control, morbidity and QoL. We elected to examine the dosimetric parameters of the treatment plans generated for the test interventions and the standard comparators, on the basis that technical developments in radiation oncology would be aimed at improving target coverage while limiting normal dose. In addition, dosimetric parameters also have the potential of being collected automatically from radiotherapy treatment planning and information systems. The quality assurance processes supporting this research ensured that all centres were credentialed before allowing patient recruitment and ensured compliance with protocol contouring, planning and dosimetry requirements to allow confidence in the identification of the parameters. It is recognised that a wealth of other factors, such as the use of combined modality therapy or individual predisposition, might alter the outcomes in individual patients, but what the Framework is designed to do is to compare potential outcomes of a standard therapy with an innovation, with these other influences remaining constant. We also recognise that these predictive indicators do not rigorously fulfil the Prentice criteria of surrogacy,12 which require the following four conditions to be satisfied: (i) that the treatment process determines the true endpoint; (ii) that the treatment process determines the surrogate endpoint; (iii) that the surrogate is prognostic for the true endpoint; and (iv) that the surrogate mediates the effect of treatment on the true endpoint. However, it is known that achieving all four criteria is difficult, and the factors identified relate to important clinical outcomes. The identification of similar predictive factors across the four protocols also suggests that these measures are likely to be applicable in other scenarios in the future in predicting the long-term outcomes of clinical advances. It is notable that the factors predictive of acute toxicity are those thought to be predictive of late morbidity outcomes, as reported, for example, by Heemsbergen et al.13 Although it is generally considered that the two outcomes are not directly linked because of different underlying pathological mechanisms, this finding may add support to their choice as predictive factors. One of the chief limitations of the project has been that we did
368
not have time to obtain prospective long-term follow-up data for all the patients enrolled on the clinical protocols (with the exception of a cohort of patients on the nasopharyngeal protocol), deriving such evidence instead from a number of sources, including systematic literature reviews, in order to populate the decisionanalytic models. However, this parallels the very nature of the project scope, which was to identify a means of early evaluation of innovative approaches without longterm follow up. A key step within the Framework is the ongoing review of its structure and function when it is used in other evaluations, also enabling it to be amended as required to improve its utility. Use of register-derived evidence will also determine the validity of the chosen predictive factors and will support modification of the Framework and models if required. There are also some limitations to be recognised in the use of NTCP parameters (and, to a lesser extent, TCP) as predictive indicators.14,15 Although the QUANTEC system4 is a significant step in the development of a comprehensive understanding of the relationships between dosimetric parameters and normal-tissue outcomes, the modelling is only as good as the data used to inform it. Key issues were the lack of uniformity in the earlier published literature regarding grading of side effects, incomplete reporting of results and the use of incompatible or ambiguous endpoints. Nonetheless, TCP and NTCP are internationally recognised parameters and provide a useful foundation for these analyses. The international community also recognises the value of radiation oncology registers in providing prospective and longitudinal evidence of health and economic outcomes that can support the modelled results obtained from use of the Framework, particularly where data are collected from multiple centres or a population base, representing real-life practice. Open-source-based infrastructure to facilitate data sharing between multiple centres is becoming available.16 A similar initiative is underway in North America under the auspices of the American Society for Therapeutic Radiology and Oncology to develop a National Radiation Oncology Registry17 with the goals of providing real-time comparative effectiveness data and circumventing the financial perils of technological introduction without thorough assessment.18 Another approach to gathering evidence sees the interim funding of new medical technologies for treating patients on clinical trials to generate the additional benefit and safety information needed to make an informed general-coverage decision. This approach is active in the United States, where Centers for Medicare and Medicaid and other funders treat patients under the Coverage with Evidence Development scheme. Although the ANROTAT Framework has its limitations, it has significant benefits in terms of providing timely data for decisions about public funding of new technologies, which are currently widely adopted without the required clinical evidence to endorse funding. It was successfully
© 2014 The Royal Australian and New Zealand College of Radiologists
The TROG ANROTAT framework
used to collate the necessary evidence for IMRT and IGRT for consideration by the Australian Government MSAC and could be adapted for similar health-care systems in other countries. Review of its use in future applications evaluating other innovative approaches is required to complete validation of the methodology. The use of this approach should seriously be considered as an alternative or adjunct to the randomised clinical trial pathway of obtaining the relevant evidence because of its rapidity and relatively low cost.
Conclusion Timely evaluation of the relative costs and benefits of introducing new approaches in radiation oncology may be facilitated by the use of the generic framework developed through the ANROTAT project. Use of early predictors of the late outcomes that are the ultimate determinants of cancer care success should permit appropriate reimbursement and support patient access to beneficial technologies and techniques.
Acknowledgements We acknowledge the participation of the Framework Workshops attendees: Eric Bone, Jeremy Booth, Celia Burrell, Haylea Cleaver, Peter Greer, Sally Hodgkinson, Robert Lin, John Stubbs, Chris Wratten, Phil Vial, Amanda Schell, Abel MacDonald, Scott Babington, Elizabeth Brown, Adrian Gibbs, Charles Lin, Sheryn Campbell, Sarat Chander, Alison Cray, Farshad Foroudi, Colin Hornby, Aldo Rolfo, Brindha Subramanian, Bronwyn Hilder, Craig Everitt, Maree Wood, Val Gebski, Merel Kimman, Andrew Martin, Deborah Schofield, Hannah Verry, Adrienne Kirby, Wal Crellin, Justin Dixon, Martin Ebert, Jim Frantzis, Chris Harper, Brendan Healy, Rebecca Montgomery, Clare Poprawski, Mark Sidhom, Natalia Vukolova and all staff members at the TROG Central Office, as well as all contributors to data collection in the TROG radiation oncology network. The Assessment of New Radiation Oncology Technology and Treatments project was funded by the Australian Government Department of Health and Ageing (DoHA). DoHA had oversight of the progress of the project but did not influence design, data handling or analysis and reporting.
References 1. Commonwealth of Australia. Review of health technology assessment in Australia. Canberra: Australian Government Department of Health, 2009. 2. Trans Tasman Radiation Oncology Group. The Assessment of New Radiation Oncology Technologies and Treatments (ANROTAT) Project. Final report. Newcastle, NSW: Trans Tasman Radiation Oncology Group, 2012. [Cited 1 Jul 2014.] Available from URL: http://www.TROG.com.au/ANROTAT
© 2014 The Royal Australian and New Zealand College of Radiologists
3. Trans Tasman Radiation Oncology Group. The Assessment of New Radiation Oncology Technologies and Treatments (ANROTAT). User guide to the Framework. Newcastle, NSW: Trans Tasman Radiation Oncology Group, 2012. 4. Marks LB, Yorke ED, Jackson A, Ten Haken RK, Constine LS, Eisbruch A. Use of normal tissue complication probability models in the clinic. Int J Radiat Oncol Biol Phys 2010; 76 (3 Suppl. ): S10–19. 5. Commonwealth of Australia. A guide to the development, implementation and evaluation of clinical practice guidelines. Canberra: Australian Government Department of Health, 1999. 6. Carter HE, Martin A, Schofield D, Duchesne G, Haworth A, Hornby C, Sidhom M, Jackson M. A decision model to estimate the cost-effectiveness of intensity modulated radiation therapy (IMRT) compared to three dimensional conformal radiation therapy (3DCRT) in patients receiving radiotherapy to the prostate bed. Radiother Oncol 2014. doi: 10.1016/j.radonc.2014.03.020. S0167-8140(14) 00163-7 [pii]. 7. Fraass BA, Moran JM. Quality, technology and outcomes: evolution and evaluation of new treatments and/or new technology. Semin Radiat Oncol 2012; 22: 3–10. 8. Lievens Y, Grau C. Health economics in radiation oncology: introducing the ESTRO HERO project. Radiother Oncol 2012; 103: 109–12. 9. Furlow BS. US Medicare to cut radiotherapy times and payments. Lancet Oncol 2012; 13: e334. 10. Zietman A, Ibbott G. A clinical approach to technology assessment: how do we and how should we choose the right treatment? Semin Radiat Oncol 2012; 22: 11–17. 11. van Loon J, Grutters J, Macbeth F. Evaluation of novel radiotherapy technologies: what evidence is needed to assess their clinical and cost effectiveness, and how should we get it? Lancet Oncol 2012; 13: e169–77. 12. Prentice RL. Surrogate endpoints in clinical trials: definition and operational criteria. Stat Med 1989; 8: 431–40. 13. Heemsbergen WD, Peeters ST, Koper PC, Hoogeman MS, Lebesque JB. Acute and late gastro-intestinal toxicity after radiotherapy in prostate cancer patients: consequential late damage. Int J Radiat Oncol Biol Phys 2006; 66: 3–10. 14. Jackson A, Marks LB, Bentzen SM et al. The lessons of QUANTEC: recommendations for reporting and gathering data on dose-volume dependencies of treatment outcome. Int J Radiat Oncol Biol Phys 2010; 76 (3 Suppl. ): S155–60. 15. Bentzen SM, Constine LS, Deasy JO et al. Quantitative analyses of normal tissue effects in the clinic (QUANTEC): an introduction to the scientific issues. Int J Radiat Oncol Biol Phys 2010; 76 (3 Suppl. ): S3–9.
369
GM Duchesne et al.
16. Roelofs E, Dekker A, Meldolesi E, van Stiphout RG, Valentini V, Lambin P. International data-sharing for radiotherapy research: an open-source based infrastructure for multicentric clinical data mining. Radiother Oncol 2014; 110: 370–4. 17. Palta JR, Efstathiou JA, Bekelman JE et al. Developing a national radiation oncology registry: from acorns to oaks. Pract Radiat Oncol 2012; 2: 10–17.
370
18. Nguyen PL, Gu X, Lipsitz SR et al. Cost implications of the rapid adoption of newer technologies for treating prostate cancer. J Clin Oncol 2011; 29: 1517–24.
© 2014 The Royal Australian and New Zealand College of Radiologists
