Cost-effectiveness analysis of neoadjuvant v surgery-first for resectable pancreatic cancer

Bradley, A. and Van Der Meer, R. and McKay, C. (2019) Cost-effectiveness analysis of neoadjuvant v surgery-first for resectable pancreatic cancer. In: BMJ International Forum on Quality and Safety in Healthcare, 2019-03-27 - 2019-03-29, Scottish Exhibition & Conference Centre.

[thumbnail of Bradley-etal-BMJ-IFQSH-2019-Cost-effectiveness-analysis-of-neoadjuvant-v-surgery-first-for-resectable]
Preview
 Text. Filename: Bradley_etal_BMJ_IFQSH_2019_Cost_effectiveness_analysis_of_neoadjuvant_v_surgery_first_for_resectable.pdf
Accepted Author Manuscript
Download (6MB)| Preview

Abstract

1) Conflicts of interest: None. 2) Context: The work was performed at the University of Strathclyde Business School, Management Science Department. It focused on patients with resectable pancreatic cancer at time of presentation. Borderline and locally advanced cases at presentation were not included. 3) Problem: Neoadjuvant therapy for resectable pancreatic cancer is controversial with current guidelines supporting resection followed by adjuvant therapy. However, up to 50% of patients fail to receive adjuvant therapy due to post-operative complications, early disease reoccurrence or decline in function. Benefits of neoadjuvant therapy include elimination of micrometastases, increased likelihood of obtaining multimodal treatment and R0 resection and identifying patients with more aggressive tumours in whom expensive and risky surgery would ultimately be futile. The aim of the study was therefore to assess whether neoadjuvant or upfront surgery was more cost-effective for the management of resectable pancreatic cancer. 4) Assessment of problem and analysis of its causes: Systematic reviews and meta-analysis revealed a lack of studies offering head-to-head comparison between both treatment pathways with most existing studies including borderline and locally advanced cases in the neoadjuvant arm therefore not offering a true like-for-like comparison. These findings were disseminated to the pancreatic cancer multi-disciplinary-team at conferences and research meetings. 5) Intervention: To improve the delivery of care to pancreatic cancer patients with resectable disease the UK’s first cost-effectiveness analysis of neoadjuvant therapy versus upfront surgery for resectable only cases of pancreatic cancer was undertaken using a Markov model. 6) Strategy for change: First a pooled proportion meta-analysis using Freeman-Tukey arcsine square root transformation under random effects model to account for heterogeneity synthesised existing evidence from randomized controlled trials and phase II/III trials from which to populate the Markov model. The structure of the Markov model was agreed with the West of Scotland Pancreatic Unit multi-disciplinary-team. Output from the Markov model for decision analysis was presented to this team and, following feedback sessions, results authenticated against their expert opinion. The agreed model structure was then used to perform cost-effectiveness analysis. 7) Measurement of improvement: Effectiveness was measured as quality-adjusted-life-months (QALMs). Each Markov cycle was 1month with a total follow-up time of 60 cycles or until death. Willingness-to-pay was set at £30,000 per QALY with discount for cost and benefit set at 3.5% as per NICE guidelines. Model uncertainties were tested through one and two-way deterministic and probabilistic Monte Carlo sensitivity analysis. Neoadjuvant therapy was found to be the most cost-effective option with 21.27 QALMs at a cost-effectiveness ratio of £4370.73 compared to 17.59 QALMs at a cost-effectiveness ratio of £5582.85 and incremental cost-effectiveness ration of -£1421.33 for upfront surgery. 8) Effects of changes: The benefits of this work are that money could be saved and reinvested into further improving pancreatic cancer services. This work can also be used to better inform shared decision-making with patients when choosing between treatment options. However there remains the possibility that patients with earliest resectable disease most likely to receive R0 resection and adjuvant therapy may still benefit from upfront surgery approach. 9) Lessons learnt: Quality-of-life data for pancreatic cancer is limited and this should be the focus of further research as costs and benefits in cancer care are complex with quality as well as quantity of survival time having great importance. Although neoadjuvant pathway was more cost-effective, this depended on receiving multimodal treatment. This highlights the need moving towards personalised predictive medicine to support shared decision-making in research and practice. 10) Messages for others: Cost-effectiveness analysis adds an important dimension to the debate about competing treatment options for resectable pancreatic cancer. Costs and benefits In cancer treatment are multifaceted and complex requiring greater patient and carer input in future research which should coincide with a move towards personalised predictive medicine in research to support shared clinical decision-making. 11) Involvement of patients, carers or family members in the project: The next phase of this project will involve patients and carers in assessing quality-of-life impact of interventions and impact of statistical models in supporting shared decision-making from the patient’s perspective. 12) Ethics Approval: Not applicable

CORE (COnnecting REpositories)