A Cost Utility Model Supports Changes in Post-Treatment Surveillance Associated With the 2022 American College of Gastroenterology Guidelines

Introduction: Radiofrequency ablation is a safe and effective treatment for neoplastic Barrett’s esophagus (BE), but surveillance after endoscopic eradication has only been studied observationally without any studies of this disease’s natural history. Recent natural history modeling work has allowed qualified estimation of a natural history scenario. We sought to apply our multi-state model of post-ablation natural history to study the cost effectiveness of surveillance after endoscopic eradication of neoplastic BE. Methods: We constructed a microsimulation cost-utility model of endoscopic surveillance after complete eradication of intestinal metaplasia (CEIM) with intervals as recommended by the 2022 American College of Gastroenterology BE guideline. A cohort of participants were modeled after the United States (US) Radiofrequency Ablation cohort in terms of demographics, worst prior histologic grade, and pre-treatment BE segment length. Transition probabilities for the natural history were estimates from published multi-state models. The model was Markov generalized to allow differing rates of progression based on BE characteristics as covariates to the multi-state model and for other-cause mortality to depend on age. There were states for: no recurrence, recurrence with various histologic grades, ablative re-treatment, endoscopic mucosal resection, invasive adenocarcinoma, and death. We considered a willingness-to-pay threshold of 100,000 2017 US dollars ($) per quality-adjusted life year (QALY). Model parameters and their sources are described in the Table. Results: In the base case scenario of the model, surveillance and re-treatment decreased progression to invasive esophageal adenocarcinoma at ten years by 1.2% in LGD and 13.0% in HGD/IMC (Figure). Compared to the natural history scenario, the incremental cost effectiveness ratio for surveillance at ten years was $79,125 for LGD and $10,952 for HGD/IMC. Conclusion: In the base case scenario of this cost-utility model, newly recommended surveillance intervals were highly cost effective for HGD/IMC and approached the margins of cost-effectiveness for LGD in a microsimulation cost-utility model. This supports the new guideline recommendation to decrease the frequency of post-treatment surveillance of HGD/IMC and LGD. While development of the model requires probabilistic sensitivity analysis and calibration, the model has the potential to inform the health economics of clinical processes after CEIM. Table 1. - Microsimulation model parameters in the base case scenario and their sources. Model variable Base case value Reference/details Structural model assumptions Model cycle length 3 months Model time horizon 10 years Model starting year 2017 For other-cause mortality and inflation adjustment Other cause mortality Population 2020 Social Security actuarial cohort life Tables Time for resection 1 cycle Resection for all recurrent HGD or IMC Resection success rate 100% Simplifying assumption Time for ablation 1 cycle Repeated until successful Discounting of costs/utilities 3% Per year Probabilities per cycle from the literature Death from invasive adenocarcinoma 7.63% SEER 5-Year Relative Survival Rates 2012-2018* Complete eradication of intestinal metaplasia after recurrence 57.7% Guthikonda et al., The American Journal of Gastroenterology, 2017 Cancer progression from recurrence 0.1625% Guthikonda et al., The American Journal of Gastroenterology, 2017 Utilities per cycle from the literature Surveillance after CEIM 97% Boger et al., Alimentary Pharmacology and Therapeutics, 2010 Retreatment endoscopy 94% Boger et al., Alimentary Pharmacology and Therapeutics, 2010 Esophageal adenocarcinoma 96% Boger et al., Alimentary Pharmacology and Therapeutics, 2010 Costs per cycle from the literature Surveillance endoscopy $1,019 Inadomi et al., Gastroenterology, 2009† Cost of cancer care $13,532 Inadomi et al., Gastroenterology, 2009† Cost of ablation re-treatment $4,317 Inadomi et al., Gastroenterology, 2009† Assumed half cost of initial treatment Cost of resection re-treatment $934 Filby et al., Journal of Comparative Effectiveness Research, 2017 ** *Assumes cumulative incidence = 1 – e^(-incidence x time).†Assuming 17.7% inflation.**Assumes 133% exchange rate with the British Pound. Figure 1.: Model state occupancy by months after complete eradication of intestinal metaplasia.