COSTS ASSOCIATED WITH PANCREATIC CYST SURVEILLANCE DECREASED AFTER RISK-STRATIFICATION WITH NEXT-GENERATION SEQUENCING

Introduction:
Incidental detection of pancreatic cysts (PCs) is increasing and identifying high-risk cysts is of critical importance. Prior studies suggest significant cost associated with PC surveillance, which is more pronounced with over-management. The specificity needed for surveillance to be cost-effective is lower than current guideline-directed strategies. Next-generation sequencing (NGS) has been shown to reliably detect PCs at risk of progression to advanced neoplasia. We aim to describe cost savings associated with incorporating NGS into the management of PCs.

Methods:
This single-center, retrospective cohort study included adult patients evaluated in a pancreatic cyst clinic between 2016 and 2021 and was approved by the institutional review board. Exclusion criteria were lack of NGS results, less than 12 months of pre-NGS follow-up, loss to follow-up after NGS, and surgical cyst resection. Demographic and cyst characteristics were abstracted for each patient. Costs of surveillance (including endoscopic ultrasound, and cross-sectional imaging) were estimated using the Medicare fee schedule for 2019. Cost estimates were calculated and compared before and after NGS. Statistical significance was defined as p<0.05.

Results:
A total of 95 patients were considered after applying exclusion criteria (Figure 1). Mean age was 71.2 ± 11.2 years with the majority (52.6%) female. The most common race was White (62.1%) (Table 1). After NGS, 78 patients (82.1%) were felt to have mucinous cysts and 17 (17.9%) were felt to have serous cystadenomas, benign pancreatic duct dilation, or pseudocysts. There were no significant demographic differences between those with and without mucinous cysts. Non-mucinous cysts were larger (30.6 ± 18.5 mm vs 22.5 ± 10.6 mm; P=0.03) and patients had shorter follow-up time (1405 ± 643 days vs 1772 ± 551 days; P=0.02) (Table 1). The mean total annual cost of PC surveillance was $679 ± 205. The mean cost of surveillance was $827 ± 283 prior to undergoing NGS and decreased to $536 ± 241 after undergoing NGS (p<0.001). Patients deemed to have non-mucinous cysts had a greater cost reduction after NGS ($926 ± 256 to $515 ± 336, p<0.001) than those with mucinous cysts ($806 ± 286 to $541 ± 217, p<0.001) (Figure 1).

Conclusion:
Incorporating NGS results into PC management led to a 35.2% reduction in estimated average annual costs associated with surveillance. Patients with non-mucinous cysts had higher pre-NGS costs, likely due to larger size, but also experienced a greater cost reduction after NGS. This is likely driven by reduced need for surveillance among non-mucinous cysts. In addition to the previously demonstrated clinical utility of incorportating NGS into PC management, these findings highlight a potential financial benefit. Cost analyses using real-cost data are needed to confirm the financial benefit of NGS.