• Decision Modeling to Guide Management of Pancreatic IPMNs: Immediate Surgery or Initial Surveillance?

    Greg D Sacks, Jonah Levine, Joseph R Habib, Madeleine Hunter, Ammar A Javed, Giovanni Marchegiani, Christopher L Wolfgang, R Scott Braithwaite 
    Ann Surg Oncol. 2026 Apr 21. doi: 10.1245/s10434-026-19667-1. Online ahead of print.

    Abstract

    Background: Most branch duct intraductal papillary mucinous neoplasms (BD-IPMNs) are indolent, but distinguishing those harboring high-grade dysplasia or invasive cancer remains difficult. This analysis focuses not on incidental small BD-IPMNs but on the subset whose cyst characteristics bring surgery into the decision-making discussion. Surgery prevents malignant progression but carries morbidity; surveillance avoids overtreatment but risks delayed cancer detection. Current guidelines rely on fixed thresholds that may not reflect individual variation. Our study compared immediate surgery and initial surveillance in patients with BD-IPMNs, using a decision-analytic model that incorporates patient-specific risk factors. 

     Methods: A Markov decision model compared immediate surgery with initial surveillance, incorporating age, comorbidities, and cyst location. Health states reflected progression from low-grade to high-grade dysplasia and invasive cancer, postoperative complications, recurrence, and quality-of-life decrements. Transition probabilities were derived from published studies and American College of Surgeons (ACS)-National Surgical Quality Improvement Program data. The primary outcome was quality-adjusted life-years (QALYs). 

     Results: For a 60-year-old patient with mild comorbidities and a pancreatic head BD-IPMN, immediate surgery provided 16.8 QALYs versus 16.3 with surveillance (incremental gain, 0.5 QALYs). Lifetime cancer probability was lower with surgery (24.5% vs 33.5%), as was cancer-related mortality (9.3% vs 20.3%), though surgery resulted in more resections for low-grade dysplasia (55.0% vs 15.3%). Age, baseline cancer probability, and perioperative mortality were the strongest determinants of the preferred strategy. 

     Conclusions: Among patients with BD-IPMNs being considered for surgery, immediate resection offers a modest benefit for younger, healthier individuals, whereas surveillance remains appropriate for older or comorbid patients. These findings support individualized, risk-based management rather than universal application of guideline thresholds.