Risk assessment of multi-focal side-branch IPMNs: Long-term follow-up of one thousand patients.

BACKGROUND

While the natural history of solitary side-branch intraductal papillary mucinous neoplasms (SB-IPMNs) is increasingly well-described, the prevalence, disease course, and oncologic risk of multifocal disease remain poorly defined. It is unknown whether the presence of multiple cysts confers additional oncologic risk compared to solitary lesions. Our objective was to evaluate the natural history and oncologic risk associated with multifocal SB-IPMNs compared with solitary SB-IPMNs during long-term surveillance.

METHODS

Patients with side-branch intraductal papillary mucinous neoplasms (SB-IPMNs) undergoing non-operative surveillance with ≥2 cross-sectional imaging studies ≥6 months apart were included. The number of cysts at diagnosis and at most recent surveillance was recorded by an abdominal radiologist who was blinded to pathologic outcomes. The primary endpoint was high-risk pathology (high-grade dysplasia or invasive carcinoma). Secondary endpoints included clinically relevant (CR) progression, defined as development of guideline-based worrisome features or high-risk stigmata excluding cyst size and growth, development of new cysts, and need for surgical resection.

RESULTS

Of 1,000 patients, 370 (37%) had multifocal disease. Median follow-up was 6.6 years (IQR 3.0-10.3). Rates of CR-progression (developing worrisome features/high-risk stigmata) were similar between multifocal and solitary SB-IPMN patients (9.7% vs 9.4%, p = 0.849). High-risk pathology occurred in 43 patients (4.3%): 3.2% multifocal vs 4.8% solitary (p = 0.248). On multivariable analysis, multifocality was not associated with high-risk pathology, while CR-progression strongly predicted malignant potential (HR 31.3, 95% CI 16.3-62.1). Findings were unchanged in sensitivity analyses restricted to patients with MRI and using guideline-based progression criteria including cyst size and growth.

CONCLUSION

Multiplicity of SB-IPMNs does not confer increased malignant risk at the patient level, despite nearly 40% of patients presenting with more than one cyst. These findings support a pancreas-wide susceptibility process (described as a "field defect") rather than a focal cyst-driven phenomenon. Accordingly, each cyst should be evaluated independently according to its own risk features.