MRCP surveillance of low-risk branch-duct IPMNs: longitudinal size changes and clinical outcomes in a real-world...

To evaluate cyst growth patterns, interval radiological changes, and clinical outcomes in patients with low-risk branch-duct intraductal papillary mucinous neoplasms (BD-IPMNs) without baseline worrisome features (WF) or high-risk stigmata (HRS), and to assess whether current surveillance strategies might be individualized in this subgroup.

In this retrospective single-center study, MRCP examinations performed between 2015 and 2025 were reviewed. Among 601 patients with BD-IPMN, 76 patients without baseline WF/HRS and with at least two MRCP examinations were included. The primary study endpoint was a composite of clinically actionable progression (new WF/HRS, including main pancreatic duct (MPD) dilatation 5-9 mm, or surgical intervention) or study-defined meaningful cyst growth (> 2 mm/year or ≥ 20%/year). Interobserver agreement was assessed using intraclass correlation coefficient (ICC). Descriptive analyses, sensitivity analyses, and exploratory association analyses were performed.

The cohort had a mean age of 65.6 ± 10.7 years, with 58% female patients. The median follow-up duration was 39.5 months (IQR 27.8-72.0). Although follow-up extended longer in a subset of patients, the cohort's median observation time was approximately 3.3 years., with a median of 3 MRCPs per patient. Median cyst size increased minimally from 10.0 mm to 11.0 mm, with a median annual growth of 0.11 mm/year. The composite endpoint occurred in 9 patients (11.8%), comprising 5 clinically actionable events (3 BD-IPMN-related WF and 2 MPD dilatation cases classified as WF) and 4 cases of meaningful cyst growth without other WF/HRS. No patient developed HRS or underwent surgical intervention. No patient developed remote PDAC during follow-up. Interobserver agreement was excellent (ICC ≥ 0.93).

Low-risk BD-IPMNs demonstrated minimal growth and infrequent clinically actionable progression during MRCP surveillance. These findings are consistent with an indolent course, particularly in patients with small cysts, but should be interpreted in the context of limited event counts, non-standardized contrast use, selection bias, and a median follow-up of approximately 3.3 years. Larger multicenter studies with standardized imaging and longer follow-up are needed before specific surveillance de-escalation strategies can be recommended.