Assessing the risk of malignant transformation of intraductal papillary mucinous neoplasms (IPMN) is crucial for determining the optimal therapeutic approach, whether conservative or surgical. Early diagnosis and surveillance of features suggestive of malignancy rely mainly on imaging, particularly endoscopic ultrasound (EUS) and pancreatic magnetic resonance imaging (MRI).

The aim of this study was to evaluate the concordance between pancreatic MRI and EUS in detecting Worrisome Features and High-Risk Stigmata in patients with IPMN.

This study reports results from a prospective, single-center, descriptive and analytical cohort conducted over a 36-month period starting in May 2024.Adult patients (≥18 years) who underwent both pancreatic MRI and EUS and who presented imaging features consistent with IPMN were included.

Exclusion criteria were: contraindications to either EUS or MRI, or pancreatic lesions not compatible with IPMN.

Radiological Worrisome Features were defined as the presence of at least one of the following: main pancreatic duct (MPD) diameter between 5–9 mm, MPD caliber change with parenchymal atrophy, lymphadenopathy, non-enhancing mural nodule, focal wall thickening.

High-Risk Stigmata were defined as: MPD diameter >10 mm, presence of an enhancing solid component.

Statistical analysis was performed using JAMOVI software.

A total of 17 patients were included: 58.8% male and 41.2% female, with a mean age of 70.8 ± 7.96 years.Among them, 52.9% were diabetic and 23.5% had undergone cholecystectomy. None had a history of acute pancreatitis.

Indications for performing MRI and EUS were: cholestatic jaundice (58.8%), abdominal pain (23.5%), and incidental discovery (17.6%).

Lesion location: 35.3% in the pancreatic head on MRI versus 29.4% on EUS, 23.5% in the tail on MRI versus 11.8% on EUS, 11.8% in the body on MRI versus 5.9% on EUS.Multifocal lesions were found in 2 patients on MRI and in 7 patients on EUS.

When Worrisome Features were absent on MRI, they were also absent on EUS in only 77.8% of cases, with no statistically significant difference (p = 0.057).When High-Risk Stigmata were absent on MRI, they were present on EUS in 27.3% of cases, with a statistically significant difference (p = 0.009).

Based on these results, EUS appears to be more sensitive than pancreatic MRI for detecting High-Risk Stigmata in IPMN.However, diagnostic performance depends greatly on operator expertise, which supports the complementary use of both modalities—at least during the initial evaluation of IPMN and during follow-up, especially when Worrisome Features are present on either imaging technique.