Can radiomics push the limits of current IPMN malignancy assessment and help avoid unnecessary resection?

Excerpt

Pancreatic cysts are commonly encountered in clinical practice, with a prevalence exceeding 50% in modern MRI systems [ 1]. The increased detection of these cysts has led to a surge of interest in intraductal papillary mucinous neoplasms (IPMNs). Although the precise proportion of these cysts representing IPMNs remains unclear, they are nonetheless the most prevalent type among surgically resected pancreatic cysts [ 2]. It is well-established that IPMN is a precursor lesion for pancreatic malignancy, following a sequence from adenoma to carcinoma. Although IPMNs have historically been subjected to routine resection due to their potential for malignancy, current treatment strategies vary based on risk assessment, with resection reserved for selected cases because of surgical hazards. Esteemed societies, such as the International Association of Pancreatology (IAP), European Study Group on Cystic Tumors of the Pancreas, American College of Radiology (ACR), American Gastroenterology Association (AGA), and American College of Gastroenterology, have issued guidelines for IPMN management [ 2‐ 6]. Several studies have recurrently reported that enhancing nodules and main pancreatic ductal dilatation are strong indicators of malignancy [ 7, 8]. Consequently, these features appear consistently across the guidelines [ 2‐ 6]. However, it is important to recognize that the presence of an enhancing nodule indicating abundant tumor volume and ductal dilatation suggesting excessive mucin production do not necessarily imply an invasive potential to traverse the basement membrane of the pancreatic duct and subsequently infiltrate the stroma. As a result, cases of pathologically benign IPMN presenting with imaging findings suggestive of malignancy are sporadically encountered in clinical practice. This slightly reduced specificity of imaging poses a challenge in current IPMN practice. For IPMN cases lacking imaging findings suggestive of malignancy, the guidelines uniformly recommend vigilant follow-up. Surveillance duration varies across guidelines: the AGA and ACR generally propose 5–10 years of surveillance [ 4, 5], whereas others advocate lifetime surveillance [ 2, 3, 6], acknowledging the potential emergence of invasive IPMN after 5 or more years of vigilance [ 9]. …