AGA Clinical Practice Update on Surveillance of Pancreatic Cystic Lesions and Hepatocellular Carcinoma in Older Adults: Expert Review.

Kyoto 2024 (IAP) is still the core algorithm for IPMN, and nothing in it has changed. MD-, mixed- and BD-IPMN all go through one HRS/WF algorithm. MDCT, MRI/MRCP and EUS are treated as equivalent for diagnosis, with tumor markers and HbA1c added. Since 2024, EUS imaging and EUS-FNA cytology count toward HRS/WF when EUS is done. High-risk stigmata are: obstructive jaundice with a head cyst; an enhancing mural nodule >=5 mm or a solid component; MPD >=10 mm; or suspicious/positive cytology. Worrisome features are: pancreatitis; CA19-9 >37 U/mL; new-onset or worsening diabetes within 1 year; cyst >=30 mm; enhancing mural nodule <5 mm; thickened or enhancing cyst wall; MPD 5-9 mm; abrupt change in duct caliber with distal atrophy; lymphadenopathy; or growth >=2.5 mm/year. HRS should prompt consideration of resection, weighed against fitness and life expectancy; it does not mean automatic surgery. WF should prompt EUS and multidisciplinary review. EUS-FNA is used only when its result would change management. Surveillance of non-resected BD-IPMN depends on size: <20 mm at 6 months, then every 18 months; 20 to <30 mm every 6 months twice, then every 12 months; >=30 mm every 6 months. Under Kyoto, surveillance may stop for cysts <20 mm with no change and no WF after 5 years, and should stop if the patient is unfit for surgery or life expectancy is <10 years. Continued surveillance remains an option because the risk of a separate pancreatic cancer keeps accumulating. NEW in 2026: an AGA Clinical Practice Update (Das, Serper, Vege; expert opinion, not formally graded) adds stopping rules by age and comorbidity for any pancreatic cystic lesion. (1) At any age, patients with significant comorbidity (age-adjusted Charlson index >=3-5) and an incidental, especially low-risk, cyst should be counseled against routine surveillance. (2) Over 65, cysts <=15 mm stable for 5 years with no WF/HRS should be considered for stopping after shared decision-making. 'Stable' means growth under 3 mm/year and under 6 mm over 2 years. (3) Over 75, cysts <30 mm stable for 5 years with no WF/HRS call for a discussion of whether surveillance is worth continuing, and it may stop if the patient accepts the small residual risk. (4) If comorbidity develops that rules out surgery, stopping should be discussed, even when WF/HRS were found earlier, provided there is no evidence of HGD or cancer. (5) After resection with no HGD or cancer, the remnant follows the same rules, counting surveillance time before surgery, and stopping is discussed after 5 years of stability. (6) After resection with HGD, MRI/MRCP every 6 months for 2 years and then yearly while the patient can still have a completion pancreatectomy. If invasive cancer was found, the remnant follows oncology guidelines. In both cases, stopping is discussed with oncology after 5 years of remnant stability with no WF/HRS. The resect-all-MCNs rule and the familial/germline (CAPS) screening pathway come from other guidelines, not from either of these documents.