Highlight
In a large single-center surveillance cohort, patients with intraductal papillary mucinous neoplasm (IPMN) had an approximately 10-fold higher risk of pancreatic ductal adenocarcinoma (PDAC) or high-grade dysplasia/malignancy than an age- and year-matched Surveillance, Epidemiology, and End Results (SEER) population comparator.
Risk was not uniform. Main pancreatic duct (MPD) dilation of 5 mm or greater and body mass index (BMI) of 30 or greater identified particularly high-risk groups, while patients younger than 60 years and those with cysts 1 cm or smaller had risks closest to population baseline.
Even so-called guideline-negative cysts, defined here as smaller than 3 cm and without MPD dilation or a solid component, still carried an elevated standardized incidence ratio (SIR) of 8.40, underscoring that “low-risk” does not mean “no-risk.”
The study supports moving beyond a one-size-fits-all surveillance model toward more personalized imaging intensity based on demographic and radiographic risk stratification.
Background
IPMNs are mucin-producing pancreatic cystic neoplasms recognized as precursor lesions to invasive pancreatic cancer. Their increasing detection is largely driven by widespread abdominal imaging and an aging population. Clinically, the challenge is familiar: some IPMNs remain indolent for years, whereas others progress to high-grade dysplasia or invasive carcinoma despite regular imaging surveillance.
This uncertainty has shaped current practice. Major guidelines generally recommend prolonged, and often lifelong, surveillance for fit patients with presumed branch-duct or mixed-type IPMN who do not meet immediate surgical criteria. Yet surveillance is burdensome, costly, anxiety-provoking, and imperfect. Repeated MRI, MRCP, CT, and endoscopic ultrasound can identify evolving morphologic risk features, but interval cancers still occur. Meanwhile, many patients never experience clinically meaningful progression.
A central unanswered question has been how much excess pancreatic cancer risk is truly attributable to the presence of IPMN, rather than simply reflecting the background risk of an older imaging-detected population. The present study addresses that gap by comparing observed malignancy events during IPMN surveillance with expected events derived from a matched SEER reference cohort.
Study Design
Choubey and colleagues performed a retrospective study using an institutional pancreas cyst database from a high-volume center. The database captured patient demographics, serial radiographic cyst characteristics, and pathologic outcomes during surveillance.
The cohort included 1,494 patients undergoing surveillance for IPMN. Across follow-up, these patients underwent 11,107 imaging studies and contributed 7,681 person-years at risk. Median surveillance duration was 55.3 months. The median age was 66 years, 64% of patients were female, and median index cyst size was 1.4 cm, indicating that many entered surveillance with relatively small lesions.
The principal analytical approach was calculation of the standardized incidence ratio. Observed malignancy events in the IPMN cohort were compared with expected PDAC events derived from an age- and calendar year-matched SEER population. This strategy estimates the degree to which malignancy risk in a surveillance cohort exceeds background population risk.
The reported malignancy events included 26 total events: 16 IPMN with high-grade dysplasia and 10 PDAC, compared with 2.44 expected events in the matched SEER comparator. The paper therefore frames risk in terms of malignancy arising during surveillance rather than only invasive PDAC as a narrowly defined endpoint.
Key Findings
Overall risk is substantially higher than background population risk
The cumulative incidence of malignancy at 60 months was 0.97% with a confidence interval of 0.53% to 1.70%. On first reading, that absolute risk appears modest. However, the key comparative finding is relative rather than absolute: the SIR for malignancy during surveillance was approximately 10-fold higher than expected in the matched general population, with a confidence interval of 7.2 to 15.6.
This distinction matters clinically. The study does not suggest that most patients with IPMN will develop invasive cancer in the near term. Instead, it shows that patients with IPMN are meaningfully enriched for pancreatic malignancy risk compared with age-matched peers, even under active surveillance.
Risk varies sharply by imaging phenotype and patient factors
The strongest radiographic signal was MPD dilation of 5 mm or greater, which was associated with an SIR of 26.5 (95% confidence interval 11 to 64). This is biologically plausible and clinically consistent with existing guideline frameworks, in which duct dilation is a major high-risk feature because it may reflect main-duct involvement, greater epithelial instability, or occult advanced neoplasia.
Obesity also emerged as a notable risk modifier. Patients with BMI 30 or greater had an SIR of 20.6 (95% confidence interval 11.4 to 37). This is important because BMI is not traditionally emphasized as a central determinant of IPMN surveillance intensity, even though obesity is a recognized risk factor for pancreatic cancer more broadly. The finding suggests that host factors and cyst biology likely interact.
Lower-risk subgroups approach, but do not reach, population baseline
Patients with IPMN 1 cm or smaller had an SIR of 3.28 (95% confidence interval 0.82 to 13.1), and those younger than 60 years had an SIR of 6.27 (95% confidence interval 1.58 to 24.9). These were the subgroups closest to general-population risk, although confidence intervals were wide and still compatible with a clinically important excess risk.
These findings are directionally reassuring for younger patients and those with very small cysts, but they do not justify declaring surveillance unnecessary based on size alone. Rather, they highlight the possibility that surveillance intervals might be safely stretched in carefully selected patients, especially when radiographic stability is durable.
“Guideline-negative” does not equal negligible risk
Perhaps the most practice-relevant observation is that cysts smaller than 3 cm without MPD dilation or a solid component still had an SIR of 8.40 (95% confidence interval 5.15 to 13.7). This group would often be considered lower risk in routine practice. Yet compared with the background population, their risk remained substantially elevated.
This helps explain why clinicians remain uneasy about discharging apparently stable IPMN patients from surveillance. A lesion may lack classic worrisome features and still mark a pancreas at increased risk for neoplasia, whether through field-defect biology, unrecognized microscopic progression, or limitations of imaging resolution.
Clinical Interpretation
The study delivers an important calibration point for day-to-day practice. It supports the premise underlying long-term IPMN surveillance: these patients do face materially higher pancreatic malignancy risk than the general population. At the same time, it argues against treating all IPMNs as equivalent.
Current surveillance paradigms largely rely on cyst size, mural nodules, duct diameter, growth, and symptoms. This remains appropriate, but the present data suggest that integrating patient-level factors such as age and BMI may improve risk discrimination. If validated, a surveillance model combining imaging phenotype with host factors could help in two directions at once: intensifying follow-up in patients most likely to benefit and reducing imaging burden in those with persistently low-risk profiles.
Importantly, absolute event rates remained relatively low over 5 years. Therefore, an elevated SIR should not be misinterpreted as a rationale for broad prophylactic surgery. Pancreatic resection carries meaningful morbidity, and the competing risks of age, comorbidity, and operative harm remain central to decision-making. The practical implication is more nuanced: surveillance should be personalized, but not abandoned.
Strengths of the Study
This analysis has several strengths. First, the cohort is large for a single-center IPMN surveillance study and includes extensive longitudinal imaging, with more than 11,000 scans. Second, the use of person-years at risk and a SEER-based expected event framework provides a clinically intuitive measure of excess cancer burden. Third, the investigators report subgroup-specific SIRs that move the discussion from average risk toward actionable heterogeneity.
The study also reflects real-world surveillance in a specialized center, where imaging interpretation and multidisciplinary management are likely more consistent than in fragmented practice settings.
Limitations and Cautions
Several limitations should temper interpretation. This was a retrospective observational study from a high-volume referral center, so referral bias is likely. Patients seen at tertiary centers may have more complex cysts, greater imaging scrutiny, or different thresholds for intervention than community populations.
The outcome definition also requires careful reading. The comparison is described as PDAC risk, but the observed events included both invasive PDAC and IPMN with high-grade dysplasia. High-grade dysplasia is clinically meaningful and often treated as a near-cancer endpoint, but it is not identical to invasive cancer. This may complicate direct comparison with SEER-derived expected PDAC incidence.
Residual confounding is also inevitable. Smoking, diabetes status, family history, genetic predisposition, pancreatitis history, and detailed cyst subtype may all influence risk but may not be fully captured in the SIR framework. In addition, wide confidence intervals in some subgroups indicate limited precision, especially for patients with very small cysts or younger age.
Finally, surveillance itself can alter the natural history that is being measured. Earlier detection of high-grade dysplasia or cancer is beneficial, but it also means observed event counts partly reflect the intensity and quality of monitoring. Therefore, the results should not be interpreted as a pure estimate of untreated natural progression.
How This Fits With Existing Guidelines and Literature
The findings are broadly consistent with long-standing guideline concerns regarding MPD dilation and other high-risk stigmata. The 2017 International Association of Pancreatology Fukuoka guidelines emphasize features such as duct dilation, mural nodules, obstructive jaundice, and cyst growth as triggers for closer evaluation or surgery. The present analysis reinforces the prognostic weight of ductal involvement.
At the same time, the study challenges overly simplistic reassurance for small branch-duct lesions. Prior guidelines, including those from the American Gastroenterological Association, have sought to reduce over-surveillance in lower-risk cysts, but this paper reminds clinicians that population-relative risk remains elevated even when classic high-risk imaging features are absent.
In that sense, the study supports neither maximalist surveillance for everyone nor wholesale de-escalation. Instead, it supports precision surveillance: dynamic, risk-adapted, and integrated with surgical fitness, life expectancy, and patient preference.
Practice Implications
For clinicians, several practical messages emerge. Patients with MPD dilation of 5 mm or greater likely warrant particularly careful multidisciplinary review, because their excess malignancy risk appears substantial. Obesity may deserve greater attention during risk counseling and interval planning, especially in conjunction with concerning imaging findings.
For patients with very small, stable cysts and no ductal or solid features, these data offer partial reassurance but not exoneration. Longer surveillance intervals may eventually be reasonable in selected individuals, but the study does not support stopping surveillance solely because the lesion is small.
For health systems and guideline developers, the work argues for next-generation models that estimate individualized risk rather than relying on binary feature lists alone. A validated calculator incorporating imaging features, demographics, metabolic factors, and perhaps biomarkers could improve both cancer detection and resource stewardship.
Conclusion
This SEER-matched study shows that patients under surveillance for IPMN carry a markedly elevated malignancy risk compared with the general population, with an overall approximately 10-fold excess. The most striking risk enrichment was seen with MPD dilation of 5 mm or greater and BMI of 30 or greater, while younger age and cysts 1 cm or smaller approached lower relative risk. Even cysts lacking traditional high-risk features remained meaningfully above baseline population risk.
The main clinical lesson is not that all IPMNs are dangerous, but that IPMN surveillance should become more personalized. Patients are not equally threatened, and they likely should not all be imaged in the same way or at the same frequency. Future work should validate multivariable risk models, integrate biomarkers and molecular data, and determine whether tailored surveillance can preserve early detection while reducing unnecessary imaging and intervention.
Funding and ClinicalTrials.gov
The abstract provided does not report a funding source. No ClinicalTrials.gov registration is applicable for this retrospective observational study, and none is listed in the citation provided.
References
1. Choubey AP, Chou JF, Alessandris R, Gonen M, Armstrong MT, Manin E, Saadat LV, Flood J, Balachandran VP, Drebin JA, Kingham TP, D’Angelica MI, Jarnagin WR, Wei AC, Rolston VS, Schattner MA, Do RKG, Soares KC; MSKCC Pancreas Cyst Collaborative. Pancreas Cancer Risk During Intraductal Papillary Mucinous Neoplasm Surveillance – A SEER-Based Comparison. Ann Surg. 2026 May 25. doi: 10.1097/SLA.0000000000007100. Epub ahead of print. PMID: 42183647.
2. Tanaka M, Fernández-del Castillo C, Kamisawa T, Jang JY, Levy P, Ohtsuka T, Salvia R, Shimizu Y, Tada M, Wolfgang CL. Revisions of international consensus Fukuoka guidelines for the management of IPMN of the pancreas. Pancreatology. 2017;17(5):738-753.
3. Vege SS, Ziring B, Jain R, Moayyedi P; Clinical Guidelines Committee; American Gastroenterological Association. American Gastroenterological Association Institute guideline on the diagnosis and management of asymptomatic neoplastic pancreatic cysts. Gastroenterology. 2015;148(4):819-822.
