GI & Hepatology News Home / Diabetes duration tied to risk of pancreatic cystic neoplasms in nationwide cohort

Diabetes duration tied to risk of pancreatic cystic neoplasms in nationwide cohort

A population-based Korean cohort study found increasing pancreatic cyst risk with worsening glycemic status, with the highest risk among individuals with diabetes for 5 years or longer.

Share

Longer-standing diabetes was associated with a higher risk of developing pancreatic cystic neoplasms (PCNs) in a nationwide cohort study of almost 4 million Korean adults followed for a median of about 10 years. The association persisted after adjustment for major confounders and appeared stronger in adults younger than 60 years, in men, and in current smokers with diabetes, according to results published in JAMA Open Network.
 
"Health care professionals should consider these factors, especially when evaluating pancreatic findings in patients with diabetes and additional risk factors," wrote Sang Hyub Lee, MD, PhD, of the Department of Internal Medicine and Liver Research Institute, at Seoul National University College of Medicine, Korea, and colleagues.
 
They used the Korean National Health Insurance Service database to identify adults aged 20 years or older who underwent standardized health examinations in 2009 and were followed through December 31, 2020. To focus on incident disease, the analysis excluded individuals with prior pancreatic cyst diagnoses during a washout period spanning 2002 through the index examination, those diagnosed with a pancreatic cyst within 1 year of the examination (a 1-year lag period), and those with missing data. To reduce inclusion of nonneoplastic cysts linked to pancreatitis, the study excluded participants with diagnostic codes for acute pancreatitis, chronic pancreatitis, alcoholic pancreatitis, pseudocysts, or pancreatic fluid collection diagnosed prior to baseline or after baseline but before cyst/PCN occurrence.
 
Participants were categorized by glycemic status at baseline as having normoglycemia, impaired fasting glucose, diabetes for less than 5 years, or diabetes for 5 years or longer.
 
Diabetes and diabetes duration were identified using a combination of ICD-10 codes, prescription records for antidiabetic medications, and fasting glucose measurements obtained during the health examination; diabetes duration was determined by the interval from first antidiabetic prescription to the date of the baseline examination.
 
Over a total observation period of roughly 38.9 million person-years, 31,877 participants developed PCNs, representing 0.8% of the cohort. Compared with normoglycemia, the adjusted hazard ratio for developing PCNs was 1.06 among those with impaired fasting glucose, 1.23 among those with diabetes for less than 5 years, and 1.37 among those with diabetes for 5 years or longer.
 
Incidence rates per 1,000 person-years rose across these categories, from 0.72 in normoglycemia to 0.89 in impaired fasting glucose, 1.25 in diabetes for less than 5 years, and 1.82 in diabetes for 5 years or longer.
 
Subgroup analyses suggested heterogeneity in relative risk. In participants younger than 60 years with diabetes, the adjusted hazard ratio for PCN occurrence was 1.34, compared with 1.21 in those 60 years or older with diabetes. Among men with diabetes, the adjusted hazard ratio was 1.32, compared with 1.20 among women with diabetes. Among current smokers with diabetes, the adjusted hazard ratio was 1.40, compared with 1.22 among never-smokers with diabetes and 1.25 among former smokers with diabetes. The authors also reported that insulin users had higher risk of pancreatic cyst occurrence than individuals using oral antidiabetic medications only, particularly among those with diabetes duration of 5 years or longer, based on supplemental analyses.
 
During follow-up, 1,315 of the 31,877 individuals who developed pancreatic PCNs were subsequently diagnosed with pancreatic cancer, corresponding to 4.1% of those with PCNs, based on claims coding. The authors noted that claims-based identification and the absence of imaging data limited subtype classification and the ability to assess cyst characteristics such as size, location, and high-risk imaging features. They also cautioned that the study design could not establish causality and that further research incorporating imaging and longitudinal clinical data is needed to clarify the clinical significance of pancreatic cysts in individuals with diabetes.
 
"From a clinician standpoint, the most important message is that although diabetes is associated with increased PCN incidence, the absolute risk remains very low. The overall rate was 0.82 per 1000 person-years, increasing to 1.82 per 1000 person-years in individuals with 5 or more years of diabetes. In other words, even after a decade of follow-up, PCNs remain uncommon in all glycemic categories," wrote Stefano Crippa, MD, PhD, and Lorenzo Piemonti, MD, PhD, in an accompanying editorial. "Thus, while the finding of an association is compelling, the absolute incidence and the cyst phenotype captured in this study remain uncertain. This limitation does not limit the epidemiologic value but should temper clinical extrapolation, noted Drs. Crippa and Piemonti, of Vita Salute San Raffaele University, Milan, Italy.
 
The investigators and invited commentary authors reported having no conflicts of interest.

Summary content