TOPLINE

In a case-control study, higher levels of the immune protein MMP12 were associated with an increased risk for pancreatic cancer, particularly among individuals diagnosed within 8 years of blood collection, with the strongest associations observed in those diagnosed within 4 years.

METHODOLOGY

• Researchers conducted a nested case-control study including 406 patients with incident pancreatic cancer (median age, 57 years; 48% men) from 10 European countries who were diagnosed between 1995 and 2012, with a median follow-up duration of 8.6 years.
• Patients were matched with 406 control individuals on the basis of age and sex. Blood samples were collected and stored at -190 °C, and 92 immune proteins were measured from the Olink immuno-oncology panel using the proximity extension assay technology.
• Associations between protein levels and the risk for pancreatic cancer were evaluated. Subgroup analyses were conducted on the basis of the lag time to the diagnosis of pancreatic cancer, categorised as 0-4, 4-8, 8-12, and more than 12 years between blood collection and diagnosis.

TAKEAWAY

• After adjustment for smoking status, alcohol consumption, self-reported diabetes, and BMI, higher levels of MMP12 were associated with an increased risk for pancreatic cancer (odds ratio [OR], 1.29; P = .02), and higher levels of interleukin (IL)-12 (OR, 0.81; P = .04), CD28 (OR, 0.74; P = .03), PD-L2 (OR, 0.66; P = .01), and IL-6 (OR, 0.86; P = .05) were associated with a lower risk for pancreatic cancer in the individual analysis.
• In the multivariable analysis, only high levels of MMP12 (OR, 1.56; P = .001) remained significantly associated with an increased risk for pancreatic cancer. IL-6 showed an inverse association with the risk for pancreatic cancer (OR, 0.84; P = .05), although this finding was not consistent with previous studies.
• Lag-time analysis showed that the association of MMP12 with the risk for pancreatic cancer was strongest for cases diagnosed within 4 years (OR, 1.89; 95% CI, 1.28-2.80) and remained significant for those diagnosed within 4-8 years (OR, 1.37; 95% CI, 1.01-1.86).
• After correction for multiple testing using the Benjamini-Hochberg false discovery rate method, none of the proteins showed statistically significant associations with the risk for pancreatic cancer.

IN PRACTICE

The results suggest "potential for MMP12 as a biomarker for early inflammatory perturbances related to pancreatic cancer development," the authors wrote.

SOURCE

The study was led by Verena A. Katzke and Yue Chen, Division of Cancer Epidemiology, German Cancer Research Center (DKFZ), Heidelberg, Germany. It was published online on June 23, 2026, in the International Journal of Cancer.

LIMITATIONS

Pancreatic cancer staging and grading information at the time of diagnosis was missing. Results for proteins with a large proportion of measurements below the detection limit, such as CD28 with 35% below the limit of detection, were less reliable. With the current sample size, some associations with pancreatic cancer may have been overlooked, whereas the observed associations might have represented chance findings.

DISCLOSURES

The study did not receive any specific funding. The authors reported having no relevant conflicts of interest.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.