Assessing the performance of methylated DNA markers in blood for detection of pancreatic ductal adenocarcinoma

June 18, 2021

Pancreatic cancer is a leading cause of cancer mortality. The global incidence of pancreatic cancer is increasing, and the majority of patients are diagnosed at an advanced stage when currently available therapies have limited impact on survival. Currently, the most widely used blood-based tumor marker for pancreatic cancer, the carbohydrate antigen 19-9 (CA 19-9) test, fails to detect disease in a subset of patients with pancreatic cancer, especially those patients with early-stage disease.

Mayo Clinic researchers previously identified and validated methylated DNA markers (MDMs) that are strongly associated with pancreatic ductal adenocarcinoma (PDAC) and precancerous lesions (high-grade dysplasia) in pancreatic tissue. Building on their previous work, Shounak Majumder, M.D., and colleagues conducted a case-control study to assess the performance of MDMs in plasma for detection of PDAC. The results of the study were published in Clinical Cancer Research in 2021. Dr. Majumder is a gastroenterologist at Mayo Clinic's campus in Rochester, Minnesota, and served as lead author of the Clinical Cancer Research article.


Using selection criteria applied to results from prior tissue experiments, the researchers identified 13 MDMs to study (GRIN2D, CD1D, ZNF781, FER1L4, RYR2, CLEC11A, AK055957, LRRC4, GH05J042948, HOXA1, PRKCB, SHISA9 and NTRK3), and they optimized the assays for a plasma application. They then assayed 340 plasma samples from 170 patients with confirmed PDAC and 170 healthy controls using a target enrichment long-probe quantitative amplified signal method.

Plasma samples were divided into a training set comprising 120 advanced-stage PDAC (stages 3 and 4) cases and 120 healthy controls and a test set with 50 early-stage PDAC (stages 1 and 2) cases and 50 controls. Within both sets, cases and controls were balanced on age and sex.

Researchers applied random forest modeling to the set of 120 advanced-stage PDAC cases and 120 healthy controls to train a prediction algorithm to 97.5% specificity. They then applied the locked algorithm derived from this training set to the set of 50 early-stage PDAC cases and 50 controls. In the last phase of the analysis, researchers combined and cross-validated data from all 340 plasma samples (170 with PDAC and 170 controls).


From their analysis of the 170 PDAC cases and 170 controls, Mayo Clinic researchers found that the combined MDM-CA 19-9 panel distinguished pancreatic cancer cases from cancer-free controls with a high degree of accuracy.

"The combined plasma MDM-CA 19-9 panel performed significantly better, when compared with either biomarker individually," explains Dr. Majumder.

The following data highlights some of the researchers' important findings:

  • The cross-validated area under the receiver operating characteristic curve (AUC) for the combined MDM-CA 19-9 panel training set was 0.99 (0.98 to 1), 0.93 (0.89 to 0.96) for the MDM panel alone and 0.91 (0.87 to 0.96) for the CA 19-9 test alone.
  • In testing the early-stage PDAC sample set, the AUC for the combined MDM-CA 19-9 panel was 0.90 (0.84 to 0.97), 0.87 (0.79 to 0.94) for the CA 19-9 test alone and 0.84 (0.76 to 0.92) for the MDM panel alone.
  • Using the combined data sets (170 with PDAC and 170 controls), the cross-validated AUC was 0.97 (0.94 to 0.99) for the combined MDM-CA 19-9 panel, 0.9 (0.86 to 0.94) for the MDM panel alone and 0.89 (0.84 to 0.93) for the CA 19-9 test alone.
  • For all cancer stages combined, the cross-validated sensitivity of the combined MDM-CA 19-9 panel was 92% (83% to 98%), with an observed specificity of 92% (81% to 100%).

"The development of a novel blood test that accurately detects pancreatic cancer across all stages of the disease, either stand alone or in combination with CA 19-9, will be clinically impactful," says Dr. Majumder. "We plan to conduct additional studies to further validate and optimize this promising new blood-based diagnostic approach. The plasma MDMs may have additional potential for multicancer screening tests and for pancreas cancer screening in high-risk individuals."

For more information

Majumder S, et al. High detection rates of pancreatic cancer across stages by plasma assay of novel methylated DNA markers and CA 19-9. Clinical Cancer Research. 2021;27:2523.