Skip to main content

Table 3 Biomarker modeling in training set (Sample Set B)

From: Validation of four candidate pancreatic cancer serological biomarkers that improve the performance of CA19.9

Biomarker combinationa

AUCb of combination

Lower 95% confidence interval

Upper 95% confidence interval

Specificity at 9% sensitivity

Sensitivity at 95% specificity

CA19.9 + SYCN + REG1B

0.926

0.880

0.965

0.509

0.739

CA19.9 + SYCN + AGR2

0.919

0.869

0.959

0.515

0.771

CA19.9 + SYCN

0.918

0.864

0.958

0.502

0.771

CA19.9 + SYCN + LOXL2

0.918

0.871

0.959

0.479

0.779

CA19.9 + REG1B + LOXL2

0.879

0.814

0.936

0.219

0.727

CA19.9 + REG1B

0.878

0.815

0.933

0.261

0.689

CA19.9 + AGR2 + REG1B

0.877

0.816

0.932

0.268

0.688

CA19.9 + LOXL2

0.844

0.773

0.908

0.065

0.744

CA19.9 + AGR2 + LOXL2

0.844

0.772

0.910

0.065

0.744

CA19.9 + AGR2

0.835

0.762

0.902

0.046

0.707

CA19.9

0.833

0.757

0.902

0.048

0.699

SYCN + REG1B + LOXL2

0.826

0.747

0.896

0.250

0.395

SYCN + REG1B

0.823

0.745

0.893

0.248

0.356

SYCN + AGR2 + REG1B

0.819

0.741

0.891

0.261

0.359

SYCN + AGR2 + LOXL2

0.800

0.716

0.875

0.149

0.306

SYCN + AGR2

0.794

0.712

0.874

0.159

0.276

SYCN + LOXL2

0.793

0.712

0.870

0.179

0.282

REG1B

0.790

0.705

0.862

0.300

0.326

SYCN

0.790

0.703

0.868

0.201

0.220

AGR2 + REG1B

0.781

0.697

0.859

0.289

0.312

REG1B + LOXL2

0.779

0.694

0.854

0.186

0.369

AGR2 + REG1B + LOXL2

0.779

0.694

0.853

0.256

0.345

AGR2 + LOXL2

0.675

0.589

0.764

0.106

0.175

AGR2

0.671

0.576

0.763

0.087

0.130

LOXL2

0.540

0.439

0.644

0.033

0.218

  1. a Biomarker models were generated for each of the above combinations for the PDAC (n = 82) versus the disease-free (n = 47) groups of Sample Set B and ordered from greatest to lowest AUC. Confidence intervals (CI) for AUC were calculated using DeLong’s method. The models from the combinations of two or three markers were then validated in the PDAC versus healthy groups of Sample Set A (Table 4); b AUC, area under the receiver operating characteristic curve; PDAC, pancreatic ductal adenocarcinoma.