Table 2 Performance of AFP and candidate biomarkers distinguishing HCC cases from liver cirrhosis cases and healthy normal controls in the serum validation cohorts.

From: Quantitative proteomics identifies a plasma multi-protein model for detection of hepatocellular carcinoma

Diagnostic modelsa

AUC (95% CI)

Pb

Sensitivity (95% CI)

Specificity (95% CI)

Cutoffc

Pd (vs. AFP)

AFP

0.831 (0.716–0.913)

 < 0.01

75.0% (47.60–92.70%)

85.4% (72.2–93.9%)

4.97 ng/mL

–

AKR1B10

0.891 (0.788–0.955)

 < 0.01

75.0% (47.6–92.7%)

89.6% (77.3–96.5%)

122.29 ng/mL

0.46

CTSA

0.894 (0.792–0.957)

 < 0.01

100.0% (79.4–100.0%)

64.5% (49.5–77.8%)

0.56 ng/mL

0.38

AFP + CTSA

0.932 (0.841–0.980)

 < 0.01

100.0% (79.4–100.0%

77.1% (62.7–88.0%)

0.12

0.09

CTSA + AKR1B10

0.952 (0.867–0.990)

 < 0.01

93.7% (69.8–99.8%)

83.3% (69.8–92.5%)

0.11

0.08

AFP + AKR1B10

0.958 (0.876–0.992)

 < 0.01

87.5% (61.7–98.4%)

89.6% (77.3–96.5%)

0.15

0.035

AFP + AKR1B10 + CTSA

0.969 (0.892–0.996)

 < 0.01

93.7% (69.8–99.8%)

85.4% (72.2–93.9%)

0.12

0.029

  1. AFP alpha-fetoprotein, AKR1B10 aldo–keto reductase family 1 member B10, AUC areas under the receiver operating characteristic curve, CI confidence interval, CTSA cathepsin A, ROC receiver operating characteristic.
  2. aThe multi-protein diagnostic models were constructed by binomial logistic regression analyses.
  3. bP values for the significance of differences between AUCs and 0.5 by Wilcoxon rank-sum test.
  4. cThe optimal cut-point values in ROC analyses were defined by Youden index.
  5. dP values for the significance of AUC differences between candidate biomarkers and AFP by Wilcoxon rank-sum test.
Back to article page