Biomarker Identification and Pathway Analysis by Serum Metabolomics of Lung Cancer

Table 5

ROC curves of potential serum biomarker levels for differentiating the control group from the PRLC group.

Marker	Cutoff value	Sensitivity (%)	Specificity (%)	AUC	p value^*	95% CI^a
Phosphorylcholine	78.32	90.00	80.00	0.874	<0.001	0.780–0.969
Glycerophospho-N-arachidonoyl ethanolamine	1752.60	96.67	90.00	0.983	<0.001	0.960–1.007
-Linolenic acid	805.17	76.67	93.33	0.889	<0.001	0.806–0.972
-Hydroxyisobutyric acid	365.23	80.00	80.00	0.860	<0.001	0.764–0.956
9,12-Octadecadienoic acid	13.56	66.67	83.33	0.704	<0.001	0.562–0.847
Oleic acid	402.22	70.00	86.67	0.749	<0.001	0.614–0.884
CEA	2.54	76.67	93.33	0.867	<0.001	0.765–0.969
CYFRA21-1	2.02	56.67	96.67	0.803	<0.001	0.690–0.916

 ^*Asymptotic significance, null hypothesis: true area = 0.5.
^a95% confidence interval of the difference.