Skip to main content
. 2017 Jan 24;7:41151. doi: 10.1038/srep41151

Table 2. Serum Markers in Lung Cancer and Controls in the Training Set.

Marker Control/Casea Median Range Pb Adjusted effect (95% CI)c AUC (95% CI)
EGF Control 336.86 40.13–1187.06      
(pg/mL)  Healthy 247.60 40.13–972.90      
   Benign 419.67 43.25–1187.06      
  LC 571.70 40.75–1716.30 0.001 0.238 (0.140–0,336)* 0.698 (0.615–0.773)
sCD26 Control 471.50 122.00–1092.00      
(ng/mL)  Healthy 509.00 237.00–886.00      
   Benign 431.50 122.00–1092.00      
  LC 356.00 136.00–1192.00 0.001 −0.097 (−0.148–0.047)* 0.711 (0.629–0.785)
CAL Control 127.87 7.56–421.23      
(ng/mL)  Healthy 105.47 7.56–362.29      
   Benign 158.13 33.13–421.23      
  LC 221.21 48.33–438.32 0.001 0.263 (0.190–0.336)* 0.759 (0.679-0.827)
MMP-1 Control 5459.61 1186.61–23960.37      
(pg/mL)  Healthy 4664.94 1186.61–23635.33      
   Benign 6420.51 1207.70–23960.37      
  LC 7133.87 1450.34–41668.33 0.047 0.089 (−0.008–0.187) 0.597 (0.511-0.679)
MMP-7 Control 21761.35 5026.14–79977.27      
(pg/mL)  Healthy 21237.20 10383.69–60968.91      
   Benign 22286.74 5026.14–79977.27      
  LC 26710.65 5383.18–79809.13 0.013 0.020 (−0.040–0.081) 0.624 (0.539-0.705)
MMP-9 Control 177.60 52.79–3611.59      
(ng/mL)  Healthy 177.19 52.79–743.25      
   Benign 181.74 57.41–3611.59      
  LC 311.86 21.06–1914.00 0.001 0.270 (0.167–0.373)* 0.729 (0.648–0.801)
CEA Control 837.26 170.84–4070.71      
(pg/mL)  Healthy 929.50 171.03–4070.71      
   Benign 559.51 170.84–2828.15      
  LC 2051.64 141.16–136039.19 0.001 0.494 (0.333–0.656)* 0.744 (0.663–0.814)
CYFRA 21.1 Control 227.05 0.00–19314.33      
(pg/mL)  Healthy 0.00 0.00–16592.63      
   Benign 1052.47 0.00–19314.33      
  LC 3007.09 0.00–173410.17 0.001 1.080 (0.587–1.573)* 0.734 (0.653–0.805)

Abbreviations: LC = Lung Cancer.

aSample size in training set: Control n = 72 (Healthy n = 36, Benign n = 36), LC n = 68.

bMann-Whitney U test for comparison between the cancer and control group corrected by Benjamini-Hochberg method to control familywise error under multiple comparisons.

cAdjusted effects and 95% confidence intervals of the case/control status on each of the log-transformed markers considered as outcome in lineal regression model adjusted for gender, age and smoking. *P-value < 0.001.