Skip to main content

Table 1 Characteristics of patients in institution 1

From: CT-based deep learning model for the prediction of DNA mismatch repair deficient colorectal cancer: a diagnostic study

Variable

Training cohort (n = 1124)

Internal validation cohort (n = 482)

dMMR (n = 273)

pMMR (n = 851)

p value

dMMR (n = 117)

pMMR (n = 365)

p value

Age

54.0 (42.8, 64.0)

60.0 (50.0, 68.0)

< 0.001

57.0 (44.7, 65.3)

60.0 (51.0, 67.3)

0.022

Gender

  

0.762

  

0.241

 Male

169

516

 

64

224

 

 Female

104

335

53

141

 

cT stage

  

0.145

  

0.019

 1–2

3

19

 

3

14

 

 3

168

550

62

241

 

 4a

94

243

52

107

 

 4b

8

39

0

3

 

cN stage

  

< 0.001

  

0.005

 0

69

250

 

17

89

 

 1

108

443

66

214

 

 2

96

158

34

62

 

Location

  

< 0.001

  

< 0.001

 Right

168

284

 

84

119

 

 Left

105

567

33

246

 

Tumor marker

      

 CEA (IQR)

12.1 (1.4, 4.5)

80.3 (2.1, 13.0)

0.164

10.6 (1.6, 4.3)

23.7 (2.3, 11.1)

0.241

 CA125 (IQR)

19.4 (8.6, 17.8)

26.3 (8.7, 21.8)

0.017

17.1 (8.2, 17.3)

32.6 (9.1, 23.4)

0.120

 CA199 (IQR)

150.9 (2.7, 20.2)

536.4 (4.1, 30.1)

0.247

27.0 (2.8, 14.8)

156.4 (3.9, 24.3)

0.243

 CA153 (IQR)

10.4 (6.3, 13.0)

10.4 (6.3, 12.2)

0.990

9.1 (6.0, 12,1)

10.4 (6.6, 12.2)

0.202

  1. IQR: interquartile range; dMMR: DNA mismatch repair deficient; pMMR: mismatch repair-proficient; cT stage: baseline clinical tumor stage; cN stage: baseline clinical lymph node stage; CEA: carcinoembryonic antigen; CA199: cancer antigen 199; CA125: cancer antigen 125; CA153: cancer antigen 153