Table 2.

Univariate and multivariate analyses of various predictive factors for pCR in different cohorts

		Univariable analysis	Multivariable analysis
		P valuea	P valueb	OR	95% CI
Beijing training cohort
Age	≥60/<60	0.406
Sex	Male/female	0.156
Tumor location	Upper, middle/lower	0.820
Tumor differentiation	Moderately, poorly/well differentiated	0.345
Clinical TNM stage	II/III	0.820
Chemotherapy regimenc	1/2, 3	0.629
Discriminant score	High/low	<0.001	<0.001	171.873	20.259–1458.122
Beijing validation cohort
Age	≥60/<60	0.562
Sex	Male/female	0.579
Tumor location	Upper, middle/lower	0.240
Tumor differentiation	Moderately, poorly/well differentiated	0.489
Clinical TNM stage	II/III	0.125
Chemotherapy regimenc	1/2, 3	0.309
Discriminant score	High /low	<0.001	<0.001	153.824	12.166–1944.861
Entire Beijing cohort
Age	≥60/<60	0.592
Sex	Male/female	0.226
Tumor location	Upper, middle/lower	0.554
Tumor differentiation	Moderately, poorly/well differentiated	0.120
Clinical TNM stage	II/III	0.317
Chemotherapy regimenc	1/2, 3	0.398
Discriminant score	High /low	<0.001	<0.001	26.169	9.075–75.458
External validation cohort
Age	≥60/<60	0.957
Sex	Male/female	0.222
Tumor location	Upper, middle/lower	0.153
Tumor differentiation	Moderately, poorly/well differentiated	0.010	0.145	0.235	0.034–1.645
Clinical TNM stage	II/III	0.043	0.068	0.111	0.010–1.177
Chemotherapy regimenc	1/2, 3	0.629
Discriminant score	High/low	<0.001	0.001	56.133	4.804–655.843

pCR pathological complete response, OR odds ratio, CI confidence interval

^aχ² or Fisher exact tests

^bLogistic regression analysis with a forward stepwise procedure and likelihood ratio test

^c1, platinum/paclitaxel; 2, platinum/fluorouracil; 3, platinum/others