Skip to main content

Table 2 Univariate and multivariate analysis for pCR-score and important factors

From: Deep learning-based predictive biomarker of pathological complete response to neoadjuvant chemotherapy from histological images in breast cancer

Factors Univariate analysis Multivariate analysisa Multivariate analysisb
OR (95% CI) P OR (95% CI) P OR (95% CI) P
pCR-score 3.516 (2.003–6.173) < 0.001 4.045 (1.822–8.980) 0.001
Subtypes < 0.001 0.003 0.084
 HR +/HER2- 1 1 1 1 1 1
 HR +/HER2 +  8.939 (1.861–42.944) 0.006 8.049 (1.484–43.664) 0.016 2.899 (0.395–21.282) 0.295
 HR-/HER2 − 5.900 (1.041–33.477) 0.045 4.798 (0.733–31.387) 0.102 2.327 (0.262–20.671) 0.449
 HR-/HER2 +  32.778 (7.031–152.815) < 0.001 21.497 (4.089–113.00) < 0.001 12.923 (1.783–93.638) 0.011
sTILs density 0.002 0.041 0.060
 Low 1 1 1 1 1 1
 Moderate 7.673 (2.302–25.581) 0.001 5.374 (1.311–22.021) 0.019 8.256 (1.404–48.546) 0.020
 High 14.250 (2.143–94.741) 0.006 9.206 (0.939–90.247) 0.057 7.379 (0.511–106.471) 0.142
T stage 0.754 (0.290–1.962) 0.563 1.515 (0.428–5.356) 0.519 2.219 (0.487–10.119) 0.303
Ki67 4.571 (0.571–36.610) 0.152 1.166 (0.116–11.737) 0.897 1.859 (0.142–24.352) 0.636
Nuclear grade 1.987 (0.753–5.240) 0.165 1.024 (0.283–3.708) 0.971 1.076 (0.233–4.956) 0.926
  1. aMultivariate analysis excluding the pCR-score
  2. bMultivariate analysis including the pCR-score