Skip to main content
. 2020 Dec 12;21(24):9460. doi: 10.3390/ijms21249460

Table 2.

Cox proportional hazard regression models for the association between clinical predictor variables, plasma cytokines, the interaction between individual TILs and CTC EMT, and PFS.

Model Variable HR 95% CI p
1 LN+ 6.446 2.56–16.24 <0.001
Ki-67 > 20% 12.00 4.22–34.11 <0.001
Cytokines * 5.928 2.12–16.58 0.001
CD3high and CTC EMT positivity 5.277 2.09–13.30 <0.001
2 LN+ 4.256 1.79–10.10 0.001
Ki-67 > 20% 8.482 3.25–22.13 <0.001
Cytokines * 6.387 2.35–17.38 <0.001
CD8high and CTC EMT positivity 3.655 1.49–9.00 0.005
3 LN+ 3.796 1.50–9.58 0.005
Ki-67 > 20% 7.251 2.59–20.30 <0.001
Cytokines * 5.172 1.80–14.87 0.002
CD45ROhigh and CTC EMT positivity 4.922 1.83–13.23 0.002

Categorical variables entered in step 1: Age categories; T-stage; Histological grade; Ki-67, cut-off 20%; HER2 status; HR status, cut-off 1%; Grade; N-stage; * TGF-β3high, MCP-3low, and IL-15low; interaction of CD3, CD8, and CD45RO values with CTC EMT positivity.