Table 2 Summary of Cox regression analysis of RPV in three datasets. RPV was used as a continuous variable in the Cox regression analysis

From: A mathematical-descriptor of tumor-mesoscopic-structure from computed-tomography images annotates prognostic- and molecular-phenotypes of epithelial ovarian cancer

 

Variables

Univariate

Multivariable

HR (95% CI)

p value

HR (95% CI)

p value

 

HH discovery (n = 136)

RPV

4.08 (2.48–6.71)

3.37e-08

3.86 (2.30–6.46)

3.04 × 10−7

Stage

2.03 (1.37–3.00)

0.000426

1.88 (1.24–2.86)

0.00305

Residual disease

1.75 (1.03–2.99)

0.0393

1.40 (0.803–2.44)

0.235

Agea

1.25 (0.741–2.11)

0.404

1.47 (0.865–2.51)

0.154

HH validation (n = 77)

RPV

2.05 (1.01–4.18)

0.0485

5.08 (1.03–25.2)

0.0465

Stage

1.32 (0.775–2.24)

0.309

1.32 (0.664–2.64)

0.425

Residual disease

1.78 (0.777–4.08)

0.173

1.28 (0.514–3.21)

0.593

Agea

2.10 (0.940–4.68)

0.0704

3.44 (1.19–9.94)

0.0228

HH cohort combinedb (n = 213)

RPV

2.94 (2.02–4.26)

1.54 × 10−8

3.32 (2.16–5.10)

4.91 × 10−8

Stage

1.82 (1.33–2.48)

0.00017

1.75 (1.24–2.50)

0.0017

Residual disease

1.72 (1.11–2.69)

0.0163

1.36 (0.855–2.15)

0.196

Agea

1.46 (0.951–2.24)

0.0835

1.74 (1.10–2.76)

0.0183

TCGA validation (n = 70)

RPV

4.94 (2.06–11.8)

0.00034

6.21 (2.06–18.7)

0.00117

Stage

1.75 (0.913–3.34)

0.0921

1.03 (0.309–3.44)

0.960

Residual disease

1.34 (0.480–3.74)

0.576

1.45 (0.414–5.05)

0.564

Agea

1.08 (0.435–2.66)

0.874

0.500 (0.154–1.63)

0.249

  1. HR hazard ratio, CI confidence interval, RPV radiomic prognostic vector, HH Hammersmith Hospital, TCGA the Cancer Genome Atlas
  2. aAge has been dichotomized at 60 years
  3. bCombining HH discovery and HH validation datasets
Back to article page