Skip to main content

Table 4 Univariate and multivariate Cox models to identify associations between clinical/ radiological factors and OS

From: Prognostic value of the tumor-to-liver density ratio in patients with metastatic colorectal cancer treated with bevacizumab-based chemotherapy. A post-hoc study of the STIC-AVASTIN trial

 

Univariate analysis (n = 79)

Multivariate analysis (n = 66)

HR [95%CI]

p

HR [95%CI]

p

Age

 

0.349

  

< 60 years

1

   

[60 years; 70 years[

0.98 [0.55–1.77]

   

≥ 70 years

1.53 [0.80–2.90]

   

ECOG-PS

 

0.195

  

0

1

   

1

1.40 [0.84–2.35]

   

Primary tumor site

 

0.125

  

Colon

1

   

Rectum

1.57 [0.90–2.74]

   

Metastases

 

0.034

 

0.041

Synchronous

1

 

1

 

Metachronous

0.52 [0.28–0.99]

 

0.53 [0.28–1.00]

 

Mean tumor density

 At baseline

 

0.065

  

 ≤ 63 HU

1

   

 > 63 HU

0.62 [0.37–1.03]

   

 At the first restaging

 

0.167

  

 ≤ 63 HU

1

   

 > 63 HU

0.68 [0.39–1.19]

   

TTLD ratio

 At baseline

 

0.024

  

 ≤ 0.6

1

   

 > 0.6

0.55 [0.33–0.92]

   

 At the first restaging

 

0.064

  

 ≤ 0.6

1

   

 > 0.6

0.61 [0.36–1.04]

   

 Morphologic response1

 

0.334

  

 1

1

   

 2 or 3

1.32 [0.75–2.29]

   

 RECIST-151

 

0.031

 

0.037

 < -15%

1

 

1

 

 ≥ -15%

1.79 [1.06–3.00]

 

1.75 [1.05–2.94]

 

 TTLD ratio change1

 

0.663

  

 < -10%

1

   

 ≥ -10%

1.12 [0.67–1.85]

   
  1. TTLD Tumor-to-liver density, ECOG-PS Eastern Cooperative Oncology Group – Performance, RECIST-15 Response Evaluation Criteria in Solid Tumors using a modified cut-off value of -15%
  2. 1Recorded at the first restaging
Back to article page

Cancer Imaging

ISSN: 1470-7330

Contact us