Skip to main content

Table 3 Factors associated with overall survival (OS) in canine invasive mammary carcinomas (n = 150)

From: The dog as a naturally-occurring model for insulin-like growth factor type 1 receptor-overexpressing breast cancer: an observational cohort study

Criteria

OS: Univariate analysis

OS: Multivariate analysis

(log-rank test) N = 150

(Cox regression model) N = 150

HR

95 % CI

p-value

HR

95 % CI

p-value

Age

  

0.005

  

0.002

<11 yrs

1.00

-

 

1.00

-

 

≥11 yrs

1.66

1.17–2.37

 

1.79

1.24–2.60

 

Multifocality

  

0.04

  

0.96

Unifocal

1.00

-

 

1.00

-

 

Multicentric

1.89

1.01–3.54

 

1.02

0.51–2.04

 

Lymph node status

  

<0.001

   

N0

1.00

-

 

-

-

-

N1

3.53

1.79–6.93

    

Histological grade

  

0.006

  

0.53

Grade I

1.00

-

-

1.00

-

-

Grade II

1.67

0.93–2.99

0.09

1.33

0.70–2.53

0.38

Grade III

2.37

1.35–4.18

0.003

1.42

0.77–2.62

0.26

Lymphovascular invasion

  

<0.001

  

0.01

No LVI

1.00

-

 

1.00

-

 

LVI

2.53

1.78–3.60

 

1.71

1.14–2.56

 

Surgical margins

  

<0.001

  

0.006

Complete excision

1.00

-

 

1.00

-

 

Incomplete excision

2.32

1.62–3.33

 

1.81

1.18–2.75

 

Muscle infiltration

  

0.001

  

0.77

No

1.00

-

 

1.00

-

 

Yes

1.86

1.27–2.71

 

1.07

0.70–1.63

 

Peritumoral Inflammation

  

0.02

  

0.04

No

1.00

-

 

1.00

-

 

Yes

1.53

1.08–2.17

 

1.53

1.03–2.29

 

ER

  

0.03

   

≥ 10 %

1.00

-

 

-

-

-

< 10 %

1.48

1.04–2.11

    

Immunophenotype

  

0.03

  

0.17

Luminal

1.00

-

 

1.00

-

 

Triple negative

1.54

1.05–2.26

 

1.34

0.88–2.04

 

IGF1R

  

<0.001

  

0.002

weak (0–1+)

1.00

-

-

1.00

-

-

moderate (2+)

1.31

0.81–2.10

0.27

1.40

0.85–2.33

0.19

strong (3+)

2.62

1.63–4.20

<0.001

2.74

1.63–4.62

0.002

  1. Univariate (log rank test) and multivariate survival analyses (Cox proportional hazard regression)
  2. HR Hazard Ratio, 95 % CI 95 % Confidence Interval, ER Estrogen Receptor, IGF1R Insulin-like Growth Factor type 1 Receptor, LVI Lymphovascular Invasion
  3. When several significant prognostic factors overlapped, only one was selected for the multivariate analysis (LVI was chosen between lymph node status and LVI because it could have been determined in all cases and immunophenotype was preferred to ER expression)