Table 3 Univariate and multivariate Cox proportional hazards regression analyses of clinical variables associated with recurrence risk.

From: Preoperative radiomics models for predicting risks of distinct recurrence patterns in pancreatic ductal adenocarcinoma based on contrast enhanced CT

Characteristics

Univariate analysis

Multivariate analysis

Hazard ratio (95% CI)

P value

Hazard ratio (95% CI)

P value

Age (years)

0.99 (0.97 – 1.01)

0.351

  

BMI (kg/m2)

1.00 (0.93 – 1.06)

0.887

  

WBC (× 10⁹/L)

0.93 (0.85 – 1.02)

0.116

  

MCV (fL)

1.00 (0.99 – 1.01)

0.759

  

MCH (pg)

1.00 (1.00 – 1.00)

0.791

  

MCHC (g/L)

1.01 (1.00 – 1.02)

0.118

  

Platelet (× 10⁹/L)

1.00 (1.00 – 1.00)

0.208

  

Lymphocyte (× 10⁹/L)

1.00 (0.96 – 1.05)

0.859

  

Neutrophil (× 10⁹/L)

0.93 (0.87 – 0.99)

0.046

1.10 (0.96 – 1.25)

0.157

Eosinophils (× 10⁹/L)

1.02 (0.91 – 1.15)

0.726

  

Basophil granulocyte (× 10⁹/L)

1.00 (0.99 – 1.02)

0.536

  

Mononuclear cell count (× 10⁹/L)

1.14 (0.95 – 1.38)

0.170

  

Platelet volume (fL)

1.00 (0.97 – 1.03)

0.863

  

Platelet distribution width (%)

0.97 (0.92 – 1.03)

0.329

  

AST (U/L)

1.00 (1.00 – 1.00)

0.357

  

Total bilirubin (μmol/L)

1.00 (1.00 – 1.00)

0.357

  

Direct bilirubin (μmol/L)

1.00 (1.00 – 1.00)

0.189

  

AFP (ng/mL)

0.99 (0.94 – 1.04)

0.740

  

Carcinoembryonic antigen (ng/mL)

1.01 (1.00 – 1.02)

0.088

  

Ventosity

No

 

1.00 (Reference)

  

Yes

1.10 (0.84 – 1.45)

0.488

  

Diabetes mellitus

    

No

 

1.00 (Reference)

 

1.00 (Reference)

Yes

0.76 (0.55 – 1.05)

0.094

0.60 (0.42 – 1.05)

0.107

Adjuvant therapy

No

 

1.00 (Reference)

  

Yes

1.26 (0.92 – 1.71)

0.144

  

CA19-9 (U/mL)

    

 ≤ 150

 

1.00 (Reference)

 

1.00 (Reference)

150–1000

1.22 (0.91 – 1.63)

0.181

1.25 (0.89 – 1.75)

0.195

 > 1000

1.51 (0.99 – 2.29)

0.055

1.72 (1.09 – 2.72)

0.019

Tumor size (cm)

 ≤ 2

 

1.00 (Reference)

  

2–4

1.42 (1.07 – 1.90)

0.017

  

 > 4

2.06 (1.01 – 4.22)

0.050

  

Tumor location

Head/Uncinate

 

1.00 (Reference)

  

Body/Tail

1.02 (0.76 – 1.36)

0.898

  

Bile duct dilatation

No

 

1.00 (Reference)

  

Yes

0.89 (0.68 – 1.16)

0.392

  

Pancreatic duct dilatation

No

 

1.00 (Reference)

  

Yes

1.09 (0.83 – 1.44)

0.545

  

AJCC stage

IA

 

1.00 (Reference)

 

1.00 (Reference)

IIA

1.64 (0.84 – 3.21)

0.149

1.28 (0.64 – 2.56)

0.488

IB

2.81 (1.31 – 6.01)

0.008

2.07 (0.93 – 4.63)

0.076

IIB

2.78 (1.43 – 5.40)

0.003

2.56 (1.28 – 5.12)

0.008

III

2.84 (1.42 – 5.70)

0.003

2.72 (1.31 – 5.63)

0.007

Surgical type

Open surgery

 

1.00 (Reference)

  

Laparoscopic surgery

1.38 (0.94 – 2.02)

0.100

  

Suspected vascular invasion

No

 

1.00 (Reference)

  

Yes

1.17 (0.86 – 1.59)

0.316

  

Suspected lymph node involvement

No

 

1.00 (Reference)

  

Yes

1.68 (1.28 – 2.22)

 < 0.001

  

Peripheral organ invasion

No

 

1.00 (Reference)

  

Yes

1.16 (0.87 – 1.55)

0.321

  

Tumor enhancement pattern

Peritumoral enhancement

 

1.00 (Reference)

 

1.00 (Reference)

No peritumoral enhancement

1.36 (1.03 – 1.79)

0.031

1.42 (1.04 – 1.94)

0.028

  1. Hazard ratios (HRs) with 95% confidence intervals (CIs) were estimated using Cox proportional hazards models. Univariate analyses were performed for each variable individually. Variables considered clinically relevant were subsequently entered into a multivariate Cox regression model using bidirectional stepwise selection, allowing for both forward inclusion and backward elimination. Variable retention in the final multivariate model was determined based on their independent contribution after adjustment for other covariates, rather than univariate statistical significance alone.
Back to article page