Skip to main content

Table 2 Multiadjusted comparison a of sociodemographic, treatment and surgeon characteristics among prostate cancer patients after prostatectomy (n = 1’254) according to surgical margin status

From: Determinants and effects of positive surgical margins after prostatectomy on prostate cancer mortality: a population-based study

Characteristics

Surgical margin status

Adjusted

P-value

Odds ratio

 

Positive

Negativeb

  
 

(cases)

(controls)

  
 

N

%

N

%

(95% CI)

 

Period

      

  1990-1994

29

50.0

29

50.0

1

0.025

  1995-1999

105

42.9

140

57.1

1.05 (0.53-2.06)

 

  2000-2004

201

40.8

292

59.2

1.47 (0.73-2.99)

 

  2005-2008

144

31.4

314

68.6

0.92 (0.44-1.92)

 

Type of surgery

      

  Perineal/Retropubic

196

45.9

231

54.1

1

0.443

  Laparoscopic

62

39.2

96

60.8

0.96 (0.61-1.53)

 

  Robot-assisted

35

32.1

74

67.9

0.96 (0.55-1.69)

 

  Unknown

186

33.2

374

66.8

0.77 (0.52-1.12)

 

Clinical T

      

  cT0-cT1

83

31.8

178

68.2

1

0.038

  cT2

163

35.7

293

64.3

0.93 (0.65-1.31)

 

  cT3

79

63.2

46

36.8

1.73* (1.05-2.85)

 

  cTx

154

37.4

258

62.6

1.18 (0.82-1.70)

 

Pathological T

      

  pT0-pT2

239

27.5

629

72.5

1

<0.001

  pT3-pT4

240

62.2

146

37.8

2.68 (2.00-3.59)

 

PSA level

      

  0-9

217

32.8

445

67.2

1

0.036

  10-20

98

51.0

94

49.0

1.52* (1.06-2.17)

 

  21+

57

64.0

32

36.0

1.80* (1.06-3.05)

 

  Unknown

107

34.4

204

65.6

1.03 (0.74-1.42)

 

Gleason

      

  2-6

218

30.5

497

69.5

1

<0.001

  7

195

45.2

236

54.8

1.61*** (1.23-2.12)

 

  8-10

60

61.9

37

38.1

2.25*** (1.39-3.63)

 

  Unknown

6

54.5

5

45.5

1.87 (0.46-7.71)

 

Tumour percentage c

     

  ≤10%

202

26.7

554

73.3

1

<0.001

  >10%

252

57.9

183

42.1

2.90*** (2.21-3.81)

 

  Unknown

25

39.7

38

60.3

1.37 (0.78-2.43)

 

Total

479

38.2

775

61.8

  
  1. Geneva Cancer Registry 1990–2008.
  2. aLogistic regression adjusted for all variables (clinical and pathological T status were entered separately, due to their co-linearity; the results presented derive from the model with clinical T).
  3. bNegative and close.
  4. cTumour percentage of the prostate was calculated as tumour volume divided by the prostate volume.
  5. CI: confidence interval.
  6. *p < 0.05; ***p < 0.001.