Skip to main content

Table 2 Multivariate analyses in models including significant univariate analyses for all patients (Cox regression, backward conditional)

From: Disease-specific outcomes of Radical Prostatectomies in Northern Norway; a case for the impact of perineural infiltration and postoperative PSA-doubling time

Characteristic BF (170 events)† CF (36 events) PCD (15 events)*
HR CI95% p HR CI95% p HR CI95% p
pT-stage    0.001 NS    NS   
  pT2 1         
  pT3a 1.70 1.14-2.54 0.010       
  pT3b 2.40 1.45-3.97 0.001       
Preop PSA     NE    NS   
  PSA < 10 1         
  PSA > 10 1.39 1.01-1.91 0.047       
Gleason    0.032    0.019    0.087
  3 + 3 1    1    1   
  3 + 4 1.05 0.70-1.56 0.81 2.45 0.78-6.90 0.09 3.71 0.41-33.2 0.242
  4 + 3 1.55 0.97-2.47 0.07 2.87 0.91-9.10 0.07 10.47 1.21-90.7 0.033
  4 + 4 1.42 0.68-2.97 0.36 2.73 0.52-14.2 0.23 7.43 0.46-121 0.159
  ≥9 2.39 1.31-4.35 0.004 6.74 2.21-20.6 0.001 15.26 1.65-141 0.016
PNI    0.090    0.012    0.047
  No 1    1    1   
  Yes 1.35 0.95-1.92   2.48 1.23-5.04   3.17 1.02-9.87  
Non-apical PSM±    0.003    0.002 NS   
  No 1    1      
  Yes 1.70 1.20-2.40   3.22 1.56-6.64     
Apical PSM±    0.031 NE    NS   
  No 1         
  Yes 0.69 0.49-0.97        
  1. Significant p-values in bold (threshold p ≤ 0.05).
  2. Abbreviations: BF biochemical failure, CF Clinical failure, LVI lymphovascular infiltration, NE not entered, due to non-significance in the univariate analyses, NS not significant, the characteristic is removed by the backward conditional analysis due to insignificance, PCD prostate cancer death, PNI Perineural infiltration, Post op RT postoperative radiotherapy, Preop preoperative, PSA Prostate specific antigen, PSM Positive surgical margin.
  3. Tumor Size, pN-stage and LVI were removed by the backward conditional model due to insignificance in all models.
  4. ±Only the subgroups (apical/non-apical PSM) of PSM were entered.
  5. *Due to the low number of events the model was carefully analyzed in advance with the inclusion and removal of variables in an enter analysis to find the most significant in advance before doing the final model with the three variables; Gleason score, perineural infiltration and positive non-apical margin.