Cytoreductive radical prostatectomy versus no local treatment for patients with metastatic prostate cancer
Patients
Men with metastatic prostate cancer
Intervention
Cytoreductive radical prostatectomy
Comparison
No local treatment (androgen deprivation therapy with or without chemotherapy)
Outcome
Absolute effect*
Relative effect(95% CI)
Certainty of evidence (GRADE)
No local treatment
Cytoreductive radical prostatectomy
Difference: patients per 100
Cancer-specific mortality
54
26
HR 0.39(0.22 to 0.69)
⊕⊕⊕Ο1,2MODERATE
Difference: 28 fewer(Margin of error: 38 fewer to 13 fewer)
All-cause mortality
67
43
HR 0.50(0.44 to 0.57)
⊕⊕ΟΟ1LOW
Difference: 24 fewer(Margin of error: 28 fewer to 20 fewer)
Disease progression
55
27
HR 0.39(0.21 to 0.73)
⊕⊕ΟΟ1,2,3LOW
Difference: 28 fewer(Margin of error: 40 fewer to 11 fewer)
Severe local complications**
35
7
RR 0.20(0.06 to 0.64)
⊕⊕ΟΟ1,3,4,5LOW
Difference: 28 fewer(Margin of error: 33 fewer to 13 fewer)
Grade IIIb complications***
24
4
RR 0.18(0.02 to 1.36)
⊕⊕ΟΟ1,3,5,6LOW
Difference: 20 fewer(Margin of error: 23 fewer to 9 more)
Development of castration resistance
73
64
HR 0.78(0.48 to 1.27)
⊕ΟΟΟ1,3,4,6VERY LOW
Difference: 9 fewer(Margin of error: 26 fewer to 8 more)
Margin of error: 95% confidence interval (CI).HR: Hazard ratio.RR: Risk ratio.GRADE: Evidence grades of the GRADE Working Group (see later).*The risk with No local treatment is based on the risk in the control group of the observational study. The risk with Cytoreductive radical Prostatectomy (and its margin of error) is calculated from relative effect (and its margin of error).** Severe complications in the group with cytoreductive radical prostatectomy were stenosis and severe urinary incontinence, and in the group with no local treatment were ureter obstructions and urinary retentions requiring catheterization and/or transurethral resection of the prostate.*** Grade IIIb complications were measured using the Clavien-Dindo classification. These complications are related to additional interventions requiring general anesthesia.1 Observational studies2 The certainty of the evidence was raised one level because of the large effect size3 The certainty of the evidence was downgraded one level for publication bias since only small studies were included in this analysis.4 The certainty of the evidence was downgraded one level due to risk of bias since one of the studies was rated at moderate risk of bias using the Newcastle-Ottawa scale.5 The certainty of the evidence was raised two levels due to a very large effect size.6 The certainty of the evidence was downgraded one level for imprecision since the 95% CI includes a null benefit of the intervention in relation to the comparator.Control risk of dying at three years (all-cause mortality rate) = 67%Control risk of dying at three years (cancer-specific mortality rate) = 54%Control risk of dying or with evidence of disease progression at three years (disease progression rate) = 55.1%Control risk of dying or developing castration resistance at three years (development of castration resistance rate) = 73.2%Control risk of developing complications represented as severe local complications and grade IIIb complications in the group with no local treatment.
