EURURO Vol. 72 No. 6

6–8 wk after surgery (all

0.02). Moreover, although no

differences were observed in the use of ADT immediately

after surgery between patients in the no RT and postopera-

tive RT groups (21% vs 23%;

= 0.3), the use of late ADT at

progression significantly differed between the two groups

(29% vs 53%;

0.001).

3.2.

Uni- and multivariable analyses predicting CSM

Median follow-up for survivors was 110 mo (interquartile

range: 98–121). Overall, 49 and 77 patients experienced

CSM and OCM, respectively. The resulting 10-yr CSM-free

survival rate was 88%

( Fig. 1 )

. At multivariable analyses,

pathologic grade group 4 (hazard ratio [HR]: 2.72; 95%

confidence interval [CI]: 1.43–5.14;

= 0.01) and pT3b/4

tumor stage (HR: 2.34; 95% CI: 1.21–4.49;

p =

0.01) were

independently associated with CSM

( Table 2

). Pathologic

grade group, pathologic stage, nodal status, surgical

margins, and immediate ADT were included in a model

to predict the 10-yr CSM risk in patients with PSA

persistence after RP. The coefficients to calculate the risk

of CSM are depicted in Supplementary Table 1. A novel

nomogram was then developed to facilitate individual

estimation of the risk of CSM at 10-yr follow-up (Supple-

mentary Fig. 1). At internal validation, the discrimination

accuracy of this model based on pathologic characteristics

and administration of immediate ADT was 67% in our

cohort. Supplementary Figure 2 depicts the calibration

plot.

3.3.

Effect of PSA levels at 6–8 wk after RP on CSM

At univariable Cox regression analyses, the level of

detectable PSA as measured at 6–8 wk after RP was

significantly associated with the risk of CSM (HR: 1.72; 95%

CI: 1.07–2.76;

= 0.02) for the overall cohort, such that we

Table 1 – Descriptive statistics of 496 patients with clinically localized prostate cancer treated with radical prostatectomy and extended

pelvic lymph node dissection between 1994 and 2014, who experienced prostate-specific antigen (PSA) persistence

Overall

No RT

Postoperative RT

value

(

= 496)

(

= 245, 49.4%)

(

= 251, 50.6%)

Year of surgery

Median (IQR)

2004 (1998–2010)

2000 (1996–2009)

2005 (2000–2010)

0.001

Age at surgery (yr)

Median (IQR)

64 (58–68)

65 (59–69)

63 (57–67)

0.002

Preoperative PSA (ng/ml

) a

Median (IQR)

8.5 (5.5–14.3)

7.9 (5.1–13.5)

9.2 (5.9–15.2)

0.02

Preoperative D’Amico risk group (%)

Low

81 (16)

50 (20)

31 (12)

0.001

Intermediate

230 (46)

125 (51)

105 (42)

High

185 (37)

70 (29)

115 (46)

Surgical technique (%)

ORP

409 (83)

207 (85)

202 (81)

0.1

RARP

87 (18)

38 (16)

49 (20)

Grade group at final pathology (%)

131 (26)

96 (39)

35 (14)

0.001

115 (23)

50 (20)

65 (26)

86 (17)

36 (15)

50 (20)

46 (9.3)

21 (8.6)

25 (10)

118 (24)

42 (17)

76 (30)

Pathologic stage (%)

231 (47)

138 (56)

93 (37)

0.001

T3a

124 (25)

52 (21)

72 (28)

T3b

134 (27)

52 (21)

82 (33)

7 (1.4)

3 (1.2)

4 (1.6)

pN1 (%)

114 (23)

42 (17)

72 (29)

0.01

Positive surgical margins (%)

246 (50)

88 (36)

158 (63)

0.001

Number of removed lymph nodes

Median (IQR)

9 (5–15)

9 (5–12)

9 (6–18)

0.06

Number of positive lymph node

s b

Median (IQR)

2 (1–5)

3 (1–5)

0.4

First PSA after surgery

Median (IQR)

0.3 (0.2–0.6)

0.2 (0.2–0.5)

0.3 (0.2–0.8)

0.001

ADT concomitant to RT (%)

58 (12)

–

58 (23)

Immediate ADT without RT (%)

51 (10)

51 (21)

–

Late ADT at progression (%)

204 (41)

71 (29)

133 (53)

0.001

IQR = interquartile range; ADT = androgen deprivation therapy; ORP = open radical prostatectomy; RARP = robot-assisted radical prostatectomy; NA = not

applicable; RT = radiotherapy.

ADT during RT in the RT group.

Immediate ADT after surgery in the no RT group.

Salvage ADT at progression.

Missing in 32 patients.

In node-positive patients.

E U R O P E A N U R O L O G Y 7 2 ( 2 0 1 7 ) 9 1 0 – 9 1 7

912