PRETREATMENT PSA VELOCITY AS A PREDICTOR OF DISEASE OUTCOME FOLLOWING RADICAL RADIATION THERAPY

1
PRETREATMENT PSA VELOCITY AS A PREDICTOR OF
DISEASE OUTCOME FOLLOWING RADICAL RADIATION
THERAPY
D. Palma, S. Tyldesley, P. Blood, M. Liu, G.
Duncan, J. Morris, T. Pickles, Prostate Cohort
Outcomes Initiative British Columbia Cancer
Agency, British Columbia, Canada
In the entire cohort, the two PSAV groups did not
differ in DSS (Figure 4) or OS (Figure 5)
PSAV 2
Median bDFS 97 mo
The median age at diagnosis was 72 years, and
median follow-up for survival was 7.6 years.
Median PSA prior to treatment was 8.2 ng/mL. The
median PSAV was 0.3 ng/mL per year. 301 patients
had a PSAV 2 and 172 had a PSAV gt2 (Figure 1).
Patients with prostate cancer undergoing external
beam radiation therapy (EBRT) are stratified into
risk groups by pre-treatment risk factors,
including Gleason score, tumor stage, and PSA.
The importance of pre-treatment PSA velocity
(PSAV) for patients undergoing EBRT is less
certain. A study of 358 men by DAmico et al1
demonstrated reduced bDFS, DSS, and OS in
patients with a PSAV gt2 ng/mL per year, compared
to those with a PSAV 2 ng/mL per year. This has
not yet been independently validated.
Survival
PSAV gt 2
Median bDFS 68 mo
Log-rank p 0.003
Time (years)
Figure 2. bDFS by PSAV group
Log-rank p 0.55
Log-rank p 0.99
On multivariate analysis, PSAV was no longer a
significant predictor of bDFS (Table 1)
Figure 4. DSS by PSAV group
Figure 5. OS by PSAV group
PSAV Histogram
PSAV 2
PSAV gt2
A pretreatment PSAV greater than 2 ng/mL in the
year prior to treatment is associated with a
reduced bDFS. When other pretreatment risk
factors are taken into account, PSAV is no longer
a significant predictor of bDFS in the entire
cohort. However, in men with high-risk disease,
PSAV did predict bDFS on univariate and
multivariate analysis. PSAV does not predict DSS
or OS in this cohort.
301
172
Patients
To determine if a pretreatment PSAV gt2 ng/mL per
year is associated with bDFS, DSS, or OS in men
undergoing radical RT for prostate cancer in
British Columbia.
PSAV
Table 1. Multivariate analysis of predictors of
bDFS.
Figure 1. Distribution of PSAV values (ng/mL/yr)
In men with high-risk disease (Figure 3), a PSAV
gt2 was associated with reduced bDFS on univariate
and multivariate analysis, but there were no
differences in DSS or OS.
An annual PSAV gt2 ng/mL was significantly
associated with a higher pre-treatment PSA
(plt0.001) and higher risk group (p0.002).
473 patients with T1-T4/NX/MX prostate cancer
treated with radiation therapy between 1993 and
2001 formed the study cohort. No men received
neoadjuvant or adjuvant hormones. PSAV was
calculated using two or more PSA levels in the
year prior to treatment. Biochemical failure was
defined as a PSA rise to a level of 2 ng/mL or
more above the post-EBRT nadir PSA (Phoenix
definition).
Time to biochemical recurrence differed
significantly between the two groups (Figure 2).
Gleason score, tumor stage and pretreatment PSA
were also significantly associated with bDFS
(plt0.001).
Log-rank p 0.002
1. D'Amico, A.V., et al., Pretreatment PSA
velocity and risk of death from prostate cancer
following external beam radiation therapy. JAMA
2005. 294(4) p. 440-7.
PSAV 2
Median bDFS 68 mo
Survival

• Contact info
• David Palma
• BC Cancer Agency
• 600 West 10th Ave Vancouver BC, V5Z 4E6
• dpalma_at_bccancer.bc.ca

PSAV gt 2
Median bDFS 38 mo
Time (years)
Figure 3. bDFS for high-risk patients by PSAV
group