This document summarizes several studies on prostate cancer screening and treatment that will be discussed at an American Urological Association panel. One study found that a single blood test before age 50 could predict long-term risk of prostate cancer death, with 44% of deaths occurring in men with above-average PSA levels. Another study found that while prostate cancer was rare in men with low PSA at age 60-70, continued screening could identify most high-risk cases. A third study found limitations to using PSA velocity but that closely following patients despite initial negative biopsies may be important. The panel will discuss refining PSA use and interpretation to better determine which cancers require treatment.
1. EMBARGOED FOR RELEASE UNTIL MONDAY, MAY 16, 2011 AT 9:00 A.M.
Contact: Wendy Waldsachs Isett, AUA
410-977-4770, wisett@AUAnet.org
PROSTATE CANCER: TO TREAT, NOT TO TREAT AND WHEN TO TREAT?
Panel to Address Key Questions about Low-Risk Prostate Tumor Management
Washington, DC, May 16, 2011—As physicians and researchers debate the merit of the prostate-specific antigen
(PSA) test, questions have arisen about the test’s ability to accurately identify the presence of prostate cancer,
as well as how the test may be interpreted and better used to determine which prostate cancers require
treatment and which do not. New research presented at the 2011 AUA Annual Meeting will bring light to the
innovative possibilities for the use of PSA. A special panel, to be held on Monday, May 16, 2011 at 9:00 a.m., will
discuss with members of the media the following studies:
How Soon Can We Identify Men at High Risk for Prostate Cancer Death? An Early Surveillance Strategy for
Prostate Cancer (#986): A single blood test before the age of 50 could predict a man’s long-term risk of prostate
cancer death, according to research from Memorial Sloan-Kettering Cancer Center in New York and Lund
University in Malmo, Sweden. Using data from the Preventive Project, a cardiovascular study enrolling men ages
33 to 50 between 1974 and 1986, and a combination of case-note review or death certificate data, researchers
identified 141 men in the study who had subsequently died of prostate cancer. Nearly half (44 percent) of the
deaths occurred in men whose PSA score fell in the top 10 percent (≥1.5 ng/ml). Researchers expanded their
analysis to the top quartile of men, measuring free PSA and human glandular kallikrein 2 (hK2), and found that
these markers helped to identify an additional 2.4 percent of deaths in the top 10 percent of risk. These data
suggest that early analysis of PSA, free PSA and hK2 may provide critical insight into a man’s risk of developing
aggressive, life-threatening disease, enabling urologists to better assess when early intervention may be
necessary.
Can a Single PSA Measurement at Age 60-70 Years Identify Men Who Need No Further Prostate Cancer
Testing? (#2025): Eliminating prostate cancer testing after the age of 60 may be an option for some men, but
others could benefit from continued testing, according to new data being presented by Johns Hopkins
researchers. Using data from the Baltimore Longitudinal Study of Aging, researchers identified 448 men with
PSA measurements between the ages of 60 and 70, including 199 with a PSA less than 1 ng/ml. They reviewed
PSA trajectory and its relationship to later diagnosis of prostate cancer (including high-risk disease, defined by
PSA greater or equal to 20 ng/ml, Gleason 8-10 or confirmed prostate cancer death). In the 199 men with low
PSA (median age of 61.9 at time of test), 13 were later diagnosed – four with significant disease. These data
indicate that it may not be advisable to apply a universal cut-off point for PSA testing.
Possible Pitfalls in Using Prostate-Specific Antigen Velocity for Detection of Prostate Cancer (#2032): Prostate-
specific antigen velocity (PSAV) can be a strong derivative in improving the performance of the PSA blood test as
a marker for prostate cancer but has limited sensitivity and specificity, according to new data from researchers
2. at Northwestern University. Researchers examined patients in two categories: those with elevated PSAV and no
cancer on biopsy, and those with low PSAV with biopsy-detected cancer. Of those patients with low PSAV with
biopsy-detected disease, 4.6 percent had a Gleason 8-10 tumor, 30 percent had slow-growing tumors with a
Gleason score less than 6 and 54 percent had Gleason 7 tumors. Of those patients with elevated PSAV and
negative biopsy, 58 percent were later diagnosed with biopsy-detected disease, suggesting a need to closely
follow a patient’s PSA despite an initial negative biopsy.
The Worst Cancers Send Early PSA Signals that Would Allow Early Detection if Monitoring Focused on
Increasing PSA (#1197): PSAV and its rate of increase over time may be a key marker in identifying aggressive
disease and could provide valuable insight in how to interpret the PSA test, according to researchers from
Medical University Innsbruck in Austria and the University of California, San Francisco. Study authors analyzed
pre-diagnosis PSA history from 94 prostate cancer patients who, following surgical treatment, suffered extra-
capsular extension (EE) or recurrence, and calculated annual growth rate in cancer PSA for each. Of the men
with EE or recurrence, 98.9 percent had an increasing or constant PSAV and 95 percent had a PSA annual growth
rate of 10 percent or more, suggesting that, in men with a current PSA less than 4.0 ng/ml, increased scrutiny of
annual growth rates of 1.0 ng/ml or more may be warranted.
“Debate has been ongoing about the use of the PSA test in the detection of prostate cancer but the question
may not be whether we use the test but, rather, how we use it,” said Christopher Amling, MD, who moderated
the briefing for media. “These studies shed important light on how we might refine our use and interpretation of
the PSA test.”
NOTE TO REPORTERS: Experts are available to discuss this study outside normal briefing times. To arrange an
interview with an expert, please contact the AUA Communications Office at the number above or e-mail
wisett@AUAnet.org.
About the American Urological Association: Founded in 1902 and headquartered near Baltimore, Maryland, the American
Urological Association is the pre-eminent professional organization for urologists, with more than 17,000 members
throughout the world. An educational nonprofit organization, the AUA pursues its mission of fostering the highest standards
of urologic care by carrying out a wide variety of programs for members and their patients.
###
3. 986
HOW SOON CAN WE IDENTIFY MEN AT HIGH RISK FOR PROSTATE CANCER DEATH? AN EARLY
SURVEILLANCE STRATEGY FOR PROSTATE CANCER
Andrew Vickers*, Caroline Savage, New York, NY, Thomas Bjork, Axel Gerdtsson, Jonas Manjer, Peter Nilsson,
Anders Dahlin, Anders Bjartell, Malmo, Sweden, Peter Scardino, David Ulmert, Hans Lilja, New York, NY
INTRODUCTION AND OBJECTIVES: The Preventive Project (MPP) in Malmö, Sweden was a cardiovascular study
enrolling a representative cohort (74% participation) of men aged 33 - 50 in 1974 – 1986. Until recently, rates of
prostate-specific antigen (PSA) testing in Sweden have remained low; retrieval and analysis of archived blood
plasma from MPP was previously used to demonstrate a strong association between PSA levels measured at age
44-50 and risk of advanced prostate cancer diagnosis up to 30 years subsequently. Our current objective was to
determine the degree of risk concentration, that is, the proportion of cases found in men with the highest PSA
levels.
METHODS: Death from prostate cancer was determined by case note review (74%) or from death certificate data
(26%). A nested case-control design was used, with three controls matched to each death. Lorenz curve
methodology was used to determine risk concentration.
RESULTS: The median follow up was 27
years. A total of 141 men died from prostate
cancer. A single PSA at age 44 – 50 was
strongly predictive of subsequent prostate
cancer death at a median follow-up of 27
years (area-under-the-curve: 0.72). The
figure shows the Lorenz curve for risk
concentration: 44% of deaths occurred in
men with the top 10% of PSA (≥ ≈1.5 ng /
ml). As alternative strategy, we took the top
quartile of men (PSA ≥ ≈1 ng / ml), measured
free PSA and hK2, and took those at highest
risk on the basis of a combined model of all
markers. Doing so would lead to
identification of an additional 2.4% of deaths
in the top 10% of risk. We repeated all
analyses using metastasis as the endpoint
(215 events); the results were similar: for example, 42% of metastases occurred in top 10% of PSA with an
additional 2.5% metastases were identified by free PSA and hK2.
CONCLUSIONS: Long-term risk of prostate cancer death can be predicted on the basis of a single blood test before
the age of 50. Targeting a small proportion (10%) of men in an early intervention strategy – regular PSA screening
with compliance monitoring, and possibly chemoprevention – could prevent almost half of all prostate cancer
deaths.
Figure. Lorenz curve for death from prostate cancer based on a single PSA at age 44 - 50. The x axis shows the
percentage of the population with PSA above the indicated levels, hence the percentages run from 100 down to 0.
The y axis shows the number of deaths that would be included (or missed) if we consider only men with PSA above
any given level.
Source of Funding: Supported by the National Cancer Institute (grant numbers R33 CA 127768-02 and P50-
CA926290); the Swedish Cancer Society (3455); the Swedish Research Council Medicine (20095); Fundacion Federico
SA, The Tegger Foundation, and the Sidney Kimmel Center for Prostate and Urologic Cancers.
4. 2025
Can a Single PSA Measurement at Age 60-70 Years Identify Men Who Need No Further PSA
Testing?
Stacy Loeb*, John B. Eifler, E. Jeffrey Metter, Luigi Ferrucci, H. Ballentine Carter, Baltimore, MD
INTRODUCTION AND OBJECTIVES: For men ages 50 to 69, there is level 1 evidence that serial PSA
testing reduces prostate cancer-specific mortality. However, Vickers et al. (BMJ 2010) recently
suggested in a case-control analysis that men with a single PSA measurement <1 ng/ml at age 60
may not require additional screening. This practice might be reasonable if these men are not at risk
for the development of prostate cancer (particularly, significant disease); whereas, if additional
screening would help identify clinically significant prostate cancer continued PSA testing might be
beneficial.
METHODS: From the Baltimore Longitudinal Study of Aging, we identified 448 men with PSA
measurements at age 60-70 years, including 199 with a PSA level <1 ng/ml. In these men we
examined the PSA trajectory and its relationship to the later diagnosis of overall and high-risk
prostate cancer (defined as PSA >20 ng/ml, Gleason score 8 to 10, or confirmed prostate cancer
death).
RESULTS: In the overall population, the median age was 62.2 years (range, 60.1-69.9) at the PSA
test, and the median PSA level was 1.1 ng/ml (range, 0.05-18.6). In the 199 men with a PSA <1
ng/ml, the median age was 61.9 years (60.1-69.7) at the PSA test. Of these men, 13 (6.5%) were
later diagnosed with prostate cancer, of which 4 (31%) were high-risk. Figure 1 shows the PSA
trajectory in the non-cancer, non high-risk cancer, and high-risk groups.
CONCLUSIONS: Although high-risk prostate cancer was infrequent in men with a PSA <1 ng/ml at
age 60 to 70, continued PSA screening could have identified the majority of these cases.
Source of Funding: This research was supported in part by the Intramural Research Program of the
NIH, National Institute on Aging.
5. 2032
POSSIBLE PITFALLS IN USING PROSTATE SPECIFIC ANTIGEN VELOCITY FOR DETECTION OF
PROSTATE CANCER
Gregory Auffenberg*, Joshua Meeks, Phillip R. Cooper, Qiaoyan Hu, Cheng Li, William J.
Catalona, Chicago, IL
INTRODUCTION AND OBJECTIVES: Prostate Specific Antigen (PSA) is a well-documented marker
for prostate cancer (CaP). PSA velocity (PSAV) is a derivative that has been used to increase PSA
performance characteristics for CaP detection. Elevated PSAV may be an early sign of underlying
CaP, but has somewhat limited sensitivity and specificity. Explanations for false-positive PSAV
include missed CaP on biopsy; prostatitis; or assay standardization bias. Reasons for false-
negative PSAV include very low volume CaP or high grade CaP (severely de-differentiated cells
produce less PSA); prior, resolved prostatitis; or assay standardization bias.
METHODS: We assessed patients who underwent prostate biopsy after prior screening with ≥3
PSA blood tests. PSAV was calculated using linear regression. Elevated PSAV was defined as
change in PSA ≥0.35 ng/mL/yr. After stratifying based on PSAV, we specifically examined: 1)
patients with an elevated PSAV but no CaP on biopsy and 2) patients with low PSAV but CaP on
biopsy.
RESULTS: Between 2003 and 2010, 1358 patients underwent prostate biopsy after previous
screening with ≥3 PSA tests. 106 had PSAV ≥0.35 ng/mL/yr and negative biopsy. Of these, 62
(58%) had CaP on biopsy at a later date, suggesting negative biopsy at the time of elevated PSAV
may have been due to missed early, low volume CaP. 10 (9.4%) biopsies revealed prostatic
intraepithelial neoplasia (PIN), also suggesting a possibly missed small CaP. 12 biopsies (11.3%)
had signs of prostatitis. In the remaining 22 (20.7%), cause for elevated PSAV was not readily
apparent; transient inflammation, prostatic trauma, or assay standardization bias may have
played a role. 480 patients with PSAV <0.35 had CaP on biopsy. Of these, 22 (4.6%) had a
Gleason 8-10 tumor. 464 were treated with radical prostatectomy. Surgical pathology for 266
(55.4%) revealed total CaP volume <5% of submitted tissue. Of the remaining patients,
30% had possibly slow growing Gleason <6 tumors involving 6-20% of submitted tissue, and 54%
had Gleason 7 tumors. In some cases, assay standardization bias or prior resolved prostatitis
may have caused the spuriously low PSAV.
CONCLUSIONS: Nearly 80% of patients with false-positive PSAV had an easily explained
confounder – missed CaP or prostatitis. Nearly 60% of patients with false-positive PSAV were
eventually found to have CaP, underscoring the need to closely follow a patient’s PSA despite
initial negative biopsy. False-negative tests can often be explained by low volume or high
Gleason grade disease, prior resolved prostatitis, or possible PSA assay standardization bias.
Source of Funding: Supported in part by the Urological Research Foundation, Prostate SPORE
grant (P50 CA90386-05S2) and the Robert H. Lurie Comprehensive Cancer Center Grant (P30
CA60553)
6. 1197
The Worst Cancers Send Early PSA Signals that Would Allow Early Detection if Monitoring Focused on
Increasing PSA
Jasmin Bektic*, Innsbruck, Austria, Peter Carroll, Matthew Cooperberg, San Francisco, CA, Helmut
Klocker, Eberhard Steiner,Viktor Skradski, Wolfgang Horninger, Tom Neville, Innsbruck, Austria
INTRODUCTION AND OBJECTIVES: We hypothesized that most cancers with extra-capsular extension
(EE) and/or recurrence (R) after treatment could have been caught earlier if monitoring focused on
increasing PSA.
METHODS: We analyzed 94 cancer patients treated with surgery from the Tyrol Screening Project, UCSF
and CaPSURE who had EE/R after treatment and had adequate PSA history before diagnosis (at least 5
tests over 4 years) that started at a PSA of 3.0 or less. The increase in PSA from low to a pre-biopsy high
was measured. The annual growth rate in cancer PSA (PSAgr) was estimated using a consistent
exponential trend plus no-cancer baseline. Trend curvature was assessed as increasing, constant
(linear) or decreasing PSA velocity. Hard
to react to sudden jumps in PSA were
noted if PSAgr exceeded 250% in the last
year before detection.
RESULTS: Median increase in PSA was 3.4,
FIG 1, with 100% 0.3 or more, 95% 1.0 or
more and 80% 2.0 or more. Median PSAgr
was 43%, FIG 2, with 95% PSAgr 10% or
more and 80% PSAgr 19% or more.
Curvature was ambiguous for 8.7%.
Trends for the remaining men showed:
92.9% with increasing PSAV, 6.0%
constant PSAV (linear) and only 1.2% (1
man) had decreasing PSAV. Jump was
ambiguous for 1.1%. Of the remaining
men, only 2.2% had a jump and 97.8% did
not.
CONCLUSIONS: The worst cancers (EE/R)
usually send a strong PSA signal that could
have been detected early with monitoring
for increasing PSA. Only 2.2% had a jump
too steep for early detection. Only 5% had
a PSA increase of less than 1.0. The
dominant pattern of increasing PSA was
clear: 98.9% had increasing or constant
PSAV and 95% had a PSAgr of 10% or
more. For men with current PSA less than 4.0, these results suggest increased screening scrutiny
whenever PSA increases by 1.0. If only two or three years of PSA tests are available then screening
intensity should increase whenever PSAV exceeds 0.20. At a median PSAgr of 43% for the worst cancers,
a PSAV of 0.20 is consistent with an exponential increase in PSA of 1.0 from cancer over ten years.
Source of Funding: none