Transcript PSA Testing 101

PSA Testing
Pitfalls & Utility
Stanley H. Weiss, MD, FACP
Professor of
Preventive Medicine and Community Health,
UMDNJ-New Jersey Medical School
Director & PI, Essex County Cancer Coalition
[email protected]
October 29, 2011
PSA Testing Pitfalls & Utility
Presentation prepared in collaboration with:
Daniel M. Rosenblum, PhD
Assistant Professor,
UMDNJ-New Jersey Medical School
Two uses for PSA test
• Screening for disease:
– in asymptomatic men &
– in symptomatic men
• Surveillance:
– in follow-up of patients previously
diagnosed with prostate cancer
2011-Oct-29
© Stanley H. Weiss, 2011
3
PSA Test – Screening Pros & Cons
• In my opinion, the standard PSA
blood test is not a very good
screening test for prostate cancer.
• It has limited sensitivity, specificity,
and predictive value.
2011-Oct-29
© Stanley H. Weiss, 2011
4
PSA Test – Screening Pros & Cons
• Nevertheless: The issues are —
–Can it save lives?
–If it saves lives,
is it cost-effective?
• Number needed to screen to save a life [NNS]
• Number needed to treat to save a life [NNT]
– How can we try to improve screening?
– How long will it be until we’ll “know” the
proposed new approach(es) will work?
2011-Oct-29
© Stanley H. Weiss, 2011
5
Probabilities of Death Due to PCa —
Control vs. Screening Ratios & Differences
Ratios: probability of death in
screening arm / probability of
death in control arm
Differences: probability of death in
control arm – probability of death
in screening arm
3.5000
0.0100
3.0000
0.0080
2.5000
0.0060
2.0000
0.0040
1.5000
0.0020
0.0000
1.0000
-0.0020
0.5000
Years: 0
0.0000
0
2011-Oct-29
5
Years 10
15
PLCO
ERSPC
Göteborg
© Stanley H. Weiss, 2011
5
10
15
PLCO non-comorbid
ERSPC piecewise exponential
6
What Are the Data?
How Should They Be Interpreted?
Concept
What is a
ratio?
Formula Problems
N2 / N1 At small N1
(denominator), can
be misleading
What is a
N1 – N2
difference?
2011-Oct-29
At small N, trends
may not be apparent
© Stanley H. Weiss, 2011
7
Probabilities of Death Due to PCa —
Control vs. Screening Ratios & Differences
Ratios: probability of death in
screening arm / probability of
death in control arm
Differences: probability of death in
control arm – probability of death
in screening arm
3.5000
0.0100
3.0000
0.0080
2.5000
0.0060
2.0000
0.0040
0.0020
1.5000
0.0000
1.0000
-0.0020
0.5000
Years: 0
0.0000
0
2011-Oct-29
5
Years 10
15
PLCO
ERSPC
Göteborg
© Stanley H. Weiss, 2011
5
10
15
PLCO non-comorbid
ERSPC piecewise exponential
8
Screening: 3 criteria
A good screening test:
1. must accurately identify
• persons with the condition (sensitivity —
if you have the disease, the test says so)
and
• those without (specificity — if you don’t
have the disease, the test says that too);
2. must identify those with the disease early
enough to make a difference in treatment
outcomes (survival and/or quality of life);
AND …
2011-Oct-29
© Stanley H. Weiss, 2011
9
Screening: 3 criteria (cont.)
3. preferably
• has high sensitivity (is very likely to indicate
you have the condition if you really do),
• has high specificity (is very likely to indicate
you don’t have the condition if you really
don’t), and
• must be
• acceptable for the population screened,
• rapid, and
• ideally, noninvasive.
2011-Oct-29
© Stanley H. Weiss, 2011
10
Screening for Prostate Cancer:
3 testing approaches
for detection
1. Digital Rectal Exam (DRE)
2. Prostate-Specific Antigen (PSA)
blood test
3. Trans-rectal ultrasonography
(TRUS)
2011-Oct-29
© Stanley H. Weiss, 2011
11
1. The Digital Rectal Exam (DRE) is a mainstay of
PCa screening, but the examiner cannot feel the
whole prostate nor small lesions
2011-Oct-29
© Stanley H. Weiss, 2011
12
2. Blood serum test – for ProstateSpecific Antigen (PSA) – takes a
small sample of blood and tests
for the amount of a substance
that is mainly produced by the
prostate gland
3. Trans-rectal ultrasonography
(TRUS) – limited value for
screening; history, exam or lab
findings may indicate a need for
TRUS evaluation
Additional details about each follow the presentation in the supplemental slides
2011-Oct-29
© Stanley H. Weiss, 2011
13
What are the risks & benefits of
screening?
• The value of screening can be proven only by
showing a reduction in the chance of dying of
prostate cancer without an unacceptable reduction in
quality of life (from both the screening and increased
use of treatments that can have negative side
effects).
• Conclusive evidence for the value of prostate cancer
screening has not yet accumulated.
• But some evidence does indicate that screening
offers the possibility to diagnose early prostate
cancer and to reduce deaths from this disease.
2011-Oct-29
© Stanley H. Weiss, 2011
14
Risks of Screening
• On the other hand, screening also detects
cancers that do not threaten the patient’s life.
• Finding such cases cannot be avoided at present.
• When screening the general population for PCa
by PSA, over 50% of the PCa’s detected will be
minimal cancers (Draisma 2003).
• As immediate treatment of these has not been
shown to be beneficial, detection and diagnosis
of some tumors may be unnecessary and
counter-productive, as in some patients there will
be treatment-associated morbidity (and, rarely,
even mortality).
2011-Oct-29
© Stanley H. Weiss, 2011
15
PSA: (prostate specific antigen)
PSA
• A protein
• Almost exclusively produced by the epithelial cells of
the prostate in normal and in pathologic conditions
such as infection, urinary retention, enlargement of the
prostate, and prostate cancer.
• Approximately 40% of patients with organ-confined
prostate cancer show no elevation of serum PSA.
• Unresolved: At what PSA value should more invasive
examinations — such as prostate biopsies — be
conducted? Not yet clear…
2011-Oct-29
© Stanley H. Weiss, 2011
16
Issues in the Optimal Implementation
of PSA Testing
• PSA – cutoff value issues:
Should it be age and/or race dependent?
Should clinical findings be incorporated into
testing algorithms as well?
• PSA – clinical issues: (fix)
(e.g., PSA rises with BPH and transiently with
acute prostatitis)
• Distinction between two uses of PSA testing:
– in symptomatic persons (routine clinical care) vs.
– prevention by population screening (main concern
of present discussion)
2011-Oct-29
© Stanley H. Weiss, 2011
17
Issues in the Optimal Implementation
of PSA Testing
• PSA velocity issues
Baseline value(s) – at what age?
How often to screen?
What is a “rapid” rise?
2011-Oct-29
© Stanley H. Weiss, 2011
18
The Clinician-Patient Conversation
In order to have an evidence-based discussion of
benefits, it is necessary to have data to support
screening.
In its absence, clinicians speculated loosely
about:
• potential benefit (“We might catch prostate
cancer early enough to save your life”) and
• potential harm (“Screening might result in
burdensome interventions with serious
complications”)
Part of this is paraphrased from Brett AS & Ablin RJ, “Prostate-Cancer
Screening — What the USPSTF Left Out.” N Engl J Med online 10/27/2011.
2011-Oct-29
© Stanley H. Weiss, 2011
19
The Clinician-Patient
Conversation (continued)
Needs to include a discussion of
morbidity & mortality associated with
screening, since:
• a positive test leads to additional
diagnostic tests and possibly treatment
that also carries associated risks,
• not just benefits
2011-Oct-29
© Stanley H. Weiss, 2011
20
Early Treatment – The Controversies
• A randomized trial has demonstrated that radical
prostatectomy can decrease the chance of dying of
prostate cancer as compared to delayed treatment.
• This benefit has been attributed in part to
suppression of the male hormone, testosterone.
• However, even in the delayed treatment group 8
years later, only ~ 25% are at risk of developing
metastatic disease.
•Thus, a majority of those with prostate
cancer die with, not from, the disease.
2011-Oct-29
© Stanley H. Weiss, 2011
21
Early Treatment – The Controversies
•Still impossible to determine up front which
cases will not progress.
•It has not been shown that the same favorable
results of surgery can be achieved when
prostate cancer has been detected by
screening.
•Uncertainty therefore remains.
• Active Surveillance —
which is more active than “Watchful Waiting” —
is a treatment option for low-Gleason score PCa that
reduces effects of aggressively treating every diagnosed
case
2011-Oct-29
© Stanley H. Weiss, 2011
22
Benefits of Screening
•Screening can find potentially lethal cancers at an early,
still curable stage as well as provide an opportunity for
earlier, and possibly life-prolonging, treatment of additional
tumors.
•Men who decide to be screened take a chance, and need
to be informed about – and balance – the potential risks
and benefits of screening and subsequent treatment. The
decision currently must be individualized, and men who
choose to be screened should not be denied the early
diagnostic tests.
•Data concerning cost efficacy – an important determinant
of public policy recommendations – are limited and
controversial.
2011-Oct-29
© Stanley H. Weiss, 2011
23
European Randomised Study of
Screening for Prostate Cancer (ERSPC)
www.erspc.org
•ERSPC established >10 years ago
•Largest randomized study (220,000 men
in eight western European countries*) on
screening for prostate ca
•Prostate cancer – 2nd leading cause of
cancer death in men in Western Europe
and the U.S.
* Netherlands, Sweden, Finland, Belgium, France, Spain, Italy and
Switzerland
2011-Oct-29
© Stanley H. Weiss, 2011
24
European Randomised study of
Screening for Prostate Cancer (ERSPC)
Provides some evidence-based advice to the pivotal
question – does screening lead to an improvement of
cancer-specific survival?
Initial results: 20 percent
reduction in the rate of death
from prostate cancer after first
ten years of follow-up
2011-Oct-29
© Stanley H. Weiss, 2011
25
Caveats – ERSPC
• The 20% reduction in rate of death due to prostate
cancer underestimates true effect since some
controls were also screened (“contamination”) and
some assigned to screening did not get screened
(“noncompliance”), but …
• Model-based adjustment to correct for
contamination only indicates that screening
reduces mortality by 27%, and …
• Model-based adjustment to correct for both
contamination and noncompliance indicates that
screening reduces mortality by 31%
Roobol MJ, et al., Eur Urol 2009:56(4);584-591
2011-Oct-29
© Stanley H. Weiss, 2011
26
More Caveats – ERSPC
• Initial data analysis indicates need to screen 1410
men and treat 48 with cancer to save 1 life, but …
• Further modeling* indicates numbers needed to
screen (NNS) & treat (NNT) to prevent 1 prostate
cancer death decrease over time:
Number of
Years Elapsed
8
9
10
11
12
NNS
NNT
2,506
1,254
837
628
503
86
43
29
22
18
Definition: The number needed to screen
(NNS) is the number of people who would
need to be screened to prevent one person
from dying.
(Similarly for the number needed to treat,
abbreviated NNT.)
NNS is computed by taking the reciprocal of
the difference between the fraction of people
who die in the screening arm and the fraction
of people who die in the control arm.
* Loeb S et al., J Clin Oncol 29(4):464-467, 2011
2011-Oct-29
© Stanley H. Weiss, 2011
27
More Caveats – ERSPC
• Limited # of biopsy samples
• Absence of A-A
• Limited follow-up
2011-Oct-29
© Stanley H. Weiss, 2011
28
Prostate, Lung, Colorectal, and Ovarian
(PLCO) Cancer Screening Trial
• Objective: The PLCO hopes to provide some answers
about the effectiveness of prostate cancer screening.
• Designed as a 17-year project of the National Cancer
Institute (NCI).
• An initial report appeared in the New England Journal of Medicine
online March 18, 2009 (in print – March 26 issue), coinciding with
presentation of the ERSPC data at the European Association of
Urology meeting in Stockholm, Sweden.
• PLCO: Six annual screenings for prostate cancer.
• FINDINGS: More diagnoses of the disease,
but did NOT lead to fewer prostate cancer deaths.
• BUT there are limitations to the study…
2011-Oct-29
© Stanley H. Weiss, 2011
29
Caveats – PLCO
• Wrong PSA cutoff
• > 40% screened within 3 yrs of enrollment
• Half of “non-screened controls” really got PSA
during trial, and 15% assigned to screening did
not get screened, greatly REDUCING its power —
only 33 percentage-point difference in screening
between control and screening arms*
• Relatively short follow-up — indeed, 2009
publication of results to that date mandated by
monitoring board, not by any achieved results
* Schröder FH & Roobol MJ, Eur Urol 2010;58(1):46-52
2011-Oct-29
© Stanley H. Weiss, 2011
30
Caveats – PLCO
• Most men not biopsied when advised (limiting
detection of early-stage PCa in screening arm)
• Limited & variable # of biopsies
• Variable care
• Very few A-A
2011-Oct-29
© Stanley H. Weiss, 2011
31
Caveat to Caveats – PLCO
• Despite the above weaknesses, in men with no or
minimal comorbidities, after ten years number
needed to screen was 723, number needed to
treat was 5 to prevent one prostate cancer death
• But men with comorbidities who were diagnosed with
PCa were less likely to receive curative treatment than
men without comorbidities, and hence perhaps more
likely to progress to death due to PCa, potentially
reducing their benefit from screening and reducing
the overall demonstrated value of screening in the
PLCO
Crawford ED et al., J Clin Oncol 29(4):355-361, 2011. Note that these authors
are mostly the same ones who initially reported the negative results of the
whole PLCO for prostate cancer screening.
2011-Oct-29
© Stanley H. Weiss, 2011
32
Randomized Screening Trials
2011-Oct-29
© Stanley H. Weiss, 2011
33
Summary of Concerns
• There are limitations to the protocols that may have
reduced the efficacy of screening
• Based on our knowledge of the epidemiology of
prostate cancer, in each trial the follow-up period is
too short to have expected to see the full potential
benefits, and much too short to calculate
cost/benefit ratios
• The US PLCO trial was significantly under-powered
to have been able to demonstrate a positive result
due to PSA screening of controls – reminiscent of
the classic epidemiologic limitations of the negative
cardiovascular “MRFIT” trial 2 decades ago
2011-Oct-29
© Stanley H. Weiss, 2011
34
The Göteborg Trial
• A smaller Swedish screening trial
consisting of roughly 20,000 men
designed with similar criteria as the
ERSPC
• Current data were reported following a
total of 14 years of screenings and
follow-up
Hugosson J et al., Lancet Oncol 2010(Aug); 11(8): 725-732
– Some of the data from the trial were
incorporated into the ERSPC
2011-Oct-29
© Stanley H. Weiss, 2011
35
The Göteborg Trial
• Final results indicate a roughly 50%
decrease in mortality from prostate cancer
in the screening group vs. the control
group.
– The benefit was greatest 10+ years from the
beginning of the trial
– Half of the attendees who died of PCa in the screening
group were diagnosed in the first round of screening,
and many of these men were 60+ years of age at entry
– The number of men from the control group who may
have received independent screening was not known or
was not included – a bias that may underestimate the
benefits of screening
– 293 screened and 12 diagnosed or treated to prevent 1
death from prostate cancer
2011-Oct-29
© Stanley H. Weiss, 2011
36
The Göteborg Results
Cumulative
Risk of
Diagnosis:
Cumulative
Risk of
Death:
2011-Oct-29
▼
© Stanley H. Weiss, 2011
The
approx
ten-year
point is
marked
by an
arrow
37
Probabilities of Death Due to PCa —
Control vs. Screening Ratios & Differences
Ratios: probability of death in
screening arm / probability of
death in control arm
Differences: probability of death in
control arm – probability of death
in screening arm
3.5000
0.0100
3.0000
0.0080
2.5000
0.0060
2.0000
0.0040
1.5000
0.0020
0.0000
1.0000
-0.0020
0.5000
Years: 0
0.0000
0
2011-Oct-29
5
Years 10
15
PLCO
ERSPC
Göteborg
© Stanley H. Weiss, 2011
5
10
15
PLCO non-comorbid
ERSPC piecewise exponential
38
Ratios: probability of death in screening arm /
probability of death in control arm
3.5000
PLCO
3.0000
PLCO noncomorbid
2.5000
ERSPC
2.0000
ERSPC
piecewise
exponential
Göteborg
1.5000
1.0000
0.5000
0.0000
0
2011-Oct-29
5
Years
10
© Stanley H. Weiss, 2011
15
39
Differences: probability of death in control arm –
probability of death in screening arm
PLCO
0.0090
0.0080
PLCO noncomorbid
0.0070
0.0060
ERSPC
0.0050
0.0040
0.0030
ERSPC
piecewise
exponential
0.0020
Göteborg
0.0010
0.0000
-0.0010
Years:
2011-Oct-29
0
5
10
© Stanley H. Weiss, 2011
15
40
Current Screening Guidelines
from Major U.S. Organizations
• American Urological Assoc Best Practice Statement
– Individual decision for those with 10yr life expectancy
– Baseline PSA at 40
– PSA at subsequent intervals based on PSA level and risk factors
• American Cancer Society
– Advises against routine screening
– PSA should be offered as option
• Age 45 in those with risk factors (FH, AA)
• Age 40 in those at highest risk (multiple family members or a family
member diagnosed at a young age)
• US Preventive Services Task Force
– Do not screen routinely over age 75
– Inadequate evidence regarding younger ages
– Current (October 2011) draft recommendation against PSA-based
screening of any asymptomatic men
2011-Oct-29
© Stanley H. Weiss, 2011
41
Dr. Weiss’s Take on the
BIG PICTURE
• PSA not for everyone
• Need more evidence to make
good decisions
– Studies evolving for whites, namely the
studies we’ve just explored as well as
others
– No adequate studies commenced yet
for other racial and ethnic groups!!
2011-Oct-29
© Stanley H. Weiss, 2011
42
Dr. Weiss’s Take on the
BIG PICTURE
• Need improved guidance for
clinicians as to what to do and in
whom to do it
• There is some evidence of lifesaving potential with PSA screening
• The costs (e.g., the NNS and NNT)
may be much less than recent highly
publicized estimates, and thus it
may be cost-effective!
2011-Oct-29
© Stanley H. Weiss, 2011
43
Supplemental
Slides
2011-Oct-29
© Stanley H. Weiss, 2011
44
DRE: (digital rectal examination)
• Tendency to detect larger tumors with DRE
• Low chance of detecting clinically insignificant tumors with
DRE, but risk depends strongly on the PSA level.
• Limitation: small multi focal lesions with aggressive biologic
potential are not detected with DRE alone.
• The DRE is subjective = variable between different
examiners.
• Several studies have questioned the use of DRE in screening
programs and found little or no additional beneficial effect of a
DRE in men with PSA levels ≥ 4.0 ng/ml (Catalona 1994,
Rietbergen 1997).
• DRE may provide an additional value in detecting clinically
significant cancer in men with a low [?“normal”} range of PSA
(< 4.0 ng/ml) (Eastham 1999, Han 2004).
2011-Oct-29
© Stanley H. Weiss, 2011
45
TRUS: (transrectal ultrasound)
• Similar to the DRE, the interpretation of TRUS
is highly dependent on the investigator.
• Several studies have shown that the value of
TRUS has LIMITED value as a screening test to
detect cancer,
• But is indispensable for guiding prostatic
biopsies and assessing the prostate volume.
2011-Oct-29
© Stanley H. Weiss, 2011
46
PSA-related tests
F/T PSA ratio:
• Objective: To (try to) increase the specificity of PSA as a
screening tool derivates from PSA are studied.
• Total PSA consists of complex PSA (cPSA) and
free PSA (fPSA).
• cPSA is serum PSA that is bound to circulating proteins.
• The proportion of circulating cPSA is higher in patients
with carcinoma than in those with benign enlargement.
• Studies comparing the diagnostic efficacy of cPSA with
total PSA and the free to total (F/T) ratio so far report
inconsistent results.
2011-Oct-29
© Stanley H. Weiss, 2011
47
PSA-related tests (continued)
proenzyme PSA (pro-PSA):
• Form of free PSA
• Elevated in cancerous prostate tissue
• Results from a multi-center study have validated proPSA
as a detector of early stage prostate cancer.
• Findings suggest that proPSA may be associated with
aggressive and significant prostate cancer, worthy of
further investigation
• See: cebp.aacrjournals.org/content/19/5/1193.abstract?etoc
2011-Oct-29
© Stanley H. Weiss, 2011
48
PCA3 Screening
• PCA3 is a non-coding mRNA molecule that is
believed to be prostate specific.
– It is highly over-expressed in cancerous prostate
cells relative to benign tissue
– Present in urine (no blood test necessary)
• Potential to be used as supplement for PSA
testing
– PSA has a 21% specificity but a 87% sensitivity for
prostate cancer
– Conversely, a test for PCA3 was reported to have a
sensitivity of only 49%, but a specificity of 78%
– Additional studies are needed
2011-Oct-29
© Stanley H. Weiss, 2011
49
Loeb S, Vonesh EF, Metter EJ, Carter HB, Gann PH, Catalona WJ
What Is the True Number Needed to Screen and Treat to
Save a Life With Prostate-Specific Antigen Testing?
Purpose The European Randomized Study of Screening for Prostate Cancer (ERSPC)
reported a 20% mortality reduction with prostate-specific antigen (PSA) screening. However,
they estimated a number needed to screen (NNS) of 1,410 and a number needed to treat
(NNT) of 48 to prevent one prostate cancer death at 9 years. Although NNS and NNT are
useful statistics to assess the benefits and harms of an intervention, in a survival study setting
such as the ERSPC, NNS and NNT are time specific, and reporting values at one time point
may lead to misinterpretation of results. Our objective was to re-examine the effect of varying
follow-up times on NNS and NNT using data extrapolated from the ERSPC report.
Materials and Methods On the basis of published ERSPC data, we modeled the cumulative
hazard function using a piecewise exponential model, assuming a constant hazard of 0.0002
for the screening and control groups for years 1 to 7 of the trial and different constant rates of
0.00062 and 0.00102 for the screening and control groups, respectively, for years 8 to 12.
Annualized cancer detection and drop-out rates were also approximated based on the
observed number of individuals at risk in published ERSPC data.
Results According to our model, the NNS and NNT at 9 years were 1,254 and 43,
respectively. Subsequently, NNS decreased from 837 at year 10 to 503 at year 12, and NNT
decreased from 29 to 18.
Conclusion Despite the seemingly simplistic nature of estimating NNT, there is widespread
misunderstanding of its pitfalls. With additional follow-up in the ERSPC, if the mortality
difference continues to grow, the NNT to save a life with PSA screening will decrease.
J Clin Oncol 29(4) 464-467 (February 1, 2011)
2011-Oct-29
© Stanley H. Weiss, 2011
50
Cumulative hazard functions of prostate cancer-specific
mortality from the ERSPC with a piecewise exponential model
Loeb S et al. JCO 2011;29:464-467
2011-Oct-29
©2011
by American Society of Clinical Oncology
© Stanley H. Weiss, 2011
51
Loeb S, Vonesh EF, Metter EJ, Carter HB, Gann PH, Catalona WJ
What Is the True Number Needed to Screen and Treat to
Save a Life With Prostate-Specific Antigen Testing?
Table 1. Modeled Results Assuming a Piecewise Exponential Model
Ye ar
Group
0
1
2
3
4
5
6
7
8
9
10
11
12
Control
Screening
Control
Screening
Control
Screening
Control
Screening
Control
Screening
Control
Screening
Control
Screening
Control
Screening
Control
Screening
Control
Screening
Control
Screening
Control
Screening
Control
Screening
Hazard
Function
0
0
0.0002
0.0002
0.0004
0.0004
0.0006
0.0006
0.0008
0.0008
0.001
0.001
0.0012
0.0012
0.0014
0.0014
0.00242
0.00202
0.00344
0.00264
0.00446
0.00326
0.00548
0.00388
0.0065
0.0045
Cons tant
No. of
Dropout
Hazard
Survival
Patie nts
NNS
Rate
Rate
at Ris k
0.0002
1
0
89,353
—
0.0002
1
0
72,890
—
0.0002
0.9998
0.02186
87,421
—
0.0002
0.9998
0.02267
71,256
—
0.0002
0.9996
0.02186
85,530
—
0.0002
0.9996
0.02267
69,659
—
0.0002
0.9994
0.02186
83,681
—
0.0002
0.9994
0.02267
68,097
—
0.0002
0.9992
0.02186
81,871
—
0.0002
0.9992
0.02267
66,570
—
0.0002
0.999
0.02186
80,101
—
0.0002
0.999
0.02267
65,078
—
0.0002
0.9988
0.04277
76,747
—
0.0002
0.9988
0.04986
61,913
—
0.0002
0.9986
0.04277
73,534
—
0.0002
0.9986
0.04986
58,902
—
0.00102
0.99758
0.37661
50,458
—
0.00062
0.99798
0.35528
41,289 2,506
0.00102
0.99657
0.37661
34,623
—
0.00062
0.99736
0.35528
28,943 1,254
0.00102
0.99555
0.37661
23,758
—
0.00062
0.99675
0.35528
20,288
837
0.00102
0.99453
0.37661
16,302
—
0.00062
0.99613
0.35528
14,221
628
0.00102
0.99352
0.37661
11,186
—
0.00062
0.99551
0.35528
9,969
503
NNT
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
86
—
43
—
29
—
22
—
18
Hazard
Ratio
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0.61
1
0.61
1
0.61
1
0.61
1
0.61
Cum ulative
Hazard
Ratio
—
—
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0.83
1
0.77
1
0.73
1
0.71
1
0.69
Abbreviations: NNS, number needed to screen; NNT, number needed to treat.
J Clin Oncol 29(4) 464-467 (February 1, 2011)
2011-Oct-29
© Stanley H. Weiss, 2011
52
If You Choose to be Screened,
What’s the Best Way?
• Risk calculator or nomogram …
2011-Oct-29
© Stanley H. Weiss, 2011
53
2011-Oct-29
© Stanley H. Weiss, 2011
54
Example:
2011-Oct-29
© Stanley H. Weiss, 2011
55