AACC WORKSHOP

1
AACC WORKSHOP 2102LECTURE 1Tumor
Markers How We Use Them In The Clinic And New
DevelopmentsFacultyEleftherios P. Diamandis,
MD, PhD, FRCPC(Workshop Leader
ediamandis_at_mtsinai.on.ca)Daniel
W.Chan,Ph.D(dchan_at_jhmi.edu)Martin Fleisher,
PhD(fleishem_at_mskcc.org)

AACC 2005
2
TUMOR MARKERS
HOW WE USE THEM IN THE CLINIC AND NEW
DEVELOPMENTS930 - 1015AM - Tumor
Markers Historical Perspective and Current
Clinical Applications
E.P. Diamandis1015 -
1045AM - Guidelines for Use of Tumor
Markers M. Fleisher 1045 - 1100AM
-Break1100 - 1200PM Guidelines for Use of
Tumor Markers (Continued)M. Fleisher1200 -
1230PM Questions / Discussion 1230 - 130PM
- Lunch130 - 230PM - Tumor Markers-Present
and Future E.P.Diamandis230 -
245PM - Clinical Cases D.W.Chan245 -
300PM Break300 -400 PM - Clinical Cases
(Continued) D.W.Chan400 - 430PM Panel
Discussion / Questions All Faculty
AACC2005
3
Where could you find the presentations?

• Go to my website
• www.acdclab.org
• Click on Workshops
• Find AACC 2005 workshops
• Follow the instructions as to how to download the
  files

AACC2005
4
Available by AACC Press (published 2002) Special
discount if ordered at this meeting
AACC 2005
5
Tumor Markers - General
InformationbyEleftherios P. Diamandis





AACC
2005
6
AACC 2005
7
AACC 2005
8
Death Rates from Cancer and Heart Disease in USA
from 1975-2001
Cancer The 1 Killer
Jemal A et al. CA Cancer J Clin 20055510-30
9
1 Killer
10
Annual Cancer Death Rates Among Males for
Selected Cancer Types, US 1930-2001
Jemal et al. CA Cancer J Clin 2005 55 10-30
11
What is a tumor marker?

• a substance present in or produced by a tumor
• OR
• a substance produced by the host in response to
  the tumors presence
• present in cells, tissues, bodily fluids

12
What is a tumor marker used for?

• Use
• to determine the presence of a tumor
• Measurement
• qualitative or quantitative methods
  chemical immunological molecular biological
  

13
Historical Background
14
The Ideal Tumor Marker (TM)

• Measured easily, reliably and cost-effectively
  using an assay with high analytical sensitivity
  and specificity.
• Quantitative level of TM reflects tumor burden
  with high diagnostic sensitivity (few FNs) with
  specificity (few FPs).
• Test results influence patient care and
  especially outcome.

15
Clinical Uses of Tumor Markers

• Screening - limited
• Diagnosis - limited
• Prognosis - limited
• Tumor staging - limited
• Tumor localization / radiotherapy - limited
• Monitoring the effectiveness of therapy -
  important
• Detecting tumor recurrence or remission -
  important

16
Types of Tumor Markers
Hormones (hCG calcitonin gastrin prolactin
growth hormone, etc.) Enzymes (acid
phosphatase alkaline phos- phatase PSA)
Proteins Glycoproteins (CA 125 CA 15.3
CA 19.9, etc.) Oncofetal antigens (CEA, AFP)
Receptors (ER, PR, EGFR) Oncogenes (Ras Myc
abl-bcr) Tumor suppressor genes (BRCA1 p53
Rb)

AACC 2005
17
Usefulness of Laboratory Tests
Affected by analytical and diagnostic sensitivity
and specificity AND a tests positive
predictive value (PPV).
PPV Probability that an individual with a
positive test result has the disease
for which the test was developed to assist
in diagnosing. PPV, unlike NPV, markedly
influenced by the prevalence of disease in
the population tested. AACC 2005
18
Analytical vs Diagnostic Sensitivity
Analytical sensitivity minimum detectable
concentration (MDC) of an antigen that an
immunoassay can reliably distinguish from
zero concentration with 95 confidence.
Diagnostic sensitivity proportion of
individuals with disease who yield a positive
test for antigen i.e. TP / (TP
FN). AACC 2005
19
Analytical vs Diagnostic
Specificity Analytical specificity ability of
an immunoassay to measure only the analyte of
interest (freedom from interferences)
expressed typically as cross-reactivity. Di
agnostic specificity proportion of
individuals without disease who yield a
negative test for antigen i.e. TN / (TN
FP). AACC 2005
20
Calculation of Predictive Value
Parameters (2x2 Table)___________________________
_____________________Assumptions Disease
prevalence 1 Population tested
100,000 No. w/disease (0.01 x 100,000)
1,000 Test sensitivity 95 Test
specificity 95-------------------------
------------------------------------------------T
est Results No. with Disease No. w/o Disease
TotalPos 950 (TP) 4,950 (FP)
5,900Neg 50 (FN) 94,050 (TN)
94,100Total 1,000
99,000 100,000----------------------------------
---------------------------------------Sensitivit
y TP/ (TP FN) 950/ (950 50) 950/ 1,000
95Specificity TN/ (TN FP) 94,500/
(94,050 4,950) 94,050/ 99,000 95PPV TP/
(TP FP) 950/ (950 4,950) 950/ 5,900
16.10NPV TN/ (TN FN) 94,050/ (94,050
50) 99.95 or 100Eff (TP TN)/ Grand Tot
(950 94,050)/ 100,000 95,000/ 100,000
95

AACC 2005
21
Effect of Prevalence of Disease on PPV
and NPV__________________________________________
_______Disease PPV of pos NPV of
negprevalence, test result, test result,
0.1 1.9 99.9 1
16.1 99.9 10 67.9 99.4 50
95 95.0 100
100.0 n.a._____________________________________
____________ For a test with 95 diagnostic
sensitivity and 95 diagnostic
specificity n.a. - not applicable

AACC 2005
22
How To Improve PPV To improve the
PPV of a laboratory test, clinicians should
increase the prevalence of disease by
appropriate selection of patients for whom the
test is ordered. AACC 2005
23
Critical Clinical Questions1.
Does a patient have cancer?2. If yes, which
organ is afflicted?3. Is the cancer localized
or disseminated?4. How aggressive is the
cancer?5. Will the patient likely relapse or
not relapse if no adjuvant therapy is given
post-primary treatment?6. Will the patient
respond better if one treatment type is
given instead of another one?7. Can cancer
relapse post-primary therapy be detected
before the patient has symptoms?8. If yes, can
the patient get a benefit with early
treatment of relapse?
AACC 2005
24
Clinical Uses of Tumor
Markers Screening for cancer -
limited Diagnosing cancer - limited Evaluating
cancer prognosis-limited Tumor
staging-limited Detecting Tumor Recurrence or
remission-important Localizing tumor and
directing chemo- or radio-therapeutic agents-
limited Monitoring the effectiveness
of cancer therapy-important

AACC 2005
25
Outcome Screening for cancer
using assays for TMs should increase the
early detection of cancer and the
long-term survival of treated patients
and decrease their morbidity and
mortality. None of the currently available
TMs are perfect (i.e. 100 diagnostic
sensitivity and 100 diagnostic
specificity). AACC
2005
26
Reference (Normal) Ranges
For Cancer Markers Healthy
subjects (how many?) Sex Relevant age
population Benign conditions Selection
of patient population (stage, grade,
histological type, etc.) Selection of
appropriate cutoff - maximize sensitivity -
maximize specificity - maximize predictive
value AACC 2005
27
Receiver Operating Characteristic (ROC) Curve

• Plot of
• sensitivity (true positive rate) y-axis
• vs.
• 1-specificity (false positive rate) x-axis
• Displays performance of a test over the entire
  range of cutoff values.
• You can select a cutoff value that offers the
  desired sensitivity and/or specificity.
• Predictive value of various markers on the same
  plot.
  

28
AACC 2005
29
1
2
4
0.2
6
0.3
10
0.4
20
0.6
0.8
1.2
30
Discovery of New Tumor MarkersVery
good prospects Human Genome Project will
unravel all genes and all proteins
Tissue-specific expression will be established
(microarrays 2 years) Differential
expression of genes in normal vs diseased
tissues (microarrays 2 years) All proteins
and antibodies will become available (3-5
years) Disease profiling with biomarkers
(microarrays mass spectrometry)
AACC 2005
31
The FuturePredictions about the
future will be presented in another
lecture130-230 PM AACC 2005