Tumor Markers
Michael A. Pesce, PhD
Department of Pathology
Columbia-Presbyterian Medical Center
IDEAL TUMOR MARKERS
• Be specific to the tumor
• Level should change in response to tumor size
• An abnormal level should be obtained in the presence
of micrometastases
• The level should not have large fluctuations that are
independent of changes in tumor size
• Levels in healthy individuals are at much lower
concentrations than those found in cancer patients
• Predict recurrences before they are clinically detectable
• Test should be cost effective
COMMON TUMOR MARKERS
Analyte Cancer Use
CEA Monitor colorectal, breast, lung cancer
CA-125 Ovarian cancer monitoring
CA15-3, 27. 29 Monitor recurrences of breast cancer
AFP Germ cell tumors, liver cancer
Total PSA Screen and monitor prostate cancer
Free PSA Distinguish prostate cancer from BPH
HCG Germ cell and trophoblastic tumors
Hormone
receptors Breast cancer therapy
NMP 22, BTA
FDP Monitor recurrences of bladder cancer
SCREENING TESTS
• Cancer must be common
• The natural history of the cancer should be
understood
• Effective treatments must be available
• The test must be acceptable to both patients and
physicians
• The test must be safe and relatively inexpensive
CEA
• Described by Gold and Freedman in1965 as a
marker for Colorectal Cancer
• Molecular mass of approximately 200,000 kA
• Glycoprotein with a carbohydrate composition
ranging from 50 - 85% of molecular mass
• CEA levels 5 - 10 times upper limit of normal
suggests colon cancer
• CEA is not used to screen for colon cancer
CEA Distribution In Healthy Individuals and
Patients with Non-Malignant Conditions
% Distribution of CEA
ng/mL ng/mL ng/mL
Healthy Subjects 0-3.0 3.1-10 >10.0
Non Smokers 96 4 0
Smokers 80 19 1
Non-Malignant Diseases
Cirrhosis 53 42 5
Ulcerative Colitis 65 26 9
Rectal polyps 78 19 3
Pulmonary 52 39 9
Gastrointestinal 76 21 3
CEA Distribution In Patients
With Malignant Disease
% Distribution of CEA
0-3 3.1-10 >10
ng/mL ng/mL ng/mL
Colorectal 28 20 52
Breast 50 27 23
Ovarian 80 16 4
Pulmonary 39 29 32
CA-125
• CA-125 glycoprotein molecular weight 200-1,000 kda
• Introduced in 1983 by Bast for ovarian cancer
• In the US, in 1998 25,400 new cases will be diagnosed
and 14,500 women will die as a result of this disease
• 70% of the women with ovarian cancer are over the age
of 50
• One half of patients with stage 1 ovarian cancer have
elevated CA-125 levels and a five year survival rate of
90%. In late stage disease, the five year survival rate is
from 4-30%
• Worldwide incidence is highest in industrialized
countries and lowest in Japan and India
SYMPTONS OF OVARIAN CANCER
• ASCITES
• ABDOMINAL and PELVIC PAIN
• ABNORMAL UTERINE BLEEDING
• GASTROINTESTINAL DISCOMFORT
• WEIGHT LOSS
• URINARY FREQUENCY
RISK FACTORS
INCREASED RISK DECREASED RISK
Family History Oral Contraceptive
Advanced Age Breast Feeding
Infertility
Tubal Ligation
Nulliparity
CA-125 Distribution In Healthy Subjects and Patients with
Non-Malignant Conditions
% Distribution of CA-125
<35 35-65 >65
u/mL u/mL u/mL
Healthy Individuals 98 1.7 1.3
Non-Malignant Conditions
Pregnancy 73 22 5
Cirrhosis 30 13 57
Pulmonary Disease 94 0 6
Pelvic Inflammatory Disease 76 3 21
Endometriosis 86 11 3
Ovarian Cysts 90 7 3
Uterine Fibroids 77 13 10
Breast Fibroids 100 0 0
CA-125 Distribution In Patients With
Malignant Disease
% Distribution of CA-125
Cancers <35 35-65 >65
u/mL u/mL u/mL
Ovarian 14 9 77
Lung 56 19 25
Breast 82 8 10
Endometrial 70 8 22
Cervical 66 15 19
Colorectal 76 11 12
CEA - CAP Proficiency Survey
#Labs Mean
ng/mL
Abbott AXSYM 689 11.14
Bayer ACS 180 73 8.63
Bayer Centaur 240 8.50
Bayer Immuno-1 34 7.71
Beckman Access 201 7.86
DPC Immulite 2000 78 11 .60
Roche Elecsys/E170 105 11.27
Tosoh AIA-Pack 33 18.76
J&J Vitros ECI 97 13.17
TROPONIN I - CAP Proficiency Survey
# Labs Mean
ng/mL
Abbott AXSYM 1228 3.93
Dade Dimension 706 0.43
Dade Stratus 229 0.40
Beckman Access 307 0.24
Bayer Centaur 144 1.22
Bayer ACS 180 173 1.12
J&J Vitros ECI 81 0.16
DPC Immulite 23 2.90
GUIDELINES FOR ORDERING/
INTERPRETING TUMOR MARKER TESTS
• Never rely on the result of a single test
• Order every test from the same laboratory
• Consider half-life of the tumor when interpreting the
result
• Consider how the Tumor Marker is removed or
metabolized
• Consider Hook Effect
• Consider presence of HAMA antibodies
MULTIPLE MYELOMA
Multiple Myeloma - proliferation of a single clone of
plasma cells that produces a monoclonal protein
Annual Incidence - 4 in 100,000
Number of cases per year - 13,000
Represents 1% of all malignant diseases
Median age at diagnosis - 65 years
Median survival - 3 years
ETIOLOGY OF MULTIPLE MYELOMA
Radiation Exposure
Agriculture - Farming & Pesticide Use
Chemicals - Benzene
Not Related to Smoking or Alcohol Consumption
Monoclonal Gammopathy of
Undetermined Significance
Defined as the presence of a serum monoclonal protein at
low levels
Number of cases per year - 750,000-1,000,000
54% Men 46% Women
Occurs in 2% of persons over 50 years, 3% over 70 years
Median age at diagnosis - 72 years
Median survival - 12 years
M
Y
E 12
L
O
M 10
A
8
P
E
6
R
1
4
0
0 2
,
0
0 0 M,White M, Black F, White F, Black
0
SYMPTOMS OF MULTIPLE MYELOMA
Bone Pain - Back, Ribs
Osteoporosis
Bone Fracture
Fatigue
Anemia
Renal Failure
Infections
Clinical Laboratory in Multiple Myeloma
-Biochemical -
Serum monoclonal proteins >3.0 g/dL
Polyclonal Immunoglobulin Decreased
Proteinuria, Bence-Jones Protein present in urine
BUN, Creatinine
Calcium ,N
- Hematological -
Hemoglobin Decreased
Anemia - Normochromatic, Normocyte
ESR Increased
Major Laboratory Features of MM
% of MM
Findings patients affected
Hemoglobin <12 g/dL 65
Serum calcium >11.0 mg/dL 14
Serum creatinine > 1.7 mg/dL 23
Serum M-protein present 92
Urine M-protein present 75
M-protein in serum or urine (or both) 99
Bone lesions 75
Bone marrow plasma cells >10% 90
Monoclonal Gammopathy of Undetermined Significance
Serum monoclonal protein <3.0 g/dL
Stability of monoclonal protein during long term follow-up
<10% Plasma cells in bone marrow
None or a small amount of Bence-Jones protein in urine
Absence of lytic bone lesions
Serum calcium, BUN, creatinine - Normal
Hemoglobin - Normal
Laboratory Data at Diagnosis for MGUS
Serum concentration of the uninvolved
immunoglobulin is decreased in 38% of patients
Urine - Kappa Bence-Jones Protein 21%
Lambda Bence-Jones Protein 10%
For most of the patients, the concentration of the
Bence-Jones protein was <150 mg/24 hr
No light chain was detected in 69% of the patients
Distribution Frequency of Monoclonal Proteins In MGUS
Biclonal 2%
IgM 14%
IgG 73% IgA 11%
Clinical Course of 241 Patients with MGUS
M Protein >3.0 g/dL No Increase in M
No Myeloma (23) Protein (46)
10%
19%
47%
24%
Died of Unrelated Developed Myeloma &
Causes (173) Unrelated Diseases (59)
Normal SPE
Albumin 1 2
Monoclonal
Albumin decreased
gammopathy
Sharp peak in gamma region
Albumin 1 2
