General Population Screening Transvaginal Ultrasound

Transvaginal ultrasound (TVUS) has been evaluated for use in ovarian cancer screening in the general population. Because of the close proximity of the probe to the ovaries, TVUS offers excellent resolution of ovarian morphology. Volume, outline, papillations, and complexity of ovarian masses can be used to raise suspicion of cancer. Benign ovarian lesions are common, however, resulting in false-positives that may necessitate invasive surgery for asymptomatic women. In addition, TVUS as an initial screening test is expensive. Figure 6-1 shows an ultrasound image with color Doppler of the typical appearance of an involuting corpus luteal cyst.

Van Nagell and colleagues14 conducted a screening study to examine whether TVUS could be used to detect ovarian cancer at an earlier stage and to decrease ovarian cancer mortality. They reported that between 1987 and 1999, 14,469 asymptomatic women underwent annual screening with TVUS. Women with an abnormal TVUS had a repeat TVUS in 4 to 6 weeks, and women with a persistently abnormal scan were advised to undergo surgery. Eligible women included all women 50 years of age and older and women 25 years or older with a first- or second-degree relative

Figure 6-1. Typical appearance of an involuting corpus luteal cyst.

Frq 7.0 MHz

Gn 21

DR 72

AO 100%

2<CF

Frq

6.3 MHz

Gn

40

L/A

1(S

AO

too %

PRF

0.8 kHz

IWF

70 Hz

- SiP

3/14

4"

with ovarian cancer. One hundred eighty (1.2%) patients underwent surgery for suspicious findings on ultrasound, with 17 ovarian cancers detected: 11 stage I, 3 stage II, and 3 stage III. Sensitivity was 81%, and the negative predictive value was at 99.7%. The PPV was 9.4%, close to the clinically acceptable goal of a 10% PPV. These researchers also updated their data to include 25,327 women screened from 1987 to 2005. Three hundred sixty-four women (1.4%) underwent surgery, with 35 primary invasive ovarian cancers detected. Sensitivity for all stages was 85%, specificity was 98.7%, positive predictive value was 14% and negative predictive value was 99.9%.15 Despite this encouraging outcome, interpretation of ultrasonography is observer dependent, and it is not certain that community-based trials could match the expertise of this group in eliminating false-positives. In addition, the cost of annual TVUS would be prohibitive in our current health system.

CA-125 and Multimodal Approaches

CA-125 is the most extensively studied tumor marker for ovarian cancer. Figure 6-2 shows the genetic structure of the CA-125 marker. CA-125 is a high-molecular-weight mucin found in mullerian-derived epithelium, namely, fallopian tube, endometrium, and endocervix. Normal surface epithelium does not express CA-125, but it is elevated in 80% of patients with epithelial ovarian cancer and in over 90% of patients with advanced-stage disease.16 CA-125 received FDA approval for use in monitoring patients with ovarian cancer for disease persistence and recurrence.17 It is not approved as a screening tool for early detection of ovarian cancer.

Several issues limit the usefulness of CA-125 as a screening tool for ovarian cancer. First, although over 90% of advanced-stage patients display CA-125 elevations, only 50% to 60% of patients with stage I disease display elevations. Second, tumors with mucinous histologies are less likely to be associated with a CA-125 elevation.18 Third, CA-125 has inadequate specificity, particularly in pre- and perimenopausal women. False-positive elevations are seen with benign ovarian cysts, endometriosis, adeno-myosis, fibroids, diverticulitis, and liver cirrhosis, in addition to other benign and malignant conditions.

A Swedish study published by Einhorn and colleagues19 in 1992 examined 5550 healthy asymptomatic women through the Stockholm Population Registry to determine whether CA-125 is a useful initial screening test for ovarian cancer. All participants had a CA-125 level drawn. Women with elevated CA-125 levels were age-matched to an equal number of women with normal CA-125 levels, and these women underwent pelvic examinations, transabdominal ultrasounds, and serial CA-125 levels. Six of the 175 women with elevated CA-125 levels were diagnosed with ovarian cancer; conversely, ovarian cancer was diagnosed in three of the controls, all of whom were younger than 50 years of age. Using a threshold of 35 |l/mL for the

CA-125, full length

Amino-terminal domain

5)

(A-

—' ^S,T,P-rich repeats^

TMv

Cloned fragment 1.1 kB

Cloned fragment, 1.6 kB

1 1

Fusion construct "MucTM"

Scale: 1000 bp per dash —

Figure 6-2. Genetic structure of the CA-125 marker.

Figure 6-2. Genetic structure of the CA-125 marker.

CA-125, specificity was found to be 98.5% versus 94.5%, respectively, for women age 50 years or older and women younger than 50. The authors concluded that CA-125 levels showed promise as a good screening tool for women older than 50 but that larger studies were needed.

Most current screening studies use both serum tumor markers and radiologic imaging. In the largest study to date using CA-125, Jacobs and colleagues20 randomized 22,000 postmenopausal women in the United Kingdom into a group who had no screening and a group who had annual CA-125 concentration measured. When an elevated CA-125 level was found, a transabdominal ultrasound was performed. Surgical intervention was undertaken when the ultrasound was abnormal. Over 3 years of annual screening, 468 women had elevated CA-125 levels and subsequently had ultrasounds. Of these, 29 had abnormal ultrasounds and therefore underwent surgery, revealing six incident ovarian cancers. This multimodal approach demonstrated a positive predictive value of 20.7%. These results demonstrate that a combination technique could be a valid approach to ovarian cancer screening.

TVUS and CA-125 have been studied in a multicenter Prostate, Lung, Colorectal and Ovarian (PLCO) cancer screening trial in the United States.21,22 This prospective randomized study (Table 6-2) enrolled over 37,500 women, and each woman was assigned to either a control group who received no screening or a screened group who received both TVUS and CA-125. The ovarian cancer screening arm was designed to evaluate whether annual screening with CA-125 and TVUS can reduce mortality rates from ovarian cancer. Eligible women had to be 55 to 74 years old and had to have no diagnosis of lung, colorectal, or ovarian cancer. Women in the intervention arm were screened for ovarian cancer annually for 6 years with CA-125 and TVUS annually for 4 years. (The screening arm originally included bimanual physical examination of the ovaries, but this procedure was dropped from the evaluation after 5 years, when it was found that no ovarian cancers had been detected with this single modality.) Any abnormal test result was followed by a referral to a gynecologic oncologist. Enrollment began in 1993 and was completed in 2001. Participants will be followed for at least 13 years from entry.

Findings from the initial screening revealed an abnormal ultrasound in 4.7% of participants and an abnormal CA-125 in 1.4%. Abnormal results in 1703 women

Table 6-2. Current Prospective Randomized Ovarian Screening Trials for Low-Risk Women

Inclusion Criteria

Aims

Comments

Prostate, Lung,

Age 55-74

1.

Annual CA-125 and

Accrual complete.

Colon, Ovarian

transvaginal ultrasound for

Patients will be

Cancer

4 years, followed by

followed for at least

Screening Trial

annual CA-125 for 2 years*

13 years.

(PLCO)

2.

Control group—no screening

o 1 o

United Kingdom

Age 50-74,

1.

Multimodal group: Annual

Accrual complete.

Collaborative

postmenopause

CA-125 using risk of

Patients will be

Trial of Ovarian

ovarian cancer algorithm^

followed for 6

Cancer

2.

Annual TVUS+

years.

Screening

3.

Control group: no screening

(UKCTOCS)

*Follow-up for abnormal screening results at discretion of physician.

fAbnormal screening results trigger (a) repeat CA-125 and more detailed TVUS in multimodal group and (b) more detailed TVUS in TVUS group.

*Follow-up for abnormal screening results at discretion of physician.

fAbnormal screening results trigger (a) repeat CA-125 and more detailed TVUS in multimodal group and (b) more detailed TVUS in TVUS group.

resulted in 571 surgeries. These operations resulted in the detection of 20 invasive ovarian, fallopian tube, or primary peritoneal cancers; 1 granulosa cell tumor; and 9 low malignancy potential tumors. Over 80% of the invasive cancers were stage III or IV. The positive predictive value for each single modality was poor, achieving only 3.7% for an abnormal CA-125 and 1% for an abnormal TVUS. When both tests were abnormal, the positive predictive value was 23%, that is, four to five operations for each case of ovarian cancer detected, but 60% of cases would not have been detected. The study is designed to demonstrate a 30% reduction in mortality rate if the patients are followed for at least 13 years from entry. Further follow-up will provide information on how this protocol detects new, incident ovarian cancers, whereas the initial data showed the preexisting (prevalent) cancers.

Serial measurements of CA-125 have been proposed as a modality to offer a more specific determination of ovarian cancer risk than a fixed CA-125 cutoff (Fig. 6-3). Skates and colleagues23 published a study that analyzed data from Jacobs' prospective trial.20 Researchers used 33,621 CA-125 samples from 9233 women in whom two or more serial samples were analyzed. Using a longitudinal change point model, the study concluded that serial CA-125 levels with the risk calculation significantly improved the detection of ovarian cancer, with a specificity of 98% and a sensitivity of 86% over a 62% sensitivity using a fixed cutoff for CA-125. Patients with ovarian cancer exhibited progressively increasing CA-125 levels, whereas patients with benign gynecologic conditions, nongynecologic disease, or no detectable illness had levels that remained constant over time, even when elevated (Fig. 6-4). The ability of elevated CA-125 to detect ovarian cancer has been evaluated prospectively in the United Kingdom Collaborative Trial of Ovarian Cancer Screening (UKCTOCS), an ongoing randomized controlled study (see Table 6-2) designed to establish the impact of ovarian cancer screening in the general population on ovarian cancer mortality.24 Conventional pelvic examination is being compared with annual TVUS and with annual CA-125 followed by TVUS only if the CA-125 is rising.

The primary end point is to evaluate ovarian cancer mortality. In addition, the study also examines the morbidity of ovarian cancer screening, determines the resource implications of screening, and assesses the feasibility of population

Figure 6-3. Serial measurements of CA-125 may offer a more detailed determination of ovarian cancer risk. (From Skates SJ, Menon U, MacDonald N, et al: Calculation of the risk of ovarian cancer from serial CA-125 values for preclinical detection in postmenopausal women. J Clin Oncol 21[10 Suppl]:206-210, 2003: Fig. 3.)

Figure 6-3. Serial measurements of CA-125 may offer a more detailed determination of ovarian cancer risk. (From Skates SJ, Menon U, MacDonald N, et al: Calculation of the risk of ovarian cancer from serial CA-125 values for preclinical detection in postmenopausal women. J Clin Oncol 21[10 Suppl]:206-210, 2003: Fig. 3.)

OVARIAN CANCER

BENIGN GYNECOLOGIC

100 35 10 1

OVARIAN CANCER

2 24 36 48 Months since May '85

2 24 36 48 Months since May '85

12 24 36 48 Months since May '85

NONGYNECOLOGIC DISEASE

NO DISEASE

2 24 36 48 Months since May '85

2 24 36 48 Months since May '85

12 24 36 48 Months since May '85

Figure 6-4. Longitudinal CA-125 values in women with ovarian cancer, benign gynecologic disease, nongynecologic disease, and no disease. (Courtesy of Steve Skates.)

screening. The multicenter trial involves 13 hospitals in the United Kingdom and has enrolled more than 200,000 postmenopausal women 50 to 74 years of age who are not considered high risk based on family history. A total of 50,000 women undergo an annual CA-125 blood test followed by TVUS when the CA-125 level is rising (CA-125-TVUS), 50,000 undergo annual TVUS, and 100,000 in the control arm are followed by their family physicians. Women fill out questionnaires to assess the behavioral and psychosocial responses of women to ovarian cancer screening. Figure 6-5 is a flow chart for UKCTOCS.

Encouraging preliminary results have been reported from an evaluation of prevalent cases detected in the first 2 years of the UKCTOCS study.25 Among the screen-detected cancers, 48% were stage I/II in both the TVUS alone and CA-125-TVUS arms—approximately twice the percentage found with conventional diagnosis. Remarkably, sensitivity for detecting ovarian cancer at any stage was greater in the CA-125-TVUS arm (89%) than in the TVUS arm (74%). The specificity and PPV of CA-125-TVUS were also superior to those with annual TVUS alone. In the first year after diagnosis, CA-125-TVUS prompted 2.8 operations per case of ovarian cancer compared with 36.2 for TVUS alone. The difference in PPV may relate to detection of benign disease with TVUS that is not associated with a rising CA-125. Finally, the estimated lead time for the CA-125-TVUS arm was 1.8 to 2.6 years and 1.1 to 1.6 with TVUS alone, consistent with the potential value of an annual screen. Whether a sufficient number of incident cases will be detected at early stage to impact on survival still needs to be determined.

A single-arm study conducted by the U.T. MD Anderson Ovarian Cancer Specialized Program of Research Excellence has been modeled after the CA-125-TVUS arm

200,000

>50-74

years

postmeno-

pausal

women

_^

Ultrasound group annual TVUS 50,000

Multimodal group annual CA-125 50,000

Ultrasound group annual TVUS 50,000

Control group

100,000

-i Normall

Abnormal

Normal

Abnormal

'1

Level II screen

.r

All women followed up via the NHS Cancer Registry as well as postal questionnaires

Level II screen: Detailed scan in both groups + repeat CA-125 in mutimodal group

Figure 6-5. Flow chart for UKCTOCS (United Kingdom Collaborative Trial of Ovarian Cancer Screening).

of the UKCTOCS study.26 Over the past 7 years, 2573 apparently healthy women have been screened annually at six sites in the United States, with 8172 CA-125 determinations using the Risk of Ovarian Cancer (ROC) algorithm. Fewer than 2% of the women have been referred for TVUS on the basis of a rising CA-125 level, and five patients have been referred for operations that have detected three ovarian cancers: stage IA borderline and stage IIA and stage IIC invasive cancers. From these data, researchers estimate that the PPV is not less than 14%. Among the women screened to date, this screen failed to detect one stage I borderline cancer but has not missed an invasive ovarian cancer. Although this is a much smaller study, the results to date are consistent with those obtained in the UKCTOCS trial. Despite these promising early results, however, screening for ovarian cancer in women at conventional risk should be limited to clinical trials.

Multiple Markers and New Technologies for Biomarker Discovery

Whatever the outcome of trials based on CA-125 as an initial step, a screening strategy with greater sensitivity is required because 20% of ovarian cancers fail to express CA-125. Given the heterogeneity of the disease at a cellular and molecular level, no single tumor marker is likely to be adequately sensitive to detect early-stage ovarian cancer. Consequently, multiple tumor markers have been evaluated (Box 6-2). Many of these tumor markers do complement CA-125 (Figs. 6-6 and 6-7). Considering a screen positive when any one of several biomarkers is positive can increase sensitivity, but specificity decreases. For example, a Lewis X mucin determinant (OVX1), the cytokine macrophage colony-stimulating factor (M-CSF), and CA-125 were evaluated in 89 serum samples from subjects known to have stage I ovarian cancer.27 Sensitivity was improved from 69% with CA-125 alone to 84% using the three-combination analysis; however, specificity decreased from 99% to 84%. Novel mathematical techniques, including artificial neural network analysis (ANN) and a mixture of multivariate normal distributions (mixed multivariate analysis, MMA), can increase sensitivity using multiple biomarkers without sacrificing specificity. When four conventional serum biomarkers (CA-125II, CA-72-4, CA-15-3, and M-CSF) were compared with CA-125II alone for distinguishing stage I disease from healthy controls, specificity could be maintained at 98%, and sensitivity increased from 48% with CA-125II to 72% (ANN)28 or 75% (MMA)29 with the panel of biomarkers.

Box 6-2. Tumor Markers That May Be Useful in Screening for Ovarian Carcinoma

Alpha-l-antitrypsin

Galactosyltransferase

M-CSF

BHCG

HE4

Mesothelin

CA15-3

HER-2/neu

Mucin-like carcinoma antigen

CA19-9

Human milk fat globule protein

Osteopontin

CA50

Human milk globule 2

Ovarian serum antigen

CA54-61

IL-2 receptor

OVXI

CA72-4

IL-6

p110 epidermal growth factor receptor

CA-125

IL-8

Placental alkaline phosphatase

CA-195

IL-10

Prostasin

Cathepsin L

Inhibin

Sialyl TN

Carcinoembryonic antigen

Kallekrein-6

Soluble Fas ligand

Ceruloplasmin

Kallekrein-10

Tetranectin

CRP

Lipid-associated sialic acid

Tumor-associated trypsin inhibitor

CYFRA21-1

Lysophosphatidic acid

Tumor necrosis factor receptor

Dianon marker 70/K

Matrix metaloproteinase 2

Urinary gonadotropin peptide

From Chu CS, Rubin SC: Screening for ovarian cancer in the general population. Best Pract Res Clin Obstet Gynaecol 20:307-320, 2006, page 312.

From Chu CS, Rubin SC: Screening for ovarian cancer in the general population. Best Pract Res Clin Obstet Gynaecol 20:307-320, 2006, page 312.

Over the past decade, novel markers have been discovered using monoclonal antibodies raised against ovarian cancer tissue, lipid analysis, gene expression arrays, and proteomic techniques. Mesothelin (soluble mesothelin-related protein, SMRP), an adhesion molecule found both on ovarian cancer and normal mesothelial cells, was originally detected empirically using monoclonal antibodies. Mesothelin is elevated in a majority of ovarian cancers and complements CA-125 for detecting early-stage disease.30 Interestingly, SMRP can be detected in the urine, and, when corrected for glomerular filtration rate, urinary SMRP levels detect 40% of stage I patients.31

Lysophosphatidic acid (LPA) is a lipid of low molecular weight that is found in the ascites fluid and plasma of most patients with ovarian cancer. LPA stimulates calcium influx, proliferation, and drug resistance of ovarian cancer cells. Although there was initial enthusiasm about LPA for detecting women with stage I ovarian cancer,32 confirming studies have not been published.

Overexpression of several potential biomarkers has been detected with gene expression arrays, including HE4, kallikreins, prostasin, osteopontin, vascular endothelial growth factor (VEGF), and interleukin 8 (IL-8). After CA-125, HE4, a human whey protein, has been the object of most intense study.33 HE4 is slightly less sensitive than CA-125 for detecting early-stage disease but has greater specificity for distinguishing malignant from benign pelvic masses, particularly in premenopausal women. Kallikreins include a family of 15 secreted serine proteases of approximately 30 kD that include prostate-specific antigen (PSA). Several may prove useful as bio-markers for the prognosis or detection of ovarian cancer.34,35 All the kallikreins are tandemly localized on chromosome 19q13.4 and were initially isolated by cloning the area. Kallikrein 6 and 10 are being investigated for usefulness as serum tumor markers for ovarian cancer.

Proteomic techniques have been used for early detection of ovarian cancer in two ways: to identify a distinctive pattern of peptide and protein expression in serum or a b c d a b c d

Figure 6-6. Immunostaining for CA-125 and other potential markers. Strong staining (A), patchy (B), and negative staining (C) for CA-125 (x20). Positive immunostainings for HK6 (D), HK10 (E), osteopontin (F), claudin 3 (G), and DF3 (H) (x20). Positive immunostaining for MUC1 (I), negative staining for MUC1 (J), positive immunostaining for VEGF (K), negative immunostaining for VEGF (L), positive immunostaining for mesothelin (M), negative immunostaining for mesothelin (N), positive immunostaining for HE4 (O), negative immunostaining for HE4 (P), positive immunostaining for CA-19-9 (Q), and negative immunostaining for CA-19-9 (R) (x20). (From Rosen DG, Wang L, Atkinson JN, et al: Potential markers that complement expression of CA125 in epithelial ovarian cancer. Gynecol Oncol 99:267-277, 2005: Fig. 1.)

Figure 6-6. Immunostaining for CA-125 and other potential markers. Strong staining (A), patchy (B), and negative staining (C) for CA-125 (x20). Positive immunostainings for HK6 (D), HK10 (E), osteopontin (F), claudin 3 (G), and DF3 (H) (x20). Positive immunostaining for MUC1 (I), negative staining for MUC1 (J), positive immunostaining for VEGF (K), negative immunostaining for VEGF (L), positive immunostaining for mesothelin (M), negative immunostaining for mesothelin (N), positive immunostaining for HE4 (O), negative immunostaining for HE4 (P), positive immunostaining for CA-19-9 (Q), and negative immunostaining for CA-19-9 (R) (x20). (From Rosen DG, Wang L, Atkinson JN, et al: Potential markers that complement expression of CA125 in epithelial ovarian cancer. Gynecol Oncol 99:267-277, 2005: Fig. 1.)

in urine and to discover individual peptides or protein biomarkers and then to develop conventional immunoassays that can be performed in combination. A comparison of the proteomic techniques used for discovery and validation of markers is summarized in Table 6-3.

Surface-enhanced laser desorption and ionization time-of-flight mass spectrometry (SELDI-TOF) and matrix-assisted laser desorption and ionization time-of-flight mass spectrometry (MALDI-TOF) have been used to analyze the pattern of peptides in

CA19-9 HE4 MES MUC1

.2 CLDN3 m

VEGF OPN HK6 HK10

0 10 20 30 40 50 60 70 80 90 100 Percentage of expression

Figure 6-7. Expression of biomarkers in 65 CA-125-deficient cases. (From Rosen DG, Wang L, Atkinson JN, et al: Potential markers that complement expression of CA125 in epithelial ovarian cancer. Gynecol Oncol 99:267-277, 2005: Fig. 3.)

0 10 20 30 40 50 60 70 80 90 100 Percentage of expression

Figure 6-7. Expression of biomarkers in 65 CA-125-deficient cases. (From Rosen DG, Wang L, Atkinson JN, et al: Potential markers that complement expression of CA125 in epithelial ovarian cancer. Gynecol Oncol 99:267-277, 2005: Fig. 3.)

sera from healthy women and from patients with ovarian cancer.36 Very high sensitivity and specificity have been reported.37 Over the past 6 years, the computer algorithm used to analyze these data has evolved, and no final formulation has been validated. Moreover, other investigators have had difficulty in confirming the original analysis and have identified significant flaws in the methods used.38,39

SELDI-TOF and MALDI-TOF both use mass spectrometry to generate a "bar code" of specific peptides in body fluids; methodologic differences between the two techniques are listed in Table 6-4. Proteins to be analyzed are co-crystallized with ultraviolet (UV)-absorbing compounds and vaporized by pulses of a UV laser beam. The energy-absorbing molecules transmit part of the energy to each peptide, resulting in positive ions of different charge and mass. The ionized peptides released by the laser are then accelerated in an electric field. A distinctive mass-to-charge ratio is deduced from the velocity of each peptide. In SELDI-TOF, several different types of array surfaces concentrate proteins of different types. Details of the commonly used arrays are summarized in Table 6-5.40

A number of variables affect the spectrum of peptide detected with SELDI-TOF or MALDI-TOF. Preanalytical conditions such as gender, age, and general health of the donor, as well as methods used for collection, handling, and storage of specimens all can affect the pattern of peptides in the sample to be tested, thus introducing bias. Most of the SELDI-TOF instruments used in biomarker analysis suffer from signal drift over a period of time, partly because of deterioration in laser power. The various corrections made during the analysis are often subjective and operator dependent. Moreover, serum or plasma—the most commonly tested body fluids—contain thousands of different proteins and peptides, which vary widely in abundance. Because SELDI-TOF can identify high-abundance molecules most readily, putative markers for early diagnosis that are present in low amounts may be missed. Since tumors often produce specific proteins in very small amounts, the high-abundance proteins can mask detection of useful biomarkers. Removal of abundant proteins before analysis can improve sensitivity, but these special techniques add complexity and potential variability to the entire procedure. Overall, lack of sensitivity, interference from nonspecific protein species, and poor reproducibility are important

Table 6-3. A Comparison of Proteomic Techniques Used in Ovarian Cancer Screening

Technique

Use

Method

Advantage

Disadvantage

2D-PAGE

Discovery

Separation of proteins based on charge and size

Robust and reproducible method for biomarker discovery Thousands of proteins can be separated on the gel

Low-abundance proteins and highly acidic or basic proteins may not be visualized Different gels needed for comparing samples Artifacts due to gel-to-gel variations Low throughput; needs much starting material and automation is difficult

DIGE

Discovery

Equal amount of proteins from different samples are labeled with Cy2, Cy3 or Cy5 dyes and run on the same gel

Protein expression differences from different samples can be studied in the same gel Robust technique that is sensitive, and quantitation of proteins is possible Detects PTM and alternative spliced forms

Can visualize high abundance proteins only Low throughput Laborious

iCAT

Discovery

Two protein samples are labeled with normal and heavy versions of hydrogen Cysteine residues bind to the reagent

High-throughput method, which allows direct identification of proteins by MS-MS analysis Automation is possible

Poor detection of alternative splice forms Two variations only—light and heavy Can identify cysteine-containing proteins only

iTRAQ

Discovery

Uses 4 tag reagents that bind to the amine groups in peptides Following digestion, and labeling, samples are subjected to LC-MS/MS

Global labeling and direct identification of proteins Absolute quantitation

Low throughput, time-consuming, and chances of experimental variation

Protein arrays

Validation

Antigen-antibody based binding on spotted arrays

Very high throughput, highly quantitative, detects PTM and splice variants

Must have high-affinity antibodies

SELDI-TOF

Discovery

Depending on the type of chip surface used, specific proteins will be attached to the spot, which is then analyzed by laser desorption and time of flight

High throughput, small amount of sample, more reproducible than 2DE, analysis of crude samples

Low resolving power, not standardized

ELISA

validation

Antigen-antibody based test on microwells

Quantitative, inexpensive

Time-consuming, singleprotein detection

SELDI-TOF, surface-enhanced desorption and ionization-time of flight.

Table 6-4.

Differences Between SELDI-TOF and MALDI-TOF

SELDI-TOF

MALDI-TOF

Analytes directly applied to the chips

Analytes premixed with matrix and dried on a passive surface

Proteins interact with the chromatographic surface and get sequestered according to the type of surface

On-the-spot washing with appropriate buffers

Pre-target deposition sample clean-up not necessary

Chip surface provides good support for co-crystallization of the matrix and protein(s)

No sample loss, good reproducibility

Higher throughput capability, requires significantly lower

No interaction with the surface Washes not possible

Essential to reduce chemical noise or ion suppression (e.g., prefractionation)

Matrix is either applied onto the surface, or proteinmatrix interaction occurs outside

Sample loss amounts of the sample

Has small range of sensitivity, offers higher resolution at High accuracy in the low-molecular-weight range low mass ranges, and is easy to use

MALDI-TOF, matrix-assisted laser desorption and ionization-time of flight; SELDi-TOF, surface-enhanced laser desorption and ionization-time of flight.

Table 6-5. Summary of Available Types and Applications of SELDI-TOF Arrays

Type of Array

Description

Applications

IMAC30

Immobilized metal affinity capture array with a nitrilo-acetic acid (NTA) surface Activated with transition metals prior to use (copper, gallium)

Histidine-tagged protein capture, metal-binding proteins, phosphoprotein profiling and protein purification

Q10

Wrong anion exchange array with quaternary amine functionality

Positively charged that acts as a strong anion exchanger

Protein profiling and protein purification

H50

Bind proteins through reverse phase or hydrophobic interaction chromatography

Protein profiling of hydrophobic or membrane-bound proteins

CM10

Weak cation exchange array with carboxylate functionality Negatively charged that acts as a weak cation exchanger

Protein profiling and protein purification

H4

Mimic reversed phase chromatography with C16 functionality

Typical applications include protein profiling of hydrophobic or membrane-bound proteins, peptide analysis and on-chip desalting

RS100

Contain a preactivated coating with carbonyl diimidazole Allow binding of proteins to the chip surface via any free-amine groups present on the surface of the protein or antibody

The reaction is a nondenaturing process so the antibody or protein remains in its correct or active confirmation

DNA: protein, protein: protein interactions, protein: ligand or affinity capture experiments

SELDi-TOF, surface-enhanced laser desorption and ionization-time of flight.

SELDi-TOF, surface-enhanced laser desorption and ionization-time of flight.

limitations that need to be overcome before patterns of peptides can be used to screen the general population.

The second approach to proteomic techniques appears more promising in the short run. SELDI, for example, has been used to identify individual serum biomarkers that distinguish patients with early-stage ovarian cancer from healthy individuals.41 Of these, alterations in three biomarkers consistently distinguished ovarian cancer patients from healthy individuals: apolipoprotein A1, transthyretin, and CTAPIII.42 A combination of CA-125 and these three proteomic markers provides 87% sensitivity at 98% specificity.

Multiplex Assays

Given the large number of potential markers, identifying an optimal panel poses a significant logistical challenge. Supplies of sera in tissue banks are limited, and each conventional assay can require as much as 100 to 200 ||L. Multiplex assays, such as those developed by Luminex Corporation, circumvent this problem through miniaturization, permitting simultaneous assay of multiple biomarkers (more than 20) using as little as 50 ||L of serum. Small polystyrene microspheres (approximately 6 |im) are internally dyed with different ratios of two spectrally distinct red fluoro-phores to create up to 100 different shades of beads, each with a unique spectrum. Each shade of bead can be coated with a specific antibody that recognizes a particular biomarker. Second antibodies are then identified that bind to distinct epitopes on each biomarker. Fluorescent probes are linked to each second antibody, creating multiple heterologous "double determinant" assays. The beads are allowed to react with serum or plasma. After washing, fluoresceinated secondary antibodies are added and the doubly fluorescent beads are then analyzed by flow cytometry. The shade of red indicates the biomarker being measured, and the shade of green indicates its quantity. The ability to mix multiple conjugated beads and labeled antibodies with each sample permits the simultaneous assay of multiple biomarkers with small volumes of serum. This method43 has been used to distinguish sera from early-stage ovarian cancer. A combination of five markers (IL-6, IL-8, VEGF, soluble-EGF [endothelial growth factor], and CA-125) could distinguish early-stage ovarian cancer from women with no cancer with a sensitivity of 84% and a specificity of 95%.

Validation of Biomarker Panels

Potential panels of biomarkers identified with a set of "training" sera must be tested on a completely independent "validation" set of sera from early-stage and late-stage ovarian cancer patients and from relevant groups of healthy controls. This rigorous validation is important before the test becomes available for clinical use. Several potential candidate biomarkers have yet to undergo rigorous validation. One panel of biomarkers—leptin, prolactin, osteopontin, IGF-II [insulin-like growth factor II], M-CSF, and CA-125—was recently marketed as Ovasure (LabCorp) to detect early-stage disease in high-risk individuals. Given that the single publication describing this panel included only a small number of early-stage patients and used the validation set rather than the training set to choose the optimal biomarkers,45 the Society of Gynecologic Oncologists released a statement saying that "additional research was needed." After communication with the FDA, the test has been withdrawn from the market.

Ideally, biomarker panels should not only identify women with stage I disease but also detect very early ovarian cancer before conventional clinical diagnosis. Samples of serum from women destined to develop ovarian cancer have been preserved from the PLCO and UKCTOCS trials and provide an important resource for biomarker validation. A recent collaboration of the PLCO with the Early Detection Research

Network (EDRN) and the Ovarian SPOREs (Specialized Programs of Research Excellence) at four institutions tested more than 50 markers in preclinical serum specimens. Results should be published in the near future.

10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook


Post a comment