Proteomic biomarkers in combination with CA 125 for detection of epithelial ovarian cancer using prediagnostic serum samples from the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial†
The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations suggest endorsement by the US Government.
When epithelial ovarian cancer is detected at an early stage (I-II), the 5-year survival rate is between 70% and 90%; whereas, when it is detected in late stages (III-IV), the 5-year survival rate slips to <30%. In a previous report, the authors observed that proteomic biomarkers and cancer antigen 125 (CA 125) exhibited a sensitivity of 84% at a specificity of 98% for identifying sera from patients who had stage I disease at the time of surgery, significantly improving the sensitivity of CA 125 alone. The challenge, however, is to detect ovarian cancer before clinical diagnosis. The current study was part of a large effort to compare different multimarker biomarker panels for the early detection of ovarian cancer. Several biomarkers were evaluated alone and in combination with CA 125 in prediagnostically collected sera from women in the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial.
Proximal prediagnostic sera from 118 women with ovarian cancer (cases) and from 951 age-matched women (controls) (8 controls per case, including 4 randomly selected from the general population, 2 with CA 125 levels ≥35 U/mL, and 2 with a positive family history of breast/ovarian cancer) were analyzed using the CA 125 immunoassay and surface-enhanced laser desorption and ionization time-of-flight mass spectrometry to measure 7 proteins (apolipoprotein A1, truncated transthyretin, transferrin, hepcidin, β-2 microglobulin, connective tissue activating protein III), and interalpha-trypsin inhibitor heavy-chain 4). Data were analyzed by 2 statistical strategies that combined the 7 markers and CA 125 into 1 predictive score for disease classification.
CA 125 levels were elevated (≥35 U/mL) in 61.5% of 65 patients who had CA 125 data available from samples that were collected <12 months before cancer diagnosis; however, levels of the additional 7 biomarkers were not different between cases and the 3 control groups individually or combined. Two panels that combined CA 125 and the 7 biomarkers failed to improve the sensitivity of CA 125 alone.
In contrast to earlier findings from analyzes of postdiagnostically collected sera, the addition of 7 biomarkers to CA 125 did not improve sensitivity for preclinical diagnosis beyond CA 125 alone. Cancer 2012;. © 2011 American Cancer Society.
Despite recent advances in ovarian cancer treatment, new methods of early detection remain of paramount importance, because they have the potential to clearly improve long-term cancer survival. Currently, >66% of ovarian cancer cases are detected at an advanced stage, resulting in poor overall 5-year survival rates of 10% to 30%.1 This is in contrast to patients with stage I/IIA disease, who have an approximately 90% 5-year survival rate. Longitudinal studies are ongoing in several countries to evaluate screening strategies using cancer antigen 125 (CA 125) and/or transvaginal sonography (TVS) and their impact on overall cancer detection and mortality. Recently, through screening with CA 125 and TVS in the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial, investigators were able to identify both early and late-stage malignancies; however, the predictive value of each was surprisingly low because of a high proportion of false-positive diagnoses.2 Other existing serum markers that have been identified to date have not proven adequate sensitivity or specificity for screening. Thus, new biomarkers that could improve the early detection for ovarian cancer are critically needed. Such methods could be used, either alone or in combination with existing methods, to improve diagnostic performance.
Proteomic methods have been used to detect several promising serum biomarkers that can distinguish patients with early stage ovarian cancer from healthy individuals.3 A 7-biomarker panel that included transthyretin (TT), apolipoprotein A1 (Apo-A1), beta-2 microglobulin (β2M), transferrin (TRFR), hepcidin (HEPC), connective tissue activating protein III (CTAPIII), and interalpha-trypsin inhibitor heavy-chain 4 (ITIH4), has been evaluated not only for early detection4 but also for distinguishing malignant from benign pelvic masses.5 Immunoassays for Apo-A1, TT, and ITIH4 in combination with CA 125 improved upon the sensitivity of CA 125 alone for detecting early stage disease.4 Sensitivity increased from 65% with CA 125 alone to 74% with the addition of Apo-A1, TT, and ITIH4, when specificity was held constant at 97%. Subsequently, Moore et al6 replicated these findings using postdiagnostic/pretreatment serum samples from ovarian cancer cases and hospitalized controls who were diagnosed with benign ovarian cysts or digestive diseases. The specificity of a panel of TT markers and Apo-A1 was 96.5%, but sensitivity was only 52.34%. However, when combined with CA 125, the specificity remained high (94.3%), and 78.6% sensitivity was achieved. In another recent study,7 a biomarker panel comprised of Apo-A1, TT, and CTAPIII achieved a sensitivity of 84% at a specificity of 98% for detecting early stage ovarian cancer.
All prior studies were performed with sera obtained at the time of clinical diagnosis. Ideally, a screening strategy should detect ovarian cancer before it becomes symptomatic or readily palpable. Consequently, as part of a large multicenter study to compare different multimarker panels for the early detection of ovarian cancer, we evaluated the discriminatory power of a 7-biomarker panel using prospectively collected, prediagnostic serum samples from 118 asymptomatic women enrolled in the PLCO Cancer Screening Trial.8 For each case, 8 matched, healthy controls were used as a comparison group. In addition to 4 population controls that had been described in a previous report,9 our analysis also included 2 high-risk comparison groups; 2 matched controls who had ever had an elevated CA 125 level (≥35 U/mL), and 2 women who had a family history of breast or ovarian cancer. The sensitivity and specificity of this expanded marker panel alone and in combination with CA 125 was evaluated for the ability to discriminate women with ovarian cancer from healthy women using prediagnostic serum samples and to determine whether this combined marker panel could perform better than serum CA 125 alone, as we observed previously in investigations using postdiagnostic serum samples and controls with benign ovarian or digestive diseases.6
MATERIALS AND METHODS
Phase 3 studies were coordinated by PLCO investigators.2, 8 Between 1998 and 2006, 118 cases of invasive ovarian, primary peritoneal and fallopian tube cancers were identified and histologically confirmed. Women with tumors of borderline malignancy were excluded as well as those with a history of cancer at baseline. Controls were selected from a pool of healthy individuals who remained cancer-free and were matched to cases by 5-year age categories at blood draw, calendar year of blood draw, and calendar year (in categories) of cancer diagnosis. For each case, 8 healthy controls were selected. Four controls per case were selected randomly from a pool of all eligible controls and are referred to as general population controls. Two controls per case were selected among women who reported having a positive family history of breast or ovarian cancer. Two additional controls per case were selected who were healthy but, at some time point during the PLCO trial sampling period, had an elevated serum CA 125 level (≥35 U/mL), allowing examination of the biomarker panel among a high-risk group of women who would be more likely to be targeted for ovarian cancer screening than healthy women in the general population. These women with at least 1 positive CA 125 test result continued to undergo annual screening. Sixty replicate pairs were randomly inserted into the batches for blinded quality control, and 10 pairs had CA 125 levels >25 U/mL. For all cases, the serum sample most proximate to the case diagnosis was selected for this investigation. Biomarker information was missing from 1 case, and CA 125 levels were missing from 1 of 51 cases who were diagnosed >1 year before blood draw, leaving 50 cases in each group for analysis. Two cases were missing CA 125 levels in the group diagnosed within 1 year of blood draw, leaving 65 of 67 cases for analysis. Two controls with high CA 125 levels during the PLCO trial and 6 population controls had missing biomarker data. Additional details of the study design have been described previously.8 The collection of biospecimens was approved by the National Cancer Institute Special Studies Institutional Review Board (OH-C-N041) at the US National Institutes of Health and by the local institutional review board for each screening site. Informed consent was obtained from all women who provided blood samples to be stored for future research. Approval for this study to use biologic specimens to evaluate biomarkers for early detection of ovarian cancer was granted through a peer review process administered by the PLCO Etiologic and Early Marker Studies program.
Sample Storage and Processing
All aliquoting for phase 3 studies investigating early detection markers for ovarian cancer was coordinated by PLCO investigators using a common sampling plan for several studies. The serum sample closest to and before diagnosis was selected for laboratory analysis. The interval from the date of blood draw to diagnosis ranged from 12 days to 2898 days, and 67 samples were collected within 1 year. Blood was drawn before case diagnosis and was processed within 2 hours. Other details of sample storage and processing have been described. During processing, all specimens were organized so that equal proportions of cases, high-risk subgroup controls, general population controls, and quality-control replicates were included per batch of 96 samples. Aliquots were shipped overnight on dry ice by Federal Express to The University of Texas MD Anderson Cancer Center (MDACC) (Houston, Tex) and to Vermillion, Inc. (Freemont, Calif) for processing and were stored at −70°C before processing.
CA 125 was measured and recorded by PLCO investigators using the CA 125II radioimmunoassay (Centocor, Inc., Malvern, Pa) according to manufacturer's instructions.2, 8 Apo-A1 (mg/dL), TT (mg/dL) and ITIH4 (pg/μL) were measured as described previously.4, 6, 10 Levels of CTAPIII, TRFR, HEPC (pg/μL), and β2M (measured as the total ion current) were estimated based on a relative quantification method that compared peak heights observed in study samples with references on a standard curve. All samples were analyzed using mass spectrometry assays with surface enhanced laser desorption/ionization mass spectrometry, as described previously (Bio-Rad Laboratories, Inc., Hercules, Calif).
Arrays were read in a PCS 4000 Protein Chip Reader (Ciphergen Biosystems, Inc., Fremont, Calif) and a time-lag, focus-in, linear laser desorption/ionization-time-of-flight mass spectrometer. Instruments were calibrated daily. Spectra were acquired in the positive-ion mode. The sampling rate was set at 800 megahertz. Ions were extracted using a 3.4-kilovolt extraction pulse and were accelerated to a final velocity using 25-kilovolt acceleration potential. The system used a pulsed nitrogen laser at a repetition rate of 20 hertz. Laser pulse energy from 1500 to 2000 nanojoules was delivered into a 100-μM diameter area. This illuminated area was rastered across a 2-mm diameter sample spot. An automated analytical protocol was used to control the data acquisition process throughout the analysis. Each spectrum averaged at least 1000 laser shots and was externally calibrated against a mixture of known peptides/proteins.
Raw data obtained from the PCS4000 Protein Chip Reader first were smoothed by a fixed-width, moving average filter of 25 data points. Subsequently, a convex hull baseline subtraction algorithm was applied to the smoothed data. Data were then internally normalized using total ion current with the Ciphergen Express 3.0 Software (Ciphergen, Inc.). Peaks corresponding to each biomarker were manually labeled and their intensities recorded from the protein chip array data while blinded to disease status. Samples were analyzed in triplicate. After spectral data were collected, archive spectral data was imported into OvaCalc Software (version 3.2.7; Vermillion, Inc.) using Protein Chip Data Manager Software to detect peak intensities and perform calculations for data processing including; baseline subtraction (15 times expected peak width based on mass), spectral filtering (0.2 times expected peak width), starting mass (blanking mass + 20%), and normalization to a factor of 1 (1/average raw ion current). After all 4 data processing steps, data were reported as a “normalized” total ion current. Six of the 7 biomarkers (excluding β2M) were processed further by comparing the total ion current normalized to a referent standard curve to provide absolute quantification.
Levels of the biomarkers across groups were compared using the Kruskall-Wallis median analysis of variance (PROC NPAR1WAY; SAS version 9; SAS Institute, Inc., Cary, NC). Logistic regression models were used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) of associations between marker levels categorized into quartiles (based on control distributions) and ovarian cancer cases (PROC LOGISTIC). In addition to marker levels, logistic models included CA 125 as either a continuous or categorical variable (≥35 U/mL) and age (in categories). Several classification models were built using k-nearest-neighbor classifiers with PROC DISCRIM.11, 12 This nonparametric method was used because it does not make assumptions about marker distributions within each group. These models combined the marker levels (as continuous variables) and CA 125 (as either continuous or a categorical variable; 35 U/mL vs <35 U/mL). Misclassification rates for the k-nearest neighbor models were assessed using leave-1-out cross-validation to obtain unbiased estimates of the predictive capability of each model.13
In a second combined marker analysis, a linear prediction model that included only the proteomic biomarkers was derived using sera from 41 patients with early stage ovarian cancer and 99 healthy individuals from the MDACC sample bank. Linear discriminant analysis14 with bootstrap was used to produce a single-valued index.10 For the current study, the unblinded portion of the PLCO samples (n = 561, including 59 cases) was used to derive a new model. The new model, which was a combination of the previously derived linear prediction model, CA 125, and the 7-biomarker panel, was tested on the remaining blinded portion of the PLCO samples.
Characteristics of study participants are provided in Tables 1 and 2 for all ovarian cancer cases and for the 3 control subgroups. Case and control groups were balanced for age, race, smoking history, and use of hormone replacement therapy. Among the 118 ovarian cancer cases, 57 patients (47.3%) had CA 125 levels ≥35 U/mL. Histologic features for the case group are provided for the 67 women who were diagnosed within 1 year of blood draw and for the 51 women who were diagnosed >1 year before blood draw. In both groups, high-grade serous cancers that presented at a late stage (III-IV) were the most prevalent. Elevated CA 125 levels (≥35 U/mL) were observed in 40 of 65 serum samples (61.5%) from women who were diagnosed within 1 year of phlebotomy but in only 1 of 50 serum samples (2%) from women who were diagnosed >1 year after the last blood draw.
Table 1. Characteristics of the Study Population: Cases and Controls
|Age at serum draw, y|| || || || || || || || |
|Race|| || || || || || || || |
|Elevated CA 125b|| || || || || || || || |
| Yes (≥35 U/mL)||41||34.7||90||37.8||0||0||7||1.5|
|Family history of breast/ovarian cancer|| || || || || || || || |
Table 2. Characteristics of the Study Population: Cases Only
|Histologic subtype: Total||51||43.2||67||56.8|
| Serous cystadenoma||28||54.9||40||59.7|
| Serous cystadeocarcinoma||1||2||1||1.5|
| Mucinous cystadenocarcinoma||1||2||0||0|
| Endometrioid tumor||6||11.8||7||10.4|
| Clear cell cystadenocarcinoma||1||2||3||4.5|
| Undifferentiated carcinoma||1||2||1||1.5|
| Adenocarcinoma, NOS/carcinoma NOS||11||21.6||11||16.4|
| Granulosa cell tumor, malignant||1||2||2||3|
|Grade|| || || || |
| Gx: Cannot be assessed||3||5.9||1||1.5|
| G1: Well differentiated||5||9.8||2||3|
| G2: Moderately differentiated||9||17.6||13||19.4|
| G3/G4: Poorly differentiated4||25||49||41||61.2|
|Stage|| || || || |
|Elevated CA 125b|| || || || |
| Yes (≥35 U/mL)||1||2||40||59.7|
In Table 3, means, medians, and standard deviations of the 7 biomarkers markers and CA 125 levels are presented. P values are given for comparison of cases diagnosed with ovarian cancer within 1 year of diagnosis with each of the control subgroups and all controls combined. Only CA 125 levels were statistically higher among this subset of cases compared with all controls and for each control subgroup. This difference was not observed in serum samples from cases who were diagnosed >1 year before diagnosis, because both the mean and median levels of CA 125 were much lower in this subgroup of cases compared with those who were recently diagnosed. Table 4 lists the ORs for biomarkers (in quartiles) mutually adjusted and also adjusted for CA 125 as a dichotomous variable for the ovarian cancer cases diagnosed within 1 year before blood draw compared with all controls and with individual control subgroups. Only levels of CA 125 were significantly elevated among all controls combined and across control subgroups (excluding healthy controls who were selected for having a family history of breast and ovarian cancers, because there were no controls with elevated CA 125 levels).
Table 3. Unadjusted Group Means and Median Test for Marker Differences by Disease Status and Across Control Group Subtypes
|Serum CA 125, U/mL||12.6±7.2||11.0||320.6±759.8||58.0||16.2±12.6||11.0||30.8±15.2||30.0||11.0±5.6||9.0||11.5±6.5||10.0|
Table 4. Odds Ratios and 95% Confidence Intervals for Individual Markers in Multivariate Models Among Cases Diagnosed <1 Year of Blood Draw Compared With all Controls and Control Subgroups
|CA 125, U/mLc|| || || || || || || || || || |
| <35||Ref|| || ||Ref|| ||NA|| ||Ref|| || |
| ≥35||13.94||7.91-24.56||<.0001||2.56||1.41-4.63||NA|| ||111.11||4348-250.00||<.0001|
|Apo-A1, mg/dL|| || || || || || || || || || |
| Q1||Ref|| || ||Ref|| ||Ref|| ||Ref|| || |
| Q2||0.51||0.21-1.24|| ||0.53||0.21-1.32||0.50||0.18-1.30||0.48||0.19-1.23|| |
| Q3||1.20||0.57-2.51|| ||1.37||0.63-2.97||0.80||0.34-1.85||1.08||0.49-2.38|| |
|β2M, TIC|| || || || || || || || || || |
| Q1||Ref|| || ||Ref|| ||Ref|| ||Ref|| || |
| Q2||0.88||0.39-2.01|| ||1.03||0.44-2.38||0.82||0.32-2.08||0.67||0.28-1.61|| |
| Q3||1.00||0.43-2.34|| ||1.16||0.48-2.79||0.80||0.31-2.04||0.86||0.35-2.13|| |
|CTAPIII, pg/μL|| || || || || || || || || || |
| Q1||Ref|| || ||Ref|| ||Ref|| ||Ref|| || |
| Q2||0.87||0.37-2.05|| ||0.99||0.41-2.39||0.88||0.34-2.28||0.70||0.28-1.75|| |
| Q3||1.24||0.55-2.78|| ||1.33||0.58-3.12||1.16||0.47-2.88||0.95||0.44-2.51|| |
|HEPC, pg/μL|| || || || || || || || || || |
| Q1-Q2||Ref|| || ||Ref|| ||Ref|| ||Ref|| || |
| Q3||0.80||0.39-1.63|| ||0.82||0.39-1.72||0.85||0.38-1.89||0.70||0.33-1.49|| |
|ITIH4, pg/μL|| || || || || || || || || || |
| Q1-Q2||Ref|| || ||Ref|| ||Ref|| ||Ref|| || |
| Q3||1.31||0.64-2.71|| ||1.45||0.70-3.04||1.02||0.45-2.30||1.18||0.55-2.56|| |
|TRFR, pg/μL|| || || || || || || || || || |
| Q1||Ref|| || ||Ref|| ||Ref|| ||Ref|| || |
| Q2||0.60||0.25-1.40|| ||0.60||0.25-1.45||0.54||0.21-1.39||0.57||0.23-1.42|| |
| Q3||0.57||0.25-1.32|| ||0.56||0.24-1.31||0.61||0.24-1.56||0.54||0.22-1.33|| |
|TT, mg/dL|| || || || || || || || || || |
| Q1||Ref|| || ||Ref|| ||Ref|| ||Ref|| || |
| Q2||0.51||0.21-1.23|| ||0.54||0.22-1.34||0.44||0.17-1.20||0.49||0.19-1.25|| |
| Q3||0.98||0.48-2.03|| ||0.96||0.45-2.02||1.09||0.47-2.51||1.04||0.47-2.29|| |
In Table 5, the cross-validated sensitivity and specificity estimates for 5 prediction models are presented for cases that were diagnosed within 1 year of blood draw and for controls. Model 1 classified women only based on having elevated CA 125 (<35 U/mL vs ≥35 U/mL). Models 2 through 4 were estimated using the k-nearest neighbor algorithm. Model 2 only included continuous CA 125 levels; Model 3a included all continuous biomarkers levels and CA 125 level as a dichotomous variable, comparing cases with all controls combined; Model 3b included all of the above with the exception that cases were compared with controls, and each control group was treated separately in the classification; Model 4a included all biomarkers and continuous CA 125 levels with cases compared with all controls; and Model 4b was the same as Model 4a but treated each control group separately in the classification. When CA 125 was considered as a dichotomous variable, the sensitivity of the models was higher than when continuous levels were used (sensitivity, 61.5% [95% CI, 48.6%-73.4%] vs 41.5% [95% CI, 29.4%-54.4%]); however, the specificity was 100% compared with 89.8% (95% CI, 87.7%-91.7%) when dichotomized. Models that included the 7 biomarkers and CA 125 exhibited much lower sensitivity than models that included only CA 125, whether it was used as a categorical or continuous marker, although the specificity remained high (≥98.5%) in each case.
Table 5. Cross-Validated Sensitivity and Specificity Estimates for Various Prediction Models Using the k-Nearest Neighbor Algorithm
|1||40/65||61.5 (48.6-73.4)||854/951||89.8 (87.7-91.7)|
|2||27/65||41.5 (29.4-54.4)||951/951||100 (99.6-100)|
|3a||5/65||7.7 (2.6-17.1)||929/943||98.5 (97.5-99.2)|
|3b||8/65||12.3 (5.5-22.8)||929/943||98.5 (97.5-99.2)|
|4a||12/65||18.5 (9.9-30)||939/943||99.6 (98.9-99.9)|
|4b||18/65||27.7 (17.3-40.2)||939/943||99.6 (98.9-99.9)|
The second analytical approach that combined CA 125 and the biomarker panel also failed to improve on the performance of CA 125 alone. For the unblinded training samples, at a fixed specificity of 95%, the sensitivity for CA 125 was 32.2% (95% CI, 21.3%-44.6%), whereas the sensitivity of the multivariate model was 35.6% (95% CI, 24.3%-48.1%). By using the same cutoffs, the sensitivity of CA 125 in the blinded test set was 35.1% (95% CI, 23.7%-47.8%), yet the specificity dropped to 92.5% (95% CI, 89.9%-94.7%). The model's sensitivity on the test set was only 26.3% (95% CI, 16.2%-38.5%) at a specificity of 93.4% (95% CI, 90.9%-95.4%).
This study was performed in the context of a cooperative trial with other Specialized Programs of Research Excellence (SPORE), Early Detection Research Network, and PLCO investigators that evaluated some 35 different biomarkers in proximal prediagnostically collected serum samples from 118 women who subsequently developed ovarian cancer. In evaluating samples that were obtained before conventional diagnosis, the addition of 7 biomarkers to CA 125 in a combined multimarker panel did not improve sensitivity over that obtained with CA 125 alone at 98% specificity. CA 125 was used originally to monitor women with verified ovarian cancer,15, 16 because changes in CA 125 levels can track the progression or regression of disease during treatment with up to 90% accuracy. Persistent elevation of CA 125 after surgery and chemotherapy indicates the presence of residual disease. When multiple serum samples are evaluated, rising levels of CA 125 can detect recurrent ovarian cancer with an average lead time of at least 3 months. CA 125 also can be elevated many months before cancer diagnosis.16 In the current study, using 1 prediagnostic serum sample from PLCO trial participants, CA 125 levels were elevated (≥35 U/mL) in 61% of patients who were sampled within the 12 months before diagnosis and in 31% of patients who were sampled >1 year before diagnosis. Observation of elevated CA 125 levels in a significant fraction of patients >12 months before diagnosis can be considered an encouraging outcome.
Despite the observed lead time in the prediagnostic samples, the positive predictive value of CA 125 was only 3.7% in the PLCO trial. Although the specificity of CA 1252 can approach 99% in postmenopausal women, given the very low prevalence of ovarian cancer (1 in 2500 among postmenopausal women), 99.6% specificity is required to achieve a positive predictive value of 10%, which means that there are <10 surgeries for each diagnosed ovarian cancer case. The specificity of CA 125 was improved by combining biomarker results with TVS. In the PLCO trial, if CA 125 was elevated and a TVS was abnormal, then the positive predictive value increased to 23.5%, although 60% of the invasive cancers would not have been detected.2
The specificity of CA 125 also can be increased by studying changes in levels over time. An algorithm using age and change point analysis to determine whether levels have increased beyond the patient's own baseline compared with annual determinations17 is being evaluated in a study of 202,638 British woman (the United Kingdom Collaborative Trial of Ovarian Cancer Screening).18 TVS results combined with rising CA 125 levels with time produced a sensitivity of 89.5%, a specificity of 99.8%, and a positive predictive value of 35%. Overall, 48% of the prevalent cancers were detected in stage I or II, which is twice the fraction expected using conventional diagnostic methods. In the United States, there was a similar screening trial of 3252 postmenopausal women who were followed annually with CA 125 using the same algorithm coordinated by the Ovarian SPORE at MDACC.19 Less than 1% of patients were referred for TVS each year, and <3% were referred over multiple years. Overall, the positive predictive value for the entire screen was 37%, which was consistent with the British study.
Only 80% of invasive ovarian cancers express significant quantities of CA 125.15 Consequently, numerous biomarkers have been evaluated to complement or replace CA 125. By using multiplex assays, some 96 biomarkers have been tested for the ability to distinguish healthy individuals from women with stage I ovarian cancer.20 A 4-biomarker panel that included CA 125, human epididymous protein 4 (HE4), carcinoembryonic antigen, and soluble vascular adhesion molecule produced 86% sensitivity for early stage (I/II) disease and 95% sensitivity for late-stage (III/IV) disease at 98% specificity. Similar sensitivity and specificity were observed at the time of conventional diagnosis with the proteomic biomarkers evaluated in the current study. A panel of Apo-A1, TT, and CTAPIII produced 87% sensitivity at 98% specificity for distinguishing stage I ovarian cancer from healthy individuals.7
The discrepancy in the results in prediagnostic samples for the 3-biomarker panel compared with our previous studies using postdiagnostic samples4, 6 may relate to the findings that Apo-A1, TT, and CTAPIII are each acute-phase reactants that are down-regulated in response to cancer and that a significant volume of tumor may be required to trigger this reaction. Although stage I disease, by definition, is limited to the ovaries, the volume of cancer can be substantial but must be sufficient to produce symptoms or a readily palpable pelvic mass to result in conventional diagnosis. In preclinical disease, the volume of tumor may not be sufficient to evoke an acute-phase response and alter levels of proteins associated with this process. Another issue with the markers in our panel (with the exception of CA 125) is that they were lower in cancer cases but also can be reduced as a result of storage time, sample processing, and other host factors (described in our previous report),6 which could hinder the identification of small-volume disease, which was a goal of the current study. Finally, in the current study, we clearly attempted to reduce confounding and bias in ways that were not possible in our previous study, in which we matched cases to hospital controls rather than healthy women, and we were unable to control sample processing or match on storage time, which may have the biased results.
Although a marginally better performance was observed for the identification of cases at least 6 months before diagnosis using an all-site multimarker panel (which included CA 125, HE4, tumor-associated glycoprotein 72 [CA 72-4], substance P-like immunoreactivity, andβ2M) were observed compared with CA 125 alone, the increase was not statistically significant.21 In addition to the current study, 5 additional panels were evaluated, none of which improved on the results with CA 125 alone.8 Considering the failure of multiple biomarkers to improve upon CA 125 in prediagnostic samples, new approaches are badly needed for biomarker discovery. One weakness of the current study is that we were unable to evaluate markers in nonwhite populations because of a very small number of nonwhite cases in the PLCO trial. The results of this combined effort will likely reshape our approach to biomarker discovery and validation. In addition to searching for protein analytes, autoantibodies also may be sought. Finally, previous studies have had limited success in identifying and evaluated autoantibodies of human proteins expressed in bacteria or insect cells. Recent advances in expressing human proteins in human cells could allow the identification of new epitopes that are selective for altered tertiary structure and glycosylation status of selected protein targets.
This work was supported by funds from The University of Texas MD Anderson SPORE in Ovarian Cancer from the National Cancer Institute (grant P50 CA83639); by philanthropic support from Golfers Against Cancer, the Tracey Jo Wilson Foundation, and the Mossy Foundation; and by the Intramural Research Program of the National Institutes of Health, National Cancer Institute, Division of Cancer Epidemiology and Genetics.
CONFLICT OF INTEREST DISCLOSURES
R.C.B. receives royalties for the discovery of CA 125 and has served on the Scientific Advisory Boards of Fujirebio Diagnostics, Inc., and Vermillion, Inc.