| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Oncology, Markers, Clinical Correlates |
Departments of Surgery [T. E. G., S. R., L. A. K., J. D. L.], Human Genetics [T. E. G.], Pathology [S. D. F.], and Biostatistics [W. E. G.] and the University of Pittsburgh Cancer Institute, Pittsburgh, Pennsylvania 15213
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Experimental Design: CEA expression was compared with clinical outcomes to determine correlation with disease recurrence. For quantitative data, an optimum CEA expression level cutoff value was defined as the value that most accurately classified patients on the basis of disease recurrence. Kaplan-Meier survival curves were generated, and multivariate analyses were performed to evaluate the prognostic value of QRT-PCR.
Results: CEA expression levels were above the optimum cutoff level in 12 tissue blocks, resulting in the identification of 11 CEA-positive patients. Of these patients, 9 suffered disease recurrence and 2 remain disease free. Of the 19 CEA-negative patients, there was 1 disease recurrence. The sensitivity and specificity for predicting disease recurrence were 90 and 90%, respectively. Kaplan-Meier analysis showed that CEA positivity resulted in significantly lower disease-free and overall survival (P <0.0001 and 0.0006 respectively). In multivariate analyses, CEA positivity measured by QRT-PCR was the strongest independent predictor of disease recurrence among other clinical and pathological factors examined.
Conclusions: QRT-PCR offers significant benefits over standard RT-PCR and identifies node-negative patients at high risk for recurrence.
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The main problems with current lymph node evaluation are sampling error and poor sensitivity for detecting individual tumor cells or small tumor foci. Histological examination only samples a very small percentage of each lymph node, and it has been calculated that a pathologist has only a 1% chance of detecting a micrometastatic focus of three tumor cell diameter (6) . Immunohistochemical staining for tumor markers improves the sensitivity of micrometastasis detection and, when combined with serial sectioning to reduce the sampling error, results in upstaging of a significant number of patients. This technique has been used in esophagus cancer, and 17% of histologically negative nodes were found to contain micrometastatic disease (7) . In this report, IHC3 -positive lymph nodes correlated with disease recurrence, but these findings were questioned in a later study in which IHC did not show any prognostic value (8) . Other studies have used molecular methods such as RT-PCR to detect micrometastases. RT-PCR is capable of detecting the mRNA for tumor markers, such as CEA, cytokeratin 19, cytokeratin 20, and others, in a variety of tissues, including blood, bone marrow, and lymph nodes, that are histologically cancer free (9, 10, 11, 12, 13) . In esophageal cancer, RT-PCR has been used on several small series of patients, but the clinical significance of RT-PCR-positive nodes is not known because little clinical follow-up has been reported (14, 15, 16) . In other tumor types, studies have shown that RT-PCR improves sensitivity (14, 15, 16) , but poor specificity and false-positive results in control lymph nodes from patients without cancer have made the clinical application of this information difficult (17 , 18) . False positives are, at least in part, attributable to the previously described phenomenon of ectopic gene expression, which results in very low background levels of expression of most genes in all tissue types (19 , 20) . Thus, although previous studies have used qualitative, gel-based RT-PCR methods, it is now becoming apparent that this simple plus/minus method for detection of tumor markers is not always a reliable sign of micrometastases. With the introduction of the fluorescent 5' nuclease assay, QRT-PCR is now a relatively simple technique. We hypothesized that a quantitative analysis would offer significant benefits over gel-based RT-PCR and would allow us to accurately predict disease recurrence in histologically node-negative esophageal cancer patients.
The objectives of our study were 3-fold: (a) we determined the ability of QRT-PCR (21) to accurately distinguish between background gene expression of CEA in lymph nodes and clinically relevant levels that are diagnostic of true micrometastasis; (b) we used real-time QRT-PCR to analyze lymph nodes from 30 node-negative esophageal cancer patients and correlated the results with disease recurrence; and (c) we compared QRT-PCR with standard gel-based RT-PCR on the same samples. We found that QRT-PCR can easily discriminate background expression from true metastatic disease, QRT-PCR is both sensitive and specific for predicting disease recurrence in node-negative esophageal cancer patients, and QRT-PCR has greater specificity than gel-based RT-PCR. Thus, we believe that quantitation addresses all of the problems that have kept RT-PCR from becoming a useful clinical test for micrometastatic disease.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Tissue blocks from 10 primary tumors (8 adenocarcinoma and 2 squamous cell carcinoma) and 4 histologically positive lymph nodes were obtained as positive controls. Negative control (benign) lymph nodes were obtained from patients who underwent esophageal surgery for causes unrelated to cancer (hernia repairs and antireflux procedures) and whose lymph nodes were removed incidentally.
|
QRT-PCR.
QRT-PCR was carried out using the 5' nuclease assay and an Applied Biosystems 7700 Sequence Detection Instrument (TaqMan). CEA expression was measured relative to the endogenous control gene, ß-Gus, using the comparative CT method described previously (22
, 23)
. All QRT-PCR assays were carried out at two RNA inputs, 400 and 100 ng, and duplicate reactions were set up for each concentration. Thus, the reported CEA expression levels are an average of four independent QRT-PCR reactions. RT-negative controls were run for all samples using 400 ng of RNA but omitting the reverse transcriptase. Template-negative controls were also run on each PCR plate. A calibrator RNA sample was amplified in parallel on all plates to allow comparison of samples run at different times (22
, 24)
.
For maximum sensitivity and to eliminate the risk of cross contamination, we used a single-tube QRT-PCR procedure described previously (25)
. In this procedure, physical separation of the reverse transcription reaction mixture (RT) and the PCR primers using a wax layer results in a more specific and sensitive RT-PCR. The PCR primers were pipetted into the PCR plate in a 10-µl volume. One Ampliwax PCR gem 50 (Applied Biosystems) was then placed in each well, and the plate was heated to 80°C for 2 min and cooled to 4°C to produce the wax barrier. A 40-µl upper layer was then pipetted into each well. The final concentrations of the reaction components were as follows: 1x PCR buffer A, 300 nM each deoxynucleotide triphosphate, 3.5 mM MgCl2, 0.4 unit/µl RNase Inhibitor, 1.25 units/µl Superscript II reverse transcriptase (Life Technologies, Inc., Gaithersburg, MD), 0.06 unit/µl Amplitaq Gold (Applied Biosystems), 20 nM reverse transcriptase primer, 200 nM of each PCR primer, 200 nM probe (ß-Gus primers and probe were at 100 nM), and 100 or 400 ng total RNA. The oligonucleotide sequences used are shown in Table 1
. All RT-PCR reactions were carried out on the ABI 7700 with the following thermocycler conditions: 48°C hold for 40 min, 95°C hold for 12 min followed by 45 cycles of 95°C for 15 s, and 64°C (60°C for ß-Gus) for 1 min. Data were analyzed using Sequence Detection Software (Applied Biosystems) with thresholds set at 0.03 for CEA and 0.045 for ß-Gus.
|
Statistical Analysis.
Comparisons of CEA levels in control tissues samples (Fig. 1)
were conducted with the Mann-Whitney U Test. For pathologically negative lymph nodes from esophageal cancer patients, the primary end point was disease recurrence measured from the time of surgery to the time of diagnosed recurrence or last date of follow-up. The highest CEA levels from each patient’s tissue blocks were used to construct a ROC curve using recurrence as the gold standard. The CEA expression level cutoff value was identified that produced the most accurate classification, and that level was used to classify patients as QRT-PCR positive or negative for risk of recurrence. Because a second set of patients was not available for prospective validation of the cutoff, the cutoff selection procedure was evaluated statistically by cross validation, and the SDs of ROC curve statistics were calculated by bootstrap resampling (26)
.
|
All comparisons between models were based on differences between likelihood ratio tests for nested models. The adequacy of the proportional assumption was checked two ways: by plotting the log cumulative hazard by log time and by plotting Schoenfeld residuals and estimating the correlation between the regression coefficient and time.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
.
Quantitative Analysis of CEA Expression.
Initially, RNA was isolated and analyzed from three sources, distinct
from the node-negative study group, and included primary esophageal tumors, lymph nodes that were histologically positive for metastases (N1), and benign lymph nodes from patients without cancer. CEA expression was detected in all tumors, and N1 nodes and in 50% of benign lymph nodes (Fig. 1)
. Individual, pairwise comparisons were carried out using a Mann-Whitney U test, and expression levels in both tumor and N1 nodes were found to be significantly higher than in benign lymph nodes (P = 0.002 and 0.0021, respectively). CEA expression was slightly higher in tumor samples than in N1 lymph nodes, but this was not statistically significant (P = 0.171).
|
. The area under the ROC curve was 0.88 with a 95% confidence of 0.71 to 0.97, indicating that classification accuracy is significantly better than chance alone. Of the 140 tissue blocks analyzed, 12 (9%) had CEA expression levels above the cutoff point, and a total of 11 patients (34%) had one or more blocks with expression above the cutoff. These 11 patients were classified as having QRT-PCR evidence of occult micrometastases. Sections from all but 1 (patient 28) of the "most positive" tissue blocks were negative for immunohistochemical staining with antibodies against CEA.
|
Occult Metastases and Recurrence.
Of the 30 patients studied, 10 suffered disease recurrence and died by the end of the study. Of the remaining 20 patients, 7 died from other causes and 13 remain alive with no evidence of recurrent disease. One patient suffered an early anastomatic recurrence that was treated with photodynamic therapy and radiotherapy. This patient was later diagnosed with, and died of, small cell lung cancer. Using the most accurate cutoff for the quantitative assay, a total of 11 patients were classified as QRT-PCR positive, and 9 of these suffered recurrence. The same 9 patients were identified as RT-PCR positive using the gel-based assay, along with 6 other patients who did not recur. Sensitivity and specificity for predicting disease recurrence using the QRT-PCR assay was 90 and 90%, with a positive predictive value of 82%. Using the gel-based assay, the sensitivity and specificity were 90 and 70% with a positive predictive value of 60%.
Fig. 3
shows disease-free survival for the patient cohort classified by RT-PCR (A) and QRT-PCR (C) as well as overall survival (B and D). Disease-free survival of patients who were RT-PCR negative was 94% when using either assay. Patients with a positive classification had poorer prognosis by either method. The log-rank test indicates that both QRT-PCR (P = < .0001) and gel-based RT-PCR (P = .0038) are significant predictors of recurrence in node-negative esophageal cancer patients. Overall survival was also worse in QRT-PCR- or RT-PCR-positive patients (P = 0.0006 and 0.106, respectively).
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In the current study, we evaluated the use of TaqMan RT-PCR to quantitatively assay CEA expression and detect occult micrometastases in lymph nodes of histologically N0 esophageal cancer patients. This quantitative analysis enabled us to define a clinically relevant CEA expression level cutoff and overcome the problems associated with background, or ectopic, CEA expression reported by others (17
, 18
, 31)
. Using this approach, 11 of 30 patients were identified as being QRT-PCR positive, and 9 of these patients suffered disease recurrence. Only 1 of the 19 patients who were QRT-PCR negative suffered recurrence during the course of this study. Interestingly, of the 20 patients who did not recur, 11 had detectable CEA expression higher than that seen in control lymph nodes (higher than background levels), and two were above the cutoff level for predicting disease recurrence. In some cases, this could indicate the presence of limited nodal disease that was cured by surgery, because even pN1 patients can have an expected 5-year survival of 
20% (4)
. CEA expression in the remaining samples could possibly be a result of either individual disseminated tumor cells that are unable to survive and are possibly undergoing apoptosis, cell-free RNA in the lymph system as a result of tumor cell apoptosis at the primary site, or contaminating cells introduced inadvertently by the surgeon.
Both disease-free survival and overall survival were significantly higher in the QRT-PCR-negative patients compared with QRT-PCR-positive patients. Furthermore, disease-free survival in the QRT-PCR-positive group was only 27% at 3 years, indicating that micrometastatic disease may be as clinically significant as histological N1 disease. With the exception of one patient, only one positive tissue block was identified per patient, indicating limited disease spread. All positive lymph nodes were locoregional and would therefore confer N1 status, rather than M1a, as defined by celiac or cervical lymph node involvement. The limited nodal involvement in these histologically N0 patients emphasizes the need for adequate lymph node sampling, from different nodal stations, during staging procedures. Although the subset of patients with node-negative disease has been relatively small in the past, the dramatic increase in esophageal cancer has led to surveillance programs that are identifying patients with early stage disease more frequently. Methods for accurate staging of these patients will thus become even more important.
Comparison of our quantitative data with gel-based analysis of the same samples showed that QRT-PCR improves the test specificity while maintaining the same high sensitivity. QRT-PCR also gives objective results, which are amenable to automated, hands-free analysis, with minimal risk of PCR product cross-contamination. Such automation will be essential if QRT-PCR is to become a routine clinical assay. In addition, the quantitative procedure allows for use of rigorous controls to confirm the accuracy and reliability of the assay from run to run. For example, in our experiments, a calibrator sample was run on all PCR plates to correct for day to day variability. Using this methodology, reproducibility tests indicate that the 95% confidence limits on our measurements are ±0.511 cycles (data not shown). The ability to accurately assess the reproducibility of the quantitative assay will be essential if RT-PCR is ever to be used in a clinical setting. In a gel-based assay, this level of assay verification is not attainable.
We acknowledge that the predictive ability of classifying patients by their CEA levels determined by QRT-PCR will be optimistic because the cutoff value was evaluated in the same patients in whom it was determined. It is likely that if the same procedure for the most accurate cutoff were applied to a second set of patients, the classification would be less successful. To this end, we conducted a reanalysis of the sensitivity, specificity, and classification accuracy by cross-validation (n = 10000). The cross-validation estimates of specificity were 0.82 and accuracy of 0.82 (compared with 0.90 and 0.90, respectively, in the original sample). We conclude that classification success of the original sample was slightly overstated, but even after correction for this optimism, classification by QRT-PCR remains an improvement over the gel-based assay.
In conclusion, we have demonstrated that QRT-PCR can detect, with high specificity, micrometastatic disease in histologically negative lymph nodes of esophagus cancer patients. We have also shown that the presence of micrometastatic disease is a strong, independent predictor of cancer recurrence and that quantitation is superior to standard RT-PCR assays. Quantitative RT-PCR should be able to identify which patients with early stage esophageal cancer are at high risk for recurrence and who might benefit from additional therapy. Finally, this quantitative approach should result in similar benefits in other tumor types.
|
FOOTNOTES |
|---|
1 This work was supported in part by the George Heckman Memorial Institutional Research Grant (60-002-40-IRG) from the American Cancer Society. 
2 To whom requests for reprints should be addressed, at Division of Thoracic Surgery, Suite C-800, Presbyterian University Hospital, 200 Lothrop Street, Pittsburgh, PA 15213. Phone: (412) 692-4652; Fax: (412) 692-4650; E-mail: Godfreyte{at}msx.upmc.edu
3 The abbreviations used are: IHC, immunohistochemistry; RT-PCR, reverse transcription-PCR; QRT-PCR, quantitative RT-PCR; CEA, carcinoembryonic antigen; ß-Gus, ß-glucuronidase; ROC, receiver-operator characteristic; TNM, tumor-node-metastasis.
Received 4/ 9/01; revised 9/18/01; accepted 9/19/01.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. A. Klein and N. H. Stoecklein Lessons from an Aggressive Cancer: Evolutionary Dynamics in Esophageal Carcinoma Cancer Res., July 1, 2009; 69(13): 5285 - 5288. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.-H. Yuan, X.-F. Liang, W.-H. Jia, J.-L. Huang, M. Wei, L. Deng, L.-Z. Liang, X.-Y. Wang, and Y.-X. Zeng Molecular Diagnosis of Sentinel Lymph Node Metastases in Cervical Cancer Using Squamous Cell Carcinoma Antigen Clin. Cancer Res., September 1, 2008; 14(17): 5571 - 5578. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. G. Nicastri, J. T. Doucette, T. E. Godfrey, and S. J. Hughes Is Occult Lymph Node Disease in Colorectal Cancer Patients Clinically Significant?: A Review of the Relevant Literature J. Mol. Diagn., November 1, 2007; 9(5): 563 - 571. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Xi, D. G. Nicastri, T. El-Hefnawy, S. J. Hughes, J. D. Luketich, and T. E. Godfrey Optimal Markers for Real-Time Quantitative Reverse Transcription PCR Detection of Circulating Tumor Cells from Melanoma, Breast, Colon, Esophageal, Head and Neck, and Lung Cancers Clin. Chem., July 1, 2007; 53(7): 1206 - 1215. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Liu, M. Jiang, J. Zhao, and H. Ju Circulating Tumor Cells in Perioperative Esophageal Cancer Patients: Quantitative Assay System and Potential Clinical Utility Clin. Cancer Res., May 15, 2007; 13(10): 2992 - 2997. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Lagarde, F. J.W. ten Kate, J. B. Reitsma, O. R.C. Busch, and J. J. B. van Lanschot Prognostic Factors in Adenocarcinoma of the Esophagus or Gastroesophageal Junction J. Clin. Oncol., September 10, 2006; 24(26): 4347 - 4355. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Xi, W. Gooding, K. McCarty, T. E. Godfrey, and S. J. Hughes Identification of mRNA markers for molecular staging of lymph nodes in colorectal cancer. Clin. Chem., March 1, 2006; 52(3): 520 - 523. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S L Kong, M Salto-Tellez, A P K Leong, Y H Chan, and E S C Koay Discordant quantitative detection of putative biomarkers in nodal micrometastases of colorectal cancer: biological and clinical implications J. Clin. Pathol., August 1, 2005; 58(8): 839 - 844. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Raja, J. Ching, L. Xi, S. J. Hughes, R. Chang, W. Wong, W. McMillan, W. E. Gooding, K. S. McCarty Jr, M. Chestney, et al. Technology for Automated, Rapid, and Quantitative PCR or Reverse Transcription-PCR Clinical Testing Clin. Chem., May 1, 2005; 51(5): 882 - 890. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. L. Ferris, L. Xi, S. Raja, J. L. Hunt, J. Wang, W. E. Gooding, L. Kelly, J. Ching, J. D. Luketich, and T. E. Godfrey Molecular Staging of Cervical Lymph Nodes in Squamous Cell Carcinoma of the Head and Neck Cancer Res., March 15, 2005; 65(6): 2147 - 2156. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Xi, J. D. Luketich, S. Raja, W. E. Gooding, V. R. Litle, M. C. Coello, S. D. Finkelstein, M. L. Chestney, R. J. Landreneau, S. J. Hughes, et al. Molecular Staging of Lymph Nodes from Patients with Esophageal Adenocarcinoma Clin. Cancer Res., February 1, 2005; 11(3): 1099 - 1109. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. T. Becker, C. G. Shores, K. K. Yu, and W. G. Yarbrough Molecular Assay to Detect Metastatic Head and Neck Squamous Cell Carcinoma Arch Otolaryngol Head Neck Surg, January 1, 2004; 130(1): 21 - 27. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Nagatani, Y. Shimada, Z. Li, J. Kaganoi, T. Kan, M. Maeda, G. Watanabe, and M. Imamura Validation of Intra-operative Detection of Paratracheal Lymph Node Metastasis Using Real-time RT-PCR Targeting Esophageal Squamous Cell Carcinoma Jpn. J. Clin. Oncol., November 1, 2003; 33(11): 549 - 555. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X. Jiao, A. Eslami, O. Ioffe, K. F. Kwong, M. Henry, Q. Zeng, Y. Refaely, W. Burrows, Z. Gamliel, and M. J. Krasna Immunohistochemistry analysis of micrometastasis in pretreatment lymph nodes from patients with esophageal cancer Ann. Thorac. Surg., October 1, 2003; 76(4): 996 - 1000. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. F. Shariat, M. P. Roudier, G. E. Wilcox, M. W. Kattan, P. T. Scardino, R. L. Vessella, S. Erdamar, C. Nguyen, T. M. Wheeler, and K. M. Slawin Comparison of Immunohistochemistry with Reverse Transcription-PCR for the Detection of Micrometastatic Prostate Cancer in Lymph Nodes Cancer Res., August 1, 2003; 63(15): 4662 - 4670. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. N. Myers and B. R. Gastman Neck Dissection: An Operation in Evolution: Hayes Martin Lecture Arch Otolaryngol Head Neck Surg, January 1, 2003; 129(1): 14 - 25. [Full Text] [PDF]
|
|
|
|
|
|
S. Raja, T. El-Hefnawy, L. A. Kelly, M. L. Chestney, J. D. Luketich, and T. E. Godfrey Temperature-controlled Primer Limit for Multiplexing of Rapid, Quantitative Reverse Transcription-PCR Assays: Application to Intraoperative Cancer Diagnostics Clin. Chem., August 1, 2002; 48(8): 1329 - 1337. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
