| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 2003;125:1091-1102
© 2003 The American Association for Thoracic Surgery
General Thoracic Surgery |
From the Center for Swallowing and Esophageal Disorders, Departments of Thoracic and Cardiovascular Surgery,a Biostatistics and Epidemiology,b Hematology and Medical Oncology,c Gastroenterology,d and Anatomic Pathology,e The Cleveland Clinic Foundation, Cleveland, Ohio.
Read at the Eighty-first Annual Meeting of The American Association for Thoracic Surgery, San Diego, Calif, May 6-9, 2001.
Received for publication April 12, 2001. Revisions requested Aug 7, 2001; revisions received May 30, 2002. Accepted for publication June 24, 2002. Address for reprints: Thomas W. Rice, MD, The Cleveland Clinic Foundation, 9500 Euclid Ave/Desk F25, Cleveland, OH 44195 (E-mail: ricet{at}ccf.org).
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and methodsResultsDiscussionAppendix 1Appendix: DiscussionReferences |
|---|
T2) or invading beyond (>T2) the esophageal wall. In cN0, increasing cT was predictive of progressively poorer survival. For each category of pT N0, cT accurately predicted survival, except for pT3, which was underestimated (P < .0001). In cN0, downstaging by induction therapy was beneficial only if tumors invaded beyond the wall (
cT3, P = .0003). In cN1, it was beneficial only when downstaging was synchronous in cT3/T4 (P < .001). |
Introduction |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAppendix 1Appendix: DiscussionReferences |
|---|
The categorization of TNM staging of esophageal carcinoma reflects disease progression: increasing depth of invasion by the primary tumor (T) to regional lymph node metastases (N) to distant metastases (M). However, in treatment decisions the hierarchy of staging criteria is reversed. Discovery of distant metastases (M1) has dismal prognosis and generally renders patients inoperable
1; discovery of regional lymph node metastases (N1), with its grave prognosis, makes esophagectomy alone ineffective, leading to a search for successful multimodality therapies. 2
When esophagectomy was the only therapeutic option, accurate clinical determination of the depth of tumor invasion (cT) was not necessary because the pathologic depth of tumor invasion (pT) provided all the information available for prognostication. Today, the complete spectrum of esophageal carcinoma, from high-grade dysplagia to distant metastatic disease, is seen. Also, many therapeutic options are available. Thus, in patients without distant or regional lymph node metastases, treatment decisions must be made on the basis of cT.
Therefore, the purposes of this study were to evaluate the effectiveness of clinical staging of depth of tumor invasion (cT), the relationship of cT to survival, the benefits of downstaging cT, and the role of cT in treatment decisions.
|
Patients and methods |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAppendix 1Appendix: DiscussionReferences |
|---|
3
|
3 EUS definition of T focused on the fourth ultrasonographic layer to define cT1 to cT3 (Figure 1). cT4 was defined by EUS evidence for invasion of adjacent structures. In these patients EUS and computed tomographic scanning staged all but 7 patients as having cM0 disease.
|
|
). The pT categories were p high-grade dysplasia if malignant cells were confined to the epithelium and did not breach the basement membrane, pT1 if the tumor did not invade beyond the submucosa, pT2 if the tumor invaded the muscularis propria but not beyond, pT3 if the tumor invaded the periesophageal tissue, and pT4 if the tumor invaded adjacent structures (Figure 1). 3 If no residual tumor was found after induction therapy, the patient's status was pT0. At resection, all accessible distant sites were evaluated. Twenty-five patients had pM1a (7%) tumors, and 10 had pM1b (3%) tumors.
Treatment
Of the 337 patients, 209 underwent esophagectomy alone (Table 1). Because this group of patients underwent esophagectomy immediately after clinical staging with no intervening therapies, this provided information about the effectiveness of clinical staging of T in terms of referent values (accuracy, sensitivity, specificity, positive predictive value, and negative predictive value). It also provided survival data after esophagectomy alone in relation to cT.
The remaining 128 patients received induction therapy after clinical staging and before esophagectomy (Table 1). This group provided no information about referent values; rather, it provided information relevant to the benefit and role of downstaging T.
Esophagectomy alone
Of the 209 patients undergoing esophagectomy alone, 68% had resection with 2-field lymphadenectomy by means of thoracotomy. Thirty-one percent had transhiatal esophagectomy with lymph node sampling (Table 1). R0 resections were achieved in 92%.
Induction therapy and esophagectomy
Of the 128 patients receiving induction therapy, 16 patients before 1991 received preoperative chemotherapy only; this included 2 planned courses of cisplatin, etoposide, and doxorubicin. 4 Since 1991, patients were offered induction chemoradiotherapy for cN1 and cT3 or cT4 cancers. However, treating physicians had the latitude to treat patients with cT1 N0 (n = 3) and cT2 N0 (n = 8) cancers with induction therapy. Induction therapy included 2 courses of chemotherapy, either cisplatin and 5-fluorouracil 5 or cisplatin and taxol. 6 Concurrent with chemotherapy, accelerated fractionation radiation therapy (1.5 Gy twice daily to a total dose of 45 Gy) was administered in a split fashion over 4
weeks. Esophagectomy was performed 4 to 6 weeks later. Ninety-seven percent of patients had esophagectomy with 2-field lymphadenectomy via thoracotomy, and 3% had transhiatal esophagectomy with lymph node sampling (Table 1). R0 resections were achieved in 95% of patients.
These treatment protocols had approval and were reviewed annually by The Cleveland Clinic Foundation's Institutional Review Board. Written informed consent was obtained from all patients before treatment began.
Study end point and patient follow-up
The end point of the study was all-cause mortality. Patients were followed by means of periodic clinic visits and cross-sectional systematic follow-up in February 2001. At that time, 2 patients previously followed for 2.8 and 7.8 years, respectively, could not be located. A search of the Social Security Death Index did not find the Social Security numbers deactivated; it was therefore assumed that the patients were alive 6 months before February 2001 (September 1, 2000). Mean follow-up was 26 ± 15 months (43 ± 35 months in patients alive at last follow-up).
Data analysis
Descriptive
Descriptive statistics are summarized as means and SDs for continuous variables and as frequencies and percentages for categorical variables.
Effectiveness of clinical staging of T
cT was compared with pT in the group undergoing esophagectomy alone (n = 209). For this comparison, clinical and pathologic T stages were grouped as follows: (1) T2 or less, T2, and greater than T2 and (2) T2 or less and greater than T2.
Conventional diagnostic characterization is in terms of the dichotomy of "positivity" and "negativity." Because cT is a graded variable, positivity and negativity had to be defined for each tumor depth. In this comparison positivity meant patients had the pT stage being considered, and negativity meant patients did not have the pT stage being considered (either lower or higher pT). True positivity meant patients with the pT stage being considered were correctly clinically staged. False positivity meant patients without the pT stage being considered were incorrectly clinically staged. True negativity meant patients without the pT stage being considered were correctly clinically staged. False negativity meant patients with the pT stage being considered were incorrectly clinically staged. Thus, sensitivity was defined as true-positives divided by the sum of true-positives and false-negatives, specificity as true-negatives divided by the sum of true-negatives and false-positives, predictive value of a positive test as true-positives divided by the sum of true-positives and false-positives, predictive value of a negative test as true-negatives divided by the sum of true-negatives and false-negatives, and accuracy of staging T as the sum of true-positives and true-negatives divided by the total number of patients undergoing esophagectomy alone.
Relationship of cT to survival
Prognostication by clinical staging was evaluated by survival analysis of the group undergoing esophagectomy alone (n = 209). Nonparametric estimates of survival were obtained by the Kaplan-Meier method. A parametric method was used to resolve the number of phases of instantaneous risk of death (hazard function) and to estimate their shaping parameters. 7 Estimates of survival were considered reliable to 10 years. Multivariable analysis of survival in this group was necessary to demonstrate the prognostic value of clinical staging compared with pathologic staging and to provide a risk-adjusted comparison with induction therapy.
For multivariable analyses, only variables known before esophagectomy and the planned surgical approach were examined. Variables included demography (Table 1), depth of tumor invasion by EUS (cT), preoperative assessment of mediastinal nodes (cN), planned surgical approach (Table 1), and date of operation. Depth of tumor invasion was analyzed both as an ordinal variable and as a set of dichotomous variables for each cT level; these produced equivalent models. Two-way interactions of all variables with cN and surgical approach and 3-way interactions with cN, surgical approach, and the other variables were examined.
Continuous and ordinal variables were assessed univariably by decile risk analysis to suggest transformations of scale required to ensure that relationships of these measured variables were well calibrated to model outcome.
Multivariable analysis was performed simultaneously for 2 hazard phases identified from the data. Both a guided technique of entry of variables into the multivariable models 8 and bootstrap bagging with 1000 resamplings were used. 9 A P value of .1 for retention of variables in final models was used.
Patients with pN0 M0 tumors undergoing esophagectomy alone were studied to evaluate the relationship of cT to pT in terms of prognostication. For each patient, a survival curve was predicted on the basis of cTN from the multivariable equation. For each category of pT, these survival curves based on cT were averaged and compared with the actual Kaplan-Meier estimates. 10 To test for differences between expected and observed survival, cumulative hazard at the time of each patient's follow-up (or death) was calculated from this multivariable equation. The sum of the cumulative hazard values was the number of expected deaths on the basis of cT. A 
2 goodness-of-fit test was used to compare the number of expected and observed deaths. 10
Benefit of downstaging cT
The benefits of downstaging cT by induction therapy were derived from survival analysis. Methods of survival analysis within the induction therapy group (n = 128), including multivariable analysis, mirrored those for the group undergoing esophagectomy alone (see above).
To evaluate the benefit of downstaging cT, we compared actual survival of these patients with predicted survival had they undergone esophagectomy alone. For each patient, a survival curve was predicted on the basis of cTN from the multivariable equation of esophagectomy alone. For selected categories of cT and pT, these survival curves were averaged and compared with the actual Kaplan-Meier estimates and tested for differences between expected and observed survival as described above. 10
cT and treatment decisions
The benefit of staging T and the use of induction therapy to downstage T were integrated with the results of our study on staging and downstaging N1 tumors. 2 A multivariable analysis of survival in patients undergoing induction therapy was performed to assess the independent benefit of downstaging T. Unlike the analyses above, this analysis included pathologic stage and number and percentage of positive nodes (Tables 1 and 2). In addition, variables were formed for the differences between cN and pN and between cT and pT, their interactions with cN and pN, and interactions with surgical approach.
|
Results |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAppendix 1Appendix: DiscussionReferences |
|---|
T2) from those invading beyond (>T2), clinical staging by means of EUS was 87% accurate, 82% sensitive, 91% specific, 89% positively predictive, and 86% negatively predictive.
|
|
|
) based on cT accurately predicted survival (P > .15, Figure 3), except for pT3, which was underestimated (P < .0001). In this group 20.6 deaths were predicted, but only 13 were observed. The excess deaths were predicted in the subgroup of patients with pT3 N0 M0 tumors who were understaged (ie, patients in whom N1 nodes had been predicted by EUS [cN1] but who had no regional lymph node metastases at operation).
|
|
|
|
pT3) had poor survival despite induction therapy and esophagectomy. If there is synchronous downstaging of cT and cN, there is a homogenization of survival (Figure 7).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAppendix 1Appendix: DiscussionReferences |
|---|
When palliation with single-modality therapy was the aim of treatment, there was no need for clinical staging. Clinical staging has become important since the introduction and increasing use of multimodality therapy. The effect of treatment decisions in neoadjuvant therapy, made on the basis of clinical staging, is difficult to assess because therapy might downstage tumors or there might be disease progression during therapy. In addition, the ability to clinically stage the tumor after neoadjuvant therapy might be lost. 11 Therefore, it is crucial to have effective and accurate clinical staging and a clear relationship of staging to prognosis to assess the benefit of current treatment and to direct future treatment decisions.
Effectiveness of clinical staging of T
There is a misconception that accuracy of cT determination by EUS increases incrementally with deeper tumor penetration. 12 In a meta-analysis the reported accuracy was 84% for T1, 73% for T2, 89% for T3, and 89% for T4 tumors. 13 A review of the literature shows variation in accuracy with T: 75% to 82% for T1, 64% to 85% for T2, 89% to 94% for T3, and 88% to 100% for T4 tumors. 14 The least reliable of EUS cT determinations is cT2. 15 Except for T2 tumors, EUS accuracy increases with increasing T.
In this study the determination of T by EUS was accurate, specific, and negatively predictive. For tumors invading beyond the esophageal wall (>T2), EUS was also sensitive and positively predictive. However, EUS sensitivity in screening for less than T2 and T2 esophageal tumors was poor because of high false negativity. The positive predictive value of EUS determination of T2 was also poor because of high false positivity.
The interplay of staging criteria and EUS anatomy partially accounts for the problem with T2. The muscularis propria is critical in differentiating T1, T2, and T3 tumors. Tumors are defined as T1 if there is no invasion of the muscularis propria, T2 if invasion is into the muscularis propria, and T3 if invasion is beyond the muscularis propria. For clinical evaluation, the fourth ultrasonographic layer is interpreted as the muscularis propria. However, studies have demonstrated that the third ultrasonographic layer is the submucosa and the acoustic interface between the submucosa and muscularis propria. The fourth ultrasonographic layer is the muscularis propria minus the acoustic interface between the submucosa and the muscularis propria. 16,17 Therefore, the boundary necessary to completely differentiate T2 from T3 tumors is contained in the third ultrasonographic layer. Because 2 boundaries must be evaluated in the determination of T2 and errors might occur at each, the inaccuracy is potentially double that of T1 and T3 tumors.
However, for the clinically pertinent distinction of tumors confined to the esophageal wall (T2) from those invading beyond (>T2) the esophageal wall, EUS is highly accurate, specific, sensitive, positively predictive, and negatively predictive.
Relationship of cT to survival
Clinical staging of T is prognostic. Clinical staging of tumors confined to the esophageal wall (<cT2) and pathologic staging of these same tumors are similarly predictive of prognosis. Clinical staging of tumors invading beyond the esophageal wall (>cT2) underestimates survival compared with pathologic staging. The interplay of T and N in these locally advanced tumors accounts for this underestimation of survival. The majority of patients (80%) with tumors invading beyond the esophageal wall (>T2) will have regional lymph node metastases. 18 However, in the minority without regional lymph node metastases (pN0), clinical overstaging of regional lymph node status (cN1), not inaccuracy in predicting depth of tumor invasion (>cT2) accounts for the underestimation of survival.
Benefit of downstaging cT
For patients with cN0 tumors, survival was enhanced by induction therapy only for those with tumors invading beyond the wall (>cT2) who were downstaged. Induction therapy might reduce survival for patients with cT2 N0 or less tumors. For patients with cN1 tumors downstaged to pN0 by induction therapy, survival was enhanced compared with that in patients not receiving induction therapy or those not downstaged by induction therapy. 2 For tumorss invading beyond the wall (>cT2), if there was downstaging of N but not T or vice versa, the survival advantage of induction therapy was lost.
cT and treatment decisions
Treatment decisions should be based on cT only in patients with cN0 M0 tumors. Patients with esophageal carcinomas that have not metastasized to regional lymph nodes (cN0) or breached the esophageal wall (cT2) have good survival when treated with esophagectomy alone. Every effort should be made to stage patients clinically and to identify those with esophageal carcinomas confined to the esophageal wall (cT2 N0 M0) because they are the best surgical candidates and might be harmed by aggressive multimodality therapy. Once tumors have spread beyond the esophageal wall (>cT2) or metastasized to regional (cN1) or nonregional (cM1a) lymph nodes, survival is drastically reduced. Only patients with advanced tumors whose disease is downstaged to pT2 N0 or less with induction therapy will benefit from this therapy.
Limitations
This study is a small, nonrandomized, single-institution, prospective experience. There might have been inaccuracies of clinical staging and treatment biases over time. Analysis was made difficult by selection of induction therapy for individual patients. Therefore, inferences from application of information on surgical intervention alone to the induction therapy group are not as robust as they would be had this been a randomized trial. In addition, patients in whom induction therapy was used almost universally received a thoracotomy, and this factor in itself adds risk to the treatment program, 19 only one component of which is induction therapy.
|
Appendix 1 |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAppendix 1Appendix: DiscussionReferences |
|---|
Method of assessing linkage
High-grade dysplasia and cT1 groups were coalesced to facilitate the analyses. Clinical stage was then numerically scored: 1 = cT1 or less; 2 = cT2; 3 = cT3; and 4 = cT4. A similar score was assigned to pT, except that absence of disease (pT0) was scored as zero. Change in T was obtained by subtracting the cT score from the pT score. Therefore, a negative difference indicated pT was less than cT (downstaging or overstaging), and a positive difference indicated pT was greater than cT (disease progression or understaging).
Change in T was analyzed both univariably and multivariably with respect to sex, race, cT, pT, cN, pN, pM, cell type, tumor differentiation, and surgical margin. Either the Wilcoxon rank sum test (dichotomous variables) or the Kruskal-Wallis test (variables with >2 categories) was used to assess the association of these variables to change in T.
For multivariable analysis, change in T was analyzed as an ordinal variable with 5 levels (-3, -2, -1, 0, and 1). 20 Particular attention was paid to 4 possible combinations of N: cN0/pN0, cN1/pN0, cN1/pN1, and cN0/pN1. The assumption of proportional odds was reasonable, and therefore multivariable ordinal logistic regression was used, which produces a different intercept for each level of change of T. Except for combinations of N, variables were retained in the analysis whose P value was greater than .1.
Linkage of T and N
cT was greater than pT in 46% of patients (downstaged or overstaged), cT was equal to pT in 47%, and cT was less than pT in 7% (disease progression or understaged). The higher cT, the lower pT was after induction therapy (pT-cT more negative, Appendix Table 1).
The higher pT, the more likely the tumor had remained unchanged or progressed (pT-cT more positive). The difference between cT and pT was similar irrespective of whether there was cN1 (P = 1). However, if the patient had pN0 tumors after induction therapy, the magnitude of T decrease was greater (-1.4 for pN0 vs -0.1 for pN1). The magnitude of T decrease was greatest in patients with pN0 tumors who had been downstaged from cN1 (ie, downstaging of T and N were synchronous); patients who had or who eventually developed pN1 tumors had, on average, no change in T. Although the numbers are small, it is suggestive (P = .06) that if distant metastatic disease were present (M1), induction therapy did not affect T. Patients with squamous cell cancers were more likely than those with adenocarcinomas to have T decrease with induction therapy. Multivariable analysis (Appendix Table 2) confirms these observations from univariable analysis and indicates the effect of these variables on the magnitude of change in T.
Thus, the greater the depth of tumor invasion (cT), the greater the downstaging of T. If pN1 tumors were found, it was highly likely that there was no T downstaging. With less certainty, the model predicted that T was downstaged in patients with squamous cell cancer more than in patients with adenocarcinoma. The analysis also suggested that linkage among the above factors accounted for the downstaging of cN1 to pN0 (P = .5 for any additional information).
|
|
|
Appendix: Discussion |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAppendix 1Appendix: DiscussionReferences |
|---|
To a mere mortal physician, the statistics are, to say the least, daunting in the article, and I will trust Dr Rice's and Dr Blackstone's analysis of the variables. However, I do have several questions regarding the conclusions that were drawn from this analysis.
In the article a final conclusion is that patients with clinically staged T1 or T2 N0 esophageal carcinoma might not benefit from induction therapy and, in fact, might be harmed. However, as you noted, the accuracy of staging T2 in the article was only 81%, with a positive predictive value of only 23%. The analysis of all these patients was also based on clinical staging rather than pathologic staging of the lymph nodes, a method that might additionally misstage 20% of the patients. Thus, do you think that this potential 20% error rate in accurately defining malignant lymph nodes and the inaccuracy of defining true T2 carcinomas weaken your conclusions? Could your conclusions and analysis be enhanced if the data were based on true pathologic staging, as might be gained first by fine-needle EUS analysis and then selective use of thoracoscopy or laparoscopy patients within clinically staged N0 disease? Can such a conclusion really be drawn from a cohort of only 11 patients?
Another interesting correlation that you have identified was the poor prognosis of those patients who continued to have N1 carcinoma after induction therapy and those patients who continued to have T3 carcinoma, N1 carcinoma, or both, either by progression of disease or lack of response. The poor prognosis of patients with residual malignant lymph nodes after induction therapy has also been seen with non-small cell lung cancer after induction therapy. Given these results, would you consider deferring surgical resection in those patients undergoing induction therapy in whom you can prove either a lack of T downstaging or persistence of malignant lymph nodes?
Finally, clearly T and N staging are related. As in most solid tumors, as you have shown, an increasing T stage will result in a higher percentage of patients with positive lymph nodes. However, despite this linkage, the correlation is not linear, and significant variability in the biologic activity and characteristics of tumors will result in earlier and/or later lymphatic or systemic failure, or both. Ultimately, molecular and biologic differences in tumor disease must play a major role in the overall prognosis and TNM stage of any tumor. Do you envision a staging system in the future that includes a biologic modification for molecular and biologic parameters?
Dr Rice. Concerning why we did not use true pathologic stage, we have to make decisions based on clinical stage. There is no doubt that there is an improvement in stratification if pT is used.
Fine-needle aspiration is now the standard of care. If there is an EUS-identified large node, the patient should undergo fine-needle aspiration. However, this paper discusses depth of tumor invasion, and in the analysis EUS is very good at identifying invasion beyond the esophageal wall. Whenever we use EUS, we are looking for T3 and N1. EUS and EUS/fine-needle aspiration are excellent at this identification.
As to the application of thoracoscopy, I think thoracoscopy and laparoscopy should be reserved for those patients in whom clinical stage is not obtainable by EUS or is in doubt. That is very few patients. Therefore, I would suggest that we stage all our patients with EUS and EUS/fine-needle aspiration and reserve invasive staging for those patients who have particular clinical problems.
The question about whether we know that our patients have not responded to induction therapy (ie, they have persistent N1 or T3 disease) is interesting. I think it is something we now have to address: What are we going to do with our patients who have induction therapy and persistent disease? Indeed, those patients probably should not proceed to resection.
I think that the staging system is a great advance that has allowed us to understand this disease, but it is imperfect. As we know, staging is an ongoing process. It is constantly evolving, and there is a definite possibility that molecular substaging will be added to it.
|
Acknowledgments |
|---|
|
References |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAppendix 1Appendix: DiscussionReferences |
|---|
Related Article
This article has been cited by other articles:
|
|
 |
|
  C. Rizzetto, S.R. DeMeester, J.A. Hagen, C.G. Peyre, J.C. Lipham, and T.R. DeMeester En bloc esophagectomy reduces local recurrence and improves survival compared with transhiatal resection after neoadjuvant therapy for esophageal adenocarcinoma. J. Thorac. Cardiovasc. Surg., June 1, 2008; 135(6): 1228 - 1236. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. F. Bruzzi, R. F. Munden, M. T. Truong, E. M. Marom, B. S. Sabloff, G. W. Gladish, R. B. Iyer, T.-S. Pan, H. A. Macapinlac, and J. J. Erasmus PET/CT of Esophageal Cancer: Its Role in Clinical Management RadioGraphics, November 1, 2007; 27(6): 1635 - 1652. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Murakami, M. Kenjo, T. Uno, M. Oguchi, M. Shimada, T. Teshima, and the Japanese Patterns of Care Study Working Subgro Results of the 1999 2001 Japanese Patterns of Care Study for Patients Receiving Definitive Radiation Therapy without Surgery for Esophageal Cancer Jpn. J. Clin. Oncol., July 23, 2007; (2007) hym055v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. W. Rice, D. P. Mason, S. C. Murthy, G. Zuccaro Jr, D. J. Adelstein, L. A. Rybicki, and E. H. Blackstone T2N0M0 esophageal cancer J. Thorac. Cardiovasc. Surg., February 1, 2007; 133(2): 317 - 324. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. W. Gee and D. W. Rattner Management of Gastroesophageal Tumors Oncologist, February 1, 2007; 12(2): 175 - 185. [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]
|
|
|
|
 |
|
 K. L. Wang, T.-T. Wu, E. Resetkova, H. Wang, A. M. Correa, W. L. Hofstetter, S. G. Swisher, J. A. Ajani, A. Rashid, S. R. Hamilton, et al. Expression of Annexin A1 in Esophageal and Esophagogastric Junction Adenocarcinomas: Association with Poor Outcome Clin. Cancer Res., August 1, 2006; 12(15): 4598 - 4604. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. J. Lightdale and K. G. Kulkarni Role of Endoscopic Ultrasonography in the Staging and Follow-Up of Esophageal Cancer J. Clin. Oncol., July 10, 2005; 23(20): 4483 - 4489. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. C. Malaisrie, B. Untch, G. V. Aranha, N. Mohideen, A. Hantel, and J. Pickleman Neoadjuvant Chemoradiotherapy for Locally Advanced Esophageal Cancer: Experience at a Single Institution Arch Surg, May 1, 2004; 139(5): 532 - 539. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. A. Hagen and T. R. DeMeester Staging of esophageal carcinoma J. Thorac. Cardiovasc. Surg., May 1, 2003; 125(5): 988 - 991. [Full Text] [PDF]
|
|
|
|
|
|
T. W. Rice, E. H. Blackstone, L. A. Rybicki, D. J. Adelstein, S. C. Murthy, M. M. DeCamp, and J. R. Goldblum Refining esophageal cancer staging J. Thorac. Cardiovasc. Surg., May 1, 2003; 125(5): 1103 - 1113. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||
