| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Division of Thoracic Surgery, Department of Cardiothoracic Surgery, New York Presbyterian Hospital, Weill Cornell Medical College, New York, New York
Accepted for publication April 1, 2011.
* Address correspondence to Dr Stiles, Division of Thoracic Surgery, Department of Cardiothoracic Surgery, Ste M404, Weill Medical College of Cornell University, 525 E 68th St, New York, NY 10021 (Email: brs9035{at}med.cornell.edu).
Presented at the Forty-seventh Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 31–Feb 2, 2011.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Methods: Surgically treated patients with cT2-T3N0M0 esophageal cancer were reviewed. Adequacy of lymph node dissection was assessed by guidelines applied to clinical stage. Survival was determined by Kaplan-Meier analysis. Univariate and multivariate analyses were done for predictors of N+ and survival.
Results: We identified 102 patients, 51 cT2N0 and 51 cT3N0, 39 (38%) of whom had induction therapy. Despite being clinically node negative, 61 patients (60%) had nodal metastases. Applied to cT classification, adequate nodal dissection was achieved in 64 patients (63%). Transthoracic esophagectomy was more likely than transhiatal esophagectomy to achieve adequate nodal dissection (69% versus 31%, p = 0.005). Adequate nodal dissection was more likely to document pN+ disease in both the surgery alone group (70% versus 50%, p = 0.13) and induction therapy group (71% versus 33%, p = 0.02). Five-year overall survival was 44% with surgery alone and 55% with induction therapy. On multivariate analysis, pN+ was the strongest predictor of overall survival (relative risk 2.73, confidence interval: 1.29 to 5.78).
Conclusions: Most cT2-T3N0M0 patients have pN+ disease. Despite induction therapy, more than 50% have persistent nodal disease. Transthoracic esophagectomy is more likely to detect pN+ disease and more likely to meet criteria of adequate nodal dissection than is transhiatal esophagectomy. Therefore, the majority of patients with cT2-T3N0M0 should be considered for neoadjuvant protocols and should be treated by transthoracic resection whenever possible.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Preoperative Therapy and Surgical Treatment
Patients receiving neoadjuvant therapy were treated by preoperative chemotherapy alone. Preoperative therapy consisted of two to four cycles of platinum with either a taxane or epirubicin and capecitibine. All patients subsequently underwent surgical resection, the majority by a transthoracic approach. In the chest, all lymph nodes in the middle and lower mediastinum, including the periesophageal, parahiatal, subcarinal, and aortopulmonary nodes were removed en bloc with the esophagus [3]. For patients undergoing three-field lymph node dissections, the right and left recurrent laryngeal nodes were also dissected. A few patients were resected using a transhiatal approach. Regardless of surgical approach, an upper abdominal and retroperitoneal node dissection was performed for all patients.
Adequacy of Lymph Node Dissection
Adequacy of lymph node dissection (LND) was extrapolated from the recommendations for optimal lymph node dissection proposed by the World Wide Esophageal Cancer Collaboration (WECC) [4]. Using Forest plot analysis, the WECC investigators recommended removing a minimum of 20 lymph nodes for all pT2 cancers and 30 lymph nodes for all pT3 cancers. Seeking a more clinically relevant application of these WECC criteria, we examined whether the nodal dissection guidelines would be useful in optimizing nodal staging when based on the clinically determined, rather than the pathologically determined, T classification.
Statistical Analysis
Descriptive statistics (including frequency, percent, mean, median, standard deviation, range) are presented for demographic and clinical factors. Variables of interest were examined by the 
2 test or Fisher's exact test for categorical variables, and the two-sample t test or Wilcoxon rank sum test for continuous variables. Overall survival for the cohort (from date of surgery) was evaluated by Kaplan-Meier survival analysis. The log-rank test was employed to compare overall survival between different cohorts of patients. All p values are two-sided with statistical significance evaluated at the 0.05 alpha level. Ninety-five percent confidence intervals (CI) for adjusted odds ratios (from the multivariate model) and for Kaplan-Meier survival probability estimates were calculated to assess the precision of the obtained estimates.
Factors identified at p less than 0.20 by univariate analysis were selected for inclusion in a Cox regression model to assess the independent effect of these variables on overall survival. Potential collinearity between univariate predictors was assessed by the Pearson (or Spearman rank) correlation coefficient (for continuous predictors) or the kappa statistic (for categorical predictors) before the final specification of the multivariate model. If two variables were highly correlated, only one of the variables was evaluated in the multivariate model. All analyses were performed in SAS version 9.2 (SAS Institute, Cary, NC), SPSS version 18.0 (SPSS Inc, Chicago, IL), and STATA version 11.0 (StataCorp, College Station, TX).
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
|
|
|
Adequacy of Lymphadenectomy
The median number of lymph nodes resected was 27 (range, 4 to 82). Transthoracic resection yielded a higher number of resected lymph nodes than transhiatal resection (median 29 versus 18.5, p < 0.001). When the adequacy of LND was examined using the WECC guidelines applied to the cT classification, 64 of 102 patients met the WECC adequacy criteria. Adequate LND was more likely to be achieved in cT2 patients than in cT3 patients (76.5% versus 49.0%, p = 0.004). A transthoracic approach (n = 86) was more likely to achieve adequate LND as compared with a transhiatal approach (n = 16 [68.6% versus 31.3%, p = 0.005). Similarly, an en bloc resection (n = 79) was more likely to yield an adequate number of nodes when compared with nonradical resection (n = 23 [70.9% versus 34.8%, p = 0.003]). Notably, for clinical T3 tumors, a transhiatal approach failed to meet WECC criteria in each of 8 cases, and a nonradical transthoracic resection yielded an adequate number of nodes in only 1 of 4 cases. Adequate (versus inadequate) LND was more likely to document pN+ disease in both the surgery alone group (69.8% versus 50.0%, p = 0.129) and the induction therapy group (71.4% versus 33.3%, p = 0.02) when applied to clinical T classification.
Effect of Preoperative Therapy
Although the incidence of nodal metastases was not significantly different, there was a trend toward a reduction in nodal metastases when patients treated with induction therapy followed by surgery were compared with patients who were treated with surgery alone (53.8% versus 63.5%, p = 0.334). Interestingly, nodal metastases were less frequently present after preoperative chemotherapy although the majority of patients receiving such treatment (71.8%) had cT3 tumors and would have been expected to have higher rates of nodal disease. When the analysis of the effect of neoadjuvant therapy was stratified by clinical T classification, there was no difference in rates of nodal metastases among patients with cT2 lesions treated with or without preoperative therapy (induction 63.6%, versus surgery alone 55.0%, p = 0.609). However, for patients with cT3 lesions, there was a significantly lower incidence of nodal disease among patients treated with induction therapy, as compared with patients treated with surgery alone (50% versus 78.3, p = 0.038).
Survival
Median follow-up for 59 surviving patients was 40.9 months. For the entire cohort, the 5-year estimate of overall survival was 48.0%, with no significant difference between cT2N0 and cT3N0 patients (Fig 1
). There was also no significant difference between patients operated on after neoadjuvant treatment and patients undergoing surgery alone (Fig 2
), although the trend favored induction therapy (5-year overall survival 56.2% versus 44.2%, hazard ratio [HR] 0.78, p = 0.465). Stratification of the comparison by clinical T classification showed comparable survival for both groups of cT2 and cT3 patients treated with and without neoadjuvant treatment. On univariate analysis (Table 4), factors most predictive of survival included grade of differentiation (HR for poor differentiation 1.629, 95% CI: 0.884 to 3.002), pT classification (HR for pT3-T4 2.629, 95% CI: 1.368 to 5.053), and pN classification (HR for N+ 2.914, 95% CI: 1.395 to 6.09). Histologic classification and the use of induction therapy were less significant predictors. On multivariate analysis (Table 5) of factors identified as significant on univariate analysis (p < 0.20), pN+ was the strongest predictor of overall survival (HR 2.73, 95% CI: 1.29 to 5.78). The pT classification was collinear with pathologic N status (
>0.4) and was therefore excluded from the final model. For patients with pN0 disease, 5-year overall survival was 72.1%, compared with 35% for pN+ patients (Fig 3
).
|
|
|
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
In the current study, we sought to characterize the outcomes of cT2N0 and cT3N0 esophageal cancer patients treated with a several different strategies. Although the retrospective nature of our study and the relatively low proportion of our patients with such clinical designations prevent us from defining any absolute treatment algorithms for these patients, several interesting observations were made in this cohort that will likely influence our clinical practice patterns. Most notable was the high rate of nodal metastases found in these patients (59.8%). Although several patients in the series were treated before the adoption of both EUS and PET scans for routine clinical staging, more than 90% had at least one of those studies, and 51.8% of patients had both preoperatively. Patients with both staging procedures performed (in addition to the standard of CT and upper endoscopy) were, nonetheless, likely to be understaged. In this group, more than half (60.3%) had a higher pT classification than suggested by clinical staging or were found to have nodal metastases. That was despite the use of induction therapy for many of these patients (46.0%), known to frequently downstage both T and N classifications [7].
If one considers only patients undergoing surgery alone, more reliable estimates of nodal metastases can be determined. In our cohort, such patients with cT2N0 esophageal cancer (n = 40) were found to have nodal metastases 55% (n = 22) of the time, whereas patients with cT3N0 esophageal cancer (n = 23) were found to have nodal involvement in 78% (n = 18) of cases. This suggests that should upfront surgical resection be applied in patients with cT2-3N0 esophageal cancer, nearly two thirds of patients would be suboptimally treated by current standards of care [8]. It should be noted that clinical staging accuracy reported here does not reflect the true sensitivity and specificity of EUS, as we only evaluated a very narrowly defined group from which we were unable to calculate true and false negatives for T classification and true and false positives by N classification.
Defining the presence or absence of nodal disease is critical to determine prognosis in this cohort of patients. Indeed, in our series, the status of the lymph nodes is the most important determinant of overall survival. Therefore, it is imperative to use a treatment strategy for these patients that optimizes the chances of detecting nodal metastases. We and others have previously sought to define the optimum LND for esophageal cancer patients [4, 9, 10]. Most notably, the WECC defined optimum LND based on pT classification [4]. Here we applied those recommendations to the clinical T classification in an attempt to guide treatment choices. Approximately two thirds of our patients (63%) had an optimum LND. Notably, a transthoracic surgical approach, using an en bloc technique in particular, significantly improved the adequacy of LND compared with a transhiatal approach. Importantly, when applied to cT classification, an optimum LND was more likely to document pN+ disease in both patients undergoing surgery alone and patients undergoing induction therapy followed by surgery. We suggest that a transhiatal resection is generally insufficient in these patients, assuming that they are able to tolerate a transthoracic approach, particularly patients thought to be T3.
Apart from the surgical approach, another frequent dilemma in this patient population with cT2-3N0 esophageal cancer is whether to treat with neoadjuvant therapy before surgical resection. Some have argued against induction therapy for cT2N0 patients [11]. Conversely, patients with cT3N0 esophageal cancer have been included in recent trials involving neoadjuvant therapy for esophageal cancer or for gastroesophageal cancer [12-14]. Although many of these trials have reported a positive result in favor of neoadjuvant therapy, the absolute benefit for N0 patients is difficult to predict. In our study, for the cohort as a whole, there was a trend toward improved overall survival among patients receiving neoadjuvant treatment followed by surgery compared with patients undergoing surgery alone, although this difference was not statistically significant (56.2% versus 44.2%, p = 0.465). This difference becomes more compelling, however, when one considers that 72% of the neoadjuvant group were cT3 compared with only 37% of the surgery alone group. Neoadjuvant therapy did not predict improved survival in either of the cT2 or cT3 subgroups alone; however, there were likely not enough patients in each subgroup to appropriately power the analysis.
We, therefore, believe that a reasonable argument can be made in favor of neoadjuvant therapy before surgical resection for cT2-3N0 patients. High rates of nodal metastases in these patients undergoing resection without induction therapy argue that nodal disease is the rule, rather than the exception. There should be little controversy with cT3 tumors. These patients should be treated with induction therapy owing to their high rate of occult nodal disease. The same rationale can be applied to cT2 tumors, for which the probability of nodal metastases is greater than 50%. Some have expressed concern over the possibility of overstaging and argue that adjuvant therapy can be used effectively in this population if nodal disease is found at operation [11]. The current evidence supporting the use of adjuvant therapy is weak, however, with most of the evidence based upon the MacDonald trial (SWOG 9008), which has its own limitations [15]. Unfortunately, adjuvant therapy often cannot be begun in a timely fashion. After esophagectomy, patients are often quite frail, and despite protestations to the contrary, do not tolerate adjuvant therapy easily. We and others have shown that downstaging of the T or N classification of esophageal cancer with neoadjuvant therapy is a powerful predictor of overall survival [7, 16, 17]. As we demonstrate in the current manuscript, patients who are pN0 have a doubling in survival compared with patients who are N+: 72.1% versus 35% (p = 0.003). It is out policy, therefore, to consider all of these patients for a neoadjuvant therapy protocol, preferably in the context of a clinical trial.
In summary, patients with clinically staged T2-3N0 esophageal cancer are often understaged, despite the performance of preoperative PET and EUS. More half of these patients will be found to have lymph node metastases. Because the presence or absence of nodal disease is the critical determinant of survival in this patient population, it is imperative to perform an adequate LND. Surgical approaches that enhance the extent of LND such as transthoracic esophagectomy and en bloc resection are more likely to meet international standards for the number of lymph nodes resected and are more likely to identify nodal metastases. Owing to the high rate of occult nodal disease in patients with cT2-3N0 esophageal cancer, consideration should be given to enrolling these patients in neoadjuvant treatment protocols.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
DR STILES: Well, let me think about that one. Underdiagnosis in EUS and PET in just the cT2 patients, is that the question?
DR HOFSTETTER: Yes, of those patients who underwent EUS and PET.
DR STILES: I have to remember. I don't know that we broke it down by just clinical T classification for both—we had looked at EUS and PET in the whole cohort of all the surgery alone patients, but I don't think we further broke it down to just look at the cT2s in that group using both modalities. For EUS alone, as I said, T2 classification was only correct about 30% of the time.
DR HOFSTETTER: Right. So as far as our practice goes, the interesting therapeutic decision-making process comes at the early T2s versus the more high-risk T2s. And last year we published a paper evaluating a specific clinical nomogram that could potentially differentiate higher risk T2 patients. Among the clinically significant factors, one with the highest risk factor was length. Specifically, length over 2 cm for a clinical T2 increased the risk of positivity for lymph nodes up to 70% versus 30% for tumors less than 2 cm in length. So I was wondering if you had thought about including length in your univariate and multivariate analysis?
DR STILES: We haven't. That's a great point. We didn't initially have length in our database, but it's something that we've grown more interested in as well.
DR HOFSTETTER: And then finally, over a period of time as you're examining patients, I recognize that it's difficult to get an enlarged sample size to look at this kind of patient. The clinical T2 or T3N0 patient is the rare patient. However, to get an adequate sample size, you were collecting patients that are, I think, 1992 to present, in whom staging criteria are vastly different. And I wonder if you just included patients that are of a modern era whether some of your conclusions would be different. Thanks again for a well-presented paper.
DR STILES: Thanks. I suspect that you're probably right. Dr Altorki had actually mentioned to me that you had talked about trying to get a bunch of institutions together to pursue that question, which I think is a great idea.
DR HOFSTETTER: Thank you.
DR MARK KRASNA (Towson, MD): I also enjoyed the paper. Just a comment and maybe one critique for the manuscript. I think it would be really helpful if you took a subset analysis looking at only those patients who had undergone esophageal ultrasound. I realize it's a span of time that included some patients who didn't have EUS. But both in regard to a question that was asked earlier and as Wayne alluded to now, EUS is the clinical standard test in defining whether the patient is going to be a treated as a certain T stage or not. It is true that the endosonographers have problems differentiating T2s and T3s; so the real cutoff is whether it is a T1 versus a T2, T3. I think in order to really look clearly at the data, it would be very helpful to look just at those patients whose had EUS.
The comment is that your institution has shown us, therefore, that you're now going to take all your T2s, not just your T3s, for radical node dissection, which I think is a great idea because you're going to get better lymph node staging. What I'd like to perhaps ask you to consider, or challenge you to think about, is, if you're now willing to consider T3 patients for neoadjuvant therapy because of the high risk of nodal disease, why not start to consider T2 patients also for neoadjuvant therapy? I enjoyed your paper.
DR STILES: Thanks for your comments, Dr Krasna. We are definitely considering T2 patients for induction therapy. As Dr Hofstetter alluded to, there is a range of T2 patients. With EUS we have a concern about how good it is and where it comes from. We tend to get them from lots of different endosonographers. One important question is do they actually biopsy nodes or do they just look and say that it is node negative? Trying to gather all that information certainly factors into how we treat those patients.
We feel pretty strongly that some T2 patients might benefit from neoadjuvant therapy. There was a paper a couple of years ago from the Cleveland Clinic where they suggested that they didn't do as well, but it was a fairly small group. But again, it's a group of patients with a survival around 50%, so I think we should give them every chance that we can. We certainly think that we can optimize their chemotherapy in the preoperative period, as opposed to postoperatively.
DR TRAVES CRABTREE (St. Louis, MO): Brendon, great presentation, and I'm going to contradict Dr Krasna a little bit. We published our small series, which most centers are, with clinical T2N0 disease, and the ones who were appropriately staged at the time of surgery had about a 90% 5-year survival, while the ones who were subsequently upstaged pathologically, had a 5-year survival of 40%. So I think we should be better at staging the patient rather than just saying we're going to give all clinical T2N0 patients induction therapy. We also looked at some different factors in terms of standardized uptake value uptake on the PET, the differentiation of the tumor, and found some trends that might predict which clinical T2N0 patients are going to be upstaged or appropriately staged.
I know some people have alluded to it, but can you comment how many people got FNAs at the time of their EUS, and if they didn't do FNAs with the EUS, what kind of clinical criteria did they use for diagnosing nodal disease? Thank you, Brendon, and congratulations.
DR STILES: Sure. Those are great points, Traves. I think, unfortunately, our endosonography is so heterogenous and comes from so many different places, we don't have it captured very well in the database. In putting this together, we had tried to go back and review each of the reports just to make sure that these really were clinical T2-3N0s. But in terms of confidently knowing how many got FNAs, what kind of quality, and what kind of consistency there was there, we don't have a good handle on it.
I think that the point that you make is a good one, and again, it just illustrates that maybe we should all get our T2N0s together and take a look. There are lots of smaller case series out there, but what we really need is to pool the results to gain some power.
DR BORIS SEPESI (Rochester, NY): I wanted to ask you a couple of questions. What was the regimen for neoadjuvant chemotherapy in your patients? We use ECX (epirubicin, cisplatin, and xeloda) chemotherapy preoperatively and observed complete response in a few patients. Most institutions, however, use 5-fluorouracil and cisplatin combined with radiation therapy. Various chemotherapy regimens may have different influence on the primary tumor as well as on the lymph nodes.
And my second question is, what is your bias for using chemotherapy versus chemoradiation therapy?
DR STILES: Sure. Most of these patients got cisplatin and 5-FU. Very few got ECX. Everybody always asks the chemotherapy versus the chemoradiation thing. We could debate that many times. We generally do a more aggressive local operation, so that we don't think that radiation plays as much of a role in that setting.
One shouldn't confuse a complete pathologic response from chemoradiation with a systemic response. I've always said when you see Nasser Altorki do a three-field esophagectomy, it's a complete pathologic response because it takes everything out locally. So if they've had chemotherapy with a good clinical response and everything is gone, in our minds a lot of times that's the equivalent.
Now, the other point with that is that people die of systemic disease with esophageal cancer, so we try to get as much chemotherapy in them as we can. Typically, we find that is easier with chemotherapy alone versus chemoradiation. Not to say that we're not starting to think about radiation, particularly given the recent report from the CROSS trial. For appropriate patients, we'll consider chemoradiation now.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. T. Martin, M. Worni, J. B. Zwischenberger, B. Gloor, R. Pietrobon, T. A. D'Amico, and M. F. Berry The Role of Radiation Therapy in Resected T2 N0 Esophageal Cancer: A Population-Based Analysis Ann. Thorac. Surg., February 1, 2013; 95(2): 453 - 458. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Q. Zhang, C. M. Hooker, M. V. Brock, J. Shin, S. Lee, R. How, N. Franco, H. Prevas, A. Hulbert, and S. C. Yang Neoadjuvant Chemoradiation Therapy Is Beneficial for Clinical Stage T2 N0 Esophageal Cancer Patients Due to Inaccurate Preoperative Staging Ann. Thorac. Surg., February 1, 2012; 93(2): 429 - 437. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
