- Email: [email protected]
Utility of Restaging Endoscopic Ultrasound After Neoadjuvant Therapy for Esophageal Cancer
Jeffrey M. Griffin, MD, MS, Carolyn E. Reed, MD, and Chadrick E. Denlinger, MD Division of Cardiothoracic Surgery, Medical University of South Carolina, Charleston, South Carolina
Background. Currently, the most accurate staging test for patients with esophageal cancer is endoscopic ultrasound (EUS). At many institutions, patients who have completed neoadjuvant chemotherapy and radiotherapy for esophageal cancer undergo restaging EUS before proceeding to surgical resection. The benefit of this restaging procedure remains controversial. Methods. We retrospectively studied consecutive patients who had pre-resection restaging EUS after receiving neoadjuvant treatment to assess accuracy of EUS restaging and determine whether it predicted survival. Results. Final pathologic data were available for 73 patients who underwent restaging EUS (3 patients had missing T or N stage at one time point). Median time from restaging EUS to resection was 20 days. Restaging EUS accurately predicted pathologic T status in 26 of 72 patients (36%), N status in 44 of 71 (62%), and detected a
complete pathologic response in 2 of 19 (10.5%). EUS inappropriately classified 10 patients as T0 N0. Agreement between EUS and pathologic staging was poor for T ( ⴝ 0.14) and N status ( ⴝ 0.24). Median time from resection to death or last follow-up was 20 months. Pathologic T and N status were each significant predictors of survival (p ⴝ 0.049 and p ⴝ 0.0004, respectively). There were nonsignificant trends toward better survival for lower EUS T (p ⴝ 0.32) and N status (p ⴝ 0.0946). Conclusions. Restaging by EUS before resection did not accurately predict pathologic stage in patients with esophageal cancer who received neoadjuvant treatment. As a result of this investigation, our institution no longer routinely performs restaging EUS.
T
with esophageal cancer. In some centers, restaging EUS is also used to assess response to neoadjuvant therapy. The utility of restaging EUS has been questioned in numerous reports during the past decade, largely due to difficulty distinguishing residual primary tumor from radiation effect and fibrosis. However, several reports have described the accuracy of restaging EUS as being comparable to that of PET, and this modality continues to be used for restaging in some centers [5]. The objective of the current study was to determine the accuracy of restaging EUS in predicting pathologic T and N status among patients with esophageal cancer treated with neoadjuvant chemotherapy and radiotherapy. Secondary objectives were to determine whether restaging EUS was predictive of long-term survival or altered clinical decision-making.
he incidence of esophageal cancer has increased significantly in the United States during the past 3 decades [1]. Specifically, the incidence of esophageal adenocarcinoma has increased nearly 400% in the past 3 decades, whereas the incidence of esophageal squamous cell cancers has decreased modestly. Patients with earlystage disease typically remain asymptomatic; therefore, most patients diagnosed with esophageal cancer present with advanced-stage disease. Endoscopic ultrasound (EUS) has been shown to be the most accurate method for assessing depth of tumor invasion and involvement of regional lymph nodes in patients with esophageal cancer and is routinely used to determine the T and N status of esophageal cancers [2– 4]. Patients diagnosed with advanced-stage esophageal cancer, defined by T stage exceeding T3 or N stage exceeding N1, are treated with neoadjuvant chemotherapy and radiotherapy before surgical resection in most centers in the United States. Before undergoing surgical resection, patients are restaged to determine whether there has been local disease progression or development of distant metastases. Computed tomography and positron-emission tomography (PET) imaging are routinely used for restaging of patients
Accepted for publication Dec 19, 2011. Presented at the Fifty-eighth Annual Meeting of the Southern Thoracic Surgical Association, San Antonio, TX, Nov 9 –12, 2011. Address correspondence to Dr Denlinger, 25 Courtenay Dr, Ste 7018, Charleston, SC 29425; e-mail: [email protected].
© 2012 by The Society of Thoracic Surgeons Published by Elsevier Inc
(Ann Thorac Surg 2012;93:1855– 60) © 2012 by The Society of Thoracic Surgeons
Material and Methods This study was approved by the Institutional Review Board at the Medical University of South Carolina.
Study Design A retrospective review of a prospectively collected esophagectomy database was conducted. The database included all patients at the Medical University of South Carolina who underwent an esophagectomy. Patients who received neoadjuvant chemotherapy and radiotherapy, followed by esophagectomy for esophageal cancer between January 1998 and July 2010, were included in 0003-4975/$36.00 doi:10.1016/j.athoracsur.2011.12.095
GENERAL THORACIC
Utility of Restaging Endoscopic Ultrasound After Neoadjuvant Therapy for Esophageal Cancer
GENERAL THORACIC
1856
GRIFFIN ET AL RESTAGING EUS FOR ESOPHAGEAL CANCER
Statistical Analysis
Table 1. Demographic and Clinical Variables Variables Age, years Days from restaging to resection Sex Male Female Histopathologic type Adenocarcinoma Squamous cell Surgical approach Ivor Lewis Three-field Transhiatal Not resected (aortic invasion) Pathologic T N M classification pT0 N0 pT0 N1 pT1 N0 pT1 N1 pT2 N0 pT2 N1 pT3 N0 pT3 N1 pT4 Nx
Ann Thorac Surg 2012;93:1855– 60
Median (range) or No. (%) 62 (40–79) 22 (2–78) 64 (87.7) 9 (12.3) 66 (90.4) 7 (10.6) 55 (75.3) 16 (22.0) 1 (1.3) 1 (1.3) 20 (27.4) 3 (4.1) 11 (15.1) 7 (9.6) 6 (8.2) 2 (2.7) 11 (15.1) 12 (16.4) 1 (1.4)
this review. Routine preoperative evaluation at our institution during this interval included a restaging EUS after completion of neoadjuvant therapy before surgical resection. A record review was conducted to obtain restaging EUS T N classification, pathologic T N classification, and the date of the last clinical follow-up. The Social Security Death Master File was queried to determine survivorship or date of death as of March 2011.
Neoadjuvant Therapy, Surgical Treatment, and Staging All patients in the database with locally advanced esophageal cancer were treated with neoadjuvant therapy consisting of a platinum-based chemotherapy regimen and 45 to 50 Gy of radiotherapy. Patients were then reevaluated 4 to 6 weeks after completing neoadjuvant therapy for restaging and final preoperative workup. Restaging evaluation during this interval included EUS in almost all cases. T and N status was assigned by the endoscopist and documented on the EUS procedure note. Most patients were resected using Ivor Lewis esophagectomy. A smaller number underwent three-field and transhiatal approaches. Patients were staged according to the Sixth Edition of the American Joint Committee on Cancer Staging Manual and Handbook (AJCC 6th).
Descriptive statistics, including frequency, percentage, median, and range, are presented for demographic and clinical variables. T and N status in each patient, as determined by restaging EUS, was compared with the final pathologic status for all patients. Overall agreement between EUS and pathologic staging was assessed with statistics. Survival from the date of surgical resection was analyzed by Kaplan-Meier survival analysis. The logrank test was used to test for differential survival by pathologic T status and by restaging EUS T status. All analyses were performed in SAS 9.2 software (SAS Institute, Cary, NC).
Results Our prospectively collected esophageal cancer database included 104 patients with locally advanced esophageal cancer treated with neoadjuvant chemoradiotherapy, followed by surgical resection, between January 1998 and July 2010. Of these, the analysis excluded 31 patients for the following reasons: 28 did not have restaging EUS records, 2 did not undergo restaging EUS, and surgical resection in 1 patient was aborted due to intraoperative finding of liver metastases. The remaining 73 patients (64 men, 9 women) comprised our analysis cohort. Patient demographics and clinical variables are summarized Table 1. Patients were a median age of 62 years (range, 40 to 79 years). Histopathologic classification was adenocarcinoma in 66 (90.4%) and squamous cell carcinoma in 7 (9.6%). The surgical approach was Ivor Lewis esophagectomy in 55 patients, three-field in 16, and transhiatal in 1. One patient was not resected due to intraoperative finding of aortic sheath invasion but was included in the analysis and assigned a pathologic TN status of pT4 Nx. As reported in Table 2, there was poor agreement between restaging EUS and pathologic T status ( ⫽ 0.1368). Restaging EUS accurately predicted the pathologic T status in only 36% of the patients: it overstaged 46% and understaged 17%. More important, 19 patients Table 2. Comparison of Endoscopic Ultrasound and Pathologic T Status Based on the American Joint Committee on Cancer Staging Manual (Sixth Edition) Pathologic Stage EUS Stage T0 T1 T2 T3 T4 Total a
pT0
pT1
pT2
pT3
PT4
Total
2 2 5 12 1 22
3 4 6 5 0 18
0 0 5 3 0 8
7 1 0 15 0 23
0 0 0 1 0 1
12 7 16 36 1 72a
One patient with missing EUS T status not included.
Shaded cells indicate accurate staging by EUS. Cells to the right represent understaging by EUS and cells to the left indicate overstaging by EUS. EUS ⫽ endobronchial ultrasound.
1857
Table 3. Comparison of Endoscopic Ultrasound and Pathologic N Status Based on the American Joint Committee on Cancer Staging Manual (Sixth Edition)
1.0
T2 T3
Pathologic Stage EUS Stage
pN0
pN1
Total
28 19 47
8 16 24
36 35 71a
N0 N1 Total a
Two patients with missing N status not included.
T0 T1
0.8 0.6 0.4 0.2 0.0
Shaded cells indicate accurate staging by EUS: 8 patients were understaged by EUS and 19 patients were overstaged by EUS. EUS ⫽ endoscopic ultrasound.
0
20
23 18 8 23
11 11 4 10
Number at risk
Table 4. Comparison of Restaging Endoscopic Ultrasound and Pathologic Stage According to the American Joint Committee on Cancer Staging Manual (Sixth Edition) Pathologic Stage EUS Stage a
0 I IIa IIb III Incompleteb Total
0
a
2 1 4 3 7 2 19
I
IIa
IIb
III
Iva
Total
3 1 4 2 0 ... 10
4 1 6 1 3 ... 15
0 0 3 4 4 ... 11
3 0 1 0 7 ... 11
0 0 3 1 3 ... 7
12 3 21 11 24 2 73
a
b Indicates complete response. Patients incompletely restaged by EUS as T1 Nx M0 (n ⫽ 1) and Tx N1 M0 (n ⫽ 1). Shaded cells indicate accurate staging by EUS. Cells to the right represent understaging by EUS, and cells to the left indicate overstaging by EUS.
EUS ⫽ endoscopic ultrasound.
T0 T1 T2 T3
5 7 1 5
80
4 3 1 2
Fig 1. Probability of overall survival is shown according to pathologic T classification.
dence interval) was 74 months (38 to 105 months) for patients classified as pN0 and 13 months (7 to 26 months) for those classified as pN1 (p ⫽ 0.0004). The pathologic T stage is dependent on the presence of viable tumor cells within the esophageal wall and the depth of penetration of these cells. More important, the current staging system does not include the degree of tumor response to induction therapy in determining the final pathologic stage. Therefore, a tumor with a nearly complete pathologic response with a small focus of viable cells penetrating through the muscularis propria is considered a T3 lesion. A small focus of viable cells may not be appreciated by the restaging EUS. To determine whether the restaging EUS may correlate more closely with an overall treatment response, rather than final pathologic tumor stage, Kaplan-Meier survival curves
1.0 Kaplan-Meier Esmates
in the entire cohort were pCR, with no residual disease identified in the resected specimen. Only 2 of these patients were identified as being complete pathologic responders on the restaging EUS. The relationship between nodal status as determined by restaging EUS and pathology is summarized in Table 3. EUS accurately predicted pathologic N status in 62% of patients. However, overall agreement between restaging EUS N status and the final pathologic N status was poor ( ⫽ 0.2360). Table 4 reports the relationship between EUS restage and final pathologic stage. In 7 patients, positive celiac axis lymph nodes were identified in the pathologic specimen (AJCC 6th edition, stage IVa). Restaging EUS did not detect the celiac lymph node involvement in any of these patients. Median duration of follow-up for the entire cohort was 20 months. Figure 1 shows Kaplan-Meier estimates of overall survival stratified by pathologic T classification. Median survival was 74 months for patients pathologically classified as pT0, 78 months for those classified as pT1, 27 months for those classified as pT2, and 23 months for those classified as pT3 (log-rank 2 test of equality across groups, p ⫽ 0.049). Figure 2 shows survival based on pathologic N status. The median survival (95% confi-
40 60 Survival (Months)
N0 N1
0.8 0.6 0.4 0.2 0.0 0
20
48 24
28 8
Number at risk N0 N1
40 60 Survival (Months) 16 2
80
10 1
Fig 2. Probability of overall survival is shown according to pathologic N classification.
GENERAL THORACIC
GRIFFIN ET AL RESTAGING EUS FOR ESOPHAGEAL CANCER
Kaplan-Meier Esmates
Ann Thorac Surg 2012;93:1855– 60
GRIFFIN ET AL RESTAGING EUS FOR ESOPHAGEAL CANCER
T0 T1
Ann Thorac Surg 2012;93:1855– 60
T2 T3
0.8 0.6 0.4 0.2 0.0 0
20
12 7 16 36
9 4 7 15
Number at risk T0 T1 T2 T3
40 60 Survival (Months) 3 3 4 8
80
3 1 2 5
Fig 3. Probability of overall survival is shown according to restaging endoscopic ultrasound T classification.
were plotted according to the repeat EUS T stage. Figure 3 shows survival based on restaging EUS T status. The median survival was 78 months for those restaged by EUS as T0, 17 months for those restaged as T2, and 27 months for those restaged as T3 (p ⫽ 0.32). Median survival could not be estimated for the T1 group because of the small number of patients at risk. Most patients in this group were still alive at the end of follow-up. Figure 4 shows survival based on restaging EUS N status. The median (95% confidence interval) survival was 44 months (25 to 105 months) for patients classified as N0 by EUS and 17 months (13 to 66 months) for those classified as N1 (p ⫽ 0.0946).
Comment There was poor overall agreement between restaging EUS and pathologic T classification, with substantial numbers of patients being understaged or overstaged. Agreement was higher for N classification (62%), although a substantial percentage of patients were still understaged or overstaged. In this cohort of patients, the sensitivity of restaging EUS to detect M1a status (defined by celiac axis node positivity in the AJCC 6th edition) was 0%. Celiac axis node positivity now constitutes N1 disease in the AJCC 7th edition, but it is still an important prognostic factor and is predictive of overall survival [6, 7]. Moreover, if the 7 patients with positive celiac axis nodes were reclassified as N1 (according to the AJCC 7th edition), the accuracy of EUS for predicting pathologic nodal status would be substantially lower than our estimate of 62%. The overall survival of patients with esophageal cancer in the present study is similar to previously published results when stratified by pathologic TN status or overall pathologic stage. This suggests that our data are a rep-
resentative sample of the larger population of locally advanced esophageal cancer patients referred for neoadjuvant therapy and surgical resection. Restaging EUS was especially inaccurate at predicting final pathologic T classification, and there was not a statistically significant survival difference by EUS T status. However, there was a trend toward improved survival among patients with lower EUS restaging T classifications (p ⫽ 0.0549). One explanation may be that some of the patients understaged as T0 or T1 by EUS had an overall good response to neoadjuvant therapy but were ultimately pathologically classified as pT2 or pT3 due to isolated foci of residual carcinoma deep within the esophageal wall. As described previously, these nearly complete responders would be expected to have improved overall and diseasefree survival. Unfortunately, only the most recent pathology reports in the current patient cohort included a treatment response estimate, and we were not able to compare the degree of treatment response with EUS restaging. Other studies have previously described the diminished accuracy of EUS after neoadjuvant therapy [8 –12]. Until recently, however, restaging EUS was still routinely used at our institution to rule out disease progression over the course of the neoadjuvant treatments and to rule out the presence of celiac lymph node involvement. In addition, other institutions continue to use restaging EUS after neoadjuvant treatments for the same reasons, despite data suggesting its inaccuracy [5]. Esophagectomy is an extensive operation with significant associated morbidity and mortality rates. An overall perioperative mortality of 3% to 4% has been reported for patients undergoing esophagectomy at high-volume centers and approaches 20% at lower-volume centers [13– 15]. Perioperative complications contributing significantly to death after esophagectomy include anastomotic leaks, pneumonia and respiratory failure, recurrent laryngeal nerve injury, cardiac arrhythmia, and chylothorax. In addition, patients who do well after an esopha-
1.0 Kaplan-Meier Esmates
1.0 Kaplan-Meier Esmates
GENERAL THORACIC
1858
N0 N1
0.8 0.6 0.4 0.2 0.0 0
20
36 36
22 13
Number at risk N0 N1
40 60 Survival (Months) 11 7
80
7 4
Fig 4. Probability of overall survival is shown according to restaging endoscopic ultrasound N classification.
gectomy make significant lifestyle changes related to early satiety, chronic gastroesophageal reflux, and dumping syndrome. Owing to the relatively high operative risk of death and significant complications associated with esophagectomy, careful selection of patients who are likely to benefit the greatest from this operation is paramount. Accurate preoperative assessment of patients who have had a complete pathologic response to neoadjuvant therapy or, conversely, identification of patients who have had disease progression during neoadjuvant therapy, may guide treatment by avoiding surgical risk in patients who not likely to receive additional benefit from an esophagectomy. Restaging modalities used for this purpose include computed tomography, PET, endoscopy, and in some centers, restaging EUS. Although computed tomography and PET imaging are recognized as providing superior assessment of distant metastases, EUS has been reported to be more likely to recognize progression of nodal disease or persistent disease in celiac lymph nodes, which carries a more dismal prognosis [2, 5, 16]. The identification of persistent celiac lymph node involvement in patients with esophageal cancer may contribute to a more educated discussion with the patient about the risks and benefits of a planned surgical resection. This study has several limitations, including the usual problems associated with a retrospective study. In addition, the relatively small number of patients included may have resulted in an imprecise estimation of the true accuracy of restaging EUS and may have also resulted in inadequate power to detect survival differences across restaging EUS T N groups. The trends toward improved survival with lower restaging EUS T and N status might have become significant if larger numbers of patients were included. Approximately 30% of patients were excluded from the analysis because their EUS restaging documentation was missing. This occurred most often early in the series. Information contained in missing reports could have altered the results, but there did not appear to be anything unique about these excluded patients compared with those comprising the study cohort. In addition, all patients in our cohort received neoadjuvant chemotherapy and 45 to 50 Gy of radiotherapy. Thus, we are not able to comment on the accuracy of restaging EUS among patients who receive only chemotherapy as neoadjuvant treatment. Finally, this series consisted of a highly selected group of patients with locally advanced esophageal cancer who received neoadjuvant therapy and proceeded to resection. For example, any patients found to have positive celiac axis lymph nodes or aortic invasion on restaging EUS would not have been offered resection, and hence would not have been included in the series. Therefore, we cannot estimate the true sensitivity of restaging EUS to detect unresectable disease in the larger population of patients with esophageal cancer. These limitations notwithstanding, the analysis of our own data revealed that restaging EUS had very
GRIFFIN ET AL RESTAGING EUS FOR ESOPHAGEAL CANCER
1859
poor correlation with the final pathologic stage. In addition, restaging EUS was unable to reliably detect the presence of disease within the celiac lymph nodes. Even though the restaging EUS does appear to provide some prognostic information, the results of this test have not influenced our clinical decisions. Therefore, as the result of analyzing our own data, we no longer perform restaging EUS after treatment with neoadjuvant chemotherapy and radiotherapy in patients with esophageal cancer.
References 1. Devesa SS, Blot WJ, Fraumeni FJ. Changing patterns in the incidence of esophageal and gastric carcinoma in the United States. Cancer 1998;83:2049 –53. 2. Botet JF, Lightdale CJ, Zauber AG, Gerdes H, Urmacher C, Brennan MF. Preoperative staging of esophageal cancer: Comparison of endoscopic US and dynamic CT. Radiology 1991;181:419 –25. 3. Tio TL, Cohen P, Coene PP, Udding J, den Hartog Jager FC, Tytgat GN. Endoscopic and computed tomography of esophageal carcinoma. Preoperative classification compared to the new (1987) TNM system. Gastroenterol 1989;96: 1478 – 6. 4. Vilgrain V. Mompoint D, Palazzo L, et al. Staging of esophageal carcinoma: comparison of results with endoscopic sonography and CT. AJR Am J Roentgenol 1990; 155:277– 81. 5. Eloubeidi MA, Cerfolio RJ, Bryant AS, Varadarajulu S. Efficacy of endoscopic ultrasound in patients with esophageal cancer. Eur J Cardiothorac Surg 2011;40:636 – 41. 6. Rice TW, Blackstone EH, Rybicki LA, et al. Refining esophageal cancer staging. J Thorac Cardiovasc Surg 2003;125: 1103–13. 7. Trovo M, Bradley J, El Naqa I, et al. Esophageal carcinoma with celiac nodal metastases; curative or palliative? J Thorac Oncol 2008;3:751–5. 8. Koch J, Halvorsen RA. Staging of esophageal cancer: computed tomography, magnetic resonance imaging, and endoscopic ultrasound. Semin Roentgenol 1994;29:364 –72. 9. Laterza E, de Manzoni G, Guglielmi A, et al. Endoscopic ultrasonography in the staging of esophageal carcinoma after preoperative radiotherapy and chemotherapy. Ann Thorac Surg 1999;67:1466 –9. 10. Zuccaro G, Rice TW, Goldblum J, et al. Endoscopic ultrasound cannot determine suitability for esophagectomy after aggressive chemoradiotherapy for esophageal cancer. Am J Gastroenterol 1999;94:906 –12. 11. Chak A, Canto MI, Cooper GS, et al. Endosonographic assessment of multimodality therapy predicts survival of esophageal cancer patients. Cancer 2000;88:1788 –95. 12. Kelly S, Harris K, Berry E, et al. A systematic review of the staging performance of endoscopic ultrasound in gastroesophageal carcinoma. Gut 2001;49:534 –9. 13. Ferguson MK, Martin TR, Reeder LB, Olak J. Mortality after esophagectomy: risk factor analysis. World J Surg 1997;21: 599 – 603. 14. Miller JD, Jain MK, de Gara CJ, et al. Effect of surgical experience on results of esophagectomy for esophageal carcinoma. J Surg Oncol 1997;165:20 –1. 15. Begg CB, Cramer LD, Hoskins WJ, Brennan MF. Impact of hospital volume on operative mortality for major cancer surgery. JAMA 1998;280:1747–51. 16. Christie NA, Rice TW, DeCamp MM, et al. M1a/M1b esophageal carcinoma; clinical relevance. J Thorac Cardiovasc Surg 1999;118:900 –7.
GENERAL THORACIC
Ann Thorac Surg 2012;93:1855– 60
GENERAL THORACIC
1860
GRIFFIN ET AL RESTAGING EUS FOR ESOPHAGEAL CANCER
Ann Thorac Surg 2012;93:1855– 60
DISCUSSION DR ROBERT J. CERFOLIO (Birmingham, AL): Congratulations. Two quick questions. I disagree with your last conclusion. We still use restaging endoscopic ultrasound (EUS), because if you have a patient with residual celiac disease—and I was surprised that you had 0 out of 7—it is valuable and may lead, in selected patients, to second-line chemotherapy instead of resection. And you should use the same endosonographer so they can make the decision every time that it is the basis for recalcitrant celiac disease, and if the patient is and elderly or high risk that is my out, I am not going to do the operation or if he is a young guy in good shape, I would consider a second-line of chemo, although I don’t have any data for it. So I still think it does provide clinical direction. So I disagree with you. I think computed tomography (CT)-positron-emission tomography (PET) is better than repeat EUS for this we have written a paper it was in the Journal of Thoracic and Cardiovascular Surgery and presented it at the Western a few years ago. But you didn’t tell us about PET. Do you have any comments on PET? DR GRIFFIN: Not at this time, but I would like to thank you for your point, though, about the celiac nodal disease. And I think one of the big limitations of our study was it really only includes patients who underwent resection. So I can tell you that 0 of 7 patients with celiac nodal disease were detected in this series. However, those patients who had positive celiac nodes didn’t proceed to resection and therefore weren’t included in the database. DR CERFOLIO: Therefore, repeat EUS changed your management, so you better change the conclusion in your paper or at least the last bullet on the last slide you presented.
he wants resection, as they all do because they are convinced it is their only change for cure, your out is biopsy-proven multiple recalcitrant lymph nodes—you need repeat EUS for that. DR GRIFFIN: Thank you. DR BENJAMIN D. KOZOWER (Charlottesville, VA): Have you done any quality control for your EUS and compared your EUS results with your pathologic results? Because the first thing I think of when I see how poorly EUS performed in this study is, who is doing your EUS staging? DR GRIFFITH: That thought has actually crossed our minds, too. Maybe we are just especially terrible at it. I don’t really think we have access to all those initial stages that were provided by the sonographer prior to the completion of neoadjuvant therapy. But you are right; it may be something that is worth looking into. DR SHANDA HALEY BLACKMON (Houston, TX): Along the same lines as your patient that had the nodal disease that wasn’t sent for esophagectomy, what about those patients that had aortic involvement and never went for resection? That is a real utility of postneoadjuvant treatment EUS. We picked up 2 patients that we would have taken for surgery by CT or PET, but by EUS we found aortic involvement and we ended up not taking them for surgery. And I know that you may have missed 1 patient that EUS said didn’t have aortic involvement and you took him for surgery and then found out and then stopped. DR GRIFFIN: Correct.
DR JOSHUA ROBERT SONETT (New York, NY): Cerf, so would you restage EUS if the patient did not have celiac nodes preoperatively?
DR BLACKMON: But you probably need to state in your paper how many people were found to have aortic involvement and never went to surgery, because that is a real utility of EUS. But I would like to congratulate you on a very well presented paper.
DR CERFOLIO: That is a good point in those patients. I am not sure if it really is clinically helpful.
DR GRIFFITH: Thank you very much.
DR GRIFFIN: What about with the new staging system, where positive celiac nodes are no longer considered N1A disease? DR CERFOLIO: Well, if their repeat PET showed multiple nodal stations, the issue was always tissue and I would like to stick the needle in it if it is not peritumoral, and then the repeat EUS helps you, provided there is recalcitrant nodal disease, and it’s a reason not to offer a potential operation. If it is a 78-year-old guy that had radiation chemo, he appears weak after chemorads, but
DR SONETT: I would be worried about a T4 by EUS and the aorta really being a T4 by EUS, and I would, at a minimum, start with some type of video, be it robot or thoracoscopic, to really prove that it is or not aortic, because I certainly have seen overcalls by EUS. DR CERFOLIO: We have had false positive repeat EUS after radiation chemo. Initially they are active, but if they are repeat, it is even harder post-op.
Background. Currently, the most accurate staging test for patients with esophageal cancer is endoscopic ultrasound (EUS). At many institutions, patients who have completed neoadjuvant chemotherapy and radiotherapy for esophageal cancer undergo restaging EUS before proceeding to surgical resection. The benefit of this restaging procedure remains controversial. Methods. We retrospectively studied consecutive patients who had pre-resection restaging EUS after receiving neoadjuvant treatment to assess accuracy of EUS restaging and determine whether it predicted survival. Results. Final pathologic data were available for 73 patients who underwent restaging EUS (3 patients had missing T or N stage at one time point). Median time from restaging EUS to resection was 20 days. Restaging EUS accurately predicted pathologic T status in 26 of 72 patients (36%), N status in 44 of 71 (62%), and detected a
complete pathologic response in 2 of 19 (10.5%). EUS inappropriately classified 10 patients as T0 N0. Agreement between EUS and pathologic staging was poor for T ( ⴝ 0.14) and N status ( ⴝ 0.24). Median time from resection to death or last follow-up was 20 months. Pathologic T and N status were each significant predictors of survival (p ⴝ 0.049 and p ⴝ 0.0004, respectively). There were nonsignificant trends toward better survival for lower EUS T (p ⴝ 0.32) and N status (p ⴝ 0.0946). Conclusions. Restaging by EUS before resection did not accurately predict pathologic stage in patients with esophageal cancer who received neoadjuvant treatment. As a result of this investigation, our institution no longer routinely performs restaging EUS.
T
with esophageal cancer. In some centers, restaging EUS is also used to assess response to neoadjuvant therapy. The utility of restaging EUS has been questioned in numerous reports during the past decade, largely due to difficulty distinguishing residual primary tumor from radiation effect and fibrosis. However, several reports have described the accuracy of restaging EUS as being comparable to that of PET, and this modality continues to be used for restaging in some centers [5]. The objective of the current study was to determine the accuracy of restaging EUS in predicting pathologic T and N status among patients with esophageal cancer treated with neoadjuvant chemotherapy and radiotherapy. Secondary objectives were to determine whether restaging EUS was predictive of long-term survival or altered clinical decision-making.
he incidence of esophageal cancer has increased significantly in the United States during the past 3 decades [1]. Specifically, the incidence of esophageal adenocarcinoma has increased nearly 400% in the past 3 decades, whereas the incidence of esophageal squamous cell cancers has decreased modestly. Patients with earlystage disease typically remain asymptomatic; therefore, most patients diagnosed with esophageal cancer present with advanced-stage disease. Endoscopic ultrasound (EUS) has been shown to be the most accurate method for assessing depth of tumor invasion and involvement of regional lymph nodes in patients with esophageal cancer and is routinely used to determine the T and N status of esophageal cancers [2– 4]. Patients diagnosed with advanced-stage esophageal cancer, defined by T stage exceeding T3 or N stage exceeding N1, are treated with neoadjuvant chemotherapy and radiotherapy before surgical resection in most centers in the United States. Before undergoing surgical resection, patients are restaged to determine whether there has been local disease progression or development of distant metastases. Computed tomography and positron-emission tomography (PET) imaging are routinely used for restaging of patients
Accepted for publication Dec 19, 2011. Presented at the Fifty-eighth Annual Meeting of the Southern Thoracic Surgical Association, San Antonio, TX, Nov 9 –12, 2011. Address correspondence to Dr Denlinger, 25 Courtenay Dr, Ste 7018, Charleston, SC 29425; e-mail: [email protected].
© 2012 by The Society of Thoracic Surgeons Published by Elsevier Inc
(Ann Thorac Surg 2012;93:1855– 60) © 2012 by The Society of Thoracic Surgeons
Material and Methods This study was approved by the Institutional Review Board at the Medical University of South Carolina.
Study Design A retrospective review of a prospectively collected esophagectomy database was conducted. The database included all patients at the Medical University of South Carolina who underwent an esophagectomy. Patients who received neoadjuvant chemotherapy and radiotherapy, followed by esophagectomy for esophageal cancer between January 1998 and July 2010, were included in 0003-4975/$36.00 doi:10.1016/j.athoracsur.2011.12.095
GENERAL THORACIC
Utility of Restaging Endoscopic Ultrasound After Neoadjuvant Therapy for Esophageal Cancer
GENERAL THORACIC
1856
GRIFFIN ET AL RESTAGING EUS FOR ESOPHAGEAL CANCER
Statistical Analysis
Table 1. Demographic and Clinical Variables Variables Age, years Days from restaging to resection Sex Male Female Histopathologic type Adenocarcinoma Squamous cell Surgical approach Ivor Lewis Three-field Transhiatal Not resected (aortic invasion) Pathologic T N M classification pT0 N0 pT0 N1 pT1 N0 pT1 N1 pT2 N0 pT2 N1 pT3 N0 pT3 N1 pT4 Nx
Ann Thorac Surg 2012;93:1855– 60
Median (range) or No. (%) 62 (40–79) 22 (2–78) 64 (87.7) 9 (12.3) 66 (90.4) 7 (10.6) 55 (75.3) 16 (22.0) 1 (1.3) 1 (1.3) 20 (27.4) 3 (4.1) 11 (15.1) 7 (9.6) 6 (8.2) 2 (2.7) 11 (15.1) 12 (16.4) 1 (1.4)
this review. Routine preoperative evaluation at our institution during this interval included a restaging EUS after completion of neoadjuvant therapy before surgical resection. A record review was conducted to obtain restaging EUS T N classification, pathologic T N classification, and the date of the last clinical follow-up. The Social Security Death Master File was queried to determine survivorship or date of death as of March 2011.
Neoadjuvant Therapy, Surgical Treatment, and Staging All patients in the database with locally advanced esophageal cancer were treated with neoadjuvant therapy consisting of a platinum-based chemotherapy regimen and 45 to 50 Gy of radiotherapy. Patients were then reevaluated 4 to 6 weeks after completing neoadjuvant therapy for restaging and final preoperative workup. Restaging evaluation during this interval included EUS in almost all cases. T and N status was assigned by the endoscopist and documented on the EUS procedure note. Most patients were resected using Ivor Lewis esophagectomy. A smaller number underwent three-field and transhiatal approaches. Patients were staged according to the Sixth Edition of the American Joint Committee on Cancer Staging Manual and Handbook (AJCC 6th).
Descriptive statistics, including frequency, percentage, median, and range, are presented for demographic and clinical variables. T and N status in each patient, as determined by restaging EUS, was compared with the final pathologic status for all patients. Overall agreement between EUS and pathologic staging was assessed with statistics. Survival from the date of surgical resection was analyzed by Kaplan-Meier survival analysis. The logrank test was used to test for differential survival by pathologic T status and by restaging EUS T status. All analyses were performed in SAS 9.2 software (SAS Institute, Cary, NC).
Results Our prospectively collected esophageal cancer database included 104 patients with locally advanced esophageal cancer treated with neoadjuvant chemoradiotherapy, followed by surgical resection, between January 1998 and July 2010. Of these, the analysis excluded 31 patients for the following reasons: 28 did not have restaging EUS records, 2 did not undergo restaging EUS, and surgical resection in 1 patient was aborted due to intraoperative finding of liver metastases. The remaining 73 patients (64 men, 9 women) comprised our analysis cohort. Patient demographics and clinical variables are summarized Table 1. Patients were a median age of 62 years (range, 40 to 79 years). Histopathologic classification was adenocarcinoma in 66 (90.4%) and squamous cell carcinoma in 7 (9.6%). The surgical approach was Ivor Lewis esophagectomy in 55 patients, three-field in 16, and transhiatal in 1. One patient was not resected due to intraoperative finding of aortic sheath invasion but was included in the analysis and assigned a pathologic TN status of pT4 Nx. As reported in Table 2, there was poor agreement between restaging EUS and pathologic T status ( ⫽ 0.1368). Restaging EUS accurately predicted the pathologic T status in only 36% of the patients: it overstaged 46% and understaged 17%. More important, 19 patients Table 2. Comparison of Endoscopic Ultrasound and Pathologic T Status Based on the American Joint Committee on Cancer Staging Manual (Sixth Edition) Pathologic Stage EUS Stage T0 T1 T2 T3 T4 Total a
pT0
pT1
pT2
pT3
PT4
Total
2 2 5 12 1 22
3 4 6 5 0 18
0 0 5 3 0 8
7 1 0 15 0 23
0 0 0 1 0 1
12 7 16 36 1 72a
One patient with missing EUS T status not included.
Shaded cells indicate accurate staging by EUS. Cells to the right represent understaging by EUS and cells to the left indicate overstaging by EUS. EUS ⫽ endobronchial ultrasound.
1857
Table 3. Comparison of Endoscopic Ultrasound and Pathologic N Status Based on the American Joint Committee on Cancer Staging Manual (Sixth Edition)
1.0
T2 T3
Pathologic Stage EUS Stage
pN0
pN1
Total
28 19 47
8 16 24
36 35 71a
N0 N1 Total a
Two patients with missing N status not included.
T0 T1
0.8 0.6 0.4 0.2 0.0
Shaded cells indicate accurate staging by EUS: 8 patients were understaged by EUS and 19 patients were overstaged by EUS. EUS ⫽ endoscopic ultrasound.
0
20
23 18 8 23
11 11 4 10
Number at risk
Table 4. Comparison of Restaging Endoscopic Ultrasound and Pathologic Stage According to the American Joint Committee on Cancer Staging Manual (Sixth Edition) Pathologic Stage EUS Stage a
0 I IIa IIb III Incompleteb Total
0
a
2 1 4 3 7 2 19
I
IIa
IIb
III
Iva
Total
3 1 4 2 0 ... 10
4 1 6 1 3 ... 15
0 0 3 4 4 ... 11
3 0 1 0 7 ... 11
0 0 3 1 3 ... 7
12 3 21 11 24 2 73
a
b Indicates complete response. Patients incompletely restaged by EUS as T1 Nx M0 (n ⫽ 1) and Tx N1 M0 (n ⫽ 1). Shaded cells indicate accurate staging by EUS. Cells to the right represent understaging by EUS, and cells to the left indicate overstaging by EUS.
EUS ⫽ endoscopic ultrasound.
T0 T1 T2 T3
5 7 1 5
80
4 3 1 2
Fig 1. Probability of overall survival is shown according to pathologic T classification.
dence interval) was 74 months (38 to 105 months) for patients classified as pN0 and 13 months (7 to 26 months) for those classified as pN1 (p ⫽ 0.0004). The pathologic T stage is dependent on the presence of viable tumor cells within the esophageal wall and the depth of penetration of these cells. More important, the current staging system does not include the degree of tumor response to induction therapy in determining the final pathologic stage. Therefore, a tumor with a nearly complete pathologic response with a small focus of viable cells penetrating through the muscularis propria is considered a T3 lesion. A small focus of viable cells may not be appreciated by the restaging EUS. To determine whether the restaging EUS may correlate more closely with an overall treatment response, rather than final pathologic tumor stage, Kaplan-Meier survival curves
1.0 Kaplan-Meier Esmates
in the entire cohort were pCR, with no residual disease identified in the resected specimen. Only 2 of these patients were identified as being complete pathologic responders on the restaging EUS. The relationship between nodal status as determined by restaging EUS and pathology is summarized in Table 3. EUS accurately predicted pathologic N status in 62% of patients. However, overall agreement between restaging EUS N status and the final pathologic N status was poor ( ⫽ 0.2360). Table 4 reports the relationship between EUS restage and final pathologic stage. In 7 patients, positive celiac axis lymph nodes were identified in the pathologic specimen (AJCC 6th edition, stage IVa). Restaging EUS did not detect the celiac lymph node involvement in any of these patients. Median duration of follow-up for the entire cohort was 20 months. Figure 1 shows Kaplan-Meier estimates of overall survival stratified by pathologic T classification. Median survival was 74 months for patients pathologically classified as pT0, 78 months for those classified as pT1, 27 months for those classified as pT2, and 23 months for those classified as pT3 (log-rank 2 test of equality across groups, p ⫽ 0.049). Figure 2 shows survival based on pathologic N status. The median survival (95% confi-
40 60 Survival (Months)
N0 N1
0.8 0.6 0.4 0.2 0.0 0
20
48 24
28 8
Number at risk N0 N1
40 60 Survival (Months) 16 2
80
10 1
Fig 2. Probability of overall survival is shown according to pathologic N classification.
GENERAL THORACIC
GRIFFIN ET AL RESTAGING EUS FOR ESOPHAGEAL CANCER
Kaplan-Meier Esmates
Ann Thorac Surg 2012;93:1855– 60
GRIFFIN ET AL RESTAGING EUS FOR ESOPHAGEAL CANCER
T0 T1
Ann Thorac Surg 2012;93:1855– 60
T2 T3
0.8 0.6 0.4 0.2 0.0 0
20
12 7 16 36
9 4 7 15
Number at risk T0 T1 T2 T3
40 60 Survival (Months) 3 3 4 8
80
3 1 2 5
Fig 3. Probability of overall survival is shown according to restaging endoscopic ultrasound T classification.
were plotted according to the repeat EUS T stage. Figure 3 shows survival based on restaging EUS T status. The median survival was 78 months for those restaged by EUS as T0, 17 months for those restaged as T2, and 27 months for those restaged as T3 (p ⫽ 0.32). Median survival could not be estimated for the T1 group because of the small number of patients at risk. Most patients in this group were still alive at the end of follow-up. Figure 4 shows survival based on restaging EUS N status. The median (95% confidence interval) survival was 44 months (25 to 105 months) for patients classified as N0 by EUS and 17 months (13 to 66 months) for those classified as N1 (p ⫽ 0.0946).
Comment There was poor overall agreement between restaging EUS and pathologic T classification, with substantial numbers of patients being understaged or overstaged. Agreement was higher for N classification (62%), although a substantial percentage of patients were still understaged or overstaged. In this cohort of patients, the sensitivity of restaging EUS to detect M1a status (defined by celiac axis node positivity in the AJCC 6th edition) was 0%. Celiac axis node positivity now constitutes N1 disease in the AJCC 7th edition, but it is still an important prognostic factor and is predictive of overall survival [6, 7]. Moreover, if the 7 patients with positive celiac axis nodes were reclassified as N1 (according to the AJCC 7th edition), the accuracy of EUS for predicting pathologic nodal status would be substantially lower than our estimate of 62%. The overall survival of patients with esophageal cancer in the present study is similar to previously published results when stratified by pathologic TN status or overall pathologic stage. This suggests that our data are a rep-
resentative sample of the larger population of locally advanced esophageal cancer patients referred for neoadjuvant therapy and surgical resection. Restaging EUS was especially inaccurate at predicting final pathologic T classification, and there was not a statistically significant survival difference by EUS T status. However, there was a trend toward improved survival among patients with lower EUS restaging T classifications (p ⫽ 0.0549). One explanation may be that some of the patients understaged as T0 or T1 by EUS had an overall good response to neoadjuvant therapy but were ultimately pathologically classified as pT2 or pT3 due to isolated foci of residual carcinoma deep within the esophageal wall. As described previously, these nearly complete responders would be expected to have improved overall and diseasefree survival. Unfortunately, only the most recent pathology reports in the current patient cohort included a treatment response estimate, and we were not able to compare the degree of treatment response with EUS restaging. Other studies have previously described the diminished accuracy of EUS after neoadjuvant therapy [8 –12]. Until recently, however, restaging EUS was still routinely used at our institution to rule out disease progression over the course of the neoadjuvant treatments and to rule out the presence of celiac lymph node involvement. In addition, other institutions continue to use restaging EUS after neoadjuvant treatments for the same reasons, despite data suggesting its inaccuracy [5]. Esophagectomy is an extensive operation with significant associated morbidity and mortality rates. An overall perioperative mortality of 3% to 4% has been reported for patients undergoing esophagectomy at high-volume centers and approaches 20% at lower-volume centers [13– 15]. Perioperative complications contributing significantly to death after esophagectomy include anastomotic leaks, pneumonia and respiratory failure, recurrent laryngeal nerve injury, cardiac arrhythmia, and chylothorax. In addition, patients who do well after an esopha-
1.0 Kaplan-Meier Esmates
1.0 Kaplan-Meier Esmates
GENERAL THORACIC
1858
N0 N1
0.8 0.6 0.4 0.2 0.0 0
20
36 36
22 13
Number at risk N0 N1
40 60 Survival (Months) 11 7
80
7 4
Fig 4. Probability of overall survival is shown according to restaging endoscopic ultrasound N classification.
gectomy make significant lifestyle changes related to early satiety, chronic gastroesophageal reflux, and dumping syndrome. Owing to the relatively high operative risk of death and significant complications associated with esophagectomy, careful selection of patients who are likely to benefit the greatest from this operation is paramount. Accurate preoperative assessment of patients who have had a complete pathologic response to neoadjuvant therapy or, conversely, identification of patients who have had disease progression during neoadjuvant therapy, may guide treatment by avoiding surgical risk in patients who not likely to receive additional benefit from an esophagectomy. Restaging modalities used for this purpose include computed tomography, PET, endoscopy, and in some centers, restaging EUS. Although computed tomography and PET imaging are recognized as providing superior assessment of distant metastases, EUS has been reported to be more likely to recognize progression of nodal disease or persistent disease in celiac lymph nodes, which carries a more dismal prognosis [2, 5, 16]. The identification of persistent celiac lymph node involvement in patients with esophageal cancer may contribute to a more educated discussion with the patient about the risks and benefits of a planned surgical resection. This study has several limitations, including the usual problems associated with a retrospective study. In addition, the relatively small number of patients included may have resulted in an imprecise estimation of the true accuracy of restaging EUS and may have also resulted in inadequate power to detect survival differences across restaging EUS T N groups. The trends toward improved survival with lower restaging EUS T and N status might have become significant if larger numbers of patients were included. Approximately 30% of patients were excluded from the analysis because their EUS restaging documentation was missing. This occurred most often early in the series. Information contained in missing reports could have altered the results, but there did not appear to be anything unique about these excluded patients compared with those comprising the study cohort. In addition, all patients in our cohort received neoadjuvant chemotherapy and 45 to 50 Gy of radiotherapy. Thus, we are not able to comment on the accuracy of restaging EUS among patients who receive only chemotherapy as neoadjuvant treatment. Finally, this series consisted of a highly selected group of patients with locally advanced esophageal cancer who received neoadjuvant therapy and proceeded to resection. For example, any patients found to have positive celiac axis lymph nodes or aortic invasion on restaging EUS would not have been offered resection, and hence would not have been included in the series. Therefore, we cannot estimate the true sensitivity of restaging EUS to detect unresectable disease in the larger population of patients with esophageal cancer. These limitations notwithstanding, the analysis of our own data revealed that restaging EUS had very
GRIFFIN ET AL RESTAGING EUS FOR ESOPHAGEAL CANCER
1859
poor correlation with the final pathologic stage. In addition, restaging EUS was unable to reliably detect the presence of disease within the celiac lymph nodes. Even though the restaging EUS does appear to provide some prognostic information, the results of this test have not influenced our clinical decisions. Therefore, as the result of analyzing our own data, we no longer perform restaging EUS after treatment with neoadjuvant chemotherapy and radiotherapy in patients with esophageal cancer.
References 1. Devesa SS, Blot WJ, Fraumeni FJ. Changing patterns in the incidence of esophageal and gastric carcinoma in the United States. Cancer 1998;83:2049 –53. 2. Botet JF, Lightdale CJ, Zauber AG, Gerdes H, Urmacher C, Brennan MF. Preoperative staging of esophageal cancer: Comparison of endoscopic US and dynamic CT. Radiology 1991;181:419 –25. 3. Tio TL, Cohen P, Coene PP, Udding J, den Hartog Jager FC, Tytgat GN. Endoscopic and computed tomography of esophageal carcinoma. Preoperative classification compared to the new (1987) TNM system. Gastroenterol 1989;96: 1478 – 6. 4. Vilgrain V. Mompoint D, Palazzo L, et al. Staging of esophageal carcinoma: comparison of results with endoscopic sonography and CT. AJR Am J Roentgenol 1990; 155:277– 81. 5. Eloubeidi MA, Cerfolio RJ, Bryant AS, Varadarajulu S. Efficacy of endoscopic ultrasound in patients with esophageal cancer. Eur J Cardiothorac Surg 2011;40:636 – 41. 6. Rice TW, Blackstone EH, Rybicki LA, et al. Refining esophageal cancer staging. J Thorac Cardiovasc Surg 2003;125: 1103–13. 7. Trovo M, Bradley J, El Naqa I, et al. Esophageal carcinoma with celiac nodal metastases; curative or palliative? J Thorac Oncol 2008;3:751–5. 8. Koch J, Halvorsen RA. Staging of esophageal cancer: computed tomography, magnetic resonance imaging, and endoscopic ultrasound. Semin Roentgenol 1994;29:364 –72. 9. Laterza E, de Manzoni G, Guglielmi A, et al. Endoscopic ultrasonography in the staging of esophageal carcinoma after preoperative radiotherapy and chemotherapy. Ann Thorac Surg 1999;67:1466 –9. 10. Zuccaro G, Rice TW, Goldblum J, et al. Endoscopic ultrasound cannot determine suitability for esophagectomy after aggressive chemoradiotherapy for esophageal cancer. Am J Gastroenterol 1999;94:906 –12. 11. Chak A, Canto MI, Cooper GS, et al. Endosonographic assessment of multimodality therapy predicts survival of esophageal cancer patients. Cancer 2000;88:1788 –95. 12. Kelly S, Harris K, Berry E, et al. A systematic review of the staging performance of endoscopic ultrasound in gastroesophageal carcinoma. Gut 2001;49:534 –9. 13. Ferguson MK, Martin TR, Reeder LB, Olak J. Mortality after esophagectomy: risk factor analysis. World J Surg 1997;21: 599 – 603. 14. Miller JD, Jain MK, de Gara CJ, et al. Effect of surgical experience on results of esophagectomy for esophageal carcinoma. J Surg Oncol 1997;165:20 –1. 15. Begg CB, Cramer LD, Hoskins WJ, Brennan MF. Impact of hospital volume on operative mortality for major cancer surgery. JAMA 1998;280:1747–51. 16. Christie NA, Rice TW, DeCamp MM, et al. M1a/M1b esophageal carcinoma; clinical relevance. J Thorac Cardiovasc Surg 1999;118:900 –7.
GENERAL THORACIC
Ann Thorac Surg 2012;93:1855– 60
GENERAL THORACIC
1860
GRIFFIN ET AL RESTAGING EUS FOR ESOPHAGEAL CANCER
Ann Thorac Surg 2012;93:1855– 60
DISCUSSION DR ROBERT J. CERFOLIO (Birmingham, AL): Congratulations. Two quick questions. I disagree with your last conclusion. We still use restaging endoscopic ultrasound (EUS), because if you have a patient with residual celiac disease—and I was surprised that you had 0 out of 7—it is valuable and may lead, in selected patients, to second-line chemotherapy instead of resection. And you should use the same endosonographer so they can make the decision every time that it is the basis for recalcitrant celiac disease, and if the patient is and elderly or high risk that is my out, I am not going to do the operation or if he is a young guy in good shape, I would consider a second-line of chemo, although I don’t have any data for it. So I still think it does provide clinical direction. So I disagree with you. I think computed tomography (CT)-positron-emission tomography (PET) is better than repeat EUS for this we have written a paper it was in the Journal of Thoracic and Cardiovascular Surgery and presented it at the Western a few years ago. But you didn’t tell us about PET. Do you have any comments on PET? DR GRIFFIN: Not at this time, but I would like to thank you for your point, though, about the celiac nodal disease. And I think one of the big limitations of our study was it really only includes patients who underwent resection. So I can tell you that 0 of 7 patients with celiac nodal disease were detected in this series. However, those patients who had positive celiac nodes didn’t proceed to resection and therefore weren’t included in the database. DR CERFOLIO: Therefore, repeat EUS changed your management, so you better change the conclusion in your paper or at least the last bullet on the last slide you presented.
he wants resection, as they all do because they are convinced it is their only change for cure, your out is biopsy-proven multiple recalcitrant lymph nodes—you need repeat EUS for that. DR GRIFFIN: Thank you. DR BENJAMIN D. KOZOWER (Charlottesville, VA): Have you done any quality control for your EUS and compared your EUS results with your pathologic results? Because the first thing I think of when I see how poorly EUS performed in this study is, who is doing your EUS staging? DR GRIFFITH: That thought has actually crossed our minds, too. Maybe we are just especially terrible at it. I don’t really think we have access to all those initial stages that were provided by the sonographer prior to the completion of neoadjuvant therapy. But you are right; it may be something that is worth looking into. DR SHANDA HALEY BLACKMON (Houston, TX): Along the same lines as your patient that had the nodal disease that wasn’t sent for esophagectomy, what about those patients that had aortic involvement and never went for resection? That is a real utility of postneoadjuvant treatment EUS. We picked up 2 patients that we would have taken for surgery by CT or PET, but by EUS we found aortic involvement and we ended up not taking them for surgery. And I know that you may have missed 1 patient that EUS said didn’t have aortic involvement and you took him for surgery and then found out and then stopped. DR GRIFFIN: Correct.
DR JOSHUA ROBERT SONETT (New York, NY): Cerf, so would you restage EUS if the patient did not have celiac nodes preoperatively?
DR BLACKMON: But you probably need to state in your paper how many people were found to have aortic involvement and never went to surgery, because that is a real utility of EUS. But I would like to congratulate you on a very well presented paper.
DR CERFOLIO: That is a good point in those patients. I am not sure if it really is clinically helpful.
DR GRIFFITH: Thank you very much.
DR GRIFFIN: What about with the new staging system, where positive celiac nodes are no longer considered N1A disease? DR CERFOLIO: Well, if their repeat PET showed multiple nodal stations, the issue was always tissue and I would like to stick the needle in it if it is not peritumoral, and then the repeat EUS helps you, provided there is recalcitrant nodal disease, and it’s a reason not to offer a potential operation. If it is a 78-year-old guy that had radiation chemo, he appears weak after chemorads, but
DR SONETT: I would be worried about a T4 by EUS and the aorta really being a T4 by EUS, and I would, at a minimum, start with some type of video, be it robot or thoracoscopic, to really prove that it is or not aortic, because I certainly have seen overcalls by EUS. DR CERFOLIO: We have had false positive repeat EUS after radiation chemo. Initially they are active, but if they are repeat, it is even harder post-op.