|
|
||||||||
Ann Thorac Surg 2002;73:1035-1040
© 2002 The Society of Thoracic Surgeons
a First Division of General Surgery, University of Verona, Verona, Italy
b Division of Medical Oncology, University of Verona, Verona, Italy
Accepted for publication November 19, 2001.
* Address reprint requests to Dr de Manzoni, 1a Chirurgia Clinicizzata, Ospedale di Borgo Trento, Piazzale Stefani 1, 37126 Verona, Italia, Italy
e-mail: chirurgia.urgenza{at}univr.it
| Abstract |
|---|
|
|
|---|
Methods. The results of 96 patients who underwent resection with curative intent for gastric cardia cancer at the First Division of General Surgery, University of Verona, from January 1988 to February 2000, were analyzed statistically with special reference to Siewert type.
Results. Despite a high number of curative resections (85.4%), the 5-year survival rate was poor (24%) for all Siewert types (p = 0.8), and for early tumors (51%) also. Chance of cure was limited to pN0 and pN1 patients. Multivariate analysis showed that microscopic or macroscopic residual tumor and pN-positive categories had a significantly higher risk of death (risk ratio, 2.18 and 2.68, respectively) and the pN2 and pN3 category had the most negative prognostic factor (risk ratio, 7.6).
Conclusions. The long-term prognosis for gastric cardia cancer remains poor and is independent of Siewert type, with cure limited to pN0 and pN1 patients.
| Introduction |
|---|
|
|
|---|
In this study, cancer of the cardia was defined according to the classification of Siewert and associates [5], recently approved during the 2nd International Gastric Cancer Congress in Munich in April 1997. This classification is based on the topographic characteristics of adenocarcinomas involving the gastroesophageal junction and is divided into three types. Type I is a tumor with the center lying 1 to 5 cm above the anatomic gastroesophageal junction; type II is a tumor with the center lying 1 cm above to 2 cm below the anatomic gastroesophageal junction; and type III is a tumor with the center lying 2 to 5 cm below the anatomic gastroesophageal junction.
At present, only a few reports have studied surgical strategy and prognosis according to the Siewert classification [68].
The aim of this study was to analyze a 12-year experience of a single institution in the management of adenocarcinoma of the cardia to assess whether lymph node involvement, T status, and Siewert classification were identifiable as prognostic factors affecting survival and long-term outcome.
| Material and methods |
|---|
|
|
|---|
From 1994 patients were prospectively classified according to the classification of Siewert and colleagues [5]. Previously, patients were classified according to the classification of the Japanese Society on Esophageal Disease [9]. As the collected data included overall length of extension, and length of invasion of stomach and esophagus, it was therefore possible to accurately convert these cases into the Siewert classification.
The choice of surgical approach was based on the location of the tumor center, as described by Siewert and Stein [10], and on the length of invasion of the esophagus, with the principal aim of achieving complete surgical resection. No patients underwent neoadjuvant or adjuvant radiochemotherapy
The standard procedure for type I tumors was a right thoracotomy with a subtotal esophagectomy with proximal gastric resection. The procedure of choice for type III tumors was a total gastrectomy with transhiatal resection of the distal esophagus. A transthoracic approach was necessary in two cases to achieve a clear oral margin. Type II tumors were treated by subtotal esophagectomy, as in type I cancers, or by total gastrectomy with transhiatal or transthoracic resection of the distal esophagus, as in type III tumors; the type of resection was chosen in order to obtain clear proximal and distal margins. The mode of reconstruction after subtotal esophagectomy was a gastric tube, whereas a jejunal Roux-en-Y reconstruction was performed after total gastrectomy.
Lymph node dissection was classified according to the Japanese Gastric Cancer Association rules: D1 lymphadenectomy (resection of perigastric nodes in position 1 to 4), D2 lymphadenectomy (resection of nodes in position 1 to 11), and D3 lymphadenectomy (resection extended to the nodes in position 12 to 16) [11]. The preferred lymph nodal dissection was a standard mediastinal and D2 abdominal for type I tumors and D2 or D3 abdominal lymphadenectomy comprising nodes of the lower posterior mediastinum for type II and type III tumors. D1 dissection was chosen in the cases of palliative resection or high-risk patients. Lymph nodes around the root of the mesenteric vein (station 14) and along the middle colic artery (station 15) were not routinely dissected. Lymphadenectomy at the splenic hilum was limited to patients with enlarged nodes or tumors reaching the gastric fundus, and it was achieved by pancreas-preserving splenectomy [12]. Perigastric lymph nodes were dissected from the excised specimens by the surgeons immediately after resection, assigned to the appropriate station according to the Japanese Gastric Cancer Association classification, and sent for histologic examination. To avoid errors caused by difficulties in assigning the lymph nodes to the correct lymph node stations after en bloc resection, the N2 and N3 nodes were subdivided by the surgeon himself during lymphadenectomy as described in a previous study [13].
Tumors were staged according to the 1997 pathologic classification (pTNM) of the International Union Against Cancer [14]. The histologic classification followed the criteria of Lauren [15].
After discharge from the hospital all patients were personally followed up at 4 months and subsequently at 6-month intervals. Median follow-up was 38 months (range, 4 to 144 months), and no patients were lost to the follow-up procedure.
In 82 cases there was pathologic confirmation of curative resection (classified as R0), whereas residual tumor was present in 14 cases, either microscopic (R1) or macroscopic (R2). The 14 cases with R1 or R2 resection were excluded from the survival analysis. All survival analyses were calculated, including operative deaths. Survival curves were estimated with the Kaplan-Meier method and compared by the log-rank test; deaths from causes other than gastric cardia cancer were censored. Multivariate analysis was performed by Cox regression model. Variables were chosen according to the current literature. In particular, two demographic variables (sex and age), three well-known prognostic factors (residual tumor, depth of tumor invasion, and nodal involvement), and a novel prognostic factor, whose relevance has not been elucidated yet, were introduced in the model. Tumor stage was not entered as a variable because it is highly collinear with pT and pN classes. The
2 test was used to compare the frequencies of different ordinal data.
| Results |
|---|
|
|
|---|
|
A curative resection (R0) was achieved in 82 cases (85.4%), whereas 14 cases showed residual tumor (microscopic [R1] in 10 cases and macroscopic [R2] in 4 cases). Median survival for the R1 and R2 patients was 7 months and for R0 was 19 months. The 5-year survival rate for patients with curative resection (R0) was 24%, whereas no patients with residual tumor (R1 and R2) survived beyond 29 months (p = 0.0001).
Among the 96 patients, 21 had type I tumor, 34 type II, and 41 type III. Tumor characteristics are summarized in Table 2, and survival curves are shown in Figure 1 (p = 0.8) according to the Siewert types.
|
|
|
|
At univariate analysis the pN category was shown to be a significant prognostic factor. The 5-year survival rate for pN0 tumors was 50% compared with 13% for tumors with one or more metastatic nodes (p = 0.0001). Further stratification showed median survivals of 58, 26, 14, and 6 months for pN0, pN1, pN2, and pN3, respectively (p = 0.0001). A 5-year plateau of 24% was reached only by pN1 tumors (Fig 3). The remaining pN categories had a similar poor outcome.
|
|
| Comment |
|---|
|
|
|---|
In the literature long-term survival rates in curatively resected patients ranged between 22% and 38% [6, 1820]. Consistent with these figures, the 5-year survival rate of the R0 patients in this series was 24%. Moreover, this subset represented 85% of patients who were operated on, a percentage among the highest reported [6, 8, 1820]. On the other hand, with univariate and multivariate analysis, R1 and R2 patients showed both a very poor median survival and a significantly higher risk of death.
Prognosis of the adenocarcinoma of the cardia is severe, because lymph node involvement and advanced depth of invasion of the visceral wall are mostly present at onset [7, 2123].
Consistent with this, 81% of our patients had nodal involvement; half of them had seven or more positive lymph nodes (pN2 and pN3). Overall, the subgroup with positive lymph nodes had a 5-year survival rate of only 13% compared with 50% of the pN0 subset. Multivariate analysis confirmed that pN-positive patients, considered as a whole, had a risk of death nearly three times higher than pN0 patients.
Stratification among pN-positive patients showed no long-term survivors in the pN2 and pN3 subset, whereas a 5-year plateau of 24% was reached only by the pN1 subset. This long-term survival percentage appears similar, if not somewhat better, when compared with the 16% to 18% reported by other authors [7, 19]. Multivariate analysis confirmed that pN2 or pN3 status was the most negative prognostic factor (Table 4).
If the 34% reported by Kodera and coworkers [24] is excluded, the percentage of early cardia tumor has been reported in the range between 10% and 18% [710, 25]; accordingly, in this study the percentage of early tumors was 14.6%. It is of note that even in the pT1 subset, 43% of our patients had lymph node involvement, compared with the 17% to 25% reported by other authors [17, 26, 27]. Data regarding survival in the pT1 subgroup is scarce and somewhat inconsistent; 5-year survival rates of 36% [28], 51% [19], and 85% [7] have been reported in the major series. These results are therefore comparable with our experience (51%).
The pT2 and pT3 subsets had similar outcome, both in terms of median survival and long-term survival rate. This result confirms previous observations [29] suggesting that given the lack of the serosa in the cardia region, tumors with transmural growth and invasion of perigastric fat (pT2) should be grouped with pT3.
Moreover, although radical resection was possible in 6 of the 9 pT4 patients, median survival was very poor, suggesting that the actual role of resection alone in this group, if any, remains controversial.
On the whole, with multivariate analysis, the pT category appears to be less important than pN category; also with univariate analysis, a survival longer than 3 years was restricted only to pN0 and pN1 subgroups, irrespective of the pT status.
Interestingly, at approximately 5 years, we observed some late relapses; this finding stresses that a follow-up longer than 5 years is needed before a firm assessment can be made about the long-term cure rate.
In summary, the Siewert classification proved to be a useful guideline in planning surgical treatment, a high curative resection rate was achieved, and the overall prognosis remains poor and chances of cure are limited to pN0 and pN1 patients. To assess whether results can be improved, a phase II study based on preoperative concomitant chemoradiation has been started at our institution.
| References |
|---|
|
|
|---|
Related Article
This article has been cited by other articles:
![]() |
C. Pedrazzani, G. de Manzoni, D. Marrelli, S. Giacopuzzi, G. Corso, A. M. Minicozzi, B. Rampone, and F. Roviello Lymph node involvement in advanced gastroesophageal junction adenocarcinoma J. Thorac. Cardiovasc. Surg., August 1, 2007; 134(2): 378 - 385. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Pedrazzani, G. de Manzoni, D. Marrelli, and F. Roviello It Is Time for a Proper Staging System for Adenocarcinoma of the Gastroesophageal Junction J. Clin. Oncol., March 1, 2007; 25(7): 907 - 908. [Full Text] [PDF] |
||||
![]() |
C. Pedrazzani, Prof. G. deManzoni, D. Marrelli, S. Giacopuzzi, G. Corso, M. Bernini, and Prof. F. Roviello Nodal Staging in Adenocarcinoma of the Gastro-Esophageal Junction. Proposal of a Specific Staging System Ann. Surg. Oncol., February 1, 2007; 14(2): 299 - 305. [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] |
||||
![]() |
K. Ott, W. A. Weber, F. Lordick, K. Becker, R. Busch, K. Herrmann, H. Wieder, U. Fink, M. Schwaiger, and J.-R. Siewert Metabolic Imaging Predicts Response, Survival, and Recurrence in Adenocarcinomas of the Esophagogastric Junction J. Clin. Oncol., October 10, 2006; 24(29): 4692 - 4698. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. R. Shen, S. D. Cassivi, C. Deschamps, M. S. Allen, F. C. Nichols III, W. S. Harmsen, and P. C. Pairolero Surgical treatment of tumors of the proximal stomach with involvement of the distal esophagus: A 26-year experience with Siewert type III tumors J. Thorac. Cardiovasc. Surg., October 1, 2006; 132(4): 755 - 762. [Abstract] [Full Text] [PDF] |
||||
![]() |
F. Pasini, G. de Manzoni, C. Pedrazzani, A. Grandinetti, E. Durante, M. Gabbani, A. Tomezzoli, C. Griso, A. Guglielmi, G. Pelosi, et al. High pathological response rate in locally advanced esophageal cancer after neoadjuvant combined modality therapy: dose finding of a weekly chemotherapy schedule with protracted venous infusion of 5-fluorouracil and dose escalation of cisplatin, docetaxel and concurrent radiotherapy Ann. Onc., July 1, 2005; 16(7): 1133 - 1139. [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 |