Cancer spread to lymph nodes survival rate

Ann Surg. 2004 Apr; 239(4): 483–490.

Abstract

Objective:

To evaluate the outcomes with 2 and 3 lymph node dissection for patients with squamous cell carcinoma of the lower thoracic esophagus at a single institution.

Background:

Extensive lymph node dissection, including the upper mediastinum, for carcinoma of the lower thoracic esophagus is advocated as a standard surgical procedure with curative intent in Japan. However, its efficacy remains controversial.

Methods:

From January 1988 to December 1997, 532 patients with carcinomas of the thoracic esophagus underwent transthoracic esophagectomy and extensive lymph node dissection with curative intent at the National Cancer Center Hospital, Tokyo. Of these, 495 (93%) had squamous cell carcinomas. A total of 156 (29%) with tumors of the lower thoracic esophagus were retrospectively analyzed.

Results:

Of the 156 patients, 55 (35%) underwent 2-field and 101 (65%) underwent 3-field lymph node dissection. The operative morbidity and 30-day and in-hospital mortality rates were 68.0%, 1.3%, and 2.6%, respectively. The overall 5-year survival rate for the entire series was 49.3%. One hundred and seven (69%) had lymph node metastases. Upper and/or middle mediastinal lymph node metastases occurred in 42% of the series. The 5-year survival rate for patients with lymph node metastases in the upper and/or middle mediastinum was 23.3%. Among them, the values after 2- and 3-field lymph node dissection were 5.6% and 30.0%, respectively (P = 0.005). Thirteen (27%) of 48 patients with upper and/or middle mediastinal lymph node metastases treated with 3-field dissection had simultaneous cervical lymph node metastases and their 5-year survival rate was 23.1%.

Conclusion:

The 3-field approach for extensive lymph node dissection provides better survival benefit for patients with squamous cell carcinoma of the lower thoracic esophagus compared to 2-field lymph node dissection when lymph node metastases are present in the upper and/or middle mediastinum.

Carcinomas of the thoracic esophagus throughout the world have remained in dismal prognosis despite improvements of surgical technique, perioperative care, and multi-modality treatment approach. During the past 2 decades, prevalence of carcinomas, especially adenocarcinoma, of the lower thoracic esophagus has increased drastically in the Western countries.1,2 Many are associated with gastroesophageal reflux and Barrett esophagus. In the East, the most frequent location of esophageal carcinomas is the middle thoracic esophagus and histologic type is mainly squamous cell carcinoma that originates from esophageal squamous epithelium; most are associated with alcohol and tobacco abuse.3–5

Differences of the tumor characteristics between the Western and Eastern countries cause various attitudes in the surgical approach to esophageal carcinomas. The majority of Western surgeons have advocated limited surgical resections such as transhiatal esophagectomy.6 Because they consider esophageal carcinomas to have poor prognosis or being already systemic when lymph node metastases exist, the primary goal of surgical intervention is palliative, with low operative morbidity and mortality rates. Furthermore, controversy has persisted about the extent of resection. Transhiatal resection only performs sampling the lower mediastinal or celiac axis nodes.7 Esophagectomy with extensive lymphadenectomy such as en bloc resection does not remove the upper mediastinal lymph nodes as a standard practice.8

In the East, especially in Japan, extensive lymph node dissection, including not only the abdominal and lower mediastinal but also the upper, middle mediastinal and occasionally cervical lymph nodes, has been advocated as a standard surgical procedure with curative intent, because systematic dissection of metastatic nodes may improve survival and lead to potential cure.9 We perform a right transthoracic esophagectomy with extensive lymph node dissection for all surgical candidates with carcinomas of the thoracic esophagus regardless of tumor location.

The aim of the present study was to evaluate our results with 2-field and 3-field lymph node dissection for patients with squamous cell carcinoma of the lower thoracic esophagus.

MATERIALS AND METHODS

Patients

From January 1988 to December 1997, 532 patients with carcinomas of the thoracic esophagus underwent esophagectomy with extensive lymph node dissection via right thoracotomy as a standard surgical procedure at the National Cancer Center Hospital, Tokyo, Japan. The year 1988 was chosen as the beginning of this study because the UICC-TNM staging system revised the T category from length to depth of the primary tumor in 1987.

Four hundred ninety-five patients (93%) had squamous cell carcinomas, and 156 (29% of entire series) had tumors of the lower thoracic esophagus. The records of all of these cases (138 male and 18 female) were analyzed. Ages ranged from 42 to 86 years, with a mean of 62.1 and a median of 62 years. Preoperative evaluation was performed for all patients with a barium swallow examination, endoscopy with biopsy, computed tomography scans from the neck to the abdomen, ultrasonography of the neck and the upper abdominal compartment, and endoscopic ultrasonography. Distant organ metastasis, except in the cervical or celiac nodes, was not evident in any of the patients on preoperative evaluation. Preoperative and postoperative staging was based on the 1997 UICC-TNM classification.10 Metastasis in the cervical or celiac nodes was classified into M1 disease according to the TNM classification. Among the 156 patients, 4 received preoperative chemotherapy in a clinical trial because of presence of intramural metastases.11

In our institute, 3-field lymph node dissection has been carried out for patients with carcinomas of the thoracic esophagus as a standard surgical procedure by a group of surgeons; another group performed 2-field approach as a standard resection except carcinomas of the upper thoracic esophagus until March 2000. Patients who visited our outpatient service on Monday or Thursday were treated by the group of proponents of the 2-field approach while those who presented on Tuesday, Wednesday, or Friday underwent the 3-field dissection. However, surgeons of the 2-field group performed cervical lymphadenectomy when patients were diagnosed or clinically positive for cervical nodal metastasis.

Thirteen patients received postoperative adjuvant chemotherapy in another clinical trial, and postoperative radiation therapy was performed for 8 patients because of residual tumors.

Surgical Procedure

All patients underwent right transthoracic esophagectomy with extensive lymphadenectomy,9 with either the 2-field or 3-field approach. Our 2-field lymph node dissection included total mediastinal, perigastric, and celiac lymphadenectomy. Thee-field lymph node dissection adds removal of lymph node in the supraclavicular and cervical paratracheal regions to 2-field approach. Gastrointestinal continuity was restored with a stomach in 145 patients: 127 through a retrosternal route, 10 through a posterior mediastinal route, and 8 through a subcutaneous route. Colon interposition was performed for the remaining 11 patients because of previous gastric surgery for peptic ulcers in 7 and gastric cancer in 1, and simultaneous total gastrectomy for gastric cancers in 3 patients, 5 through a retrosternal, 1 through a posterior mediastinal, and 5 through a subcutaneous route. Anastomoses of 153 patients were performed at the neck, and 3 patients underwent anastomosis in the right thoracic cavity.

All patients were extubated in the operating room after surgery, and returned to the intensive care unit for 4 days on average. Analgesia with morphine was provided through an epidural catheter for the first 5 postoperative days, and postoperative bronchoscopic lavage was also performed for a few days.

Pathologic Assessment of the Resected Specimens

Pathologic evaluation was performed to identify the depth of invasion of primary lesions and to assess additional lesions in the resected specimens. The entire resected esophagus was examined, with 5-μm sections stained with hematoxylin and eosin for microscopic examination. All removed lymph nodes, identified according to the anatomic location, were formalin fixed and processed to provide 2.5-μm sections for staining with hematoxylin and eosin.

Follow-up

All data were entered prospectively into a database, and all surviving patients were followed for at least 3 years after surgery. The median follow-up period of all patients was 45 months (range, 0.4–151), that for the 66 survivors being 83 months (range, 37—151). Survival time was measured as the period from the date of surgery until death or until the most recent follow-up investigation, with none lost to follow-up. Information about the cause of death was available for all patients.

Statistical Analysis

Survival curves were calculated according to the Kaplan-Meier method, including all causes of death, and log-rank statistics were used for comparisons. The χ2 test was employed for comparisons of proportions. All probabilities were 2-tailed, with a P value less than 0.05 regarded as statistically significant. The statistical calculations were conducted with SPSS 10.0J (SPSS Inc, Chicago, IL) and Stat View 5.0J (Abacus Concepts Inc, Berkeley, CA).

RESULTS

Preoperative Characteristics According to Lymph Node Dissection

Of the 156 patients, 55 (35% of this series) underwent 2-field and 101 (65%) 3-field lymph node dissection. Relationships between preoperative characteristics and lymph node dissection are listed in Table 1. Six patients were diagnosed as having T4 tumors. One patient of the 2-field group demonstrated direct invasion of the primary tumor to the liver. Among 5 patients with T4 tumors in the 3-field group, direct invasion to the lung was diagnosed in 4 and to the aorta in 1.

TABLE 1. Preoperative Characteristics According to Lymph Node Dissection Approach

Operative Outcomes

The mean ± SD for duration of surgery was 456 ± 87 minutes in the 2-field group and 487 ± 84 minutes in the 3-field group. Operative blood loss was 530 ± 247 mL in the 2-field group and 540 ± 356 mL in the 3-field group.

Postoperative complications are listed in Table 2. Fifty patients of this series had an uncomplicated postoperative course. Thus, the operative morbidity was 68.0%, with a 2.6% (4 patients) in-hospital mortality rate. Two patients (1.3%) died of postoperative complications within 30 days of surgery. The incidences of postoperative complications did not differ between the groups undergoing 2-field and 3-field lymph node dissection.

TABLE 2. Postoperative Complications

The overall 5-year survival rate for the entire series was 49.3%. Survival curves of patients after 2-field and 3-field lymph node dissection are shown in Figure 1, the 5-year survival rates being 45.0% and 51.7%, respectively (P = 0.406).

FIGURE 1. Survival curves of patients with squamous cell carcinomas of the lower thoracic esophagus after 2-field or 3-field lymph node dissection.

Status of Lymph Node Metastases and Survival Rates

Survival rates according to the status of lymph node metastases are summarized in Table 3. All patients with lower mediastinal lymph node metastases had simultaneous perigastric nodal involvement. There were 66 patients with lymph node metastases in the upper and/or middle mediastinum, including 3 patients with simultaneous abdominal paraaortic nodal involvement. The 5-year survival rate for these 16 patients, who had lymph node metastases in the upper and/or middle mediastinum but not the other regions, was 30.0%. The value for the remaining 50 patients, who had simultaneous lymph node metastases in the abdomen, was 21.1%. One patient among this series, who had cervical nodal involvement alone without having any other lymph node metastasis, is still alive 4 years after surgery without recurrence of disease.

TABLE 3. Status of Lymph Node Metastases and Survival Rates of Patients With Squamous Cell Carcinoma of the Lower Thoracic Esophagus Treated With Extensive Lymph Node Dissection

Pathologic Characteristics

Pathologic characteristics according to lymph node dissection are summarized in Table 4. Of 4 patients with pathologic T4 tumors, 2 were diagnosed as clinical T3 tumors. The primary tumor directly invaded into the aorta, liver, and lung in 1 patient each in the 2-field dissection group. Those with liver and lung involvement were resected completely with co-resection of the invaded organ. A patient treated with 3-field dissection, in whom the primary tumor had directly invaded into the lung, pericardium, and left main bronchus simultaneously, underwent co-resection of both lung and pericardium. However, the tumor was left grossly in the left main bronchus.

TABLE 4. Pathological Characteristics According to Lymph Node Dissection

Of 3 patients with M1b disease in the 2-field group, 1 had left paratracheal lymph node metastasis that was diagnosed positive preoperatively and resected at the time of anastomosis in the neck; the remaining 2 patients had paraaortic nodal metastases in the abdomen. Of 15 with M1b disease in the 3-field group, 14 had nodal metastases in the neck and 1 of the paraaorta in the abdomen. Three with nodal metastases of the paraaorta died of recurrent disease at 8, 21, and 31 months after surgery.

Complete resection (R0 resection) of the original tumor was accomplished in 95% of patients treated with both 2- or 3-field dissection. Of 2 with microscopically residual tumors (R1 resection) after 2-field dissection, in 1 the vertical margin of the primary tumor positive and in the other the distal margin was positive because of intramural metastasis. Of 4 with R1 resection in the 3-field group, 3 were positive of the proximal surgical margin of the resected esophagus because of multiple primary lesions and the remaining 1 was distal surgical margin positive because of intramural metastasis.

Pathologic Characteristics and Survival Rates According to Lymph Node Dissection

Pathologic characteristics and survival rates according to lymph node dissection are shown in Table 4. There was a statistically significant difference between patients with upper and/or middle mediastinal lymph node metastases undergoing 2-field and 3-field dissection (P = 0.005) (Fig. 2). Thirteen (27%) of 48 patients with upper and/or middle mediastinal lymph node metastases treated with 3-field dissection had simultaneous cervical lymph node metastases and their 5-year survival rate was 23.1%.

FIGURE 2. Survival curves of patients with lymph node metastases of the upper and/or middle mediastinum treated with 2-field and 3-field lymph node dissection.

The 5-year survival rates for patients with R0 resection after 2-field and 3-field dissection were 47.6% and 52.3%, respectively. Two with microscopically residual tumors (R1 resection) after 2-field dissection died 11 and 38 months after surgery. Of 4 with R1 resection after 3-field dissection, 2 died after 5 and 16 months, and remaining 2 were still alive at 8 and 12 years after surgery. Of 2 with macroscopically residual tumors (R2 resection), the 1 undergoing 2-field died after 8 months, and the other died 11 months after 3-field dissection.

The patients with stage 0 disease treated with 2-field and 3-field dissection are still alive 10 and 11 years after resection, respectively. The 5-year survival rates after 2-field dissection were 70.7% for stage I, 56.3% for stage II, 26.3% for stage III, and 14.3% for stage IV. Those after 3-field dissection were 75.2% for stage I, 71.1% for stage II, 33.4% for stage III, and 27.8% for stage IV.

The distribution of subdivisions of stage IV differed between the 2-field and 3-field groups. Cervical nodal involvement was classified as M1b disease according to the 1997 UICC-TNM staging system. Of 7 patients treated with 2-field dissection, 4 were with stage IVA and 3 were stage IVB (1 M1b-neck and 2 M1b-abdominal paraaorta). Of 21 with stage IV treated with 3-field dissection, 6 were stage IVA and 15 were stage IVB (14 M1b-neck and 1 M1b-abdominal paraaorta).

DISCUSSION

The present study of extensive lymph node dissection for patients with squamous cell carcinoma of the lower thoracic esophagus demonstrated a high frequency of lymph node metastases in the upper and/or middle mediastinum. Furthermore, 3-field dissection provided better survival benefit for patients with upper and/or middle mediastinal lymph node metastases than 2-field dissection.

In Japan, the most common histologic type of carcinoma of the thoracic esophagus is the squamous cell carcinoma, accounting for over 90% of the total cases. The lower thoracic esophagus was the site for 30% of all patients treated with extensive lymph node dissection in the period of this study. In the Western world, a drastic increase in adenocarcinoma of the lower thoracic esophagus and esophagogastric junction has been reported during the last 2 decades. Orringer and his associates reported that adenocarcinomas of the lower thoracic esophagus or esophagogastric junction accounted for 73% and 69% of the total tumors seen from 1976 to 1998.12 Clearly, the tumor location and histologic type of carcinomas of the thoracic esophagus differ between the Western world and Japan.

The lymphatic drainage system of the esophagus, which is well developed in the submucosal layer and forms an intensive longitudinal extension, causes a unique pattern of lymph node metastasis.13 Lymphoscintigrams of the esophagus reveal uptake in the cervical, upper mediastinal, and perigastric nodes.14 Akiyama reported that the most frequent sites of lymph node metastases were the perigastric nodes in patients with squamous cell carcinoma of the lower thoracic esophagus, and the upper mediastinal nodes in patients with squamous cell carcinoma of the upper thoracic esophagus.3 Metastases in both the upper mediastinal and perigastric lymph nodes occurred similarly with high frequencies in patients with squamous cell carcinoma of the middle thoracic esophagus, spread thus being in both upward and downward directions.

Differences in tumor location and sites of lymph node metastases between the Western world and Japan have caused different surgical approaches for tumors of the lower thoracic esophagus. In Japan, transthoracic esophagectomy with extensive lymph node dissection has been carried out as a standard surgical procedure with curative intent. In the Western world, with the recent increase of adenocarcinoma of the lower thoracic esophagus and esophagogastric junction, the surgical approach has changed. The majority of Western surgeons have more favored a transhiatal approach without thoracotomy rather than transthoracic esophagectomy, because it is controversial whether transthoracic esophagectomy with extensive lymph node dissection carries a survival benefit.12,15 The lack of data on the benefit of extensive lymph node dissection for adenocarcinoma of the lower thoracic esophagus and esophagogastric junction discourages extension of lymph node dissection. In our series of squamous cell carcinoma, lymph node metastases in the upper and/or middle mediastinum from the lower thoracic esophageal lesions were present in 42% of patients after extensive lymph node dissection, and a 23.3% 5-year survival rate obtained. Less extensive surgery without removal of these lymph nodes might leave tumors. While the impact of microscopically residual tumor (R1 resection) on survival is controversial, gross residual tumors in lymph nodes may mean a poor prognosis. Despite the lack of prospective randomized controlled trials comparing the different degrees of lymphadenectomy, the survival rates of our series are substantially superior to those obtained with less extensive lymph node removal.8,12

Even in Japan, adding a dissection of cervical nodes as a standard surgical procedure is still controversial. Proponents of 2-field lymph node dissection for carcinomas below the tracheal bifurcation, especially for carcinomas of the lower thoracic esophagus or esophagogastric junction, consider that cervical lymphadenectomy is unnecessary. They argue that since the frequency of lymph node metastases to the cervical regions is quite low, the removal of cervical nodes might not improve survival looking at all patients with tumors below the tracheal bifurcation.16 There was no statistically significant difference between the overall survival rates after 2-field and 3-field lymph node dissection in the present series. However, it is noteworthy that 3-field dissection gave better survival for those with upper and/or middle mediastinal lymph node metastases, the 27% affected with simultaneous cervical nodal involvement having a 5-year survival rate of 23%. Upper mediastinal lymph nodes, especially recurrent laryngeal nerve nodes, have continuity to cervical paratracheal nodes.17,18 Furthermore, metastases to these nodes are associated with deep cervical involvement.19 Because the 3-field approach offers thorough removal of lymph nodes not only in the neck but also in the cervicothoracic region, this procedure can provide better survival of patients with upper and/or middle mediastinal lymph node metastases, as compared with 2-field dissection.

Transthoracic esophagectomy with extensive lymph node dissection has remained a highly morbidity procedure, despite improvement of surgical skill, and high postoperative complication rates have been reported.20,21 Vocal cord palsy, caused by thorough lymph node dissection along the recurrent laryngeal nerves, occurred in 12% of our series. However, the incidence of this postoperative complication after 3-field dissection was lower than with the 2-field approach. The meticulous dissection of lymph nodes along the recurrent laryngeal nerves via both transthoracic and cervical approaches could rather avoid injury. Furthermore, the 1.3% 30-day and 2.6% in-hospital mortality rates seem to be acceptable, contrasting with the 68% morbidity rate as compared with less extensive resection.22,23 Nevertheless, further efforts in reducing postoperative morbidity following extensive lymph node dissection must be continued because patients without nodal involvement can suffer complications from lymphadenectomy.

In conclusion, we advocate 3-field lymph node dissection rather than the 2-field approach for patients with carcinomas of the lower thoracic esophagus for survival benefit without increasing postoperative complications. Since the size of our series was limited by the fact of the single institution nature of the study, a multi-institutional prospective randomized controlled trial should now be performed for confirmation the results of this study.

​

TABLE 5. Pathological Characteristics and 5-Year Survival Rates According to Lymph Node Dissection

Footnotes

Reprints: Hiroyasu Igaki, MD, Esophageal Surgery Division, Department of Surgery, National Cancer Center Hospital,5-1-1, Tsukiji, Chuo-ku, Tokyo 104–0045, Japan. E-mail: pj.og.ccn@ikagiih.

REFERENCES

1. Pera M, Cameron AJ, Trastek VF, et al. Increasing incidence of adenocarcinoma of the esophagus and esophagogastric junction. Gastroenterology. 1993;104:510–513. [PubMed] [Google Scholar]

2. Hesketh PJ, Clapp RW, Doos WG, et al. The increasing frequency of adenocarcinoma of the esophagus. Cancer. 1994;58:1574–1579. [Google Scholar]

3. Akiyama H, Tsurumaru M, Udagawa H, et al. Radical lymph node dissection for cancer of the thoracic esophagus. Ann Surg. 1994;220:364–373. [PMC free article] [PubMed] [Google Scholar]

4. Morita M, Kuwano H, Ohno S, et al. Multiple occurrence of carcinoma in the upper aerodigestive tract associated with esophageal cancer: reference to smoking, drinking and family history. Int J Cancer. 1994;58:207–210. [PubMed] [Google Scholar]

5. Yokoyama A, Muramatsu T, Omori T, et al. Alcohol and aldehyde dehydrogenase gene polymorphisms and oropharyngolaryngeal, esophageal and stomach cancers in Japanese alcoholics. Carcinogenesis. 2001;22:433–439. [PubMed] [Google Scholar]

6. Orringer MB, Sloan H. Esophagectomy without thoracotomy. J Thorac Cardiovasc Surg. 1978;76:643–654. [PubMed] [Google Scholar]

7. Orringer MB, Marshall B, Iannettoni MD. Transhiatal esophagectomy for treatment of benign and malignant esophageal disease. World J Surg. 2001;25:196–203. [PubMed] [Google Scholar]

8. Skinner DB. En bloc resection for neoplasms of the esophagus and cardia. J Thorac Cardiovasc Surg. 1983;85:59–69. [PubMed] [Google Scholar]

9. Kato H, Tachimori Y, Watanabe H, et al. Lymph node metastasis in thoracic esophageal carcinoma. J Surg Oncol. 1991;48:106–111. [PubMed] [Google Scholar]

10. International Union Against Cancer. TNM Classification of Malignant Tumors, 5th ed. New York: Wiley-Liss, Inc., 1997. [Google Scholar]

11. Hokamura N, Kato H, Tachimori Y, et al. Preoperative chemotherapy for esophageal carcinoma with intramural metastasis. J Surg Oncol. 2000;75:117–121. [PubMed] [Google Scholar]

12. Orringer MB, Marshall B, Iannettoni MD. Transhiatal esophagectomy: clinical experience and refinements. Ann Surg. 1999;230:392–403. [PMC free article] [PubMed] [Google Scholar]

13. Liebermann-Meffert D, Duranceau A: Anatomy of the esophagus. In Zuidema GD, Orringer MB, eds. Shackelford's Surgery of the Alimentary Tract, 4th ed. Philadelphia: W. B. Saunders, 1996:3–21. [Google Scholar]

14. Terui S, Kato H, Hirashima T, et al. An evaluation of mediastinal lymphoscintigraphy for carcinoma of the esophagus studied with 99mTc rhenium sulfur colloid. Eur J Nucl Med. 1982;7:99–101. [PubMed] [Google Scholar]

15. Millikan KW, Silverstein J, Hart V, et al. A 15-year review of esophagectomy for carcinoma of the esophagus and cardia. Arch Surg. 1995;130:617–624. [PubMed] [Google Scholar]

16. Ide H, Endo M, Hanyu H, et al. Lymph node metastases of the thoracic esophageal cancer. Shujutu [Operation] 1974;18:1355–1364 [Japanese].

17. Haagensen, CD. The Lymphatics in Cancer. Philadelphia: W.B. Saunders Co, 1972. [Google Scholar]

18. Matsubara T, Ueda M, Yanagida O, et al. How extensive should lymph node dissection be for cancer of the thoracic esophagus? J Thorac Cardiovasc Surg. 1994;107:1073–1078. [PubMed] [Google Scholar]

19. Tabira Y, Yasunaga M, Tanaka M, et al. Recurrent nerve nodal involvement is associated with cervical nodal metastasis in thoracic esophageal carcinoma. J Am Coll Surg. 2000;191:232–237. [PubMed] [Google Scholar]

20. Fujita H, Kakegawa T, Yamana H, et al. Mortality and morbidity rates, postoperative course, quality of life, and prognosis after extended radical lymphadenectomy for esophageal cancer: comparison of three-field lymphadenectomy with two-field lymphadenectomy. Ann Surg. 1995;222:654–662. [PMC free article] [PubMed] [Google Scholar]

21. Bhansali MS, Fujita H, Kakegawa T, et al. Pattern of recurrence after extended radical esophagectomy with three-field lymph node dissection for squamous cell carcinoma in the thoracic esophagus. World J Surg. 1997;21:275–281. [PubMed] [Google Scholar]

22. Stark SP, Romberg MS, Pierce GE, et al. Transhiatal versus transthoracic esophagectomy for adenocarcinoma of the distal esophagus and cardia. Am J Surg. 1996;172:478–482. [PubMed] [Google Scholar]

23. Karl RC, Schreiber R, Boulware D, et al. Factors affecting morbidity, mortality, and survival in patients undergoing Ivor Lewis esophagectomy. Ann Surg. 2000;231:635–643. [PMC free article] [PubMed] [Google Scholar]

Articles from Annals of Surgery are provided here courtesy of Lippincott, Williams, and Wilkins