Ann Thorac Surg 2009;88:1647-1653. doi:10.1016/j.athoracsur.2009.05.081
© 2009 The Society of Thoracic Surgeons
Original Articles: General Thoracic
Colon Interposition After Esophagectomy With Extended Lymphadenectomy for Esophageal Cancer
Shinji Mine, MDa,*,
Harushi Udagawa, MD, PhDa,
Kenji Tsutsumi, MD, PhDa,b,
Yoshihiro Kinoshita, MD, PhDa,
Masaki Ueno, MD, PhDa,
Kazuhisa Ehara, MDa,
Syusuke Haruta, MDa
a Department of Gastroenterological Surgery, Toranomon Hospital, Tokyo, Japan
b Department of Surgery, Sayama Hospital, Saitama, Japan
Accepted for publication May 27, 2009.
* Address correspondence to Dr Mine, Department of Gastroenterological Surgery, Toranomon Hospital, 2-2-2, Toranomon, Minato-ku, Tokyo, 105-8470, Japan (Email: mineshin{at}rc4.so-net.ne.jp).
 |
Abstract
|
|---|
Background: The purpose of this retrospective study was to investigate the feasibility of colon interposition procedures after esophagectomy with extended lymphadenectomy.
Methods: Between 1990 and 2008, 95 consecutive patients underwent colon interposition after esophagectomy with extended lymphadenectomy for esophageal cancer in our Institution. We reviewed clinical data and long-term survival, and also investigated the association between anastomotic leakage and clinicopathologic findings.
Results: We applied three-field lymphadenectomy to 71 patients and two-field to 24 patients, by a right thoracotomy. Ninety-two patients underwent reconstruction by a retrosternal route, and a posterior mediastinal route was applied to only three patients. We performed hand-sewn anastomosis in the neck in all cases. Three patients required microvascular surgery. Sixty-one patients (64%) experienced postoperative morbidity, most commonly pulmonary complications. Anastomotic leakage occurred in 12 patients (13%). No colon conduit necrosis was detected. Overall mortality, including hospital mortality, was 5.3%. Dysphagia (39%) and diarrhea (38%) were common and stricture was low (6%) after discharge. The overall 5-year survival rate was 43%. During the latter period (1998 to 2008), when ileocolon grafts evolved as the primary choice for interposition, the rate of leakage decreased from 17% (1990 to 1997) to 5.4%. No mortality was recorded during the latter period.
Conclusions: Results from this study demonstrate that colon interposition after esophagectomy with extended lymphadenectomy is feasible and can have a favorable outcome.
|
Introduction
|
|---|
In cases of esophageal cancer requiring reconstruction after esophagectomy, the stomach is the first choice as an esophageal substitute because of its facility, plasticity, and rich submucosal vascular network [1]. In cases with a history of gastrectomy, concurrent gastric disease, or cancer involvement of the stomach, the colon or jejunum is used instead of the stomach as an esophageal substitute. As a conduit, a colon graft is generally preferred over the jejunum because the colon can be brought easily up to the neck and because procedures involving the jejunum are more likely to require microvascular surgery [2].
Although we have reported that extended lymphadenectomy for esophageal cancer results in improved postoperative survival rates [3], esophagectomy with extended lymphadenectomy is still considered a highly invasive gastroenterological surgery with high morbidity and mortality [4]. Whether colon interposition after esophagectomy with extended lymphadenectomy increases the risk of surgery is unknown. Few previous reports have examined the postoperative complications after such surgeries [5].
In this study, we investigated the feasibility of colon interposition procedures after esophagectomy with extended lymphadenectomy by conducting a 19-year retrospective analysis of early morbidity, mortality, postoperative symptoms, and long-term outcome involving cases with colon interposition. In addition, we looked for factors that were associated with the occurrence of anastomotic leakage, because leakage is a severe postoperative complication and causes the deterioration of the quality-of-life in the patients.
|
Patients and Methods
|
|---|
Patients
This study was approved by the Toranomon Hospital Institutional Review Board of Clinical Research. The need for informed consent from patients was waived because of its retrospective design. Between January 1990 and September 2008, 1,146 patients with esophageal cancer underwent esophagectomy at the Department of Gastroenterological Surgery, Toranomon Hospital, Tokyo, Japan. Among these, 119 consecutive patients (10%) had reconstruction using colon interposition. Of these 119 patients, 95 patients (80%) who underwent esophagectomy with extended lymphadenectomy were enrolled in this retrospective study. The remaining 24 patients, who underwent esophageal blunt dissection without thoracotomy or with localized lymphadenectomy, were excluded. Tumor stage was described according to the tumor-nodes-metastasis (TNM) classification of 2002 [6].
Surgical Procedures
For each case, the extent of lymphadenectomy depended on the location and depth of the tumor, the condition of the patient (including performance status), and the results of preoperative examinations for comorbid disease. Esophagectomy was performed generally with three-field lymphadenectomy through a right thoracotomy. However, if the risk for three-field lymphadenectomy was high based on the patient's condition or comorbid disease, or if the tumor was located in the lower esophagus was deemed superficial, we would then perform an esophagectomy with two-field lymphadenectomy through a right thoracotomy. In five recent esophagectomy cases using video-assisted thoracotomy, the same extent of lymphadenectomy was performed as in previous open thoracotomy operations. In cases with a history of distal gastrectomy, remnant gastrectomy for the purpose of lymphadenectomy was also performed. Consequently, all the patients in this study became total gastrectomized after esophagectomy.
Between 1990 and 1997, the type of colon graft (ileocolon, ascending-transverse colon or left colon) was selected based on colon vessel findings during surgery. In more recent reconstructions conducted between 1998 and 2008, ileocolon interposition evolved as the first choice for colon graft, and ascending-transverse colon or left colon interposition was only used when an ileocolon graft was not available due to a deficit of connections between ileocolic and right colic vessels.
For ileocolon interposition, we adequately mobilized the region from the terminal ileum to the ascending colon from the retroperitoneum, and then detected colic vessels using transillumination. Ileocolic vessels and marginal vessels of ileum were then clamped together to assess flow from right colic vessels. Then, we divided just the root of the ileocolic artery and vein after the confirmation of the pulsation at the tip of the graft. In relatively rare cases in which the length of the ileocolon graft was inadequate to reach the neck, right colic vessels were clamped together. In cases where blood flow for the colon graft could not be assessed, the ileocolic vessels and marginal artery were clamped with the terminal ileum for approximately 10 to 20 minutes, after which blood flow could often be detected by palpation of colon graft vessels. In more recent procedures we used an ileocolon graft in which the length of the ileum was approximately 20 cm from the ileocecal valve, which generally corresponded to the region supplied from the ileocolic artery. The length of colon was 15 cm in the ileocolon graft. We preferred the Billroth 2 method in reconstruction (Fig 1). If the depth of the tumor was considered superficial and the patient had no history of gastrectomy, we interposed the ileocolon graft between the esophagus and duodenum. When the patient had a history of gastrectomy in the Billroth 2 or Roux-en-Y method, we used an ileocolon graft in the Roux-en-Y method. When applying a left colon interposition, the left colic artery was used as a feeding artery. When using an ascending-transverse colon graft, the middle colic artery was a feeding artery.
View larger version (32K):
[in this window]
[in a new window]
|
Fig 1. (A) The continuous arrowed line shows the cut end of the mesenterium in preparing ileocolon graft. If the length of graft is not adequate, we cut the right colic vessels at the root (dotted line). (B) The diagram shows the reconstruction form using the ileocolon graft in Billroth 2 type. (RCA = right colic artery; ICA = ileocolic artery.)
|
|
In all cases, the anastomosis to the esophagus was sewn by hand using two layers of interrupted suture in the neck. No stapler device or intrathoracic anastomosis was used. Esophago-ileo anastomosis and esophago-colic anastomosis were performed end-to-side. In most patients, anastomosis was checked on postoperative day 9 using radiography with contrast medium. Patients demonstrating thin leakage by radiography and no clinical symptoms were diagnosed with minor leakage, whereas patients with clinical symptoms of leakage or a definite pool of contrast medium by radiography were diagnosed with major leakage.
We conducted a retrospective review of hospital medical charts, including clinicopathologic features, operative data, operative morbidity, and postoperative symptoms after discharge. Patients were followed either until the time of death or until December 2008. In regard to investigations for postoperative symptoms, we excluded patients who experienced hospital mortality, very early tumor recurrence, or were followed mainly at other outpatient clinics. Finally, data of 80 patients were available for postoperative symptoms. If a patient was recorded to suffer from the symptom at least once in the medical charts, we determined the symptom was positive.
Statistical methods were applied to detect associations between anastomotic leakage and any of the following factors: general characteristics of the patient population, reasons for colon interposition, preoperative comorbid disease (including heart disease, pulmonary disease, diabetes mellitus, or liver dysfunction), with or without neoadjuvant chemotherapy or chemoradiotherapy, operative data, extent of lymphadenectomy, or type of colon graft. Three patients who could not be examined for leakage due to severe postoperative morbidity, including early mortality, were excluded from this analysis. Survival rates were estimated using the Kaplan-Meier method. Associations between leakage and clinical factors were analyzed using 
2 tests. All statistical analyses were performed using StatView, version 5 (SAS institute, Cary, NC). The p values less than 0.05 were considered statistically significant.
|
Results
|
|---|
General characteristics of the patient population are shown in Table 1. The patient group included 94 men and 1 woman, and ranged in age from 50 to 78 years, with an average of 63 years. More than half of the patients had middle thoracic esophageal cancer, and almost all cases showed pathologic findings of squamous cell carcinoma (typical in Japan). Among those receiving colon interposition, 56 patients had a history of gastrectomy for gastric cancer or gastroduodenal ulcer. Of these 56 patients, 49 had undergone distal gastrectomy and 7 had undergone total gastrectomy. Twenty-eight patients had synchronous gastric cancer and 7 patients had synchronous gastric ulcer. Four patients had an advanced esophageal cancer invading the stomach directly, and so the stomach was not available as a substitute (Table 2).
Surgical findings are shown in Table 3. We performed esophagectomy with three-field lymphadenectomy in 71 patients (75%) and with two-field lymphadenectomy in 24 patients (25%). A retrosternal route of reconstruction was chosen for most cases (92 patients, 97%), with only 3 patients undergoing reconstruction by a posterior mediastinal route. No colon interpositions followed a subcutaneous route. For esophageal substitutes, 62 patients received an ileocolon graft, 7 patients received an ascending-transverse colon graft, and 26 received a left colon graft. Three of 95 patients (3.2%) experienced blood flow failure with a colon graft, and microvascular surgery was conducted during the operation to achieve a supercharge in 1 patient and a superdrainage in the other 2 patients.
The incidence of postoperative morbidity is shown in Table 4. Among the 12 patients with anastomotic leakage (13%), 7 had major leakage and 5 had minor leakage. One patient with major leakage contracted postoperative pneumonia that led to multiple organ failure and consequent in-hospital death on the 55th postoperative day. All of the remaining 11 patients with major or minor leakages recovered successfully with conservative treatment, and none required additional surgical procedures. No colon conduit necrosis occurred. Among the 3 patients who underwent microvascular surgery for improvement of blood flow during the colon graft operation, none experienced colon graft necrosis or anastomotic leakage.
Pulmonary complications (including pneumonia, pulmonary edema, and acute respiratory distress syndrome) were common, occurring in 31 patients (33%). In addition, 12 patients (13%) experienced vocal cord paralysis, 6 patients (6.3%) experienced bowel obstruction (3 required surgery), and 1 patient was diagnosed as suffering from tracheal necrosis and ultimately died. Fifteen patients (16%) experienced some surgical site infection, including wound infection or abdominal abscess. Collectively, 61 patients (64%) experienced some form of postoperative morbidity, whereas 34 patients (36%) showed none. The 30-day mortality rate was 2.1%, with 1 patient death after tracheal necrosis, and another due to postoperative pulmonary embolization. Three patient deaths contributed to a hospital mortality rate after 30 days of 3.2%. Eventually, the overall mortality rate was 5.3%. During the latter period (1998 to 2008), when ileocolon grafts evolved as the primary choice for interposition, the rate of anastomotic leakage decreased from 17% (1990 to 1997) to 5.4%. Moreover, no mortality was recorded during the latter period.
Postoperative symptoms are shown in Table 5. Most popular complaints were dysphagia (40%) and diarrhea (39%). The incidence of strictures in anastomosis was low (6%). These strictures improved after a few dilatations by balloon (medium 1 [1–4]), and no patient required surgical procedures for strictures.
We tested the significance of correlations between the incidence of anastomotic leakage and other case factors using 2 square tests (Table 6). Patients during the latter period (1998 to 2008) showed a lower incidence of anastomotic leakage than those in the former period (1990 to 1997), but the relationship was not statistically significant (p = 0.07). Although patients without any comorbid disease showed a decreased tendency for leakage than patients with comorbid disease, this association also was not statistically significant (p = 0.07). Ileocolon graft patients did demonstrate a significantly lower rate of leakage than patients receiving ascending-transverse colon graft or left-colon grafts (p = 0.02). For assessment of long-term outcome, 95 patients were followed for a median time period of 31 months (0 to 157 months). As of December 2008, 68 patients had died and 25 were living. Two other patients who were missing later records survived for at least 42 and 50 months after their operations, respectively. Overall, 3- and 5-year survival rates for all patients in this study were 54% and 43%, respectively (Fig 2).
|
Comment
|
|---|
Since 1911, there have been frequent reports demonstrating the outcome of colon interposition as an esophageal substitute [2, 7–14]. Although patient numbers for each these previous reports ranged from 32 to 347 individuals, most studies included patients with benign esophageal diseases or those having undergone several types of esophagectomy, including transhiatal, Ivor Lewis, or three-field lymphadenectomy. The present study focuses on esophageal cancer patients who underwent esophagectomy with extended lymphadenectomy through a right thoracotomy; in this regard, our study of 95 cases exceeds all others to date.
Although colon interposition as an esophageal substitute is often considered more surgically complex than gastric pull-up and has been reported to demonstrate a comparatively higher rate of morbidity and mortality than gastric pull-up [9], other reports from high-volume centers demonstrate no significant difference in operative morbidity and mortality rates between colon interposition and gastric pull-up [2, 15]. According to these reports, the occurrence of anastomotic leakage as a consequence of colon interposition for esophageal cancer ranged from 0% to 46%, and operative mortality rate ranged from 0% to 17% (Table 7). Similar to previous reports, the incidences of anastomotic leakage and operative mortality in the present study were 13% and 5.3%, respectively. There was a 43% 5-year overall survival rate. Based on these retrospective findings, we consider the application of colon interposition after esophagectomy with extended lymphadenectomy feasible.
Although there was no incidence of colon graft necrosis among our 95 patients, previous studies reported colon graft necrosis in 0 to 9.4% of patients (Table 7). We attribute our present outcome to several factors, including overall attention and care to procedures such as preserving feeding arteries and veins, dividing just the root of the ileocolic artery to preserve vessel communication, confirming pulsation of the feeding artery at the tip of the colon graft, and preventing colon graft injury during the lift to the neck.
In addition, our choice of route in reconstruction was consistent, with a retrosternal route being used in almost all cases (92 patients, 97%) and no reconstructions by a subcutaneous route. Although reconstruction by a subcutaneous route is considered less risky after the occurrence of anastomotic leakage or colon conduit necrosis, a previous report identified the subcutaneous route as a significant risk factor toward graft ischemia [16]. The reason for avoiding a posterior mediastinal route is that we do not desire to locate esophageal substitutes in tumor beds. We have not chosen a subcutaneous route for the deteriorated cosmetic change.
We also used consistent preparation for microvascular procedures, which is indispensable for rare cases of blood-flow failure at the colon graft during surgery [17, 18]. However, because this study included only three cases (3.5%) of microvascular anastomosis for the purpose of supercharge or superdrainage, we consider the routine use of microvascular surgery during colon interposition unnecessary.
From the viewpoint of anastomotic leakage, in this study the statistical analysis results showed that the use of an ileocolon graft was significantly related to a reduced frequency of anastomotic leakage. Other clinical factors, including without comorbid disease, with neoadjuvant treatment, and posterior operation period were likely to be associated with a low frequency of anastomotic leakage, but it was not statistically significant. We are unable to explain why the incidence of leakage was likely to be low in the patients with neoadjuvant chemotherapy or chemoradiotherapy.
Choice of graft site (ileocolon, ascending-transverse colon, or left colon) is generally decided among surgeons or departments. During recent years (1998 to 2008), our facility adopted the ileocolon as the preferred choice for graft interposition for the following reasons. First, a similar diameter of the cervical esophagus and the ileum improves the facility of anastomotic procedures relative to the colon. Second, the ileum is less bulky than the colon, which often can cause cervical wound bulging. Third, we favor the ileocecal valve because it functionally inhibits regurgitation of digestive juices [19]. Finally, the use of ileocolon grafts allows alternative grafts, including ascending-transverse colon or left-colon grafts, to be used should the first graft fail [20]. Procedures in the opposite temporal order are considered to be difficult. After switching to the consistent use of ileocolon grafts (1998 to 2008), we performed 37 colon interpositions and only 2 ascending-transverse colon reconstructions. During this period, none of the 39 patients required microvascular surgery for conduit blood flow failure.
During the latter period (1998 to 2008), the rates of anastomotic leakage and operative mortality decreased to 5.4% and 0%, respectively. This improvement of the outcome may be attributed to several factors. One factor is our choice of ileocolon graft as an esophageal substitute as mentioned above, and the other factors are considered to be the developments of surgical procedures and perioperative managements. One of the major improvements in surgical procedures is the induction of video-assisted thoracoscopic esophagectomy. These procedures are generally considered minimally invasive surgeries. However, only 5 patients underwent video-assisted thoracoscopic esophagectomy in this study, so we could not evaluate this procedure in regard to less-invasiveness.
Using colon interposition as an esophageal substitute, the assessment of quality-of-life for postoperative patients is considered to be important. In the present study, the frequencies of dysphagia and diarrhea were comparatively high; on the other hand, the frequency of stricture were low. The symptoms of dysphagia, diarrhea, reflux, and dumping syndrome made improvements during postoperative courses in our patients. No patient required reconstruction for improvement of postoperative complaint. A previous report showed that colon interposition with preservation of the stomach was superior to gastric pull-up for an esophageal substitute in quality-of-life [19]. However, in this study we did not compare colon interposition with gastric pull-up because each group had a different background in our Institution. All of the patients who underwent colon interposition as an esophageal substitute in this study were accompanied by a total gastrectomy.
In conclusion, results from this retrospective study demonstrate that colon interposition after esophagectomy with extended lymphadenectomy, as conducted routinely by our facility, is feasible and can have a favorable outcome. In particular, we demonstrated that the adoption of a specific procedure for colon interposition (involving ileocolon graft, retrosternal route, cervical hand-sewn anastomosis, and no routine microvascular surgery) in recent years (1998 to 2008) reduced the occurrence rate of anastomotic leakage and conduit necrosis to very low levels. The outcomes of postoperative symptoms in esophagectomy with colon interposition were acceptable. Although many facilities still perform colon interposition during radical esophagectomy only when the stomach is not available as a substitute, results of the present study maintain and extend our application of colon interposition after esophagectomy. Specifically, if the expected prognosis of patients is good due to early-stage esophageal cancer and no comorbid disease, we intend to consider colon interposition reconstruction for the preservation of stomach function and the improvement of quality-of-life. On the other hand, the patients who require esophagectomy and are not able to use the stomach as an esophageal substitute like the cases in this study, probably should undergo the reconstruction with ileocolon interposition by VATS, with the aim of less invasiveness in the near future.
|
References
|
|---|
- Akiyama H, Miyazono H, Tsurumaru M, Hashimoto C, Kawamura T. Use of the stomach as an esophageal substitute Ann Surg 1978;188:606-610.[Medline]
- Davis PA, Law S, Wong J. Colonic interposition after esophagectomy for cancer Arch Surg 2003;138:303-308.[Abstract/Free Full Text]
- Akiyama H, Tsurumaru M, Udagawa H, Kajiyama Y. Radical lymph node dissection for cancer of the thoracic esophagus Ann Surg 1994;220:364-373.[Medline]
- Isono K, Sato H, Nakayama K. Results of a nationwide study on the three-field lymph node dissection of esophageal cancer Oncology 1991;48:411-420.[Medline]
- Doki Y, Okada K, Miyata H, et al. Long-term and short-term evaluation of esophageal reconstruction using the colon or the jejunum in esophageal cancer patients after gastrectomy Dis Esophagus 2008;21:32-38.
- Sobin LH, Witterkind C. TNM classification of malignant tumors6th ed.. NewYork: John Wiley; 2002.
- Kelling G. Oesophagoplastic mitt Hilfe des Querkolon Zentralbl Chir 1911;38:1209-1212.
- Vuilliet H. De l'oesophagoplastic et des diversos medifications Semin Med 1911;31:529-530.
- Cerfolio RJ, Allen MS, Deschamps C, Trastek VF, Pairolero PC. Esophageal replacement by colon interposition Ann Thorac Surg 1995;59:1382-1384.[Abstract/Free Full Text]
- DeMeester TR, Johansson KE, Franze I, et al. Indications, surgical technique, and long-term functional results of colon interposition or bypass Ann Surg 1988;208:460-474.[Medline]
- Fürst H, Hüttl TP, Löhe F, Schildberg FW. German experience with colon interposition grafting as an esophageal substitute Dis Esophagus 2001;14:131-134.[Medline]
- Knezevi
JD, Radovanovi NS, Simi AP, et al. Colon interposition in the treatment of esophageal caustic strictures: 40 years of experience Dis Esophagus 2007;20:530-534.[Medline] - Isolauri J, Markkula H, Autio V. Colon interposition in the treatment of carcinoma of the esophagus and gastric cardia Ann Thorac Surg 1987;43:420-424.[Abstract/Free Full Text]
- Kolh P, Honore P, Degauque C, Gielen JL, Gerard P, Jacquet N. Early stage results after oesophageal resection for malignancy-colon interposition vs. gastric pull-up Eur J Cardiothorac Surg 2000;18:293-300.[Abstract/Free Full Text]
- Hagen JA, DeMeester SR, Peters JH, Chandrasoma P, DeMeester TR. Curative resection of esophageal adenocarcinoma Ann Surg 2001;234:520-531.[Medline]
- Moorehead RJ, Wong J. Gangrene in esophageal substitute after resection and bypass procedures for carcinoma of the esophagus Hepatogastroenterology 1990;37:364-367.[Medline]
- Fujita H, Yamana H, Sueyoshi S, et al. Impact on outcome of additional microvascular anastomois-supercharge-on colon interposition for esophageal replacement: comparative and multivariate analysis World J Surg 1997;21:998-1003.[Medline]
- Shiraka Y, Naomoto Y, Noma K, et al. Colonic interposition and supercharge for esophageal reconstruction Langenbecks Arch Surg 2006;391:19-23.[Medline]
- Metzger J, Degen L, Beglinger C, von Flüe M, Harder F. Clinical outcome and quality of life after gastric and distal esophagus replacement with an ileocolon interposition J Gastrointest Surg 1999;3:383-388.[Medline]
- Popovici Z. A new philosophy in esophageal reconstruction with colon. Thirty-years experience. Dis Esophagus 2003;16:323-327.[Medline]
Related Article
-
Invited Commentary
- Donald E. Low
Ann. Thorac. Surg. 2009 88: 1653-1654.
[Extract]
[Full Text]
[PDF]
This article has been cited by other articles:
|
|
|
|
 
D. E. Low
Invited Commentary
Ann. Thorac. Surg.,
November 1, 2009;
88(5):
1653 - 1654.
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
