REVIEWS
Evidence-Based Surgical Treatment of Esophageal
Cancer: Overview of High-Quality Studies
Sjoerd M. Lagarde, MD, PhD, Bart C. Vrouenraets, MD, PhD,
Laurents P. S. Stassen, MD, PhD, and J. Jan B. van Lanschot, MD, PhD
Department of Surgery, Saint Lucas Andreas Hospital, Amsterdam; Department of Surgery, Reinier de Graaf Group, Delft; and
Department of Surgery, Erasmus Medical Center, Rotterdam, The Netherlands
(Ann Thorac Surg 2010;89:1319 –26)
© 2010 by The Society of Thoracic Surgeons
O
nly few adequately powered randomized controlled
trials (RCTs) have been performed in esophageal
cancer (EC) surgery. In a systematic review of other
surgical procedures, less than 5% of reported studies had
a prospective design [1–3]. Possibly, surgeons may not
consider RCTs as a feasible strategy to resolve specific
surgical questions. Indeed, standardization of surgical
techniques, defining end points and study designs that
are acceptable to both surgeons and patients may be
problematic. Another problem is the number of resections performed. Although high-volume centers achieve
better results [4, 5], only few hospitals perform esophageal surgery with sufficient frequency to exclude technical factors and overcome a learning curve in performing
novel operations or techniques. Moreover, compared
with pharmaceutical trials, fundraising for surgical trials
is difficult. Finally, a commonly held view is that the
timing is never right for RCTs. On the one hand, surgeons are reluctant to randomize while a procedure is
being developed, and on the other hand they are reluctant after a steady state has been reached because they
are already using the procedure on consecutive patients
and are convinced of its value [6]. The present review
gives an overview on evidence-based studies that focus
on surgical topics related to the treatment of EC.
Material and Methods
Systematic reviews and RCTs are considered the highest
level of evidence [1–3]. For the purpose of this overview,
the literature search was restricted to this highest level of
evidence related to the surgical treatment of EC. NonsurAddress correspondence to Dr Lagarde, Department of Surgery, Saint
Lucas Andreas Hospital, Jan Tooropstraat 164, Amsterdam, 1061 AE, The
Netherlands; e-mail: [email protected].
© 2010 by The Society of Thoracic Surgeons
Published by Elsevier Inc
gical issues, such as (neo)-adjuvant chemo(radio)therapy,
preoperative staging, and the value of a palliative resection are not addressed.
All available systematic reviews and RCTs were independently selected by two investigators (SML, BCV) by
using PubMed, MEDLINE, and the Cochrane Controlled
Trial register through November 2008. Electronic links to
related articles and references of selected articles were
hand searched as well. The present systematic review
only includes studies published in English. The following
search terms were used: esophageal cancer, adenocarcinoma/squamous cell carcinoma, and esophagus.
Searches were restricted to randomized controlled trials,
systematic reviews, and meta-analyses. Ideally, level 1a
evidence was reviewed, if available, and if not, then
evidence from the next level was reviewed [1–3].
Results
Surgical Approach
TRANSTHORACIC VERSUS TRANSHIATAL ESOPHAGECTOMY. Two
major surgical strategies can be used to optimize outcome after esophagectomy. To improve cure, an en bloc
transthoracic resection or an extended resection with
two-field lymphadenectomy have been proposed. Alternatively, early postoperative morbidity and mortality
may be decreased by limiting the extent of dissection
using a transhiatal technique. English-language literature was systematically reviewed. With only three small
RCTs (including a total of 138 patients) published at that
time [7–9], 50 studies were reviewed including 7,527
patients. There were no important differences between
perioperative complications, although blood loss was
significantly higher after transthoracic resection. Transthoracic resections had a higher risk of pulmonary complications, chylous leakage, and wound infection. Anastomotic leakage and vocal cord paralysis were more
frequent after transhiatal resections. Intensive care unit
stay and hospital stay was significantly longer after
transthoracic resection. Furthermore, in-hospital mortality was significantly higher after transthoracic resections
(9.2% vs 5.7%). There was no significant difference in
5-year survival (23.0% vs 21.7%) after transthoracic and
transhiatal resections, respectively. Although transthoracic resections had significantly higher early morbidity
and mortality rates, it was concluded that 5-year survival
was comparable [10]. However, this meta-analysis merely
0003-4975/10/$36.00
doi:10.1016/j.athoracsur.2009.09.062
REVIEWS
Evidence-based medicine is the conscientious, explicit,
and judicious use of best available evidence in making
decisions for individual patient care. The present review
gives an evidence-based review of esophageal cancer
surgery. The literature search was restricted to the highest level of evidence on the surgical treatment of esophageal cancer.
1320
REVIEW
LAGARDE ET AL
EVIDENCE-BASED SURGICAL TREATMENT OF ESOPHAGEAL CANCER
compared surgical access rather than extent of dissection,
limiting these conclusions.
Subsequently, a RCT was conducted. Patients were
randomized to either transhiatal (106 patients) or transthoracic resection with two-field en bloc lymphadenectomy (114 patients) for adenocarcinoma of the distal
esophagus or cardia. Perioperative morbidity was higher
after transthoracic esophagectomy without significant
difference in-hospital mortality. After transhiatal and
transthoracic resection, 5-year survival was 34% and 36%,
respectively. The median number of quality-adjusted
life-years after transhiatal resection was not significantly
different (p ⫽ 0.26). The cost of treatment with transthoracic resection was 56% higher. It was concluded that
transhiatal esophagectomy was associated with lower
morbidity than transthoracic esophagectomy. There appeared a trend toward improved long-term survival at 5
years with the extended transthoracic approach [11, 12].
Randomization was stratified according to tumor location. In a subgroup analysis based on post-surgery classification, the long-term benefit of transthoracic esophagectomy was more substantial in patients with
esophageal tumors. In 90 patients with an adenocarcinoma located in the esophagus, a survival benefit of 14%
was seen with the transthoracic approach compared with
the transhiatal approach (51% vs 37%). No overall survival benefit for either surgical approach was seen in 115
patients with junctional or cardiac tumors [12]. Therefore,
the authors now consider transthoracic esophagectomy
standard treatment for otherwise fit patients with potentially curable EC, whereas transhiatal esophagectomy is
the preferred approach in patients with junctional or
cardiac cancer [13].
Two-Field Versus Three-Field Lymphadenectomy
REVIEWS
Only one small RCT (62 patients) compared extended
cervical and superior mediastinal lymphadenectomy
with conventional two-field dissection and demonstrated
a slightly (not significantly) better survival in the extended lymphadenectomy group at 2 years (83% vs 66%)
and 5 years (65% vs 48%), respectively. The patients
undergoing three-field dissection had significantly more
phrenic nerve palsy (13% vs 0%) and tracheostomies
(53% vs 10%), but fewer anastomotic leakages (6% vs
20%). The authors stated that larger RCTs were needed to
establish the value of cervical lymphadenectomy [14].
Therefore, there is no evidence to support the routine use
of a three-field lymphadenectomy.
Two-Stage or Synchronous Two-Team
Transthoracic Approach
Hayes and colleagues [15, 16] randomized 27 patients
with esophageal carcinoma to a standard two-stage
Lewis-Tanner transthoracic subtotal esophagectomy or a
synchronous two-team approach. Although the synchronous operations were completed significantly more
quickly (230 vs 305 minutes), they produced a higher
incidence of complications (in 7 vs 4 patients after standard esophagectomy), with significantly more blood
transfusions (5 vs 3 units), and 3 postoperative deaths
Ann Thorac Surg
2010;89:1319 –26
after the two-team approach. Therefore, use of the twostage procedure is recommended.
Minimally Invasive Resection
Esophagectomies are associated with high risk postoperative morbidity. Therefore, minimally invasive esophagectomy is gaining in popularity, despite concerns of the
learning curve and oncological results. No RCTs have
been performed to compare minimally invasive surgery
with open surgery. However, Gemmill and McCulloch
[17] systematically reviewed the English-language literature. The review included relatively weak case studies
and case-matched studies. There were 1,398 patients
described who underwent some form of minimally invasive esophagectomy. There were 32 (2.3%) of these 1,398
patients who died within 30 days after the esophagectomy. There were 628 (46%) who had a reported complication. Anastomotic leakage affected 106 (7.7%) of 1,381
patients. Respiratory infections occurred in 167 (13%) of
1,268 patients. Mean blood loss was 316 mL. The duration
of operation was 281 minutes 56 (4.9%). There were 1,138
esophagectomies that were converted to open surgery.
Only in 222 patients, the complete resection rate was
stated clearly. Overall the complete resection rate was
91.0%, and the mean number of resected nodes was 17.6
(607 patients). It can be concluded that the quality of the
studies is poor; therefore, the data must be analyzed with
caution. The reports describe a variety of different techniques. However, in experienced hands, minimally invasive surgery seems feasible and safe, but its generalizability should be questioned. Potential bias is that
patients selected for minimally invasive surgery are not
representative of the population of patients with cancer
(eg, patients with smaller tumors and avoidance of candidates with serious co-morbidity). Moreover, surgeons
whose results were unsatisfactory may have been less
inclined to publish their results (publication bias). Until
large randomized studies are available, minimally invasive resection should be seen as investigational.
Technical Surgical Aspects
The stomach is the method of
reconstruction in most patients. The esophageal substitute can be placed in the anatomical prevertebral position or in an extra-anatomical (retrosternal or subcutaneous) position. Extra-anatomical reconstruction could offer
the advantage that a recurrent intrathoracic tumor mass
will not invade the neo-esophagus. However, it may lead
to increased anastomotic leakage rates and worse longterm functional results. A meta-analysis of RCTs to
determine the effect of the route of reconstruction on
patient outcomes was performed. Six RCTs containing
342 patients were reviewed [18 –22, 23]. Relative risk,
expressed as posterior versus anterior mediastinal route
(treatment vs control) was 0.56 (95% CI, 0.17–1.82) for
mortality, 1.01 (95% CI, 0.35–2.94) for anastomotic leaks,
0.43 (95% CI, 0.17–1.12) for cardiac complications, and
0.67 (95% CI, 0.34 –1.33) for pulmonary complications.
Systematic qualitative review of the data did not suggest
any difference in other perioperative outcomes or con-
ROUTE OF RECONSTRUCTION.
REVIEW
LAGARDE ET AL
EVIDENCE-BASED SURGICAL TREATMENT OF ESOPHAGEAL CANCER
duit function for the two routes of reconstruction. Therefore, posterior and anterior mediastinal routes of reconstruction are considered to have similar functional
outcome and quality of life. However, they erroneously
included the RCT of Zieren and colleagues [23], who did
not randomize for the route of reconstruction. More
RCTs would be needed to definitively resolve the controversies [24]. At present, the commonly used prevertebral
route seems safe and suffices for most situations. When a
macroscopically incomplete (R2) resection is done, there
are theoretical grounds to support extra-anatomical reconstruction, because the risk of developing a symptomatic locoregional recurrence is high [25].
Cervical or Thoracic Anastomosis?
Cervical anastomosis allows for a larger proximal margin
of resection and is believed to result in less dangerous
leakage, and an increased risk of injury to the recurrent
laryngeal nerve compared with intrathoracic anastomosis. However, the technique of first choice remains controversial. Thus far, three RCTs have been published.
Chasseray and colleagues [26] compared stapled intrathoracic with cervical anastomoses after esophagectomy. Transfusion requirements and operating time were
similar for the 49 patients having a thoracic anastomosis
(TA) and the 43 patients who had a cervical anastomosis
(CA). A CA resulted in a greater median margin of
macroscopically normal esophagus above the tumor (4.0
vs 1.5 cm for TA; p ⬍ 0.05). One patient from each group
had involvement of the resected proximal anastomotic
doughnut. Overall, 41 patients (44%) sustained one or
more complications (22 vs 19 patients). Leakage was
significantly more frequent after cervical anastomosis (11
vs 2 patients; p ⬍ 0.02). Thirty-day mortality rates were
comparable: 7 deaths occurred after TA (14.3%) and 4
after CA (9.3%). Length of hospital stay was similar.
Postoperative strictures occurred in 14% of TA and 23%
of CA patients (p ⫽ not significant) and were most
common after an anastomotic leak. Median survival time
was comparable. The authors concluded that the greater
length of tumor-free esophagus removed did not result in
improved survival period, but was associated with a
higher incidence of anastomotic leakage. This study is
difficult to evaluate because of the variety of techniques
used.
Ribet and colleagues [27] randomized 60 consecutive
patients with cancer of the thoracic esophagus to undergo a cervical anastomosis (30 patients) or a thoracic
anastomosis (30 patients). The surgical technique was
identical. Proximal resection margins were more frequently involved in patients undergoing TA (10 vs 3
patients after CA). Morbidity consisted of 8 anastomotic
leaks in the neck (6 of which were subclinical) and 3 in
the chest (with one subclinical). Respiratory complications (21 vs 11 patients; p ⫽ 0.01) and recurrent laryngeal
trauma (6 vs 1 patient) were more frequently seen in
patients having CA. Postoperative mortality was comparable (5 deaths after CA and 4 after TA). Median survival
was comparable (9 months vs 12 months). Although
1321
mortality was comparable, the authors concluded that
subclinical leaks, respiratory morbidity, and recurrent
laryngeal nerve trauma were more prevalent after CA.
However, it remained unclear as to what extent the neck
dissection added to the morbidity.
Walther and colleagues [28] compared a manually
sutured esophagogastric anastomosis in the neck with
stapled anastomosis in the chest after transthoracic
esophagectomy. Apart from anastomosis, surgical procedures were identical. No cervical lymphadenectomy was
carried out. There were 83 patients who were randomized to receive an anastomosis in the neck (41 patients) or
in the chest (42 patients). To evaluate selection bias,
patients undergoing esophageal resection during the
same period but not randomized (n ⫽ 29) were also
followed and compared with those in the study (n ⫽ 83).
The anastomosis was checked for leakage by roentgenograms with water-soluble contrast medium 5 days postoperatively. Anastomotic leakage was defined as extravasation of contrast or clinical symptoms of leakage, or
both. Objective measurements of anastomotic level and
diameter were assessed with an endoscope and balloon
catheter at 3, 6, and 12 months after surgery. Two patients
(1.8%) died in the hospital, and the remaining 110 patients were followed until death or for a minimum of 60
months. No differences were seen in operating time. The
5-year survival rate was comparable (29% for TA vs 30%
for CA). The leakage rate was 1.8% (one radiologic and
one clinical) with no relation to mortality or anastomotic
method. All patients in the randomized group had tumor-free proximal and distal resection margins, but 1
patient in the nonrandomized group had tumor infiltrating into the proximal resection margin. At 3, 6, and 12
months after the operation, there was no difference in
anastomotic diameter (p ⫽ 0.771). Both increased with
time (p ⫽ 0.004). Experience of dysphagia, number of
anastomotic dilatations and body weight were comparable. With similar results in randomized and nonrandomized patients, study bias was eliminated. The authors
concluded that when neck and chest anastomoses are
performed in a standardized way after esophageal resection, they are equally safe. The additional esophageal
resection of 5 cm in the neck group did not increase
tumor removal or survival; on the other hand, it did not
adversely influence morbidity, anastomotic diameter, or
eating as reflected by body weight development.
Considering the latter trial as the best available evidence at the present time, one could conclude that there
are no major differences in clinical outcomes for anastomoses in the neck or chest after transthoracic esophagectomy. Hospital mortality was not influenced by the site of
the anastomosis in any of the RCTs. One can hardly
imagine that the resection of 5 cm of (uninvolved) proximal esophagus will result in a detectable survival benefit. The increased incidence of anastomotic complications
after neck anastomosis is not uniformly confirmed.
Therefore, it seems prudent to advice surgeons to use the
anastomotic site with which they are best familiar.
REVIEWS
Ann Thorac Surg
2010;89:1319 –26
1322
REVIEW
LAGARDE ET AL
EVIDENCE-BASED SURGICAL TREATMENT OF ESOPHAGEAL CANCER
Ann Thorac Surg
2010;89:1319 –26
REVIEWS
Hand-Sewn or Stapled Anastomosis?
Hand-Sewn: One-Layer or Two-Layer Anastomosis?
Anastomoses can be fashioned by hand or mechanical.
Early (leakage) and late (stricturing) patient outcomes
may be influenced by the technique. A meta-analysis of
RCTs to determine the effect of the anastomotic technique on patient outcomes was performed. Five RCTs
containing 467 patients were reviewed [29 –33]. Only one
RCT compared hand-sewn and stapled esophagogastric
anastomoses in the neck. Relative risk (expressed as
hand-sewn versus stapled) was 0.45 (95% CI, 0.20 –1.00)
for operative mortality, 0.79 (95% CI, 0.44 –1.42) for anastomotic leaks, 0.60 (95% CI, 0.27-1.33) for anastomotic
strictures, 0.99 (95% CI, 0.55–1.77) for cardiac morbidity,
and 0.93 (95% CI, 0.63–1.37) for pulmonary morbidity.
Outcomes examined by systematic qualitative review
demonstrated that the hand-sewn methods led to a
longer operation time in three of four studies, but only
one was significantly longer. Although both techniques
gave similar results for anastomotic outcomes, the stapled
method seemed to be associated with an increased operative mortality, a finding which is difficult to explain [34].
More recently, Hsu and colleagues [35] randomized 63
patients with squamous cell cancer. Patients were randomized to receive either hand-sewn (32 patients) or
circular stapled (31 patients) cervical anastomosis. Transthoracic esophagectomy with mediastinal lymphadenectomy was performed in 59 patients, 4 patients underwent
transhiatal resection. The hand-sewn anastomosis was
accomplished using a double layer of interrupted sutures. The mean operating time was 37 minutes longer
when the hand-sewn method was used (524 vs 447 min;
p ⬍ 0.001). Anastomotic leakage (7 in the hand-sewn
versus 8 patients in the stapler group), hospital mortality
(4 vs 3 patients), and benign esophageal stricture (4 vs 5
patients) were comparable in both groups. The authors
concluded that using a circular mechanical stapler shortens operating time with comparable outcome to the
hand-sewn technique for cervical esophagogastric
anastomoses.
Recently, another RCT was undertaken in 117 patients
with squamous cell carcinoma of the thoracic esophagus
who all underwent Ivor-Lewis esophagectomy. The results of this study showed that both the hand-sewn
method and the staple method were safe. Hand-sewn
anastomosis took 15 minutes longer to perform. The
stapled method had a higher incidence of anastomotic
stricturing in patients with a small diameter of the
esophagus [36]. Another recently performed small RCT
with 32 patients showed similar results. In this small
study the incidence of recurrent laryngeal nerve injury
was higher after handsewn anastomosis, and operation
times were longer. This did not lead to a difference in
postoperative symptoms and long-term survival [37].
In summary, the difference in operating times between
hand-sewn and stapled groups is not consistent in the
literature, probably because different methods of handsewn suturing were used. It seems that both methods
have comparable results in experienced hands. Costs
were not mentioned in these studies.
Zieren and colleague [23] reported an RCT comparing
one-layer and two-layer cervical anastomoses after
esophagectomy. After transthoracic resection (24 patients) and palliative transhiatal resection (16 patients),
the gastric tube was placed retrosternally and after “curative” transhiatal resection (67 patients) in the posterior
mediastinum. Anastomotic leakage and late stricture
were the study endpoints. On postoperative day 7, a
water-soluble contrast study was carried out. The groups
were comparable regarding patient characteristics. After
54 one-layer and 53 two-layer procedures, the rates of
anastomotic leakage were comparable (10 patients in
each group [19%]). After a mean follow-up of 44 weeks,
13 of 51 patients (25%) undergoing one-layer anastomosis
and 28 of 50 (56%) undergoing the two-layer procedure
complained of cervical dysphagia and required dilatation
(p ⬍ 0.01). Strictures were malignant in 2 and 4 of these
patients, respectively. The risk of developing a fibrotic
(benign) stricture during the first year after the operation
was 30% for one-layer anastomosis and 64% for two-layer
anastomosis (p ⬍ 0.05). The authors concluded that based
on comparable leakage rates, one-layer anastomosis is
superior to the two-layer procedure because of the lower
incidence of a benign, fibrotic stricture.
Single-Layer Anastomosis: Continuous
or Interrupted Sutures?
Bardini and colleagues [38] carried out an RCT to compare the efficacy of a single layer of continuous absorbable monofilament with that of a single layer of interrupted Polyglactin sutures in the performance of cervical
esophagogastric anastomoses. Forty-two patients were
enrolled in the study (21 patients per group). A gastrografin swallow study was performed on postoperative
day 10. There was no hospital mortality. One asymptomatic anastomotic leak (4.8%) and two early anastomotic
strictures requiring dilation (9.5%) occurred in patients in
whom the interrupted technique was used. No leaks or
strictures were observed in the continuous technique
group. Significantly less operative time to perform the
anastomosis was required for the continuous technique
compared with the interrupted technique (10 min vs 16
min; p ⬍ 0.0001), and the cost of the suture material was
reduced markedly due to the lower number of sutures
needed for each anastomosis. It was concluded that
either a continuous or an interrupted single-layered
esophagogastric anastomosis in the neck can give satisfactory results after esophagectomy. The continuous
technique has the advantages of being somewhat timesaving and cheaper.
The Diameter of the Anastomosis
A wide cross-sectional area at the anastomotic site might
help in reducing the incidence of anastomotic leakage
and strictures. There were 100 consecutive patients who
were randomized. Fifty patients underwent end-to-side
(two-layer) cervical esophagogastric anastomosis after
removal of a 3 ⫻ 2 cm gastric crescent from the anterior
REVIEW
LAGARDE ET AL
EVIDENCE-BASED SURGICAL TREATMENT OF ESOPHAGEAL CANCER
wall of the gastric tube. In the control group, an end-toside (two-layer) cervical esophagogastric anastomosis
was constructed on the anterior wall of the gastric tube
without removal of a gastric tube crescent. All patients
underwent water soluble contrast esophagography on
postoperative day 8. Anastomotic stricture was defined as
difficulty in swallowing solids or failure to pass a
12.8-mm endoscope through the anastomosis that necessitated bougie dilatation. The incidence of anastomotic
leakage in the experimental arm was significantly less in
comparison with the control group (2 patients [4.3%] vs
10 patients [20.8%]; p ⫽ 0.03). Similarly, anastomotic
stricture formation was significantly lower in the study
group (4 patients [8.5%] vs 14 patients [29.2%]; p ⫽ 0.02).
They concluded that a wide cross-sectional area at the
anastomotic site by removal of a gastric crescent resulted
in significantly less anastomotic complications [39].
Width of the Gastric Tube
The (near)-total stomach can be used as the esophageal
substitute after esophagectomy because of its abundant
blood supply, its elasticity and reservoir function, and the
advantage of only one anastomosis. The advantages of a
narrow gastric tube are its excellent elasticity and the
ease with which it can be pulled up into the neck without
tension. On the other hand, the whole stomach may have
the advantage of the intact submucosal network, as
reported by Pierie and colleagues [40]. Recently, in an
RCT containing 44 patients, the size of the gastric tube on
the blood flow at the anastomotic site, the frequency of
leakage from esophagogastrostomy, and the postoperative nutritional status was evaluated. The patients were
randomized to either reconstruction with a subtotal
stomach (22 patients) or a slender gastric tube (22 patients) after esophagectomy (level 2b evidence). Esophagogastrostomies were hand-sewn in a one-layer method.
Of 22 patients, although leakage occurred in 1 with
subtotal stomach (5%) and 5 with slender gastric tube
reconstruction (25%), the difference did not reach statistical significance (p ⫽ 0.19). The mean blood flow at the
anastomotic site, as measured with laser flow meters was
comparable between the two types of esophageal substitutes. There was also no difference noted in the nutritional status at 6 and 12 months after the operation, as
expressed by body mass index, serum albumin level, or
total lymphocyte count. However, as stated by the authors, the number of patients in this study was small, and
therefore its power was not enough to justify the nodifference outcome (risk of type II statistical error), especially for the frequency of leakage [41].
The Use of Pedicled Omentum in Anastomosis
Anastomotic leakage is a life-threatening complication.
The use of a pedicled omentum in the esophagogastric
anastomosis may lead to less anastomotic complications
due to rapid adhesions and new blood vessel networks to
the underlying tissues, thereby helping to seal microscopic leaks and helping with tissue remodeling. The
mobile pool of histiocytes, monocytes, and granulocytes
in the omentum, contain the local infective process and
1323
thereby protect the anastomosis. Using a pedicled omental wrap has been evaluated in an RCT. There were 194
patients who were randomized in two groups. In 97
patients, a pedicled omentum was wrapped around the
esophagogastric anastomosis, and 97 patients did not
have an omental graft. The use of a pedicled omental
wrap was associated with a significant reduction in
anastomotic leakage (3.7% vs 14.3%). However, mortality
rates in the two groups of patients were not significantly
different. The findings in this study should be taken with
a little caution because hand-sewn anastomoses were
performed with silk, which is known to increase the risk
of infection because of potential bacterial harboring [42].
The Use of a Pyloroplasty
Because gastric emptying may be impaired after esophagectomy with gastric reconstruction, a pyloroplasty or
pyloromyotomy may be performed. Urschel and colleagues [43] performed a meta-analysis of RCTs to determine the effect of pyloric drainage. Nine RCTs could be
identified, including a total of 553 patients [44 –52]. Early
outcomes assessed included operative mortality, anastomotic leakage, pulmonary morbidity, pyloric drainage
complications, fatal pulmonary aspiration, and gastric
outlet obstruction. Only three trials provided data on
these early postoperative outcomes. The trials were not
uniform in their patient inclusion criteria, outcome assessment, method of esophagectomy, extent of resection,
pathology, route of reconstruction, specific characteristics of gastric conduit, and technique of pyloric drainage
(pyloroplasty or pyloromyotomy). The diagnostic criteria
for early postoperative gastric outlet obstruction often
included clinical findings, such as aspiration, vomiting,
postprandial fullness, and regurgitation, and radiological
findings of gastric dilatation. The relative risk, expressed
as pyloric drainage versus no drainage, was 0.92 (95% CI,
0.34 –2.44) for operative mortality, 0.90 (95% CI, 0.47–1.76)
for esophagogastric anastomotic leaks, 0.69 (95% CI,
0.42–1.14) for pulmonary morbidity, 2.55 (95% CI, 0.34 –
18.98) for pyloric drainage complications, 0.25 (95% CI,
0.04 –1.60) for fatal pulmonary aspiration, and 0.18 (95%
CI, 0.03– 0.97) for early postoperative gastric outlet obstruction (favors pyloric drainage). Systematic semiquantitative review showed a nonsignificant trend favoring pyloric drainage for the late outcomes like gastric
emptying, food intake and nutritional status, and obstructive upper gastrointestinal symptoms. For the late
outcome of bile reflux, however, there was a nonsignificant trend favoring the no-drainage group. The scintigraphic gastric emptying time, expressed as a ratio (pyloric drainage/no drainage), was 0.53 (shorter emptying
with pyloric drainage). The authors concluded that pyloric drainage procedures reduce the occurrence of early
postoperative gastric outlet obstruction after esophagectomy with gastric reconstruction, without affecting other
early or late patient outcomes. Further trials with much
larger numbers of patients would be needed to clarify the
possible beneficial or detrimental impact of pyloric drainage on other early and late postoperative outcomes [53].
REVIEWS
Ann Thorac Surg
2010;89:1319 –26
1324
REVIEW
LAGARDE ET AL
EVIDENCE-BASED SURGICAL TREATMENT OF ESOPHAGEAL CANCER
The Use of Neck Drainage
To evaluate the role of a closed-suction drain for esophagogastric anastomosis in the neck, an RCT in 40 patients
who underwent esophagectomy with a (left-sided) cervical anastomosis was conducted. Half of the patients had
a drain inserted at the end of operation. The output of
neck drains was recorded every 4 hours; the neck wounds
were inspected daily for any evidence of hematoma or
seroma formation. The drain was removed when its
output was less than 10 mL per day. The median duration
of drainage was 46 hours (range, 36 to 88 hours). The
median total output of neck drains was 63 mL (range, 15
to 210 mL). Hematoma or seroma formation did not occur
in any patients in either group. Anastomotic leakage did
not occur in any patient. The benefits of a drain could not
be demonstrated, and therefore the investigators concluded that routine drainage was not necessary. Again,
the number of patients in this trial may have been too
small to detect a difference in outcome. In view of the low
risk on hematoma or seroma formation in the neck, or
both, such a difference would be of little clinical relevance [54].
The Use of Fibrin Glue to Reduce Lymph Leakage
REVIEWS
Fibrin glue has been shown to be effective in improving
postoperative chylothorax after various thoracic procedures and in reducing lymphorrhea after axillary dissection. The value of using fibrin glue after esophagectomy
was investigated in a randomized controlled trial. A
series of 43 consecutive patients with thoracic EC who
underwent extended esophagectomy were prospectively
randomized to two groups; in 21 patients 3 mL of fibrin
glue was applied to the dissected mediastinum, and in 22
patients fibrin glue was not applied. The daily volume
from the thoracic drain was significantly larger (p ⬍ 0.05)
on postoperative day 1 in patients who had the fibrin glue
applied. Also the cumulative drainage volume was significantly larger on postoperative days 4 to 6, and 9,
suggesting that the cumulative drainage volume in this
group was consistently larger. Therefore, the authors
concluded that the application of fibrin glue to the
dissected mediastinum seems to induce postoperative
lymph leakage and may thus be responsible for prolonged chest tube insertion in some patients. Hence, the
use of fibrin glue can not be recommended for reducing
lymph leakage [54].
Comment
In summary, with EC surgery there long has been a
debate about the best surgical approach. Only one RCT
addresses the concerns of the optimal extent of resection,
in which the results suggest that fit patients are best
treated by a transthoracic esophagectomy with extended
en bloc (two-field) lymphadenectomy. For less fit patients
or patients with junctional or cardiac tumors, transhiatal
esophageal resection could suffice. Future and ongoing
RCTs should establish the role of three-field lymphadenectomy. Currently, there is no evidence to support the
Ann Thorac Surg
2010;89:1319 –26
routine use of a three-field lymph node dissection for EC.
When performing a two-field resection, a synchronous
two-team approach is associated with more complications. Although the results of minimally invasive surgery
are promising, randomized studies are lacking in evidence, and therefore this type of surgery should be seen
as investigational.
Esophagectomy is associated with a high rate of morbidity and considerable mortality. Most RCTs in this field
concern technical aspects of the procedure, mainly to
optimize the results (and minimize complications) of
esophageal reconstruction. Diverse technical adaptations
have been subject of investigation; however, again, limited randomized trials have been performed. Available
evidence suggests that posterior and anterior mediastinal
routes of reconstruction have similar functional outcome
and quality of life after esophagectomy for cancer. There
were no major differences found in clinical outcomes for
anastomoses in the neck or chest after transthoracic
esophagectomy. In addition, results of the hand-sewn
and stapled anastomosis are comparable in experienced
hands. When using a hand-sewn anastomosis, a continuous or an interrupted single-layered esophagogastric
anastomosis in the neck can give satisfactory results after
esophagectomy, although the latter takes more time and
is slightly more costly. Therefore, it seems prudent to
advice surgeons to use the technique in which they are
best familiar.
Although leakage rates of one-layer anastomosis are
similar to rates of the two-layer procedure, the first
technique is superior due to a lower incidence of benign
strictures. For cancers located in the thoracic esophagus,
a wider anastomosis might be of benefit, and the use of
pedicled omentum might decrease the rate of anastomotic leakage, but this should be confirmed in larger
trials. Routine pyloroplasty and placement of a neck
drain are unnecessary. There is no indication to use fibrin
glue to prevent lymph leakage.
Altogether, the present review in EC surgery indicates
that there is only limited high-quality evidence available.
There is a need for large, multicenter randomized studies
concerning many aspects of EC surgery.
References
1. Sackett DL, Rosenberg WM, Gray JA, et al. Evidence based
medicine: what it is and what it isn’t. BMJ 1996;312:71–2.
2. Sackett DL, Rosenberg WM, Gray JA, et al. Evidence based
medicine: what it is and what it isn’t. Clin Orthop Relat Res
2007;455:3–5.
3. Ubbink DT, Legemate DA. Evidence-based surgery. Br J
Surg 2004;91:1091–2.
4. van Lanschot JJ, Hulscher JB, Buskens CJ, et al. Hospital
volume and hospital mortality for esophagectomy. Cancer
2001;91:1574 – 8.
5. Wouters MW, Wijnhoven BP, Karim-Kos HE, et al. Highvolume versus low-volume for esophageal resections for
cancer: the essential role of case-mix adjustments based on
clinical data. Ann Surg Oncol 2008;15:80 –7.
6. Horton R. Surgical research or comic opera: questions, but
few answers. Lancet 1996;347:984 –5.
REVIEW
LAGARDE ET AL
EVIDENCE-BASED SURGICAL TREATMENT OF ESOPHAGEAL CANCER
7. Goldminc M, Maddern G, Le PE, et al. Oesophagectomy by
a transhiatal approach or thoracotomy: a prospective randomized trial. Br J Surg 1993;80:367–70.
8. Chu KM, Law SY, Fok M, et al. A prospective randomized
comparison of transhiatal and transthoracic resection for
lower-third esophageal carcinoma. Am J Surg 1997;174:
320 – 4.
9. Jacobi CA, Zieren HU, Muller JM, et al. Surgical therapy of
esophageal carcinoma: the influence of surgical approach
and esophageal resection on cardiopulmonary function. Eur
J Cardiothorac Surg 1997;11:32–7.
10. Hulscher JB, Tijssen JG, Obertop H, et al. Transthoracic
versus transhiatal resection for carcinoma of the esophagus:
a meta-analysis. Ann Thorac Surg 2001;72:306 –13.
11. Hulscher JB, van Sandick JW, de Boer AG, et al. Extended
transthoracic resection compared with limited transhiatal
resection for adenocarcinoma of the esophagus. N Engl
J Med 2002;347:1662–9.
12. Omloo JM, Lagarde SM, Hulscher JB, et al. Extended transthoracic resection compared with limited transhiatal resection for adenocarcinoma of the mid/distal esophagus: fiveyear survival of a randomized clinical trial. Ann Surg 2007;
246:992–1000.
13. Hulscher JB, van Lanschot JJ. Individualised surgical treatment of patients with an adenocarcinoma of the distal
oesophagus or gastro-oesophageal junction. Dig Surg 2005;
22:130 – 4.
14. Nishihira T, Hirayama K, Mori S. A prospective randomized
trial of extended cervical and superior mediastinal lymphadenectomy for carcinoma of the thoracic esophagus. Am J
Surg 1998;175:47–51.
15. Hayes N, Shaw IH, Raimes SA, et al. Comparison of conventional Lewis-Tanner two-stage oesophagectomy with the
synchronous two-team approach. Br J Surg 1995;82:426.
16. Hayes N, Shaw IH, Raimes SA, et al. Comparison of conventional Lewis-Tanner two-stage oesophagectomy with the
synchronous two-team approach. Br J Surg 1995;82:95–7.
17. Gemmill EH, McCulloch P. Systematic review of minimally
invasive resection for gastro-oesophageal cancer. Br J Surg
2007;94:1461–7.
18. Imada T, Ozawa Y, Minamide J, et al. Gastric emptying after
gastric interposition for esophageal carcinoma: comparison
between the anterior and posterior mediastinal approaches.
Hepatogastroenterology 1998;45:2224 –7.
19. van Lanschot JJ, van BM, Oei HY, et al. Randomized comparison of prevertebral and retrosternal gastric tube reconstruction after resection of oesophageal carcinoma. Br J Surg
1999;86:102– 8.
20. Gawad KA, Hosch SB, Bumann D, et al. How important is
the route of reconstruction after esophagectomy: a prospective randomized study. Am J Gastroenterol 1999;94:1490 – 6.
21. Bartels H, Thorban S, Siewert JR. Anterior versus posterior
reconstruction after transhiatal oesophagectomy: a randomized controlled trial. Br J Surg 1993;80:1141– 4.
22. Coral RP, Constant-Neto M, Silva IS, et al. Comparative
anatomical study of the anterior and posterior mediastinum
as access routes after esophagectomy. Dis Esophagus 2003;
16:236 – 8.
23. Zieren HU, Muller JM, Pichlmaier H. Prospective randomized study of one- or two-layer anastomosis following oesophageal resection and cervical oesophagogastrostomy. Br J
Surg 1993;80:608 –11.
24. Urschel JD, Urschel DM, Miller JD, et al. A meta-analysis of
randomized controlled trials of route of reconstruction after
esophagectomy for cancer. Am J Surg 2001;182:470 –5.
25. van Lanschot JJ, Hop WC, Voormolen MH, et al. Quality of
palliation and possible benefit of extra-anatomic reconstruction in recurrent dysphagia after resection of carcinoma of
the esophagus. J Am Coll Surg 1994;179:705–13.
26. Chasseray VM, Kiroff GK, Buard JL, et al. Cervical or
thoracic anastomosis for esophagectomy for carcinoma. Surg
Gynecol Obstet 1989;169:55– 62.
1325
27. Ribet M, Debrueres B, Lecomte-Houcke M. Resection for
advanced cancer of the thoracic esophagus: cervical or
thoracic anastomosis? J Thorac Cardiovasc Surg 1992;103:
784 –9.
28. Walther B, Johansson J, Johnsson F, et al. Cervical or thoracic
anastomosis after esophageal resection and gastric tube
reconstruction: a prospective randomized trial comparing
sutured neck anastomosis with stapled intrathoracic anastomosis. Ann Surg 2003;238:803–12.
29. Suturing or stapling in gastrointestinal surgery: a prospective randomized study. Br J Surg 1991;78:337– 41.
30. Laterza E, de’ MG, Veraldi GF, et al. Manual compared with
mechanical cervical oesophagogastric anastomosis: a randomised trial. Eur J Surg 1999;165:1051– 4.
31. Law S, Fok M, Chu KM, et al. Comparison of hand-sewn and
stapled esophagogastric anastomosis after esophageal resection for cancer: a prospective randomized controlled trial.
Ann Surg 1997;226:169 –73.
32. Craig SR, Walker WS, Cameron EW, et al. A prospective
randomized study comparing stapled with handsewn oesophagogastric anastomoses. J R Coll Surg Edinb 1996;41:
17–9.
33. Valverde A, Hay JM, Fingerhut A, et al. Manual versus
mechanical esophagogastric anastomosis after resection for
carcinoma: a controlled trial. Surgery 1996;120:476 – 83.
34. Urschel JD, Blewett CJ, Bennett WF, et al. Handsewn or
stapled esophagogastric anastomoses after esophagectomy
for cancer: meta-analysis of randomized controlled trials.
Dis Esophagus 2001;14:212–7.
35. Hsu HH, Chen JS, Huang PM, et al. Comparison of manual
and mechanical cervical esophagogastric anastomosis after
esophageal resection for squamous cell carcinoma: a prospective randomized controlled trial. Eur J Cardiothorac
Surg 2004;25:1097–101.
36. Luechakiettisak P, Kasetsunthorn S. Comparison of handsewn and stapled in esophagogastric anastomosis after
esophageal cancer resection: a prospective randomized
study. J Med Assoc Thai 2008;91:681–5.
37. Okuyama M, Motoyama S, Suzuki H, et al. Hand-sewn
cervical anastomosis versus stapled intrathoracic anastomosis after esophagectomy for middle or lower thoracic esophageal cancer: a prospective randomized controlled study.
Surg Today 2007;37:947–52.
38. Bardini R, Bonavina L, Asolati M, et al. Single-layered
cervical esophageal anastomoses: a prospective study of two
suturing techniques. Ann Thorac Surg 1994;58:1087–9.
39. Gupta NM, Gupta R, Rao MS, et al. Minimizing cervical
esophageal anastomotic complications by a modified technique. Am J Surg 2001;181:534 –9.
40. Pierie JP, de Graaf PW, van Vroonhoven TJ, et al. The
vascularization of a gastric tube as a substitute for the
esophagus is affected by its diameter. Dis Esophagus 1998;
11:231–5.
41. Tabira Y, Sakaguchi T, Kuhara H, et al. The width of a gastric
tube has no impact on outcome after esophagectomy. Am J
Surg 2004;187:417–21.
42. Bhat MA, Dar MA, Lone GN, et al. Use of pedicled omentum
in esophagogastric anastomosis for prevention of anastomotic leak. Ann Thorac Surg 2006;82:1857– 62.
43. Urschel JD, Blewett CJ, Young JE, et al. Pyloric drainage or
no drainage in gastric reconstruction after esophagectomy: a
meta-analysis of randomized controlled trials. Dig Surg
2002;19:160 – 4.
44. Fok M, Cheng SW, Wong J. Pyloroplasty versus no drainage
in gastric replacement of the esophagus. Am J Surg 1991;162:
447–52.
45. Gupta S, Chattopadhyay TK, Gopinath PG, et al. Emptying
of the intrathoracic stomach with and without pyloroplasty.
Am J Gastroenterol 1989;84:921–3.
46. Hsu HK, Huang MH, Chien KY, Liu RS, Yeh SH. Functional
evaluation of using the stomach as an esophageal substitute.
J Surg Assoc ROC 1984;6:186 – 8.
REVIEWS
Ann Thorac Surg
2010;89:1319 –26
1326
REVIEW
LAGARDE ET AL
EVIDENCE-BASED SURGICAL TREATMENT OF ESOPHAGEAL CANCER
47. Huang GJ, Zhang D, Zhang D. A comparative study of
resection of carcinoma of the esophagus with and without
pyloroplasty. Esophageal Disorders 1985;8:383–7.
48. Kao CH, Chen CY, Chen CL, et al. Gastric emptying of the
intrathoracic stomach as oesophageal replacement for oesophageal carcinomas. Nucl Med Commun 1994;15:152–5.
49. Kobayashi A, Ide H, Eguchi R, et al. [The efficacy of pyloroplasty affecting to oral-intake quality of life using reconstruction with gastric tube post esophagectomy]. Nippon Kyobu
Geka Gakkai Zasshi 1996;44:770 – 8.
50. Mannell A, McKnight A, Esser JD. Role of pyloroplasty in the
retrosternal stomach: results of a prospective, randomized,
controlled trial. Br J Surg 1990;77:57–9.
Ann Thorac Surg
2010;89:1319 –26
51. Zieren HU, Muller JM, Jacobi CA, et al. [Should a pyloroplasty
be carried out in stomach transposition after subtotal esophagectomy with esophago-gastric anastomosis at the neck? A
prospective randomized study]. Chirurg 1995;66:319 –25.
52. Chattopadhyay TK, Gupta S, Padhy AK, et al. Is pyloroplasty
necessary following intrathoracic transposition of stomach?
Aust N Z J Surg 1991;61:366 –9.
53. Choi HK, Law S, Chu KM, et al. The value of neck drain in
esophageal surgery: a randomized trial. Dis Esophagus 1998;
11:40 –2.
54. Tachibana M, Kinugasa S, Yoshimura H, et al. Does fibrin glue
reduce lymph leakage after extended esophagectomy? Prospective randomized clinical trial. World J Surg 2003;27:776 – 81.
REVIEWS