Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
1877—V. Czerny: first successful resection of the cervical esophagus for carcinoma
1913—F. Torek: first successful transthoracic resection of the esophagus
1913—W. Denk: cadaver and experimental animal studies on the transhiatal resection of the esophagus
1933—T. Ohsawa: first report on transthoracic esophageal resection and esophagogastrostomy
1933—G. Turner: first transhiatal resection
1938—W. Adams and D. Phemister: first single stage transthoracic resection and reconstruction in the United States
1946—I. Lewis: esophageal resection and esophagogastrostomy via a right thoracotomy and laparotomy
1976—K. McKeown: description of a three-hole esophagectomy
1978—M. Orringer: popularizes transhiatal esophagectomy in the Western hemisphere
1992—A. Cushieri: first report on thoracoscopic esophagectomy
2003—J. Luketich: popularizes total thoracoscopic and laparoscopic esophagectomy
Milestones in esophageal reconstruction
1879—T. Billroth: attempt of reconstruction with skin
1886—J. Mikulicz: Reconstruction of the cervical esophagus by skin flaps
1905—C. Beck and A. Carrel: experimental animal study on tubulization of the greater curvature of the stomach
1906—A. Carrel: successful transplantation of autologous small bowel into the neck of dogs
1907—C. Roux: first use of a presternal jejunal loop combined with skin tube for benign esophageal stricture
1911—H. Vuillet and G. Kelling separately introduced colon as a substitute: first attempt of two-stage resection followed by colonic interposition
The end of the 19th century was characterized by a true race to perform the first esophagectomy. Although Czerny had already performed a successful partial resection of the cervical esophagus for a cancer as early as in 1877, further attempts to successfully resect an intrathoracic esophageal cancer had failed.
It is Franz Torek who is to be credited on the first successful transthoracic (and transpleural) resection of the esophagus in 1913. Reconstruction was not attempted and the patient was fed using a rubber tube connecting the proximal esophagostomy with a gastrostomy; she lived for 13 years.
Further attempts made in the following years were mostly unsuccessful due to lack of technology to adequately ventilate the lungs. Only after the introduction of safe oro-tracheal intubation in the late 1920s by Rowbotham and Magill could surgeons undertake such a complex operation as a transthoracic esophagectomy more safely. Over the following decades pioneers, such as Denk, Ohsawa, Grey Turner, Adam and Phemister, Sweet, Ivor Lewis, McKeown, Belsey, and Orringer, further developed and refined the surgical techniques as we use them today. However, postoperative mortality remained high well into the 1970s. Better insights in medical operability and better perioperative management hallmarked the 1980s and 1990s where operative mortality was brought down to below 5%, and is now approximately 1% to 2% in many centers of experience.
Better selection in terms of oncologic operability through the introduction of the computed tomography scan, the positron emission tomography scan, and endoscopic ultrasound have resulted in a sharp decrease of futile exploratory thoracotomies. Better surgical techniques and the advent of induction therapy substantially increased the R0 resection rates in locally advanced (T3) carcinoma reaching greater than 90% today, in many published series.
Consequently, the resulting long-term survival has been constantly on the rise over the last decades, reaching an overall 5-year survival today between 35% and 45%.
As a result of these improvements in oncologic outcome, increasing attention is now being paid to the functional outcome both short and long term. This is reflected by the high number of publications in recent years focusing on quality of life (QOL).
Ideally the conduit substituting for the esophagus should mimic as closely as possible the function of a normal esophagus to preserve QOL. It should allow the undisturbed transport of the alimentary bolus from mouth to stomach, providing an adequate antireflux mechanism that protects the lungs from aspiration, but allowing the possibility to belch or vomit when necessary.
An equally important objective is to reduce mortality and morbidity inherent to major surgery, such as resection and reconstruction for cancer of the esophagus. In this respect, the choice of the type of conduit to restore continuity may play an important role. Indeed one can imagine that there is a potentially higher risk in postoperative morbidity, eventually mortality, when using a vascularized, so-called supercharged, long-segment, jejunal graft requiring five different anastomoses by two different teams, as compared to a gastric pull-up requiring only one anastomosis.
Furthermore, suboptimal nutritional, immunological, cardiopulmonary status, old age, or impaired performance status (e.g., due to severe arthrosis) may play a role in the choice of the substitute. In this context, the surgeon will preferably make use of a gastric pull-up with one single anastomosis versus a colon interposition requiring three anastomoses, which carries a higher potential for infectious complications.
Organ availability may play a role in the choice of the substitute; for example, previous total gastrectomy or partial or total colectomy, which will preclude the use of the stomach or the colon, respectively, as a substitute.
During surgery, the surgeon may encounter unforeseen situations, for example, former division of the right gastroepiploic artery as a consequence of a right colectomy with lymphadenectomy for colon cancer. Such a finding at the time of surgery will oblige the surgeon to switch to another conduit (e.g., the jejunum). Unforeseen metastatic intramural deposits higher up in the esophagus may require the surgeon to switch the level of the anastomosis from the chest to the neck.
Finally, a critical determinant is the surgeon's expertise, and that of the whole team involved in the pre- and postoperative management.
In other words, esophagectomy followed by reconstruction is one of the most complex and difficult operations on the alimentary tract, which requires an experienced surgical team familiar with all available conduits, and who are able to adapt to every situation in order to offer the patient the best possible type of reconstruction.
Historically, the first attempts for reconstruction were tried by Billroth as early as 1879. Skin was the first material used for reconstruction, but the stomach, colon, and jejunum—in order of their frequency of use—became the three classic substitutes over time. They are mostly used as a single pedicled transposition, but can be used as free vascularized grafts (in particular the jejunum) or occasionally as composite replacements.
Combining conduit choice with a multitude of different access routes, including the recent minimally invasive techniques, and different levels of anastomosis, it is clear that there are a myriad of options available when planning an esophagectomy and reconstruction for cancer. A tailored approach for each individual patient guided by an experienced surgeon is the key to success.
Today, gastroplasty is by far the most preferred conduit for replacing the esophagus in over 95% of cases. It is indeed the quickest and “simplest” organ surgery for reconstruction. It has a robust arterial and venous supply and submucosal plexus. After mobilization, the stomach will receive its blood supply from the right gastroepiploic artery with the venous blood drained via the right gastroepiploic vein ( Fig. 41.1 ). There is only the need for a single anastomosis. As the stomach is a very flexible organ, it can easily reach the neck for a cervical or even hypopharyngeal anastomosis. Its main disadvantage is the potential of reflux and related aspiration problems.
Regarding the “robustness” of the blood supply, the surgeon must be aware of some anatomic variations at the level of the gastroepiploic arcade. In some instances, the right gastroepiploic artery may end about half way up the greater curvature connecting to the left gastroepiploic artery only through delicate arterioles in the omentum, which need to be respected during mobilization ( Fig. 41.2 ). Moreover, the pioneering work by Liebermann-Meffert et al., using corrosion casts on cadaver specimens, clearly shows that the submucosal plexus is thinning out near the top of the fundus, and that there is a watershed zone with a clear decrease of the intertwining connections between the submucosal microcirculation of the left side (lesser curvature) and the right side (greater curvature).
Studies have shown that after full mobilization of the greater curvature of the stomach and ligation of the left and right gastric artery, the oxygen tension at the top of the fundus decreases substantially, and, after transposing the conduit up to the neck, it falls to approximately 50%. Efforts have been made to prevent this effect by preoperative conditioning of the vascular supply; that is, ligating or embolizing the left gastric artery a few days before the planned esophagectomy and reconstruction. However, the results have been equivocal and one prospective, randomized trial failed to show any advantage. Therefore, it is of paramount importance not to traumatize the top end of the stomach in order to avoid anastomotic leak or eventually fundic necrosis.
In general, most surgeons prefer to start with a laparotomy/laparoscopy in order to first inspect the abdominal cavity to exclude the presence of unforeseen metastasis before mobilization of the stomach. Alternatively, one may prefer to start with a thoracotomy/video-assisted thoracoscopic surgery (VATS) in order to assess resectability, particularly in the case of a questionable T3-4 tumor. Mobilization of the stomach is started by opening the gastrohepatic ligament so that the hiatus and the right pillar of the right crus come into view. In some patients, there will be a separate left hepatic artery present. If this artery is small, a couple of millimeters in diameter, it can be ligated without causing harm to the liver. A larger size left hepatic artery needs to be dissected carefully down to its origin, usually at the left gastric artery, and both need to be preserved in order to avoid major, possibly lethal, hepatic necrosis.
This is followed by the mobilization of the greater curvature by dividing, distally to the gastroepiploic vascular arcade, the omentum and its feeding vascular branches, as well as the attachments to the transverse colon in this area. Usually, the easier point to start the dissection of the greater curvature is about halfway up the greater curvature.
In some cases, due to previous episodes of pancreatitis for example, the retrogastric area will be obliterated making the dissection more difficult. The utmost care has to be taken not to damage the gastroepiploic vessels. Especially in laparoscopic mobilization, care should be taken never to grab the gastroepiploic vessels nor the greater curvature of the stomach with the laparoscopic forceps because of the earlier described very important but fragile microcirculation. In laparoscopic mobilization, it is very helpful to pull on the lesser curvature, as it will be stapled out afterward.
The dissection is further continued upward along the greater curvature and the gastroepiploic arcade is divided at the level of the inferior pole of the spleen at the point where the left gastroepiploic artery ends in the splenic artery.
In a number of patients, the right gastroepiploic artery interrupts somewhere midway or two-thirds of the way down on the greater curvature ( Fig. 41.2 ). The connections with the left gastroepiploic artery or the short gastric vessels here come through the omental intertwining small vessels. At this level, it is important to stay away from the greater curvature and dissect more on the periphery of the omentum so these connections can be preserved. Coming at the level of the short gastric vessels connecting the spleen and the proximal greater curvature, it is again preferable to stay as close to the spleen as possible when dividing the vessels. Again this allows preservation of the small delicate connections within the submucosal plexus. Preserving as much as possible of the omentum at that level also allows it to be used to protect the anastomosis later on. Obviously, great care is taken not to damage the spleen. Today, with the help of ultrasonic devices, the risk of damage requiring eventually iatrogenic splenectomy has substantially decreased and, in fact, should no longer occur. Once the short gastric vessels are divided, the gastrodiaphragmatic area is dissected so that, after retraction of the fundus to the right, the left pillar of the right crus will come into view, allowing further dissection of the entire retrogastric area covering the hiatus.
The hiatus can now be opened by incising the phrenoesophageal ligament (if not already done from above during the thoracic portion of the case). In case of a gastroesophageal junction (GEJ) tumor, a rim of the diaphragmatic hiatus is excised and left in continuity with the esophagus in order to ensure a complete R0 resection. Now the inferior mediastinum becomes widely opened and accessible so the distal esophagus can be dissected out and a tape passed around it.
Next comes the ligation of the left gastric artery. This is done after the dissection of lymph nodes around the left gastric artery and the celiac axis extending over the common hepatic and the splenic artery.
Both the left gastric artery and accompanying vein are ligated and divided separately. The further dissection is continued toward the right toward the pylorus―first at the side of the greater curvature, again making sure not to damage the right gastroepiploic vessels, then on the lesser curvature where the right gastric artery is ligated and divided.
In order to obtain maximal mobility and length, it may be useful to perform a Kocher maneuver elevating the duodenum and the head of the pancreas anteriorly off the inferior vena cava. By doing so, the pylorus easily can be brought up to the level of the esophageal hiatus.
Most surgeons today prefer to create a rather narrow gastric tube with a width of about 4 cm (which is about on the watershed zone between the earlier described left and right microcirculation). To do so, the lesser curvature needs to be resected. This is also an essential oncologic act since the lymph nodes on the lesser curvature are known to be at risk for metastatic involvement—definitely in the lower half and in GEJ cancers.
The creation of a gastric tube has been enormously facilitated by the introduction of the staplers. If the access route is a left thoracoabdominal one, the stapling starts from above, at the top of the fundus. A linear cutting stapler will be placed at the top end of the fundus about 5 cm laterally from the GEJ ( Fig. 41.3 ). The staplers are then placed in a vertical direction. Several staplers will be necessary. The end result is a long gastric tube with a width of approximately 4 to 5 cm. If the access is through a laparoscopic approach, the stapling starts from the lesser curvature at the level of the crow's foot, about 4 cm proximal to the pylorus.
Many surgeons will oversew the staple line with separate or running sutures. This will help not only to protect against leakage but also (arguably) to prevent early postoperative dilatation of the tube ( Fig. 41.4 ). However, it can shorten the gastric tube and when performed laparoscopically, it is very time consuming.
The gastric tube is temporarily fixed to the divided lesser curvature with two stay sutures and the whole complex of esophagus, lesser curvature, and gastric tube will be pulled up via the cervical or thoracic incision and exteriorized. It is important to make sure that the tube is not twisted around its axis, with the staple line being on the medial side.
After exteriorizing the whole complex, the anastomosis will be constructed.
In case of an Ivor Lewis approach, the anastomosis will be fashioned in the chest via a right thoracotomy or VATS. In a McKeown fashion, the anastomosis will be in the neck, preferably on the left side.
The anastomosis is the Achilles heel of the operation. Anastomotic leak once carried a high mortality. Fortunately, today, mortality has become very low but its relation to early and late morbidity is a source of concern and will be discussed in Chapter 43 .
Because of these serious consequences, over time different techniques, either handsewn, mechanical, or combined, have been developed. This chapter will describe cervical handsewn and semimechanical anastomosis, and the intrathoracic stapled anastomosis.
In the setting of a McKeown (three-hole) intervention, the patient is positioned and draped for laparotomy, or laparoscopy and cervicotomy, so that the surgical team can proceed simultaneously with the abdominal phase together with a left-sided cervicotomy. After the transection of the platysma and omohyoid muscles, the medial border of the sternocleidomastoid muscle is identified and retracted laterally. After division of the inferior thyroid artery, access to the cervical esophagus is achieved. There, the esophagus can be hooked with the surgeon's finger, exteriorizing the esophagus from the neck and chest. In the same movement, the gastric conduit is carefully brought up through the diaphragmatic hiatus into the posterior mediastinum and eventually into the neck. Attention should be paid during this maneuver to not injure the left recurrent laryngeal nerve.
To prevent damage to the vascular pedicle, the conduit should be carefully guided by one hand of the surgeon through the hiatus (or in the case of a laparoscopic procedure, by guiding the vascular pedicle through the hiatus without grasping it with the forceps when pulling up the conduit with the other hand). Care is taken not to axially twist the gastric tube at the cervical level.
At the thoracic inlet, there has to be sufficient width (about three fingers) to allow free passage into the cervical field without compression of the tube. Once the gastric tube is brought in the operative field in the neck, the temporary stay sutures are cut. The proximal esophagus is now lying side by side with the gastric conduit ready to construct an end (esophagus) to side (gastric tube) anastomosis. The place to incise the gastric tube is chosen well away from the lateral staple line. After reflecting the esophagus upward, the same is done on the esophageal wall.
On the outer sides of the esophagus and gastric tube, two sutures of nonresorbable monofilament 3-0 are placed between the muscularis of the esophagus and the seromuscular layer of the stomach. Separate sutures or continuous running suture will complete the posterior outer layer of the anastomosis. Care is taken to avoid any tension between the two conduits ( Fig. 41.5 ).
Then with the electrocautery an incision of approximately 2 cm (maximum 3 cm) is made, widely opening the lumen of the stomach. This incision is made at a minimum distance of 1 cm away from the posterior outer layer.
The same maneuver is done on the esophageal wall comprising the whole width of the esophagus and again a minimum of 1 cm away from the outer layer. The slimy mucous content of the gastric tube and esophagus are carefully removed with suction and swabs, avoiding any spillage in the operative field, and both lumina are disinfected with a povidone-iodine (Betadine) swab.
Two 3-0 monofilament resorbable sutures are placed in each corner from the esophagus outside to inside, then to the stomach inside to outside, to start off the posterior inner layer. After tying the stitches, the inner layer is further completed out from one corner toward the opposite corner with a running suture similar to a vascular anastomosis. After finishing the posterior inner row, the nasogastric tube, partially withdrawn into the proximal esophagus, is advanced through the anastomosis well into the gastric tube.
At this point the residual anterior wall of the esophagus is transected so the anterior part of the anastomosis can start beginning from the opposite corner and running in a similar way back to the corner where the suturing initially started. Some surgeons will prefer to use separate sutures. One has to make sure to always incorporate both the mucosal and seromuscular layer into each stitch in such a way that the mucosal layer is “tucked away” underneath the seromuscular layer. This results in a smooth watertight anastomosis. Five to six mattress sutures taking muscularis on the esophageal side and seromuscularis on the gastric side complete the anterior outer layer. A mattress suture helps to have a better grip on the muscularis layer of the esophagus, which is rather fragile. For this reason, it is advisable not to tie the mattress sutures immediately one after the other, but rather to gently pull them up all together to equally spread out the traction over all sutures before tying them.
Mostly the length of the gastric tube and the chosen location of the anastomosis allow for the resection of a redundant proximal part. Otherwise, a blind sac would form, which can act as a pseudodiverticulum, possibly impairing the passage of food down the tube. This resection is done using a stapler, taking care that the line of transection is about 2 cm away from the anastomosis line to avoid ischemia with the risk of necrosis of the gastric wall in between ( Fig. 41.6 ). The staple line is oversewn via a running nonabsorbable 3-0 monofilament suture. Usually, some omentum is available to wrap around the suture line as a protection against potential leaking. It is of paramount importance to avoid any traumatization of the tissues by clamps, forceps, or other instruments.
As the anastomosis is now finished, the esophagus-gastric tube complex is gently pushed back into the thoracic inlet and, after leaving behind a Redon-type drain, the cervicotomy is closed in layers.
If the length of the conduit is sufficient (>5 cm overlap), a semimechanical Orringer or modified Collard end-to-side anastomosis is preferred ( Fig. 41.7 ). To avoid leakage, from traumatizing the tissues the tip of the gastric conduit will be resected with another linear stapler as described. It starts by placing five separate 3-0 nonresorbable stitches, in the form of a pentagon, between the muscular esophageal wall and the gastric serosal layer, with the tip being at the deepest point. The gastric tube is incised at the base of the pentagon (see Fig. 41.7 ). Next, monofilament 3-0 resorbable sutures are placed in the corners outwards. Then, two monofilament 4-0 resorbable sutures are placed in the middle of the incision, bringing the base of the pentagon together and aligning the gastric and esophageal walls. In between these sutures, a 45-mm linear stapler is fired over a distance of ~35 mm ( Fig. 41.8 ). This will create a V -shaped back wall, allowing a wider passage ( Fig. 41.9 ). This technique prevents narrowing down during the cicatrization of the anastomosis, resulting in less dysphagia for semi-solid and solid food, thus limiting the need for repetitive anastomotic dilatations and improving the patient's QOL. The base of the back wall, lateral to the stapler line, is completed with separate 4-0 monofilament sutures. Then, the nasogastric tube can be pushed through the anastomosis, down in the gastric conduit, to decompress the stomach after surgery. The front wall of the anastomosis is closed with a continuous two-layer suture with the earlier placed monofilament resorbable suture for the inner layer and the nonresorbable suture for the outer layer, as shown in Fig. 41.6 . The tip of the gastric tube is resected by using a linear stapler as described. After careful hemostasis, a small Redon-type drain is placed in the neck to prevent the accumulation of blood, and the incision is closed. This semimechanical anastomosis results in improved dysphagia scores for solids and semi-solids, and significantly reduces the need for dilations, in particular repeat dilatations.
Whether performing a handsewn or semimechanical anastomosis, before closing the cervicotomy, a mini-tracheostomy tube may be inserted through the cricothyroid ligament under tracheoscopic guidance from the anesthesiologist. A mini-tracheostomy facilitates the aspiration of secretions, which can be of particular assistance to patients who have insufficient strength to cough them up. Then, the platysma muscle is closed, followed by the skin suture.
In recent years, as a result of the increased interest in minimally invasive esophagectomy (MIE), there has been a growing trend to perform the anastomosis in the chest. The advantage is that it saves time, but obviously an intrathoracic location of the anastomosis must be oncologically safe; that is, at least 5 cm proximal to the upper pole of the tumor. Furthermore, the anastomosis should be located high in the apex of the right chest to reduce the risk of reflux and related aspiration.
There are two variations on how to proceed. The first option is not to fully complete the gastric tube during the abdominal phase and, after bringing up the mobilized stomach in the chest, to open the lesser curvature remnant to introduce the circular stapler and to resect the lesser curvature remnant after firing the stapler.
The second variant is to first complete the creation of the gastric tube and after bringing it up in the chest, to open the proximal part of the tube to introduce the circular stapler, resecting the top end of the tube after firing the stapler.
First, the anvil of the circular stapler, usually a 25-mm or 28-mm diameter is introduced in the esophagus. To do so, a purse string suture is applied around the cut end of the transected esophageal wall after making sure the frozen section is free of tumor. This purse string is inserted loosely using a monofilament nonabsorbable 3-0 suture ( Fig. 41.10A ). Second, the detached anvil is introduced. This can be somewhat tricky requiring the help of one or two Babcock-type clamps to exert counter traction. The purse string is now securely tied around the central rod of the anvil. Often after this maneuver, there may still be some tissue protruding that requires the placement of a second purse string.
In both variants, a gastrotomy is then performed to introduce the head of the circular stapler with the tip of the shaft fully retracted. The head is carefully positioned well away from the vertical staple line and ensures that the gastric conduit is not twisted. By turning the screw system of the gun, the central rod with pin is pushed through the gastric wall. Then the anvil and shaft are clicked together with the help of clamps ( Fig. 41.10B ). Further turning on the screw mechanism on the stapler device approximates the head and anvil tightly. The stapler is then firmly fired and removed and the two “doughnuts” are inspected for completeness. The esophageal doughnut is submitted for final pathologic examination of the margin. A nasogastric tube is passed through the anastomosis and the gastric remnant is resected using a linear stapler ( Fig. 41.10C ).
The principle of the reversed gastric tube was described by Beck and Carrel as early as in 1905, and later by Jianu, but it is mainly the Rumanian surgeon, Gavriliu, who popularized the technique in the 1980s. Claimed advantages are that it preserves part of the stomach and related gastric function and that it has the ability to reach the pharynx. In this respect, it also has been advocated for benign diseases. For malignant diseases it is rarely used nowadays.
The blood supply is based on the left gastroepiploic artery, which arises from the splenic artery and from the short gastric vessels, and requires a careful dissection in the hilum of the spleen. In essence, the liberation of the greater curvature is as described earlier. Typically, the right gastroepiploic artery is divided about 4 cm proximal to the pylorus ( Fig. 41.11 ). The stomach is now divided starting at the greater curvature with a linear cutting stapler placed vertically, the subsequent staplers being placed in parallel with the greater curvature at a distance of approximately 3 to 4 cm away from the border and up to about two-thirds the length of the greater curvature. If a greater length is needed, a modification can be done by incorporating the pylorus into the tube ( Fig. 41.12 ). The branches from the right gastroepiploic artery to the pylorus are preserved and the artery itself is divided close to its origin from the gastroduodenal artery. To make the conduit, the first stapler is placed at the lesser curvature at the level of the antrum and then further worked out in parallel with the greater curvature. The duodenum is transected with a linear cutting stapler just below the pylorus. Bringing up a Roux-en-Y jejunal limb to the distal stomach in order to avoid biliary reflux restores continuity with the duodenum.
An interesting variant is the split nonreversed gastric tube or split stomach ( Fig. 41.13 ).
Approximately 4 to 5 cm proximal to the pylorus, the shaft with a pin of a 28-mm circular stapler is pushed through both the anterior and posterior walls of the stomach. After clicking the anvil in the central rod the gun is fired, resulting in a 28-mm circular opening in the stomach. This opening suffices to introduce a linear cutting stapler cephalad. The end result is an isoperistaltic nonreversed gastric tube in connection with the antrum. The lesser curvature retains its continuity with the cardia and esophagus, the left gastric artery remaining intact and the gastric tube being pedicled on the right gastroepiploic artery, so only the left gastroepiploic artery and the short gastric vessels are taken down. This technique can be used in the rare case a bypass is needed for an unresectable cancer.
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here