February 7, 2024
Esophagectomy is a complex, morbid procedure that is usually performed to treat esophageal cancerâ€”a poor prognosis cancer with a median survival of 11 months.1 Postoperative complications are common, affecting 59%â€“64% of patients, with â€śprofoundâ€ť short- and long-term negative effects, including early death.2-4 Despite optimal treatment, a minority of patients survive beyond 5 years, often with significantly compromised health-related quality of life.4,5
The ideal esophagectomy would improve survival via reduced complications and good oncologic outcomes, while maximizing the patientâ€™s remaining quality of life. Research suggests that standardizing the surgical approach may be an important prognostic factor.1
In this article, we present our technical adaptations to the total minimally invasive Ivor Lewis esophagectomy (MIE) and our perioperative management strategy (see Figure 1). In particular, we emphasize mediastinal pleural envelope closure over a slender conduit with early structured oral nutrition, obviating the need for jejunostomy feeding tubes.
This approach is highly team-based and depends on regionalization of esophagectomy care to three specialized centers.Ěý Many aspects would not be possible without careful coordination and teamwork from nursing staff, registered dietitians (RDs), and inpatient and outpatient multidisciplinary care teams.6
The key component of our total minimally invasive Ivor Lewis esophagectomy is mediastinal envelope closure over a slender gastric conduit, facilitating early oral intake (see Figure 2, below).
Figure 2: Early oral intake facilitated by operative modifications
In the laparoscopic phase, we first perform a standard hiatal dissection with mobilization of the stomach. Most of the omentum is removed from the gastric conduit, taking care to avoid the right gastroepiploic artery. A thin omentum both facilitates closure of the mediastinal envelope and minimizes shunting to inferior omental branches from the right gastroepiploic artery in the setting of ischemia. We then tubularize the stomach along the lesser curvature to allow for pull-up of the conduit into the thoracic cavity. We do not place a jejunostomy tube and omit a pyloric emptying procedure completely.
With the patient in left lateral decubitus (right video-assisted thoracoscopic surgery or VATS), the mediastinal pleura overlying the esophagus anteriorly is divided superiorly along the length of the mediastinum. Care is taken to preserve the divided pleural edge as the pleura and underlying connective tissue will ultimately be closed around the newly formed conduit.
Inferior to the posterior hilum, the inferior pulmonary ligament is composed of apposed bi-layered pleura that attach medially to the esophagus and inferiorly to the diaphragm. Beginning at the hilum, the divided pleura is a single layer of mesothelium. After neoadjuvant chemoradiation, this layer is often conveniently thickened, further buttressing the mediastinal envelope closure. Division of the pleura is carried out superiorly above the ligated azygous vein to the beginning of the thoracic inlet.
After full esophageal mobilization circumferentially, the esophagus is transected, and a 28 mm EEAâ„˘ stapler anvil is inserted into the proximal esophagus. To create a clean anastomosis, the anvil is secured with two consecutive chromic ENDOLOOPâ„˘ suture ties and excess tissue distal to the ENDOLOOP ties is removed. The gastric conduit is pulled up into the thoracic cavity with the staple line maintained to the patientâ€™s right to avoid torsion. A gastrotomy is made along the lesser curvature of the stomach near the staple line for EEA stapler insertion, and the spike is deployed at the cephalad aspect of the greater curvature, taking care to resect an often devascularized gastric fundus.
After docking the EEA spike and anvil, the esophagogastric anastomosis is created. The gastric conduit (external diameter <4 cm) is further slenderized on the lesser curvature side with exclusion and removal of the gastrotomy with the discarded specimen. The remaining omentum is interposed between the anastomosis and bronchi to protect against tracheoesophageal fistula formation in case of an anastomotic leak or dehiscence postoperatively.
We then close the entire mediastinal pleural envelope from the thoracic inlet above the anastomosis to the diaphragm with interrupted 2-0 silk sutures spaced 2 cm apart. If tension limits adequate envelope closure, the parietal and visceral pleura can be undermined on either side, similar to pleural tenting. Interrupted sutures are used to allow for potential mediastinal drainage between sutures and prevent unravelling if the conduit distends upon oral liquid initiation postoperatively. A 24 French BlakeÂ® Drain is placed over the closed mediastinal envelope near the anastomosis, along with a nasogastric tube (NGT) and a single 28 French chest tube. Before closure, we perform an intercostal nerve blockade with liposomal bupivacaine for postoperative pain control, avoiding thoracic epidurals mainly because of the associated hypotension.
We hypothesize that the slender conduit and mediastinal pleural envelope closure are advantageous for several reasons. Slender tubularization more closely resembles the native esophageal anatomy with a luminal diameter of 2â€“3 cm (see Figure 3, below). The largely denervated gastric conduit is prone to dilation with loss of muscle tone. Significant luminal size discrepancies between the downstream pyloric sphincter and the upstream conduit may exacerbate conduit dilation and intraluminal stasis.
Figure 3: Completing slender tubularization of the conduit after esophagogastric anastomosis
Exposure of a dilated conduit in the right hemithorax to the respiratory cycle also may increase aspiration risk. Closure of the mediastinal envelope offers several potential anatomic advantages. In the absence of mediastinal envelope closure, the conduit is suspended from the anastomosis and a few other points of tension. Anastomotic tension, exacerbated by gravity in the upright esophagectomy patient, impedes tissue perfusion and can be mitigated by creating a thinner conduit and subsequent mediastinal envelope closure. The mediastinal pleural envelope closure also maintains the conduit in a linear cephalocaudal position, promoting streamlined emptying through gravity.
Finally, the pleura is a natural protective barrier with blood supply from both the intercostal and pulmonary vasculature; the mediastinal envelope may function similarly to an omentoplasty without diverting gastroepiploic blood flow. We have found that partial closure of the mediastinal envelope (e.g., below the level of the anastomosis) does not reduce postesophagectomy complications as effectively as full closure of the mediastinal envelope.
The primary condition limiting mediastinal envelope closure is excess tension. Occasionally we cannot perform complete mediastinal envelope closure. When this is the case, we prefer to leave the entire mediastinal envelope open as we have found that partial closure can lead to increased anastomotic leaks (in press). Bulky conduits and excess omentum can impede closure, while preoperative radiation occasionally makes the pleura too friable. Although conduit revisions can be more complex with a slender conduit, we find that a slender conduit is less ischemia-prone, possibly because the right gastroepiploic arteryâ€™s branches have a shorter distance to travel circumferentially with a reduced volume of tissue to perfuse.
In a multicenter retrospective cohort study of adult patients undergoing minimally invasive esophagectomy (in press), we found that complete closure of the mediastinal envelope led to improved outcomes compared to nonclosure or partial closure. Partial closure is defined as closure only below the level of the ligated azygous, leaving the anastomosis exposed.
The complete mediastinal envelope closure led to a statistically significant decrease in anastomotic leaks (2% vs. 14.7%, p = 0.007); delayed gastric emptying (6.1% vs. 20.6%, p = 0.02); need for pyloric dilation (15.6% vs. 32.4%, p = 0.03); length of stay (2 days vs. 4 days, p <0.001); and 90-day hospital readmission (11.6% vs. 35.3%, p = 0.001).
In particular, patients with no or partial closure of the mediastinal envelope had 3.74 times increased odds of anastomotic leak (p = 0.007). Anastomotic leak is associated with increased length of stay, stricture formation, morbidity, and mortality, but consistently effective interventions are limited. One meta-analysis found only omentoplasty (relative risk [RR]: 22%) and early NG tube removal (within 2 days) or no NG tube (RR: 38%) reduce the risk of anastomotic leak.7 The techniques presented here may offer alternative solutions to reduce common and highly morbid complications for patients after esophagectomy.
In our MIE, we emphasize an efficient operation with deliberate surgical maneuvers. MIE offers several advantages over open esophagectomy, including reduced perioperative blood loss, respiratory infections, and length of stay with improved 1-year quality of life, but MIE also has a longer median operative duration in the literature by more than 30 minutes (MIE: 326 minutes vs. Open: 295 minutes).8 Extended operative and anesthesia time may be uniquely harmful to esophagectomy patients for numerous reasons, including increased blood loss, tenuous conduit perfusion, and protracted single-lung ventilation.
In the laparoscopic phase, in addition to omitting routine jejunostomy tube placement, we do not perform pyloric emptying procedures (i.e., pyloromyotomy, pyloroplasty, or BotoxÂ®). Pyloric drainage procedures in MIE have been associated with increased postesophagectomy symptoms and the need for subsequent pyloric dilation.9 Pyloric stenosis is addressed postoperatively as needed, but with the slender conduit and improved emptying due to mediastinal envelope closure, fewer patients require pyloric dilation.
In the thoracoscopic phase, we specifically reduce single-lung ventilation time by using two ENDOLOOP ties to secure the stapler anvil, instead of a purse-string suture with an Endo Stitchâ„˘ or laparoscopic needle driver. In another multicenter retrospective cohort study of 368 patients (in press), where approximately one-third of MIE patients had operative durations <4 hours (240 minutes), we found that prolonged operations were associated with increased postoperative complications, specifically respiratory infections and anastomotic leaks. The relationship between operative duration and MIE complications is likely bi-directional. We do not endorse extreme speed, but recommend that surgeons prioritize purposeful, streamlined operative techniques.
Careful perioperative management of esophagectomy patients is essential, particularly with regard to nutrition. Esophageal cancer has the highest median pre-diagnosis weight loss of all cancers2 and >10% reduction in body weight is common preoperatively.10 After esophagectomy, weight loss and malnutrition continues to be a central issue. With loss of the stomach reservoir and accompanying malabsorption, studies indicate that two-thirds of patients lose more than 10% of their preoperative bodyweight and 20% lose more than 20% by 6-months postesophagectomy.1
Enhanced recovery after surgery (ERAS) pathways are multidisciplinary in nature and, on our team, a dedicated outpatient RD is a key decision-maker.2 Ideally, esophagectomy is performed 6-8 weeks after the last radiotherapy session in our program, but the RD coordinates the timing of surgery with the team based on the patientâ€™s nutritional status as well. Surgery will be delayed if nutritional status is extremely poor.
Beginning 4 weeks before the anticipated operative date, the RD provides the patient with our MIE-focused diet book and meets with him or her weekly, providing tailored education, discussing nutritional goals, and monitoring their current weight. Following each appointment, the surgeon, physician assistant, and RD jointly evaluate the patientâ€™s surgical fitness in a virtual multidisciplinary conference.
Preoperatively, most patients can drink clear liquids until 2 hours before surgery. The surgical approach is as described earlier in this article. Postesophagectomy, patients recover on medical-surgical nursing floors without routine use of intensive care. Floor nurses and nursing supervisors are thoroughly educated on, and uphold the use of, our MIE ERAS pathway. Postoperatively, patients immediately are maintained with balanced crystalloid fluids infusing at 125 cc/hour and multimodal intravenous pain medication as needed.
Beginning on postoperative day (POD) 1, the inpatient RD meets with the patient daily until discharge. Consistent with recommended esophagectomy ERAS protocols, we remove the NGT and 28 French chest tube early on POD 1 on morning rounds with immediate oral initiation of a clear liquid diet along with a protein-dense, reduced-carbohydrate meal-replacement supplement. Although this supplement is not a clear liquid, it became a central component of our ERAS protocol after careful discussion with nutrition about the specific benefits for esophagectomy patients. It has the highest nutritional density and decreases dumping syndrome with 30 g of protein, 6 g of carbohydrates, and essential micronutrients.
We do not routinely perform an esophagram or obtain daily labs. Pain is minimized by minimally invasive incisions (see Figure 4, below), long-acting liposomal bupivacaine, multimodal oral pain medication, and early chest tube removal. Most patients are discharged on POD 2 with a Blake drain and multimodal pain control (acetaminophen, gabapentin, ibuprofen) to reduce opioid requirements.
Figure 4: Location of minimally invasive incisions
Upon discharge, the surgeon will call and/or text the patient daily for 10â€“14 days. Patients follow the MIE diet book provided preoperatively and resume their weekly virtual outpatient appointments with the team RD for a minimum of 4 weeks. On POD 5, patients are advanced to a full liquid diet, and on POD 9, patients are advanced to soft food with removal of the right pleural Blake drain in clinic. Finally, on POD 10, patients advance to regular diet.
Esophagectomy ERAS pathways vary widely. In comparison to our pathway, one highly regarded cancer center starts jejunostomy tube feeds on POD 2, removes chest tubes and NG tubes on PODs 3-4 in the absence of conduit stasis or dilatation concerns, initiates a clear liquid diet on POD 5, and generally discharges patients on POD 7 with removal of the jejunostomy tube 3 weeks after discharge.11
We believe our approach works well for several reasons. Intensive nutrition management for 8 weeks perioperatively not only improves short-term malnutrition and associated complications, but also provides lifelong education and nutrition management strategies for patients and their caregivers Intraluminal nutrients may induce localized blood flow despite autonomic denervation12, meaning early oral nutrition may stimulate greater blood flow to the conduit than jejunal tube feeds. One multicenter randomized controlled trial (NUTRIENT II) demonstrated comparable complication rates and improved long-term survival with oral intake beginning POD 1, while others have demonstrated reduced complications, including anastomotic leak.13,14 Critical care data indicate that oral nutrition should be prioritized over other nonoral enteral feeding.13 Meanwhile, jejunostomy tubes reduce quality of life, frequently malfunction, necessitating urgent outpatient management15 and only defer, rather than prevent, weight loss postesophagectomy.14
The consummate MIE pathway will provide superior oncologic outcomes, minimize complications, and provide the best quality of life. Even with optimal treatment, a minority of patients survive beyond 5-years postesophagectomy and one-third of patients with a complete pathologic response will recur in a median of 11.6 months.16 Five years of data on our operative and perioperative approach highlight good short- and long-term outcomes compared to those reported in the literature, including a median length of stay of 3 days (vs. 7 days), a low proportion of anastomotic leaks (2.1% vs. 10-14%)3,8, and 30-day readmissions (9.9% vs. 9%â€“15%) with a median survival of 4.6 years.17
When the life expectancy of most esophageal cancer patients is measured in months rather than years, each day matters. We must avoid the urge to practice defensively when there is no clear benefit. By closely approximating the original esophagus anatomy with mediastinal pleural envelope closure over a slender conduit and prioritizing early oral nutrition with close personalized monitoring and education in the perioperative setting, we are able to attain good outcomes and approach humanization of the esophagectomy.
Dr. Haley Tupper is a general surgery resident at the University of California, Los Angeles.
Dr. Jeffrey Velotta is a cardiothoracic surgeon and a surgical lead for the esophageal cancer program at Kaiser Permanente Northern California. He also is a clinical professor in the Department of Clinical Science at the Kaiser Permanente Bernard J. Tyson School of Medicine in Pasadena, CA, and clinical assistant professor of surgery at the University of California San Francisco School of Medicine.