The first, and arguably most frustrating, technical failure that will be discussed is a leak. This involves the extravasation of ingested food or gastrointestinal (GI) fluids out of the GI tract directly into the cervical neck area or mediastinum. Although, a leak may occur at any point along the tract where it was manipulated and freed from the surrounding structures. There are several common etiologies for leaks following surgery including true anastomotic leaks, leaks from an enterotomy, and gastric conduit necrosis. Leaks in general are particularly devastating as they have historically resulted in up to 35% mortality [1]. Although, with advances in leak surveillance and treatment including endoscopic stenting procedures and the ability to start and continue jejunal feeds or total parenteral nutrition (TPN) there has been a decline in the associated mortality to 3% [2].

Presentation of a leak ranges from repairable intraoperative identification to devastating postoperative sepsis. Excluding those that are found with routine intraoperative and postoperative surveillance, the majority of patients will present with fever (57%), leukocytosis (52%), pleural effusion (70%), and septicemia (70%). Less common signs include pneumothorax (35%), wound infection (13%), coughing bile (4%), and dysphagia (4%) [3]. These patients can become very ill fairly quickly and often require prolonged intensive care unit (ICU) stays and re-exploration procedures [4].

The causes for leaks following surgery are plentiful. The anastomosis between the esophagus and the stomach comes in several varieties including end to end and side to side approaches [5]. Both of which may be made using stapling devices or hand-sewn techniques and are susceptible to technical failures [6]. There is also extensive manipulation of the esophagus as it is freed from the surrounding structures, which places it in danger of an injury that may go unrecognized. Twenty-one percent of patients who experienced a hypotensive episode during surgery ended up with leaks versus 3% in patients who did not have such an episode [7]. Another potential source of leak is gastric tip or gastric conduit necrosis.

Evaluation of a leak depends on the stage and clinical status of the patient. As mentioned before, the earliest presentation is while the patient is still undergoing surgery when the anastomosis is tested with insufflation. The escape of air denotes a leak and at this time corrective measures can be undertaken to ensure it is closed prior to completing the surgery. Hence, intraoperative endoscopic evaluation is routinely done in some centers to evaluate for leaks [8]. Postoperative endoscopy is championed by some clinicians as a safe and reliable way to evaluate for leaks 1 week after surgery, with repeat endoscopy of patients with necrotic findings [9]. It is argued that the procedure is safe and is 100% sensitive for detecting leaks and can be performed on all patients, including those who remain intubated [10]. It is also routine at some centers for the anastomosis to be evaluated radiographically between 5 and 7 days following surgery with a contrast swallow where a leak will present as contrast extravasation during the study [8]. Although some would argue that the rates of false positive (4.7%) and false negative (5.4%) studies is too high for routine use [11]. Also, drains left in close proximity to the anastomosis and surgical field will give advanced warning of a leak by draining GI contents. This becomes especially apparent if a patient is given a swallow trial with a colorful liquid and the drainage begins to change color and/or increases.

Gastric conduit necrosis is a complication that involves the necrosis of part or all of the gastric tube that connects with the esophagus. The necrosis is most commonly noted at the tip that connects with the esophagus as it is the farthest removed from the preserved right gastroepiploic blood supply [12]. The presentation of gastric tip necrosis is extremely variable ranging from asymptomatic patients where it is only identified with endoscopy to persistent coffee ground nasogastric tube material, fevers, leukocytosis, an otherwise unexplained acidosis, and eventually sepsis.

Gastric tube necrosis seems to have several etiologies including improper alignment of the gastric conduit where the anastomosis is made under tension or a twist is present, creation of a tube that is too narrow, strangulation within the hiatus, failure to preserve the gastroepiploic vessels, and an anastomosis created in an area with poor blood supply [13–15].

The evaluation of gastric tip necrosis is done via an endoscopic approach where the necrotic mucosa can be directly visualized and assessed for leaks. Studies show that not all cases of identified gastric tip necrosis will lead to leaks, however they may warrant closer observation and follow up endoscopic procedures to monitor for progression [9, 16, 17]. Prompt diagnosis is important so proper treatment can be initiated.

There are several other GI related complications that arise from surgical failures that, while maybe not as devastating as a leak or necrosis, can significantly impact a patient’s life. Strictures plague up to 27% of patients [18]. Presentation of strictures range from subclinical, found only on endoscopic or radiographic evaluation, to symptomatic patients presenting with dysphagia. Strictures are related to the type of anastomosis performed with 40% of end-to-end (ETE) anastomosis having strictures, whereas 18% of end-to-side (ETS) anastomoses developed strictures [19]. It was once thought that end to side or side to side anastomoses were more prone to complications, however it is now generally accepted that they are safe and effective [20, 21]. Hand sewn anastomoses also appear to be correlated with a higher rate of stricture, 20.3%, than stapled anastomoses, 6.3% [22]. It is worth noting that late stricturing, greater than 1 year after surgery, are very often, if not always related to recurrent cancer [18].

Dysphagia is also a risk of MIE that can be found with or without the aforementioned stricture. There also seems to be no difference in rates of dysphagia between thoracic and cervical anastomoses [23]. Radiographic, manometry, and visual evaluations show several possible etiologies including poor coordination of the deglutitive response, decreased pharyngeal pressure, decrease in total laryngeal elevation, increased pharyngeal transit time, and hypertensive peristalsis [24–26]. It is possible that all patients have some degree of impairment of their swallowing mechanism following cervical anastomosis [27].

Delayed gastric emptying (DGE) is also a common issue found after surgeries that alter the structure and innervation of the stomach. In the case of esophagectomies between 10 and 50% of patients are affected [28]. It is thought that damage to the vagus nerve and alteration of the stomach’s ability to achieve pressures greater than the pylorus often contributes to this entity. Patients typically present with regurgitation, reflux, aspiration, early satiety, chest pressure, and halitosis. Evaluation can be done with gastric emptying studies. The technical failure associated with DGE is the inability to strike a balance between a gastric conduit that is wide enough to prevent gastric necrosis yet narrow enough to allow stomach pressures to overcome the pylorus [29–31].

Dumping syndrome (DS) is the complete opposite problem where hypertonic stomach contents are evacuated from the stomach into the intestines too quickly. This results in a patient who presents with a combination of diarrhea, bloating, abdominal pain, and assorted hypovolemic symptoms. The etiology of DS is similar to that of DGE in that denervation and alteration of the stomach architecture can cause food to progress too rapidly. One technical aspect of surgery known to result in DS is intraoperative or postoperative pyloroplasty which is employed at some centers to assist with symptoms of DGE. There is no consensus on performing pyloroplasty routinely during surgery as there is no significant increase in post-op DGE in patients who did not have an operative pyloroplasty along with their esophagectomy [32].

Reflux following esophagectomy is relatively common and clinically apparent in as many as 70% of surgical patients [33]. Patients suffering from reflux run the whole range from subclinical esophagitis, found only on endoscopy, to reflux that causes severe impairment in daily activities. This reflux is thought to be related to the loss of the lower esophageal sphincter, poor gastric emptying, disruption of the diaphragm, and diminished length of the esophagus. It has been shown on several occasions that the degree of esophagitis is not well correlated with symptoms; however, the degree of reflux is correlated with the severity of esophagitis noted on biopsy [34–36]. Evaluating reflux is challenging since not all patients with reflux will have symptoms but endoscopy with biopsy is the gold standard for surveillance and determining the extent of damage and epithelial cellular change. Esophageal pH monitoring and impedance along with a contrast radiograph are also options for evaluating patients for reflux [36].

Aspiration is a very common finding after esophagectomy and comes in two varieties, vocal and silent. Vocal aspiration is easier to diagnose as it presents with the usual clinical signs of aspiration including a cough [8]. Many, if not all patients have some degree of impaired swallowing that may lead to aspiration for at least a short time following an esophagectomy [38, 39]. From a surgical perspective, this is related to modifiable factors including extended operative time and vocal cord paralysis [40]. Evaluation of aspiration is done routinely with fluoroscopy as contrast can be seen entering the trachea and circumventing the epiglottis, bedside swallow evaluation by a speech pathologist, or direct visualization with an endoscope [39]. This is an important topic as it appears aspiration and the degree of severity is directly related to the development of pneumonia [41].

Pneumonia following esophagectomy is devastating with a 20% mortality rate and is considered to be the most lethal complication [37]. It is also highly correlated with short and long term mortality, reducing the 5 year overall survival rate by 26.7% [42]. Pneumonia presents in the typical fashion with cough, leukocytosis, fever, and is often evaluated with radiographs, blood and sputum cultures, and clinical correlation. The etiology of this complication is not fully understood however there are related technical issues that can be avoided. First and foremost it is important to remember that not all technical failures occur within the operating room. Thorough investigation with swallow evaluations, radiographic or bedside and endoscopy prior to resuming a diet can reduce the likelihood of pneumonia to 9% [43]. It has also been theorized that thoracic dissection in the prone position versus the left lateral decubitus position may reduce pneumonia due to partial inflation of the lung, but there has been no compelling or statistically significant evidence to date and the practice has not been widely adopted [44]. Finally, higher rates of blood loss and longer operating times also seem to be related to adverse pulmonary outcomes including pneumonia [38, 45].

Chylothorax is an uncommon but potentially lethal complication of esophagectomy, affecting between 1.9 and 3.8% of patients [3, 46]. Presentation involves dyspnea, with decreased breath sounds on the ipsilateral side, high chest tube output, hypovolemia, malnutrition, and even immunosuppression that may progress very rapidly to a sepsis picture [47]. With respect to the technical surgical failures as the cause, it is generally assumed that damage to either the thoracic duct or smaller lymphatic vessels that are divided and not properly sealed are at fault. It is important to note that there is evidence that no difference exists between transthoracic and transhiatal approaches [48]. Some researchers have been advocating the use of prophylactic thoracic duct ligation during the initial surgery as a way to prevent chylothorax formation and have reported a significant reduction [49]. The evaluation of a chylothorax involves monitoring chest tube output, greater than 400 milliliters per day is generally considered high output, although greater than 1 liter of output is not unheard of [3]. The fluid will typically be milky to brownish/greenish in color and laboratory tests will show higher than peripheral blood numbers of lymphocytes, a high concentration of triglycerides (> 1.24 mmol/L) and chylomicrons [50, 51].

Damage to the recurrent laryngeal nerve (RLN) is a major factor in post esophagectomy morbidity and mortality with rates reported as high as 50% in patients receiving cervical anastomoses [52]. Injuries are mostly unilateral (83%) with a propensity for the side of the cervical incision (83%) [21]. What makes this injury particularly dangerous is the association with aspiration pneumonia. Patients typically present with hoarseness or aphonia [21]. Recurrent laryngeal nerve damage is due to trauma to the nerve during surgery, including transection, stretching, or thermal injury. In minimally invasive esophagectomies with cervical anastomosis one surgical technique that has been shown to decrease the propensity for RLN injury is a thorascopic approach with dissection close to the esophagus when above the azygous vein [30]. Patients with suspected RLN injury are evaluated with direct visualization using a laryngoscope.