CHAPTER 36 Surgical Treatment of Benign Esophageal Diseases
The esophagus actively transports solids and liquids from the pharynx to the stomach. It has no digestive, absorptive, metabolic, or endocrine functions. A muscular tube subtended by two sphincters performs this rudimentary transfer task. Despite simplicity in esophageal function and form, surgical treatment of benign esophageal disorders is challenging. Few options are available to repair damaged sphincters; disorders of the esophageal body are rarely amenable to surgical correction. Often, progressive disease and failed surgical therapy result in a nonrepairable esophagus. The only treatment option is resection and replacement. Successful surgical therapy requires a sound understanding of esophageal anatomy, physiology, investigative techniques, and disease processes.
The esophagus is lined with stratified, nonkeratinizing squamous epithelium (Fig. 36-1), isolated from the remainder of the esophageal wall by a basement membrane. Immediately beneath the basement membrane is the lamina propria, a thin layer of loose connective tissue with a complex of collagen and elastic fibers. It contains a network of endothelium-lined channels, both capillaries and lymphatics. The muscularis mucosae supports the lamina propria and is composed of a longitudinal layer of smooth muscle. This continuous muscle layer pleats the inner layers of the esophagus into a series of folds that disappear with distension. The epithelium, lamina propria, and muscularis mucosae comprise the esophageal mucosa.
(The Cleveland Clinic Center for Medical Art and Photography. © 2009. All rights reserved.)
The submucosa is composed of connective tissue that contains a network of blood vessels and lymphatics. Elastic fibers and collagen combine to make this the strongest esophageal layer. Submucosal glands are mixed types, producing a combination of serous and mucous secretions. These submucosal glands are unique to the esophagus and allow differentiation of the esophagus from the stomach in instances of glandular epithelial metaplasia. Ducts from these glands pierce the mucosa to drain into the esophageal lumen.
The muscularis propria is the muscular sleeve that provides the propulsive force necessary for swallowing. There are two layers of muscle: an inner circular layer and an outer longitudinal layer. The proximal 4% to 5% of the esophagus is composed completely of striated muscle and the distal 54% to 62% completely of smooth muscle.1 Smooth muscle first appears in the anterior circular layer. The transition from striated to smooth muscle in the circular muscle layer is gradual, and the 50% point is approximately 5 cm from the cricopharyngeus muscle.
The cricopharyngeus (upper esophageal sphincter [UES]) is a continuous transverse band of muscle originating from the cricoid cartilage (Fig. 36-2). Superiorly, the muscle of the cricopharyngeus blends with the inferior pharyngeal constrictor muscle. A posterior defect, Killian’s triangle, is an inverted fan–shaped weakness in the inferior constrictor at the superior border of the cricopharyngeus. Inferiorly, the cricopharyngeus merges with the inner, circular layer of the muscularis propria. The longitudinal muscle layer of the muscularis propria originates from the lateral aspect of the cricoesophageal tendon. Posteriorly, these anterior and lateral components converge to meet at the midline. Thus, the proximal 1 to 2 cm of the posterior cervical esophagus is composed only of inner circular muscle, creating a potential for a mirror-image triangular area of weakness called Laimer’s triangle.
Contraction of the longitudinal muscle fibers of the esophageal body produces esophageal shortening. The inner circular muscle is arranged in incomplete rings producing a helical pattern. Muscle layers are equal and uniform in thickness until the distal 3 to 4 cm of the esophagus. Here, the inner circular layer thickens and divides into incomplete horizontal muscular clasps on the lesser curve aspect of the distal esophagus, and oblique fibers on the greater curve aspect. These become gastric sling fibers (see Fig. 36-2). Although no complete circular bands exist at the lower esophageal sphincter (LES), it is this area of rearranged circular fibers that corresponds to the high-pressure zone of the LES.
The esophagus lies in a bed of fat, neurovascular, and connective tissue and elastic fibers termed the adventitia. This layer of loose connective tissue surrounding the esophagus contains lymphatics and regional lymph nodes, blood vessels, and nerves. Unlike the stomach, small bowel, and colon, it has no serosa, except in its short abdominal segment.
Lymphatics begin as blind endothelium-lined saccules in the lamina propria just below the epithelium and basement membrane. Recent immunohistochemical study of the esophageal wall using lymphatic endothelial marker D2-40 has provided new insight into the lymphatic anatomy of the esophageal wall (see Fig. 36-1).2 There is a dense longitudinal plexus of lymphatic vessels in the lamina propria. Rare perforating lymphatics have been found draining into a sparse circumferential lymphatic network in the outer margin of the submucosa. Perforating lymphatics from the submucosal plexus, usually running with an artery and vein, penetrate the inner circular layer of the muscularis propria. Here, they drain into a circumferential intramuscular plexus that accompanies the artery, vein, and nerves of this space. Afferent lymphatics usually accompanied by an artery and a vein drain the intramuscular plexus into the lymphatic channels in the adventitia. No direct connection from the lamina propria network and the thoracic duct has been identified.2 Existence of direct routes from mural lymphatics to the thoracic duct, without a relay through regional lymphatics and lymph nodes, has been documented by many authors; however, the exact patterns and occurrence of these pathways are highly variable.3–5
The arterial supply of the esophagus is parasitic. It is derived from blood vessels supplying other organs in the neck, chest, and abdomen. Generally, these vessels divide at a distance from the esophagus and send small segmental branches to that segment of the esophagus. Esophageal blood supply has three principal sources. The superior and inferior thyroid arteries supply the cervical esophagus. The proximal and middle thoracic esophagus receives blood from branches of the bronchial arteries. The only dedicated esophageal arteries are one or two branches that arise from the anterior aspect of the aorta below the tracheal carina. In a third of autopsy specimens, no esophageal artery could be identified.6 These esophageal arteries directly supply the lower thoracic esophagus. The lower thoracic esophagus and abdominal esophagus receive arterial branches from the left gastric and, occasionally, the splenic arteries. The combination of a segmented arterial supply derived from multiple sources and a rich intramural vascular plexus ensures an excellent esophageal blood flow and permits extensive esophageal mobilization without esophageal arterial insufficiency or ischemia. Because esophageal arteries branch from larger arteries some distance from the esophagus, stripping of the esophagus from its bed during transhiatal (blunt) esophagectomy is possible without direct ligation of the esophageal arterial supply. Arterial spasm provides adequate hemostasis; thus, significant bleeding does not complicate this procedure.
Subepithelial esophageal venules drain into a substantial submucosal venous plexus that extends the length of the esophagus.7 There are venous connections between the lower thoracic and abdominal esophagus and the portal venous system. Venules then pierce the muscularis propria to drain into veins on the surface of the esophagus. Regional drainage is directed to the inferior thyroid and brachiocephalic veins in the neck, the azygos and hemiazygos veins in the chest, and the left gastric and splenic veins in the abdomen.
Both parasympathetic and sympathetic nerves innervate the esophagus. Branches of the vagus nerve supply parasympathetic fibers that are motor to the muscle coat and secretomotor to the submucosal glands. The cervical and thoracic sympathetic chain and the celiac plexus provide sympathetic fibers that promote contraction of sphincters and relaxation of the esophageal body muscle, increase peristaltic and glandular activity, and cause vasoconstriction. These fibers enter the esophageal wall with the blood supply and form fibers and ganglia within it. The myenteric (Auerbach’s) plexus is positioned between the longitudinal and circular layers of the muscularis propria, and it controls these muscles. The submucosal (Meissner’s) plexus controls the muscularis mucosae and submucosal glands.
The esophagus spans the lower neck, thoracic cavity, and upper abdomen (Fig. 36-3). The anatomy of the esophagus is best divided into fifths: cervical, upper thoracic, middle thoracic, lower thoracic, and abdominal esophagus. The anterior wall of the cervical esophagus is in intimate contact with the posterior membranous trachea. The recurrent laryngeal nerves course anteriorly and laterally in the tracheoesophageal groove. The carotid sheaths bind the cervical esophagus laterally. The posterior wall of the cervical esophagus lies on the vertebral bodies.
The thoracic esophagus occupies the posterior mediastinum and passes anteriorly to the vertebral bodies. The upper thoracic esophagus lies posteriorly to the trachea and is bound laterally by the mediastinal pleura. In its lower left aspect, it is sandwiched between the azygos vein on the right and the aortic arch on the left. The middle thoracic esophagus lies behind the pulmonary hilum and between the azygos vein and descending aorta. The lower thoracic esophagus has the same lateral and posterior boundaries, but it lies behind the pericardium. The thoracic duct is situated between the azygos vein and the descending thoracic aorta, and posteriorly and to the right of the lower and midthoracic esophagus. At approximately the level of the fourth thoracic vertebra, it crosses the midline to become a left-sided structure.
The abdominal esophagus is cradled in the muscular esophageal hiatus. The inferior vena cava is on the right posterolateral aspect, the abdominal aorta on the left posterolateral aspect. Superiorly, the left lateral segment of the liver overlies the esophagus and the esophagogastric junction.
Swallowing has three phases: oral, pharyngeal, and esophageal. The action of swallowing is voluntarily initiated and is followed by a cascade of involuntary muscle activities that propels the swallowed bolus aborally. The esophageal phase of swallowing commences with the relaxation of the UES during the initiation of pharyngeal contraction. Food is pushed by pharyngeal contraction, and its transit is facilitated by negative intrathoracic pressure. Duration of UES relaxation is between 0.5 and 1 second. After passage of the bolus, the UES contracts, reaching twice resting pressure. The primary peristaltic wave is a progressive contraction activated by voluntary swallowing. With an antegrade peristaltic contraction speed of 2 to 7 cm/sec, a wave carries the bolus into the stomach. The strength of the primary peristaltic contraction increases with propagation along the esophagus. In more than 90% of wet swallows, a primary wave normally follows. If impaction occurs, esophageal distension produces closure of the UES, and a secondary peristaltic wave begins at the site of obstruction and passes distally. Tertiary contractions are nonperistaltic contractions occurring spontaneously between swallows, and they are ineffective in antegrade bolus transit.
Resting pressure of the LES exceeds intragastric pressure and prevents reflux of gastric contents into the distal esophagus. Within 2 seconds of pharyngeal contraction, the LES relaxes to near intragastric pressure for 7 to 10 seconds. The LES then contracts to greater than resting pressure for 8 to 12 seconds before return to resting pressure.
The symptoms most commonly associated with esophageal diseases are heartburn, regurgitation, dysphagia, and odynophagia. Other symptoms that may be associated with esophageal disease are sore throat, hoarseness, cough, bad taste, globus, hiccup, aspiration, wheezing, chest pain, nausea, vomiting, choking, hematemesis, and melena. Symptom composition depends on the esophageal disorder and is discussed in various sections of this chapter. Physical examination of the esophagus is indirect and focuses on head and neck, thoracic, and abdominal findings.
Box 36-1 lists systemic diseases with esophageal manifestations. Some are discussed later in this chapter. These potential underlying disorders should be considered while obtaining the patient’s history and performing the physical examination.
Box 36–1 Systemic Diseases of the Esophagus
A three-phase study assessing mucosa, contour, and function of the esophagus is optimal.8 First, the mucosa is examined in the double-contrast phase, in which the patient, in the upright position, ingests high-density barium and CO2 tablets (Fig. 36-4). Next, esophageal function is assessed with the patient in the right anterior oblique (RAO) position and ingesting low-density barium in single swallows at 20- to 30-second intervals (Fig. 36-5). The examination is videotaped. The value of attempting to elicit reflux in this phase is questionable, because 20% of normal individuals have radiologic reflux.9 Barium tablets, or barium-coated marshmallows or solids, may demonstrate abnormalities not visualized by liquid barium studies. The final phase, the full-column technique, is performed with the patient in a semiprone RAO position and with low-density barium. Multiple quick swallows produce a column of barium that fully distends the esophagus. This optimizes imaging of the distal esophagus and can demonstrate small hiatal hernias, subtle strictures, or distal rings (Fig. 36-6). The esophagus is allowed to empty, and the remaining barium coating the esophageal wall provides a mucosal relief study, now rarely performed.
Figure 36–4 Barium esophagram: mucosa. Double-contrast phase of the barium esophagram provides mucosal definition. A, Patient with a hiatal hernia and peptic stricture; no significant ulceration is seen. B, Patient with a columnar cell–lined esophagus, distal peptic stricture, ulcers, and nodules.
Figure 36–5 Barium esophagram: function. Single swallows every 20 to 30 seconds with the patient in the right anterior oblique semiprone position assesses esophageal function. A, Patient with diffuse esophageal spasm (cork-screw esophagus). B, Patient with abnormal motility and a midthoracic diverticulum.
Figure 36–6 Barium esophagram: contour. The full-column phase of the barium esophagram fills and fully distends the esophagus, providing an opportunity to examine the esophageal contour. A, Patient with an obstructed esophagus due to a peptic stricture and associated nonreducible hiatal hernia. B, Patient with a Schatzki’s ring. C, Patient with achalasia.
A timed barium esophagogram is a simple test of esophageal emptying (Fig. 36-7). After ingestion of a premeasured amount of barium, usually 250 mL, spot films are taken at 1-, 2- and 5-minute intervals and, if necessary, at 10 minutes and 20 minutes after barium ingestion. This allows simple quantification of esophageal emptying and is useful for evaluating motility disorders and to follow therapy.10,11
Figure 36–7 Timed-barium esophagram. A, Before Heller’s myotomy, this patient was able to ingest only 70 of the 250 mL of barium requested. Height (barium-coated saliva not included) and width of the column are measured at 1, 2, and 5 minutes after ingestion. Note markedly delayed esophageal emptying and incomplete emptying at 5 minutes. B, After Heller’s myotomy, ingestion of 70 mL of barium resulted in trace barium in the esophagus at 1 and 2 minutes and clearance by 5 minutes.
Esophagoscopy is used to visually assess mucosal and structural esophageal abnormalities. Biopsies of epithelial abnormalities such as esophagitis, mucosal nodules, columnar cell–lined segments, and strictures are an integral part of flexible fiberoptic esophagoscopy. However, the biopsies are limited to the mucosa. Indirect evidence of deeper mural abnormalities or extraesophageal lesions may be appreciated by extrinsic compression or displacement of the overlying epithelium.
Esophageal manometry is most commonly performed in patients with esophageal dysphagia, in patients with chest pain after excluding cardiac disease, and prior to antireflux surgery. Water-perfused fine capillary tubes or solid-state microtransducers with three to eight pressure sensors at variable distances apart have traditionally been used. However, a catheter has been developed with 36 solid-state pressure sensors 1 cm apart that offers high-resolution manometry (HRM) from the pharynx to the proximal stomach (Manoscan, Sierra Scientific, Los Angeles). It makes accurate catheter placement easier and markedly shortens procedure length by eliminating the LES pull-through. A disadvantage is its higher cost. In addition, it has not been demonstrated that the new observations with HRM translate to better outcomes.
LES resting pressure and length, but most importantly completeness of LES relaxation, are measured (Table 36-1). HRM allows more accurate measurement of LES relaxation than traditional manometry by compensating for the LES movement with swallows. Esophageal body parameters to assess are antegrade esophageal peristalsis, contraction amplitude, and duration. HRM allows generation of a spatiotemporal plot, where segments of the peristaltic wave-front can be separately analyzed (Fig. 36-8).12 Direct comparison with traditional manometry shows that HRM is more accurate at predicting abnormal barium bolus transport, especially with mild or moderate esophageal dysmotility.13 Reclassification of esophageal motility disorders based on HRM has been suggested,12,14 but further reports, including outcome studies, are needed to confirm validity. UES measurements have not been found to be clinically useful.
Figure 36–8 A, High-resolution manometry (HRM) tracing derived from 36 pressure sites 1 cm apart. HRM tracing on the left is divided into anatomic segments from pharynx to stomach on the right. Pharynx is from 16 to 18 cm, and pharyngeal contraction (black arrow) occurs as the upper esophageal sphincter (UES) (18 to 20 cm) relaxes, after a 5-mL swallow, and the UESp (UES pressure) falls from 50 to 0 mm Hg. Striated muscle contraction wave begins as the UES recovers, and it advances from 21 to 24 cm, where a transition zone (24 to 26 cm) of lower pressure occurs. Smooth muscle peristaltic contraction wave extends from 26 to 44 cm, is antegrade (3 cm/sec [normal, <8]), and normal in amplitude (50 to 75 mm Hg [normal, <180]), duration, and intrabolus pressure. The lower esophageal sphincter (LES) extends from 44 to 47 cm, relaxes, just after a swallow occurs, from 15 mm Hg to near 0, and then overshoots to 30 mm Hg. B, Traditional esophageal manometry, with four sites 5 cm apart, shows antegrade contraction waves at 5, 10, and 15 cm above the LES, and normal LES relaxation.
Ambulatory pH monitoring detects and quantifies acid gastroesophageal reflux. Because this test is typically performed without acid-suppression medication, proton pump inhibitors should be discontinued for 1 week, histamine 2 (H2) blockers for 24 hours, and antacids for 8 hours. Transnasal monitoring is done for 24 hours with the thin pH catheter placed 5 cm above the LES, located by manometry. Patients are instructed to have a “typical day” regarding activity and eating. Because symptom correlation is an important component of this test, the patient presses a symptom button when a symptom is felt.
A pH below 4 has arbitrarily been chosen to define an acid reflux episode. The normal parameters for 24-hour pH monitoring, based on this reference, have been defined (Table 36-2). Total acid exposure time, expressed as a percentage of study time, is the best discriminator between normal and abnormal.15,16 Composite scores, such as the DeMeester score and the frequency-duration index, are no better than simple measured parameters in identifying abnormal reflux. The symptom index relates symptoms to reflux events and is calculated by dividing symptom episodes with reflux by the total number of symptom episodes multiplied by 100%; 50% is the optimum threshold.17
Wireless pH monitoring is a recent technological advance. The Bravo delivery system “pins” a 6 × 5.5 × 25-mm capsule to the esophagus 6 cm above the squamocolumnar junction, identified by endoscopy. Manometry is not required. Improved patient tolerance allows increased activities and improved food intake; moreover, monitoring is traditionally extended to 48 hours (Fig. 36-9).18 A 2008 American Gastroenterological Association Institute position paper states that wireless pH monitoring has superior sensitivity for detecting pathologic acid reflux.19 Drawbacks are cost, occasional premature detachment, and severe pain requiring endoscopic removal in less than 2% of cases. Like nasal pH monitoring, it is unable to detect nonacid reflux.
When an ionic fluid bolus traverses an electrode pair, impedance to current flow decreases. Impedance pairs placed on a pH catheter at multiple sites can detect retrograde fluid flow throughout the esophagus as acid reflux if the pH is less than 4, or nonacid reflux if pH is greater than 4.20 An expert panel concluded that impedance-pH monitoring when the patient is off proton pump inhibitor (PPI) therapy is the best test to detect all reflux episodes in a patient21 and thus is the best test for symptom association in patients with gastroesophageal reflux disease (GERD). However, it has limitations: a nasal catheter is necessary, and inaccuracies in current software require manual data correction of the study.
Impedance-pH monitoring when the patient is on PPI therapy has been used when suspected esophageal or extraesophageal GERD symptoms (or both) persist despite PPIs (Fig. 36-10). Small case series in patients with a positive symptom index have shown good results after fundoplication.22 However, we believe that long-term controlled studies, performed in centers with experience in impedance-pH monitoring, are needed to identify the role of such monitoring when only symptom association and nonacid reflux are found.
Figure 36–10 Simultaneous impedance and pH monitoring shows two nonacid liquid reflux episodes in a patient receiving acid suppression. The first episode shows retrograde liquid reflux to 9 cm above the lower esophageal sphincter (LES) (down arrows), and subsequent clearance after 2 minutes at all sites (up arrows). The second episode shows retrograde liquid reflux to the most proximal site, and cough occurs 40 seconds later.
Endoscopic ultrasound (EUS) provides definition of the esophageal wall and periesophageal tissue not possible by routine fiberoptic esophagoscopy. It is indispensable in evaluating abnormalities of the esophageal wall and in diagnosing nonmucosal esophageal tumors. Ultrasound endoscopes scan the wall with ultrasound waves from 7.5- and 12-MHz. The probes, which can be passed through the biopsy channel of flexible endoscopes, can be used to evaluate esophageal strictures that prevent passage of standard ultrasound equipment. The esophagus and periesophageal tissues are seen as five alternating layers of different echogenicities (Fig. 36-11). This examination also images periesophageal structures, including regional lymph nodes. Both the layer of origin and the ultrasound characteristics of a mass are critical for diagnosing benign esophageal tumors. Periesophageal masses and regional lymph nodes can also be studied. EUS-directed fine-needle aspiration (FNA) provides cytologic and pathologic assessment of esophageal tumors, periesophageal masses, and regional lymph nodes.
Figure 36–11 Esophageal wall is visualized as five alternating layers of differing echogenicity by esophageal ultrasound (EUS). The first (inner) layer is hyperechoic (white) and represents the superficial mucosa (epithelium and lamina propria). The second layer is hypoechoic (black) and represents the deep mucosa (muscularis mucosae). The third layer is hyperechoic and represents the submucosa. The fourth layer is hypoechoic and represents the muscularis propria. The fifth layer is hyperechoic and represents the paraesophageal tissue. The thickness of the layers as indicated by ultrasound does not equal the actual thickness of the anatomic layers.
The Bilitec 2000 fiberoptic spectrophotometer detects the presence of bilirubin, the principal component of bile, by absorption of the band of light (450 nm) characteristic for bilirubin.23 Because impedance-pH monitoring can detect nonacid reflux, the role of Bilitec in GERD is limited. Moreover, it is not widely available. However, measuring duodenogastroesophageal reflux may be useful in the patient with previous esophagogastric surgery and with suspected symptoms from bile reflux.
In impedance-manometry, multiple impedance electrode pairs on a manometry catheter allow transport of a swallowed bolus to be assessed as it traverses the esophagus (Fig. 36-12). Bolus transit can be compared with simultaneous esophageal peristalsis and contraction amplitude.24 Among 350 patients studied at a single center, all those who had achalasia and scleroderma had abnormal bolus transit, and 95% or more of those who had normal manometry, nutcracker esophagus, and LES dysfunction (high or low) had normal bolus transit. However, only 50% of patients with esophageal spasm and ineffective esophageal motility had normal bolus transit.25 Impedance-manometry may define different treatment strategies for patient subpopulations. For example, patients with abnormal rather than normal bolus transit may develop dysphagia after antireflux surgery, or patients with esophageal spasm and with poor as opposed to normal bolus transit may respond better to approaches improving esophageal emptying. Clinical studies with outcome data will assess if impedance-manometry improves patient management in these and other situations.
Figure 36–12 Simultaneous impedance and manometry showing normal bolus transit after each of three 5-mL saline swallows. Bolus entry occurs at each of four impedance sites as impedance decreases (see first vertical line in second swallow). Antegrade decrease in impedance in the four sites shows normal bolus advance through the esophagus. After bolus persistence with a low impedance value, a rise in impedance toward baseline occurs at each site from bolus clearing (see second vertical line in second swallow) due to simultaneous esophageal contraction at the same level. Antegrade increase in the four impedance sites shows normal bolus clearing from the esophagus. Normal lower esophageal sphincter relaxation and overshoot is seen.
Herniation of abdominal contents through the esophageal hiatus is a common occurrence. With provocative maneuvers that increase intra-abdominal pressure, 55% of patients undergoing barium esophagram were found to have herniation of the stomach into the chest.26 Symptoms are secondary to reflux, incarceration or strangulation of herniated organs, or compression of thoracic structures. There are four types of hiatal hernia, each with its own symptom presentation. Type I, or sliding hiatal hernia, is the most common (Fig. 36-13, and see Fig. 36-6A). Herniation of the esophagogastric junction into the posterior mediastinum occurs because of thinning and elongation of the phrenoesophageal ligament. There is no potential for incarceration. The majority of patients with type I hiatal hernias are asymptomatic. If symptoms occur, they are related to GERD. Type II, or rolling, hiatal hernias are uncommon (Fig. 36-14). They result from a defect in or isolated weakness of the phrenoesophageal ligament, allowing a portion of the stomach to herniate through the hiatus while the esophagogastric junction remains anchored in the abdomen. Symptoms of gastric obstruction, strangulation, anemia, and, less commonly, shortness of breath and arrhythmia result from gastric herniation through the hiatus and presence of the stomach in the chest. Type III, or mixed, hiatal hernias are the second most common type (Fig. 36-15). Patients may present with reflux or with symptoms of type II hernias, or with both. As type III hernias increase in size, there may be organoaxial volvulus with the potential for strangulation. In many patients, these hernias may be a progression of type I hernias.27 Type IV hiatal hernias contain the stomach and other abdominal contents such as colon, spleen, small bowel, and pancreas (Fig. 36-16). The term paraesophageal hernia is sometimes used to describe any type II, III, or IV hiatal hernia.
Figure 36–16 Barium esophagram (A) demonstrates a type III hiatal hernia with organoaxial rotation. Three days later, barium is seen in the colonic diverticulum on preoperative chest radiograph, posteroanterior view (B) and lateral view (C). Therefore, this is a type IV hiatal hernia.
Symptomatic hiatal hernias should be repaired. Repair of asymptomatic types II and III hernias is controversial. The potential for strangulation and gastric necrosis has been advocated as the prime reason to repair paraesophageal hernias in all patients, particularly because 50% mortality was initially reported when this complication occurred.28 However, strangulation is an uncommon occurrence without antecedent symptoms; therefore, this is an overestimate, and careful follow-up of asymptomatic patients is a viable alternative to repair in all patients.
Repair follows the principles of surgical management of GERD. Addition of a fundoplication is controversial, but it is definitely indicated if symptomatic GERD is present. Laparoscopic repairs have been reported to have an earlier and increased rate of failure.29 A meta-analysis showed this rate to be 25%, at variable follow-up periods.30 This high incidence of recurrence has prompted the use of a variety of mesh reinforcement of laparoscopic hiatal reconstruction. Although mesh has been reported to decrease recurrence at 6 months,31 the clinical significance of this finding32 and long-term durability33 are in question. In most patients, addition of a gastrostomy or gastropexy is not required. Compared with patients with GERD, patients with type III hiatal hernias are older and have more comorbidities. Adjusting for these factors demonstrates that patients with type III hernias are more likely to have pulmonary, thromboembolic, and bleeding complications in their postoperative course than those with type I hiatal hernias.34
GERD was defined at the recent Montreal consensus conference as “a condition which develops when the reflux of stomach contents causes troublesome symptoms and/or complications.”35 Typical symptoms of GERD are heartburn, regurgitation, and dysphagia. Extra-esophageal (or atypical) symptoms or disorders having an established association with GERD are cough, laryngitis, and asthma. Other atypical symptoms or disorders (pulmonary fibrosis, pharyngitis, otitis media, and sinusitis) are often linked to reflux and are proposed associations, but data are insufficient to establish causation.35 It is important that abdominal pain, gas, and bloating not be misinterpreted as GERD symptoms.
Complicated GERD includes reflux esophagitis, esophageal stricture, and Barrett’s esophagus. The mucosal response to acid may produce intestinal metaplasia (Barrett’s esophagus), and damage to the submucosa and muscularis propria can result in a short esophagus (see Figs. 36-4 and 36-6A).Two feared complications of GERD are peptic stricture and the development of high-grade dysplasia or intramucosal cancer in Barrett’s esophagus.
The LES and diaphragmatic hiatal mechanism are the major components of the reflux barrier.36,37 A hiatal hernia is seen in 50% to 90% of GERD patients.38–41 However, reflux events are most commonly caused by transient LES relaxations, although reflux over a low basal LES is also important.42 Poor acid clearance from peristaltic dysfunction and delayed gastric emptying contribute to prolonged acid exposure in some patients.
The mainstay of therapy for patients with GERD is medical management. PPIs can heal esophagitis in more than 90% of patients; randomized studies do not demonstrate a superiority of surgery over medical therapy.43–45 Atypical symptoms or disorders respond less predictably to PPIs, especially in the absence of typical symptoms. Controlled trials of PPIs showing benefit in patients with laryngitis or asthma have been predominantly in those having concomitant typical symptoms. However, a recent controlled trial in patients with chronic posterior laryngitis symptoms and endoscopic findings, but no typical GERD symptoms, found no benefit with esomeprazole.46
Surgical management of uncomplicated GERD should be considered in patients who are PPI intolerant and who have typical GERD symptoms responsive to PPI, and persistent typical GERD symptoms despite PPI, especially volume regurgitation.47 The following preoperative GERD evaluation is suggested. First, endoscopy should be performed to assess for presence of esophagitis, Barrett’s mucosa, stricture, hiatal hernia, and alternative upper gastrointestinal diagnoses such as ulcer disease. Second, manometry should be performed to locate and measure LES, to exclude another diagnosis such as esophageal spasm, and to assess peristalsis. Third, a barium esophagram should be performed to assess the anatomy of the esophagus and esophagogastric junction, mucosal changes, and esophageal function. Fourth, ambulatory pH monitoring off therapy should be performed to confirm excessive acid exposure. The presence of typical symptoms that respond to PPI therapy, and of abnormal acid exposure as determined by pH monitoring, are reliable predictors of successful surgical treatment of GERD.48,49 In patients with suspected gastric drainage abnormalities, a nuclear medicine gastric emptying study is required.
Surgical management is less likely to be effective in patients with persistent atypical symptoms or disorders in the absence of typical GERD symptoms. Thus, antireflux surgery primarily for atypical symptoms or disorders should be reserved for patients with concomitant typical GERD symptoms, and in whom causality of reflux for the atypical symptom or disorder has been established to the greatest degree possible during the preoperative evaluation just described. Although small case series have shown good results for antireflux surgery in patients with persistent cough on PPIs and reflux by impedance-pH monitoring,50 long-term controlled studies are needed to validate this approach.
Complicated GERD is first managed with aggressive medical therapy; however, peptic esophageal stricture and chronic Barrett’s ulcer may require surgery. The columnar cell–lined esophagus is not in itself an indication for surgery. It is debatable that surgery can reverse the changes of columnar lining. Partial reversal, although interesting, is not a compelling argument for surgical correction of GERD.51 Effective reflux control may reduce the incidence of malignant degeneration in the columnar cell–lined esophagus. However, failure of antireflux repairs is frequently seen in patients with adenocarcinoma of the esophagus. Patients with a columnar cell–lined esophagus have the most disordered physiology, largest hernias, and most disrupted hiatal mechanisms.52 Therefore, either prevention of or halting malignant degeneration in a columnar-lined esophagus is not an indication for surgery. If a patient with a columnar cell–lined esophagus has antireflux surgery, the need for endoscopic surveillance is not eliminated. Thus, the patient with a columnar cell–lined esophagus has the same indications for surgery as the patient with a squamous cell–lined esophagus.
The principles of antireflux surgery are restoration of the intra-abdominal length of esophagus, reconstruction of the esophageal hiatus, and reinforcement of the LES.53 These can be accomplished by a number of approaches and techniques. Possible approaches, in descending order of morbidity, are thoracoabdominal, thoracotomy, laparotomy, and laparoscopy. facility with all of these approaches is necessary.
Restoration of the intra-abdominal esophagus requires reducing the hiatal hernia and mobilizing the esophagus. This portion of the operation necessitates recognition of the short esophagus.54 Failure to lengthen the short esophagus by extensive esophageal mobilization to the aortic arch or by addition of a Collis gastroplasty will result in a repair under tension and early failure. The short esophagus should be suspected by history of a stricture or previous dilation or findings of a long segment of columnar cell–lined esophagus, a large type I hiatal hernia (>4 cm), a type III hiatal hernia, or failure of the hernia to reduce below the diaphragm on upright barium esophagram.
The importance of reconstruction of the hiatus cannot be underestimated. It plays a role in reflux prevention equal to that of the LES.55 The recent addition of mesh reinforcement of the hiatal closure ignores the past history of antireflux surgery and the dynamic nature of this structure. Complete mobilization of the hiatal crura and careful removal of the hernia sac will allow primary suture reconstruction of the hiatus. Failure to reconstruct the hiatus is a common cause of failure of laparoscopic antireflux surgery.56 The principal reason for reherniation after an otherwise adequate repair is tension on the repair, either because the short esophagus was ignored or because it was not recognized, and not an unbuttressed reconstruction of the hiatus.
The final step in the surgical correction of GERD is reinforcing the LES by constructing a fundoplication, which may be either total or partial. The Nissen fundoplication, a 360-degree total fundoplication, completely encircles the esophagus. Partial fundoplication is typically 270 degrees and is either anteriorly placed (Belsey Mark IV repair), or posteriorly placed (Toupet repair). Theoretically, the tradeoff is between better reflux control with a total fundoplication and less dysphagia with a partial fundoplication. Some surgeons tailor the fundoplication to suit the peristaltic activity of the esophageal body.57 High-resolution manometry should allow a more precise cutoff at which the severity of peristaltic dysfunction requires partial fundoplication. Until then, we consider total fundoplication to be appropriate for all except the patient with an aperistaltic esophagus.15,58–61 The use of a partial fundoplication for complicated GERD is associated with an increased incidence of failed repair.62 Dysphagia after surgery for GERD is usually transient with a properly constructed fundoplication. Prolonged dysphagia is indicative of a malformed fundoplication, either long, tight, or twisted. Return of normal esophageal motility with resolution of GERD after fundoplication is hypothetically possible. It is more likely that the amplitude of the peristaltic wave will increase than that failure of propagation will be corrected. Therefore, any fundoplication should be constructed under the assumption that the peristaltic abnormalities will not improve. Finally, the potential for postprandial symptoms of gas bloat and early satiety must be mentioned to any patient being considered for fundoplication.
An often ignored but essential part of the physical examination is measuring and recording weight and height and calculating body mass index (BMI). Overweight (BMI, 25 to 29) and obese (BMI, 30 to 34) GERD patients should be counseled on weight loss and encouraged to reach their ideal weight before elective surgery. Because obesity and GERD are interrelated,63,64 successful and sustained weight loss may eliminate the need for surgery. Although there is disagreement about the impact of obesity on the outcome of antireflux surgery,65–70 the health benefits of weight loss in severely (BMI, 35 to 39) and morbidly (BMI, ≥40) obese patients with GERD should make weight loss surgery the operation of choice in these patients.
Antireflux operations are not infinitely durable. It is important that all patients are instructed to avoid activities that excessively increase intra-abdominal pressure and to maintain their ideal weight.71 Unrealistic patient expectations and widespread application of laparoscopic fundoplication without careful patient selection in low-volume centers by inexperienced surgical teams have produced poor results and have had a negative impact on this operative approach.72–74