Paraesophageal hiatal hernia is defined as the cephalad migration of the stomach and potentially other abdominal viscera away from their normal intraperitoneal locations into the mediastinum through an enlarged esophageal diaphragmatic hiatus. This defect occurs as a result of variable destruction of the normal attachments of the gastroesophageal junction (GEJ) and the barriers between the viscera of the abdomen and the thorax. It is estimated to occur in fewer than 1% of the population and is most likely an acquired defect given its greater frequency in the elderly population. Factors thought to contribute to the development of hiatal hernia include genetics, pregnancy, and factors that increase intraabdominal pressure, such as chronic cough, obesity, and repeated intraabdominal straining with physical activity, such as that associated with chronic constipation or childbirth.

Ninety-five percent of paraesophageal hiatal hernias are of the sliding type and are denoted type I. In type I hiatal hernia, the GEJ slides up into the lower mediastinum because of defect or deterioration of the phrenoesophageal ligament (Fig. 130-1). Type II–IV hiatal hernias account for the remaining 5%. Type II hiatal hernia involves a focal defect in the anterior aspect of the phrenoesophageal ligament and is very rare. The GEJ remains attached to the preaortic fascia and median arcuate ligament. Rather than the GEJ serving as the leading edge of the herniation, as in type I hernias, here, the gastric fundus serves as the leading edge of the herniation. As the hiatus defect enlarges, the stomach rolls into the chest along a longitudinal axis, producing organoaxial rotation.

Type III hiatal hernia is a combination of types I and II. Both the GEJ and the fundus herniate into the mediastinum. This is the most common form of paraesophageal hernia. Depending on the patient population and the method of determining the precise location of the GEJ, type III hiatal hernias comprise approximately 80–90% of all paraesophageal hernias. The hallmark of this disease is a foreshortened esophagus, which occurs as a result of the chronic reflux, in which inflammation, edema, and esophageal submucosal scarring cause esophageal remodeling. In addition, the adhesions associated with the hernia sac and the chronic position of the stomach and mediastinum create tension that "pulls" the distal esophagus deeper into the chest. Type IV hernias are very large. Here, other abdominal organs, such as the pancreas, spleen, omentum, and/or the small and/or large intestine, herniate with the stomach into the mediastinum.

It is difficult to determine the frequency of the complications that occur from paraesophageal hiatal hernias. According to many reports, most patients present without symptoms and have only incidental findings on radiologic examination or endoscopy performed for other reasons. However, these reports are misleading because most patients with type III hernias are revealed to be symptomatic when thoroughly and properly questioned. Consequently, the true prevalence of paraesophageal hiatal hernia is obscured, and it is difficult to make clear-cut recommendations for management. At one time, all patients with paraesophageal hiatal hernia were advised to undergo surgical repair. Recently, that policy has been questioned.^1–3 Based on the bulk of the literature, the presumed morbidity and mortality associated with hernia repair are greater than the risk associated with nonsurgical management. However, with the advent of minimally invasive techniques and better understanding of postoperative management, the morbidity and mortality have improved. Hence the best treatment approach to asymptomatic paraesophageal hiatal hernia remains a controversy.

SYMPTOMS AND SIGNS

Patients with paraesophageal hiatal hernia present most commonly with a "presumed" lack of symptoms. On specific questioning, however, over 90% of patients will reveal signs and symptoms indicative of hernia, which they have overlooked. The most common symptoms are related to obstruction and include resulting chest/abdominal pain, emesis, and dysphagia. Substernal or epigastric pain or heartburn is the most common symptom (60–70%). Other symptoms include discomfort particularly accentuated after meals with common radiation to the left,^4,5 regurgitation (50–70%), dysphagia (30–50%), nausea and retching not uncommonly accompanied by the inability to vomit (30%), anemia, usually iron-deficiency type (20–40%), and gastrointestinal bleeding (5–40%). Patients may develop acute symptoms from strangulation, obstruction, bleeding, and perforation. The inability to pass a nasogastric tube may be an indication of esophageal or gastric obstruction. Ulcerative esophagitis may occur in 20% and resulting stricture in 15%. It may be difficult and often unhelpful to perform manometry because of the altered anatomy. However, in patients who have been studied with manometry, as many as 30–60% had hypomotility, and more than 50% had hypotensive lower esophageal sphincter (see also Part 3 on esophageal motility and Part 4 on esophageal reflux disorders). These symptoms and signs can help to direct further clinical investigations.

Case series reported in the 1980s raised concern about the life-threatening complications of the paraesophageal hernia. Pearson and colleagues reported that 43 of 53 patients treated conservatively developed gastric volvulus, resulting in acute dysphagia, substernal or epigastric pain, and substernal distress.⁶ Bleeding occurred from ulceration or gastritis owing to trauma or ischemia from the compromised blood supply. In addition, Pearson and colleagues found incarceration of hernia contents in 24 of 53 patients. Acute dyspnea was observed to develop from the enlarging stomach, thereby compromising cardiopulmonary function. Another confirming report by Walther and colleagues found a 30% incidence of gastric volvulus, bleeding, and strangulation.⁷

The ominous findings of acute perforation and strangulation resulting in peritonitis, mediastinitis, and empyema increase the likelihood of death, morbidity, and long-term disability. It is usually presumed that these patients develop symptoms suddenly without warning, but this presumption is likely not the case.

DIAGNOSTIC EVALUATION

Radiologically, the most common presentation of paraesophageal hiatal hernias is the presence of retrocardiac shadows or masses on chest x-ray (Fig. 130-2) or chest CT scan and a frequent finding of air-fluid levels. These findings should prompt a contrast-enhanced upper gastrointestinal series, which provides the greatest information. The upper gastrointestinal series can identify the anatomy of the stomach and esophagus, the relationship of each to other structures, the location of the GEJ and its relationship to the diaphragmatic hiatus. Additionally, it provides data about esophageal motility, gastric emptying, esophageal stricture, presence of masses and ulcers of the esophagus and stomach, and presence of gastric volvulus. By esophageal contrast study, the finding of esophageal shortening has a positive predictive value of 37%.⁸ The presence of a gastric herniation measuring 5 cm or more in size is also an unreliable determinant of the shortened esophagus.^8,9 The upper gastrointestinal endoscopy is a useful procedure and should be performed routinely in patients with paraesophageal hernia to determine the location of the GEJ and evidence for reflux (such as esophagitis, ulceration, or stricture), gastritis, masses, and Barrett's esophagus. Manometry and pH catheters are difficult to place because of the distorted anatomy and thus are rarely indicated. Since many surgeons perform an antireflux procedure in conjunction with hiatal hernia repair (see Chap. 31), the use of pH determination as a preoperative requirement is unlikely necessary. Manometry has been used to assess for a shortened esophagus and has a positive predictive value of only 25%.⁸ In patients who have been carefully studied, approximately 30–60% were found to have disordered motility. Abnormal acid exposure was found in 40 of 44 patients studied.⁴ Of those who do not have preoperative evidence for gastroesophageal reflux and who do not have a fundoplication, 20% will develop reflux postoperatively.⁷ As a consequence, most surgeons perform an antireflux procedure with the hernia repair, and thus routine performance of a pH determination is unnecessary because it does not alter the surgical plan. Chest CT scan also provides information concerning the anatomy and relationship to other structures. Before surgery, it may be helpful to evaluate for other disease processes commonly found in the elderly, along with associated anatomic distortions. In addition, CT scanning provides a better three-dimensional sense of the diaphragmatic defect. Although it is not used routinely in patients with paraesophageal hiatal hernias, it may be helpful when more information is needed prior to the surgical repair.

TREATMENT OPTIONS

The indications for surgery are controversial. Before the mid-1990s, the mere diagnosis of paraesophageal hiatal hernia was sufficient indication to operate. This was based on the belief that nonsurgical management would result in a 30% risk for acute gastric dilatation, perforation, strangulation, and hemorrhage.^10,11 In one series of patients, adverse outcome occurred in as many as 50% of patients treated conservatively,¹² and the resulting death rate was reportedly as high as 27%.¹⁰ The principle of repairing all paraesophageal hernias was challenged in 1993 by Allen and colleagues, who found that symptoms rarely worsened in 23 patients who were managed nonsurgically for a median of approximately 6½ years, and only 1 patient died from aspiration.¹ Whether these findings were the result of improvements in medical management or of observational bias, it challenged the dogma of surgical repair. Moreover, despite the advanced age of most patients with hiatal hernia, improvements in intra- and postoperative management of thoracic surgery patients in the 1980s reduced the likelihood of postoperative death and complications, even after perforation.⁶ In 1999, without any data to support his claim, Floch stated in an editorial published in the Journal of Clinical Gastroenterology that acutely symptomatic patients should be treated electively with nasoenteric tube decompression and that the likelihood of catastrophic event was "no longer accurate."² This article has been quoted or misquoted by others as support for elective management as well. However, it does address the fact that even in cases of acute strangulation, without signs or symptoms of acute perforation, the patient can be decompressed and repaired electively rather than emergently. Further influencing the management of paraesophageal hernia, Stylopoulos and colleagues performed a thorough analysis of the prevailing literature as recently as 2003.³ Using a Markov Monte Carlo analytical model based on the assumption of complete reporting of all paraesophageal hernias and complications thereof, they identified 203 patients who underwent emergent surgical repair from the Nationwide Inpatient Sample of the Healthcare Cost and Utilization Project. Among these patients, they identified an additional "watchful waiting" group based on six small reports in the literature, five of which were used for their analysis. From this evaluation, they concluded that the "watchful waiting" approach was safe for asymptomatic patients over 65 years of age. Unfortunately, Stylopoulos and colleagues failed to address potentially significant reporting bias because the true prevalence of paraesophageal hernia is unknown, and the risk factors for poor outcome and/or debility or increased cost of care have not been defined. Simply stated, our knowledge of the natural course of these patients is insufficient to support such conclusions, and currently, we lack an evidence-based approach. The elderly with little comorbidity may benefit from surgical repair, if operated early. Moreover, the morbidity associated with worsening herniation and the potential for repeated hospitalizations and cost of care may be reduced with surgical management as opposed to continued observation. Hence the latter analysis raises questions but is not sufficient to set a standard of care.

Current recommendations call for repair of all paraesophageal hiatal hernias in symptomatic patients, with the caveat that all patients who are thoroughly questioned likely will be symptomatic. However, symptom palliation rather than surgical repair is preferable in patients who have significant comorbidity or for those whose likelihood of survival is less than 5–10 years. When a nonoperative approach is recommended, the patient, the referring physician, and the family/support system should be informed about the signs and symptoms of acute gastric volvulus, strangulation, bleeding, ulceration, perforation, mediastinitis, and peritonitis.

SURGICAL TECHNIQUE

The goals of repair of the paraesophageal hiatal hernia are to alleviate symptoms, minimize the morbidity and mortality associated with the surgical procedure, reduce the cost of care, and minimize the likelihood of recurrence. The recurrence rate is reportedly high if evaluated by esophagogram and/or esophagoscopy rather than symptoms. For patients who are systematically evaluated by esophageal contrast study, there is a 12–42% recurrence within 1–2 years, and this rate continues to increase with time, although most will recur within the first year.^13–16 Only 30–40% of patients who recur are symptomatic. The goals of repair are to (1) reduce the hernia contents into the abdomen, (2) remove the hernia sac, (3) obtain an intraabdominal esophageal length of 3–4 cm, (4) repair the hiatal defect, and (5) prevent reherniation and gastroesophageal reflux.

In most patients, the hernia contents can be easily surgically reduced into the abdomen. Adhesions and the attenuated gastric wall can make this maneuver more difficult, requiring special care to prevent injury. The intrathoracic peritoneal-lined hernia sac is surrounded by scar and should be excised completely. Presence of the sac and the surrounding scar tissue results in chronic tension on the GEJ and the sac contents preventing sufficient intraabdominal esophageal length (Fig. 130-3). Furthermore, without removing the sac, it is difficult to identify the precise location of the GEJ, an important component of performing an adequate repair. In clinical trials, failure to remove the sac increased the likelihood for recurrence¹⁷ and also resulted in the development of symptomatic mediastinal fluid collections that may require reoperation to remove.¹⁸ If necessary, small portions of the most cephalad portions of the sac may be retained safely.

The 3- to 4-cm intraabdominal length of the distal esophagus is necessary to achieve a tension-free hernia reduction and repair. To achieve this goal, circumferential dissection of the esophagus to the level of the aortic arch is advised. Occasionally, this requires division of the middle esophageal artery and left inferior bronchial artery. If this can be achieved adequately irrespective of approach, that is, laparoscopy, thoracotomy, or laparotomy, few patients (20% or fewer) will require other techniques to achieve adequate esophageal length.^19–22 As an alternative, some surgical groups advocate the routine use of an esophageal lengthening procedure, such as a Collis-Nissen gastroplasty. A potential but rare problem can arise if the acid-secreting gastric mucosa is placed at or above the hiatus repair, which may lead to postoperative esophagitis and/or dysphagia/motility disorder. Two laparoscopic Collis gastroplasty techniques have been described, one requiring an end-to-end stapler^23,24 and the other a wedge resection of the stomach on the greater curvature side just distal to the GEJ^24,25 (Fig. 130-4). Another means of accomplishing several centimeters of esophageal length is transection of the posterior branch of the vagus nerve or, if necessary, both branches. This procedure has been performed in patients without increasing the likelihood of gastric outlet obstruction or dumping syndrome.²⁶

For defects larger than 5-6 cm, the so-called giant hiatal hernias, reapproximating the diaphragmatic esophageal hiatus may be difficult. The surrounding diaphragm muscle and support fascia are attenuated and often do not hold sutures well. After surgical repair, the closure is under continuous dynamic stress. To provide a stable repair, different techniques have been used (Fig. 130-5). Both anterior and posterior cruroplasty have been performed, but most advocate posterior cruroplasty because the closure is achieved more easily without tension, and the esophageal length is more likely to be sufficient because the esophagus transcends the diaphragm at the apex of the diaphragmatic curvature rather than at the base (Fig. 130-6). The hiatus closure is performed with a 60F bougie inserted into the esophagus to aid identification (Fig. 130-7). Simple vertical or horizontal mattress sutures are placed in areas of visible tension and along the alternating muscle anatomic planes of the esophageal hiatus, in some cases unfurrowing the rolled-up hiatus muscular defect. Pledgets²⁷ and synthetic or biologic reinforcement has been advocated. Both polypropylene and polytetrafluoroethylene have been described, but there have been reports of erosions, ulcerations, and esophageal strictures.^28–30 However, using these tissue substitutes to reinforce the hiatus repair has reduced the recurrence rate by more than 60%.^31,32 More recently, biologic materials such as the ligamentum teres,³³ AlloDerm (LifeCell Corp., Branchburg, NJ),³⁴ and Surgisis (Cook Biotech, Wilson-Cook, Inc., West Lafayette, IN)^35–38 have been used in various ways to provide support. Use of biologic reinforcement does not appear to cause stricture or erosion.³⁸ In a multicenter prospective, randomized trial, Surgisis was used in 51 of 108 patients.³⁹ At 6 months, the recurrence rate in the Surgisis group was 9% compared with 24% in the primary closure or non-Surgisis groups.

Another component of the repair is gastric fixation, which can be performed by gastropexy or fundoplication with varying rates of success. The posterior gastropexy, or Hill repair (see Chap. 32), provides suture anchoring of the GEJ to the median arcuate ligament over the aorta. The anterior gastropexy can be accomplished with several suture ligatures to the anterior abdominal wall and also can be performed by placing a gastrostomy.⁴⁰Both techniques provide an anterior fixation to the abdominal wall, but the gastrostomy provides postoperative decompression of the stomach and, if necessary, a ready access for postoperative feeding in patients who many times are malnourished preoperatively. Gastropexy has been used as an adjunct to maintain the stomach in the abdomen but has a high rate of recurrence when used alone.^27,41 Pliability of the stomach results in reherniation. If gastropexy is used with an antireflux procedure, it puts stress on the wrap that may result in an ineffective antireflux mechanism. On the other hand, fundoplication has the advantages of fixing the stomach below the diaphragm, as well as creating an antireflux mechanism. Most paraesophageal hiatal hernia patients have a defective lower esophageal sphincter, and many complain of reflux.⁴² With the extensive surgical dissection required to remove the sac and lengthen the esophagus, the normal phrenoesophageal attachments are destroyed. This combination of factors suggests that performance of an antireflux procedure, such as a Nissen fundoplication (see Chap. 31), is necessary. Performing a Nissen fundoplication does not appear to impair esophageal body motility in these patients.⁴³ In patients who have evidence of esophageal dysmotility preoperatively, some surgeons advocate a partial fundoplication, such as a Toupet or Dor fundoplication. There is evidence, on the other hand, that partial fundoplication in comparison with full fundoplication worsens the likelihood of postoperative complications with impaired esophageal motility.⁴⁴

All these steps can be accomplished through different routes: thoracotomy, laparotomy or laparoscopy, or thoracoscopy. Thoracotomy has advantages in that it provides exposure to the distal two-thirds of the esophagus for sufficient esophageal mobilization and vagus nerve preservation, an excellent view of the hernia and the hernia contents for careful hernia reduction and resection of the sac, and a simple means to perform a Collis gastroplasty. The problems associated with thoracotomy include increased postoperative pain, increased length of stay, increased cost, and the potential for pulmonary complications. Laparotomy provides easy, wide access for the repair and has minimal morbidity and mortality, but the visibility may be compromised, especially in obese patients. Both approaches appear to have a similar mortality rate of 1–2% but a morbidity rate of 20% and similar recurrence rates. Although the first minimally invasive repairs were performed thoracoscopically,^45,46 the laparoscopic approach is simpler and provides excellent visibility, similar to the advantages of thoracotomy, but without the thoracotomy-associated morbidity. Hospital stays are shorter, there is less blood loss, and the return to preoperative function is faster in patients repaired by the laparoscopic approach.^47,48 Quality of life after laparoscopic repair appears to be superior to the open approach,⁴⁹ and long-term quality-of-life parameters appear to be similar to those of age-matched patients who have not had surgical repair.⁵⁰ In earlier series, recurrence rates appear to be higher with the laparoscopic approach,¹³ but with greater knowledge and experience, the recurrence rate should be less than or similar to those of the open approach.²⁶

An important characteristic of performing a successful endoscopic repair is the experience and knowledge of the surgeon performing the procedure in both the open surgical technique and the laparoscopic approach. Typically, five trocars are placed. One is placed at the right subcostal margin in the anterior axillary line for a liver retractor, and a second is placed supraumbilical in location for the videoscope. The remaining three are used for grasping and dissecting. After the hernia contents are taken down and the hernia sac is resected, a portion of the hernia sac is left attached to the distal esophagus and proximal stomach as a "handle" for manipulating the stomach and distal esophagus to avoid trauma from grasping the organs directly. Once esophageal mobilization is complete, the repair is performed in similar fashion to the open procedure. Reducing intraabdominal CO₂ below 15 mm Hg of pressure and placing the ports high on the abdominal wall may permit adequate reach into the mediastinum for esophageal mobilization and hernia sac removal. The short gastric arteries are taken with a Harmonic scalpel. Complete dissection of the anterior GEJ fat pad will assist in determining adequacy of esophageal length. The mean operating time is 2–4 hours, with conversion rates of approximately 2–10%.⁵¹

We have applied a robotic approach⁵² to the repair of the giant paraesophageal hiatal hernia. The robotic approach has several advantages: The three-dimensional surgeon-directed visibility allows for detailed characterization of the defect and permits dissection high into the mediastinum and precise placement of sutures during the repair. With greater experience, the lack of tactile feedback can be overcome visually (see Chap. 141).

POSTOPERATIVE AND OUTPATIENT MANAGEMENT

A nasogastric tube is used in most cases and left in place for the first 24 hours both to reduce the likelihood of immediate postoperative gastric distention and to provide access for a Gastrografin esophageal contrast study to evaluate for potential leaks, delayed gastric or esophageal emptying, or obstruction. If no leak abnormality is found, the nasogastric tube is removed, and a thin-barium swallow is performed. Performing both procedures is believed to provide greater sensitivity for identifying esophagogastric leak and a better anatomic assessment of the repair. If delayed emptying or obstruction is discovered, the nasogastric tube is left in place, and the study is repeated a few days later. With no anatomic defect noted, clear liquids can be initiated, and the patient can be advanced to a full liquid prior to discharge, if tolerated. Patients may remain on a full liquid diet for the next 1–2 weeks after discharge, and then their diet is advanced to a regular diet. The median hospital stay is 3–4 days.

Esophageal leak occurs in 1–3% of patients, and the procedure has an expected mortality rate of fewer than 1–2%. Excellent results are achieved in 85%, with only 3% of patients achieving a poor result. Approximately 6% of patients will require esophageal dilation subsequently, and approximately 2.5% will require reoperation for recurrent hiatal hernia. It is not clear whether a routine outpatient esophagram should be performed. Likely, the best strategy is to thoroughly evaluate patients postoperatively for any signs or symptoms of recurrence. Patients who have asymptomatic recurrences may be followed carefully.⁵³

EDITOR'S COMMENT

Laparoscopic surgery has revolutionized the field of hiatal hernia management and has increased the volume of cases referred to surgeons. There is still some concern about long-term recurrence some series have noted with the laparoscopic approach. Whether this is related to the learning curve or to fundamental shortcomings of this approach, i.e., too may "shortcuts," remains to be seen. Those who apply the same principles regardless of approach–open versus laparoscopic–such as good mobilization, hiatal reconstruction with buttressing, and tension free wraps with appropriate use of gastroplasty, have excellent results.

–LZ

REFERENCES