41: Surgery and Endoscopic Interventions for Barrett Esophagus

Barrett esophagus (BE) is a condition in which the stratified squamous epithelium of the esophagus is replaced by intestinal columnar epithelium, a process called intestinal metaplasia (IM). Dr. Norman Barrett, an Australian thoracic surgeon, characterized the findings in an article that he published in 1950.¹ In 1953, the association was made between BE and gastroesophageal reflux disease (GERD) by the English thoracic surgeon Dr. Philip Allison.² The etiology of BE is accepted to be due to chronic exposure of the esophageal epithelium to gastroduodenal refluxate. The importance of BE is its potential to progress and develop into adenocarcinoma. The association between BE and esophageal adenocarcinoma was established in a case report in 1975 by Dr. Alan Farnsworth, a cardiac surgeon, in which he identified adenocarcinoma in a segment of BE.³

Multiple factors have been identified that lead to an increased risk of developing BE. These include age over 50 years, male gender, white race, elevated body mass index (BMI), intra-abdominal distribution of body fat, hiatal hernia, and chronic GERD.⁴ BE is diagnosed by endoscopic biopsy and histologic examination which identifies the IM (Fig. 41-1).

Progression of BE to cancer is recognized as a continuum through dysplasia to cancer. As such, the histologic examination of this disease is often classified into BE without dysplasia, BE with low-grade dysplasia (LGD), and BE with high-grade dysplasia (HGD). Progressive severity of dysplasia leads to a higher risk for the development of esophageal cancer. BE without dysplasia has a 0.5% per year risk of developing into esophageal cancer, whereas BE with LGD has a 1.4% per year risk and BE with HGD has a 6% per year risk.⁵ It is this strong association between BE and esophageal adenocarcinoma that has driven the aggressive surveillance and treatment of BE.

The goal of surveillance is to reduce mortality of a disease by early diagnosis. In the case of BE, the goal of surveillance is to reduce the mortality due to esophageal cancer. The risk of developing esophageal cancer in the general Western population is low; therefore, routine endoscopic screening has not been recommended. For patients who have risk factors for BE as described earlier, there is some data supporting screening endoscopy. Therefore, the American College of Gastroenterology recommends that screening endoscopy in this patient population should be individualized.⁶

For patients already diagnosed with BE, it is a common practice to perform endoscopic surveillance to search for signs of progression to cancer. Although this makes intuitive sense, there is a lack of randomized clinical trials to identify the true effectiveness of endoscopic surveillance in patients with BE. There have, however, been population studies that have shown the practice to lead to earlier detection and associated improved survival rates of esophageal adenocarcinoma in patients with BE who had endoscopy prior to diagnosis of cancer.⁷ Therefore, the current recommendations by the American Gastroenterological Association (AGA) are as follows: for patients with BE without dysplasia, endoscopic surveillance should be performed every 3 to 5 years. For patients with BE and LGD, endoscopic surveillance is recommended every 6 to 12 months. For patients with BE and HGD who do not undergo intervention, surveillance is recommended every 3 months.⁶ It should be noted that in order for endoscopic surveillance to be effective, adequate surveillance involves taking four quadrant biopsies every 1 to 2 cm along the affected length of esophagus in addition to biopsies of any irregular appearing mucosa (Table 41-1).

The goals for treatment of BE are (1) to decrease or eliminate gastroduodenal reflux into the esophagus, primarily to improve the associated symptoms of GERD and potentially to limit progression or cause regression of BE and (2) to directly eliminate Barrett metaplasia and dysplasia (via resection or ablation), thereby decreasing the risk for development of esophageal adenocarcinoma. The methods by which to achieve these goals are continuously evolving and in certain clinical scenarios, still remain somewhat controversial.

Control of Reflux

Medical Therapy

There are several ways to reduce and eliminate gastroduodenal reflux. Medical therapy, in the form of proton pump inhibitors, has been shown to be very effective for neutralizing gastric acid and controlling acid reflux symptoms. In addition, there are several studies that have shown that medical therapy can cause modest regression of BE. Regression is often defined in several ways, including decrease in the length of BE, down-grading of Barrett dysplasia, and complete elimination of IM. The regression rates have been reported to be from 19% to 41% depending on the definition of regression. Practitioners should be aware, however, that there can be progression of BE while on medical therapy. Reported rates of progression to dysplasia range from 1.8% up to 31%. Similarly, rates of progression from BE to esophageal adenocarcinoma have been reported to be from 1.6% to 7.4% while on medical therapy. Therefore, acid and symptom control with proton pump inhibitors alone should be used in conjunction with endoscopic surveillance to monitor for regression or progression of disease.

Antireflux Surgery

Antireflux surgery has been offered to patients who have failed medical treatment, usually defined as breakthrough symptoms, or progression or continuation of esophageal damage due to reflux. The technical details of the various types of antireflux operations are discussed in detail in other chapters. Laparoscopic Nissen fundoplication is the most common operation performed, but all operations described for the treatment of GERD may be used in patients with BE.

The use of antireflux surgery to change the natural history of BE is controversial. While there are no randomized controlled trials that show superiority of laparoscopic antireflux surgery over medical therapy for causing regression of BE or preventing the development of esophageal adenocarcinoma, there are some retrospective studies that suggest there may be a benefit of surgery.⁸ In a recent review of the literature, we found that on average, 11.1% of patients who underwent medical therapy progressed to dysplasia compared to only 3.4% of patients who had antireflux surgery.⁹ Similarly, 4.1% of patients undergoing medical therapy for BE progressed to adenocarcinoma compared to only 0.7% of patients who had antireflux surgery. Although these are selective case studies and few were established to compare medical and surgical therapy, the trends do suggest a possible benefit from antireflux surgery. Despite the improved symptomatic responses to surgery compared to medical therapy and the apparent decreased rates of regression of BE and progression to adenocarcinoma, the regression is still incomplete and progression to adenocarcinoma still occurs.¹⁰ Therefore, surveillance endoscopy should still remain as an important component of treatment for BE after antireflux surgery (Table 41-2).

Elimination of Barrett Metaplasia and Dysplasia

Once dysplasia develops, the risk of cancer increases exponentially, therefore, more aggressive treatments are recommended. Not very long ago, esophagectomy was recommended as the only treatment of choice in patients with BE and HGD and still appears in the Society of Thoracic Surgeons guidelines.¹¹ In the last decade, however, advanced endoscopic tools have been developed that can provide less invasive ways to adequately treat BE and have made esophagectomy rarely necessary.

Endoscopic Ablation

Endoscopic ablation therapy for BE was first reported in 1992. Laser therapy was studied early on, yet remains largely experimental, given its ineffectiveness in treating large areas of disease, associated perforation rate, and need for multiple treatments to achieve elimination. Multipolar electrocoagulation (MPEC) involves application of a 50-watt probe to an area of metaplasia until a coagulum forms. Studies have shown high success rates for eliminating nondysplastic BE ranging from 89% to 100%, but also have associated high complication rates. Dysphagia occurred in approximately 19% of patients and odynophagia occurred in approximately 16% of patients undergoing MPEC. For this reason, MPEC is not routinely used for treatment of BE.¹² Argon plasma coagulation (APC) is a cauterization tool that uses ionized argon gas to transmit a constant current to a superficial layer of mucosa. Results have shown 89% to 100% elimination of Barrett mucosa, but also a high 3% to 11% recurrence rate after treatment. This is thought to be due to the relatively low depth (2–3 mm) of treatment. Although the penetration depth prevents complications such as perforations from occurring, there have been reports of “buried dysplasia/neoplasia.” Given better alternative endoscopic ablation technologies on the market, APC is not frequently performed.¹² Photodynamic therapy (PDT) involves administration of a systemic light-sensitizing agent (porfimer sodium or 5-aminolevulinic acid) followed by endoscopic application of diffusing fiber optics against the target tissue to create necrosis. PDT has been compared with APC in a randomized trial. The result showed only 50% complete eradication of BE compared to 97% for APC.¹³ In addition, a higher stricture rate was seen with PDT ranging from 30% overall to 50% after receiving two treatments¹² (Fig. 41-2).

Radiofrequency ablation (RFA) has been shown to be an effective method to destroy dysplastic Barrett's epithelium, and is currently the most common method to do so. A randomized controlled multicenter clinical trial recently showed that ablation can both reduce progression and permanently eliminate Barrett's metaplasia.¹⁴ In this study, there was 90.5% complete eradication of dysplasia in patients with LGD compared to 22.7% of patients who had a sham procedure. For patients with HGD, 81% of patients had complete eradication compared to 19.0% in the sham group. Patients who had RFA also had less disease progression (3.6% vs. 16.3%) and fewer cancers (1.2% vs. 9.3%) compared to sham procedures. It is also a relatively safe procedure with one multicenter prospective study that reported no strictures 2.5 years after having focal ablation.¹⁵ Although there is an apparent high success rate, several limitations exist. Large surface areas of BE and circumferential BE are difficult to treat with RFA compared with focal BE because of the potential for postablation strictures. In addition, some reports have suggested that Barrett tissue that is deeper than the treatment depth of RFA may become “buried” and overgrown with normal squamous epithelium, thus harboring a deeper focus of dysplastic cells that cannot be accessed with surveillance biopsies. Although this is rare and has not been seen as commonly compared to APC, it is still a concern with this treatment modality (Fig. 41-3).

Cryoablation is a newer technology originally introduced in 1997, which has undergone some recent refinements that holds some promise for ablation of Barrett tissue. This technique involves using liquid nitrogen to freeze the tissue resulting in intracellular destruction while preserving the extracellular matrix, and may allow for deeper, submucosal ablation. A recent study showed 97% complete eradication of HGD, 87% had eradication of all dysplasia, and 57% complete resolution of BE.¹⁶ Randomized clinical trials are currently in progress to evaluate its efficacy compared to other ablative therapies.

Endoscopic Ablation for Nondysplastic BE

Ablation as an alternative for surveillance may be considered for nondysplastic BE. This practice is controversial, but the argument often used is the same one that is used for polypectomy for colonic adenomas—it is a lesion that has some propensity for progressing to cancer; therefore, if the lesion can be safely eradicated, then the risk of cancer can be modified. A multicenter cohort study that followed patients with a diagnosis of nondysplastic BE found that 0.5% per patient-year progressed to esophageal adenocarcinoma.¹⁷ Given the relatively low risk of complications related to RFA and the potential sampling error related to surveillance, ablation may be presented as a reasonable alternative to surveillance.¹⁵ In addition, several studies have reported complete clinical response (no residual IM) with RFA for nondysplastic BE.¹⁸

Endoscopic Mucosal Resection

Endoscopic mucosal resection (EMR) is a technique that provides both diagnostic and therapeutic benefits for patients with BE. The EMR technique is described in detail in Chapter 173. By taking full mucosal resections of esophageal mucosa, depth and extent of metaplasia and dysplasia can be accurately assessed. This is a clear benefit over ablative therapies in which histopathologic specimens are not obtainable. In patients who undergo EMR for HGD or early cancer, depth of invasion and surgical staging can be performed. In patients who have early-stage esophageal cancer that is confined to the mucosa, EMR has been shown to be an acceptable curative tool, given the low rate of lymph node metastases in these lesions. EMR does bear the risk of causing strictures in up to 50% of patients in addition to causing bleeding or perforation. Therefore, patients with long-segment BE (>3 cm in length) or multifocal BE spanning across a larger distance are typically not amenable to EMR, given the very high likelihood of developing strictures.

EMR in conjunction with RFA has been shown to have excellent results while maintaining a lower complication profile. EMR with RFA was shown to completely eliminate BE with HGD or intramucosal adenocarcinoma in excess of 94% of patients.¹² A recent multicenter trial demonstrated complete eradication of BE in 96% of patients who underwent EMR with RFA, but had only a stricture rate of 14% compared to a stricture rate of 88% for patients who underwent widespread EMR for eradication of BE.¹⁹ This high success rate in combination with accurate histopathologic staging of BE has made EMR followed by RFA the primary method of treatment in patients who have HGD or intramucosal adenocarcinoma (Fig. 41-4).

Surgical Resection

With improvement in endoscopic therapies such as EMR for BE, there became fewer indications for surgical resection. The presence of esophageal adenocarcinoma with invasion beyond the mucosa is currently an indication for esophagectomy, with or without neoadjuvant or adjuvant therapy depending on final staging. For patients without invasive cancer, esophagectomy is becoming reserved for patients with specific characteristics of BE, who are also physically able to tolerate this potentially highly morbid procedure. There are several techniques for performing esophagectomy and that are discussed in other chapters. We will discuss specific scenarios when esophagectomy may be indicated for BE.

Long-Segment BE with High-Grade Dysplasia

Long-segment BE is defined as involvement of a segment of esophagus that is greater than 3 cm in length. It has been shown that long segment BE has a higher risk for metachronous neoplasms. For this reason, surveillance is difficult for patients with long-segment BE. In addition, the increased segment length is associated with a higher complication rate, thus endoscopic treatment is less ideal. To eliminate a long segment, ablation or EMR must be performed over a long longitudinal distance. Therefore, the stricture rate already associated with endoscopic techniques becomes magnified.⁴ For this reason, esophagectomy should be considered for patients with long-segment BE with HGD and who are young or healthy enough to tolerate this operation, especially if the HGD is multifocal.

Multifocal Barrett Esophagus with High-Grade Dysplasia

Multifocal BE, like long-segment BE, has been shown to have a higher risk for metachronous neoplasms. It also presents an endoscopic challenge for all the same reasons. Accurate surveillance is technically challenging because the multifocality of the disease may result in underdiagnosing involved segments of BE. The rate of stricture with endoscopic therapies is mirrored as in long-segment BE; the rates are increased, given the multiple sites need for EMR and ablation. Therefore, esophagectomy should also be considered for patients with multifocal BE with HGD, barring any physical reasons for which the patient would be unable to undergo surgery.

Nodular BE

Nodular has been shown to have a higher incidence of concurrent esophageal cancer. The depth of penetration of nodular Barrett also makes it difficult to adequately treat with EMR, ablation, and even a combination of both. These techniques all have elevated risks as they are applied to deeper and deeper layers of tissue. This does not preclude endoscopic therapy, however. Given the higher cancer rate and difficulty to completely eradicate nodular with endoscopic techniques, esophagectomy should be considered in these patients, especially for recurrent esophageal nodules in areas of resection or ablation.

Treatment of BE has evolved significantly in the last 15 years. Surveillance for BE is still debated, though it is agreed upon that once Barrett's has been identified, surveillance be continued in a systematic fashion to identify progression to dysplasia and to detect early esophageal adenocarcinoma. The role of antireflux surgery for controlling reflux-associated symptoms, and decreasing disease progression remains widely debated, but should be considered primarily for controlling symptoms.

Barrett metaplasia without dysplasia or neoplasia can be treated with both medical therapy and antireflux surgery. Both methods have shown similar rates of disease regression. There is some evidence to support antireflux surgery to limitation of progression of BE to dysplasia and neoplasia, though there are no definitive randomized clinically conducted trials to support this (nor are there likely to be any).

Endoscopic therapy has become the treatment of choice for with HGD over the last decade. These new endoscopic technologies have entered the field and has replaced esophagectomy as effective and less morbid treatment strategies. The current treatment of choice that has the most data to support its effectiveness is EMR followed by RFA to control all of the abnormal epithelium.

Although becoming increasingly rare, there are still certain unique clinical scenarios in which esophagectomy for BE is indicated. These typically are young patients with multifocal, long-segment BE with HGD, and nodular BE.

Barrett esophagus or esophagitis is a complication of reflux disease and arguably an indication for both antireflux surgery and life-long surveillance. There is more than one type of BE, each with a different propensity for developing cancer and different long-term sequelae. Treatment therefore must be individualized and longer-term studies are needed to better inform the medical community as to the best treatments. In the meantime, patients with BE need to be educated about the disease, its risks, and the therapeutic options. Esophageal surgeons should be familiar with current therapies for BE so they can provide the best options for their patients.