High-grade dysplasia is defined histologically as cytologic malignant changes that are confined to the epithelium and do not penetrate the basement membrane (Figure 3). High-grade dysplasia is thought to represent the last step in the neoplastic sequence prior to invasive cancer; it can exist as an isolated lesion or can occur in association with invasive cancer. It should therefore be regarded as both a precursor to, and a marker for, invasive cancer. The natural history of high-grade dysplasia is, in general, not well understood. The cumulative incidence of progression from high-grade dysplasia to esophageal cancer ranges from 16% to 59% in studies spanning 5 to 8 years of follow-up.[38,39]
Detection of high-grade dysplasia is a challenge. Presently, esophagoscopy employing the Seattle protocol (four-quadrant jumbo biopsies taken every 1 cm) remains the best method of detection. Once the diagnosis of high-grade dysplasia is established, confirmation of the diagnosis by an independent pathologist is performed due to the significant interobserver variation in assessing high-grade dysplasia biopsy specimens. Once the diagnosis has been confirmed, several management options exist, including endoscopic surveillance, mucosal ablation, and esophagectomy.
Not all patients with high-grade dysplasia have or will develop cancer. This has led several experts to recommend intensive endoscopic surveillance (every 3–6 months), with invasive therapies being withheld until there is definitive evidence of adenocarcinoma. Since there are no distinctive gross features of high-grade dysplasia at the time of endoscopy, endoscopists rely on random four-quadrant biopsy protocols of the affected segment. Unfortunately, by the time biopsy specimens reveal high-grade dysplasia, 40% to 60% of patients will already have an invasive malignancy.
Extensive work is underway to evaluate methods of enhancing the yield of surveillance protocols utilizing chromendoscopy, high-magnification endoscopy, optical coherence tomography, light-induced fluorescence endoscopy (''LIFE''), and narrow-band imaging. At this time, however, none of these modalities have been clearly established as appropriate for broad use in clinical practice. Existing data fail to clearly demonstrate that current intensive surveillance protocols are cost-effective or that they improve mortality from esophageal cancer.
Another option in the treatment of high-grade dysplasia is endoscopic ablation. The goals of endoscopic ablation include destruction of the abnormal Barrett's epithelium, while preserving the overall integrity of the esophagus, minimizing morbidity, and offering a better quality of life. Multiple techniques have been developed to selectively eradicate Barrett's mucosa in an effort to minimize the risk of progression to cancer. Among the most promising modalities are photodynamic therapy (PDT), radiofrequency ablation (RFA), and endoscopic mucosal resection (EMR).
• Photodynamic Therapy—Photodynamic therapy employs a photosensitizing drug that is absorbed and retained at higher concentrations in neoplastic tissue. Light of the proper wavelength (630 nm for porfimer sodium) delivered at the time of endoscopy to the targeted tissues produces an oxidative photochemical reaction that results in mucosal destruction. The endoscopic ablation of Barrett's esophagus with PDT has been evaluated in several studies.
Wang and coworkers from the Mayo Clinic reported their 2-year follow-up of 146 patients with Barrett's esophagus and pathology ranging from no dysplasia to cancer. At the time of last surveillance, nearly one-half the total population of patients had complete ablation of Barrett's esophagus while others had varying degrees of residual metaplasia. Only four patients (2.7%) went on to develop cancer.
Overholt and coworkers have presented the preliminary results of an international, multicenter, randomized, partially blinded trial of PDT and omeprazole(Drug information on omeprazole) (n = 138) vs omeprazole alone (n = 70) in 208 patients with Barrett's esophagus showing high-grade dysplasia. In the PDT/omeprazole group, 77% of patients achieved complete ablation of high-grade dysplasia, compared with 39% in the omeprazole group (P < .0001). In addition, 52% of PDT/omeprazole recipients achieved complete replacement of all Barrett's metaplasia/dysplasia compared with only 7% in the omeprazole group (P < .0001). Although cancer-free survival was substantially improved in the PDT/omeprazole group (P < .0014), 13% of these patients advanced to adenocarcinoma during a mean follow-up of 24.2 months. Strictures occurred in 36% of patients.
The University of Pittsburgh experience includes 50 high-risk patients with either high-grade dysplasia or localized esophageal cancer, with a mean follow-up of 28.1 months. Sixteen patients (32%) are alive with no evidence of disease, 30% are alive with residual or recurrent disease, and 38% have died with recurrent disease. The overall survival at 36 months was 31%. Strictures occurred in 42% of the patients. The intent-to-treat success rate was 38% in cases of high-grade dysplasia and 30% in focal carcinoma.
The primary long-term side effect of PDT, especially with porfimer sodium, is formation of esophageal strictures, which can occur in 20% to 53% of patients. Such strictures are usually responsive to endoscopic dilation. Other complications following PDT include cutaneous photosensitivity, chest pain, nausea, pleural effusions, Candida esophagitis, atrial fibrillation, and odynophagia. Esophageal perforation and tracheo-esophageal fistulas have been reported but are rare (incidence < 1% in most large studies).
The use of steroids has been postulated to reduce stricture formation in the setting of PDT, but Panjehpour and associates found no benefit in the use of oral steroids to prevent the formation of strictures. Another important finding in patients undergoing PDT is the presence of residual patches of subsquamous intestinal metaplasia, occurring in 5% to 50% of cases. Though the biologic significance of residual subsquamous Barrett's remains to be established, concern arises regarding the possibility of ongoing dysplastic progression in these hidden nests of metaplastic tissue.
• Radiofrequency Ablation—Radiofrequency ablation of Barrett's esophagus is also gaining increased popularity. The ablation system (Halo360 System, BARRX Medical Corporation, Inc, Sunnyvale, Calif) employs balloon-based ablation catheters positioned over a wire. A sizing balloon is initially deployed to measure the inner diameter of the targeted segment. The appropriately sized ablation balloon is then deployed, containing a 3-cm bipolar microelectrode consisting of 60 electrode rings. The electrode rings completely encircle the balloon and alternate in polarity, spaced 250 µm apart. Radiofrequency energy is then delivered to preset levels (10 J/cm2 × 2) at 300 watts in less than 1 second. The catheter can be repositioned to treat the entire length of Barrett's esophagus (Figure 4).
The largest published study to date prospectively evaluated RFA in treating 70 patients with Barrett's esophagus (no dysplasia). At a mean 12-month follow-up, metaplasia was completely eradicated in 70%, and partially eliminated in 25%. Adverse events, including chest discomfort and fever, were rare. The investigators noted no strictures or residual subsquamous metaplasia.
As experience with RFA is acquired, there is an increasing interest in using the procedure, given the promising early results, rare complications, and relative ease of deployment. Prospective, comparative studies will be necessary to elucidate the potential benefit of RFA in the treatment of Barrett's esophagus.
• Endoscopic Mucosal Resection—Endoscopic mucosal resection has also been established as a viable treatment option in patients with nodular or focal high-grade dysplasia. The technique involves raising the mucosal/submucosal target area by intramural saline and/or suction and subsequent snare resection (Figure 5).
The safety of this procedure was evaluated by Seewald and coworkers, who performed EMR on five patients with multifocal high-grade dysplasia or cancer and seven patients with high-grade dysplasia. Complete resections were achieved in all cases, and there were no recurrences in 9 months, although two patients had strictures that were treated with dilation. Giovannini and associates also achieved complete, circumferential resection in 18 of 21 patients with high-grade dysplasia or cancer. All were disease-free after 24 months, except for two local recurrences that were again treated with EMR.
May and coworkers evaluated the results of EMR in the treatment of 115 patients with high-grade dysplasia or early adenocarcinoma. Complete local remission was achieved in 98% of patients, with an associated 3-year survival rate of 88%. Interestingly, a 30% rate of metachronous lesions was observed during the follow-up period, and all but one were treated successfully with repeat EMR. This finding underscores the need for close endoscopic follow-up in patients undergoing local ablative approaches for Barrett's esophagus and associated dysplasia and/or focal intraepithelial carcinoma.
Buttar and associates combined EMR with PDT in treating 17 patients with superficial cancer. Importantly, cancer was seen at the margins of resection in 3 (18%) of the 17 patients. However, 16 (94%) patients achieved and remained in remission at a median follow-up of 13 months. EMR improved staging in 47% of the patients, and esophageal strictures were seen in 30% of the patients after EMR/PDT.
In another large study of 100 consecutive patients evaluating the efficacy of EMR in the treatment of early-stage adenocarcinoma, Ell and coworkers achieved a complete local remission in 99% at an average of 1.9 months. During a mean follow-up of 37.6 months, 11% of patients were found to have recurrent or metachronous carcinomas, all of which were treated successfully with repeat EMR. The estimated 5-year survival rate was 98% in this series.
EMR is now being performed with good success both focally and circumferentially for high-grade dysplasia and early cancer. Complication rates associated with EMR are as high as 30% to 50%, with bleeding being the most common event. Esophageal stricture (especially when EMR is combined with PDT) is a late complication, reported in up to 30% of cases. Perforation risk is less than 1%.
While mucosal ablation techniques are potentially less morbid then esophagectomy, their efficacy remains to be proven in controlled trials. The vast majority of available literature is derived from single-institution, retrospective series. All described ablative techniques have revealed a cohort of patients (2%–60%) that have residual subsquamous Barrett's epithelium. Subsquamous Barrett's epithelium theoretically remains at risk for progression to adenocarcinoma despite continued surveillance endoscopy. At this time, endoscopic ablative therapy for Barrett's esophagus should only be considered appropriate for patients with high-grade dysplasia who are unwilling or unable to undergo esophagectomy.
Esophagectomy is the only therapy that can eliminate the risk of progression of dysplasia to cancer. As mentioned above, up to two-thirds of patients with high-grade dysplasia are found to be harboring a cancer in the resected specimen. It is impossible to predict with any certainty who has a malignancy or who will develop one. Fortunately, high-grade dysplasia is highly curable with surgery, with 5-year survival rates exceeding 90%. However, operative mortality rates of 3% to 12% have been reported, and serious postoperative complication rates range from 30% to 50%.
Centers with extensive experience in esophageal resection have minimized the rates of morbidity and mortality, thus maximizing the benefit:risk ratio. Tremendous advances in laparoscopic surgery have made it possible to perform esophagectomy via minimally invasive techniques. Minimally invasive esophagectomy has been developed as a safe and feasible alternative to the traditional open techniques.[57,58] In our recently published experience with 222 minimally invasive esophagectomies, mortality was only 1.4%. Due to the mucosal localization of disease in Barrett's esophagus, minimally invasive esophagectomy for high-grade dysplasia represents an ideal approach for this lesion. This strategy may minimize the morbidity associated with esophageal resection, while providing definitive therapy for even long segments of Barrett's esophagus.
The Practice Parameters Committee of the American College of Gastroenterology currently recommends that patients with chronic reflux symptoms undergo screening endoscopy for Barrett's esophagus (Figure 6). Barrett's esophagus without evidence of dysplasia can be safely followed with endoscopy at 3-year intervals. The presence of low-grade dysplasia warrants yearly surveillance endoscopy with biopsy. Pathologically confirmed high-grade dysplasia is managed by esophageal resection, mucosal ablation with follow-up surveillance, or intensive endoscopic surveillance (every 3–6 months). Esophagectomy remains the gold-standard treatment for high-grade dysplasia.
As techniques and technology improve, it is likely that ablative strategies will become an important part of the armamentarium in the treatment of Barrett's esophagus. Prospective, controlled studies will be necessary to properly define the optimal role of ablative therapy in this setting.
Financial Disclosure: The authors have no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.