Treatments for Barrett’s Esophagus

Fig. 22.1

(a) Gross appearance of esophageal specimen with Barrett’s metaplasia (salmon-colored patches) with adjacent native squamous epithelium (white). (b) Endoscopic appearance of Barrett’s changes (pink patches) extending proximally from the EG junction

In addition to white light endoscopy , the use of narrow-band imaging has been investigated for enhanced detection of mucosal abnormalities that might indicate Barrett’s-associated dysplasia (Fig. 22.2). Narrow-band imaging filters white light to wavelengths specific for hemoglobin absorption, thus highlighting mucosal vasculature, and in this way has been shown in increase the detection of dysplasia [11–13].

../images/473262_1_En_22_Chapter/473262_1_En_22_Fig2_HTML.png — Fig. 22.2

Narrow-band imaging showing Barrett’s metaplasia (a) and Barrett’s esophagus with high-grade dysplasia (b)

Complicating an already complex range of diseases, histologic progression from no dysplasia to low-grade dysplasia to high-grade dysplasia to invasive esophageal adenocarcinoma is not necessarily stepwise and linear, with one study showing that half of patients who developed high-grade dysplasia or esophageal adenocarcinoma had only non-dysplastic Barrett’s metaplasia seen on previous biopsies [14]. This finding underlies the importance of adequate endoscopic inspection, surveillance protocols, and accurate histologic assessment by expert pathologists (Fig. 22.3), as patients who have adenocarcinoma detected through surveillance EGD demonstrate consistently improved survival relative to patients whose cancer was not detected through surveillance protocols [15–17].

../images/473262_1_En_22_Chapter/473262_1_En_22_Fig3_HTML.jpg — Fig. 22.3

(a) Histologic appearance of non-dysplastic Barrett epithelium showing mucinous glandular metaplasia with characteristic goblet cells. (b) Histologic appearance of Barrett’s esophagus with high-grade dysplasia, including cellular crowding due to abnormal proliferation and increased nucleus/cytoplasm ratio

Endoscopic Interventions

Endoscopic interventions include both excisional techniques, such as endoscopic mucosal resection and endoscopic submucosal dissection, and ablative techniques, including radiofrequency ablation and cryoablation. The primary indications for ablative therapies are for treatment of flat Barrett’s dysplasia or for treatment of areas of residual disease following endoscopic resection of visible dysplastic lesions. The indications for each endoscopic technique will be reviewed, with the standard approach favoring resection of any visually identified lesions, such as nodularity or ulcerations, and ablation of the remainder of the identified dysplastic Barrett’s segment.

Photodynamic Therapy (PDT)

The first endoscopic ablative therapy to gain widespread acceptance was photodynamic therapy. This was investigated in a randomized control trial where patients with high-grade dysplasia were randomized to receive either porfimer sodium PDT with omeprazole or omeprazole alone [18]. At 5 years, 77% of patients treated with PDT achieved eradication of high-grade dysplasia , while only 39% of patients receiving omeprazole alone had regression of disease. Fifteen per cent of patients who received PDT progressed to esophageal adenocarcinoma , while 29% of patients treated with only omeprazole progressed to development of cancer. Retrospective analysis of patients with high-grade dysplasia who underwent PDT or esophagectomy found similar rates of overall and cancer-free survival at 5 years [19]. Despite its early success, PDT has become less popular in the current era due to the cost of porfimer sodium, the prolonged posttreatment photosensitivity, and the rate of esophageal strictures following treatment, which has been reported to be as high as 50%.

Argon Plasma Coagulation (APC)

APC is another early endoscopic technique used for eradication of Barrett’s esophagus and involves coagulation of adjacent tissue by ionized argon gas injected at target areas from the tip of an endoluminal probe. Advantages of APC include the relatively low cost of argon gas and the no-contact technique, which might result in greater safety of the procedure. It has been successfully used for treatment of Barrett’s esophagus even in the absence of dysplasia and is delivered over the course of multiple sessions. In one series of 50 patients treated with APC and followed for 1 year, 34 patients had more than 90% eradication of Barrett’s lesions, while 16 patients had persistent Barrett’s changes after a median of four treatments [20]. Fifteen of the 34 patients who had macroscopically cleared their disease, however, had persistent buried glands under new squamous epithelium following treatment . At 1-year follow-up, 6 of these patients had persistent buried glands, while 2 out of 19 patients without prior buried glands developed subsquamous glands within 1 year. Given the risk for progression of these glands to high-grade dysplasia or adenocarcinoma, these patients require ongoing surveillance endoscopies to evaluate for progression of disease beneath the regenerated squamous cell lining, which is harder to detect than surface disease. Adverse effects following APC included posttreatment chest pain and transient dysphagia or odynophagia, but there were no posttreatment strictures seen at the time point of 1 year [20]. A separate study of 32 patients evaluated the long-term results of APC and demonstrated that two -thirds of patients who had complete eradication of Barrett’s changes following treatment maintained this result at long-term follow-up [21]. This study, however, did not demonstrate a protective effect of APC against development of adenocarcinoma in patients treated for non-dysplastic Barrett’s esophagus. This was attributed to the retrospective design of the study and inclusion of older APC devices and lower dosage of proton pump inhibitors; however, this result has supported the recommendation that patients with non-dysplastic Barrett’s esophagus should not be referred for endoscopic ablative therapies in most cases. The incidence of buried glands reported for APC in this study was 19%, which is markedly better than the reported rate for PDT, which has been documented as up to 51% [21]. However, this is still significantly higher than the reported rate for radiofrequency ablation (0.9%), which has become the gold standard treatment for eradication of dysplastic Barrett’s lesions.

Radiofrequency Ablation (RFA)

Radiofrequency ablation was developed for treatment of Barrett’s esophagus-associated dysplasia and has become standard therapy for treatment of symptomatic non-dysplastic and low-grade dysplastic Barrett’s esophagus, as well as high-grade dysplasia or carcinoma in situ. RFA gained popularity after it was tested against proton pump inhibitor therapy alone in a study called the AIM-Dysplasia trial [22]. In this study , patients with dysplastic Barrett’s changes were randomized to RFA with omeprazole or omeprazole only. 2.4% of patients treated with RFA with omeprazole progressed to cancer at 1 year, as compared with 19% of patients who progressed in the omeprazole only arm. A stricture rate of 7.6% was observed following RFA treatment.

RFA is performed by initially assessing the extent of Barrett’s metaplasia using white light endoscopy, as well as any visually identified lesions such as nodules or ulcerations [23]. Visible lesions are dealt with using EMR or other resection strategies, but in the absence of any identified lesions, the complete Barrett’s segment is treated with RFA. The RFA probe inserts into the esophagus adjacent to the endoscope and is comprised of a copper electrode sheet mounted on the surface of a balloon (Barrx 360 Express RFA balloon catheter, Medtronic Inc). It is positioned roughly 1 cm cranial to the proximal-most extent of the identified Barrett’s segment. The balloon is inflated, and good mucosal contact is confirmed, after which RF energy is deployed from the electrode across the surface of the balloon over approximately 1 second. The balloon automatically deflates after discharge of energy, and a circumferential burn is visible (Fig. 22.4a). The balloon is then advanced distally, and the process is repeated, avoiding overlap in segments, until the esophagogastric junction is reached. At this point, the catheter is removed, and the ablated segments are mechanically debrided using a transparent cap mounted on the tip of the endoscope (Fig. 22.4b), after which the full ablation is repeated for a second round. The total recommended energy delivered is 10 J/cm². Previous models of this catheter that are still commercially available may lack the automatic inflation of this device and therefore require pretreatment sizing of the esophageal diameter with an initial sizing balloon prior to deploying the RFA treatment balloon. Alternatively, rectangular ablation catheters can be positioned at the end of the endoscope through a hinge attached to a rubber cap that is placed on the tip of the endoscope (Barrx 90 RFA focal catheter, Barrx Ultra Long RFA focal catheter, Barrx 60 RFA focal catheter, Medtronic Inc) or can be placed through the working channel of the endoscope (Barrx Channel RFA endoscopic catheter, Medtronic Inc).

../images/473262_1_En_22_Chapter/473262_1_En_22_Fig4_HTML.jpg — Fig. 22.4

Appearance of esophageal mucosa following radiofrequency ablation using balloon applicator (a) and subsequent debridement of necrotic tissue (b)

RFA treatment has been associated with post-procedure chest discomfort, as well as a stricture rate of 6–11.8% [3, 10, 22]. Recurrence rates following RFA have been reported to be 8–10% in randomized control trials but as high as 26–33% in retrospective studies [10]. For this reason, active surveillance following RFA treatment is necessary, in addition to posttreatment use of proton pump inhibitors. Recurrent or persistent lesions can be treated with further ablative therapy or excisional techniques .

Cryoablation

In contrast to heat-based ablative techniques such as PDT, APC, or RFA, cryotherapy utilizes a cold-based technique to induce cellular necrosis and sloughing through tissue disruption caused by repeated freeze–thaw cycles. Cryotherapy can be applied to the esophagus through a spray, with either liquid carbon dioxide or liquid nitrogen (CryoSpray, CSA Medical, Baltimore, MD), which is applied through a low-flow continuous delivery system using a noncontact method [24]. Flow of liquid nitrogen across the cryoprobe is 4–6 L over 20 seconds, achieving a temperature of −196 °C. A separate decompression catheter is required to evacuate the gas due to the rapid expansion of liquid nitrogen. In contrast, the carbon dioxide spray catheter delivers 6–8 L of gas per minute, resulting in temperatures ranging from −70 to −78 °C. Either gas is applied in two 20-second application cycles or four 10-second cycles to induce a freeze-thaw cycle in the adjacent tissue. Recently, a delivery balloon that can be deployed through the working channel of the endoscope has been developed to allow focal ablation of an area roughly 2 cm² through release of nitrous oxide at −85 °C. The advantage of the cryoballoon is immediate venting of gas back through the balloon into the catheter. Cryotherapy has been shown to eradicate 81% of high-grade dysplasia and 91% of low-grade dysplasia in initial prospective studies [25]. Cryotherapy has been suggested as follow-up therapy for treatment-resistant disease following RFA, as its mechanism of crystallization followed by subsequent necrosis may allow for deeper penetration into tissues. In one study, 16 patients who had persistent dysplasia after three RFA treatments, progression of dysplasia while receiving RFA treatment, or treatment failure as reported by the endoscopist were treated with cryospray therapy using liquid nitrogen [26]. Seventy five percent of these patients achieved complete eradication of dysplasia, while 31% achieved complete eradication of intestinal metaplasia. However, a recent prospective single-center analysis using carbon dioxide cryopsray following endoscopic resection of any visible lesions was less promising, with complete eradication of intestinal metaplasia in only 11% of patients [27]. This failure of therapy in the majority of patients included in this study has been attributed to the use of carbon dioxide rather than liquid nitrogen, though the results were concerning enough to terminate the trial prematurely. Nevertheless, a follow-up retrospective study of 64 patients who underwent cryotherapy with a carbon dioxide cryospray, including 28 patients who had undergone prior PDT or RFA and 16 patients who had undergone prior EMR, showed complete eradication of Barrett’s metaplasia in 67% of patients [28]. Studies are currently underway to evaluate the cryoballoon focal ablation system, which has already passed safety and feasibility studies, with 100% of patients showing complete eradication of the treated Barrett’s areas in targeted trials [29, 30].

Complications of cryotherapy include chest pain and discomfort, which have been reported in 17.6% of patients [31]. Pain scores, however, seem to be less than those reported for RFA, though no direct comparative studies of the two modalities exist at present [24]. Stricture rate is reported to range from 3 to 9%, and perforation is a rare but reported event, which may be more of a concern in theory given the distention of the GI tract from gas released with use of the cryosprays. A durable response to cryosprays has been observed at 5 years, with retrospective studies reporting complete eradication of dysplasia in 88% of patients at 5 years and complete eradication of intestinal metaplasia in 75% of treated patients at 5 years [32].

The risk of buried metaplasia or subsquamous Barrett’s changes has been reported to be higher for cryotherapies relative to RFA. Ablative therapies are effective by causing necrosis and sloughing of the surface cells of the esophagus and subsequent replacement of these cells with normal native squamous cells lining the esophagus. However, if not all dysplastic cells are eradicated by ablative therapies, new squamous epithelium can be grown on top of residual dysplastic cells and allow sub-squamous high-grade dysplasia or adenocarcinoma to develop. Because these abnormal cells lie beneath normal-appearing squamous epithelium, they can evade visual surveillance as well as superficial biopsies and may not be detected until a later stage. While this can occur with any ablative therapy, the reported rate of sub-squamous metaplasia following RFA is 0.9% [33], while the reported rate following cryotherapy is as high as 9.1% [27, 30, 34–36].

Endoscopic Mucosal Resection (EMR) and Endoscopic Submucosal Dissection (ESD)

Endoscopic mucosal resection can be used to resect visible lesions, nodules, or ulcerations of the esophagus, short-segment Barrett’s esophagus with dysplasia, superficial adenocarcinoma (T1a), and esophageal squamous cell carcinoma. The EMR procedure may involve either the cap-assisted mucosectomy or the ligation-and-snare/multiband ligation technique. During cap-assisted mucosectomy, saline is injected into the submucosal space under the target lesion to elevate the mucosa. A snare is then used to surround the area and strangulate the base, and the lesion is then suctioned into a specialized cap on the tip of the endoscope. During multiband ligation, the target area is suctioned into a cap at the tip of the endoscope, and a rubber band is deployed around the base of the target tissue (Fig. 22.5). This tissue is then resected using a snare positioned below the base of the rubber band. EMR is highly effective at removing Barrett’s lesions; however, it cannot be used on lesions spanning over 50% of the circumference of the esophagus to avoid debilitating stricture formation. Additionally, the EMR technique can resect small lesions in their entirety, but larger lesions often require piecemeal resection, which can result in tissue distortion that affects histopathologic evaluation and can leave positive margins that require repeat resection or even follow-up esophagectomy for complete removal of malignant tissue. The advantage of EMR over ablative therapies is that histopathologic samples are sent to pathology and tissue can be evaluated for depth of invasion, while ablative therapy destroys the surface tissue to allow for reepithelialization with squamous epithelium but does not provide tissue samples for pathologic diagnosis. Following EMR, patients should remain on proton pump inhibitors to promote healing of ulcerations following the resection, as well as undergo repeat endoscopy 8 weeks after the procedure. EMR has been reported to effectively eradicate superficial neoplastic tissue in 91–98% of cases and eradicate the dysplastic Barrett’s segment in 80% of cases [37, 38]. It has therefore become the first-line treatment for superficial esophageal adenocarcinoma (T1a), followed by surveillance endoscopy to evaluate for recurrence or residual disease [39]. Some groups advocate for concurrent RFA ablation of the complete Barrett’s segment at the time of EMR of visible lesions [40], though these techniques are more commonly performed in a sequential fashion, with initial resection of visible dysplastic or neoplastic lesions followed by ablation of the residual Barrett’s segment. Reported complications following EMR include bleeding in 10% of cases [37, 41, 42], perforation in 3–7% of cases [43–45], and stricture formation in 17–37% of cases [46], though this is directly related to the length and circumference of the resected mucosal region. Strictures following EMR are managed with endoscopic dilatation.