Endoscopic Therapy Using Radiofrequency Ablation for Esophageal Dysplasia and Carcinoma in Barrett’s Esophagus




Radiofrequency ablation (RFA) is a novel and promising treatment modality for treatment of Barrett’s esophagus (BE) with high-grade dysplasia or early carcinoma. RFA can be used as a single-modality therapy for flat-type mucosa or as a supplementary therapy after endoscopic resection of visible abnormalities. The treatment protocol consists of initial circumferential ablation using a balloon-based electrode, followed by focal ablation of residual Barrett’s epithelium. RFA is less frequently associated with stenosis and buried glandular mucosa as are other ablation techniques and has shown to be safe and effective in the treatment of patients with BE and early neoplasia. In this article, the technical background, current clinical experience, and future prospects of RFA are evaluated.


Barrett’s esophagus (BE) is a premalignant condition in which the normal squamous cell lining of the esophagus is replaced by columnar epithelium containing specialized intestinal metaplasia (IM), defined by the presence of goblet cells. The most important risk factor for BE is chronic gastroesophageal reflux disease (GERD). BE is found in approximately 10% of patients undergoing endoscopy for chronic GERD symptoms and therefore it is hypothesized that the transition from squamous mucosa into mucus-secreting columnar epithelium is an adaptation of esophageal mucosa caused by the erosive effect of the refluxate. Via a gradual process from no dysplasia to low-grade (LGD) and high-grade dysplasia (HGD), BE can develop into adenocarcinoma. The estimated annual incidence of adenocarcinoma in BE is 0.5%. When esophageal cancer enters a symptomatic stage, disease is usually locally advanced and has a poor prognosis with a 5-year survival of approximately 20%. For patients with HGD or early cancer, previously surgical resection was the gold standard. Surgical esophagectomy is, however, associated with a mortality rate of 3% to 5% and serious complications occur in 40% to 50% of patients. In the subset of patients with “early neoplasia”, defined as HGD or intramucosal cancer, lymph node involvement is rare: 0% for HGD and up to 2% for intramucosal cancer. As a result, endoscopic treatment of early neoplastic lesions is a viable alternative to surgery for these patients.


Endoscopic treatment modalities


Endoscopic treatment comprises endoscopic resection (ER) and endoscopic ablation techniques. The most important advantage of ER above ablation therapy is that ER results in a resection specimen for histopathological assessment, contributing to optimal staging and patient selection. Although ER is a safe and effective treatment modality, with 5-year disease-specific survival of 95%, metachronous lesions arising in the residual BE segment can be found after ER in 30% of patients.


To eliminate the malignant potential after ER, eradication of the residual BE epithelium is advocated. To accomplish this, treatment protocols combining ER and ablation therapy have been developed. Photodynamic therapy (PDT) using amino levulinic acid (ALA) or porfimer sodium (Ps) and argon plasma coagulation (APC) are ablation techniques that have been administered for this purpose, but both techniques have disappointingly high rates of recurrent early neoplasia and residual IM. In both PDT and APC, ablation depth is not controlled and may vary, resulting in residual BE that still carries preexistent oncogenetic abnormalities. Sometimes residual BE glands become hidden underneath neosquamous epithelium (ie, “buried Barrett’s glands”), and some fear that these may progress to dysplasia or cancer without being detected endoscopically. Another complication from porfimer sodium (Ps-PDT) and APC is a high stenosis rate, which may be a result of submucosal scarring owing to uncontrolled ablation depth. Finally, Ps-PDT is associated with significant photosensitivity for a period of up to 6 weeks after treatment.


An alternative to ablation therapy is the complete removal of the BE by radical endoscopic resection (RER), generally performed in multiple treatment sessions. Small-sized single-center studies have reported low recurrence rates of early neoplasia after RER varying from 0% to 9% after up to 28 months of follow-up, and newly formed squamous epithelium after RER has been shown free of oncogenetic abnormalities. However, symptomatic stenosis develops in up to 56% of patients, requiring multiple dilations sessions.


The latest treatment modality for complete removal of BE is radiofrequency ablation (RFA). This novel ablation technique has shown excellent results for selected patients with nondysplastic BE, LGD, and HGD with or without prior ER with a follow-up of almost 2 years. Regarding the safety and efficacy profile, RFA compares favorably to other ablation techniques such as PDT and APC.




Radiofrequency ablation: the HALO system


The HALO system (BÂRRX Medical, Sunnyvale, CA) ( Fig. 1 ) for RFA is designed to control tissue penetration depth of the RF energy of 0.5 mm, thereby realizing a uniform ablation depth that is operator independent, owing to the use of bipolar electrodes and the automated delivery of a preset amount of RF energy. The HALO system consists of a circumferential and a focal ablation device for stepwise ablation of the BE epithelium. For larger areas of BE, circumferential ablation is performed using the HALO 360 -device, a balloon-based catheter with spindle-shaped electrodes covering a length of 3 mm. The HALO 360 -balloon is available in several diameters: 18, 22, 25, 28, 31, and 34 mm. Ablation using the HALO 360 -balloon is performed with an energy density of 12 J/cm 2 and 40 Watt/cm 2 . For focal ablation of residual BE epithelium, the endoscope-mounted HALO 90 -device is developed. The HALO 90 -device is equipped with an articulating surface containing an electrode array (13 mm wide × 20 mm long) to target small areas of BE with an energy density of 15 J/cm 2 and a power of 40 W/cm 2 .




Fig. 1


The HALO system for stepwise circumferential and focal radiofrequency ablation (RFA). ( Upper left ) The HALO 360 generator with integrated pressure:volume system, used to inflate the sizing and ablation catheters, to calculate esophageal inner diameter, and to deliver radiofrequency energy to the ablation catheter. ( Upper right ) The HALO 360+ ablation catheter. ( Lower left ) The HALO 90 energy generator for delivery of radiofrequency energy. ( Lower right ) The HALO 90 catheter fitted on the tip of an endoscope, without impairing the endoscopic view or function. ( Reproduced from www.endosurgery.eu ; with permission. Courtesy of ER and RFA Training Program, Amsterdam, The Netherlands; with permission.)




Radiofrequency ablation: the HALO system


The HALO system (BÂRRX Medical, Sunnyvale, CA) ( Fig. 1 ) for RFA is designed to control tissue penetration depth of the RF energy of 0.5 mm, thereby realizing a uniform ablation depth that is operator independent, owing to the use of bipolar electrodes and the automated delivery of a preset amount of RF energy. The HALO system consists of a circumferential and a focal ablation device for stepwise ablation of the BE epithelium. For larger areas of BE, circumferential ablation is performed using the HALO 360 -device, a balloon-based catheter with spindle-shaped electrodes covering a length of 3 mm. The HALO 360 -balloon is available in several diameters: 18, 22, 25, 28, 31, and 34 mm. Ablation using the HALO 360 -balloon is performed with an energy density of 12 J/cm 2 and 40 Watt/cm 2 . For focal ablation of residual BE epithelium, the endoscope-mounted HALO 90 -device is developed. The HALO 90 -device is equipped with an articulating surface containing an electrode array (13 mm wide × 20 mm long) to target small areas of BE with an energy density of 15 J/cm 2 and a power of 40 W/cm 2 .




Fig. 1


The HALO system for stepwise circumferential and focal radiofrequency ablation (RFA). ( Upper left ) The HALO 360 generator with integrated pressure:volume system, used to inflate the sizing and ablation catheters, to calculate esophageal inner diameter, and to deliver radiofrequency energy to the ablation catheter. ( Upper right ) The HALO 360+ ablation catheter. ( Lower left ) The HALO 90 energy generator for delivery of radiofrequency energy. ( Lower right ) The HALO 90 catheter fitted on the tip of an endoscope, without impairing the endoscopic view or function. ( Reproduced from www.endosurgery.eu ; with permission. Courtesy of ER and RFA Training Program, Amsterdam, The Netherlands; with permission.)




Patient selection for endoscopic management of early neoplasia


To correctly select the subgroup of patients suitable for RFA treatment, a thorough work-up and staging protocol is mandatory. In general, patients with early neoplasia in BE are considered potential candidates for endoscopic treatment when the lesion is limited in size (<2 cm), there are no signs of deep submucosal invasion, and no suspicious lymph nodes are found on endoscopic ultrasound (EUS). The tumor characteristics are of importance in staging, as the risk for lymph node involvement increases with increasing infiltration depth, poor differentiation grade, and lymphatic/vascular-invasive tumor growth.




Imaging endoscopy


Optimal endoscopic treatment of patients with early BE neoplasia starts with at least one high-resolution endoscopy by an experienced endoscopist. During inspection, visible abnormalities are classified according to the Paris classification. The Prague-CM-classification is used to describe the BE segment, including the length of the circumferential segment (C), and the maximal extent of the BE segment (M). Subsequently, targeted biopsies are obtained from visible abnormalities, followed by four-quadrant biopsies of every 1 to 2 cm of the BE segment (Seattle protocol).




Staging procedures: endoscopic ultrasound and computerized tomography


Patients with early neoplasia in BE generally undergo EUS for T and N staging, which is superior to computerized tomography (CT) scanning. EUS has a high negative predictive value (>95%) for local lymph node involvement and for tumor infiltration in the deeper layers of the esophageal wall (≥T2). In case of suspicious lymph nodes, fine-needle aspiration (FNA) may be performed to exclude malignant disease. The accuracy of EUS in differentiating mucosal from submucosal invasion is, however, relatively poor and EUS does not perform better than simple endoscopic inspection of the lesion by an experienced endoscopist. In addition, the high negative predictive value for local lymph node involvement may simply reflect the low prevalence of positive lymph nodes in these lesions. Recent studies have therefore questioned the value of EUS in the work-up of early BE neoplasia. Many centers still perform a thoracic and abdominal CT scan for patients with early BE cancer to exclude distant metastases. This is however controversial, as the distant metastases are virtually absent T1 lesions and thus rarely change the TNM stage.




Endoscopic resection


ER is the most important tool for accurate assessment of invasion depth and other histologic tumor characteristics (differentiation grade, lymphatic invasion). ER therefore serves both a diagnostic as well as therapeutic purpose. Diagnostic in the sense that it correctly identifies patients with submucosal invasion, poor differentiation grade, and/or lymphatic invasion; therapeutic in the sense that it allows removal of all visible (ie, nonflat mucosa) lesions thus rendering the remaining BE segment flat and thereby suitable for RFA. The two most widely used techniques are the ER-cap technique and the Multi Band Mucosectomy (MBM) technique; both are safe and effective for “en bloc” as well as “piecemeal” resections.




Histopathological assessment


For the histopathological evaluation of biopsies or resection specimens, the revised Vienna classification is used. After ER, patients are suitable candidates for further endoscopic management, if the resection specimen shows negative vertical resection margins, no submucosal tumor infiltration, well- or moderate tumor differentiation (G1-G2), and no lymphatic/vascular invasion. Because early BE neoplasia is not frequently encountered by community pathologists, it is imperative that the histopathological evaluation is performed by a pathologist with experience in this field.




Work-up for radiofrequency ablation


Before RFA, all carcinoma has to be completely removed and the epithelium should be endoscopically flat. This is required because the HALO electrodes for RFA are designed to perform controlled superficial ablation of 0.5 mm and are unlikely to be effective for neoplasia in thickened mucosa. Most RFA studies have excluded patients with invasive cancer at baseline and biopsies should therefore be taken from the residual BE segment after ER. Also, in most studies, an additional endoscopy was performed 4 to 6 weeks after the ER session to exclude visible abnormalities and to obtain four quadrant biopsies from the BE segment to rule out invasive cancer. At that time, however, the edges of the ER scar may be slightly elevated because of reactive changes, which may be misinterpreted as visible abnormalities. The most optimal time to exclude the presence of residual visible abnormalities and to obtain biopsies from the residual BE is thus immediately after the ER in the same session.


When the mucosa is flat and biopsies contain no invasive cancer, the patient is scheduled for an RFA session 6 to 8 weeks after ER.




Performing circumferential ablation: use of the HALO 360 -device


RFA procedures can be performed on an outpatient basis using conscious sedation. At the initial ablation session, the extent of the BE segment determines if the patient will be treated with the HALO 360 -balloon or with the focal HALO 90 -device. When the circumferential BE segment is smaller than 2 cm in length, or consists only of tongues and/or islands smaller than 2 cm, the patient can be primarily treated with focal ablation. Most frequently, patients initially require circumferential ablation, which involves the following steps:



  • (1)

    Inspection and recording of esophageal landmarks



Before ablation, the esophagus is inspected to confirm the flat aspect of the mucosa, to exclude the presence of esophagitis, and to record the esophageal landmarks according to the Prague CM classification. After inspection, the esophagus wall is sprayed with 20 mL of acetylcysteine (1%, watery solution) to remove excessive mucus, followed by flushing with plain water.



  • (2)

    Sizing procedure ( Fig. 2 ).




    Fig. 2


    The sizing procedure using the HALO system. The HALO 360 generator with integrated pressure:volume system automatically inflates and deflates the sizing balloon, to calculate esophageal inner diameter. The esophageal inner diameter is measured for each centimeter of the Barrett’s segment from proximal to distal, to adequately select the size of the ablation balloon. ( Reproduced from www.endosurgery.eu ; with permission. Courtesy of ER and RFA Training Program, Amsterdam, The Netherlands; with permission.)



Subsequently, the esophageal inner diameter (EID) is measured, to select an appropriate ablation balloon, allowing for optimal contact between electrode and esophageal wall without applying too much pressure. A 4-cm long noncompliant sizing balloon-catheter is connected to the generator with an interposed air filter. After calibration, the sizing balloon is inserted over a 0.035-in guide wire (Amplatz extra stiff, Cook Endoscopy, Limerick, Ireland), omitting lubricant jelly for introduction, exclusively using water. Sizing is performed for every centimeter of the BE segment, starting 5 cm above the most proximal extent of the BE, using the centimeter scale on the catheter shaft for reference (discordant to the centimeter scale of the endoscope). For sizing, the balloon is automatically inflated to 4 psi (0.28 atm) after pressing a foot switch, and automatically deflated after measurement. The mean EID over 4 cm is then calculated with a pressure-volume algorithm and displayed on the generator. When a strong increase in the measured EID indicates that the sizing balloon has reached the hiatal hernia or stomach, the sizing balloon is removed, leaving the guide wire in place.



  • (3)

    Selection of the appropriate ablation balloon:



The smallest measured EID defines the appropriate balloon diameter. For patients without prior ER, the recommended balloon diameter is the one closest to the smallest measurement (eg, if the smallest EID is 24 mm, a 25-mm ablation balloon is chosen; if the EID is 23 mm, the 22-mm ablation balloon is appropriate). For patients with a prior ER or stenosis, a balloon that is two sizes smaller than the smallest measurement has to be selected, to reduce the risk for laceration (eg, if the smallest EID is 26 mm, a 22-mm ablation balloon should be used). When a suitable ablation balloon size has been selected, the ablation balloon catheter is inserted over the guide wire followed by introduction of the endoscope alongside for visualization of the ablation procedure.



  • (4)

    First ablation pass ( Fig. 3 ).




    Fig. 3


    The ablation procedure using the HALO system. Ablation is performed from proximal to distal starting 1 cm above the most proximal extent of the Barrett’s segment. ( Upper ) The ablation balloon is inflated and ablation is performed upon activation using the foot switch. ( Middle ) After ablation, the Barrett’s epithelium shows a white debris, which helps identifying the subsequent BE zone when advancing the deflated ablation balloon distally. ( Lower ) Subsequent ablations are performed allowing for a maximal overlap of 0.5 to 1.0 cm, until the complete Barrett’s segment has been ablated. ( Reproduced from www.endosurgery.eu ; with permission. Courtesy of ER and RFA Training Program, Amsterdam, The Netherlands; with permission.)



For ablation with the HAL0 360 -device, the surface is treated with two energy deliveries of 12 J/cm 2 with cleaning of the ablation zone in between ablation passes, referred to as the “double” regimen. Ablation is performed from proximal to distal starting 1 cm above the most proximal extent of the BE segment, allowing for a maximal overlap of 0.5 to 1.0 cm. Before ablation, the endoscope and ablation catheter are fixed to the bite block to avoid dislocation of the balloon by inflation. After inflation of the ablation balloon, as initiated by the endoscopist using the foot switch, suctioning is applied for optimal contact between electrode and epithelium. Next, the endoscopist starts the ablation, resulting in RF energy delivery and automatic deflation of the balloon. Ablation renders the BE epithelium white, which helps in identifying the subsequent BE zone when advancing the deflated ablation balloon distally. After ablation of the complete BE segment, the ablation balloon, guide wire, and endoscope are removed simultaneously.



  • (5)

    Cleaning procedure:



After the first ablation pass, the coagulum is cleaned off the ablation surface and the balloon electrode before the second ablation pass. Cleaning in between the ablation passes has been shown to increase the efficacy of the initial ablation session from 90% to 95%. Debris is gently pushed off the ablated area using the rim of the small flexible cap (eg, HALO cap; BÂRRX Medical or Olympus zoom cap, MB0-046; Olympus, Tokyo, Japan), mounted on the tip of the endoscope, followed by forcefully spraying with plain water using a spraying catheter (eg, Olympus PV-5-1; Olympus) and a high-pressure pistol (eg, Alliance; Boston Scientific, Limerick, Ireland). Cleaning of the balloon electrode is performed outside the patient using plain water and a gauze, being careful to wipe “with the grain” of the electrode array.



  • (6)

    Second ablation pass:



After the cleaning procedure, the guide wire and ablation balloon are reinserted followed by the endoscope, as before, and a second ablation pass is performed from proximal to distal, resulting in a brownish discoloration of the previously ablated BE epithelium.




Performing focal ablation: use of the HALO 90 -device


Two to 3 months after initial ablation, upper endoscopy with narrow band imaging (NBI) is performed to assess BE regression. Patients with a residual circumferential BE segment with a length of more than 2 cm and/or multiple isles or tongues are subjected to a second circumferential ablation session. In case the circumferential extent of residual BE is less than 2 cm in length, and/or there is an irregular gastroesophageal junction, and/or tongues and islands are smaller than 2 cm in size, this can be treated with the HAL0 90 -device for focal ablation. The focal ablation procedure is performed as follows:



  • (1)

    Inspection and recording of esophageal landmarks:



Again the esophagus is inspected to confirm the flat aspect of the mucosa and to exclude the presence of esophagitis. Additionally, the presence of a Zenker’s diverticulum has to be excluded because this may complicate the introduction of the HAL0 90 -device. After recording of the esophageal landmarks, the esophagus wall is sprayed with acetylcysteine (1%) and flushed with plain water.



  • (2)

    Introduction of the HALO 90 -catheter ( Fig. 4 ).




    Fig. 4


    ( A ) The leading edge of the HALO cap is visible proximal to the arytenoids. ( B ) A biopsy forceps is blindly advanced behind the arytenoids into the proximal esophagus. ( C ) The endoscope is angulated downward, causing the leading edge of the HALO cap to touch the shaft of the biopsy forceps. ( D ) After gently advancing the endoscope, using the biopsy forceps for guidance, the proximal esophagus is entered. ( Reproduced from www.endosurgery.eu ; with permission. Courtesy of ER and RFA Training Program, Amsterdam, The Netherlands; with permission.)



The HALO 90 -catheter is attached to the distal end of the endoscope, and the electrode cap placed at the 12 o’clock position in the endoscopic field of view. The HALO 90 -device is inserted without using lubricant jelly but only water to facilitate introduction. After advancing the HAL0 90 -electrode surface over the tongue, the top of the electrode is deflected downwards when the laryngeal cavity is visualized, to allow the electrode to pass behind the arytenoids. Subsequently, the patient is asked to swallow and the catheter is gently advanced. When introduction of the HALO 90 -device is difficult, a biopsy forceps or the spraying catheter can be used to guide the instrument into the esophagus. With the endoscope in the hypopharynx, the forceps is advanced under direct view dorsal to the arytenoids deep into the esophagus. The endoscope is angled down so that the leading edge of the HALO 90 -device can slide over the forceps while the endoscope is gently advanced. Other techniques for difficult intubations include Savary dilation (45 Fr) followed by leaving the guide wire in place to facilitate passage of the HALO 90 -device.


Sep 12, 2017 | Posted by in GASTOINESTINAL SURGERY | Comments Off on Endoscopic Therapy Using Radiofrequency Ablation for Esophageal Dysplasia and Carcinoma in Barrett’s Esophagus

Full access? Get Clinical Tree

Get Clinical Tree app for offline access