**8. Endoscopic resection**

Endoscopic mucosal resection (EMR) describes any technique which removes a complete area of mucosa. However, the term is somewhat misleading and many authors recommend a switch to the term 'endoscopic resection' (ER), as the aim of EMR should involve complete excision of the mucosal and submucosal layers down to the muscularis propria.

The technique involves raising an area of mucosa using suction or by submucosal injection, and then snaring it off (in a similar manner to a colonic polyp). It is a useful technique for removing focal areas of HGD or early cancer, and as well as being therapeutic, can provide important diagnostic information.

Fig. 2. Endoscopic resection of an early invasive cancer

## **8.1 Diagnostic endoscopic resection**

Although not yet recognised by BSG guidelines, ER has now become an important diagnostic technique in patients with HGD / early cancer. ER preserves tissue architecture so that a full histopathological assessment can be made. (Odze and Lauwers, 2008) An ER specimen can be more easily orientated than a mucosal point biopsy and should contain a significant portion of submucosa allowing accurate assessment of the depth of invasion of IMC. A retrospective study of 150 EMR specimens (focal lesions) found that following analysis of EMR specimens, initial diagnoses (based on point mucosal biopsies) where changed in 49% of cases, leading to a change in management plan in 30%. (Peters et al., 2008) Mino-Kenudson et al recently demonstrated that interobserver reporting agreement between pathologists was improved when reporting EMR specimens rather than point biopsies, particularly when differentiating between intramucosal and submucosal carcinoma – a key distinction when planning a treatment strategy. (Mandal et al., 2009) ER specimens also enable improved assessment of other important prognostic factors such as the grade of cellular differentiation and the presence of lymphovascular invasion. ER has been shown to be the best technique for assessment of visible mucosal abnormalities within Barrett's oesophagus. However, the technique does have complications, and these must be considered when performing an ER for diagnostic purposes.

#### **8.1.1 Therapeutic ER**

As a therapeutic technique, oesophageal ER has been assessed in a number of large studies, although no randomised controlled trials currently exist. Its first description in HGD and

Endoscopic mucosal resection (EMR) describes any technique which removes a complete area of mucosa. However, the term is somewhat misleading and many authors recommend a switch to the term 'endoscopic resection' (ER), as the aim of EMR should involve complete

The technique involves raising an area of mucosa using suction or by submucosal injection, and then snaring it off (in a similar manner to a colonic polyp). It is a useful technique for removing focal areas of HGD or early cancer, and as well as being therapeutic, can provide

Although not yet recognised by BSG guidelines, ER has now become an important diagnostic technique in patients with HGD / early cancer. ER preserves tissue architecture so that a full histopathological assessment can be made. (Odze and Lauwers, 2008) An ER specimen can be more easily orientated than a mucosal point biopsy and should contain a significant portion of submucosa allowing accurate assessment of the depth of invasion of IMC. A retrospective study of 150 EMR specimens (focal lesions) found that following analysis of EMR specimens, initial diagnoses (based on point mucosal biopsies) where changed in 49% of cases, leading to a change in management plan in 30%. (Peters et al., 2008) Mino-Kenudson et al recently demonstrated that interobserver reporting agreement between pathologists was improved when reporting EMR specimens rather than point biopsies, particularly when differentiating between intramucosal and submucosal carcinoma – a key distinction when planning a treatment strategy. (Mandal et al., 2009) ER specimens also enable improved assessment of other important prognostic factors such as the grade of cellular differentiation and the presence of lymphovascular invasion. ER has been shown to be the best technique for assessment of visible mucosal abnormalities within Barrett's oesophagus. However, the technique does have complications, and these must be

As a therapeutic technique, oesophageal ER has been assessed in a number of large studies, although no randomised controlled trials currently exist. Its first description in HGD and

excision of the mucosal and submucosal layers down to the muscularis propria.

**8. Endoscopic resection** 

important diagnostic information.

Fig. 2. Endoscopic resection of an early invasive cancer

considered when performing an ER for diagnostic purposes.

**8.1 Diagnostic endoscopic resection** 

**8.1.1 Therapeutic ER** 

early mucosal cancers in Barrett's oesophagus was published in 2000, although ER was described for early oesophageal SCC as far back as the early 1990s.

A recent study reported that ER achieved remission in 82.5% of patients with HGD. However, over a 12 month period of follow-up, metachronous lesions or disease recurrence were identified in 14% of patients, necessitating re-treatment. A further study using ER, photodynamic therapy (PDT) or a combination, showed overall complete disease remission in 98% of patients, but metachronous cancer was identified in 31% over a 34 month post treatment surveillance period. (Pech et al., 2007)

Extensive, multifocal disease is more difficult to manage endoscopically using ER. Several studies have described circumferential ER of extensive Barrett's segments but, despite experienced hands, these extensive resections have been associated with significant complication rates (bleeding 33%, strictures 17-26% and perforation 3%) and large studies with prolonged follow-up have not been conducted. (Pech et al., 2007; Seewald et al., 2003)In addition, significant recurrence rates have been reported and in up to 25% of cases, complete resection of the Barrett's segment was impossible despite several staged attempts at treatment. (Pech et al., 2007)

As discussed previously, oesophageal lesions that invade into the lamina propria but are confined to it (do not invade the submucosa) T1m1-3 (T1a) lesions have a 5% chance or less of nodal involvement. Recent data have suggested that shallow mucosal invasion T1sm1 also has a significantly lower risk of nodal metastases than other grades of submucosal invasion. (Prasad et al., 2007; Gondrie et al., 2008) Whereas deeper invasion into the submucosa (T1sm2-3) sees this risk rise to 20-45%. (Peyre et al., 2008) In early cancers with a low risk of lymphatic spread, ER offers a curative, minimally invasive treatment option, which may be particularly appropriate in older patients at higher risk of operative morbidity / mortality. Average 3-year survival rates of more than 80% have been reported for IMC treated by ER.

In recent years there has been a move towards combination therapy in an attempt to reduce recurrence rates. Areas of focal dysplasia (or IMC) could be treated with ER, followed by complete ablation of the entire Barrett's segment using APC, PDT, or RFA.

#### **8.2 Mucosal ablation therapy**

Ablation techniques aim to destroy the lining of the oesophagus and promote regenerative growth of normal squamous mucosa. Techniques for mucosal ablation include photodynamic therapy, thermal ablation, radiofrequency ablation and argon plasma coagulation, all of which must be used in combination with acid suppression.

There are so far no randomised trials comparing these treatments against each other. In addition, the natural history of regenerated squamous epithelium is not fully known, (although there certainly appears to be a substantial reduction in malignant potential) so further long-term studies are still awaited. (Overholt et al., 2005)

#### **8.2.1 Radiofrequency ablation**

RFA is a relatively new technique which can be used to ablate circumferential (HALO360) or focal (HALO90) Barrett's oesophagus. Circumferential ablation is performed using a balloon to apply radiofrequency energy evenly to the oesophageal lining.

The length of the Barrett's segment is first measured endoscopically. N-acetylcysteine is then used to wash saliva, mucus and gastric juice from the oesophagus and a guidewire is placed into the gastric antrum. The endoscope is removed and a sizing balloon is inserted

Endoscopic Detection and Eradication of Dysplastic Barrett's Oesophagus 161

complications have been reported with 5-ALA and further work is therefore required to

APC (the most commonly used form of thermal ablation) uses a jet of ionized argon gas (plasma) directed through an endoscopic probe to ablate short segments of Barrett's or areas of persistent disease following other ablative treatment. A trial by Pech et al retrospectively assessed disease recurrence in patients treated by EMR with or without subsequent APC ablation of the residual Barrett's segment. Rates of recurrence fell from 33.3% to 17.6% with the inclusion of APC ablation. Other studies support these results and confirm low

In some cases following ablative therapy for Barrett's oesophagus, a normal squamous epithelium may re-grow over a portion of Barrett's tissue. Endoscopically this appears normal, but these buried Barrett's glands may retain malignant potential. Endoscopists must be aware of this when surveying patients who have previously undergone ablative endotherapy and for this reason life-long endoscopic surveillance is recommended for these

HGD is associated with early invasive malignancy in up to 30% of cases, and carries a significant long-term chance of malignant progression. In addition, recurrence rates following ablative therapies are significant and endoscopic surveillance must be lifelong. For these reasons, surgical excision of the entire Barrett's segment must still be considered

Oesophagectomy is the only potentially curative treatment once lymph nodes are involved. It also aims to remove the entire Barrett's segment minimising the chance of recurrence or missed metachronous lesions. Recent centralisation of cancer services has improved operative mortality to 5% or less in most specialist units. However, for patients without proven invasive cancer, this still remains a considerable risk. In addition, morbidity following oesophagectomy remains considerable although minimally invasive and vagal sparing surgery aims to minimise this and improve long-term functional outcomes. (Ell et

Patients with LGD should undergo repeat endoscopy with adherence to a 'gold standard' biopsy regimen 8-12 weeks after the commencement of PPI therapy. A repeat endoscopy should then be performed at 6 months, and if LGD persists, endoscopy should be repeated 6-monthly unless regression to normal Barrett's or squamous epithelium occurs, at which

In some cases of multifocal, persistent LGD, endoscopic mucosal ablation therapy could be considered, particularly if there is a strong patient desire for intervention (BSG Working Party, 2005) (although evidence for this statement is limited and widespread treatment of

time surveillance can be reduced to 2 yearly intervals. (BSG Working Party, 2005)

clarify the most effective drug with the least side-effects.

patients, even in the absence of residual Barrett's oesophagus.

**11. Management of low-grade dysplasia** 

LGD is not cost-effective and is not recommended).

the 'gold standard' treatment for young, fit patients with multifocal HGD.

**8.2.3 Argon plasma coagulation** 

complication rates.

**9. Buried glands** 

**10. Oesophagectomy** 

al., 2007)

over the guide wire and inflated once in the distal oesophagus. In long segment Barrett's oesophagus several measurements are taken and subsequently, an appropriately sized ablation catheter is selected (based on the smallest oesophageal diameter measurement). The catheter is then passed over the guide wire and positioned at the proximal extent of the Barrett's segment. The endoscope is re-passed to ensure correct positioning of the catheter and the balloon is then inflated and a standardised dose of energy is delivered (which has a power density sufficient to ablate down to the muscularis mucosae, 700-1000µm deep). After a short period of treatment (<5s) the catheter is passed distally to the next portion of the oesophagus to be treated, trying to minimise overlap between zones by endoscopic visualisation. Once the entire Barrett's segment has been ablated the catheter is removed and the endoscope is reinserted in order to debride the ablated mucosa. The procedure is then repeated so that the whole Barrett's segment receives two treatments.

Complications are rare but include significant bleeding (1-2%), stricture formation (6%) and perforation (very rare). (Shaheen et al., 2009) Repeat OGD is recommended after 2 – 3 months and any residual focal Barrett's oesophagus can then be treated using HALO90 RFA.

The only RCT, by Shaheen et al in 2009, demonstrated successful resolution of dysplastic Barrett's oesophagus following treatment with RFA.(Shaheen et al., 2009) Complete eradication of LGD was seen in 90.5% (ablation group) compared to 22.7% (control group) (P<0.001). Complete eradication of HGD occurred in 81.0% (ablation group) versus 19.0% (control group) (P<0.001). RFA also decreased the likelihood of disease progression (3.6% vs. 16.3%, P=0.03) and cancer (1.2% vs. 9.3%, P=0.045).

Recent NICE guidelines (June 2010) recommend that clinicians in the UK consider endoscopic ablation therapy (preferentially RFA) along with EMR, for treatment of HGD and IMC, particularly in patients not suitable for oesophagectomy.

#### **8.2.2 Photodynamic therapy**

Porfimer sodium photodynamic therapy (PDT) has been approved by the US Food and Drug Administration (FDA) and (provisionally) by NICE for treatment of HGD in Barrett's oesophagus.

The procedure involves systemic (intravenous) administration of a photosensitising agent (porfimer sodium) which is retained selectively by dysplastic cells. After about 48 hours the patient undergoes an upper endoscopy and a laser is used to excite a cytotoxic reaction in dysplastic Barrett's cells, leading to their destruction. There is now strong evidence that PDT can prevent the progression of disease in patients with Barrett's HGD. (Overholt et al., 2007) A five year randomised multicentre trial by Overholt et al demonstrated that PDT was significantly more effective at eradicating HGD than omeprazole only (odds ratio 2±0.7). It also significantly lengthened the time taken to progress to malignancy and reduced the overall risk of malignant progression by half. (Overholt et al., 2007) Following PDT, patients are required to continue life-long surveillance, and repeat ablation may become necessary.

Side-effects of PDT include nausea and chest pain in the first day or two after treatment. In the longer term, oesophageal strictures may occur in up to a quarter of patients. Oesophageal perforation has also been described (very rarely). In addition, due to the photosensitising affect of porfimer sodium, patients are required to minimise light exposure to their skin for up to 4-6 weeks after the treatment.

Several trials in Europe have used 5-ALA as the photosensitising agent in an effort to reduce skin sensitivity and oesophageal strictures. However, additional blood pressure and cardiac complications have been reported with 5-ALA and further work is therefore required to clarify the most effective drug with the least side-effects.
