Barrett's oesophagus

"Endoscopically-visible segment of metaplastic columnar mucosa in the oesophagus, with intestinal metaplasia proven histologically."

• Note differing definitions between UK and USA - requirement of intestinal metaplasia (as indicated by goblet cells) is uncertain. USA says goblet cells required, UK says not required.

Risk factors

• Complication of GORD - GORD is the ONLY cause of BO

Screening

• Not currently recommended in Australia, but can be done overseas as per below
• Indications for screening to look for Barrett's: (if multiple risk factors and have had GORD for >5 years)
  • Age >50
  • Male
  • White
  • Hiatal hernia
  • Obesity
  • Nocturnal reflux
  • Smoker
  • First-degree relative with Barrett's and/or adenocarcinoma
• Only need to screen each patient once.
• If they have oesophagitis, give PPI for 3/12 and screen again

Pathophysiology

• Dysplasia in Barrett's is characterised by cytologic malignant changes including atypical nuclei, increased mitoses, and lack of surface maturation
• Dysplasia is most common closer to GOJ in a long segment of Barrett's
• HGD is distinguished from LGD by more prominent cytologic or architectural derangements
• Remains dysplasia as long as it is confined to the epithelium without invasion of the basement membrane, regardless of the degree of abnormality
• Histological grading is based on the Vienna classification system - non-dysplastic, indeterminate, LGD, HGD, intra-mucosal carcinoma
  • Lots of intra-observer variation. In particular, LGD from general pathologists is downgraded to non-dysplastic by specialist pathologists in 73% of cases
  • Indefinite for dysplasia means equivocal for inflammation or dysplasia - 11.4% pooled annual risk for progressing to LGD and 0.6% for progression to carcinoma
• Precursor for oesophageal adenocarcinoma
  • Non-dysplastic - 0.3 per 100 person-years
  • LGD - 0.5
  • HGD - 6.6
  • Overall incidence of oesophageal adenocarcinoma in BE patients is about 40x the general population
• Long segment is >=3cm - significantly higher risk of progression to carcinoma
• Short tongues <1cm/'variable Z-line' were previously called ultra-short segment Barrett's, but do not constitute increased malignancy risk and no need to biopsy if no concerns for dysplasia

Evaluation:

• Careful endoscopy - white light and NBI are both useful
  • Suspect when GOJ (proximal extent of rugae) is >=1cm distal to the z-line
  • Barrett's has a characteristic salmon pink colour and velvety texture
  • Firstly identify the segment of columnar epithelium, and also look for areas of irregularity within it
  • Use Prague classification system (see 'endoscopy' topic)
  • Note hiatal hernias
  • Seattle protocol for biopsies
• Oesophagitis + Barrett's
  • If LA C or D oesophagitis is found, the pathologist will have trouble identifying dysplasia and a diagnosis of 'indeterminate' is likely
  • If no concerns for cancer, put them on high-dose PPI and re-scope in 6-8 weeks
• Irregular z-line:
  • Found in 10-15% of patients
  • When the SCJ lies above the GOJ, but without any confluence, forming tongues shorter than 1cm, and thus does not fulfil criteria for BO
  • It can harbour histological intestinal metaplasia, but the progression rate to BO is low, let alone to malignancy, so no further evaluation or screening is recommended

Principles of management:

1. Therapy for GORD
2. Prevention of progression/treatment of BE

Management and surveillance:

Flat or nodular lesion present within Barrett's:

• • Refer to centre with EMR/ESD for resection

Barrett's <1cm

• • No increased risk of malignancy
  • No clear guidelines, but probably don't need to do another scope

Non-dysplastic BE: manage as for GORD

• • Symptom relief
  • Continue PPI indefinitely
  • Don't recommend anti-reflux surgery just because of theoretical risk of progression - incidence of cancer appears to be low
  • If can't control symptoms with lifestyle and medication, then they need anti-reflux surgery (exactly same as uncomplicated GORD)
  • Radiofrequency ablation is unproven in this population
  • Surveillance:
    • Non-dysplastic BE - 3-5 yearly - but could do much more frequently to limit progression to that which can be treated endoscopically - some UGI surgeons do yearly scopes
      • If short segment (<3cm), thoroughly biopsied, and no dysplasia, surveillance every five years (10% risk of dysplasia)
      • Long segment every three years (31% risk of dysplasia)
    • Consider stopping surveillance in elderly or unfit patients, similar to NBCSP

Indefinite for dysplasia

• • BD PPI
  • Re-scope Seattle protocol 3-6/12
  • If still indefinite, review with experienced pathologist and either surveillance every 12 months or referral to Barrett's centre
  • Most will regress to non-dysplastic

Low-grade dysplasia

• • Repeat endoscopy ASAP with Seattle protocol if you didn't do it initially - need to exclude more advanced disease
  • Need to review slides with experienced pathologist
  • Either start on PPI or increase existing dose
  • American College of Gastroenterology suggests either endoscopic eradication (then follow-up endoscopy in 6 months) or surveillance every six months for one year and then annually until reversion
    • Most sensible way is probably to check persistence, and refer for ablation if persistent
  • Can be offered anti-reflux surgery - given that most patients with LGD regress after surgery
  • If persistent, may need mucosal ablation

High-grade dysplasia

• • Review slides with experienced pathologist
  • Indication for intervention - ideally RFA
  • Most get endoscopic therapy, followed by surveillance gastroscopy three months later
  • Careful repeat biopsy - look out for focus of invasive carcinoma - use every imaging modality available - white light, NBI, acetic acid
  • Small lesions - endoscopic resection - ideally take full thickness of mucosa and submucosa
  • Can evaluate lesions larger than 1-2cm for invasion using EUS - key point would be whether it is T1a or T1b
  • If HGD or T1a (without LVI), options are oesophageal preservation therapy vs oesophagectomy
    • Oncologically equivalent, but long-term surveillance and multiple initial scopes required, and recurrence is a risk
    • Oesophagectomy historically gets good results in this group, with mortality under 1%
    • However, endoscopic treatment and surveillance is much better now

Endoscopic therapy

• Goals are removal of BE epithelium, while preserving overall integrity of the oesophagus and accurately staging dysplasia or invasive disease
• Should have close surveillance and PPI afterwards
• Beware the risk of 'buried glands' - areas of Barrett's epithelium hidden beneath areas of new squamous epithelium, even if the risk of malignancy in those glands appears to be low
• Radiofrequency ablation
  • Most commonly used endoscopic therapy
  • This is the HALO device, although I think it has been renamed now to 'Barrx'
  • Much more effective than photodynamic therapy with a lower stricture and overall complication rate
  • Delivered with either a circumferential balloon or an electrical plate. The treated mucosa is replaced by neosquamous mucosa. The standard ablation program uses two double pulses of 12 J/sq cm.
  • AIM-II trial - 81% of patients with HGD and 90% of patients with LGD had eradication of dysplasia. 4% had their dysplasia progress to a higher grade.
• Argon plasma coagulation
• Cryotherapy
  • Generally well tolerated with little pain and low stricture rates
  • Does not require the probe to be in contact with tissue
  • Use either a balloon or endoscopic spray catheter to deliver liquid nitrogen
  • Eradicated LGD and HGD at 91% and 81% respectively in one study
  • Can be used as salvage therapy after failed RFA
• Photodynamic therapy
• Endoscopic mucosal resection
  • Best for visible segments
  • Commonly use 'Duette' system (Cook)
• Endoscopic submucosal dissection
  • Better with >2cm
  • Best option for cancer
• Surveillance after eradication
  • Previous HGD: 1, 2, 3, 4, 5, 7, 10 years after treatment
  • Previous LGD: 1, 3, and 5 years after treatment