Barrett's Oesophagus

Intestinal metaplasia (IM) of the lower oesophagus after reflux injury.
Endoscopically evident columnar epithelium extending into oesophagus and (in US) IM goblet cells on histology.

top D I A B M I M home

Found in 10-15% of pts gastroscoped for GORD (Sabiston)
- but most have no symptoms of GORD
- those who have both tend to be the severe ones.
1 in 10 people with erosive esophagitis
1 in 3 with peptic stricture.

Neonates - 90s
- average 55yrs
- peaks at 48-80
- and minor peak at 0-15


MF 3:1.


West / Europeans
Marked increase in last 5 decades.
top D I A B M I M home

Asymptomatic Barrett's-like patches occur rarely

Traumatic following chronic GORD
- up to 20% of GORD pts; those younger and w worse reflux
- often associated with hiatal hernia
- bile reflux and peptic ulcers also found to be associated
- associated with high BMI, metabolic syndrome / visceral fat.
- those with acid secretion
Post sphincter-myotomy (for achalasia)
top D I A B M I M home



Irritation by acid
--> mucosal metaplasia and infiltration
--> replacement of normal lining
--> selective advantage of abnormal tissue
This is the theory.
- in reality, the cellular pedigree is less certain: ?direct metaplasia or gastric infiltration.
When associated with GORD tends to be severe, with ulcers, strictures & bleeding.

1. Specialist intestinal metaplasia
- villiform surface and intestinal crypts with mucus-secreting columnar and goblet cells.
- most common
- this one associated with the dysplasia and carcinoma.
2. Gastric-fundic epithelium
3. Junctional-type epithelium

Short-segment Barrett's
Up to 18% of pts undergoing gastroscopy have been shown to have metaplasia microscopically at the OGJ.
Unknown if this translates into a risk factor for cancer or not.

Natural History
Up to 20% of pts with Barrett's will already have dysplasia at presentation , 7% adenoCa.
- 1% progress to HGD / yr, 0.5% to cancer


3-5% of Barrett's pts / yr progress to dysplasia.
- ~20% progression risk from low grade to high grade.
Precursor of invasive malignancy.
Histologically classified as low grade, high-grade of indefinite for dysplasia
 - indefinite due to inflammation; PPI 40 mg bd and repeat in 4-6w.
Not an ideal marker, however, for a number of reasons:
- small foci of dysplasia are easily missed
- histological interpretation can be subjective (poor inter-pathologist grading; need 2 paths)
- natural history of dysplasia is not conclusive
But ~1/4 of pts with high-grade dysplasia have cancer within 1.5 years.
- 50% at 3 yrs, and 80% at 8 yrs


Barrett's increases risk 30-40x
- especially in smokers and drinkers
8-10% of group seen by surgeons already have it at time of diagnosis.
- this point prevalence overestimates the general incidence amongst Barrett's patients, which is ~1 cancer per 441 pt-years of follow-up.
Usually adenocarcinoma, occasionally SCC
Also rarely - 'adenoid cystic carcinoma'

What happens to Barrett's after anti-reflux surgery?
May regress, but ineffective at preventing malignant degeneration.

Segment length
Short segment <3cm
- twice as common
Long >3cm
- associated with severity of symptoms and progression.

The actuarial survival of pts with Barrett's, without known cancer vs citizens in the same country without Barrett's is the same.
- at the population level, survival is not affected by the diagnosis of Barrett's.
The group that develop cancer are clearly another matter.

Carcinogenesis relates to loss of tumor suppressor genes eg p53, APC, RBI.
Oncogenes c-erbB-2, K-ras, Cyclin D1 are implicated.
Growth factors eg EDGF-Receptor implicated.

May ooze, haemorrhage or perforate.
As above
Colon Cancer
- also strangely slightly increased incidence
top D I A B M I M home



Barret's is a sign of severe GERD
Hx of heartburn, acid regurge
Bitter taste at back of mouth
Difficulty swallowing, especially solids
(Barrett's Ulcer)
- burning discomfort / pain
- radiating to back if severe
As per complication

top D I A B M I M home

Need to be alert for short segment disease
- irregular flame-shaped spread of gastric mucosa onto white oesophageal mucosa
- ?stricture, hiatal hernia, ulcer, tumour
- must biopsy from 4 quadrants to confirm at 2cm
- indocarmine spray down endoscope can detect intestinal metaplasia and possibly dysplasia
LOS pressure decreased, disturbed peristalsis
99Tc taken up by gastic mucosa
top D I A B M I M home


Target GORD and the Barrett's

Primary Screening?
Could be an argument for white men >50yrs, but no recommendations exist.


Treat aggressively the GORD = main therapeutic goal.
Diet change (low fat)
Small meals
Can avoid coffee, tea, alcohol, cola, citrus, tomatoes & chocolate.

PPIs etc - see GORD .
No evidence for impact on cancer.

Algorithm for Management (Sabiston/Cameron)
Start surveillance after GORD inflammation controlled.
- including 4-quadrant biopsy every 2 cm
Findings on surveillance
i) no dysplasia (87%)
--> ~q3 yearly surveillance.
ii) indefinite
--> review by 2nd pathologist + re-biopsy after PPIs high dose for 6/52.
iii) low grade dysplasia (11%)
--> endoscopy at 3m, then q6 -12 monthly, unless radioablation performed.
iv) high grade dysplasia (1.2%) - cancer risk: 50%
--> review by 2nd pathologist.
1. Surveillance --> 4-quadrant biopsies every 1 cm every q3monthly (mapping, 2 cm blocks, 4 per block; Seattle protocol)
2. EMR / Ablation --> endoscopic mucosal resection of visible mucosal irregularities and /or ablation / resection.
3. Surgical resection

Which option for high-grade dysplasia
Patient, disease, surgeon factors.
Is patient young and fit or old and co-morbid?
Is disease short or long segment? Is it multi or single focus?
- if long and multiple, then high risk of Ca already --> prefer surgery and EMR may be contraindicated (below)
- if short and single, then endoscopic approaches now preferable.


Non-dysplastic lesions
RCT shows that anti-reflux surgery is as good or better than PPIs in this group... pH studies better though QOL same;
But be aware that these patients often have severe GERD, hiatal hernia, stricture, poor motility or shortened oesophagus
--> successful surgery is a challenge.
- shortening suspected if video barium oesophagogram for stricture, h. hernia >5cm.
Manometry can show dysmotility, then question is whether to wrap 360 (and risk dysphagia) or lesser wrap (and risk not controlling reflux).
Nissen Fundo on current evidence may outperform PPIs in terms of reduced progression and regression of non-dysplastic metaplasia.
- 30% will regress.
All patients with Barretts should be referred to an esophageal surgeon for a discussion.                                                                                                                                                                               

HALO (Barrx Medical)
- heat-generating balloon conducts high-freq radiowaves
- start by sizing the lesion.
- positioned over metaplastic tissue; inflate to flatten folds
--> short high-power burst of energy that achieves uniform ablation of the muscularis mucosa.
- Results suggest 52% clinical response at 1 yr, and 90% of those with residual disease get clearance on repeat dosing.
--> successful reversion to squamous epithelium in 97% of patients with nondysplastic BE (including 'touch-up' doses).
- Low stricture rate of just 0-1%
Bottom Line
Ideal for patients with non-dysplastic disease.
Controversy is that common use trials not yet performed; only expert centers trained and equipped.

Reflux surgery will not prevent regression of abnormal mucosa.
Previously, oesophagectomy was advised for Barrett's with severe dysplasia or early intramucosal carcinoma.
- however the high morbidilty and mortality, and QOL issues make this view challenging
--> pretty controversial situation.
Now, several alternative methods are also available:
- the HALO is promising, alone or in combination with endoscopic resection.
--> Perhaps 80-90% eradication of dysplasia (low or high) at 1 year.
--> however, await long-term follow up and continued technological evolution.
- endoscopic resection has emerged.
--> almost complete remission after mean 1.9 treatments at 12 months (Ell et al 2007)
--> 5-yr survival 98; endoscopies at 1,2,3,6,9,12 mo, then q6mo until 5 yrs.
-->11% metachronous tumour rate required retreatment.

Endoscopic resection
Endoscopic resection is important. now offered as a potentially curative Rx.
- understand this:
-- LN met risk in dysplasia is practically zero; this is epithelial neoplasia
-- LN risk in intramucosal carcinoma is ~2%
-- LN risk in carcinoma penetrating submucosa is 20-30%.
--> ER is a potentially curative therapy for cancers limited to the mucosa only.
i) focal HGD or T1A intramucosal tumour.
ii) <2 cm
iii) polypoid or flat with ulcers
iv) well or moderately differentiated
v) no evidence of lymphovascular invasion
- EUS for depth of invasion can be used, but note cannot differential intramucosal from intrasubmucosal.
--> T2 or T3 is an absolute contraindication
- be careful, take time to identify all mucosal abnormalities; e.g. h endoscopy, narrow band imaging, dyes.
- isolate lesion with submucosal injection; e.g. with 10-20 mL of saline with 1:100,000 adrenaline
- then snare and burn remove.
--> treat bleeding with dilute adrenaline injections, cautery or clips.
--> if +ve margins, either repeat or esophagectomy.
--> note risk of perforation.


Mortality in this context is around 1% in literature (i.e. lower than <4% rate typically cited for the procedure in high-volume centers) and morbidity still ~20%
physiological outcome also pretty good; 75% return to normal diet.
Should take the entire columnar segment of oesophagus.
- the 'field' is premalignant
- as many as 50% of pts with high grade dysplasia have an unrecognised adenoCa somewhere in the specimen.
- multicentric cancers are common, in 13-37%.

Bottom Line
Refer dysplasia to a high-volume center for consideration of endoscopic resection or perhaps HALO / combination therapy.
Oesophagectomy still a highly valid treatment option, considered for patients with contraindications to EMR, interested in piece of mind, and not intensive endoscopic follow up.

top D I A B M I M home

Sabistons 17th