Chapter 10 – Histology of Gastroesophageal Reflux Disease and Barrett’s Oesophagus

Abstract

Gastro-oesophageal reflux disease (GORD) is a condition that develops as a result of retrograde flow (reflux) of gastro-duodenal contents into the oesophagus, with or without mucosal tissue damage. The main symptoms include heartburn without a specific frequency or duration. Histological mucosal changes include mild basal cell hyperplasia, scattered intraepithelial eosinophils, dilated intracellular spaces, and lamina propria papillae reaching the upper one-third of the squamous epithelium. Other histological findings that can be present in the setting of GORD are balloon cells, keratinocyte vacuolization, intraepithelial lymphocytosis, vascular lakes, and parakeratosis. Ulceration can also occur if GORD is severe. Barrett’s oesophagus (BO) is an acquired condition defined by endoscopically visible metaplastic columnar epithelium more than 1 cm above the gastro-oesophageal junction (GOJ) with biopsy confirmation of columnar metaplasia. Presence or absence of intestinal metaplasia (IM) should be documented as it harbours an increased risk of progression to oesophageal adenocarcinoma (OAC). The risk of progression of BO to OAC increases if there is dysplasia in addition to columnar metaplasia. Endoscopic examination in conjunction with histopathological evaluation and documentation of presence or absence of dysplasia is the current recommended method of surveillance.

Chapter 10 Histology of Gastroesophageal Reflux Disease and Barrett’s Oesophagus

Naziheh Assarzadegan and Maryam Kherad Pezhouh

Introduction

Gastro-oesophageal reflux disease (GORD) is a common chronic disorder of the gastrointestinal (GI) tract.¹ GORD is defined as retrograde flow (reflux) of gastroduodenal contents into the oesophagus, with or without mucosal tissue damage.¹ GORD is one of the most common disorders of the upper GI tract worldwide.² The prevalence of GORD in the Western world is about 10%–20%. This is much higher than in Asia, where the prevalence of GORD is less than 5%.¹ Furthermore, the prevalence of GORD is higher in North America than in Europe.¹ In the United States, approximately 18.1%–27.8% of the population report symptoms of GORD weekly, and about 40% report symptoms occasionally.² Population-based studies indicate that the incidence of GORD is on the rise in different regions globally.³ Obesity, excess consumption of alcohol and caffeine and smoking are the main risk factors for GORD. ³ GORD is also frequently seen in patients with hiatal hernia, scleroderma, and chronic obstructive pulmonary disorder (COPD).³ Some drugs and hormonal products increase the risk of GORD. These include benzodiazepines, anticholinergics, calcium channel blockers, nitrates, dopamine, nicotine, theophylline, glucagon, oestrogen, progesterone, and certain prostaglandins.³ However, in many patients there is no clear predisposing factor.

In adults, a clinical diagnosis of GORD, without the need for further investigations, is usually possible. The classic symptoms of GORD include heartburn and regurgitation, which occur in approximately 70% of patients with GORD.² Chest pain may indicate GORD, but only after exclusion of a cardiac cause.² GORD-related chest pain does not have any standardised frequency or duration. Other extra-oesophageal symptoms of GORD include chronic cough, hoarseness, dental erosions, asthma, sinusitis, laryngitis, and sleep disturbances.² Atypical symptoms of GORD such as epigastric pain, dyspepsia, bloating, nausea, and eructation may also occur in some patients.²

In most cases, characteristic symptoms are sufficient for a diagnosis of GORD, and further investigative tests are not necessary.² Symptom relief following a short trial of proton pump inhibitor (PPI) therapy (PPI test) can help to make a diagnosis of GORD, especially in patients presenting with atypical symptoms.²

Diagnostic tests are indicated in patients who present with alarming symptoms such as GI bleeding, progressive dysphagia, vomiting, unexplained weight loss, or anaemia.² These symptoms need further investigation, as they raise a strong suspicion of other disorders such as eosinophilic oesophagitis or a possible underlying malignant condition or complication. Patients presenting with these alarm symptoms often require endoscopic examination.² Endoscopy is also indicated in patients at risk of Barrett’s oesophagus (BO) and in patients whose symptoms of GORD remain unresponsive to PPI therapy.² Patients at risk of BO are selected for endoscopy on the basis of risk factors such as long-standing GORD, age more than 50 years, male sex, white ethnicity, high body mass index (BMI), obesity with high intra-abdominal fat, and hiatal hernia.⁴^, ⁵ However, there should be a lower threshold for performing an endoscopic evaluation if a patient has a family history of at least one first-degree relative with BO or with oesophageal adenocarcinoma.⁵ Endoscopy has high specificity for the diagnosis of GORD, but it has low sensitivity. Barium swallow studies and oesophageal manometry are usually not helpful in establishing a diagnosis of GORD. However, these studies may be useful to evaluate complications such as oesophageal stricture or to rule out achalasia.²

Histological Definition of GORD

Although Ismail-Beigi et al. documented the abnormal histological features of GORD in 1970,⁶ Fiocca et al. produced the first consensus guideline for the histological recognition of GORD 39 years later in 2009.⁷ Histologically, the lining of the oesophagus consists of stratified squamous epithelium. Normally the basal layer is around one to two cell layers thick (Figure 10.1) and the vascular papillae are confined to the lower one-third of the epithelium.

Figure 10.1 Unremarkable oesophageal squamous mucosa. The basal layer is only a few cells thick and the vascular papillae are confined to the lower third of the epithelial thickness.

One of the histological features associated with GORD is mild basal cell hyperplasia, defined as more than 15% of epithelial thickness or more than five to six cell layers thick (Figure 10.2A and B). The upper limit of the basal layer is defined as the level above which the nuclei are separated by a distance greater than the nuclear diameter.⁷ Another histological feature of GORD is the presence of scattered intraepithelial eosinophils, i.e. around one to two eosinophils per high-power field. However, the presence of eosinophilic microabscesses or eosinophil degranulation should raise suspicion for eosinophilic oesophagitis. Less commonly, scattered neutrophils can also be present in GORD, while severe reflux injury can lead to erosion and ulceration with prominent neutrophilic infiltration. Nevertheless, the presence of a superficial band-like neutrophilic infiltrate and an increase in intraepithelial lymphocyte numbers should alert the pathologist to rule out Candida oesophagitis.⁸ Another histological feature of GORD is elongation of the lamina propria papillae to more than two-thirds of the epithelial thickness (or more than 50% of the thickness, according to other authors) (Figure 10.2C). Spongiosis or dilated intercellular spaces are another feature of reflux disease (Figure 10.2D).

(A) Note the basal cell hyperplasia. The normal basal cell layer should be around one or two cells thick. In addition, there are scattered eosinophils (arrows) and focal parakeratosis on the surface (arrowheads).

(B) This example highlights the basal hyperplasia. The basal layer constitutes up to half of the thickness of the epithelium. Note the presence of dilated intercellular spaces.

(C) Reflux-associated changes including elongated vascular papillae and basal hyperplasia. Furthermore, note the parakeratosis present on the left upper corner.

(D) Basal hyperplasia, dilated intercellular spaces, and vacuolated squamous cells (arrows) in the oesophagus mucosa with GORD.

Figure 10.2 Reflux-associated changes.

Other histological findings that often occur in GORD include balloon cell change, keratinocyte vacuolisation, an increase in intraepithelial lymphocytes, vascular lakes, and parakeratosis. Balloon cell change is likely to be an early change wherein the squamous cells appear to be distended with abundant pink cytoplasm as a result of accumulation of plasma cell proteins (Figure 10.2C). Parakeratosis can occur as a component of GORD, but the presence of focal parakeratosis warrants a search for fungal organisms. Severe erosive oesophagitis leading to stricture formation was one of the complications of GORD in the past. Treatment with PPIs and lifestyle modifications is now often successful. Fact Sheet 10.1 lists the main histological features of GORD and Table 10.1 defines some of the terms in more detail.

Table 10.1 Definitions of the main histopathological features of GORD

Criteria	Definition
Basal cell hyperplasia	>15% or 5 to 6 layers The upper limit of the basal layer is defined as the level above which the nuclei are separated by a distance greater than the nuclear diameter
Intraepithelial eosinophils	>1 to 2 eosinophils per × 40 high-power field
Spongiosis	Dilated intercellular spaces
Elongated vascular papillae	Vascular papillae reach the upper one-third of the epithelium

A novel histological definition of normal oesophageal state and GORD was proposed by Chandrasoma.⁹ This concept recognises the fact that normal oesophagus is lined by stratified squamous epithelium and normal proximal stomach is lined by oxyntic mucosa. Other columnar epithelia such as cardiac type gastric mucosa, cardio-oxyntic mucosa, and intestinal mucosa (if present) are always located between squamous epithelium and oxyntic mucosa, giving rise to a new histological concept of ‘the squamo-oxyntic gap’. This concept is based on autopsy studies that showed that most patients without clinical GORD had no cardiac or intestinal metaplasia.¹⁰^, ¹¹ Based on the squamo-oxyntic gap, the oesophageal squamous epithelium in the normal state attaches directly to the gastric oxyntic mucosa lining the proximal stomach. Thus, in the normal state, there is zero squamo-oxyntic gap and there are only two epithelial types.⁹ On the contrary, the histological presence of cardiac, oxyntocardiac, or intestinal epithelia between the squamous epithelium and the gastric oxyntic mucosa that makes the squamo-oxyntic gap greater than zero indicates cellular damage caused by GORD.⁹ Furthermore, the prevalence of intestinal metaplasia is directly proportional to the length of the squamo-oxyntic gap.⁹ This concept is still not widely accepted and needs more validation through in-depth study of the normal status of proximal gastric mucosal histology.

Barrett’s Oesophagus

Barrett’s oesophagus (BO) is a condition wherein columnar epithelium replaces the normal oesophageal squamous epithelium, i.e. columnar metaplasia.¹² Norman Barrett, a British surgeon, first described this condition in 1950.¹³ The worldwide prevalence of BO is between 0.5% and 2.0%.¹² In Western countries, BO affects approximately 2%–7% of the population.¹⁴ Prevalence of BO in the general population in Europe is between 1% and 2%.¹⁵^, ¹⁶ In the United States, an estimated prevalence of BO is 5.6%.¹⁷ Endoscopically, BO is apparent in about 1%–2% of the patients receiving endoscopy for any clinical indication.¹² The condition primarily affects the elderly population in developed countries.¹² The incidence of BO in men is significantly higher than in women.¹² The risk factors for BO include behavioural and anthropomorphic factors such as GORD, obesity and tobacco smoking.¹² Chronic or long-standing GORD is the strongest risk factor for BO. In patients with GORD, the prevalence of BO is about 5%–15%.¹² Other potential risk factors include type 2 diabetes mellitus, sleep apnoea, and metabolic syndrome.¹⁸^–²⁰

BO is a precursor of oesophageal adenocarcinoma (OAC), the incidence of which has increased drastically in the past few decades.¹² BO is also linked to a proportion of ‘gastro-oesophageal junction’ (GOJ) adenocarcinomas.¹⁴ Compared to the general population, the estimated risk of OAC is 30- to 125-fold higher in patients with BO.¹² The estimated annual incidence of OAC in patients with BO is approximately 0.12%–0.13%.²¹ OAC has a poor prognosis, with a 5-year survival rate of less than 20%.¹² The current surveillance method includes endoscopic examination along with histological evaluation of the BO and determination of the presence or absence of dysplasia.⁵

Over the past decade, significant advances have occurred in the biological and pathological understanding of the various characteristics of the oesophagus and the GOJ in response to mucosal damage caused by chronic GORD. These advances have enabled a better understanding of the pathogenesis of BO and its progression to OAC. In the next section, we will summarise the histopathological features of BO, along with the diagnostic criteria for BO in the United States and Europe.

Histological Features of Barrett’s Oesophagus

Disagreements regarding distal oesophageal histology and BO continue to the present. The American College of Gastroenterology (ACG) defines BO as ‘the condition in which any extent of metaplastic columnar epithelium that predisposes to cancer development replaces the stratified squamous epithelium that normally lines the distal oesophagus’.²²^–²⁴ In 2015, the ACG released an updated clinical guideline on the diagnosis of BO. According to the new ACG guideline, the diagnosis of BO should be made when there is extension of salmon-coloured mucosa into the tubular oesophagus at least 1 cm proximal to the GOJ in conjunction with biopsy confirmation of intestinal metaplasia (IM).²² ACG also recommends that endoscopic biopsies should not be performed in the presence of a normal Z line or a Z line with less than 1 cm of variability.²² However, British Society of Gastroenterology (BSG) guidelines only require visual presence of columnar-lined epithelium on endoscopic examination and evidence of columnar metaplasia in corresponding biopsy specimens, irrespective of IM.⁵ Presence of IM or goblet cells is not a requirement in the BSG guideline. Recently, the European Society of Gastrointestinal Endoscopy (ESGE) issued a position statement to homogenise recommendations and patient management. In their statement, the diagnosis of BO is appropriate if the distal oesophagus is lined by columnar epithelium with a minimum length of 1 cm and there is IM on histopathological examination.²⁵ The difference between ACG and BSG is mainly due to the different risk of malignant transformation between columnar epithelium with or without intestinal metaplasia. In some reports, the risk of malignant progression to OAC is significantly higher in columnar (gastric) epithelium with IM compared to gastric metaplasia without IM.²¹^, ²⁶^, ²⁷ However, some studies do not support this notion.²⁸^, ²⁹ Therefore, the BSG guideline opts for broader diagnostic criteria. The latter are reasonable for several reasons, including the conflicting evidence about the risk of progression, the possibility of underestimation of IM due to sampling error, and the fact that our understanding of the cellular and molecular basis for malignant progression to OAC continues to evolve.³⁰ Moreover, the international Benign Barrett’s and Cancer Taskforce (BOBCAT) consensus defined BO as presence of columnar epithelium above the GOJ with clear documentation of the presence or absence of IM.³¹

As mentioned earlier, the BSG guideline defines BO as endoscopically visible metaplastic columnar epithelium extending more than 1 cm above the GOJ with biopsy confirmation of columnar metaplasia.⁵ Normally, the oesophagus has a lining of stratified squamous epithelium. Compact submucosal glands and related squamous-lined ducts also form part of the normal oesophageal histology. Histologically, Barrett’s oesophagus is regarded as the end-result of the metaplastic transformation of the typical oesophageal squamous epithelium into columnar epithelium.¹⁴ In general, BO comprises two epithelial compartments histologically. These epithelial compartments are the surface epithelium and the glandular compartment.¹⁴ This surface epithelial unit of BO is known as ‘pit or crypt epithelium’ because it exhibits features of the gastric foveolar epithelium (pit) as well as the colonic crypt epithelium. ¹⁴ In advanced stages of BO with IM, also known as fully developed BO, this functional epithelial unit presents more features of colonic crypt epithelium than of gastric pit epithelium. In BO, there can be three different types of columnar epithelium. These types are (1) cardia type, (2) gastric fundic type, and (3) intestinal type with goblet cells.²³ Of these three types of columnar metaplasia of the oesophageal mucosa, the third type with IM has been directly associated with a high risk of malignant progression.²¹

In BO, both the surface epithelium and the crypt epithelium consist of a mosaic of different cell types.¹⁴ Mucinous cells (normally found in the gastric epithelium) and goblet cells (intestinal cells) form part of the surface and crypt epithelium in BO (Figure 10.3A and B). Less frequently, intestinal cells such as endocrine cells, enterocytes and Paneth cells also constitute the epithelial compartments in BO.¹⁴ Cells with combined features of intestinal and gastric mucosa (multilayered epithelium) can also be present in BO epithelium (Figure 10.4). The multilayered epithelium is composed of superficial mucinous columnar epithelium with four to eight layers of basal squamous epithelium and is believed to be a precursor of BO. ³² The percentage of these cell types in the epithelial compartment probably depends on the time required for BO development and the stage of BO. However, the factors responsible for the presence of various cell types in BO remain unknown.¹⁴

Figure 10.3

(A) Barrett’s oesophagus without intestinal metaplasia. Notice that the diagnosis of BO is based on endoscopic correlation.

(B) Barrett’s oesophagus with focal intestinal metaplasia, negative for dysplasia. The 4 lines are present (see Figure 10.5) and the nuclei have open chromatin with abundant cytoplasm and surface maturation.

Figure 10.4 Multilayered epithelium. This epithelium is characterised by presence of four to eight basally located squamous epithelial cells with superficial mucinous epithelium and is believed to be a precursor of Barrett’s oesophagus.

The glandular compartment located underneath the crypt epithelium is composed of pure oxyntic glands, pure mucous glands, or a combination of both types of gland.¹⁴ Among BO patients, the number of glands, their location and their type are highly variable. Some evidence suggests that the amount and type of glands vary according to the natural progression of BO to cancer and according to the location of the BO lesion in the oesophagus. For example, oxyntic glands are more frequent in BO of the distal oesophagus or GOJ region, while pure mucous glands are more common in proximal BO mucosa.¹⁴

Along with the epithelial changes, BO shows mesenchymal and stromal changes that include duplication of the muscularis mucosae, increased number of lymphatics and blood vessels, and alterations in the inflammatory cells.¹⁴

Diagnosing BO requires endoscopic correlation, and the pathologist’s role initially is to document the type of mucosa present. Assigning the term BO should be limited to biopsies from the tubular oesophagus located more than 1 cm above the GOJ with columnar metaplasia. Presence or absence of IM and of dysplasia should also be recorded.

Histopathology of Dysplasia in Barrett’s Oesophagus

Progression of BO to oesophageal adenocarcinoma (OAC) is a series of molecular and histological events.¹⁴ The sequence of events begins with metaplasia, then advances through several stages of dysplasia, ultimately developing into adenocarcinoma. Therefore, risk assessment of patients with BO typically involves routine endoscopic surveillance, along with histological evaluation of oesophageal mucosal biopsies to determine the type and grade of dysplasia.³³ Determining the degree of dysplasia can be challenging, and several studies have shown poor interobserver agreement between pathologists when assessing the degree of dysplasia in BO.³⁴^, ³⁵ One of the issues on which the BSG and ACG agree is that a second GI pathologist should review cases of suspected dysplasia for confirmation of the diagnosis.⁵^, ²²

Dysplasia is morphologically defined as neoplastic epithelium that is limited to the basement membrane. Several types of dysplasia can occur in BO. Certain subtypes of BO dysplasia, such as serrated and gastric dysplasia, have been described relatively recently. Therefore, they are not as well characterised as intestinal-type dysplasia.¹⁴ Some patients present with no particular type of dysplasia (neither gastric, nor serrated, nor intestinal). On the contrary, many patients may show evidence of a combination of different types of dysplasia. Irrespective of the type of dysplasia in BO, it is graded as either negative, low grade, or high grade, and the option of indefinite for dysplasia is also available. ‘Negative for dysplasia’ implies non-neoplastic epithelium, which can be regenerating. When a definite diagnosis for a lesion with epithelial atypia is not possible, the term ‘indefinite for dysplasia’ is applicable. The endoscopic surveillance frequency of patient with non-dysplastic BO is every 3–5 years.²² However, patients with low-grade dysplasia should be surveyed every 6 months and, based on the BSG guideline, no ablation therapy is currently recommended.⁵ On the other hand, the recent ACG guideline recommends endoscopic ablation therapy in BO and low grade dysplasia.²² Treatment of BO with high-grade dysplasia involves endoscopic or surgical management based on staging, comorbidities, nutritional status, and patient preferences.⁵ The histopathological classification and the criteria used to grade dysplasia in BO are discussed in the text that follows.

Negative for Dysplasia

The term ‘negative for dysplasia’ describes non-dysplastic columnar epithelium that maintains its cellular architecture. The non-dysplastic surface epithelium should maintain an apical mucin vacuole followed by a layer of cytoplasm and then the nucleus at the base of the cell. This architecture was recently described as ‘4 lines’ (Figure 10.5);³⁶ the gastric foveolar mucin vacuole constitutes line 1, the base of the mucin vacuole is line 2, and line 3 is the cytoplasm underneath. Line 4 is the nuclei, showing no stratification. Presence of all four lines implies a non-dysplastic BO.³⁶ BO biopsies that are ‘negative for dysplasia’ usually demonstrate surface maturation. This is because the cell nuclei gradually become smaller as the cells come nearer to the surface. The cytoplasmic volume of the cells is increased and the nuclei become less hyperchromatic. Glands in non-dysplastic BO are usually round and evenly spaced. They also present with ample intervening lamina propria.²³ In both the surface epithelium and the basal crypts, the nuclear and nucleolar membranes are smooth. If mitotic figures are present, they are usually limited to the basal compartment. The non-dysplastic epithelium may also show mild cytological atypia, with scattered mitotic figures, mild enlargement of the nuclei, and nuclear hyperchromasia.²³ The cytological atypia is more pronounced in the basal crypts and in the surface epithelium that is directly in the vicinity of the squamous mucosa.²³ In cases with inflammation and/or ulceration, reparative changes may be present. Such changes include nuclear and nucleolar enlargement, along with loss of superficial mucin. However, the nuclear and nucleolar membranes in such cases remain smooth with ample lamina propria between the glands.²³

Figure 10.5 Barrett’s oesophagus without dysplasia showing four lines arrows. The first line is defined as the gastric foveolar mucin vacuole constitute, line 2 is the base of the mucin vacuole, line 3 is the cytoplasm underneath and line 4 is the nuclei with no stratification.

Only gold members can continue reading. Log In or Register to continue