This chapter serves as a guide to gastrointestinal (GI) endoscopy for the histopathologist for benign pathology and outlines the value to endoscopists of GI biopsies. It sets out to explain to a histopathology audience the decision-making process that endoscopists undertake when performing an endoscopic procedure and, crucially, what they need in return in a pathology report. There are explanations as to the thought processes that lead to endoscopic biopsies both in macroscopic disease and, perhaps more importantly, when appearances are normal. There is a guide to each anatomical area detailing how, where, and when to take samples from each region from the oesophagus to the rectum, followed by detailed up-to-date explanations on the latest guidelines for endoscopic sampling for each disease state. For each anatomical area there are practice point summaries and for some diseases detailed endoscopic pictures. The goal of this chapter is to enhance the pathologist’s knowledge of endoscopy, benign GI disease, and national and international sampling guidelines. Finally, we hope it serves as a guide for enhanced discussion between endoscopists and pathologists and encourages a collaborative approach to disease diagnosis.
Prior to the development of the first fibre-optic endoscopes in the 1960s, gastroenterology, in common with other traditional medical specialties, relied on biochemical and radiological techniques to investigate the gastrointestinal (GI) tract. Histological confirmation of a disease process usually required the presence of a surgeon, was invasive, and carried inherent risks. Because of a rapid improvement in technology, we are now able to sample the entire GI tract from the mouth and anus through to the ileal–jejunal junction. In addition, endoscopy can now sample beyond the mucosa into the submucosa and extra-luminal structures, because of the development of endoscopic ultrasound. Therefore, in many ways the history of GI pathology has mirrored the development of GI endoscopy. With the exponential growth in the use of endoscopy since the early 1970s, there is a continually increasing need to interpret the GI pathophysiology and immunology through the lens of a 1 to 2 mm sample of mucosa embedded in formalin. Like any investigative process, dialogue between the endoscopist framing the question and the pathologist attempting to answer it is crucial. The keys to success are clear communication, meticulous specimen labelling, sufficient clinical information, and forums for feedback and discussion such as multidisciplinary meetings, research, or simply coffee and a chat. All of these combine to embed the histopathology report within the clinical environment.
This chapter sets out to explain the value of histology to a gastroenterologist or surgeon, particularly in benign disease, and sets out the latest guidelines for endoscopic sampling in the upper, mid, and lower GI tract. Finally, it aims to explain to a histopathological audience the decision-making process that endoscopists follow when performing an endoscopic procedure and, crucially, what the endoscopist needs in return in a pathology report.
What Is the Clinical Relevance of Histopathology in Benign Gastrointestinal Disease?
Endoscopists take biopsies of the GI tract for a huge variety of reasons. The indications vary depending on the experience of the endoscopist but can be broken down into three broad categories.
1. Macroscopic disease. An experienced endoscopist taking a biopsy usually has a reasonable idea of the pathology. This process starts before the procedure with a clinical history, examination, and perhaps biochemical or radiological investigation. In diseases such as inflammatory bowel disease (IBD), the macroscopic appearances are often so characteristic that in combination with the pre-endoscopic history and biochemical investigations the diagnosis is clear. However, there is always a differential diagnosis and the histopathological confirmation remains crucial. Importantly, a diagnosis of IBD will probably remain a lifelong label; will underpin management decisions such as immunomodulators, biologics, and surgery for years to come; and may have a profound psychosocial impact on the patient. Other benign conditions that the endoscopist may feel confident about diagnosing macroscopically are peptic ulcer disease, colonic polyps, and GI cancers.
2. Microscopic diseases. There are several conditions with subtly abnormal or entirely normal macroscopic appearances but with a clinical history and supplementary investigations that may point to a high level of suspicion of abnormality. For example, in mild coeliac disease the typical features of scalloping of the second part of the duodenum may be difficult for an endoscopist to detect. Although there may be a typical history of bloating, discomfort, and diarrhoea associated with the ingestion of wheat and with serological markers that make the diagnosis highly likely, the gold standard ultimately remains histological. In other conditions such as eosinophilic oesophagitis or collagenous colitis, there is no serum or radiological marker and a diagnosis is entirely reliant on the histopathologist’s skill. The value of the biopsy in these conditions is even greater than in the scenario in which there are obvious macroscopic changes. Histology can play a role in both the diagnosis and monitoring of these diseases, and repeated sampling through the lifetime of the patient may be a component of management.
3. Reassurance. Despite the broad range of GI conditions, the most common histopathological result is likely to state that the microscopic appearances are within the normal range. Why is this the case? Endoscopy is an invasive procedure, relatively uncomfortable, costly to the provider, and carries a small but calculable risk to the patient. Symptoms have often been ongoing for many years and at the time of endoscopy the clinician and patient are looking for confirmation that the mucosa of the GI tract is both macroscopically and microscopically normal. There are several conditions in which confirmation that the mucosa is normal at a microscopic level is useful, e.g. duodenal biopsies to exclude coeliac disease in iron deficiency anaemia (IDA), or right and left colonic biopsies to exclude microscopic colitis in chronic diarrhoea. Although in both these conditions the yield from histology is likely to be relatively low, the physician and patient are looking for reassurance. Therefore, the accompanying histopathology request form should reflect this low level of suspicion.
When and where to take biopsies in different regions of the GI tract is the subject of further discussion in the text that follows (see also Practice Points 1.1).
What Does the Gastroenterologist Want from the Histopathologist?
What the gastroenterologist wants depends on the circumstances of sampling. The endoscopist should of course submit sufficient clinical information on the accompanying request form to allow the pathologist to frame the findings within the clinical context. Ideally, a pathology request form will contain a question from the endoscopist to the pathologist – rather than ‘? Coeliac’ or, worse, simply ‘D2’, the clinician should write ‘Raised TTG, iron and folate deficiency, macroscopically normal duodenum; does the patient have coeliac disease?’ This allows the pathologist to refute or confirm the hypothesis rather than simply list a long differential for raised intraepithelial lymphocytes. What the gastroenterologist appreciates from the pathologist is to document the changes seen microscopically and to form a differential diagnosis but then, if possible, to conclude with the most likely diagnosis. At times, this can be difficult, but a report’s conclusion includes, ideally, a pathologist’s best guess or an idea as to what the diagnosis is, taking into account the information from the endoscopy report and/or the request form. If possible, the histological diagnosis should also consider the interpretation of previous biopsies. For example, a patchy colitis with features of chronicity may be Crohn’s disease, but if there are 10 previous reports documenting ulcerative colitis then it is likely that this is simply ‘treated UC’, which can often have a patchy appearance. Ideally, there is a forum for ongoing dialogue once a report has been finalised. Although a multidisciplinary team meeting (MDTM) is a common forum for the management of GI malignancy, MDTMs for non-neoplastic disease such as IBD remain hugely useful, and involvement of histopathologists enhances these meetings (Practice Points 1.2).
Document the microscopic features.
Offer a differential diagnosis.
Conclude with the most likely diagnosis, or with a best guess.
Take into account information from the endoscopy report and/or request form.
Consider previous biopsy findings and clinical picture.
Support multidisciplinary team meetings (MDTMs) for non-neoplastic disease.
Endoscopic Mucosal Tissue Sampling in Common Clinical Situations
The oesophagus is easy to access in the majority of gastroscopies but is often negotiated rapidly en route to the stomach. Consequently, oesophageal lesions can be missed.1 Routine biopsies from the oesophagus are not standard practice, unless there is a history of dysphagia. Sampling-specific lesions can be difficult because of the narrow confines of the oesophagus, the peristaltic movement of the oesophagus, and the quality of the stratified squamous epithelium. To help the pathologist, all samples should be labelled clearly with their site of origin. Typically, the label is the distance measured in centimetres from the tip of the scope to the incisors using the markings that run along the side of the scope. If relevant or required, further information can be given stating whether a lesion is seen on the anterior, posterior, left, or right wall of the oesophagus when the scope is in the neutral position (Fact Sheets 1.1 and 1.2).
Narrow band imaging to document extent (Prague classification)
Possible dysplasia is described (Paris classification) and sampled separately
If no suspicious lesions: quadrantic biopsies from 1 cm above the gastric folds and then at 2-cm intervals
If previous dysplasia: biopsies at 1-cm intervals
Suspected Eosinophilic Oesophagitis
Macroscopic: trachealisation, linear furrowing, white patches
Biopsies: at least six from at least two different sites (upper, middle, or lower oesophagus)
Gastro-oesophageal Reflux Disease
Common reason for endoscopy; mainly to exclude Barrett’s oesophagus and dysplasia
Biopsy is unnecessary for uncomplicated GORD or mild oesophagitis
Biopsy is usual for grade D oesophagitis to exclude dysplasia
Repeat endoscopy 6–8 weeks after GORD therapy to exclude Barrett’s oesophagus and neoplasia
Biopsy all ulcers to exclude dysplasia or malignancy
Follow-up endoscopy after 6 weeks following high-dose PPI.
Causes: GORD, IDA, EE, neoplasia, post-procedure
Biopsy if neoplasia is suspected
Crohn’s disease, tuberculosis, human papilloma virus, candidiasis, connective tissue disorders, sarcoidosis, and Kaposi’s sarcoma may require biopsy
The definition of Barrett’s oesophagus (BE) is an oesophagus in which any portion of the normal distal squamous epithelial lining has been replaced by metaplastic columnar epithelium which is clearly visible endoscopically (≥1 cm) above the gastro-oesophageal junction and is confirmed histologically from oesophageal biopsies2 (Figure 1.1). Although to a pathologist this may seem a fairly straightforward diagnosis, the presence or absence of a hiatal hernia, reflux oesophagitis, and difficulties defining the top of the gastric folds can all make establishing and classifying BE hazardous for an inexperienced endoscopist. The current guidelines encourage endoscopists to take their time assessing the mucosa, to wash the oesophagus carefully, to use narrow band imaging to document the disease according to the Prague classification, and to look for obvious signs of dysplasia.3 Suspicious areas identified as possible dysplasia should be described according to the Paris classification, sampled separately, and labelled clearly with site of origin to allow follow-up sampling. Accurate documentation guides subsequent endoscopic therapy such as endoscopic mucosal resection (EMR) and radiofrequency ablation (often documented as HALO®).3 If no suspicious lesions are identified, then quadrantic biopsies should be taken starting from 1 cm above the top of the gastric folds and then every 2 cm, according to the Seattle protocol.4 In patients with previous high- or low-grade dysplasia, samples at 1 cm intervals are appropriate.
Figure 1.1 Endoscopic appearance of Barrett’s oesophagus. The normal stratified squamous epithelium of the oesophagus has been partially replaced by proximally extending columnar epithelium, which is visible endoscopically.
Eosinophilic oesophagitis (EE) is a condition diagnosed histologically and defined by the presence of>15 eosinophils per high-power field (hpf) together with other characteristic histological features in the appropriate clinical setting (see Chapter 11). Clinically it is characterised by symptoms of odynophagia, dysphagia, and, in extreme cases, food bolus obstruction.5 Its incidence is rising through a combination of increased awareness and screening and is probably in line with overall rises in the incidence of the atopic conditions to which it is related. Endoscopically the features include trachealisation, linear furrowing, and white patches denoting immune cell aggregation, although up to 15% of patients have no endoscopic abnormalities (Figure 1.2). The sensitivity of one biopsy is 55% for detecting EE and this rises to 100% for six biopsies. Accordingly, the new British Society of Gastroenterology (BSG) guidelines suggest taking six biopsies from at least two different sites (upper, middle, or lower) when EE requires exclusion.3
Figure 1.2 Endoscopic appearances of eosinophilic oesophagitis. There is trachealisation of the tubular oesophagus with linear furrows and white exudates, which are the hallmark endoscopic features of eosinophilic oesophagitis.
Gastro-oesophageal Reflux Disease
Gastro-oesophageal reflux disease (GORD) is a common reason for endoscopic referral, typically to exclude Barrett’s oesophagitis or cancer. The incidence of GORD varies with country, probably reflecting variations in diet, obesity, alcohol consumption levels, and smoking rates. Estimated rates are 18%–27% in North America and 8%–25% in Europe.6 Routine biopsy is not necessary for uncomplicated GORD or for mild evidence of Los Angeles grade A–C oesophagitis. Biopsies should be performed for patients with evidence of Los Angeles grade D oesophagitis to look for the presence of dysplasia and these patients should have a repeat endoscopy 6–8 weeks after high-dose therapy with a proton pump inhibitor (PPI) to exclude Barrett’s oesophagitis and to survey for dysplasia7 (Figure 1.3).
Figure 1.3 Endoscopic appearances of grade D reflux oesophagitis in the distal oesophagus. There is significant erythema involving more than 75% of the oesophageal circumference with associated superficial ulceration and friability of the oesophageal mucosa.
Oesophageal ulceration is defined as a break in the mucosa of>5 mm. Biopsies should be performed for all oesophageal ulcers to look for dysplasia or malignancy and patients should have a follow-up endoscopy 6 weeks later following high-dose PPI.8
The finding of oesophageal strictures always raises the possibility of malignancy. If there is a suspicion of neoplasia, samples should always be taken. Other common causes of strictures are GORD, IDA, and EE. Strictures can also occur after surgery, radiofrequency ablation, radiotherapy, or endoscopic mucosal resection. There is weak evidence to suggest that even benign-looking strictures require sampling when they are first encountered (Figure 1.4). However, the sampling should not precede therapeutic procedures such as dilatation because the biopsy procedure could increase the risk of perforation.3
Figure 1.4 Endoscopic appearances of a benign oesophageal stricture at the gastro-oesophageal junction. The stricture has a smooth, regular contour and no depressed component. These features support a benign process.
The stomach is a large organ, easily accessed with the endoscope. Accurate labelling of samples aids the pathologist and informs future endoscopy. Briefly stated, the stomach is divided into the cardia, fundus, body, incisura, antrum, and pylorus. Further labelling may denote anterior or posterior and lesser or greater curve (Fact Sheet 1.3).
Features that raise the possibility of H. pylori infection include gastritis, duodenitis, and gastric and duodenal ulceration
Initial diagnosis relies on the urease test
If there is a suspicion of H. pylori infection in a patient receiving a PPI, biopsies from the antrum and from above the incisura are taken
Biopsies from the edge of gastric ulcers to exclude dysplasia
Biopsy at presentation to exclude malignancy and to confirm polyp type; further routine surveillance unnecessary
Targeted biopsies of suspicious lesions plus two non-targeted biopsies from the antrum, body, and incisura (Sydney)
Iron Deficiency Anaemia
Common reason for endoscopy
Biopsies of antrum and body to exclude gastric atrophy
Typically, initial diagnosis of Helicobacter pylori (H. pylori) infection relies on a point of care urease-based test in the endoscopy department. To avoid a false-negative result, PPIs should be discontinued 2 weeks prior to the gastroscopy.3 If patients are taking a PPI at the time of endoscopy and there is a suspicion of H. pylori infection, confirmatory biopsies should be taken to look for the presence of the bacteria. Whenever possible, a urease test should be performed, as it is fast (result within a few minutes), inexpensive, sensitive, and specific.2 For a urease test, one or two biopsies should be taken 5 cm proximal to the pylorus, in the lesser curvature near the incisura angularis or the greater curvature opposite the incisura angularis. In the presence of a PPI (that should ideally be stopped 1 week prior to endoscopy), bismuth, or antibiotics, the sensitivity of a urease test decreases dramatically.9 In this circumstance, adherence to a biopsy protocol such as the three-biopsy protocol or the Sydney protocol may be appropriate. The three-biopsy protocol comprises one biopsy from the greater curvature of the body, one from the greater curvature of the antrum, and one from the incisura angularis. The Sydney protocol comprises five biopsies encompassing the lesser and greater curvature of the antrum (2–3 cm from the pylorus), lesser and greater curvature of the body (8 cm distal to the cardia), and incisura angularis. These protocols appear equal as regards H. pylori detection rates (100% from both protocols in a retrospective study of 46 patients),10 so the three-biopsy protocol is often preferred. Typical endoscopic findings that would raise the suspicion of H. pylori infection include gastritis, duodenitis, and gastric and duodenal ulceration (Figure 1.5). The latest recommendations from the BSG suggest that H. pylori testing should also be performed in the presence of IDA.3
Figure 1.5 Endoscopic appearances of gastritis. There are multiple superficial erosions with relative hypertrophy of the pre-pyloric folds in the distal stomach.
Peptic ulcer disease can present with central epigastric pain that improves after meals, and occasionally with anaemia and GI bleeding (haematemesis, melaena from a bleeding ulcer). The characteristics of gastric ulcers, including site, size, morphological appearance, and suspicion of malignancy, should be documented carefully (Figure 1.6). Biopsies from the edge of gastric ulcers are necessary to exclude dysplasia. Testing for H. pylori should be performed.11