19 The Contribution of Histopathology to Endoscopy
According to Rudolf Virchow (1858), the etiology of all diseases can be boiled down to alterations of single cells. 1 Thus morphology and visualization is the basis by which diseases can be identified. It all began in about 1595 with Hans Janssen, who probably built the first microscope. Janssen’s creation was nearly 45 cm in length and could magnify objects three- to ninefold. 2 During the 17th century, scientists in The Netherlands, England, and Italy began to dissect insects and plants and thus the microscope became the foundation for the natural sciences. At the end of the 18th century, Bichat was regarded as the founder of histology with his descriptions of 21 different tissues. 3 The first professional histologic laboratory was opened midway through the 19th century in Prague (1851) followed by one in Leipzig (1875). 4 , 5 Prerequisites were not only the optics but also the development of specimen fixation, embedding and preparing paraffin blocks, cutting with a microtome, and, last but not least, histochemical staining. The majority of histologic staining methods were described until the 1930s. Immunohistochemical methods were introduced in the 1980s, from 1990 on in situ hybridization, and from 2000 on, molecular pathology became more and more sophisticated and relevant for treatment and diagnosis.
Parallel to these developments, optical endoscopic visualization improved markedly in the last few decades. Endoscopy and histopathology share the morphologic analysis as their basis and these share the characterization of epithelia, vessels, infections, and neoplasia, among others. Whereas identification of color and pit pattern is the domain of endoscopy, histopathology can deliver risk factors and precise analysis of cells and tissue and depth of infiltration to build the basis for modern diagnosis (▶Fig. 19.1) and treatment options (▶Fig. 19.2).
Endoscopy and histopathology are interrelated. Sampling without prior precise surface analysis will not lead to representative biopsy material whereas histopathology allows more exact analysis of inflammation, possible infection, and better classification of tumors. Finally, by demonstration of mutations such as RAS in colorectal carcinoma, histopathology at the subcellular level has a direct consequence for therapy. The main contribution of histopathology has to be seen in the fact that histopathology delivers the precise etiology of changes seen endoscopically or by other methods. Therefore, it is essential that pathologists always report cause and etiology of their findings in their histopathologic diagnosis. Histopathology is important for screening for neoplasia, confirmation of clinical diagnosis, early diagnosis of neoplasia, differential diagnosis between benign and malignant, diagnosis of metabolic disease drug-induced changes, parasitic, bacterial, viral infections, inflammatory diseases, detection of immunopathologic diseases, intraoperative consultation for further treatment, and, last but not least, science.
Close collaborations is warranted between endoscopists and pathologists for a precise diagnosis. For colorectal biopsies, the benefits of such a collaboration have been clearly demonstrated (▶Fig. 19.3). 6 It makes no sense at all to hide key information from the pathologist with the misguided notion of introducing objectivity to histologic diagnosis. 7 The endoscopist should provide the pathologists with a copy of the endoscopic report (optimum) but at least with the endoscopic diagnosis (minimum), preferably with images (in case of unexpected or unusual findings). In the United States this is currently not the case; pathologists rarely receive any endoscopic images. Pertinent clinical data include duration of symptoms, previous medications, and general clinical information such as prior diagnoses.
On receipt of a sample, standard protocol is for the pathologist to describe the received material. In the case of biopsies, the number of biopsy fragments received is recorded, but more detailed descriptions of the fragments are usually not necessary. The size of the fragments is mentioned. In all other cases, a gross examination is performed to ensure precise information of the resection margin status. In specimens removed for neoplasia, measurements of the tumor and its relationship to the specimen’s margins are recorded to ensure proper future therapeutic steps and allow proper staging. For cases for which clarification about the sample is required, the pathologists need to contact the clinician. Regularly scheduled clincopathologic conferences ensure proper flow of information, specimens, and diagnoses. Endoscopists and pathologists need to understand the requirements of the other part and certainty of diagnostics (▶Fig. 19.1, ▶Fig. 19.4).
Proper orientation is important to avoid artefacts that can interfere with the pathologist’s ability to interpret the histopathologic findings. Endoscopic resections should be carefully oriented in the endoscopy suite to ensure proper analysis of the resection margins. Specimens can be orientated by thread marks or latex colors or even India ink. Pinning these samples on thick paper or cork helps as well. Stretching a specimen prior to fixation should be definitively avoided. Specimens should be loosely pinned on paper or cork. The reasoning is that specimens can shrink up to 50% during fixation with formalin and this may cause artefacts that hamper the histologic diagnosis. Pinning through the lesion itself and close to the edges of the lesion should be avoided as well.
In addition, it should be noted that adequate fixation also is essential for all samples, including those obtained for cytologic evaluation (endoscopic ultrasound–fine-needle aspiration [EUS-FNA]). Some methods, such as electron microscopy, require special fixatives, whereas frozen sections and some molecular analyses require fresh nonfixated material. Immunohistochemical analyses can deliver false results if the incorrect fixative was used. If the endoscopist is unclear about the fixative, the pathologist’s office should be contacted for clarification.
19.2 Clinical Impact of Histopathology by Segment within the Gastrointestinal Tract
The most frequent indication for taking esophageal biopsies is refractory gastroesophageal reflux disease (GERD).
However, there are some limitations to the clinical impact of such biopsies: first, there are no clear recommendations where exactly to take them and how many biopsies should be obtained. Secondly, using biopsy protocols and clear histopathologic grading schemes increases the utility of such samples. It has been shown that when clear recommendations are introduced, histopathologic results correlate with endoscopic findings in GERD. 8 , 9 However, such protocols have not been widely introduced and thus, in classical reflux disease with no suspicion for mimickers or neoplasia, biopsy results do not deliver information beyond the endoscopic diagnosis since the treatment and dosage of drugs are based on the combination of typical symptoms and lesions for prescribing the full dosage of proton pump inhibitors (PPIs) compared to cases with no mucosal breaks (nonerosive reflux disease [NERD]) for which patients are prescribed half the standard dosage of PPI. It is noteworthy that the vast majority of patients with GERD present with endoscopically negative reflux disease rather than with mucosal breaks.
It should be noted that in addition to GERD, esophageal inflammatory conditions also include various infections, drug-associated injury, neoplastic lesions, and even esophageal involvement in chronic inflammatory bowel disease (IBD). In such cases, histology has a clinical impact since the etiology of the inflammation can often be given and thus sufficient treatment options discussed. Biopsies in patients in whom there is a clinical suspicion for GERD should be taken (if at all) within a range of 0.5 cm above the Z-line and preferably in line with the lesser curvature of the stomach since histologic changes are known to be more striking there than elsewhere.
Cases suspicious for Barrett’s esophagus (BE) should be biopsied according to the Seattle protocol, namely with four biopsies in every 2 cm. Unfortunately, this protocol is rarely followed in routine cases, but, when strictly adhered to, the Seattle protocol ensures a high probability of detecting dysplasia, if any. No biopsies should be taken from a straight, normal-appearing Z-line. Definitions of BE differ worldwide. In the United States, biopsies from columnar tongues longer than 1 cm are recommended. If goblet cells (intestinal metaplasia) are detected histopathologically, the diagnosis of BE is proven. However, there are variable different definitions worldwide with and without goblet cells with and without columnar metaplasia identified endoscopically (e.g., British Barrett’s guidelines, German Barrett’s guidelines, Japanese Barrett’s guidelines, European Barrett’s guidelines, Czech Barrett’s guidelines). Pathologists are generally unaware of the endoscopic appearance and may state that the specimen is “suggestive of BE” when goblet cells are present. However, if there are tongues less than 5 mm or an irregular Z-line, the patient likely has “intestinal metaplasia of the Z-line.” It is important that gastroenterologists correctly convey this information to their patients.
In patients for whom there is a clinical suspicion for eosinophilic esophagitis (EOE), biopsies should be taken from the distal, middle, and upper esophagus (two each in different containers) to allow differentiation of eosinophils in GERD within the distal esophagus from eosinophilic esophagitis, in which more proximal eosinophils are expected. The eosinophilic infiltrates in EOE can be rather discontinuous, focal, and sometimes even very sparse (< the requested 20/mm2) especially in patients taking PPI or partially treated with steroids and thus multiple samples are ideal.
In patient in whom there is a suspicion of moniliasis/candidiasis, biopsies are recommended from the whitish sloughed material with and without mucosal breaks; in other words, from the edge or erosions/ulcers that encompass both the exudate/granulation tissue as well as the intact epithelium. The same applies to cases with suspicion for viral infections such as herpes simplex virus (HSV, which is detected in the edges of a necrosis of an ulcer) and cytomegalovirus (CMV, which is principally detected in the endothelial cells in granulation tissue or the center of an ulceration).