Preparation of a full-thickness bowel biopsy

Béla Veress
University of Lund
Department of Pathology
Malmö University Hospital
SE-205 02 Malmö, Sweden
E-mail: bela.veress@skane.se

The surgical theatre

Full-thickness intestinal biopsies comprising all layers of the intestinal wall should be taken.

The blood circulation to the area selected for biopsy should be maintained intact as long as possible in order to avoid ischaemic damage.

The biopsy should be cut out either parallel or perpendicularly to the long axis of the intestine. In the colon both taenia and intertaenial parts of the wall should be included.

A sharp surgical knife should be used, not laser knife or diathermy to avoid thermal damage.

The central portion of the biopsy must not be touched, drawn, or seized by any instrument during the whole process in the surgical theatre to prevent mechanical damage.

Figure 1.

The biopsy specimen should be placed on a piece of cork with the mucosa downwards, i.e. touching the cork and the peritoneal surface face-up. The specimen should be fastened to the cork by pins immediately at the margin of the muscle coat (Figure 1).

It is important not to stretch or draw the specimen with the pins but to let it remain in its "natural size" in order to avoid stretching artefacts.

The optimal size of the biopsy is 15 x 10 mm measured on the peritoneal surface area (NOT mucosal surface area). The minimum size is 10 x 5 mm.

Figure 2.

The cork plate with the biopsy should be placed in 10% buffered formalin with the specimen downwards and the specimen-free surface facing the air. Then the plate should be covered by a paper cloth to press down the cork in order to ensure optimal fixation (Figure 2). The specimen should be transported to the pathology laboratory as soon as possible.

The pathology laboratory

The optimal length of fixation is 24 hours at room temperature. "Overfixation" should be avoided since it may diminish the number of available epitopes, hence jeopardizing the results of immunohistochemistry.

At cutting the specimen is divided into three pieces of different size: two pieces for light microscopy, and one for electron microscopy.

Figure 3.

1.	About half of the biopsy is cut into two slices: one parallel and the other transversal to the long axis of the specimen. These slices will be oriented at embedding in paraffin to allow transverse sectioning of the whole bowel wall with all its layers for light microscopy, (Figure 3).
2.	About one third of the biopsy is embedded also in paraffin but for sectioning parallel to the serosal surface. This is to ensure examining the largest possible area of the myenteric plexus. The mucosa and submucosa can be removed by using a scalpel under the stereo-microscope. This step, however, is not advisable in the case of suspected Hirschsprung's disease when the en face examination of the submucosa makes it easier to find giant ganglia of Meissner. The biopsy should be embedded so that the sectioning starts from the peritoneal surface. one of the surfaces can be marked with Indian ink in order to facilitate orientation at embedding (Figure 3).
3.	The muscle coat of the remaining part of the specimen is prepared for electron microscopy by dividing it into mm-sized pieces before embedding into resin.

Sectioning

1.	Transversal cuts: Sections should be obtained from at least three different depths of the paraffin block. At each level, 3-5 sections are taken up for object glasses. These sections will be stained with haematoxylin and eosin. Apart from these, between levels 2 and 3, sections should be taken up and reserved for special staining (4 sections) and for immunohistochemistry on special glasses (10 sections).
2.	Parallel cuts: The myenteric plexus is located between the circular and longitudinal muscle layers. Therefore, the subserosa and the larger part of the longitudinal muscle layer should be removed before starting the serial sectioning of the intermuscular layer. The approximate depths of this layer is determined under the light microscope at the examination of the transversal sections by measuring the thickness of the subserosa and the outer muscle layer. The whole intermuscular layer should be sectioned and 4-6 sections are reserved for immunohistochemistry.
3.	Electron microscopy: Sections are prepared for electron microscopy whenever light microscopy gives insufficient evidence for diagnosis, when special features need to be analysed, or for research purposes.

Staining

1.	Transversal sections are routinely stained with haematoxylin och esosin (H&Eo). When special tissue-components or structures should be analysed the following stainings are performed: (a) "inclusions" (PAS, PAS-diastase, Gömöri's methenamine silver, von Kossa, and immunohistochemistry for cytomegalovirus/herpes virus); (b) fibrosis (one of the trichrome-stainings); (c) amyloid (Kongo-red and polarisation microscopy); (d) mast cells/eosinophils (Giemsa); (e) chromatolysis (cresyl-violet). Immunocytochemistry always includes desmin, actin, smooth muscle actin (alfa-actin), S100 and neuron specific enolase (NSE). If requires also CD3, CD20, CD43, and CD45 for lymphoid cells.
2.	Parallel sections. One section each from the two peripheral parts and one section from the middle of the intermuscular layer are stained with H&Eo, one section with cresyl-violet, the others are reserved for immunohistochemistry.
3.	Immunocytochemistry. The following cellular components or cells are always analysed: (a) intermediate muscle filaments (actin, alpha-actin, desmin); (b) neurons (neuron specific enolase, neurofilament, bcl-2); (c) Schwann cells/glial cells (S 100); (d) lymphocytes (CD45).
4.	Silver impregnation. This method on 50 µm thick frozen sections was used previously but after introduction of immunohistochemistry on parallel sections from paraffin blocks the method has been abandoned.

Morphometry

The thickness of the circular and longitudinal muscle layers is measured in the microscope by using an ocular micrometer. If necessary, the size of smooth muscle cells and neurons can also be measured. The number of ganglia and neurons can be counted and reported as number/mm in transversely cut sections and number/mm-square in parallel sections.

The counting of intraepithelial lymphocytes is best performed on CD45-stained sections. At least 300 epithelial cells should be counted and the lymphocyte count is expressed as lymphocytes per 100 epithelial cells.

Microscopic analysis

Microscopy always starts with examining the transversely cut sections. At this stage it is decided whether the preparation of parallel sections are necessary or not (see above Sectioning, #2).

All tissue components must be examined starting from the epithelium and ending at the peritoneum. Furthermore, every section should be examined for the detection of patchy lesions (e.g. ganglionitis).

Special points of interest are: the number of intraepithelial lymphocytes, neurons in the l. propria, vacuoles or inclusions within smooth muscle cells, inclusions in neurons, "swollen"- or "shrunken"-degeneration of neurons, neuron/axon vacuolisation, gliosis, lymphocytes/eosinophils/mast cells around and within ganglia and along small nerve fibres, fibrosis, architecture (layers, orientation) of the muscles, and thickness of the circular and longitudinal muscle layers.

Regarding the inflammatory conditions the following points should be kept in mind:

(a) normally there are no lymphocytes, eosinophils or mast cells within the ganglia (neutrophils can be found in any location and regarded as the consequence of the surgical procedure if there is no sign of tissue damage);

(b) the inflammatory infiltrate within the ganglia is often discrete and always patchy. The examination of parallel sections is very much recommended if inflammation is suspected;

Pitfalls

Mechanic, thermic, and ischaemic damages should be taken into consideration at the interpretation of the microscopic morphology. Such damage can mimic cellular degeneration. Shrunken, eosinophilic cytoplasm with "pyknotic" nucleus of a well defined group of muscle cells with absence of intermediate filaments or, the contrary, diffuse very dark staining for all kinds of filaments should be interpreted with cautiousness. Special attention should be given to the examination of the neighbouring fibrous soft tissue. The edges of the biopsy must not be considered at all. Neuronal changes in the vicinity of such areas should be regarded as artefacts.

Medical history data are important in the interpretation. Previous abdominal operation(s) can result in adhesions with secondary hypoxic changes of the neurons or a compensatory hypertrophy, hyperplasia or extra muscle layer of the muscle coat. Therefore, the pathologist must be informed about previous operations.

General

Preparation of a full-thickness bowel biopsy

Scientific