Techniques for the standard histological and ultrastructural assessment of nerve biopsies

1 INTRODUCTION

Sural nerve biopsy remains the primary window to the peripheral nervous system for the detailed examination of morphological changes and an important differential diagnostic tool in many cases of peripheral neuropathies. However, due to the advances in molecular genetics and imaging and in the establishment of blood biomarkers, nerve biopsies are performed less frequently nowadays than in the past. In our institutions, nerve biopsies have declined in number by 50% compared with the 1980s and 1990s, with numbers having now reached a lower plateau.

Nerve biopsy should only be performed following a complete clinical, electrophysiological and laboratory examination. The biopsies should be processed by a laboratory with sufficient experience in the techniques described in the present review. Nerve biopsy may then be very helpful in disclosing inflammation or other important interstitial pathology such as amyloidosis or immunoglobulin deposits associated with dysproteinemic neuropathy. It may also give clues to the underlying gene defect in unsolved cases of familial neuropathy, and may reveal combined pathology, for example, microangiopathic/diabetic and inflammatory. Furthermore, biopsy tissue is invaluable in the research on these diseases, as the tissue changes provide important clues towards the understanding of the underlying pathological processes. Archival specimens, especially paraffin and resin blocks, can be retrieved virtually unaltered after many years and can be a valuable source of information, for example, in cases of hereditary neuropathy.

As nerve biopsy is not as widely applied any more as it used to be, it becomes even more important to instruct clinicians as well as researchers on the benefits and pitfalls of this technique. Here, we provide an overview on the current status and methodology of nerve biopsy that is addressed to clinicians, neuropathologists and basic research scientists alike.

2 THE NERVE BIOPSY AND ITS WORKUP 2.1 Biopsy sites

The biopsy is generally obtained from the sural nerve because this nerve is easily accessible from the surgical point of view and because most neuropathies have a length-dependent distribution, invariably involving the sural nerve. Moreover, since the sural nerve is purely sensory in more than 90% of patients and contains only a minimal part of motor fibers in the remaining,1 sural nerve harvesting in most cases causes only a small area of numbness, without motor deficits.2 Conduit repair has been discussed.2, 3

The sural nerve usually consists of 5 to 10 nerve fascicles containing numerous large and small myelinated as well as unmyelinated nerve fibers (Figure 1A). It is easily studied by the electrophysiologist, which is relevant because it allows the choice of an affected nerve and eventually the selection of the side, which seems more likely to provide diagnostic information. When a vasculitis is suspected, for example, it is possible that one side is much more affected than the other, and in this case, it is generally advisable to harvest the most damaged nerve. On the other hand, in a symmetric polyneuropathy, it may be better to avoid a terminally depleted nerve since this can be less informative.4

image

(A) Two normal sural nerve fascicles. Semithin section, toluidine blue. Scale bar = 250 μm. (B) Considerable reduction in myelinated nerve fiber density and prominent microangiopathic, obliterative thickening of the walls of endoneurial blood vessels (arrows) in a 50 years old male patient with chronic sensorimotor neuropathy. Semithin section, toluidine blue. Scale bar = 100 μm. (C-E) Amyloid neuropathy. (C) Homogeneous staining of amyloid deposit (arrow) associated with an endoneurial vessel in a semithin section. 71 years old male patient with ATTR amyloidosis. Only few remaining myelinated nerve fibers. Toluidine blue. Scale bar = 60 μm. (D) Strong thioflavin S labelling of endoneurial amyloid in a 75 years old male patient. Scale bar = 125 μm. (E) Preferential loss of small myelinated nerve fibers in a biopsy obtained from a 79 years old female patient with amyloidosis associated with lambda light gammopathy. Arrow: myelin debris. Semithin section, toluidine blue. Scale bar = 60 μm

Some authors support the choice of the superficial peroneal nerve along with specimens from the adjacent peroneus brevis through the same surgical procedure.5, 6 This might be very useful when a vasculitis is suspected, because it increases the number of arterioles and small vessels available for examination, thus increasing the diagnostic yield.5

When a neuropathy involves mainly the upper limbs, for example when leprosy occurs without skin lesions, a superficial radial nerve biopsy may be considered. However, its removal implies a sensory loss of the dorsal surface of the hand; therefore, it is advisable to perform a fascicular biopsy.7 Finally, the biopsy of the obturator nerve branch to the gracilis muscle might be helpful in the differential diagnosis between motor neuropathies and lower motor neuron diseases.8, 9

2.2 Complications

Normally, sural nerve biopsy is followed by a permanent numbness in the area supplied by the nerve just below the lateral malleolus. However, in many patients affected by a severe neuropathy, the sensory impairment may already be advanced such that the effects of nerve transection go largely unnoticed. In rare cases, sural nerve biopsy may be followed by neuroma formation with persistent neuropathic pain.10 Other risks, common to any surgical procedure, include bleeding, delayed healing, infection and scarring. In the case of superficial peroneal nerve biopsy, if underlying peroneus brevis muscle is also biopsied, there may be a slight weakness in foot dorsiflexion.

Since the sural nerve is particularly susceptible to mechanical damage, undesired damage to the specimen may occur following crush, stretch or excessively long surgical procedures, thus leading to artefacts. In teased fiber preparation, the use of forceps can lead to artificial widening of the paranode. Fortunately, most of the artefacts do not compromise the prospects of reaching a diagnosis.

2.3 Processing and fixation

Tables 1 and 2 provide an overview of the nerve biopsy workup and of commonly applied techniques and targets of examination. The fresh, unfixed nerve specimen should be processed as soon as possible after removal and should be divided into fractions, each at least 1.5 to 2.0 cm long, for formalin and glutaraldehyde fixation and optionally for freezing in liquid nitrogen-cooled isopentane. The proximal and distal part is recommended for cryo- and formalin fixation, respectively, as these methods are somewhat less sensitive to the inevitable compression artefacts at the edges of the specimen. The frozen proximal part can be cut with a cryostat and used immediately for a preliminary diagnosis and stored at −80°C for later immunohistochemistry/immunofluorescence or as a protein, DNA and RNA source for molecular pathology, genetic or biochemical examination. The distal segment is to be fixed in 4% phosphate-buffered formalin solution, preferably overnight, then further dissected and embedded in paraffin arranged in a way to yield both cross and longitudinal sections. The remaining middle segment should be fixed with 3.9% phosphate-buffered glutaraldehyde solution, postfixed with 1% osmium tetroxide for contrast enhancement and embedded in epoxy resin for semithin (approx. 1 μm thickness) and ultrathin (less than 100 nm thickness) sectioning. Semithin sections are stained with toluidine blue and/or paraphenylenediamine for fine detail light microscopy. Ultrathin sections are contrasted with heavy metal stains (such as uranyl acetate and lead citrate) and examined with an electron microscope.

TABLE 1. Nerve biopsy workup image TABLE 2. Overview of commonly applied techniques and targets of examination Initial evaluation FFP Hematoxylin and Eosin (H&E)

Epineurium

Perineurium

Vascular pathology

Endoneurial fibrosis, edema

Inflammatory or neoplastic infiltration

GA Toluidine blue and/or paraphenylendiamine stained semithin sections (cross, as well as longitudinal)

Nerve fiber density and distribution

Myelin alterations (hypo- or hypermyelination, focally folded myelin, tomacula, etc.)

Onion bulbs

Bands of Büngner, myelin debris/myelinophagia

Segmental demyelination on longitudinal section

Axonal degeneration/atrophy

Clusters of regeneration

Vascular alterations including microangiopathy, arteriosclerosis, vessel wall necrosis, calcifications, recanalisation, inflammatory infiltration, etc.

Histochemistry FFP Turnbull blue Iron deposition Congo red Amyloid accumulation Thioflavin S Gomori trichrome Myelin abnormalities Luxol fast blue van Gieson Collagen and elastic fiber alterations Immunohistochemistry FFP Neurofilament Neurofilament accumulation, fiber loss PGP9.5 Fiber loss, axonal swellings Immunoglobulin (IgE, IgM, IgG, IgD, IgA) Ig deposition κ- and λ-light chain Amyloid neuropathy Transthyrethin TTR amyloidosis EMA Cells with perineurial differentiation S100 Schwann cell marker Myelin basic protein (MBP) Myelin sheath pathology LCA, CD3, CD4, CD8, CD20 Inflammatory infiltration CD68 Macrophage marker Smooth muscle antigen (SMA) Vessel wall alterations CD34 Electron microscopy GA Ultrathin sections

Myelin changes (decompaction, macrophage-mediated demyelination, hypo- or hypermyelination, etc.)

Axonal degeneration/loss, spheroids/altered autophagy

Accumulations of cell organelles or other material (axonal or Schwann cell)

Mitochondrial pathology

Nodes of Ranvier

Unmyelinated nerve fiber alterations

Abbreviations: FFP, Formalin-fixed, paraffin-embedded tissue; GA, 3.9% glutaraldehyde-fixed, resin-embedded tissue. 2.4 Paraffin and resin section histology

The initial evaluation of the nerve should start with H&E stained paraffin sections, which provide a good overview of interstitial pathology of the epi-, peri- and endoneurium. Fibrosis and inflammatory or neoplastic infiltrates as well as granulomas can be observed here, as well as vascular alterations such as microangiopathy, (fibrinoid) vessel wall necrosis and vascular obstruction and recanalization. Further features include (endoneurial) edema, calcification, especially of vessel walls and of the perineurium.

With the advent of effective treatments of amyloidosis, the detection of amyloid deposits has become even more important. Congo red-stained sections should be examined using polarized light to detect the characteristic green birefringence. The thioflavin S stain (Figure 1D) has become part of our standard repertoire because it is more sensitive and reliable than the Congo red stain, at least in our hands (JW, IK, SN); however, it does require fluorescence microscopy. Other tinctorial stains such as Gömöri trichrome, Ladewig, elastica van Gieson, and myelin stains such as luxol fast blue as well as stains for acid-fast bacilli (Ziehl-Neelsen, see Figure 2A; Wade-Fite) are considered optional.4

image

(A) Lepromatous neuropathy. Acid-fast bacilli (arrows) in endoneurial cells (Schwann cells and/or macrophages). Paraffin section, Ziehl-Neelsen. Scale bar = 100 μm. (B) Chronic vasculitic neuropathy associated with rheumatoid arthritis in a 46 years old male patient. Patchy, focally accentuated loss of myelinated nerve fibers. Semithin section, toluidine blue. Scale bar = 200 μm. (C) Thickened wall of an epineurial blood vessel infiltrated by LCA-immunoreactive cells (brown) in a case of chronic vasculitic neuropathy in a 70 years old female patient. Paraffin section. Scale bar = 60 μm. (D) Recanalized blood vessel in the sural nerve of a 43 years old male patient with chronic vasculitic neuropathy. Paraffin section, smooth muscle antigen staining. Scale bar = 60 μm. (E) Renaut bodies (arrows). 50 years old male patient. Semithin section, toluidine blue. Scale bar = 150 μm. (F) Compression artefact leading to myelin dislocation, focally mimicking tomacula. Semithin section, toluidine blue. Scale bar = 100 μm

For the assessment of nerve fiber pathology, paraffin sections are not suitable because they show prominent myelin swelling and splitting; axons are difficult to discern. In some cases, nerve biopsies are sent primarily to general pathology labs and are formalin-fixed and paraffin-embedded in their entirety. If such material is then forwarded to us for expert review in the framework of the German Reference Center for Nerve and Muscle Pathology (JW), we use neurofilament and protein gene product 9.5 (PGP9.5) immunohistochemistry, which may facilitate a rough estimation of nerve fiber density, but does not provide any more detailed information on nerve fiber changes.

2.5 Resin semithin sections

Semithin sections of the glutaraldehyde-fixed, resin-embedded material offer far better morphology, especially in terms of axonal and myelin pathology, but also regarding vessel wall alterations, than paraffin sections (Figures 1 and 2). For each biopsy, we examine cross sections from two blocks and longitudinal sections from one block at the minimum. In semithin cross sections of well-fixed nerve tissue, alterations of axon and myelin can be assessed reliably. These include myelinated nerve fiber loss, axonal atrophy/shrinkage, hypertrophy/giant axons, axonal dystrophy and regeneration as well as myelin breakdown, de- and remyelination, tomacula, focally folded myelin and onion bulb formations. Longitudinal sections are especially well suited for the observation of node of Ranvier alterations, segmental de- and remyelination and focal myelin abnormalities (eg, tomacula, focally folded myelin). Most of these and other nerve fiber changes characteristic or even specific for certain types or groups of hereditary neuropathies are best recognized in semithin sections combined with electron microscopy.11, 12 Morphometry is used mainly in scientific settings.6 Because they provide excellent morphological detail, semithin sections are actually well-suited for morphometrical studies; morphometry might be applied more widely in the future with the introduction of digital pathology technology.

In cases of amyloid neuropathy, a preferential loss of small myelinated fibers may be present in semithin sections that are not always accompanied by overt deposits of congophilic material in paraffin sections. In such cases, serial paraffin sections stained with Congo red and thioflavin S should be examined. Occasionally, amyloid may only be found in the semithin sections as irregularly shaped, homogenously toluidine blue-positive extracellular endoneurial deposits often associated with blood vessels. Actually, many alterations of blood vessels and other interstitial pathology are easier to recognize in semithin resin than in paraffin sections. Taken together, these advantages have led to the conclusion that resin histology is an essential component of the nerve biopsy workup.4, 13

2.6 Immunohistochemistry

We perform standardized automated paraffin immunohistochemistry (IHC) using antibodies against leukocyte common antigen (LCA; CD45), CD3 (T cells), CD8 (cytotoxic T cells) and CD68 (monocytes/macrophages) in every case. In many cases, serial section IHC using LCA (Figure 2C), CD8 and CD68 antibodies is performed if the initial IHC yielded equivocal evidence of inflammation only and if clinical findings and/or the presence of focally accentuated nerve fiber loss in the semithin sections suggest a neuritic or vasculitic process. Smooth muscle antigen immunohistochemistry may be used to detect vessel wall alterations, especially in cases of suspected vasculitis (Figure 2D). Further optional stains include CD4 (T helper cells) and CD20 (B cells and B cell lymphoma cells), neurofilament and PGP9.5 for axons (see above), immunoglobulin and light-chain antibodies in cases of monoclonal gammopathy or amyloidosis, transthyretin and gelsolin in cases of hereditary amyloidosis, S100 and myelin basic protein (MBP) for Schwann cells/myelin sheaths and epithelial membrane antigen (EMA) for perineurial cells. In most laboratories, direct immunofluorescence, previously used to highlight epineurial deposits of immunoglobulin, complement and fibrinogen, has not been used since years, following the introduction of anti-MAG and anti-gangliosides antibodies plasma dosage.14

2.7 Teased fiber preparations

Teasing allows to examine individual nerve fibers with consecutive internodes and has a high sensitivity in identifying demyelination or axonal impairment.15 A fragment measuring 1 to 1.2 cm is fixed in 2.5% glutaraldehyde immediately after biopsy, then incubated in osmium tetroxide 4% overnight. Thereafter the nerve is conserved in glycerin.

With the help of a dissecting microscope and curved, pointed forceps, connective layers are stripped away, leaving 20 to 30 fibers that are aligned on glycerinated glass slides. Teased fiber preparation allows to visualize Wallerian degeneration (Figure 3B), axonal swellings, segmental demyelination (Figure 3C), remyelination and to identify redundant myelin figures, including tomacula and outfoldings.6, 15, 16 Moreover, the frequency of all of these features should be recorded to calculate the percentage of fibers carrying each abnormality. Even though many centers consider teased fiber preparation too time-consuming, we still value this technique as an important contribution.

image

(A-C) Teased-fiber preparation showing from top down: (A) three consecutive portions of the same single, normal fiber with several nodes of Ranvier; (B) three consecutive, overlapping portions along the length of the same, single teased fiber undergoing Wallerian degeneration in a patient with vasculitic neuropathy; (C) three consecutive portions along the length of the same, single teased fiber showing segmental demyelination (segment marked by arrow heads).Scale bar = 100 μm. (D-F) Electron microscopy of ultrathin sections. (D) Normal unmyelinated axons (arrow) in a Remak bundle. Scale bar = 2 μm. (E) Remak bundle devoid of unmyelinated axons. Bundles of collagen fibers are enwrapped by Schwann cell processes (the so-called collagen pockets; arrows). Scale bar = 2 μm. (F) Focally folded myelin and incipient onion bulb formation in the case of CMT4C due to SH3TC2 mutation. Arrows: Redundant Schwann cell process (black) and basal lamina (white). Scale bar = 1.5 μm

2.8 Electron microscopy

Conventional light microscopy cannot visualize most ultrastructural details and at most can only hint the presence of certain ultrastructural abnormalities. Changes detectable by transmission electron microscopy (TEM) include abnormal myelin folding (Figure 3F), un- or decompacted myelin, macrophage-mediated demyelination, protein aggregates, organelle accumulations and abnormal autophagic vacuoles as well as abnormal inclusions such as zebra and tuff stone bodies in metachromatic leukodystrophies and trilaminar structures in adrenoleukodystrophy/adrenomyeloneuropathy.17, 18 Loss of unmyelinated fibers (Figure 3B) and other alterations of this nerve fiber population, for example, the abnormal processes and basal laminae of Schwann cells in Remak bundles found in CMT4C due to SH3TC2 mutation,19 are detectable only by TEM as they are beyond the resolving abilities of bright field microscopy. For the diagnosis of small fiber neuropathy, examination of skin biopsies is an established method, which is discussed elsewhere in comparison to nerve biopsy.6, 20

2.9 Unfixed fresh frozen tissue

Frozen sections have been frequently used in the past to detect certain types of antigens in sural nerve biopsies (ie, IgM deposits in myelinated fibers). The development of antibodies that work well with paraffin sections largely reduced the use of frozen sections due to the inferior morphological detail in the latter. Frozen sections may still be used to detect some antigens, like lipid-based ones, which are sensitive to alcohol treatment of the tissue, and for research purposes. The frozen material may also be very helpful for molecular pathology, genetic and biochemical studies.

2.10 Artefacts, pitfalls and things to keep in mind

In many cases, combined pathologies may occur, most frequently combinations of microangiopathy with inflammatory changes or toxic or hereditary neuropathy. Less frequently, a combination of hereditary neuropathy with inflammation is found. Up to 4% of biopsies labelled as sural nerve actually consist of a smaller vein that has been mistaken for the nerve.4 “Toothpaste” artefacts mimicking tomaculous fibers may arise due to compression of the nerve during surgical removal or afterwards; such alterations are especially prominent in small biopsies. Hyperosmolar fixation leads to shrinkage of the nerve fibers and of the fascicles as a whole. Initial fixation in formalin or freezing before or after fixation causes major nerve fiber damage. And, last but not least, Renaut bodies (Figure 2E), which rather frequently occur in nerve biopsies, may be mistaken for nerve infarcts, a rare finding. Renaut bodies correspond to subperineurial deposits of extracellular matrix containing elastic fiber components21 and sparse elongated cells and arise due to chronic mechanical stress.

2.11 The nerve biopsy report

The report should include a short summary of the relevant clinical aspects of the case and a description of the specimen received and of its workup. Features to be routinely examined and reported are listed in Table 3.

TABLE 3. Features to be described in a nerve biopsy report (Weis et al, modified) Status of the epineurium including blood vessels Alterations of the perineurium (thickening, fibrosis, calcification) Endoneurial edema, fibrosis Endoneurial blood vessel pathology, esp. basal lamina reduplication/thickening Density of large and small myelinated nerve fibers Extent of axonal degeneration and atrophy Axonal swelling and dystrophy, giant axons Frequency of bands of Büngner and macrophages containing myelin debris Macrophages (CD68 staining): diffuse increase, clusters Regeneration clusters Demyelinated/remyelinated fibers; tomacula Onion bulb formations Inflammatory infiltrates (LCA, CD3, CD8 immunohistochemistry) Presence/absence of amyloid (Congo red, thioflavin S or T, transthyretin immunohistochemistry) ACKNOWLEDGEMENT

Open access funding enabled and organized by Projekt DEAL.

AUTHOR CONTRIBUTIONS

Joachim Weis: Writing of the manuscript and providing cases. Istvan Katona: Writing of the manuscript and providing cases. Stefan Nikolin: Providing cases. Lucilla Nobbio: Nerve biopsy processing and manuscript correction. Valeria Prada: Nerve biopsy processing and manuscript correction. Marina Grandis: Writing of the manuscript and providing cases. Angelo Schenone: Writing of the manuscript and providing cases.

留言 (0)

沒有登入
gif