The mechanism of tissue destruction in amyloidosis of nerve, as in other organs, is not known. Each amyloidogenic protein (immunoglobulin light chains, or mutated-TTR, GEL, APOA1) is not expressed in the perikaryon or soma of small and large myelinated and unmyelienated fibers nor in Schwann cells. We do not see within the neural tube at or beneath basement membranes or within Schwann cell cytoplasms deposition of amyloid. Rather each fibril forming protein has humoral circulation with deposition in the interstitium of affected nerve fascicules. Endoneurial microvessels most commonly have amyloid deposition with associated apparent spread of amorphous acellular amyloid (Fig. 31.1). There is typically axonal degeneration at the level of the amyloid deposition but also often remotely to the discovered protein. It has been postulated that, due to its tendency to involve blood vessel walls, amyloids’ effects might be ischemic, as was initially put forth by Kernohan and Woltman in 1942 [31]. A compressive or structural effect has been suggested by pathologic studies demonstrating physical distortion of nerve fibers by endoneurial amyloid deposits, such as was described by Dyck and Lambert in 1969 [11]. Although rare patients are seen with massive amyloid deposition [19] and poor neurologic course, there seems to be little correlation with extent of deposition and neuropathy severity. This lends credence to the potential for a distal neural apoptotic effect of amyloid fibrils or their precursor proteins and/or its associated pathologic protein microenvironment [32, 33].

Amyloid infiltration of nerve is associated with primary axonal degeneration. Rare examples of demyelination are known to have occurred, though in at least one case demyelination may have been the result of a paraneoplastic effect of an IgM monoclonal gammopathy in association with antibodies to myelin-associated glycoproteins [34] (Fig. 31.3). An increased rate of axonal degeneration and an increased number of empty fibers are the most common abnormalities seen on teased fiber preparation (Fig. 31.4). Additionally, frequently seen in teased fibers is a flocculent deposition at various portions of the nerve fibers (Fig. 31.3).

Depending on chronicity of disease, either active axonal degeneration or resultant fiber loss may be relatively prominent. Semithin sections stained with methylene blue typically show decreased myelinated fiber density (Fig. 31.1). Amyloid infiltrates connective tissues and aggregates in blood vessel walls, more frequently in the endoneurium (Fig. 31.1), though rarely deposits can be seen in nerve arterioles in the epineurium. The amorphous, acellular deposits of amyloid are relatively eosinophillic and PAS positive. Congo red preparations demonstrate congophilia and apple-green birefringence of deposits, and metachromasia can be seen with methyl violet preparations (Fig. 31.1). The absence of congophilia or metachromasia in amorphous, acellular deposits argues strongly against the presence of amyloid, and can be suggestive of amyloid mimicking conditions such as non-amyloidal monoclonal immunoglobulin deposition disease [35], see below section, (Fig. 31.5), or thickening of the basement membrane of endoneurial blood vessels, as can be seen in chronic diabetes [1]. Due to directionality of beta-pleated sheets, cross sections of spherical or tubular deposits of amyloid can demonstrate alternating areas of apple-green birefringence, in some cases manifesting as a Maltese cross appearance of deposits under polarized light (Figs. 31.1 and 31.2) [36]. The examination of Congo red stained slides by fluorescence microscopy using a TRITC (tetramethylrhodamine isothiocyanate) filter shows marked hyperluminescence in areas of amyloid deposits (see Fig. 17.​3c, Chap. 17). Collections of perivascular inflammatory cells may be seen in the epineurium and endoneurium, though the frequency and etiologic significance of this finding have not been described (Fig. 31.4).

Amyloidomas are focal, macroscopic aggregate of amyloid, and an uncommon manifestation of amyloidosis in general. Cases affecting peripheral nerve are exceedingly rare. In several reported cases, nerve roots are affected by amyloidomas arising from vertebral bodies [18, 22, 37]. Only very few cases of amyloidoma arising primarily from peripheral nerve have been reported, involving the lumbosacral plexus (Fig. 31.2), lumbar nerve root (Fig. 31.3), Gasserian ganglion, sciatic nerve, cervical root, brachial plexus, and infraorbital nerve [16, 17, 19–21, 38, 39].

None of the above described approaches distinguish type of amyloid. One of the most difficult issues arising after the discovery of histologic amyloid in nerves is the correct typing of the specific amyloid proteins. Historically, immunohistochemical staining (anti-TTR, -lambda-light chain, and kappa-light chain) is performed but is challenging. The difficulties in immunocharacterization have been highlighted also in other tissues where misdiagnosis of amyloid type was common and for complex reasons [40, 41]. Contributing to the difficulty in immunocharacterization is the following: (1) antigenic epitopes may be lost by formalin cross-linking in amyloid; (2) circulating contaminant TTR and AL-light chain proteins may produce false-positive results; (3) comparing staining intensities between the different amyloid antibody stains may be unreliable; (4) Pathologic consideration based on clinical, family history and monoclonal serum positivity may be misleading in pathologic interpretation.

We have recently validated application in nerve the use of liquid chromatography tandem mass spectrometry (LC-MS/MS) with laser microdissection (LMD) of amyloid deposits from formalin-fixed, paraffin-embedded sections [4]. This approach eliminates the described problems with immunophenotyping and can be done independent of often misleading clinical information. Our initial validated studies showed that 21 persons with either sural nerve biopsies or proximal root biopsy with amyloid were able to have defined AL, TTR, or GEL cause, whereas earlier attempts at immunophenotyping had failed. Included in the important result was that not only can the predominant amyloid protein be defined, but the specific TTR mutation is determined. This is important as TTR is a normal circulating protein and frequently older persons have circulating incidental circulating monoclonal proteins. The approach has had immediate clinical application in routine practice with large potential for future research with the proteomic microenvironment defined by the approach. Based on our experience we have streamlined interpretation away from immunophenotyping.

Among systemic acquired amyloidoses, only immunoglobulin light chain amyloidosis (AL) has been thought to cause neuropathy. This dogma will require further review with the proliferation of proteomic analysis of amyloid by mass spectrophotometry [4, 42]. Lambda-light chains are more commonly involved than kappa-light chains, roughly by twofold [1, 43]. Monoclonal gammopathy, in amyloidosis as in other conditions, develops in the setting of plasma cell dyscrasia, multiple myeloma, or MGUS-associated lymphoma including Waldenströms [44, 45]. When monoclonal proteins are identified by serum or urine analysis in patients with amyloidosis, AL variety is often given major consideration. However, studies have emphasized that the presence of monoclonal protein, particularly in older patients, may be incidental [41, 46]. Therefore, caution should be used in assuming the causality of circulating monoclonal proteins when amyloid is found in peripheral nerves. Conversely, circulating monoclonal proteins may be small, and their detection dependent on the sensitivity of the laboratory methods utilized. The inclusion of serum light chain ratio and 24-h urine analysis provides increased sensitivity beyond that of serum protein electrophoresis and immunofixation [44, 47].

The variability of peripheral nervous system involvements in AL amyloid is most dramatic compared to familial forms of systemic amyloidosis affecting nerve. Asymmetric and variable combinations of motor, sensory, and autonomic nerves, roots, and plexus can be involved. However, most commonly, AL causes polyneuropathy involving in length-dependent severity (distal worse than proximal) small nerve fibers, often earlier or more apparently than large nerve fibers, and lower limbs earlier than upper limbs, manifesting in neurologic deficits distally more than proximally (i.e., length dependent). However, isolated mononeuropathy, plexopathy, and radiculopathy are possible, i.e., amyloidoma [16, 17, 19–21, 38, 39] (Figs. 31.2 and 31.3). Involvement of autonomic ganglia and peripheral c fibers are common, and dysautonomic features may include gastrointestinal dysmotility with intractable diarrhea, orthostatic hypotension, urinary incontinence, and erectile dysfunction. Simultaneously, multiple organ systems may be involved, including cardiovascular, renal, hepatic, gastrointestinal, and skin [1, 2, 24, 38, 48–50].

Occasionally, isolated peripheral nervous system amyloidosis may precede spread to other organ systems by years. Presentation with somatic (sensory loss and weakness) non-autonomic peripheral neuropathy in amyloidosis portends on the whole better prognosis. Heart failure and gastrointestinal bleeding being the most common causes of death in systemic AL amyloidosis, though early involvement of autonomic nerves (orthostasis, cardioadrenergic) may also contribute specifically to poorer prognosis [45, 51, 52].

It is important to be aware of a very rare pathologic mimicker of amyloidosis having intraneural deposition of most commonly IgM macroglobulin in patients with IgM paraproteinemic neuropathy, occurring in monoclonal gammopathy of undetermined significance and Waldenstrom’s macroglobulinemia. Few cases are reported in the literature [53–55]. Described patients have presented with asymmetric onset sensory, distal large- and small-fiber deficits with painful dysesthesias which evolve to length-dependent involvements. Notably, there may be relative absence of dysautonomia and systemic organ involvement until very late in the disease course, distinguishing the condition clinically from most cases of peripheral nerve amyloidosis. The mechanism of nerve damage as in amyloidosis remains unexplained—compressive, microvascular, inflammatory, and immune-mediated mechanisms have been considered.

Histopathologically, deposits are seen like amyloid to involve endoneurial blood vessels, with infiltration throughout the endoneurium, often tracking subperineurially (Fig. 31.5). Deposits are eosinophillic, PAS positive, and very similar morphologically to amyloid, though lacking the tinctorial qualities of amyloid on methyl violet and Congo red preparations (Fig. 31.5). Epineurial perivascular mononuclear cell collections may be prominent, suggesting an inflammatory or immune-mediated process. Teased fiber preparation often shows increased numbers of empty nerve strands, floccular aggregates adhering to fibers, and segmental demyelination. Mass spectroscopy of laser microdissected deposits demonstrates that amyloid-like material is composed of μ-heavy chain, λ-light chain, and J-joining chain without serum amyloid protein (SAP) and ApoE, suggesting deposition of IgM pentameric molecules in the absence of amyloid-associated proteins [56]. Deposits show a granulofibrillar structure on electron microscopy, which is slightly distinct from the crisscrossing fibrillar structure of amyloid (Fig. 31.5).

FAPs have been subcategorized by the three variant proteins associated with amyloid neuropathy: transthyretin, gelsolin, and apolipoprotein A1 (apo A1). Beta2-microglobulin amyloid deposition, most commonly in the context of chronic hemodialysis, is frequently associated with carpal tunnel syndrome with compressive median neuropathy at the wrist. However, this is known to be due to extraneural deposition within the carpal tunnel and is not as such considered an amyloid neuropathy. More recently, a single family with an apparent, hereditary, systemic amyloidosis, derived from a variant of Beta2-microglobulin, was reported with autonomic and subsequent symmetric sensorimotor axonal polyneuropathy [57].