Figure 72-1 The chest wall of the patient described had normal pinprick sensibility near the spine, but posterior to the posterior axillary line, she had blunted sensation to pinprick (region with diagonal lines), and over the anterior chest wall, she had severely impaired pain and temperature sensibility (cross-hatched region). This pattern reflects length-related sensory neuropathy in the intercostal nerves.

Her examination demonstrated marked loss of pinprick and thermal sensibility in the feet, hands, and anterior to the anterior axillary line over the chest, as illustrated in Figure 72-1, left panel. The loss of pin sensibility was graded, so that the pinprick felt “sharp” behind the posterior axillary lines, was blunter down the sides (see Fig. 72-1, right panel), and felt “dull” over the anterior chest (see Fig. 72-1, left panel). She had loss of joint position and vibratory sensibilities in all limbs, weakness of the intrinsic muscles of the foot, bilateral foot drop, and wasting and weakness in the intrinsic muscles of the hand. Her tendon reflexes were absent at the ankles and elicitable only with reinforcement elsewhere. She had marked autonomic insufficiency including cardiovascular and sudomotor modalities.

Her evaluation led to the diagnosis of severe diabetic polyneuropathy, complicated by spontaneous neuropathic pain and by loss of protective sensibility in the distal distributions of the longest nerves innervating the legs, arms, and chest wall.

COMMENT

Neurologists are trained to search out patterns that have localizing and pathophysiologic specificity. It is axiomatic that the stocking-and-glove pattern of sensory loss suggests the diagnosis of a peripheral neuropathy. Similarly, early weakness of intrinsic muscles of the foot, footdrop, and weakness in the hand muscles is consistent with many peripheral neuropathies. This length-related or length dependent pattern of clinical involvement finds a satisfying pathologic correlate in the pathologic concept of dying back, articulated by John Cavanagh in the 1960s.^1,² Cavanagh’s concept of dying back has three elements: long axons are more susceptible than short ones³; the degeneration begins in the nerve terminals or very distal most regions of axons; and with time, axonal loss progresses from distal to more proximal regions by means of sequential episodes of wallerian-like degeneration, and the process begins in shorter nerves.

This patient points up the fact that this spatial pattern of length-related involvement can even apply to the intercostal nerves. The sternal shield or parasternal stripes represent the distal most regions of the intercostal sensory fibers. With time, the loss of sensation and the neuropathic pain over the chest wall became wider, that is, progressed back toward the spine. She also reminds us that spontaneous neuropathic pain can be severe even in regions with profound loss of nociception.

The spatiotemporal pattern of length-related axonal loss is the most frequent pattern of involvement in peripheral neuropathies, but the underlying pathophysiology is not fully understood. Molecular genetic analyses of heritable neuropathies have identified a daunting number of genetic defects associated with length-related axonopathies. These defects fall into three broad classes: demyelinating disorders of nerve, defects likely to alter axonal transport, and putative abnormalities affecting mitochondrial function.

Mutations in myelin constituents

On the surface, it is surprising that many of the causes of length-related axonal degeneration are mutations affecting the myelin-forming cells. The genetic defects underlying most forms of Charcot-Marie-Tooth type 1(CMT)—the demyelinating forms of CMT—are in proteins of the myelin sheath (PMP22, P0, PLP) or the Schwann cell (connexin 32). These proteins are not in neurons or their axons. The demyelination resulting from these mutations underlies the reduction in conduction velocities characteristic of CMT1, and the pathologic changes of demyelination, remyelination, and the whorls of supernumerary Schwann cells (onion bulbs), are present throughout the peripheral nervous system (PNS). However, even though the myelinated internodes bear the mutations, the classical clinical descriptions of the 19th century identified and underlined the distally predominant muscle wasting. The reductions in muscle bulk in the foot and calf muscles have led generations of medical students to describe the legs of individuals with CMT as stork-legs or “inverted champagne bottle legs.” In CMT1, the length-related clinical pattern* correlates with length-related axonal loss.^4,⁵

By what mechanisms do primary disorders of myelin result in axonal degeneration? Myelination is energetically advantageous for impulse conduction,⁶ and demyelination increases axonal vulnerability in part by increasing energy requirements for impulse conduction. In addition, axonal vulnerability is likely to be increased by disordered glial-axonal interactions. Cell-cell interactions in the PNS are usually discussed in the framework of axonal signaling to Schwann cells that influence migration, entry into or exit from the cell cycle, and the decision to myelinate or remain unmyelinated. One molecule, neuregulin I type 3, has emerged as a dominant influence in each of these areas. However, the Schwann cells also signal to the axons.⁷ This has been best documented in the effects of myelination on the axonal cytoskeleton. Myelinated segments have larger calibers, greater spacing of such cytoskeletal elements as neurofilaments, greater phosphorylation of the higher molecular weight members of the neurofilament “triplet” of proteins NF-M and NF-H,^8,⁹ and changes in the relative phosphorylation of tau and other microtubule-associated proteins. Importantly, these Schwann cell influences maintain normal axonal transport.⁸ In demyelinated internodes, materials that are axonally transported accumulate, and the rate of fast transport is reduced.⁸

One molecule involved in glial-axonal signaling is the myelin-associated glycoprotein (MAG). MAG contributes to axonal protection.¹⁰ Mice lacking MAG undergo slowly evolving distally predominant length-dependent degeneration of axons in the central nervous system (CNS) and the PNS.¹¹ They also have cytoskeletal spacing and phosphorylation patterns characteristic of nonmyelinated axons, even though they have morphologically normal myelin.¹¹ In vitro MAG protects DRG neurons against the neurotoxicities of vincristine, acrylamide, and an “inflammatory soup” from activated CD8 T cells.¹⁰

The means by which MAG signals to the axon is unresolved. MAG is best known for its ability to impair the outgrowth of axons. This property depends, at least in part, on the Nogo receptors NgR1 and NgR2 on axons. The axonal protection by MAG is independent of these receptors.¹⁰ It requires arginine 118 in the MAG molecule,¹⁰ a site that is also necessary for interaction with specific axonal gangliosides (GT1b and GD1b).¹² Axonal protection is reduced by mutagenesis of MAG in this region and by elimination of these axonal gangliosides in mice by knocking out the enzymatic machinery to produce these gangliosides, by sialidase treatment, or by pharmacologic treatment to block ganglioside production.^10,¹² This same site on MAG is also required for interactions with integrins such as ß1-integrin.¹³ Finally, glial molecules other than MAG almost certainly influence axonal transport and axonal survival.

It is attractive to postulate that long axons have, by definition, more sites of cytoskeletal vulnerability to demyelination that can impair axonal transport than do shorter fibers with fewer internodes. Thus, long fibers may cumulate a greater aggregate burden of mild transport deficits and energy deficits that over time leads to “dying back” degeneration. The MAG story makes the important point that influences from the myelin-forming cell can influence the fundamental biology of the axon in ways that determine its survival or degeneration in disease. These influences are likely to contribute to the axonal loss that underlies the symptoms and signs in demyelinating neuropathies.

Only gold members can continue reading. Log In or Register to continue