Autosomal Dominant Inherited Neuropathies With Prominent Sensory Loss and Mutilations: A Review

Hereditary sensory neuropathies (HSNs) are rare disorders characterized by progressive distal sensory loss, predominantly affecting the lower limbs. Foot ulcers, severe skin and bone infections, arthropathy, and amputations are frequent and feared complications. Occasionally, patients complain of spontaneous shooting or lancinating pain. Autonomic fibers can be affected to a variable degree. Patients with HSN can also have severe distal weakness, and some HSN variants have therefore been classified among the hereditary motor and sensory neuropathies (HMSNs). Molecular genetic studies of autosomal dominant inherited neuropathies with prominent sensory loss and ulceromutilating features have assigned the genetic loci for HMSN type 2B (Charcot-Marie-Tooth syndrome type 2B) and HSN type 1 to chromosomes 3q13-22 and 9q22.1-22.3, respectively. However, some families with HSN have been excluded for linkage to these loci, suggesting further genetic heterogeneity. Recently, disease-causing mutations in the SPTLC1 gene have been identified in patients with HSN type 1. In this review, we discuss the hallmark features associated with the distinct genetic subtypes of autosomal dominant inherited HSN and provide genotype-phenotype correlations.

Inherited neuropathies are a clinically and genetically heterogeneous group of disorders of the peripheral nervous system. They are subdivided into 3 main categories based on the predominant involvement of motor and/or sensory nerves. In the most common variants, the hereditary motor and sensory neuropathies (HMSNs), both motor and sensory nerves are clearly affected,¹ and the associated phenotype is classic Charcot-Marie-Tooth syndrome (CMT). Distal hereditary motor neuropathies (distal HMN, spinal CMT, and distal spinal muscular atrophy)² caused by an exclusive degeneration of motor nerve fibers are much rarer. Finally, neuronal degeneration, mainly affecting the peripheral sensory nerves and sometimes involving the autonomic nerves, results in hereditary sensory neuropathy (HSN) or hereditary sensory and autonomic neuropathy (HSAN), respectively.³ However, marked interfamilial and intrafamilial variability of clinical, electrophysiological, and morphological features may hamper a correct classification of individual cases. Recent molecular genetic studies have significantly advanced our understanding of these disorders and have resulted in a more rational classification of inherited peripheral neuropathies.^4-6

In this review, we focus on autosomal dominant inherited peripheral neuropathies that show striking sensory loss and are often complicated by severe infections and amputations. Based on the literature and our experiences, we provide detailed descriptions of the currently known genetic subtypes and corresponding phenotypes. The challenge of a rational classification of these disorders is discussed.

The first description of familial neurotrophic plantar ulcers can be traced to a report by Nelaton in 1852.⁷ Subsequently, additional kinships were reported. Most early descriptive diagnoses referred to the ulcers as the most striking feature of the disease or speculated about the underlying disease mechanism. Thus, historically, the disorders were named hereditary perforating ulcers,⁸ familial trophoneurosis,⁹ familial syringomyelia,¹⁰ hereditary sensory radicular neuropathy,^11,12 or ulcerative and mutilating acropathy,¹³ among others. In the dermatological literature, the term Thevenard syndrome¹⁴ is still used for familial forms, whereas isolated cases of neuropathic ulcerations of unknown origin are classified as Bureau-Barrière syndrome.¹⁵

In 1975, Otha and Dyck¹⁶ proposed a descriptive classification of the inherited peripheral neuropathies and thereby introduced the term HSN. Dyck updated this classification in 1984. The autosomal dominant variant and most common subtype is HSN type 1 (HSN1).³ Based on the presence of additional features, Dyck further subdivided HSN1 in 1993.³ These additional features were then thought to result from the expression of different genes, different alleles of a single gene, modifying genes, or environmental factors. In the presence of autonomic features, the disorder was named HSAN.³ Donaghy et al¹⁷ suggested another classification based on the mode of inheritance, but the term HSN is still more commonly used.¹⁸ The presence of severe motor deficits in some kinships was described long ago^11-13,19 and has been a source of confusion because it blurs the boundaries between HMSN and HSN.²⁰

The disorder usually starts within the second to third decade of life but may occasionally have a later onset. The first signs are often foot deformity and sensory loss. Pes cavus, pes planus, and/or hammertoes may be present since childhood. Badly fitting shoes promote the early occurrence of callus, particularly at the convexity of curled toes or other pressure points.³ Skin changes, such as recurrent paronychia and onychomycosis, are early and frequent findings. Often, patients do not notice sensory loss for a long time. However, some patients may complain of loss of pain and temperature sensation¹² or may have loss of all sensory modalities at the onset of the disease.^13,21-23 In the early stages of the disease, sensory abnormalities are often limited to the toes, but they then spread to the distal parts of the lower limbs, occasionally extending up to the knees. Many patients only become aware of the disease when they notice painless injuries and burns or when they seek medical advice for slowly healing foot ulcers. Once infections occur, complications such as osteomyelitis and osteonecrosis may follow.^3,21 These may sometimes lead to shortening of the feet due to bone loss and can necessitate amputations of the toes or even parts of a limb at more proximal sites. Pain is frequently reported and is usually related to local events but can also occur spontaneously.³ In some kinships, this spontaneous pain has a lancinating or shooting character and often occurs in the feet, legs, thighs, hands, and even shoulders.^8,11,12 Occasionally, spontaneous pain is the most prominent sign of the disease. Progressive hearing loss or even complete deafness may occur.³ In addition, many patients exhibit, to a variable degree, symmetrical distal muscle wasting and weakness, resulting in severe gait disturbances.^{11,13,19,21-24} Weakness may occasionally be an initial sign of the disorder. As the disease progresses, hands can show distal sensory loss and muscle wasting,^6,22,24 and, in severe cases, neuropathic arthropathy can develop.³ Disturbed sweating (hypohidrosis, hyperhidrosis, or anhidrosis) is the most frequent manifestation of autonomic nervous system dysfunction,³ whereas hypotension, erectile dysfunction, and disturbed lacrimation are rare. Progression is usually slow, but, occasionally, patients may have severe proximal limb weakness and require a wheelchair.³ As in most other autosomal dominant disorders, the severity of the disease can be extremely variable, even within one family. Whether burning feet syndrome represents a very mild variant of HSN remains elusive at present.²⁵

The clinical findings of HSN consist of changes in sensation, muscle strength, reflexes, and autonomic function. All sensory modalities can be affected, but sometimes loss of pain and thermal sense is striking, suggesting a predominant small-fiber neuropathy. However, complete loss of touch-pressure sense occurred while joint position sense was affected in some kindreds.³ Interestingly, vibration sense can be preserved for a long time, even when other sensory modalities are already severely abnormal (personal observation, M.A.-G., 1999). Peroneal muscles may be preferentially wasted and weak.^{11,13,19,22-24} Patellar tendon reflexes are often preserved or even brisk but disappear with progression of the disease,^3,21 whereas Achilles tendon reflexes are usually diminished or absent. Pyramidal tract signs are not a common feature, although spasticity of the lower limbs has been reported in a form of HSN with spastic paraplegia.²⁶

Most patients with HSNs have an axonal neuropathy.³ Changes in nerve conduction studies are generally more severe in lower than in upper limbs. Sensory nerves are primarily affected and sural nerve responses may be absent. Compound motor action potential amplitudes are reduced, and motor nerve conduction velocities (NCVs) are slightly slowed. With complete denervation of the distal muscles, compound motor action potentials may not be recordable. In some kinships, severe dispersion of compound motor action potentials and moderate slowing of motor NCVs suggested an additional underlying demyelinative process that has also been demonstrated in studies of nerve biopsy specimens.^21,23 In a large family with CMT type 2B (CMT2B), it has been shown that it is rarely possible to identify mutation carriers at the preclinical stage by NCV studies.²³ Electromyography shows chronic neurogenic alterations, sometimes with spontaneous activity such as fibrillations, sharp waves, and complex repetitive discharges.

Jughenn et al²⁷ were the first to determine that syringomyelia does not occur in patients with HSN1. Unequivocal abnormalities in the peripheral nerves in a postmortem study included the absence of almost all myelinated fibers, an increased number of Schwann cell nuclei, thickened perineuria, and a discontinuity of myelin. Subsequently, Denny-Brown¹¹ described a decrease of dorsal root ganglion cells and the loss of fibers in the dorsal root entry zone and the posterior columns. All these findings suggested that degeneration of the dorsal root ganglia cells was the primary pathological event in HSN. Studies by Dyck et al^3,28 focused on the pathological alterations of the sural nerve and clearly demonstrated that fibers of all size classes are affected. However, small fibers may be preferentially involved compared with large fibers. In addition, fibers appeared to undergo atrophy, myelin wrinkling, demyelination, and remyelination as well as axonal degeneration.

In 1995, Kwon et al⁶ described a single large kinship in which 10 family members were affected by a neuropathy with prominent distal sensory loss. Six patients reported recurrent foot ulcers, and 4 had amputations of the toes because of recurrent infections. In addition, all affected individuals had prominent symmetrical distal motor weakness and wasting.^6,24 Thus, the disorder in this family was not classified as HSN1 but instead was grouped among the axonal forms of HMSN (CMT) and genetically categorized as CMT2B (Online Mendelian Inheritance in Man [OMIM] entry 600882; available at http://www.ncbi.nlm.nih.gov/omim/). This classification has been questioned.²⁰ The gene locus was mapped to a 30 centimorgan (cM) interval on chromosome 3q13-q22.⁶ So far, 2 other families, 1 Scottish and 1 Austrian, have been identified and the genetic interval was reduced to a region of 10 cM.^22,23 In both families, a distinct clinical CMT2B pattern was documented, consisting of prominent motor and sensory disturbances with ulceromutilating complications. In the 3 CMT2B families described, no patients complained of lancinating pain, and sensory loss equally affected all modalities. Studies of NCV showed axonal degeneration, and demyelination was observed with progression of the disease. Recently, Verhoeven et al²⁹ reported that 2 missense mutations in the small GTP-ase late endosomal protein RAB7 cause CMT2B.

In 1996, Nicholson et al⁵ reported 1 large and 3 small Australian kinships with HSN1 (OMIM entry 162400) and mapped the gene locus to an 8-cM region on chromosome 9q22. Linkage of HSN1 to chromosome 9q22.1-22.3 was confirmed in a large American family of German origin, and the HSN1 locus was further refined.³⁰ Recently, 3 different missense mutations (C133Y, C133W, and V144D) in the SPTLC1 gene (OMIM entry 605712) encoding serine palmitoyltransferase long-chain base subunit 1 have been identified in 11 families with HSN1 from Australia, England, and Austria.³¹ Interestingly, haplotype analysis of 3 Australian families of English origin and 3 English families with the same SPTLC1 mutation (C133W) demonstrated that these families had a common founder.³²

Serine palmitoyltransferase is a pyridoxal 5′-phosphate–dependent enzyme that consists of 2 subunits, serine palmitoyltransferase long-chain base subunits 1 and 2. Serine palmitoyltransferase catalyzes the first step of sphingolipid biosynthesis, including ceramide and sphingomyelin. Both subunits are required for enzyme activity.

Recent studies suggest that ceramide functions as a bioactive lipid, modulating several biochemical and cellular responses to stress, including apoptosis, cell-cycle arrest, and cell senescence. Ceramide produced by sphingomyelin catabolism is known to mediate programmed cell death, and increased de novo ceramide synthesis causes apoptosis during early neural differentiation in vitro.³¹

Measuring the rate of de novo synthesis of ceramide and a ceramide metabolite, glycosyl ceramide, in intact cells has demonstrated that HSN1 patients with SPTLC1 mutations have similar serine palmitoyltransferase activity to that in lymphoblast cell lines from healthy controls. Thus, it has been suggested that HSN1 is not caused by haploinsufficiency (decrease of normal function), but neuronal degeneration in HSN1 may be caused by an increase in serine palmitoyltransferase enzyme activity that leads, in neurons, to increased de novo ceramide synthesis and apoptosis.³¹ Alternatively, the mutant enzyme subunit protein may gain a toxic function.

Clinical features common to all chromosome 9q–linked families with HSN1 have been found. The clinical phenotype of the large Australian kinship has been originally described as hereditary sensory radicular neuropathy.¹² In the Austrian family with the C133Y missense mutation, the 6 affected individuals (4 living and 2 deceased) had prominent sensory loss with a predilection for pain and thermal sense, whereas vibration sense was preserved for a long time. Sensory disturbance was the initial sign in 5 family members who became aware of the disease as they developed painless injuries or burns. Five older patients reported attacks of shooting and lancinating pain that was difficult to control. Four family members had recurrent foot ulcerations, all of them resulting in osteonecrosis and osteolysis and subsequently necessitating amputations of bones, toes, or even legs. Pes cavus was observed in 1 patient. Three individuals reported autonomic features, consisting of hyperhidrosis. Two older patients developed hearing loss. In addition, most of our patients with HSN1 had prominent distal muscle weakness and wasting in both lower and upper limbs. This was the presenting sign in the youngest patient, who received the diagnosis of HMSN2. Studies of NCV suggested an axonal neuropathy, but compound motor action potentials and sensory nerve action potentials were often not recordable.

We have personally examined several families with CMT2B and HSN1. We found that the presence of shooting and lancinating pain is an important clinical hallmark of HSN1. Another characteristic feature of HSN1 is severe loss of pain and temperature sensation, leading to painless injuries. The degree of muscle involvement clearly fails to distinguish these 2 subtypes because it can be severe in both forms.

In 2 families from Austria²¹ and Italy,³³ linkage to the CMT2B and HSN1 loci was excluded. In the Austrian family, 9 individuals were definitively affected. Both motor and sensory signs were less pronounced, whereas foot deformity, skin changes, and ulceromutilating features were a prominent and frequent finding. All sensory modalities were equally affected. None of the patients complained of spontaneous pain, in contrast to the Italian family. The results of NCV studies were normal or slightly to moderately slowed.

Other types of inherited neuropathies, in particular HMSN1 (CMT1), may have some features in common with autosomal dominant HSN. We found 5 families with a severe demyelinating neuropathy and prominent sensory loss in which several affected family members had pronounced foot deformity, causing multiple pressure points and leading to the development of foot ulcers. Molecular genetic studies in these families revealed that the CMT1A duplication on chromosome 17p11.2 was the underlying genetic defect. One of these families had been diagnosed with HSN1 in 1977 based on the presence of severe sensory loss and recurrent ulceromutilating complications in affected individuals.³⁴ Similar observations have been reported in 2 families with CMT1A from England.³⁵

At present, no specific treatment is available for inherited neuropathies with prominent sensory loss and ulceromutilating features. The main goal is the prevention of foot complications. Patients must be carefully instructed about the risk of developing foot ulcers and infections. Unnoticed sensory loss, which can lead to painless injuries and burns, must be detected by careful neurological assessment. Patients should wear comfortable shoes, and pressure points caused by foot deformity can be avoided with orthopedic shoes. Furthermore, the shoes should be carefully inspected for foreign bodies before being worn. Long walks should be avoided, and daily inspection of the feet is necessary to recognize fissures or other skin changes. Good regular care of the feet is indispensable. If an ulcer or infection occurs nevertheless, it must be treated adequately and immediately. Weight bearing should then be discontinued. The ulcer should be cleaned, and treatment with antibiotics might become necessary. Radiography should be performed early to exclude an underlying osteomyelitis.

Genetic counseling is required before testing preclinical cases. At-risk individuals should be informed about their 50% risk of having inherited the disease gene from an affected parent. Molecular genetic testing is needed to confirm the diagnosis. Known CMT1 mutations should be excluded first in patients with demyelinating neuropathy. Axonal neuropathies can be screened for mutations within the SPTLC1 gene. Also, patients with the phenotype of CMT2B can now be tested for mutations in the RAB7 gene.

Genetic heterogeneity has been shown in autosomal dominant inherited peripheral neuropathies with predominant sensory loss and ulceromutilating features. However, no satisfying nomenclature of these disorders is available at present. Both HMSN1 (CMT1) and HMSN2 (CMT2) can be complicated by ulceromutilations, whereas patients with HSN1, although usually characterized by severe sensory loss, ulcers, and amputations, may also present with prominent motor involvement and no ulcers.^11,13,19 The situation is also complicated by intrafamilial phenotypic variability. Therefore, classification of these disorders can no longer be based on the type of affected nerve fibers. Classification of the first CMT2B family,⁶ described as an axonal type of HMSN (CMT), was criticized.²⁰ However, recently it has been suggested to classify HSN1 as a form of HMSN2 (CMT2)^4,36 because motor deficits are often present. In this novel classification, CMT2B accounts for CMT2 with prominent sensory deficits.⁴ A classification according to the most prominent clinical feature, ie, the ulceromutilations, as has been used in the past, still makes sense. It highlights the most prominent clinical characteristics of the disease, but it fails in those patients within a family who do not have ulceromutilating complications.

Molecular genetic linkage studies in more families are needed to confirm or identify new genes and to document phenotypic variation. Such studies require a very careful ascertainment of at-risk family members.³⁷Table 1 provides key signs and symptoms of the different types of these diseases that should be assessed. Table 2 gives an overview of the most important differential diagnoses in sporadic cases. The identification of additional genes involved in these disorders and of the underlying abnormalities will lead to a more satisfying nomenclature.

Corresponding author: Michaela Auer-Grumbach, MD, Institute of Medical Biology and Human Genetics, Karl-Franzens University Graz, Harrachgasse 21/8, A-8010 Graz, Austria (e-mail: michaela.auergrumbach@kfunigraz.ac.at).

Accepted for publication February 27, 2002.

Author contributions: Study concept and design (Drs Auer-Grumbach, De Jonghe, Timmerman, and Wagner); acquisition of data (Dr Auer-Grumbach); analysis and interpretation of data (Drs Auer-Grumbach, Verhoeven, Hartung, and Nicholson); drafting of the manuscript (Drs Auer-Grumbach, De Jonghe, Timmerman, and Wagner); critical revision of the manuscript for important intellectual content (Drs Auer-Grumbach, Verhoeven, Hartung, and Nicholson).

This study was supported by grants FWF P13563-BIO and FWF P15378 from the Fonds zur Förderung der wissenschaftlichen Forschung, Vienna, Austria. Drs Verhoeven and Timmerman are postdoctoral fellows of the Fund for Scientific Research, Flanders, Belgium.