INVITED REVIEW ABSTRACT: Neuropathy is a common complication of end-stage kidney disease (ESKD), typically presenting as a distal symmetrical process with greater lower-limb than upper-limb involvement. The condition is of insidious onset, progressing over months. and has been estimated to be present in 60%–100% of patients on dialysis. Neuropathy generally only develops at glomerular filtration rates of less than 12 ml/min. The most frequent clinical features reflect large-fiber involvement, with paresthesias, reduction in deep tendon reflexes, impaired vibration sense, muscle wasting, and weakness. Nerve conduction studies demonstrate findings consistent with a general- ized neuropathy of the axonal type. Patients may also develop autonomic features, with postural hypotension, impaired sweating, diarrhea, constipa- tion, or impotence. The development of uremic neuropathy has been related previously to the retention of neurotoxic molecules in the middle molecular range, although this hypothesis lacked formal proof. Studies utilizing novel axonal excitability techniques have recently shed further light on the patho- physiology of this condition. Nerves of uremic patients have been shown to exist in a chronically depolarized state prior to dialysis, with subsequent improvement and normalization of resting membrane potential after dialysis. The degree of depolarization correlates with serum Kϩ, suggesting that chronic hyperkalemic depolarization plays an important role in the develop- ment of nerve dysfunction in ESKD. These recent findings suggest that maintenance of serum Kϩ within normal limits between periods of dialysis, rather than simple avoidance of hyperkalemia, is likely to reduce the inci- dence and severity of uremic neuropathy. Muscle Nerve 35: 273–290, 2007

UREMIC NEUROPATHY: CLINICAL FEATURES AND NEW PATHOPHYSIOLOGICAL INSIGHTS

ARUN V. KRISHNAN, MB, BS, PhD, and MATTHEW C. KIERNAN, MB, BS, PhD

Prince of Wales Medical Research Institute and Prince of Wales Clinical School, University of New South Wales, Barker Street, Randwick, Sydney, NSW 2031, Australia

Accepted 24 October 2006

End-stage kidney disease (ESKD) occurs when required in order to remove waste products and nephrons are irretrievably impaired to the extent excess fluid. that the retention of metabolic waste products, salt, Uremic neuropathy has for many decades been and water becomes potentially fatal.128 ESKD may recognized as a common complication of ESKD.7,72,145 occur due to either a primary renal disorder or as a Although the advent of dialysis and transplant pro- complication of a multisystem disorder. The com- grams has led to reductions in the rate of severe neu- mon causes of ESKD remain diabetes, glomerulone- ropathy, the prevalence of this condition remains phritis, and hypertension.136 When renal function high.113 This review covers the clinical and electrophys- reaches critically low levels, renal replacement ther- iological features of uremic neuropathy, with a focus apy either in the form of dialysis or transplantation is on advances in understanding the pathophysiology of this condition.

Available for Category 1 CME credit through the AANEM at www. aanem.org. HISTORICAL ASPECTS Abbreviations: ADH, activity-dependent hyperpolarization; CSF, cerebro- spinal fluid; CMAP, compound muscle action potential; CTS, carpal tunnel The possibility of peripheral neuropathy in patients syndrome; EPO, erythropoietin; ESKD, end-stage kidney disease; NSS, neu- treated with hemodialysis was first raised shortly after ropathy symptom score; PTH, parathyroid hormone Key words: end-stage kidney disease; middle molecules; peripheral neurop- the introduction of the first formal hemodialysis pro- athy; potassium; uremic neuropathy gram.172 The first clinical documentation of neuropa- Correspondence to: M.C. Kiernan; e-mail: [email protected] thy was provided in 1961 in two young male patients © 2006 Wiley Periodicals, Inc. 209 Published online 28 December 2006 in Wiley InterScience (www.interscience. with hereditary interstitial nephritis and deafness. wiley.com). DOI 10.1002/mus.20713 The development of neuropathy in these cases, how-

Uremic Neuropathy MUSCLE & NERVE March 2007 273 ever, was attributed to the underlying hereditary disor- der, rather than viewed as a complication of ESKD. Following this report, Asbury et al.6 provided extensive clinical and pathological findings in four men who developed neuropathy as a consequence of ESKD of varying etiologies. All four patients had clinical features of renal disease for many years be- fore the development of neuropathy, which mani- fested as a symmetrical length-dependent sensorimo- tor neuropathy. Nerve biopsies established axonal degeneration, maximal distally, with sparing of prox- imal nerve segments and nerve roots. Moreover, there was no evidence to suggest nerve compression, inflammation, or the superimposition of a systemic disease process, such as diabetes or amyloid, leading to the conclusion that the development of neurop- athy was a consequence of the underlying renal dis- order. Early clinical neurophysiological investigations in ESKD patients demonstrated reductions in motor nerve conduction velocity in symptomatic and asymptomatic patients.158,185 Jebsen et al.,82 studying the natural history of uremic neuropathy, compared clinical and nerve conduction findings in patients treated conservatively to those receiving dialysis ther- apy. Whereas the development of neuropathy in the conservatively treated group was related to deterio- rating renal function, those patients treated with long-term dialysis manifested improvement in both clinical and neurophysiological parameters. Follow- ing these early reports and in light of the increasing FIGURE 1. Prominent wasting of intrinsic hand muscles in a use of dialysis and renal transplantation therapies, patient with established uremic neuropathy. In addition to the prominent wasting and resultant clinical weakness, the patient greater attention has been focused on uremic neu- complained of numbness and had impaired proprioception. ropathy, with numerous studies reporting high rates of neuropathy in ESKD patients, generally relating the development of neuropathy to the severity of renal failure.5,20,82,107,144,146,149,158,185,187 Of particular lower-limb than upper-limb involvement. The condi- note, studies by Nielsen144,149–152 and Bolton et tion is of insidious onset, progressing over months, al.20,21 in the 1970s demonstrated nerve conduction and has been noted to have a male predominance. It slowing in clinically unaffected nerve segments, with generally only develops at glomerular filtration rates correlation between the extent of renal impairment of less than 12 ml/min.39 The most frequent clinical and degree of conduction slowing, as well as im- features are those of large-fiber involvement, with provement in neurophysiological parameters follow- paresthesias, reduction in deep tendon reflexes, ing renal transplantation. These studies provided impaired vibration sense, weakness, and muscle wast- clinical evidence to suggest that a uremic toxin was ing (Fig. 1). In the 1970s, Nielsen144,145 demon- responsible for the development of neuropathy in strated the presence of neuropathic symptoms in ESKD patients, a hypothesis that was to become a over 50% of patients with ESKD. Other studies have major focus of future neurophysiological research in demonstrated prevalence rates varying from 60% to this condition. 100%, depending on the diagnostic criteria ap- plied.1,23,122,191 Laaksonen et al.122 staged the clinical severity of INCIDENCE AND CLINICAL FEATURES uremic neuropathy in 21 ESKD patients, using a Peripheral neuropathy in ESKD generally presents modified version of the neuropathy symptom score as a distal symmetrical polyneuropathy with greater (NSS) developed by Dyck et al.,57,58 and combined

274 Uremic Neuropathy MUSCLE & NERVE March 2007 this assessment with results of nerve conduction In a recent study, amplitude of the sural sensory studies. The NSS quantified symptoms that were nerve action potential was found to be the most sensi- grouped into three categories to reflect alteration in tive indicator of uremic neuropathy, being reduced in motor, sensory, and autonomic systems. Within each 50% of ESKD patients.113 Other groups have con- group, further subsets were used to group symptoms firmed similar findings, demonstrating reductions in according to the region affected and the presence of sensory and motor response amplitudes in addition to positive or negative symptoms. Using the NSS and abnormalities of late responses.1,4,49,122,130,137,157,191 Re- the staging procedure previously used in studies of duction in peroneal nerve motor conduction veloc- 46,137,148 diabetic patients,57,58 81% of ESKD patients received ity and prolongation of tibial F-wave minimum 122 a diagnosis of neuropathy. Stage 1 neuropathy latencies have been established as sensitive indica- (asymptomatic neuropathy) was diagnosed in 19%, tors of neuropathy in ESKD patients. Prolongation of stage 2 neuropathy (symptoms nondisabling) was soleus H reflexes has also been demonstrated in pa- present in 48%, and stage 3 neuropathy (disabling tients without clinical evidence of neuropathy, suggest- ing that this parameter may be more sensitive in de- symptoms) was noted in 14%. In a more recent tecting early neuropathy.68,191 study,113 93% of ESKD patients had neuropathic Studies of quantitative sensory testing in ESKD symptoms on NSS testing, with 72% diagnosed with patients have demonstrated increased vibratory per- stage 2 neuropathy and 21% with stage 3 neuropa- ception thresholds, most marked in the lower thy, despite all patients meeting currently accepted limbs,147 whereas somatosensory-evoked potentials 143 guidelines of dialysis adequacy. in ESKD patients demonstrate abnormalities of con- duction along both the distal and proximal segments CLINICAL AND NEUROPHYSIOLOGICAL of peripheral somesthetic pathways, but less com- FINDINGS IN GENERALIZED monly along intracranial sensory pathways.25,153,166 UREMIC NEUROPATHY A study of single-fiber electromyography demon- Early studies of uremic neuropathy utilizing nerve strated normal fiber densities in motor units of ESKD biopsy techniques revealed prominent axonal de- patients.186 This finding suggested that reinnervation, generation, most severe in the distal parts of nerve characterized by increased fiber density, had failed to trunks. Although initial studies suggested that demy- occur. However, this was accompanied by increased elination was a significant feature of uremic neurop- jitter, possibly reflecting peripheral demyelination in athy,5,54 subsequent reviews demonstrated that de- the setting of axonal degeneration. A further single- myelination was secondary to axonal loss and that fiber EMG study established that jitter abnormalities 106 proximal segments of the nerves were relatively improved following a year of dialysis. spared.3,55,67,187 These findings supported the con- Early studies of nerve excitability, utilizing a lim- cept that uremic neuropathy was a dying-back neu- ited range of excitability parameters, demonstrated ropathy, with metabolic failure of the neuron caus- an elevated threshold for excitation even when nerve conduction values were normal, in addition to dem- ing distal axonal degeneration.55 onstrating prolongation of absolute and relative re- Numerous neurophysiological series have been fractory periods.25,125,182,208 As a consequence, it was undertaken in patients with uremic neurop- concluded that the safety factor for neural transmis- athy and have demonstrated findings consistent sion at the nodes of Ranvier would be lowered. Un- with a generalized neuropathy of the axonal expectedly, uremic nerves retained vibratory percep- 1,4,13,20,41,46,49,68,80,107,113,121,130,148,155,156,179,183,185,191 type. tion and their sensory response amplitudes for a Early studies focused on motor nerve conduction pa- longer period than control nerves when rendered rameters and demonstrated slowing of conduction ve- ischemic.43 Uremic nerves also behaved differently locity in patients prior to the development of clinical when temperature was lowered, with a less rapid rise 45,158 neuropathy. Subsequent studies demonstrated ab- in response amplitude compared to controls.24,25 normalities of nerve conduction148 with generalized In addition to the slowly progressive sensorimo- slowing in both sensory and motor nerves, accompa- tor axonal neuropathy, a more rapidly progressive nied by reduction in sensory response amplitudes. Mo- motor neuropathy has been described. A small num- tor response amplitudes tend to remain relatively pre- ber of ESKD patients with diabetes have also been served, although abnormalities in lower-limb motor shown to develop a subacute neuropathy progress- nerves were noted in some patients, accompanied by ing over a few months, with severe muscle weakness. neurogenic changes in distal lower-limb muscles on In this group of patients, nerve conduction studies electromyography.148,149 may demonstrate features of either a demyelinating

Uremic Neuropathy MUSCLE & NERVE March 2007 275 or axonal neuropathy.26,27,165 Although the presence of diabetes complicates assessment of nerve conduc- tion data, the absence of preexisting neuropathic symptoms and the clinical improvement noted fol- lowing dialysis or renal transplantation suggest a metabolic basis for the neuropathy, related to the underlying ESKD. Analysis of cerebrospinal fluid (CSF) is rarely helpful, as CSF protein concentration is frequently elevated in ESKD patients and may simulate the albuminocytologic dissociation that is characteristic of Guillain–Barre´ syndrome.23 Small-fiber neuropathy may develop as a clinical entity in ESKD patients. Lindblom and Tegner124 demonstrated abnormalities of thermal sensation in 30% of ESKD patients and concluded that small- fiber neuropathy may exist as a distinct entity in these patients. These results, however, differed from those of other groups who demonstrated minimal impairment of thermal sensation in ESKD.56,188 In a study of 20 ESKD patients, abnormalities in standard nerve conduction studies were demonstrated in 16 patients,4 whereas abnormal thermal thresholds were found in only 6 patients and, when present, did not correlate with clinical evidence of polyneurop- athy.4 Such findings are consistent with those of pathological studies that demonstrated greater vul- FIGURE 2. Forearm arteriovenous fistula (basilic vein transposi- nerability of larger-diameter fibers in ESKD pa- tion) in a patient referred for neurophysiological assessment of tients.55 right hand pain and numbness. Subsequent nerve conduction studies established a generalized axonal neuropathy, with su- MONONEUROPATHIES IN ESKD peradded features consistent with carpal tunnel syndrome. Mononeuropathies are a frequent clinical complica- tion in ESKD patients and most typically occur in the blood from the distal limb.66,70,105,119,131,141,201 The median, ulnar, and femoral nerves.39 placement of Brescia–Cimino arteriovenous fistulas Carpal tunnel syndrome (CTS) is the most com- between the radial artery and cephalic vein has been mon mononeuropathy in ESKD, with prevalence associated with the development of both the clinical 15,17,52,66,67,76,81,170 ␤ rates varying from 6% to 31%. 2- features of CTS and subclinical neurophysiological Microglobulin amyloidosis is a major factor underly- abnormalities in median or ulnar nerve territo- ing the development of CTS in ESKD patients,63 a ries.70,105 A recent study demonstrated CTS in 30.5% complication noted in patients on long-term hemo- of limbs with fistulas compared to 12.2% on the dialysis. Amyloid deposits have been identified in contralateral side.66 Although the site of the fistula synovial specimens from dialysis patients with CTS63 had no effect on the development of CTS, a signifi- and an increase in the rate of CTS has been demon- cant correlation was noted between the age of the strated with increasing hemodialysis duration.63 fistula and the presence of CTS. ␤ Strategies geared at reducing the levels of 2-micro- With regard to treatment of CTS in ESKD, the globulin, such as the use of high-flux biocompatible outcome of median nerve decompression appears ␤ membranes and 2-microglobulin adsorption col- inferior in this group compared to patients with umns, have resulted in reduced rates of CTS devel- idiopathic CTS.76,102,173 In addition, recurrence rates opment and ultimately improvement in symp- are higher. In ESKD patients with recurrent CTS, an toms.63,118,192 endoscopic approach may prove to be effective in Other factors that may contribute to the in- relieving persistent symptoms.211 creased incidence of CTS in ESKD patients include Ulnar neuropathy is also a common occurrence uremic tumoral calcinosis47,203 and the placement of in ESKD, affecting up to 51% of patients undergoing arteriovenous fistulas (Fig. 2), inducing a “steal” of hemodialysis.142 Causes include external compres-

276 Uremic Neuropathy MUSCLE & NERVE March 2007 sion at the elbow during prolonged dialysis thera- polyneuropathy,89,200 but others have shown no cor- py142 and other risk factors as outlined for CTS, relation between autonomic dysfunction and periph- particularly arteriovenous fistulas and uremic tu- eral nervous system abnormalities.44,180,195 The moral calcinosis.62,69 mechanisms underlying the development of uremic In addition to their possible contributions to the autonomic neuropathy remain unknown, although development of CTS and ulnar neuropathy, upper- an association with hyperparathyroidism has been limb arteriovenous fistulas may be associated with an suggested.178,194 acute-onset neuropathy, first described by Bolton et Studies utilizing objective measures of auto- al.22 and later termed ischemic monomelic neurop- nomic function, including R–R interval variation athy.206 In this condition, acute limb ischemia devel- as a measure of parasympathetic function and sus- ops due to shunting of arterial blood away from the tained handgrip and sympathetic skin response as distal parts of the limb.135 The severity of ischemia measures of sympathetic function, have estab- typically affects nerves, without causing changes in lished abnormalities in up to 62% of ESKD pa- other tissues such as muscle, that possess a higher tients on dialysis treatment.44,121,178,200 However, threshold for ischemic injury. Diabetic patients are these abnormalities frequently occur in the ab- particularly vulnerable, especially those with preex- sence of clinical symptoms of autonomic dysfunc- 162 isting peripheral vascular disease or neuropathy. tion.59,164 Parasympathetic dysfunction has been The symptoms are those of multiple upper-limb shown to occur with greater frequency than sym- mononeuropathies, with distal sensory loss and pathetic dysfunction, which is generally more com- weakness in the muscles of the forearm and hand. mon in diabetic ESKD patients.73,164,168,195,196,200,212 Electromyography demonstrates neurogenic abnor- The contribution of autonomic dysfunction to malities in distal limb muscles with sparing of prox- the development of intradialytic hypotension re- imal musculature.206 There may be predilection for mains a matter of ongoing debate, with some studies median nerve involvement and the presence of con- suggesting a possible association89,168 and others sug- duction block in this nerve has been described as a gesting no significant relationship.123,195 A recent sign of reversible injury.87 Early ligation or revision review of the literature on the use of the oral alpha- of the fistula frequently leads to significant clinical and electrophysiological improvement.87,161 1-adrenoceptor agonist midodrine in the treatment Femoral neuropathy in ESKD patients is a well- of intradialytic hypotension suggested a beneficial recognized complication of renal transplantation, effect, although the authors drew attention to the with an incidence of ϳ2%.83,134,175 Prolonged use of fact that most studies were not randomized and had 137 self-retaining retractors in the region of the femoral small sample sizes. nerve during renal transplantation may cause nerve compression and neuropraxic injury.175 Rapid recov- EFFECTS OF DIALYSIS AND TRANSPLANTATION ery is expected in cases of neurapraxia but pro- ON UREMIC NEUROPATHY longed ischemia may lead to axonal loss.210 The Early reports investigating the effects of hemodialysis prognosis is generally favorable, with most patients on uremic neuropathy suggested that some patients 175 achieving complete recovery, although residual with mild neuropathy recovered completely with ad- deficits may be present when significant axonal loss equate dialysis.72,82,107,185 In fact, failure to improve 134 has occurred. was considered to be an indictor of insufficient dial- ysis. These reports, however, did emphasize that the AUTONOMIC NEUROPATHY IN ESKD extent of improvement was likely to be related to the Autonomic neuropathy may develop in ESKD pa- severity of neuropathy and that patients with severe tients, manifesting as postural hypotension, im- neuropathy were unlikely to experience any signifi- paired sweating, diarrhea, constipation, or impo- cant recovery.82,107,185 tence.14,197 In a study of 36 ESKD patients, More recent studies, however, have demon- gastrointestinal autonomic symptoms were evident strated that improvement in neuropathy with dialysis in 42% and impotence in 45%.200 Although postural is an uncommon event.13,41,151,155 Although these hypotension was an uncommon clinical finding, studies suggest that dialysis retards the progression 36% complained of episodes of postural dizziness, of neuropathy in most patients, in some cases a which was most prominent in elderly ESKD patients. gradual deterioration of neuropathy may oc- Some studies have suggested that autonomic neu- cur.13,41,151,155 A comparison of hemodialysis and ropathy occurs as a manifestation of generalized peritoneal dialysis with regard to neuropathy pro-

Uremic Neuropathy MUSCLE & NERVE March 2007 277 gression has demonstrated no significant difference for neuropathy was better dialyzed by the perito- between the two dialysis forms.184 neum than by the cellophane membranes used in Renal transplantation remains the only known hemodialysis. On this basis, the most likely group of cure for uremic neuropathy,20,21 with clinical im- substances was thought to be the “middle mole- provement in sensory and, to a lesser extent, motor cules,” substances with a molecular weight of 300– function occurring within a few days of transplanta- 12,000 Da,193 given that such substances were known tion.80 Serial nerve conduction studies following to be poorly cleared by hemodialysis membranes. transplantation demonstrated a correlation between Marked elevations in the concentrations of mid- the improvement in nerve conduction and biochem- dle molecules have been demonstrated in ESKD pa- ical parameters, suggesting that metabolic phenom- tients, a finding not observed in healthy controls.60,61 ena may underlie the rapid improvement.156 Even Examples of such molecules include parathyroid with severe neuropathy, improvement in symptoms hormone (PTH) and ␤-2-microglobulin (␤-2M), the and signs may occur within 1 month of transplanta- levels of which are elevated in patients with ESKD.193 tion, although in some patients the recovery is pro- Further studies demonstrated that the use of thinner longed or remains incomplete.21,152,183 dialysis membranes and longer dialysis times, strate- Dialysis and transplantation are less beneficial for gies that would have greater benefits for the clear- patients with autonomic neuropathy compared to ance of middle molecules compared to small mole- large-fiber neuropathy. An early study suggested that cules, led to significant reductions in the rates of autonomic function may be improved with dialysis,74 severe neuropathy.9,64 A study using a hemodialysis but a more recent report failed to show any signifi- membrane highly permeable to middle molecules cant benefit.198 Although renal transplantation may also demonstrated a dramatic reduction in the de- lead to improvement or normalization of autonomic velopment of neuropathy.129 function,127,164 the time course of such improvement These early studies, however, were hampered by is often slow and may be incomplete, with significant a number of difficulties, not least of which was the changes often occurring after 4–8 years.177,199 inability to measure middle molecule levels.10 An- Recent evidence suggests that treatment with other major shortcoming of the hypothesis has been erythropoietin (EPO) may prove beneficial in ESKD the lack of conclusive evidence that any single mol- patients with neuropathy71,174 as well as for patients ecule in the middle molecular range is actually neu- 104,193 with neuropathy due to other etiologies.91,189 Treat- rotoxic. In a study of nerve conduction velocity ment with EPO improved motor nerve conduction following renal transplantation, correlation was velocity in ESKD patients, but had no effect on sen- noted between the postoperative concentration of sory indices. In vitro studies have shown that EPO myoinositol, a middle molecule, and median nerve 156 receptors are present on Schwann cells and in dorsal sensory conduction velocity. Although myoinosi- root ganglion neurons.79,90 Upregulation of EPO re- tol levels are elevated in ESKD patients, there is little 193,207 ceptors occurs after axonal injury, mediated by re- convincing evidence for a neurotoxic effect. lease of nitric oxide, and administration of exoge- The only middle molecule for which some evi- nous EPO is associated with reduction in limb dence of neurotoxicity exists is PTH, with some studies weakness and neuropathic pain behavior.90 suggesting a link between PTH and the neurological complications of ESKD.132,176 PTH has been shown to prolong motor nerve conduction velocities in animal DIALYZABLE TOXINS AND THE 65 MIDDLE MOLECULE HYPOTHESIS studies, although human studies of the effect of PTH on peripheral nerves have yielded conflicting results, Hegstrom et al.72 postulated that uremic neuropathy with variable changes in motor nerve conduction ve- occurred due to accumulation of a dialyzable sub- locity in patients with ESKD.8,53,169 stance on the basis of their observational studies that Despite the shortcomings of the middle molecule demonstrated improvement in neuropathy in two hypothesis, the hypothesis that a dialyzable toxin subjects with long-standing ESKD following com- may be involved in the pathophysiology of this con- mencement of dialysis therapy. Later studies demon- dition remains prevalent. More recently, it has been strated that patients treated with peritoneal dialysis suggested that the following criteria should be met had lower rates of uremic neuropathy despite the in order for a substance to be truly regarded as a fact that these patients frequently had higher blood uremic neurotoxin: (1) it must be an identifiable urea and creatinine concentrations.10 The lower chemical; (2) it should be elevated in the blood of neuropathy rate in the peritoneal dialysis group was uremic patients; (3) there should be a direct positive thought to indicate that the substance responsible relationship between blood level and neurological

278 Uremic Neuropathy MUSCLE & NERVE March 2007 FIGURE 3. (A) Saltatory conduction, with the action potential jumping from one node of Ranvier to the next. (B) Different channels are distributed unevenly along the axonal membrane: Naϩ channels are found in high concentrations at the node, as are slow Kϩ channels. ϩ ϩ ϩ Fast potassium channels (K f) are almost exclusively paranodal. Inward rectifier channels (Ih), permeable to both K and Na ions act to limit axonal hyperpolarization, whereas the Naϩ/Kϩ pump serves to reverse ionic fluxes that may be generated through activity.

dysfunction; (4) it should cause neurological dys- information regarding membrane potential and ax- function in animals at appropriate blood levels; and onal ion function may be gained from nerve excit- (5) its removal from the blood should abolish the ability studies.37,40 Axonal excitability can be investi- dysfunction.38 The middle molecule hypothesis fails gated using threshold tracking, where “threshold” to satisfy a number of these criteria, most impor- indicates the stimulus current required to produce a tantly criterion 3, as there is very little evidence to target potential that can be adjusted online by com- suggest that such molecules are actually neurotoxic. puter (i.e., tracked) to assess excitability. The recent Despite the evidence that a dialyzable toxin may development of automated protocols has facilitated underlie the development of uremic neuropathy, the use of excitability techniques in the clinical set- the mechanism of this neurotoxicity remained un- ting.93,95,108 Rather than relying on a single parame- clear. The possibility that the neurotoxic effect may ter, excitability techniques provide information re- be due to alteration in membrane excitability was garding alterations in membrane potential and first proposed by Nielsen who, drawing on evidence axonal ion channel function based on coherent from in vitro studies of muscle and red blood cells in changes in a number of different indices. Nerve ESKD patients,18,205 proposed that one or more of excitability measures have been used to study periph- these toxins may cause neuropathy by inhibiting ac- eral nerves in patients with neuropathy and have ϩ ϩ tivity of the axonal Na /K pump. This energy- provided information about disease pathophysio- ϩ dependent pump is electrogenic, with three Na logy.42,85,88,96,110,111,120 Prior to discussing the ϩ ions being pumped out for every two K ions changes in axonal excitability that develop in pa- pumped into the axon,159 leading to a net deficit of tients with ESKD, a general understanding of nerve positive charge on the inner aspect of the axonal physiology is critical for the further discussion re- ϩ ϩ membrane. Paralysis of the Na /K pump abolishes lated to the pathophysiological mechanisms involved the direct contribution of the hyperpolarizing pump in the development of neuropathy. current to the membrane potential and leads to an ϩ accumulation of extracellular K that causes further MOLECULAR STRUCTURE OF THE AXON ϩ ϩ depolarization.84,163 The Na /K pump is therefore AND SALTATORY CONDUCTION of critical importance in maintaining normal ionic Transmission of impulses in myelinated axons oc- gradients, which are essential for axonal survival. curs by means of saltatory conduction (Fig. 3), with Disruption of these gradients may cause reverse op- ϩ ϩ action potentials advancing between successive eration of the Na /Ca2 exchanger, leading to in- ϩ nodes of Ranvier: creased levels of intracellular Ca2 and axonal loss.48 Although it is not possible to measure membrane “Like a kangaroo travelling at speed, the action potential directly in human axons in vivo, indirect potential advances at near-uniform velocity, but it

Uremic Neuropathy MUSCLE & NERVE March 2007 279 is powered by discrete kicks of inward membrane old current for a target potential declines as the current at nodes of Ranvier.”33 stimulus duration is increased.28,138,139,204 Calcula- ␶ tion of SD in human subjects may be performed The chief role of the axon is that of impulse conduc- using the ratio between the stimulus–response tion, which depends on the electrical cable structure curves for two different stimulus durations using and voltage-dependent ion channels of the axonal Weiss’ formula,204 shown below for threshold cur- membrane. Much of the knowledge about axonal rents for stimuli of 0.2 ms (I 0.2) and 1.0 ms (I 1.0) membrane structure and ion channel function comes duration: from studies in nonmammalian axons (squids, verte- brates). Only over the last few decades have techniques ␶ ϭ 0.2͑I –I ͒/͓I –͑0.2 ϫ I ͔͒ such as patch-clamping allowed investigation of mam- SD 0.2 1.0 1.0 0.2 malian axons, and only over more recent years have Both strength–duration time constant and rheobase human axons been studied in vitro. (the threshold for a stimulus of infinitely long dura- In myelinated axons from peripheral nerve, ϩ tion) are properties of the nodal membrane, depen- voltage-sensitive Na channels are clustered at high dent on passive membrane properties and a local re- densities (up to 1,000/␮m2) in the nodal axon, com- ϩ sponse mediated by persistent Na channels.36 pared to the internodal region (25/␮m2).202 The ϩ Alterations in membrane potential will affect both pa- high density of Na channels at the node reflects the rameters, with depolarization leading to prolongation need of saltatory conduction for a large inward cur- of strength–duration time constant and reduced rheo- rent at the node (Fig. 3). When the nodal membrane base; and hyperpolarization causing a shorter is depolarized, an inward current is established, car- ϩ ϩ strength–duration time constant and increased rheo- ried by Na ions. The Na conductance is voltage base. When measured in isolation, strength–duration sensitive and regenerative: it increases with depolar- time constant, however, is of limited utility as a marker ization, and this in turn leads to greater depolariza- of membrane potential, as it is also affected by other tion as well as depolarization of the next node.75 This factors including demyelination and discrete changes explosive process would end with the whole axon ϩ ϩ in nodal Na conductances.28,100 depolarized, were it not that the Na channels im- ϩ Threshold electrotonus is the only clinical tech- mediately started closing again, due to Na channel nique available to assess alterations in both nodal inactivation. ϩ and internodal conductances. This method mea- Na channels are membrane-spanning protein sures the threshold changes produced by long-dura- molecules, containing a pore unit (␣-subunit) ϩ tion depolarizing and hyperpolarizing currents.30,37 through which Na ions can diffuse almost freely in These conditioning polarizing currents are set to the open state.171 The fast activation process (from defined percentages of the unconditioned threshold resting closed to open state) and the slower inacti- current, but importantly remain subthreshold in vation process (from open to inactivated state) con- that they do not trigger an action potential.37 The sist of conformational changes by the channel pro- response of the threshold current to the subthresh- tein, both driven by the changes in voltage gradient old conditioning currents is tested at varying condi- across it. A variety of toxins and drugs bind to the ϩ tioning–test intervals before, during, and after the ␣-subunits of Na channels. Most toxins bind to sites conditioning currents. By convention, threshold that are involved in activation and inactivation, ex- changes are plotted as threshold reductions (see cept for tetrodotoxin and its derivatives that occlude ϩ Figs. 5, 6), such that depolarizing responses are plot- the outer pore of the Na channel.100,154 Some of ted in an upward direction and hyperpolarizing re- these binding sites are also the target of mutation in ϩ sponses in a downward direction. In addition to hereditary Na channelopathies.101 providing information regarding internodal conduc- tances, threshold electrotonus is also sensitive to ASSESSMENT OF NERVE EXCITABILITY IN A changes in axonal membrane potential,11,92 with CLINICAL SETTING membrane depolarization causing a “fanning-in” ap- Assessment of nerve excitability using automated pearance of threshold electrotonus,86 whereas hy- protocols (Fig. 4) includes assessment of both nodal perpolarization leads to a “fanning-out” (Fig. 5). and internodal axonal properties. The activity of Information regarding axonal ion channel func- ϩ nodal persistent Na conductances may be assessed tion may also be gained through assessment of re- by measurement of the strength–duration time con- covery cycle parameters. Following conduction of a stant and rheobase. The strength–duration time single impulse, myelinated axons go through a ste- ␶ constant ( SD) refers to the rate at which the thresh- reotypical series of excitability changes known as the

280 Uremic Neuropathy MUSCLE & NERVE March 2007 FIGURE 4. (A) Arrangement for nerve excitability studies in ESKD patients undergoing hemodialysis. The current required to produce the desired compound motor action potential (CMAP) amplitude is determined using a computerized threshold-tracking program that is run by a personal computer. Recordings are amplified and digitized using an analog-to-digital board. Stimulus waveforms are converted to current with a purpose-built isolated linear bipolar constant current stimulator. The threshold tracking software consists of an automated tracking system in which the amplitude of the test stimulus is automatically increased or decreased by a percentage step after each response, depending on the difference between the recorded response and the target response, generally set to 40% of maximal CMAP amplitude for motor axons, as shown in (B). The protocol incorporates a proportional tracking system, in which the change in the stimulus is proportional to the difference between the actual and target responses. recovery cycle (Figs. 5, 6). For a period of 0.5–1.0 ms are associated with changes in latency, which is in- after an impulse, axons are completely inexcitable creased during the refractory period, decreased dur- and cannot generate another impulse regardless of ing superexcitability, and increased during late sub- the strength of the depolarizing stimulus. This pe- excitability.16,181 riod is known as the absolute refractory period. The The relative refractory period results from in- ϩ axon then enters a period of relative refractoriness activation of nodal transient Na channels. It is that can be measured either as the increase in cur- prolonged by membrane depolarization and re- rent required to produce a potential of a specified duced by hyperpolarization. Refractoriness may size, known as refractoriness, or as the duration of therefore be used a measure of membrane poten- the relative refractory period until threshold has tial, although it is essential to take into account returned to baseline, usually 3–4 ms. the effect of temperature, given that cooling leads This period of refractoriness is followed by a to an increase in refractoriness.94 Furthermore, period of superexcitability (or supernormality), dur- measures of refractoriness may be unreliable in ing which there is a reduction in threshold occurring situations of impaired distal transmission, as may over a 10–15-ms interval. Finally, there is a late phase occur with axonal demyelination, neuromuscular of raised threshold known as late subexcitability, junction abnormalities, muscle disease, or any ending around 100 ms. These changes in threshold other factor that reduces the security of impulse

Uremic Neuropathy MUSCLE & NERVE March 2007 281 FIGURE 5. Six plots of excitability parameters recorded from tibialis anterior for a single representative patient (continuous lines with circles) prior to dialysis with 95% confidence intervals (broken lines) for a single subject.95 (A) Absolute stimulus–response relationship for stimuli of 0.2 ms duration (line without filled circle) and 1-ms duration (line with filled circle). The filled circle on the 1-ms response curve corresponds to the threshold for a CMAP 50% of maximum and the broken ellipse corresponds to the 95% confidence limits for a single subject. (B) Normalized stimulus–response relationship. (C) Current–threshold relationship, reflecting rectifying properties of the axon following polarizing currents. Threshold changes to hyperpolarizing current are represented on the left and to depolarizing current on the right. (D) Distribution of strength–duration time constants of nine populations of axons, from 5% to 95% of maximal CMAP in groups of 10%. (E) Threshold electrotonus. Changes are plotted as threshold reductions, with depolarization represented in an upward direction and hyperpolarization in a downward direction. (F) Recovery cycle, showing the relative refractory period, superexcitability, and late subexcitability. There is “fanning-in” of threshold electrotonus,76,82 an increase in refractoriness, and reductions in superexcitability and late subexcitability (reproduced with permission from Oxford University Press; Krishnan et al.,113 fig. 1).

ϩ transmission.37,114,120,190 Alteration in refractori- Na conductances.100,103,126 In a recent study, an ness may also occur secondary to changes in nodal increase in refractoriness was demonstrated in pa- ϩ Na conductances. Reductions in refractoriness tients treated with the chemotherapeutic agent have been demonstrated in diabetic and toxic neu- oxaliplatin, in the absence of significant changes ropathies, consistent with a reduction in the nodal in other excitability parameters, suggesting that

282 Uremic Neuropathy MUSCLE & NERVE March 2007 FIGURE 6. Comparison of threshold electrotonus (A,C) and recovery cycle (B,D) in ESKD patients (continuous lines with circles) pre- and postdialysis, with 95% confidence intervals for normal controls (broken lines). Predialysis traces demonstrate “fanning-in” of threshold electrotonus, increased refractoriness, and reduced superexcitability and late subexcitability. Abnormalities have largely resolved 1 h postdialysis (reproduced with permission from Oxford University Press; Krishnan et al.,113 fig. 2).

the neurotoxicity of this agent may be mediated by is unchanged. It may therefore be used to differen- ϩ blockage of nodal transient Na channels.110 tiate between pure depolarization and depolariza- Superexcitability is due to a depolarizing afterpo- tion secondary to nerve ischemia in which there is no tential that results from the capacitative charging of significant change in late subexcitability due to com- ϩ the internode by the action potential12 and subse- pensatory changes in extracellular K ions.92 quently discharges through high resistance pathways 40,99 under or through the myelin sheath. Recent NERVE EXCITABILITY STUDIES IN ESKD studies have also suggested that activation of nodal ϩ Na conductances may be a further contributing Nerve excitability studies in ESKD patients (Fig. 4) factor.133 As for refractoriness, superexcitability also have recently demonstrated significant alterations in varies with membrane potential, with depolarization membrane potential prior to hemodialysis, with re- leading to a reduction in superexcitability and hy- covery in the postdialysis period.97,113,115–117 Prior to perpolarization causing an increase. Late subexcit- dialysis, measures of nerve excitability were signifi- ϩ ability is determined by activation of nodal slow K cantly abnormal in ESKD patients compared to con- channels and the difference between membrane po- trol data.93,95,108 Stimulus–response curves were ϩ tential (Er) and the K equilibrium potential (Ek) shifted to the right, indicating that axons were of ϩ and increases with depolarization if extracellular K high threshold. This was accompanied by a fan-

Uremic Neuropathy MUSCLE & NERVE March 2007 283 Potassium satisfies criteria that have been sug- gested for a substance to be accepted as a uremic neurotoxin.38 It is an identifiable chemical that is elevated in the serum of ESKD patients and causes neurological dysfunction in both humans and ani- mals. It is also a critical determinant of axonal rest- ing membrane potential.97 Moreover, a direct rela- ϩ tionship exists between serum levels of K and FIGURE 7. Relationship between predialysis depolarizing thresh- neurophysiological parameters, and its removal old electrotonus at the 90–100 ms interval (TEd 90–100ms) and Kϩ for pooled data from median sensory studies (open square) leads to considerable improvement in these indi- and lower-limb motor studies (TA: filled diamond; EDB, open ces.97,113,115,117 The resting potentials of both nerve diamond).113,117 and muscle membranes are largely determined by ϩ ϩ K , with relative changes in K capable of depolar- izing membranes. It may be argued that the abnormalities of serum ning-in of threshold electrotonus curves,86,92 which ϩ K noted in excitability studies are the consequence assess alterations in both nodal and internodal con- of a transient homeostatic disturbance, rapidly cor- ductances and are sensitive indicators of change in rected by dialysis, and therefore unlikely to play a membrane potential (Figs. 5, 6). Measurement of major role in the development of chronic irrevers- recovery cycles of excitability demonstrated that re- ible neuropathy. Against such an argument, pro- fractoriness, due to inactivation of voltage-gated ϩ longed exposure to hyperkalemia in ESKD patients transient Na channels, was increased and superex- ϩ seems likely, given that postdialysis rebound of K is citability, due to the depolarizing afterpotential,12 a well-recognized phenomenon,2,51 with hyperkale- was reduced. The findings were consistent with pre- mia typically recurring within6hofadialysis session dialysis axonal depolarization in ESKD patients.92 due to reequilibration between intracellular and ex- Measures of motor and sensory nerve excitability tracellular fluid compartments.19 Such prolonged have been assessed in relation to changes in serum ϩ 2ϩ hyperkalemia may cause disruption of normal ionic levels of potential neurotoxins, including K ,Ca ϩϩ gradients and activate damaging Ca -mediated urea, uric acid, and middle molecules such as PTH processes, leading to axonal loss.48 Furthermore, and ␤-2M. Predialysis excitability abnormalities were ϩ ϩ given the importance of K in mediating these ab- noted to be strongly correlated with serum K in all normalities, current measures of dialysis adequacy,50 studies, suggesting that hyperkalemic depolarization which are based solely on blood urea concentrations, may underlie the development of uremic neuropa- may be inappropriate for determining the adequacy thy (Fig. 7). Furthermore, abnormalities of excitabil- ϩ of a dialysis regimen to prevent neurotoxicity. A ity became apparent with serum K concentrations better indication of adequate dialysis might be the in the high normal range, well below the levels re- ϩ maintenance of serum K within normal limits be- quired to produce cardiac toxicity.38,115 The excit- tween periods of dialysis, which may require more ability abnormalities in ESKD patients were also dif- ϩ attention to dietary restriction of K intake in some ferent from those noted in patients with diabetic patients.113 neuropathy,111 another common metabolic neurop- athy, suggesting that the abnormalities noted in ؉ ؉ ESKD patients were not purely a consequence of AXONAL NA /K PUMP FUNCTION IN ESKD PATIENTS structural changes. ϩ ϩ Predialysis excitability recordings also demon- Inhibition of the Na /K pump by uremic neurotox- strated an apparently paradoxical reduction in late ins, previously proposed as the mechanism underlying ϩ subexcitability, determined in part by nodal slow K the development of uremic neuropathy,148 may induce channels. Since late subexcitability depends on the membrane depolarization.16,31,32,35,92 Of further rele- ϩ difference between the resting potential and the K vance, previous studies have demonstrated that alter- equilibrium potential and actually increases with de- ations in membrane potential and intra- and extracel- ϩ ϩ ϩ ϩ polarization if extracellular K is unchanged, the lular K concentration have a direct effect on Na /K reduction in this parameter was thought to provide pump function.78,77,109,140,159,160 further evidence that predialysis axonal membrane Nerve excitability measurements provide a means depolarization in patients with ESKD was due to of detecting changes in membrane potential caused ϩ ϩ ϩ effects mediated by serum K .97,113 by activation of the Na /K pump.29,34,98 Vagg et

284 Uremic Neuropathy MUSCLE & NERVE March 2007 the highly depolarized levels of the ischemic period. Importantly, however, the rapid return of threshold to baseline levels in the postischemic period con- ϩ ϩ firmed that the Na /K pump was functioning well in ESKD.

CONCLUSION Neuropathy is a common complication of ESKD, occurring in the majority of patients undergoing dialysis. At present, renal transplantation remains the only known cure for uremic neuropathy.20,21 Recent nerve excitability studies have suggested that hyperkalemia may underlie the development of neu- ropathy and have argued against any dysfunction of ϩ ϩ the axonal Na /K pump in the development of ϩ this condition. Excess K fits the profile of the neu- rotoxin responsible for uremic neuropathy better FIGURE 8. Comparison of excitability changes following maximal than middle molecules, parathyroid hormone, or voluntary contraction in ESKD patients predialysis (filled dia- any other organic substance that has been previously mond) and normal controls (open diamond). The filled arrow indicates the time of contraction. (A) Changes in normalized linked to the development of uremic neuropathy. threshold for a 1-ms stimulus duration. (B) Strength–duration Recent findings from nerve excitability studies in ␶ time constant, SD. (C) Superexcitability. (D) Submaximal CMAP ESKD patients suggest that maintenance of serum amplitude. Threshold and CMAP amplitude changes are normal- ized to precontraction values and are expressed as mean data Ϯ SEM. Although baseline superexcitability was significantly less in ESKD patients, there were no significant differences between patients and controls in the magnitude of excitability changes induced by activity,99 arguing against Naϩ/Kϩ pump dysfunction in this condition (reprinted with permission from International Federation of Clinical Neurophysiology; Krishnan et al.,115 fig. 3).

al.190 showed that the activity-dependent hyperpolar- ization (ADH) of motor axons induced by voluntary activity causes threshold changes that can be used to ϩ ϩ assess Na /K pump function. Assessment of activ- ity-dependent excitability changes prior to hemodi- alysis in 10 ESKD patients demonstrated quantita- tively similar changes in ESKD patients and controls, arguing against any significant reduction in the ax- ϩ ϩ onal Na /K pump in ESKD (Fig. 8).115 ϩ ϩ A further study also assessed Na /K pump func- tion in six ESKD patients prior to dialysis by moni- toring the excitability changes that occurred before, during, and after 13 min of nerve ischemia.116 With the onset of ischemia, a short-lived threshold reduc- tion was followed by a rapid increase in threshold FIGURE 9. Excitability parameters recorded from tibialis anterior for ESKD patients and controls before, during, and after the (Fig. 9). Whereas normal controls manifested a post- ischemic period (indicated by filled horizontal bar). Mean data Ϯ ischemic threshold increase due to increased activity SEM. During ischemia refractoriness increased and superexcit- ϩ ϩ of the Na /K pump and consequent membrane ability decreased in both groups, consistent with axonal depolar- hyperpolarization,112 a paradoxical reduction in ization. The reversal of these changes in the postischemic period threshold was noted in ESKD patients. This pattern is consistent with axonal hyperpolarization. There is a smaller ϩ ϩ threshold reduction during ischemia in ESKD patients and a suggested that the axonal Na /K pump had a sta- prominent ischemic threshold increase beyond baseline values, a bilizing effect on the axon and was attempting to feature not observed in control subjects (reproduced with permis- return membrane potential toward baseline from sion from Oxford University Press; Krishnan et al.,116 fig. 3).

Uremic Neuropathy MUSCLE & NERVE March 2007 285 ϩ K within normal limits between periods of dialysis, 19. Blumberg A, Roser HW, Zehnder C, Muller-Brand J. Plasma potassium in patients with terminal renal failure during and rather than simple avoidance of hyperkalemia, is after haemodialysis; relationship with dialytic potassium re- likely to reduce the incidence and severity of uremic moval and total body potassium. Nephrol Dial Transplant neuropathy. 1997;12:1629–1634. 20. Bolton CF, Baltzan MA, Baltzan RB. Effects of renal trans- Grant support was received from the Australian Brain Founda- plantation on uremic neuropathy. A clinical and electro- tion, Sylvia and Charles Viertel Charitable Foundation, National physiologic study. N Engl J Med 1971;284:1170–1175. Health and Medical Research Council of Australia and the Aus- 21. Bolton CF. Electrophysiologic changes in uremic neuropa- tralian Association of Neurologists. We thank Hugh Bostock and thy after successful renal transplantation. Neurology 1976; David Burke for help in the development of the excitability stud- 26:152–161. 22. Bolton CF, Driedger AA, Lindsay RM. Ischaemic neuropathy ies, and Bruce Pussell and John Charlesworth for assistance in in uraemic patients caused by bovine arteriovenous shunt. studies in uremic patients. J Neurol Neurosurg Psychiatry 1979;42:810–814. 23. Bolton CF. Peripheral neuropathies associated with chronic renal failure. Can J Neurol Sci 1980;7:89–96. 24. Bolton CF, Carter K, Koval JJ. Temperature effects on con- REFERENCES duction studies of normal and abnormal nerve. Muscle 1. Ackil AA, Shahani BT, Young RR, Rubin NE. Late response Nerve 1982;5:S145–147. and sural conduction studies. Usefulness in patients with 25. Bolton CF, Young GB. Neurological complications of renal chronic renal failure. Arch Neurol 1981;38:482–485. disease. Stoneham, UK: Butterworth-Heinemann; 1990. 2. Ahmed J, Weisberg LS. Hyperkalemia in dialysis patients. 26. Bolton CF, McKeown MJ, Chen R, Toth B, Remtulla H. Semin Dial 2001;14:348–356. Subacute uremic and diabetic polyneuropathy. Muscle 3. Ahonen RE. Peripheral neuropathy in uremic patients and Nerve 1997;20:59–64. in renal transplant recipients. Acta Neuropathol 1981;54:43– 27. Bolton CF, Remtulla H, Toth B, Bernardi L, Lindsay RM, 53. Maryniak O, et al. Distinctive electrophysiological features of 4. Angus-Leppan H, Burke D. The function of large and small denervated muscle in uremic patients. J Clin Neurophysiol nerve fibers in renal failure. Muscle Nerve 1992;15:288–294. 1997;14:539–5542. 5. Appenzeller O, Kornfeld M, MacGee J. Neuropathy in 28. Bostock H. The strength–duration relationship for excita- chronic renal disease. Arch Neurol 1971;24:449–461. tion of myelinated nerve: computed dependence on mem- brane parameters. J Physiol (Lond) 1983;341:59–74. 6. Asbury AK, Victor M, Adams RD. Uremic polyneuropathy. 29. Bostock H, Grafe P. Activity-dependent excitability changes Arch Neurol Psychiatr 1963;8:413–428. in normal and demyelinated rat spinal root axons. J Physiol 7. Asbury AK. Recovery from uremic neuropathy. N Engl J Med (Lond) 1985;365:239–257. 1971;284:1211–1212. 30. Bostock H, Baker M. Evidence for two types of potassium 8. Avram MM, Feinfeld DA, Huatuco AH. Search for the ure- channel in human motor axons in vivo. Brain Res 1988;462: mic toxin. Decreased motor-nerve conduction velocity and 354–358. elevated parathyroid hormone in uremia. N Engl J Med 31. Bostock H, Baker M, Grafe P, Reid G. Changes in excitability 1978;298:1000–1003. and accommodation of human motor axons following brief 9. Babb AL, Popovich RP, Christopher TG, Scribner BH. The periods of ischaemia. J Physiol (Lond) 1991;441:513–535. genesis of the square meter-hour hypothesis. Trans Am Soc 32. Bostock H, Baker M, Reid G. Changes in excitability of Artif Intern Org 1971;17:81–91. human motor axons underlying post-ischaemic fascicula- 10. Babb AL, Ahmad S, Bergstrom J, Scribner BH. The middle tions: evidence for two stable states. J Physiol (Lond) 1991; molecule hypothesis in perspective. Am J Kidney Dis 1981; 441:537–557. 1:46–50. 33. Bostock H. Impulse propagation in experimental neuropa- 11. Baker M, Bostock H. Depolarization changes the mechanism thy. In: Dyck PJ, Thomas PK, Griffin JW, Low PA, Poduslo JF, of accommodation in rat and human motor axons. J Physiol editors. Peripheral neuropathy. Philadelphia: Saunders; (Lond) 1989;411:545–561. 1993. p 109–120. 12. Barrett EF, Barrett JN. Intracellular recording from verte- 34. Bostock H, Bergmans J. Post-tetanic excitability changes and brate myelinated axons: mechanism of the depolarizing af- ectopic discharges in a human motor axon. Brain 1994;117: terpotential. J Physiol (Lond) 1982;323:117–144. 913–928. 13. Bazzi C, Pagani C, Sorgato G, Albonico G, Fellin G, D’Amico 35. Bostock H, Burke D, Hales JP. Differences in behaviour of G. Uremic polyneuropathy: a clinical and electrophysiolog- sensory and motor axons following release of ischaemia. ical study in 135 short- and long-term hemodialyzed patients. Brain 1994;117:225–234. Clin Nephrol 1991;35:176–181. 36. Bostock H, Rothwell JC. Latent addition in motor and sen- 14. Bellinghieri G, Santoro D, Lo Forti B, Mallamace A, De sory fibres of human peripheral nerve. J Physiol (Lond) Santo RM, Savica V. Erectile dysfunction in uremic dialysis 1997;498:277–294. patients: diagnostic evaluation in the sildenafil era. Am J 37. Bostock H, Cikurel K, Burke D. Threshold tracking tech- Kidney Dis 2001;38:S115–117. niques in the study of human peripheral nerve. Muscle 15. Benz RL, Siegfried JW, Teehan BP. Carpal tunnel syndrome Nerve 1998;21:137–158. in dialysis patients: comparison between continuous ambu- 38. Bostock H, Walters RJ, Andersen KV, Murray NM, Taube D, latory peritoneal dialysis and hemodialysis populations. Am J Kiernan MC. Has potassium been prematurely discarded as Kidney Dis 1988;11:473–476. a contributing factor to the development of uraemic neu- 16. Bergmans J. The physiology of single human nerve fibres. ropathy? Nephrol Dial Transplant 2004;19:1054–1057. Vander, Belgium: University of Louvain; 1970. 39. Brouns R, De Deyn PP. Neurological complications in renal 17. Bicknell JM, Lim AC, Raroque HG Jr, Tzamaloukas AH. failure: a review. Clin Neurol Neurosurg 2004;107:1–16. Carpal tunnel syndrome, subclinical median mononeuropa- 40. Burke D, Kiernan MC, Bostock H. Excitability of human thy, and peripheral polyneuropathy: common early compli- axons. Clin Neurophysiol 2001;112:1575–1585. cations of chronic peritoneal dialysis and hemodialysis. Arch 41. Caccia MR, Mangili A, Mecca G, Ubiali E, Zanoni P. Effects Phys Med Rehabil 1991;72:378–381. of hemodialytic treatment on uremic polyneuropathy. A 18. Bittar EE. Maia muscle fibre as a model for the study of clinical and electrophysiological follow-up study. J Neurol uraemic toxicity. Nature 1967;214:310–312. 1977;217:123–131.

286 Uremic Neuropathy MUSCLE & NERVE March 2007 42. Cappelen-Smith C, Lin CS, Kuwabara S, Burke D. Conduc- 64. Ginn HE, Bugel HJ, James L, Hopkins P. Clinical experience tion block during and after ischaemia in chronic inflamma- with small surface area dialyzers (SSAD). Proc Clin Dial tory demyelinating polyneuropathy. Brain 2002;125:1850– Transplant Forum 1971;1:53–60. 1858. 65. Goldstein DA, Chui LA, Massry SG. Effect of parathyroid 43. Castaigne P, Cathala HP, Beaussart-Boulenge L, Petrover M. hormone and uremia on peripheral nerve calcium and mo- Effect of ischaemia on peripheral nerve function in patients tor nerve conduction velocity. J Clin Invest 1978;62:88–93. with chronic renal failure undergoing dialysis treatment. 66. Gousheh J, Iranpour A. Association between carpel tunnel J Neurol Neurosurg Psychiatry 1972;35:631–637. syndrome and arteriovenous fistula in hemodialysis patients. 44. Chang MH, Chou KJ. The role of autonomic neuropathy in Plast Reconstr Surg 2005;116:508–513. the genesis of intradialytic hypotension. Am J Nephrol 2001; 67. Graf RJ, Halter JB, Pfeifer MA, Halar E, Brozovich F, Porte D 21:357–361. Jr. Glycemic control and nerve conduction abnormalities in 45. Chaumont P, Lefebvre J, Lerique JL. Explorations e´lec- non-insulin-dependent diabetic subjects. Ann Intern Med trologiques au cours des insufficicances re´nales graves. Rev 1981;94:307–311. Neurol 1963;108:199. 68. Halar EM, Brozovich FV, Milutinovic J, Inouye VL, Becker 46. Codish SD, Cress RH, Bolton CF. Uremic neuropathy. VM. H-reflex latency in uremic neuropathy: correlation with N Engl J Med 1971;285:752–753. NCV and clinical findings. Arch Phys Med Rehabil 1979;60: 47. Cofan F, Garcia S, Combalia A, Segur JM, Oppenheimer F. 174–177. Carpal tunnel syndrome secondary to uraemic tumoral cal- 69. Hamilton DV, Evans DB, Henderson RG. Ulnar nerve lesion cinosis. Rheumatology (Oxford) 2002;41:701–703. as complication of Cimino–Brescia arteriovenous fistula. 48. Craner MJ, Hains BC, Lo AC, Black JA, Waxman SG. Co- Lancet 1980;2:1137–1138. localization of sodium channel Nav1.6 and the sodium-cal- 70. Harding AE, Le Fanu J. Carpal tunnel syndrome related to cium exchanger at sites of axonal injury in the spinal cord in antebrachial Cimino–Brescia fistula. J Neurol Neurosurg Psy- EAE. Brain 2004;127:294–303. chiatry 1977;40:511–513. 49. D’Amour ML, Dufresne LR, Morin C, Slaughter D. Sensory 71. Hassan K, Simri W, Rubenchik I, Manelis J, Gross B, Shasha nerve conduction in chronic uremic patients during the first SM, et al. Effect of erythropoietin therapy on polyneurop- six months of hemodialysis. Can J Neurol Sci 1984;11:269– athy in predialytic patients. J Nephrol 2003;16:121–125. 271. 72. Hegstrom RM, Murray JS, Pendras JP, Burnell JM, Scribner 50. Daugirdas JT. Dialysis adequacy and kinetics. Curr Opin BH. Two year’s experience with periodic hemodialysis in the Nephrol Hypertens 2000;9:599–605. treatment of chronic uremia. Trans Am Soc Artif Intern 51. De Nicola L, Bellizzi V, Minutolo R, Cioffi M, Giannattasio P, Organs 1962;8:266–280. Terracciano V, et al. Effect of dialysate sodium concentra- 73. Heidbreder E, Schafferhans K, Heidland A. Autonomic neu- tion on interdialytic increase of potassium. J Am Soc Neph- ropathy in chronic renal insufficiency. Comparative analysis rol 2000;11:2337–2343. of diabetic and nondiabetic patients. Nephron 1985;41:50– 52. Delmez JA, Holtmann B, Sicard GA, Goldberg AP, Harter 56. HR. Peripheral nerve entrapment syndromes in chronic he- 74. Heidbreder E, Schafferhans K, Heidland A. Disturbances of modialysis patients. Nephron 1982;30:118–123. peripheral and autonomic nervous system in chronic renal 53. Di Giulio S, Chkoff N, Lhoste F, Zingraff J, Drueke T. failure: effects of hemodialysis and transplantation. Clin Parathormone as a nerve poison in uremia. N Engl J Med Nephrol 1985;23:222–228. 1978;299:1134–1135. 75. Hille B. Ionic channels of excitable membranes. Sunder- 54. Dinn JJ, Crane DL. Schwann cell dysfunction in uraemia. land, MA: Sinauer Associates; 1992. J Neurol Neurosurg Psychiatry 1970;33:605–608. 76. Hirasawa Y, Ogura T. Carpal tunnel syndrome in patients on 55. Dyck PJ, Johnson WJ, Lambert EH, O’Brien PC. Segmental long-term haemodialysis. Scand J Plast Reconstr Surg Hand demyelination secondary to axonal degeneration in uremic Surg 2000;34:373–381. neuropathy. Mayo Clin Proc 1971;46:400–431. 77. Hodgkin AL, Keynes RD. Movement of cations during recov- 56. Dyck PJ, Johnson WJ, Nelson RA, Lambert EH, O’Brien PC. ery in nerve. Symp Soc Exp Biol 1954;8:423–437. Uremic neuropathy. III. Controlled study of restricted pro- 78. Hodgkin AL, Keynes RD. Active transport of cations in giant tein and fluid diet and infrequent hemodialysis versus con- axons from Sepia and Logilo. J Physiol (Lond) 1955;128:28– ventional hemodialysis treatment. Mayo Clin Proc 1975;50: 60. 641–649. 79. Hoke A, Keswani SC. Neuroprotection in the PNS: erythro- 57. Dyck PJ, Sherman WR, Hallcher LM, Service FJ, O’Brien PC, poietin and immunophilin ligands. Ann N Y Acad Sci 2005; Grina LA, et al. Human diabetic endoneurial sorbitol, fruc- 1053:491–501. tose, and myo-inositol related to sural nerve morphometry. 80. Ibrahim MM, Barnes AD, Crosland JM, Dawson-Edwards P, Ann Neurol 1980;8:590–596. Honigsberger L, Newman CE, et al. Effect of renal transplan- 58. Dyck PJ. Detection, characterization, and staging of polyneu- tation of uraemic neuropathy. Lancet 1974;2:739–742. ropathy: assessed in diabetics. Muscle Nerve 1988;11:21–32. 81. Jain VK, Cestero RV, Baum J. Carpal tunnel syndrome in 59. Ewing DJ, Winney R. Autonomic function in patients with patients undergoing maintenance hemodialysis. JAMA 1979; chronic renal failure on intermittent haemodialysis. 242:2868–2869. Nephron 1975;15:424–429. 82. Jebsen RH, Tenckhoff H, Honet JC. Natural history of ure- 60. Furst P, Asaba M, Gordon A, Zimmerman L, Bergstrom J. mic polyneuropathy and effects of dialysis. N Engl J Med Middle molecules in uraemia. Proc Eur Dial Transplant 1967;277:327–333. Assoc 1975;11:417–426. 83. Jog MS, Turley JE, Berry H. Femoral neuropathy in renal 61. Furst P, Zimmerman L, Bergstrom J. Determination of en- transplantation. Can J Neurol Sci 1994;21:38–42. dogenous middle molecules in normal and uremic body 84. Kaji R, Sumner AJ. Ouabain reverses conduction distur- fluids. Clin Nephrol 1976;3:178–188. bances in single demyelinated nerve fibers. Neurology 1989; 62. Garcia S, Cofan F, Combalia A, Campistol JM, Oppenheimer 39:1364–1368. F, Ramon R. Compression of the ulnar nerve in Guyon’s 85. Kaji R, Bostock H, Kohara N, Murase N, Kimura J, Shibasaki canal by uremic tumoral calcinosis. Arch Orthop Trauma H. Activity-dependent conduction block in multifocal motor Surg 2000;120:228–230. neuropathy. Brain 2000;123:1602–1611. 63. Gejyo F, Narita I. Current clinical and pathogenetic under- 86. Kaji R. Physiology of conduction block in multifocal motor standing of beta2-m amyloidosis in long-term haemodialysis neuropathy and other demyelinating neuropathies. Muscle patients. Nephrology 2003;8(Supp):S45–49. Nerve 2003;27:285–296.

Uremic Neuropathy MUSCLE & NERVE March 2007 287 87. Kaku DA, Malamut RI, Frey DJ, Parry GJ. Conduction block 108. Krishnan AV, Lin CS-Y, Kiernan MC. Nerve excitability prop- as an early sign of reversible injury in ischemic monomelic erties in lower limb motor axons:evidence for a length- neuropathy. Neurology 1993;43:1126–1130. dependent gradient. Muscle Nerve 2004;29:645–655. 88. Kanai K, Kuwabara S, Arai K, Sung JY, Ogawara K, Hattori T. 109. Krishnan AV, Colebatch JG, Kiernan MC. Hypokalemia in- Muscle cramp in Machado-Joseph disease: altered motor duces activity-dependent conduction block. Neurology 2005; axonal excitability properties and mexiletine treatment. 65:1309–1312. Brain 2003;126:965–973. 110. Krishnan AV, Goldstein D, Friedlander M, Kiernan MC. 89. Kersh ES, Kronfield SJ, Unger A, Popper RW, Cantor S, Oxaliplatin-induced neurotoxicity and the development of Cohn K. Autonomic insufficiency in uremia as a cause of neuropathy. Muscle Nerve 2005;32:51–60. hemodialysis-induced hypotension. N Engl J Med 1974;290: 111. Krishnan AV, Kiernan MC. Altered nerve excitability prop- 650–653. erties in established diabetic neuropathy. Brain 2005;128: 90. Keswani SC, Buldanlioglu U, Fischer A, Reed N, Polley M, 1178–1187. Liang H, et al. A novel endogenous erythropoietin mediated 112. Krishnan AV, Lin CS, Kiernan MC. Excitability differences in pathway prevents axonal degeneration. Ann Neurol 2004;56: lower-limb motor axons during and after ischemia. Muscle 815–826. Nerve 2005;31:205–213. 91. Keswani SC, Leitz GJ, Hoke A. Erythropoietin is neuropro- 113. Krishnan AV, Phoon RK, Pussell BA, Charlesworth JA, Bos- tective in models of HIV sensory neuropathy. Neurosci Lett tock H, Kiernan MC. Altered motor nerve excitability in 2004;371:102–105. end-stage kidney disease. Brain 2005;128:2164–2174. 92. Kiernan MC, Bostock H. Effects of membrane polarization 114. Krishnan AV, Kiernan MC. Axonal function and activity- and ischaemia on the excitability properties of human motor dependent excitability changes in myotonic dystrophy. Mus- axons. Brain 2000;123:2542–2551. cle Nerve 2006;33:627–636. 93. Kiernan MC, Burke D, Andersen KV, Bostock H. Multiple 115. Krishnan AV, Phoon RK, Pussell BA, Charlesworth JA, Bos- measures of axonal excitability: a new approach in clinical tock H, Kiernan MC. Neuropathy, axonal Naϩ/Kϩ pump testing. Muscle Nerve 2000;23:399–409. function and activity-dependent excitability changes in end- 94. Kiernan MC, Cikurel K, Bostock H. Effects of temperature stage kidney disease. Clin Neurophysiol 2006;117:992–997. on the excitability properties of human motor axons. Brain 116. Krishnan AV, Phoon RK, Pussell BA, Charlesworth JA, Bos- 2001;124:816–825. tock H, Kiernan MC. Ischaemia induces paradoxical changes 95. Kiernan MC, Lin CS, Andersen KV, Murray NM, Bostock H. in axonal excitability in end-stage kidney disease. Brain Clinical evaluation of excitability measures in sensory nerve. 2006;129:1585–1592. Muscle Nerve 2001;24:883–892. 117. Krishnan AV, Phoon RK, Pussell BA, Charlesworth JA, Kier- 96. Kiernan MC, Guglielmi JM, Kaji R, Murray NM, Bostock H. nan MC. Sensory nerve excitability and neuropathy in end- Evidence for axonal membrane hyperpolarization in multi- stage kidney disease. J Neurol Neurosurg Psychiatry 2006;77: focal motor neuropathy with conduction block. Brain 2002; 548–551. 125:664–675. 118. Kuchle C, Fricke H, Held E, Schiffl H. High-flux hemodial- 97. Kiernan MC, Walters RJ, Andersen KV, Taube D, Murray ysis postpones clinical manifestation of dialysis-related amy- NM, Bostock H. Nerve excitability changes in chronic renal loidosis. Am J Nephrol 1996;16:484–488. failure indicate membrane depolarization due to hyperkal- 119. Kumar S, Trivedi HL, Smith EK. Carpal tunnel syndrome: a aemia. Brain 2002;125:1366–1378. complication of arteriovenous fistula in hemodialysis pa- 98. Kiernan MC, Lin CS, Burke D. Differences in activity-depen- tients. Can Med Assoc J 1975;113:1070–1072. dent hyperpolarization in human sensory and motor axons. 120. Kuwabara S, Ogawara K, Sung JY, Mori M, Kanai K, Hattori J Physiol (Lond) 2004;558:341–349. T, et al. Differences in membrane properties of axonal and 99. Kiernan MC, Burke D, Bostock H. Nerve excitability mea- demyelinating Guillain–Barre´ syndromes. Ann Neurol 2002; sures: biophysical basis and use in the investigation of pe- 52:180–187. ripheral nerve disease. In: Dyck PJ, Thomas PK, editors. 121. Laaksonen S, Voipio-Pulkki L, Erkinjuntti M, Asola M, Falck Peripheral neuropathy. Philadelphia: Elsevier Saunders; B. Does dialysis therapy improve autonomic and peripheral 2005. p 113–130. nervous system abnormalities in chronic uraemia? J Intern 100. Kiernan MC, Isbister GK, Lin CS, Burke D, Bostock H. Acute Med 2000;248:21–26. tetrodotoxin-induced neurotoxicity after ingestion of puffer 122. Laaksonen S, Metsarinne K, Voipio-Pulkki LM, Falck B. Neu- fish. Ann Neurol 2005;57:339–348. rophysiologic parameters and symptoms in chronic renal 101. Kiernan MC, Krishnan AV, Lin CS, Burke D, Berkovic SF. failure. Muscle Nerve 2002;25:884–890. Mutation in the Naϩ channel subunit SCN1B produces par- 123. Ligtenberg G, Blankestijn PJ, Boomsma F, Koomans HA. No adoxical changes in peripheral nerve excitability. Brain change in automatic function tests during uncomplicated 2005;128:1841–1846. haemodialysis. Nephrol Dial Transplant 1996;11:651–656. 102. Kim SJ, Shin SJ, Kang ES. Endoscopic carpal tunnel release 124. Lindblom U, Tegner R. Thermal sensitivity in uremic neu- in patients receiving long-term hemodialysis. Clin Orthop ropathy. Acta Neurol Scand 1985;71:290–294. 2000;:141–148. 125. Lowitzsch K, Gohring U, Hecking E, Kohler H. Refractory 103. Kitano Y, Kuwabara S, Misawa S, Ogawara K, Kanai K, period, sensory conduction velocity and visual evoked poten- Kikkawa Y, et al. The acute effects of glycemic control on tials before and after haemodialysis. J Neurology Neurosurg axonal excitability in human diabetics. Ann Neurol 2004;56: Psychiatry 1981;44:121–128. 462–467. 126. Mackel R, Brink E. Conduction of neural impulses in dia- 104. Kjellstrand CM. Do middle molecules cause uremic intoxi- betic neuropathy. Clin Neurophysiol 2003;114:248–255. cation? Am J Kidney Dis 1981;1:51–56. 127. Mallamaci F, Zoccali C, Ciccarelli M, Briggs JD. Autonomic 105. Knezevic W, Mastaglia FL. Neuropathy associated with Bres- function in uremic patients treated by hemodialysis or CAPD cia-Cimino arteriovenous fistulas. Arch Neurol 1984;41: and in transplant patients. Clin Nephrol 1986;25:175–180. 1184–1186. 128. Mallick NP, Gokal R. Haemodialysis. Lancet 1999;353:737– 106. Konishi T, Nishitani H, Motomura S. Single fiber electro- 742. myography in chronic renal failure. Muscle Nerve 1982;5: 129. Man NK, Granger A, Rondon-Nucete M, Zingraff J, Jungers 458–461. P, Sausse A, et al. One year follow-up of short dialysis with a 107. Konotey-Ahulu FI, Baillod RA, Comty CM, Heron JR, Shal- membrane highly permeable to middle molecules. Proc Eur don S, Thomas PK. Effect of periodic dialysis on the periph- Dial Transplant Assoc 1973;10:236–246. eral neuropathy of end-stage renal failure. Br Med J 1965;2: 130. Mansouri B, Adybeig B, Rayegani M, Yasami S, Behshad V. 1212–1215. Uremic neuropathy and the analysis of electrophysiological

288 Uremic Neuropathy MUSCLE & NERVE March 2007 changes. Electromyography Clin Neurophysiol 2001;41:107– 152. Nielsen VK. The peripheral nerve function in chronic renal 115. failure. 8. Recovery after renal transplantation. Clinical as- 131. Martinelli P, Baruzzi A, Montagna P, Ravasio A, Poppi M. pects. Acta Med Scand 1974;195:163–170. Carpal tunnel syndrome in a patient with a Cimino–Brescia 153. Obeso JA, Marti-Masso JF, Asin JL, Remirez MV, Irigoyen V, fistula. Eur Neurol 1981;20:478–480. Iragui M, et al. Conduction velocity through the somesthetic 132. Massry SG. Parathyroid hormone: a uremic toxin. Adv Exp pathway in chronic renal failure. J Neurol Sci 1979;43:439– Med Biol 1987;223:1–17. 445. 154. Ogata N, Ohishi Y. Molecular diversity of structure and 133. McIntyre CC, Richardson AG, Grill WM. Modeling the ex- ϩ citability of mammalian nerve fibers: influence of afterpo- function of the voltage-gated Na channels. Jpn J Pharmacol tentials on the recovery cycle. J Neurophysiol 2002;87:995– 2002;88:365–377. 1006. 155. Ogura T, Makinodan A, Kubo T, Hayashida T, Hirasawa Y. 134. Meech PR. Femoral neuropathy following renal transplanta- Electrophysiological course of uraemic neuropathy in hae- tion. AustNZJSurg 1990;60:117–119. modialysis patients. Postgrad Med J 2001;77:451–454. 135. Miles AM. Vascular steal syndrome and ischaemic 156. Oh SJ, Clements RS Jr, Lee YW, Diethelm AG. Rapid im- monomelic neuropathy: two variants of upper limb isch- provement in nerve conduction velocity following renal aemia after haemodialysis vascular access surgery. Nephrol transplantation. Ann Neurol 1978;4:369–373. Dial Transplant 1999;14:297–300. 157. Panayiotopoulos CP, Lagos G. Tibial nerve H-reflex and 136. Meguid El Nahas A, Bello AK. Chronic kidney disease: the F-wave studies in patients with uremic neuropathy. Muscle global challenge. Lancet 2005;365:331–340. Nerve 1980;3:423–426. 137. Mitz M, Prakash AS, Melvin J, Piering W. Motor nerve con- 158. Preswick G, Jeremy D. Subclinical polyneuropathy in renal duction indicators in uremic neuropathy. Arch Phys Med insufficiency. Lancet 1964:731. Rehabil 1980;61:45–48. 159. Rakowski RF, Gadsby DC, De Weer P. Stoichiometry and 138. Mogyoros I, Kiernan MC, Burke D. Strength-duration prop- voltage dependence of the sodium pump in voltage- erties of human peripheral nerve. Brain 1996;119:439–447. clamped, internally dialyzed squid giant axon. J Gen Physiol 139. Mogyoros I, Lin C, Dowla S, Grosskreutz J, Burke D. 1989;93:903–941. Strength–duration properties and their voltage dependence 160. Rang HP, Ritchie JM. On the electrogenic sodium pump in at different sites along the median nerve. Clin Neurophysiol mammalian non-myelinated nerve fibres and its activation by 1999;110:1618–1624. various external cations. J Physiol (Lond) 1968;196:183–221. 161. Redfern AB, Zimmerman NB. Neurologic and ischemic 140. Morita K, David G, Barrett JN, Barrett EF. Posttetanic hyper- ϩ ϩ complications of upper extremity vascular access for dialysis. polarization produced by electrogenic Na -K pump in liz- J Hand Surg (Am) 1995;20:199–204. ard axons impaled near their motor terminals. J Neuro- 162. Riggs JE, Moss AH, Labosky DA, Liput JH, Morgan JJ, Gut- physiol 1993;70:1874–1884. mann L. Upper extremity ischemic monomelic neuropathy: 141. Naito M, Ogata K, Goya T. Carpal tunnel syndrome in a complication of vascular access procedures in uremic dia- chronic renal dialysis patients: clinical evaluation of 62 betic patients. Neurology 1989;39:997–998. hands and results of operative treatment. J Hand Surg (Br) 163. Ritchie J, Straub R. The hyperpolarization which follows 1987;12:366–374. activity in mammalian non-medullated fibres. J Physiol 142. Nardin R, Chapman KM, Raynor EM. Prevalence of ulnar (Lond) 1957;136:80–97. neuropathy in patients receiving hemodialysis. Arch Neurol 164. Rockel A, Hennemann H, Sternagel-Haase A, Heidland A. 2005;62:271–275. Uraemic sympathetic neuropathy after haemodialysis and 143. National Kidney Foundation. NKF-DOQI clinical practice transplantation. Eur J Clin Invest 1979;9:23–27. guidelines for hemodialysis adequacy. Am J Kidney Dis 1997; 165. Ropper AH. Accelerated neuropathy of renal failure. Arch 30(Suppl 2):S15–66. Neurol 1993;50:536–539. 144. Nielsen VK. The peripheral nerve function in chronic renal 166. Rossini PM, Treviso M, Di Stefano E, Di Paolo B. Nervous failure. II. Intercorrelation of clinical symptoms and signs impulse propagation along peripheral and central fibres in and clinical grading of neuropathy. Acta Med Scand 1971; patients with chronic renal failure. Electroencephalogr Clin 190:113–117. Neurophysiol 1983;56:293–303. 145. Nielsen VK. The peripheral nerve function in chronic renal 167. Said G, Boudier L, Selva J, Zingraff J, Drueke T. Different failure. I. Clinical symptoms and signs. Acta Med Scand patterns of uremic polyneuropathy: clinicopathologic study. 1971;190:105–111. Neurology 1983;33:567–574. 146. Nielsen VK, Winkel P. The peripheral nerve function in 168. Sato M, Horigome I, Chiba S, Furuta T, Miyazaki M, Hotta chronic renal failure. 3. A multivariate statistical analysis of O, et al. Autonomic insufficiency as a factor contributing to factors presumed to affect the development of clinical neu- dialysis-induced hypotension. Nephrol Dial Transplant 2001; ropathy. Acta Med Scand 1971;190:119–125. 16:1657–1662. 147. Nielsen VK. The peripheral nerve function in chronic renal 169. Schaefer K, Offermann G, von Herrath D, Schroter R, Stol- failure. IV. An analysis of the vibratory perception threshold. zel R, Arntz HR. Failure to show a correlation between Acta Med Scand 1972;191:287–296. serum parathyroid hormone, nerve conduction velocity and 148. Nielsen VK. The peripheral nerve function in chronic renal serum lipids in hemodialysis patients. Clin Nephrol 1980;14: failure. V. Sensory and motor conduction velocity. Acta Med 81–88. Scand 1973;194:445–454. 170. Schwarz A, Keller F, Seyfert S, Poll W, Molzahn M, Distler A. 149. Nielsen VK. The peripheral nerve function in chronic renal Carpal tunnel syndrome: a major complication in long-term failure. VI. The relationship between sensory and motor hemodialysis patients. Clin Nephrol 1984;22:133–137. nerve conduction and kidney function, azotemia, age, sex, 171. Schwarz JR, Eikhof G. Na currents and action potentials in and clinical neuropathy. Acta Med Scand 1973;194:455–462. rat myelinated nerve fibres at 20 and 37 degrees C. Pflugers 150. Nielsen VK. The peripheral nerve function in chronic renal Arch 1987;409:569–577. failure. IX. Recovery after renal transplantation. Electro- 172. Scribner BH, Buri R, Caner JE, Hegstrom R, Burnell JM. The physiological aspects (sensory and motor nerve conduction). treatment of chronic uremia by means of intermittent he- Acta Med Scand 1974;195:171–180. modialysis: a preliminary report. Trans Am Soc Artif Intern 151. Nielsen VK. The peripheral nerve function in chronic renal Organs 1960;6:114–122. failure. VII. Longitudinal course during terminal renal fail- 173. Sekiya H, Sugimoto N, Kariya Y, Hoshino Y. Carpal tunnel ure and regular hemodialysis. Acta Med Scand 1974;195: pressure in patients with carpal tunnel syndrome due to 155–162. long-term hemodialysis. Int Orthop 2002;26:274–277.

Uremic Neuropathy MUSCLE & NERVE March 2007 289 174. Shaheen FA, Mansuri NA, al-Shaikh AM, Sheikh IA, Huraib 193. Vanholder R, De Smet R, Hsu C, Vogeleere P, Ringoir S. SO, al-Khader AA, et al. Reversible uremic deafness: is it Uremic toxicity: the middle molecule hypothesis revisited. correlated with the degree of anemia? Ann Otol Rhinol Semin Nephrol 1994;14:205–218. Laryngol 1997;106:391–393. 194. Vita G, Savica V, Calabro R, Padovano I, Manna L, Toscano 175. Sharma KR, Cross J, Santiago F, Ayyar DR, Burke G 3rd. A, et al. Uremic vagal neuropathy: has parathyroid hormone Incidence of acute femoral neuropathy following renal a pathogenetic role? Funct Neurology 1986;1:253–259. transplantation. Arch Neurol 2002;59:541–545. 195. Vita G, Dattola R, Calabro R, Manna L, Venuto C, Toscano 176. Slatopolsky E, Martin K, Hruska K. Parathyroid hormone A, et al. Comparative analysis of autonomic and somatic metabolism and its potential as a uremic toxin. Am J Physiol dysfunction in chronic uraemia. Eur Neurol 1988;28:335– 1980;239:F1–12. 340. 177. Solders G, Tyden G, Persson A, Groth CG. Improvement of 196. Vita G, Dattola R, Puglisi RM, Marabello L, Venuto C, To- nerve conduction in diabetic neuropathy. A follow-up study scano A, et al. Autonomic nervous system dysfunction in 4 yr after combined pancreatic and renal transplantation. chronic uraemics on haemodialysis. Funct Neurol 1989;4: Diabetes 1992;41:946–951. 195–197. 178. Spallone V, Mazzaferro S, Tomei E, Maiello MR, Lungaroni 197. Vita G, Messina C, Savica V, Bellinghieri G. Uraemic auto- G, Sardella D, et al. Autonomic neuropathy and secondary nomic neuropathy. J Auton Nerv Syst 1990;30:S179–184. hyperparathyroidism in uraemia. Nephrol Dial Transplant 198. Vita G, Savica V, Puglisi RM, Marabello L, Bellinghieri G, 1990;5(Suppl 1):128–130. Messina C. The course of autonomic neural function in 179. Stanley E, Brown JC, Pryor JS. Altered peripheral nerve chronic uraemic patients during haemodialysis treatment. function resulting from haemodialysis. J Neurol Neurosurg Nephrol Dial Transplant 1992;7:1022–1025. Psychiatry 1977;40:39–43. 199. Vita G, Savica V, Milone S, Trusso A, Bellinghieri G, Messina 180. Stojcheva-Taneva OO, Polenakovic MH. Autonomic neurop- C. Uremic autonomic neuropathy: recovery following bicar- athy in hemodialysis patients treated with recombinant hu- bonate hemodialysis. Clin Nephrol 1996;45:56–60. man erythropoietin. Int J Artif Organs 1996;19:574–577. 200. Wang SJ, Liao KK, Liou HH, Lee SS, Tsai CP, Lin KP, et al. 181. Stys PK, Ashby P. An automated technique for measuring the Sympathetic skin response and R-R interval variation in recovery cycle of human nerves. Muscle Nerve 1990;13:750– chronic uremic patients. Muscle Nerve 1994;17:411–418. 758. 201. Warren DJ, Otieno LS. Carpal tunnel syndrome in patients 182. Tackmann W, Ullerich D, Cremer W, Lehmann HJ. Nerve on intermittent haemodialysis. Postgrad Med J 1975;51:450– conduction studies during the relative refractory period in 452. sural nerves of patients with uremia. Eur Neurol 1974;12: 202. Waxman SG, Ritchie JM. Molecular dissection of the myelin- 331–339. ated axon. Ann Neurol 1993;33:121–136. 183. Taylor N, Halar EM, Tenckhoff H, Marchioro TL, Masock 203. Weiber H, Linell F. Tumoral calcinosis causing acute carpal AJ. Effects of renal transplantation on motor nerve conduc- tunnel syndrome. Case report. Scand J Plast Reconstr Surg tion velocity. Arch Phys Med Rehabil 1972;53:227–231. Hand Surg 1987;21:229–230. 184. Tegner R, Lindholm B. Uremic polyneuropathy: different effects of hemodialysis and continuous ambulatory perito- 204. Weiss G. Sur la possibilite´ de rendre comparables entre eux neal dialysis. Acta Med Scand 1985;218:409–416. les appareils servant l’excitation e´lectrique. Arch Ital Biol 185. Tenckhoff HA, Boen ST, Jebsen RH, Spiegler JH. Polyneu- 1901;35:413–446. ropathy in chronic renal failure. JAMA 1965;192:1121–1124. 205. Welt LG, Sachs JR, McManus TJ. An ion transport defect in 186. Thiele B, Stalberg E. Single fibre EMG findings in polyneu- erythrocytes from uraemic subjects. Trans Assoc Am Phys ropathies of different aetiology. J Neurol Neurosurg Psychi- 1964;77:169–181. atry 1975;38:881–887. 206. Wilbourn AJ, Furlan AJ, Hulley W, Ruschhaupt W. Ischemic 187. Thomas PK, Hollinrake K, Lascelles RG, O’Sullivan DJ, Bail- monomelic neuropathy. Neurology 1983;33:447–451. lod RA, Moorhead JF, et al. The polyneuropathy of chronic 207. Wills MR, Savory J. Biochemistry of renal failure. Ann Clin renal failure. Brain 1971;94:761–780. Lab Sci 1981;11:292–299. 188. Thomas PK. Screening for peripheral neuropathy in patients 208. Wright EA, McQuillen MP. Hypoexcitability of ulnar nerve treated by chronic hemodialysis. Muscle Nerve 1978;1:396– in patients with normal motor nerve conduction velocities. 399. Neurology 1973;23:78–83. 189. Umapathi T, Chaudhry V. Toxic neuropathy. Curr Opin 209. Wu FF, Ryan A, Devaney J, Warnstedt M, Korade-Mirnics Z, Neurol 2005;18:574–580. Poser B, et al. Novel CLCN1 mutations with unique clinical 190. Vagg R, Mogyoros I, Kiernan MC, Burke D. Activity-depen- and electrophysiological consequences. Brain 2002;125: dent hyperpolarization of human motor axons produced by 2392–2407. natural activity. J Physiol (Lond) 1998;507:919–925. 210. Yazbeck S, Larbrisseau A, O’Regan S. Femoral neuropathy 191. Van den Neucker K, Vanderstraeten G, Vanholder R. Pe- after renal transplantation. J Urol 1985;134:720–721. ripheral motor and sensory nerve conduction studies in 211. Yoshida A, Okutsu I, Hamanaka I, Motomura T. Results of haemodialysis patients. A study of 54 patients. Electromyogr endoscopic management of primary versus recurrent carpal Clin Neurophysiol 1998;38:467–474. tunnel syndrome in long-term haemodialysis patients. Hand 192. van Ypersele de Strihou C, Jadoul M, Malghem J, Maldague Surg 2004;9:165–170. B, Jamart J. Effect of dialysis membrane and patient’s age on 212. Zoccali C, Ciccarelli M, Mallamaci F, Maggiore Q. Parasym- signs of dialysis-related amyloidosis. The Working Party on pathetic function in haemodialysis patients. Nephron 1986; Dialysis Amyloidosis. Kidney Int 1991;39:1012–1019. 44:351–354.

290 Uremic Neuropathy MUSCLE & NERVE March 2007