Autonomic and Peripheral (Sensorimotor) Neuropathy in Chronic Liver Disease: A Clinical and Electrophysiologic Study

Transcription

1 Autonomic and Peripheral (Sensorimotor) Neuropathy in Chronic Liver Disease: A Clinical and Electrophysiologic Study VINAY CHAUDHRY, 1 ANDREA M. CORSE, 1 RICHARD O BRIAN, 1 DAVID R. CORNBLATH, 1 ANDREW S. KLEIN, 2 AND PAUL J. THULUVATH 3 Peripheral neuropathy has been reported in association with chronic liver disease. However, the precise incidence, severity and characteristics of neuropathy, and the relationship of neuropathy to different etiologies of liver disease have not been defined. In this study, 58 patients with advanced liver disease were evaluated in detail for the presence of neuropathy. Peripheral (sensorimotor) neuropathy was found in 71% and autonomic neuropathy was found in 48% of the patients. Although the majority of patients were asymptomatic, neurological examination showed distal sensory loss to pain, or vibration or distal loss of reflexes in 17 patients (29%). Sensory neuropathy was seen more commonly than motor axonal polyneuropathy on nerve conduction studies. Quantitative sensory testing was frequently abnormal (62%) and cooling thresholds were more affected than vibration thresholds. Overall, the pattern of neuropathy in patients with liver disease conformed to the pattern expected in dying back or length-dependent neuropathy. The neuropathy was most severe in patients with advanced hepatic decompensation. Comparison of causes of liver disease showed no significant differences in the severity of neuropathy among the different etiologies. In conclusion, axonal sensory-motor polyneuropathy and autonomic neuropathy are commonly seen in patients with end-stage liver disease of different causes. (HEPATOLOGY 1999;29: ) Although peripheral neuropathy is reported to occur in patients with liver disease, there is controversy regarding a cause and effect relationship. 1,2 While some authors doubt the existence of hepatic neuropathy, 3 others report an incidence ranging from 19 to 100%. 4-8 Most authors agree that the neuropathy seen in liver disease is generally mild or subclinical but detailed characterizations of the nature of neuropathy are lacking. 4-8 Additionally, autonomic neuropathy has been reported in patients with alcoholic and nonalcoholic liver disease 9-15 though this has not been characterized in the context of peripheral neuropathy. Abbreviations: NCS, nerve conduction study; QST, quantitative sensitive testing; AFT, autonomic function testing; TNS, total neuropathy score. From the Departments of Neurology, 1 Surgery, 2 and Gastroenterology, 3 The Johns Hopkins University School of Medicine, Baltimore, MD. Received January 8, 1999; accepted March 29, Address reprint requests to: Paul J. Thuluvath, M.D., The Johns Hopkins University and School of Medicine, 600 North Wolfe Street, 1830 Building Room 429, Baltimore, MD PJTHULUV@WELCHLINK.WELCH.JHU.EDU; fax: Copyright 1999 by the American Association for the Study of Liver Diseases /99/ $3.00/ There has been no systematic characterization of peripheral neuropathy in the context of coexisting liver failure in the past 20 years. Because there have been many refinements in the diagnosis of peripheral neuropathy (e.g., quantitative sensory testing) during this period, we decided to reexamine the association of peripheral neuropathy and chronic liver disease. This prospective study was designed (1) to systematically define the prevalence of peripheral neuropathy and autonomic neuropathy in patients with liver disease, (2) to correlate the severity of neuropathy with the severity of liver disease, (3) to determine whether neuropathy is related to the cause of liver disease, and (4) to determine the prevalence of autonomic neuropathy and its relationship to sensory-motor neuropathy in liver disease patients. PATIENTS AND METHODS Fifty-eight patients (28 men, 30 women) with end-stage liver disease, all awaiting liver transplantation, were evaluated for the presence of peripheral neuropathy. Causes of the liver disease and other demographic data are as shown in Table 1. The severity of liver disease was graded using the Child-Pugh 16 classification: 26 were class A, 19 class B, and 13 class C. Three patients had diabetes (non insulin-dependent) and 12 patients had alcohol-induced liver disease; no other patient had a history or laboratory evidence of other causes of neuropathy including vitamin B 12, thiamine, human immunodeficiency virus, familial, toxic, or paraproteinemias. Patients with insulin-dependent diabetes were excluded from the study. Patients with alcohol-induced liver disease had a minimal abstinence period of 6 months before they were evaluated. All patients in Child-Pugh class A category had a history of decompensation secondary to variceal bleeding before being placed on the transplant list. All patients underwent a standardized neurological evaluation including a history and neurological examination, which was relevant to peripheral and autonomic nervous systems, nerve conduction studies (NCS), quantitative sensory tests (QST), and autonomic function tests (AFT). A history to elicit sensory or motor symptoms of neuropathy and a neurological examination testing for pin vibration sensibilities, strength, and deep tendon reflexes were performed by one of the authors (V.C., A.M.C., R.O.B., or D.R.C.). Nerve conduction studies were done for the right sural, peroneal, and median nerves using our standard laboratory techniques. For sensory nerves (sural and median) peak-to-peak amplitudes and conduction velocities were measured. For motor nerves (peroneal and median) baseline-topeak amplitudes, conduction velocities, and distal and F-wave latencies were measured. The right tibial H-reflex latency was also measured. Quantitative sensory tests were done with the CASE IV (WR Medical Electronics Co., Stillwater, MN) machine recording the cooling threshold at the dorsum of the foot and hand and the vibration threshold at the great toe and the index finger on the nondominant side. For each patient, symptoms, signs, abnormalities in NCS and in

2 HEPATOLOGY Vol. 29, No. 6, 1999 CHAUDHRY ET AL TABLE 1. Demographic Data and Results Etiology of Hepatic Disease No. of Patients Median Age (range) Years Sex (M:F) Child-Pugh (A:B:C) TNS Mean (range) Neuropathy (%) AFT (% abnormal) Alcohol-induced (35-63) Cryptogenic (47-66) Viral hepatitis (34-62) Primary biliary cirrhosis 5 57 (47-65) Sclerosing cholangitis (39-58) Others (24-65) Total (24-66) *Autoimmune, hemachromotosis, liver cancer, sarcoid, Budd-Chiary. 8:4 3:5:4 7.8 (0-14) 2:4 1:3:2 503 (1-13) 15:8 13:6:4 5 (0-18) 0:6 3:1:1 3 (0-5) 1:5 4:1:1 5.3 (0-19) 2:4 3:2:1 5.6 (0-11) 28:30 26:19: (0-19) QST were given a severity score and then added up to obtain a total neuropathy score (TNS), a validated instrument, 17,18 which ranged froma0to40. AFT included determination of blood pressure and heart rate changes on the tilt table, heart rate variation with deep breathing, and the relative risk ratio with the Valsalva maneuver. Abnormalities in AFTs were given a point each. Hence, the maximum score for AFT abnormality was 3. The presence of neuropathy was defined by the presence of abnormalities in at least 2 of 5 categories of symptoms, signs, NCS, QST, or AFT. 19,20 The severity of neuropathy was based on the TNS, with higher numbers representing worse neuropathy. To determine the statistical relationship between severity of liver disease, different causes of liver disease, and severity of neuropathy, Student s t test, ANOVA, and Pearson correlation were performed as appropriate. RESULTS Seventy-one percent of the patients had evidence of neuropathy (Fig. 1). Thirteen of 58 patients (22%) had symptoms consisting of mild cramps, paresthesias, or numbness. In no patient were the sensory symptoms disabling. One patient complained of generalized weakness and this patient was diagnosed as having Guillain-Barré syndrome. Seventeen patients (29%) had abnormalities on examination consisting of distal sensory loss to vibration in 12, and pin in 7; reduced or absent ankle reflexes in 12 patients; and toe extensor weakness (MRC 4) in 1 patient and proximal (4 MRC) and distal weakness (MRC 4 ) in 1 patient. Twenty-five patients (43%), had reduced sensory or motor amplitudes on the NCS. The results of the NCS are summarized in Table 2. The most common abnormality was reduced or absent sural sensory amplitudes in 20 patients, followed by reduced peroneal motor amplitude in 15 (Fig. 2). Median sensory amplitude was reduced in 9 patients and median motor amplitude was decreased in 2 patients. Tibial H-reflex was absent or abnormal in 17 patients (29%). Peroneal conduction velocity was reduced below the lower limit of normal in 14 patients (24%); in 6 patients (10%) the conduction velocity was reduced less than 85% of the lower limit of normal (Fig. 2). F-wave latency was abnormal in 22 (38%) of the patients, but in no patient was the latency increased to greater than 120% of the upper limit of normal (Fig. 2). Peroneal distal latency was prolonged in only 2 patients. Nerve conduction study in the median sensory and motor nerves showed focal reduction of conduction velocity across the carpal tunnel segment in 19 patients. The NCS in all patients were consistent with a distal sensory-motor axonal neuropathy except in 1 patient with sclerosing cholangitis who presented with back pain and mild diffuse weakness of 3 to 4 weeks duration. She was areflexic with distal sensory loss and diffuse weakness (proximal 4 distal 4). NCS showed markedly demyelinating features: the median nerve distal and F-wave latencies were markedly prolonged (18.3 ms, 49.5 ms, respectively) and the conduction velocity reduced (26 m/s, respectively). This patient was diagnosed as having Guillain-Barré syndrome. The abnormalities in the quantitative sensory test are illustrated in Fig. 3. TABLE 2. Results of Nerve Conduction Studies Test Number of Patients With Abnormal Test (%) FIG. 1. Percent of cirrhotic patients with at least one abnormality in symptoms, signs, NCS, AFT, or QST. The total represents percent of patients with abnormalities in two or more of the categories of symptoms, signs, NCS, AFT,orQST. Sural nerve sensory amplitude 20 (34) Median nerve sensory amplitude 9 (16) Peroneal nerve motor amplitude 15 (26) Median nerve motor amplitude 2 (3) Tibial H-reflex 17 (29) Peroneal nerve conduction velocity 14 (24) F-wave latency 22 (38)

3 1700 CHAUDHRY ET AL. HEPATOLOGY June 1999 FIG. 2. Top (A and B): Sensory and motor amplitudes in the 58 patients studied. Patients below the dotted line, lower limit of normal (LLN) for our laboratory, have electrophysiological evidence of sensory or motor axonal loss. Bottom (C and D): Motor conduction velocity (CV) and F-wave latency values in the peroneal nerve for the 58 patients. The dotted line represents the LLN for the CV (C) and the upper limit of normal (ULN) for the F-wave (D), respectively; and values below (C) or above (D) the dashed line would be considered in the demyelinating range for our laboratory. Thirty-five patients (60%) had abnormality of cooling threshold in either the hand or foot. Nineteen patients (33%) had abnormality of the vibration threshold either at the great toe or the index finger. Thirty-six patients had elevation of either cooling or vibration thresholds. Figure 4 illustrates the severity of neuropathy as graded by TNS. Fifteen patients (26%) had a TNS value of 0 or 1 and were asymptomatic and had normal examination and normal NCS, QST, and AFT; 20 patients (35%) had a TNS value between 2 and 5 and had subclinical neuropathy detected mainly by laboratory testing; 11 patients (19%) had a TNS of 6 to 10 and had mild neuropathy; 10 patients (17%) had a TNS of 11 to 15 and had moderate neuropathy; and 2 patients (3%) had a TNS between 16 and 20 with severe neuropathy. The total neuropathy score was higher in patients with more severe liver disease compared with patients with less severe disease. The mean TNS of Child-Pugh class A grade was 4.4, for class B was 4.6, and for class C was 9.2. Although the differences between Child-Pugh class A and B patients were not significant, those between class A and C patients (P.001) and class B and C patients (P.01) were statistically significant (Fig. 5). Furthermore, a statistically significant correlation (P.02) was obtained using the 2-tailed Pearson correlation of TNS against Child-Pugh score. Using ANOVA, no statistically significant correlation was obtained between the different causes of liver disease and the severity of neuropathy (Table 1). The abnormalities detected by autonomic function testing battery are shown in Fig. 3. The most frequent abnormality was reduced heart rate variation with deep breathing in 26 patients (45%). Abnormality of the relative risk ratio with Valsalva maneuver was present in 22 patients (38%) whereas an abnormal decrease in blood pressure with tilt table was present in 5 patients (9%). One or more of the three tests were found to be abnormal in 28 patients (48%). Ninety percent of the patients who had autonomic neuropathy also had evidence of somatic neuropathy as determined by abnormalities in one or more of the categories of symptoms, signs, nerve conduction, or QST abnormalities. No patient with abnormalities in the AFT had symptoms related to the autonomic dysfunction. Like the TNS, autonomic dysfunction was more pronounced in patients with more severe liver disease (Child- Pugh class C) compared with those with less severe liver disease (Child-Pugh class A and B), and these differences were statistically significant (P.01). Thirty-eight percent of Child-Pugh class A patients, 52% of Child-Pugh class B patients, and 75% of Child-Pugh class C patients showed evidence of autonomic neuropathy; the mean AFT score for Child-Pugh class A was 0.5, for B was 1, and for C was 1.4. DISCUSSION We prospectively evaluated 58 patients with end-stage liver disease and found that peripheral neuropathy was frequent, occurring in 71% of these patients. In most patients, the neuropathy was subclinical or mild with minimal symptoms. Examination showed distal sensory loss and loss of distal reflexes consistent with a length-dependent neuropathy. Electrophysiology confirmed this and suggested that the primary process is axonal degeneration. Only one patient had severe demyelinating physiology but this patient had coincidental Guillain-Barré syndrome. Median neuropathy at the wrist (carpal tunnel entrapment) was common occurring in 33% of the patients. Quantitative sensory testing showed abnormalities of small-fiber function (cooling threshold) (60%) greater than large fiber function (vibration threshold) (33%). Forty-eight percent of patients also showed evidence of autonomic neuropathy. The severity of sensory-motor and

4 HEPATOLOGY Vol. 29, No. 6, 1999 CHAUDHRY ET AL FIG. 5. Total neuropathy score is correlated with the severity of hepatic disease as defined by Child-Pugh classification of A, B, and C. Patients with Child-Pugh class C appeared to have a greater total neuropathy score compared with those in the class A and B categories. FIG. 3. (A) Percent of patients with abnormalities with the CASE IV quantitative sensory testing for cooling and vibratory threshold in the upper or lower limbs are shown here. Values at or above the 95th percentile are considered abnormal. The cool or vib category is for patients with abnormality in one or both of the two categories of cooling or vibration threshold respectively. (B) Autonomic function testing abnormalities with the tilt table, relative risk ratio after Valsalva (Val) maneuver, and heart rate variation with deep breathing (HR DB). FIG. 4. The distribution of the total neuropathy score, calculated by adding the measurements of symptoms, signs, NCS, and QST, is shown for the 58 patients. The number on the bar shows the number of patients. A score of 0 to 2 is considered within normal limits. autonomic neuropathy correlated with the severity of liver disease. In the literature, there is a wide range in the reported incidence of peripheral neuropathy in chronic liver disease with some reports of none 3 to over 90%. 4-8 This discrepancy is primarily because of the methods used to detect and quantify the presence and severity of neuropathy and the variable criteria used to define neuropathy 17,19-22 ; the differences in the severity of liver disease may also be another factor. A consensus panel on peripheral neuropathy in diabetics suggested that to fully and sensitively classify the severity and type of neuropathy, various testing methodologies should include at least one measure from each of the following: electrophysiology, QST, symptom scoring, neurological examination, and autonomic testing. 19,20 Our study uses all of these modalities to evaluate peripheral neuropathy in chronic liver disease patients. As reported in the literature, most patients were either asymptomatic or had minor symptoms. 4-7,23 Only 1 patient, who had Guillain- Barré syndrome, had significant weakness. In the others, the neurological findings were distal in a stocking-glove distribution, as one would expect for an axonal neuropathy. This can be inferred from previous reports as well, although descriptions of this have not been detailed. 4-7,23 Our detailed electrophysiology study confirms that most patients have a length-dependent axonal neuropathy. This is contradictory to the conclusions drawn in the previous electrophysiology reports in which the neuropathy was called demyelinating based primarily on reduced conduction velocity in the median, ulnar, and peroneal nerves. 5-8 The reported velocities were rarely in the demyelinating range and can be secondary to large fiber axonal loss. Furthermore, in these studies, entrapment neuropathies were not excluded as possible explanations for the reduced conduction velocities. We showed that carpal tunnel entrapment of the median nerve can occur in up to one third of these patients. Previous reports also concluded that neuropathy was demyelinating based on sural nerve biopsy specimens that showed thinly myelinated fibers and short internodes. 4-6,23 However, because no active demyelination or inflammatory cells were reported, thinly myelinated fibers and small internodes may have been a reflection of axonal degeneration followed by regeneration. 24,25 It has been shown by Dyck et al. 26 that nonrandom demyelination may take place as an early phase

5 1702 CHAUDHRY ET AL. HEPATOLOGY June 1999 of axonal degeneration. Because this study did not include nerve biopsies, we cannot conclusively reconcile with this previous observation. However, the clinical presentation, neurological findings, and electrophysiology studies seen in this study were those of a dying-back axonal neuropathy. 27 Because some of the systemic illnesses that cause liver dysfunction also are independent causes of peripheral nerve dysfunction, a cause and effect relationship between liver disease and neuropathy has been questioned. 1,2 Alcoholinduced cirrhosis and neuropathy are the prime examples of this with porphyria, polyarteritis nodosa, certain intoxications, primary biliary cirrhosis, and amyloidosis being other examples. 1,2,28-30 In this study, neuropathy was seen irrespective of the cause of liver disease, and there was a significant correlation of the severity of neuropathy to the severity of liver disease. These observations suggest that metabolic dysfunction caused by the liver disease rather than the etiology of liver disease is the primary determinant of polyneuropathy. Although diabetes and alcohol use can independently cause a neuropathy, most patients did not have these confounding causes of neuropathy and yet developed significant neuropathy. The most compelling argument for hepatic dysfunction causing polyneuropathy is from the cryptogenic group in which neuropathy was most prevalent. There is an increase in literature on hepatitis B and C causing an independent neuropathy. Hepatitis C, especially when associated with cryoglobulinemia, can cause a neuropathy. However, patients usually present with a fulminating vasculitis and mononeuropathy multiplex syndrome, 31,32 neither of which was seen in this study. Furthermore, none of the hepatitis C patients described in this study had clinically significant cryoglobulinemia or vasculitis. In this study, the prevalence of neuropathy in patients with chronic hepatitis C virus was similar to other groups. In one study, portosystemic shunting was thought to be one of two important factors in the genesis of hepatic neuropathy, the other being heptocellular damage. 23 However, both Kardel et al. and Chari et al. 5 found no differences in the patients with cirrhosis who had or had not undergone portocaval shunt. Furthermore, an experimental study of portocaval anastomosis in rats also favored hepatocellular failure as the principal pathophysiological mechanism in hepatic neuropathy. 33 Guillain-Barré syndrome and chronic demyelinating neuropathies have been reported to occur in the setting of viral hepatitis. 1,2,34-38 One of the patients studied presented with Guillain-Barré syndrome in the setting of sclerosing cholangitis. The clinical and electrophysiological findings in this patient were typical for an acute acquired demyelinating neuropathy, which is known to have an immune basis. Therefore, an autoimmune mechanism rather than hepatocellular damage is more likely to explain this patient s disease. In patients with primary biliary cirrhosis, a distinct type of sensory neuropathy has been reported. 39,40 In agreement with these reports, three of the five patients with primary biliary cirrhosis in our study had abnormalities only detectable by QST. Only one patient was symptomatic with paresthesias and was defined to have a neuropathy. One patient showed evidence of median neuropathy at the wrist. Recent studies have addressed the relationship between autonomic neuropathy and chronic liver disease We found frequent abnormalities of heart rate variation with deep breathing and with the Valsalva maneuver suggesting the presence of autonomic neuropathy with predominant parasympathetic dysfunction. Patients did not show orthostatic decreases in blood pressure, suggesting relatively intact sympathetic function. This is consistent with previous observations by Thuluvath and Triger 9 and Kempler et al., 12 who also determined that autonomic function was abnormal both in alcohol and nonalcohol categories of chronic liver disease. 9,12,13,41,42 Autonomic neuropathy appears to be part and parcel of a generalized sensory-motor polyneuropathy; the majority of patients with autonomic dysfunction had evidence of a somatic neuropathy (26 of 29). This has not been emphasized in previous reports. Lastly, like previous reports, we also found that the prevalence and severity of autonomic dysfunction was related to the severity of hepatic dysfunction and was independent of cause of liver disease. 14 We conclude that a length-dependent sensory-motor neuropathy and an autonomic neuropathy occur frequently in patients with end-stage hepatic disease. The neuropathy tends to correlate with the severity of liver disease suggesting that liver dysfunction is the primary cause of neuropathy. The effect of peripheral and autonomic neuropathy on the longterm prognosis and on predicting survival or predicting likelihood of major neurological impairment after liver transplant and its potential reversibility after liver transplant need to be addressed in future studies. REFERENCES 1. Asbury A. Hepatic neuropathy. In: Dyck PJ, Thomas PK, Lambert EH, eds. Peripheral neuropathy. 2nd ed. Philadelphia: Saunders, 1984; Asbury A. Neuropathies with renal failure, hepatic disorders, chronic renal insufficiency, and critical illness. In: Dyck PJ, Thomas PK, Griffin JW, Low PA, Poduslo J, eds. Peripheral Neuropathy. 3rd ed. Philadelphia: Saunders, 1993; Thomas PK. Metabolic neuropathies. In: Aguayo AJ, Karpati G, eds. Current Topics in Nerve and Muscle Research. Amsterdam: Excerpta Medica, 1979: Dayan AD, Williams R. Demyelinating peripheral neuropathy and liver disease. Lancet 1967;2: Chari VR, Katiyar BC, Rastogi BL, Bhattacharya SK. Neuropathy in hepatic disorders. A clinical, electrophysiological and histopathological appraisal. J Neurol Sci 1977;31: Knill-Jones RP, Goodwill CJ, Dayan AD, Williams R. Peripheral neuropathy in chronic liver disease: clinical, electrodiagnostic, and nerve biopsy findings. J Neurol Neurosurg Psych 1972;35: Kardel T, Nielsen VK. Hepatic neuropathy. A clinical and electrophysiological study. Acta Neurol Scand 1974;50: Seneviratne KN, Peiris OA. Peripheral nerve function in chronic liver disease. J Neurol Neurosurg Psych 1970;33: Thuluvath PJ, Triger DR. Autonomic neuropathy and chronic liver disease. Q J Med 1989;72: Trevisani F, Sica G, Bernardi M. Autonomic neuropathy in advanced liver disease. HEPATOLOGY 1996;24: Oliver MI, Miralles R, Rubies-Prat J, Navarro X, Espadaler JM, Sola R, Andreu M. Autonomic dysfunction in patients with nonalcoholic chronic liver disease. J Hepatol 1997;26: Kempler P, Varadi A, Szalay F. Autonomic neuropathy in liver disease. Lancet 1989;2: Johnson RH, Robinson BJ. Mortality in alcoholics with autonomic neuropathy. J Neurol Neurosurg Psych 1988;51: Hendrickse MT, Thuluvath PJ, Triger DR. Natural history of autonomic neuropathy in chronic liver disease [see comments]. Lancet 1992;339: Hendrickse MT, Triger DR. Peripheral and cardiovascular autonomic impairment in chronic liver disease: prevalence and relation to hepatic function. J Hepatol 1992;16: Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60:

6 HEPATOLOGY Vol. 29, No. 6, 1999 CHAUDHRY ET AL Chaudhry V, Eisenberger MA, Sinibaldi VJ, Sheikh K, Griffin JW, Cornblath DR. A prospective study of suramin-induced peripheral neuropathy. Brain 1996;119: Cornblath DR, Chaudhry V, Carter K, Lee D, Seysedadr M, Miemicki K, et al. Total neuropathy score: validation and reliability study. Submitted. 19. Dyck PJ, Karnes JL, O Brien PC, Litchy WJ, Low PA, Melton LJ. The Rochester Diabetic Neuropathy Study: reassessment of tests and criteria for diagnosis and staged severity. Neurology 1992;42: Dyck PJ. Detection, characterization, and staging of polyneuropathy: assessed in diabetics. Muscle Nerve 1988;11: Dyck PJ, Kratz KM, Lehman KA, Karnes JL, Melton LJ, O Brien PC, Litchy WJ, et al. The Rochester Diabetic Neuropathy Study: design, criteria for types of neuropathy, selection bias, and reproducibility of neuropathic tests. Neurology 1991;41: Feldman EL, Brown MB, Stevens MJ, Canal N, Thomas PK, Greene DA. A practical two-step quantitative clinical and electrophysiological assessment for the diagnosis and staging of diabetic neuropathy. Diabetes Care 1994;17: Chopra JS, Samanta AK, Murthy JM, Sawhney BB, Datta DV. Role of porta systemic shunt and hepatocellular damage in the genesis of hepatic neuropathy. Clin Neurol Neurosurg. 1980;82: Midroni G, Bilbao JM. Examination of the peripheral nerve biopsy. In: Midroni G, Bilbao JM, eds. Biopsy Diagnosis of Peripheral Neuropathy. Newton: Butterworth Heinemann, 1995; Fullerton PM, Gilliatt RW, Lascelles RG, Morgan-Hughes JA. The relation between fiber diameter and internodal length in chronic neuropathy. J Physiol 1965;178:26P-28P. 26. Dyck PJ, Johnson WJ, Lambert EH, O Brien PC. Segmental demyelination secondary to axonal degeneration in uremic neuropathy. Mayo Clin Proc 1971;46: Spencer PS, Schaumburg HH. Central-peripheral distal axonopathy the pathogenesis of dying-back polyneuropathies. In: Zimmerman H, ed. Progress in Neuropathology. Volume 3. New York: Grune and Stratton, 1976; Thorner PS, Bilbao JM, Sima AA, Briggs S. Porphyric neuropathy: an ultrastructural and quantitative case study. Can J Neurol Sci 1981;8: Becker DM, Kramer S. The neurological manifestations of porphyria: a review. Medicine 1977;56: Barohn RJ, Sanchez JA, Anderson KE. Acute peripheral neuropathy due to hereditary coproporphyria. Muscle Nerve 1994;17: Apartis E, Leger JM, Musset L, Gugenheim M, Cacoub P, Lyon-Caen O, Pierrot-Deseilligny C, et al. Peripheral neuropathy associated with essential mixed cryoglobulinemia: a role for hepatitis C virus infection? J Neurol Neurosurg Psychiatry 1996;60: David WS, Peine C, Schlesinger P, Smith SA. Nonsystemic vasculitic mononeuropathy multiplex, cryoglobulinemia, and hepatitis C. Muscle Nerve 1996;19: Hindfelt B, Holmin T. Experimental porta-caval anastomosis and motor nerve conduction velocity in the rat. J Neurol 1980;223: Berger JR, Ayyar R, Sheremata WA. Guillain-Barré syndrome complicating acute hepatitis B. A case with detailed electrophysiological and immunological studies. Arch Neurol 1981;38: Niermeijer P, Gips CH. Guillain-Barré syndrome in acute HBs Agpositive hepatitis. Br Med J 1975;4: Tsukada N, Koh CS, Inoue A, Yanagisawa N. Demyelinating neuropathy associated with hepatitis B virus infection. Detection of immune complexes composed of hepatitis B virus surface antigen. J Neurol Sci 1987;77: Tsukada N, Koh CS, Owa M,Yanagisawa N. Chronic neuropathy associated with immune complexes of hepatitis B virus. J Neurol Sci 1983;61: Inoue A, Tsukada N, Koh CS, Yanagisawa N. Chronic relapsing demyelinating polyneuropathy associated with hepatitis B infection. Neurology 1987;37: Charron L, Peyronnard JM, Marchand L. Sensory neuropathy associated with primary biliary cirrhosis. Histologic and morphometric studies. Arch Neurol 1980;37: Illa I, Graus F, Ferrer I, Enriquez J. Sensory neuropathy as the initial manifestation of primary biliary cirrhosis. J Neurol Neurosurg Psychiatry 1989;52: Duncan G, Johnson RH, Lambie DG, Whiteside EA. Evidence of vagal neuropathy in chronic alcoholics. Lancet 1980;2: Novak DJ, Victor M. The vagus and sympathetic nerves in alcoholic polyneuropathy. Arch Neurol 1974;30: