Chronic and acute (arrows) loss of myelinated axons, 1-µm plastic section. Bar indicates 10 µm (methylene blue, original magnification ×390).
Loss of primarily unmyelinated axons with clusters of remaining Schwann cell processes (arrows). Bar indicates 1 µm (transmission electron microscopy, original magnification ×6475).
Customize your JAMA Network experience by selecting one or more topics from the list below.
Wöhrle JC, Spengos K, Steinke W, Goebel HH, Hennerici M. Alcohol-Related Acute Axonal Polyneuropathy: A Differential Diagnosis of Guillain-Barré Syndrome. Arch Neurol. 1998;55(10):1329–1334. doi:10.1001/archneur.55.10.1329
Chronic axonal polyneuropathy is a well-known clinical sequela of excessive alcohol consumption; however, acute axonal polyneuropathy related to alcohol abuse is less well recognized.
To describe alcohol-related acute axonal polyneuropathy in 5 chronic alcoholics who developed ascending flaccid tetraparesis and areflexia within 14 days.
Case series with clinical, laboratory, electrophysiological, and, in 1 patient, biopsy data.
All 5 patients consumed a daily average of 250 g of alcohol, and 4 had lost a substantial amount of weight recently. Additional clinical features included painful paresthesia, myalgia, and glove and stocking–type sensory loss. Repeated cerebrospinal fluid examinations failed to show the marked increase of protein concentration with normal cell count typical of Guillain-Barré syndrome, although the protein level was mildly elevated in 1 patient. Blood laboratory findings were consistent with long-standing alcohol abuse. Compound muscle and sensory nerve action potentials were absent or reduced, while conduction velocities were normal or mildly reduced. Three to 4 weeks after onset, needle electromyography displayed moderate to severe fibrillations and positive sharp waves in addition to normal motor unit potentials, indicating an acute axonal polyneuropathy; this was confirmed by sural nerve biopsy in 1 patient.
Excluding other factors, we assume that in these patients the combination of alcohol abuse and malnutrition caused severe acute axonal polyneuropathy. Its distinction from Guillain-Barré syndrome is important because treatment requires balanced diet, vitamin supplementation, and abstinence from alcohol, while immunotherapy may not be indicated.
IN CONTRAST to the chronic distal symmetrical neuropathy with initially insidious development of sensory symptoms, acute courses of polyneuropathy have been infrequently described in patients with chronic alcohol abuse. Chronic alcoholic neuropathy involves sensory more than motor nerve fibers, and electrophysiological and histological studies demonstrated axonal degeneration as the primary pathophysiological mechanism.1 Case series including more acutely developing alcoholic neuropathy reported intervals from the onset to the nadir of symptoms from a few days to "several weeks" and described axonal involvement on electrophysiological and histological examinations.2-4 Some of these patients manifested an ascending symmetrical paresis and areflexia, and thus partly fulfilled clinical criteria of the Guillain-Barré syndrome (GBS).5 "Classic" GBS is caused by peripheral nerve demyelination,6,7 but an axonal form of GBS was recently recognized.8,9 One report of GBS in chronic alcoholism did not give clear electrophysiological evidence of a demyelinating or axonal neuropathy.10
During a 3-year study, we observed 5 patients with a history of chronic alcohol abuse who, within 14 days, developed a severe ascending flaccid tetraparesis with areflexia, sensory loss in a glove and stocking–type distribution, painful paresthesia, and myalgia. They were all referred under the provisional diagnosis of acute GBS. Apart from clinical and laboratory data, we report the results of nerve conduction studies (NCSs), electromyography (EMG), and, in 1 patient, findings of sural nerve biopsy, indicating acute axonal polyneuropathy in all patients.
Between February 1, 1992, and January 31, 1995, 5 patients with chronic alcoholism (mean age, 37 years) were referred to the University Department of Neurology, Klinikum Mannheim of the University of Heidelberg, Mannheim, Germany, with the provisional diagnosis of acute GBS. Within a maximum of 14 days, they developed severe ascending flaccid tetraparesis with areflexia. Their average daily alcohol intake had been 250 g for more than 3 years. Four patients reported substantial loss of weight during the preceding 3 months. Medical history showed previous alcohol-related disease in 4 patients; however, preexisting symptoms or signs of peripheral neuropathy were denied by all patients and their referring physicians. A preceding infection was reported only by patient 4, who had suffered from an upper respiratory tract infection 2 weeks before the onset of neurological symptoms (Table 1).
All patients underwent neurological examination, and paresis was graded by means of a summary score of the British Medical Research Council grades: upper limbs (0-30): sum of Medical Research Council grades for right and left arm abduction, forearm flexion, and wrist extension; lower limbs (0-30): sum of Medical Research Council grades for right and left hip flexion, knee extension, and foot dorsiflexion. Additionally, routine blood tests, including creatine kinase; thyroid hormones; δ-aminolevulinic acid; serum electrophoresis and immunoelectrophoresis; vitamin B1, B6, B12, and folate levels; and antiganglioside GM1 antibodies, were done. Cerebrospinal fluid (CSF) was analyzed, including protein and immunoglobulin electrophoresis and oligoclonal bands, on admission and after 10 to 14 days. Motor NCSs of median and peroneal nerves, sensory NCSs of median and sural nerves, and needle EMG were repeatedly performed (Dantec counterpoint EMG device; Dantec Medical A/S, Skovlunde, Denmark), according to standard techniques.11 In 1 patient, we performed a biopsy of a sural nerve.
Four patients had symmetrical tetraparesis (patients 1, 3, 4, and 5) and 1 patient had severe paraparesis (patient 2). They had all reached their maximum of disability by 14 days or less after onset of first neurological symptoms. They all were bedridden at that time. Muscle tone was flaccid, and deep tendon reflexes were either absent or markedly depressed in involved limbs. Multimodal sensory loss of glove and stocking–type distribution was accompanied by painful paresthesia or myalgia. Limb or gait ataxia was prominent in 3 patients. Ataxia was of sensory type, and 1 patient additionally showed features of cerebellar ataxia. No patient displayed cranial nerve involvement, urinary or fecal incontinence, or central nervous system signs. No patient required assisted mechanical ventilation. Autonomic disturbances were present in all patients, consisting of hyperhidrosis or tachycardia. Tachycardia resolved within 2 to 3 weeks after admission; however, hyperhidrosis persisted throughout the period of observation. General physical examinations and investigations disclosed acne or spider nevi in 3, liver cirrhosis in 3, fatty liver in 2, splenomegaly in 3, and esophageal varicosis in 2 patients (Table 2).
Impaired liver function was documented by an increased γ-glutamyltransferase level in all patients, and hypoproteinemia (total protein, fibrinogen) or coagulation deficit (international normalized ratio) in 4 patients. Macrocytic anemia was present in 4 patients, although only 2 patients displayed vitamin deficiency, 1 being deficient in vitamin B6 and 1 other in folate. In all patients, creatine kinase levels, glucose profiles, results of thyroid function tests, and δ-aminolevulinic acid levels were normal; monoclonal bands on electrophoresis and immunoelectrophoresis and anti-GM1 antibodies were negative. All patients had a normal CSF cell count, normal or slightly elevated CSF protein level, normal CSF–serum albumin quotient, and neither oligoclonal bands nor autochthonous IgG production on repeated examinations (Table 3).
In every patient, repeated motor and sensory NCSs showed the most pathological results during week 3 to 6 after onset. Compound muscle action potentials of peroneal or median nerves were absent, of low amplitude, or accompanied by mildly reduced nerve conduction velocity in 4 patients (Table 4). Distal latencies were normal or mildly prolonged. The distal latency of a very low compound muscle action potential of 1 peroneal nerve could have been suggestive of some degree of demyelination only in patient 2. F-wave latencies were normal in 6 of 10 motor nerves, mildly prolonged in 2 (1 median and 1 peroneal), and absent in 2 peroneal nerves. All sensory nerves tested showed abnormal results of NCSs, with reduced amplitudes or absent sensory nerve action potential, 2 to 4 weeks after onset. Sensory nerve conduction velocities were reduced or normal. During week 3 to 4 after onset, EMG evidence of acute denervation had already occurred in 4 patients in the form of fibrillations and positive sharp waves. After more than 6 weeks, patient 3 also developed pathological spontaneous activity indicative of acute denervation, and later all patients who could be reexamined had EMG signs of reinnervation. This pattern of neurographic and EMG abnormalities was indicative of acute axonal polyneuropathy.
Sural nerve biopsy was performed 47 days after onset in patient 3 (Figure 1 and Figure 2). Considerable loss of myelinated axons of large and small calibers and wallerian degeneration of large myelinated axons, but no further disease-specific features, were found on light microscopy. Additionally, electron microscopy disclosed several sets of Schwann cell processes without axons after loss of primarily unmyelinated nerve fibers. There were no signs of primary demyelination or inflammation.
All patients received oral supplementation of vitamins (thiamine hydrochloride, pyridoxine hydrochloride, cyanocobalamin, and folate); despite normal vitamin B12 serum levels, patient 3 also received intramuscular cyanocobalamin injections because of his history of gastrectomy. Immunotherapy was given only in patient 4, who received 1 course of plasma exchange, and in patient 5, who received 1 course of intravenous IgG. In both patients, we learned the true extent of alcohol abuse only some days after admission. Daily physiotherapy was administered and continued after discharge to rehabilitation hospitals. Within months, all patients regained the ability to walk, while living abstinent from alcohol. The pain syndromes lasted between 3 and 7 weeks. Patients 1, 3, and 5 reached a functional classification score of 2 after 6, 11, and 1.5 months, respectively; patient 2 reached a score of 3 after 4.5 months; and patient 4 reached a score of 1 after 4.5 months. Unrelated to neuropathy, patient 4 died of esophageal hemorrhage after 15 months.
Our 5 patients with a long-standing history of severe alcohol abuse developed an ascending, flaccid tetraparesis and areflexia within a maximum of 14 days, thus apparently fulfilling major clinical criteria of GBS.5 In addition, painful paresthesia and myalgia and severe multimodal sensory loss were prominent clinical features of all patients, associated with sensory ataxia in some cases. Other alcohol-related illnesses were present, indicating the severity of alcohol abuse. Autonomic disturbances on admission most likely represented physical signs of alcohol withdrawal. There were no clinical suggestions of preexisting peripheral neuropathy. Routine blood investigations documented chronic liver damage but no further cause of acute neuropathy. Macrocytic red blood cells were a frequent finding of chronic alcohol intoxication; however, deficiency in folic acid was present in only 1 patient. One other patient was deficient in vitamin B6, and no patient was deficient in vitamin B1 or B12. Studies of CSF showed a normal or only mildly elevated (patient 5) protein level in all patients. Serial NCSs and EMG clearly indicated an acute axonal neuropathy as the cause of weakness in all patients. Acute axonal degeneration but also chronic nerve changes were confirmed by sural nerve biopsy in 1 patient. Treatment consisted of a balanced diet, vitamin supplementation, and physiotherapy; in 2 patients, 1 course of immunotherapy was given before the full extent of alcohol abuse was recognized. All patients regained their ability to walk within months of abstinence from alcohol.
In a large series of 1030 consecutive patients with chronic alcohol abuse in a city hospital, Victor12 reported that 9% of patients suffered from peripheral neuropathy. In more than half of the patients, the neuropathies were subclinical, and in the rest it was mostly insidious in onset, although some patients had a more rapid onset. Detailed clinical descriptions of acute courses of peripheral neuropathy in alcoholic patients are few.2,4,10 Walsh and McLeod2 reported on 7 patients with acute polyneuropathy progressing over several weeks with predominant sensory symptoms and signs. Only 1 patient showed flaccid tetraparesis. They all showed severe axonal damage on electrophysiological examination and involvement of large and small fibers on nerve biopsy. Tabaraud et al4 collected a series of 8 patients over 10 years who had developed severe tetraparesis and severe sensory impairment within a few days to 3 weeks, similar to our patients. Electrophysiological and histological evidence pointed toward an axonal polyneuropathy; the CSF was described as normal. Of the 6 alcoholic patients who were reported by Pastena et al10 to suffer from acute GBS, the electrophysiological results given unfortunately did not allow differentiation between acute axonal polyneuropathy and acute GBS.
The present group of patients had several features distinct from GBS. Severe objective sensory loss is rather uncommon in GBS.6 Power of respiratory muscles was preserved and cranial nerve function remained unaffected, unlike the findings in many patients with GBS with severe tetraparesis.6 In contrast to GBS, where CSF protein level is abnormal in 80%,13 the present group of patients had normal or only slightly elevated CSF protein levels and did not develop the typically marked increase of protein concentration with normal cell count of GBS. Early in the course, NCSs and EMG clearly indicated an acute axonal neuropathy as the cause of weakness in all patients. The typical findings of GBS with conduction block at distal or proximal sites and sites of entrapment, or length-dependent block,7 were not encountered. F-waves were affected only in nerves with a markedly reduced compound muscle action potential. Early in the course, ie, 10 to 18 days after onset of symptoms, sensory nerve action potentials could not be elicited or were small and with low nerve conduction velocity in 3 patients who could be examined at that time. By week 4, EMG showed moderate or severe pathological spontaneous activity in 4 of 5 patients. In GBS, sensory conduction does not become abnormal before week 3 or 4; at most, only mild abnormal spontaneous activity can be observed by week 4 to 5.14
We performed a biopsy of the sural nerve in only 1 patient. The loss of large and small myelinated fibers and of unmyelinated fibers and the wallerian degeneration were in keeping with other reports of alcoholic axonal polyneuropathy.2,15 These changes indicated acute axonal degeneration, apart from chronic subclinical nerve degeneration of indeterminable duration. Primary demyelination or inflammation was not encountered. Although the timing of our biopsy was as late as 47 days after onset, at least some degree of demyelination would have been expected to be present together with macrophage invasion and axonal degeneration if the primary abnormality had been inflammatory demyelination.16
In 1986, Feasby et al8 described an acute axonal form of GBS in 5 patients in whom inexcitable motor nerves were the electrophysiological hallmark. Clinically, these patients had very rapidly (within 8 days) progressed to a quadriplegic state with bulbar involvement in 4 patients and mechanical ventilation in 3 patients. They mostly showed the marked increase of CSF protein concentration with normal cell count typical of GBS. In a subsequent series of 4 patients with very severe GBS, histological studies of nerve biopsy specimens after 15 to 29 days confirmed axonal degeneration in 3 patients and primary demyelination in 1 patient.9 Our patients did not deteriorate to such a severe clinical state; apart from 1 patient with a slightly increased CSF protein level, they had normal protein concentrations in CSF; therefore, they seem to differ from patients with axonal GBS. Epidemic axonal GBS of northern China17 also seems to be different from the neuropathy of our cases because of its lack of sensory involvement and very severe clinical course, often leading to respiratory failure. However, the existence of a mild form of axonal GBS remains questionable9 and cannot be totally excluded in our patients, as no single laboratory variable can positively confirm this diagnosis. An association of Campylobacter jejuni infection with GBS and subsequent axonal degeneration has increasingly been recognized and was recently confirmed in a prospective case-control study.18 Evidence of molecular mimicry for GM1 ganglioside epitopes and Campylobacter bacteria has been brought forward by studies of serum samples from patients with GBS.19 Our patients were all negative for anti-GM1 antibodies, and they had no history of preceding diarrhea; however, Campylobacter antibody titers were not determined.
The pathophysiological mechanisms of polyneuropathy in patients with chronic alcohol abuse are still a matter of debate. Animal studies gave controversial evidence for a direct toxic effect of alcohol.1 Poor correlations between vitamin levels, the nutritional state, and polyneuropathy of alcoholic patients are known, and when sural nerve histological findings were compared, axonal degeneration was more severe in malnourished alcoholic patients than malnourished patients who had undergone gastrectomy, thus implying a direct toxic effect of alcohol.15 A dose-dependent toxic effect of alcohol was statistically described in a case-control study on autonomic and peripheral neuropathy.20 Nutritional imbalance was present in 4 of our 5 patients who had lost 10 to 15 kg of weight during the preceding 3 months; 2 of them had low serum vitamin levels. All patients had consumed large amounts of alcohol over years, and it could not be determined with certainty whether they had additionally increased their intake and thus exerted an additional toxic effect before the onset of acute polyneuropathic symptoms. The preexistence of subclinical chronic polyneuropathy, presumably caused by alcoholism, cannot methodologically be ruled out in our patients and seems likely for the patient who underwent sural nerve biopsy. However, the acute GBS-like syndrome has to be considered separately from such a subclinical polyneuropathy. For want of other evidence, we assume that for the clinical syndrome of acute axonal polyneuropathy to develop, the combination of toxic effects of alcohol and nutritional deficiency had reached a critical threshold.
The clinical, electrophysiological, and laboratory data of our patients allowed the diagnosis of alcohol-related acute axonal polyneuropathy. Its distinction from "classic" demyelinating GBS or axonal GBS is clinically important because of the need for different management, including a balanced diet, vitamin supplementation, and abstinence from alcohol, while specific immunotherapy, including plasmapheresis, is probably unnecessary.
Accepted for publication January 28, 1998.
Reprints: Johannes C. Wöhrle, MD, Department of Neurology, Klinikum Mannheim of the University of Heidelberg, D-68135 Mannheim, Germany (e-mail: firstname.lastname@example.org).