Patient 7. Photomicrographs of representative transverse sections of the spinal cord. A, Almost complete loss of motor neurons in the anterior horn of the thoracic segment (Klüwer-Barrera, original magnification ×20). B, Only a few dark motor neurons in the anterior horn of the lumbar segment are visible (hematoxylin-eosin, original magnification ×200).
Niewiadomska M, Kulczycki J, Wochnik-Dyjas D, Szpak GM, Rakowicz M, Łojkowska W, Niedzielska K, Inglot E, Wiȩclawska M, Glazowski C, Tarnowska-Dziduszko E. Impairment of the Peripheral Nervous System in Creutzfeldt-Jakob Disease. Arch Neurol. 2002;59(9):1430-1436. doi:10.1001/archneur.59.9.1430
The clinical manifestations of Creutzfeldt-Jakob disease (CJD) primarily reflect involvement of the central nervous system. The coexistence of CJD with peripheral nervous system involvement has also been reported.
To analyze peripheral neuron electrophysiologic changes and to compare these data with neuropathologic features of spinal motor neurons in patients with definite CJD.
Design and Patients
Electrophysiologic examinations were performed on 16 patients with sporadic CJD. The diagnosis was confirmed by neuropathologic examinations (15 patients) or by intravital detection of the 14-3-3 protein in the cerebrospinal fluid (1 patient). The spinal cord was neuropathologically examined in 8 patients.
Department of Clinical Neurophysiology, I Neurological Department, Institute of Psychiatry and Neurology, Warsaw, Poland.
Main Outcome Measures
Electromyography, compound muscle and sensory nerve action potentials, distal latencies, F waves, peripheral motor and sensory conduction velocity, and spinal motor neuron numbers and morphologic characteristics.
All patients had signs of central nervous system damage typical of sporadic CJD. Only 3 patients had clinical signs of peripheral nervous system involvement. Electrophysiologic examinations confirmed peripheral nervous system damage in these patients and revealed preclinical peripheral nervous system impairment in 11 more patients. In 1 patient, electrophysiologic examination revealed features of motor neuron disease; in 9, axonal disease; and in 4, axonal-demyelinating neuropathy. Neuropathologic examination results confirmed severe loss of spinal motor neurons in 1 patient with motor neuron disease and revealed the features of motor neuron chronic disease in 4. In 2 of them, electrophysiologic data were normal.
In sporadic cases of CJD, peripheral nervous system impairment should be considered to be an integral component of disease.
CREUTZFELDT-JAKOB disease (CJD) is a rare neurodegenerative disease belonging to the transmissible spongiform encephalopathies.1- 3 It is believed that CJD is transmitted by prion proteins (PrP 27-30), pathologically conformed proteins that are a product of partial physiologic proteolysis of cell membrane proteins.2- 4 The types of prion diseases found in humans are sporadic CJD, iatrogenic CJD, familial CJD, Gertsman-Sträussler-Scheinker disease, fatal familial insomnia, as well as kuru.2,3,5 In 1996, a new variant of CJD resulting from bovine spongiform encephalopathy transmitted to humans was described.6,7
Sporadic CJD occurs most frequently.2,3,8 Incidences of sporadic CJD are about 1 per million per year worldwide. About 5%-15% of cases are familial.3,5 The genetic background is characterized by mutations within several codons of the PrP gene on the short arm of chromosome 20.2,5,8
Typically, cases of sporadic CJD begin in the seventh decade of life, but the range of onset is wide (16-82 years).8 The mean duration of the disease is 5 months; however, it may be longer.3,8 Clinical manifestations primarily concern the central nervous system. Fast-progressing dementia is the first and most dominant feature. It is often accompanied by extrapyramidal, cerebellar, and pyramidal signs, myoclonus, cortical blindness, and akinetic mutism.3,8,9 Sporadic epilepsy occurs in certain cases.3,5,9
The cerebrospinal fluid (CSF) is usually normal.3,5,9 Computed tomography and magnetic resonance imaging usually show cerebral atrophy.5,9,10 In the late stage of the disease, electroencephalography shows periodic synchronous triphasic sharp waves against a background of progressively suppressed cortical activity.3,5,11
Neuropathologic identification of CJD depends on 3 factors in postmortem examinations of the brain: spongiform changes, neuronal loss, and reactive astrocytosis.12 It has recently been found that immunoassay for the presence of the 14-3-3 protein in CSF and molecular genetic evaluation may be diagnostically important.4,5,13,14
Certain patients develop muscular atrophy and fasciculations, suggesting amyotrophic lateral sclerosis.15- 17 The coexistence of CJD with axonal or demyelinating sensory-motor neuropathy has been reported.15,18- 21 It is suggested that in some patients, damage of this type may be a preclinical characteristic of the disease,15 and electromyography (EMG) should become a routine procedure in the clinical diagnostics of CJD.22 Bearing this in mind, we decided to analyze the results of routine electrophysiologic examinations of the peripheral nervous system (PNS) of our patients diagnosed as having CJD.
Interest in the epidemiologic problems surrounding CJD within Poland prompted the I Neurological Clinic (Warsaw) Institute of Psychiatry and Neurology to begin admitting patients from throughout Poland who had dementia and possible symptoms of CJD. Sixteen consecutive patients with a definite diagnosis of CJD (11 women and 5 men, aged 20-75 years) were observed from 1984 to 2000. All of these patients had sporadic CJD. In 15 of 16 patients, the CJD diagnosis was confirmed by a neuropathologic examination of the brain. In one case, intravital detection of the 14-3-3 protein in the CSF substantiated the diagnosis of CJD (Creutzfeldt-Jakob Disease Laboratory of the University of Göttingen, Göttingen, Germany).
Each patient was regularly subject to clinical neurologic examinations (same protocol, same neurologist) to document the progress of the disease. In addition to routine analyses of blood and urine, patients underwent electroencephalogram, magnetic resonance imaging, or computed tomography, CSF analysis, EMG, and neurographic tests. The results of these last 2 analyses provided the basis for this article. None of the 16 patients were diagnosed as having diabetes, uremia, hepatopathia, alcoholism, or any other factors causing damage to the PNS.
Four patients were subject to genetic analyses (DNA tests performed post mortem). Other analyses were not performed. Consent to perform tests was obtained from relatives and/or guardians of the patients.
The investigations were carried out according to the commonly applied routine methods23- 28 with the use of Mystro 25 (Medelec Ltd, Old Woking, Surrey, England) and Counterpoint MK II (Dantec Medical A/S, Skovlunde, Denmark). Disposable concentric needles and/or surface silver electrodes were used. Throughout the disease duration, 13 patients were examined once, and the remaining 3 were examined twice.
We attempted to examine 2 proximal muscles (deltoideus and rectus femoris) and 2 distal muscles (abductor pollicis brevis and extensor digitorum brevis) as well as their 4 motor nerves. Within the sensory nervous system, 2 to 4 peripheral nerves were analyzed (usually both sides). The severe clinical state of the patients often caused the examinations to be limited. This is why EMG was quantitatively performed on 7 of 16 patients (24 muscles). Analyses of F waves and their conduction velocity (F-CV) were only performed in 19 nerves in 11 of 16 patients. Nevertheless, all patients were subject to the following:
Resting EMG (29 proximal and 36 distal muscles) was obtained.
Neurographic examinations of motor fibers (nervus [n] axillaris, n femoralis, n medianus, n peroneus) were performed, with evaluations of amplitude and distal latency of compound muscle action potentials [CMAPs] in 59 muscles).
Motor conduction velocity (MCV) was estimated in 34 nerves.
Neurographic examination of sensory fibers (n medianus and/or n ulnaris; n peroneus superficialis and/or n suralis) was performed. The amplitude of sensory nerve action potentials (SNAPs) and sensory conduction velocity (SCV) was evaluated (88 nerves).
The results of the examinations were compared with our own standards26,27 or standards adopted from the literature.23,25,28 The electrophysiologic criteria of demyelination and axonal degeneration presented in our study were taken from Oh.25 The values of MCV, SCV, and F-CV that were less than 60% of the normal means and values of distal latency that were greater than 150% of the normal means were acknowledged as a severe slowing of conduction (result of demyelination). Respectively, the values that were 60% to 80% and 120% to 150% of the normal means were defined as moderate slowing of conduction. This was assumed to be a result of axonal degeneration of fibers of larger diameter.24,25 The hallmark of axonal degeneration is the diminution of the amplitude of CMAPs or SNAPs, normal or moderate slowing of CV, and normal or moderate prolonged distal latency in CMAPs.25 Significant reduction of the amplitude of CMAPs and SNAPs was defined as being less than 40%, a moderate 40% to 55% of normal means.
Indicators of damaged spinal motor neurons were defined by widespread features of motor axonal degeneration,25 with a reduced number of motor unit action potential parameters and denervation in EMG, but with normal SNAPs examinations.25,29
In all of the postmortem examinations, 12 segments of cerebral lobes, basal ganglia, and cerebellum were taken from material fixed in paraffin. In 8 of these patients, spinal cord segments were examined histopathologically on 1 to 3 levels, with particular attention paid to the number and morphologic characteristics of the motor cells of the anterior horns. Paraffin segments were stained using the hematoxylin-eosin, periodic acid–Schiff, and the Klüver-Barrera, Kanzler-Arendt, and Bielschowsky methods. All spinal cord segments were further stained with the 3F4 monoclonal antibody against human PrP.
The number of motor cells of both anterior horns of the spinal cord was evaluated in the cervical, thoracic, and lumbar segments under a magnification of ×200. Based on calculations, the mean number of motor neurons per 1 anteromedial nucleus of the anterior horn on one side of the spinal cord was estimated on each of the 3 analyzed levels.
Table 1 presents the clinical features observed in the patients during electrophysiologic PNS examinations. In most patients, the onset of CJD began in their 70s. In patients 7 and 15, the onset of the disease was at an early age. In 11 of 16 patients, the duration of the disease was short (2-6 months). In 3 others, the duration was 9 to 10.5 months, and in the remaining 2, the maximum duration was 24 months.
All patients already had advanced clinical manifestations of damage to the central nervous system, but only 3 patients showed clinical signs of PNS involvement; these included atrophy of the distal muscles (patients 6 and 14). General muscle atrophy and lack of deep tendon reflexes were found in patient 7. In this patient, at the earlier stages of the disease, reflexes were preserved and were more pronounced on the right side. No signs of muscle atrophy, fasciculations, or sensory disturbances were observed at that time. Nevertheless, slight spastic signs were then recorded. Table 1 also presents the results of tests other than EMG. Four patients examined by DNA tests displayed no signs of mutation on the PrP gene.
The presence of active denervation in the EMG (positive sharp waves, fibrillations) was found in 7 patients. This occurred more often in distal muscles than in proximal ones (17% and 25% of the muscles examined). In 3 patients (6, 7, and 14), denervation was general and in 4 patients (1, 5, 8, and 11), individual muscles were affected by denervation. In 3 patients (5, 6, and 14), denervation was accompanied by fasciculations.
Table 2 presents neurographic data. The diminution of amplitude and the prolongation of the distal latency CMAPs occurred more often and were more advanced than the generally moderate slowing of MCV. The diminution of the amplitude also dominated the parameters of SNAPs. Lack of nerve excitability was diagnosed in 1 peroneal nerve (patient 15). We ascertained a lack of SNAPs in single nerves in 4 patients (1, 5, 11, and 14). There was a severe slowing of MCV or SCV observed only in single nerves. Significant prolongation of distal latency and significant reduction of CMAP amplitude occurred more often in peroneal nerves than in median nerves. None of the patients showed signs of conduction block.
Electromyographic characteristics of reinnervation (prolongation of the mean duration and increase of the amplitude and percentage of multiphasic MUAPs) were found in only 3 muscles (12%) in 3 patients (2, 4, and 8) of 7 examined (quantitative EMG).
In only 1 patient (11, n peroneal) was there moderate slowing of F-CV together with the normal peripheral MCV, which may suggest root lesion.
Table 3 presents the final conclusions drawn from the electrophysiologic examinations of the PNS in each patient. The state of the PNS in 2 of the examined patients (12%) was normal. Preclinical damage to the PNS was found in 11 patients (69%). Findings from electrophysiologic testing on 3 patients (6,7, and 14) (19%) with clinical muscle atrophy confirmed damage to the PNS. In 5 patients, the damage was general (polyneuropathy). In 6 others, the distribution of changes was varied (mononeuropathy multiplex), and in 2 patients (10 and 15), changes were found only on single nerves. Damage to the PNS swayed from a severe state with active denervation to a mild state, although the damage to the central nervous system during the examination was significantly advanced and similar among all of the patients (Table 1). It is important to note that in patients examined twice (same muscles and nerves) (patients 6 and 8), the development of PNS changes was observed during longer durations of disease. Moderate changes were found within shorter durations of disease (Table 3) and on earlier examinations of the PNS.
One patient (patient 7) (6%) revealed features of motor neuron disease. Nine patients (56%) showed clear electrophysiologic signs of axonal neuropathy. Severe prolonged distal latency and/or severe slowing of CV in some nerves allowed us to conclude axonal demyelinating neuropathy in 4 patients (25%).
In all 15 patients, examination of the brain revealed images typical of spongiform encephalopathy: changes of the character of the spongy state (spongiosis) in the cerebral cortex and in the area of the basal nuclei, neuron atrophy, and proliferation and hyperplasia of the astrocyte glia. Cortical changes were of different location and intensity. In 6 patients, spongiform changes were present in the cortex of all cerebral lobes; in 7 patients, this was particularly evident in the occipital lobe, while in 2, it was evident in the frontal and parietal lobes. Particularly marked spongiform cortical changes were disclosed in 7 patients. The changes were particularly intense in 6 patients, while in 2 patients, the intensity of these changes was mild. Spongiosis was accompanied by pronounced astrocyte cortical gliosis, which in 5 patients was particularly evident in the white matter. None of the patients showed signs of kuru plaques in the cerebellum. The neuropathologic picture allowed us to confirm the clinical suspicion of the sporadic form of CJD in all of the patients studied.
In 1 patient (7, with selective damage to the motor neuron on EMG), a spinal cord examination revealed a lowering of the number of anterior horn neurons, especially in the anteromedial nuclei, to 2 to 3 remaining neurons on the average and to 4.7 in the cervical segment and 8.2 in the lumbar segment (Table 4 and Figure 1). A considerable portion of the preserved neurons showed features of chronic damage with very scant glial reaction. The patient also had marked changes in the pyramidal tracts of the spinal cord in the form of spongiform degeneration and disappearance of myelin from the lateral and partly anterior columns of the spinal cord as well as in the higher segments of the pyramidal tracts in the pons.
In the remaining patients, the mean number of motor neurons found in individual segments of the spinal cord was within normal limits. However, the morphologic characteristics of motor neurons was not normal in 4 patients (Table 4). Shrinkage and pyknotic appearance of cells (chronic disease) were observed. These changes were found in patients with signs of damage to peripheral neurons (patients 4 and 11) as well as in patients in whom the results of electrophysiologic examination were within normal limits (patients 9 and 12). In patient 9, numerous amyloid bodies were revealed in the spinal cord and the investigation with the 2F4 monoclonal antibody disclosed a substantial number of amorphous PrP deposits (more rarely resembling in their shape kuru plaques). In some patients (1, 3, and 5), the appearance and number of motor cells in anterior horns were within normal limits despite the electrophysiologic changes found.
The literature on CJD usually discusses the coexistence of PNS involvement with typical signs of central nervous system damage. Most often, there is discussion of amyotrophy, an objective sign of lower motor neuron lesions8,9,16,17,22 In cases of sporadic CJD, symptoms of amyotrophy appear more often in the later (12% of patients) than in the earlier stages of the disease (0.5% of patients).8 The progression of changes during the time of the disease was confirmed electrophysiologically in our patients who were examined twice. Amyotrophy has been noted in familial CJD16,22 but not in the new variant of CJD.7,30 Our study revealed clinical signs of muscular atrophy in 3 (19%) of 16 patients.
Clinical examinations of the sensory nervous system are limited because of problems associated with the patients' dementia. Objective data can be established only through electrophysiologic examinations, which may detect preclinical dysfunction of the PNS.25 We found preclinical changes in 11 of 16 patients. Thus, PNS involvement was electrophysiologically confirmed in 14 (88%) of 16 patients.
Neuropathologic examinations of anterior horn cells and of spinal ganglia are, unfortunately, rarely performed. On observation, some results were normal,1 whereas others revealed a loss of and degenerative changes to anterior horn cells.9,17,18,22,31
Our material analyses of the anterior horn cells in 3 of 8 patients showed normal results; in 2 of these patients, however, electrophysiologic changes were severe, and in 1, slight. Anterior horn cells were analyzed on 1 (2 patients) or 2 levels (1 patient). The electrophysiologic estimation was composed of 4 levels of spinal innervations. The neuropathologic examinations may not have been preferable to the level of motor neuron pools analyzed electrophysiologically.
Features of chronic motor neuron disease were found in 5 of 8 patients. In one of them, we additionally found a PrP deposit in the spinal cord. In 2 patients, electrophysiologic testing showed normal results. Electromyography and neurography were performed just before the patients' death; however, the disease lasted only 2 to 3 months. It is probable that we would have witnessed further development of PNS deterioration if the duration of the disease had been longer.9
A significant loss of anterior horn cells was discovered in 1 of 8 patients (beside chronic damage of spinal motor neurons). Here, the electrophysiologic result indicated a motor neuron disease. This explained the muscle atrophy as well as the lack of reflexes in the patient. Moreover, changes were found in the pyramidal tracts of the spinal cord and in the higher levels of the brain, asserting the slight spastic signs observed in the earlier stages of the disease.
In a careful study on a CJD case with polyneuropathy, Antoine et al18 found PrP deposits in the brain but did not find any in the spinal cord. They noted some vacuoles in spinal gray matter. They were able to present varying demyelination and remyelination in teased fiber preparations from L-5 ventral and dorsal roots. Moreover, in the roots, the sciatic nerve, and the proximal part of the superficial peroneal nerve, 8% to 12% of fibers underwent valerian degeneration; 27% were in the distal part of this nerve. This seems to suggest a dying-back process. The axonal degeneration of the dying-back type begins in the most peripheral ramification of axons, but it is believed that the body of motor neurons and neurons of sensory ganglia may be the original site of damage.24,32 Electrophysiologic characteristics compatible to axonal degeneration, including a motor neuron lesion, prevailed in our study (63%). Axonal polyneuropathy was described in sporadic CJD31,32 as well as in cases with point mutation in codon 200 of the PrP gene.19
The demyelinating damage to the PNS can be found in sporadic21 and familial CJD15 as well as in cases with revealed mutation of the PrP gene.18 These cases are cited as "atypical." Conclusions from the electrophysiologic examinations in 4 (25%) of our patients allowed us to identify axonal demyelinating neuropathy. Electrophysiologic features of demyelination were found most often in the peroneal nerve and demyelination was always accompanied by significant reduction of the amplitude of CMAPs. This suggests a secondary demyelination in relation to axonopathy of the dying-back type, which generally manifests itself in the longest peroneal nerve.24,33 The prevalence of muscle denervation rather than reinnervation signs on EMG seems to additionally confirm the presence of a progressing degenerative process underlying the CJD. From various studies published between 1968 and 2000, Worral et al22 gathered and analyzed 50 cases of sporadic and familial CJD with amyotrophy and fasciculations, partially confirmed by EMG testing. These authors assert that every case of CJD should be documented by accurate clinical, EMG, and postmortem spinal cord examinations. Our perceptions are the same.
Our electrophysiologic analyses have shown that damage to the PNS in sporadic CJD occurs more frequently than is normally assumed. Among the 16 patients diagnosed as having sporadic CJD, we found damage to the PNS in 88% of the patients. In the remaining 12%, features of chronic disease in motor cells were revealed by neuropathologic examinations. On the basis of these findings, we conclude that PNS involvement is an integral part of sporadic CJD.
Accepted for publication March 25, 2002.
Author contributions:Study concept and design (Drs Niewiadomska, Kulczycki, Wochnik-Dyjas, and Szpak); acquisition of data (Drs Niewiadomska, Kulczycki, Wochnik-Dyjas, Szpak, Rakowicz, Łojkowska, Niedzielska, and Tarnowska-Dziduszko, and Mss Inglot and Wiȩclawska, and Mr Glazowski); analysis and interpretation of data (Drs Niewiadomska and Wochnik-Dyjas); drafting of the manuscript (Mss Inglot and Wiȩclawska, Mr Glazowski, and Dr Tarnowska-Dziduszko); critical revision of the manuscript for important intellectual content (Drs Niewiadomska, Kulczycki, Wochnik-Dyjas, Szpak, Rakowicz, Łojkowska, and Niedzielska); statistical expertise (Dr Kulczycki and Mr Glazowski); administrative, technical, and material support (Dr Niewiadomska and Mss Inglot and Wiȩclawska); study supervision (Drs Niewiadomska, Wochnik-Dyjas, and Szpak).
Corresponding author and reprints: Maria Niewiadomska, MD, Department of Clinical Neurophysiology, Institute of Psychiatry and Neurology, 9 Sobieskiego, 02-957 Warsaw, Poland (e-mail: firstname.lastname@example.org).