Amyotrophic Lateral Sclerosis Mimic Syndromes: A Population-Based Study

Background  The Irish ALS Register is a population-based register of the epidemiological characteristics of amyotrophic lateral sclerosis (ALS) in the republic of Ireland.

Objective  To describe the clinical and demographic details of those patients included in the Irish ALS Register who were incorrectly diagnosed as having ALS (patients who were ultimately rediagnosed as having an "ALS mimic syndrome").

Methods  The medical records of each patient referred to the register are routinely reviewed and, where possible, patients are examined by our group during their illness.

Results  Between January 1, 1993, and December 31, 1997, 32 patients (representing 7.3% of 437 referrals) were rediagnosed as having a condition other than ALS. The median age at onset for these 32 patients was 56.0 years (range, 19.5-85.8 years) for men and 53.5 years (range, 39.5-70.4 years) for women. Twenty-nine patients (91%) presented with symptoms referable to the limbs, and the remainder presented with symptoms involving the bulbar musculature. Multifocal motor neuropathy was the most common condition mistaken for ALS, accounting for 7 cases (22%), followed closely by Kennedy disease (4 cases [13%]). Factors leading to diagnostic revision included evolution of atypical symptoms, results of specific investigations, and failure of symptoms to progress. Twenty-seven (84%) of the patients with an ALS mimic syndrome fulfilled the El Escorial criteria for either "suspected" or "possible" ALS, 4 (13%) met the criteria for probable ALS, and 1 (3%) had definite ALS.

Conclusions  The application of the El Escorial diagnostic criteria may facilitate early recognition of non-ALS cases. Misdiagnosis of ALS remains a common clinical problem despite the increased availability of investigations and a greater awareness among neurologists of potential diagnostic pitfalls.

THE TERM amyotrophic lateral sclerosis (ALS) mimic syndrome has been used to describe a diverse group of conditions, the presentation and clinical features of which may resemble those of ALS at the outset. It has been estimated that up to 10% of patients initially diagnosed as having ALS are ultimately rediagnosed as having a disease other than ALS.^1,2 Thus, the correct diagnosis of ALS remains a complex clinical problem despite the increasing availability of improved diagnostic tests,³ the development of specific research criteria for diagnosis (the El Escorial criteria⁴), and a greater awareness among neurologists of the existence of ALS mimic syndromes.^5-7 With the exception of a similar study¹ of ALS mimic syndromes from Scotland, to our knowledge, there have been few published studies of ALS mimic syndromes and their importance in clinical practice.

The Irish ALS Register has accumulated data on all cases of ALS since January 1, 1993, permitting ongoing study of the incidence and prevalence of ALS in the republic of Ireland. These data have recently been published.⁸ In this study, we describe patients referred to the Irish ALS Register who were incorrectly diagnosed as having ALS by referring physicians (those patients who were ultimately rediagnosed by our group as having an ALS mimic syndrome following review). In addition, we have compared the clinical and demographic characteristics of patients with ALS and patients with ALS mimic syndromes and have examined the accuracy of the El Escorial diagnostic criteria⁴ in distinguishing ALS mimic syndromes from true ALS.

The strength of this research lies in the existence of a complete register of all patients diagnosed as having ALS in the republic of Ireland since 1993. Multiple sources of information are used to ensure complete data collection for the Irish ALS Register, details of which have been published elsewhere.⁸ Neuromedical care is provided by a few consultant neurologists, all of whom collaborate in the enrollment of patients in the register. Most patients suspected of having ALS are seen by a neurologist at some stage of their illness. Although the diagnosis is determined for many by general physicians, most patients see neurologists for ongoing management. Occasional patients, especially those older than 75 years, are treated by a geriatrician. However, most of these patients undergo neurophysiological tests as part of their diagnostic workup, and all neurophysiologists working in Ireland collaborate in recruitment for the register.

During the 5-year study, clinical data were referred to the Irish ALS Register for 437 patients who were diagnosed as having ALS. Of these, 124 (28.4%) attended the ALS clinic run by one of us (O.H.) and, therefore, were examined regularly. Because of the demographic features of Ireland, it was not possible for our group to routinely examine the other 313 patients. Instead, these patients were followed up by their referring neurologist and physicians, who organized further management and confirmatory investigations as appropriate. Regular telephone contact was maintained with these patients through a research nurse (B.C.), in association with the Irish Motor Neuron Disease Association, Dublin. This nurse is also responsible for the continuous updating of the Register of Motor Neuron Disease, thereby permitting the clinical course of these patients to be closely followed.

The El Escorial diagnostic criteria for ALS⁴ are applied to all patients referred for inclusion in the register. Although early case ascertainment is encouraged, in practice, patients are referred to the register at different stages of their illness and at different levels of certainty of the diagnosis of ALS. However, all patients included in the register have been diagnosed by a referring physician or neurologist as having ALS and have undergone neurophysiological studies before the inclusion of their clinical details in the register. Patients were included in this study of ALS mimic syndromes if they were informed at some stage of their illness that their diagnosis was (or was likely to be) ALS, but this diagnosis was then revised at a later date, either by the referring neurologist or physician or by our group. In addition to the diagnostic criteria, patients included in the register must have established residency in the republic of Ireland for at least 1 year before the diagnosis of ALS being made. Unlike ALS, many of the ALS mimic syndromes do not have formal internationally agreed diagnostic criteria.

Eight patients with progressive muscular atrophy, 7 with progressive bulbar palsy, and 1 with primary lateral sclerosis were enrolled in the Irish ALS Register between January 1, 1993, and December 31, 1997. These patients have not been included in the present study.

Data are stored on a database,⁹ allowing for organized retrieval of data. Statistical analyses are performed using statistical software.¹⁰ A separate paper file is maintained for each patient. Ninety-five percent confidence intervals are estimated from the cumulative Poisson distribution.

In total, 437 patients were registered in the Irish ALS Register from January 1, 1993, to December 31, 1997. Of these, 404 patients had been newly diagnosed as having ALS after January 1993, and there were 33 prevalent cases at the time of register establishment. Following a careful review of patient medical records, neurophysiological test results, and where possible, the results of clinical examination by our group, 32 patients (7.3%) were ultimately rediagnosed as having conditions other than ALS. In each case, the patient had been diagnosed as having ALS by a neurologist and had neurophysiological investigation results suggestive of ALS before inclusion in the register. A diagnosis was determined for 18 patients (56% of misdiagnosed patients) before the establishment of the register, and they were still diagnosed as having ALS at that time. These patients were subsequently rediagnosed by our group as having diseases other than ALS after January 1993.

The demographic and clinical features of the patients with an ALS mimic syndrome and the patients with ALS enrolled in the register are outlined in Table 1. There was a preponderance of men among the patients with an ALS mimic syndrome compared with the patients with ALS. Among the patients with an ALS mimic syndrome, there was a considerably longer period from the onset of symptoms to the establishment of the correct final diagnosis, compared with the median of 8.1 months for the patients with ALS. Of the 32 patients with an ALS mimic syndrome, 29 (91%) presented with symptoms referable to the limbs. Bulbar symptoms were the initial symptoms in the 3 remaining patients (9%).

The revised diagnoses of the 32 patients with ALS mimic syndromes are listed in Table 2. Multifocal motor neuropathy (MFMN) was the most common condition misdiagnosed as ALS, accounting for 7 cases (22%). The clinical features of these patients are outlined in Table 3. Four patients (13%) referred to the register with a diagnosis of ALS were ultimately rediagnosed as having Kennedy disease (Table 4). Clinically, all 4 men with Kennedy disease were characterized by slowly progressive lower motor neuron signs commencing in their limbs and spreading over several years to involve the bulbar musculature. Perioral fasciculations were prominent in 3 of the 4 patients, and a family history consistent with X-linked recessive inheritance was found in 3 of the 4 patients. Sensorimotor neuropathy accounted for 4 cases (13%). In each case, the patient presented with purely motor findings and was reexamined on development of sensory symptoms. A second neurophysiological investigation revealed unequivocal evidence of sensory nerve involvement, rendering the diagnosis of ALS highly unlikely. Cervical spondylitic myelopathy only accounted for 1 case (3%); the low occurrence of this condition was most likely due to diagnosis by magnetic resonance imaging at an early stage in the patient's workup.

An 85-year-old man referred to the register by a general physician was ultimately rediagnosed as having hyperthyroidism. The confusion arose from the patient's presentation with generalized twitching of muscles and dysphagia thought to be secondary to pseudobulbar palsy. Examination 6 months later by a neurologist revealed a goiter, which accounted for the patient's swallowing difficulties, and an elevated thyroxine level. His symptoms resolved within 3 months of appropriate antithyroid treatment. A 65-year-old male smoker presented with right hand weakness and lower limb fasciculations without sensory symptoms or ptosis. An electromyogram demonstrated a lower brachial plexus lesion and a denervation pattern in all 4 limbs consistent with a diagnosis of ALS. Eighteen months later, the patient complained of paresthesia affecting the right ring finger and was markedly cachectic despite normal bulbar function. A chest x-ray film revealed a large apical lung carcinoma and the overall diagnosis was changed to Pancoast tumor. A 57-year-old man with a history of old polio was diagnosed as having ALS based on new-onset upper limb weakness, generalized fasciculations, and a denervation pattern in all 4 limbs on electromyography. This diagnosis was later revised to postpolio syndrome. There was a single case of multiple sclerosis confused for ALS. In 8 patients (25%), the new diagnosis was uncertain, but it was the opinion of the patients' physicians and of our group that ALS was no longer the correct diagnosis. These patients had developed signs (cerebellar signs, visual disturbances, and extrapyramidal features) incompatible with a diagnosis of ALS.

Factors that led to revision of the diagnosis of ALS fell into 3 broad categories, namely, evolution of atypical symptoms, failure of symptom progression, and results of specific investigations. These investigations (androgen receptor gene screening and magnetic resonance imaging) were initiated by our group following review of the medical records and/or the patient or by the patient's referring neurologist following review. In most cases, a combination of factors lead to diagnostic revision, the most common scenario being a clinical suspicion confirmed by specific investigations. In 4 cases, the patient developed symptoms that were atypical for ALS. A 50-year-old man with Kennedy disease developed detrusor instability 3 years after being diagnosed as having ALS; and the other 3 cases were patients with sensorimotor neuropathy in which the motor signs preceded the sensory signs. These patients included a 76-year-old woman who presented with lower limb motor neuron signs and developed sensory signs 1 year later, a 60-year-old man who presented with right foot drop and right hand wasting and subsequently complained of paresthesia 7 months later, and a 58-year-old man with rapidly progressive upper and lower limb motor neuron weakness who noted distal numbness at a later date.

Failure of symptoms or disability to progress as rapidly as expected was the most common reason for diagnostic revision, accounting for 11 (34%) of the 32 cases. These 11 patients included 3 with Kennedy disease, 3 with MFMN, 2 with motor neuropathy, and 1 each with Pancoast tumor, postpolio syndrome, and hereditary spastic paraparesis. The lack of clinical progression was also evident from the prolonged survival of patients with an ALS mimic syndrome. In contrast to the median survival of 2.2 years from the time of diagnosis among the patients with ALS, there have been only 4 deaths among the 32 patients with an ALS mimic syndrome. These deaths occurred in 1 patient with Kennedy disease, 1 with postpolio syndrome, and 2 with no definitive neurologic diagnosis at the time of death. No autopsies were performed in either case.

Table 2 outlines the El Escorial categories of the 32 patients with an ALS mimic syndrome at the time of their diagnosis. Twenty-seven (84%) of the 32 patients met the criteria for suspected or possible ALS, 4 (13%) fulfilled the El Escorial criteria for probable ALS, and 1 (3%) had definite ALS by El Escorial criteria. Patients with an ALS mimic syndrome rarely changed El Escorial category when observed over time. However, the clinical features of 2 patients with possible ALS evolved sufficiently with time to fulfill the diagnostic criteria for probable ALS and for definite ALS. These 2 patients were ultimately rediagnosed as having cervical spondylitic myelopathy and sensorimotor neuropathy, respectively.

Misdiagnosing ALS has many far-reaching implications for the patient and the neurologist. Potentially curative treatments exist for certain ALS mimic syndromes, but delay in commencement of these therapies may have a detrimental effect on outcome. For example, 90% of patients with MFMN respond to immunoglobulin infusions.¹¹ In our study, patients with an ALS mimic syndrome and their caretakers had been informed before referral to the register of the likely diagnosis of ALS and the associated fatal prognosis. The psychological stresses of being diagnosed as having a condition such as ALS,¹² together with implications for life and medical insurance and employment status, are clearly enormous. Furthermore, these may not be completely resolved even when the patients and their families have been fully informed of their more benign diagnosis at a later date. In addition, as accurate diagnosis is important in the context of clinical trials, reliability regarding the appropriate inclusion and exclusion criteria for trials is of critical importance, thus highlighting the importance of early recognition of conditions masquerading as ALS. Finally, delay in the diagnosis of inheritable diseases such as X-linked spinal and bulbar muscular atrophy may have genetic implications not only for the patients themselves but also for the patients' families.¹³

The proportion of patients with an ALS mimic syndrome (7.3%) enrolled in the Irish ALS Register is lower than that found in similar studies. This difference is more pronounced if the 18 patients who were diagnosed as having ALS before the establishment of the register are excluded. Haverkamp et al² reported that, of their 1200 patients referred to a monthly ALS clinic, 5% had possible postpolio syndrome and an additional 4.2% had concurrent medical problems that could mimic ALS. The Scottish Motor Neuron Disease Register,¹ which is identical in several methodological respects to the Irish ALS Register, noted that 53 (9.7%) of 546 individuals ultimately had a diagnosis other than ALS during the 3-year study from 1989 to 1992. As this study only enrolled patients in whom ALS was newly diagnosed (incident) and failed to account for prevalent patients alive on the date of establishment of the register, the figure of 9.7% is likely to be an underestimation of the total number of patients in Scotland with an ALS mimic syndrome. The corresponding figure for the incident cases in the Irish ALS Register is 3.8%.

The low occurrence of mimic syndromes in this study compared with that in the Scottish study was primarily a function of the small number of patients with cervical myelopathy (3% in the Irish study and 20% in the Scottish study). This may be attributed to our increased use of available technology, particularly magnetic resonance imaging scanning, compared with the Scottish group, which relied more heavily on clinical diagnostic criteria. Alternatively, there may be differences in the radiological criteria used in diagnosing this disorder. The postoperative improvement of the Irish patient diagnosed as having cervical spondylitic myelopathy confirmed the radiological diagnosis. In addition, the Scottish register accepted referrals from general practitioners in addition to hospital-based physicians and neurologists, thus increasing the risks of a false-positive diagnosis.

In the present study, MFMN was the most frequent source of diagnostic confusion, accounting for 7 (22%) of the 32 cases. This is in contrast to the Scottish register, where MFMN occurred in only 2 (4%) of 53 cases.¹ It is apparent from our study that confusion with ALS lay in the combination of fasciculations with retained or even brisk reflexes in the same limb. This was coupled with the limited availability of detailed neurophysiological characteristics in Ireland. Anti–GM1 antibody titers, which when elevated can provide a useful although not diagnostic marker for the presence of MFMN,¹⁴ were not routinely measured by nonneuromuscular neurologists or general physicians referring patients to the Irish register. The high incidence of MFMN in this study thus highlights the importance of performing further serologic testing and extensive neurophysiological studies on patients with pure lower motor neuron syndromes, in an attempt to demonstrate conduction block. Furthermore, an empirical trial of intravenous human immunoglobulin may be reasonable in patients with a solely lower motor neuron disorder or with retained reflexes, especially those of long duration and slow progression.¹⁵ All 7 patients with MFMN met the clinical criteria for possible or suspected ALS according to the El Escorial criteria, which is in keeping with a previous report.¹⁶ This again underlies the absolute necessity of performing adequate neurophysiological studies in such patients.

The phenotype of the 4 patients with X-linked spinal and bulbar muscular atrophy enrolled in the Irish register varied considerably, and suggestive clinical clues, such as gynecomastia, tremor, and impotence, were frequently absent. As suggested by other researchers,¹⁵ the most consistent clinical features in our 4 patients were facial fasciculations and a positive family history of Kennedy disease. However, the significance of these findings in our patients was not sufficiently appreciated by their referring physicians, thus leading to their erroneous referral to the register as patients with bulbar ALS.

In operating the Irish register, we examined all new referrals as rigidly as possible. Our observations suggest that application of the EL Escorial diagnostic criteria can improve early recognition of these non-ALS cases. However, critical review of clinical data, especially neurophysiological test results, is required to confirm or refute the diagnosis. Furthermore, patients with features atypical of ALS or those with inconsistent investigative results should be frequently and thoroughly reviewed. The increase in general awareness among Irish physicians of the necessity to perform further supportive investigations may also contribute to the lower rate of ALS mimic syndromes in the latter years of the Irish register compared with that found in other studies, although the number of patients who had been misdiagnosed remained unacceptably high.

In conclusion, the misdiagnosis of ALS by general physicians and nonneuromuscular neurologists remains a notable problem. We suggest that the clinical application of improved technology can facilitate early detection of ALS mimic syndromes, although it remains necessary to educate nonneuromuscular neurologists and general physicians of the importance of recognizing mimic syndromes. We have found that most investigating physicians and neurologists do not assiduously apply investigations aimed at identifying possible mimic syndromes, leading often to incorrect diagnoses and unnecessary hardship for patients.

Accepted for publication May 7, 1999.

This study was supported by the Irish Motor Neuron Disease Association and the Irish Brain Research Foundation (Dr Traynor and Ms Corr), Dublin, Ireland.

We thank the consultant neurologists, neurophysiologists, and primary care physicians who collaborate in recruitment for the Irish ALS Register; and Robert Brown, MD, DPhil, for his review of the manuscript.

Reprints: Orla Hardiman, MD, MRCPI, Department of Neurology, Beaumont Hospital, Beaumont Road, Dublin 9, Ireland (e-mail: ohard@iol.ie).