A Long-term Prospective Study of the Natural Course of Sporadic Adult-Onset Lower Motor Neuron Syndromes

Objective  To determine the natural course of sporadic adult-onset lower motor neuron syndrome in a long-term prospective study of patients with the syndrome.

Design  Inception cohort with a follow-up of 72 months.

Setting  Three university hospitals in the Netherlands (referral centers for neuromuscular diseases).

Patients  Thirty-two patients were classified as having the following phenotypes according to previously defined criteria: progressive muscular atrophy (PMA; 10 patients), segmental distal muscular atrophy (8 patients), and segmental proximal muscular atrophy (14 patients). A disease duration of at least 4 years was chosen to exclude most patients with amyotrophic lateral sclerosis (ALS).

Main Outcome Measures  Muscle strength, functional impairment, and respiratory function were assessed at 0, 6, 12, 18, and approximately 72 months.

Results  The diagnosis had to be changed to ALS in 3 patients (classified at inclusion as PMA in 2 patients and segmental proximal muscular atrophy in 1) owing to the development of upper motor neuron signs in 2 patients and familial ALS in 1. The remaining 8 patients with PMA showed further deterioration, and the other 24 patients remained more or less stable during long-term follow-up. Respiratory insufficiency developed in 6 of the 11 patients with ALS or PMA, 5 of whom died.

Conclusions  Patients with lower motor neuron syndromes and a disease duration of at least 4 years usually have a favorable prognosis if muscle involvement has a segmental distribution. In patients with a generalized phenotype, progression is relentlessly progressive and eventually leads to death due to respiratory insufficiency.

Approximately 10% of patients with motor neuron disease have lower motor neuron signs only.^1-5 Whether this syndrome is a distinct nosological entity separate from amyotrophic lateral sclerosis (ALS) has been debated since its first description as progressive muscular atrophy (PMA) in 1850.^6,7 A substantial number of patients presenting with a lower motor neuron syndrome eventually develop signs of upper motor neuron (UMN) involvement,⁵ have a disease progression similar to ALS,⁸ or reveal UMN pathological features at autopsy.^9-11

However, a subset of patients with a lower motor neuron syndrome may have a disease course that is slowly progressive or stationary for years or for decades.^1,2,6 Identifying these patients is important in view of prognostication. However, reports on the natural course of sporadic adult-onset lower motor neuron syndromes are sparse, and their interpretation is hampered owing to a retrospective design, lack of standardized neurological assessment, and clinical heterogeneity because multifocal motor neuropathy, spinal muscular atrophy type IV, and Kennedy disease were not excluded. A retrospective study carefully excluding these mimics suggested that the prognosis is good for most patients with a disease duration of at least 4 years.¹²

At present, it is not known how long patients must be followed up before the diagnosis of PMA or of more favorable diagnoses, such as segmental spinal muscular atrophy, can be established. In a long-term prospective study, we assessed muscle strength, functional impairment, and respiratory function in patients with a sporadic adult-onset lower motor neuron syndrome of at least 4 years' duration to determine the natural disease course.

In 1998, the Academic Medical Centre in Amsterdam and the University Medical Centre Utrecht started a prospective study of patients with an adult-onset lower motor neuron syndrome. All patients underwent diagnosis by experienced neurologists at 1 of 2 university ALS clinics. Selection consisted of screening the medical records of our outpatient neuromuscular departments for the following diagnoses: lower motor neuron disease, PMA, progressive spinal muscular atrophy, and focal or segmental (spinal) muscular atrophy. In addition, patients with a new diagnosis at the outpatient clinics of the tertiary referral centers at the Academic Medical Centre and the University Medical Centre Utrecht were asked to participate in the study. Six patients were recruited from other university hospitals. We also included 3 self-referred members of the Dutch Neuromuscular Diseases Association who had a diagnosis of PMA. All patients underwent a standardized neurological, laboratory, and electrophysiological examination before inclusion. Inclusion criteria were (1) being older than 18 years at onset, (2) evidence of lower motor neuron involvement on neurological examination (weakness and atrophy), and (3) electrophysiological evidence of lower motor neuron involvement on standardized needle electromyography (EMG) as described earlier.^12,13 Exclusion criteria were (1) a disease duration of less than 4 years from the time of onset of weakness to exclude most patients with an ALS or ALS-like disease course; (2) a family history of lower motor neuron syndromes; (3) a deletion in the survival motor neuron 1 gene or an expansion of CAG repeats (>40) in the androgen receptor gene; (4) a history of acute poliomyelitis, spinal radiculopathy, diabetic amyotrophy, thyrotoxicosis, or hyperparathyroidism; (5) clinical signs of UMN involvement (pseudobulbar symptoms, clonus of the masseter reflex, hyperreflexia, or extensor plantar response); (6) sensory signs on neurological examination; (7) tracheostomy or intermittent ventilatory assistance; (8) structural lesions (tumors, intervertebral disc herniation, vascular lesions, or syringomyelia) on magnetic resonance imaging or computed tomography with myelography of the spinal cord; and (9) motor conduction block on extensive standardized nerve conduction studies according to previously defined criteria.¹⁴

The distribution of electrophysiological abnormalities was determined by bilateral concentric needle EMG. In each limb region, EMG was performed in 1 muscle as described previously.^12,13 A finding in the muscle was considered abnormal if there was spontaneous muscle fiber activity (fibrillations, positive sharp waves, or complex repetitive discharges) in at least 1 insertion, a severely reduced pattern on maximal voluntary effort mainly consisting of long-lasting polyphasic or giant (>7 mV) motor unit potentials, or no insertional and no motor unit potential activity.

To rule out other diseases, we measured the erythrocyte sedimentation rate and levels of hemoglobin, hematocrit, thyrotropin, phosphate, calcium, and, if the calcium level was elevated, parathyroid hormone. In addition, we performed serum immunoelectrophoresis with immunofixation and measured levels of serum IgM anti-GM1 antibodies.¹⁵

The study was approved by the medical ethics committees of the Academic Medical Centre and the University Medical Centre Utrecht, and we obtained written informed consent from all participants. Patients were seen at inclusion and at 6, 12, 18, and 72 months after inclusion. After the 18-month visit, patients were followed up by telephone to inquire about the development of symptoms of respiratory insufficiency and survival. The last visit took place on average 73 months after inclusion in the study (range, 66-84 months). Muscle strength, reflexes, respiratory function, and functional impairment were measured by the same investigator at each visit (R.M.B.-V. or J.V.).

Muscle strength was measured by Medical Research Council (MRC) grading,¹⁶ modified to a 9-point scale.¹⁷ The muscle groups were subdivided into upper, middle, and lower cervical and lumbosacral regions for a total of 12 limb regions per patient. A mean MRC score per limb region and the total mean MRC score for each patient were calculated as described earlier.^8,18 We considered a limb region to be affected when 1 or more muscle groups per region had an MRC score of 4+ or lower. We calculated the number of affected limb regions at each visit. Muscle tone was assessed using a modified Ashworth scale.¹⁹ Before the study, examiners practiced performing and interpreting reflexes in a standardized manner to minimize interobserver bias. The biceps, triceps, knee, and ankle reflexes were scored on both sides according to the National Institute of Neurological Disorders and Stroke myotatic reflex scale (0, reflex absent; 1, reflex small, less than normal; 2, reflex in lower half of normal range; 3, reflex in upper half of normal range; 4, reflex enhanced [subclonus]; and 5, reflex [clonus]).²⁰ A reflex score of 4 or 5 was defined as hyperreflexia and was therefore an exclusion criterion. Respiratory function was assessed by measuring slow vital capacity (VC),²¹ and results were expressed as a percentage of the predicted normal VC. A value of more than 80% of the predicted level was considered normal. Functional impairment was assessed using the Amyotrophic Lateral Sclerosis Functional Rating Scale²² (ALSFRS), consisting of 4 items for bulbar muscles (ALSFRS bulbar score, which includes speech, salivation, swallowing, and breathing), 3 items for arm functioning (ALSFRS arm score), 1 item for truncal muscles, and 2 items for leg functioning (ALSFRS ambulation score). Each item is scored from 0 (unable to attempt the task) to 4 (normal function), yielding a maximal score of 40. Patients who were no longer able to visit the outpatient clinic owing to advanced functional impairment were followed up by telephone.

At baseline, patients were considered to have a lower motor neuron syndrome and were classified as having the following phenotypes: PMA with generalized weakness (>50% of limb regions affected), segmental distal muscular atrophy (predominantly distal weakness of the arms), and segmental proximal muscular atrophy (predominantly proximal weakness of the arms). After follow-up, patients who had developed UMN signs were considered to have ALS, and the remainder were classified into 1 of the 3 described phenotypes.

We used the Mann-Whitney test to compare baseline characteristics. The results of the assessment of muscle strength, number of affected limb regions, functional impairment, and respiratory function before and after follow-up were compared using the Wilcoxon matched pairs test. P < .05 was considered significant. To estimate the rate of decline for the variables, we used a random-effects model for repeated measurements (SAS, version 8, PROC MIXED; SAS Institute Inc, Cary, North Carolina), and an interaction term of disease duration was entered as a covariate in the random-effects model. Of patients who died before the end of the study, the results of their last visit were carried forward.

Phenotypes at inclusion were PMA in 10 patients, segmental distal muscular atrophy in 8, and segmental proximal muscular atrophy in 14. After follow-up, 3 patients were considered to have ALS. The remaining patients developed no UMN signs, and the phenotypes were classified as PMA in 8 patients, segmental distal muscular atrophy in 8, and segmental proximal muscular atrophy in 13. The baseline characteristics of the 32 patients who were included are shown in Table 1. In the patients with ALS, the age at disease onset was higher and the median disease duration was shorter than in those with PMA. In the patients with PMA, weakness and functional impairment were more pronounced than in the groups with segmental muscular atrophy. Eight patients underwent a muscle biopsy (4 with PMA, 2 with segmental distal muscular atrophy, and 2 with segmental proximal muscular atrophy). There were no distinguishing features between the subgroups because all of the biopsy results showed neurogenic characteristics with type grouping.

All patients considered to have ALS after follow-up had developed respiratory insufficiency. One of these patients died suddenly, presumably because of a cardiac problem, 29 months after inclusion in the study and 13 years after the onset of weakness. Two patients died of respiratory failure, one 84 months after inclusion in the study and 15 years after the onset of weakness and the other 48 months after inclusion in the study and 10 years after the onset of weakness.

Three of the 8 patients with PMA developed respiratory insufficiency during follow-up, and 2 of them died. In 1 patient, VC decreased from 101% at inclusion in the study to 92% at 18 months. In the subsequent months, symptoms and signs of respiratory insufficiency worsened, and the patient died of respiratory insufficiency 33 months after inclusion in the study and 11 years after the onset of weakness. In the second patient, VC decreased from 87% at inclusion in the study to 77% at the last visit at 82 months. At that time he had developed dysarthria and dysphagia but no signs and symptoms of respiratory insufficiency. This patient died of pneumonia 88 months after inclusion and 34 years after the onset of weakness. The third patient had a VC of 38% at inclusion in the study, and noninvasive nocturnal ventilation was initiated several months after inclusion and 11 years after the onset of weakness. Respiratory functioning and VC have remained stable since then.

None of the patients with segmental distal muscular atrophy or segmental proximal muscular atrophy developed respiratory insufficiency during follow-up. One patient with PMA and 1 with segmental proximal muscular atrophy died during follow-up of unrelated causes (pancreatic tumor and stroke, respectively).

For all patients, the variables during follow-up are given in Table 1, Table 2, Figure 1, and Figure 2. No significant interaction between disease duration and increase in functional impairment or decline in muscle strength or VC was found, indicating that patients with a relatively short disease duration did not progress faster than patients with a long disease duration.

One patient with PMA at baseline and another patient with segmental proximal muscular atrophy at baseline developed hyperreflexia in the arms and legs 12 years after onset of weakness. In the patient diagnosed as having PMA at baseline, the diagnosis was changed to familial ALS after his sister developed a bulbar type of ALS. One patient who developed ALS had bulbar signs and symptoms at inclusion, which increased during follow-up (Figure 2B). The decrease of muscle strength during follow-up was more pronounced in the arms than in the legs (Table 2), and the increase of functional impairment was most pronounced in the ALSFRS arm score in patients with ALS than in the other subgroups (Figure 2C). All patients had abnormal EMG findings in all cervical limb regions, and 2 of the 3 patients had abnormal EMG findings in the thoracic region at the first visit (Table 2).

Muscle weakness, functional impairment, and the number of affected limb regions increased significantly (Figures 1A-B and 2A). At study inclusion, weakness in the legs was more severe than in the arms, but during follow-up the decline in muscle strength was more pronounced in the arms than in the legs and more pronounced in distal rather than in proximal limb regions (Table 2). Four patients had bulbar signs and symptoms at inclusion, and in 2 of them these signs and symptoms increased during follow-up (Figure 2B). The decline of the ALSFRS total score (Figure 2A) could be attributed to the 3 items that assessed arm function because only the ALSFRS arm subscale score declined significantly during follow-up (Figure 2C). Electromyographic abnormalities were found in a mean of 71% of all limb regions and in 63% of the thoracic regions at the first visit (Table 2).

No significant change in muscle strength, the number of affected limb regions, or VC was found during follow-up (Figure 1A-C). The decline of the total ALSFRS score was of borderline significance (P = .06) (Figure 2A). Weakness progressed to 1 additional limb region in a segmental pattern in 2 patients, specifically to an adjacent, more proximal spinal cord segment in one and to a lower cervical region on the contralateral side in the other. Electromyographic abnormalities were found in a mean of 40% of the cervical limb regions (excluding the limb region that exhibited the first symptoms), a mean of 29% of the lumbosacral limb regions, and 14% of the thoracic region at the first visit (Table 2).

The decline in muscle strength and VC and the increase in functional impairment were not significant during follow-up (Figures 1A-C and 2A). In 4 patients, there was segmental spreading of muscle weakness because more distal muscle groups were affected on the first affected side and more proximal muscle groups were affected on the contralateral side (Table 2) at the last visit. Abnormal EMG results were found in a mean of 50% of the cervical limb regions (excluding the limb region with first symptoms), a mean of 17% of the lumbosacral limb regions, and 31% of the thoracic region at the first visit (Table 2).

In the present study, a well-defined group of 32 patients with a sporadic adult onset of lower motor neuron syndrome and a disease duration of at least 4 years was followed up for a mean period of 6 years. Two patients with a generalized phenotype (ie, PMA) and 1 with segmental proximal muscular atrophy eventually were diagnosed as having ALS because of the development of UMN signs or the occurrence of ALS in a family member. In line with this diagnosis, they developed respiratory insufficiency. Of the remaining patients, those with a generalized phenotype (PMA) had a progressive disease course with increasing muscle weakness, number of affected regions, and functional impairment, and almost half developed respiratory insufficiency. Those with the segmental phenotypes did not develop respiratory insufficiency or significant changes in muscle strength, functional impairment, or VC. After follow-up, 3 patients were classified as having ALS, 8 as having PMA (from 10 at baseline), 8 as having segmental distal muscular atrophy (similar to baseline), and 13 as having segmental proximal muscular atrophy (from 14 at baseline).

In a previously published study focused on prognostic factors in 37 patients with a lower motor neuron syndrome and a disease duration of less than 4 years,⁸ 13 of 25 patients with PMA developed UMN signs and were subsequently diagnosed as having ALS. Twenty-six patients (with ALS and/or PMA) eventually died. In addition, of the 8 patients with segmental distal and segmental proximal muscular atrophy, 5 developed PMA and 1 developed ALS; of these, 5 died of respiratory failure. Of the 4 patients with segmental distal muscular atrophy, 1 developed ALS. Patients with a low VC at baseline and a sharp decline of VC during the first 6 months in particular had a relentlessly progressive course. The 3- and 5-year survival rates of 67% and 45%, respectively, were comparable to studies on survival in patients with ALS.⁴ In the present study of patients with a lower motor neuron syndrome of more than 4 years' duration, 3 of the 32 patients developed ALS, the remaining 8 patients with PMA showed further deterioration after a median disease duration of 17 years, and the other 21 patients did not show much disease progression during long-term follow-up. Thus, the prognosis is much more favorable in patients with a disease duration of at least 4 years, in particular in those with focal muscle involvement. Nevertheless, ALS may still develop, as is demonstrated by the patient with segmental proximal muscular atrophy who was found to have hyperreflexia in all 4 limbs after a disease duration of 12 years. Therefore, patients with an adult-onset lower motor neuron syndrome of more than 4 years' duration also should undergo clinical follow-up, including monitoring of respiratory function. In each patient group, the EMG showed abnormalities in a proportion of all regions (varying from 8% to 100%) at the first visit. Therefore, in an individual patient with at least 4 years' duration of disease, the EMG cannot be used to predict prognosis.

Both segmental muscular atrophy forms showed a disease course with very slow progression. Only 1 patient in our group with segmental distal muscular atrophy had an age at onset of 18 years. We included him in the study after a disease duration of 30 years. At that time, the upper, middle, and lower cervical limb regions were all markedly affected, with the lower cervical regions most severely. Therefore, a diagnosis of juvenile muscular atrophy of the distal upper extremity (Hirayama disease) was considered unlikely.^23,24 We did not find a difference in the rate of progression between segmental distal and segmental proximal muscular atrophy. This finding refutes other studies that reported a stationary disease course in segmental distal muscular atrophy²³ but a progressive disease course in segmental proximal muscular atrophy.²⁵ A spontaneous arrest of symptoms and signs in segmental distal muscular atrophy was reported within 5 years after the onset of disease in 73% of the patients in one study²⁴ and within 6 years in 91% of the patients in another study.²⁶ In a previous study of 10 patients with segmental proximal muscular atrophy, weakness had progressed throughout the arm in 7 patients and to the legs in 3 patients after a disease duration of up to 7 years.²⁵ None of these patients developed respiratory insufficiency, which is similar to our findings. Three of the 13 patients with segmental proximal muscular atrophy had the clinical phenotype of flail arm syndrome with severe, disabling, and proximal muscle weakness in both arms.²⁷ Nevertheless, we recommend monitoring these patients because 1 of our patients with a segmental muscular atrophy and 1 patient in the study by Katz et al²⁵ developed ALS.

In conclusion, the present study demonstrates that, in patients with a sporadic lower motor neuron syndrome with adult onset and a disease duration of at least 4 years, there is usually a favorable prognosis if muscle involvement is focal, although segmental spreading of weakness may occur and, rarely, ALS may develop in the occasional patient with segmental proximal muscular atrophy. In patients with generalized weakness and a phenotype of PMA, evolution into ALS or a relentlessly progressive course eventually leading to death due to respiratory insufficiency is the rule rather than the exception.

Correspondence: R. M. Van den Berg-Vos, MD, PhD, Department of Neurology, St Lucas Andreas Hospital, PO Box 9243, 1006 AE Amsterdam, the Netherlands (r.vandenberg-vos@slaz.nl).

Accepted for Publication: November 12, 2008.

Author Contributions:Study concept and design: Van den Berg-Vos, Visser, Kalmijn, de Visser, de Jong, de Haan, Franssen, Wokke, and Van den Berg. Acquisition of data: Van den Berg-Vos, Visser, Franssen, and Van den Berg. Analysis and interpretation of data: Van den Berg-Vos, Visser, Kalmijn, Fischer, de Visser, de Jong, de Haan, Franssen, Wokke, and Van den Berg. Drafting of the manuscript: Van den Berg-Vos, Visser, de Visser, Franssen, Wokke, and Van den Berg. Critical revision of the manuscript for important intellectual content: Visser, Kalmijn, Fischer, de Visser, de Jong, de Haan, Franssen, Wokke, and Van den Berg. Statistical analysis: Van den Berg-Vos, Kalmijn, Fischer, and de Haan. Obtained funding: de Visser, de Jong, and Van den Berg. Administrative, technical, and material support: Visser, Franssen, and Van den Berg. Study supervision: Kalmijn, Fischer, de Visser, de Jong, Franssen, Wokke, and Van den Berg.

Financial Disclosure: None reported.

Funding/Support: This study was supported by a grant from the Prinses Beatrix Fonds.