SESSION 2B MND PHENOTYPES

C9 ALS MIMIC SYNDROMES

TALBOT K

University of Oxford, United Kingdom

E-mail address for correspondence: [email protected]

Keywords: mimic ALS, diagnosis

The core clinical syndrome of Charcot amyotrophic lateral sclerosis is unmistakeable to clinical neurologists once the disease has evolved to encompass mixed upper and lower motor neuron signs with clear progression. However, despite evidence that the rate of diagnostic error is low in specialist settings, the grave implications of making the diagnosis of ALS still leads to considerable anxiety amongst both specialists and non-specialists. Diagnostic doubt can contribute to management delays, impede open communication between doctor and patient and promote excess investigation. Particular areas of difficulty arise in young (<40 years) people with ALS, atypical regional syndromes (flail limb, isolated corticobulbar palsy), very slowly progressive cases and the existence of an ALS-like paraneoplastic syndrome. Using a systematic approach based on mode of presentation, this talk will review conditions which can mimic ALS and suggest a practical approach to identifying which cases require more extensive investigation.

C10 NATURAL HISTORY AND PROGNOSIS OF THE FLAIL ARM AND FLAIL LEG SYNDROMES

WIJESEKERA L¹, TALMAN P², MATHERS S², WILLEY E¹, GALTREY C¹, ELLIS C¹, SHAW C¹, AL-CHALABI A¹, LEIGH N¹

¹Kings College London, United Kingdom, ²Calvary Health Care Bethlehem, Melbourne, Australia

E-mail address for correspondence: [email protected]

Keywords: Flail arm, Flail leg, Prognosis

Objective: We sought to define the clinical, prognostic and biological significance of the flail arm and flail leg variants of amyotrophic lateral sclerosis (ALS) in comparison with typical forms of sporadic ALS.

Methods: We analysed clinical phenotypes in relation to survival in two clinic cohorts at the King's MND Care and Research Centre London, UK and the Bethlehem MND service in Melbourne, Australia. Using a standard protocol we categorised patients into phenotypes including bulbar onset and classical limb onset ALS, the flail arm syndrome (FA) and the flail leg syndrome (FL). 1137 cases were analysed from the King's cohort and 432 from the Bethlehem sample. Survival times from disease onset were analysed using the Kaplan- Meier method and the Cox proportional hazards model. Demographic and clinical characteristics between groups were compared using Oneway ANOVA with Dunnett's t test for post hoc comparisons.

Results: In the King's cohort, median survival for limb onset ALS was 34 months and for bulbar onset ALS it was 27 months (p < 0.001). In contrast, median survival for the FA syndrome was 61 months (p < 0.001 compared to limb onset) and for the FL syndrome, 69 months (p <0.001 compared to limb onset). Five year survival rates in the King's cohort were 9% for bulbar onset, 20% for limb onset, 52% for FA syndrome, and 64% for FL syndrome. In the Bethlehem cohort median survival for limb onset ALS 31 months, bulbar onset 27 months (p = 0.001 compared to limb onset), FA syndrome 66 months (p < 0.001 compared to limb onset) and the FL syndrome 71 months (p = 0.001 compared to limb onset). Five year survival rates were 9% for bulbar onset, 19% for limb onset, 52% for FA and 77% for FL variants. The ratio of men to women was 4:1 in the FA group compared to 1.5:1 in bulbar and limb onset cases. Age of onset was not significantly different between FA cases and limb onset ALS (Mean 57.3 years; p = 0.89) or FL cases and limb onset ALS (Mean 55 years; p = 0.90). 25% of flail arm cases and 33% of flail leg cases did not fall within the 2000 Revised El Escorial criteria. Median delay in diagnosis time was significantly longer in both flail arm (17 months; p < 0.001) and flail leg subgroups (25.5 months; p < 0.001) compared to limb onset ALS (11 months). Median time to spread to a second region was significantly longer in flail arm (29 months; p < 0.001) and flail leg cases (33 months; p < 0.001) in comparison to limb onset ALS (8 months). The median time from onset to NIV use was 22 months for limb onset ALS, and was significantly longer in the FA syndrome (median 51 months, p < 0.001) and the FL (median 52.5 months, p < 0.001).

Conclusions: Median and five year survival estimates for the FL and FA phenotypes were significantly longer than for more typical forms of ALS. These syndromes also show longer diagnostic delays, time to requiring ventilation remain localised for longer to the region of onset for longer than typical ALS. Our findings suggest that these syndromes have both clinical importance and biological relevance.

C11 THE CLINICAL FEATURES THAT DISTINGUISH PLS FROM ALS

GORDON PH, CHENG B, KATZ IB, ROWLAND LP, MITSUMOTO H

Columbia University, New York, United States

E-mail address for correspondence: [email protected]

Keywords: primary lateral sclerosis, PLS, Diagnosis

Background: Previously we reported that patients with clinically pure PLS, with no evidence of LMN dysfunction on examination or EMG after 4 years, had a significantly better prognosis and better function than patients with ALS. Patients with “UMN-dominant ALS,” defined by predominantly UMN signs and minor LMN signs, had findings and prognosis intermediately between PLS and typical ALS. Prior to the fourth year from symptom onset, the diagnosis of PLS could not be made reliably because many patients developed LMN signs before that time. Predicting which patients with solely UMN dysfunction are destined to develop LMN signs is difficult, but important to patients and to the design of clinical trials. We undertook the current project to determine how clinical features can be used to suggest a diagnostic category and provide an indication of eventual outcome. The specific objective was to analyze which examination findings help distinguish between categories at the first visit and during follow-up.

Methods: We reviewed the records of 21 patients (9 PLS and 13 UMN-dominant ALS) who had repeated follow-up for more than 4 years, the earliest time of reliable diagnosis, as well as those of 10 randomly selected control patients with ALS seen in 1984–2007. ANOVA F-tests for continuous variables and chi-square tests for categorical variables assessed differences in baseline data among the three diagnostic categories. Linear and logistic regression analyses with subject random effects assessed the relation between examination data over time and diagnostic group.

Results: At first examination, only the lowest score on the weakest muscle (p < 0.001) and the site of onset (p = 0.041) discriminated between eventual diagnostic group; PLS patients were stronger and more likely to have limb-onset than either the UMN-dominant or ALS groups. Across all visits, group assignment significantly predicted FVC (p = 0.026) and ALSFRS-R (p = 0.009) scores. UMN-dominant and ALS groups were more likely to have muscle atrophy (p = 0.009) or lost weight (p = 0.004), even when controlled for dysphagia (p = 0.020), and ALS patients were more likely to have hyporeflexia (p = 0.001) than the other groups. An MRC score of 4 or less on any muscle was associated with the diagnosis of ALS (p = 0.0001), but not PLS.

Conclusions: Features at baseline most suggestive of eventual LMN signs are muscle weakness and bulbar onset. Later, weight loss, reduced FVC, and limb muscle weakness are significant predictors of LMN dysfunction, and the transition away from the diagnosis of PLS. In particular, an MRC score of 4 or less in any muscle is associated with the eventual diagnosis of ALS, and its worse prognosis, but not PLS. We suggest that patients with only upper motor neuron signs have periodic EMG, muscle strength testing, weight measurement and FVC testing, because a change in any can signal the development of LMN signs and eventual ALS. Patients for whom these findings are normal after 4 years from symptom onset can be reassured by the diagnosis of clinically pure PLS, and may form a homogeneous group for enrollment in clinical trials.

C12 PROGRESSIVE MUSCULAR ATROPHY (PMA) IS SLOW AMYOTROPHIC LATERAL SCLEROSIS (ALS)

KIM WK¹, LIU X², SANDNER J³, PASMANTIER M³, MITSUMOTO H³

¹Hallym University College of Medicine, Seoul, Republic of Korea, ²Department of Biostatistics Columbia University, New York, ³The Eleanor and Lou Gehrig MDA/ALS research center, Neurological Institute, Columbia University, New York, NY, United States

E-mail address for correspondence: [email protected]

Keywords: PMA, ALS, survival

Background: PMA is clinically characterized by pure lower motor neuron disease. It remains uncertain if PMA is part of or distinct from ALS. Only the absence of clinical upper motor neuron (UMN) signs distinguishes PMA from ALS. While PMA is thought to have a protracted disease course [1], a recent study emphasizes progression in the disease course [2].

Objectives: To investigate the proportion of patients with PMA who develop UMN signs in the disease course and to identify the outcome and clinical management in patients with PMA to find out how different they are from those of patients with ALS.

Methods: We reviewed the medical records of patients diagnosed with motor neuron disease (MND) between 2000 and 2007. We identified 91 patients diagnosed with PMA and 871 patients with ALS using follow up data. We excluded patients with flail arm/leg syndrome, primary lateral sclerosis, and other unclassified MND or cases with insufficient data. Kaplan-Meier survival curve for time from onset of weakness to death was estimated and the difference in survival curve between groups was examined by log rank test. Cox proportional hazards model was used to assess the effect of risk factors on the survival time.

Results: PMA had a higher proportion of males than ALS (73.6% vs. 54.9%, p = 0.0006) and an older age of onset than ALS (mean of 63.4 years vs. 59.9 years; p = 0.0067). PMA patients seemed to live longer with median survival time of 48.3 months vs. 36 months for ALS patients (log rank test, p = 0.01). The Cox model suggested that the hazard of death increased with age of onset in both groups (p < 0.005). Furthermore, ALS patients were at a higher risk of death than PMA patients after controlling gender and the age of onset (HR = 1.74, p < 0.0001). Twenty (22%) of these PMA patients developed UMN signs during the follow-up period (up to 68.7 months). Between PMA with and without UMN signs, no differences were found in gender, body regions involved at the onset and diagnosis, forced vital capacity, ALSRFS-R at diagnosis and changes of those during the follow-up. No convincing differences were found between the two groups in survival time. The use of percutaneous endoscopic gastrostomy (PEG) was 10% vs. 15%, while noninvasive ventilation (NIV) was used in 30% vs. 60%. Among patients with ALS, 14.5% used PEG, and 29.8% used NIV (the data after 2003).

Conclusion: PMA, when diagnosed, regardless of subsequent development of UMN signs, had a longer survival than ALS. PMA was progressive, care and management required in these patients did not differ from patients with ALS. Considering recently accumulated knowledge in pathology, neuroimaging, and experience in familial SOD mutation cases, PMA appears to be slowly progressive ALS, rather than a distinct entity (The study was supported by MDA Wings Over Wall Street.)

C13 DIFFERENTIATION OF HEREDITARY SPASTIC PARAPARESIS FROM PRIMARY LATERAL SCLEROSIS IN SPORADIC ADULT-ONSET UPPER MOTOR NEURON SYNDROMES

BRUGMAN F¹, VELDINK J¹, FRANSSEN H¹, DE VISSER M², DE JONG JMBV², FABER CG³, KREMER BHP⁴, SCHELHAAS J⁴, VAN DOORN PA⁵, VERSCHUUREN JJGM⁶, BRUYN RPM⁷, KUKS JBM⁸, ROBBERECHT W⁹, WOKKE JHJ¹, VAN DEN BERG L¹

¹Rudolf Magnus Institute of Neuroscience, University Medical Centre ²Department of Neurology, Academic Medical Centre, Amsterdam, ³Department of Neurology, Maastricht University Medical Centre, Maastricht ⁴Department of Neurology, Radboud University Medical Centre, Nijmegen, ⁵Department of Neurology, Erasmus University Medical Centre, Rotterdam ⁶Department of Neurology, Leiden University Medical Centre, Leiden, ⁷Department of Neurology, Diakonessenhuis Utrecht, Utrecht, ⁸Department of Neurology, University Medical Centre Groningen, Netherlands ⁹Department of Neurology, University Hospital Gasthuisberg, Leuven, Belguim

E-mail address for correspondence: [email protected]

Keywords: primary lateral sclerosis, hereditary spastic paraparesis, upper motor neuron

Background: Differentiation between sporadic presentations of hereditary spastic paraparesis (HSP) and primary lateral sclerosis (PLS) is important for genetic counseling and for determining prognosis, as HSP generally has a more favorable prognosis than PLS and progression to ALS, as may occur in PLS, is not to be expected in HSP. To separate HSP from PLS previous studies used clinical features, such as age at onset < 40 years, evidence of mild dorsal column impairment, and symptoms of urinary urgency, but there is little evidence to support this.

Objectives: To study whether clinical characteristics can be used to differentiate sporadic presentations of hereditary spastic paraparesis (HSP) from primary lateral sclerosis (PLS) in patients with apparently sporadic adult-onset upper motor neuron (UMN) syndromes.

Methods: Patients with a gradually progressive, adult-onset, UMN syndrome of at least three years duration were included. Exclusion criteria were a positive family history and other causes of UMN loss. Mutation screening of the spastin gene (SPG4), the paraplegin gene (SPG7) and the seipin/BSCL2 gene (SPG17) was performed in all patients.

Results: Included were 52 patients with a phenotype similar to typical HSP (involvement of the legs only), 36 patients with a phenotype suggestive of PLS (bulbar region involvement) and 16 patients with a phenotype of UMN involvement of arms and legs that was difficult to classify as a specific phenotype. Causative mutations were identified in 14 patients (7 SPG4 and 7 SPG7; 13 with leg involvement only and one with symptoms in arms and legs). Although age at onset was lower in the patients with confirmed HSP (39 years, range 29–69), there was considerable overlap with the subgroup of patients with bulbar involvement (52 years, range 32–76). No differences were found for evidence of mild dorsal column impairment (decreased vibratory sense, or abnormal leg SSEP), symptoms of urinary urgency or mild EMG abnormalities. Disease onset in arms or bulbar region, development of bulbar region involvement or marked asymmetry during the disease course may support a diagnosis of PLS.

Conclusions: In patients with an apparently sporadic adult-onset UMN syndrome and symptom onset in the legs, differentiation of sporadic presentations of HSP from PLS based on clinical characteristics is unreliable and therefore depends on genetic testing, although development of bulbar region involvement or marked asymmetry during the disease course may support a diagnosis of PLS.

C14 DEFINING SURVIVAL AS AN OUTCOME MEASURE IN ALS

GORDON PH¹, CORCIA P³, LACOMBLEZ L², POCHIGAEVA K², ABITBOL JL⁴, CUDKOWICZ M⁵, LEIGH N⁶, MEININGER V²

¹Columbia University, New York, United States, ²Fédération des maladies du system nerveux, Hopital de la Salpétrière, Paris, France, ³CHRU Bretonneau, Tours, France, ⁴Trophos Biopharmaceuticals, Marseille, France, ⁵Harvard Medical School, Boston, MA, United States, ⁶Institute of Psychiatry, King's College London, United Kingdom

E-mail address for correspondence: [email protected]

Keywords: endpoints, outcome measures, survival rate

Background: Which primary outcome measure best defines disease progression in ALS is debated. Functional endpoints are used, but consensus guidelines suggest that survival should be the primary endpoint for phase III trials and survival remains the standard for drug approval by regulatory agencies. Survival analyses have been used in different trials, but measuring the time to death in ALS is complex. Nutritional and respiratory interventions have become more sophisticated; survival rates could differ depending on the type of intervention used, both between and within trials. Further, the definition of survival has changed from trial to trial, with some studies including time to tracheostomy or prolonged non-invasive ventilation as part of the survival outcome. The objective of this study was to examine the impact of respiratory interventions on survival as an outcome measure and to define survival rate for trials in ALS.

Methods: We reviewed the data from three phase III clinical trials conducted between 1997 and 2005, and examined differences in time to death, tracheostomy, and permanent assisted ventilation (PAV) where data were available. Respiratory interventions and their relation to vital capacity (VC) were assessed with descriptive statistics, chi-square and Fisher's exact tests for categorical variables, and t-tests for continuous variables. Kaplan-Meier methods estimated differences in survival time between interventions. A power analysis generated sample size estimates for different endpoints.

Results: Approximately 82% of patients died of respiratory failure; 18% of patients died from other causes, usually sudden death. The use of respiratory interventions across centers ranged from 0–6.6% (p = 0.001) of patients for tracheostomy, and 11.1–23.1% (p = 0.05) of patients for non-invasive ventilation. Twenty-two percent of patients with tracheostomy had VC ≥ 50% at the time of the procedure. In one trial, the mean survival time was 457.9±3.1 days at 18 months as defined by a combined outcome and 467.2±2.9 days with death alone as the end point (p = 0.017). An estimated sample size to detect a 10% difference between groups was 488 patients for the combined outcome and 472 patients for death alone.

Conclusions: Tracheostomy and PAV are not equivalent to death in ALS. The use of respiratory interventions differs between centers leading to variability in combined outcome assessments. The time to endpoint can differ significantly depending on its definition, and combining outcomes does not reduce the estimated sample size of a trial. Death rate alone is the least variable and most easily identifiable measure of survival rate in ALS.

C15 HOW WELL CAN WE PREDICT ONE AND TWO YEAR SURVIVAL FOLLOWING DIAGNOSIS OF ALS?

MILLER RG, MOORE DH, THE WALS STUDY GROUP

California Pacific Medical Center, San Francisco, United States

E-mail address for correspondence: [email protected]

Keywords: Prognosis, predicting survival, outcome measures

Background: Previous studies have assessed correlations between survival and baseline ALSFRS (both in its original and revised format), baseline FVC and other factors such as age, sex, onset site and time from first symptom to diagnosis. These studies have focussed on dividing patients in good and poor prognosis groups rather than individual patient prognosis.

Objective: To assess the feasibility of using a large collection of data for several sources to predict survival in individual ALS patients.

Methods: We assembled a large pool of placebo patients from seven clinical trials and one clinical practice. We used a Cox proportional hazards (PH) model to select baseline characteristics to be used in predicting survival at 6, 12, 18 and 24 months following diagnosis or enrollment in a clinical trial. Characteristics under study included age, sex, site of first symptom, initial ALSFRS and FVC as a percentage of expected for normals as well as a measure of rate of disease progression based on time since first symptom and current ALSFRS. This rate is defined as 40 (or 48 for ALSFRSr) minus initial ALSFRS divided by time since first symptom measured in years.

Results: Using the combined data from 2356 patients and a Cox PH model, we found that age, initial FVC% and ALSFRS rate of progression were significant predictors of survival. Unfortunately, we also found that survival varied significantly from study to study even after adjustment for differing baseline patient characteristics. The estimated probability of surviving 12 months ranged from 63% to 92% for a 58 year old male, with initial FVC 86% and a 8.7 unit average drop per year in ALSFRS during the time from first symptom to enrollment. These estimates are based on data from 6 studies with minimum 12 month followup. At two years the range was 36% to 72% based on data from 4 studies.

Conclusions: Age, initial FVC% and initial ALSFRS expressed as a rate of progression are significant predictors of survival for patients with ALS. However, the proportion of patients expected to survive 12 months varied from study to study. The sources of this variability are not explained by the data gathered from these studies.