Measuring Disease Progression in Primary Lateral Sclerosis

Abstract

Quantitative measures of disease severity are essential outcome measures for clinical trials. The slow progression of disease in primary lateral sclerosis (PLS) requires clinical measures that are sensitive to changes occurring within the time frame of a clinical trial. Proposed clinical outcome measures include the PLS functional rating scale (PLSFRS), burden scores derived from clinical examination findings, and quantitative measures of motor performance. The PLSFRS has good inter-rater reliability and showed greater longitudinal change over 6- and 12-months compared to the revised ALS functional rating scale. Examination-based upper motor neuron burden (UMNB) scales also have good reliability, and longitudinal studies are in process. Quantitative measures of strength, dexterity, gait, and speech have the potential to provide objective and precise measures of clinical change, but have been the least studied in persons with PLS.

Introduction

Reproducible, sensitive outcome measures are essential to assess effects of interventions in clinical trials. The slow progression of symptoms in primary lateral sclerosis (PLS) presents a challenge for developing outcome measures that reflect progression of disease over the time frame of a typical clinical trial. Consensus guidelines exist for outcome measures for clinical trials in amyotrophic lateral sclerosis (ALS) (1), but many of these – such as survival or respiratory function – are not relevant to PLS. There have been relatively few longitudinal studies in PLS with measures of disease progression. Measures that quantify disease severity include clinimetric scales of functional activity, scales based on findings on neurological examination, and quantitative measures of motor performance.

The novel Primary Lateral Sclerosis Function Rating Scale (PLSFRS)

For clinical trials in ALS, the ALS Functional Rating Scale-Revised (ALSFRS-R) is the most widely validated and broadly used clinimetric scale (2–5). The PLS COSMOS study found that the ALSFRS-R is much less sensitive when applied to patients with PLS. The 5-site PLS COSMOS study aimed to assess the natural history of patients with PLS using the ALSFRS-R. Patients with PLS only declined an average of 0.05 points per month over a two-year observation period, whereas patients with ALS declined at an average of 1.40 points per month.(6) None of the patients in the PLS-COSMOS study, which followed 50 PLS patients for three years, required non-invasive ventilation (Mitsumoto, personal communication). This substantial difference between disease progression of ALS and PLS patients demonstrated the need for a more sensitive clinimetric scale, similar to the ALSFRS-R, for future clinical trials in PLS.

Consequently, a group of investigators across 21 sites (the PLSFRS Study Group) was established to develop a novel PLS disease-specific clinimetric scale, the PLS Functional Rating Scale (PLSFRS). Using patient feedback and clinical observation, investigators added two intermediate levels of function to the first 10 questions of the ALSFRS-R, increasing the maximum possible score from 48 to 68 points. These additional levels of changes in function were inserted to capture early disease changes, adding a level between 2 and 3 points and between 3 and 4 points on the original ALSFRS-R. The final two items of the ALSFRS-R, that assess orthopnea and ventilation support, were unaltered to allow for comparisons between the ALSFRS-R and PLSFRS.

The ensuing PLSFRS validity study enrolled 77 PLS patients that were followed for one year. Internal consistency and construct validity, as compared against the Patient Reported Outcome (upper/lower extremity) and the Schwab-England Activities of Daily Living (ADL) Scale, and test-retest reliability (inter-, intra-, and telephone) were established. When compared to baseline measurements, the average scores of the PLSFRS and the ALSFRS-R both decreased significantly over six months and one year; however, the PLSFRS was more sensitive than the ALSFRS-R in capturing these changes. Between baseline and six months, the average percent change on the ALSFRS-R was 7% compared to a 10% change on the PLSFRS. This 3% difference was significant (p<0.0001). Additionally, between baseline and one year, the average percent change on the ALSFRS-R was 11% compared to the PLSFRS, which was 15%. This 4% difference was also significant (p<0.0001). The scores in each sub-domain of the PLSFRS (bulbar, fine motor, gross motor, and respiratory) showed significant decreases over one year, which was consistent with the Patient Reported Outcome scores for both upper and lower extremities and the Schwab-England ADL Scale (7).

This study demonstrated that the novel PLSFRS is a valid and reliable method to assess progression in PLS patients. When compared to the ALSFRS-R, the PLSFRS is more sensitive in detecting very subtle disease-specific changes. This scale can be used to shorten clinical trial periods to one year or even six months, considerably reducing study durations that would otherwise require much longer periods of time. To further evaluate its usefulness, the PLSFRS must be administered in clinical settings, natural history studies, and eventually implemented in early clinical trials. PLSFRS is shown in Table 1.

Table 1.

PLS Functional Rating Scale (reproduced with Permission (7))

1. SPEECH  6 normal speech processes  5 occasional speech disturbance, not detectable to others  4 detectable speech disturbance  3 alteration in habits to improve speech  2 intelligible with repeating  1 speech combined with non-vocal communication  0 loss of useful speech	7. TURNING IN BED AND ADJUSTING BEDCLOTHES  6 normal  5 slow, slight difficulty  4 somewhat slow or clumsy, needs no help  3 nearly always slow and clumsy, may require intermittent assistance  2 can turn alone or adjust sheets with great difficulty  1 can initiate, cannot turn or adjust sheets  0 helpless
2. SALIVATION  6 normal  5 possible excess saliva, may notice increased frequency of swallowing or coughing  4 slight but definite excess of saliva in mouth, may have nighttime drooling  3 definite nighttime drooling, may have day time drooling  2 moderately excessive saliva, may have minimal drooling  1 marked excess of saliva with some drooling  0 marked drooling, requires constant tissue	8. WALKING  6 normal  5 slow, may notice occasional tripping or change in gait  4 early ambulation difficulties  3 mixture of walking independently and with assistance  2 walks with assistance  1 non-ambulatory functional movement only  0 no purposeful leg movement
3. SWALLOWING  6 normal eating habits  5 early modifications to eating habits, may have occasional gagging or coughing, no episodes of choking  4 early eating problems, occasional choking  3 regular episodes of choking, most foods are problematic  2 dietary consistency changes  1 needs supplemental tube feedings  0 NPO (exclusively parenteral or enteral feedings)	9. CLIMBING STAIRS  6 normal  5 exercising caution and focus  4 slow  3 intermittent unsteadiness or fatigue  2 mild to moderate unsteadiness or fatigue, may use hand rail  1 needs assistance  0 cannot do
4. HANDWRITING  6 normal  5 slower, but not sloppy, requires more focus  4 slow and sloppy, all words legible  3 modifications to writing style (size, cursive versus print, etc.)  2 not all words legible  1 able to grip pen, unable to write  0 unable to grip pen	R-1. DYSPNEA  6 none  5 occurs occasionally with activities more strenuous than walking  4 occurs when walking  3 occurs intermittently with activities other than walking  2 occurs with one or more; eating, bathing, dressing  1 occurs at rest, either sitting or lying  0 significant difficulty, considering mechanical support
5. CUTTING FOOD AND HANDLING UTENSILS  6 normal  5 occasional difficulty cutting certain types of foods, no help needed  4 somewhat slow and clumsy, needs no help  3 using a modified method for cutting foods or handling utensils  2 can cut most foods, slow or clumsy, some help needed  1 foods cut by someone else, can still feed slowly  0 needs to be fed	R-2 ORTHOPNEA  4 none  3 some difficulty sleeping, d/t shortness of breath, does not routinely use more than two pillows  2 needs extra pillows to sleep (>2)  1 can only sleep sitting up  0 unable to sleep
6. DRESSING AND HYGIENE  6 normal  5 slower, but completely independent  4 independent self-care with effort or decreased efficiency  3 modifications to the frequency and/or completion of tasks, considering the use of one or more assistive devices  2 intermittent assistance or substitute methods  1 needs attendant for self-care  0 total dependence	R-3 RESPIRATORY INSUFFICIENCY  4 none  3 intermittent use of BiPAP  2 continuous use of BiPAP at night  1 continuous use of BiPAP day and night  0 invasive mechanical ventilation by intubation/trach ** If R-3 < 2, score R-2 as 0 (unable to sleep) ** TOTAL PLSFRS SCORE: ____of 68

Intermediate levels of function added to the ALSFRS-R are indicated in colored text

Scales to quantify Upper Motor Neuron Function

Scales using clinical examination findings to grade the upper motor neuron burden (UMNB) have been used in several imaging studies for ALS (8–11). These scales typically are heavily weighted toward measures of hyperreflexia, but may include measures of muscle tone or movement speed (9, 12–14). Recently one such scale that includes several types of UMN measures, the Penn Upper Motor Neuron Score© (PUMNS), underwent testing to validate its reliability (14).

The Penn Upper Motor Neuron Score©

The Penn Upper Motor Neuron Score© (PUMNS) was developed at the Penn Comprehensive ALS Center at the University of Pennsylvania to quantify the clinical burden of upper motor neuron dysfunction. Data from approximately 1800 patients seen at the Penn Comprehensive ALS Center were used to develop the PUMNS, of whom about 5% had PLS. The PUMNS was partially modeled after existing UMNB scales in the literature (9, 11, 15), and two previously validated scales, the Ashworth Spasticity Scale (16, 17) and CNS-Lability Scale (18) (CNS-LS) for the measurement of pseudobulbar affect were included as elements in the score. The PUMNS was designed with a few specific goals in mind: 1) to have face validity; 2) to be easily and rapidly administered as part of a usual and customary neurologic examination; 3) where possible to be based on previously used methods to quantify UMN disease; 4) to represent all segments of the body; 5) to reduce the potential for error as much as possible; and 6) to be reproducible with respect to inter-rater and intra-rater reliability.

Face validity is evident as the PUMNS score is comprised of checking reflexes, pathological reflexes, tone, and a standardized assessment for pseudobulbar affect (PBA), all well accepted UMN signs. The PUMNS takes less than five minutes to perform. Because reflexes and tone may be affected by weakness, an assessment of strength is required to determine whether a reflex should be considered brisk in the specific clinical setting. The PUMNS includes items for the bulbar segment and each limb. A thoracic item was excluded because it was found to be unreliable in early utilization of the score. Each item is a binary choice rather than a continuous variable in order to reduce inter-rater variability and increase accuracy of the score. This method proved to be effective.(19) A validation study has shown excellent inter-rater and intra-rater reliability (14).

The PUMNS has been used to correlate UMNB with radiologic findings(19) and pathologic findings in ALS (20). In ALS, the PUMNS score does not necessarily increase over time because the accumulation of LMN disease results in inability to elicit clinical signs of UMN disease. To calculate an individual’s maximum PUMNS, the highest value ever obtained from each segment is tallied for the total maximum derived score. To date, preliminary analysis has not found the PUMNS to be a good predictor of ALSFRS-R, NIV, PEG, or survival in ALS. The PUMNS was developed to be a quick, reliable, measure of clinical UMN disease that can be used longitudinally. Work is in progress on longitudinal analyses using the PUMNS in ALS patients and to examine its suitability for longitudinal studies in patients with PLS (Elman, personal communication).

The MGH-Upper Motor Neuron Burden Scale

The UMNB scale used at the Massachusetts General Hospital (MGH-UMNB) is based on reflex testing of bulbar, cervical and lumbosacral segments (Table 1). The scale ranges from 0–45. Deep tendon reflexes are graded 0–4 for physiologically present reflexes and 0–1 for presence or absence of pathological reflexes (21, 22). Table 2 shows the elements of the PUMNS and the clinical MGH-UMNB scale side-by-side.

Table 2.

Scoring examination findings in the Penn Upper Motor Neuron Score© and the MGH-Upper Motor Neuron Burden Score

UMNB Scale	Total Score	Sub-domains	Sub-domain components	Item scoring
Penn Upper Motor Neuron Score© (Woo JH et al, 2014)	0–32	Reflexes (bilateral)	a. Hyperactive (Biceps, Triceps, Patellar, Ankle)	0 if absent, diminished or normal 1 if pathologically brisk (≥ 3)/retained in weak muscle
			b. Pathological (Jaw jerk, Facial, Palmomental sign, Hoffman sign, Finger flexors, Crossed adductor, Clonus in arm or leg, Babinski sign)	0 if absent 1 if present
		Spasticity	Modified Ashworth scale (MAS)	0 if no increase in tone 1 if MAS 2 or 3 2 if MAS 4 or 5
		Pseudobulbar affect	CNS-Lability Scale (CNS-LS)	0 if CNS-LS < 13 1 if CNS-LS ≥13
Upper Motor Neuron Burden Score (Zurcher N, et al, 2015)	0–45	Reflexes (bilateral)	a. Hyperactive (Biceps, Brachioradialis, Triceps, Patellar, Ankle)	0 if Absent 1 if Diminished 2 if Normal 3 if pathologically brisk/retained in MRC grade ≤2 muscle 4 if Clonus
			b. Pathological (Jaw jerk, Hoffman, Babinski)	0 if Absent 1 if Present

The MGH-UMNB and PUMNS have been examined in multi-modal brain imaging studies at MGH, including MRI and Positron Emission Tomography using a marker of glial activation ([¹¹C]PBR28-PET). A cohort of 53 ALS participants who participated in a [¹¹C]PBR28-PET/MRI brain neuroimaging biomarker study at MGH, underwent either or both MGH-UMNB and PUMNS clinical scoring of UMNB at the same timepoint as their MRI/PET scans. Both MGH-UMNB and PUMNS scores showed statistically significant correlations with [¹¹C]PBR-PET uptake in the motor cortical regions on whole brain voxel-wise analyses (21). Both MGH-UMNB and PUMNS scores correlated significantly with [¹¹C]PBR-PET uptake (MGH-UMNB r= +0.60, p<0.0001; PUMN r= +0.61, p=0.0004) as well as fractional anisotropy (MGH-UMNB r= +0.40, p=0.006; PUMN r= +0.52, p=0.005) measures in the bilateral motor cortices in region of interest (ROI) analyses using a previously published automated processing pipeline (23). This study suggests that clinically observed upper motor neuron dysfunction may be attributable to glial activation and white matter axon loss, localized to the bilateral motor cortices in ALS. Furthermore, both MGH-UMNB and PUMNS scales are similar in their sensitivity of identifying participants with motor cortical dysfunction in ALS. In ALS, the MGH-UMNB did not predict need for mobility aids (Dr. Babu, personal communication). However, in a cohort of ten PLS participants the MGH-UMNB scale did not correlate with imaging measures of [¹¹C]PBR-PET uptake (r = +0.26; p = 0.46) or diffusion tensor imaging measures of white matter fractional anisotropy (r = −0.04; p = 0.89) (24).

Quantitative Measures of Motor Performance

Because PLS causes weakness and decreased motor control affecting speech, the speed and accuracy of movements, and mobility, quantitative measures of movement or performance of motor tasks offer another way to measure the severity of upper motor neuron dysfunction. An advantage of motor performance measures compared to functional rating scales is that they produce objective, interval level data. Rating scales are subjective and less sensitive to increments of change. For example, if each item on a rating scale typical has only 5 grades, each grade on the scale, at best, covers 20% of the total scale (assuming steps between grades are equal). However, the steps between grades on ordinal scales are often disparate. Individual item grades may be misleading and indicate plateaus or sharp declines despite steady progression of the disease. A rating scale score is usually expressed as the sum of all items (despite items being ranks and not integers). Additionally, functional rating scales are subjectively scored and can be influenced by multiple intervening variables (i.e. personal, environmental, cultural factors). In contrast, quantitative measures of movement provide interval level data that are objective and sensitive to change. There are several quantitative measurement tools that may be applicable to measuring disease severity as disease progresses in PLS.

Instrumented strength testing using a hand-held or fixed load cell

In patients with ALS, quantitative measures of isometric strength decline over time with UMN-associated weakness (25). In PLS, weakness typically progresses very slowly, and strength measures over a long period would be required to appreciate disease progression. In studies of patients with ALS, instrumented strength testing using a load cell was found to produce more sensitive, objective data compared with manual muscle testing(26) and correlated with ability to ambulate (27). Devices to measure strength were shown to have good inter- and intra-rater reliability (26, 28). Although quantitative muscle testing has not been reported in patients with PLS, its use in long-term studies of other neuromuscular disorders (29) suggests that it may be useful for detecting subtle strength changes over time in PLS.

Finger and foot tapping

Decreased speed and accuracy of movement is another sign of upper motor neuron loss that can be measured quantitatively. Measuring the number of finger or foot taps within a specified time is a quick, sensitive indication of motor control that does not require expensive equipment. Finger tapping was shown to correlate with imaging measures of motor cortex function, including cerebral prefusion in ALS (30) and spectroscopic imaging measures of NAA/Cr in both ALS (31) and PLS (32) and with transcranial magnetic thresholds in PLS.(32) Longitudinal measures of finger tapping in PLS were found to decline in the first years after diagnosis followed by stabilization later in disease (33).

Pegboard Tests

Although their use has not been reported in patients with PLS, timed pegboard tests are another method for quantifying hand dexterity (34). The Purdue Pegboard test is one such quick, reliable, and well-validated measure of hand dexterity (35). The test consists of counting the number of pegs that can be picked up from the dish and placed in consecutive holes on the board within a specified time.

Speech Rate tests

Slowed, dysarthric speech can be measured by counting the number of times that a speech sound (i.e. ‘pa’ or ‘pata’) is repeated in a specified time. This method of measuring speech rates is a quick, reliable, and sensitive method to measure dysarthria and requires no equipment except a stopwatch feature on an electronic device. Timed reading of a standard passage has been recommended for speech evaluation in ALS (36) and is another measure that could be followed longitudinally in patients with PLS.

Timed Up and Go (TUG)

TUG is a standardized test that measures the time needed to rise from a standard chair, walk 10 feet, return and sit back down into the chair. The TUG test is extensively used by rehabilitation professionals and has been well validated as a measure of gross functional mobility in several neurological disorders (37). Although TUG test results have not been reported in patients with PLS, the TUG test was found to be a predictor of falls in ALS (38).

Cognitive and Behavioral Evaluation

PLS patients have a low incidence of dementia in comparison to ALS patients, but several studies have reported mild executive function or behavioral impairment (reviewed in De Vries et al (39)). Because many cognitive tests are timed or require the use of the hands, test batteries designed to minimize effects of motor deficits, such as the ALS Cognitive Behavioral Screen (ALS-CBS; (40)) or Edinburgh Cognitive and Behavioural ALS screening tool (ECAS (41)) are preferred to screen for cognitive impairment. The ALS-CBS and ECAS have been used for longitudinal studies in patients with ALS (42–44), but to date have not been studied longitudinally in patients with PLS.

Discussion

Each of the methods for measuring disease progression has its advantages, disadvantages, and unknowns. Advantages of the PLSFRS are its brevity and ease of administration. It can easily be administered while interviewing a patient during a clinic visit. The PLSFRS was shown to have a 3–4% decline/month among a cohort of patients with PLS, a significantly faster rate of change than the ALSFRS-R. However, the greater slope of change over time was accompanied by increased variability and it is not known whether the slope remains linear throughout the duration of disease. UMNB scales also have the advantage that they add little time to a patient’s clinic visit, as the values are obtained in the course of a neurological examination. Challenges for using UMNB scales as an outcome for clinical trials in PLS include: (a) Many participants may have ceiling or near-ceiling scores for reflexes and spasticity without much dynamic range in the total scores of these scales to allow correlation analyses with other electrophysiological or imaging or clinical outcomes, (b) clinically meaningful changes in UMNB scores may occur too slowly to observe measurable change over the typical time frame of a clinical trial and (c) concomitant medications such as baclofen or dextromethorphan-quinidine may mask upper motor neuron signs. It is not known whether UMNB scores change in a linear fashion over time. UMNB scales based on reflexes alone showed a ceiling effect in a small longitudinal study of patients with PLS (45). It is also unclear whether counting points for each hyperactive reflex provides a linear measure, since reflexes in the same limb are often correlated in PLS. As a clinical trial outcome, quantitative measures of motor performance offer the advantage of being objective and providing interval-level data. Tapping tests, timed gait, and speech rate measurements are simple, easy to perform and sensitive to early change. In contrast, quantitative measures of strength require specialized instrumentation and training. The time needed to carry out these quantitative tests is somewhat of a disadvantage, although simple timed measures add very little time to a clinic visit. It is not known whether motor performance declines in a linear way throughout the course of disease in PLS. It is also not known whether other aspects of movement besides measurements of speed and strength would better quantify disability in PLS. The possible involvement of descending motor pathways other than the corticospinal tract was raised in the Second International Conference on Primary Lateral Sclerosis. Descending motor pathways controlling axial muscles, balance, and startle could present a different burden that could be quantified in PLS. The observation has been made that the apparent slowness of movement in PLS results from delayed initiation of movement, not slowing in the execution of individual strokes. Delay in ankle dorsiflexion at each step, for example, summates to produce slow gait. Work showing a delay in the initiation of anti-saccades in PLS patients was noted to be consistent with this observation (46).

Summary: Gaps and Future Needs

Progress has been made to develop a PLS disease-specific clinimetric scale that measures functional progression and to validate its reliability when administered in person, by phone, and longitudinally. The PLSFRS was designed to be administered by trained study evaluators or physicians. It was validated in North American patients with PLS, with longitudinal phone follow-up by investigators at the Columbia University site. Further validation in a larger, internationally diverse cohort is needed, as well as studies examining whether the PLSFRS correlates with other UMN measures. The PUMNS and MGH-UMNB scales, both based on clinician examination, have been used in cross-sectional neuroimaging studies as measures of clinical severity (19–22), and studies to measure their reproducibility are underway. As noted above, longitudinal studies are needed to assess the sensitivity of UMNB scales to change over time in patients with PLS, particularly over the typical time frame of a clinical trial. To date, relatively little work has been published using quantitative measures of motor performance to assess disease progression in PLS.

Lastly, there is a clear need to include outcome measures from the perspective of a person with PLS. To date, relatively little attention has been paid to using patient-centered outcomes or quality of life measurements in studies of PLS. Additionally, it will be important to determine whether functional scales, UMNB scales, or quantitative measures of movement predict clinical milestones, such as the need for using a walker, a wheelchair, loss of ambulation, or need for communication devices. A longitudinal multi-national observational study using the PLSFRS, upper motor neuron scales, and simple quantitative motor measures would be optimal. We recommend that at least some sites should evaluate how the outcome measures presented here correlate with biomarkers such as neuroimaging and neurophysiology measures of motor cortex and corticospinal integrity.