Table 2 Responses to statement included in Round 1 (16 respondents)
1a. Which of the following NPC Severity Scales is the most useful as a practical measure of disease severity in normal clinical practice? | 17-domain NPCCSS [18] | 5-domain NPCCSS [16] | Disability scale [4] | Disease-specific disability scale [19] | NPC-cdb scale [20] | Functional disability scale [3] | None | Other | ||||||||||
18.75% (3) | 43.75% (7) | 12.5% (2) | 18.75% (3) | 0% (0) | 18.75% (3) | 0% (0) | 6.25% (1) | |||||||||||
1b. Please explain the reason for your answer | Summary of key insights: The 5-domain NPCCSS and Disease-specific disability scale were highlighted by multiple respondents as being simple, quick to administer and complete in any clinical environment in a routine clinical exam, with no additional work, tools or expertise required The increased validity of the 17- and 5-domain NPCCSS scales, given their recent use by multiple groups for large cohorts of NPC patients and in clinical trials, was cited in further support of their use While the time-effectiveness and accuracy of the 5-domain NPCCSS scale in clinical practice were acknowledged, its limitations were also flagged in terms of evaluation of certain subsets of patients, e.g. those with mainly psychiatric involvement or experiencing seizures The granularity of scores and the comprehensiveness of the 17-domain NPCCSS scale was appreciated by multiple respondents Notably, the accuracy of the description of eye movement impairment was questioned across all of the scales The challenges of capturing progression in late-onset patients with more slowly-progressing disease when using these scales was also raised, with the suggestion for greater granularity of scoring across domains | |||||||||||||||||
2a. In the context of routine clinical practice, if you had to limit measurements to only 5 of these domains, which would you select? | Ambulation | Cognition | Eye movement | Fine motor | Hearing | Memory | Seizures | Speech | Swallowing | Other | ||||||||
100% (16) | 100% (16) | 12.5% (2) | 93.75% (15) | 6.25% (1) | 12.5% (2) | 12.5% (2) | 81.25% (13) | 87.5% (14) | 25% (4) | |||||||||
2b. Please let us know which are the minimum number of domains you feel would be sufficient to reflect disease burden and progression in everyday practice and why | 1–4 domains | 5-domains | 7–9 domains | |||||||||||||||
18.75% (3) | 43.75% (7) | 12.5% (2) | ||||||||||||||||
3a. Which of the following NPC Severity Scales is the most useful as a measure of disease severity for enrolment, in the context of a research study or clinical trial? | 17-domain NPCCSS [18] | 5-domain NPCCSS [16] | Disability scale [4] | Disease-specific disability scale [19] | NPC-cdb scale [20] | Functional disability scale [3] | None | Other | ||||||||||
43.75% (7) | 37.5% (6) | 0% (0) | 12.5% (2) | 12.5% (2) | 6.25% (1) | 6.25% (1) | 6.25% (1) | |||||||||||
3b. Please explain the reason for your answer | Summary of key insights: The 17-domain NPCCSS scale was most popular among respondents in the context of clinical trial enrolment; it was seen as the most refined scale with the broadest coverage of the disease and the largest score range in each domain (5 instead of 4 or less). However, it was noted that the scale could be improved with respect to the linearity of the rating in some domains Granularity was seen as critical to measuring change and baseline assessment within clinical trials; it should be as comprehensive as possible while remaining quantifiable As more data becomes available, e.g., genomic data, there may be a need to reconsider which parameters are most important and whether preferred scales need to be amended accordingly Simplicity was seen as valuable for multi-center trials. The simplicity of the 5-domain NPCCSS scale, as well as its proven correlation with the 17-domain scale, may drive the preference for use in some trials Additionally, the question of which parameters can be expected to change in a clinical trial should be considered, as they determine both the endpoint and inclusion criteria and the identification of patients who can demonstrate measurable progression. Given the heterogeneity of the condition, general scores may not be suitable for every trial | |||||||||||||||||
4a. In the context of trial enrolment, if you had to limit measurements to only 5 of these domains, which would you select? | Ambulation | Cognition | Eye movement | Fine motor | Hearing | Memory | Seizures | Speech | Swallowing | Other | ||||||||
100% (16) | 93.75% (15) | 25% (4) | 93.25% (15) | 0% (0) | 18.75% (3) | 18.75% (3) | 87.5% (14) | 87.5% (14) | 6.25% (1) | |||||||||
4b. Please let us know which are the minimum number of domains you feel would be sufficient to reflect disease burden and progression in a clinical trial setting and why | 1–4 domains | 5-domains | 7–9 domains | |||||||||||||||
12.5% (2) | 43.75% (7) | 12.5% (2) | ||||||||||||||||
5a. Do you think that the ASIS score (Annual Severity Incremental Score), is a suitable measure to capture the rate of disease progression for trial enrolment and/or clinical trial outcome measures? | Yes, it is suitable for trial enrolment | Yes, it is suitable for clinical trial outcome measures | No, it is not suitable for either | |||||||||||||||
68.75% (11) | 62.5% (10) | 12.5% (2) | ||||||||||||||||
5b. Please explain the reason for your answer | Summary of key insights: Multiple respondents highlighted that as ASIS is a general scale and should only be a secondary outcome measure. It is not as sensitive as other scales, particularly over a potentially short period of a clinical trial Broadly, its value for both prospective and retrospective measures was recognised by the majority of respondents, particularly in regard to quantifying progression in a respective age group over multiple years of treatment The need for more data on its use was highlighted by two respondents It was seen by two separate respondents as a better indicator of disease progression than age of onset and arguably the best scale available for this | |||||||||||||||||
6a. Do you think an NPC severity score is a suitable endpoint for a clinical trial? If 'yes', which of the following NPC severity score systems is optimal? | 17-domain NPCCSS [18] | 5-domain NPCCSS [16] | Disability scale [4] | Disease-specific disability scale [19] | NPC-cdb scale [20] | Functional disability scale [3] | None | Other | ||||||||||
31.25% (5) | 50% (8) | 0% (0) | 6.25% (1) | 6.25% (1) | 12.5% (2) | 0% (0) | 25% (4) | |||||||||||
6b. Please explain the reason for your answer | Summary of key insights: To market an expensive drug, a sponsor will need to demonstrate a positive impact on the dynamics of a composite clinical progression score The challenge of conducting an outcome trial of sufficient duration (probably > 24mo) to see a robust statistically significant clinical effect in any of the scales with a reasonable number of participants was raised by more than one respondent A severity score was seen as a suitable outcome measure if the data are collected properly and in a rigorous and consistent manner across sites and with proper (and fairly simple) training. Otherwise, data are less reliable and more objective measures are needed, such as MRI, BAEPS, oxysterols, and videos with blind raters, of walking and the 9HPT, as suggested by other respondents To support reproducibility and reliability across trial sites, limiting the severity score to the 5 major domains was seen as sensible. These need to be guided with precise assessments (named tests) and be age/cognition dependent The 5-domain scale addresses the five most important domains, based on clinician and family opinion, and does not include items that can vary due to other treatments and thus act as confounders | |||||||||||||||||
7a. Which do you think are the key domains to capture as a clinical trial outcome measure? Please select all that apply | Ambulation | Cognition | Eye movement | Fine motor | Hearing | Memory | Seizures | Speech | Swallowing | Other | ||||||||
93.75% (15) | 75% (12) | 25% (4) | 100% (16) | 12.5% (2) | 25% (4) | 25% (4) | 87.5% (14) | 81.25% (13) | 12.5% (2) | |||||||||
7b. Please provide any further insights | Summary of key insights: The top 5-domains chosen by the group were seen as the most relevant to describe neurological disease progression. However, it was suggested the impact of seizures needs to be accounted for, as well as the quality of life of the patient and their caregivers Sophisticated computer assessment to measure speech in trials was suggested for consideration Until an effective disease modifying therapy becomes available, deciding what to measure in clinical trials remains a challenge. The solution proposed was to start by measuring everything and adapting endpoints dependent on the findings, particularly with different age groups involved | |||||||||||||||||
8a. Do you think that the adoption of a single severity scale in all scenarios is optimal, even if this means losing some refinement? If 'yes', which of the following NPC Severity Scales would you recommend? | 17-domain NPCCSS [18] | 5-domain NPCCSS [16] | Disability scale [4] | Disease-specific disability scale [19] | NPC-cdb scale [20] | Functional disability scale [3] | None | Other | ||||||||||
12.5% (2) | 50% (8) | 0% (0) | 6.25% (1) | 0% (0) | 12.5% (2) | 25% (4) | 0% (0) | |||||||||||
8b. Please provide any further insights | Summary of key insights: This was the most divisive question for the group, with many calling for greater consistency and optimisation of a single multi-domain scale on a global scale, while others suggested the use of a single scale would be too reductive. The following (sometimes conflicting) considerations were put forward: In the absence of a proven composite score that can work in all settings, the use of different scales in clinical trials should be at the liberty of each investigator/sponsor Neither clinical research nor clinical practice should be compromised by a one size fits all approach. This would be regression to the least common denominator Losing refinement of scales may be acceptable in some clinical routine practices but not in a trial setting. Even though an extensive set would be optimal the practicability may be less likely Alternatively, it may be appropriate to consider that if a scale cannot be implemented in routine clinical practice, it is not justifiable to use in a trial It is critically important to try to standardize scoring and implementation to make datasets comparable The 5-domain NPCCSS scale would be best suited to all three settings | |||||||||||||||||
9a. Are a limited number of domains sufficient to meet needs in all scenarios? If 'yes', tick the limited number of domains you believe would be sufficient | Ambulation | Cognition | Eye movement | Fine motor | Hearing | Memory | Seizures | Speech | Swallowing | Other | ||||||||
50% (8) | 43.75% (7) | 12.5% (2) | 50% (8) | 0% (0) | 0% (0) | 6.25% (1) | 37.5% (6) | 43.75% (7) | 6.25% (1) | |||||||||
9b. Please provide any further insights | Summary of key insights: Many respondents did not agree with the question that a limited number of domains could be sufficient to meet the needs all scenarios It was suggested that there should be a focus on domains where change can be expected with therapy and a domain where changes can be quantified. Measuring everything at baseline within clinical trials would show where changes occur The five identified domains are almost always all involved as the disease progresses. Only a small percentage of patients experience hearing loss and seizures, memory is a part of general cognition and too hard to separate from this domain, and eye movement change are difficult to measure In very young children, an additional developmental scale (e.g. Bayleys, Kauffman, etc.) should be used and, in adults, a dementia scale should be used One respondent suggested that it remains unclear if breaking down scores into domains is particularly helpful, while the dynamics of additive sum scores (across domains or without domains) is what matters for outcome trials An additional suggestion included videoing of the walk-test and 9HPT as functionally most relevant; with analysis performed by blinded raters on a 0 + / − 3 scale | |||||||||||||||||