1 2 aim

#apaperaday: “If you cannot measure it, you cannot improve it”. Outcome measures in Duchenne Muscular Dystrophy: current and future perspectives

In today’s #apaperaday, Prof. Aartsma-Rus reads and comments on the paper titled: “If you cannot measure it, you cannot improve it”. Outcome measures in Duchenne Muscular Dystrophy: current and future perspectives

Today’s pick has dragon origami and is from Acta Neurologica Belgica, a perspective paper on outcome measures in Duchenne to measure efficacy of treatments by Benemei et al. DOI: 10.1007/s13760-024-02600-2

There are many therapies in clinical development for Duchenne and to assess whether they are effective (have an impact on Duchenne disease progression) you need outcome measures that objectively measure something that is relevant to the patients.

For Duchenne this is challenging as there is variability between patients, with different ages of loss of ambulation, onset of cardiomyopathy, rate of upper limb function decline, that are influenced also by development (or not) of scoliosis, contractures, obesity etc.

Another challenge is that only limited natural history exists. There is cross-sectional data, but longitudinal data where patients are followed for a long time are very limited. As Duchenne has an ambulatory and non-ambulatory phase, one outcome measure won’t work for everyone.

There will be a ceiling for upper limb functions in ambulant patients and non-ambulant patients cannot perform the ambulatory outcome scales. Biomarkers are suggested by the authors as a more objective scale – functional outcome tests will have a motivation aspect.

However, biomarkers are challenging as e.g. CK, a marker for muscle damage, is high in younger patients, but then declines when patients lose muscle. This makes it difficult to interpret a decrease (muscle getting worse or better?).

MRI can be used to measure fat infiltration, but here there are logistic challenges as not all hospitals have an MRI machine, and those who have, will have different machines and interpretation needs to be done centrally. Older patients also do not tolerate MRI well some told me.

Authors discuss the different scales that are used in trials such as the item scales North Star Ambulatory Assessment (NSAA) and Performance Upper Limb (PUL) and also the 4 stair climb and the 6-minute walk test. For the latter, the challenge is that many aspects influence outcome.

So stratification for many aspects is needed. For the PUL there are multiple versions, and the 2.0 is recommended as this has less of a ceiling effect. Patient-reported outcome measures are discussed too. Multiple scales exist, but these were often not developed for Duchenne.

Also, the challenge is that they do not account for patients adapting to a chronic illness. Finally, there is limited correlation between functional outcomes and patient-reported outcome scales. Still, ideally, you would take along both – provided the scale is relevant for Duchenne.

Finally, authors discuss digital outcome measures such as wearable. The stride velocity 95th percentile (SV95c) has been accepted by the EMA for Duchenne, but this can only be applied to ambulant patients. The wearable captures activity at home, but comes with a burden for families.

Data needs to be uploaded, equipment charged, etc. Authors discuss that the fact that many Duchenne patients have cognitive problems and behavioral issues hampers with doing functional tests. Furthermore, they discuss the duration of trials: usually 48 weeks, which is too short.

I fully agree here. The treatments for Duchenne aim to measure a slower decline – this is very difficult to detect (more difficult than an improvement from baseline) and a 48-week trial is too short. I understand this is not a nice message because it means longer placebo trials.

Authors outline efforts to collect and collate natural history and placebo group data such as DRSC and cTAP. They stress that finding a minimally clinically important difference is more challenging when a treatment slows down decline.

I agree with the authors, but want to add also that biomarkers can indeed be very useful but it is impossible to validate them lacking a therapy that works. Most Duchenne treatments are approved based on dystrophin restoration, without functional effects confirmed.

If you want to know whether a biomarker is predictive of future functional benefit, you need first to have a functional/clinical benefit. Hopefully, the number of trials that show this (for now, givinostat only) will increase so this work can be done properly.

This will take longer placebo-controlled trials to be able to detect clinical benefit (note the givinostat trial was 72 weeks and after 48 weeks no clinical benefit was seen). However, this will hopefully lead to the identification of biomarkers that allow future shorter trials.