Your Data at Work: Studying Loss of Walking Using DuchenneConnect
This summary was provided by Richard Wang, PhD, at UCLA’s Department of Human Genetics. Dr. Wang is a member of Dr. Stanley Nelson’s research team and one of their areas of interest is Duchenne muscular dystrophy.
We are excited to share information from our article recently published in PLoS Currents Muscular Dystrophy: “Online Self-Report Data for Duchenne Muscular Dystrophy Confirms Natural History and Can Be Used to Assess for Therapeutic Benefits”. You can view the original article here but for this post, we’ll skip the technical jargon and highlight some of the interesting findings made possible by your participation in DuchenneConnect!
When trying to find treatments for any diseases, the gold standard is the “randomized controlled clinical trial”. The simplest type of randomized controlled trial would give a drug under development to half the participants and a placebo to the other half. The problem with this type of study in the Duchenne community is that it is difficult to get enough people for these studies and these studies are very expensive. A web based questionnaire, however, is relatively easy to do and much less expensive than a randomized controlled trial.
For our study, we used the DuchenneConnect registry to study factors that affect the age when someone loses ambulation. We chose to use the age at which boys and men with Duchenne require full time use of a wheelchair for several reasons. One, researchers have shown that this measurement is useful and robust. Two, parents and patients generally remember this date accurately and three, it is a simple measurement for helping to figure out the effects of medications and supplements.
We used data collected from you through DuchenneConnect from 2007 until 2011. We did have some quality control measures in place to make sure we had high quality data. For instance, we removed profiles where the diagnosis was not Duchenne, the gender was female or not known, or the respondent came from a country where health care standards differed from the United States. (FYI, the reason we had to remove females and people with Becker was that their symptoms often look a bit different. Don’t worry – we care a lot about the women and girls in DuchenneConnect and people with Becker in the Registry and we plan more studies in the future.) In the end, we were left with data from 1057 men and boys (384 of whom currently use a wheelchair full time) to analyze. Figure 1 (in the journal article) illustrates the process we used to get our final group that we studied.
Comparison with Natural History Studies
In order to figure out if our data from DuchenneConnect was of high quality, we compared it with other natural history studies. Overall, our data from DuchenneConnect was very similar to three large databases (Dutch Leiden database, the UMD-DMD database and the MD STARnet cohort). For instance, the average age of Duchenne diagnosis hovered around 4 years old in all databases and the percentages of the types of genetic changes or mutations within the dystrophin gene (deletions, duplications, point mutations) were very similar across all groups. The age at loss of ambulation (Figure 2) also agrees with other studies.
Effect of Steroids
Many studies have shown that corticosteroids have a positive impact on boys with Duchenne. We wanted to see if we could replicate this result using our data set. Our reasoning was that if we can find the same result, then we can have more confidence in any other results we derive from DuchenneConnect. To do that, we looked at the 384 boys who use a wheelchair full time and split them in three groups: (1) those currently using steroids, (2) those never using steroids and (3) those who have used steroids in the past but have stopped. Then this information was plotted on a graph, see Figure 3A. From prior studies, we know that without steroids, most boys will start to use a wheelchair to get around beginning at the age 9-10. This is exactly what we find in the DuchenneConnect group. Boys never on steroids (green line) and boys who have been off steroids (brown line) both started using a wheelchair to get around full time at around age 10. Current steroid users started using a wheelchair full time at a later age, roughly 3 years. At this point, we did a statistical test to support our intuition and found that the differences between the groups were in fact significant. Putting all this together, we have found the same benefit of steroids as other randomized controlled trials!
It’s very exciting to see that we can replicate such a finding using nothing other than a web based survey where you volunteer to give your information. In fact, if we did not already know steroids were beneficial, we could have discovered that fact using the DuchenneConnect method.
Deflazacort versus Prednisone
As most of you are aware, the most commonly prescribed steroids are prednisone and deflazacort. Roughly 58% of the DuchenneConnect population uses deflazacort. When we compared prednisone users versus deflazacort users, we see that deflazacort seems to prolong ambulation by about one year (Figure 3B).
We also compared dosing strategies. The way people take steroids can vary quite a bit: daily, every other day, 10 days on and 10 days off, etc. To simplify, we created two groups: (1) daily users and (2) less-than-daily. Looking at Figure 3D, we see a possible trend where deflazacort may be more effective when given daily (solid line). However we did not find that the different groups were statistically different from each other. One explanation may be that our sample size is too small. We only have about 30 individuals on less-than-daily deflazacort. As more individuals participate in the registry, we can hopefully figure out whether daily versus less-than-daily makes a difference.
Clearly there are limitations to using registry data instead of doing a clinical trial. We can only observe information as it comes to us instead of designing the study exactly as we would like. Also, our questions may be lacking. For example, we did not ask everyone what dose of steroids they take (but DuchenneConnect is adding this question to the registry now). Having more detailed information such as the dose would give us an idea about what works and what doesn’t, and would open up new questions such as side effects of medications. The good news is that we might be able to obtain this information just by asking the participants, which is something not always possible with traditional trials.
The Use of Supplements
Guidelines for treatment of Duchenne leave some unanswered questions. We have some guidelines for steroids, but are there any other supplements that might make a difference? This is an exciting question we can begin to tackle with DuchenneConnect. Physicians and parents give different supplements based on their own line of reasoning (for example, calcium helps with bone strength) but the value of this type of data becomes powerful when we look across a larger group like the one in DuchenneConnect.
We looked at all supplements used by at least 20 individuals in the registry and compared them with the probability of walking until 13 years old. We compared boys on steroids alone vs. boys on steroids plus a supplement. For example, we looked at steroid-only boys versus steroids plus vitamin D users. To our surprise, vitamin D and Coenzyme Q10 were significantly associated with a delay in the use of a wheelchair full time (Table 2). This would give support to the idea that vitamin D and Coenzyme Q10 are effective in prolonging walking ability. Like any good bunch of scientists, however, we are aware that other factors may be at play here and we are missing some information. For example, we didn’t ask about dose or frequency of the supplements. But this finding certainly qualifies for follow up research.
Why this is exciting
One of the challenges for rare diseases like Duchenne comes from getting big enough numbers of participants for research and clinical studies. Randomized controlled trials are the gold standard for any sort of research or clinical trial but they are expensive, geographically limited and require several years to complete.
DuchenneConnect is an alternative model. In just five years, the Registry has collected the largest group of people with Duchenne and Becker muscular dystrophy to date. And that group has provided a lot of data! The hope is that this web survey model can provide a rapid turnaround from question to answer through a close partnership between patients, clinicians and researchers.
Using nothing more than a survey and patient reported data from a large number of individuals, we showed that we can replicate results from focused randomized controlled trials on the benefit of steroids. We have also found new insights, such as the increased benefit of deflazacort compared to prednisone, and the possible beneficial roles for vitamin D and Coenzyme Q10.
As the DuchenneConnect registry continues to grow, we expect to be able to better answer questions that are important to patients, their families and clinicians. Currently, we estimate 10% - 15% of the US Duchenne population is registered in DuchenneConnect, leaving an enormous potential for growth. Continued participation as well as new enrollment will hopefully allow us to further explore the role of different potential therapies.