Super Smart Health

The Rise and Role of Patient-Generated Research

by Daniel Friedland | Dec 08, 2008 08:39:00 PM

ALS (Amyotrophic Lateral Sclerosis), or “Lou Gehrig’s Disease,” is a relentlessly progressive disease affecting the nervous system and leading to an inevitable decline in voluntary muscle use around the body. Currently, only one drug, riluzole, has been FDA approved to treat ALS and it has limited efficacy.

So when a prestigious medical journal published a study in February 2008 titled, “Lithium delays progression of amyotrophic lateral sclerosis”, it offered a glimmer of hope to all those who suffer with the disease.

The study, led by Francesco Fornai from Italy, was a small randomized controlled trial that compared 16 patients with ALS who took lithium plus riluzole with 28 patients who took riluzole alone. Over the 15-month duration there were no deaths in the group of patients taking lithium. In contrast about 30% of patients died in the group that did not. The progression of ALS symptoms also slowed dramatically in patients taking lithium.

The Rise of Patient-Generated Research

The seeds of promise from any first study, particularly a small one, begs the confirmation of others – but these studies often take time to do. In this era of Health 2.0 (see previous blog) the results of follow up research in this case came far more quickly than anyone would previously have imagined. Most remarkably, this research was initiated and conducted by an ALS patient and a family member of another patient leveraging the power of online communities.

In January, 2008, within 2 months of the first announcement of the Italian study, Karen Felzer, a PhD earthquake researcher from the U.S. Geological survey, whose father was diagnosed with ALS, and Humberto Macedo, a 42-year-old computer systems analyst and father of six from Brazil who has ALS, recruited 191 patients with ALS from the Internet to follow-up with a 6-month study on the safety and efficacy of lithium.

The researchers collaborated with the online community, PatientsLikeMe to collect and publish the data, a good fit for this type of study since it enables users to share information on a number of neurological conditions, including ALS, and provides tools for patients to describe their symptoms in detail.

In the study 60% of the patients self-elected to take riluzole with the lithium, while 40% decided to take lithium by itself. Patients then self-reported the progression of their symptoms using a well-validated instrument called the ALS Functional Rating (ALSFRS-R). Patients electing to take lithium were compared to 66 similar patients in the PatientsLikeMe database who did not take lithium.

By 6 months more than a third of patients had stopped taking lithium because of side effects, lack of efficacy, or doctor’s advice and only 50 patients reported their outcomes on the 6-month ALSFRS-R evaluation.

The findings of the study are available at the PatientsLikeMe research on Lithium and ALS section as well as the original website of the study. In contrast to the Italian study and to the disappointment of many, the study found no effect of lithium in slowing the progression of ALS.

So what does this all mean? How do we reconcile the findings of these two studies and what is the role for this new breed of patient-generated research?

The Role of Patient-Generated Research

At first glance it seems easy to dismiss the patient-generated research as an unreliable upstart. After all, this type of research falls lower down on the hierarchy of studies referred to in Evidence-Based Medicine (for more on this, see the section on Classification of Medical Research ). And with a large drop-out of patients, the study is certainly flawed.

The Italian study was an experimental design, a randomized controlled trial (RCT), considered the gold standard in research methodology.

The patient-generated study simply observed what happened to patients who elected to take lithium (i.e., it was observational research). In studies such as these it is difficult to tell whether the findings are due to the drug or the differences in the characteristics between the groups compared, the natural history of disease or a number of other issues, some of which have been nicely laid out in the the following blog. So does this type of research have any merit at all?

It adds another dimension to the clarifying the puzzle, and has value, as long as one knows how to jigsaw the pieces together.

While RCT’s are the gold standard, it is important to remember that these studies are usually conducted on highly selected patient populations under ideal circumstances. Observational studies, on the other hand, provide additional insight, because they peer into the experience of real world populations.

Rather than dismissing observational studies, any differences found between this type of research and the more highly regarded experimental designs should provide a spark for important questions. Often real world studies provide insight into the challenges of implementing therapy or what happens in different patient populations.

For example in the case of these lithium studies we would ask whether the lack of effect in the patient- generated research occurred because the patients did not have enough lithium in their bloodstream. In the RCT, researchers carefully titrated the dose of drug to blood levels between 0.4 and 0.8 mmol/l. In the patient-generated research, about two thirds of the population had blood levels below 0.4 mmol/l. Aha!

Yet the progression rates in the observational study were also not altered in the patients who obtained a 0.4 mmol/l blood level of lithium. Further, it did not seem to depend on whether they took riluzole with the lithium, or whether patients were early or late in disease progression at the time that lithium was started.

The unexplained discordant findings between these two types of research invite other trials to clarify the potential efficacy of lithium in ALS, especially since the initial RCT itself had its own limitations. It was a small study and only “single-blinded.” (Patients did not know whether they were receiving lithium, but one of the researchers did, and this opens the door to potential bias.)

Thankfully, another RCT is on the way. The new trial is sponsored by the ALA Association and will be led by Dr. Merit Cudkowicz, an Associate Professor at Harvard Medical School. It will randomize 84 patients to lithium or placebo and this time it will be “double-blinded.” (Neither patients nor researchers will know who is taking what medication).

Part of the story then is that real world patient-generated research can raise questions about the efficacy of a drug. Still we have to remember this kind of research is observational and does not yet incorporate statistical methods, academic rigor, or a thorough review to rise to the level of observational research published in peer reviewed medical journals. As such it is unlikely to ever provide the definitive word on efficacy. Yet, by stimulating questions it does contribute to clarifying the picture on efficacy. The fully realized potential of this type of research, however, probably lies elsewhere.

The Promise of Patient-Generated Research

Currently, the greatest challenge in evaluating drugs does not relate to establishing its efficacy, but rather its safety. The gold standard RCT, typically randomizing 200 to 300 patients, can give a good indication as to whether a drug is more effective than a comparitor.

However, the size of these studies is hopelessly inadequate to evaluate potentially serious or life threatening events that occur only infrequently. One usually requires thousands, or even tens of thousands of patients, to thoroughly evaluate safety concerns. This is very expensive to do in an experimental research setting and is therefore rarely done.

So what is done? We have to rely on observational research. Ideally, we observe what happens in large populations. And this is where the full power of Health 2.0 and online communities comes in. Networking the collective wisdom of the masses, and now with sites like PatientsLikeMe providing patients the ability to record their health experiences in great detail, we have the opportunity to pick up safety signals like never before.

With its speed, efficiency and reach the trend towards patient-generated research will inevitably continue, as will the sophistication with which this research is done. Given the power to contribute, patients frustrated with bureaucracy and current speed of research will not want to wait for answers. They will want to take an active role in moving the science along, and will likely do so.

Post a comment

Thanks for your comment!

Your Name

Your Email Address (required, but not displayed)

Your Web Address (optional)

Your Comment