What is meant by progression? – Part 2
To view Part 1, please click here.
Relapses are the most obvious symptom of MS and an indication that inflammation is occurring in the central nervous system. But they’re only part of the story. The other part is neurodegeneration, which is the main process causing long-term disability. When evaluating the effectiveness of an MS drug, relapses are the low-hanging fruit: more obviously apparent and less ambiguous than progression. Everyone can agree (more or less) if a relapse has or hasn’t occurred.
But the value of relapse reduction as the sole benefit of a drug has come into question over the past decade. In the first few years after diagnosis, suppressing inflammation probably provides long-term benefits – a reason to start treatment as early as possible. But during the course of the disease, relapses become a poor indicator of how a person is doing. Once the secondary-progressive MS (SPMS) phase has begun, relapse frequency decreases but disability increases – illustrating the growing disconnect between inflammation (relapses) and neurodegeneration (disability).
As new drugs – Tysabri, Gilenya, Tecfidera, Aubagio – have become available, it has become increasingly important to show that treatment will slow the development of disability rather than just reduce the relapse frequency. The difficulty, of course, is that disability is a long-term outcome which gradually develops over decades in people with relapsing-remitting MS whereas drug trials last only one or two years. So is there some measuring stick that researchers can use to demonstrate that MS drugs are reducing the rate or severity of disease progression?
The measure that’s used is “sustained disability progression”, sometimes called “sustained accumulation of disability” or something similar. What this measure tries to do is provide a snapshot of the pace of accumulating disability in the short term to represent what will happen in the long term. The tool that’s used is the EDSS scale, with successive scores showing neurological disability over a period of time.
MS is a disease of fluctuation: good and bad days, on and off symptoms. If you’re trying to detect the signal of how someone is doing, these day-to-day variations constitute background noise. To cancel out the noise caused by relapses and the disease itself, researchers have devised cut-off values for the amount of impairment and the amount of time to define sustained disability progression. In clinical trials, the most common definition is a 1-point deterioration in a person’s EDSS score (the amount of impairment) that hasn’t improved three months later (the amount of time). This three-month window is intended to exclude short-term worsening caused by relapses.
Of course fiddling with these values will produce very different findings. For example, if the impairment criterion has a lower cut-off value (a 0.5-point change), then more people will appear to have disability progression compared to a higher cut-off value (a 1.5-point change). Similarly, extending the observation period would allow more people to heal the damage caused by relapses, so an EDSS change measured at six or 12 months (rather than three months) would be expected to identify fewer cases of progression – although arguably the cases it did identify would be more likely to be “true progression” rather than short-term or EDSS progression. (Clinical trials typically measure 3- or 6-month sustained progression.)
Since sustained disability progression is an artificial construct, it’s fair to ask if it really represents true progression. The answer, it seems, is no.
A new study at Harvard has looked at various definitions used for sustained disability progression to see if they accurately identified true progression (Healy and colleagues. Mult Scler Int 2013;2013:189624; free full text at www.ncbi.nlm.nih.gov/pmc/articles/PMC3608311/pdf/MSI2013-189624.pdf). The study examined data for 929 people with clinically isolated syndrome (CIS) or MS. The main EDSS cutoff values they looked at were a 0.5-point change, a 1-point change, and a 1.5-point change. The timeframe they used in most cases was six months.
If a trial defined EDSS progression as a 0.5-point change, then 42% of people would be said to have sustained progression. If the amount were a 1-point change, it would be 30%; and if it were a 1.5-point change it would be 16%. So the answer is very different depending on how you pose the question. If you ask how many people are showing evidence of progression, more people will seem to be progressing and at a much faster rate if you call progression a 0.5-point change rather than a 1.5-point change.
But even these numbers are misleading. When the researchers extended the average observation time to 3-4 years, only a minority of people continued to show evidence of progression. At six months, 42% of people had a 0.5-point change – but at the end of the observation period, only 44% of this group (18% in total) continued to show progression. In other words, most cases of “sustained progression” were not sustained. Similarly, using the 1-point cutoff value, only 44% of the 30% who seemed to be progressing (i.e. 13%) had sustained progression. And 51% of those with a 1.5-point change (or 8% in total) had sustained progression. So when someone in a trial is said to have sustained disability progression, in most cases what’s being measured isn’t true progression. A ballpark estimate would be that perhaps 10% of people actually have some measure of worsening function that will persist. Given enough time, most people will be able to heal much of the damage caused by their relapses – at least in the early stages of their MS.
Another finding from this analysis was that people with less disability at the start of treatment showed faster rates of disability (no matter what criteria were used) compared to people with more disability at the start. In other words, it’s easier to show that a person with no or minimal disability (EDSS 0 or 1) has sustained worsening of 1 point on the EDSS compared to someone with moderate disability (EDSS 3) at the start.
This has two important implications. People with minimal disability might be expected to be earlier in their disease course and have better healing, so flagging them as having “sustained progression” is probably going to be even more wrong-headed. Secondly, drug trials enroll people very soon after diagnosis when many have minimal disability, so they may skew the estimates even more – over-representing how many people are actually progressing.
So some caution is needed when looking at trial results. For example, in the DEFINE trial of Tecfidera, 32% of people had an EDSS score less than 2.0 at the start of the study (Gold and colleagues. N Engl J Med 2012;367:1098-107). Even more (37%) had an EDSS less than 2.0 in the CARE-MS I trial of alemtuzumab (Cohen and colleagues. Lancet 2012;380:1819-1828). This group might boost the percentage of people reported to be progressing.
In DEFINE, 16-18% of people taking Tecfidera had “sustained progression” compared to 27% taking a placebo using a 3-month cutoff. The Harvard paper suggests that less than 10% of these people would actually have progressed over the next few years. CARE-MS I used more stringent criteria for sustained progression so only 8% and 11% had sustained progression. Even so, perhaps only a minority of these people had true progression; the rest would get better whether treated or not. This isn’t meant to single out these two trials. The problem lies in “sustained disability progression” as an indicator of what’s really going on. Other groups have suggested the same thing – that there’s no valid way of assessing progression in the short term (Ebers and colleagues. Neurology 2008;71:624-631). Instead, what this measuring stick is really about relapses, random variation and error.
The Harvard group concluded that sustained progression – whatever the criteria used to define it – “has uncertain clinical meaning in terms of long-term disease progression.” There is a great need to determine if MS treatments in development will slow the progression of disability. Unfortunately, the only way to do that at the moment is to give someone a drug for 20 years and hope that the treatment worked.
Share this article
Facebook Twitter pin it! Email