This blog is different from most of my blogs because I'm reporting on a case study rather than a clinical trial. A case study is the experience of one person who got a treatment, as reported in a medical journal. Even the smallest (phase-I) trials usually have 10 or more people enrolled. But this journal article reports on a single person's experiences with a new treatment.
I'll discuss exactly what happened in more detail below, but the summary is that a 19 year-old woman was diagnosed with type-1 diabetes. She quickly started taking Fenofibrate, which is an FDA approved drug usually used to reduce cholesterol and triglycerides (fatty acids) in the blood. The results were shocking: her need for external insulin quickly dropped to zero. After approximately 14 months she stopped testing her blood glucose daily. She did not need to inject or inhale insulin for 21 months, at which time she still was not using external insulin. (!)
Results
The table below shows two separate data points on each day after treatment started. The blue line is injected insulin levels per day, and use the blue numbers on the left. The red line is blood glucose numbers, and use the red numbers on the right. For blood glucose, they use European units, but the 5 is about 90 in American units, 10 is 180, and 15 is 270.
Image is from the journal article and is presented for educational purposes only.
Starting on day 19, she stopped taking any insulin at all, with one exception: while backpacking in Sri Lanka she was admitted to a hospital with a high fever. The doctors told her to take insulin as a precaution, and she took 2 units. You can see that on the 133rd day.
Prior to treatment, her blood glucose numbers were high, as would be expected of someone newly diagnosed. Her A1c was 13.5 at diagnosis. However, her blood glucose numbers rapidly normalized. After she stopped taking insulin, her blood glucose averaged about 90, with A1c around 5.7%.
As you look at this data, which is amazing, it is important to remember that it is from just one person.
Background on Fenofibrate
Fenofibrate is commonly prescribed for high cholesterol and high triglycerides. In 2017, it was the 70th most commonly prescribed medication in the United States with more than eleven million prescriptions. However, it is not approved for use in children under 18. It is also sometimes prescribed for diabetic retinopathy (eye damage from long term diabetes), and (off label) for gout.
With a 45+ year history of use in Europe and 25+ years in the US, my assumption is that Fenofibrate is a safe and well understood drug. As with all drugs, it has side effects, some serious. However, this drug has been in use for a long time, by a lot of people, who typically use it for a long period of time, so it starts out with as good a reputation as you could hope for. It has no black box warnings, which are the most serious FDA warnings. To the best of my knowledge, it has no safety controversies (no class action law suits, no exposés on 60 minute, etc).
The Clinical Trial
Obviously, the women described in this case study has had the onset of
her type-1 diabetes delayed for 21 months (and counting), or possibly it
has been entirely prevented, as long as she keeps taking Fenofibrate.
So the key question is: how common is this result? Is this one person a
fluke, or is this the result we should expect from treatment with this
drug soon after T1D diagnoses? The only way to answer that question is
through a clinical trial.
A phase-I clinical trial has already (as of 16-April-2020) been started. It is being run by Dr. Flemming Pociot at the Steno Diabetes Center in Copenhagen, Denmark. This trial will enroll people between 16 and 46 years old who are within 6 weeks of their first injection of insulin (ie. just diagnosed). The primary end point is how much natural insulin the person is creating, as measured by C-peptides. Secondary end points include A1c numbers and insulin use. Patients will be followed for 2 years. There are expected to be 10 children and 48 adults in this study.
Discussion
This follow up clinical trial is exactly what I would expect. It has more than enough people and they will be followed long enough to get a good feel for the treatment. It's a randomized, controlled, double blind trial, so the highest quality. This trial is a safe and conservative way to be sure, in three years, if this treatment is worth an even larger study or not.
Normally, I would be all in favor of a rigorous study like this. And I am in favor of doing this study, but I also think it is time to be more aggressive. The results from this single case study are so strong and so unexpected, that I think we should also start a quicker, smaller trial, so we can get some data in 6 months rather than 3 years.
The bottom line, is that (because of the 3 year duration and the double blinding), we will not know if this works for 3 years. However, in the case study, the patient stopped using insulin after only 19 days, and that is a huge, good effect that would benefit honeymooners immediately, and would be obvious if it happened even in a smaller trial.
So I hope that another research team does a second study, starting right now, and running at the same time as the slow, careful, high quality study described above. This second study would be about 20 people, run maybe 3-6 months, and be open label / no control group. The advantage of this study is speed. If this study (which I'll call a "pilot study") is as successful as the case study, then it will be plenty big enough and long enough to show an effect. In this case study, the patient's insulin use stopped at 19 days, so even a 90 day trial (which is short) would have more than enough time to see that effect.
It is important to remember that some people naturally go through a period during their honeymoon where they don't have to inject insulin (called "spontaneous remission"), or who need to inject very little insulin ("spontaneous partial remission"). I have not been able to find a paper that has a specific number of how often it happens, but it is described in the literature. This period lasts for a few weeks or a few months, so is much shorter than the 19 months seen in this case study. However, since it is seen, it is important that the pilot study be large enough to make sure we are seeing a result of the treatment, and not this "spontaneous remission".
Furthermore, the drug is inexpensive and safe. Even if the drug has no effect at all, the danger of taking a common drug for a few months is tiny. On the other hand, if the drug has a good effect, it is already approved for use, and could start having that good effect for people with honeymoon T1D years before the slower study would complete.
This is a sort of "fail fast" plan which is common in Silicon Valley web startups. I'm usually not in favor of "fail fast" for medical research because of the impact on safety. However, in this case, the drug is well known, and the advantage of a quicker result is large (for honeymooners). Indeed, I would argue that the fast plan is safer, because if it doesn't work, the people will have gotten the drug for a shorter period of time, and if it does work, they can continue with an "off label" prescription knowing that it does work.
And I want to emphasize, that this plan is not extreme. With the publication of a (very) successful case study, the next obvious step is a pilot or phase-I study. For T1D, such studies are commonly about 20 people, commonly 3-6 months, and often don't have control groups. So they are exactly what I am proposing. The difference here is that the original researchers are running a larger, longer, more complex study: something closer to a phase-II study.
Questions and Answers About The Case Study
Are we sure she had type-1 diabetes?
It is a good question. Since all our data comes from one person, if her diagnosis is a mistake, then we have nothing. When she was diagnosed, her A1c was 13.5% and her max blood glucose level was around 540. When a 19 year old shows up with those numbers, I don't think anyone would fail to diagnose type-1 diabetes. In addition she had "minor diabetic ketoacidosis", "polydipsia" (drinking a lot) and "polyuria" (peeing a lot), all of which are classic T1D symptoms. She tested positive for two T1D autoantibodies. Although none of her 1st or 2nd degree relatives had type-1 diabetes, her mother had a different autoimmune condition.
Are we sure she did not have MODY diabetes?
MODY, also called monogenic diabetes is a group of rare diabetes which are caused by specific genetic defects. It is not one disease (like "type-1" or "type-2") but a class of related diabetes. By some estimates, 1% of the people who think they have type-1 diabetes, actually have a form of MODY diabetes. These people sometimes need insulin to treat their diabetes, but sometimes oral drugs can be used instead, and sometimes no drugs are needed, except in times of stress (such as pregnancy or flu). The journal article did not discuss any genetic testing for MODY diabetes.
Has this happened before?
There have been a few cases over the years where someone diagnosed with type-1 diabetes went into remission for a long period of time. None of them involved Fenofibrate. You can read three examples below:For me, these show why a clinical trial is needed. There is always the chance that this remission was just a random happening, and had nothing to do with the Fenofibrate she was taking.
Might This Cure People with Established Type-1 Diabetes?
I think it is premature to discuss this at all. The case study was someone who starting taking Fenofibrate within days of diagnosis, so was very much a honeymooner. But in terms of guessing about the future, there is both positive and negative information. On the positive side, when it was tested on NOD mice, it cured about half of the mice who already had T1D, and that is a hopeful sign. On the negative side, because this is a widely used drug, I'm sure it's been used by people with long established T1D, and I would think that someone would have noticed if it cured them. Similarly, it is surely used on many people who have type-2 diabetes (who often have high cholesterol or high triglycerides), and it's not known to help their blood glucose levels.
Even more on point: there is a long running clinical trial where people with established type-1 diabetes are given Fenofibrate to stop the progression of retinopathy (eye problems). This study started in 2016, and is expected to run until 2025. Blood glucose, A1c, and amount of injected insulin are not end points for this study, however, I would expect that if many patients started using much less insulin (and especially stopped taking insulin at all) someone would have noticed.
But, as with honeymooners, the only way to know if Fenofibrate will work on established T1Ds is to test it in a clinical trial specifically designed to look at this result as a primary end point. Right now, no such clinical trial has been registered.
Why did this person start taking Fenofibrate right after being diagnosed?
Basically, the woman's father saw a paper which showed that Fenofibrate had good results when used on NOD mice in their honeymoon phase, and so she started taking it. After 19 days she had this wonderful result, and never stopped taking it. My understanding is that this was done under the Danish version of "off label use". The family contacted the researchers who had done the NOD mice study and they wrote up this case study.
Can I take Fenofibrate if I have been recently diagnosed with Type-1 Diabetes?
No. And yes. (And this is not medical advice!) Fenofibrate is a prescription medicine, so you cannot just go out and buy it. However, it is approved for use in the US since 1993 (and longer in Europe), so your doctor can prescribe it for any purpose. In the US, this is called "off label use". I am in no way suggesting that anyone should take Fenofibrate. That is a question that only your medical team can answer.
Is this for real?
I don't usually
discuss researchers when I discuss clinical trials. My opinion has
always been that it is the results of the research that matters and not
the previous achievements of the researchers.
However,
when I report on especially good results from early research, I am
often asked "are these guys for real?" Of course, "for real" means
different things for different people. However, the three authors of
this case study are all at the Copenhagen University Hospital in
Denmark. The have published 200+ 25+, and 10+ papers respectively. One
is a full professor, another the chief of endocrinology. The person
running the clinical trial (who is not one of the authors of the case
study), has published over 190 papers, and is at the Steno Diabetes
Center, which is world famous.
In short, these researchers are very
much "for real" by my standards. But the real question is, and always will be, will they be successful? Will the results from this single person case study be replicated in clinical trials?
More Reading
Background Material:
Joshua Levy
http://cureresearch4type1diabetes.blogspot.com
publicjoshualevy at gmail dot com
All the views expressed here are those of Joshua Levy, and nothing here is official JDRF or JDCA news, views, policies or opinions. My daughter has type-1 diabetes and participates in clinical trials, which might be discussed here. My blog contains a more complete non-conflict of interest statement. Thanks to everyone who helps with the blog.
