Wednesday, February 25, 2009

Following Research into Type-1 Diabetes Cures

Every now and then, I get email from someone asking about all the diabetes research they hear about. Often these are parents of kids who have just been diagnosed with type-1 diabetes. They are being flooded by news that makes it sound like a cure is just around the corner, but they intuitively know that's not true, and are confused about why the news seems so positive, but the results so meager.

At other times, someone will email me a web page that reports some progress on research into a cure for type-1 diabetes, and that person will be all excited. And they will want to know how soon this cure will be ready to use, and why I'm not reporting on it more aggressively. I'll look at the exact same web page (or newspaper article), and think to myself "what's the big deal?"

This posting is my suggestions for what research is worthwhile following, and which is not; which research is likely to lead to a cure sooner rather than later. I'm not claiming that this is the only way to do it, but it works for me:

Even if you limit your reading to treatments aimed at curing type-1 diabetes, there is still a lot out there, and it is still hard to know what you should care about and what you should not. So I suggest that you read my earlier blog entry on "Understanding the Research Funnel" here:
It contains background information on how research progresses.

Rule 1: Ignore animal research.

I suggest that you completely ignore all research which has only been done on animals, and also research which has only been done in petri dishes or animal tissues ("in vitro"), for these reasons:
  1. Human studies (called "clincial studies") must be done after animal studies and before people can use the cure. But human studies take at least 10 years. So anything done in animal studies is more than 10 years away from widespread human use. You will hear about it going through human trials for years before you can use it.
  2. Less than half of the things that are successfully tested on animals are ever tested on humans at all. So even if something is totally successful in animals, there is a less than 50/50 chance it will be tested on people.
  3. I don't know exactly, but it appears that less than 1% of the treatments which are successful in animals are also successful in people. So a successful animal test, you wait over 10 years and have a 99% chance of not working on people.
So, for these reasons, and a few others as well, I don't think it's a good idea to spend much time following animal research. It is just to unlikely to pan out, and too far away from use. Therefore, I generally recommend that people ignore research results in animals. Instead, focus on the research that has already made it to human (or "clinical") trials.

In my mind, every web page, every newspaper article, and every blog entry which discusses research that has been done only on animals, should contain the following warning:

Warning: this research has been done on animals, not people. Even if wildly successful, it has a less than 1% chance of ever being useful for people, and is more than 10 years away from general use. Do not get your hopes up. This is a long way from you being able to use it, and may never work!

Rule 2: Not all Results are Created Equal

I'm completely results oriented. When I read news about research, I focus on the data results. Everyone's got a great idea, and almost everyone's got a sob-story about why you should give them money, and explanations about why they haven't gotten farther than they have. But none of that stuff matters to curing type-1 diabetes. All that matters is 1. results, 2. progress through the different phases of research leading to FDA approval, and 3. money (and money is only important because it helps with 1 and 2).

So with that in mind, this is what I look for in a newspaper article, press release, web page, blog entry or anything else:
  1. Human Results that cure diabetes
    • Lower BG numbers
    • Lower A1C numbers
    • Using less insulin

  2. Results that the FDA Regulates
    • The start or end of a phased human trial
    • Approval to start a new phased human trial, or continue/expand an existing one.

  3. Scientific Results that might lead to a cure
    • Changes in the immune system
    • Changes in sugar metabolism etc.

  4. Operational Progress
    • Funding
    • Development Deals
I focus on these because they directly result in a cure for diabetes (group 1), and will make that cure generally available in the US (group 2). I track group 3 only because they might lead to group 1 and 2 results in the future, and 4 because it funds 1, 2 and 3.

Think about it: "what is a cure"? The answer for me is that you have normal BG numbers, low A1C, and don't have to use insulin. Those are all the group 1 tests. What is "generally available in the US"? That means it has gone through a normal FDA approval process, so passed all three phases of human trials. That's the group 2 tests.

So, the first thing to look for in a news article, press release, web page, or anything else is hard numbers for the group 1 results. Did the treatment result in lower BGs? Lower A1Cs? Less insulin usage? If that data isn't available, then I take my excitement down a step.

The second thing to look for is any change in the phase status of the human trials. Are they starting phase-I trails? Have the finished phase-II trials? Those sorts of milestones are easy to track, and do show important progress. [d1]

The last scientific results that I look for are other findings that the researchers think are important. Usually these are small changes in the immune system that the researchers think are important, but that don't effect A1C or BG numbers. These are -- by far -- the least important, because years before they will generally available, they will have to be refined into group 1 results, and I'll pay more attention to them at that point. [d2]

Fourth, I try to keep track of what I call "operational progress" which is separate from scientific progress, which is described above. Operational progress refers to progress which is not based on science, but which (I hope) is motivated by scientific progress. This can cover many different things, but mostly it involves money. If a company gets $10 million dollars, that means another company has reviewed it's research and thinks it is likely to work (and be profitable). If a company signs a marketing agreement, again that is good news, because it means someone else has reviewed the work, and thinks it is promising. [d3]

Rule 3: Accept no excuses!

If someone isn't making progress or the they don't have good results in terms of BG, A1C or insulin usage, they they will certainly have excuses as to why they don't have these things. My advice is to ignore these excuses, and just note the fact that they don't have results.

I don't have time to talk about all the hundreds of excuses different people have come up with when they haven't done the studies and don't have the results. But let me give you two examples. You will often hear people say "I've got a cure for type-1, but the big pharma companies make too much money off it and no one will fund me!" Not true. There are lots of small pharma companies that don't make a dime off of type-1 treatment, who would love to make a mint by curing type-1. Add to that the NIDDK and other US government agencies who would love the publicity of curing type-1. Not just the US government, but the UK, and many other governments would love to fund such a thing. Then add in the non-profits; not just JDRF, but ADA, DRI, JLN, and many more. There are many sources of funding besides "big pharma".

A related argument is the "my cure is generic so no one will make money so no one will fund me" Wrong again. First of all, there are many drug companies that only produce generics and would love to produce more of them. Second, government agencies and non-profits don't care if some company is going to make money or not. In fact, they would probably prefer funding research into a generic drug [d4]. And so on....

Rule 0: Only follow research that is peer-reviewed and published in an indexed journal.

We are very lucky in the world of type-1 diabetes, that we don't need to worry (too much) about outright frauds and quacks. So we don't need to worry (too much) about someone claiming to cure type-1 diabetes, who actually isn't. However, if you want to apply these rules to other diseases, you might not be so lucky. Some diseases have a real problem with frauds and quacks [d5] and a quick and easy way to filter these out is to insist on peer reviewed research published in indexed journals. There are scores of peer reviewed, indexed journals so if the research can not be published in any of them, it is a fair bet that there are problems with it. And as you apply this rule, remember Rule 3: Accept No Excuses! Every quack who can't get his stuff published in a peer reviewed, indexed journal will have lots of excuses about why not. Ignore them all.

Some extra discussion

[d1] I also keep track of predictions, hopes, and plans about these phases. So I look for statements like "this phase-II trial will finish in 2011", or "we plan (or hope) to start phase-III tests in 2009". I don't track these because I consider them progress by themselves (after all, anyone can hope to start something next year). I track them so I can look back and see if the researchers have a history of accuracy in their future predictions.

[d2] Beware especially of researchers who talk about group 3 results because they don't have any group 1 results! Or researchers who have been trumpeting the same group 3 results for years without translating them into group 1 results.

[d3] Although be careful of all of these. Some companies publicize every C-level hire, as though it showed their research was working. Or treat being bought as a sign of success (which is sometimes true, but not always).

[d4] A perfect example of this is Teva, which recently put $10 million into Andromedia which owns DiaPep227. Teva is a large scale producer of generic drugs (maybe the world's largest).

[d5] In general diseases which naturally change are the most subject to quackery and fraud. Type-2 diabetes and Autism are both examples. Both get better and worse over the years for a large number of poorly understood reasons. So this means the quack (or fraud) can sell something and if the patient naturally happens to improve, the quack gets to take credit for it! Type-1 doesn't work that way, because (except during the honeymoon phase) the patient will rapidly get very ill if standard treatment is replaced by quackery. And if quackery is added to regular treatment, most people can keep track of their own insulin use, and realize that it is not helping.

Joshua Levy

No comments: