Saturday, August 29, 2009

Update on Gitelman's ATG Phase-II Clinical Trial

Dr. Gitelman at UCSF has started a Phase-II human trial aimed at using ATG to cure (or lessen) type-1 diabetes. The goal of this study is to preserve remaining beta cells for people within 100 days from diagnosis. This study started in August 2007, and is expected to run until June 2011. So far it has enrolled 11 out of an expected 66 patients. It's a four day treatment (in hospital), and 12 follow up visits over a 2 year period. They are recruiting at a total of 9 sites (growing to 13), which are listed in the clinical trial page linked below.

Here is a quick summary of how they hope ATG (Thymoglobulin) will work:

Type 1 diabetes is caused by destruction of insulin-producing beta cells by T cells, part of the immune system. Thymoglobulin is a mixture of different antibodies that target T cells. This mixture includes the anti-CD3 antibody (and anti-CD3 treatments are in Phase-III trials already, having completed successful Phase-I and II trials). The researchers hope that this multi-faceted approach will be even more successful and have a longer lasting effect than with anti-CD3 alone, and may treat diabetes by several mechanisms. First, it lowers the number of T cells, so there are fewer to attack the beta cells. It also seems to alter the T cells remaining behind, rendering them less likely to be destructive. Following this depletion, the T cells that grow back in the following weeks may be reset and have a healthier balance (meaning that a special type of T cells, called regulatory T cells, will help keep the destructive T cells in check).

Using ATG to cure type-1 diabetes is an idea that flows from two sources. The first source is a line of clinical trials aimed directly at curing type-1 with ATG. Eisenbarth's work of long ago, a Phase-I trial done in Europe, etc. But it also has another source, which is Burt's Brazilian research as described below.

Discussion: How to Follow Up Burt?

One of the questions that I'm occasionally asked is this: when will there be a follow up to Burt's Brazilian research? When will there be a Burt, phase-II? This is a natural question, because Burt has -- by far -- the best results of any one. Many patients on that trial went years without using external insulin. No other study comes close. So an obvious question is: how do you move forward with that research.

Burt's research is the opposite of most. Most researchers use the smallest possible doses during phase-I. Phase-I is targeted at safety, so they use small doses to assure safety even at the expense of effectiveness. So for most studies, for phase-II studies, they move forward by raising the dose to try to make a safe treatment a more effective one. But Burt's research is the opposite. The effectiveness is the strongest yet, but there are real questions about safety. So you would NOT follow it up by raising the doses!

One way to follow up Burt's relative success, is by turning down the dose. Trade off a little effectiveness for added safety. Oversimplified a little: Burt's research involves using three drugs (ATG, GCSF, and cyclophosphamide), and reinjecting the patient's own precursor bone marrow cells (previously removed). Dosing with just ATG can be viewed as a "kinder, gentler" Burt. Especially since the cyclophosphamide is the most risky drug of the bunch. (I'll be posting on two GCSF clinical trials in the coming weeks.)

Since I'm a software engineer, I'll use a software analogy: Burt is doing a "hard reboot", Gitelman is trying for a "soft reboot".


One of the complexities of following the research, is that ATG is referred to by many different names by different groups. For example: Thymoglobulin, rATG, hATG, ATGAM, antithymocyte globulin, etc.

The cells that Burt is removing and then reinjecting are sometimes called "precursor bone marrow cells" and other times "adult stem cells".

The lead investigator for this clinical trial is Dr. Gitelman at UCSF. For those of you at Bearskin Meadows: yes, this is the same "Dr. Steve" who is often up there. I'm indebted to Dr. Gitelman for his insights into Burt's research and how it relates to other research, including his own. All mistakes, opinions, and oversimplifications are my own.

Patient oriented information on this trial:

Previous Blog postings on this (blog on ATG) (blog on Burt) (blog on Gitelman)

Web pages (but these need updating): (Gitelman) (Burt)

Clinical Trial records for this study: (Phase-II ATG) (Phase-I ATG) (Burt)

Joshua Levy

Sunday, August 23, 2009

Update on Pescovitz's Rituximab Phase-II trial

Pescovitz (and team from TrialNet) have finished a phase-II human trial of Rituximab, and the results have been accepted for publication at a big-name scientific journal. Unfortunately, that journal has embargoed the result data until publication, so I don't actually know the details of how well it worked. The good news, is that the study's authors did an oral presentation at ADA 2009, and (I'm told) that the results were good, but not as good as with the anti-CD3 monoclonal antibody drugs from Macrogenics or ToleRx. Those drugs preserved beta cell functionality for at least a year after treatment. Basically, untreated people lost half their ability to produce insulin in the year after onset of type-1 diabetes, but treated people retained about 95% of their ability. Rituximab was not as successful as that, but did preserve some insulin production.

Rituximab is a monoclonal antibody, a product of Genetech (now a division of Roche), and already approved by the US FDA for rheumatoid arthritis and several cancers. You can read more about it here:

It works by attacking B cells that have the CD20 marker. These B cells are part of the immune system and completely different that the beta cells (sometimes called B cells) in the pancreas. The immune system has many different types of both B cells and T cells. Most drugs currently being tested to cure type-1 diabetes target T cells. (For example: MacroGenics, ToleRx, Diamyd, Faustman, ATG, Abatacept, etc.) Rituximab is targeting a completely different type of cell (B rather than T). It is thought to work because B cells may interact with T cells influencing their targeting to attack beta cells. So this study is interesting, both because the results are good, but also because it potentially opens up a whole new area to understand as part of the cause of type-1 diabetes, and that means a whole new area where a cure or treatment might be found.

A follow on trial of Rituximab is currently being designed. It will probably involve repeated dosing of the drug. Obviously, I'll post when that clinical trial starts.

Discussion (Why many different treatments might be more effective.)

Right now, we have at least 5 drugs in phase-II or III human trials that, when given during the honeymoon phase, result in the patient loosing only half as much insulin production capacity, as if they were not given the drug. These include Rituximab, Diamyd's GAD65, several different anti-CD3s, and (maybe) Alpha Interferon. There are three ways of looking at these results:

The pessimist: "I don't want five half-way cures. I want one actual cure. Come back when you have something that works."

The engineer: "If it works half way now, then maybe we can improve it a little so it is three quarters working in a year or two, and 90% working a year or two after that and...."

The optimist: "Maybe we don't need a single cure. Maybe the proper combination of the many half cures we have will work right now, or will work for some people."

So if you're an optimist, then having several different treatments for type-1 diabetes which each target a different part of the autoimmune attack, is a good thing. We're used to thinking of a cure as being one pill or one operation or one something. In fact, for many of the more complex diseases (such as cancers, etc.), it is a cocktail of drugs which work together to have the best effect. And type-1 diabetes is surely a complex disease!

Of course "drug cocktails" are harder to test and harder to get approval for, and generally slower to market. In the past there has been discussion of pairing something that stops the autoimmune attack with something that regrows beta cells. That is one type of "drug cocktail", but it also may be that the thing that stops the autoimmune attack with the fewest side effects is itself a mixture of several drugs which target different parts of the autoimmune response.

Thanks to Dr. Gitelman for insightful discussions of this research. All mistakes here are my own.

Clinical Trial Record for the Rituximab trial is here:

Joshua Levy

Sunday, August 16, 2009

Brod Completes Phase-II trial of Alpha Interferon

Brod has finished a Phase-II clinical trial of oral interferon alpha, as a possible honeymoon cure of type-1 diabetes, and has published the results. The study involved 128 people, some of whom got 5k units of interferon, others got 30k, and others got none at all; daily doses for one year. After a year, natural insulin production in response to a meal was measured. The experiment was random assignment and double blind.

The results were this: The untreated group lost 56% of their insulin production one year after diagnosis. The group treated with 30k lost 46% but the group treated with 5k lost only 29%. So the best-to-worst summary is that the treated group lost about half as much insulin production as the untreated group. This result is very similar to Diamyd's and ToleRx's Phase-II results.

The press release describes Brod's theory this way:
Brod's theory is that autoimmune diseases, which occur when the body is attacked by its own immune system, are actually an alpha interferon immunodeficiency syndrome. Interferons are a group of proteins produced by cells in response to an attack by a virus.
My translation of this, is that he thinks that type-1 (and other autoimmune diseases) are caused by a lack of alpha interferon. Therefore, dosing with alpha interferon is an obvious path to a cure.

This clinical trial was funded by Diabetes Action Research and Education Foundation, the US Gov and the Children's Hospital of Minnesota Foundation.

Press release is here:

Abstract of the results is here:

Clinical Trial Record is here:
(By the way: this is one of the worst clinical trial records I've seen. Especially since the US Gov is a major contributor to this work. There is no end date, Brod is not listed as primary investigator, no description of how many patients got which treatment, etc.)
There is also another clinical trial record here, but I'm not sure if this is the same clinical trial, a related one, or a totally different one:

Joshua Levy