August 2010

Using Phase II Failures

Link Medicine
Link Medicine is currently testing a failed cancer drug for use in treating Alzheimer's disease.

We now know that many quite different diseases share common pathways and processes in the cell. Cancer is a disease of abnormal cell survival; in Alzheimer’s disease the survival pathways have failed. Alzheimer’s patients have significantly lower risk of many cancers. What if the cure for Alzheimer’s disease is sitting on some drug company’s shelf, as a potential cancer drug that failed in Phase II? (A biotech company called Link Medicine is currently testing one such failure to find out.) Gaucher disease and Parkinson’s disease both involve lysosomal damage and display aggregates of a protein called alpha-synuclein; Gaucher carriers are at elevated risk for Parkinson’s. What if a drug intended to cure Gaucher disease, one that failed in Phase II, is actually a treatment for Parkinson’s? (Another biotech company, Amicus Therapeutics, is beginning to investigate that possibility.) Recent studies show that people diagnosed with psoriasis are at greater risk of developing heart disease; in fact, in patients with severe psoriasis who are younger than 50 years old, the risk is comparable to that seen in diabetes. How many Phase II-failed psoriasis drugs have ever been tested in heart disease clinical trials?

I could list literally a dozen more examples, but you get the idea: because we balkanize biomedical research, biomedical-research funding and pharmaceutical industry drug development according to phenotypically classified diseases, the possibility that a drug that has failed the efficacy test in one disease might be efficacious in a completely different one has not permeated the culture. Yet, we should not be surprised if such cross-disease activity occurs, because these Phase II failures got as far as human clinical trials for a reason: they hit a target (or targets) in cell culture and animal models, and produced an effect. That they failed to do so in the human disease is an indictment of our disease models, not the biochemical and cellular data that showed they did something. What we need to do is test them on other diseases - a battery of other diseases - perhaps first in cell culture (iPS cells?) and animal models and, if they show an effect, then directly in Phase II clinical trials for the new disease. (It may even be that we can proceed immediately to Phase II studies without the animal model testing (after all, our animal models aren’t very good) if we have mechanistic or other data to suggest that efficacy is possible in the unrelated disease.) The problem is that such tests usually cannot be done by the original developer, because no drug company has programs in all the major human diseases. And they certainly aren’t going to let their competitors test them. So who will do the tests?

Academic labs are the perfect answer. In most cases, they discovered the disease targets and developed the disease models in the first place. Many of these labs are already trying to find compounds that show efficacy in those models, in the hope that a pharmaceutical or biotech company will become interested in developing them further. But they lack libraries of compounds to test that are known to be safe in humans and that are guaranteed to interact with something, anything, in the cell. The Phase II failures are a perfect library to test.

If the tests are successful, who would take the next steps of funding the new Phase II clinical trials? It makes sense for such funding to come from the government, and there is a new program that might be an interesting way to do it. In the new health-care reform bill recently signed into law by President Barack Obama, there is an amendment that authorizes the NIH to establish a $500 million a year program called the Cures Acceleration Network, whose mission is to aid in establishing partnerships between academic labs and industry that would accelerate the finding of cures for untreatable human illnesses. I have discussed this amendment in more detail in an article in Genome Biology (1). At present, there is no agreement on just how to fulfill that charge. I think finding new indications for some of the Phase II failures would be a great way to do that. Since they already have passed Phase I, they are very much closer to approval than any set of random compounds from other libraries, or even compounds that are currently in preclinical testing; all that is needed is to find the right disease for them, if one exists. $500 million per year would fund a number of Phase II clinical trials, as well as a grant program to identify academic labs to test the compounds in disease models for those cases where a second disease indication is not obvious from the biology. If efficacy is established against a new disease in humans, then the government could give the company that originally developed the compound the right of first refusal on an option to fund Phase III trials and then to market the drug. If that company was not interested, the government could hold a competition to select another company that would take the compound forward. In any case, the academic lab that established the new disease indication would get some royalties from the sales, as would the original developer.

Why should pharmaceutical companies be interested, given how jealously they guard their secrets? After all, the probability that one of their compounds will show any efficacy in a new disease is still quite low. However, there are some incentives, including especially good publicity, that might be persuasive - after all, the pharmaceutical industry has taken a big beating lately in the court of public opinion. It also might be possible to find a legislative fix for some of their problems that could be traded for their participation in a Phase II failures program. And steps could certainly be taken to protect the confidentiality of some of the information about the compounds in question, at least for a time. We can find the right incentives if we try.

So I want the Phase II failures. I REALLY want the Phase II failures. I want them for my own research and for your research. I want them because they could make a difference for a host of unmet medical needs. And here’s my last question, aimed at patient advocacy groups and scientific societies and medical school deans and biotechnology associations and government officials everywhere: who wants to help me get them?


1. Petsko, G. A. (2010) The Devil’s in the Details. Genome Biol. 11, 117.

Gregory A. Petsko ( is the Gyula and Katica Tauber professor of biochemistry and chemistry at Brandeis University.

* This article originally appeared in BMC Biology (2010) 8, 61 and was reproduced with permission from BioMed Central.

NEXT PAGE 1 | 2 | 3

First Name:
Last Name:

Comment on this item:
Our comments are moderated. Maximum 1000 characters. We would appreciate it if you signed your name to your comment.


There aren't any comments on this item yet. Tell us what you think!


Page 1 of 1

found= true898