Personalized Medicine Relies on Animal Research Too


Dr. Greek writes about personalized medicine as if one could do this work without relying on animal research at all.


For example, he writes:


When will personalized medicine become a reality?

We are already seeing it, with breast cancer being a prime example. Breast cancer treatment is now determined in part based on a patient’s genetic makeup. About 25-30 percent of breast cancer patients overexpress the HER2 oncogene, which is a gene involved in the development of cancer. The overexpression results in an increase in the replication of the cancer cells. Physicians are now able to identify which breast cancer patients overexpress HER2 and give them Herceptin, a monoclonal antibody that inhibits HER2.


This is true..  but where did Herceptin come from?   Does he know?


The basic research that led to the development of Herceptin goes back to work by Milstein and Kohler who discovered the potential for using antibodies to fight disease.    They developed the first methods to produce monoclonal antibodies using mice.   Both Milstein and Kohler went on to win the Nobel Prize partly for this work.


Harold Varmus (now back as Director of the National Cancer Institute) showed that disturbances in some gene families could turn the cells cancerous.  He also went on to win the Nobel Prize for this work.  Robert Weinberg subsequently discovered in rats that a mutant gene (named “neu”) encoding a tyrosine kinase produced cancer features in cells.   


Later, Axel Ullrich and collaborators at Genentech cloned the human HER2/neu gene.  Work at UCLA Dennis Slamon and colleagues showed HER2 over-expression in 25% of patients with aggressive breast cancer.


The group at Genentech then developed the mouse 4D5 (parent of Herceptin) show that 4D5 could suppress the growth of HER2 tumor cells as well as enhance the ability of the host immune system to kill them.   A collaboration between UCLA and Genentech then demonstrated that radio-labeled 4D5 localized HER2-expressing tumors in human patients.


With these information obtained from animal experiments, Genentech created Herceptin by humanizing the 4D5 mouse antibody directed at HER2.   


So, you see..  Herceptin was derived from a mouse antibody.


Let me repeat: a mouse antibody!


Clinical trials in humans subsequently showed the effectiveness of Herceptin to treat HER2 positive breast cancer.


Perhaps, Dr. Greek and other animal rights activists should carefully listen to the experts that were actually involved in the process of developing Herceptin (a drug he appears to thinks highly of) which, indeed, benefits so many women battling breast cancer.   A drug derived from mice.

Here is what Robert Weinberg had to say about Dr. Greek’s views on research:

“Dr. Greek says the silliest things, [..] implying that people are not studying human tumors, and implying that the kinds of experiments that one can do in mice can be done as well in humans -- truly mindless! “


I couldn’t have said it better.


Quote: “Dr. Greek says the silliest things, [..] implying that people are not studying human tumors, and implying that the kinds of experiments that one can do in mice can be done as well in humans -- truly mindless! “

The skepticism expressed at this quote, are you saying it is more effective to study human cancer in mice? Don't mice and humans react differently to both disease and medication?

Hasn't the BMJ British Medical Journal (2006, December 18) come out saying 4/5 animal experiments are not only useless, but dangerously useless, finding not only flaws in the way the testing was done, but also fraud, and where animal tests are done, human trials are also conducted at the same time, which would imply that human trials are conducted irrespective of animal testing, or else wouldn't these researchers wait until the animal tests were in before moving on to humans? Or else what is the point of animal testing?

RedGlitter, one of the flaws in Pound et al's analysis was that they assumed that all animal studies are performed prior to clinical trials. While most animal studies are carried out prior to clinical trails it is not uncommon for subsequent animal studies to examine use of the therapy for other conditions , modified versions of the therapy etc. Sometimes they are performed to examine in more detail aspects of the therapy that were identified in clinical trials, for example particular responses, adverse reactions etc.

There's no need to invoke fraud, indeed the people performing these studies could hardly hide the fact that the drug is in clinical trials (or clinical use) from those reviewing their grant applications.

RedGlitter, the BMJ article you refer to (presumably Pound P. et al (2006) doi:10.1136/bmj.328.7438.514) was itself flawed in some of its conclusions. The 8 reviews of 6 treatments that the authors examined were all performed because the treatments in question had failed in clinical trials. It's a bit like analysing flight safety only by examining air crash investigation reports..sure it's useful but hardly the whole picture. In all cases when the results of the pre-clinical animal studies that were performed were pooled and reviewed carefully the outcome of the animal experiments actually agreed with the poor outcome in the human clinical studies.

While the authors of the BMJ piece do suggest that animal experiments are not reliable the actual evidence they present indicates that the animal experiments were reliable, but that poor design (small underpowered studies being a key problem) and poor subsequent analysis of the pre-clinical data resulted in the animal studies being misinterpreted as supporting the decision to proceed to clinical trials when in fact they did not.

One of the things we have learned for studies such as that by Pound et al. is that most of the apparent failure of animal studies have been due not to any inherent weakness in the animal models themselves but due to flaws in the design, interpretation and reporting of experiments, just as was the case with clinical trials before the Cochrane collaboration started to review clinical trial design.

While I don't agree witheverything Pound et al say I'll give them credit for one thing, they have provided the impetus for animal researchers to toughen up standards for experimental design, similar to the tightning of clinical trial rules a few years ago, the new ARRIVE guidelines are a good example of this. Guidelines such as these should ensure that in future animal research makes an even greater contribution to medical progress.

I'm afraid I couldn't find RedGlitter's BMJ reference in the issue cited, so please be more accurate when attributing such quotes to the BMJ. There was an interesting article in that issue on sword swallowing though..
Sword swallowing and its side effects
Brian Witcombeand Dan Meyer
BMJ2006;333:1285 doi:10.1136/bmj.39027.676690.55
Scientific rigour really needs to be supported by solid evidence from an analysis of studies 'for' and 'against' a hypothesis. Having an opinion and backing that up with one (not accurate) reference is hardly good practice!