About this Author
DBL%20Hendrix%20small.png College chemistry, 1983

Derek Lowe The 2002 Model

Dbl%20new%20portrait%20B%26W.png After 10 years of blogging. . .

Derek Lowe, an Arkansan by birth, got his BA from Hendrix College and his PhD in organic chemistry from Duke before spending time in Germany on a Humboldt Fellowship on his post-doc. He's worked for several major pharmaceutical companies since 1989 on drug discovery projects against schizophrenia, Alzheimer's, diabetes, osteoporosis and other diseases. To contact Derek email him directly: Twitter: Dereklowe

Chemistry and Drug Data: Drugbank
Chempedia Lab
Synthetic Pages
Organic Chemistry Portal
Not Voodoo

Chemistry and Pharma Blogs:
Org Prep Daily
The Haystack
A New Merck, Reviewed
Liberal Arts Chemistry
Electron Pusher
All Things Metathesis
C&E News Blogs
Chemiotics II
Chemical Space
Noel O'Blog
In Vivo Blog
Terra Sigilatta
BBSRC/Douglas Kell
Realizations in Biostatistics
ChemSpider Blog
Organic Chem - Education & Industry
Pharma Strategy Blog
No Name No Slogan
Practical Fragments
The Curious Wavefunction
Natural Product Man
Fragment Literature
Chemistry World Blog
Synthetic Nature
Chemistry Blog
Synthesizing Ideas
Eye on FDA
Chemical Forums
Symyx Blog
Sceptical Chymist
Lamentations on Chemistry
Computational Organic Chemistry
Mining Drugs
Henry Rzepa

Science Blogs and News:
Bad Science
The Loom
Uncertain Principles
Fierce Biotech
Blogs for Industry
Omics! Omics!
Young Female Scientist
Notional Slurry
Nobel Intent
SciTech Daily
Science Blog
Gene Expression (I)
Gene Expression (II)
Adventures in Ethics and Science
Transterrestrial Musings
Slashdot Science
Cosmic Variance
Biology News Net

Medical Blogs
DB's Medical Rants
Science-Based Medicine
Respectful Insolence
Diabetes Mine

Economics and Business
Marginal Revolution
The Volokh Conspiracy
Knowledge Problem

Politics / Current Events
Virginia Postrel
Belmont Club
Mickey Kaus

Belles Lettres
Uncouth Reflections
Arts and Letters Daily
In the Pipeline: Don't miss Derek Lowe's excellent commentary on drug discovery and the pharma industry in general at In the Pipeline

In the Pipeline

« How Much Is Wrong? | Main | A Tool Compound's New Personality »

January 7, 2014

Adoptive T-Cell Therapy for Cancer: The Short Version

Email This Entry

Posted by Derek

If you're looking for a good overview of the modified-T-cell immune therapies for cancer that have been making such headlines recently, this writeup at Nature is for you. It also makes clear that getting this to work across a broader number of tumor types is not going to be a job for the faint of heart. If the antigen that you're training all those T-cells to go after happens to exist somewhere else in the body (other than on the surface of the tumor cells), or if you have some cross-reactivity that you hadn't realized, the patients are going to be in for a very hard time indeed.

Designing effective treatments requires finding cell-surface antigens for T cells to target without damaging normal tissue. This specificity may prove more difficult than researchers initially thought. Many antigens found in cancer are also expressed in normal tissue — HER2, for example, which is the target of the antibody-based therapy trastuzumab (Herceptin), is also expressed in heart cells. Before researchers can make progress, they must understand how extensively each candidate target is expressed in all tissues of the body.

Recent studies have highlighted what can happen when T cells unexpectedly attack normal tissue. In a clinical trial of TCR-engineered T cells, researchers at the US National Cancer Institute were targeting the cancer-specific antigen MAGE-A3 when two of their nine patients slipped into a coma and died. It turns out that the cells also recognized another member of the MAGE-A family that the researchers later discovered is expressed in low levels in brain tissue. Another type of MAGE-A3-specific TCR caused two patients to die from heart failure when the TCR bound to a similar protein, called Titin, which is expressed on heart cells. Adaptimmune, the company based near Oxford, UK, that developed the T-cell receptor, has implemented more extensive safety testing techniques in an attempt to prevent unexpected reactions in the future.

I'm quite excited about this whole field, but there are very strong reasons why it's been tried (so far) almost entirely on people who are near death. The immune system is to be feared.

Comments (15) + TrackBacks (0) | Category: Cancer


1. johnnyboy on January 7, 2014 2:39 PM writes...

In the development of therapeutic antibodies, the preclinical safety work includes tissue cross-reactivity studies, in which you look at whether the antibody binds unexpectedly to any other tissue than the target. A similar approach should probably be carried out for these T-cell methods, particularly if you are somehow bothered by killing people with your approach...

Permalink to Comment

2. Hap on January 7, 2014 2:55 PM writes...

I wouldn't have thought that "Get started living or get started dying" would be quite so immediate. Though it's cancer, and it seems like if there's no chance the treatment can't kill you, it won't work.

Permalink to Comment

3. cynical1 on January 7, 2014 3:30 PM writes...

The question I have is why the treatment was cross reactive in 2/9 as opposed to 9/9 with the MAGE-A3 therapy. Were these two patients the only ones who expressed the other MAGE-A family protein in their brains? I sort of doubt it.

I suspect that when this approach runs its course, we might learn as much about autoimmunity as we do about cancer.

Permalink to Comment

4. Brett on January 7, 2014 3:52 PM writes...

I figure that's what it will eventually be if things work out in the best possible way (i.e. immunotherapies work well and wipe out most forms of cancer down the line). 2-3 weeks of hell from your immune system going into overdrive and killing tumor cells, followed by cancer-remission status.

Permalink to Comment

5. pete on January 7, 2014 7:08 PM writes...

Differences in vascular leakiness related to differences in tumor stage/invasiveness could be contributing to the safety problem here. That is, perhaps the unfortunate 2/9 patients had more significant disruption to their blood-brain barrier, thereby leading to greater brain immune-exposure as compared to the other patients.

Similarly, variable leakiness of the vasculature in other organ systems (due to in variation cancer phenotype, location, severity) might expose other tissues to surprising, unwanted immune attack that might not have been anticipated from drug safety pre-clin. & clinical studies.

Very exciting approach(es) - but could be a tough road, indeed.

Permalink to Comment

6. qetzal on January 7, 2014 9:47 PM writes...

Keep in mind these are autologous cell products, created individually for each patient by gene modifying their T cells ex vivo so they'll recognize the target antigen. So each product is different, and may have different potency. Sort of like individual variability squared.

Permalink to Comment

7. The cooler king on January 7, 2014 10:02 PM writes...

"The immune system is to be feared."
True enough, but so is FOLFIRINOX and some other current standard of care.

Permalink to Comment

8. Thoryke on January 7, 2014 10:15 PM writes...

I'm reminded of the cytokine storm problem encountered in the TeGenero trial of TGN1412:

How much can in silico or lab-on-a-chip testing help avoid these situations?

Permalink to Comment

9. mmmorgoth on January 8, 2014 4:52 AM writes...

Is it possible to have a drug that binds to two targets? I remember reading something like that on this site.

That way you only kill hose cells that have both, or using some tricks one but not the other.

Permalink to Comment

10. John Schilling on January 8, 2014 10:56 AM writes...

I understand why the first reaction is to try this sort of therapy only on people who are already near death. But it seems to me that the therapy would be far more useful to people who are still fairly healthy. Less likely to kill them, in that their bodies are better capable of withstanding an immune system gone haywire, and more likely to be essentially cured in that their cancer has not yet developed the broad genomic complexity that seems likely to provide a reservoir of resistant tumor cells without the target antigen.

Not that this is a problem limited to adaptive T-cell therapy, of course.

Permalink to Comment

11. RickW on January 8, 2014 11:57 AM writes...

The unique thing about TCR engineering, as opposed to monoclonals or even the CD19 engineered T cells, is that these TCRs will behave differently in every individual. Natural TCRs are selected based on an individual's HLA type and are subjected to multiple forms of immune tolerance induction. All of that is bypassed when you plug in an engineered TCR. No amount of preclinical testing will solve this.

Maybe this could be fixed by testing each patient with an engineered T cell that is crippled in some way but can be followed for undesired reactivity, or by adding a kill switch that works very quickly.

Permalink to Comment

12. cynical1 on January 8, 2014 12:50 PM writes...

@ pete: Perivascular infiltration of T-cells into the CNS takes place all the time and is part of a healthy immune response. The CNS is definitely not the immune privileged compartment that most of us were taught way back when. So if you eat some salmonella at lunch, your immune system is going to go looking for it in your brain. If the target antigen is there, then you're hosed. (And even if it's not but those T-cells confuse something else for it, you're still hosed.) However, if they don't find what they're looking for, those T-cell specific responses dissipate over a 24 hour period and disappear from the CNS. So, vascular leakage doesn't explain the 2/9. The other 7 also had a MAGE-specific T-cell response directed into their CNS. The question is whether it was a difference in antigen presentation or protein expression to the cross-reactive MAGE-related protein. I'd probably start by looking at their MHC-class 2 haplotypes and see if those 2/9 differed from the other 7 assuming that everyone has low-level expression of that protein. But I'm a chemist, not an immunologist.

Permalink to Comment

13. anonymous on January 8, 2014 2:56 PM writes...

@8 Thoryke - Not that it matters, but WBC count in the article itself (figure 3) is off by 10^6 (one-thousandth of a cell per microliter - why not?) and by 10^3 in supplementary.

Permalink to Comment

14. qetzal on January 8, 2014 3:27 PM writes...

@ RickW:

Engineered TCRs are still HLA restricted. In the NIH MAGE-A3 study, for example, all the patients were HLA-A*0201 to match the specificity of the TCR. You wouldn't give HLA-A2 restricted TCRs to a patient who was not HLA-A2 positive.

@ cynical1:

Since the T cells in that study were all acting through HLA-A*0201, I don't think MHC haplotypes are a likely factor. There were plenty of other differences that could be relevant. The 2 patients who died plus the 1 additional patient with CNS tox received more T cells than any other patient. Most of the trial patients had melanoma, but one of the deaths was a patient with esophageal cancer. Who knows how different their medical histories were.

And again, the cell products themselves were variable, due to uncontrollable aspects of the patient-specific manufacturing. Since each patient's starting cells are different, the final products from two different patients will vary much more than, say, two different lots of a therapeutic antibody. In this specific case, the study authors noted that the products given to the 3 affected patients had higher percentages of so-called naïve T cells.

Bottom line, lots of things could have contributed.

Permalink to Comment

15. RickW on January 8, 2014 4:50 PM writes...

@ qetzal,
Sorry, I wasn't clear. The problem is the other HLA alleles present in the patient. The engineered TCR hasn't been through negative selection in that patient, and therefore could be alloreactive (like graft versus host disease) or reactive against another similar peptide presented by allo MHC. I'm not sure this explains any of the negative consequences seen so far, but it seems impossible to rule out for future patients.

Permalink to Comment


Remember Me?


Email this entry to:

Your email address:

Message (optional):

The Last Post
The GSK Layoffs Continue, By Proxy
The Move is Nigh
Another Alzheimer's IPO
Cutbacks at C&E News
Sanofi Pays to Get Back Into Oncology
An Irresponsible Statement About Curing Cancer
Oliver Sacks on Turning Back to Chemistry