Sunday, September 16, 2012

Genetically Modified T-Cells for CLL (and evetually NHL)

(Sept update of prior post after I heard Carl June present at German CLL conference last weekend)

Last summer there was quite a splash when researchers at University of Pennsylvania used genetically manipulated T-Cells to fight off CLL.  The studies were simultaneously published here in New England Journal of Medicine and here in Science.  This past weekend, I had the opportunity to hear Carl June update us on the progress of this technology so I wanted to update this post in light of what I learned.

Looks like Novartis is going to purchase the technology and in my mind that is the best possible outcome for making this move more quickly.  While I love the science done in academic settings - those settings are not designed to bring these breakthroughs to patients on a broad scale.  I know many of the folks at Novartis and they are very good at advancing technology quickly.


I've had a lot of patients ask me about these studies so I thought it might be interesting to describe the technology at play.

The "brains" of the immune system is probably the T-Cells with the B-Cells being second in command.  T-Cells help coordinate many aspects of immunity and heavily rely upon their "T-Cell Receptor" (TCR) which is unique to each individual T-Cell to figure out when to engage a target or not.  B-Cells also have a "B-Cell Receptor" aka. antibody / (BCR).

We've gotten pretty good making artificial antibodies and they are a good therapy (rituxan, GA-101, herceptin, ofatumumab, etc).  So far we have no idea how to make a good T-Cell Receptor.  Much to my surprise though, you can take the part of the BCR that hangs out outside of a cell and the part of the TCR - stick them together as a "fusion gene,"  and the thing works as a "chimeric receptor" (I forget the greek story of the half lion, half human - called a chimera - I may also have my story completely wrong but the idea is the same).

Now you can make your artificial B cell receptor - plug it onto a T-Cell receptor and make T-Cells do things you want them to do - like get rid of cancerous B cells (aka CLL/NHL).  There were a number of other key features (costimulatory molecules, lymphoid depletion, etc) but for now it is reasonable to focus on the main technology.

Viola - it works!  The have now treated about 11 patients at Penn.  Most were CLL but a few were Acute Leukemia.  If it works in ALL, that is very interesting.  That disease is extremely aggressive and new therapies are needed there even more than CLL.

But there are potentially problems.
1)  To get this new T-Cell chimeric receptor into the T-Cell, you need to infect the cell with a virus that has some resemblance to HIV.  It is heavily modified, but you can get the picture, there is need for caution.

2) So far, many of the patients who respond favorably to the therapy undergo a very medically frightening reaction when the T cells attack the B cells.  In some cases it has been life threatening and required ICU level support.  As if that is not bad enough, you cannot predict when this will happen.  In some patients it happens a few days later, in others, it may be 3-4 weeks later.

3)  The manufacturing of the genetic modified T cells may be finicky and it is possibly the reason this does not work in all patients - when they try to achieve scale to make this work for lots of patients, this may be a big problem.

4)  If this works, it permanently depletes all health and cancerous B cells.  So much for B cell immunity.  Furthermore, it looks like plasma cells (the professional antibody producing cells) also slowly disappear.  Therefore you may need monthly IVIG shots for the rest of your life.

5)  What happens if the T-Cells and B-Cells fight to a draw?  Do the T-Cells keep proliferating?  What happens then?  Right now there is no "off switch."  You might be able to program in a "suicide gene" but that was not a part of this treatment.  Talking with some folks at the conference, it sounds like that may be a future consideration with these treatments.

6)  Gene therapy periodically causes a whole different kind of cancer - a really nasty one.  Hard to exchange a slow disease for a chance of a  fast one unless that slow one has ran out of other good options.  They are using a "lentivirus" platform instead of a "retrovirus" platform.  They feel this minimizes this problem but I'm not convinced.  Either way, the FDA requires that patients get observed for 15 years since that is the standard for "gene therapy" studies.

7)  Timing - this is not likely to be available for another 5-10 years at best.  In the meantime, new CLL drugs like ABT-199, ibrutinib, CAL-101, GA-101 and so forth are making CLL such a different disease that it may render this approach less interesting.  It could still be extremely valuable in NHL where we need more breakthroughs.

So who will this be appropriate for?

In CLL, I would send someone for this if they required a stem cell transplant.  This may be quite a bit better than getting someone else immune system as it will get rid of the graft vs host disease.  This may be great therapy for the patient with NHL - particularly DLBCL that has relapsed.  Furthermore, it may make some older patients eligible for immune intervention who may otherwise be ineligible for a transplant.  I would NOT think this is a good front line therapy for anyone - not any time soon at least.

Many of these limitations may have answers and I am excited that a company is willing to invest in the idea - it often moves ideas a lot faster than the grant cycles of the universities.  Hopefully this goes someplace exciting but for now it is still a ways off for mainstream CLL treatment.  Huge gratitude to the brave patients who are the first ones on this experimental therapy - they may be paving the way for many others.  Will keep you posted as I hear more....