This study is a case study from this study (lol):

Chimeric Antigen Receptor-Modified T Cells in Chronic Lymphoid Leukemia

I imagined this scene happened in one of their lab meetings:

I didnt go in-depth as to how everything was working before, so I thought I would do it now!

This paper is a case report of an individual who had a B-cell cancer. The patient had tried everything and anything recommended to him by scientists (no mention in the report of him disappearing for 6 months to Mexico for ‘vitamin c’ therapy, or whatever stupid thing wooers push to ‘cure’ cancer) and he was still dying. Fast.

So he agreed to an experimental procedure designed by scientists.

Fact #1– B-cells and especially B-cell lymphomas express a protein, CD19.
Humans will not have antibodies to human CD19 (your immune system normally has ways of protecting itself from autoimmunity), but its really easy to get mice to make anti-human-CD19 antibodies.

Theoretically, these antibodies could be humanized (humans dont like mouse antibodies), produced industrially, and infused into cancer patients as a treatment, as the antibodies would tag the cancer for destruction.

In fact, that is exactly what a drug, Rituximab, does (it targets a different protein, though). And, unfortunately, this patient tried that, and it didnt work.

Fact #2– T-cells could also be used to kill cancers (technically, ‘persuade the cancer to commit suicide’).
T-cells have receptors on their surface that ‘read’ the proteins presented in cells MHC molecules. Normally, if a cell is infected with say, a virus, the cell will put up ‘VIRUS’ flags on their MHC molecules, and a T-cell will cruise by and say ‘Hey… that aint right…’ and kill the virally infected cell (MHC posts on ERV).

The problem is, cancers are basically *you*, just you gone wrong. And again, your immune system has ways of preventing autoimmunity. And, everyones MHC proteins are different– even if we could theoretically get your T-cells to attack cells making CD19, the flag my cells put up in my MHC would be different than yours. Treating a cancer this way would have to be super personalized based on individuals genetic profiles. HUGE pain in the butt, if not unwieldy to the point of being impossible except for the most common MHC types.

Solution– Combine Fact #1 and Fact #2 into one treatment.
Scientists in this study took the ‘arms’ off of mouse anti-CD19 antibodies (F(ab)2), and attached them to signaling domains normally found around the T-cell receptors (CD3 and a CD137 just to make the signal stronger).

So a T-cell expressing this protein would have the arms on an anti-B-cell-cancer on the outside, with the super powerful ‘KILL THIS CELL!!’ signaling domain on the inside. If the outside comes into contact with a cancer cell, the T-cell will kill it, and we wont have to deal with any of the issues of ‘my MHC is different than your MHC’!

But how the hell are you supposed to get a T-cell to express this kind of thing? Half antibody/Half TCR isnt normal at ALL. Well, they took the genetic information required to make this chimera (half super-specific anti-cancer antibody, half super powerful anti-cancer signaling domain) and put it into HIV-1. Well, a gutted out HIV-1 that can only reverse transcribe and integrate– so, when one of these anti-cancer-HIVs infects a cell, it will permanently integrate the genes necessary to fight the cancer, NOT ‘HIV-1’. You physically cannot get ‘AIDS’ from this kind of therapy.

But the cells that kill infected cells are CD8+ T-cells.

HIV-1 likes to infect CD4+ T-cells.

To get around this, they put on the outside of the anti-cancer-HIV molecules VSV-G. Know how viruses have ‘keys’ that only fit in certain ‘locks’ on the surface of your cells? Some cells cannot be infected with certain viruses because they do not have the right ‘lock’ for the viruses ‘key’? VSV-G is a skeleton key. Gets into everything.

SO! Here is what the scientists did–

  • They took blood cells out of the cancer patient.
  • They purified the T-cells from the blood cells.
  • They infected those blood cells with an HIV-1 virus with VSV-G on its surface, that contained the genetic information for making an anti-cancer antibody/TCR.
  • They then put the T-cells back into the cancer patient.
  • The T-cells with the chimeric receptor killed the cancer.
  • All of it.
  • It also killed all of his non-cancerous B-cells (dude has no antibodies) but he is not having any recurrent infections or anything weird. Hes gardening and hanging out and doing just fine. No cancer. Not dead.

If Ive said it once, Ive said it a million times:



  1. #1 Usually Anon
    September 15, 2011

    Neat. 2 questions: Will the number of modified CD8+ T-cells ever (within a normal human lifetime) drop low enough to allow B cells to survive; and did the therapy kill all the Progenitor B cells? (In other words, can that part of his immune system eventually grow back?)

  2. #2 CarlosT
    September 15, 2011

    What are the chances of further refining this approach so it kills only the cancerous cells?

  3. #3 ERV
    September 15, 2011

    Usually Anon– Well, they have tried things like this before, and the ‘problem’ has been that the modified cells havent lived very long. In those instances, the healthy cells have bounced back, but then there is always a risk the cancer will too. I think this is the first time they have gotten a super strong, maintained response like this.

    But yes, if these guys were removed, Im pretty sure the B-cells would bounce back, but I think the big problem would be the loss of memory. The B-cell that ‘remembers’ measles from the MMR vaccine you got when you were a baby? Dead. Your immune system would have to relearn everything if this treatment really is slaughtering the B-cell population.

    CarlosT– Its going to be hard, and maybe based on luck, a bit. Like I said, the problem is, cancer is *you*. Any protein a tumor is making can be found in normal tissue somewhere. But maybe, some tumors express something that is only found in say, embryonic tissue, so its not around anymore in a normal adult– that would be a super target (unless you were a female who wanted to get pregnant some day). But if a tumor isnt expressing a protein like that, I think it will be really hard to avoid bystander effects…

  4. #4 Yuri
    September 15, 2011

    Long-time reader here, and fellow HIV researcher, though my immunology knowledge is lacking thanks to the structure of my program, which leads me to ask: any thoughts on how, if this therapy consists of sustained T-cell growth (um, isn’t that cancer?) which kill B-cells, 1) how are the cells not poliferating to the point of being cancer-like (i.e. damaging), and 2) how is the patient A-OK with no B-cells? O_o

  5. #5 mo
    September 15, 2011

    I’m not sure, but I think B-memory-cells don’t express CD19. Should do a literature dive. If that’s true these modified cells would only kill some of immature types and fully developed B-cells like effectors – And of course the memory cells that differentiate to effectors in the case of a recurrent pathogen, depleting the specific memory cells in the case of the persistence of the T-cell-vaccine šŸ˜

    I wished they had combined their approach with this holy-shit-I-can’t-believe-that’s-working approach:

    1) they should behave like normal activated T-cells, killing their target until target is gone, then dying with some of them maybe becoming memory cells.

    2) Antibodies are long-lasting, so the patient should be fine for some months/years. Problems will probably come later.

  6. #6 mo
    September 15, 2011

    Ok, I was wrong, memory B-cells do express CD19 and are thus killed. BUT! Fully differentiated plasma cells downregulate CD19. Duh, makes sense for their mode of action.

    So if the vaccine persists: no new adaptive humoral immune functions for you, no memory response, but decades of antibodies against all the stuff you encountered earlier.
    Seems like a sensible compromise to death for someone in their fifties, especially because his “new adaptive immune functions”-fu will be weak by then and the T-cells are more important anyway.

  7. #7 fnxtr
    September 15, 2011

    Effin’ brill.
    I, too, am wondering about the effect of having no B-cells. The old man in the plastic bubble?

  8. #8 Paul
    September 16, 2011

    It’s a fascinating study.

    The Weizman Wave had a post on some of the early work – way back in the 1980’s – that paved the way for this trial

    while I have contributed my own tuppence worth at

  9. #9 the cynic
    September 18, 2011

    Wow. This is the sort of idea so crazy, so out of the box, that although it works in theory, you never expect to actually see it.

    My only reaction: this is effing cool!

  10. #10 slothrop
    September 20, 2011

    I’ve read that they will give these patients intravenous immunoglobulin to deal with the loss of their B-cells:

    “The treatment wiped out all of the patientsā€™ B-cells, both healthy ones and leukemic ones, and will continue to do for as long as the new T-cells persist in the body, which could be forever (and ideally should be, to keep the leukemia at bay). The lack of B-cells means that the patients may be left vulnerable to infection, and they will need periodic infusions of a substance called intravenous immune globulin to protect them.

    So far, the lack of B-cells has not caused problems for Mr. Ludwig. He receives the infusions every few months. He had been receiving them even before the experimental treatment because the leukemia had already knocked out his healthy B-cells.”

  11. #11 The Math Skeptic
    September 21, 2011

    That is pretty amazing work, using HIV to kill cancer. However, I can’t help imagining it in the pre-credit sequence of a post-apocalyptic zombie movie as the “thing that went utterly wrong.”

New comments have been temporarily disabled. Please check back soon.