[cancer] Some science neepery on what we’re trying to do here at NIH

A week or two ago, my friend Chris Johnson asked me a question via email.

Did the NIH researchers describe how the hell they managed to count the number of distinct mutations among the cancer cells? I was amazed by that. I was at least as amazed by their ability to customize your T cells for each of those strains. How the hell do they do that stuff? In my ignorance, I’d’ve sworn the technology to do that either didn’t exist, or, at least, that it would be so slow and difficult that it couldn’t be employed in practice. I love being wrong, and finding the world is more amazing than I suspected, but if you’re able to shed any light on those processes in future postings, I’m all eyes.

My not-necessarily-scientifically accurate answer was sufficiently interesting to me that I’m reposting it here, as it may be of interest to some of you reading. If you’re a clinician or a researcher who can clear up some of my misconceptions, please feel free to comment.

The mutation count comes from comparing the Whole Genome Sequencing (WGS) data on my tumor tissue to the WGS data on my germline tissue. In and of itself, that’s not terribly hard if you have the right data in hand in the first place, which we’ve possessed since last spring thanks to the crowd-funded process I went through then.

The part that is terribly hard is selecting for the appropriate TIL cells. They’re sensitizing my B cells (recovered through apheresis) to the known mutations through a process which I cannot even begin to describe. I’m not sure if they’re culturing all 100+ known mutations, or just a subset which is likely to be more key.

They will then have separate B cell cultures where each of those mutations is active, which would normally be detected through the presence of a specific protein. (In a very fundamental sense, that’s what any mutation does, it alters the production of some specific protein, which in turn can alter enzymes, cell structures, etc.)

At that point, they introduce my TIL cells, cultured from the tumor harvest, into each B cell culture, and note which B cell cultures stimulate the most TIL cell activity. They then retrieve the TIL cells from that culture and set them aside for the expansion growth, which is the 10^7 or 10^8 multiplication of TIL cells required before infusing them back into me.

The theory being that the subset of my own TIL cells used for the expansion growth and subsequent infusion will be the ones most likely to seek out tissue in my body harboring the tested-for mutations. That is to say, mutations found only in my tumor tissues.

This makes a couple of assumptions. One, that all my tumor tissues carry the same mutations. Since I have had at least five or six generations of tumors in my body, each of which may have experienced its own genetic drift through further mutation, this is not necessarily true, though it is likely enough. Generation 1 was the original presentation of my primary cancer, resected in 2008. Generation 2 was my initial metastatic presentation, resected in 2009. Et cetera, until the crop of tumors I have now which represent Generations 5/6 or 6/7, depending on how you count.

Two, this process assumes that those mutations can be found nowhere else in my body. The germline tissue which was typed was from my liver. However, there is increasing evidence suggesting that human DNA can be far more chimeric than we normally understand it to be. So, for example, what if my colon tissue, where the cancer originated, harbors some of those same mutations, even though the liver tissue from which my WGS assay was done did not? Would the sensitized TIL cells attack otherwise healthy colon tissue in that case, creating a permanent gastrointestinal autoimmune disorder for me?

Issues such as those I mention in the assumptions above are why this is a highly experimental procedure, rather than normal clinical medicine. That and the fact that the hard cost of this process is astronomical, and can probably not be duplicated at more than a handful of other institutions in the world, if that. On a purely procedural basis, much of what’s being done to me could be handled by any competent institutional oncology department, but the whole business with the genetic selection and managing the TIL cell growth requires a level of expertise and equipment that is incredibly expensive and vanishingly rare. Also, the holistic and theoretical basis for managing the treatment and evaluating the outcomes requires a similar level of expertise.

4 thoughts on “[cancer] Some science neepery on what we’re trying to do here at NIH

  1. Dan Gollub says:

    I read your link salad link about the possibility of exploiting the metabolic quirk of cancer and it gave me an idea. Disable the anerobic mechanism via some agonist for glucose, perhaps. The subject in the experiment will get very sick. But perhaps the cancer cells will die, and then the anaerobic process can be restored. I’ll write a story about this.

    1. Jay says:

      Good luck with that! Sounds interesting.

  2. Guess says:

    Great post. I Had to google what TIL Cells and B cells were and read it a couple of times to keep up.

    Is this simiiar to other studies where they train White Blood cells to attack cancer cells? I sent you a link to a New York Times article about this.

    you can get through the new york time paywall (and all paywalls) by going to news.google.com and searching for it there then following the link.

    just do a search on this link without the http from google news.


    You can get through paywalls to all newspapers by going through google news.

    1. Jay says:

      Yep, pretty much that ‘training white blood cells’ thing, though there’s multiple paths towards that goal.

Comments are closed.