The Novartis/Penn team working on CAR-T therapies reports a long-term follow-up on their patients (well, as long-term as its possible to get in this field). Here’s the paper, here at Science Translational Medicine, and here’s the press release. The upshot is, that if you’re cured by this technique, you seem to stay cured. The first patient to receive the treatment has now gone five years with no sign of leukemia.
Out of the 14 patients in that trial, 8 responded, and four of those appear to have been complete remissions. There were four partial responses (with effects measured in months), and six patients who did not appear to respond at all. In those, the modified T-cells did not seem to expand in vivo, which would account for the lack of response. (If they had expanded and still nothing happened, that would have been a puzzle indeed). The key is to figure out why the treatment failed in these cases (so far, one gets the impression that this is a mystery), and how to fix that. Combining this technique with other therapies hasn’t really been tried yet, either.
So there’s definitely something major here, and definitely room to improve it. That’s what Novartis, and Juno, and Kite (and who knows who else) are frantically working to do right now. The biggest prizes are to extend this idea to more tumor classes, and to make it scalable to larger numbers of patients. The second one is a big challenge, and the first is even bigger. People have been antigen-hunting for quite a while now, looking for something that’s fit to turn the immune system on in other types of tumors, and it’s slow going. But you have to be sure, because if the immune system can strip a couple of pounds of aberrant white blood cells out of your system, so vigorously that it nearly overloads your kidneys, then it can do the same with lots of other tissues, too, some of which you might prefer to keep. Ripping out a couple of pounds of motor neurons or Kupffer cells would be suboptimal, to say the least.