I am skeptical that CAR-T therapy will go anywhere with solid tumors. Blood cancers and solid tumors are extremely different - they are basically different dise

I am skeptical that CAR-T therapy will go anywhere with solid tumors. Blood cancers and solid tumors are extremely different - they are basically different diseases. First, blood cancers have very few mutations (< 10); solid tumors may have tens of thousands. This means that blood cancers are a lot less heterogeneous than solid tumors. CAR-T cells are basically like a very specific kind of gun - you can say, "kill anything with protein X". But we've tried targeted therapies in solid tumors; they don't work, because the tumors usually have enough genetic heterogeneity that they can just evolve around the therapy. This is likely to be the case with CAR-T therapy as well. Solid tumors also have the advantage of not being in the blood (obviously) - this makes them much harder to target with anything, especially circulating cells, which would have to extravasate and infiltrate the tumor itself. Blood is very accessible - you can just inject stuff into it. Heck, we have machines that can take blood out of the body, spin out whatever fractions we want, and put it back in all in one go. Finally, blood cancers also have no microenvironment; they are circulating in the blood. A solid tumor can recruit and reprogram all sorts of quisling cells to cloak itself. These cells do all kinds of immunosuppressive things that we don't understand. The main problem with the solid tumor is its complexity and its adaptive potential. CAR-T cells are a magic bullet - but the magic bullet can only kill one thing, not the many-headed thing that is a tumor. To beat the solid tumor I believe we need to fight it with something that has the adaptive potential of the tumor itself - i.e., the adaptive immune system. For this reason I think we're more likely to see success by recruiting the native immune system. So far we've done this via checkpoint blockade. This also has its limits - unhooking immune checkpoints is like unchaining a mad rottweiler. It might kill the tumor, but it might just as well kill you. Where I'd like immuno-oncology to go is understanding the mechanisms of immune evasion (i.e., what the tumor is doing to suppress the immune system, which should be mopping up cancerous cells, but isn't) and suppressing those. That way we can recruit the immune system to do its job without also causing it to damage the rest of the body. But we don't even have an understanding of the immune makeup of the tumor yet, let alone how the tumor is recruiting and reprogramming immune cells to suppress the immune response. If we get there I think we'll see some real progress, but it will take a lot of work. It remains to be seen if we have the patience to do that work before we give up and move on to the next drug fad.