These engineered viruses are remarkable feats of biotechnology.
I have to wonder if they are great treatments though. The issue is complexity. The virions have to achieve contact with the target cell, penetration, avoidance of host immunity, delivery of whatever their effector function is, and hopefully then killing the cancer cell. They also need to do this for a substantial portion of the billions of cancer cells in the patient (not all cancer cells, bystander killing is part of the way they work), which display considerable heterogeneity. This is a pretty complex chain of events, which is a disadvantage when dealing with an evolvable target like cancer. Other (effective) antibody or small molecule therapies have simpler mechanisms of action. The viruses we know and love have evolved over long periods to infect small populations of homogeneous cells, quite different to the task here. I guess we will see.
If there is niche for the virus in the cancer cells then the virus could in theory and adapt as well. It depends on how the virus was constructed.
In addition to directly killing tumor cells, a small number of dead cells could be processed into antigens by a macrophage nearby and presented to a T-cell.
If in some cases the virus can get in to the tumor cell but can't kill it, it's possible the virus can still be detected and presented on MHC Class I (unless it's suppressed) of the tumor cell by neighboring Natural Killer cells.
That's why it's important to reverse the tolerogenic environment that the tumor creates around itself. The viral infection will hopefully go a long way towards waking up the immune system up.
Real time adapting viruses are not really something that exists, unless I'm mistaken. They would be very hard to build not least because the problem is poorly specified and would introduce an unpredictable element which is undesirable for a medical therapy.
Yeah using viruses to potentiate anti-tumor immunity is an interesting idea. But returning to my main point, this is creating additional complexities. Most successful cancer therapies tend to do one thing well. Reliable synergy between treatments is actually quite rare. Peter Sorger has published some interesting stuff about this point. The reason combination therapies work better is because it is harder for a cancer cell to be resistant to everything, not because the drugs synergise to kill otherwise resistant cells.
MHC Class I mediated killing is done by effector T cells, not NK cells. Cancers of course learn to turn off peptide presentation via MHC, which stops anti-PD-1 drugs from working. Interestingly many real viruses stop the cells they infect from displaying peptides to escape T cell mediated destruction.
There are already studies of treatments which ramp up the immune response without killing off cancer cells much eg intratumoral STING agonists, mRNA compounds which induce manufacture of chemokines. They don't work that well, so I think the cancer killing capacity of the virus is important. Otherwise there are simpler more 'drug-like' ways to attract more effector cells to the tumor.
Sorry, I mis-remembered my immunology. NK cells kill in the absence of MHC Class I, which is something that would be really nice if it could be turned into therapy.
I have to wonder if they are great treatments though. The issue is complexity. The virions have to achieve contact with the target cell, penetration, avoidance of host immunity, delivery of whatever their effector function is, and hopefully then killing the cancer cell. They also need to do this for a substantial portion of the billions of cancer cells in the patient (not all cancer cells, bystander killing is part of the way they work), which display considerable heterogeneity. This is a pretty complex chain of events, which is a disadvantage when dealing with an evolvable target like cancer. Other (effective) antibody or small molecule therapies have simpler mechanisms of action. The viruses we know and love have evolved over long periods to infect small populations of homogeneous cells, quite different to the task here. I guess we will see.