Why are all the gene-based cancer breakthroughs I've seen in the news for blood diseases? Is it easier to manipulate genes in blood?
With solid tumors, you have to deal with the complexities of the tumor microenvironment. The tumor microenvironment is the complex environment of the cancer tumor tissue. Of particular note, decreased lymphatic drainage, increased connective tissue, and aberrant angiogenesis (growth of new blood vessels) create very high interstitial fluid pressures within the tumor, limiting the ability of drugs or cells to penetrate within the tumor and "work their magic" by reaching the cells/target molecules of interest, etc.
Blood cancers are a good proof-of-principle for many cancer therapies; translation to solid tumors is difficult (see challenges with CAR-T cell immunotherapy).
As for gene manipulation in blood--yes, it is relatively easy to manipulate genes in blood. Hematopoetic Stem Cells (which differentiate into all different kinds of blood cells) can be withdrawn without too much difficulty, manipulated in the lab, and then reinjected into the patient.
ps: I remember seeing more 'barbaric' version of the idea, namely metal pipes planted in the patient as a direct path. I was wondering about something more organic.
I'm not personally familiar with work in developing capillary devices, but it wouldn't surprise me at all if this was an approach under development--it certainly makes a lot of sense.
The reason blood cancers are first with these techniques boils down to having better known antigen targets to target compared to solid tumors where known antigens have more cross reactivity to other tissues (so less specificity, more side-effects).
Still good to differentiate from "giving a drug".
Otherwise, you get the side effect of destroying your own body, the case of a woman a few years ago, who died having her lungs destroyed by immunotherapy. Lungs had the same antigen.
This morning I read something hopeful in this direction - UCSF researchers designed an AND gate, so now they could trigger a T cell attack only when 2 specific antigens are present, making it much more reliable. And I could imagine these can be chained, further narrowing it down. It needs genetic sequencing to isolate.
https://www.ucsf.edu/news/2016/01/401471/tricked-out-immune-...
At the moment this is a terrible manual and time consuming process, so it is expensive.
Trading cancer for SLE or rheumatoidal arthritis etc.