Any solution that involves stem cells needs to also massively reduce errors in DNA replication, otherwise cancer would occur.
Their larger size is due to more cells (not bigger cells), so they would be more at risk of cancer. But they've evolved to have more TP53 genes, so their lifetime risk of cancer is around 5%:
https://www.nih.gov/news-events/nih-research-matters/how-ele...
As well as all that, elephant cells are intrinsically more cancer resistant as you point out.
Unfortunately untrue. You are referring to https://en.wikipedia.org/wiki/Peto%27s_paradox but as you say 5% of Elephants still die of cancer. Elephants have evolved other mechanisms to protect against cancer because they have more cells (more copy errors) https://www.sciencedaily.com/releases/2021/02/210204131438.h...
When thinking of cancer, try to think of the incentives of the organism to avoid copying errors, which are eventually deadly. So many deaths are direct effects of copying errors (cancer), that the human body gives most cells a lifetime of ~40 duplications (https://en.wikipedia.org/wiki/Hayflick_limit measured by telomeres) before killing off cells (cell senescence). Senescence causes the majority of other deaths i.e. the root cause of death is often indirectly due to the body’s mechanisms to prevent cancer. The mechanism to cause senescence has to be robust to error, so that mutations are unlikely to defeat cell death. A metaphor: the body tries to keep the tree of all cell duplications as a balanced tree - and cancer is when one node has enough error that it can reproduce and unbalance the tree.
https://royalsocietypublishing.org/doi/10.1098/rstb.2016.044... links to better information, and it has a lovely summary: “We assume that selection favours short telomeres as a mechanism to protect against cancer. At the same time, selection favours long telomeres as a protective mechanism against DNA damage and replicative senescence.” We have evolved compromises that protect us against DNA copying errors, but those compromises are imperfect (from a longevity POV).
Cancer is an area where there is extensive misinformation as to the root causes, and “the cure for cancer” is non-sensical: the cure would require eliminating DNA errors in all cells within the body. Currently we can fix some individual errors (specific cancer therapies), or attack dividing cells.
After I found out about this, I now view all other cancer research suspiciously. Them trying to maybe cure one specific type of cancer by blocking some specific protein now feels like they're trying to fix a bug in a react web app by carefully loading a crudely written kernel extension into the OS. Yes, it's technically possible, but wouldn't it make more sense to operate on the right abstraction level?
We can work on different abstraction levels because in the worst case, we can just crack open a hex editor and write raw assembly or mess with the OS directly. We've built on that to get C, and Java, and Python, and all that other good stuff, but we can go all the way to the bottom if we need to.
Now imagine you're working on a computer with proprietary hardware. The easiest way to interact with it is via a provided high level language, like Lua. This is eating, drinking, sleeping, exercising. Now imagine that you crack open the assembly because something is just not working right. Some illness or something. And you come to realize there are thousands of different opcodes (proteins, enzymes, genes, etc.). Not only that, but sometimes they work differently depending on where in the computer they are located or the time of day or whatever (organs, menstrual cycle, circadian rhythm, etc.). You can make some good guesses on what a couple of them do, but you are not sure of the entirety of their effects for the majority of them.
So you're looking at the interfaces and you figure out if you send the right data to this and that place, you can get it to work right, maybe with some acceptable or unforeseen consequences. This is medicine, and it is a miserable way to do things due to the complete lack of understanding and control, much like driving your car from the backseat with a 10 foot pole while facing backwards.
So you've discovered a lot of hacks, made a lot of medicine, collected a lot of 10 foot poles. And as you look at your ever expanding collection, now easily in the thousands, you begin to get the creeping feeling that (a) you will never find a panacea, (b) the system you are working on is a disgusting mess of patchwork jobs on patchwork jobs developed over millions of years, and (c) there is probably an underlying system directing all of this, but its complexity might not even be better than what you're operating on now, and you definitely lack the technology to utilize it properly anyway.
And what you're doing is saving and bettering people's lives on the whole. And then some big shot comes in and says "Ackchyually, why don't you just access the lowest layer directly? Pretty sus." Try to have some perspective and humility on how horrifically complicated biology is. Life was evolved, not designed. Life isn't a computer. The relative simplicity and sanity of technological systems is a blessing absent in biology.
Biology is not programming, except when it is like programming, or when it involves programming. But that could also be said about golf.
A lot of people have imagined that. But nobody does.
Potentially cancerous cells are constantly being produced in the body and then killed off by the young and functional immune system.
Keep the immune system young and functional and most of the cancer concerns are not terribly concerning.