I beg to differ... cancer is solely caused by (a combination of misfortunate) gene or gene methylation alterations, which disable the controlled death program of a cell. Controlled cell death after replication is normal, essential, happens every day and does not kill the entire organism at all, you just keep on living with copies of your initial cells until the telomeres are exhausted. However, if a cell (copy) becomes "immortal" due to a damaged death program, it uncontrollably replicates itself, expatiates, displaces healthy tissue, alterates vital organs and increasingly consumes space. Actually you do not die from cancer cells but from their secondary health effects such as circulatory disorders, hypoxia or organ failure.
Cancer cells appear more often than you would believe, but your immune system (T-cells) detects and kills most of the mutations immediately. But in rare cases, the cancer cell is able hide or fool the T-cell. Modern immune therapy aims at this mechanism, it tries to adjust the T-cells to the target gene alterations. This therapy would be able even to cure terminal cancer (and it already has in first clinical trials!) having said this, the hit ratio is still not formidable and the T-cells sometimes attack healthy tissue as well. Which means some patients show critical inflammatory response. We are at the beginning here and it will likely take yet another 10 years until immune therapy can cure (or at least chronify) some (but not all) kinds of cancer.
If your body wants to live, it has to maintain its construction plan and eliminate any 'anarchic' cells. It is all about killing the immortal cells ... radiotherapy does the same, by wittingly altering all gene information of all cells in an area, hoping that an already modified cell will be even less functional thereafter, and die after all. Breaking a butterfly on a wheel, but we don't have anything else available at present.
"Reversing" the gene (or methylation) alteration is not so easy. That is something which requires advanced technologies such as CRISPR which will take even longer for deployability than the immune therapy. Currently it is too dangerous to use it, as you don't know if it really cuts a DNA at the right position. CRISPR is cheap, fast and brilliant, but the hit ratio still below 20% for many applications. A lot of improvement to be done in the next 20 years ahead.
I would love to see HCF1 protein folding and interaction simulations to help my little boy... someday.