Isn't the evolution of cancer cells rather more constrained than your average pathogen? A normal pathogen can infect many thousands or millions of hosts, and can evolve immunities over many years, decades or centuries. A cancer is limited to a single host, and cannot pass on any of its tricks; it must start from scratch each time.
If we had medicine as potent as penicillin that could attack or inhibit certain cancer cells, it might be that the cancer simply wouldn't have the time to fight back before it's removed entirely.
Cancer cells are rather more difficult to distinguish from normal cells than bacteria, of course, but it seems reasonable to assume we'd have less problems with cancer cells evolving defences.
If we had medicine as potent as penicillin that could attack or inhibit certain cancer cells, it might be that the cancer simply wouldn't have the time to fight back before it's removed entirely.
Cancer cells are rather more difficult to distinguish from normal cells than bacteria, of course, but it seems reasonable to assume we'd have less problems with cancer cells evolving defences.