The fusion reaction needs containment, the right fuel, and massive amounts of energy in
just the right place. It's phenomenally difficult to marshal the containment necessary to make the energy from a fusion reaction go into more fusion and not, say, warming up the test chamber a fraction. Yes, the fact that there's not much fuel is important, but not as important as the finickiness of making the plasma do what you want in the first place.
One of the major problems (arguably the problem) with tokamak or similar fusion reactors is that if the plasma ever touches the edge of the vessel, it immediately cools down and stops being plasma. It can't fuse with itself any more, much less trigger anything else to start. Heavier atoms need more energy to fuse, that's why you see hydrogen, helium, lithium bandied about in these discussions. The amount of energy needed to fuse the atoms in the walls of a fusion reactor is literally supernova-scale. We're not talking "there's an engineering tolerance built in for safety", rather "as a civilisation it's not immediately conceivable how we might generate amounts of energy that large".
Given that the majority of the earth is made out of elements that are inconveniently heavy, runaway fusion is absolutely, definitely, totally, completely not a problem.