Actually most of the heat transfer really is in those 14 Mev neutrons and the "cooling" neutral helium that is left to drift down the bottom of the torus.

Neutron deposit heat directly in the reactor structure, which must all be cooled. And their is a "heat conveyor belt" with water of helium at the bottom of fusion torus, whether it's JET, Tore Supra, or ITER. And last time I checked, to my great surprise, heat extraction WAS NO LONGER a challenge. It is solved at ITER target powers.

Challenge N°1 is materials durability under neutron bombardment. Unless it's greatly improved, fusion reactors won't have enough uptime to be viable, and will produce just as much waste as fission, except it's structural not fuel.

Tritium breeding is not even a challenge anymore, hope is given up. It's now a known limitation of the deployment speed of the technology: a fusion reactor needs an initial tritium capital to put in its first blankets, and more tritium will be bred out of it only  very slowly. Mankind would NEVER be capable of bootstrapping more than 1 fusion reactors per two decade and per fusion reactor already running. And all the initial stock built and painfully maintained over decades by all fission reactors would go into the first commercial scale fusion reactor. There was a very interesting discussion of this on The Oil Drum a few weeks ago.

Being an ugly malevolent nuclear lobbyist, I think we should move to thorium breeders (and multiply fuel availability by 300 compared to PWR), but it is interesting to note that they too have widescale bootstrapping contingencies, albeit not as stringent.

Pierre