There have been several thermal runaway incidents in LENR. The most famous was early in the days of Pons-Fleishmann.
I would think if someone can engineer a material that promotes sustained plasmoid reaction while controlling how they wander around and that won't melt or disintegrate, you could get a a rod that heats up to say 2,000F and stays that way for months or years.
Perhaps with the application of the right current or magnetic field you could shut it off.
But it might be as difficult to engineer such a material as it is to get containment fusion.
But if you could get pellets to burn in succession, you'd still have a great power source.