Plausible fusion drive systems envision an ingition mechanism similar to that used by ITER - multiple lasers on a Deuterium target. The plasma created is momentarily held in an open ended magnetic pinch-bottle effect, with the highly energetic plasma guided at large velocity out the back end of the sucker.
Well, that's the idea anyways.
Multiple problems suggest themselves, and one benefit.
The benefit is not having the requirement to hold the millions of degrees plasma more than momentarily, as it's not created to create heat (=power production), but simply for the plasma itself; a very different pinch effect than utilized for current fusion experiments.
Guiding the plasma may be a problem. Will you get significant ISP, or just a big poof of rapidly expanding, cooling plasma?
Also, any ignition system would by needs have to be capable of continual operation - injecting Deuterium targets precisely and perfectly each time - to accelerate. E.g. a long series (minutes, hours, days, etc.) of constant pulses, rather than one long continuous burn.