If you want to confine your plasma in a given volume by a B-field, then it's possible (this is the concept of magnetic bottles, tokamaks or stellerators).
Yeah, this is well verified in laser produced plasma experiments. Explosive plasma having directed velocity can be converted into converging flow and can be highly decelerated by using strong magnetic field (strength of the magnetic field needed depends on the initial kinetic energy of the plasma). You can see a recent communication Plechaty et al., PRL111, 185002 (2013).
Yes, it depends in the cyclotron radius which should be very small in comparison to the mean free path and the dimension of your device. This is the typical situation in Hall-effects ion source where the magnetic field topology is chosen in such a way to confine the plasma with a closed-drift within the azimuthally direction.
I think that if your plasma reach the magnetic field from a tangential direction to the drift axisymmetrical magnetic field you can reduce drastically the velocity of your beam and trapped in the azimuthally direction. However, turbulence/instabilities and third dimension effects will appear quickly to reduce the plasma density and particles could escape from the axial direction.