No. Definitely not. A nuclear decay will typically lower the mass-energy of a system, while many high-energy collisions will add to the mass-energy. In these systems, the mass-energy is converted to/from kinetic energy.
It is possible to say that energy is conserved locally. What this means is that it's possible to use equations such that the change in energy in an infinitesimally-small region is equal to the energy flowing into that region. The caveat here is that this only works reliably for flat space-time, but since you can always use coordinates where the local space-time is flat, this is always possible to do locally. This is why we generally consider energy to be conserved in reactions on the Earth: the space-time curvature is small, and regardless we typically deal with reactions that are tiny in size. Furthermore, we can write down a gravitational potential energy which allows us to account for the energy changes induced by gravity (this isn't always possible).
But when you've got a curved space-time, much of the time it's impossible to say that energy is conserved across the entire system, regardless of whether it's closed or not. As Bandersnatch pointed out, global energy conservation just doesn't work, because there isn't a unique definition of total energy for a curved space-time.