The benign termination of a runaway electron (RE) beam has emerged as a central thrust of disruption research. Experiments at both JET [^1] and DIII-D [^2] have demonstrated that a global MHD instability is able to expel the vast majority of REs and distribute them over a broad area of the vessel wall thus avoiding localized damage. While encouraging, important questions remain with regard to extrapolating this scheme to reactor scale plasmas. Here we seek to address two critical aspects of this problem. The first seeks to identify an approximate lower bound on the size of the remnant RE seed that remains confined after the MHD instability. It is found that in plasmas containing significant amounts of high-Z material such as Neon, strong pitch-angle scattering leads to a substantial number of magnetically trapped energetic electrons. These trapped energetic electrons remain well confined in the presence of a stochastic magnetic field, and hence are lost on a far longer timescale compared to passing REs during a global MHD instability. Once the flux surfaces reform, this remnant population of electrons can be detrapped and thus provide a small seed that may be subsequently amplified by the avalanche of REs. The second thrust utilizes a recently developed fluid-kinetic hybrid framework of RE and bulk plasma dynamics to assess under what scenarios this remnant trapped electron population is able to lead to a partial reformation of the RE beam. Critical quantities that emerge from the analysis are the amount of poloidal magnetic flux present in the system before the global MHD instability is triggered, along with the efficiency through which the consumption of this poloidal flux amplifies the remnant seed population. It is found that a partial reformation of the RE beam is possible for plasmas containing RE plateau currents in excess of two mega Amperes across a range of conditions. [^1]: Reux et al., Phys. Rev. Lett. (2021), [^2]: Paz-Soldan et al. Nucl. Fusion (2021)