One year ago, we presented a new approach to treat hadronic interactions for the initial stage of nuclear collisions. It is an effective theory based on the Gribov-Regge formalism, where the internal structure of the Pomerons at high energies is governed by perturbative parton evolution, therefore the name "Parton-Based Gribov-Regge Theory". The main improvement compared to models used so-far is the appropriate treatment of the energy sharing between the different elementary interactions in case of multiple scattering. It is clear that the above formalism is not yet complete. At high energies (RHIC, LHC), the multiple elementary interactions (Pomerons) can not be purely parallel, they interact. So we introduce multiple Pomeron vertices into the theory.