On the nonexistence of global solutions to a nonlinear Euler-Poisson-Darboux equation☆