Nonparametric estimation of a state entry time distribution conditional on a “past” state occupation in a progressive multistate model with current status data