It is assumed that for most naked viruses, the egress of newly formed virions involves only cell lysis, but a study by Karla Kirkegaard, formerly of Stanford, show that nonlytic release via phagosomes was also possible.
A important link to understanding this other method of viral exit is the fact that all eukaryotic +ssRNA viruses replicate genetic material on membranes in the cytoplasm (e.g. mitochondrial, ER, Golgi).
Kirkegaard postulates that polioviruses and some other nonlytic viruses can induce and be ecapsulated by cytoplasmic double-layered membranes, in which they mature and are eventually shipped out to the extracelluar fluid.
Despite the normal use of this autophagy pathway to degrade intracellular components, the digestive properties of said double-layered membranes seem to be ineffective or inhibited by nonlytic viral invaders.
Inhibition of cellular autophagy resulted in a lower yield intracellular viruses (because they are dependent on these membrane surfaces) but an even lower yield of extracellular virus; this is consistent with the non-lytic escape hypothesis.
Furthermore, experiments in which cells were polarized such that vesicles would only exit in one direction and infected with poliovirus showed virion emergence from the predicted side of the membrane.