Another major difference between lytic strains and nonlytic strains is that, although they both have the potential to kill infected cells, the mechanisms by which they accomplish this result are different. The production of infectious progeny virus particles by lytic strains gives them the ability to kill host cells fairly quickly. The budding of progeny viruses that contain activated hemagglutinin-neuraminidase and fusion protein molecules in their outer coats causes the plasma membrane of NDV-infected cells to fuse with the plasma membrane of adjacent cells, leading to the production of large, inviable fused cells known as syncytia. Reviewed in [13,18,30] The more efficiently a lytic strain can replicate inside a host cell, the more quickly it can kill that cell. The preferential killing of cancer cells by a lytic virus is known as oncolysis; thus, lytic strains of NDV are also called oncolytic strains. Nonlytic strains of NDV kill infected cells more slowly, with death apparently the result of viral disruption of normal host cell metabolism. Reviewed in 
The specific mechanism by which nonlytic NDV strains cause cell death in cancer cells has not been completely elucidated, but in Vero cells (derived from kidney epithelium) it was determined that NDV caused cell death by decreasing DNA content, increasing the ratio of Bax to Bcl-2, increasing p53 level, and increasing caspase expression, resulting in apoptosis.[39,40,41]
As indicated previously, both lytic strains and nonlytic strains have been investigated for their anticancer potential. In fact, the major differences between the two strain types have been exploited to develop three different approaches to cancer therapy:
- The infection of cancer patients with a lytic strain of NDV.
- The use of oncolysates, i.e., preparations containing plasma membrane fragments from NDV-infected cancer calls, as anticancer vaccines.
- The use of intact cancer cells infected with a nonlytic strain of NDV as whole cell vaccines.
One proposed advantage of the first approach is that virus replication may allow the spread of cytotoxic viruses to every cancer cell in the body; Reviewed in [8,34] however, the production of virus-neutralizing antibodies by the immune system might limit this possibility. Reviewed in [6,8,13,30,30,40] The rationale for the second and third approaches is that tumor-specific antigens (i.e., proteins or other molecules that are generally located in the plasma membrane of cancer cells and that are either unique to cancer cells or much more abundant in them) may be better recognized by the immune system if they are associated with virus antigens (i.e., virus proteins that have been inserted into the plasma membrane of host cells). Reviewed in [8,12,13,28,32,34,42,43,44,45,46,47,48] If this enhanced recognition takes place, then it may increase the chance that cancer cells, whether they are virus infected or not, will be recognized as foreign by the immune system and be destroyed. Reviewed in [8,12,28,47,48]