One of the reasons that the immune system degenerates with age is, and I greatly simplify the reality on the ground in saying this, that only a limited number of immune cells can be supported at once. As the years roll on more and more of the immune cells known as T cells become memory T cells, dedicated to remembering specific threats. That leaves less and less room for naive T cells that can go out and stomp on new threats. Thus as your immune system becomes ever more knowledgable, it also becomes less and less effective at its primary missions - including destroying pathogens, destroying senescent cells, and destroying early stage cancers.
Now in theory, an old immune system that is top-heavy in memory T cells could be at least partially restored (remember that there are other issues and degenerations beyond the one I discuss here) by destroying all the unwanted memory cells. In recent years researchers have destroyed and then used stem cells to recreate the entire immune system in human trial patients, and have done this to essentially remove misconfigured immune cells that were the source of an autoimmune disorder. If that can be done, then it should certainly to be possible to take one of the new generation of targeted cell destruction technologies developed in the cancer researcher community and use to it destroy only a specific population of T cells.
Recall that T cells, like all cells, are richly decorated with specific receptor molecules on the surface, molecules going by such descriptive names as CD28, CD8, or CD4. A cell with a CD8 receptor is called CD8+ for example, though cells can have many combinations of receptors (e.g. CD4+/CD8+/CD28-). Researchers will generally refer to a cell type by the receptors they are interested in at that moment rather than list them all. The surface receptors of T cells are well understood, and would make good targets for engineered viruses or nanoparticles designed to seek out specific cell receptors and kill the cell bearing them.
Is the excess population of memory T cells in an old immune system needed for anything else, however? Would destroying them remove needed portions of the immune system's programming and thus lead to issues more critical than those suffered before the procedure? Recent research suggests that this is not the case:
Highly differentiated CD28 effector T cells which accumulate in a variety of diseases and also with increasing age contribute to inflammatory processes, limit immunological space and diversity, and are associated with immunological dysfunction and reduced responses to vaccination. Elimination of CD28 T cells has been suggested as a measure for immunological rejuvenation but may lead to the loss of important T cell specificities.
we show that the same clonotypes are present in CD8(+)CD28(+) naive/early memory and CD8(+)CD28 effector T cells. Therefore, CD28 cells do not seem to contain clones which are not present in the residual population. The elimination of effector T cells would not lead to the loss of important specificities, as relevant clonotypes could be recruited and propagated from naive or early memory T cell subsets in the case of exposure to pathogen.
Which sounds promising from the point of view of moving right on ahead with the destruction of the dead wood. I expect it will not be too many more years before researchers are performing selective culls on old immune systems in order to see if this strategy can indeed produce significant rejuvenation to a more youthful level of performance.
Weinberger B, Welzl K, Herndler-Brandstetter D, Parson W, & Grubeck-Loebenstein B (2009). CD28(-) CD8(+) T cells do not contain unique clonotypes and are therefore dispensable. Immunology letters PMID: 19715728