display | more...
In the writeup above, fhayashi gives a great description of what monoclonal antibodies are, but never mentions what they are good for. Monoclonal antibodies are useful in medicine because they enable the immune system to fight a specific pathogen without expending the time and energy required to make its own antibodies.

Some modern monoclonals are even better, because they can fight a specific pathogen without any help from white blood cells. They do this by carrying a toxin or radioactive isotope, rather than the white blood cell binder that's part of a normal antibody. These elements damage the pathogen when the antibody attaches to it, not involving the immune system at all. Normal antibodies (and monoclonal antibodies acting in the normal way) serve as attractor sites for white blood cells which eventually come and destroy the pathogen.

Since monoclonals were invented in 1975 (resulting in the 1984 Nobel Prize mentioned above), and seem to have all the advantages over vaccination, you may wonder why they aren't in drastically wider use. As it turns out, monoclonals simply haven't proven viable in human trials until eight or nine years ago, for a variety of reasons. First, the monoclonals developed in the 80s were themselves interpreted as pathogens by the immune system, and dealt with accordingly. Thus they were worse than useless, since the immune system inevitably killed some of its own antibodies while purging the monoclonals. Also, many monoclonals had been developed using mouse antibodies, which are more often than not rejected by the body in the human anti-mouse antibody (HAMA) reaction.

B lymphocytes, a variety of white blood cell, are isolated from mice that have been exposed to an antigen/pathogen. These lymphocytes are then fused with a human lymphocyte isolated from myeloma, a cancer of the bone marrow. These "hybridomas" have the antibody properties of the mouse cells along with the ability to reproduce infinitely from the cancer cells, so they are perfect to grow a large culture of. The culture is exposed to the original antigen, and those hybridomas which bind to it (i.e. those which contain the antibody against it) are isolated. They can then be cultured in a petri dish or in living mice, with monoclonal antibodies eventually harvested from either.

Unfortunately, these harvested antibodies will still suffer from the problems above, since they are made completely of mouse protein. One way to get around this is to fuse the antigen binding regions from the mouse antibody with the basic structure of a human antibody, resulting in a so-called chimeric antibody which is around 66% human. Another, related, method is taking the antigen binding sites themselves off of the mouse's binding regions, and attaching them to a fully formed human antibody. This is known as a humanized antibody, and is 90% human protein.

Making human monoclonals seems an even better approach than rebuilding mouse ones, but has turned out to be very difficult in practice. Researchers have (finally, after 25 years of work) fused human B lymphocytes with immortalized cells, but haven't tested antibodies produced by these new cultures. Taking the opposite approach, two companies have engineered mice to produce human antibodies rather than murine ones. Some antibodies made this way are currently undergoing human drug trials.