In article <816180076.3745 at njosborn.demon.co.uk> you write:
>>So why can't we prevent metastasis simply by passive immunity-
>injecting antibodies into someones bloodstream which bind any free
>cancer cells preventing them infecting some other organ. I can only
>believe that you would need such a large dose of antbody as most will
>bind to the tumour itself.
The problem isn't so much getting enough antibody (though that's a
technical hitch); rather it's a question of finding an antibody that will
recognize a tumour cell and not a normal cell. Since tumour cells really
are noral (i.e. not 'foreign') cells, almost all of their cell-surface
markers are identical to those of normal cells. There may be
differences in levels of some proteins, which doesn't help much with most
antibodies - you need an all-or-nothing effect for this to work well.
There are also fairly often differences in some internal proteins of the
tumour cells - that is, proteins which are not expressed on the cell
surface, and which therefore are not recognized by antibodies. But
often, tumour cells can differ from normal cells in that they lack some
normal proteins (i.e. regulatory proteins) - obviously this doesn't
provide a handle for an antibody.
There are some tumour cells which express cell-surface proteins different
enough from normal cells that this might work: however, most of these are
quite rare and unusual tumours. (Also, they tend to be unique to the
individual. That is, there are very few cases where you could whip up a
batch of antibodies and have them work in more than one patient.) There's
a great deal of interest and ongoing research into identifying more
potential targets. As you might imagine, it isn't simple to identify
antibodies that might work this way.
A related area that's also quite hot is the same concept applied to
cytotoxic T lymphocytes. CTL recognize internal proteins and in some
cases can pick up smaller differences than can antibodies, so it's
possible that more tumours will be susceptible to this approach. In this
case, there are some hints that it might be more broadly applicable - that
is, some antigens have been identified where, at least in theory, the same
batch of CTL could be used in many different patients.
On the other hand, it's much, much more difficult to grow the CTL in any
kind of appropriate numbers, and there are many more technical hitches
than with the antibody-based approach.
So the bottom line is that your general concept isn't a bad one, and may
be used clinically in a few special cases now. (I'm not sure if clincal
trials have begun or not; I think they're very close if not.) However
there are major problems preventing it from being a broadly applicable
approach.
This is a very superficial reply; there is a lot of literature on the
subject. I've left out several other potential problems and different
approaches.
Ian
--
Ian York (iayork at panix.com)
Dana-Farber Cancer Institute, 44 Binney St., Boston MA 02115
Phone (617)-632-3921 Fax (617)-632-2627
