A big part of the problem in medicine is the lack of good targeting technology.  This thought occurred to me as I was lamenting how slowly my steroids were reversing an autoimmune hemolytic anemia today.  If only I could target the exact antibody that was consuming the red cells! Was the antibody directed against a red cell membrane protein, or a lipid component?  I have no way of finding that out clinically, let alone selecting that bad antibody for destruction.  As I explained to the patient, the immune system's making too much of the bad antibody, and not enough of the good antibodies.

We have broad techniques to carpet-bomb antibody production: steroids.  IVIg.  Rituxan.  These techniques have their drawbacks.  Steroids have horrible metabolic side effects as well as osteoporosis.  IVIg is expensive and the effect doesn't last that long.  Also, some bags of IVIg are probably better than others, depends on the kinds of antigens the donors have been exposed to.  Rituxan can lead to profound immunosuppression, analogous to HIV infection.  So what's a guy to do?

We face the same problem in oncology too.  We have the tools to kill cells, but we lack the technology to deliver that kill shot to the right cell most of the time.  Defining cancer on a molecular level is pretty tough when you get down to it, though it's not for lack of trying.  Even if we create a molecular definition, which we have in a paucity of cancer types, we often lack the vehicle to deliver the payload to the target.  A nuclear weapon is pretty useless if it blows up over the North Pole (though that analogy might not hold up in a few years).

I really think it will take nanotechnology to get us to the next level of targeting.  We will need comprehensive profiling of the entire organism on a molecular level to understand what distinguishes cancer from normal.  We thought that microarray profiling was going to be the answer, but so far, looks like checking thousands of cancer cells mixed in with non-cancer cells is too "dirty" to give us a clear picture of the targets we need (with some exceptions).

We thought the targets would be cell receptors.  We are currently midway through the receptor targeting era in oncology, where molecules home in on growth factor receptors or their downstream signaling pathways in an attempt to selectively disable cancer cells.  Success has been moderate, and toxicity still vexing, despite claims that efficacy would be excellent and toxicity minimal.  Whole departments of "translational" medicine have yet to yield marketable drugs.

I think RNAi is exciting for this reason--there are many new RNAi molecules that correspond to the neoplastic phenotype.  Perhaps these are the targets we have been looking for in oncology.  Now, how long will it take pharmacology to catch up to RNA-ology in designing molecules to enter cancer cells and knock out the bad RNAi's?  Or will this too turn out to be only helpful in some particular cancers driven by RNAi changes?

Seems like for every new "targeted" therapy we develop, certain cancers respond particularly well.  Why are only myeloma and mantle cell lymphoma treatable with Velcade?  What is it about MDS that makes it particularly amenable to treatment with Dacogen?

Until we understand the entire system of molecular regulation in the cell, we will have to be satisfied with slow progress with one or two successes for every new treatment paradigm that emerges.