Targeted Cancer Drugs, Succeeding and Failing

 by Derek Lowe

You get widely varied perspective on targeted cancer therapeutics depending on where you get your information from. Cancer therapy centers in the US like to advertise as if they can zero right in on a particular patient's particular sort of cancer and then reach up on the shelves to pick out which of the thousands of alternatives will be most efficacious. Now it's true that we can do more of that than we used to, but the real-world situation is messier: there are indeed targeted cancer therapies out there, but they don't cover very broad swaths of the cancer landscape. It's hard to feel fortunate with a cancer diagnosis, but if you end up with one that lands squarely on one of the targeted drugs, then fortunate is the only way to describe it.

Meanwhile, inside the industry narrowly targeted therapies are (and long have been) a real balancing act. Market size versus efficacy will do that, because outright cures are worth a great deal, but are extremely rare, and the scale slides down from there. If there are only a small number of patients who will ever be taking the drug, it's going to be expensive (and it means that those few had better be pretty well served by it). But if there's a huge population of patients, you can be sure that some of them will respond better to it than others, generally due to factors beyond anyone's control (if they're understood at all). Mechanistically, there are some tumor types that really do seem to be driven by a particular pathway, and if you hammer that pathway then patients have a very strong chance of benefiting from the treatment. The BCR-Abl therapies directed towards acute lymphoblastic leukemia (ALL) are an example of this. But other tumor types are unfortunately more versatile (and often contain a wider variety of mutated cells), and these are the very-bad-prognosis ones like glioblastoma and pancreatic cancer. It is highly unlikely that you'll find a single pathway to target for those.

This new paper illustrates some of the details. It's looking at a particular kinase inhibitor (dasatinib, Sprycel) that is very effective indeed against Philadelphia-chromosome-positive CML and is also included in ALL therapies, both because of its Abl kinase inhibitor activity. (Nice note about its generic name here!) Dasatinib was considered a second-generation drug after imatinib (Gleevec), a prototype in this area, and has been particularly useful in some imatinib-resistant cases. It also is a strong inhibitor of Src kinase, and this led to a whole string of clinical trials against solid tumors where Src is known to be important. But in contrast to the Ph+ leukemia trials, these have been pretty unrewarding (Ph- blood cancers likewise). Table 2 of the paper has the grim statistics - failure after failure in trials against various sorts of lung and breast cancer and several others.

But as the paper argues, no one had shown that Src activity was a primary driver of any particular tumor type (against, as opposed to the Philadelphia chromosome Abl activity in those leukemia types). Dasatinib is outright cytotoxic to those latter cells, and the resulting elimination of them is a big factor in its clinical success. But shutting down Src doesn't really kill cancer cells - it slows down their growth and cell division, sure, and you'd hope that such cytostatic activity would be worth something, but the clinical results say otherwise. (I'm reminded once again about the similarities between oncology treatments and antibacterials, where there has been a longrunning debate about the usefulness of bacteriostatic agents as opposed to the bacteriocidal ones).

The authors stress that the preclinical cell and animal model tests of dasatinib really did predict its activity in the clinic, although at the time many people talked about a disconnect. The problem was that slowing solid tumor growth was considered enough of a reason to be optimistic, and it really wasn't. That means, they say, that oncology research teams really should value selective cytotoxicity and prioritize targets and compounds that show evidence of it, rather than chasing after cytostatic agents that might never show enough power on their own. But that "on their own" is there for a reason, because something like dasatinib can be a valuable thing to throw into a cocktail that features other cytotoxic agents (and indeed, there are several examples of this). You'd rather work on something that can be a frontline therapy, other things being equal, but we have a number of these compounds out there and they can still be put to use.

https://www.science.org/content/blog-post/targeted-cancer-drugs-succeeding-and-failing