Inhibiting HSP90 to treat cancer: a strategy in evolution.

Since the identification of the first HSP90 inhibitor almost two decades ago, there has been substantial progress made in the development of potent and selective molecules that inhibit this chaperone and that have anticancer activity. In turn, these compounds have been invaluable for probing how HSP90 supports the profound changes in cellular physiology that characterize the malignant state. Unfortunately, when used as single agents HSP90 inhibitors have demonstrated disappointing activity against advanced cancers in most of the clinical trials reported to date. This problem may be due to the major pharmacological liabilities of the first-generation HSP90 inhibitors that have been most extensively studied. We suggest, however, that it may well be intrinsic to the target itself. Systemically tolerable exposure to HSP90 inhibitors may not be highly cytotoxic for the majority of common clinical cancers. Instead, HSP90 inhibitors might better be used to enhance the activity of other antineoplastic agents while simultaneously reducing the capacity of tumors to adapt and evolve drug resistance; the overall result being more durable disease control. This review will focus on these fundamental issues with the goal of suggesting ways to make the clinical development of HSP90 inhibitors become less empiric and ultimately more successful.