Previously unrecognized and potentially consequential challenges facing Hsp90 inhibitors in cancer clinical trials.

Targeting the heat shock protein-90 (Hsp90) chaperone machinery in various cancers with 200 monotherapy or combined-therapy clinical trials since 1999 has not yielded any success of food and drug administration approval. Blames for the failures were unanimously directed at the Hsp90 inhibitors or tumors or both. However, analyses of recent cellular and genetic studies together with the Hsp90 data from the Human Protein Atlas database suggest that the vast variations in Hsp90 expression among different organs in patients might have been the actual cause. It is evident now that Hsp90β is the root of dose-limiting toxicity (DLT), whereas Hsp90α is a buffer of penetrated Hsp90 inhibitors. The more Hsp90α, the safer Hsp90β, and the lower DLT are for the host. Unfortunately, the dramatic variations of Hsp90, from total absence in the eye, muscle, pancreas, and heart to abundance in reproduction organs, lung, liver, and gastrointestinal track, would cause the selection of any fair toxicity biomarker and an effective maximum tolerable dose (MTD) of Hsp90 inhibitor extremely challenging. In theory, a safe MTD for the organs with high Hsp90 could harm the organs with low Hsp90. In reverse, a safe MTD for organs with low or undetectable Hsp90 would have little impact on the tumors, whose cells exhibit average 3-7% Hsp90 over the average 2-3% Hsp90 in normal cells. Moreover, not all tumor cell lines tested follow the "inhibitor binding-client protein degradation" paradigm. It is likely why the oral Hsp90 inhibitor TAS-116 (Pimitespib), which bypasses blood circulation and other organs, showed some beneficiary efficacy by conveniently hitting tumors along the gastrointestinal track. The critical question is what the next step will be for the Hsp90 chaperone as a cancer therapeutic target.