Robert A. Nagourney, M.D.
Two related clinical trials were reported in the last several months describing the use of heat shock protein 90 (HSP90) inhibitors in lung cancer. Both trials fell short of their pre-specified endpoints casting a pall upon these drugs. However, the study of HSP90 inhibitors should not be abandoned based on these findings, as this is a fertile area of investigation and offers opportunities for the future.
Human cells marshal many defenses against stress. Thermal injury can damage basic cellular functions by denaturing (inactivating) proteins. The machinery of cells is largely comprised of protein enzymes. Excessive heat coagulates proteins much the same way the albumin of an egg turns white during cooking. The loss of fluidity and function ultimately results in cell death. The heat shock proteins come to the rescue by shepherding these proteins away from injury and protecting them from denaturation. There are many different heat shock proteins found in human cells, but one of the most abundant and active in cancer cells is known as HSP90 for its molecular weight in the range of 90-kilodaltons. Over the last two decades, investigators have explored the use of small molecules to inhibit these important proteins. Among the first compounds to be isolated and applied were derivatives of Geldanamycin. Although Geldanamycin itself is a poison that causes severe liver damage, its derivative 17-AAG, also known as Tanespimycin, has successfully entered clinical trials.
The current studies examined two other HSP90 inhibitors. One Retaspimycin, has been developed by the Infinity Pharmaceuticals. This clinical trial combined Retaspimycin with Docetaxel and compared results with Docetaxel alone in 226 patients with recurrent lung cancer. None of the patients had received Docetaxel prior to the trial. Drugs were administered every three weeks and the efficacy endpoint was survival with a subset analysis focused upon those with squamous cell cancer. The trial fell short of its pre-designated endpoint. Interestingly, the study failed to provide benefit even in patients who were specifically targeted by their tumor’s expression of the K-RAS, p53 or by elevated blood levels of HSP90, the putative biomarkers for response.
The second trial examined a different HSP90 inhibitor developed by Synta Pharmaceuticals. The drug Ganetespib was combined with Docetaxel and the combination was compared with Docetaxel alone. The results just reported indicate that the combination provided a median survival of 10.7 month, while Docetaxel alone provided a median survival of 7.4 month. Although this represented a three-month improvement, it did not meet the pre-specified target.
Taken together these results could dampen enthusiasm for these agents. This would be unfortunate, for this class of drugs is active in a number of human tumors.
Through our EVA-PCD functional profile we have observed favorable activity and synergy for the HSP90 inhibitor Geldanamycin and its derivative 17-AAG as we reported at the American Association for Cancer Research meeting in 2005 (Nagourney RA et al Proc. AACR, 2005). More importantly, 17-AAG (Tanespimycin) provided objective responses in 22 percent and clinical benefit in 59 percent of patients with recurrent HER2 positive breast cancer after these patients had failed therapy with Herceptin (Modi S. et al, Clinical Cancer Research August 2011). This clearly supports the role of HSP90 inhibition in breast cancer and would suggest that other more carefully selected target diseases could benefit as well.
The function of HSP90 is not completely understood as it influences the intracellular trafficking of dozens of proteins. One of the complexities of this class of drugs is that they protect and enhance the function of both good and bad proteins. After all, the HSP90 protein doesn’t know which proteins we as cancer doctors would like it to protect.
When we apply EVA-PCD analysis to these and other related classes of compounds, we focus our attention upon the downstream effects, namely the loss of cell survival. That is, whatever proteins are influenced, the important question remains “did that effect cause the cells to die?”
Classes of compounds with nonspecific targets like the HSP90 inhibitors will surely be the most difficult to characterize at a genomic or proteomic level: What protein? What gene? Functional platforms like the EVA-PCD offer unique opportunities to study these classes of agents. We are convinced that the HSP90 inhibitors have a role in cancer therapy. It would be unfortunate if these setbacks led us to “throw the baby out with the (hot) bathwater,” thus, slowing or preventing their use in cancer treatment.