Supplement: HSP90 Inhibition Turns Up the Heat on Tumors


April 1, 2018 (Vol. 38, No. 7)

Vicki Glaser Writer GEN

If “Cold” Tumors Become “Hot,” They Respond Better to Checkpoint Blockade Immunotherapy

The emergence and stunning early success of checkpoint blockade, a therapy that helps the immune system recognize and attack cancer cells, has garnered well-deserved headlines. Checkpoint blockade, however, benefits only a minority of cancer patients. While these patients harbor tumors that are responsive to checkpoint drugs such as anti-PD-1 or anti-CTLA-4 monoclonal antibodies, many patients have conditions that are refractory to checkpoint inhibition—melanoma, non-small cell lung cancer, gastrointestinal tumors, and other types of cancer.

Clinicians are left looking for solutions that would broaden the population that would benefit from checkpoint blockade. One potential solution is to flip a switch that would turn so-called “cold” tumors into “hot” tumors. Cold tumors are refractory to checkpoint blockade. (They may respond initially but become unresponsive). Hot tumors are responsive.

Recent research points to heat shock protein (HSP) 90 inhibition as a potential mechanism for enhancing cancer immunotherapy. In vitro studies in human melanoma cell lines have shown that HSP90 inhibition can enhance T cell–mediated killing of tumor cells via upregulation of interferon response genes. For example, in one recent paper (Mbofung RM, et al. 2017. Nat. Commun.), studies in mouse cancer models showed that HSP90 inhibition could enhance responses to immune checkpoint blockade therapy.

“I found these data quite compelling,” says Steven Hall, Ph.D., president and CEO of Esanex, which is developing SNX-5422, an ester prodrug of the active compound SNX-2112. “We now need to replicate some of the published work with different structural classes of HSP inhibitors.”

In recent years, numerous and varied HSP90 inhibitors have gone through preclinical testing and at least early-stage clinical trials as single-agent therapies in various types of cancer with mixed and generally less than compelling results. “We need to move beyond the paradigm of doing studies in molecularly driven cancer. None of these drugs has even gotten close to regulatory approval,” insists Rodabe N. Amaria, M.D., assistant professor, department of melanoma medical oncology, University of Texas MD Anderson Cancer Center.

“The idea of branching out and using novel combinations with immunotherapy is really the smartest way to go,” she continues. “It’s going to take some work, but there is a niche in cancer to try to find drugs specifically for the checkpoint-refractory population. Most of the Phase I trials we are designing are specific for that patient population. If this really does work and it can make a cold tumor into a hot tumor, then that is definitely a way forward to FDA approval.”

Illuminating a Path Forward

In the study published in Nature Communications in 2017, a research team from MD Anderson Cancer Center screened a diverse collection of 850 bioactive compounds to assess their ability to enhance the killing of primary melanoma cell lines by autologous T cells. HSP90 inhibitors as a group were among the most highly active compounds in the screen, and subsequent experiments demonstrated the upregulation of a family of interferon response genes (interferon-induced proteins with tetratricopeptide repeats, or IFITs) as the mechanism underlying their effect on T cell–mediated killing.

Furthermore, tumor-bearing mice treated with a combination of the first-generation HSP90 inhibitor ganetespib and anti-CTLA4 or anti-PD1 monoclonal antibodies had a better antitumor response than did mice given either treatment alone. The researchers also showed that the combination of HSP90 inhibition and checkpoint blockade with anti-CTLA-4-enhanced CD8 T-cell function.

“I think it is reasonable to design a Phase I trial,” says Dr. Amaria. “The main issue is what drug will you use.” Reviewing the existing HSP90 inhibitors, she admits that she is unsure whether any has proven that it is superior to any another. She notes, however, that an oral drug would be easier to administer and more convenient for patients.

“The data that have been generated so far are really strong,” Dr. Amaria asserts, but before a Phase I combination immunotherapy trial is initiated with any existing HSP90 inhibitor, the drug should undergo the same type of modeling studies described in the Nature Communications paper to prove that its efficacy is comparable to that of ganetespib.

New Role for HSP Inhibition

In bacteria, HSPs are activated under conditions of physiological stress. They have a broader range of activities in human cells. The primary role of HSPs is to facilitate the proper folding of cellular proteins. If not properly folded, the proteins will be unstable and subject to degradation by the ubiquitin-mediated pathway. This function is even more essential in tumor cells in which the mutation burden makes it more likely that proteins will misfold.

“HSP90 plays a pivotal role in maintaining transformation and in elevating the survival and growth potential of cancer cells, since its essential chaperoning functions are exploited to facilitate the acquisition and maintenance of the malignant phenotype,” wrote the authors of another recent paper (Sidera K, Patsavoudi E. 2014. Recent Pat. Anticancer Drug Discov.). “In addition, cancer cells experience various types of stress in their microenvironment such as acidosis, hypoxia, and nutrient deprivation and therefore exhibit higher requirements for HSP90 function in order to tolerate alterations and survive these hostile, stressful conditions.”

HSPs do not exist or function in isolation. Each is part of a multicomponent complex that incudes co-chaperones. The composition and structure of the HSP complex allows for the binding of a client protein and proper positioning into its active conformation. An HSP inhibitor can compete for the HSP’s active site or block binding of a critical co-chaperone, for example, preventing the HSP complex from carrying out its function.

Recent data have shown that HSP90 inhibitors do not all work the same way and are not equally effective. There are some similarities depending on the client proteins, but with some client proteins, HSP90 may be affected more by one inhibitor than another, likely due to how the inhibitors bind.

If a clinical trial is a reasonable next step, then it should start with a Phase I study to show that the drug selected is safe and well-tolerated at plasma levels that are sufficient to inhibit HSP90. Ideally, a randomized, blinded Phase II study would follow. If such a study compared PD-1 immunotherapy alone to combination therapy with anti-PD-1 checkpoint blockade and HSP inhibition, it would allow tumor response, patient survival, and progression-free survival to be assessed. In addition, a biopsy cohort could be studied to demonstrate that HSP90 inhibition can turn a cold tumor into a hot tumor able to respond to checkpoint blockade.

Current Development, Future Promise

According to Dr. Hall, the first generation of HSP90 inhibitors had toxicity issues that limited the frequency with which they could be dosed. Due to their ocular toxicity and other adverse effects, it was not possible to assess the potential effectiveness of these early compounds. Subsequent regimens using synthetic small molecules were better tolerated. The availability of orally administered small molecule HSP90 inhibitors also simplified later regimens.

A crucial stumbling block in the early development of HSP90 inhibitors as anticancer monotherapies was the lack of data on target tumor indication or patient selection for clinical testing. “The lack of efficacy led many companies to walk away from this area,” Dr. Hall recalls.

Among the HSP90 inhibitors that had progressed to at least Phase I clinical trials, but ultimately were put on the shelf, are 17-AAG from Bristol-Myers Squibb, AUY922 from Vernalis, 17-DMAG from Kosan Biosciences, STA-9090 from Synta Pharmaceuticals, and KW-2478 from Kyowa Hakko Kirin. (STA-9090 progressed to a Phase III trial).

XL888, from Exelixis, is in institution-sponsored trials in combination with chemotherapeutic agents in advanced gastrointestinal cancer and melanoma, and AT-13387, from Astex Pharmaceuticals, is in development by the National Cancer Institute, with multiple ongoing trails in advanced, metastatic, and recurrent solid tumors.

Esanex has completed two clinical studies with SNX-5422: one to treat neuroendocrine tumors in combination with everolimus, and the other to treat epidermal growth factor receptor (EGFR)-wild-type non-small cell lung cancer, which showed a nearly 40% objective response rate in combination with carboplatin and paclitaxel. Based on the positive results from both of these trials, the company plans to move forward toward comparative trials in these indications.

“We are now thinking about merging our existing clinical data with the new information on the role of HSP90 inhibitors in stimulating the immune response to tumors,” states Dr. Hall. “We are in discussions about a trial that would combine an HSP inhibitor with a checkpoint inhibitor.”

Success will depend largely on picking the right drug combinations with synergistic mechanisms of action, the right tumors, and the right patient populations for clinical testing.

Technologies Impacting Immunotherapy

Cliff Hoyt, Ph.D., and Anis Khimani, Ph.D.

Immunotherapeutic strategies are becoming a primary focus of cancer research because of their significant efficacy and lasting benefits. For example, anti-PD-1 and PD-L1 drugs look promising. However, response rates to monotherapy remain in the 20–40% range, driving the need for a better understanding of the interactions between cancer and the immune system. Such an understanding could lead to new immunotherapy strategies, including those for combination therapies.

It is imperative to identify biomarkers that reveal, with cell-level specificity, the biological response to immunotherapy. Such biomarkers could stratify patient populations, for example. The need for such biomarkers is growing, especially given the rapid introduction of new therapies.

Technological advancements have enabled the identification of such biomarkers at the molecular level, at the cellular level, and in intact tissue sections.

PerkinElmer has taken a holistic approach for translational research to evaluate numerous parameters within a single sample. Multispectral imaging enables detection of up to seven protein or RNA markers while retaining spatial context and cellular arrangement.

For example, PerkinElmer’s Vectra® Polaris™ platform has been used by one team of researchers to evaluate several types of tumor-infiltrating lymphocytes. This work included the measurement of spatial parameters. (For details, see Feng Z, Puri S, Moudg T, et al. 2015. J. Immunother. Cancer 3: 47.)

Simultaneous analysis of expression and spatial parameters via unsupervised hierarchical clustering improved the stratification of patients into classes of positive and negative outcome. Furthermore, multiparametric profiling apparently pointed to tumor-responsive T cells, suggesting new therapeutic strategies via checkpoint blockade.

The tissue-based multiplex platform has been used in preclinical and clinical research, upstream drug discovery, and workflow development. In addition, the platform has integrated multimode detection solutions such as AlphaLISA® technology that allow it to advance immuno-oncology research and screen for inhibitors of PD-1 and PD-L1 binding.
Cliff Hoyt, Ph.D., is an oncology fellow, tissue applications and collaborations, and Anis Khimani, Ph.D. (, serves as strategy leader and director of applications, both at 


Lung cancer tissue stained for PD-L1 and several immune cell markers and imaged using PerkinElmer’s Vectra® Polaris™ platform.

This site uses Akismet to reduce spam. Learn how your comment data is processed.