A group led by researchers at the Memorial Sloan-Kettering Cancer Center has identified a mutation in the androgen receptor that could drive resistance to second-generation antiandrogen drugs such as
Medivation Inc.'s Xtandi enzalutamide. The team's subsequent rational drug design studies yielded a series of molecules that could overcome the resistance mechanism.1

The researchers now are testing their lead molecule in mouse xenograft models for prostate cancer. If the results are promising, MSKCC could consider running a clinical trial.

Antiandrogen drugs are one of the standard treatment options for prostate cancer, but many patients progress to castration-resistant disease in 12-18 months.2
Acquired mutations in the androgen receptor (AR) are one of the key events that can cause resistance to antiandrogen drugs.3-5

Xtandi is a second-generation oral androgen receptor antagonist that received FDA approval in August 2012 to treat metastatic castration-resistant prostate cancer (CRPC) in patients who previously received docetaxel. Last month, the EMA's Committee for Medicinal Products for Human Use recommended approval of an MAA for Xtandi.

In the U.S., clinicians already are seeing patients whose tumors initially respond to Xtandi but then develop resistance.

"We wanted to know how this could happen," said Charles Sawyers, chair of the Human Oncology and Pathogenesis Program at MSKCC and an investigator at the Howard Hughes Medical Institute. "We suspected that mutations in the androgen receptor might be one mechanism and designed a screen to look specifically for them."

The researchers developed a mutagenesis screen to identify cells with mutations in AR that confer resistance to a particular drug. When applied to Xtandi, the screen implicated the F876L point mutation in AR as the dominant mutation that confers resistance to the drug.

In a series of validation studies using Xtandi-sensitive prostate cancer cell lines and mouse xenograft models, the researchers found that the F876L mutant AR could spontaneously emerge following prolonged exposure to Xtandi or ARN-509, a second-generation AR antagonist from Aragon Pharmaceuticals Inc. that is in Phase II testing for CRPC.

Xtandi and ARN-509 are both bisaryl-thiohydantoin AR antagonists. Sawyers co-discovered both molecules and is a cofounder of Aragon.

Follow-up structural modeling studies and in vitro assays showed that Xtandi and ARN-509 both antagonized wild-type AR but actually activated the F876L mutant receptor. Not surprisingly, both were ineffective at killing prostate cancer cells that expressed the mutant receptor.

Modeling studies suggested that modifications to the central B-ring structure in the Xtandi scaffold might restore antagonistic activity against F876L mutant AR.

The researchers used the insights gained from these assays and structural modeling studies to design and synthesize analogs from the Xtandi scaffold that have an additional cyclic hydrocarbon ring-dubbed the D-ring-attached to the molecule's central B-ring.

In prostate cancer cell lines, three of the resulting D-ring-substituted molecules from the series inhibited growth of prostate cancers that expressed the F876L mutant AR, whereas Xtandi did not.

Results were published in eLife. The MSKCC group collaborated with researchers at the Toyota Technological Institute at Chicago and The University of Chicago on the study.

"The key finding is the discovery of a mutation in AR that causes resistance to enzalutamide," said Sawyers, the co-corresponding author. "Now that we know about it, we can start to look for it in patients. We also showed that certain chemical modifications can be made to enzalutamide that overcome the drug resistance."

"This work paves the way for the development of a new series of antiandrogens and further supports the case that the androgen receptor could remain a critical and valid target even in patients who have late-stage disease," said Jeffrey Hager, senior director of biology at Aragon. "The data from this study suggest that among the patients that progress after initially responding to enzalutamide or ARN-509, there exists a subset who have tumors that remain fully addicted to the androgen receptor."

Hager said the findings also improve the field's understanding of ligand-specific mutations in AR by providing insights on how such a mutation can affect the activity of a given class of antiandrogens and how to modify such molecules to circumvent the mutation. "The androgen receptor is known to acquire ligand-selective mutations that can flip a molecule from being an antagonist into an agonist," he said.

Importantly, Hager added that such mutations do not typically impart agonist activity across all chemical classes of antiandrogens. Indeed, in vitro cellular assay data showed that structurally distinct first-generation AR antagonists such as Casodex bicalutamide and the generic hydroxyflutamide did not activate F876L mutant AR.

The assay also showed that mutant ARs activated by Casodex or hydroxyflutamide were not activated by Xtandi and ARN-509.

AstraZeneca plc markets Casodex to treat prostate cancer.

Clinical occurrences

The next step is determining the relevance of the F876L mutation in a clinical setting.

"It will be important to determine whether the identified mutation is found in patients who are being treated with enzalutamide or ARN-509 and if so the frequency at which it occurs," said Hager. "The latter will have a strong influence on how excited drug developers such as Aragon will be to develop compounds that can circumvent the resistance mutation."

Hager added that any compounds that do block the mutant receptor could be accompanied by a companion diagnostic to identify patients who are most likely to respond and to guide treatment decisions. He also said it would be a good idea to do sequencing studies on tumor samples from patients receiving second-generation antiandrogens to determine if there are mutations in AR that can spontaneously arise in the clinic but not in preclinical models.

"For such mutations, one will then need to go back to preclinical studies to elucidate how they affect the activity and function of the androgen receptor," he told SciBX.

Aragon has an ongoing internal program focused on understanding potential resistance mechanisms against second-generation antiandrogens such as ARN-509 and Xtandi, which parallels the work of Sawyers' group. Hager said the company is not disclosing additional details.

Meanwhile, Sawyers' group at MSKCC is evaluating DR103, the lead D-ring-substituted AR antagonist identified from the study, in mouse xenograft models. "If the results from these studies look promising, we will consider scaling up for more preclinical testing and a possible clinical trial at MSKCC," he told SciBX.

MSKCC has filed two provisional patent applications covering the AR mutation and the new chemical entities described in the eLife paper. MSKCC declined to disclose licensing details.

Medivation did not respond to requests for comment.

Lou, K.-J. SciBX 6(18); doi:10.1038/scibx.2013.429
Published online May 9, 2013


1.   Balbas, M.D. et al. eLife; published online April 9, 2013; doi:10.7554/eLife.00499
Contact: Charles L. Sawyers, Memorial Sloan-Kettering Cancer Center, New York, N.Y.
e-mail: sawyersc@mskcc.org
Contact: Yang Shen, Toyota Technological Institute at Chicago, Chicago, Ill.
e-mail: yangshen@ttic.edu

2.   Gulley, J. et al. Clin. Adv. Hematol. Oncol. 1, 49-57 (2003)

3.   Brooke, G.N. & Bevan, C.L. Curr. Genomics 10, 18-25 (2009)

4.   Feldman, B.J. & Feldman, D. Nat. Rev. Cancer 1, 34-45 (2001)

5.   Shi, X.-B. et al. Cancer Res. 62, 1496-1502 (2002)


Aragon Pharmaceuticals Inc., San Diego, Calif.

AstraZeneca plc (LSE:AZN; NYSE:AZN), London, U.K.

European Medicines Agency, London, U.K.

Food and Drug Administration, Silver Spring, Md.

Howard Hughes Medical Institute, Chevy Chase, Md.

Medivation Inc. (NASDAQ:MDVN), San Francisco, Calif.

Memorial Sloan-Kettering Cancer Center, New York, N.Y.

Toyota Technological Institute at Chicago, Chicago, Ill.

The University of Chicago, Chicago, Ill.