November 2013, Vol 2, No 7
Bench to Bedside, Gene to Drug Discovery Defines New Era of Precision Medicine
The next decade in anticancer drug discovery promises to be a complicated era of attempts to further define and overcome tumor heterogeneity, cancer evolution, and drug resistance, said Paul Workman, PhD, DSc, at the second annual Global Biomarkers Consortium conference. Killing off multiple cancer cell populations early in the disease through combinations of drugs may represent the future of personalized cancer treatment.
Anticancer drug discovery starts at the molecular level, where the abnormalities in genes and biochemical pathways that cause cancer are identified, allowing for the discovery of gene products that become potential drug targets.
Building on the prior decade of characterization of large-scale genomics, “we now have an incredible amount of information to exploit,” said Workman, head, Division of Cancer Therapeutics, The Institute of Cancer Research, London, United Kingdom.
“Although the amount of genetic complexity in tumors has proven to be huge, we can reduce it down to maybe a dozen or so survival pathways,” he said, which means that the challenge, although daunting, is not insurmountable.
The goal in discovery is to exploit oncogene, non-oncogene, and metabolic addictions; microenvironment dependencies; and vulnerabilities of cancer cells to discover innovative small-molecule drugs and essential biomarkers that will constitute the personalized cancer medicine of the future.
Discovery of genes, drugs, and biomarkers is an integrated process in the quest for precision diagnosis and treatment of cancer. Requirements for a drug discovery project start with evidence of a cancer gene followed by development of an instructive gene algorithm that will serve as a target for the drug. In selecting the target, the clinical path and predictive biomarkers must be known. Patient selection markers must also be actionable. Ultimately, “we’re finding a weak spot in the otherwise strong phenotype of a cancer,” said Workman.
“The discovery of [tumor] heterogeneity means that we are going to have to revisit combinations, either longitudinally by switching the coded genome sequence, or vertically by combining multiple approaches, as the way forward,” he said. To use combinations optimally, drugs must be available for identified molecular pathways and cancer hallmark traits (ie, self-sufficiency in growth signals, their resistance to inhibitory signals and their own programmed cell death, their ability to multiply, a limitless reproductive potential, and sustained angiogenesis).
He and colleagues conducted a primary computation analysis of biological and chemical space to prioritize targets for therapeutic exploration and to identify potentially druggable cancer-associated proteins that had been poorly explored. They discovered that of the 500 cancer genes known, 95% did not have a drug developed to act on it. “That’s a devastatingly low basis from which to build armamentariums of combinations,” said Workman.
They recommended prioritizing targets based on biology, 3-dimensional structure, and druggability. Forty-six genes have no associated chemical matter (ie, inhibitor) despite being predicted as druggable. “A massive amount of the cancer genome is waiting for drug discoveries,” he said.
Histone H3.3 Mutation as a Driver of Pediatric Glioblastoma
One example of a discovery project is the attempt to find a drug that targets a histone 3.3 mutation in pediatric glioblastoma. The breakthrough came when novel, specific mutations in the histone H3.3 gene in pediatric glioblastoma were discovered. Key residues in these genes are associated with posttranslational modification of the histone tail. K27M was predicted to be repressive, while G34 was activating for gene expression. These predictions proved correct. Gene set enrichment analysis confirmed that G34 mutations define distinct expression subgroups of pediatric glioblastoma.
Chromatin immunoprecipitation sequencing for gene localization of H3Kme3 marks and RNA polymerase II identified MYCN – a neuroblastoma-derived proto-
oncogene – as the top hit. These results were confirmed by quantitative polymerase chain reaction. Knockdown of MYCN was found to reduce cell viability in G34-mutant pediatric glioblastoma, but MYCN is not druggable directly, said Workman.
A robotic siRNA kinome screen revealed that the kinases CHK1 and AURKA stabilize MYCN and thus may be candidate targets in the treatment of G34-mutant pediatric glioblastoma.
Heat Shock Protein 90 as Cancer Chaperone
Heat shock protein 90 (HSP90) represents a cancer-selective network target to overcome tumor resistance. HSP90 is especially important for folding and activation of oncogenic client proteins, including mutant/amplified oncoproteins and drug-resistant forms. Inhibitors of HSP90 hit multiple cancer targets, pathways, and hallmark traits; hence they can act as network drugs, said Workman.
HSP90 inhibitors are now in phase 2/3 clinical trials, showing particular promise in HER2+ breast cancer and EGFR-mutant, ALK-rearranged non–small cell lung cancer (NSCLC), consistent with activity where the key oncogenic driver is an HSP90 client.
The HSP90 inhibitor AUY922 exploits heightened dependence of oncogenic client proteins on HSP90. The ERBB2 client protein is highly dependent on HSP90. This protein is rapidly and extensively depleted by AUY922 in ERBB2+ breast cancer cells. In phase 1 studies, ERBB2+ breast tumor xenograft was highly responsive to AUY922. Activity has been confirmed in trastuzumab-refractory ERBB2+ breast cancer as well as in EGFR-mutant and ALK-translocated NSCLC. This agent has a broader potential to treat prostate cancer, he said.
HSP90 may represent part of a 2-drug combination used early in the treatment of cancer to kill multiple cancer cell populations at once and prevent resistant tumor cells from arising, said Workman.
“As a drug developer, identifying network combination targets is the key to overcome heterogeneity,” he said. “I think resistance will be overcome by adaptive combination treatments in the future, and biomarkers are going to be essential to guide therapy.”
Clinical informatics – the use of data from the electronic medical record (EMR) to inform the development and improvement of evidence-based medicine, the dissemination and implementation of health outcomes research, and translational research initiatives – can facilitate the validation and translation of biomarkers, said Julie Lynch, PhD, RN, MBA, at [ Read More ]
Incorporating personalized medicine into everyday oncology clinical practice will require new paradigms in an effort to match cancer patients with the best therapies and attempts to treat solid tumors at an earlier stage with targeted agents, said Razelle Kurzrock, MD, at the second annual Global Biomarkers Consortium conference. “In a [ Read More ]