Precision Medicine-An Inevitable (Though Challenging) Goal!

Andre Goy, MD

Editor-in-Chief of

Oncology & Biotech News

Chairman and Director Lymphoma Division Chief John Theurer Cancer Center at HackensackUMC Chief Science Officer and Director of Research and Innovation Regional Cancer Care Associates Professor of Medicine, Georgetown University

The concept of linking chromosomes to cancer was first proposed by Boveri in 1902. The notion that cells carry surface receptors was established in 1905 by Ehrlich, who also coined the term chemotherapy. At that time, however, surgery and radiation were the only strategies used in the treatment of tumors, including in lymphosarcoma. The era of cancer chemotherapy did not officially begin until the 1940s, with the first use of nitrogen mustard in lymphoma (after its declassification as a warfare agent) by Gilman and Goodman at Yale and folic acid antagonists by Sidney Farber in Boston.

From then on, many new compounds (mostly cytotoxics) were established using the standard clinical trial approach. Combination regimens appeared in the mid-1960s, thanks to Holland, Freireich, and Frei, who hypothesized that cancer chemotherapy should follow the strategy of antibiotic therapy for tuberculosis. Subsequently, all combination chemotherapy regimens were developed empirically by using drugs with distinct mechanisms of action affecting DNA (directly or indirectly) or cells’ microtubules. Retrospectively, it seems oncologists had hit a wall in terms of results for several decades, though dramatic progress was being made toward understanding the mechanisms underlying cancer.

The enormous progress of cell biology applied to cancer cells led to the publication of the initial classic six hallmarks of cancer in 2000, which then appeared much more complex in the updated version a decade later (Cell. 2011;144[5]:646- 674). Meanwhile, it was conceptualized then that cancer cells relied on deregulated pathways—particularly either loss of tumor suppressor genes or gain of oncogenes function, leading to the concept of oncogene addiction—which in itself provided an obvious rationale for targeted intervention.

The Human Genome Project, completed in 2003, mapped the 25,000 genes in the genetic code and ushered in the era of personalized medicine. The era of targeted therapy was launched with targeting HER2 with trastuzumab in breast cancer (Slamon et al. N Engl J Med. 2001;344[11]:783-792.) and with targeting BCRABL with imatinib in chronic myeloid leukemia (Druker et al. N Engl J Med. 2001;344[14]:1031-1037). Since then, an impressive number of targeted approaches have been developed that either help pick a drug (eg, crizotinib for ALK-positive NSCLC, erlotinib for EGFR-positive NSCLC, PARPi in BRCA1/2-mutated malignancies, vemurafenib in BRAF V600E-positive melanoma) or identify patients for whom a drug is unlikely to benefit (eg, patients with KRAS-positive colorectal cancer are unlikely to benefit from anti- EGFR therapy).

Though these examples are quite dramatic, unfortunately they represent only a true minority of patients. The Human Genome Project will definitely remain as a landmark in the evolution of medicine at large, though a dramatic acceleration has taken place since then. The high throughput technologies developed then have led to the genomics of cancer, which have revealed a daunting diversity, both interpatient, but also within a patient (clonal heterogeneity).

The heterogeneity of cancer arises in large part from genetic variation, where the random mutation frequency in human cancer cells is over 100- to 500-fold greater than in adjacent tumor cells. The mutational landscape (insertions, deletions, fusions, etc) of cancer is extremely complex (beyond the point of this editorial) and multilayered -omics have emerged from DNA, RNA, and epigenetic differences, to metabolomics and eventually, functional proteomics. This remarkable tumor heterogeneity will likely represent a major challenge to personalized medicine and biomarkers development.

It is clear, however, that even for tumors with thousands of mutations, a limited number of them are critical in cancer cell behavior—driver mutations—while many others are considered passenger mutations—potential noise in data interpretation. While the price of whole genome sequencing has dramatically improved—Illumina launched its HiSeq X Ten sequencer this year, which delivers the first $1000 genome at 30x coverage— the difficulty resides in the interpretation of the data from complex bioinformatics to actionable items in routine practice. Massive parallel sequencing or next generation sequencing offers a platform that can deliver answers (profiles) within 2 weeks and usually focuses on mutations (a few hundred) believed to be relevant in cancer cell biology.

While the technology keeps evolving rapidly and might make it easier (smaller amount of materials needed or profiles based on circulating tumor cells [CTCs]), it becomes clear that an enormous effort is needed to implement these changes in the current and future management of cancer patients. The goal is to ultimately be able to integrate all these -omics data through more crossdisciplinary biology approaches and computational science—the foundation of the promising and elegant field of systems biology led by Leroy Hood (see our upcoming “Applications of Systems Biology in Cancer” conference on May 12, 2014, at the John Theurer Cancer Center).

Emerging efforts to develop molecularly relevant clinical trials still face a lot of challenges: from access to tissue (with repeat biopsies critical to understanding clonal evolution and resistance), to defining relevant subsets of important mutations (to reduce noise and improve feasibility), to eventually analyzing pathways/ networks that hopefully will guide our decisions in the future. There is no question that the era of precision medicine has arrived, and we need to embrace it and resolve together the challenges that come with it, including at the regulatory level.

The ultimate integration of -omics data at large will help improve cancer patients’ outcome at several levels: (1) early detection (before any symptoms or clinical abnormality) using, for example, CTCs, serum proteins, or exosomes signatures (2) patient stratification at diagnosis (eg, identify patients who will do well with a classic regimen versus offering a new option) (3) build clinical trials that are molecularly relevant (eg, adaptive trials) (4) assessment of disease progression, response, or resistance to therapy (5) eventually redefining the classification of the tumor itself through the ongoing work of large consortia (TCGA and ICGC).

Implementing precision medicine requires a global effort from clinicians, insurance providers, government regulators, and the pharmaceutical industry. Finally simplifying the drug development process (away from phase III trials) by focusing on targeted populations will help reduce costs, while giving the right drug to a given patient, and will translate into a better “mileage.” This in itself will help control cancer care cost by reducing the use of empiric and palliative ineffective therapies. “Smarter medicine is better medicine!”