DeVita. Cancer

15

Precision Medicine in Oncology

James H. Doroshow

INTRODUCTION Novel insights into the biology of cancer have been translated into improvements in clinical care at an accelerating pace over the past 15 to 20 years. 1 The introduction of increasingly sophisticated mo- lecular tools with which to interrogate both individual cancers and populations of patients with cancer has led to a steady stream of new diagnostic and therapeutic interventions that have altered the natu- ral history of several solid tumors and hematopoietic malignancies. 2,3 From the concurrent approval of trastuzumab and a companion diagnostic for breast cancer patients whose tumors overexpress the HER2 gene in 1998 4 to the recent (May 2017) approval by the U.S. Food and Drug Administration (FDA) of the anti–programmed cell death protein 1 (anti-PD-1) antibody pembrolizumab for the treat- ment of adults and children with solid tumors that demonstrate high microsatellite instability or mismatch repair deficiency independent of cancer histology, 5 the application of molecular tumor characteri- zation in the clinic has facilitated introduction into oncologic prac- tice of novel small molecules and immunotherapies applicable to a variety of malignancies without substantive prior therapeutic options, effectively ending the development of nonspecific cytotoxic agents. 6,7 The approach to precision cancer medicine exemplified by these recent developments, and described in a recent National Academy of Medicine report, 8 can be defined as follows: an intervention to prevent, diagnose, or treat cancer that is based on a molecular and/ or mechanistic understanding of the causes, pathogenesis, and/or pathology of the disease. Where the individual characteristics of the patient are sufficiently distinct, interventions can be concentrated on those who will benefit, sparing expense and side effects for those who will not. The first generation of clinical trials evaluating the feasibility of matching a range of treatments to specific molecular features that can be measured in an individual patient’s malignancy has provided some notable successes as well as demonstrated certain limitations of this approach that will require improvements in pre- clinical modeling, diagnostic assay development, and clinical trial design to overcome. 9,10 However, viewed from the perspective of the therapeutics discovery model prevalent during the last half of the 20th century, 11 reorientation of that paradigm to focus on the need to develop innovative tools and molecules for translating modern cancer biology into the clinic for individual patients (“precision oncology”) has already paid substantive dividends. 12 This dramatic change in the development of cancer therapeutics is, perhaps, best exemplified by the fact that of the more than 50 new systemic treat- ments for cancer approved by the FDA in the past 5 years (2013 to 2017) that were not formulation variants, only 2 were cytotoxic agents, and these possessed pleiotropic mechanisms of action. APPROACHTO PRECISION MEDICINE IN ONCOLOGY The overall approach to the development of novel diagnostics and therapeutics for one patient at a time, as outlined in Figure 15.1, depends on the interrogation of improved preclinical models that

can be used to validate molecular targets at a functional level and permit the evaluation of mechanism(s) of action in vivo by assays that can be definitively transferred from the preclinical to the clin- ical setting. In this way, tumor tissues obtained prior to treatment can be used to assign specific patients to clinical trial arms based on the molecular characteristics of their disease. Patients could then be monitored both clinically and radiographically to deter- mine drug exposure levels and the efficacy of treatment, while subsequently undergoing posttreatment molecular analyses of specimens of tumor or the tumor microenvironment (obtained either by direct biopsy or by monitoring circulating tumor cells [CTCs], circulating tumor DNA, or circulating endothelial cells [CECs]), and by functional molecular imaging to demonstrate en- gagement of essential therapeutic targets. It is important to point out that molecular monitoring in this fashion is not exclusively performed using genomic analy- ses; rather, conceptually, a variety of molecular characterization methodologies (transcriptomic, proteomic, immunohistologic, epigenetic, or image-guided), in addition to somatic and germline genomics, are used based on the disease and biologic contexts under examination. 13,14 In fact, it is likely that with technologic advances over time dynamic/functional tumor profiling may allow the type of biochemical pathway analysis that is required for opti- mal therapeutic decision making. 15 Next-Generation DNA Sequencing for Precision Oncology The rapid improvement in massively parallel DNA sequencing capacity over the past decade has made it possible to acquire enor- mous amounts of tumor DNA sequence data at rapidly declining cost over shorter time frames; this remarkable advance has occurred together with the development of improved bioinformatic tools for analysis of an increasing body of genomic information. 16 With this capacity, the somatic mutational frequencies (at the  2% level) of the most common human tumors have now been characterized in the primary disease setting, 17 and studies of the mutational land- scape in metastatic cancers have recently become available. 18,19 In this context, the use of pretreatment multigene profiling to direct therapeutic choice has progressed rapidly over the past 5 years. 20 There are multiple clinical programs at major U.S. and international medical centers that use gene panels measuring de- fined mutational variants to assist clinical decision making. Some sites apply locked-down algorithms, and others use expert tumor boards to guide the choice of specific targeted agents for individ- ual patients based on their mutational profiles. 21,22 Currently, most commercially available and academic sequencing efforts focus on panels of genes and mutations that can be examined at great depth of coverage within a less-than-2-week time frame using both fresh and formalin-fixed, paraffin-embedded specimens to examine from 20 to many hundreds of genes (with much larger numbers of variants). 23 However, one of the major current limitations of next-generation sequencing (NGS) panels is that the range of genes

Copyright © 2018 Wolters Kluwer Hea th, Inc. Unauthorized reproduction of he article is prohibited.

examined is, for the most part, limited to genomic alterations that

186