In recent years, precision medicine has rapidly altered the oncology diagnostic and treatment spectrum.1 Technological advances for characterizing patients’ genomic, proteomic, metabolomic, and cellular profiles, combined with the development of large biological databases along with computational tools to analyze them, have allowed clinicians to tailor treatment strategies to precisely target the molecular alteration underlying individual patients’ disease.1,2 This may prevent or delay the need to use treatments with less favorable tolerability profiles that are also potentially more costly.3
The recent advances in precision medicine have helped pave the way for the US Precision Medicine Initiative, which launched in 2015 and has helped usher in continued innovation in the field.2 Oncology has been designated as the near-term focus of the Precision Medicine Initiative,2 guided by new knowledge of the oncogenic mechanisms of cancer that has already influenced risk assessment and therapeutic strategies. The development and expanding use of targeted therapies to counter specific molecular drivers have already yielded compelling results, further supporting molecular signatures as representatives of a promising treatment decision tool.2
Although new discoveries in precision medicine have led to significant steps forward in diagnosis and treatment in oncology, precision medicine is nonetheless a relatively new concept and practice in the field. Prior to 2005, Sanger sequencing was used to sequence a few genes at a time.4 Sanger sequencing had low throughput, and it also could not detect structural rearrangements.4 Early companion diagnostics for targeted therapies mostly were single-gene assays, but increased knowledge of tumor molecular biology has resulted in the testing of patients for a wider number of genes.5
Next-generation sequencing (NGS) overcame the limitations of Sanger sequencing.5 NGS can detect genetic alterations, such as rearrangements, copy number alterations, insertions, and deletions; it also comes at a lower cost.5 NGS techniques are high throughput and are able to analyze multiple DNA sequences in parallel. Their multigene panels represent a targeted approach for sequencing a number of genes simultaneously.6 These panels can be individualized to the tumor types, and they differ according to the genes being tested.4 For example, targeted breast/ovarian panels may test for 6 or more genes, whereas a comprehensive panel that tests for numerous cancer susceptibility genes may test for 70 or more.6 Clinicians can tailor a multigene panel to the genes of interest. Panel options are also updated frequently as new genes are discovered.6 Genomic advances lead the way to the development of treatments that target different molecular subclasses of tumors,7 and they delineate subgroups of patients who are more likely to benefit from targeted agents.8 These tests can also identify actionable driver mutations and underlying mechanisms of drug resistance to reveal patients who are likely to be resistant to therapy.8 With NGS, a patient’s cancer genome can be deciphered efficiently, allowing cancer biologic systems to be studied at a pace that was not possible previously.8
The number of FDA-approved companion molecular diagnostic assays is growing.9 So far, most of these approved assays inform the use of a specific class of agents for specific cancers. However, the FDA recently approved comprehensive NGS diagnostic 468- and 324-gene panels that capture actionable molecular alterations of targets that can be treated with approved therapies or with therapies under clinical development.10-12 The approval of these NGS panels suggests that NGS testing is becoming a standard of care in oncology.10 Certainly, the paradigm is shifting from single-gene assays to inform a single therapeutic option in specific cancers to an NGS approach that can detect the patient’s entire mutational catalog.10
Results from observational studies have shown progression-free survival (PFS) benefits from using molecular-guided treatments.13-15 In one study, investigators found PFS benefits in the community setting among patients with advanced cancer who received precision medicine treatments compared with matched historical controls who received standard treatments (hazard ratio, 0.47; 95% CI, 0.29-0.75).14
Approaches to biopsies are also evolving. Liquid biopsies have been developed as a way to detect potentially actionable molecular alterations when tissue biopsies are not feasible. Liquid biopsies detect alterations in circulating tumor DNA (ctDNA), which are small fragments of DNA released by malignant lesions.16 In solid tumors, including breast cancer, high rates of concordance between liquid and tissue biopsies have been demonstrated.16 Results of one study showed that 65% of cancers had detectable ctDNA alterations—the majority of which were actionable with an FDA-approved drug.17 In addition to their use when tissue biopsies cannot be performed, liquid biopsies can allow repeat genomic profiling to detect tumor response, evolution, and resistance, as well as to interrogate the genomic signatures from multiple metastases.16
In addition to genomic data, other forms of precision data are on the horizon, including transcriptomics, epigenomics, proteomics, metabolomics, and digital pathology from the tumor, surrounding tissues, circulating blood, and other body fluids.18 Other information, including electronic health record (EHR) data, radiographic features (radiomics), and patient-reported outcomes (personomics), also contributes to a comprehensive profile of the patient’s condition.18,19 To interpret increasingly complex and large-scale biological databases, big data analytic approaches are needed to provide data quality control, analysis, model building, interpretation, and validation.19
Precision Therapeutics in Oncology
There are many examples of molecular-guided therapies in oncology. The field has witnessed rapid advancements in molecular target validation and development of targeted therapies in breast cancer and melanoma, in particular.
Breast cancer continues to be the most commonly diagnosed malignancy among women in the United States.20 The discovery and validation of estrogen receptor (ER) expression as a therapeutic target were followed by the recognition of human epidermal growth factor receptor 2 (HER2) overexpression as another molecular therapeutic target.8 HER2 overexpression is found in 15% to 30% of patients with metastatic breast cancer and carries a poor prognosis.21 The advent of HER2-targeted agents such as trastuzumab, trastuzumab emtansine, pertuzumab, lapatinib, neratinib, and afatinib was a significant therapeutic breakthrough.21
Beyond ER and HER2, advances in high-throughput genomic technologies have enabled a better understanding of the molecular pathways that lead to tumor progression. Examples of these molecular alterations in breast cancer can be seen in Table 1.21-30 Molecular alterations leading to tumor progression and/or drug resistance can be detected by various technologies,22 and some can be targeted by currently approved therapies.21,25,26
Other molecular targets are being investigated in breast cancer as well. For example, studies of AKT inhibitors are being evaluated in patients with AKT1 mutations.22 When the number of genomic alterations and validated treatment targets becomes large enough, multiplex NGS technologies may replace the companion diagnostic approaches based on Sanger single-gene sequencing.22 In addition to target validation, another area under active investigation addresses the question of whether genome-guided therapies improve outcomes in breast cancer. A randomized phase 2 clinical trial in metastatic breast cancer is under way to compare standard maintenance chemotherapy versus targeted maintenance treatment based on molecular anomalies identified by high-throughput multiplex genome analysis (NCT02299999).31 Studied outcomes include PFS and overall survival.31 Results of this study will inform whether the use of high-throughput genome analysis is a valuable therapeutic decision tool in advanced breast cancer.
One of the applications of liquid biopsies under development is profiling of nipple aspirate fluids.32 Nipple aspirate samples reflect the entire ductal-alveolar tree and can provide a complete biological profile of the patient’s breast cancer.32 A recent study demonstrated the proof of concept in using nipple aspirate fluid to discover breast cancer biomarkers.33 When fully validated, liquid biopsies of nipple aspirate fluids may offer a more efficient alternative to detection of breast cancer markers and growth factors in the blood. Besides providing diagnostic and prognostic information, nipple aspirate fluid biopsies may also overcome the massive dilution when blood is used for biomarker detection.33
Melanoma is another malignancy that has seen a rapid advancement in the understanding of its molecular biology and the development of targeted therapies.34 Melanoma has a remarkably high rate of acquired mutations compared with other tumor types, including in the epidermal growth factor receptor, insulinlike growth factor receptor, and mitogen-activated protein kinase pathways.34 Immune dysregulation also plays an important role in the pathophysiology of melanoma. Expression of the cytotoxic T-lymphocyte—associated protein 4 and programmed cell death protein 1 receptors on the T-cell surface facilitate the inactivation of T cells, thus preventing the immune system from destroying cancer cells.35 Molecular targets in melanoma and their associated FDA-approved therapies, as well as those under investigation, are further described in Table 2.34,36-44
Although precision medicine has the potential to contribute to high-value healthcare by improving outcomes and decreasing cost, its full potential can be realized only when regulatory, economic, and technical barriers are addressed.1 Currently, the landscape of genomic data is fragmented and not well incorporated into clinical workflows, making it difficult for clinicians to select among the various diagnostic methods and treatment options.45
Improved integration of cancer genomic data into EHRs will further support clinical decision making.46 Key challenges of integrating cancer genomic data into EHRs include the complexity of translating raw genomic data into clinically meaningful and actionable results, the need for human curation in the interpretation of reported genetic mutations, and the lack of consistent standards in nomenclature in reporting genetic mutants.46 Platforms that overcome these barriers and integrate clinical and molecular data with workflow management are in development, and at least 1 of these platforms has been piloted within a community health system.46,47 The Electronic Medical Records and Genomics Network Consortium provides recommendations on how to integrate genomic information into EHRs.19 In addition, Health Level Seven International has proposed interoperability standards so that clinicians can use genomic information from other clinics and hospitals.19
Even if just genomic data are considered, tumors often have multiple driver mutations; thus, oncologists frequently need to combine different streams of data to make treatment decisions.45 Some major cancer care institutions have developed various systems to incorporate complex genomic results into patient care, but scaling these programs to other healthcare systems may be challenging.18 Efforts are under way to develop algorithms through bioinformatics to find gene signatures and targeted regimens that match the molecular profile of a specific patient’s tumor.18
Another implementation challenge is the cost associated with testing for genomic alterations and subsequent targeted therapy. An analysis of billing for NGS panels submitted to payers suggests that only one-third of patients receive reimbursement.18 Despite the high cost of targeted therapies, precision medicine allows clinicians to tailor treatments to individual patients and thus reduce the costs of ineffective therapies.48 Other uncertainties associated with the economic evaluation of precision medicine include methodological uncertainties, structural uncertainties (eg, choice of decision analytic model), and parameter uncertainties (eg, uncertainty in clinical effectiveness of a strategy).49 Methods and cost-effectiveness may also differ significantly based on testing strategies, NGS technologies, and cancer type.50 Therefore, much work remains regarding the economic impact of precision medicine, particularly as the technology and testing methods continue to evolve.51
The lack of clarity on policies regarding precision medicine presents another challenge. More than 10 guidelines from professional/medical associations exist pertaining to the standardization of NGS for clinical use.52,53 A clear need exists for collaboration among different stakeholders to avoid conflicting recommendations and policy redundancy so that precision oncology can be more efficiently implemented.54 Initiatives to facilitate the implementation of genomics in clinical practice are under way, including the National Human Genome Research Institute’s Implementation of Genomics in Practice project and the National Academies of Sciences, Engineering, and Medicine’s roundtable report on applying implementation science to genomic medicine.54-56
Advances in genomic medicine are revolutionizing our understanding of and treatments for cancer. Molecular-guided therapy in oncology represents a proactive approach to treatment and has the potential to reduce the time and financial expenditures of ineffective treatments, thus increasing the patient’s duration and quality of life.3 Both breast cancer and melanoma have well-validated molecular targets that are actionable with FDA-approved therapies. These advances are also changing the diagnostic and treatment platforms of other forms of cancer.
Looking ahead, clinical studies such as the National Cancer Institute Molecular Analysis for Therapy Choice and the American Society of Clinical Oncology Targeted Agent and Profiling Utilization Registry will test the performance, including safety and efficacy, of using targeted agents based on genomic alterations in patients with advanced cancer, regardless of cancer type.57,58 Results of these studies will likely provide credence to the broader use of high-throughput genomic analyses as a treatment decision tool in oncology. Also, as NGS technology continues to evolve, the complex mutational landscape of a patient’s cancer may also be assessed longitudinally through repeated liquid and tissue biopsies during and after treatments, to define patients at risk of relapse.10 Preliminary efforts are also in progress to address implementation challenges (eg, workflow, EHR integration, economic evaluations) so that precision medicine can be more broadly applied in oncology clinical practice.