Pharm Med DOI 10.1007/s40290-017-0192-8
REVIEW ARTICLE
Integrating Genomics into Drug Discovery and Development: Challenges and Aspirations Rajiv Raja1 • Young S. Lee1 • Katie Streicher1 • James Conway1 • Song Wu1 Sriram Sridhar1 • Mike Kuziora1 • Hao Liu1 • Brandon W. Higgs1 • Philip Z. Brohawn1 • Carlos Bais1 • Bahija Jallal1 • Koustubh Ranade1
•
Ó Springer International Publishing AG 2017
Abstract Molecular biomarkers are increasingly being used to identify subgroups of patients that have a higher chance of benefiting from targeted therapies. Identification of predictive biomarkers and development of companion diagnostics to accompany targeted agents have been shown to significantly improve the efficacy and approval rate of these novel therapies, making treatment decisions more personalized to individual patients. Mutations of epidermal growth factor receptor (EGFR) and rearrangements of anaplastic lymphoma kinase (ALK) in non-small-cell lung cancer and of BRAF in melanoma provide great examples of driver mutations defining patient subgroups that respond to specific therapeutic agents. Recent advances in genomic technologies such as next-generation sequencing offer new opportunities for discovery and development of targeted therapies. They also pose numerous challenges in implementing molecularly guided precision medicine in clinical care. In this article, we review how molecular diagnostics have evolved over recent decades, discuss types and capabilities of clinically applicable genomic technologies, and highlight examples of companion diagnostics that have gained regulatory approval. Finally, we discuss technical and regulatory challenges associated with incorporating next-generation genomic technologies into clinical practice and consider potential ways to overcome these challenges to enable precision medicine.
& Rajiv Raja
[email protected] 1
MedImmune, LLC., 1 Medimmune Way, Gaithersburg, MD 20878, USA
Key Points Molecular biomarkers are highly valuable in the identification of patient subgroups that benefit from targeted therapies. Recent advances in genomic technologies enable highly multiplexed biomarker analysis for comprehensive disease profiling. As challenges related to complex, highly multiplexed technologies are addressed, these new genomic technologies will enable disease-based diagnostic panels that will revolutionize personalized medicine.
1 Introduction Historically, drug development and clinical practice have focused on the general characteristics of a disease rather than accounting for individual variability. This type of drug development strategy that targets an entire disease population assumes that the patient population is quite homogenous in pathology and etiology. Accordingly, this strategy has led to disappointing outcomes in numerous clinical trials in diseases with a high degree of patient heterogeneity. It is now widely accepted that a one-sizefits-all approach for drug development is extremely costly and inefficient (reaching 2558 million USD after adjusting for failures [1]) while also leading to suboptimal treatment responses that expose patients to unnecessary safety risks [2].
R. Raja et al.
During the past decade, approximately 90% of drug development projects failed to obtain an approval for clinical use. Between 2006 and 2015, a cumulative success rate of drug development from phase I to approval was 9.6% [3]. Despite the recent upward trend, clinical development success rates still remain lower than industry expectations (11.6% in 2011–2014 [4]). In 2010, an economic model of the drug development industry estimated that only 8% of new molecule entities (NME) will make it to the market and that this process will take an average of 13.5 years for each NME launched [5]. More recent studies on clinical trial success rates reported similar findings (Fig. 1a). Success rates of clinical trials for small-molecule drugs ranged from 6.2 to 10.0%, and biologics had slightly higher success rates, ranging between 11.5 and 18.0% [3, 4, 6]. Low success rates and long drug development timelines have prompted a need to refine/develop new strategies to deliver more effective, safer drugs at higher success rates by characterizing subsets of patients who benefit from certain treatments compared to those who do not. The concept of seeking more individualized treatment options for different patient sub-groups rather than treating the population as a whole can be broadly described as precision medicine. Precision medicine provides one potential solution to increase the success rates of clinical trials by enabling clinicians to make accurate and precise clinical decisions using comprehensive health information that includes genetic/genomic data, environmental exposures, and demographic and lifestyle information that has been collected over time. Recent advances in genomics, information technology, and bioinformatics have been major driving forces in the ability to implement precision medicine in a clinical setting. Specifically, high-throughput sequencing technology allows for the collection and evaluation of genetic, epigenetic, and gene expression data at the whole genome level. Furthermore, advances in information technology allowing for the storage and retrieval of large volumes of data combined with modern bioinformatics tools that enable the integration of large sets of demographic, clinical, and multi-omic data have led to a deeper understanding of molecular mechanisms underlying pathogenesis and provided information on key biomarkers that can be used to connect the right patients to the right drugs. A biomarker is defined as ‘‘a characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention’’ [7]. A genomic biomarker is a biomarker evaluated using DNA and/or RNA [8]. The presence or absence of these molecular biomarkers provides information on the expression, function, or regulation of a gene. DNA can be used to
Fig. 1 Success rates of drug development programs in the USA. a Comparison of success rates of small-molecule drug candidates versus biologics, and b success rates of clinical trials with and without biomarkers. aBiotechnology Innovation Organization et al. [3], b Smietana et al. [4], cFalconi et al. [6]. NSCLC non-small cell lung cancer
detect single nucleotide variations (SNVs), copy number variations (CNVs), insertions or deletions, alterations in repeat sequences, methylation, or gene fusions. RNA can be used to measure the altered expression levels of genes in patients compared with healthy individuals. Genomic technologies such as quantitative reverse transcription polymerase chain reaction (QRT-PCR), next-generation sequencing (NGS), and digital PCR have the capability of analyzing genomic variants and mutations across a large number of genes quickly, accurately, and reproducibly.
Genomics in Drug Discovery and Development
Identifying and connecting genetic/genomic changes with relevant disease phenotypes offers several potential utilities in facilitating precision medicine. Genes or functional pathways linked to a certain disease phenotype can be used as diagnostic biomarkers or to identify risk factors. Such genes or pathways may also be utilized as new targets for drug development. Additionally, altered genes or pathways may determine the overall outcome or prognosis of the patient or be predictive of treatment response to different lines of therapy. Owing to recent evolution in genomic technologies and the value of these datasets in understanding disease phenotypes and discovering relevant biomarkers, nucleic acids (DNA or RNA) hold the greatest potential for successfully implementing precision medicine strategies into clinical trials. Precision medicine—driven by new biomarker discoveries—has demonstrated the ability to increase the success rates of clinical trials. Clinical trials in biomarker-selected patients had a phase III success rate of 76.5%, whereas non-biomarker-led trials only had a phase III success rate of 55% [3]. Similar findings were reported in an independent study that examined the success rates for non-smallcell lung cancer (NSCLC) drug development during the period between 1998 and 2012 [6]. The cumulative success rate for advanced or metastatic (stage IIIb–IV) NSCLC clinical trials was found to be lower than the industry estimation (11 vs. 16.5%) without the utilization of biomarkers. However, the inclusion of patient selection biomarkers substantially increased the clinical trial success rate for NSCLC from 11 to 62% (Fig. 1b). These findings clearly indicate that identifying a targeted, well-defined patient population can have a profound impact on success rates for drug development. Examples in this review illustrate the value of using genomic/genetic biomarkers to increase the probability of clinical success when applied to target identification, efficacy, and safety in a variety of therapeutic areas.
2 Genetics- and Genomics-Based Drug Target Identification and Efficacy Biomarkers 2.1 Cystic Fibrosis Genetic variation that is causally associated with disease development may identify tractable drug targets [9]. Association of genetic changes with human disease suggests that targeting the altered downstream gene/pathway could lead to clinical benefit for patients with these genetic variations. There are a few examples of US FDA-approved drugs that were developed from human genetic evidence. An early example is ivacaftor, developed by Vertex and approved by the FDA in 2012 to treat cystic fibrosis in
patients with specific genetic mutations in CFTR (cystic fibrosis transmembrane conductance regulator) [10, 11]. Genetic variants in CFTR have been shown to be associated with impaired function of the CFTR channel and are thus causally associated with the severity of cystic fibrosis [12]. Ivacaftor was developed to target CFTR harboring specific loss-of-function mutations within *5% of all patients by acting as a potentiator that increases the likelihood of an open CFTR channel. 2.2 Hypercholesterolemia or Clinical Atherosclerotic Cardiovascular Disease Perhaps the most well-known example of human genetics driving target identification is the proprotein convertase subtilisin/kexin type 9 (PCSK9) gene in subjects at risk of high low-density lipoprotein cholesterol (LDL-C). Both protective loss-of-function and risk-associated gain-offunction genetic variants have been characterized for PCSK9 [13]. Additionally, a healthy individual with extremely low plasma LDL-C (14 mg/dl) was characterized as a compound heterozygote with two inactivating mutations in PCSK9 [14]. Thus, the genetic characteristics of PCSK9 fulfill key aspects of human genetic target validation: (1) protective- and risk-associated genetic variants and (2) nearly full inhibition of the target in the absence of toxic effects. The PCSK9 inhibitor alirocumab was developed by Regeneron/Sanofi and approved by the FDA in July 2015, and evolocumab was developed by Amgen and approved by the FDA in August 2015. 2.3 Oncology Within the last 30 years in the field of oncology, signaling pathway dysregulation has been a focus for new drug target development. Cancer cells proliferate through survival mechanisms such as genetic mutations or aberrations within oncogenes or tumor suppressors, which modulate critical cell functions involved in cell growth and survival, apoptosis, and cell cycle. Epidermal growth factor, mitogen-activated protein kinase (MAPK), or hedgehog signaling have been identified as some of the major pathways regulating these fundamental processes, where protein-altering mutations can drive tumorigenesis. Therefore, therapies targeting these signaling pathways have shown promise in various tumor types [15–17]. 2.3.1 Epidermal Growth Factor Receptor (EGFR) Among growth factor pathways, epidermal growth factor receptor (EGFR) is a key entry point to cell proliferation. Mutations in this receptor drive uncontrolled growth and division. Tyrosine kinase mutations in EGFR are observed
R. Raja et al.
in approximately 15% of NSCLC adenocarcinoma patients within the USA and occur more frequently in smokers [18]. In Asian populations, the incidence of EGFR mutations is substantially higher (up to 62%) [19]. Panitumumab is a fully human monoclonal antibody, manufactured by Amgen and marketed as Vectibix, which was approved by the FDA for the first time in 2016 for the treatment of EGFR-expressing metastatic colorectal cancer with disease progression. Panitumumab was the first monoclonal antibody to demonstrate the use of KRAS mutation as a predictive biomarker. Cetuximab is chimeric monoclonal antibody targeting EGFR, which was developed by ImClone (Eli-Lilly). In 2012, the FDA granted approval for cetuximab for use in KRAS wild-type, EGFR-expressing metastatic colorectal cancer measured with the diagnostic test kit therascreenÒ KRAS RGQ PCR Kit. Among small molecules, gefitinib, osimertinib, erlotinib, and afatinib are EGFR inhibitors marketed by AstraZeneca, Roche, and Boehringer Ingelheim; gefitinib was first approved by the FDA in 2003 and erlotinib was first approved in 2005, while afatinib and osimertinib were approved by the FDA more recently in 2013 and in 2015, respectively. The cobasÒ EGFR Mutation Test v2 from Roche was developed for detection of EGFR mutations in NSCLC patients who would qualify for erlotinib or osimertinib treatment, while the therascreenÒ EGFR RGQ PCR Kit from Qiagen is used to guide treatment decisions for gefitinib or afatinib. Typically, EGFR-targeted treatments are initially effective in patients whose tumors have sensitizing mutations. However, after 8–16 months, resistance develops and disease progression occurs. Examples of recurrent EGFR mutations driving disease progression include exon 19 deletion and mutations L858R and T790M. EGFR T790M mutation is found in 60–70% of NSCLC patients who have progressed on treatment with EGFR-tyrosine kinase inhibitors (TKIs) [20, 21]. To effectively treat these patients who have disease progression, next-generation therapies targeting specific EGFR mutations have been developed, such as osimertinib, which is an EGFR inhibitor developed by AstraZeneca and approved by the FDA in 2015. This EGFR inhibitor targets NSCLC patients harboring the EGFR T790M mutation, leading to an almost 80% response rate [22, 23]. Complementary to EGFR mutations are those in a downstream mediator of this pathway, KRAS. Recurrent mutations in this gene (codons 12 or 13) cause dysregulation of this pathway, thereby bypassing inhibition strategies targeting EGFR. Therefore, colorectal patients with wild-type KRAS alleles at these recurrent loci are eligible for cetuximab or panitumumab treatment. Similar tests from Qiagen (therascreenÒ KRAS RGQ PCR Kit) and Roche (cobasÒ KRAS Mutation Test) have been developed to test this KRAS mutation status to determine patient eligibility.
2.3.2 BRAF The MAPK pathway is an important driver in melanoma, with the BRAF gene as one of the key components to this pathway. The BRAF V600E mutation constitutively activates the gene and thus drives approximately *60% of advanced melanoma cases [24]. Two BRAF inhibitors, vemurafenib and dabrafenib, were approved by the FDA for the treatment of late-stage melanoma in patients with the V600E mutation. Vemurafenib received FDA approval in 2011 and is marketed by Genentech, whereas dabrafenib was developed by GSK and was approved by the FDA in 2013. Though both therapies have demonstrated effectiveness in this indication, patient resistance often occurs within 6–7 months. To overcome this, the MEK inhibitors cobimetinib and trametinib are typically used in combination with one of these BRAF inhibitors [25–28]. Trametinib is marketed by GSK and was approved by the FDA in 2013; cobimetinib received an FDA approval in 2015 and is being marketed by Genentech. 2.3.3 PTCH1 The PTCH1 receptor in the hedgehog pathway is frequently mutated in basal cell carcinoma (BCC) at 73–75% [29, 30] as well as medulloblastoma at 45% [31]. Vismodegib, marketed by Genentech, is the first small molecule developed to target the hedgehog signaling pathway, which is leading to a 30–40% response rate in BCC and medulloblastoma. Vismodegib was approved by the FDA in 2011 for the treatment of BCC and medulloblastoma. 2.3.4 BCR-ABL One of the most successful drug development stories driven by a genetic-based companion diagnostic is the smallmolecule inhibitor imatinib/Gleevec, which was first approved by the FDA in 2001 for the treatment of chronic myelogenous leukemia (CML). Most CML cases are associated with a translocation between the Abelson murine leukemia (ABL) gene on chromosome 9 with the breakpoint cluster region (BCR) gene on chromosome 22, resulting in the fusion protein BCR-ABL, and acting as a constitutively active tyrosine kinase found only in cancer cells. BCR-ABL protein is measured by the co-localization of genomic probes specific to the BCR and ABL genes using fluorescent in situ hybridization (FISH) or by amplifying the region around the splice junction between BCR and ABL using PCR. The existence of BCR-ABL predicts response to imatinib/Gleevec, a selective TKI of this fusion protein that has demonstrated substantial and durable responses in CML. Dasatinib and nilotinib are two additional BCR-ABL inhibitors that were developed by
Genomics in Drug Discovery and Development
BMS and Novartis individually. Dasatinib was approved for first-line treatment of Philadelphia chromosome-positive (Ph?) CML in 2010 and for acute lymphoblastic leukemia (ALL) in 2013. Nilotinib received accelerated approval in 2007 for the treatment of imatinib-resistant CML. Nilotinib is 10- to 30-fold more potent than imatinib in inhibiting BCR-ABL tyrosine kinase activity; therefore, it has demonstrated activity in CML that is resistant to treatment with imatinib [32]. 2.3.5 EML4-ALK A recurrent gene fusion between EML4 and ALK in 6.7% of NSCLCs [33] results in highly expressed fusion protein ALK and a constitutive oncogenic signal. In August 2011, the FDA approved crizotinib (marketed by Pfizer) to treat certain late-stage NSCLCs that express abnormal forms of the ALK gene. This approval required a companion molecular test for the EML4-ALK fusion gene (Ventana ALK D5F3 companion diagnostic assay). 2.3.6 17p Deletion in CLL Chromosome 17p deletion occurs in up to 10% of chronic lymphocytic leukemia (CLL) but carries the worst prognosis [34]. In April 2016, the FDA approved the first BCL2 inhibitor, venetoclax, for CLL patients with a chromosome 17p deletion. This genetic aberration is identified using an FDA-approved companion diagnostic Vysis CLL FISH probe kit.
3 Genetics- and Genomics-Based Biomarkers of Safety 3.1 Dermatology Genetics-based biomarkers have also indicated adverse responses to approved treatments. Two such approved biomarkers include DPYD and CYB5R1-4. Fluorouracil is a generic chemotherapeutic agent that was originally produced by Spectrum Pharmaceuticals and approved by the FDA in 1962 [35]. It is used for the treatment of numerous cancers as well as dermatologic conditions (actinic keratosis, extra-mammary Paget’s disease confined to the epidermis, Bowen’s disease, porokeratosis, and genital warts) [36]. Individuals with a deficiency in the enzyme dihydropyrimidine dehydrogenase (DPD) are more likely to experience severe side effects to fluorouracil treatment [37]. DPD deficiency can be assessed using a blood-based biomarker measuring levels of the DPYD gene, which can help to determine the likelihood of DPD deficiency and subsequently the potential risk for receiving fluorouracil.
3.2 Gastroenterology Another FDA-approved biomarker is cytochrome b5 (CYB5)-R1-4 for determining potential adverse reactions to the drug metoclopramide [38]. Metoclopramide (ANI Pharmaceuticals Inc., Baudette, MN, USA) is used for the treatment of gastroesophageal reflux in people who do not respond to other forms of treatment. Deficiency in the nicotinamide adenine dinucleotide (NADH)-cytochrome b5 reductase (encoded for by the CYB5R1, CYB5R2, CYB5R3, and CYB5R4 genes [39]) and glucose-6-phosphate dehydrogenase enzymes has been associated with the development of metoclopramide-induced methemoglobinemia, a disorder affecting the balance of iron in blood that subsequently affects oxygen release to tissues [40, 41]. Levels of CYB5 reductase can be measured in whole blood samples [40], while mutations in the CYB5R3 gene (which causes autosomal recessive congenital methemoglobinemia) can result in reduced activity or stability of the CYB5R3 enzyme [42]. 3.3 Immunosuppressants in Transplantation, Inflammation, and Oncology Thiopurine derivatives such as thioguanine, azathioprine, and 6-mercaptopurine (6-MP) are commonly used as immune suppressants in transplant patients as well as in the treatment of rheumatoid arthritis (RA), inflammatory bowel disease, and some cancers. A serious adverse effect observed in certain patients given thiopurine-based drugs is myelosuppression, the cause of which has been linked to reduced activity of the enzyme thiopurine methyltransferase (TPMT), which plays a role in the metabolism of thiopurines [43]. Between 3 and 10% of patients administered a thiopurine-based drug experience myelosuppression due to heterozygous mutations in the TPMT gene leading to reduced TPMT activity. Homozygous TPMT mutations are extremely rare; however, these individuals are at the highest risk of severe myelosuppression [44]. While three mutations in TPMT account for 90% of TPMT polymorphisms, other lower frequency mutations do exist. Several whole blood-based PCR tests for TPMT mutations have been developed, but not all known mutations are assayed in a single test. Alternatively, biochemical assays for enzyme activity in patient samples are available, and while they may detect enzyme deficiencies caused by mutations not assayed in the PCR test, they are more difficult to perform. A literature survey comparing genetic and biochemical TPMT testing identified a wide range of reported sensitivities and specificities attributed to the low frequency of homozygous nulls and the lack of assay standards [45]. The FDA recommends determining TPMT genotype or enzyme activity prior to treatments involving azathioprine or 6-MP
R. Raja et al.
therapy but does not specify the use of a genetic or biochemical test [38].
4 Evolution of Genomic Technologies Used for Biomarker Analysis Genetic testing has traditionally been performed on one or a few loci for the purpose of identifying genetic disorders. Most of the earlier genetic tests were designed to target genotypes and karyotypes that are directly related to disease phenotypes. Classical cytogenetic methods were utilized to identify chromosome abnormalities such as Down’s syndrome or the Philadelphia chromosome, which led to the subsequent discovery of the genetic causes of many diseases. However, over the years, we have learned that most diseases are driven by complex molecular traits involving many genes, and classical genetic testing methods are inadequate to interrogate more than one or a few specific loci. A fundamental change in the techniques used was necessary to better understand the complex diseaserelated phenotypes as well as genotype–phenotype and genotype–environment interactions. To overcome the above challenges, genomic technologies have improved dramatically over the last several decades. With the invention of PCR, methods for genomic analysis based on hybridization, fragment analysis and Sanger sequencing gave way to amplification-based analysis. With the advent of microarrays in the mid-1990s and subsequent improvements in microarray technologies in the late 1990s, it became possible to analyze the whole genome or transcriptome on a single microarray [46–48]. Genomic analysis went through yet another revolution in the mid-2000s with the invention of NGS, which enabled high-throughput, massively parallel deep sequencing at a much lower cost than standard dye-terminator methods of sequencing. They have provided high-throughput, deep and relatively low cost solutions for whole genome sequencing (WGS), whole exome sequencing (WES), gene expression analysis (RNA-seq), and methylation analysis (Methyl-seq) at the whole genome level. Table 1 provides a list of wellestablished as well as emerging technologies for biomarker analysis with each of their relative strengths and weaknesses. Further, we provide a list of well-established (Table 2) and emerging (Table 3) genomic technologies, and detail the types of clinical tests with regulatory approval that have been implemented using those technologies. In spite of these new technologies, non-sequencingbased technologies such as QRT-PCR and FISH are still predominantly used to screen for specific pathological variations to make treatment decisions. Even though NGS is widely considered to be the technology of choice for
diagnostics in the future, major analytical and interpretative challenges have to be overcome before it can be used reliably for routine clinical decision making. Such challenges include the validation of large numbers of genomic aberrations within a single patient; reimbursement mechanisms for large multi-analyte high-complexity assays; worldwide availability of assay platforms in standardized format; information technology infrastructure to manage and access terabytes of data; and the ability to quickly analyze and report results that are clinically actionable. Interpreting whole genome information from a patient is not a trivial endeavor, and addressing infrastructure, informatics, regulatory, and reimbursement challenges will be required to utilize whole genome data for diagnostic purposes. When these challenges have been overcome, WGS will transform clinical care. With the emergence of multi-analyte high-complexity assays, the tool box and opportunities for identifying biomarkers that influence rational drug development and clinical decision making have greatly improved. However, the access to high-quality, well-annotated clinical samples for biomarker studies remains both crucial and a challenge. The next section discusses this challenge and the complexities of integrating biomarker evaluations into clinical trials.
5 Biomarkers Enabling Drug Development and Clinical Decision Making Biomarkers play an essential role in early- and late-stage drug development. For early trials, biomarker data provide initial evidence of on-target biological activity of the investigational drug as evidenced by pathway modulation and/or target engagement; inform go/no-go decisions; and aid in the identification of dose responses to guide drug dose determination. Correlative studies in early trials are also key to identifying and/or confirming preliminary biomarker hypotheses. For late-stage development, pre-defined candidate predictive biomarkers in combination with investigational use only (IUO) specific diagnostic assays are necessary for prospective validation of the effectiveness of the targeted therapy in subpopulations of patients most likely to derive increased benefit from these drugs and for maximizing the therapeutic effect of targeted drugs. 5.1 Relevance and Challenges of Biomarker Evaluation in Early Clinical Trials Before large randomized trials to investigate the efficacy of a new drug are conducted, it is necessary to confirm safety, to identify an active drug dose (during dose escalation), and to verify that the investigated agent is indeed
Genomics in Drug Discovery and Development Table 1 Overview of current nucleic acid technologies Technology
Multiplexing capacity
Sample throughput
Instrument run time
Dynamic range
Advantages
Disadvantages
FISH
Limited
Low
1.5–2 days
Narrow (2–2.5 logs)
Multicolor detection possible; higher sensitivity than IHC
Specialized equipment required; highly trained and skilled person required for interpretation of results; subjectivity in data interpretation
CISH
Limited
Low
1.5–2 days
Limited (1–2 logs)
Ability to test analytes and tissue morphology simultaneously; no requirement for specialized fluorescence microscopes; Lower cost than FISH
No multicolor detection; requires a chromogenic procedure
Real-time PCR
Low
High
2h
Broad (7–10 logs) [49]
High sensitivity and specificity; rapid and robust; very broad dynamic range of detection; no post-PCR processing; automated analysis; good scalability
Analog signal; quantification is based on a standard curve (heavy reliance on the accuracy of standard curve); higher coefficient of variation at lower template copy number; prior sequence knowledge of targets required to design primers/probes
Digital PCR
Low
High
3–5 h
Good (7 logs with 1M chambers) [50]
High sensitivity and specificity; no need to rely on references or standards; high tolerance to inhibitors; capable of analyzing complex mixtures; rare variant detection; absolute quantification of nucleic acids
Lower analytical sensitivity compared with qPCR due to the restricted reaction volume; a different dilution strategy is required for high and low abundance analytes; longer turnaround time than real-time PCR; technical variations in handling small volumes of liquid
Microarray
High
Low
3–4 days
Narrow (2–3 logs) [51]
Global genomic profiling at low cost; easier data interpretation compared to RNA-seq; well-established methods
Narrow dynamic range; crossplatform variability; analog signal; variation in hybridization efficiency and signal normalization; expression profiling is limited to the genes included on the array; small changes in mRNA levels (\2-fold) may not be accurately measured, especially for genes expressed at low levels
Sanger sequencing
Limited
Limited
0.5–3 h
N/A
High accuracy; long read lengths; good calls in repeats and homopolymer regions
High per-base cost; low throughput
Solid-phase microarraybased
Low
Low
2.5–6.5 h
N/A
Ease of use; no instrument maintenance (no fluidics)
High cost; open amplification platform (risk for contamination)
R. Raja et al. Table 1 continued Technology
Multiplexing capacity
Sample throughput
Instrument run time
Dynamic range
Advantages
Disadvantages
Luminex beadbased liquid hybridization capture
Medium
High
2h
Good (3.5–4.5 logs) [52]
High reproducibility; low multiplex cost
Better suited to batch testing of samples rather than the single sample testing; low to medium resolution (may require further testing by sequence-based typing to obtain high resolution results); high instrument maintenance
NanoString digital colorcoded barcode technology
Medium
Limited
2.5 h
Good (5–6 logs) [53]
Digital signal; direct detection of molecules; no amplification; high resolution; low limit of detection; fusion gene detection; short hands-on time; easy workflow
High cost; specialized instruments; high maintenance needs; data analysis; dynamic range lower than qPCR or digital PCR; prior sequence knowledge required
NGS
High
High
1–14 days ? data analysis
Unlimited
High throughput; wide (unlimited) dynamic range; high reproducibility; digital signal
Higher error rate compared to Sanger sequencing; Short reads; High instrument cost; complexity of data analysis; despite low per-base cost, overall cost per run is high
Thirdgeneration sequencing
High
Low
0.5–4 h (PacBio); 50 h (Oxford NanoPore)
Unlimited
SMS; real-time sequencing; long read length
Technology not wellestablished yet; high instrument cost; low read output; high error rate
CISH chromogenic in situ hybridization, FISH fluorescence in situ hybridization, IHC immunohistochemistry, mRNA messenger RNA, N/A not applicable, NGS next-generation sequencing, PCR polymerase chain reaction, qPCR quantitative PCR, SMS single molecule sequencing
modulating its intended target in relevant samples from treated patients. Evaluation of the status of pharmacodynamic biomarkers between pre-treatment and on-treatment samples offers an opportunity to confirm the mechanism of action (MOA) of the investigated therapeutic agent and can also help (in combination with safety, pharmacokinetic, and/or clinical data) to select an appropriate drug dose for evaluating efficacy in subsequent and larger efficacy clinical studies. In addition, negative pharmacodynamic biomarker data (e.g. lack of evidence of target modulation) can also be extremely useful as it may inform early no-go drug development decisions.
5.2 Evaluation of Biomarkers in Blood-Derived Samples The samples used for most pharmacodynamic studies are blood, plasma, serum, and/or peripheral blood mononuclear cells (PBMCs). Longitudinal sample collections (multiple time-points) of these blood-derived samples are relatively easy to obtain and are useful to evaluate drug effects in the periphery. With the emergence of robust and sensitive omics technologies, the possibilities for
investigating pharmacodynamic biomarker modulation and the MOA of new drugs have been greatly expanded. For example, pre- versus on-treatment changes in whole blood gene expression can be easily monitored through blood collection in PAXgene tubes followed by isolation of RNA and gene expression analysis using microarrays, QRTPCR, NanoString, or RNA-seq technology. Target/pathway modulation and/or the effects of the drug in immune cells can be investigated in PBMCs via flow cytometry. Changes of circulating protein levels (including growth factors and chemokines) can be investigated from plasma and/or serum samples using specific enzyme-linked immunosorbent assays (ELISAs), multiplex assays, or mass spectrometry proteomics. Changes in metabolites can be investigated with specific metabolite assays or with more highthroughput metabolomics methods. When the therapeutically relevant cells are present in the periphery, target engagement (modulation of the target in the intended therapeutic site of action) can also be evaluated using these samples. Finally, baseline samples can also be used to evaluate or investigate predictive biomarkers. However, it is important to keep in mind that the biological parameter(s) measured in peripheral samples do not necessarily reflect the biology of the disease-affected organ.
Genomics in Drug Discovery and Development Table 2 Clinical applications of traditional nucleic acid technologies Technology
Platform
Manufacturer
FDA approval/clearance
Clinical assay
FISH
DuetTM System
BioVew Ltd
K050840
UroVysionTM Bladder Cancer Recurrence Kit (K011031)
K130775
VysisÒ ALK Break Apart FISH Probe Kit (P110012)
CISH
Real-time PCR
VysisÒ AutoVysionTM System
Vysis, Inc.
DEN040010
VysisÒ PathVysionÒ HER-2 DNA Probe Kit (P980024)
No instrument specified
Abott Molecular, Inc.
P150041
VysisÒ CLL FISH Probe Kit (P150041)
No instrument specified
ARUP Laboratories, Inc.
H140005
PDGFRB FISH Assay (H140005)
Autostainer Link 48
Agilent Technologies Dako Denmark A/S
P100024
HER2 CISH PharmDx Kit (P100024)
BenchMarkÒ XT automated slide staining instrument
Ventana Medical Systems, Inc.
P100027
INFORMÒ HER2 DUAL ISH DNA Probe Cocktail (P100027)
No instrument specified
ThermoFisher Scientific
P050040
SPOT-LightÒ HER2 CISH Kit (P050040)
Abbott m2000TM system
Abbott Molecular, Inc
K092705
Applied Biosystems 7500 FAST DX
ThermoFisher Scientific
K082562
THxIDTM BRAF Kit (P120014)
Applied Biosystems 7900HT
ThermoFisher Scientific
DEN080007
AlloMapÒ Molecular Expression Testing (K073482)
BD MAXTM Instrument
Becton Dickinson & Co
K111860
BD MAXTM GBS Assay (K111860)
K120138
BD MAXTM MRSA Assay (K120138)
Cepheid Inc.
K060540
XpertTM HemosILÒ FII & FV (K082118)
Cepheid SmartCycler Dx system
Cepheid Inc.
K062948
Cepheid Smart GBSTM Assay (K062948)
cobasÒ 4800 system
Roche Molecular Diagnostics
P110020
cobasÒ 4800 BRAF V600 Mutation Test (P110020)
P120019/P150047
cobasÒ EGFR Mutation Test v2 (P120019 & P150047)
P140023
cobasÒ KRAS Mutation Test (P140023)
Ò
Cepheid GeneXpert Dx System Ò
Microarray
IMDx Flu A/B and RSV Assay (K131584) IMDx HSV-1/2 Assay (K140198)
LightCyclerÒ Real-Time PCR System v 1.2
Roche Molecular Diagnostics
K033734
eQ-PCR LC Warfarin Genotyping kit (K073071)
Rotor-Gene Q MDx
QIAGEN
K113319
therascreenÒ RGQ PCR Kit —KRAS (P110030) & EGFR (P120022)
Spartan RX Analyzer
Spartan Bioscience, Inc.
K123891
Spartan RX CYP2C19 TEST SYSTEM (K123891)
QuantStudioTM Dx real-time PCR instrument
ThermoFisher Scientific
K123955
–
Agilent DNA microarray scanner
Agilent technologies
K101454
MammaPrintÒ (K101454)
Affymetrix GeneChipÒ Instrumentation System
Affymetrix, Inc.
DEN040012 (K080995)
Roche AmpliChip CYP450 microarray (K043576) PathworkÒ Tissue of Origin Test Kit—FFPE (K120489) Affymetrix CytoScan Dx Assay (K130313)
BeadXpressÒ System and VeraScan software v.2.0.17
Illumina, Inc.
K093128
Illumina VeraCode Genotyping Test for Factor V and Factor II (K093129)
Illumina iScan instrument
Illumina, Inc.
DEN140044
23andMe PGS Carrier Screening Test for Bloom Syndrome (DEN140044)
INFINITIÒ Analyzer
AutoGenomics, Inc.
K060564
INFINITI CYP2C19 Assay (BiolfilmChip Microarray) (K101683)
R. Raja et al. Table 2 continued Technology
Platform
Manufacturer
FDA approval/clearance
Clinical assay
Sanger sequencing
ABI 3100 Genetic Analyzer
ThermoFisher Scientific
BK030005
ViroSeq HIV-1 Genotyping System (BK030005)
ABI 3730xl
ThermoFisher Scientific
P140020
BRACAnalysis CDxTM (P140020)
OpenGeneÒ DNA Sequencing System
Siemens Healthcare Diagnostics Inc.
BK120013
TRUGENEÒ HIV-1 Genotyping Kit (BK120013)
CISH chromogenic in situ hybridization, FISH fluorescence in situ hybridization, PCR polymerase chain reaction, PGS personal genome service
5.3 Evaluation of Biomarkers in Tissue Samples Tissue samples from the relevant disease organ (e.g. tumor tissue in oncology) can also be used to evaluate pharmacodynamic and predictive biomarkers. In general, these samples are harder to obtain but are key to generate relevant in situ information regarding target engagement/modulation, provide insight into the MOA of an investigational drug, and to evaluate potential predictive biomarkers. Additionally, in oncology, biopsy samples taken at progression are important for identifying mechanisms of resistance to approved drugs [54, 55]. Tumor samples for predictive biomarker analysis can be either archival or fresh biopsies. Fresh biopsies are more likely to represent the biology of the tumor that is being treated as opposed to archival that may provide outdated information on an earlier, less advanced tumor stage at the time of diagnosis or surgery. Nonetheless, archival samples can be informative in many situations and generally contain larger pieces of tissue. Obtaining pre-treatment fresh biopsies can be medically and technically challenging or not feasible at all in certain patients and indications. Although there are several methods for collecting biopsies (fine needle aspiration, core biopsies, etc.), all biopsy procedures are highly invasive and are able to collect only small pieces of tissue. Thus, given the limited tissue available, only a restricted number of biomarker evaluations can be performed in these samples. For example, a formalin-fixed paraffin-embedded core biopsy may contain enough tissue for immunohistochemistry (IHC), QRT-PCR analysis, or targeted sequencing, but it can be challenging to perform more than one of these assays or to extract sufficient RNA and/or DNA for whole transcriptome RNA-seq or WES. Fresh frozen core biopsies may help with the issue of obtaining enough material for NGS, but extracting material for these assays may exhaust the available tissue and thus it may conflict with the need to evaluate additional biomarkers. In addition, fresh frozen sample collection is more laborious
and less common than formalin fixation in clinical practice, so implementation of thorough collections of fresh frozen biopsy samples in clinical trials may require additional logistical considerations and costs. Another factor influencing the utility of biopsy samples is the tumor content in the tissue. For example, if the biopsy contains predominantly adjacent normal tissue, then the gene expression and sequencing data obtained from this sample may not appropriately capture tumor information and can potentially be misleading. Additionally, tumor tissue is inherently heterogeneous, such that a small piece of tissue may not truly represent the overall cancer biology of a patient. This concept was recently highlighted by results indicating that mutational status can vary in different sections of the same tumor (intra-tumor heterogeneity) [56]. In addition, biopsies from different lesions from the same patients may have different genetic and molecular alterations (inter-tumor heterogeneity). Given that the molecular information gathered from a primary tumor or a metastatic lesion from the same patient may not be consistent, it is important to understand that the differences in biomarker levels between pre-treatment and ontreatment samples could simply be a consequence of intrinsic heterogeneous baseline levels of expression in these two particular tumor samples and may not be driven by drug treatment effects. However, if a consistent direction of fold expression change is observed in multiple paired biopsies, then a pharmacodynamic effect can be more reliably inferred from the data. For this reason, when pharmacodynamic effects in paired biopsies are investigated, collecting a relatively large number of such samples is advisable, as is selecting the most relevant time to collect a second biopsy to obtain meaningful results. Despite all these caveats, archival, baseline, and ontreatment biopsies are invaluable tools to evaluate pharmacodynamic and efficacy biomarkers and to select patients for specific treatments. Some illustrative examples of the clinical utility of archival biopsy samples for
Genomics in Drug Discovery and Development Table 3 Recent advances and emerging trends in nucleic acid technology Technology
Platform
Manufacturer
FDA approval/clearance
Assay description
RNA ISH
Bayer System 340 Analyzer
Bayer Diagnostics
BP000028
VersantÒ HIV-1 RNA 3.0 Assay (bDNA)
Panomics ViewRNATM RNA FISH Assay
Affymetrix, Inc.
N/A
mRNA and miRNA ISH
RNAscopeÒ Fluorescent Multiplex assay
Advanced Cell Diagnostics
N/A
RNA ISH assay for detection of target RNA within intact cells
BioMark HD
Fluidigm Corporation
N/A
Microfluidics-based
QuantStudio 3D digital PCR system
ThermoFisher Scientific
N/A
Microfluidics-based
QX100 ddPCR System
Bio-Rad Laboratories
N/A
Droplet-based
Digital PCR
RainDrop Digital PCR
RainDance
N/A
Droplet-based
BEAMing Digital PCR
Sysmex Inostics
N/A
BEAMing (beads, emulsion, amplification, magnetics)
VerigeneÒ System
Nanosphere, Inc.
K070597 K070804
VerigeneÒ F5 Nucleic Acid Test (K070597) VerigeneÒ Warfarin Metabolism Nucleic Acid Test (K070804)
eSensorÒ XT-8 Instrument system
GenMark Diagnostics
K073720
eSensorÒ Warfarin Sensitivity Test (k073720)
LUMINEX CORP
K073506
xTAG Cystic Fibrosis 60 Kit v2 (K083845);
LuminexÒ MAGPIXÒ instruments with xPONENTÒ software
LUMINEX CORP
K121894
xTAGÒ GPP (K121894)
ARIESÒ System
LUMINEX CORP
K151917
ARIESÒ HSV 1 & 2 Assay (K151906)
BioPlex 2200 System
Bio-Rad Laboratories
BK140112
BioPlexÒ 2200 HIV Ag-Ab assay (BP140111)
Hybridizationbased
NanoString nCounterÒ Dx Analysis System
Nanostring Technologies, Inc.
K130010
Prosigna Breast Cancer Prognostic Gene Signature Assay (K130010)
Secondgeneration sequencing
HiSeq 2000/2500, HiSeq 3000/4000, HiSeq-X
Illumina, Inc.
N/A
High-throughput, four-channel SBS technology
NextSeq 500/550
Illumina, Inc.
N/A
High- and mid-throughput benchtop sequencer; Two-channel SBS technology
MiSeqDx Instrument/MiSeqDx Universal Kit 1.0
Illumina, Inc.
DEN130011 (K123989)
Illumina MiSeqDxTM Cystic Fibrosis Clinical Sequencing Assay (K132750)
MiniSeq
Illumina, Inc.
N/A
Illumina’s lowest output sequencer
Ion PGMTM Dx DuetTM System
ThermoFisher Scientific
Listed
IVD NGS platform
Ion S5/Ion S5 XL
ThermoFisher Scientific
N/A
Semiconductor sequencing technology; Longer read length than MiSeq; Short turnaround time
NovaSeq 5000/6000
Illumina, Inc.
N/A
Scalable throughput; Data output per run ranges from 167 Gb and up to 6 Tb
PacBio RS-II/Sequel
Pacific Biosciences
N/A
SMRT sequencing technology
MinION/GridION
Oxford Nanopore Technologies
N/A
Miniaturized nanopore sequencing
Genia NanoTag Sequencing
Roche
N/A
Single molecule DNA sequencing by a nanopore with phosphate tagged nucleotides
Solid-phase microarraybased
Liquid bead microarray
Third-generation sequencing
Ò
Luminex 100/200 instrument
TM
eSensorÒ CF Genotyping Test (K090901) xTAG RVP FAST (K103776)
R. Raja et al. Table 3 continued Technology
Platform
Manufacturer
FDA approval/clearance
Assay description
Emerging sequencing technology
FISSEQ
Wyss Institute/ ReadCoor, Inc.
N/A
3D in situ gene sequencing (spatial gene sequencing technology)
GenalysisÒ Semiconductor Sequencing
DNA electronics
N/A
Miniaturized on-chip sequencing and analysis
FISSEQ fluorescent in situ sequencing, GPP gastrointestinal pathogen panel, ISH in situ hybridization, IVD in vitro diagnostic, miRNA microRNA, mRNA messenger RNA, N/A not applicable, NGS next-generation sequencing, PCR polymerase chain reaction, RVP respiratory viral panel, SBS sequencing by synthesis, SMRT single molecule, real-time
diagnosis in oncology are the HER2 test for HER2? breast cancer [57], EGFR mutation test for NSCLC [58], BRAF V600E for melanoma [25, 59], and PD-L1 test for 2nd line NSCLC [60, 61]. In addition, datasets from large phase 3 trials with high-quality biomarker and clinical data can be harnessed for research purposes through correlative analysis to evaluate exploratory hypotheses that may lead to the identification/validation of new targets or biomarker candidates that in turn can inform future trial design. Although the distinction between early and late biomarkers can be conceptually useful, is also important to keep in mind that biomarker data, in reality, is part of the continuous flow of scientific information that moves from early research to late-stage development and back again. Biomarker discovery and validation require a commitment to collect mandatory, well-annotated, high-quality samples. However, as mandatory tissue sample collections can slow down recruitment rates and increase the logistical complexity and costs of a trial, implementation of thorough mandatory tissue sample collection in large trials is still relatively uncommon.
rarely detectable in many other indications. Hence, the current applications of CTC analysis are mainly limited to research use. Another approach is the isolation of tumor DNA from plasma samples. A clear advantage of this methodology is that it is a highly specific way to track somatic tumorderived mutations. Although there are challenges with the sensitivity of these assays—only a small proportion of circulating cfDNA molecules for a given gene are mutant or tumor-derived—this technique has been shown to be reliable in detecting known mutations and is less invasive than tumor collection. Multiple studies have reported correlations between circulating tumor DNA (ctDNA)—a distinct and significant fraction of cfDNA in cancer patients—and tumor burden in patients undergoing targeted therapies with BRAF inhibitors [66, 67], MEK inhibitors [68], MAPK inhibitors [69], and immunotherapies [70]. The great potential of this new technology is illustrated by the recent FDA approval of a cfDNA test to detect EGFR mutations in plasma specimens to identify NSCLC patients eligible for treatment with erlotinib [71, 72].
6 Liquid Biopsies 7 Genomic Markers and Companion Diagnostics With the emergence of more sensitive and robust molecular methodologies, it is now possible to investigate specific tumor markers in circulation. Recent studies have reported the use of circulating tumor cells (CTCs), cell-free DNA (cfDNA), cell-free RNA (cfRNA), vesicles and exosomes in blood for repeated and non-invasive monitoring of disease [62–64]. For example, isolation and characterization of CTCs can be assessed from blood samples and used as a surrogate marker for treatment efficacy in cancer. Such methodologies have unique limitations such as sensitivity, specificity, and prevalence in all patients. CTCs are commonly found in indications such as prostate cancer [65] but
According to the FDA, a companion diagnostic (CDx) is a medical device, often an in vitro device, that provides information essential for the safe and effective use of a corresponding drug or biological product. The test helps a healthcare professional determine whether a particular therapeutic product’s benefits to patients will outweigh any potential serious side effects or risks. A CDx can (1) identify patients who are most likely to benefit from a particular therapeutic product, (2) identify patients likely to be at increased risk for serious side effects as a result of treatment with a particular therapeutic product, or (3)
Genomics in Drug Discovery and Development
monitor response to treatment with a particular therapeutic product for the purpose of adjusting treatment to achieve improved safety or effectiveness [73]. There are a number of unique opportunities with regards to the use of genomics technologies in personalized medicine and CDx. Currently, over 50% of CDx or human genetic testing devices approved by the FDA as of September 2016 rely on nucleic acid target amplification strategies (Fig. 2a; PCR 26%, modified PCR 19%, microarray 14%) [74, 75], and the majority of the FDAapproved CDx and genetic tests make diagnostic measurements based on fluorescent signals (data not shown). Even though the number of nucleic acid-based in vitro diagnostics (IVDs) obtaining FDA clearance doubled between 2006 and 2010 (Fig. 2b), PCR and hybridizationbased technologies have dominated the 510(k) clearance for interrogating nucleic acid targets for the last decade [76]. However, technologies and platforms are rapidly evolving in this space, creating opportunities to potentially utilize these technologies in the clinic. Evolution of platforms and chemistries have enabled the analysis of an array of biomarkers with a single platform. Ideally, these approaches will maximize the use of patient samples and leverage the power of multiplexed and multi-analyte biomarker development strategies. On top of the ability to evaluate multiple biomarkers with a single platform, the power of NGS offers a unique opportunity, driving sensitivity of biomarker detection at a genome level to levels not previously achievable. Although the adoption and use of NGS in a clinical setting has not been widely realized yet, the technology is quickly adapting to meet the needs for clinical utilization. The first NGS-based IVD, the MiSeqDx instrument (Illumina, San Diego, CA, USA), was cleared by the FDA in 2013 [77]. In addition, three NGS assays from Illumina obtained FDA 510(k) clearances in the same year (Fig. 2b). The automated nature of some of these evolving technologies renders them highly applicable in the clinical testing setting by minimizing technical variability, maximizing repeatability and efficiency, and reducing time to results. Opportunities for implementing genomics into drug development as CDx have also expanded due to the Precision Medicine Initiative, which was announced in January 2015 [78]. As a result of the initiative, the FDA has released a number of draft guidance documents related to the implementation of genomic data in drug development, including a draft co-development guidance specifically addressing CDx development principles [79–81]. The guidance and regulatory flexibility for the incorporation of genomic-based assays in drug development emphasize the value of developing these types of assays as CDx. Despite the promise, opportunities, and role of molecular diagnostics in advancing personalized medicine, there
Fig. 2 Molecular technologies utilized in FDA-cleared or approved nucleic acid in vitro diagnosticsa. a Nucleic acid-based technologies utilized in all human genetic testsb and companion diagnosticsc and cleared or approved by FDA until September 2016, b Changes in the distribution of nucleic acid technologies cleared by the FDA over the last two decadesd. aMicrobial testing devices were not included in the analysis; bUS FDA Nucleic Acid Based Tests—Human Genetic Tests [74], cUS FDA List of Cleared or Approved Companion Diagnostic Devices [75], dUS FDA 510(k) Premarket Notification Database (Jan 1996–Sep 2016) [76]
remain a number of challenges to translating biomarker discovery into companion or complementary diagnostic assays. Technical, logistic, and regulatory considerations are among these challenges. From a technical perspective, multiple technologies currently exist to address the assessment of molecular characteristics such as SNPs or messenger RNA (mRNA) expression. While advantageous in biomarker discovery studies, the complexity of methodologies and benefits/drawbacks of each technology must be strongly considered to optimize the likelihood of successfully migrating from a research-based biomarker
R. Raja et al.
assessment to an analytically validated and concordant method. Furthermore, proper implementation of technologies involve a truly collaborative effort between the pharmaceutical companies evaluating the biomarker in association with their experimental therapy and the company whose technology enables that analysis. This situation becomes more complex with the scenario of multi-analyte biomarkers, which may span more than one technology or marker, and add additional complexity to the analytical performance characterization and validation. The translation of a biomarker into a CDx assay can also provide logistical challenges. One particular aspect that is of the utmost importance is the availability of the sample from subjects enrolled in clinical trials. While most representative of the disease state, tissue samples from affected organs are often difficult to obtain and can also increase the risk to the patient from invasive procedures. While translating biomarkers to a more easily accessible tissue such as blood or circulating factors is ideal, that translation is not always feasible. Incorporating the appropriate sampling procedures in clinical trial design is therefore critical to ensuring the migration of a biomarker from research status to a companion or complementary diagnostic. Finally, the migration of a biomarker to a companion or complementary diagnostic is subject to regulatory filings that can be challenging to integrate with a therapeutic filing. Recently released draft CDx device guidance (July 15, 2016) [79] seeks to provide a therapeutic or diagnostic developer with a framework in which a CDx could be approved in conjunction with a therapeutic. This framework for diagnostic test development must be executed in concert with the process for which a therapeutic would obtain approval from the agency. Balancing the regulatory requirements and associated timelines for demonstrating analytical performance metrics in concert with aggressive clinical timelines makes coordination of development efforts and flexibility by regulatory agencies critical to the success and implementation of personalized healthcare strategies. While regulation of devices in the USA has been more stringent than in other geographies, the expanding use of biomarker-enriched trials and personalized therapy options is leading other regulatory authorities to reevaluate their existing CDx regulations. Careful attention must be paid to evolving regulatory pathways for diagnostics in these regions, as this will increase the complexity of worldwide applications of biomarker-based drug development approaches.
8 Conclusion With the invention and widespread adoption of NGS, we have entered a new era for molecular diagnostics. However, diagnostic testing today still relies heavily on
conventional technologies such as PCR and hybridizationbased technologies. As challenges related to highly multiplexed assays and complex analysis pipeline and reimbursement strategies for multiple testing are addressed, NGS-based diagnostics will gain regulatory approval and replace conventional technologies in the near future. This will enable the development of diagnostic tests that integrate all the relevant biomarkers for the diagnosis and treatment of a disease into a disease panel. Such a diseasebased diagnostics panel would ensure optimal use of patient samples and guide physicians to deliver effective therapy or sequence of therapies to patients and revolutionize personalized medicine. Compliance with Ethical Standards Funding No external funding was used in the preparation of this manuscript. Conflict of interest Rajiv Raja, Young S. Lee, Katie Streicher, James Conway, Song Wu, Sriram Sridhar, Mike Kuziora, Hao Liu, Brandon W. Higgs, Philip Z. Brohawn, Carlos Bais, Bahija Jallal, and Koustubh Ranade are employees of MedImmune/AstraZeneca and own AstraZeneca stock/stock options.
References 1. DiMasi JA, Grabowski HG, Hansen RW. Innovation in the pharmaceutical industry: new estimates of R&D costs. J Health Econ. 2016;47:20–33. 2. Schork NJ. Personalized medicine: Time for one-person trials. Nature. 2015;520(7549):609–11. 3. Biotechnology Innovation Organization, Biomedtracker, Amplion I. Clinical development success rates, 2006–2015. 2016. Available from: https://www.bio.org/sites/default/files/Clinical% 20Development%20Success%20Rates%202006-2015%20-%20 BIO,%20Biomedtracker,%20Amplion%202016.pdf. 4. Smietana K, Siatkowski M, Møller M. Trends in clinical success rates. Nat Rev Drug Discov. 2016;15(6):379–80. 5. Paul SM, Mytelka DS, Dunwiddie CT, Persinger CC, Munos BH, Lindborg SR, et al. How to improve R&D productivity: the pharmaceutical industry’s grand challenge. Nat Rev Drug Discov. 2010;9(3):203–14. 6. Falconi A, Lopes G, Parker JL. Biomarkers and receptor targeted therapies reduce clinical trial risk in non-small-cell lung cancer. J Thorac Oncol. 2014;9(2):163–9. 7. Strimbu K, Tavel JA. What are biomarkers? Curr Opin HIV AIDS. 2010;5(6):463–6. 8. U.S. Food and Drug Administration (FDA). Guidance for industry: E15 definitions for genomic biomarkers, pharmacogenomics, pharmacogenetics, genomic data and sample coding categories. 2008. Available from: http://www.fda.gov/downloads/ drugs/guidancecomplianceregulatoryinformation/guidances/ucm 073162.pdf. 9. Kamb A, Harper S, Stefansson K. Human genetics as a foundation for innovative drug development. Nat Biotechnol. 2013;31(11):975–8. 10. Sheridan MB, Hefferon TW, Wang N, Merlo C, Milla C, Borowitz D, et al. CFTR transcription defects in pancreatic sufficient cystic fibrosis patients with only one mutation in the coding region of CFTR. J Med Genet. 2011;48(4):235–41.
Genomics in Drug Discovery and Development 11. Van Goor F, Hadida S, Grootenhuis PD, Burton B, Cao D, Neuberger T, et al. Rescue of CF airway epithelial cell function in vitro by a CFTR potentiator, VX-770. Proc Natl Acad Sci USA. 2009;106(44):18825–30. 12. Kerem B, Rommens JM, Buchanan JA, Markiewicz D, Cox TK, Chakravarti A, et al. Identification of the cystic fibrosis gene: genetic analysis. Science. 1989;245(4922):1073–80. 13. Cohen JC, Boerwinkle E, Mosley TH Jr, Hobbs HH. Sequence variations in PCSK9, low LDL, and protection against coronary heart disease. N Engl J Med. 2006;354(12):1264–72. 14. Zhao Z, Tuakli-Wosornu Y, Lagace TA, Kinch L, Grishin NV, Horton JD, et al. Molecular characterization of loss-of-function mutations in PCSK9 and identification of a compound heterozygote. Am J Hum Genet. 2006;79(3):514–23. 15. Douillard JY, Siena S, Cassidy J, Tabernero J, Burkes R, Barugel M, et al. Randomized, phase III trial of panitumumab with infusional fluorouracil, leucovorin, and oxaliplatin (FOLFOX4) versus FOLFOX4 alone as first-line treatment in patients with previously untreated metastatic colorectal cancer: the PRIME study. J Clin Oncol. 2010;28(31):4697–705. 16. Messersmith WA, Ahnen DJ. Targeting EGFR in colorectal cancer. N Engl J Med. 2008;359(17):1834–6. 17. Robarge KD, Brunton SA, Castanedo GM, Cui Y, Dina MS, Goldsmith R, et al. GDC-0449-a potent inhibitor of the hedgehog pathway. Bioorg Med Chem Lett. 2009;19(19):5576–81. 18. Kawaguchi T, Ando M, Kubo A, Takada M, Atagi S, Okishio K, et al. Long exposure of environmental tobacco smoke associated with activating EGFR mutations in never-smokers with non-small cell lung cancer. Clin Cancer Res. 2011;17(1):39–45. 19. Shi Y, Au JS, Thongprasert S, Srinivasan S, Tsai CM, Khoa MT, et al. A prospective, molecular epidemiology study of EGFR mutations in Asian patients with advanced non-small-cell lung cancer of adenocarcinoma histology (PIONEER). J Thorac Oncol. 2014;9(2):154–62. 20. Oxnard GR, Arcila ME, Sima CS, Riely GJ, Chmielecki J, Kris MG, et al. Acquired resistance to EGFR tyrosine kinase inhibitors in EGFR-mutant lung cancer: distinct natural history of patients with tumors harboring the T790M mutation. Clin Cancer Res. 2011;17(6):1616–22. 21. Yu HA, Arcila ME, Rekhtman N, Sima CS, Zakowski MF, Pao W, et al. Analysis of tumor specimens at the time of acquired resistance to EGFR-TKI therapy in 155 patients with EGFRmutant lung cancers. Clin Cancer Res. 2013;19(8):2240–7. 22. Janne PA, Yang JC, Kim DW, Planchard D, Ohe Y, Ramalingam SS, et al. AZD9291 in EGFR inhibitor-resistant non-small-cell lung cancer. N Engl J Med. 2015;372(18):1689–99. 23. Xu M, Xie Y, Ni S, Liu H. The latest therapeutic strategies after resistance to first generation epidermal growth factor receptor tyrosine kinase inhibitors (EGFR TKIs) in patients with nonsmall cell lung cancer (NSCLC). Ann Transl Med. 2015;3(7):96. 24. Long GV, Menzies AM, Nagrial AM, Haydu LE, Hamilton AL, Mann GJ, et al. Prognostic and clinicopathologic associations of oncogenic BRAF in metastatic melanoma. J Clin Oncol. 2011;29(10):1239–46. 25. Long GV, Stroyakovskiy D, Gogas H, Levchenko E, de Braud F, Larkin J, et al. Combined BRAF and MEK inhibition versus BRAF inhibition alone in melanoma. N Engl J Med. 2014;371(20):1877–88. 26. Robert C, Karaszewska B, Schachter J, Rutkowski P, Mackiewicz A, Stroiakovski D, et al. Improved overall survival in melanoma with combined dabrafenib and trametinib. N Engl J Med. 2015;372(1):30–9. 27. Ascierto PA, McArthur GA, Dreno B, Atkinson V, Liszkay G, Di Giacomo AM, et al. Cobimetinib combined with vemurafenib in advanced BRAF(V600)-mutant melanoma (coBRIM): updated
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41. 42. 43.
efficacy results from a randomised, double-blind, phase 3 trial. Lancet Oncol. 2016;17(9):1248–60. Larkin J, Ascierto PA, Dreno B, Atkinson V, Liszkay G, Maio M, et al. Combined vemurafenib and cobimetinib in BRAF-mutated melanoma. N Engl J Med. 2014;371(20):1867–76. Bonilla X, Parmentier L, King B, Bezrukov F, Kaya G, Zoete V, et al. Genomic analysis identifies new drivers and progression pathways in skin basal cell carcinoma. Nat Genet. 2016;48(4):398–406. Jayaraman SS, Rayhan DJ, Hazany S, Kolodney MS. Mutational landscape of basal cell carcinomas by whole-exome sequencing. J Invest Dermatol. 2014;134(1):213–20. Kool M, Jones DT, Jager N, Northcott PA, Pugh TJ, Hovestadt V, et al. Genome sequencing of SHH medulloblastoma predicts genotype-related response to smoothened inhibition. Cancer Cell. 2014;25(3):393–405. Kantarjian H, Giles F, Wunderle L, Bhalla K, O’Brien S, Wassmann B, Tanaka C, et al. Nilotinib in imatinib-resistant CML and Philadelphia chromosome-positive ALL. N Engl J Med. 2006;354(24):2542–51. Soda M, Choi YL, Enomoto M, Takada S, Yamashita Y, Ishikawa S, et al. Identification of the transforming EML4-ALK fusion gene in non-small-cell lung cancer. Nature. 2007;448(7153):561–6. Stilgenbauer S, Zenz T, Winkler D, Buhler A, Schlenk RF, Groner S, et al. Subcutaneous alemtuzumab in fludarabine-refractory chronic lymphocytic leukemia: clinical results and prognostic marker analyses from the CLL2H study of the German Chronic Lymphocytic Leukemia Study Group. J Clin Oncol. 2009;27(24):3994–4001. Drugs@FDA: FDA Approved Drug Products. Fluorouracil. [cited 2016 October 7]; Available from: https://www.fda.gov/ downloads/drugs/guidancecomplianceregulatoryinformation/ guidances/ucm217130.pdf. Moore AY. Clinical applications for topical 5-fluorouracil in the treatment of dermatological disorders. J Dermatolog Treat. 2009;20(6):328–35. Van Kuilenburg AB, Vreken P, Beex LV, Meinsma R, Van Lenthe H, De Abreu RA, et al. Heterozygosity for a point mutation in an invariant splice donor site of dihydropyrimidine dehydrogenase and severe 5-fluorouracil related toxicity. Eur J Cancer. 1997;33(13):2258–64. U.S. Food and Drug Administration. Table of Pharmacogenomic Biomarkers in Drug Labeling. Available from: http://www.fda. gov/Drugs/ScienceResearch/ResearchAreas/Pharmacogenetics/ ucm083378.htm. U.S. Food and Drug Administration. U.S. Food and Drug Administration (FDA) label information for metoclopramide and CYB5R1, CYB5R2, CYB5R3, CYB5R4, G6PD. Available from: http://www.accessdata.fda.gov/scripts/cder/daf/index.cfm?event= overview.process&varApplNo=017854. Karadsheh NS, Shaker Q, Ratroat B. Metoclopramide-induced methemoglobinemia in a patient with co-existing deficiency of glucose-6-phosphate dehydrogenase and NADH-cytochrome b5 reductase: failure of methylene blue treatment. Haematologica. 2001;86(6):659–60. Mary AM, Bhupalam L. Metoclopramide-induced methemoglobinemia in an adult. J Ky Med Assoc. 2000;98(6):245–7. NIH US National Library of Medicine GHR. CYB5R3 gene. Available from: https://ghr.nlm.nih.gov/gene/CYB5R3. Lennard L, Van Loon JA, Weinshilboum RM. Pharmacogenetics of acute azathioprine toxicity: relationship to thiopurine methyltransferase genetic polymorphism. Clin Pharmacol Ther. 1989;46(2):149–54.
R. Raja et al. 44. Nguyen CM, Mendes MA, Ma JD. Thiopurine methyltransferase (TPMT) genotyping to predict myelosuppression risk. PLoS Curr. 2011;3:RRN1236. 45. Roy LM, Zur RM, Uleryk E, Carew C, Ito S, Ungar WJ. Thiopurine S-methyltransferase testing for averting drug toxicity in patients receiving thiopurines: a systematic review. Pharmacogenomics. 2016;17(6):633–56. 46. DeRisi J, Penland L, Brown PO, Bittner ML, Meltzer PS, Ray M, et al. Use of a cDNA microarray to analyse gene expression patterns in human cancer. Nat Genet. 1996;14(4):457–60. 47. Lipshutz RJ, Morris D, Chee M, Hubbell E, Kozal MJ, Shah N, et al. Using oligonucleotide probe arrays to access genetic diversity. Biotechniques. 1995;19(3):442–7. 48. Pease AC, Solas D, Sullivan EJ, Cronin MT, Holmes CP, Fodor SP. Light-generated oligonucleotide arrays for rapid DNA sequence analysis. Proc Natl Acad Sci USA. 1994;91(11):5022–6. 49. Abruzzo LV, Lee KY, Fuller A, Silverman A, Keating MJ, Medeiros LJ, et al. Validation of oligonucleotide microarray data using microfluidic low-density arrays: a new statistical method to normalize real-time RT-PCR data. Biotechniques. 2005;38(5):785–92. 50. Heyries KA, Tropini C, VanInsberghe M, Doolin C, Petriv OI, Singhal A, et al. Megapixel digital PCR. Nat Methods. 2011;8(8):649–51. 51. Dallas PB, Gottardo NG, Firth MJ, Beesley AH, Hoffmann K, Terry PA, et al. Gene expression levels assessed by oligonucleotide microarray analysis and quantitative real-time RT-PCR – how well do they correlate? BMC Genomics. 2005;27(6):59. 52. Luminex Corp. xMAP Technology Products and Services: Luminex Instrumentation Overview. 2016. Available from: http:// info.luminexcorp.com/download-the-life-science-product-cata log. 53. Prokopec SD, Watson JD, Waggott DM, Smith AB, Wu AH, Okey AB, et al. Systematic evaluation of medium-throughput mRNA abundance platforms. RNA. 2013;19(1):51–62. 54. Van Allen EM, Wagle N, Sucker A, Treacy DJ, Johannessen CM, Goetz EM, et al. The genetic landscape of clinical resistance to RAF inhibition in metastatic melanoma. Cancer Discov. 2014;4(1):94–109. 55. Wang X, Goldstein D, Crowe PJ, Yang JL. Next-generation EGFR/HER tyrosine kinase inhibitors for the treatment of patients with non-small-cell lung cancer harboring EGFR mutations: a review of the evidence. Onco Targets Ther. 2016;9:5461–73. 56. Gerlinger M, Horswell S, Larkin J, Rowan AJ, Salm MP, Varela I, et al. Genomic architecture and evolution of clear cell renal cell carcinomas defined by multiregion sequencing. Nat Genet. 2014;46(3):225–33. 57. Owens MA, Horten BC, Da Silva MM. HER2 amplification ratios by fluorescence in situ hybridization and correlation with immunohistochemistry in a cohort of 6556 breast cancer tissues. Clin Breast Cancer. 2004;5(1):63–9. 58. Kawaguchi T, Ando M, Asami K, Okano Y, Fukuda M, Nakagawa H, et al. Randomized phase III trial of erlotinib versus docetaxel as second- or third-line therapy in patients with advanced non-small-cell lung cancer: Docetaxel and Erlotinib Lung Cancer Trial (DELTA). J Clin Oncol. 2014;32(18):1902–8. 59. Ribas A, Gonzalez R, Pavlick A, Hamid O, Gajewski TF, Daud A, et al. Combination of vemurafenib and cobimetinib in patients with advanced BRAF(V600)-mutated melanoma: a phase 1b study. Lancet Oncol. 2014;15(9):954–65. 60. Gandini S, Massi D, Mandala M. PD-L1 expression in cancer patients receiving anti PD-1/PD-L1 antibodies: a systematic review and meta-analysis. Crit Rev Oncol Hematol. 2016;100:88–98.
61. Zou W, Wolchok JD, Chen L. PD-L1 (B7-H1) and PD-1 pathway blockade for cancer therapy: Mechanisms, response biomarkers, and combinations. Sci Transl Med. 2016;8(328):328rv4. 62. Siravegna G, Marsoni S, Siena S, Bardelli A. Integrating liquid biopsies into the management of cancer. Nat Rev Clin Oncol. 2017 (Epub 2 Mar 2017). 63. Lee JH, Long GV, Boyd S, Lo S, Menzies AM, Tembe V, et al. Circulating tumour DNA predicts response to anti-PD1 antibodies in metastatic melanoma. Ann Oncol. 2017;28(5):1130–6. 64. Minciacchi VR, Zijlstra A, Rubin MA, Di Vizio D. Extracellular vesicles for liquid biopsy in prostate cancer: where are we and where are we headed? Prostate Cancer Prostatic Dis. 2017, 1–8. 65. Leon-Mateos L, Vieito M, Anido U, Lopez Lopez R, Muinelo Romay L. Clinical application of circulating tumour cells in prostate cancer: from bench to bedside and back. Int J Mol Sci. 2016;17(9):1580. 66. Momtaz P, Gaskell AA, Merghoub T, Viale A, Chapman PB. Correlation of tumor-derived circulating cell free DNA (cfDNA) measured by digital PCR (DigPCR) with tumor burden measured radiographically in patients (pts) with BRAFV600E mutated melanoma (mel) treated with RAF inhibitor (RAFi) and/or ipilimumab (Ipi). J Clin Oncol. 2014;32(15 Suppl):9085. 67. Panka DJ, Buchbinder E, Giobbie-Hurder A, Schalck AP, Montaser-Kouhsari L, Sepehr A, et al. Clinical utility of a blood-based BRAF V600E mutation assay in melanoma. Mol Cancer Ther. 2014;13(12):3210. 68. Frenel JS, Carreira S, Goodall J, Roda D, Perez-Lopez R, Tunariu N, et al. Serial next-generation sequencing of circulating cell-free DNA evaluating tumor clone response to molecularly targeted drug administration. Clin Cancer Res. 2015;21(20):4586. 69. Gray ES, Rizos H, Reid AL, Boyd SC, Pereira MR, Lo J, et al. Circulating tumor DNA to monitor treatment response and detect acquired resistance in patients with metastatic melanoma. Oncotarget. 2015;6(39):42008–18. 70. Lipson EJ, Velculescu VE, Pritchard TS, Sausen M, Pardoll DM, Topalian SL, Diaz LA Jr. Circulating tumor DNA analysis as a real-time method for monitoring tumor burden in melanoma patients undergoing treatment with immune checkpoint blockade. J Immunother Cancer. 2014;2(1):42. doi:10.1186/s40425-0140042-0. 71. Kwapisz D. The first liquid biopsy test approved. Is it a new era of mutation testing for non-small cell lung cancer? Ann Transl Med. 2017;5(3):46. doi:10.21037/atm.2017.01.32. 72. U.S. Food and Drug Administration. cobas EGFR Mutation Test v2. 2016. Available from: https://www.fda.gov/Drugs/ InformationOnDrugs/ApprovedDrugs/ucm504540.htm. 73. U.S. Food and Drug Administration. In Vitro Companion Diagnostic Devices: Guidance for Industry and Food and Drug Administration Staff. 2014. Available from: http://www.fda.gov/ downloads/MedicalDevices/DeviceRegulationandGuidance/Guid anceDocuments/UCM262327.pdf. 74. U.S. Food and Drug Administration. Nucleic Acid Based Tests Human Genetic Tests. Available from: http://www.fda.gov/ MedicalDevices/ProductsandMedicalProcedures/InVitroDiagnos tics/ucm330711.htm. Accessed 6 Oct 2016. 75. U.S. Food and Drug Administration. List of Cleared or Approved Companion Diagnostic Devices. Available from: http://www.fda. gov/MedicalDevices/ProductsandMedicalProcedures/InVitroDia gnostics/ucm301431.htm. 76. U.S. Food and Drug Administration. 510(k) Premarket Notification Database. Jan 1996–Sep 2016. Available from: http://www. accessdata.fda.gov/scripts/cdrh/cfdocs/cfPMN/pmn.cfm. 77. FDA News Release. FDA allows marketing of four ‘‘next generation’’ gene sequencing devices. Two devices aid in screening and diagnosis of cystic fibrosis. Silver Spring: U.S. Food and Drug Administration; 2013.
Genomics in Drug Discovery and Development 78. The Precision Medicine Initiative. 2015. Available from: https:// www.whitehouse.gov/precision-medicine. 79. U.S. Food and Drug Administration. Principles for Codevelopment of an In Vitro Companion Diagnostic Device with a Therapeutic Product. 2016. Available from: http://www.fda.gov/ downloads/MedicalDevices/DeviceRegulationandGuidance/Guid anceDocuments/UCM510824.pdf?source=govdelivery&utm_ medium=email&utm_source=govdelivery. 80. U.S. Food and Drug Administration. Use of Public Human Genetic Variant Databases to Support Clinical Validity for Next
Generation Sequencing (NGS)-Based In Vitro Diagnostics. 2016. Available from: http://www.fda.gov/downloads/MedicalDevices/ DeviceRegulationandGuidance/GuidanceDocuments/UCM5098 37.pdf. 81. U.S. Food and Drug Administration. Use of Standards in FDA Regulatory Oversight of Next Generation Sequencing (NGS)Based In Vitro Diagnostics (IVDs) Used for Diagnosing Germline Diseases. 2016. Available from: http://www.fda.gov/ downloads/MedicalDevices/DeviceRegulationandGuidance/Guid anceDocuments/UCM509838.pdf.