Abstract
Biomarkers are an important tool in modern drug development. The FDA has posited that increased use of biomarkers in clinical trials can accelerate pharmaceutical industry productivity, ushering new drugs to market. Accordingly, the FDA has created two pathways for evaluation of biomarker utility. Biomarkers incorporated into clinical trials, the traditional pathway, are effectively private to a therapeutic sponsor and to the scope of the trial. By contrast, in Biomarker Qualification (“BQ”), the second pathway, a biomarker is certified as a publicly available tool. The FDA has hoped that academic, non-profit, and industry stakeholders would work together in consortia to qualify biomarkers, cumulatively generating a common resource of broad utility. In practice, utilization of Biomarker Qualification has been paltry. Incentives for BQ that align with the interests of industry resource holders are necessary to fuel increased utilization of biomarkers in clinical trials and create the communal biomarker toolkit envisioned by the FDA. A blanket extension of exclusivity for any drug successfully paired with a companion biomarker would diminish public access to medicine by encouraging spurious biomarkers and correspondingly narrowed clinical trials. As a measured alternative, an exclusive right to include a qualified companion biomarker on an FDA drug label would balance public access externalities. This exclusivity would waylay label approval, and thus marketability, of later drugs relying on the qualified biomarker for clinical safety or efficacy. Accordingly, sponsors would find no incentive to portage an ineffective or unnecessary biomarker through clinical trials, as there would be no benefit to securing exclusive rights in a tool others saw no value in using.
I. BIOMARKERS
In 2004, the Director of the Food and Drug Administration (“FDA”)'s Center for Drug Evaluation and Research (“CDER”) authored a seminal “Critical Paths” report identifying a productivity crisis in pharmaceutical industry research and development. 1 The report starkly observed that the FDA had overseen a decade of decline in the annual approval of new chemical entities. 2 Over the same period, US investment in biomedical research had increased two-fold, indicating that inadequate funding was not to blame. 3 Clinical trial practices, however, were identified as a substantial concern. 4 The report leveled criticisms that “new science is not being used to guide … technology development” and that “developers … use the tools and concepts of the last century.” 5 It challenged the FDA to modernize by embracing under-utilized tools, such as biomarkers, and by encouraging collaboration between public and private contributors. 6
The FDA's commitment to biomarker utilization solidified rapidly. A follow-up Critical Paths report in 2006 stated that developing biomarkers and streamlining clinical trials were the highest priorities in medical product development. 7 The Director observed that “[t]he absence of practical processes to establish the clinical significance of a given biomarker has severely limited the use of existing biomarkers in drug development and in the clinic.” 8 By 2007, the FDA had established a new pilot program for validating biomarkers. 9
The biomarker definition is broad and flexible. According to federal statute, a biomarker is “a characteristic (such as a physiologic, pathologic, or anatomic characteristic or measurement) that is objectively measured and evaluated as an indicator of normal biologic processes, pathologic processes, or biological responses to a therapeutic intervention.” 10 Few, if any, biological measures that reflect health status would be excluded by this definition. Reflecting common academic wisdom, one article in a prominent scientific journal simply defines a biomarker as “anything that is reliably indicative of a biological state.” 11 The FDA is thus theoretically unconstrained in its ability to undertake review of proposed biomarkers, carving a broad path for innovation.
In practice, biomarkers include “everything from pulse and blood pressure through basic chemistries to more complex laboratory tests of blood and other tissues.” 12 Molecular, histological, radiographic, and physiological biomarkers are among recognized types. 13 Many are “pharmacogenomic” biomarkers, meaning biomarkers “related to variations in DNA and RNA characteristics.” 14 For example, genes that mediate patient-specific drug metabolism can be important, e.g., in determining proper dosage. 15 Other pharmacogenomic biomarkers are specific to a drug's target. For instance, four lung cancer drugs inhibit anaplastic lymphoma kinase (ALK), which promotes cell proliferation, and are marketed with an ALK biomarker label. 16 Thus, many biomarkers embody a new era of biological sciences, rooted in laboratory techniques, such as genetic sequencing.
The FDA classifies biomarkers by the type of clinical information they provide. Seven categories are recognized, namely: (1) biomarkers of susceptibility or risk that an individual will develop a disease or respond poorly to a drug; (2) biomarkers that diagnose disease; (3) biomarkers used to monitor disease progression; (4) biomarkers that predict drug response; (5) biomarkers used to monitor drug response; (6) biomarkers of drug toxicity; and (7) biomarkers that may predict a clinical event. 17 Together, the FDA classifications describe biomarkers that predict, diagnose, or monitor diseases, drug responses and clinical events.
Biomarkers are also classified as exploratory, probable valid, or known valid. A known valid biomarker has well-established performance characteristics and accepted clinical significance. A probable valid biomarker is supported by scientific evidence, but has not yet been accepted as clinically significant. Exploratory biomarkers have met neither standard. 18
Use of pharmacogenomic biomarkers is not new. Genetic markers that inform treatment decisions trace back to at least the 1970's. 19 For example, a 1975 manuscript demonstrated that metastatic breast cancer was markedly more responsive to various endocrine therapies (e.g. adrenalectomy, castration, or estrogen therapy) if the tumor expressed a high concentration of estrogen receptor. 20 Researchers concluded that assessment of patient estrogen receptor expression was predictive of treatment outcomes and should therefore be used in patient stratification; by appropriate patient selection, “a large number of patients can … be spared unrewarding major endocrine ablative therapy.” 21
Although the medical community has utilized biomarkers for decades, the FDA's strategies for regulating biomarkers and promoting their use continue to develop.
II. PATHS TO BIOMARKER VALIDATION
Historically, biomarkers were validated by the widespread acceptance of relevant clinical, scientific, and regulatory communities. Evidence of clinical value accrued gradually through scattered scientific publications, debate, and clinical observation. 22 This time-consuming, decentralized path allows extensive community input and organic cost-sharing, but the multi-year timeframe can be prohibitive. Moreover, with input from many laboratories and methodologies, data collation and reproducibility may prove challenging. 23 Time-honored biomarkers that percolated through this process include, e.g., the centuries-old practice of measuring blood pressure as a metric of cardiovascular function. 24 Other examples are more recent, such as the use of increased blood alanine aminotransferase to detect liver damage. 25
Today, biomarker evaluation often occurs concomitantly with an Investigational New Drug Application, New Drug Application, or Biologics License Application. These paths are necessarily more concerted and streamlined than comparatively haphazard community acceptance. In these paths, a drug sponsor engages directly with the FDA to validate a particular biomarker in conjunction with a particular therapeutic product. 26 The sponsor must develop its own clinical trial data to support biomarker validation and bear all associated costs. This pathway to biomarker validation is narrow: the particular therapeutic and context of use define the scope of validation. Such limitation will apply even if all data suggests the biomarker has broader utility. 27
Following the call of the 2004 Clinical Paths Report, the FDA unveiled new “Drug Development Tools” (“DDTs”) including a first-of-its-kind Biomarker Qualification program. 28 Biomarker Qualification is defined as “a conclusion that within [a] stated context of use (“COU”), the [biomarker] can be relied on to have a specific interpretation and application in drug development and regulatory review.” The COU “describes the way the DDT is to be used and the purpose of the use.” 29
Qualified biomarkers become available to the public. Moreover, while biomarkers accepted via an IND, NDA, or BLA are validated based on trial-specific data, Biomarker Qualification provides a generalized approval for use across a class of diseases or therapies. 30 Thus, once qualified, any sponsor can use the biomarker in an IND, NDA, or BLA submission. The biomarker will not be subject to further review for clinical suitability, absent new evidence to the contrary. 31 The FDA expects that public availability of qualified biomarkers will expedite drug development by providing helpful tools to drug developers. 32
In its effort to understand the productivity crisis, FDA has theorized that the increasing demands of biomedical development are often beyond the capacity of single private entities, such as pharmaceutical companies. 33 The drug approval process “puts the entire burden of developing a biomarker on a single pharmaceutical company.” 34 By contrast, in Biomarker Qualification, “pharmaceutical companies, patient- or disease-specific foundations, health research organizations, or consortia made up of any or all of these groups, may collectively generate the data necessary for making drug-product independent decisions with respect to biomarkers used in drug development.” 35
The FDA expects that collaboration among academic, non-profit, and industry stakeholders will be further excited by the promise that qualified biomarkers become available to the public, conferring a tool to the medical community at large. It reasoned that such collaborations are workable “[b]ecause the development and qualification of new biomarkers will benefit many parties.” 36 In the FDA's view, consortia can “brin[g] together scientists from highly competitive companies and then maintain a productive environment” based on “modern project management” and the heightened efficacy of collaborative development. 37 The FDA has noted that “industry scientists are finding it very rewarding to share with their competitors … in pre-competitive areas.” 38
Biomarker Qualification is not designed to replace traditional paths to validation. Rather, “[q]ualification of a biomarker with FDA is particularly useful for biomarkers that will have a broad application.” 39 Moreover, collaborative Qualification is better suited to early-stage, pre-clinical biomarkers because competitive tensions increase with the proximity of drug discovery, clinical investment, and marketing. 40 Accordingly, “collaboration has been the exception rather than the norm.” 41 Individual companies are still expected to utilize traditional drug approval pathways, while Biomarker Qualification is generally viewed as a path for consortia or academic groups. 42
Neither the FDA's commitment to biomarker development, nor the Biomarker Qualification program in particular, has markedly improved pharmaceutical industry productivity. Rather, “the translation of basic science discoveries to drugs with obvious real-world benefits is becoming less, rather than more, frequent.” 43 Barriers may include increasing regulatory complexity, increasing clinical trial costs, insufficient engagement of non-profit stakeholders in drug development, and the difficulty of marshalling vast new reservoirs of drug- and disease-specific data into workable clinical solutions. 44 Biomarker Qualification whittles down these barriers by providing an opportunity for industry and non-industry actors to collaborate in early validation of key clinical tools, thereby reducing clinical trial burden. The program's failure to manifest practical benefits, despite strong theoretical foundations, begs review.
III. EVALUATION AND REGULATION OF BIOMARKERS
Traditional drug approval and Biomarker Qualification paths to marker validation are subject to separate regulatory schemes. Each is governed by a combination of general statutory requirements, procedures set forth in the Code of Federal Regulations (“CFR”), and more specific criteria found in agency guidance documents. Nevertheless, the agency retains a high degree of flexibility in biomarker evaluation.
A. Evaluation of Biomarkers used in Clinical Trials
No new drug can be introduced into interstate commerce unless the agency approves an application including evidence of drug safety and efficacy. 45 Statute does not set forth detailed procedures or evidentiary standards for approval, however.
Greater detail is found in the CFR. 46 For instance, 21 C.F.R. § 314.1259(b) lists 18 conditions under which an NDA may be refused. An NDA maybe refused, for example, if the FDA finds that trials “d[id] not include adequate tests … to show whether or not the drug is safe,” or if the FDA finds “a lack of substantial evidence … that the drug product will have the [claimed] effect.” 47
As may be inferred from the above examples, the CFR does little to circumscribe FDA discretion in NDA review. Indeed, the CFR recognizes that “while the statutory standards apply to all drugs, the many kinds of drugs … subject to the statutory standards and the wide range of uses for those drugs demand flexibility in applying the standards.” 48 The same section further explicates that “FDA is required to exercise its scientific judgment to determine the kind and quantity of data and information an applicant is required to provide.” 49
No statute or regulation on traditional drug approval pathways refers to biomarkers. 50 Rather, applicable standards, to the extent articulated, are found in agency guidance documents. 51
In 2005, the FDA published draft guidelines for drug-diagnostic co-development. 52 It stated that new diagnostics should be developed in parallel with clinical trials. 53 In particular, the draft showed that, prior to initiation of a phase 1 clinical trial, a sponsor should identify a marker of interest, validate the marker's relevance, and develop a diagnostic kit to standardize analysis. 54 Sponsors were to evaluate biomarker assay characteristics including accuracy, precision, specificity, sensitivity, conditions for use, sampling protocols, and expected range for measured biomarker values (where applicable). 55 A valid clinical test would “improve the benefit/risk of the drug” and meet specific statistically-defined performance standards. 56 This draft guidance appears to have been abandoned following negative response from industry. 57 Skeptics noted that parallel drug/diagnostic development was unrealistic because timelines for biomarker and drug development rarely aligned. Industry also expressed concern that the co-development model unnecessarily applied the same stringent standards to both drug approval and biomarker validation, compounding the difficulty reaching the market. 58 Abandoning the 2005 draft left both industry and agency without clear protocols, presumptively placing sponsors at the whim of the FDA. Later-finalized guidance documents suggest the net result was de facto adoption by the FDA of many provisions discussed in the draft. 59
Narrow guidance for use of pharmacogenomic markers was finalized in 2013. 60 The pharmacogenomic guidance states that if a test “will provide information … essential for the safe and effective use of a therapeutic product,” then “an FDA-approved or -cleared test will be required at the same time that the drug is approved.” 61
Perhaps the most informative guidance covers devices that perform diagnostic tests such as biomarker measurement. 62 Companion diagnostic devices are paired with therapeutic products to “provide information that is essential for the safe and effective use of a corresponding therapeutic.” 63 For instance, overexpression of the cell division protein HER2 in breast cancer is a biomarker for successful treatment with Herceptin, and the FDA has approved specific commercial assays that measure HER2 expression. 64 Like the stymied 2005 draft guidance, the FDA recommends “[the] device and its corresponding therapeutic product should be approved or cleared contemporaneously.” 65 As is the case with biomarkers themselves, the FDA reviews each device submission “within the context of, or in conjunction with, its corresponding therapeutic product.” 66 Sponsors are encouraged to engage the FDA early in development to ensure that trials will “produce sufficient data to establish the safety and effectiveness of both.” 67 The level of regulatory scrutiny required follows a “risk-based approach” that balances “the level of risk to patients” and the controls available “to provide a reasonable assurance of safety and effectiveness.” 68 After approval, the therapeutic label must include information on the diagnostic test and how the results correlate with treatment. 69 Over 30 companion diagnostic devices have been cleared or approved as “provid[ing] information that is essential for the safe and effective use of a corresponding therapeutic product,” while nearly 120 human genetic test submissions have been cleared or approved. 70
The 21st Century Cures Act, enacted December 13, 2016, will likely impact biomarker review in the context of drug approvals. The Act invites agency review of “real world evidence” in the approval of novel indications for previously approved drugs. 71 Real world evidence is defined as “data … from sources other than randomized clinical trials.” 72 Examples of real world evidence are expected to include registry data, retrospective database analysis, case reports, insurance records, electronic health records, public survey data, and records collected in the monitoring of patients. 73 The FDA administration could read the 21st Century Cures Act as an invitation to use real world evidence in new drug evaluation generally: Commissioner Gottlieb has stated that “there's nothing in our statute or regulations that prevent FDA from using a broad range of information sources as evidence.” 74 In view of the FDA's flexibility regarding evidentiary standards, incorporating real world evidence may remain a black box for some time. The 21st Century Cures Act set a two-year deadline for the FDA to establish an implementation framework. 75
B. Biomarker Qualification
Biomarker Qualification is the subject of substantial, direct FDA guidance. Guidance documents published in 2014 and 2016, respectively, set forth the Qualification process and evidentiary framework. 76
The process for Biomarker Qualification can be broken down into three stages: (1) an initiation stage; (2) a consultation and advice stage; and (3) a review stage. A biomarker submitter first files a Letter of Intent (LOI), briefly describing support for the possible biomarker. If a Qualification Review Team (“QRT”) sees potential, the submission advances to stage two. The submitter is invited to a consultation and must provide a more detailed briefing package that includes strengths, limitations, expected benefits, and unresolved issues. 77 FDA staff with appropriate expertise are assigned to the case. 78 The QRT works with the submitter to define the Context of Use (“COU”) and strengthen the proposal. 79 The COU includes the measurement to be taken, clinical utility, conditions for use, and standards for interpreting measured values. 80 The QRT may require further data, additional studies, or application of new methodologies. When the QRT and the submitter agree that the consultation and advice phase is complete, stage three final review begins. The QRT can seek expert advice, gather input from other FDA centers or officers, hold public forums, or make further requests of the submitter. 81 Qualification establishes that a biomarker can be used appropriately in a drug development program. 82
A separate 2016 framework outlines evidentiary requirements for Biomarker Qualification. 83 The framework acknowledges and confronts the lack of clear evidentiary standards for Qualification, emphasizing a need for harmonization with the demands of drug-specific development. 84 Supporting evidence must be coextensive with the proposed Context of Use. 85 However, unsurprisingly, “the amount of evidence needed … cannot currently be provided as an absolute relationship and is instead related to many factors, some of which are not easily quantified.” 86 Key evidentiary concerns identified by the document include (1) the scientific rationale underlying selection of a particular biomarker; (2) source and amount of existing data or knowledge; (3) relationship of the biomarker to clinical outcomes; (4) substantive and statistical strength of clinical trial design and resulting data; (5) comparison with existing standard protocols; and (6) technical performance of the biomarker assay. 87 Review will balance the therapeutic benefits of biomarker data against accompanying risks, such as the risks patients face from an erroneous biomarker test result. 88
The 21st Century Cures Act codifies Biomarker Qualification, but does not appear to conflict with existing FDA guidance on the subject. 89 Indeed, the Act appears to broaden the scope of Biomarker Qualification by abandoning the FDA's exclusion of “assessment of how an individual feels, functions, or survives.” 90 The apparent expansion of scope may have minimal impact, since biomarker development to date has emphasized molecular phenotypes. 91 Moreover, “feelings” or “functions” may lack sufficient analytical clarity to merit biomarker validation. The Act provides the FDA with a three-year grace period in which to issue new draft guidance. 92
C. Biomarker Information in Drug Labels
If a biomarker constitutes scientific information necessary for safe and effective use of a drug, it must be referenced in the drug label. 93 For instance, according to one FDA draft guidance document, when “the consequences of … genetic variations result in recommendations for restricted use, dosage adjustments, contraindications, or warnings, this information will be summarized in … the labeling as appropriate.” 94 Inclusion in labeling will generally depend on pivotal trial results. 95
Drug approvals supported exclusively by clinical trials in biomarker-positive subjects logically require labeling limited to biomarker-positive patients. 96 The underlying reasoning is that “there will be no effectiveness or safety information on the enrichment-marker-negative patients.” 97 Summarizing its practices, the FDA has stated that “[t]he specific patient population studied has sometimes, but certainly not always, been given as the indicated population in the Indications section of labeling in addition to its description in the Clinical Studies section.” 98
D. Patent Protection of Biomarker -Related Inventions
Commercially, incentives to seek FDA approval are typically bolstered by parallel patent protection. However, recent Supreme Court decisions have undercut available patent protections with respect to biomarkers, straining incentives for their development. In Ass'n for Molecular Pathology v. Myriad Genetics, Inc., the Supreme Court held that isolated human DNA, such as a human gene whose sequence predicts breast and ovarian cancers, is not patentable. 99 Cognizant that DNA isolation is “necessary to conduct genetic testing,” the Court wrote that a “[g]roundbreaking, innovative, or even brilliant discovery” is insufficient, by itself, to merit an exclusive right to isolate a newly found gene. 100 The Court further appeared to indicate, in dicta, that DNA isolation would only be patentable if it utilized processes that are not “well understood, widely used, and fairly uniform insofar as [used by scientists].” 101 In practice, this would make patenting pharmacogenomic assays difficult, first because DNA isolation and sequencing are already widely understood, and second because developers of new genetic biomarkers are often separate entities from developers of new isolation technologies.
A diagnostic method was specifically at bar in Mayo Collaborative Servs. v. Prometheus Labs., Inc. 102 The Supreme Court held that the relationship between (1) the concentration of drug metabolite circulating in a subject's blood, and (2) whether, based on that concentration, drug dosage should be increased or decreased, was a mere “natural law” or “phenomen[on] of nature.” 103 As a result, the process of applying such know-how, without more, was deemed unpatentable. 104 The holding underscores the Court's position that discovery of new and useful biological relationships are not themselves patentable, and cannot be rendered otherwise by “additional steps [that] consist of well understood, routine, conventional activity already engaged in by the scientific community.” 105
The impact of these cases is manifest in Ariosa Diagnostics, Inc. v. Sequenom, Inc., where the Federal Circuit, bound by Mayo, denied patentability to a new, non-invasive method of characterizing fetal DNA. 106 While previous methods involved sampling from fetal or placental tissue, with attendant risks to the fetus, the patent in suit claimed a method of amplifying fetal DNA from a pregnant female's plasma or serum. 107 However, the court held that “the existence of [cell-free fetal DNA] in maternal blood is a natural phenomenon,” and that the technical steps of detection were “well-understood, routine, and conventional.” 108 Together, Myriad, Mayo, and Ariosa represent a substantial blow to patent-based incentives for biomarker development.
IV. CURRENT UTILIZATION OF BIOMARKER VALIDATION PATHWAYS
The utilization of biomarkers in FDA approvals is rising. Between 2014 and 2016, nearly one of every four FDA approvals included a personalized medicine component. This trend is particularly striking in oncology, where over seventy percent of drugs in development include a subject-specific analytic component. 109
A. Therapeutics Approved with a Biomarker Component
Over two hundred drugs (214) are labeled with pharmacogenomic biomarker information. The drug labels include a total of 284 entries referencing 61 distinct biomarkers. The ten most-referenced biomarkers represent well over half of the entries (177). Prominent among these are CYP2D6 (60 drug labels) and CYP2C19 (22 drug labels). 110 The “CYP” designation represents a class of “cytochrome P450” genes that participate in the body's effort to breakdown and excrete foreign substances, including drugs. The rate at which this occurs can be very important to individual drug response. 111 For instance, the breast cancer drug tamoxifen is converted to a more potent form by CYP2D6, so individuals with a “CYP2D6 poor metabolizer phenotype” respond poorly to treatment. 112 CYP2D6 also converts codeine to morphine. In one instance, a mother receiving codeine for post-childbirth pain was, unbeknownst to her caregivers, an “ultrarapid CYP2D6 metabolizer.” As a result, she delivered a lethal dosage of morphine to her infant through breast milk. 113 Remarkably, twenty-five to thirty percent of prescription medications are metabolized by CYP2D6. 114 Over eighty known variants of CYP2D6, and nearly thirty of CYP2C19, are found in humans, each with the potential to impact drug potency at the individual level. 115
Many biomarkers are specific to a drug target or mechanism of disease. For example, mutation of the Epidermal Growth Factor Receptor gene (“EGFR”) drives disease pathogenesis in certain forms of non-small-cell lung cancer (“NSCLC”). 116 Historically, EGFR-targeted small molecule therapeutics drugs predate the discovery of EGFR mutations. Bifurcation of clinical results between treatment responders and non-responders led researchers to seek genetic explanations, revealing the link. Subsequently, EGFR was approved through IND submission as a biomarker in lung cancer therapy. 117
One difficulty in incorporating a biomarker into a traditional approval is the impact on study design. A traditional trial requires treatment and control groups to discern the effect of treatment. However, a trial with a biomarker component might require four experimental groups: treatment of biomarker-positive subjects, treatment of biomarker-negative subjects, control biomarker-positive subjects, and control biomarker-negative subjects. 118 The increase in experimental groups complicates study design and increases clinical trial costs. Many drug trials opt to minimize costs by excluding biomarker-negative subjects from trials altogether (leaving only two experimental groups).119,120 This is called enrichment. 121 While enrichment study designs reduce up front costs, they often exclude potential treatment beneficiaries prematurely. 122 For example, panitumumab clinical trials were limited to subjects with EGFR-expressing tumors, but drug efficacy was later shown to have no correlation with EGFR expression. 123 Meta-analysis of clinical trials supporting thirty-five approved drug/biomarker pairs showed that over two-thirds were restricted to biomarker-positive subject; only six were empowered to show treatment-specific differences between biomarker-positive and biomarker-negative subjects. 124
The confounding reality of biomarker discovery is that biomarkers are often not identified until some time after clinical trial initiation. 125 According to industry commentary, “common industry practice is … non-parallel drug and test development” where biomarkers are identified only after “clinical potential has been examined in early exploratory studies.” 126 Moreover, from a commercial perspective, “early development of a partner diagnostic test (validated pre-phase 3) is considered an expensive and high-risk endeavor in those cases when the approvability of the partner drug is uncertain.” 127
From a social perspective, traditional drug development's drawback is that a sponsor will ordinarily develop only enough data to support use of a biomarker in a particular context. Such data will not be helpful in extending the marker to other clinical settings or contexts. Other trials “would have little ability to build on that knowledge to expand the tool's use to additional settings.” 128 Thus, the sponsor is awarded a near-proprietary ability to rely on clinical trial data “retained with the specific candidate drug/biologic submission … and … not necessarily available for others to use.” 129 By contrast, industry interest in validation via Biomarker Qualification is likely quelled by the availability of qualified biomarkers for use in clinical trials by all comers.
B. Use of the Biomarker Qualification Pathway
Overall utilization of Biomarker Qualification remains dwarfed by traditional biomarker approval routes. To date, the FDA has qualified only seven biomarkers. Of these, four are for nonclinical laboratory use, and three are qualified for clinical use. An additional twenty are reportedly in varying stages of Qualification. 130
All seven qualified biomarkers were supported by collaborative arrangements, several of which were formal consortia. One collaboration united GlaxoSmithKline, Pfizer, AstraZeneca, and University College, Dublin. 131 The group achieved Qualification of circulating cardiac troponins as indicators of cardiac damage in 2012, pursuant to a 2008 request. Qualification was based entirely on data from dozens of previously published manuscripts that cumulatively evidenced the relationship. Many of the manuscripts showed circulating cardiac troponins were at least as effective as standard histopathology assays in identifying cardiac damage. The group specifically sought and received approval for use of troponins in nonclinical studies, particularly in dogs, rats, and monkeys. 132
One recently qualified biomarker predicts progressive decline in renal function of subjects with autosomal-dominant polycystic kidney disease (“ADPKD”). At the time of submission, tools for clinical development of ADPKD therapies were lacking. 133 The Polycystic Kidney Disease (PKD) Foundation collaborated with the FDA, Critical Path Institute (“C-Path”), academia, and industry in creating a PKD Outcomes Consortium (“PKDOC”). PKDOC identified Total Kidney Volume (TKV) as a possible biomarker by integrating existing data from patient registries and observational studies, and it was qualified in 2016. 134
The C-Path has been an important supporter of the FDA's biomarker efforts. C-Path is a nonprofit institute formed by collaboration between the University of Arizona and a Stanford research institute. It is neutrally funded by local governments and private foundations, without financial investment from the pharmaceutical industry. In 2005, with a five-year operating budget of ten million dollars, C-Path hired a team to work with FDA and industry scientists on projects including biomarker development. C-Path limited its efforts to projects with FDA support, two or more existing companies ready to collaborate, and a source of non-commercial external funding, with preference for projects deemed pre-competitive by virtue of distance from clinical use. 135
C-Path's first (and ongoing) project has brought together 15 global pharmaceutical companies and 250 scientists to share knowledge on the detection of drug toxicity. 136 The group has received formal FDA support for biomarkers of muscle injury (model organism use), kidney injury (model organism use), and liver injury (clinical use). 137
The glaring problem with the Biomarker Qualification program is revealed by the sheer lack of activity it has generated. Since the pilot launch in 2007, and the first Qualification in 2008, the entire program has approved only seven biomarkers (about one every two years). This does not appear to be the result of a high failure rate, as the program has only fifteen biomarkers in queue. 138 Even crediting the queue, the Biomarker Qualification program is dwarfed by approvals through traditional paths. 139
The Biomarker Qualification program is enervated by lackluster incentives. Qualification provides only sub-market approval, in that it is a step toward, but not sufficient for, market approval of a diagnostic. It merely establishes that a biomarker is suitable for use in a drug development program. 140 Thus, “once qualified, a biomarker can be used in multiple drug development programs without the need for CDER to reconsider or reconfirm its suitability.” 141
Theoretically, the purported step-wise benefit of Qualification (pre-approval for use in diverse clinical trials) promotes the pace and resourcefulness of medical research. However, the FDA is careful to delineate the conspicuously narrow advantages of Qualification. A qualified biomarker is considered “conceptually independent of the specific test that measures the biomarker.” 142 Thus, even after Qualification, successful incorporation of a biomarker into a clinical trial requires independent assay development. Moreover, Qualification “does not qualify the biomarker for use in clinical practice” and “does not imply approval or clearance of a diagnostic device or of a companion diagnostic for use in clinical practice.” 143 Accordingly, Qualification does not ease the burden of showing that the biomarker contributes to safety or efficacy in a particular treatment context.
Where precisely is the benefit to a drug application sponsor utilizing a qualified biomarker, if any? As noted above, even after Qualification, a sponsor must still (1) establish a diagnostic test for the biomarker; (2) undertake clinical trials including the biomarker; and (3) establish that use of the biomarker meets established criteria of safety and efficacy within the intended clinical use. 144 If a drug sponsor is required to develop clinical evidence that would satisfy both Qualification and drug approval standards regardless of whether Qualification was previously undertaken, the incentive for any party to engage in Biomarker Qualification is difficult to identify.
V. INCENTIVIZING BIOMARKER QUALIFICATION
The Biomarker Qualification program has failed to meet the challenge of the 2004 Critical Path Report. Rather than sparking a wave of biomarker development, it has qualified a meager handful of biomarkers at the rate of one every two years. 145 Nor is there evidence showing the FDA's biomarker efforts have resurrected pharmaceutical industry productivity. 146 The program, and incentives for participation, need to be revised.
A. Failed Incentives for Biomarker Qualification
Incentives to participate in the current Biomarker Qualification program have not generated results. In principle, Qualification “creates a collaborative setting where multiple interested parties may pool their resources and data to decrease cost, expedite drug development, and facilitate regulatory review.” 147 Possible benefits, however, are offset, in part, by loss of proprietary control. Ignoring this counter-incentive, the FDA has stated that “although the drug approval process (in an IND/NDA/BLA submission)…may be efficient for drug developers, it has limitations…[in that] confidential discussions between the FDA and the drug developer do not allow input from the larger scientific community, because information about the biomarker may not be publicly available.” 148 The FDA promotes Qualification as “facilitat[ing] widespread use of the biomarker.” 149
More importantly, any benefits that might be expected to accrue to clinical trials are too diffuse. The purported advantages of Biomarker Qualification are severely dampened by the absence of any clear illustration of how Qualification actually advantages drug approval. Indeed, the FDA has spent more effort elaborating the limitations of Biomarker Qualification than on explicating its administrative benefits.
Finally, traditional interests of neither industry nor non-industry actors are well-served by the current Biomarker Qualification program. Non-industry actors may find that academic publication alone is sufficient to meet institutional goals of disseminating knowledge. Even if motivated to contribute to clinical trials, a non-industry actor would have little reason to qualify a biomarker absent assurances that the biomarker will be picked up by clinical trials. That assurance would require an industry partnership. However, industry actors may prefer to monopolize know-how, maximize exclusivity, and leverage first-mover advantages in individually sponsored clinical trials. If a non-industry actor is best served by industry partnership, that partnership may be more profitably arranged outside of the Biomarker Qualification context. Thus, the alignment of interests that would justify Biomarker Qualification may be rare.
B. Improving Participation in Biomarker Qualification
An improved Biomarker Qualification program should provide clear incentives that appeal to both industry actors seeking to bring new therapeutics to market and non-industry actors seeking to contribute to clinical trial progress, whether acting alone or in consortia. Engaging pharmaceutical industry actors that have traditionally, and continue to be, the driving force behind U.S. pharmaceutical development is particularly critical. Optimally, Qualification will be incentivized whenever any industry or non-industry actor identifies a biomarker with general utility.
1. Clarification of Existing Incentives to Qualify Biomarkers
Biomarker Qualification program reform must first clarify the advantages of participation. The current scheme appears to be based on the supposition that traditional drug/biomarker approval pathways include steps of (1) determining whether a biomarker is suitable for use in a proposed clinical trial context; and (2) reviewing clinical trial data to determine whether its use is beneficial in practice. No such steps are clearly defined in the agency's regulatory scheme. However, the evidentiary framework for Biomarker Qualification suggests that Qualification standards also apply in drug-specific development programs, indicating a Qualification-like component. 150 Moreover, the aborted 2005 Drug-Diagnostic Co-Development Concept Paper recites distinct steps including “analytical validation” and review of “clinical performance.” 151 Thus, the FDA apparently does conceive integration of a biomarker into drug approval as including at least the steps of (1) establishing “suitability” and (2) establishing “clinical performance.” Consequently it appears that Biomarker Qualification satisfies the “suitability” determination, but not trial-specific “clinical performance” requirements. The FDA should clarify the substantive impact of Biomarker Qualification, namely satisfaction of the suitability requirement, in these or similar terms.
2. Creation of New Incentives to Qualify Biomarkers
Since clarification would not necessarily entail a substantive change in existing incentives, industry actors will likely require new, additional incentives with clear financial advantages to induce a meaningful change in participation. Moreover, the Biomarker Qualification pathway is unique because it creates a publicly available tool, and private industry actors will require compensation to expand access to otherwise closely held know-how. Industry has historically responded positively to incentives such as extended regulatory exclusivity, expedited regulatory approval, fee reductions, research grant opportunities, and tax benefits.
Regulatory exclusivity has proven to be a particularly powerful incentive for industry action. However, outright extension of drug exclusivity would be inappropriate in this context. Such outright extension would encourage “frivolous” Qualification of unnecessary, non-informative, or minimally informative biomarkers. For instance, under an outright extension regime, a savvy drug sponsor could qualify a biomarker that captures a substantial portion of a population of interest, e.g., ninety percent of subjects with a targeted form of cancer, but lacks clinical significance. The sponsor could then undertake enriched clinical trials that exclude biomarker-negative subjects, negating any opportunity for clinical data to debunk the proposed biomarker. Under a simple exclusivity regime, approval of the drug with this biomarker indication would then gain an exclusivity benefit, despite the absence of a corresponding improvement in patient care. While the sponsor would suffer a marginal narrowing of its drug label (i.e., to biomarker-positive patients), losses could be mitigated by off-label use. Thus, an outright exclusivity award for pairing of a drug with a biomarker creates adverse incentives for development of marginally useful biomarkers and narrow clinical trials.
A form of exclusivity that rewards Qualification of useful biomarkers, but deters frivolous use, can be achieved by decoupling biomarker exclusivity from drug exclusivity. In particular, a focus on drug labeling, rather than drug marketability per se, enables a socially beneficial alignment of incentives. During a period of “label exclusivity,” the right to reference a particular qualified biomarker in an approved drug label would belong exclusively to the Qualification submitter. The proposed mechanism and benefits of label exclusivity are explicated below.
Label exclusivity would leverage the important role that labels already play in biomarker diagnostics. Under statute, a drug will only be approved if its label comports with the scope of use deemed safe and effective by clinical investigation. 152 If a clinical trial relies upon a biomarker to establish safety or efficacy, the biomarker must be included on the drug label. 153 Thus, inability of a second comer to cite a qualified biomarker in a drug label is equivalent to exclusion from market if (and only if) the second comer relied upon the previously qualified biomarker to show safety or efficacy of its drug. The particular identity of the second comer's drug would be immaterial; rather, exclusivity would specifically regulate recitation of the qualified biomarker.
Returning to an example cited above, consider the breast cancer therapeutic Herceptin, which binds the cell division protein HER2. 154 Herceptin's FDA-approved label indicates, in part, that the drug is for “treatment of HER2-overexpressing metastatic breast cancer.” 155 However, Herceptin is not the only marketed drug that interacts directly with HER2. Others include Tykerb, Perjeta, and Kadcyla. 156 Under a hypothetical label exclusivity regime, if the sponsor of Herceptin (Genentech), had qualified HER2 overexpression as a biomarker, none of the later drugs, to the extent developed for treatment of HER2-overexpressing cancers, could have been brought to market until the label exclusivity period expired. Mechanistically, if a sponsor cannot obtain a label reflecting the safe and effective scope of drug use, its drug is barred from market.
A label exclusivity period of at least five years following Qualification is recommended, during which period no other party may receive approval of a drug label referencing the qualified biomarker. Existing regulatory exclusivity programs provide exclusivity periods ranging from six months to twelve years. 157 For instance, a six-month exclusivity benefit is granted to sponsors that conduct pediatric studies. 158 Toward the opposite end of the exclusivity spectrum, “orphan drugs” that treat, e.g., conditions affecting fewer than 200,000 individuals in the U.S. are entitled to seven years of exclusivity. 159 Data suggests that even the six-month pediatric benefit offers substantial incentive to drug sponsors. According to one estimate, the pediatric exclusivity benefit drove seventy-one billion dollars in pharmaceutical industry revenue over a fifteen-year period. 160 The same analysis observed that while short exclusivity periods are of substantial benefit to blockbuster drugs, gains accrue more gradually to others. 161 Accordingly, longer exclusivity periods incentivize participation by sponsors of narrower-use therapeutics, consistent with the patient-excluding function of biomarker label components. Moreover, to the extent label exclusivity would run in parallel with clinical trials, the period of exclusivity must outlast clinical trials to secure a sponsor's first mover market advantage.
In order to promote participation by academic and non-profit actors, as well as pharmaceutical producers, the label exclusivity program must provide meaningful incentives to parties that do not directly sponsor drugs. The presently proposed label exclusivity program would achieve this end by granting a single transferrable right to biomarker-submitting entities upon validation. More specifically, each submitter would be free to use or license access to its exclusive biomarker labeling right as it sees fit. Benefits of label exclusivity would accrue to a biomarker submitter regardless of that submitter's role in clinical trials. Thus, the right to use or license will provide value to both industry and non-industry biomarker submitters. Biomarker submitters with clinical trial goals will gain a regulatory edge by qualifying biomarkers they alone can reference in a label for five years. Non-industry submitters will be able to financially benefit from licensing a qualified biomarker at a commercially defined price, much as research institutions are presently able to out-license patent rights.
Importantly, Qualification of a biomarker has no value to a submitter if no other party seeks to use or license that biomarker within the exclusivity period. Thus, there is little incentive to submit a spurious biomarker. A benefit is garnered only if a biomarker is a tool so valuable that another would seek its use.
Review of approved drug labels supports the premise that valuable biomarkers are used repeatedly. As noted above, the FDA identifies 284 pharmacogenomic biomarker label components from 214 drugs, representing only 61 distinct biomarkers, of which the ten most-used represent 177 of the entries. Thus, substantial blocking or licensing value may be gained by label exclusivity. 162
A label exclusivity program would not limit biomedical research activity because the rights granted to biomarker submitters act only at the point of FDA drug label approval. Under this scheme, academic and non-profit institutions are uninhibited in their mission of advancing medicine. They are free to use the qualified biomarker as a research tool, to study its value in any desired therapeutic context, and to publish any results they may uncover. Moreover, any drug sponsor seeking to develop a therapeutic using a qualified biomarker can do so even during that biomarker's exclusivity period. The subsequent sponsor will be able to launch that drug onto the market as soon as any relevant exclusivity period expires. 163 Thus, research, public dissemination of knowledge, and drug development can be simultaneously incentivized under a label exclusivity regime.
A label exclusivity incentive will require that the FDA establish a system by which the proper “owner” of such exclusivity can be established. Because no competitive advantage presently accrues to a biomarker submitter, ownership is irrelevant to, and unaddressed by, the current regime. Thus, strategies for management of an incentivized Biomarker Qualification program are considered below.
The FDA cannot practically assess the “inventor,” nor the “inventiveness,” of each submitted biomarker, and therefore cannot define ownership in these terms. The resource demands of such review would be too burdensome. Indeed, the FDA does not review these issues in the drug approval context. Accordingly, practicality and agency policy dictate a “first-to-file” system for Biomarker Qualification. The first-to-file system will award those who transfer knowledge relating to a biomarker from, e.g., an academic or private context into the regulatory approval context. Features of a revised, first-to-file Biomarker Qualification program are proposed as follows:
First, the Qualification program should not interfere with past or current drug sponsor activities. Accordingly, a biomarker cannot be qualified if it is the subject of any previous FDA review or label by a different sponsor.
Second, the Biomarker Qualification program should not sustain any exclusivity in conflict with patent rights granted by the United States Patent and Trademark Office. Accordingly, a biomarker cannot be qualified if it is the subject of a patent issued to another and having an effective filing date earlier than that of the Biomarker Qualification submission. The strength of patent protection for biomarker-related inventions has been compromised by the Supreme Court's recent patent eligibility decisions in this area. Nevertheless, where a valid, earlier patent has issued, the FDA should recognize that ownership interest. The agency should provide an opportunity to adjudicate challenges between parties in this area. Moreover, the actual or perceived shortfall of current biomarker patent protections may increase the incentives for commercially interested parties to engage in Qualification.
Third, the breadth of exclusivity provided under Biomarker Qualification should be relatively narrow. Fortunately, the mechanism for regulating scope of Qualification is already in place. Under a revised program, as under the present one, Qualification should require the submitter to specify a particular, narrowly defined Context of Use. The biomarker and its Context of Use should not be read independently. Rather, each Qualification would be specific to the combination thereof.
Fourth, a fee payable to the FDA would both fund the Biomarker Qualification program and disincentivize frivolous submissions. The fee should be large enough to inhibit frivolous filings, but small enough to avoid placing substantial obstacles in the path of legitimate private and academic submitters. Fees could vary depending on the size of the submitting entity or real party of interest (e.g., $2,000 for individuals, $5,000 for small entities and universities, $20,000 for large entities).
An incentivized “first-to-file” system will result in a “race to the Qualification Office” among stakeholders. Such activity will add to the drug development toolkit and encourage IND, NDA, or BLA submissions by those seeking to take advantage of the label exclusivity period. In some instances, Biomarker Qualification submission will decide a competition between two sponsors for use of a biomarker. Moreover, in some instances, Qualification will be abused by a submitter that qualifies a biomarker it played little or no role in discovering. Although these potentially adverse scenarios underscore the “race to the Qualification Office,” and thus would spur activity, they also suggest a potential for inequity. The public benefit gained when a submitter enters a biomarker into the reservoir of FDA resources may, in one view, fail to offset the risk of awarding the label exclusivity benefit to a non-discovering party. Indeed, the proposed scheme does nothing to prevent this brand of abuse. However, the risks are offset in several ways. First, the period of exclusivity is finite; even at 5 years, it is, for instance, only a quarter of the presumptive twenty-year term of a patent. Second, under the Context of Use requirements, Qualification is intended to provide narrow rights. Third, the right to use the biomarker in a label is licensable, allowing motivated parties to negotiate terms. Finally, as knowledge of the Qualification program spreads, parties actively developing biomarkers will be able to submit in a timely manner, heading off competitors and unscrupulous intervenors.
This revised Biomarker Qualification program does not inhibit the current program's goal of collaborative resource development. Stakeholder collaboration in pre-competitive areas envisioned by the FDA will be no less valuable. Qualified biomarkers would still become available to the public for use in any drug approval program, albeit after a period of exclusivity. The added incentives will only spur additional participation by more parties.
VI. CONCLUSION
The recent passage of the 21st Century Cures Act provides a timely opportunity to revise the Qualification program. Under the minimal requirements of the 21st Century Cures Act, the FDA must produce updated Biomarker Qualification guidance documents by December 13, 2019. 164 Incorporating new incentives for industry and non-industry participation, particularly through a label exclusivity regime, could transform the floundering Biomarker Qualification program into a valuable tool. The FDA should take this opportunity to further promote pharmaceutical industry productivity.
Footnotes
1
U.S. D
2
C
3
C
4
Id. at 5. Over a similar period, and the cost of bring a new medicine to market had risen from about 1.1 billion (1995 to 2000) to about 1.7 billion (2000 to 2002). Id. at 4.
5
Id. at ii.
6
Id. at 15.
7
U.S. D
8
Woodcock & Woosley, supra note 1, at 6.
9
Federico Goodsaid & Felix Frueh, Biomarker Qualification Pilot Process at the US Food and Drug Administration, 9(1) AAPS J. e105, e106 (2007).
10
21st Century Cures Act § 3011, 21 U.S.C. § 357(e)(2) (2016). Prior to the 2016 passage of the 21st Century Cures Act, FDA defined a biomarker as “[a] defined characteristic that is measured as an indicator of normal biological processes, pathogenic processes, or responses to an exposure or intervention, including therapeutic interventions,” but that is not “an assessment of how an individual feels, functions, or survives.” This prior defintion excludes how an individual feels, functions, or survives as “clinical endpoints.” FDA-NIH B
11
Alla Katsnelson, Life sciences: Biomarkers on the brain, 479 N
12
Kyle Strimbu & Jorge A. Tavel, What are Biomarkers?, 5(6) C
13
FDA-NIH B
14
U.S. D
15
Id. at 24.
16
Laura Mezquita & Benjamin Besse, Sequencing ALK inhibitors: alectinib in crizotinib-resistant patients, a phase 2 trial by Shaw et al., 8(11) J. T
[hereinafter T
17
FDA-NIH B
18
Goodsaid & Frueh, supra note 9, at e106.
19
See William Leo McGuire, Current Status of Estrogen Receptors in Human Breast Cancer, 36 C
20
McGuire, supra note 10, at 642.
21
Id. at 643.
22
D
23
Id.
24
Id.; Jeremy Booth, A Short History of Blood Pressure Measurement, 70(11) P
25
The association of ALT with liver damage was first discovered in 1955, but was not formally accepted by the FDA until 2009. U.S. F
26
D
27
D
28
The DDT programs also included a qualification program for animal models of disease and a qualification for clinical outcome assessments such as a patient's overall mental state, function, and symptoms for use in, e.g., determining therapeutic value. U.S. D
29
Id.
30
Id.
31
Id.
32
Id.
33
Woodcock & Woosley, supra note 1, at 5.
34
FDA 2017 C
35
Id.
36
Woodcock & Woosley, supra note 1, at 6.
37
Id. at 10.
38
Id.
39
FDA 2017 C
40
Woodcock & Woosley, supra note 1, at 8.
41
C
42
D
43
Thomas G. Beach, A Review of Biomarkers for Neurodegenerative Disease: Will They Swing Us Across the Valley?, 6 (Suppl. 1) N
44
Id.
45
21 U.S.C. §§ 355(a)-355(b) (2017).
46
21 C.F.R. Ch. 1(D) (2017).
47
21 C.F.R. § 314.125(b)(2), (b)(5) (2017).
48
21 C.F.R. § 314.105(c) (2017).
49
Id.
50
However, the 21st Century Cures Act does reference biomarkers in two contexts. The Act introduces Biomarker Qualification and also states that development of drugs for rare diseases should “maximize the use of scientific tools or methods, including surrogate endpoints and other biomarkers.” 21st Century Cures Act §§ 3011-3012, 21 U.S.C. §§ 357, 360 (2016).
51
21 C.F.R. § 314.105(c) (2017).
52
The draft concept paper addressed the particular scenario of “a single test in conjunction with a single drug.” U.S. D
53
Id. at 10.
54
Id. at 3.
55
Id. at 7-8.
56
Id. at 16, 36.
57
L.M. Hinman et al., The drug diagnostic co-development concept paper: Commentary from the 3rd FDA-DIA-PWG-PhRMA-BIO Pharmacogenomics Workshop, 6 P
58
Id. at 378.
59
R
60
Id.
61
Id. at 7.
62
U.S. D
63
Id. at 7.
64
Id. at 6.; John J. Barron et al., HER2 Testing and Subsequent Trastuzumab Treatment for Breast Cancer in a Managed Care Environment, 14(8) T
65
U.S. D
66
A diagnostic device and corresponding therapeutic can be studied together in the same clinical trials. Id. at 6, 8, 12.
67
Id. at 6.
68
Id. at 10.
69
Id. at 11; 21 C.F.R. §§ 201.56-201.57 (2017).
70
U.S. F].
71
21st Century Cures Act § 3022, 21 U.S.C. § 355(g) (2016).
72
Id.
73
R
74
Cole Werble, Real-World Evidence: Advice, Principles and Examples Emerge from FDA, T].
75
21st Century Cures Act § 3022, 21 U.S.C. § 355(g) (2016).
76
Q
77
Q
78
Id. at 7-13.
79
Id. at 7-13, 18.
80
Id. at 18-22.
81
Id. at 7-13.
82
Id. at 3.
83
B
84
Id. at 2-3.
85
Id. at 3.
86
Id. at 6.
87
Id. at 8, 20. In somewhat more technical terms, the guidance identifies precision, accuracy, limit of detection, limit of quantitation, specificity, linearity and range, ruggedness and robustness as fundamental assessment factors. Id. at 18.
88
Id. at 13, 14. A unique feature of this guidance is that it establishes three tiers of risk/benefit (favorable, intermediate, and challenging) with increasing evidentiary demands. Id. at 16,17. Here, the FDA has provided tables identifying evaluative factors and what would satisfy “minimal” or “high” evidentiary criteria. This scheme maintains flexibility while also providing guidance that goes well beyond that available in statute, CFR, or other relevant guidance documents. Id. at 23.
89
21st Century Cures Act § 3011, 21 U.S.C. § 357 (2016).
90
Id.
91
T
92
21st Century Cures Act § 3011, 21 U.S.C. § 357 (2016).
93
21 C.F.R. § 201.56(a)(1) (2017); Robert N. Schuck & Joseph A. Grillo, Pharmacogenomic Biomarkers: an FDA Perspective on utilization in Biological Product Labeling, 18(3) AAPS J. 573, 574 (2016).
94
R
95
Schuck & Grillo, supra note 93, at 574.
96
Alexandre Vivot, Evidence for Treatment-by-Biomarker Interaction for FDA-Approved Oncology Drugs with Required Pharmacogenomic Biomarker Testing, 7 S
97
U.S. D
98
Id. at 30.
99
Ass'n for Molecular Pathology v. Myriad Genetics, Inc., 569 U.S. 576, 580 (2013).
100
Id. at 577, 585.
101
Id. at 595-596.
102
Mayo Collaborative Servs. v. Prometheus Labs., Inc., 566 U.S. 66 (2012).
103
Id. at 66, 72-73, 87.
104
Id.
105
Id. at 79-80.
106
Ariosa Diagnostics, Inc. v. Sequenom, Inc., 788 F.3d 1371, 1373 (Fed. Cir. 2015).
107
Id.
108
Id. at 1376, 1377.
109
Edward Abrahams, Landmark FDA Approval Bolsters Personalized Medicine, STAT N].
110
T
111
Navin Pinto & Eileen Dolan, Clinically Relevant Genetic Variations in Drug Metabolizing Enzymes, 12(5) C
112
Id. at 489.
113
Id.
114
Id. at 488.
115
Id. at 488-489.
116
M.D. Vincent et al., Biomarkers that Currently Affect Clinical Practice: EGFR, ALK, MET, KRAS, 19 (Suppl. 1) C
117
D
118
Vivot, supra note 96, at 2.
119
Id. at 1. One difficulty in justifying the cost of broader clinical trials, however, is the high rate of attrition of candidate drugs during clinical trials. C
120
Ethical concerns arising from the risk that a treatment is ineffective or harmful to biomarker-negative patients can also support use of enrichment. Vivot, supra note 96, at 1.
121
When a clinical trial is specific to a biomarker positive population, the approval of the drug is correspondingly narrow. Id.
122
Id.
123
Id.at 6.
124
Id.
125
Id.at 2.
126
L.M. Hinman et al., The drug diagnostic co-development concept paper: Commentary from the 3rd FDA-DIA-PWG-PhRMA-BIO Pharmacogenomics Workshop, 6 P
127
Id.
128
Q
129
FDA 2016 C
130
U.S. F]. [hereinafter L
131
Id.
132
U.S. F
133
D
134
D
135
Woodcock & Woosley, supra note 1, at 8.
136
Id. at 9; C].
137
C
138
L
139
T
140
B
141
U.S. F].
142
Id.
143
Id.
144
Id.
145
L
146
Beach, supra note 43, at S5-S6.
147
U.S. F].
148
FDA 2017 C
149
Id. at 5.
150
B
151
U.S. D
152
21 U.S.C. § 355(d) (2017).
153
Schuck & Grillo, supra note 93, 574-575; Vivot, supra note 96, at 1; R
154
Barron, supra note 64, at 760.
155
U.S. F
156
Maribeth Maher, Current and Emerging Treatment Regimens for HER2-Positive Breast Cancer, 39(3) P&T 206, 206 (2014).
157
CDER S
158
Id. at 2; 21 C.F.R. § 314.108 (2017).
159
CDER S
160
N
161
Id.
162
T
163
By contrast, a regulatory scheme under which development could only begin after the exhaustion of exclusivity would create further, substantial time delays for second comers.
164
Already, the FDA has indicated that Qualification submissions and assessments will be public in accordance with transparency provisions of the 21st Century Cures Act. U.S. F].