Thomas G. Roberts Jr.
At present, in the fourth quarter of 2004, there are more than 400 agents in clinical trials for the treatment of cancer, more than the number in any other therapeutic drug class.1 The sponsor of each of these agents ultimately seeks approval by the US Food and Drug Administration (FDA) to market their product in the United States for at least one indication.a However, each year only 2 to 7 new drugs and biologics for the treatment of cancer obtain this goal. FDA approval of the product represents the final hurdle that sponsors must clear in their time-consuming efforts to discover and optimize lead compounds, test new drugs in animals, and conduct clinical studies in humans. Without this license, the effort to develop a new drug, which may have consumed hundreds of millions of dollars and spanned more than a decade of time, is spent in vain.2, 3 Sponsors face considerable odds in their quest to gain FDA approval; the low odds of success are major contributors to the expense and risk associated with cancer drug development.4, 5 The most recent estimate of the average total cost (i.e., preclinical plus clinical costs) of developing a representative new drug from concept to FDA approval is $802 million,b with an average clinical development time of over 6 years.3, 5 Highly active products in a market without competition can generally be developed quickly, but many compounds are not highly active and require time and careful planning to demonstrate their potential.
Making decisions about the safety and efficacy of new drugs for licensure is just one of the roles of the FDA. The agency participates in the regulation of almost every step of clinical development. This regulation includes the oversight of clinical research, the evaluation of marketing claims for drugs seeking approval, and the monitoring of postmarketing activity for safety.6 Over time, the FDA has evolved its approach to these regulatory activities. During the last 15 years, in particular, decisions by the FDA have become more transparent, and the relationship between the agency and the pharmaceutical industry has become more collaborative. Sponsors now consult with the agency throughout all stages of development, particularly in defining endpoints, selecting indications, and designing clinical trials. The FDA has also shown flexibility with respect to its approval and review policies. This flexibility has been especially evident with respect to drugs intended to treat serious or life-threatening illnesses such as cancer and AIDS. In response to criticism about the pace of drug development and FDA review, programs such as Fast Track Designation, Priority Review, and Accelerated Approval have been introduced to allow patients suffering from cancer and other life-threatening illnesses to receive new medicines at relatively early stages of development. In this chapter, we discuss the role of the FDA and focus in particular on the requirements for achieving marketing approval.
THE AUTHORITY OF THE FDA
The mandate of the FDA has evolved over the 20th century to assume three fundamental assurances: safety, efficacy, and adequate and accurate labeling.6Prior to the 20th century, therapeutic products received little regulation, and fraudulent claims of medicines escaped the control of the government. In 1902 the United States became the first country to establish federal control over therapeutic products with the passage of the Biologics Control Act, whose purpose was to define and regulate the safety of biologic products. In 1938, Congress enacted the Federal Food, Drug, and Cosmetic Act, which, for the first time, required drugs to be safe for their intended use. It was not until almost 25 years later, in 1962, that Congress amended the 1938 act by codifying a requirement that therapeutic drugs also demonstrate efficacy in addition to safety. Under the amendments, Congress indicated that the FDA should require “full reports of investigations which have been made to show whether or not such drug is safe for use and whether such drug is effective in use.”7 Over the last 40 years, Congress has enacted subsequent laws governing the FDA, such as the Orphan Drug Act, the Federal Advisory Committee Act, and the Food and Drug Administration Modernization Act of 1997, statutes that are discussed later in the chapter.
In addition to laws originating in the US Congress, the FDA derives its authority and direction from multiple regulations. The executive branch develops regulations to provide interpretation of the laws. Regulations do not require congressional action, but they are subjected to a period of public review and scrutiny. When finalized, however, they are binding until revised or withdrawn. They are published in the Code of Federal Regulations. The FDA regulations stipulate, “The purpose of conducting clinical investigations of a drug is to distinguish the effect of a drug from other influences, such as spontaneous change in the course of the disease, placebo effect, or biased observation.”8 The regulations specify further that reports of “adequate and well-controlled investigations” are required to provide the primary basis for determining whether there is “substantial evidence” to support the claims of effectiveness for new drugs.
In addition to the laws and regulations governing the agency, the FDA itself can issue guidance documents to reflect its current thinking. These documents are not binding, but they can be helpful to industry in its interpretation of the laws and regulations. Examples of FDA guidance documents include “Guidance on Providing Clinical Evidence of Effectiveness for Human Drug and Biological Products,” issued in May 1998, and a series of guidances issued in conjunction with the International Conference on Harmonization on the conduct and analysis of clinical studies. When issues arise that require greater clarification, the Federal Advisory Committee Act provides the FDA authority to consult a panel of outside experts to solicit advice and recommendations on policy.6 By statute the FDA can only receive advice from chartered advisory committees. There is an Oncologic Drugs Advisory Committee (ODAC) that has a statutorily mandated pediatric oncology subcommittee. The ODAC has a committee chairperson and 12 voting members, who each serve staggered 4-year terms. At least one of the members is a patient representative and another is a statistician. The FDA typically follows the advice of the ODAC in response to questions addressed to the committee. However, the agency does not present all reviews for licensing applications to this committee and renders approval decisions based on the totality of the evidence, of which ODAC opinions represent a part.
INTERACTIONS BETWEEN THE FDA AND SPONSORS
It would be incorrect to think that sponsors and the FDA interact only at the time of FDA review of a licensing application. With increasing frequency, the FDA and drug sponsors communicate formally throughout the drug development process (Fig. 22.1).
Prior to the initiation of clinical testing of an investigational agent, a drug sponsor must file an investigational new drug (IND) application, and the FDA must grant permission to proceed with clinical studies. The IND application must include a copy of the protocol for each proposed study; a description of the physical, chemical, and biological characteristics of the agent; information on absorption, distribution, metabolism, and excretion; and an integrated summary of animal toxicology data.9 IND filing is also required for the use of approved products that a sponsor wants to study in new populations or in regimens for which the risks are unknown. The IND approval process provides federal oversight of clinical investigations and is designed to protect potentially vulnerable research subjects.
After the FDA grants IND approval, the investigational agent enters into sequential phases of clinical development, with each phase representing a distinct set of goals and challenges. Phase I trials represent the first experience with the agent, regimen, or dosing schedule in humans. The major objectives of phase I trials are to determine the maximum tolerated dose that is appropriate to use in phase II trials, characterize the toxicity profile, and collect relevant pharmacokinetic data. To meet these objectives, investigators typically treat cohorts of 3 to 6 advanced cancer patients with escalating doses of the investigational agent until a dose-limiting toxicity (DLT) is identified; additional patients are then treated at the dose just below the DLT level to confirm its safety. The purpose of phase II trials is to determine a preliminary assessment of the agent's efficacy and safety in a particular type of cancer. Phase II trials typically do not include a concurrent control group and usually enroll 15 to 50 patients. Phase III trials are large, randomized, multi-institutional, resource-intensive efforts that seek to determine if the investigational treatment shows clinically important benefit over a widely accepted standard of care.3, 10 A sponsor may refer to a phase III trial as a registration trial or pivotal trial if it plans to use the data generated by the study as the primary basis for marketing approval.
Figure 22.1 Overview of the drug development process. The multiple interactions between drug sponsors and the FDA reflect the convergence of the duties and responsibilities throughout the phases of development. IND, initial new drug application; NDA, new drug application; ODAC, Oncologic Drug Advisory Committee.
The optimal times for the sponsor to solicit FDA input occur when particular landmarks are reached in the development process. Prior to filing a formal IND application, it can be helpful to the sponsor to request a Pre-IND Meeting to solicit FDA advice regarding requirements for preclinical studies and to discuss clinical development scenarios, particularly for new molecular entities (NMEs).
After the completion of phase I trials, a sponsor may request an End of Phase I Meeting to help in the planning of phase II trials, but such meetings are not universally sought. Once data are available from phase II studies, however, the sponsor almost always requests a meeting to discuss safety issues and/or approaches that are likely to establish efficacy of its products. In this critical End of Phase II Meeting, the sponsor can discuss clinical trial design and analysis plans for proposed registration studies. The goal of a registration study is to design and complete a trial or group of trials that will produce the relevant data from which a marketing claim may be made. Table 22.1 outlines the many issues that sponsors must consider as they design and execute these pivotal studies. A analysis of NMEs approved from 1987 to 1995 demonstrated that agents that were discussed in Pre-IND and End of Phase II conferences with the FDA had shorter development times than those that were not discussed.11
At the conclusion of the End of Phase II Meeting, the sponsor can submit a detailed proposal for their critical phase III trial. This proposal is submitted for Special Protocol Assessment, with a legally mandated FDA review time of 45 days. In its assessment of the protocol, the FDA responds to specific questions the sponsor may have and provides general comments. Three types of studies are eligible for Special Protocol Assessment: animal carcinogenicity studies, product stability protocols, and clinical protocols for Phase III studies intended to form the basis for an efficacy claim. If the FDA concurs with the protocol, then a commitment exists to accept the design and size of the study as appropriate for approval, but the agency makes no commitment for ultimate approval, which ultimately depends on the data. Additional meetings are often scheduled before filing for a licensing application (Biologic Licensing Application [BLA] or New Drug Application [NDA]) to address particular chemical and manufacturing concerns (Product Meeting) and to seek input on the organization and format of the licensing application (Pre-BLA or Pre-NDA Meeting) and the product labeling.
TABLE 22.1 KEY QUESTIONS THAT SPONSORS MUST CONSIDER IN THE PLANNING OF REGISTRATION TRIALS
FDA PROGRAMS WITH SPECIAL RELEVANCE TO ONCOLOGY
The regulatory review process has undergone significant evolution since 1980, prompted by concerns that patients lacked access to innovative therapies and were impacted negatively by a lengthy review process.12, 13 AIDS advocates lobbying for early access to antiretroviral drugs provided the initial impetus for a series of programs designed to expedite the drug development process, and advocates from the cancer community joined in these efforts shortly thereafter,14especially those concerned with breast cancer. The programs initiated in response to the AIDS crisis have also become available to cancer products. Table 22.2 presents some of the reform programs introduced since 1980 with particular relevance to cancer.
In 1983, The US Congress enacted the Orphan Drug Act, allowing the FDA to provide incentives and grants for certain drugs intended to treat diseases with a prevalence of less than 200,000 people in the United States or those that affect more than 200,000 people but for which there is no reasonable expectation that the costs of development would be recovered from sales in the United States. The designation applies to a product used for a specific disease. Because there are more than 200 histological neoplastic subtypes, many cancer indications have the potential to obtain Orphan Drug designation. Orphan Drug designation qualifies the sponsor for a longer period of marketing exclusivity (7 years starting on the approval date, compared with 5 years), a 50% tax credit for money spent on clinical trials for the orphan indication, exemptions from application filing user fees, and the option to compete for FDA development grants.15 As of 2004, there have been approximately 40 cancer drugs or biologics that have been approved for marketing with Orphan Drug designation since the inception of the program. Cancer drugs and biologics comprise about 20% of all Orphan Drug approvals over this time period. Examples of approved products with Orphan Drug designation include pemetrexed (Alimta) in the treatment of malignant pleural mesothelioma, bortezomib (Velcade) in the treatment of multiple myeloma, and rituximab (Rituxan) in the treatment of non-Hodgkin's B-cell lymphoma. The Orphan Drug Program has been successful in providing incentives for sponsors to develop agents for rare diseases, but the program was not designed to address the criteria for approval nor to establish new target goals for FDA review times.
The “Fast Track” Programs
Over the period from 1992 to 1997, the US Congress enacted two laws, the Prescription Drug User Fee Act of 1992 (PDUFA) and the FDA Modernization Act of 1997 (FDAMA), which together have had a major impact on the process that the agency uses for reviewing and approving cancer drugs.16 The legislation in these statutes is far-reaching, but from the perspective of medical oncology, the three programs introduced over this time period with the greatest impact are Fast Track Designation, Priority Review, and Accelerated Approval. Each of these programs exerts its impact at different stages of the drug development process.
Fast Track Designation
The Fast Track mechanism represents a formal structure by which sponsors may interact with the FDA. As described in the FDAMA, Fast Track designation can be granted to the combination of a product and a claim that the product addresses an unmet medical need for a serious or life-threatening illness. The benefits of Fast Track designation, if granted, include scheduled meetings to seek FDA input into development plans, the option of submitting an NDA in sections rather than all components simultaneously, and the option of requesting evaluation of studies using surrogate endpoints. For the most part, these interactions were already available to sponsors prior to the FDAMA, but the Fast Track mechanism formalized the approach for a subset of drugs. Since its inception, approximately half of the Fast Track designations have gone to drugs for the treatment of cancer or HIV/AIDS. In a survey of industry representatives, respondents indicated that the FDA has been able to approve 75% of the requests for Fast Track designation within 2 months.16Initially in 1998, and then updated in 2004, the FDA issued a guidance document describing the Fast Track programs in more detail.17
TABLE 22.2 RECENT FDA POLICY CHANGES RELEVANT TO CANCER
Of the programs that have been designed to expedite the drug development process, none has had more impact or has generated more public commentary than Accelerated Approval. The Accelerated Approval regulations were added in 1992 (under Subpart H for drugs and under Subpart E for biologics). These provisions allow the FDA to approve agents intended to treat serious or life-threatening illnesses before the clinical benefit necessary to meet the standard for Regular Approval has been demonstrated. Using the Accelerated Approval mechanism, the FDA can grant a provisional approval on the basis of a surrogate measure of clinical benefit (e.g., tumor shrinkage) if the treatment is considered superior to available therapy for a serious or life-threatening illness.18 The FDA grants the approval conditionally and receives the sponsor's agreement to complete confirmatory phase IV trials in a timely manner during the postapproval period. If these trials do not confirm a clinical benefit, the drug can be withdrawn from the market. To date, however, the FDA has requested suspension of marketing of only a single product,c despite the fact that as of August 2004 sponsors had completed the phase IV trials required for an upgrade to Regular Approval in only 6 of the 23 oncology-related approvals.19 Table 22.3 lists oncology agents approved using the Accelerated Approval mechanism through August 2004. The FDA and others in the medical community have expressed concern over the failure of sponsors to complete confirmatory phase IV studies,20, 21 prompting a focus on the issue during the March 2003 meeting of the ODAC. The regulatory standard for an Accelerated Approval can be less challenging than the standard for Regular Approval, because Regular Approval is predicated on the demonstration of a clinical benefit (e.g., prolonged survival or an improved quality of life) and not on a surrogate endpoint.
TABLE 22.3 ACCELERATED APPROVALS IN ONCOLOGY THROUGH AUGUST 2004
The FDA granted its first oncology-related Accelerated Approvals in 1995 (dexrazoxane for cardiomyopathy and liposomal doxorubicin for the treatment of AIDS-related Kaposi's sarcoma). Since then approximately 30% of the agents approved in medical oncology have utilized this mechanism. A recent review of the Accelerated Approval Program from 1992 through 1997 estimated that access to the program shortened overall development time by as much as 4 years in some instances.22 It should be noted that Priority Review designation and eligibility for Accelerated Approval are assessed independently, and therefore the FDA may make discordant decisions in granting these designations. It is possible to receive Accelerated Approval and a Standard Review (10-month target), and it possible to receive Regular Approval with a Priority Review (6-month target).6
Whereas Fast Track designation applies throughout the development process, and Accelerated Approval applies following an intermediate stage of development before formal demonstration of clinical benefit, Priority Review is relevant only after a claim has been submitted to the FDA for review. As described in the PDUFA and the FDAMA, the FDA designates reviews for NDAs as either Standard or Priority. A Standard Designation sets the target date for the agency to complete all aspects of a review and to take action on the application (i.e., approve or not approve) at 10 months after the date of NDA filing. In comparison, a Priority Designation sets the target date for the FDA action at 6 months. Similar to the Fast Track program, Priority Review is intended for those products that address unmet medical needs. Fast Track designation does not automatically lead to eligibility for Priority Review, but FDA guidance states that achieving Fast Track Designation means that a product “ordinarily will be eligible for Priority Review.”17 Since 1994, approximately 50 cancer-related marketing claims involving 24 drugs have been granted Priority Review by the FDA, including such drugs as docetaxel (Taxotere) in the treatment of advanced non–small cell lung cancer after failure of prior platinum-based chemotherapy and topotecan (Hycamptin) in the treatment of patients with metastatic carcinoma of the ovary after failure of initial or subsequent chemotherapy. The quickest Priority Reviews to date were the reviews of imatinib mesylate (Gleevec) for the treatment of chronic myelogenous leukemia and of oxaliplatin (Eloxatin) in combination with 5-fluorouracil and leucovorin for the treatment of relapsed or refractory colorectal cancer; the claims for these drug uses received approval after just 10 weeks and 7 weeks of FDA review, respectively.
EVIDENCE REQUIRED FOR APPROVAL
Strategies for Registration
The FDA does not approve a drug but rather approves a claim about the use of a drug.6 However, it is common after approval for oncologists to use agents for other than the approved claim.23 A sponsor can take multiple pathways to gain approval for the first marketing claim of its drug or biologic agent. A major decision that the sponsor must make is for what line of treatment it will seek approval. A common approach is to begin therapeutic development with the use of a new single-agent as secondor third-line therapy for relapsed or refractory disease. The assumption is that activity in treatment-experienced patients will translate into clinical benefit for treatment-naïve patients. The phase II trials of gifitinib (Iressa) for the treatment of advanced non–small cell lung cancer after failure of platinumand docetaxel-based chemotherapies illustrate this approach, since the data from these trials supported an Accelerated Approval for gefitinib in the third-line setting.24, 25 In comparison, first-line approvals tend to be more difficult to achieve, since the new agent must prove at least as good as the best treatment available. In order to show either an outcome difference or establish noninferiority, trials supporting first-line approvals will typically require more patients and longer follow-up than trials in refractory or relapsed populations, unless the new agent is impressively superior to available therapy. First-line approvals are also less likely to earn an Accelerated Approval, because there is at least one first-line regimen established for most cancers, and Accelerated Approval requires a demonstration of superiority if a standard therapy exists.
A common strategy in the first-line setting is to examine whether a new agent adds any benefit (or risk) to an established regimen. The registration trial of bevacizumab (Avastin) provides a recent example of this strategy: previously untreated patients with metastatic colorectal cancer were randomly assigned to receive irinotecan, bolus fluorouracil, and leucovorin (IFL) plus bevacizumab versus IFL alone.26 Despite the risks associated with seeking a first-line indication, the effort can be worthwhile because a first-line approval will apply to a larger treatment population.
Another decision that sponsors must make is how many registration studies to undertake. In general, the FDA regards the results of a single trial supporting approval as inadequate evidence.27 However, a single study may be considered adequate for licensing when the study is a large and well designed multicenter study, the implementation is of unquestionable quality, the findings appear clinically important, the results are statistically persuasive, and the confirmation of the results would present ethical or logistical hurdles.
General Approval Considerations
From a regulatory perspective, an intervention is a substance administered to a patient or a procedure performed on a patient with the intention of altering or interfering with the natural history of a disease. If the intervention disrupts or arrests the disease process so that suffering will be relieved or survival will be extended, and if it provides an acceptable risk for further suffering or loss of life, then the intervention can be considered a safe and effective treatment and is therefore eligible for FDA approval. At the most fundamental level, the agency must determine if there are differences between the treatment and control groups and then determine if the differences are due to the intervention under review. Establishing an appropriate control is therefore critical, because it is often difficult to determine efficacy without a control. Controls can either be historical or concurrent, and sponsors using the Accelerated Approval mechanism have usually employed historical controls in their phase II trials that support approval. However, concurrent controls have multiple advantages. Specifically, concurrent controls assure consistency of diagnosis, uniform techniques and frequency of clinical assessments of response and toxicity, a common level of supportive care, and consistency of administration of an intervention. Concurrent controls also allow for the possibility of reducing bias through blinding, although it is often difficult to blind cancer trials because of side effects. The goal of clinical research is to minimize bias and uncertainty, and concurrent controls with effective randomization can minimize both. When registration studies use an active concurrent control, it is usually the best available treatment.
The regulations guide the overall process of FDA approval, but the steps and the sequence of steps that the agency can take in this process are not completely standardized and therefore allow for some flexibility. During a review of a licensing application, the separate disciplines at the FDA review all primary data. The agency determines the nature of the claim, prepares a survey of available therapies that address the same problem (focusing on the nature and duration of the treatment effect of each therapy), and reconstructs what data and analysis would support a new claim. The FDA then identifies the key elements of the submission that provide the data to support the claim. The agency reviews the study protocols (focusing on eligibility, endpoints, measured variables, and planned analysis) and conducts its own analysis with regard to each patient's meeting eligibility, having the requisite measurements, and completing the study. The treatment effect is then determined, and the FDA analysis is compared to the sponsor's planned and submitted analyses.
If the FDA analysis demonstrates that the efficacy of the experimental treatment is either inferior to a control or cannot be distinguished from placebo effect, then there is no need for further analysis, and the application is denied. If the agency determines that the efficacy of the treatment is either superior to or at least not inferior to an active control, and if dropouts and censored patients can be adequately accounted for, the FDA proceeds with a safety analysis for each study to determine an estimate of the risk-benefit ratio (Fig. 22.2). Although a discussion of noninferiority is beyond the scope of this chapter, the factors that are critical in the design and analysis of a noninferiority study are the magnitude and reproducibility of the active control effect; defining the acceptable margin of retaining the active control effect; and describing in detail the analytic plan, including provisions for missing data and censoring. After completing its analysis, the agency compares the efficacy data and safety data to the proposed marketing claims and then, if approval can be made, adjusts the claims in alignment with the conclusions supported by the data.
Figure 22.2 Flow diagram of the initial FDA review process. The first step is to determine if the intervention showed its expected activity. If the expected activity is not demonstrated, the analysis is stopped without proceeding to a safety analysis. If the expected activity is demonstrated, the agency proceeds with a safety analysis, and the intervention may be eligible for approval if a favorable risk-benefit ratio is found.
Assessment of Benefit
The assessment of risk is reasonably standard for oncology studies, and the same principles of a graded scale (e.g., NCI Common Toxicity Criteria) are applied to the oncology field as is done in other areas of medicine. The assessment of benefit on the other hand is less standardized (Table 22.4) and is often the subject of considerable dialogue and discussion.
In the early 1970s and 1980s, the FDA approved oncology drugs on the basis of response rate alone, i.e., the fraction of patients experiencing tumor shrinkage. By the mid-1980s the ODAC advised the FDA against using response rate as the sole basis for approval, since the possible benefit of a response may be outweighed by toxicity. Moreover, the correlation between response rate and survival benefit had not been established for most solid tumors.28 In response to this recommendation, the FDA adopted a new position calling for an improvement in survival or patient symptoms as the standard for Regular Approval.29 Guidance documents promulgated in the 1980s specified further that the requirement for efficacy should be demonstrated by prolongation of life, a better life (e.g., relief of symptoms), or an established surrogate for at least one of these.18
TABLE 22.4 TYPES OF ENDPOINTS USED IN CLINICAL ONCOLOGY
In 1991, the National Cancer Institute and the FDA jointly examined the potential of various endpoints to demonstrate clinical benefit and again considered the issue of whether response rates could be used as a valid endpoint for drug approval. The two agencies concluded that complete responses of reasonable duration, particularly in acute leukemia, could be a potentially valid endpoint, as long as they correlated with clinical benefits such as reduced transfusion requirements. Partial responses were also considered as possibly valid endpoints, but only after considering their duration, the associated toxicity, and the potential to relieve tumor-related symptoms.30
A summary of the endpoints used in studies to support licensure for 90 separate claims for a variety of indications between 1985 and 2003 is shown in Table 22.5. The data show that most applications have multiple endpoints, and in most cases the determination of approving a claim for marketing was made on the totality of the evidence and not on a single endpoint. If several endpoints indicate the same trend, the support for considering approval is usually increased. Table 22.6 presents a tabulation of the endpoints from Table 22.5 to indicate their frequency in the aggregate of all 90 approved claims. The data show that few endpoints were used as sole criteria for marketing approval, but response rate was used most frequently, appearing in 60% of all claims and about 75% of Accelerated Approvals. In the hematologic malignancies, complete responses of a predetermined duration can be considered of patient benefit due to absence of disease complications such as bleeding, need for transfusions, or infections. Overall survival was used as a component of 27% of all claims but as a single criterion in only 5% of claims—twice for first-line colorectal cancer, once for first-line glioblastoma multiforme, and once each for second-line non–small cell lung cancer and glioblastoma multiforme. The frequency of use of overall survival and time to disease progression is quite similar. Disease-free survival and recurrence rate have been primarily used in the adjuvant setting.
The difficulty in establishing a coherent policy on the use of response rates to support approval relates to the lack of a documented relationship between tumor response and clinical benefit for most tumors. This relationship has been evaluated adequately in only a few tumor types, often with conflicting results. For example, Buyse et al. performed a meta-analysis of 25 trials in colorectal cancer involving fluoropyrimidines and concluded that tumor response was a significant independent predictor of survival.31 In contrast, Chen et al. did not find a significant correlation of phase II response rates with median survival times in phase III trials of the same regimen in small cell lung cancer.32 Variations in the definition of response exist and may account for some differences in interpretation. Response rates may provide a reasonable surrogate for survival, but only in certain diseases and with certain drugs. At present, the FDA continues to approve agents on the basis of response rates, but most of these approvals only meet the standard for Accelerated Approvals (Table 22.3) and therefore require confirmatory trials that employ clinical benefit endpoints.
Time to Tumor Progression
Some agents, particularly those with mechanisms of actions that are not cytotoxic, may produce clinical benefit by delaying tumor progression, with relatively low response rates. For these agents, time to tumor progression (TTP) or progression-free survival (PFS) may be a more appropriate surrogate endpoint than response. TTP or PFS is typically defined as the time from enrollment to documented progression of tumor size based on imaging tests (and not on biochemical tumor markers). The difference between TTP and PFS is that, in the latter, death is considered progression. Two major benefits of using progression include the potential need for smaller sample sizes and shorter follow-up times than for overall survival studies. Delayed progression in conjunction with response rates has been considered an adequate surrogate for clinical benefit in evaluating hormonal treatments for breast cancer, supporting Regular Approvals for exemestane, toremifene, anastrozole, letrozole, and fulvestrant in randomized trials comparing each of these with tamoxifen or with another approved hormonal agent.18
The ODAC has had difficulty accepting TTP as an endpoint for first-line indications for a number of submissions.6 Their negative assessments relate to the general limitations of using TTP as an endpoint. First, TTP can only be evaluated reliably in the context of randomized trials, due to the difficulty of comparing results from historical controls in which assessment of tumor status posttherapy is not consistent across trials. In addition, there are few historical control databases that have used TTP as an endpoint. Second, even in randomized trials, TTP outcomes may be influenced by the frequency of obtaining imaging studies. If effective blinding is not performed, investigator bias may influence decisions regarding the timing of imaging studies and the interpretation of clinical data. More recent ODAC discussion of progression as a basis for approval has resulted in a reevaluation of the applicability and definition of this endpoint. The committee expressed a preference for progression-free survival over time to progression because PFS includes deaths from all causes, including disease progression and drug toxicity. The use of PFS in appropriate disease settings is attractive because of the shorter time frame to assess results compared with overall survival and because of the absence of potential confounding of a survival effect by subsequent therapy.
TABLE 22.5 ENDPOINTS FOR APPROVALS OF ONCOLOGY DRUG MARKETING APPLICATIONS JANUARY 1, 1985, TO NOVEMBER 1, 2003
TABLE 22.6 TABULATION OF REGISTRATION STUDY ENDPOINTS THAT SUPPORTED MARKETING CLAIMS FROM JANUARY 1, 1985, TO NOVEMBER 1, 2003
Relief of Tumor-Related Symptoms
Relief of tumor-related symptoms has been used as a primary or supportive basis for licensing in 6 applications for antitumor products between 1985 and 2003 (Table 22.5). Examples of symptom benefit supporting licensure include the approval of mitoxantrone in combination with corticosteroids as initial chemotherapy for the treatment of patients with pain related to advanced hormone-refractory prostate cancer; the pre-defined “clinical benefit response,” a composite endpoint of pain, weight gain, and performance status, that provided support for the approval of gemcitabine for the treatment of locally advanced or metastatic pancreatic cancer, even though a small but significant improvement in survival was the primary basis for approval; changes in respiratory symptoms in small cell lung cancer associated with topotecan use; and changes in respiratory symptoms in non–small cell lung cancer associated with the use of porfimer sodium in the photodynamic treatment of endobronchial lesions.
Of all of the possible endpoints to assess benefit, survival is the least prone to bias and the least controversial in terms of its ability to support an FDA approval. In fact, all other endpoints used to justify approval must not be used to support a treatment that has a negative impact on survival.6 Survival as an endpoint, however, is not without its limitations. The accurate recording of survival times requires long follow-up and potentially large sample sizes depending upon the magnitude of drug effect. Subsequent treatments after progression can also confound or mitigate the interpretation of survival effect, particularly in crossover study designs. In part due to these limitations, over the last 18 years overall survival has contributed to the basis for approval in only 27% of claims and formed the primary basis for approval in only 8% of claims (Tables 22.5 and 22.6).
The industry and the FDA are keenly interested in the employment of new endpoints to support approvals. Potential endpoints of particular interest include progression-free survival (including symptomatic progression), surrogate biomarkers,33 and patient-reported outcomes. The FDA is also interested in the use of interim analysis based on endpoints that differ from the endpoints planned for final analysis and in the use of methods to reduce bias such as centralized endpoint assessment committees.34
Over the past 10 to 15 years, a growing number of academic and industry groups have begun to study trends in FDA actions and have monitored statistics such as development and review times. This work reflects the growing importance of cancer drugs and biologics as an industry sector. In 2003, the total worldwide market for oncology-related products was approximately $36.8 billion, and expenditures on oncology-related products have grown to almost 10% of the total $430 billion worldwide market for pharmaceuticals.35 By 2008, the oncology market is projected to exceed $60 billion.
Regulatory review represents only 5 to 10% of total clinical development time for most new oncology products (Fig. 22.3).6 This percentage has decreased over time, from 33% in the 1960s to approximately 25% in the 1980s and to approximately 15% in the early 1990s.22 This trend is not unique to oncology products. The approval phase of drug development has fallen for drugs across all therapeutic classes. However, the trend toward decreasing review times has been most pronounced in the fields of cancer and AIDS. Of the 50 fastest drug approvals from the period 1963 to 2002, almost half involved agents used to treat cancer, AIDS, or AIDS-related opportunistic infections.36
Figure 22.3 Development and approval times for new molecular entities approved 1995–2002. Regulatory review time (approval period) now represents approximately 5 to 10% of total clinical development time for most new oncology products. (The data are from the FDA.)
Beginning in 1993, the payment of user fees by drug sponsors became linked to a series of FDA performance goals that committed the agency to incremental shortening of the review process.22 Under the most recent amendments to these laws in 2002, the FDA is mandated to review and act on 90% of Standard Review original applications within 10 months of receipt, and on 90% of Priority Review original applications within 6 months.12 There is evidence that the agency is able to meet these targets. At least one recent analysis demonstrated that from 1980 to 2001, the mean time period for approval for antineoplastic products granted Priority Review was 42% shorter than those granted Standard Review.12
There is also evidence that the Accelerated Approval mechanism is having its intended effect. In a Tufts Center for the Study of Drug Development analysis of new therapeutics approved for marketing in the United States between 1980 and 2001, antineoplastic small molecule drugs that received accelerated approval had clinical development times that were on average 43% shorter than agents that did not utilize this mechanism.12 Despite the shortened development times of molecules approved under the Accelerated Approval mechanism, cancer drugs as a class tend to have longer development times than agents in other therapeutic categories. For example, in a Tuft's Center analysis involving a sample of 38 antineoplastic drugs approved from 1980 to 2001, the length of clinical trials was 81.6 months, and the median approval time was 12.8 months (for a total development time of 101 months). In contrast, cardiovascular drugs had a median length of clinical trials of 59.6 months and a median approval time of 26.7 months (for a total of 91.7 months).12
Levels of Innovation
The approval of NMEs, defined as medications containing an active substance that has never before been approved for marketing in any form in the United States, or new biopharmaceuticals is often taken as a measure of innovation in the pharmaceutical industry.37 In general there has been a decline in the number of NMEs or biopharmaceuticals over time across therapeutic classes. In contrast, the number of NMEs approved per year in the oncology field increased substantially from 1991 to 2000, although it has plateaued since 2001 (Fig 22.4). In 2003, the FDA approved seven oncology-related NMEs, the most of any therapeutic class.
The FDA continues to strive to become more collaborative in its approach to reviewing drugs for approval and to streamline the regulatory and product development processes. In May 2003, the NCI and the FDA (both of which are Health and Human Service agencies) agreed to share knowledge and resources to facilitate the development of new cancer drugs. In particular, the agreement focused on developing markers of clinical benefit; creating a cancer bioinformatics infrastructure to improve data collection across all of the sectors involved in the development and delivery of cancer therapies; and encouraging collaborative training, rotations, and joint appointments in the two agencies.38
Figure 22.4 Number of approved new molecules for cancer use by the FDA. The number of approvals per 5-year period was relatively constant through 1990 and has increased substantially since then. (Adapted with permission from Rothenberg ML, Carbone DP, Johnson DH. Improving the evaluation of new cancer treatments: challenges and opportunities [opinion]. Nat Rev Cancer 2003;3:303–309. Data are updated through August 2004 and are available from the FDA at: http://www.accessdata.fda.gov/scripts/cder/onctools/statistics.cfm#count.)
The future will bring many challenges for the FDA in its regulation of cancer-related products. Integrating pharmacogenomic data into the review process represents one such challenge. The expanded use of imaging to answer questions of drug mechanism, distribution, and efficacy in human subjects is another accelerating trend in cancer drug development. To this point, observers in the academic community have recently called upon the agency to mandate sponsors of new targeted anticancer treatments to invest in the research that can define subsets of patients who are likely to have responses to treatment based on the molecular signatures of their tumors.17 Whether the agency will move toward requiring sponsors to invest in studies to define molecular signatures of response is unclear. Two additional challenges are the need to cooperate with the Center for Medicare and Medicaid Services and the continuing need to balance early access to potentially life saving drugs with the assurances of safety and efficacy. Guiding the FDA in these challenges will be the increasing trends toward more transparency, flexibility, and cooperation.
The author would like to thank Steven Hirschfeld of the Food and Drug Administration for his participation in helpful discussions.
The Center for Biologics Evaluation and Research:http://www.fda.gov/cber/
The Center for Drug Evaluation and Research:http://www.fda.gov/cder/
Oncology Tools Website:http://www.fda.gov/cder/cancer
Public discussion of clinical trial endpoints:http://www.fda.gov/cder/drug/cancer_endpoints/default.htm
1. IMS LifeCycle, R&D focus. In: Parexel's Pharmaceutical R&D Statistical Sourcebook 2003. Waltham, MA: Parexel.
2. Dimasi JA. New drug development in the United States from 1963 to 1999. Clin Pharmacol Ther 2001;69:286–296.
3. Roberts TG Jr, Lynch TJ Jr, Chabner BA. The phase III trial in the era of targeted therapy: unraveling the “go or no go” decision. J Clin Oncol Oct 1 2003;21:3683–3695.
4. Von Hoff DD. There are no bad anticancer agents, only bad clinical trial designs. Twenty-first Richard and Hinda Rosenthal Foundation Award Lecture. Clin Cancer Res 1998;4:1079–1086.
5. DiMasi JA, Hansen RW, Grabowski HG. The price of innovation: new estimates of drug development costs. J Health Econ 2003; 22:151–185.
6. Hirschfeld S, Pazdur R. Oncology drug development: United States Food and Drug Administration perspective. Crit Rev Oncol Hematol 2002;42:137–143.
7. Mark B. McClellan, Commissioner, Food and Drug Administration. Speech before the Food and Drug Law Institute, April 1, 2003. Available at http://www.fda.gov/speeches/speechli.htm. Accessed April 28, 2003.
8. CFR Title 21 Part 314, Subpart D, Section 126.
9. US Food and Drug Administration. Content and Format of Investigational New Drug Applications (INDs) for Phase 1 Studies of Drugs, Including Well-Characterized, Therapeutic, Biotechnology-derived Products. Rockville, MD: US Food and Drug Administration, November 1995.
10. Roberts TG Jr, Lynch TJ Jr, Chabner BA. Identifying agents to test in Phase III clinical trials. In: Figg WD, ed. Pharmacokinetics and Pharmacodynamics of Anti-cancer Drugs. Totowa, NJ: Humana Press, 2004.
11. DiMasi JA, Manocchia M. Initiatives to speed new drug development and regulatory review: the impact of FDA-sponsored conferences. Drug Inf J 1997;31:771–788.
12. Reichert JM. Trends in development and approval times for new therapeutics in the United States. Nat Rev Drug Discov 2003;2: 695–702
13. Lasagna L. Congress, the FDA, and new drug development: before and after 1962. Perspect Biol Med 1989;32:322–343.
14. Anderson LF. Cancer and AIDS groups push for changes in drug approval process. J Natl Cancer Inst 1989;81:829–831.
15. Milne CP. Orphan products: pain relief for clinical development headaches. Nat Biotechnol 2002;20:780–784.
16. Milne C-P. Fast track designation under the food and drug administration: the industry experience. Drug Inf J 2001;35:71–83.
17. US Food and Drug Administration. Fast Track Drug Development Programs: Designation, Development, and Application Review. Rockville, MD: US Food and Drug Administration, September 1998.
18. Johnson JR, Williams G, Pazdur R. End points and United States Food and Drug Administration approval of oncology drugs. J Clin Oncol 2003;21:1404–1411.
19. Roberts TG Jr, Chabner BA. Beyond fast track for drug approvals. N Engl J Med 2004;351:501–505.
20. Mitka M. Accelerated approval scrutinized: confirmatory phase 4 studies on new drugs languish. JAMA 2003;289:3227–3229.
21. Schilsky RL. Hurry up and wait: is accelerated approval of new cancer drugs in the best interests of cancer patients? J Clin Oncol 2003;21:3718–3720.
22. Shulman SR, Wood-Armany MJ. Accelerating access to cancer drugs. J Biolaw Business 1999;2(2):38–44.
23. Laetz T, Silberman G. Reimbursement policies constrain the practice of oncology. JAMA 1991;266:2996–2999.
24. Kris MG, Natale RB, Herbst RS, et al. Efficacy of gefitinib, an inhibitor of the epidermal growth factor receptor tyrosine kinase, in symptomatic patients with non-small cell lung cancer: a randomized trial. JAMA 2003;290:2149–2158.
25. Fukuoka M, Yano S, Giaccone G, et al. Multi-institutional randomized phase II trial of gefitinib for previously treated patients with advanced non-small-cell lung cancer. J Clin Oncol 2003;21:2237–2246.
26. Hurwitz H, Fehrenbacher L, Novotny W, et al. Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer. N Engl J Med 2004;350:2335–2342.
27. US Food and Drug Administration. Providing Clinical Evidence of Effectiveness of Human Drugs and Biological Products. Rockville, MD: US Food and Drug Administration, May 1998.
28. Pazdur R. Response rates, survival, and chemotherapy trials. J Natl Cancer Inst 2000;92:1552–1553.
29. Johnson JR, Temple R. Food and Drug Administration requirements for approval of new anticancer drugs. Cancer Treat Rep 1985;69:1155–1159.
30. O'Shaughnessy JA, Wittes RE, Burke G, et al. Commentary concerning demonstration of safety and efficacy of investigational anticancer agents in clinical trials. J Clin Oncol 1991;9: 2225–2232.
31. Buyse M, Thirion P, Carlson RW, et al. Relation between tumour response to first-line chemotherapy and survival in advanced colorectal cancer: a meta-analysis. Meta-Analysis Group in Cancer. Lancet 2000;356:373–378.
32. Chen TT, Chute JP, Feigal E, et al. A model to select chemotherapy regimens for phase III trials for extensive-stage small-cell lung cancer. J Natl Cancer Inst 2000;92:1601–1607.
33. Kelloff GJ, Coffey DS, Chabner BA, et al. Prostate-specific antigen doubling time as a surrogate marker for evaluation of oncologic drugs to treat prostate cancer. Clin Cancer Res 2004;10: 3927–3933.
34. US Food and Drug Administration. On the Establishment and Operation of Clinical Trial Data Monitoring Committees. This is a Draft Guidence by the FDA available at: http://www.fda.gov/ohrms/dockets/98fr/010489gd.pdf Rockville, MD: US Food and Drug Administration, November 2001.
35. Cowen SG. Pharmaceutical therapeutic categories outlook: comprehensive study. 2004. P. 375.
36. A closer look at the FDA's 50 fastest drug approvals, 1963–2001. In: Parexel's Pharmaceutical R&D Statistical Sourcebook 2002. Waltham, MA: Parexel.
37. Frantz S. 2003 approvals: a year of innovation and upward trends. Nat Rev Drug Discov 2004;3:103–105.
38. US Food and Drug Administration. NCI and FDA announce joint program to streamline cancer drug development. May 30, 2003. Available at: http://www.fda.gov/bbs/topics/NEWS/2003/NEW00912.html. Accessed July 12, 2004.
39. Rothenberg ML, Carbone DP, Johnson DH. Improving the evaluation of new cancer treatments: challenges and opportunities [opinion]. Nat Rev Cancer 2003;3:303–309.
a This chapter focuses on the laws and regulations specific to the United States FDA. Thus, the information presented in this chapter may not apply to analogous regulatory agencies, though the United States Food and Drug Administration is a partner to the International Conference on Harmonization.
b This figure accounts for the cost of developing other drugs that fail as well as the opportunity cost of capital. When postapproval development costs are added, this number increases to $897 million.
c The sponsor and FDA jointly agreed to suspend marketing for gifitinib in December of 2004, following analysis of a Phase III trial that failed to show a survival benefit for gefitinib in non small-cell lung cancer. In June of 2005 gifitinib's label was changed to indicate that it is only to be used in patients who have previously taken gifitinib and are benefitting or have benefitted.