Discussion Points
A complex system: FDA regulations are complex and include lengthy interpretations. Resubmission of applications by investigators are common at many stages in the process. Unnecessary resubmissions and reviews can add years of delay for each drug. Delays cost lives and contribute to patient suffering. Delays also add to costs. A delay of one month can cost many millions of dollars. More importantly, the prospect of delays increase the risks of investing in cancer drug research and discourages creative initiatives.
FDA drug approval standards may be too stringent: While progress has been made with the passage of PDUFA and FDAMA, dramatic examples of success exist (see AIDS example, above), it appears to the patient community that the FDA will tend to error on the side of caution, perhaps because the agency is more harmed politically by approving harmful drugs (i.e., thalidomide) than by failing to approve beneficial drugs (i.e., Idiotype vaccines).
Bureaucratic delays: The FDA is a large institution with a broad range of responsibilities. They may be under-funded and require reorganization to adjust to the increased demands caused by explosion of information about cancer and new therapeutic targets. Importantly, the important work of education and proactive guidance for drug sponsors is not covered by user fees that fund the bulk of FDA activity.
High costs makes investment risky: It costs over $500-800 million to develop a single drug for commercial availability in the U.S. Many promising therapeutics are not evaluated because potential sponsors deem it too risky to fund the testing process.
New therapies have poor odds for success: Only one of every 50,000 drugs initially screened actually becomes an approved medicine. Some of these compounds may be quite useful either alone or in combination with other agents.
Current evaluation methods take a long time: It now takes from 10 to 15 years to test drugs before they can become available to patients. However, it takes as little as one year to determine the safety of a drug. The majority of time and capital is spent on the assessment of clinical benefit.
Shortages of capital and trained people: There is a shortage of adequately trained clinical investigators and research nurses. Money is needed to fund studies and attract talent.
More cancer subtypes will not help: New diagnostic techniques will define subcategories of cancers, which will reduce the pool of eligible patients and thus make accrual in trials more difficult. Current methods of evaluating efficacy requires the evaluation of large numbers of patients for long periods of time.
There are more ideas than resources: An explosion of new and highly specific therapeutics are emerging as a result of insights about cancer and cancer-host interactions. This will place an enormous burden on the FDA to evaluate these agents in a timely manner, and require creative solutions to evaluating the safety and efficacy of new treatments.
Transparency in the drug evaluation and application process: The law that requires non-disclosure of agency decisions should be amended. The policy is in place, we're told, to protect trade secrets. But, we have doubts that publicly identifying reasons for delays or denials, such as problems with toxicity, manufacturing, or clinical outcomes, really compromises trade secrets. Greater transparency would enable advocates and drug sponsors to make informed contributions to the process; drug sponsors will be able to learn what succeeds, what fails, and why. For example: If delays are caused by a sponsor's inability to manufacture consistent product, is it productive for cancer pts to blame the FDA for foot dragging? Conversely, if a delay is from questionable agency policy, how can we redress that policy unless we see it applied? Finally, we will all gain by learning from the mistakes made in the process, such as what constitutes a flawed trial design, or how the FDA defines substantial evidence of effectiveness in particular cases.
Commentary on Transparency:
-
Statement by Jane E. Henney, M.D. Commissioner of Food and Drugs Food and Drug Administration before the Committee on Health, Education, Labor and Pensions United States Senate October 21, 1999 - FDA
Questions about statistics and biomarkers: Is the FDA providing guidance about how biomarkers might be validated as future endpoints to show clinical benefit? Have drug sponsors or the NCI asked the FDA for such guidance or made proposals? Who is most likely to conduct such studies (NCI/Drug sponsors) What would such a study entail, and do you see signs that they have begun or are likely to begin soon?
Question about statistics and historical controls: How many pts would need to be enrolled in a non-randomized trial in order for the time to progression outcome data to have statistical significance -- assuming a good deal of retrospective data exists for standard treatment?
Question about statistics and drug safety profile: Is the relative safety of a new drug and properties related to drug resistance included into statistical assessments of clinical benefit?
Question about statistics and biomarkers measured in a given study: What if biomarkers correlate with effectiveness in a given study. Can that correlation be used to support the finding of clinical benefit? If such a correlation is established, can sponsors who produce new agents that use the same mechanism of action use that marker as evidence of clinical benefit?
Question about statistics and chance: We understand that measures of response can be misleading in a study. That said, is it true that for some cancers, placebo or chance cannot account for verifiable responses to treatment in a study? If true, do statistical methods incorporate this perspective, or do all studies of all drugs for all conditions use the same assumptions about chance and placebo effects? Can a new drug that has less but significant clinical benefit --compared to standard treatments -- still readily win approval if it's toxicity is less, or that benefit applies to some patients refractory to all standard treatments?
PROPOSALS: (Under Construction)
To expand the definition of clinical benefit to include tumor stabilization, improvements quality of life, modest activity if the agent has reduced or transient toxicity and/or the new agent does not impair the pts ability to benefit from other treatments. Tumor regression, delayed time to progression, tumor stabilization, immune responses, can all be valid endpoints, especially when therapies are less toxic.
To identify and validate surrogate biomarkers that infer clinical benefit. This may aid in the identification of safer drugs that may not kill tumors but, instead, prevent tumor growth. Candidate Biomarkers include: LDH, gelectin-3, BCL-2, VEGF, bFGF, copper/zinc ratio, or some combinations of these markers.
Acknowledge the limitations of single-agent studies. A one-punch prize fighter will not win many battles. Cancer cells are adaptable and therefore it is unlikely that any single agent will improve survival.
To rapidly approve safer therapies even if they show modest activity, such as endogenous compounds like endostatin, and autologous vaccines so they may be used by patients in need and more rapidly combined with other agents.
To increase patient participation: Seek informed patients to participate in the design of clinical trials and sensitize investigators to the needs of patients when designing studies. Simplify consent forms and case report forms. Provide objective education about the merits of new trials. Require insurance providers to pay for related tests. Provide travel and lodging support, and require government reimbursements for associated tests.
To not attempt to prove survival benefit for indolent cancers. While proving survival benefit is important, it's also very difficult to prove when treating indolent cancers in which life expectancy can range from five to twenty years. Patients who progress while participating in a study will cross over -- use numerous treatments that will confound the survival assessment. Patients may avoid participating in studies that do not offer a crossover provision in some cases.
To make the drug evaluation system more transparent. More individuals will be willing to invest in companies developing cancer drugs when they know what the board of directors knows and at the same time. Patients will have more confidence in the clinical trial system when they have access to real outcome and side effect data instead of press releases that may contain biases.
To apply drug evaluation policy in ways that protect incentives to develop new drugs: When two similar drugs are in the Accelerated Approval path, the approval of the first should not adversely affect the approval of the second as this policy will reduce incentives to develop new agents.
To support novel statistical study designs and procedures to improve the efficiency of clinical trials of novel agents. This will reduce the risks of investing in cancer drug research and encourage more creative initiatives.
To specify early stopping criteria to terminate studies of toxic or ineffective agents. This will direct funds and talent to more productive areas.
To support advocacy efforts with Medicare and third-party medical insurance providers to ensure patient access to participation in Phase I, II and III clinical trials of innovative cancer therapies. This will save insurance providers money in the long run as better and safer therapies will reduce costs. It must be cheaper to treat patients right the first time, and not have to treat the side effects of toxic therapies, for example.
To encourage development of new methods to determine optimal dose: "We need new rules for this age of molecular therapeutics, Druker suggested. The most efficient method for setting trial doses may be to base ranges on insights gleaned from molecular biology rather than to base doses on older procedures developed when less information was available."
To learn from our successes and test therapies on untreated patients: Three factors were important for the success with STI571, said Druker. First, the disease involved a discrete molecular abnormality. Second, clinicians have a reliable and straightforward method of identifying eligible patients (STI571 might not work on the 5% of CML patients that don't carry the Philadelphia chromosome). Third, treatment is possible in the early phase of the disease, before secondary molecular defects have accumulated.
Utilize Fast-track (Accelerated Approval): This FDA provision can expedite the review of drugs and biologics intended to treat serious or life-threatening conditions and that demonstrate the potential to address unmet medical needs. One "unmet need" of low grade lymphoma patients is the availability of novel treatments that controls or regresses the disease while minimizing the following:
- Damage to normal cells and organs, including immune function. It is not uncommon for patients to succumb to infection that results from depressed immune function -- a side effect of the majority of standard treatments for cancer.
- Drug resistance. A drug with even limited efficacy can be extremely valuable when it does not preclude the patient from benefiting from subsequent treatments. Chemotherapy drugs often regress tumors, but they are not curative and can contribute to producing surviving cells that are more aggressive and resistant to future treatments.
|