Access to Experimental Drugs for Terminally Ill Patients

Terminally ill patients for whom conventional medicine offers little benefit are often willing to try unproven therapies. Because these individuals are often ineligible for clinical trials and compassionate use can be procedurally difficult to secure, gaining access to experimental drugs is not a viable option for many patients.¹ Congress and the US Food and Drug Administration (FDA) have attempted for decades to strike a proper balance between access and oversight, with periodic episodes of heightened awareness that include the AIDS epidemic of the 1980s, breast cancer advocacy in the 1990s, and most recently, Abigail Alliance v Von Eschenbach, a case in which the plaintiffs unsuccessfully argued that governmentally imposed restrictions on access violate the Fifth Amendment's due process requirement.²

After 2 appeals, the US Court of Appeals, DC Circuit, held that due process does not include a fundamental right to access experimental compounds, a ruling that allows Congress and the FDA to continue regulating the area as they see fit. Nevertheless, the litigation revitalized the debate surrounding access to investigational drugs and sent a clear message to Congress and the FDA that the current policy needs reform. What might be the appropriate new policy that balances access, patient protection, and scientific integrity? In this Commentary, we outline key themes for policy makers and advocates to consider. The Table summarizes a number of patient and societal concerns regarding access to experimental agents and suggests strategies for overcoming these concerns.

Minimization of harm to terminally ill patients is a primary goal of governmentally imposed restrictions on access. To achieve this goal, policy makers must begin by precisely defining the threshold level of evidence required for an experimental drug to be made available. Abigail Alliance contended that completion of phase 1 testing should be sufficient for patient access. In response to the Abigail Alliance litigation, the FDA issued a notice of proposed rule-making that appears consistent with a phase 1 threshold. The proposed rule explains that “[f]or a patient with an immediately life-threatening condition, the evidentiary burden could be very low—little if any clinical evidence to suggest a potential benefit or possibly only animal data to support safety of the use.”^{3 (p75153)} Others have contended that using phase 1 as a threshold is inappropriate because phase 1 studies are too small to adequately assess safety in a broad population and are not designed to assess effectiveness.⁴

How much evidence is necessary to release investigational drugs to any terminally ill patient? Empirically determining, for each developmental phase, the number of patients who must be treated with an experimental agent for 1 person to benefit (ie, the number needed to treat [NNT]) may be useful. Such analysis might begin by determining, for each phase of development, the probability that an agent will eventually receive FDA approval. Oncology drugs entering phases 1, 2, and 3 subsequently have FDA approval rates of 26%, 34%, and 57%, respectively.⁵ Yet, even among agents that are FDA-approved, only a fraction of patients who receive the agent experience benefit. Hence, to determine phase-specific NNT figures, these data can be combined with figures pertaining to the absolute mortality reduction seen in compounds that reach each stage of development. In essence, better phase-specific data recording and reporting could generate NNT figures that serve as an empirical guidepost in the selection of a cutoff point for access. Such data may be imprecise, but it is preferable to make policy choices based on whatever phase-specific numbers can be generated rather than intuitive assessments of appropriateness.

Once the appropriate threshold is established, policy makers should determine whether additional, centralized review of individual requests for otherwise qualified compounds is necessary. The FDA's proposed rule regarding evaluations of compassionate use petitions (quoted above) is so deferential to treating physicians that, if implemented, case-by-case review by the FDA would add little to the goal of patient protection.³ Moreover, unfettered and misguided disbursement of inappropriately injurious compounds will always be constrained, however imperfectly, by physicians' concerns about their reputations and the omnipresent specter of tort law. The FDA could, however, continue to regulate 2 crucial areas. First, it might be charged with restricting access to individuals who possess an “immediately life-threatening condition.” Second, the FDA could assess the potential impairment of clinical trial enrollment. Transparency of the decision-making procedure can be used to prevent de facto review of safety and efficacy.

There is general agreement that access to experimental drugs can only be granted if clinical trial enrollment is unimpaired.⁶ Two steps are needed to ensure that access does not impede trial enrollment. First, authorities must deny access to experimental drugs for patients who are eligible for clinical trials. Second, individuals should be adequately deterred from gaming the system by, for instance, initiating therapy with an alternative compound that renders them ineligible for a study. Congress can further ensure the generation of important, new knowledge by requiring patients who obtain access to investigational drugs to enroll in a registry that would allow for the collection of basic data on outcomes and toxicities. Care must be taken, however, to avoid misattribution of all adverse outcomes to the investigational compounds.

Manufacturers currently have limited incentives to participate in compassionate use programs. In addition to concerns about litigation, production barriers for complex agents, and the inability to profit from investigational drugs, companies rightfully are concerned that experimental agents will be used by only the sickest patients, and the disproportionate number of poor outcomes will be unfairly attributed to the drug. Abigail Alliance expressed an interest in loosening the regulations that prevent manufacturers from earning a profit prior to formal FDA approval. The FDA, however, has 2 principal interests in limiting profits on investigative drugs. First, it must encourage companies to conduct additional trials and seek approval to promote the full development of therapeutic options for future patients. Second, the FDA seeks to prevent exploitation of inequitable bargaining power—ie, inordinately high prices that take advantage of desperate patients.

With regard to the FDA's first concern, the interests of regulators and manufacturers would both be served if the profits of investigational drugs are released when the compound in question is approved by the FDA. Under such an approach, the incremental cost of producing the pharmaceutical would be immediately available to the company (as is true today), while any surplus amounts would remain in an interest-bearing escrow account until approval is granted. If FDA approval is not obtained within a specified time frame, then the proceeds would be reallocated to health-related government use.

To prevent companies from charging exploitative prices, manufacturers might be prohibited from charging more than what will ultimately be the average sales price of the drug postapproval. Strict penalties would be levied against firms that generate an average sales price that is less than the amount paid by any terminally ill patient for the compound when it was still in development. For drugs that must be administered in the physician's office or hospital setting, health care professionals would be prohibited from marking up their fees for administration-related procedures beyond what is typically billed for an FDA-approved compound or earning more than a defined spread on the pharmaceuticals.

Establishing an efficient and fair process by which all patients have access to a clinician who can help them navigate the system presents another challenge. Knowledge of FDA policies about the use of investigative treatments has historically been concentrated among physicians in academic medical centers.³ Ensuring equal access for all patients could be improved by providing modest governmental reimbursement for the submission of investigational treatment applications. Firms that demonstrate the knowledge and ability to pair patients with the right physicians and help them navigate the process would apply for certification and thereby be eligible for reimbursement. Funding could come from the proceeds of experimental compounds that fail to garner FDA approval. This model has similarities to Medicaid reimbursement for case management and Medicare's demonstration projects on care coordination, in which governmental money is used to pay administrators who help patients navigate a complex patchwork of care.

Such reimbursement, however, would be minimal or perhaps unnecessary if manufacturers are permitted to charge prices for experimental drugs that are in line with postapproval prices. Under such circumstances, industry should find it profitable to connect unsolicited drug candidates with the proper support teams. Additionally, the smaller the role of the FDA in granting individualized access to experimental compounds, the easier it will be to provide an efficient and fair application process.

The final consideration is the ethical distribution of resources. By definition, compassionate use programs involve experimental treatments rather than medically necessary interventions. Thus, employers and health plans using standard contract language should not be expected to pay for investigational drugs. Even if manufacturers do not calibrate prices according to patients' ability to pay, ethical concerns are minimal. Principles of equity do not require providing the same level of health care to all individuals.⁷ Given that 66% of oncology drugs that successfully complete phase 1 studies fail to receive FDA approval,⁵ it would be difficult to deny that health care dollars are better spent elsewhere.

This proposed framework for efficient and equitable access to experimental compounds promotes patient safety, ensures the generation of reliable scientific data, increases manufacturers' incentives to participate, and prevents exploitation of vulnerable patients. The FDA's proposed deference to patients' requests, coupled with physicians' concerns about their reputations and tort law, suggest that governmentally administered, case-by-case review of safety and efficacy is unnecessary. FDA review might be limited to determinations of whether a patient has an immediately life-threatening condition and whether access to a compound would adversely affect clinical trial enrollment. Manufacturers could be given an incentive to make experimental drugs available by conditioning profit disbursement on FDA approval and levying penalties for postapproval prices that are lower than preapproval prices. Finally, widespread geographic access to the program can be ensured by restricting the role of the FDA, allowing manufacturers to offer administrative support, and, if necessary, providing modest reimbursement for firms that help patients navigate the system. Given terminally ill patients' enormous interest in rapidly obtaining investigational drugs that offer even minute chances of improved longevity or an increased quality of life, it is imperative that society create a system of access that is fair and efficient and balances the interests of patients, manufacturers, and society.