Share this article on:

The Ins & Outs of Compassionate Use Programs

Fromer, Margot J.

doi: 10.1097/

First, let's get the terminology right: “compassionate use” is neither a legal nor regulatory term, although it is often misused by both the general public and physicians. What we are discussing here is a clinical trial for a single patient of an investigational cancer drug. It is called a single-patient IND (Investigational New Drug Application), or special exception use.

As Robert J. Temple, MD, Associate Director of the FDA Center for Drug Evaluation and Research (CDER), said in testimony before the House Committee on Government Reform:

“Compassion, an intent to help, should be, and is, an element of all drug investigation. We describe this as treatment uses because their intent is to provide treatment of patients, not primarily to evaluate the safety and effectiveness of drugs, which is the primary and usual purpose of studies under an IND.”

What Dr. Temple described was using an investigational drug to treat an individual patient who is not part of a clinical trial. This process is highly controversial, but even worse than the heated arguments that surround it are the lack of information about what it is, when it can be put into effect, and to whom it can be provided.

Single-patient IND, or treatment use of an investigational drug, is what it sounds like: a one-person clinical trial conducted to provide treatment (not to test safety and efficacy) for a single individual.

It can be, but rarely is, initiated by a commercial sponsor (the developer and manufacturer), by a physician who must then act as the principal investigator (PI) with all that entails, or a patient who must then find a physician who will agree to act as PI.

In general, the more patients in a clinical trial, the better and more useful the results. Thus, patients who might benefit from a drug under investigation should be strongly encouraged to enter a trial.

There are, however, situations in which it may be appropriate to make a particular investigational drug available to treat rather than to evaluate. These are almost always limited to the lack of a therapeutic alternative for patients with a serious or life-threatening disease.

Richard Pazdur, MD, Director of the FDA Office of Oncology Drug Products, explained that clinical trials and single-patient INDs are entirely different things. “The first asks a scientific question: Does the drug in question work, and is it relatively safe?

“Determining efficacy and treatment in a significant number of patients is the purpose of a clinical trial. The purpose of a single-patient IND is to treat one person who, for a variety of reasons, may not be eligible for a trial.”

A single-patient IND, even if there are several for a particular drug, ought not to be directed toward subverting the clinical trial process, he added.

“There should be cogent and rational reasons an individual patient gains access to an investigational drug, and it's not easy to do. The sponsor has to agree to provide the drug—and can stop at any time—the physician has to be willing to act as principal investigator, with all the paperwork that entails, and the protocol has to be approved by an institutional review board, the same as a regular trial. All this is significantly daunting.”

The FDA may not allow a single-patient IND because of safety concerns, said Dr. Temple. But if a physician makes a request for a patient for whom no effective therapy exists, if the drug is under active study and the sponsor agrees to provide it, and if there are no known significant adverse effects, the FDA will not object, he said.

Back to Top | Article Outline

Two Ways to Get Drugs Outside of Clinical Trials

Treatment use of experimental drugs falls into two broad categories: expanded access and single patient.



But regardless of the way the drug is obtained, the same regulations as those for formal clinical trials apply: There must be an investigator; informed consent; oversight by an institutional review board (IRB); and a sponsor who accepts responsibility for the protocol, communicates with the FDA, and supplies the drug.

Expanded-access protocols are treatment regimens that apply to a group of patients with similar characteristics—either Treatment INDs or Group C.

In 1987, as a result of lobbying by AIDS-activist groups, the FDA formalized the Treatment IND process to provide drugs to patients who are seriously ill and have no satisfactory alternative. The drug in question must be nearing the end of its development, and data from clinical trials must conclude that it is effective for the intended indication.

A Group C drug has reproducible efficacy in one or more tumor types. If a Group C drug is to be used in an expanded-access protocol, the FDA requires that the physician acting as PI be able to administer the drug safely without specialized supportive care facilities.

Single-patient treatment can occur in two ways.

First, the FDA can grant an exception when a patient is ineligible for a clinical trial. The sponsor provides the drug and reports to the Agency as if it were a real clinical trial.

Second, if the sponsor is unwilling to assume this responsibility, a physician may assume the role of sponsor but must obtain the drug from the manufacturer (which does not have to provide it) and apply directly to FDA for an IND.

Since information contained in an IND is confidential and proprietary, FDA cannot reveal most of it to a physician-investigator, which can hamper the process.

Back to Top | Article Outline

FDA Not Speaking with One Voice

In the interview with OT, Dr. Pazdur said he wanted to be very clear about the difference between access to and approval of a drug:

“We need to have a great deal of information about a cancer drug before it can be approved, especially about its toxicity, some of which is life-threatening,” he said. “These drugs are essentially poisons, and when we approve them they will affect thousands of people.

“It's one thing for an individual patient to say that he or she is willing to risk serious toxicity if there is no other treatment available, but it's quite another for FDA to let many people have access to the drug before approval. We just can't do that.

“People need to be rational about what they do in this regard,” Dr. Pazdur continued. “They need to understand how—and how well—a drug works, which is considerably more difficult than the willingness to risk side effects. They must know how to make good decisions about risks and benefits. The side effects of these drugs are not minor.

“Every drug is a new experience, every tumor is different, every host is different. It's FDA's job to know all this about every drug we approve. Unfortunately, the public doesn't understand this process. All they realize is that a drug exists that may be able to help them—and they want access to it. We are not in a position to help every single patient with every single dread disease. Of course, it's not fair, but it's the law.”

Terry Toigo, Director of the FDA Office of Special Health Issues, and Patricia Delaney, Director of Cancer Liaison Programs in that office, say they try to take a softer, more patient-centered approach and do everything they can to help patients get access to investigational drugs.

“First of all, it depends what we know about the drug,” Ms. Toigo said. Added Ms. Delaney: “If the patient doesn't meet the criteria for a clinical trial, we suggest that they look into either a single-patient IND or an expanded-access program. We explain the procedure, which is indeed onerous.”

Back to Top | Article Outline

Point of View?

Does the Office of Special Health Issues have a point of view about who should or should not be given access to investigational drugs?

“There's no general point of view,” Ms. Toigo said. “Each case is viewed individually. We help people find the right sources at the drug company, which is not easy, and we tell them if a drug has an expanded-access program. We also tell them how to find appropriate clinical trials in case they have not done that.”

Back to Top | Article Outline

Legal & Regulatory Issues

Prior to 1997 and the FDA Modernization Act (FDAMA), use of expanded-access and single-patient INDs was a hodgepodge of inconsistent criteria. FDAMA has clarified them somewhat, but the FDA doesn't have to grant the request—and it will not do so unless the following conditions are met:

  • ▪ The patient has no comparable alternative therapy.
  • ▪ There is evidence that the drug is safe and effective. Even for a life-threatening disease, there must be no evidence that the drug would expose a patient to serious additional risk of illness or injury.
  • ▪ Provision of the drug to a single patient will not interfere with an ongoing clinical trial. If too many people demand single use, trial enrollment will suffer.
  • ▪ The sponsor and/or clinical investigator submits FDA-required information (with an individual physician willing to act as principal investigator), and the sponsor has a sufficient supply of the drug and is willing to provide it.

These conditions are still fairly general and often engender more questions: How much evidence of anti-tumor activity is enough to support single-patient use? What is an acceptable ratio of benefit to toxicity? How strongly should the effectiveness of standard therapy be weighed against the investigational drug?

Too, it is possible that a patient could be indirectly harmed, even if the investigational treatment by itself does not cause injury, if it is used instead of a proven curative therapy, and the investigational therapy doesn't work.

Therefore, as long as there is a curative treatment, the FDA considers it unethical and unsafe to approve single-patient treatment.

Said Dr. Temple, “We must help ensure that the use of investigational drugs is carried out safely and that the known limitations of the drug are conveyed to patients.”

The FDA cannot compel a sponsor to supply an individual patient with an investigational drug, he said. The decision is solely the sponsor's, and patients may not be aware of this and can be confused and angry when a company refuses to provide it, and mistakenly may then blame the FDA.

Back to Top | Article Outline

‘Right’ to Investigational Treatment

It is not uncommon to hear cancer patients and their families claim that they have a right to investigational treatment. They feel that decisions about treatment are theirs alone and if a drug exists, not matter how new or unproven, they ought to be able to take it. But the FDA is legally barred from distributing investigational agents except in the very narrow circumstances noted.

“I don't believe cancer patients have a right to receive investigational drugs outside clinical trials,” said Ellen Stovall, President and CEO of the National Coalition for Cancer Survivorship (NCCS). “But I do think they have a right to every evidence-based treatment available.



“NCCS believes that public policy should reflect that,” she continued. “That is, if a patient is not eligible for a clinical trial or if traveling to one is impossible, then an expanded-access program might be a good solution.”

Ms. Stovall, a member of OT's Editorial Board, was critical of the way clinical trials are designed, in that eligibility requirements are often so stringent that patients who need a drug, often desperately, cannot have it.

“We need to make clinical trials smarter and more practical,” she said, “to open them up to more patients, especially those with refractory disease.”

When cancer patients question why FDA insinuates itself between them and their unfettered access to investigational drugs, their sense of disappointment is understandable, even if it is born of lack of understanding of the regulatory process.

“We believe that the independent scientific consideration provided by the agency is critical and is an essential component of patient protection in considering drugs about which relatively little is known and potential toxicity may be great,” Dr. Temple said. “Typically, physicians have very limited information about the investigational therapy being requested.”

And patients typically know even less.

When patients have exhausted standard therapy, they often crave a “last-ditch” effort, but unfortunately, this is almost never appropriate. Not only is the new therapy unlikely to be more effective in very advanced disease than standard treatment, but the patient's performance status is usually too poor to sustain the effort.

“It all comes down to designing better trials,” Ms. Stovall said. ”Patients wouldn't be clamoring for drugs if more of them were eligible for trials. We need to level the playing field.”

She added that if medical professionals did a better job of discussing end-of-life issues with patients and families, and if they provided access to good support in the form of palliative and hospice care, the desperate insistence on last-gasp drugs would likely decrease.

In the 1980s, when AIDS drugs were first being tested, patients felt they ought to be given new drugs, regardless of how untested and how risky. Many of these HIV patients knew that trials are randomized and didn't want to risk being on the control arm. Cancer patients today feel the same way, and the ethics of giving them a known drug when trial participants take their chances with randomization remains as murky as ever.

When asked if drug companies are obligated to provide investigational drugs to individual cancer patients, Ms. Stovall said there was no good answer. Patients don't have a right to what still belongs solely to the sponsor, but the clinical trial system is so flawed that manufacturers ought to do more to relieve a critical need.

Back to Top | Article Outline

Mixed-Up System

Many manufacturers do not have a standard policy or procedure for handling requests for single-patient INDs, especially when the drug is in early stages of development. This creates confusion and has led to a system in which some people get what they want and others don't.

Sponsors are not the bad guys, however. They have legitimate concerns about single-patient INDs:

  • ▪ There may be a limited supply of the drug during development. Making larger batches is expensive and not reasonable until there is good evidence that it is marketable.
  • ▪ Sponsors often have trouble filling clinical trials and don't want to jeopardize their chances for full accrual.
  • ▪ If a single-patient IND is conducted with a seriously ill patient with advanced disease (as is usually the case), adverse reactions might raise difficult-to-resolve questions about the drug's safety.

Very little is learned about a drug from a single-patient IND. The response rate in patients who have received multiple courses of therapy with a variety of treatments is low, and whatever data are derived cannot be used as part of any study performed in support of FDA approval.

Back to Top | Article Outline

Bristol-Myers Squibb

But there are companies that do their best to provide investigational drugs for cancer patients outside of formal clinical trials. Renzo Canetta, MD, Vice President of Oncology Global Clinical Research at Bristol-Myers Squibb (BMS), described some of what his company has been doing.

From 1985 to 1989, BMS gave paraplatin (Carboplatin) to 804 people, about two thirds of whom had ovarian cancer, as single-patient INDs. “These were patients who could not take cisplatin and who were not candidates for a regular trial,” Dr. Canetta said.

“We began the program when we had compiled a great deal of clinical data (in the United States and Europe), and so we felt comfortable releasing doses of the drug to the physicians of these 804 patients. At the end of the program, which we called ‘compassionate IND,’ we collected the data and published them.”

In 1991, BMS established treatment referral centers selected by NCI. They had the ability to handle drug toxicity or emergencies engendered by allergic reactions.

“There was a significant shortage, so between 1991 and 1992, we provided paclitaxel [Taxol] to 1,819 patients with refractory ovarian cancer,“ he said. “Because there were so many patients and because they were treated at the referral centers, we were able to collect data and publish the results in both Lancet and the Journal of Clinical Oncology.”

Between 1989 and 1990, AZT was the only approved drug for AIDS, but thousands of patients could not tolerate it. “There was a heavy demand for something else, so we felt we had to give didanosine or ddi [Videx] to 7,806 patients,” Dr. Canetta said.



“We wanted to do the right thing. There was a tremendous unmet need, and we felt an obligation.”

In 2003 and 2004, BMS ran a large expanded-access program in which 750 patients not in clinical trials received cetuximab (Erbitux). Even after that drug was approved and on the market, those patients continued to receive it from the company without charge.

Dr. Canetta said that in general, the company has no objection to giving investigational drugs to patients in expanded-access programs, as long as there are sufficient data about safety and efficacy. But the company rarely will provide a drug to one patient at a time—“Even if we receive requests from high places in Washington.”

He said that it isn't “honest” to simply give away drugs. “We need to have confidence in the data, not only to protect the company, but also to protect the patients,” Dr. Canetta said. “But whenever we receive a request for a single-patient IND—even when the request is obviously not medically justified, we always discuss the pros and cons among the oncologists on our staff and with the patient's own physician. Our first reaction, of course, is to direct the patient to a clinical trial.”

Back to Top | Article Outline


That sentiment was echoed by Mary Lynn Carver, Director of Brand Communications for Oncology at AstraZeneca, who described that company's policy on single-patient INDs and expanded access in the broadest possible terms:

Patients are encouraged to enroll in clinical trials, she said, but in certain circumstances it may be appropriate for an investigational drug to be made available through the company's compassionate-use program. However, this is done only “for patients for whom no other therapy is available and who are unable to participate in a clinical trial, or when there is sufficient clinical trial data to evaluate safety and efficacy,“ she said. This is usually not until phase 2 trials are complete and the decision has been made to go on to phase 3.”

If patients cannot enroll in a clinical trial, AstraZeneca will provide “compassionate use” via an expanded- access program, Ms. Carver said. But the company's policy on expanded access, however, is very nonspecific. “A procedure would be customized to meet the needs of the disease state and drug,” she related.

When asked what happens when a patient calls the company and asks for a single-patient IND, Ms. Carver said that he or she is referred to a clinical trial, but if that doesn't work out, “We would determine if it would be appropriate to set up a protocol to include other patients in a similar situation, rather than grant a single-patient IND.”

Novartis, Genentech, and GlaxoSmithKline did not respond to requests to talk about their policies.

© 2005 Lippincott Williams & Wilkins, Inc.
Home  Clinical Resource Center
Current Issue       Search OT
Archives Get OT Enews
Blogs Email us!