Compassionate use of investigational drugs requires IRB approval, regs say

The question remains: Is full-panel review necessary?

Investigators and clinicians occasionally see patients who might benefit from use of an investigational drug, but who are unable to participate in existing clinical trials.

In some cases, sponsors allow expanded treatment distribution of the study drug to needy patients who have exhausted other treatment options and cannot wait until the completion of Phase III study and approval by the U.S. Food and Drug Administration (FDA).

Although these protocols often are not viewed as research, federal regulators still consider these patients to be research subjects and such distribution still must be reviewed by local IRBs, say federal officials and others familiar with the process.

"The terms compassionate use’ and expanded access’ do not appear anywhere in the regulations," warns Joe Breault, MD, ScD, MPH, MS, FAAFP, CIM, CIP, chair of the IRB at the Ochsner Clinic Foundation in New Orleans. "The FDA and NIH [National Institutes of Health] may publish guidance documents advising IRBs how to proceed, but these are not official parts of the regulation. Under the law, these protocols are research protocols like any other."

Breault has become familiar with the process because several investigators at his institution want to use the investigational drug 5-Azacytidine (5AZ), which is being tested to treat refractory acute myelogenous leukemia and acute myelodysplastic syndrome.

There are two basic categories of treatment use distribution:

  1. Expanded access — includes the National Cancer Institute’s (NCI) Treatment Referral Center (TRC) protocols. One treatment protocol is to be used for all patients with the same diagnosis. Much like a standard clinical trial, the IRB reviews the protocol and the informed consent documents, etc., then eligible patients are enrolled to have access to the drug.
  2. Single treatment — includes compassionate use, emergency exception protocols, also known as compassionate IND (investigative new drug) programs, under FDA policies. Clinicians seeking to use the investigational drug for a patient unable to participate in a clinical trial must submit a separate application for each patient to be treated. The IRB must separately approve each patient as a separate study, examining the protocol, the informed consent documentation, etc.

Decisions about whether to make investigational drugs open to nonresearch distribution are up to the sponsors of the drug’s clinical trials — often the pharmaceutical company or a federal entity, such as the NIH or NCI.

The sponsors must obtain permission from the FDA for treatment distribution, and the type of distribution depends on the approval obtained.

For example, NCI has approval from the FDA for three types of expanded distribution: special exception protocols, Group C treatment protocols, and TRC protocols, says Ellen Fiegal, MD, acting director of the division of cancer treatment and diagnosis at the NCI.

Each type of distribution requires a different level of record keeping and reporting by investigators and different distribution mechanisms for the study drug.

NCI special exception protocols are open to any qualified investigator on a patient-by-patient basis, with the study protocol consisting of a two-page form to be filled out by the investigator. FDA provides retrospective approval after the protocol is cleared by NCI.

Group C treatment protocols are also open to any NCI-approved investigator, but come with a standard guideline for treatment, with a model informed consent document. According to the NCI, it may request the IRB requirement be waived for drugs designated eligible for Group C protocols.

For TRC protocols, the study drug is provided to designated treatment referral centers, along with a standard protocol and informed consent. The IRB at each center must approve the protocol. The TRCs then enroll patients.

However, Breault notes, even if an expanded-access TRC protocol is in place for a particular investigational drug, a single-treatment exception may be necessary for some patients.

For example, a drug may be eligible for expanded access distribution for one diagnosis, but the same drug may require single-treatment protocols for use in patients with another diagnosis.

Regardless of the category of treatment use, all applications for investigational treatment require an investigator, informed consent, a sponsor who accepts responsibility for the study and communicates with the FDA, a drug supplier (who may also be the sponsor), and oversight by an IRB, he adds.

For NCI-sponsored protocols, the study drugs must be distributed to, and administered by, an investigator registered with the institute, with the exception of TRC protocols, in which clinicians at the preapproved centers all have access to the drugs, says Fiegal.

After approval of the special exception request, investigators must submit a brief protocol for each patient that describes the treatment plan, treatment history, expected toxicities, efficacy, and monitoring procedures.

"Investigators must obtain IRB approval before treating the patient and retain documentation of this approval in the patient’s medical record," Fiegal notes.

Of course, informed consent still applies. The investigator must provide enough information to the patient, family or guardian to make an informed decision. This is documented in a written informed consent that must be signed by the patient or guardian before treatment and kept in the patient’s medical record. The informed consent should require all items required by the FDA in clinical trial informed consent documents, she says.

"On completing therapy with the special exception agent, the investigator must also provide the NCI with a report of the treatment experience [called the Report of the Independent Investigator], which describes toxicity and efficacy," she says.

Question about IRB review

For special exception drugs, IRBs often are faced with considering identical applications for treatment protocols for different patients with the same diagnosis, says Breault.

The rules stipulate that use of the drug must be separately reviewed for each patient, but some IRBs have questioned whether this merits a full-panel review in each case, or whether it might be permissible to have the full IRB review the initial protocol, with the IRB chair reviewing subsequent applications for treatment use of the same drug by the same investigator.

For Group C treatment protocols that it sponsors, the NCI permits three categories of IRB review for special exception protocols:

  • An entirely new study approved by the full IRB for each new patient.
  • Expedited review of the protocol by a few IRB members or the IRB chair.
  • Full-panel review of the protocol, with the IRB chair sending a letter of approval for each new patient enrolled in the protocol.

However, though expedited review may be permissible by the NCI or other study sponsors, it violates FDA IRB regulations prohibiting expedited review of research that poses more than minimal risk to subjects, Breault says.

Using a strict interpretation of the federal regulations, only the first option should be considered. This can be extremely burdensome, particularly for institutions that may have 50-60 patients on a compassionate use protocol, he adds.

In talks with FDA officials, some have indicated that the third option is acceptable, because it involves full-panel review of the protocol, then the IRB chair approving adding patients subject to the FDA’s final approval.

There is nothing in writing, though, he says, and no guarantee that such a practice would not draw fire during an IRB audit, Breault notes.

For more information

  • U.S. Food and Drug Administration. Briefing Document for the ODAC (Oncologic Drugs Advisory Committee) Single Patient Use of Investigational Anticancer Agents. Available on the FDA web site:
  • National Cancer Institute. Access to Investigational Drugs: Questions and Answers. Published June 6, 2002. Accessed on the NCI web site at:
  • Shalabi AM, High J, Edwards MS, et al. Obtaining investigational agents from the National Cancer Institute: When clinical trials are not an option. Oncol Pract 2000; 18: 8-14.