For investigational new drug application (IND) sponsors running clinical trials and debating when a serious adverse event can be linked to an investigational drug and is serious enough to unblind data, FDA offered new draft recommendations on Wednesday.

The new guidance on IND safety reporting follows previous guidance on Safety Reporting Requirements for INDs and BA/BE Studies from 2012 and contains specific details on: The composition and role of a safety assessment committee (and how it differs from a data monitoring committee), how to aggregate analyses to compare adverse event rates across treatment groups, how and when to unblind safety data, reporting thresholds for IND safety reporting and the development of a safety surveillance plan.

Safety Assessment Committee

FDA is calling on sponsors to develop a safety assessment committee to review safety information in a clinical development program and make recommendations to the sponsor on whether any numerical imbalance in the unblinded rates of safety events meets the criteria for IND safety reporting.

No later than 15 calendar days after determining if the committee’s information qualifies for IND safety reporting, sponsors are required to report these potential serious risks.

“The safety assessment committee, possibly together with other parties (e.g., steering committees, data monitoring committees [DMCs]), can also participate in decisions about whether the conduct of the study should be revised (e.g., change in eligibility criteria, revision of informed consent),” FDA says, noting that the roles and responsibilities of both the safety assessment committee and the individuals on the safety assessment committee should be clearly defined.

Safety assessment committees should include at least one physician familiar with the therapeutic area for which the investigational drug is being developed as well as clinicians who have general or specific (e.g., cardiology, hepatology, neurology) safety experience. Other disciplines -- such as epidemiology, clinical pharmacology, toxicology, chemistry and biostatistics -- should be considered on a regular or an ad hoc basis, FDA says.

Aggregate Analyses

As far as sponsors looking into adverse event data, FDA suggests that aggregate analyses should generally be performed across multiple studies under the IND and, as appropriate, across other INDs held by the same sponsor to determine whether the criteria for IND safety reporting have been met.

But sponsors “should not submit IND safety reports for those serious adverse events that were prospectively identified as anticipated to occur in the study population unless the evidence suggests a causal relationship between the drug and the event,” FDA says, acknowledging that that decision requires a complex judgment “that is, in most cases, not a simple application of a planned statistical analysis.”

Unblinding

FDA says it recommends “unblinding to allow a comparison of event rates and detection of numerical imbalances across treatment groups to identify important safety information. The safety assessment committee should regularly perform unblinded comparisons of rates across treatment groups for serious adverse events that are prespecified in the premarket safety surveillance plan as anticipated serious adverse events or as previously recognized serious adverse reactions listed in the protocol or the investigator brochure, as long as appropriate steps to maintain the overall study blinding are taken.”

In addition, all IND safety reports submitted to FDA and participating investigators should be unblended, the agency says, though two distinct cases should be considered.

First, some serious and unexpected adverse events are interpretable as single or small numbers of adverse events and FDA does not believe that unblinding single or small numbers of serious and unexpected suspected adverse event cases will compromise the integrity of the study, “in part because unblinding outside of the safety assessment committee should be infrequent based on the specific criteria that must be met to submit the serious and unexpected suspected adverse reactions in an IND safety report.”

The second case is when the adverse event is interpretable only by examining rates of events in treated and control groups to determine whether a specific serious adverse event is occurring more frequently in the drug treatment group or whether there is a clinically important increase in the rate of a specific previously recognized serious adverse reaction.

And as far as ensuring the integrity of trials remains intact when data is unblinded, “FDA recommends that those participating in the conduct or analysis of the study (e.g., study clinicians, statisticians, chief medical officers, clinical research associates) remain blinded to overall data, although in individual serious adverse event cases, appropriate medical care may require unblinding.”

Thresholds for IND Safety Reporting

For the purposes of IND safety reporting, FDA says, “reasonable possibility” means there is evidence to suggest a causal relationship between the drug and the adverse event.

Factors to consider when making the judgment include:

• The size of the difference in frequency between the test and control groups
• Consistent increase in multiple trials
• Preclinical evidence to support the finding
• Evidence of a dose response
• Plausible mechanism of action
• Known class effect
• Occurrence of other related adverse events (e.g., both strokes and transient ischemic attacks)

Safety Assessment for IND Safety Reporting Guidance for Industry