Final Guidance On Non-Inferiority Studies Lacks Industry-Sought Changes

One of industry's central concerns about the US FDA’s 2010 draft guidance was a "vaguely" and "subjectively" defined M2 clinical margin, but FDA did not provide a new definition in the finalized version.

Despite industry's concerns about a "vaguely" defined M2 clinical margin in a 2010 draft guidance on non-inferiority, FDA has stood pat in its wording.

Several major industry players, including the Pharmaceutical Research and Manufacturers of America, AstraZeneca PLCand Pfizer Inc. commented that the meaning of the M2 margin – defined in the draft guidance as the largest clinically acceptable difference (degree of inferiority) of the test drug compared to the active control – is both unclear and subjective. Nevertheless, FDA maintained the exact same definition in the final guidance released Nov. 7.

According to both the draft and final guidances, it is generally desirable for sponsors to choose M2, a smaller value, for the non-inferiority margin over M1. M1 is defined as "the entire effect of the active control assumed to be present in the NI study."

"Showing non-inferiority to M1 would provide assurance that the test drug had an effect greater than zero, but in many cases that would not be sufficient to conclude that the test drug had a clinically acceptable effect," the final guidance states, with marginally different wording from the draft. "Recall that the main reason an NI study is conducted is that it is not ethical to include a placebo arm. The active control has a beneficial effect, and denying that benefit to subjects with a serious illness is not ethical."

"It is therefore usual in NI studies to choose a smaller margin (M2) that reflects the largest loss of effect that would be clinically acceptable. This can be described as an absolute difference in effect (typical of antibiotic trials) or as a fraction of the risk reduction provided by the control (typical in cardiovascular outcome trials)."

AstraZeneca offered a particularly scathing assessment of the M2 definition in its comments on the draft guidance, describing it as "hopelessly subjective."

"The requirement to use a margin will lead to inconsistencies that may prevent truly superior new drugs (or truly effective drugs with improved safety) to be approved," the British drug giant wrote.

PhRMA echoed these comments, noting that the precise role of M2 is unclear, as has why it has been introduced, and how it might affect regulatory product approval.

"While there are circumstances in which clinicians and statisticians have developed a sense of what constitutes a 'clinically acceptable difference,' these notions are inherently imprecise and subjective," the drug lobby stated.

Pfizer in its comments on the draft called on FDA to work both internally and with sponsors to develop "a mutually acceptable means to utilize clinical judgement in the choice of M2," noting that "appropriate input from clinicians and careful consideration of this input in an open forum with open-minded consideration from the agency and sponsors would result in a means for developing mutually acceptable M2 values."

The draft guidance noted that M2 is a matter of "clinical judgement," and Pfizer maintained that such a term should not be dismissed in the final guidance. Clinical judgement, Pfizer says, "is core to medical practice and its use should be practiced in the development of M2 values and NI trials." FDA retained the term in the final guidance.

The Biotechnology Innovation Organization also concurred that the exact definition of M2 needs to be clarified.

FDA, however, did not change the definition of M2 in the final guidance. FDA spokesman Kris Baumgartner tells the Pink Sheet that the agency "considered all comments received to the public dockets and incorporated appropriate changes in the current final guidance."

Another element of the final guidance was that it supersedes the it supersedes a separate 2010 guidance titled "Antibacterial Drug Products: Use of Noninferiority Trials to Support Approval," which the agency announced it has withdrawn. FDA declined to comment on why the guidance was withdrawn beyond what was in the guidance.

Definition Of Effectiveness

Another one of PhRMA's central concerns that was not reflected in the final guidance was an explicit definition of "product effectiveness." PhRMA said that such a definition "would help to clarify the underlying purpose of the methods of analysis discussed within the guidance."

"PhRMA is concerned that the draft guidance implicitly defines product effectiveness only in terms of a fixed, known constant M1, which after specification may fail to accurately reflect the underlying product efficacy, and may therefore limit the accuracy of statistical analysis of product effectiveness," the drug lobby wrote.

FDA does not appear to supply an explicit definition of "product effectiveness" in the final guidance.

So What Did Change?

The final was cut to 53 pages from 63 in the draft. A large chunk of the page difference can be attributed to the shortening of examples at the end of the guidance, which aim to help illustrate the determination of the non-inferiority margin, application of methods of non-inferiority analysis and other considerations relevant to determining whether it is possible to conduct and interpret the results of a non-inferiority study. There were additionally fewer reference pages at the end of the guidance.

One clarification the guidance did make regarding M2 is its use as a "comparative effectiveness" standard. In the draft guidance, FDA included a section about the standard, where it noted that there is growing interest among third party payers and some regulatory authorities in comparative effectiveness of treatments on both cost effectiveness and medical grounds. PhRMA commented that this type of trial appears to be a special case outside the scope of the non-inferiority guidance.

FDA clarified in the final guidance that it may be important to consider whether a new product is less effective than available alternative therapies, when a less effective treatment could pose a threat to patient and public safety. The agency quoted former President Bill Clinton and Vice President Al Gore from the National Performance Review in 1995 on the regulation of drugs and medical devices. Comparative effectiveness would be appropriate, Clinton and Gore say, when the disease to be treated is life-threatening or capable of causing irreversible morbidity, such as a stroke or heart attack, or the disease is a contagious illness that poses serious consequences to the health of others, such as a sexually transmitted disease.

The agency added that comparative effectiveness studies must be distinguished from non-inferiority studies, which are the main focus of the guidance.

"The methods described in this document are intended to show that a new treatment that demonstrates non-inferiority is effective, not that it is as effective as the active comparator," the guidance states.

Much of the substance in the final guidance, however, remains unchanged from the draft. For example, sponsors should continue to use a drug with a well-defined effect as their active control in non-inferiority studies.