Amidst the current fiscal climate, certain words evoke visceral reactions ... as one example - a margin call, ie the process of an authority such as a lender making a demand of a account holding client to make good on a shrinking collateral (margin) balance, by demanding payment pronto (which may have dire consequences for a party with a diminishing account "coming due" because of taking a calculated risk, based on prior assumptions that subsequently changed). In like spirit, the review of noninferiority margin by the FDA's Anti-Infective Drugs Advisory Committee this week will recommend changes that may be viewed as precedent setting, and which are likely to have similar impact on "pharma portfolio managers" who are subject to the decisions of the agency and its advisors - with potential dramatic impact for sponsors across all therapeutic areas.
For those of us who are not statistical wizards, the concept of noninferiority margins refers to an issue of "assay sensitivity" involving the setting of expectations for the ranges of outcomes in clinical trials comparing two agents which are deemed acceptable to declare the two competing chemicals as "equivalent" (ie calculations to rule out that the test intervention is no worse than the control intervention by a chosen amount). In turn, the margin set will drive the size (and cost/duration) of such trials ... the smaller the margin set (ie 10% vs. 15%) the more patients needed in the study to prove noninferiority. Past US anti-infective trials had been constructed with margins around 15% as being a reasonable rule of thumb (as in FDA's 1998 Points to Consider document), until a 2001 expert review (and subsequent FDA communications) suggested a reduction of the acceptable margin below 15% was indicated, which in some cases did compel sponsors to extend ongoing studies to meet the expectations of the change in standards. Clearly, setting margins for any trial requires thoughtful and individual review based on disease severity, range of comparator/historical control outcomes and like considerations. In the Advisory Committee this week, the discussion will focus on margins for complicated skin and skin structure infections ... many speculate that margin call will be 10%, and that the recommendations will be used as a reference point in other therapeutic groups (a la the "Ten" Commandments), which will, in turn, create some very challenging circumstances for those considering development programs for new antimicrobials and perhaps other agents. In some cases, it is possible that larger companies faced with portfolio decisions on selecting which compounds to advance (or terminate) given limited funds (and those smaller companies which have projected their capital burn for development based on prior margin rules) may be compelled to delay or exit these areas of study. Might the law of unintended consequences be invoked here, in that making margins more rigid could force pharma's anti-infective developers to reconsider their chances of successful, timely development (and instead provoke them to polish their burger flipping skills from college years)? Or will the modifications in margins be accompanied by more creative design features (a la the Critical Path programs, which might consider modifications such as single tail statistical designs for noninferiority analyses in selected conditions, or conversely to encourage superiority trials in a "bet the ranch" approach?). Whatever the outcomes, it is hoped that those deliberating on the margin call give due consideration to long term factors which their recommendation may influence ... to paraphrase HF Emerson "when you pick up one end of a stick, the other end comes along" and this stick could impose quite a beating ... or be a framing timber of new trial constructs ... depending on what and how this discussion plays out. In a worst case scenario, the margin of noninferiority and the margin of company development budgets could become an inverse relationship for sponsors - with a chilling effect on innovation, and thus no marginal gain for patients in the end result ... which would be an unfortunate & unintended margin of error in judgement.
