Noninferiority trials

For equivalence trials, two-sided confidence intervals should be used. Equivalence is inferred when the entire confidence interval falls within the equivalence margins (p. 15). 

Confidence intervals were introduced in Chapter 8. The CIs discussed there were two-sided, even though this term was not introduced at that time. The term two-sided simply means that both the lower and the upper limit of the Cl are of interest. 

In this case, the research hypothesis states that the two drugs are equivalent, and the null hypothesis states that they are not equivalent. The locations of the lower and the upper limit of the 95% Cl determine whether or not the null hypothesis is rejected. If both the lower limit and the upper limit lie within the equivalence margin, we reject the null hypothesis and the new drug and the reference drug are declared to be equivalent. If either the lower limit or the upper limit lies outside the equivalence margin or if both limits lie outside, we fail to reject the null hypotheses, and the drugs are not declared to be equivalent. 

4 Statistical Analysis and Clinical Judgment Working Together 

The hypothesis testing approach for noninferiority trials is similar to that used in equivalence trials, as just discussed. The first step in this approach is to establish 

While a drug s efficacy is of considerable interest, other factors influence its suitability for therapeutic use and the degree to which it will be successful in the marketplace if approved. These factors include the drug s tolerability, its safety profile (and hence, in combination with its efficacy, the overall benefit-risk profile), and increasingly its cost cost is influential in the extent to which its use may be reimbursed by payers and hence the degree to which it may be prescribed. 

---

Although conventional p-values have no role to play in equivalence or noninferiority trials there is a p-value counterpart to the confidence intervals approach. The confidence interval methodology was developed by Westlake (1981) in the context of bioequivalence and Schuirmann (1987) developed a p-value approach that was mathematically connected to these confidence intervals, although much more difficult to understand It nonetheless provides a useful way of thinking, particularly when we come later to consider type I and type II errors in this context and also the sample size calculation. We will start by looking at equivalence and use A to denote the equivalence margins. 

Ringleb PA, AllenbergJ, Bruckmann H, etal. 30 day results from the SPACE trial of stent-protected angioplasty versus carotid endarterectomy in symptomatic patients a randomised noninferiority trial. Lancet 2006 368 1239-1247. 

Once the treatment effect has been determined, attention moves directly to calculation of the 95% Cl for the treatment effect. As for equivalence trials, p-values are not used in noninferiority trials. Establishing noninferiority is based entirely on the use of CIs. ICH E9 states the following ... 

As in the case of equivalence trials, it is necessary to use statistical analysis in conjunction with clinical judgment when conducting a noninferiority trial. The choice of the noninferiority margin, which must be made before the trial is conducted, is a clinical judgment. Once the data have been collected, a statistical technique is used to determine whether the new drug is deemed to be noninferior compared with the reference drug. 

Ruzyllo, W, Tendera, M., Eord, I., Fox, K. M. Antianginal efficacy and safety of ivabradine compared with amlodipine in patients with stable effort angina pectoris a 3-month randomised, double-blind, multicentre, noninferiority trial. Drugs 2007, 67, 393 05. 

Noninferiority trials are similar to equivalence trials, but require that the Investigational drug be in the worst case only trivially worse than the reference to be considered noninferior. A precise statistical definition of "trivially worse" must be agreed upon before the start of the trial. For noninferiority trials the research question is Does the new drug demonstrate efficacy that is not unacceptably worse (Fleming 2007) than the reference drug The null and alternate hypotheses corresponding to this research question take the form of ... 

Equivalence and noninferiority trials are quite different from superiority trials in their design, analysis, and interpretation (although exactly the same methodological considerations apply to collect optimum quality data in these trials). Superiority trials continue to be our focus in this book, but it is important that you are aware of other designs too. Therefore, in Chapter 12 we discuss some of the unique features of these other design types. 

The research question for a noninferiority trial is stated as Is the test drug not inferior to the control The null hypothesis to be tested in this study is ... 

Equivalence and noninferiority trials may be the only viable means to test a new drug in... 

Figure 12.4 Conclusions to be drown from the difference in population means from a noninferiority trial...

In what ways are noninferiority trials different from superiority trials ... 

Phase III studies can be tested against a placebo control with the intent of showing superiority over placebo. Another type of study design is to show equivalence or noninferiority to an approved therapy. An equivalence trial is intended to show that the response to two or more treatments differs by an amount which is clinically unimportant. A noninferiority trial demonstrates that the response to the investigational product is not clinically inferior to a comparative agent. 

In an open, randomized trial, voriconazole provided superior efficiency to C-AMB in the primary therapy of invasive aspergillosis. In a secondary analysis, survival also was superior in the voriconazole arm. Voriconazole was compared to Ambisome in an open randomized trial in the empirical therapy of neutropenic patients whose fever did not respond to more than 96 hours of antibacterial therapy. Because the 95% confidence interval in this noninferiority trial permitted the possibility that voriconazole might be... 

In a sense, virtually, all clinical trials can be considered to be safety noninferiority trials, since virtually, all trials collect safety information and therefore provide evidence as to whether or not there is an important difference in safety between treatments. However, there are two key distinguishing characteristics of the trials that we refer to as safety non-inferiority trials in this chapter a safety endpoint is designated as primary and a noninferiority margin is prespecified. In the next section, we will discuss some of the issues associated with non-inferiority trials in general, focusing on the factors that are unique to safety non-inferiority trials. 

The facf fhat an efficacy non-inferiority trial involves an indirect comparison between the experimental treatment and placebo, while a safety noninferiority trial does not, underlies nearly all of the key differences between these types of trials, as discussed in the following. 

While preservation of effect is an important consideration for efficacy noninferiority trials and is discussed in both the FDA and EMA guidance documents, it is not relevant to the design of a safety non-inferiority trial. 

---