Prescrire has participated in the public consultation organised by the European Medicines Agency (EMA) on its "Concept Paper for the Development of a Guideline on Non-Inferiority and Equivalence Comparisons in Clinical Trials", dated 12 February 2024 (1).

In the field of health care, the aim of a non-inferiority trial is to show that a treatment is not markedly less effective or markedly more harmful than the treatment to which it is compared. In other words, this type of trial is not intended to show whether the treatment represents a therapeutic advance, and questions have been raised about whether such trials are even ethical (2-4).

Yet it is essential to establish which new drugs represent a real advance over the treatment options already available. It is only through randomised  comparative clinical trials, designed to demonstrate superiority, that such an advance can be determined, provided that the trial results are based on high-level evidence. If the EMA is to improve the quality and safety of health care, first and foremost in the interests of patients, it should impose a hard and fast rule that the evaluation of new drugs must include at least two randomised trials, designed to demonstrate the drug's superiority over the standard of care, when such treatment exists.

The use of non-inferiority trials to obtain marketing authorisation must remain an exception to this rule, reserved for rare circumstances that are clearly defined and restricted in the forthcoming guideline. These trials should always compare the new drug to the standard of care, and should be used exclusively to show the non-inferiority of the drug's efficacy, provided that there are sufficient robust data suggesting that the tested drug is also less toxic or more convenient than the standard of care.

The document released by the EMA for consultation states that non-inferiority trials versus placebo are sometimes intended "to demonstrate that the new treatment is not harmful" (1). The use of non-inferiority trials for this purpose is unacceptable and must be explicitly forbidden.

Every pharmacological treatment, without exception, has adverse effects. They may be infrequent or mild, or may only develop after prolonged exposure or after a long delay (sometimes long after the end of treatment), but there are always some adverse effects. It would be fallacious to conclude that a drug has no adverse effects on the basis of a placebo-controlled trial. In addition, pharmacovigilance is essential for identifying a drug's adverse effects (5). The European guideline on non-inferiority and equivalence comparisons in clinical trials must explicitly prohibit the use of placebo as the comparator, with no exceptions.

References
1- EMA "Concept paper for the development of a guideline on non-inferiority and equivalence comparisons in clinical trials" available here https://www.europa.eu/en/documents/scientific-guideline/concept-paper-development-guideline-non-inferiority-equivalence-comparisons-clinical-trials_en.pdf
2- Prescrire Editorial Staff "Concepts and methods: non-inferiority studies" Prescrire Int 2006; 15 (85): 165.
3- Prescrire Rédaction "Remue-méninges. Un chouïa moins?" Rev Prescrire 2021; 41 (451): 397-398.
4- Garattini S, Bertele' V "Non-inferiority trials are unethical because they disregard patients' interests" Lancet 2007; 370: 1875–1877.
5- Prescrire Editorial Staff "Evaluation of treatment risks: taking clinical data, pharmacology and patient characteristics into account" Prescrire Int 2010; 19 (105): 44-45.

©Prescrire 1 September 2024

Source: "Prescrire's contribution to the EMA's consultation on the 'Concept Paper for the Development of a Guideline on Non-Inferiority and Equivalence Comparisons in Clinical Trials'" Prescrire Int 2024; 33 (262): 223. Free.

Download the full article