You are here:--Superiority vs. non-inferiority studies – can PK-PD bridge this gap?
Superiority vs. non-inferiority studies – can PK-PD bridge this gap? 2018-07-18T08:46:37+00:00

Home Discussions Clinical Superiority vs. non-inferiority studies – can PK-PD bridge this gap?

Viewing 6 posts - 1 through 6 (of 6 total)
  • Author
  • Community managerCommunity manager
    Post count: 23
    #2645 |

    In our recent webinar titled “PK-PD in support of accelerated programmes for antimicrobial development: how much is enough?” with William Hope, we received more questions from the audience than we could answer during the session. We will now follow up on some remaining questions in the forum and encourage our experts but also users to discuss these questions.

    Please see find a recording of this webinar here.

    One webinar participant asked: In clinical trials, there is some discussion about the difficulty or even inappropriateness to perform superiority studies, while non-inferiority studies have pitfalls. Can PK/PD be used to bridge this “gap” with regard to regulatory demands?

    Ursula Theuretzbacher
    Post count: 10

    A superiority trial design is not inappropriate per se when comparing two treatments and has the advantage of requiring lower patient numbers to show an effect. The main problem with the superiority design is that it can realistically only be used in case of resistance without an available active comparator or when the comparator has severe issues (e.g. colistin). Such extensively or pan resistant bacteria may not be prevalent enough to be able to recruit enough patients in clinical trials. Additionally, such resistant bacteria cause infections mainly in critically ill patients, a patient group which is notoriously difficult to recruit. PK/PD can be used to substantiate an expedited development program. If you want to include certain resistant bacteria in the label, some clinical experience has still to be shown.

    David Shlaes
    Post count: 5

    Now we’re getting into the details of trial design. First – PK/PD, as Ursula points out, can help justify a small development program whether it would be superiority or non-inferiority in design. Superiority trials will probably only be truly feasible for antibiotics using, mainly, external or historical controls. Stay tune for the next webinar in our series.

    François Franceschi
    Post count: 7

    As mentioned above, more than inappropriate to perform, the main difficulty is to recruit enough patients. A practical example of this is Plazomycin, the original design was a superiority one, but then it was approved in a non-inferiority trial because of the difficulty in enrolling.

    Matti Karvanen
    Post count: 1

    Thank you for your answers.

    Yes, I should have written unfeasible rather than inappropriate. My question stems from a discussion where some argue that non-inferiority studies are so hampered with difficulties that only superiority studies should be acceptable for proof of efficacy. And as noted here above, a superiority study would be exceedingly difficult to perform for a new antibiotic with an indication for e.g. blood-stream infections caused by CRE.

    Hence my question: could we use a non-inferiority type of trial setup and combine with (in vitro and/or in silico) PKPD to generate stronger evidence for efficacy so as to calm any worries that critics of non-inferiority studies may have. My educated guess would be that it is technically possible, but would such a set-up pass regulatory demands?

    Looking forward to the next webinar, and I guess I ought to listen to the one in April as well titled “PK-PD in support of accelerated programmes” as well.

    Ursula Theuretzbacher
    Post count: 10

    The ability to recruit patients is a complex topic and influenced by many factors, mostly epidemiology of infection caused by resistant pathogens but also organisational factors at CROs and trial sites, number of clinical trials in the same field and location, legal obligations in a specific country or engagement of PIs. In general, superiority design in a randomized controlled trial clearly shows (if successful) that a new antibiotic works against resistant strains in clinical practice when compared with a drug that is not or less effective against resistant strains. This situation is rare and would require pan drug-resistant pathogens or colistin as main active comparator. In case of carbapenem-resistant Enterobacteriaceae this is extremely difficult (see plazomicin trial) but more realistic in case of Acinetobacter (see AIDA trial) or possibly Pseudomonas. Non-inferiority design of clinical studies with urinary tract or intra-abdominal infections is relatively easy and can be done in about one year with about 600 patients. The main problem is that the results don’t tell anything about efficacy in infections in severely sick patients with highly resistant pathogens.

    PK/PD is an important part of any expedited regulatory path and gives essential information to inform a clinical program, especially prediction of effective dosing. On the other hand, this predictive and valuable tool doesn’t replace clinical unbiased experience.

    You can find a more detailed discussion about this topic here:

Viewing 6 posts - 1 through 6 (of 6 total)

You must be logged in to reply to this topic.