Blinded review of PK data: History… [Study Performance]
❝ what does PPT slides 11 means?
Profiles like the left ones of Slide 9 are physiologically simply not possible. We re-analyzed them (which is nowadays no more acceptable according to the EMA’s and the FDA’s rules). All concentrations were confirmed.
We suspected a sample mix-up at the clinical site (transferring plasma samples after centrifugation). But how to confirm that? Initially we thought of DNA testing but plasma doesn’t contain cellular stuff any more. Next we considered looking at lab-values. It would be unlikely that two subjects had similar ones. Unfortunately the anticoagulant was citrate and we could only measure γ–GT and albumine. The right panel of Slide 9 showed a similar pattern than the drug concentrations and thus, confirmed the mix-up.
Slide 10 shows what would have happen when switching the suspect values. Of course, that’s not allowed. It was a pilot study and therefore, we didn’t care.
This drug is subjected to polymorphic metabolism. The within-subject variability is low (CV ≈15%) and hence, studies are performed in small sample sizes. On the other hand, between-subject variability is large (CV ≈75%). Therefore, patients are dose-titrated for effect.
In a pivotal study such a mix-up would be a disaster. Neither excluding the suspected concentrations nor the subjects is allowed. The study would fail because the variability would go through the ceiling.
At the joint EGA/EMA Symposium on Bioequivalence (London, 1 June 2010) about the BE-guideline Gerald Beuerle of ratiopharm presented an example, where – due to an obviuous mix-up – a study would pass. The study would fail if the two suspect concentrations or the subjects would be excluded. Members of the EMA’s Pharmacokinetics Working Party (PKWP) responded in unison that it is not acceptable and the product approved. Then nothing kept me in my chair. I stood up and said, “The EMA is a serious risk to public health!” This did not win me any friends in the regulatory community, I guess.
❝ In China, it's common to discuss PK parameters (without treatment assignment, such as pre-dose concentration, abnormal PK concentration profiles, λz estimation, %AUCextrap etc.) to decide analysis set in data review meeting before database lock, is it still compliant with regulatory?
Dif-tor heh smusma 🖖🏼 Довге життя Україна!
The quality of responses received is directly proportional to the quality of the question asked. 🚮
- open-label BE trials maintain blindness kimhuang 2023-10-14 09:19 [Study Performance]
- Blinded NCA Helmut 2023-10-14 10:40
- open-label BE trials maintain blindness ElMaestro 2023-10-14 17:39