WHO: Chapeau! [BE/BA News]

posted by Helmut Homepage – Vienna, Austria, 2017-06-12 23:31 (1564 d 10:56 ago) – Posting: # 17490
Views: 7,759

Dear Ohlbe,

» The WHO is apparently opening a pilot phase to allow scaling for AUC on a case-by-case basis. Protocols and justifications are to be submitted in advance to the WHO PQP team.

THX at lot for this gem! Obviously people at the WHO are more clever than the EMA’s – by avoiding this IMHO, undesired – side effect. Kudos!

Something puzzles me when reading the respective section of TRS 992, Annex 7, Section 7.9.3 “Highly variable active pharmaceutical ingredients” again…

A “highly variable API” has been defined as an API with an intrasubject variability of > 30% in terms of the ANOVA-CV (14). Proving the bioequivalence of FPPs containing highly variable APIs can be problematic because the higher the ANOVA-CV, the wider the 90% confidence interval. Thus large numbers of subjects must be enrolled in studies involving highly variable APIs to achieve adequate statistical power.
Although there is variability in how regulatory authorities deal with the issue of highly variable APIs, the most rigorous of the current approaches involve the scaling of bioequivalence acceptance criteria based on the intrasubject standard deviation observed in the relevant parameters for the comparator product (15–17). Of the two most common assessment parameters Cmax is subject to the highest variability and hence is the parameter for which a modified approach is most needed.
For highly variable FPPs it is recommended that a three-way partial replicate (where the comparator product is administered twice) or a four-way fully replicated cross-over bioequivalence study be conducted and reference-scaled average bioequivalence be employed to widen the acceptance interval for the Cmax parameter, if the intrasubject variability for Cmax following replicate administrations of the comparator product is > 30%. If this is the case the acceptance criteria for Cmax can be widened to a maximum of 69.84–143.19%. The applicant should justify that the calculated intrasubject variability is a reliable estimate and that it is not the result of outliers.

Translation of the WHO’s terms: highly variable API = highly variable drug (HVD), highly variable finished pharmaceutical product (FPP) = highly variable drug product (HVDP).
Neither in (14) nor in any of all the other Bio-International conference I attended, 30% was set as the limit for the API. The respective section of (14) reads:

In some cases, however, it has been difficult to meet these criteria in experiments with a reasonable number of subjects. This may be, at least in part, a function of high intrasubject variability of the drug itself and/or the drug product. […]
Role of Variability of Reference Product-The panel does, however, believe that there may be some value in seriously considering the possibility of varying the confidence interval acceptance criteria in accordance with the intrasubject variability of the reference product and/or expanding bioequivalence intervals based on pharmacodynamic considerations.

(my emphases)
The distinction between HVDs (highly variable if administered IV or as a solution) and HVDPs (highly variable if administered as a pharmaceutical product) was first made by Midha a couple of years later. Whether we have to deal with a HVD or a HVDP is of academic interest only. If reference-scaling would be only acceptable for HVDs (highly variable APIs!) many HVDPs would fall through the cracks. Example: CVintra of IV diclofenac <10% but CVintra of enteric coated diclofenac ~40%. At last the last paragraph of the WHO’s GL talks about “highly variable FPPs”. Confused? So am I.

Dif-tor heh smusma 🖖
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes

Complete thread:

Activity
 Admin contact
21,693 posts in 4,535 threads, 1,541 registered users;
online 5 (1 registered, 4 guests [including 3 identified bots]).
Forum time: Friday 10:28 CEST (Europe/Vienna)

Tortured data will confess to anything.    Fredric Menger

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5