EMA: MD PK metrics [Regulatives / Guidelines]

posted by Helmut Homepage – Vienna, Austria, 2020-09-08 17:25 (82 d 06:41 ago) – Posting: # 21911
Views: 1,001

Hi Pharma_88,

» I have kept following parameters in SS study (IR VS ER).

PK metrics, pleeze.

» Primary pharmacokinetic parameters: AUC0-τ,ss


» Secondary pharmacokinetic parameters:
» Day 01 to 04-Morning dose: Cpd

OK. No comparison – just report them.

» Day 04 (Morning dose): Cτ,ss, Cmax,ss, Cmin,ss, Tmax,ss, accumulation index, %Fluctuation and Cav
» Please correct me.

If you don’t have a lag-time, Cmin,ss (the ‘true’ minimum concentration) should be very similar to both Cpd and Cτ,ss. Cmin,ss is required only for originators. Since a hybrid application is closer to a generic, you could drop it.

AFAIK, the accumulation index \(\frac{1}{1-\exp (-\lambda_\textrm{z}\cdot \tau)}\) is not required by any agency.
You would need a reliable estimate of \(\small{\lambda_\textrm{z}}\), which you may not obtain for the ER formulation in all subjects. As long as you don’t have flip-flop PK (see below), \(\small{\lambda_\textrm{z}}\) is a property of the drug and therefore, not relevant for comparing products.
Cav is not informative because it is simply AUC0–τ/τ. If you want you can report it though without a statistical comparison. The outcome will be exactly the same as for AUC0–τ. Cav is only needed for the calculation of fluctuation.

» Cpd: All pre-dose samples needs to be considered including profiling day sample or only samples before profiling day?

You have to collect at least three pre-dose samples in order to demonstrate achievement of steady-state. In your case mornings of days 2–4. See also this presentation (slide 14).

» Cτ,ss: How to calculate it? Whether morning dose of Day 04 is also need to be taken in to account to calculate.

Nope. It’s the concentration at the end of the (last) dosing interval or in your case at 96 hours. Too lazy to search but the EMA states somewhere that this sample should be collected with a time deviation of ≤10 minutes. In my studies I always used the estimated concentration at τ.

Let’s explore three examples. One-compartment model, no lag-time, D 100, V 5, t½,el 14 h, t½,abs 1 h (IR). Pseudo-steady-state reached after 5 × 14 = 70 h, results for the profile day 4.
  1.       Cpd  Ctau  AUCtau  %PTF  Cmax
    IR    9.2   9.4   400.0    87  23.9
    ER   11.5  11.7   399.0    48  19.8
    T/R  125%  125%    100%   56%   83%


    Cτ,ss > Cpd. In practice this difference likely will not be evident due to variability. If you prefer a belt plus suspenders, dose for another day. As expected the AUCs are practically identical and the ER produces less fluctuation.
    Even for a moderately slower rate of absorption it is not possible to waive the MD study because the extrapolated AUC is 45% of AUC0–∞.

  2.       Cpd  Ctau  AUCtau  %PTF  Cmax
    IR    9.2   9.4   400.0    87  23.9
    ER   13.2  13.6   395.9    28  18.2
    T/R  144%  145%     99%   32%   76%


    Similar but a further reduction in fluctuation.

  3.       Cpd  Ctau  AUCtau  %PTF  Cmax
    IR    9.2   9.4   400.0    87  23.9
    ER   13.6  14.5   377.8    13  16.6
    T/R  144%  145%     94%   15%   69%


    Now we crossed the border of flip-flop PK (kabs ≤ kel) which leads to trouble. Absorption of the ER is the slowest process and hence, steady-state is not reached.
In any case to hope to demonstrate equivalence of Cmax and Cτ is futile. Either concentrate on PTF (one-sided test for non-superiority) and/or consider a bracketing approach (see this post).

Since you mentioned before that you have to deal with a HVD(P), you can expand the limits for some PK metrics (see this post) if the study will be performed in a replicate design. You don’t need a treatment-free washout, i.e., the built-up of a subsequent steady states can overlap with the washout from the previous one. Example for a full replicate design, where T = ER and R = IR. Profile days in red.

Sequence/Day  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15
     1        T  T  T  T  T  R  R  R  R  R  T  T  T  T  T
     2        R  R  R  R  R  T  T  T  T  T  R  R  R  R  R

No fun.

Dif-tor heh smusma 🖖
Helmut Schütz

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

Complete thread:

 Admin contact
21,220 posts in 4,427 threads, 1,481 registered users;
online 13 (0 registered, 13 guests [including 8 identified bots]).
Forum time: Sunday 23:06 CET (Europe/Vienna)

Half the harm that is done in this world
Is due to people who want to feel important.    T. S. Eliot

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