Bioequivalence and Bioavailability Forum • Lottery or science?

nobody
nothing

2015-01-13 11:21
(3770 d 02:31 ago)

Posting: # 14275
Views: 10,765

Lottery or science? [PK / PD]

Hi everybody!

Read this UK EPAR by chance, stopped on page 63, Table with CIs, including for Ctrough

click me to the EPAR

Nice example that one single concentration (maybe close to LLOQ) might mess up a whole BE-study and development program.

Is this still science or is it just some lottery (inviting for every kind of, ehhhmmm, data beautification)? This Ctrough parameter for multiple-dose studies is far from good, in my opinion... Apparently the replicate design was only chosen due to Ctrough, as Cmax showed moderate CVintra, and AUC even lower.

btw: page above:
"The elimination half life of quetiapine is approx. 7 hours. It can therefore be assumed that steady state would be reached..."

Really? Isn't it flip-flop kinetics for the PR-formulation and the IR half-life isn't interesting at all (except for absorption phase)? Just asking...

Edit: Category changed. [Helmut]

—
Kindest regards, nobody

ElMaestro ★★★ Denmark, 2015-01-13 11:38 (3770 d 02:13 ago) @ nobody Posting: # 14276 Views: 9,316	Lottery or science? Post reply
	Hi Nobody, ❝ Nice example that one single concentration (maybe close to LLOQ) might mess up a whole BE-study and development program. You lost me. The product was deemed BE. Can you put a few more words to your concern? — Pass or fail! ElMaestro

nobody nothing 2015-01-13 11:45 (3770 d 02:06 ago) @ ElMaestro Posting: # 14277 Views: 9,420	Lottery or science? Post reply
	Yepp, it was bioequivalent. But is this a merrit of the formulation? Or was it purely by chance (if you have a look at the lower limit of the CI for Ctrough)? Or was it something else ? Imagine that a slightly different trough value for ONE formulation in a SINGLE volunteer might have resulted in a lower limit of CI for Ctrough of 79.98% Getting clearer now? — Kindest regards, nobody

ElMaestro
★★★

Denmark,
2015-01-13 12:05
(3770 d 01:46 ago)

@ nobody
Posting: # 14278
Views: 9,384

Lottery or science?

Post reply

Hi Nobody,

❝ Yepp, it was bioequivalent. But is this a merrit of the formulation? Or was it purely by chance (if you have a look at the lower limit of the CI for Ctrough)? Or was it something else ;-) ?

❝

❝ Imagine that a slightly different trough value for ONE formulation in a SINGLE volunteer might have resulted in a lower limit of CI for Ctrough of 79.98%

I see your point, and yes I quite agree. Hope for the best, light some candles in the church, make sacrifices on your home altar to please the higher powers. You do not always have any other good options.
It is tremendously problematic when regulators tack on new CIs that also need to pass a criterion. Sometimes I think the inclusion of such "new" metrics seem intuitive but look to be quite without a scientific basis.
Perspective: In the Advair US guidance which was published 2013 FDA required 50 (fifty!!) succesful BE-tests, including 14 confidence intervals in order for the product to be considered bioequivalent.

In such situations when primary requirements are piled skyhigh, It is virtually impossible to plan your way out of it, unless you happen to
either

have lot of $$$ so that you can power each individual test to something like 99.XYZ % so that the whole shebang passes with 80% chance or whatever your overall power aim is.
have a tremendous insight on how the individual metrics correlate and can model your way out of the trouble. Noone really has this insight to an extent where the uncertainty in the model isn't a problem.

—
Pass or fail!
ElMaestro

nobody
nothing

2015-01-13 12:23
(3770 d 01:28 ago)

@ ElMaestro
Posting: # 14279
Views: 9,331

Lottery or science?

Post reply

❝ Hope for the best, light some candles in the church

I do this on a regular basis, but it shouldn't be necessary for the scientific evaluation of a robust (!) and meaningful (!) characteristics of a PR formulation ;-)

That's the point.

If a parameter is needed to compare exposure at the end of the dosing interval, why not AUCtrough ( :D no joke)? How about: AUC -2h, -1h, 0h prior to next application? 2 additional plasma samples might make your day and safe a lot of volunteers blood (replicate design), so science meets ethics, win-win, I think.

But this Ctrough is neither meaningful nor robust (close to LLOQ and with lots of physiological variability from GI movement to whatever), it's kind of lottery that should never make it through the discussion process of a regulatory guidance document. My opinion...

—
Kindest regards, nobody

Helmut
★★★

Vienna, Austria,
2015-01-13 14:50
(3769 d 23:01 ago)

@ nobody
Posting: # 14280
Views: 9,467

Lottery: maybe. Science: no.

Post reply

Hi nobody,

❝ If a parameter is needed to compare exposure at the end of the dosing interval, why not AUCtrough ( :D no joke)? How about: AUC -2h, -1h, 0h prior to next application?

That’s an interesting idea.

❝ But this Ctrough is neither meaningful nor robust (close to LLOQ and with lots of physiological variability from GI movement to whatever), it's kind of lottery that should never make it through the discussion process of a regulatory guidance document. My opinion...

Get a decent cup/glass of you preferred drink and click this search.
We discussed that ad nauseam, not only here but at numerous conferences. I know a couple of people submitting harsh comments about it to EMA, but almost two months after the final MR-GL was adopted still none are published. I’m eager to learn what the PKWP’s “justification“ was. To my knowledge not a single [sic] paper demonstrating its “merits” is published yet. Science? Gimme a break. A hobby of one of the PKWP’s members?

—
Dif-tor heh smusma 🖖🏼 Довге життя Україна!
Helmut Schütz

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

ElMaestro
★★★

Denmark,
2015-01-13 17:45
(3769 d 20:06 ago)

@ Helmut
Posting: # 14281
Views: 9,327

Lottery: maybe. Science: occasionally

Post reply

Hi Hötzi,

❝ I’m eager to learn what the PKWP’s “justification“ was. To my knowledge not a single [sic] paper demonstrating its “merits” is published yet. Science? Gimme a break.

Diplomatic as usual. :-D

I think regulators are sitting on much more comprehensive info than just the papers that are occasionally published. In fairness, sometimes regulators know stuff you and I don't, and the guideline could simply be a consequence of that.
But in such cases it would be really prudent of the agency to publish comments and some useful degree of justification, and not just resorting to either not publishing comments and justification at all, or publishing argumentation that is conveniently vague sentences referring to reviews of the current scientific knowledge incl. study reports available to agencies etc.

We'll see....... or perhaps we won't.

—
Pass or fail!
ElMaestro

nobody
nothing

2015-01-13 18:16
(3769 d 19:36 ago)

@ ElMaestro
Posting: # 14283
Views: 9,230

Lottery: maybe. Science: occasionally

Post reply

❝ I think regulators are sitting on much more comprehensive info than just the papers that are occasionally published. In fairness, sometimes regulators know stuff you and I don't...

Might be, but can't be CIs for Ctrough, as long as they didn't calculate it from historic data (submissions). But what should prevent them from sharing their knowledge at least at some conferences? Apocryphal knowledge can't be the basis for good decisions, although it's sharply on the rise, I know...

—
Kindest regards, nobody

nobody
nothing

2015-01-13 18:12
(3769 d 19:39 ago)

@ Helmut
Posting: # 14282
Views: 9,246

Lottery: maybe. Science: no.

Post reply

Maybe we should collect some data to support the nonsense-theory for this parameter? :-)

I will post here additional studies as I find them...

The really disturbing part is: Even if someone would want to go "scientific" on new parameters for characterization of absorption, there is virtually no science on clinical Pk (anymore, at least regarding real studies, not simulation), as universities don't have the money, regulatory authorities have no capacity/budget, industry/CROs are not interested in systematic research...

Or do I get that completely wrong?

—
Kindest regards, nobody

ElMaestro
★★★

Denmark,
2015-01-13 18:46
(3769 d 19:05 ago)

@ nobody
Posting: # 14284
Views: 9,352

post additional studies

Post reply

Hi Nobody,

❝ Maybe we should collect some data to support the nonsense-theory for this parameter? :-)

❝

❝ I will post here additional studies as I find them...

Excellent idea.
Let's see the numbers. In particular, what sample size is necessary to prove BE for a drug against itself if we look at Ctrough? Once calculate please tell me what level of confidence would that sample size give us for AUCt and Cmax?

Would be great to get some insight.

—
Pass or fail!
ElMaestro

nobody
nothing

2015-01-13 19:28
(3769 d 18:23 ago)

@ ElMaestro
Posting: # 14285
Views: 9,375

post additional studies

Post reply

Here you can find at least some numbers for Ctrough in Tab 1 and 2

http://www.chinaphar.com/1671-4083/25/390.pdf

...although its only for 200mg, it's somewhat higher than LLOQ (the reason why I posted in the Bioanalytics section), but exhibits some variability. Most studies evaluate Ctrough for some days (to "evaluate" steady-state), but metrics for BE rely on only ONE concentration, hmmmm....

(And apparently t1/2 is 7 h for the ER formulation :-D

)

—
Kindest regards, nobody

Helmut
★★★

Vienna, Austria,
2015-01-14 02:42
(3769 d 11:09 ago)

@ nobody
Posting: # 14287
Views: 9,480

Historic abstract of Cmin in BE

Post reply

❝ Or do I get that completely wrong?

Not quite. I try to summarize my personal account:

C_min was never a primary PK metric in any regulation.
Some alternative PK metrics for MR products were explored in the mid 1990s. BTW, C_min was none of them (for obvious reasons I concur). I would not bet that EMA’s “experts” have read these papers. Given that C_min is a single point metric and likely the concentration with the highest variability of the entire profile powering a study for BE is a nightmare.
Is there any evidence that a product passing conventional PK metrics would fail on C_min in a properly powered BE study? I doubt it. There are more generics on the US market than in the EU. The FDA never required it. Even more: Only SD studies are required. Any problems? Is there any substantial risk?
Analyzing BE of C_min first appeared in EMA’s IR draft in 2008 after some concerns for oxycodone. However, It was dropped from the final IR GL.
At the first conference about the MR GL (Barcelona, Feb 2011) Alfredo García Arieta of the Spanish agency suggested C_τ as an additional metric in SD in order to waive MD studies.
I strongly opposed that, but …
… Paixão et al. (from the Portuguese agency and the University of Lisbon) showed in extensive simulations that it would work. They explored different type of MR formulations (IR+CR, osmotic pumps, biphasic systems). The main author received the “AAPS Outstanding Manuscript Award in Modeling and Simulation” in 2012. I was convinced.
In the meantime A. García Arieta changed his mind as well and published a bizarre cherry-picking paper. I regularly peer-review papers and this one should have been rejected because the conclusions were exactly the opposite of the reported results. Obviously a man on a mission.
Normally a draft GL is published when members of the PKWP agree at the end (or are so bugged out that they start a trial ballon). In this case there was a vote and the majority was for waiving MD studies + C_τ in SD. Can you guess who insisted in “The discussion of the opportunity of using equivalence in C_τ in single dose studies as basis for waiving the multiple dose study has been recognized. However, there is not considered to be sufficient scientific evidence at the moment to encourage this approach”?
At the next conference (Bonn, June 2013) the introduction of a new PK metric without justified relevance / discriminatory power was challenged by many participants. Alfredo told me “I don’t believe [sic] in simulations.” Henning Blume and I suggested to that it should be mandatory to submit the data, but not base the BE decision on it. A similar approach was used by the FDA when they were interested in the subject-by-formulation interaction. For two years all MR studies had to be performed in a replicate design and after analyzing them the FDA decided that S×F is not an issue.
It felt like talking to a pile of bricks.
Numerous people sent comments to EMA.
The éminence grise of the PKWP showed us the finger. Comments on the draft – which likely would show that many people did not agree – not published…

Slowly I’m loosing confidence in this entire system of (pseudo-)transparency.

Go ahead, collect data (I’m not very optimistic that you will find sufficiently powered studies). Don’t count on me. After six years of opposing C_min I’m running out of steam.

—
Dif-tor heh smusma 🖖🏼 Довге життя Україна!
Helmut Schütz

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

nobody
nothing

2015-01-14 09:42
(3769 d 04:09 ago)

@ Helmut
Posting: # 14288
Views: 9,288

Historic abstract of Cmin in BE

Post reply

I'm not on a mission :-p

I just do my job and from time to time I have a hobby. When the weather is not fine, I prefer indoor hobbies. So let' see what we can find over the time... :-D

I would a agree that a parameter should not be introduced into reg. guidance solely based on simulations. But this trash was introduced based on even less scientific justification, not to say nothing. Pretty scary... As usual: cui bono? ;-)

—
Kindest regards, nobody