Bioequivalence and Bioavailability Forum

dixit ★ India, 2012-08-28 13:33 (5047 d 23:28 ago) Posting: # 9116 Views: 5,570	build-up [Design Issues] Post reply
	Dear All, In steady-state studies, the washout period of the previous treatment can overlap with the build-up of the second treatment, provided the build-up period is sufficiently long (at least 5 times the terminal half-life). kindly let me know what does the word build up exactly mean in the above paragraph. Regards, DIXIT.

Helmut
★★★

Vienna, Austria,
2012-08-28 16:26
(5047 d 20:34 ago)

@ dixit
Posting: # 9117
Views: 4,790

Saturation phase

Post reply

Dear Dixit!

❝ […] what does the word build up exactly mean

Aka “saturation phase” (see this slide). The red line are concentrations after doses in the first period (“building up” steady state). The first profile is measured on day 5 (the fist yellow line). At the end of day 5 (120 hours) we switch directly to the other formulation (therefore, this is also called a “switch-over design”). During days 5–9 (the first green line) steady state of the second formulation is build up, whilst the first formulation is simultaneously eliminated (“washed out”). On day 10 we measure the second profile (the second yellow line). Such a design is preferable over a design with a treatment-free washout (days 6–9), where we would dose on days 10–14 (represented by the second green line) and measure the second profile on day 14 (the third yellow line). The duration of the study is shorter and chances of drop-outs lower.
Rationale: According to the superposition principle of pharmacokinetics steady state is independent from previous concentrations. In other words in steady state you see the same concentrations irrespective whether you start from zero or any other concentration level. You can see this in the plot, where the thin green lines are parallel, i.e., the steady state profile in a switch-over design is identical to the one obtained in a design with a wash-out.

—
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

The Outlaw Torn
★

Europe,
2013-06-21 15:48
(4750 d 21:13 ago)

@ Helmut
Posting: # 10850
Views: 4,291

Saturation phase

Post reply

Dear Helmut,

❝ The red line are concentrations after doses in the first period (“building up” steady state). The first profile is measured on day 5 (the fist yellow line). At the end of day 5 (120 hours) we switch directly to the other formulation (therefore, this is also called a “switch-over design”). During days 5–9 (the first green line) steady state of the second formulation is build up, whilst the first formulation is simultaneously eliminated (“washed out”). On day 10 we measure the second profile (the second yellow line).

If we are for all practical reasons already at steady state on day 5, why is a second build-up necessary? Couldn't one profile be measure on day 5 and the other, after the switch, on day 6.

Thanks,
Outlaw

Helmut
★★★

Vienna, Austria,
2013-06-21 17:03
(4750 d 19:58 ago)

@ The Outlaw Torn
Posting: # 10853
Views: 4,380

Switch-over (+example)

Post reply

Hi Outlaw,

❝ If we are for all practical reasons already at steady state on day 5, why is a second build-up necessary? Couldn't one profile be measure on day 5 and the other, after the switch, on day 6.

Nope. Would only work if the true relative BA is exactly (!) 100%. It needs five half-lives to practically¹ reach steady state (irrespective of the starting level).²
A different case are replicate designs in steady state (scaling as acceptable according to the draft MR GL). Here you could measure two subsequent profiles of the same formulation. A full replicate set-up would be:

RRRRRRTTTTTTRRRRRRTTTTTT TTTTTTRRRRRRTTTTTTRRRRRR

In Bonn last week there was a discussion whether it is necessary to sample a complete second profile if scaling is intended only for C_max and/or C_min. Most people (including members of the PK-group) opted for ‘no’.
A particular case would be scaling only for C_min. In true steady state no additional administration would be necessary; the first C_min would be C₀ and the second one C_τ. Personally I would go with six administrations instead of five, since this is closer to SS (98.44% vs. 96.88%).

“Practically” in this context means 1–½⁵=96.875% of steady state. Some people are only satisfied with seven half-lives (99.219%).
A picture tells more than a thousand words. Imagine you have two formulations which show 50% (–) and 200% (–) of the reference (–).

In SD T/R of AUC_∞ 50% and 200%. Unbiased. So far, so good. Sequences in the example for MD are RT₁ or RT₂.

If you would compare AUC_τ of the first profile of the tests after the switch (= day 6) with the reference you would estimate T/R with 77% (bias +53%) and 152% (bias –24%) – since the new steady state is not established and the remaining concentrations of the reference are still very influential. If you continue dosing and use AUC_τ of day 10 you would get 53% (bias +6%) and 203% (bias +2%). In reality differences are not so pronounced and additionally due to the randomization the small remaining bias should cancel out.

—
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

The Outlaw Torn
★

Europe,
2013-06-21 17:56
(4750 d 19:05 ago)

@ Helmut
Posting: # 10855
Views: 4,226

Switch-over

Post reply

Hi Helmut,

❝ Nope. Would only work if the true relative BA is exactly (!) 100%. It needs five half-lives to practically* reach steady state (irrespective of the starting level).

If you start at concentration zero, the absolute increase is significant (0 to 100, for example). But if you are already almost at steady state (assuming a small difference in formulations) the absolute value of reaching "true steady state for the second formulation" in the build-up is likely insignificant (96 or 97 or 98 to 100, for example). So, although it's technically/theoretically true that it takes 5 or so half lives to move from 96 or 97 or 98 to 100, it's not really clinically meaningful, is what I was trying to get to initially. But I don't believe a regulator would likely be swayed by this justification, so your point is well taken.

❝ A different case are replicate designs in steady state (scaling as acceptable according to the draft MR GL).

Thanks for the feedback. Something to consider if this "concern" arises again.

Take care and have a nice weekend,
Outlaw

Helmut
★★★

Vienna, Austria,
2013-06-21 19:53
(4750 d 17:08 ago)

@ The Outlaw Torn
Posting: # 10857
Views: 4,303

Switch-over

Post reply

Hi Outlaw,

❝ […] So, although it's technically/theoretically true that it takes 5 or so half lives to move from 96 or 97 or 98 to 100, it's not really clinically meaningful, is what I was trying to get to initially. But I don't believe a regulator would likely be swayed by this justification, so your point is well taken.

See the plot I added to my previous post in the meantime. Ratios of 0.5 or 2 are extreme, but the bias for e.g., ±30% (which we commonly see) would be still substantial. ;-)

❝ ❝ A different case are replicate designs in steady state (scaling as acceptable according to the draft MR GL).

❝

❝ Thanks for the feedback. Something to consider if this "concern" arises again.

It will. C_min is mandatory and – especially if you only have limited accumulation – it might be variable as hell. See also this post about EMA’s current thinking on SD/MD.

—
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

build-up [Design Issues]

Saturation phase

Saturation phase

Switch-over (+example)

Switch-over

Switch-over