Bioequivalence and Bioavailability Forum • R-SABE and ABEL versus Blood Volume

Lucas
★

Brazil,
2015-08-11 18:06
(3972 d 00:57 ago)

Posting: # 15231
Views: 7,376

R-SABE and ABEL versus Blood Volume [Design Issues]

Hi to everybody.

RSABE and ABEL came to aid in studies of HVDs/ HVDPs, but what must we do when we are dealing with a fixed dose combination (FDC) that has, for instance, 3 drugs with different PK, and consequently different blood sampling schedule, and one of the drugs is a HVD?

Take Clorpheniramine + Phenylephrine + Paracetamol as a example. We are worried with Phenylephrine's ISCV and want to use the ABEL approach for it, but for the other 2 drugs we don't feel that there is need for that. If I use a replicated design, even a 3-period one, the sampling schedule is very compromised. In Brazil we have a limit of 530mL of blood per study, so no more than that can be withdrawn from the subjects in the entire study.

So my question is: is it possible to not use the replicated samples for the other 2 drugs and apply conventional ABE for them, and consider the replication only for the HVD? Does that make sense? For me this seems a bit bizarre and questionable, in ethics and in GxP. Other option would be to use Balaam's 4x2 design, but I don't think that someone has budget for that... :-D

THX in advance.

Lucas

Helmut
★★★

Vienna, Austria,
2015-08-11 18:41
(3972 d 00:22 ago)

@ Lucas
Posting: # 15232
Views: 6,206

R-SABE and ABEL versus Blood Volume

Post reply

Hi Lucas,

yep, combos can be challenging when it comes to the blood volume. What I have seen is to optimize the sampling schedule for each drug and split samples accordingly. Drawbacks:

Makes only sense if the PK is substantially different.
The logistics is challenging. We color-coded all vials and even the respective rows in the CRF. Turned out to be not fail-safe.
I have never seen something like this for more than two drugs.

❝ is it possible to not use the replicated samples for the other 2 drugs and apply conventional ABE for them, and consider the replication only for the HVD?

Not sure whether I understand you. Can you post sequences / periods you have in mind?

❝ Other option would be to use Balaam's 4x2 design, but I don't think that someone has budget for that... :-D

Correct. In Balaam’s design we throw away half of the sample size in the estimation of T/R. Furthermore, I don’t know how to estimate the sample size. RT|TR|RR|TT is currently not implemented in PowerTOST sampleN.scABEL().

—
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

Lucas ★ Brazil, 2015-08-11 19:04 (3971 d 23:59 ago) @ Helmut Posting: # 15233 Views: 6,138	R-SABE and ABEL versus Blood Volume Post reply
	Hi Helmut! Do you believe that I still can't access the forum via my work computer? That's a unsolved mystery! PASS gives, for example, a result of 260 subjects to achieve 80% power with a ISCV of 40% and a true ratio of 95%, if I'm not mistaken... It's not very practical. Thx! Lucas

Helmut
★★★

Vienna, Austria,
2015-08-11 19:31
(3971 d 23:32 ago)

@ Lucas
Posting: # 15234
Views: 6,164

R-SABE and ABEL versus Blood Volume

Post reply

Hi Lucas,

❝ Hi Helmut! Do you believe that I still can't access the forum via my work computer?

If you tell…

❝ That's a unsolved mystery! :-D

It’s easier to be blacklisted than to get out of the trap. Can you e-mail your office PC’s IP again?

❝ PASS gives, for example, a result of 260 subjects to achieve 80% power with a ISCV of 40% and a true ratio of 95%, if I'm not mistaken... It's not very practical.

Confirmed:

library(PowerTOST) sampleN.TOST(CV=0.4, design="2x4x2") +++++++++++ Equivalence test - TOST +++++++++++ Sample size estimation ----------------------------------------------- Study design: Balaam's (2x4x2) log-transformed data (multiplicative model) alpha = 0.05, target power = 0.8 BE margins = 0.8 ... 1.25 Null (true) ratio = 0.95, CV = 0.4 Sample size (total) n power 260 0.805585

But: I would not assume a ratio of 0.95 for a HVD/HVDP. For 0.90 the sample size explodes to 532.

—
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

Lucas
★

Brazil,
2015-08-11 20:26
(3971 d 22:37 ago)

@ Helmut
Posting: # 15235
Views: 6,074

R-SABE and ABEL versus Blood Volume

Post reply

Thanks Helmut, I'll send you my ip later.

I forgot to ask your question. The suggestion I was given was to not use the samples in period 3 (for a 2x2x3 RTR|TRT crossover, let's say) for quantification of the drugs that aren't HVDs, only use for the drug that is in order to be to collect less blood in this period and meet the limit of blood.

So for the drugs with low variability we would have 2 periods (RT and TR), each treatment one time, like a 2x2. But this seems too mixed up for me... Gotta be 2x3 or 2x2. Or am I wrong?