Bioequivalence and Bioavailability Forum

jag009 ★★★ NJ, 2012-06-19 01:19 (4766 d 11:46 ago) Posting: # 8782 Views: 10,361	Highly soluble Highly permeable drug [PK / PD] Post reply
	Hi all, If a drug product (IR) is highly soluble and highly permeable. Can one conclude that it will have small within subject variability? Thanks John

Helmut
★★★

Vienna, Austria,
2012-06-19 03:53
(4766 d 09:12 ago)

@ jag009
Posting: # 8783
Views: 9,051

BCS Class I, variability

Post reply

Hi John!

Being in BCS I will keep one part of the variability low. But don’t forget presystemic and/or fist-pass metabolism and variability in clearance.
Examples from my files and the literature (single dose; CV_intra of C_max, № of studies):

Amitryptiline: 20–33% (7)
Doxycycline: 9–54% (43)
Metronidazole: 7–10% (7)
Prednisolone: 7–23% (4)
Verapamil: 18–35% (3)

I love doxicycline if planning an in vivo study. ;-)

—
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

jag009 ★★★ NJ, 2012-06-19 18:56 (4765 d 18:10 ago) @ Helmut Posting: # 8799 Views: 9,170	BCS Class I, variability Post reply
	Thanks Helmut, I have one molecule now with CVintra for Cmax 78%!!! Based on n=18. Yes compound undergoes presystemic and systemic metabolism. Data is all over the place. John

Helmut
★★★

Vienna, Austria,
2012-06-19 19:17
(4765 d 17:48 ago)

@ jag009
Posting: # 8801
Views: 9,134

CV 78% – RSABE

Post reply

Hi John!

❝ I have one molecule now with CVintra for Cmax 78%!!! Based on n=18. Yes compound undergoes presystemic and systemic metabolism. Data is all over the place.

Luckily for the FDA you can apply RSABE both to C_max and AUC; major obstacle with very high CVs is the GMR-restriction. According to the tables of the two Lászlós that would mean (80% power for CV_WR 80%):

                                        expected GMR

design                      0.85 0.90 0.95 1.00 1.05 1.10 1.15 1.20

3-period partial replicate  144   40   29   27   29   37   76  >201

4-period full replicate      95   29   21   20   20   25   51  >201

So it’s no too tough – unless the GMR is more than ±10% away from unity.

—
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

jag009 ★★★ NJ, 2012-06-19 21:37 (4765 d 15:29 ago) @ Helmut Posting: # 8805 Views: 9,043	CV 78% – RSABE Post reply
	Hi Helmut, Cancer patients

Helmut
★★★

Vienna, Austria,
2012-06-19 22:03
(4765 d 15:03 ago)

@ jag009
Posting: # 8807
Views: 8,941

CV 78% – parallel

Post reply

Hi John!

❝ Cancer patients

Are you by any chance referring to this post of yours? Terrible – especially #5.
In a parallel design (T/R 95%, 80% power, CV 78%) you would need 410 patients. :not really:

But the 78% was CV_intra; what was CV_total in your pilot study?

—
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

jag009 ★★★ NJ, 2012-06-20 01:07 (4765 d 11:59 ago) @ Helmut Posting: # 8809 Views: 8,973	CV 78% – parallel Post reply
	Hi helmut, Tada! You answered the correct question. A 3-way replicate design? Actually FDA has details on the study design, looks interesting. I will look up the total CV. 78% is because a weirdo, without him the CV was 58%... If I go by Laszlo's sample size table, then n~40... John

Helmut
★★★

Vienna, Austria,
2012-06-20 04:40
(4765 d 08:25 ago)

@ jag009
Posting: # 8810
Views: 9,144

CV 78% – parallel

Post reply

Hi John!

❝ A 3-way replicate design? Actually FDA has details on the study design, looks interesting. I will look up the total CV. 78% is because a weirdo, without him the CV was 58%... If I go by Laszlo's sample size table, then n~40...

Are you trying to confuse me? I would not perform a cross-over in patients, because cancer is not a stable disease. A replicate design would be even more problematic (№ of blood samples). I was asking for the CV_total because you need this value (rather than CV_intra) in designing a parallel study. You cannot scale in a parallel design. Forget Lászlós’ tables. BTW, do you have to deal with a NTID? If yes, note that FDA is considering scaling down (=narrowing) the acceptance range (Advisory Committee for Pharmaceutical Science and Clinical Pharmacology, July 26, 2011).

—
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

jag009
★★★

NJ,
2012-06-20 18:19
(4764 d 18:46 ago)

(edited on 2012-06-21 15:50)
@ Helmut
Posting: # 8817
Views: 8,923

CV 78% – parallel

Post reply

Hi Helmut,

Sorry for the confusion. I was still thinking at the time of writing, cloudy mind :-)

.

For Cmax (results with all subjects including one weirdo), the CV_total is.... Problem... MSEw = 0.410377 (From Proc GLM table Root MSE = 0.640607), Sub(seq) MS = 0.308524 (for calculating intersubject CV, from Proc GLM table). I can't freaking calculate the intersubject CV :confused:

Cmax data is all over the place (Range 3.0 ng/mL to 35 ng/mL in each treatment arm).

I can calculate the CV_total: Sqrt(e^{0.308524+0.410377/2}-1)*100 = 65.76%

Regarding the no-no in running crossover. You will be surprised as 1 or 2 CROs that my project manager spoken with claimed that they have done crossover in cancer patients in the past...

I still have to confirm that this drug (for Hogkin) is a NTID but I don't think it is.

We are trying to get by with a BCS class waiver on this drug but that route seems problematic. That's why I want to see how complicate and costly the biostudy will be.

One interesting note.. Before I was "Born" in my current company, someone ran a healthy subject crossover study with this drug and that's where I got the data from. :lookaround:

yes it's a cancer drug...

Thanks

John

Helmut
★★★

Vienna, Austria,
2012-06-20 18:47
(4764 d 18:18 ago)

@ jag009
Posting: # 8818
Views: 8,830

CV 78% – parallel

Post reply

Hi John!

❝ Sorry for the confusion. I was still thinking at the time of writing, cloudy mind :-)

Never mind. I will answer only partly (have to leave the office shortly).

❝ Regarding the no-no in running crossover. You will be surprised as 1 or 2 CROs that my project manager spoken with claimed that they have done crossover in cancer patients in the past...

I have seen such studies as well. One assumption required for a cross-over is that the conditions in all periods are the same. This is not the case if

the disease state chances independent from the drug (might influence the PK)
the drug ‘works’ – and we are back at #1.

❝ I still have to confirm that this drug (for Hogkin) is a NTID but I don't think it is.

❝ We are trying to get by with a BCS class waiver on this drug but that route seems problematic. That's why I want to see how complicate and costly the biostudy will be.

OK, you have to check that. Biowaivers are not acceptable for NTIDs.

❝ One interesting note. Before I was "Born" in my current company, someone ran a healthy subject crossover study with this drug and that's where I got the data from. :lookaround: yes it's a cancer drug...

I have done that as well. It depends one the dose and effect/AE profile. Methotrexate is one example.

—
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

jag009 ★★★ NJ, 2012-06-20 22:50 (4764 d 14:15 ago) @ Helmut Posting: # 8819 Views: 8,868	CV 78% – parallel Post reply
	Hi Helmut, See my edited post above your reply. Thanks john

Highly soluble Highly permeable drug [PK / PD]

BCS Class I, variability

BCS Class I, variability

CV 78% – RSABE

CV 78% – RSABE

CV 78% – parallel

CV 78% – parallel

CV 78% – parallel

CV 78% – parallel

CV 78% – parallel

CV 78% – parallel