Sereng
☆    

USA,
2023-05-24 17:41
(422 d 21:57 ago)

Posting: # 23566
Views: 1,857
 

 Sample Size Benefits of Replicate Design with ABE [Power / Sample Size]

Folks, I have three questions:

(1) If you conduct a full replicate crossover design BE study (T x 2, R x 2, e.g., TRTR|RTRT) is there any sample size efficiency (fewer patients required) with ABE analysis or is it no different (in terms of sample size) than a non-replicate crossover design BE study (T x1, R x 1) with ABE analysis.

(2) For demonstrably HVD (WSV >30%), why do some product specific guidance for ANDA suggest (or allow) full replicate designs and others do not mention this approach?

(3) For demonstrably HVD (WSV >30%), why do some Clinical Divisions object to use of full replicate design studies in 505(b)(2) NDAs?

Any input would be appreciated!

Biostatistically Challenged CEO
dshah
★★  

India/United Kingdom,
2023-05-25 11:56
(422 d 03:42 ago)

@ Sereng
Posting: # 23569
Views: 1,513
 

 Sample Size Benefits of Replicate Design with ABE

HI Sereng!


❝ (1) If you conduct a full replicate crossover design BE study (T x 2, R x 2, e.g., TRTR|RTRT) is there any sample size efficiency (fewer patients required) with ABE analysis or is it no different (in terms of sample size) than a non-replicate crossover design BE study (T x1, R x 1) with ABE analysis.


The number of observations will be same but the sample size will be half in fully replicate design compared to standard CO study.

❝ (2) For demonstrably HVD (WSV >30%), why do some product specific guidance for ANDA suggest (or allow) full replicate designs and others do not mention this approach?


For HVD, the fully replicate/partial replicate can be helpful for widening of 90% CI limit based on regulatory agency for particular PK parameters. Ethically, due to widening of 90% CI, the sample size will be less than if it has be to standard limit of 80-125%.

❝ (3) For demonstrably HVD (WSV >30%), why do some Clinical Divisions object to use of full replicate design studies in 505(b)(2) NDAs?

Not aware of any. Can you share the link?

Regards,
Divyen
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2023-05-25 18:27
(421 d 21:11 ago)

@ Sereng
Posting: # 23570
Views: 1,556
 

 Dropouts less problematic

Hi Sereng,

I agree with Divyen’s post.
Maybe this article helps as well. Furthermore, if in a TRTR|RTRT-design a subject drops out in period 3 or 4, he/she can be kept in the analysis – and would not be completely lost like a dropout in period 2 of a 2×2×2 Xover.

Dif-tor heh smusma 🖖🏼 Довге життя Україна! [image]
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
UA Flag
Activity
 Admin contact
23,112 posts in 4,858 threads, 1,644 registered users;
89 visitors (0 registered, 89 guests [including 16 identified bots]).
Forum time: 15:39 CEST (Europe/Vienna)

Statistics means you never have to say you’re certain.    David L. Streiner

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5