Bhupesh
☆    

India,
2009-10-12 16:14
(5281 d 14:41 ago)

Posting: # 4339
Views: 7,166
 

 Missing samples in terminal elimination phase [Study As­sess­ment]

Dear All
Regards :-)

Few days back, we did a parallel bioequivalance study for a molecule with long half life (terminal elimination half life between 120 hr - 156 hr) and have sampling points up to 720 hrs post dose.

Due to longer study period, there were some missing ambulatory samples (In one volunteer missed 3 samples between 72 - 144 hr) and in others missing some samples after 144 hr.

Tmax of the molecule was said to be in between 6 - 24 hr.

Can anyone please suggest a reliable estimation of AUC0-t considering the missing samples.

Thanks and Regards

Bhupesh


Edit: Category changed. [Helmut]
Jaime_R
★★  

Barcelona,
2009-10-12 16:51
(5281 d 14:05 ago)

@ Bhupesh
Posting: # 4340
Views: 5,810
 

 Missing samples in terminal elimination phase

Dear Bhupesh!

Why didn't you use a truncated AUC72 approach? The rationale for truncated AUC is:
  1. After 72 h the formulation has left the GI-tract and no more influence on absorption is expected after 2×tmax anyhow (search the forum and see this post).
  2. To avoid exactly the problems you are facing right now...
IMHO sampling for one month is a flawed design (missing samples are very likely).

You have some options to deal with the problem (caution: don't evaluate for BE right now. Make a decision and amend the protocol accordingly):
  • Exclude these subjects (bad, you will loose statistical power for AUC - but keep these subjects for the assessment of Cmax, which is likely the crucial metric due to higher variability). Look at your sample size estimation based on the assumed CV of AUC to get an impression whether the decreased power will be still acceptable.
  • Estimate the missing values from log-linear regression. Caution: don't use these estimated values in the calculation of the elimination phase. Regulatory acceptance: unclear.
  • Change your primary metric for extent of absoption / total exposure to AUC72.

Regards, Jaime
yjlee168
★★★
avatar
Homepage
Kaohsiung, Taiwan,
2009-10-13 00:53
(5281 d 06:03 ago)

@ Jaime_R
Posting: # 4343
Views: 5,763
 

 Missing samples in terminal elimination phase

Dear Jaime_R,

❝ Why didn't you use a truncated AUC72 approach? The rationale [...]



Since the study has been done, was it too late to switch to truncated AUC method? unless the protocol did not state the exact method... but it seems not possible.

❝ You have some options to deal with the problem (caution: don't evaluate [...]


These are very nice suggestions! :clap: These should be carefully considered when developing a BE study protocol. I've learned.

All the best,
-- Yung-jin Lee
bear v2.9.1:- created by Hsin-ya Lee & Yung-jin Lee
Kaohsiung, Taiwan https://www.pkpd168.com/bear
Download link (updated) -> here
Jaime_R
★★  

Barcelona,
2009-10-13 20:24
(5280 d 10:31 ago)

@ yjlee168
Posting: # 4349
Views: 5,646
 

 Missing samples in terminal elimination phase

Dear Yung-jin!

❝ ❝ Why didn't you use a truncated AUC72 approach?

❝ Since the study has been done, was it too late to switch to truncated AUC

❝ method?


Oh yes, sure. I should have added a :lol3: to my rhetorical question...

❝ ❝ You have some options [...]


❝ These are very nice suggestions [...] should be carefully considered

❝ when developing a BE study protocol.


Maybe. But I would put all efforts not running into such a situation, i.e. avoid sampling for one month.

Regards, Jaime
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2009-10-12 17:18
(5281 d 13:38 ago)

@ Bhupesh
Posting: # 4341
Views: 5,776
 

 Missing samples in terminal elimination phase

Dear Bhupesh

agreeing with Jaime's post one more caveat: depending on the software you use, it may be dangerous to keep these values only as 'Missing'. For instance WinNonlin always interpolates linear - leading to a positive bias of the calculated AUC. See this post and this one as well. There's a tricky (and undocumented) workaround to get a correct estimate (essentially adding two partial areas), but this is nothing for beginners. :-D

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
22,957 posts in 4,819 threads, 1,636 registered users;
91 visitors (0 registered, 91 guests [including 8 identified bots]).
Forum time: 05:56 CET (Europe/Vienna)

With four parameters I can fit an elephant,
and with five I can make him wiggle his trunk.    John von Neumann

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