jag009
★★★

NJ,
2013-06-19 19:44
(3935 d 01:22 ago)

Posting: # 10822
Views: 7,779
 

 Failing BE study due to AUCinf [NCA / SHAM]

Hi everyone,

Just out of curiosity. Has anyone seen a BE study failing due to AUCinf falling outside 80-125% and the variability in the results seem to be attributed to the determination of Kel (jumpy concentrations)? Sampling timepoint setup was adequate in terms of capturing the terminal eleminaition phase of the PK profile. I wonder if there is any way to save this study.


Thanks
John
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2013-06-19 20:15
(3935 d 00:51 ago)

@ jag009
Posting: # 10824
Views: 7,008
 

 AUC∞ = lousy metric

Hi John,

❝ Has anyone seen a BE study failing due to AUCinf falling outside 80-125% and the variability in the results seem to be attributed to the determination of Kel (jumpy concentrations)?


From a European perspective, AUC doesn’t bother me (BE only required for AUCt). Sometimes sponsors wanted me to assess BE of AUC as ‘supportive information’ (for some wacky reasons I don’t understand). As expected regularly the CV was higher. Wouldn’t surprise me if there are studies in the wild passing AUCt and failing AUC. Kamal Midha gave numerous presentations showing dependency of the variability of partial AUCs from the truncation time point. The CV is (sometimes prohibitively) large during the – early absorption phase and decreases afterwards. Sometimes the variability slightly increases again during the latest time points since the variability of concentrations approaching the LLOQ increases. AUC was generally more variable than AUCt.
Funny enough it was shown already two decades ago that AUC is a bad metric for extent of absorption.* Some of the authors were usual suspects from the FDA. From the abstract:

The accuracy, precision, and ease of use of the various measures of extent were evaluated, and statistical power analyses were performed. Among the measures tested, the most reliable were the AUC computed up to the time of the last quantifiable concentration, without extrapolation, and Cmax.

(my emphasis)
Lessons not learned.

❝ Sampling timepoint setup was adequate in terms of capturing the terminal eleminaition phase of the PK profile. I wonder if there is any way to save this study.


Duno. Maybe you can start an argument that AUC is not a reliable PK metric. According to the paper AUC is not only more variable, but may also be biased (depending on the PK model and the LLOQ). On the other hand both AUCs are required. Good luck.


  • Bois FY, Tozer TN, Hauck WW, Chen M-L, Patnaik R, Williams RL. Bioequivalence: Performance of Several Measures of Extent of Absorption. Pharm Res. 1994;11(5):715–22. doi:10.1023/A:1018932430733.

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
jag009
★★★

NJ,
2013-06-19 21:35
(3934 d 23:31 ago)

@ Helmut
Posting: # 10827
Views: 6,836
 

 AUC∞ = lousy metric

Hi Helmut,

True because AUCi is based on extrapolation of the terminal phase Kel which can cause all kind of issues when the concentrations in the elimination phase become erratic due to pk complexity and other factors.

I was just wondering if you all experienced (okay, FDA only studies) such scenarios, what kind of approach you used and what the outcome was (meaning did the agency agree to only Cmax and AUCt).

Like I said a while back a clinic I used previously would consider kel as "cannot-be-determined" if the last 2 (or 1) timepoints showed a +ve increase in concentration relative to the timepoints before.

Thanks
John
jag009
★★★

NJ,
2013-06-19 23:20
(3934 d 21:46 ago)

@ Helmut
Posting: # 10831
Views: 6,809
 

 AUC∞ = lousy metric

Hi Helmut,

❝ Duno. Maybe you can start an argument that AUC is not a reliable PK metric. According to the paper AUC is not only more variable, but may also be biased (depending on the PK model and the LLOQ). On the other hand both AUCs are required. Good luck.


Not trying to play genius but isn't that obvious since AUCinf is based on AUCt and a component which is based on extrapolation and in turn dependent on regression?

John
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2013-06-20 19:29
(3934 d 01:37 ago)

@ jag009
Posting: # 10841
Views: 6,733
 

 AUC∞ = lousy metric

Hi John,

❝ Not trying to play genius but isn't that obvious since AUCinf is based on AUCt and a component which is based on extrapolation and in turn dependent on regression?


Yes & no. The variability is higher because it depends on Clast which is expected to be the most variable data point of the profile. That’s trivial. More interesting is that AUC might also be biased especially for two-compartment PK and caused by not always ‘catching’ the true λz in the area of LLOQ.

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
jag009
★★★

NJ,
2013-06-20 22:32
(3933 d 22:34 ago)

@ Helmut
Posting: # 10843
Views: 6,616
 

 AUC∞ = lousy metric

Hi Helmut,

❝ Yes & no. The variability is higher because it depends on Clast which is expected to be the most variable data point of the profile. That’s trivial.


Agree especially when the value is low enough to approach LLoQ.

❝ More interesting is that AUC might also be biased especially for two-compartment PK and caused by not always ‘catching’ the true λz in the area of LLOQ.


Exactly but for BE purposes (I mean ANDA) which is limited to Non-compartmental analysis (correct me if I wrong), this alone will lead to biased estimation of Kel.

John
ElMaestro
★★★

Denmark,
2013-06-19 20:48
(3935 d 00:18 ago)

@ jag009
Posting: # 10826
Views: 6,855
 

 Failing BE study due to AUCinf

Hi Jag,

❝ Just out of curiosity. Has anyone seen a BE study failing due to AUCinf falling outside 80-125% and the variability in the results seem to be attributed to the determination of Kel (jumpy concentrations)? Sampling timepoint setup was adequate in terms of capturing the terminal eleminaition phase of the PK profile. I wonder if there is any way to save this study.


I have occasionally seen this with drugs requiring very low LLOQs such as inhalation drugs. It seems to me like the assays sometime underperform slightly near the LLOQ in spite of bioanalytical assays with reported scatter at LLOQ within acceptable limits.
Check your ISRs at or near the LLOQ and contrast them with the rest. How does it look?
You may sometimes trigger an audit and have the opportunity to eliminate subjects for whom manifest trouble have been observed but you need a better reason than just a desire to get rid of bad values. Also, it sounds like you may argue that variability is the problem (cf. "jumpy" concentration in the elimination phase) and therefore the study was not adequately powered in terms of AUCinf so you can repeat it with Barbara D's blessing.

Pass or fail!
ElMaestro
jag009
★★★

NJ,
2013-06-19 21:41
(3934 d 23:25 ago)

@ ElMaestro
Posting: # 10828
Views: 6,812
 

 Failing BE study due to AUCinf

Hi ElMaestro,

Fortunately I haven't encountered such problem yet but I have seen drugs with crazy AUCinfinity being reported ⇒ like T1/2 = 5000 hrs (the individual values should sig degree of variability but 5000 hrs is a bit ridiculous, if not BS) :confused:

Granted that it was a pilot study so I ended up playing with the data...

Thanks
John
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2013-06-20 19:59
(3934 d 01:07 ago)

@ jag009
Posting: # 10842
Views: 6,620
 

 BS

Hi John,

❝ […] ⇒ like T1/2 = 5000 hrs


Wow, that’s > ½ a year. ;-)

❝ […] 5000 hrs is a bit ridiculous, if not BS


If your drug didn’t belong to the Bisphosphonates, yes. They are removed by osteoclasts at an even slower rate of 10%/year.

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
jag009
★★★

NJ,
2013-06-20 22:35
(3933 d 22:31 ago)

@ Helmut
Posting: # 10844
Views: 6,603
 

 BS

Hi Helmut,

❝ ❝ […] 5000 hrs is a bit ridiculous, if not BS

❝ If your drug didn’t belong to the Bisphosphonates, yes. They are removed by osteoclasts at an even slower rate of 10%/year.


It was not. It is actually a prodrug. 70% CV, measurable concentration.

John
UA Flag
Activity
 Admin contact
22,957 posts in 4,819 threads, 1,636 registered users;
86 visitors (0 registered, 86 guests [including 7 identified bots]).
Forum time: 20:07 CET (Europe/Vienna)

Nothing shows a lack of mathematical education more
than an overly precise calculation.    Carl Friedrich Gauß

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