Vishal S
☆

India,
2021-08-02 09:20
(173 d 13:44 ago)

Posting: # 22491
Views: 1,619

partial AUC [NCA / SHAM]

Hiii

In calculation of terminal partial AUC (AUC cut off t-tlast) specified as per EMA guideline for MR formulations
in such cases what is meaning of tlast?

is this last time of blood collection or its last time of dosing interval?

for example-In study blood collection is 72.00 hrs, However dosing interval of Drug is 24.00 hrs then early partial AUC will be AUC(0-12 hrs) & what would be the Terminal Partial AUC?

Thanks and Reagrds
Vishal S

Helmut
★★★  Vienna, Austria,
2021-08-02 12:53
(173 d 10:11 ago)

@ Vishal S
Posting: # 22492
Views: 1,335

partial AUCs in SD

Hi Vishal,

» In calculation of terminal partial AUC (AUC cut off t-tlast) […] in such cases what is meaning of tlast?
» is this last time of blood collection or its last time of dosing interval?

By convention, $$t_\textrm{last}$$ is the time point of the last measurable concentration. There is no relationship with the last sampling time point. Sometimes in a SD study (issues with the LLOQ), $$t_\textrm{last}$$ can be earlier than late sampling time point(s). If you have relevant accumulation, it rarely happens after multiple doses.

» for example-In study blood collection is 72.00 hrs, However dosing interval of Drug is 24.00 hrs then early partial AUC will be AUC(0-12 hrs) & what would be the Terminal Partial AUC?

I guess you are are talking about SD studies (fasting and fed) and hope to get a waiver for the multiple dose study (‘no risk of accumulation’)? Sampling should be sufficiently long (in your case 72 h) to get a reliable estimate of $$\small{\lambda_\textrm{z}}$$ (you need it for $$\small{AUC_{0-\infty}}$$). The late cut-off is the intended dosing interval $$\small{\tau}$$ (in your case 24 h). If you have no justification for another cut off time, the GL suggests $$\small{\tau}/2$$. Hence you need these PK metrics:
1. $$\small{AUC_{0-t_\textrm{last}}}$$
2. $$\small{AUC_{0-\infty}}$$
3. $$\small{AUC_{0-\tau}}$$
4. $$\small{_\textrm{partial}AUC_{0-\textrm{cut off}}}$$
5. $$\small{_\textrm{partial}AUC_{\textrm{cut off}-{\tau}}}$$
• The first two are required in all SD studies of MR products.
• If $$\small{100\times AUC_{0-\tau}/AUC_{0-\infty}\geq 90\%}$$ and you pass BE for all PK metrics, the MD study can be waived and you’re done.
• If you fail at least one of the partial $$\small{AUC\textrm{s}}$$ and pass both $$\small{AUC_{0-t_\textrm{last}}}$$ and $$\small{AUC_{0-\infty}}$$, the SD-part is done and you continue with the MD study. There you have to demonstrate BE only for $$\small{AUC_{0-\tau}}$$.
Once you succeed in the MD study, failing $$\small{AUC\textrm{(s)}}$$ in the SD study don’t matter (they were just an option to waive the MD study): You passed the required PK metrics both in the SD study $$\small{(AUC_{0-t_\textrm{last}}}$$ and $$\small{AUC_{0-\infty}}$$) and in the MD study $$\small{(AUC_{0-\tau})}$$.

Note that waiving the MD study works only for products with a fast $$\small{t_{1/2}}$$. Any $$\small{t_{1/2}>\sim 3\,\textrm{h}}$$ (esp. controlled release with flip-flop PK) practically never works.
If you have to deal with a multiphasic release product, you have to (additionally to $$\small{C_\textrm{max})}$$ demonstrate BE of the maximum concentrations in each phase. For two phases you have to show BE for 7 (seven!) PK metrics. Good luck!

Of note, I suggest to perform the fed study first because all too often it is more difficult than fasting (products can have different food effects). I’ve seen many cases where fasting passed BE (sponsor happy) only to see the fed study terribly failing. Ended in a re-formulation and repeating both (this time in the ‘right’ order).

Dif-tor heh smusma 🖖
Helmut Schütz The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
Relaxation
★

Germany,
2021-08-02 13:36
(173 d 09:28 ago)

@ Helmut
Posting: # 22493
Views: 1,256

partial AUCs in SD

» 5. $$\small{_\textrm{partial}AUC_{\textrm{cut off}-{\tau}}}$$

This one I think should be: $$\small{_\textrm{partial}AUC_{\textrm{cut off}-t_\textrm{last}}}$$?
Vishal S
☆

India,
2021-08-02 13:41
(173 d 09:23 ago)

@ Relaxation
Posting: # 22495
Views: 1,264

partial AUCs in SD

» » 5. $$\small{_\textrm{partial}AUC_{\textrm{cut off}-{\tau}}}$$
»
» This one I think should be: $$\small{_\textrm{partial}AUC_{\textrm{cut off}-t_\textrm{last}}}$$?

Hii Helmut

Thank you for above information
as per your suggestion the partial AUCs will be AUC(0-12 hrs) & AUC12-24 hrs (Considering dosing interval of 24 hrs)
what happen if we select partial AUCs ie AUC(0-12 hrs) & AUC12-72 hrs (Considering last sample collection time point ie 72 hrs) & prove the bioequivalence is this accepted by regulatory?
Helmut
★★★  Vienna, Austria,
2021-08-02 14:05
(173 d 08:59 ago)

@ Vishal S
Posting: # 22497
Views: 1,223

partial AUCs in SD

Hi Vishal,

» as per your suggestion the partial AUCs will be AUC(0-12 hrs) & AUC12-24 hrs (Considering dosing interval of 24 hrs)

Yep. See my reply to Relaxation’s post.

» what happen if we select partial AUCs ie AUC(0-12 hrs) & AUC12-72 hrs (Considering last sample collection time point ie 72 hrs) & prove the bioequivalence is this accepted by regulatory?

I’m not an assessor. If you have no missings, a sufficient LLOQ (all late concentrations measurable), and if the assessor is a ‘tick-the-box type’ (since $$\small{_\textrm{partial}AUC_{\textrm{cut off}-t_\textrm{last}}}$$ is stated in the GL), all is good.
However, from a PK perspective (‼) it doesn’t make sense.

Dif-tor heh smusma 🖖
Helmut Schütz The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
Helmut
★★★  Vienna, Austria,
2021-08-02 13:56
(173 d 09:08 ago)

@ Relaxation
Posting: # 22496
Views: 1,270

Hi Relaxation,

» » 5. $$\small{_\textrm{partial}AUC_{\textrm{cut off}-{\tau}}}$$
»
» This one I think should be: $$\small{_\textrm{partial}AUC_{\textrm{cut off}-t_\textrm{last}}}$$?

According to the GL, yes. However, doesn’t make sense to me.1 In trying to get a waiver for the MD study I’m interested in showing BE for the partial $$\small{AUC\textrm{s}}$$, which are – hopefully – predictive of the clinical situation (multiple doses administered with $$\small{\tau}$$). If you have issues with the LLOQ in the SD study or missing sample(s) in the late part of the profile, already $$\small{AUC_{0-t_\textrm{last}}}$$ [sic] is a pile of poo. If in a subject $$t_\textrm{last(T)}\neq t_\textrm{last(R)}$$, that’s comparing apples with oranges (though similar by weight and shape, extremely different by smell, taste, touch, and texture).
IMHO, it’s high time to abandon $$\small{AUC_{0-t_\textrm{last}}}$$ in all guidelines (because biased) and move forward to the – always (‼) unbiased – $$\small{AUC_{0-t_\textrm{last(Common)}}}$$.2

Confession: In my studies of multiphasic products I chickened out and used still $$\small{AUC_{0-t_\textrm{last}}}$$ (I knew that I won’t have problems with BQLs and wanted to make assessors happy) but $$\small{_\textrm{partial}AUC_{\textrm{cut off}-{\tau}}}$$. Were accepted without problems.

1. Remember Henning’s credo of ‘Science-based Regulations’!
2. Fisher D, Kramer W, Burmeister Getz E. Evaluation of a Scenario in Which Estimates of Bioequivalence Are Biased and a Proposed Solution: tlast (Common). J Clin Pharm. 2016; 56(7): 794–800. doi:10.1002/jcph.663. free resource.

Dif-tor heh smusma 🖖
Helmut Schütz The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes Ing. Helmut Schütz 