Log in
Register|
boonchai_l ☆ Thailand, 2010-03-02 06:22 (5542 d 07:02 ago) (edited on 2010-03-02 07:23) Posting: # 4849 Views: 6,812 |
|
|
Dear all I would like to know which method you usually use for estimating the AUCt in the BE industry (not for BA or PK study)? Why? Is there any guidance suggest about this? For me, I usually use linear trapezoidal in non-compartmental model. In my country (I really don't want to tell you which country), the regulator suggests we should use log-linear trapezoidal rule. I don’t think it is suitable because the log-linear you have to check the assumption to avoid a big error. While the other one is no need. However, for both of them, you have to design suitable sampling time. The major objective of BE study is to proof whether two formulations are equivalent, while the accuracy of PK parameters is not major objective. If the accuracy of PK parameters is the most important I think we should change model to compartmental or others method. At least I think the way to compromise is using linear in ascending and log-linear in descending. Regards, BL Edit: Category changed. [Helmut] |
|
Helmut ★★★   Vienna, Austria, 2010-03-03 23:14 (5540 d 14:10 ago) @ boonchai_l Posting: # 4858 Views: 5,731 |
|
|
Dear Boonchai! ❝ I would like to know which method you usually use for estimating the AUCt in the BE industry (not for BA or PK study)? Lin-up/log-down. ❝ Why? Should be the least biased one. ❝ Is there any guidance suggest about this? I don't think so. See this rather lengthy thread. ❝ For me, I usually use linear trapezoidal in non-compartmental model. Why not? It was shown in simulations that a well-chosen design (e.g., with respect to sampling times) is more important than other factors. ❝ In my country (I really don't want to tell you which country),... OK, respected (well - your name is telling...). But it would be really interesting for others in study planning. A web-forum should not be a one-way road. The policy states: Some problems may be related to a national regulation; in such a case please include the target country. ❝ ... the regulator suggests we should use log-linear trapezoidal rule. Maybe they follow WHO's suggestions. ❝ I don't think it is suitable because the log-linear you have to check the assumption to avoid a big error. Which assumption? ❝ While the other one is no need.  OK, a few words about theoretical bias in AUC estimation:
❝ However, for both of them, you have to design suitable sampling time. The major objective of BE study is to proof whether two formulations are equivalent, while the accuracy of PK parameters is not major objective. You are right. If profiles are superimposible, it wouldn't matter at all. But in the real world there's a lot of variability in PK (between and within subjects) - the sampling schedule may lead to different biases for test and reference in the same subject. ❝ If the accuracy of PK parameters is the most important I think we should change model to compartmental or others method. At least I think the way to compromise is using linear in ascending and log-linear in descending. Disagree concerning modeling (which is not acceptable in BE anyhow). I would use the method which - based on theory - is the least biased. Yes, that's lin-up/log-down. — Dif-tor heh smusma 🖖🏼 Довге життя Україна! ![[image]](https://static.bebac.at/pics/Blue_and_yellow_ribbon_UA.png) Helmut Schütz ![[image]](https://static.bebac.at/img/CC by.png) The quality of responses received is directly proportional to the quality of the question asked. 🚮 Science Quotes |
|
d_labes ★★★ Berlin, Germany, 2010-03-04 11:36 (5540 d 01:48 ago) @ Helmut Posting: # 4859 Views: 6,595 |
|
|
Dear Boonchai, dear Helmut, ❝ ❝ Is there any guidance suggest about this? Here a (surly incomplete) survey I have done some times ago. Surely its not totally up to date. EMEA - old but yet in effect CPMP/EWP/QWP/1401/98 (2001): point 3.3 Characteristics to be investigated "The method of estimating AUC-values should be specified." - new CPMP/QWP/EWP/1401/98 Rev. 1 (2010) point 4.1.5 Characteristics to be investigated Pharmacokinetic parameters "Non-compartmental methods should be used for determination of pharmacokinetic parameters." EMEA veterinary EMEA/VPMP/016/00-corr-FINAL (2001) point 9.1 Characteristics to be analysed "The AUC t-last up to the last measured time (t-last) corresponding to a c-last exceeding the limit of quantification (LOQ) should be calculated using the linear trapezoidal method. If sampling times are spaced during the elimination phase, logarithmic trapezoidal or other method must be considered. The method of AUC calculation must be qualified; extrapolation should preferably be avoided." FDA Guidance for industry: Bioavailability and bioequvalence studies for orally administered drug products – General considerations (2003) point III:A:8 Total exposure "Area under the plasma/serum/blood concentration-time curve from time zero to time t (AUC0-t), where t is the last time point with measurable concentration for individual formulation." No method named but: FDA 1992 Under point II. General Methodology "a. Area under the plasma/blood concentration - time curve from time zero to time t (AUC0-t), calculated by the trapezoidal rule, where t is the last measurable time point." Canada Conduct and Analysis of Bioavailability and Bioequivalence Studies - Part A: Oral Dosage Formulations Used for Systemic Effects (1992) - point 7.3 Pharmacokinetic Parameters "AUCT : Area under the concentration-time curve measured to the last quantifiable concentration, using the trapezoidal rule." - point 9 Glossary of Terms "AUCT (AUC TO THE LAST QUANTIFIABLE CONCENTRATION) - This describes the AUC to the time of the last quantifiable concentration. AUCT is calculated from observed data at specific time points by the linear trapezoidal rule." WHO WHO Expert Committee on Specifications for Pharmaceutical Preparations, Fortieth Report (WHO Technical Report Series No. 937) (2006) point 6.6.4 Prameters to be assessed "The method of calculating AUC-values should be specified. In general AUC should be calculated using the linear/log trapezoidal integration method ..." Brazil Resolution - RE n. 896, of May 29, 2003 GUIDE FOR RELATIVE BIOAVAILABILITY / BIOEQUIVALENCE TESTS OF DRUG PRODUCTS point 3.1 General methodology "3.1.2.1. the area under the blood concentration versus time, calculated using the method of the trapezoids, from time zero to time t (ASC0-t), where t is the time related to the last concentration determined through experimentation" Japan Guideline for Bioequivalence Studies of Generic Products December 22, 1997 point 2.2 Parameters to be assessed "... AUC is calculated using the trapezoidal integration method ..." India GUIDELINES FOR BIOAVAILABILITY & BIOEQUIVALENCE STUDIES (2005) under PHARMACOKINETIC TERMS: "AUC0-t:Area under the plasma concentration-time curve from 0 h to the last quantifiable concentration to be calculated using the trapezoidal rule." Underlined by me. If we read FDA, Brazil, Japan, India as "linear" trapezoidal method there is clearly a regulatory majority for that. (I suspect that according to the ElMaestrophylistic axiom "The simpler the regulatory" we must read it as "linear"  ).BTW: Boonchai ❝ In my country (I really don’t want to tell you which country) ...  .— Regards, Detlew |
|
Helmut ★★★ Vienna, Austria, 2010-03-04 15:51 (5539 d 21:33 ago) @ d_labes Posting: # 4860 Views: 5,653 |
|
|
Dear D Labes, THX for your great overview! ❝ EMEA veterinary ❝ "The AUC t-last up to the last measured time (t-last) corresponding to a c-last exceeding the limit of quantification (LOQ) should be calculated using the linear trapezoidal method. ❝ If sampling times are spaced during the elimination phase, logarithmic trapezoidal or other method must be considered. […]" How does the second sentence fit to the first one? Anyhow, the current draft (March 2009) doesn't mention a specific method any more. — 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 |
|
boonchai_l ☆ Thailand, 2010-03-04 23:32 (5539 d 13:52 ago) @ Helmut Posting: # 4862 Views: 5,760 |
|
|
Dear HS, dear D Labes, First, I have to say thank you very much for your replies. They make me more understand. The main objective of my questions is that I think method selection for AUCt estimation is quite subjective and I feel moody that the Thai regulators don’t let it go. Moreover, they advise the log-linear one instead of to be the mixture of two methods. It makes me more moody. Surely, linear trapezoidal method is suitable for linear kinetic and if you use this method with non-linear kinetic, the AUCt will be overestimated. The log-linear trapezoidal method can reduce the value of AUCt but no one knows whether it will reduce the value until underestimate because no one knows the true value. AUCt from this method is a sum of each piece (area) which each piece is estimated from only 2 point (see also the log-linear formula). It is not concern the overall figure how the curve is steep or concave or it maybe mono or poly exponential curve. If it is the poly-exponential, this technique may produce large error. And all these I call them “assumption”. However, Gabrielsson state in his textbook “… the difference between the two methods is negligible as long as a reasonable sampling design …” or even many guidance (big thanks for your overview), they quite open to select method but no one state log-linear trapezoidal should be select in both of up and down. By the way, sorry that I didn’t tell the country name at first. Best regards, BL |
Ing. Helmut Schütz