- On 29 Sep 2004 at 14:36:24, "Harshu Chaobal" (harshu.-at-.u.washington.edu) sent the message

Back to the Top

Hi All,

I am comparing AUCs for subjects who each were in 4 treatment groups on

separate days - enzyme induced, enzyme inhibited, selective intestinal

inhibition, and controls. My sample times are the same for each enzyme

state, and using a noncompartmental analysis (NCA), I am extrapolating

approximately 30% of the AUCinf in the enzyme inhibited group. This

appears

to be giving me inflated AUCs, however when I deliberately limit my

data in

the NCA such that the AUC derived from the controls and other enzyme

states

is equally extrapolated (about 30%), I do not get inflated AUCs. The

last

sample time for the inhibited session is approximately equal to the

inhibited t1/2, and I limited the other enzyme state data to samples at

or

below the t1/2 as a test if the high % AUC extrapolated is the

explanation

for the inflated AUCs. This does not appear to be the case. I would

expect

the high % AUC extrapolated to result in more variance though not

necessarily bias, provided the kinetics are linear (they are). And why

did

the test run of limited data on the other enzyme data not produce the

same

bias? I believe the inhibited state AUCs to be inflated because when I

plot

all the AUC data (all 4 sessions) against a plasma concentration at a

single

sample time, all but the inhibited data falls on a line. Any thoughts? - On 29 Sep 2004 at 15:28:51, "Gordi, Toufigh" (Toufigh.Gordi.-a-.cvt.com) sent the message

Back to the Top

Dear Harshu,

I am a little confused about your statement that the kinetics of the

compound is linear on all the occasions, despite the fact that you

inhibit or induce the metabolism and thereby clearance, or manipulate

the absorption and thus bioavailability. Could you explain why you

think the kinetic is linear and what your criterias are for a linear or

non-linear kinetic? What are the relative changes in the AUCs compared

to the control group?

Toufigh Gordi - On 30 Sep 2004 at 08:47:10, "J.H.Proost" (J.H.Proost.aaa.farm.rug.nl) sent the message

Back to the Top

Dear Harshu Chaobal,

You wrote:

>I am comparing AUCs for subjects who each were in 4 treatment groups on

>separate days - enzyme induced, enzyme inhibited, selective intestinal

>inhibition, and controls. My sample times are the same for each enzyme

>state, and using a noncompartmental analysis (NCA), I am extrapolating

>approximately 30% of the AUCinf in the enzyme inhibited group. This

appears

>to be giving me inflated AUCs, however when I deliberately limit my

data in

>the NCA such that the AUC derived from the controls and other enzyme

states

>is equally extrapolated (about 30%), I do not get inflated AUCs.

IHMO, your reasoning is incorrect:

1) The term 'inflated AUC' suggests that you assume that

the calculated value of AUC is overestimated due to the

extrapolation. If you have concerns that AUC is 'inflated'

by extrapolation, you cannot simply omit data points from

the other groups to get the same level of extrapolation.

You will not correct the inflated AUC, but, at best,

inflate the AUC in the other groups to the same level. But

in real practice it is more likely that you will increase

the error in your final results. You should focus on

improving the extrapolation of the 'problem' data, e.g. by

a more sophisticated way of estimating the terminal rate

constant.

2) 'Deliberately' omitting data sound like filling your

car with water instead of petrol. This will never work.

One should only omit data if there is a serious reason why

the data are not correct, or, perhaps, if the data do not

contribute to your results. But omitting data to increase

the degree of extrapolation is definitely bad practice.

Best regards,

Hans Proost

Johannes H. Proost

Dept. of Pharmacokinetics and Drug Delivery

University Centre for Pharmacy

Antonius Deusinglaan 1

9713 AV Groningen, The Netherlands

tel. 31-50 363 3292

fax 31-50 363 3247

Email: j.h.proost.-a-.farm.rug.nl

[Are the clearance values the same across groups? Has this be adjusted

for? How about truncated AUCs (or is this dropping data ;-) - db] - On 25 Dec 2004 at 08:01:35, "patelia bipin dahyalal" (bippatel.aaa.rediffmail.com) sent the message

Back to the Top

Hi, I have a query for calculating of AUC0-inf

This is regarding calculating of AUC0-inf, during a bioequivalence

study with 36 hours time point as last sample; the concentration of

last sample (36 hours) is 0.00 ng/mL. What would be my last measurable

concentration for calculating AUC0-inf? Is it 0.00 or the concentration

found at 24.00 hours (second last sample)? In other words is AUC0-t and

AUC0-inf the same if the last sample concentration found is 0.00?

Thanks in advance

Bipin Patel - On 27 Dec 2004 at 23:36:42, "Sima Sadray" (sadrai.-at-.sina.tums.ac.ir) sent the message

Back to the Top

The following message was posted to: PharmPK

Dear Bipin,

AUC (total) = AUC1+AUC2+AUC3+....AUCn

in your example

AUCn= (C(24)+C(36)/ 2) * (t (24) - t (36)) = ( (C(24) + 0)/2) * 12=

C(24) *

6

If your limit of quantification (LOQ) is well validated.

Dr Sadray - On 27 Dec 2004 at 09:06:16, "Porzio, Stefano" (Stefano.Porzio.-at-.ZambonGroup.com) sent the message

Back to the Top

In reply to

Hi, I have a query for calculating of AUC0-inf

This is regarding calculating of AUC0-inf, during a bioequivalence

study with 36 hours time point as last sample;

[SP] AS A RULE YOU MUST CONSIDER AS "LAST POINT" THE LAST QUANTIFICABLE

SAMPLE, I.E. THE LAST SAMPLE ABOVE LLQ (LOWER LIMIT OF QUANTIFICATION).

the concentration of[SP] last sample (36 hours) is 0.00 ng/mL. What

would be my last measurable concentration for calculating AUC0-inf? Is

it 0.00 or the concentration found at 24.00 hours (second last sample)?

In other words is AUC0-t and AUC0-inf the same if the last sample

concentration found is 0.00?

[SP] OBVIOUSLY AUC-t IS CALCULATED FROM TIME 0 TO THE TIME OF LAST

POINT >ABOVE LLQ AND AUC-inf IS CALCULATED AS AUC-t+(EXTRAPOLATED AREA

FROM LAST >POINT ABOVE THE LLQ AND INFINITY)

BEST REGARDS

STEFANO PORZIO

Pharmacokinetic and Tox. Dept.

Inpharzam Ricerche SA - ZAMBON-GROUP

Taverne - Switzerland - On 28 Dec 2004 at 12:26:53, =?ISO-8859-1?Q?Helmut_Schutz?= (helmut.schuetz.-a-.bebac.at) sent the message

Back to the Top

The following message was posted to: PharmPK

Dear Bipian, Sida, and Stefano,

although statistics deals with probabilities, in PK we can be sure

(p=1),

that one particular concentration simply does not exist: zero ;-)

I would suggest to follow Stefano's advice, and calculate AUC until

the time point of the last quantifiable concentration by NCA-methods

(whatever you have stated in the protocol: in decreasing sequence of

regulatory acceptance - linear trapezoidal, log/linear trapezoidal,

Splines, Lagrange interpolation) and extrapolate from this time point

to infinity.

In the past there was a good deal of discussion around about the

"correct" calculation of AUC in NCA. It does make sense to include

the "absorption triangle" (since we know, that the pre-dose

concentration is sufficiently close to zero), but the "elimination

triangle" leading to something called "AUC-total" or "AUC-all"

simply does not make sense.

Please have a look at http://www.boomer.org/pkin/pk/AUCinf.jpg.

Example:

A drug with a half life of 12h, the last sampling time points

16h, 24h, and 36h.

Now imagine an infinitesimal difference in ka (or tlag) between

formulations (within the same subject). In one case you will get

a concentration at 24h just slightly above LOQ, whereas in the

other case you will not be able to quantify that concentration.

In the first case you will add a last "triangle" from 24h to 36h

(the "true" concentration at 36h is about LOQ/2, not zero!),

whereas your last triangle in the second case will cover 16h to 24h

(the "true" concentration at 24h is just slightly below LOQ, not

zero...).

best regards,

Helmut

--

Helmut Schutz

BEBAC

Consultancy Services for Bioequivalence and Bioavailability Studies

Neubaugasse 36/11

A-1070 Vienna/Austria

tel/fax +43 1 2311746

http://BEBAC.at Bioequivalence/Bioavailability Forum at

http://forum.bebac.at

http://www.goldmark.org/netrants/no-word/attach.html - On 28 Dec 2004 at 16:57:33, otilia.hemma.-a-.tele2.se sent the message

Back to the Top

The following message was posted to: PharmPK

Dear Helmut

Even though

C(36)=LOQ/2=almost 0

the AREA of the last triangle can be whatever

C(24)x(36-24)/2

and may be a value not negligible. Or is there something I dont get

right?

Best regards

Otilia Lillin

MSc Scientific Computing - On 29 Dec 2004 at 10:19:59, "Porzio, Stefano" (Stefano.Porzio.aaa.ZambonGroup.com) sent the message

Back to the Top

DEAR OTLILLA,

I SUSPECT YOU HAVE NO EXPERIENCE TO CALCULATE NC PARAMETERS!

IF YOU LIKE YOU CAN FOLLOW THESE SUGGESTIONS:

1. PLACE THE PROFILE IN SEMILOG GRAPHIC I.E. LOG (CONECENTRATION)

VS. TIME

2. VERIFY THAT A MINIMUM OF 3 POINTS ARE AVAILABLE TO APPROXIMATE A

STRAIGHT-LINE IN THE LAST PART OF YOUR PROFILE (YOU MUST DEFINE THE

LAST POINT AS THE FAREST POINT ABOVE LOQ, THEN GET BACK TO OTHER

POINTS)

3. DESIGNE, MANUALLY OR BY LINEAR REGRESSION, A REGRESSION LINE IN

SEMILOG SCALE FOR SELECTED POINTS (A MINIMUM OF 3 POINTS ARE USUALLY

REQUIRED TO DEFINE AN ACCEPTABLE REGRESSION)

4. CALCULATE THE SLOPE OF THIS STRAGHT LINE (EQUATION IS

Concentration = -lambda x time + constant)

5. USE lamda TO CALCULATE THE AUC FROM LAST POINT TO INFINITY AS

[CONCENTRATION OF THE LAST POINT]/lamda

6. AS [CONCENTRATION OF THE LAST POINT]YOU CAN CONSIDER THE

[EXPERIMENTAL CONCENTRATION] OR THE [CALCULATED CONCENTRATION] AS BY

YOUR REGRESSION STRAIGHT LINE (I.E. [CALCULATED CONCENTRATION] =

-lambda x [LAST TIME]+ constant)

7. CALCULATE THE AUC 0-inf AS THE SUM OF AUC 0-last + AUC last-inf

BEST REGARDS

STEFANO PORZIO

INPHARZAM RICERCHE S.A.

ZAMBONGROUP

TAVERNE - SWITZERLAND - On 31 Dec 2004 at 13:09:13, =?ISO-8859-1?Q?Helmut_Schutz?= (helmut.schuetz.-a-.bebac.at) sent the message

Back to the Top

The following message was posted to: PharmPK

Dear Otilla,

you wrote:

>Even though,

>C(36)=LOQ/2=almost 0

>the AREA of the last triangle can be whatever

>C(24)x(36-24)/2

>and may be a value not negligible.

>

I am afraid, you first line is not correct, since:

C(36) # LOQ/2 # 0!

We simply do not know the value of C(36);

the only thing we do know is:

LOQ > C(36) > 0.

It does not make sense to put so much effort in validating

the analytical method (and we are satisfied with 20% precision

and 20% inaccuracy at LOQ!), and then rather arbitrarily

setting the first value after the last quantifiable to LOQ/2

or - even worse - to zero.

Best regards,

Helmut

--

Helmut Schutz

BEBAC

Consultancy Services for Bioequivalence and Bioavailability Studies

Neubaugasse 36/11

A-1070 Vienna/Austria

tel/fax +43 1 2311746

http://BEBAC.at Bioequivalence/Bioavailability Forum at

http://forum.bebac.at

http://www.goldmark.org/netrants/no-word/attach.html

Want to post a follow-up message on this topic? If this link does not work with your browser send a follow-up message to PharmPK@boomer.org with "AUC calculation" as the subject

PharmPK Discussion List Archive Index page

Copyright 1995-2010 David W. A. Bourne (david@boomer.org)