Dr. Harish L. Rao
☆    

India,
2009-09-10 09:53
(5314 d 06:21 ago)

Posting: # 4182
Views: 17,336
 

 BLQ values in Winnonlin [Software]

Dear Group Members,

During a PK study till 72 hours,
  1. If we get BLQ values for all time points after 12 hours,
    1. Do we need to include all time points after 12 hours in the "24.0", "36.0", "48.0" and "72.0" format before Winnonlin data analysis? If so, which AUC do we need to report (AUC0-12 or AUC0-72 or both)?
    2. In case we enter the values as BLQ in Winnonlin and define Status Code for BLQ as "Missing", does Winnonlin totally omit these time points?
    3. Can we totally delete all the timepoints after 12 hours in Winnonlin and use only the data till 12 hours?

  2. If BLQ values are found after 24 hours for 2 subjects and after 36 hours for 4 subjects,
    1. Do we need to report AUC(0-12) or AUC(0-36) or both?

  3. In another rare case, if we have BLQ values in between (eg: 8 hrs.) and before 12 hours,
    1. Can we use Status Codes and treat the timepoint as "Missing"?
    2. Can we round it off to zero (eg: 8.0) and use it in Winnonlin?

Thanks & regards,
Dr. Harish L. Rao


Edit: List formated. [Helmut]
Ohlbe
★★★

France,
2009-09-11 02:18
(5313 d 13:57 ago)

@ Dr. Harish L. Rao
Posting: # 4188
Views: 15,591
 

 BLQ values in Winnonlin

Dear Dr Harish,

❝ a. If we get BLQ values for all time points after 12 hours,

❝ 1. Do we need to include all time points after 12 hours in the "24.0", "36.0", "48.0" and "72.0" format before Winnonlin data analysis?


Including them or not will make no difference. AUC calculation will stop at the last measurable concentration for each subject.

❝ If so, which AUC do we need to report (AUC0-12 or AUC0-72 or both)?


AUC0-last (from 0 to the last concentration measured for each subject, whatever the time point).

❝ 2. In case we enter the values as BLQ in Winnonlin and define Status Code for BLQ as "Missing", does Winnonlin totally omit these time points?


BLQ is not missing: you do have a value. Not a big problem for the last sampling points, but don't replace BLQ with missing if you have a measurable concentration at a later time point.

❝ 3. Can we totally delete all the timepoints after 12 hours in Winnonlin and use only the data till 12 hours?


If all time points after 12 hours are BLQ, omitting them will make no difference.

❝ b. If BLQ values are found after 24 hours for 2 subjects and after 36 hours for 4 subjects,

❝ 1. Do we need to report AUC(0-12) or AUC(0-36) or both?


None. You have to report AUC0-last.

❝ c. In another rare case, if we have BLQ values in between (eg: 8 hrs.) and before 12 hours,

❝ 1. Can we use Status Codes and treat the timepoint as "Missing"?


BLQ is BLQ is BLQ, not missing. You do have a result.

❝ 2. Can we round it off to zero (eg: 8.0) and use it in Winnonlin?


That's even what I would expect you to do. If it is the last time point: it will not be used for AUC calculation. If there are measurable concentrations at a later time point: it will be considered as 0.

Regards
Ohlbe

Regards
Ohlbe
SDavis
★★  
Homepage
UK,
2009-09-11 10:13
(5313 d 06:01 ago)

@ Ohlbe
Posting: # 4191
Views: 29,463
 

 BLQ values in Winnonlin > Status codes tool

Hi Ohlbe,
I agree with everything above except one small point where I think Dr Rao's question needs clarification.

❝ ❝ c. In another rare case, if we have BLQ values in between (eg: 8 hrs.) and before 12 hours,

❝ ❝ 1. Can we use Status Codes and treat the timepoint as "Missing"?


❝ BLQ is BLQ is BLQ, not missing. You do have a result.


Imagine his raw data stream looks like;

[image]

If Dr Rao put this data column through WinNonlin as it stands the "<1" or any other text values will in effect be treated as missing.

At 0.25 and 8 hours this may lead to a slight over estimation in AUClast but probably a good approximation, personally it would probably be my default for an analysis if I didn't have an SOP in place before I began my Study.

If you substituted all BLQ (or <1) for Zero you now will have an underestimation of AUClast, particularly between 5 and 12 hours as the curve is 'forced' through ZERO at 8 hours. This is sometimes termed an "embedded" BLQ since there are quantifiable values before and after.

The Status code option is a Tool in WNL that creates a second column where BLQ and other text values are replaced with numeric equivalents; this option can be used to make this substitution dependent on their position in the curve. (note I would retain the original column for tabulation and presentation in the report) e.g.

Before Tmax: 0 and 2.5 above, could set to "0"

After Tmax: <this is for embedded values> i.e here the 8 hours, I personally would set it to missing, so the point is effectively ignored, to avoid underestimating AUC.

First of consecutive after Tmax: after Tlast you can if you wish apply different rules for the FIRST e.g. 14 hours, and SUBSEQUENT points.

After First of Consecutive after Tmax: rule for SUBSEQUENT points.

If asked for a recommendation I would leave both of these to "0" or missing however some people use the "FIRST" function to set some 'real number' e.g half the LoQ i.e. 14 hours to "1". Their purpose here is trying to get a more representative AUC where perhaps the LoQ is high relevant to the observed conc. and this fudge factor makes the variablitly between subjects less at perhaps the first dosing level in an ascending FIM study.

One notable warning is if you do this substitution at FIRST with anything other than "0", Missing or other TEXT value, you will have added another point to your estimation of Lz - I would suggest excluding that point in the Slope selector of WNL if you do use this method.

Dr Rao - Lastly for this and any other confusion with WNL (or other tools) - why not read the manual whilst working through examples of different datasets/profiles to come to an informed decision of the method you are happiest with, and then document that as your SOP.

Simon
Senior Scientific Trainer, Certara™
[link=https://www.youtube.com/watch?v=xX-yCO5Rzag[/link]
https://www.certarauniversity.com/dashboard
https://support.certara.com/forums/
Ohlbe
★★★

France,
2009-09-12 00:19
(5312 d 15:56 ago)

@ SDavis
Posting: # 4197
Views: 15,035
 

 BLQ values in Winnonlin > Status codes tool

Dear Simon,

❝ Imagine his raw data stream looks like;


Yes, this is always a problematic situation, and we could discuss it for hours. Treating the value at 8 hours as 0 will lead to an underestimation of the AUC. But treating it as missing will mean that you just ignore your experimental values when you don't like them. This can lead to heated discussions with regulators, which can be much less comfortable than posting on a forum.

If the next concentration is just above LLOQ, this can be due to analytical variability. But if it is much higher, I'd first try to understand what's happening: sample inversion, problem at the clinic, analytical error, anything. Even if it has no impact on this study, it can help prevent the recurrence of the problem.

IMHO it is a real pity that regulators are allergic to "PK repeats". Such repeats can help understand where the problem comes from, and they can avoid situations such as the one you describe.

Regards
Ohlbe

Regards
Ohlbe
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2009-09-14 19:11
(5309 d 21:04 ago)

@ Ohlbe
Posting: # 4199
Views: 15,214
 

 Irregular profiles

Dear Ohlbe and Simon!

❝ Yes, this is always a problematic situation, and we could discuss it for hours.


Sure. Do you have some spare time?

❝ […] treating [the value at 8 hours] as missing will mean that you just ignore your experimental values when you don't like them.


I would not call this ‘disliking’. There should be an SOP in place where data are reviewed in a blinded manner by an experienced pharmacokineticist and a request for reanalysis should be initiated. The code 'Missing' WinNonlin uses to skip a datapoint is a little bit unfortunate. If a value is really missing (e.g., vial broken), WinNonlin interpolates linear between adjacent values (little bias in the absorption phase and a positive bias in the distribution/elimination phase). Kinetica has the option to interpolate logarithmically between data points (see this post). Therefore an SOP for data imputation (regard­less the software used) is not a bad idea. In Simon’s example (values close to the LLOQ) it doesn’t make a big difference if the outcome of the study is concerned, but I have seen examples where a BQL-value (‘embedded’ in Simon’s terminology) was obtained even after repeated analysis in between two values far above the LLOQ. More on this at the end of this post.

❝ This can lead to heated discussions with regulators, which can be much less comfortable than posting on a forum.


I’m sure you are right about heated discussions. But on the other hand why did it never happen to me? Pure chance or because I always state what is planned not only in an SOP but also in the protocol? I’m guessing that a lot of troubles are originating from all types of post-hoc ‘decisions’. It’s important to start such a plausibility review if a reasonable percentage of subjects are already analysed (my SOP calls for ≥⅓). It does not make sense to initiate a repeated analysis after the first subject just to find out that profiles with a highly variable time course are ‘normal’ for this particular drug/formulation/method. Such an overview of data is related to this thread about reliability in estimating the elimination.

❝ If the next concentration is just above LLOQ, this can be due to analytical variability. But if it is much higher, I'd first try to understand what's happening: sample inversion, problem at the clinic, analytical error, anything. Even if it has no impact on this study, it can help prevent the recurrence of the problem.


100% agree. We discussed a similar point in this post. We still reanalyse not only the ‘suspect value’, but neighbouring values as well. These values serve as internal validators. If they are within some predefined limits (I don’t want to give the entire SOP here – actually the limits are dependent on the inter-day variability of the method plus a little statistics) we can be more confident in accepting or rejecting the suspect value. Reanalysing a single value (even as replicates) IMHO is a flawed concept. Why should the second analysis be ‘better’ than the first one?

In your response you stated:

❝ We used to do this in a lab where I worked some years ago, but I can't remember seeing it done at any of the labs I have visited since.


Repeating single values reminds me on lab-values. A subject shows a value outside the normal range in the post-study evaluation and a follow-up is initiated. The second one is within the normal range. Great, but why do we accept the second one as the ‘true one’? Only because we ‘like’ it more? Or is the second one by definition ‘better’ than the first one?

❝ IMHO it is a real pity that regulators are allergic to "PK repeats".


Are they? Is it because repeats are generally badly documented or do you think that there are other reasons?

We should not forget that it’s always possible that something we cannot control may have happened to a particular sample preventing a ‘regular’ result. A stabiliser was forgotten to be added to the blood sample, the blood sample was not put on ice, or even the stopper of the vacutainer was contaminated with some nasty stuff degrading our analyte. All these cases will lead to a value (even if repeated) which does not fit the profile (FDA invented the term ‘irregular profile’ for such a case). IMHO it does not make sense to stick to the experimental result if it's simply not plausible. Such a value should be dropped. Maybe Pharsight should consider a code for such a value ('Rejected' instead of 'Missing'?). Numerically it could be handled like a missing one – but it should be clear in the output that an experimental result exists.

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
Ohlbe
★★★

France,
2009-09-15 00:01
(5309 d 16:14 ago)

@ Helmut
Posting: # 4200
Views: 14,974
 

 About "PK repeats"

Dear Helmut,

❝ why did it never happen to me? Pure chance or because I always state what is planned not only in an SOP but also in the protocol? I’m guessing that a lot of troubles are originating from all types of post-hoc ‘decisions’.


Agreed.

❝ We still reanalyse not only the ‘suspect value’, but neighbouring values as well. These values serve as internal validators.


Yes, I still do like this concept.

❝ ❝ IMHO it is a real pity that regulators are allergic to "PK repeats".


❝ Are they? Is it because repeats are generally badly documented or do you think that there are other reasons?


Paranoia is the other reason, I guess. Afraid that the lab might repeat some samples so that borderline 90 % CI switch from the wrong side of the border to the right. Would take quite a few repeats for AUC, but not that many for Cmax. But then we're back to the discussion on blind review by an experienced pharmacokineticist.

❝ Maybe Pharsight should consider a code for such a value ('Rejected' instead of 'Missing'?). Numerically it could be handled like a missing one – but it❝ should be clear in the output that an experimental result exists.


Suggestion supported.

Regards
Ohlbe

Regards
Ohlbe
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2010-10-21 17:51
(4907 d 22:24 ago)

@ Ohlbe
Posting: # 6069
Views: 14,327
 

 About "PK repeats"

Dear Ohlbe,

just want to continue an old discussion. ;-)

❝ ❝ We still reanalyse not only the 'suspect value', but neighbouring values as well. These values serve as internal validators.


❝ Yes, I still do like this concept.


Found an interesting White Paper* from “The 2nd Calibration and Validation Group Workshop on recent issues in Good Laboratory Practice bioanalysis“ held 17-18 April 2008 in Montreal. See R. Phull on ‘Pharmacokinetic repeats’:

The procedure presented for performing PK repeats involved assaying the anomalous value in duplicate, so that three values are available for the sample. An additional approach presented was to assay one sample on either side of the PK repeat.

(my emphasis)
So I’m not alone – but:

Following the presentation, a comment was brought forth stating that regulatory agencies recommend against performing PK repeats; therefore, it is unwise to invest resources in the process.

In other words, don’t invest your time and money in science, but stay at the lab technician’s level of maintaining HPLC’s piping.

Plumbers of the world, unite!



  • Savoie N, Booth BP, Bradley T, Garofolo F, Hughes NC, Hussain S, King SP, Lindsay M, Lowes S, Ormsby E, Phull R, Rocci ML, Vallano PT, Viau A, Zhu Z. White Paper: The 2nd Calibration and Validation Group Work­shop on recent issues in Good Labora­tory Prac­tice bio­analysis. Bioanalysis. 2009;1(1):19–30. online

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
Ohlbe
★★★

France,
2010-10-21 18:38
(4907 d 21:37 ago)

@ Helmut
Posting: # 6070
Views: 14,034
 

 About "PK repeats"

Dear Helmut,

❝ Found an interesting White Paper from "The 2nd Calibration and Validation Group Workshop on recent issues in Good Laboratory Practice bioanalysis"


Yes, this yearly event is gaining importance and audience. Nice programme each year. Next year will be on "Recent findings from regulatory agencies" !

❝ Following the presentation, a comment was brought forth stating that regulatory agencies recommend against performing PK repeats


Unfortunately. But remember the discussions at the EBF/EUFEPS meeting in Brussels in April: on the European side there is still some hope that the ban will be restricted to BE trials only, not to other PK trials. And there is still a possibility in the draft for repeats for lab investigations. So we are not completely loosing our time.

Best regards
Ohlbe

Regards
Ohlbe
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2010-10-21 19:01
(4907 d 21:14 ago)

@ Ohlbe
Posting: # 6072
Views: 14,217
 

 About "PK repeats"

Dear Ohlbe!

❝ ❝ Following the presentation, a comment was brought forth stating that regulatory agencies recommend against performing PK repeats


❝ Unfortunately. But remember the discussions at the EBF/EUFEPS meeting in Brussels in April: on the European side there is still some hope that the ban will be restricted to BE trials only, not to other PK trials. And there is still a possibility in the draft for repeats for lab investigations. So we are not completely loosing our time.


Yes, I remember. ;-)
But I must confess that I don't see the rationale behind 'a ban on BE studies'. Much more is know about the PK of the drug at that stage as compared to FIM or PK-studies of the innovator...

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
Ohlbe
★★★

France,
2010-10-21 19:32
(4907 d 20:42 ago)

@ Helmut
Posting: # 6073
Views: 14,037
 

 About "PK repeats"

Dear Helmut,

❝ But I must confess that I don't see the rationale] behind 'a ban on BE

❝ studies'.


The rationale is not scientific, it's just paranoia. The fear is that the lab could be tempted to repeat the assay if bioequivalence is not demonstrated, particularly for Cmax, to try and get "better" results. No reason to do such a thing in BA trials when developing a new drug.

There is a lot of regulatory paranoia regarding BE trials, but one has to admit that this paranoia is fed by inspection results. Remember the results of French inspections which were presented at the EGA meeting some years ago (2007 I think): basically, critical observations in half of the dossiers inspected, including fraud or some level of data manipulation in one quarter of the dossiers inspected.

Best regards
Ohlbe

Regards
Ohlbe
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2010-10-21 20:47
(4907 d 19:28 ago)

@ Ohlbe
Posting: # 6075
Views: 14,201
 

 slowly going off-topic

Dear Ohlbe!

❝ The rationale is not scientific, it's just paranoia. The fear is that the lab could be tempted to repeat the assay if bioequivalence is not demonstrated, particularly for Cmax, to try and get "better" results. No reason to do such a thing in BA trials when developing a new drug.


Well paranoia is a reason you can’t deal with easily. Psychotherapy is not very effective treating ICD-10 F20.0, F22.0, F22.8; I’ve heard some drugs might help?

❝ There is a lot of regulatory paranoia regarding BE trials, but one has to admit that this paranoia is fed by inspection results. Remember the results of French inspections which were presented at the EGA meeting some years ago (2007 I think):


Sure; Lisbon! I’m recycling one of the slides in my lectures (remember: the nice one with the hand-drawn baseline of a low QC sample in order to ‘make’ a batch valid). I’m using the heading “Don’t try to fool inspectors!”

❝ basically, critical observations in half of the dossiers inspected, including fraud or some level of data manipulation in one quarter of the dossiers inspected.


[image] Yes, scary. But these findings should support more inspections, not preventing science. The problem with paranoiacs is that they are so afraid of getting robbed that first they don’t leave the house to go shopping and later don’ leave the bed anymore because they are afraid of breaking a leg.

To quote Dirk Barends: In BE we forget the only important person, the patient!
He/she is a living person, not just α=0.05… To the right is Foucault’s pendulum. We started seeing ‘essential similarity’ from a clinical perspective (leader of the pack: Germany) and ended up with an in vivo dissolution test (measure of pharmaceutical quality: The Netherlands). Hopefully the pendulum will continue swinging.

BTW, Cmax is such a lousy PK metric that IMHO, it should be dropped from BE assessments at all – except for drugs/formulations where there are safety/efficacy issues (I want my headache to go away – now!).

It’s time to have another beer together – maybe at one of our Captain’s preferred harbor taverns. ;-)

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
ElMaestro
★★★

Denmark,
2010-10-21 18:56
(4907 d 21:19 ago)

@ Helmut
Posting: # 6071
Views: 14,033
 

 About "PK repeats"

Dear HS,

I do have a strong opinion about this topic in general but I will spare you this time* as I promised earlier not to write lengthy posts. However...

❝ An additional approach presented was to assay one sample on either side of the PK repeat.


Does this mean on either side of the anomalous value within the time series, or it is on either side in the sample tray?

Best regards,
EM.


* Hint: Think! Audit! Stop fiddling!
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2010-10-21 20:02
(4907 d 20:13 ago)

@ ElMaestro
Posting: # 6074
Views: 14,148
 

 About "PK repeats"

Dear ElMaestro!

❝ I do have a strong opinion about this topic in general but I will spare you this time* as I promised earlier not to write lengthy posts.


Hey, you made a promise nobody asked for - I love lengthy posts. :cool:

❝ ❝ An additional approach presented was to assay one sample on either side of the PK repeat.


❝ Does this mean on either side of the anomalous value within the time series, or it is on either side in the sample tray?


The former; the reason for the repeat is triggered by a plausibility review of PK data after analytical data are released (SOP in place, :blahblah:). The idea behind is not to reanalyse only the suspected value, but neighbouring concentrations as some kind of internal validation as well. Most SOPs trigger the reanalysis by looking at the time course, e.g. value <LLOQ embedded between "normal" values,… We have some statistics behind (based on the PK of the drug and the analytical variability) defining a threshold. For example if the ratio of the suspected value > than x-times or < 1/x-times the log/lin-interpolated value of the neighbours reanalysis is triggered. One example (without the actual statistics of the analytical method included):
C(t) = 100·exp(-0.5·t) + ε*
 0 100.08     limits
 1  59.37
 2  38.09
 3  21.11
 4  15.28
 6   7.13
 8   1.80   1.13-18.93
12   2.25


[image]Let's assume the suspected value is at 8 hours. The [6-12] log/lin interpolated value at 8 hours is 4.85, first we apply limits of 4.85/3 and 4.85×3 [1.62 - 14.56] and then allow for variability of the method 1.62×0.7, 14.56×1.3, getting [1.13 - 18.93].

A value outside these limits at 8 hours will trigger reanalysis.

If the neighbouring values in the repeated analysis are within the expected variability (rules similar to ISR are applicable), we can be pretty sure that the method performs well. If the suspected value is confirmed, use the original (and cross your finger for the PK analysis), if not, something happened in the analytical (or clinical) performance we can't explain. That's the problem with allowing reanalysis only for analytical reasons: It assumes that all potential causes of errors are noticed and documented. In other words, sample mix-up in the clinical part (yes, it happens!) can never trigger a confirmatory reanalysis, because no documented errors occured in the analytical part.

If we see bioanalytics isolated from the fact that we have information about PK of the drug we step into the boots of analytical chemists working in the field of environmental analysis. They have a single sample and no clue how the analyte entered the matrix. But we should know better. If we have a value which is 10times higher than adjacent values it's simply physiologically not possible. Just my 2¢.


  • µ = 0, σ² = 2: the example looks like a two-compartment model by chance (w=1/y² iterative reweighting; 1-comp AIC 4.65, 2-comp AIC -1.78). The red line is the elimination from NCA and the blue ones from modeling. So far about simulations. ;-)

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
Dr. Harish L. Rao
☆    

India,
2009-09-15 07:45
(5309 d 08:30 ago)

@ Helmut
Posting: # 4203
Views: 14,990
 

 BLQ Values in Winnonlin

Dear Ohlbe, Simon & Helmut,

Thanks a lot for all your valuable inputs.

Best regards,
Harish
UA Flag
Activity
 Admin contact
22,957 posts in 4,819 threads, 1,636 registered users;
129 visitors (0 registered, 129 guests [including 6 identified bots]).
Forum time: 15:15 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