Bioequivalence and Bioavailability Forum • Bioequivalence decision affected by ISR?

Dr_Dan
★★

Germany,
2012-12-21 10:52
(4541 d 13:00 ago)

Posting: # 9739
Views: 9,328

Bioequivalence decision affected by ISR? [Bioanalytics]

Dear all
In order to reduce the variability in outcome in bioequivalence studies all samples of one subject together are analysed in one analytical run (including calibration standards, QC samples, and study samples). The samples are handled at the same time, i.e. subsequently processed without interruption in time and by the same analyst with the same reagents under homogeneous conditions. The QC samples are divided over the run in such a way that the accuracy and precision of the whole run is ensured. So, if anything happens during bioanalysis (instrument issues, scientist performance of method, metabolite interferences, back conversion of metabolites, poor ruggedness, internal standard response, matrix effects) which could have been detected by ISR this would affect the samples derived from test treatment in the same way as the samples derived from reference treatment (at least in studies with a cross-over design), right? In consequence, even if the absolute concentration determined in the samples is not correct the relative difference of test and reference i.e. the ratio for AUC and Cmax should be the same. Therefore I argue that the bioequivalence decision based on a validated bioanalytical method will not be affected by the (missing) results of ISR.
Do you agree with these considerations?
I am looking forward to your valuable thoughts.
Kind regards
Dan

—
Kind regards and have a nice day
Dr_Dan

Ohlbe
★★★

France,
2012-12-21 11:30
(4541 d 12:21 ago)

(edited on 2012-12-21 15:13)
@ Dr_Dan
Posting: # 9740
Views: 8,419

Bioequivalence decision affected by ISR?

Post reply

Dear Dan,

❝ Do you agree with these considerations?

No, sorry I don't...

If the reason for ISR failure is metabolite back-conversion, I have two concerns:

metabolites can have a lower intra-CV than the parent (I have seen a few examples; clopidogrel, well-known for metabolite back-conversion, is one of them). If what you measure is a mix of parent and metabolite, you will get a lower intra-CV and increase your chances to demonstrate BE;
according to the BE guideline, metabolites are less sensitive to detect differences in formulation than the parent. I have seen a nice example too, where we had differences in dissolution profiles between the test and the reference; bioequivalence was demonstrated for the metabolite but not for the parent. Again if what you measure is a mix of parent and metabolite, you could get a point estimate closer to 1.

Even if ISR fails for another reason than metabolite back-conversion, I always had problems with justifications like "my method is just as bad for the reference as for the test, so there should be no impact in the end". Sorry !

I also have problems to say that an increase in the analytical variability can only decrease chances to demonstrate bioequivalence. You can't exclude that it will result in a point estimate closer to 1, or in a slightly decreased intra-CV (difficult to predict the influence of a random effect, isn't it ?). It would be great to run some sims:

take a BE trial with 24 subjects, borderline failing for Cmax (let's say 90 % CI 79-98 %). Consider these Cmax values as "true values".
add some analytical variability to the Cmax values (CV 10 %, 15 %, 20 %, 25 %, 30 %)
see how often the study passes in the end (just the 5 % alpha, or more ?). How often does the point estimate get closer to 1 and by how much, and how often is the intra-CV reduced.

To be honest I don't expect much changes for AUC, but Cmax could be more problematic.
I'm afraid I just don't know how to fire such sims. If any of our sims-specialists gets bored during the Christmas break ;-)

Regards
Ohlbe

—
Regards
Ohlbe

Dr_Dan
★★

Germany,
2012-12-21 16:31
(4541 d 07:20 ago)

@ Ohlbe
Posting: # 9743
Views: 8,238

Bioequivalence decision affected by ISR?

Post reply

Dear Ohlbe
Sorry, but I do not like/agree with your reply because you use clopidogrel as an unfair example to justify ISR and in the second part you argue with borderline results (which are out of question) in combination with back-conversion. In case of clopidogrel the amount of the metabolite within the systemic circulation is about 5000-fold! higher than of the parent compound. This is rather unique, right? In this case I agree that ISR is demanded by the Q&A document. However, a conversion of the main metabolite to clopidogrel is chemically an esterification step. Esterifications are bound to different conditions, e.g. presence of an organic acid and an alcohol as well as higher temperature. If within the sample preparation and the chromatography neither the extractions procedure and the final extract nor the mobile phase does contain the relevant molecules e.g. methanol as alcohol for esterification then you could exclude metabolite back-conversion.
What are your objections against the following hypothesis: In case the amount of metabolite(s) is not so high as in above given example and you can exclude back-conversion by the method used* and your 90% CIs are well within the acceptance range than an increase in the analytical variability can only decrease chances to demonstrate bioequivalence.
I am looking forward to your and Other reply(ies)
Kind regards
Dan

* In this case your argument that the lower CVintra of the metabolite will affect/decrease the CVintra of the parent drug does not apply

—
Kind regards and have a nice day
Dr_Dan

Ohlbe
★★★

France,
2012-12-21 20:12
(4541 d 03:40 ago)

(edited on 2012-12-21 22:08)
@ Dr_Dan
Posting: # 9744
Views: 8,222

Bioequivalence decision affected by ISR?

Post reply

Dear Dan,

❝ you use clopidogrel as an unfair example to justify ISR (...) the amount of the metabolite within the systemic circulation is about 5000-fold! higher than of the parent compound. This is rather unique, right?

Agreed, that's an extreme example. There are other examples of molecules with extreme differences between parent and metabolite, but I'm not aware of any back conversion with them.

❝ However, a conversion of the main metabolite to clopidogrel is chemically an esterification step.

More precisely, it seems to be a transesterification of the glucuronide of the carboxyacid metabolite.

❝ If within the sample preparation and the chromatography neither the extractions procedure and the final extract nor the mobile phase does contain the relevant molecules e.g. methanol as alcohol for esterification then you could exclude metabolite back-conversion.

Yes, perfectly logical. But have a look at this poster from ASMS 2009. I know, it's clopidogrel again. But it looks like strange things can happen on some SPE columns, in that case even without methanol.

❝ In case the amount of metabolite(s) is not so high as in above given example

What limit should we set: 10 % ? 1/1 ? 10-fold ? It would depend on the percentage of back-conversion (very low for clopidogrel, but can be quite high for some other metabolites). But OK, we can sort that out.

❝ and you can exclude back-conversion by the method used

That's where problems start:
- all metabolites are not always known, particularly for old molecules (nor the metabolite/parent ratio, by the way)
- the back-conversion mechanism is not always easy to predict. Back to clopidogrel, it took some time to understand how it really works and what gets back-converted (the guys in the PKWP got trapped: what they recommend has been proved not to work).

❝ and your 90% CIs are well within the acceptance range

If all of your "and" are met you indeed have chances to meet the criteria in the new Q&A and to save your application. We'll see what the Agencies think of it in the end.

❝ than an increase in the analytical variability can only decrease chances to demonstrate bioequivalence.

Still, I'm not fond of the "I'm just as bad for the reference as for the test so there is no issue" reasoning. Sorry I'm a bit dogmatic, but I'm more of the "garbage in, garbage out" kind.

Regards
Ohlbe

—
Regards
Ohlbe

Helmut
★★★

Vienna, Austria,
2012-12-21 21:19
(4541 d 02:33 ago)

@ Ohlbe
Posting: # 9745
Views: 8,232

Lousy studies?

Post reply

Dear Ohlbe,

❝ ❝ than an increase in the analytical variability can only decrease chances to demonstrate bioequivalence.

❝

❝ Still, I'm not fond of the "I'm just as bad for the reference as for the test so there is no issue" reasoning, sorry.

“To be as bad as the reference” is a common challenge in generic BE. ;-)

Joking aside, taking all of Dan’s “ands” for granted higher residual variance increases the producer’s risk. The patient’s risk is fixed. If a study passes despite a high CV, fine. But you are right that we might run into an ethical issue, e.g., if the study was planned as such and better methods would have required fewer subjects. On the other hand, the effect of analytical variability (as compared to physiologic variability) on the residual variance is generally overrated.^*

Do you remember this example (slides 46–49)? The LC/MS-MS method didn’t employ a stable isotope IS and was hit by a massive matrix effect. These were the good ol’ days of plausibility reviews – we stopped the analysis after 12 subjects and went for GC/MS. If you compare the profiles on slide 48 it’s clear that the LC/MS-MS method was awful. But have a look at the CVs and PEs on slide 49. Did it really matter? No. Actually the CV of C_max of the GC/MS method was higher.

Gaffney M. Variance Components in Comparative Bioavailability Studies. J Pharm Sci 1992;81(4):315–7. doi:10.1002/jps.2600810402.

—
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

Helmut
★★★

Vienna, Austria,
2012-12-21 16:23
(4541 d 07:28 ago)

@ Dr_Dan
Posting: # 9742
Views: 8,231

Tricky question

Post reply

Dear Dan,

your example reminds me on the first Arlington conference in 1991. One analyst claimed that he usually runs an entire BE study in one day (ultrafast GC/MS). He was pissed off by the requirement of showing batch-to-batch accuracy/precision (“What the hell? I have only one batch!”). Ended up as a fabulous shouting match with Vinod Shah. :-D

❝ Do you agree with these considerations?

Partly.

Following Ohlbe’s arguments, no. At the EPF/EUFEPS workshop in Brussels 2010 it was agreed that the first study failing ISR should not lead to rejection of the application – the issue should be investigated and resolved for subsequents ones (GL: “In case incurred sample analysis showed deviating results, this should be investigated, and adequate steps should be taken to minimize inaccuracy (and imprecision).”).
If this is regulatory practice (duno) the same logic IMHO should be applicable to old studies lacking ISR at all.
If (!) you can guarantee that you don’t have the metabolite issues Ohlbe mentioned, yes. A method is only validated with spiked samples. If the first study with a method passes ISR I – still – don’t understand the requirement to show ISR for subsequent ones with the same method.

—
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

ElMaestro
★★★

Denmark,
2012-12-23 06:10
(4539 d 17:42 ago)

@ Helmut
Posting: # 9749
Views: 8,004

Add'l requirements

Post reply

Dear all,

further: With immediate effect EMA adds a few new requirements for PK studies:

All CRO technicians must have at least a doctorate and 10 years documented work experience. Which educations correspond to acceptable functions will be assessed on a case-by-case basis but also note pt. 5.
The educating institutions for lab techs will be inspected at 20% coverage by regulators.
CRO technicians who work on the liquids must each morning prior to commencing work be dosed beta blockers or other appropriate anti-tremor medication.
Pipettes must be calibrated daily by a certified institution.
Computer systems used to calculate the 90% CI may only be used once due to the risk of RAM burning or harddisk failure which may or may not affect the system's ability to calculate. Validate before use.
The GLP study director must be an MD because they tend to know a lot about analytical chemistry.
Ramipril and Clopidogrel are specifically discouraged.
The CRO must, as part of the validation, prove 8 years stability of stock solutions. This needs to be done at low, middle and high barometric conditions.
Requirements 1-8 do not apply to originator developments, because phase 3 safety studies usually imply so many blood samples anyway.

—
Pass or fail!
ElMaestro