Nirali
★    

India,
2009-04-13 11:53

Posting: # 3515
Views: 2,713
 

 Evaluation of Reference variability [Design Issues]

Dear All,

We want to conduct BE replicate study to demonstrate that the within-subject variability for Cmax of the reference compound in the study is >30% for EU regulatory.

To check reference variablity which study design and model shall be used for statistical analysis?

Shall we use 2-period replicate design to assess the inter and intra-subject variability of reference drug, in which reference drug given in both the periods?

Thanks & Regards,
NIRALI
KR
★    

India,
2009-04-13 14:39
(edited by Ohlbe on 2009-04-14 10:39)

@ Nirali
Posting: # 3517
Views: 2,383
 

 Evaluation of Reference variability

Dear Nirali,

You can use partial replicate study in which two reference and one test formulation are given to evaluate variablity of Cmax.
For details, see the Helmut's post on "Reference scaled approach for HVD".

KR
--
Edit: Full quote removed. Please see this post! [Ohlbe]
Ravi
★    

India,
2009-04-14 06:25

@ Nirali
Posting: # 3519
Views: 2,395
 

 Evaluation of Reference variability

Dear Nirali,

To check for reference variablity you can use modified replicate design i.e. 3 period, 3 sequence and 2 treatment crossover design with sequences as RTR, RRT, TRR. But this design will provide you with intra CV for reference only. For more details please have alook at
Bioequivalence Approaches for Highly Variable Drugs and Drug Products by Sam H. Haidar.

I hope this helps.

Thanks & Regards
Ravi Pandey
Nirali
★    

India,
2009-04-17 07:38

@ Ravi
Posting: # 3551
Views: 2,317
 

 Evaluation of Reference variability

Thank you friends.

Actually, we got the proposal to evaluate reference variability using below design:
"2-period replicate design to assess the intra-subject variability of reference drug, in which reference drug given in both the periods".

Shall it be appropriate design to evaluate reference variability for EU regulatory? :confused:

Regards,
NIRALI
Ohlbe
★★★

France,
2009-04-17 10:04

@ Nirali
Posting: # 3553
Views: 2,308
 

 Evaluation of Reference variability

Dear Nirali,

» Shall it be appropriate design to evaluate reference variability for EU regulatory? :confused:

This was not the approach recommended at the EGA meeting last October. The preferred approach was to have a 3- or 4-period replicate study, with the reference administered twice, and the test once or twice. It seems that they are afraid that you might run a messy trial to show a high variability with the reference alone, and use these results to justify wider acceptance limits in a clean 2x2 trial with test and reference afterwards.

Regards
Ohlbe
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2009-04-17 13:44

@ Nirali
Posting: # 3555
Views: 2,334
 

 Evaluation of Reference variability

Dear Nirali,

I agree with Ohlbe. Same opinions were expressed at the EUFEPS meeting last January (see this thread), where almost the entire PK group of EMEA was present. The group univocally expressed their point of view that a two period replicate design (reference only) is not acceptable. See also Jan Welink's presentation (slide 22).
On the other hand I know of European companies running such studies right now (despite clear reservations by regulators). Obviously these companies are sure that they will be able to convince authorities that the performance was not sloppy. I would not bet on their success!

Cheers,
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
Nirali
★    

India,
2009-04-18 09:04
(edited by Nirali on 2009-04-18 09:19)

@ Helmut
Posting: # 3561
Views: 2,244
 

 Evaluation of Reference variability

Dear HS,

Thank You :-)

Just for further knowldedge - which statistical model can apply in case of two-period replicate design (reference only)?

Regards,
NIRALI
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2009-04-18 12:31

@ Nirali
Posting: # 3562
Views: 2,325
 

 Evaluation of Reference variability

Dear Nirali,

see this post.

Cheers,
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
Nirali
★    

India,
2009-04-18 19:34

@ Helmut
Posting: # 3564
Views: 2,245
 

 Evaluation of Reference variability

Dear HS,

Thank you very much again.


Regards,
NIRALI
Marcel
★    

2010-03-16 13:10
(edited by Jaime_R on 2010-03-16 14:54)

@ Helmut
Posting: # 4920
Views: 1,864
 

 Evaluation of Reference variability

» Same opinions were expressed at the EUFEPS meeting last January (see this thread), where almost the entire PK group of EMEAas present. The group univocally expressed their point of view that a two period replicate design (reference only) is not acceptable.

Hello all. My apologies for dragging this replicate skeleton out of the closet when it seemed like it was comfortably resting.

Was there any mention at this meeting as to whether the results of a replicate design study (finding of intra-subject variablility exceeding 30%) with a lower strength can be used to justify widen acceptance limits for a higher strength of the same drug to undergo a standard BE study?

If this is not allowed, can someone explain why not?

Thank you in advance.


Edit: Full quote removed. Please delete anything from the text of the original poster which is not necessary in understanding your answer; see also this post! [Jaime]
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2010-03-16 16:01

@ Marcel
Posting: # 4921
Views: 1,840
 

 Evaluation of Reference variability

Dear Marcel!

» My apologies for dragging this replicate skeleton out of the closet when it seemed like it was comfortably resting.

No problem - if you think we need some more forensics, go for exhumation!

» Was there any mention at this meeting as to whether the results of a replicate design study (finding of intra-subject variablility exceeding 30%) with a lower strength can be used to justify widen acceptance limits for a higher strength of the same drug to undergo a standard BE study?

No. Or to be more precise, nobody asked that question, but I would bet the answer would have been a straight "No!" as well.

» If this is not allowed, can someone explain why not?

I would say there are two reasons.
  • The general reason, that no reference to previous studies is allowed when it comes to widening of the acceptance range. CVintra,ref >30% must be demonstrated within the submitted study.
  • There is also a pharmacokinetic reason. Drugs with nonlinear PK may show different variability throughout the dose range. Think about e.g., saturation effects. Example: ASS -> salicylic acid.

Cheers,
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
ElMaestro
★★★

Belgium?,
2010-03-16 17:13

@ Helmut
Posting: # 4922
Views: 1,877
 

 Evaluation of Reference variability

Dear Marcel and HS,

» I would say there are two reasons.
  • The general reason, that no
    » reference to previous studies is allowed when it comes to widening of the
    » acceptance range. CVintra,ref >30% must be demonstrated within
    » the submitted study.
    »
  • There is also a pharmacokinetic reason. Drugs with nonlinear PK may
    » show different variability throughout the dose range. Think about
    » e.g., saturation effects. Example: ASS -> salicylic acid.

Warning: This post is not fully thought through yet.
The comment from HS implies we would no longer be able to talk about highly variable drugs per se; we would have to think along the lines of "drug X is a highly variable drug at dose Y" etc. Perhaps we could not even talk about highly variable drugs, but about highly variable drug doses.
I am not sure this is fully in line with current thinking; in the PK-subgroup's earlier definition of a highly variable drug there was no mention of such considerations. Granted, this does not mean that they don't exist.

A further implication: Let's say we want to develop a generic of Schützomycin. The product is available in one strength, posology is 1 tablet daily. Now we also have an LC/MS/MS assay we cannot improve further upon and which has a pretty high LLOQ so we decide to do a BE study in healthy volunteers each being dosed 2 tablets. The ethics board accepts it, because Schützomycin is a nice drug with little safety concern.
So we do our study and find high variability for the product (haha, if a statistical menthod was only available nnwwsnm), and blah blah we end up with equivalence. We submit the dossier and we feel good. On Day 70 we get a short message from CA: "Potential serious risk to public health: Because high variability may be a dose dependent phenomenon, the applicant should provide evidence that the drug is not only highly variable at double the recommended dose."

This would be quite an evil one to deal with. Again, I have not thought it through, I am just relaying my initial thoughts to HS' post.

EM.
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2010-03-16 18:59

@ ElMaestro
Posting: # 4923
Views: 1,900
 

 Some desultory thoughts

Hi ElMaestro!

» Warning: This post is not fully thought through yet.

Warning: So is my reply.

» ... we would no longer be able to talk about highly variable drugs per se;

It was always important to make a distinction between HVDs (Highly variable Drugs) and HVDPs (Highly variable Drug Products).

» ... we would have to think along the lines of "drug X is a highly variable drug at dose Y" etc. Perhaps we could not even talk about highly variable drugs, but about highly variable drug doses.

Yes, probably.

» I am not sure this is fully in line with current thinking;

Both the Q&A-document (2006) and the new BE-GL talk about HVDs/HVDPs. And CV>30% has to demonstrated in a replicate design study. If we want to claim a wider acceptance range, for some drugs/formulations IMHO a historical justification would also do the job. After so many studies we all know that PPIs - especially in fed state - are HVDPs. But we need the replicate design anyhow to perform the scaling (theoretically, because we don't have a statistical method :angry:).
Marcel's question dealt with a historical reference (CV>30% after another dose). My two answers were based on:
- no historical reference anyway
- a worst case scenario (well, I still think that may be possible)

» A further implication: Let's say we want to develop a generic of Schützomycin. The product is available in one strength, posology is 1 tablet daily. Now we [...] decide to do a BE study in healthy volunteers each being dosed 2 tablets. The ethics board accepts it, because Schützomycinis a nice drug with little safety concern.

Fine to hear that Schützomycin is a nice drug. When I read the GL, I was wondering how anyone can take the responsibility of dosing more than the approved daily dose for analytical reasons. 'Anyone' in that case = sponsor + investigator + IEC. The drug was shown to be nice at the registered strength and OAD. No data for doubled dose. There are some anecdotal reports that Schützomycin if administered off-label is toxic to the sensory cells of the ear, sometimes causing complete hearing loss.

» [...] CA: "Potential serious risk to public health: Because high variability may be a dose dependent phenomenon, the applicant should provide evidence that the drug is not only highly variable at double the recommended dose."
»
» This would be quite an evil one to deal with.

This would be the absolute show-stopper.

Cheers,
Helmut Schütz
[image]

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
Marcel
★    

2010-03-17 12:47

@ Helmut
Posting: # 4929
Views: 1,847
 

 Evaluation of Reference variability

Thanks to both of you for the feedback. It's all very enlightening.

» » The general reason, that no reference to previous studies is allowed when it comes to widening of the acceptance range. CVintra,ref >30% must st be demonstrated within the submitted study.

Isn't a significant chunk of the logic for conducting BE studies of non-linear PK drugs at the higher/lower strength due to this "sensitivity" to distinguish between formulation effects being the greastest at that particular strength? Doesn't that imply it is done so because the variability at the highest/lowest strength is theorectically greatest and, therefore, if one can pass BE at this dose, then BE can be implied for the other strengths? Shall I tack on another run-on sentence?

I don't mean to kick a dead horse, but I'm wondering why scientifically (irrespective of historical precedent or variability throughout the dose range--as good as these answers are) one cannot use the results from one study to design another given the drugs are the same (both test and reference). What I'm getting at here is that if one suspects a drug may be highly variable, given the current interpretation of the regulations, one has to do one of two...no, make that three things:
  1. Scrap the originally planned standard BE study for the higher strength (in my case) and conduct it in replicate design.
  2. Burst the development budget by conducting a replicate design first and then designing a standard BE study with the appropriate confidence intervals next.
  3. Down a bottle of Schützomycin and hope the bacterial infection that set in one's scalp from all the head scratching is sensitive to this antibiotic (hopefully, a generic whose BE to the reference product was conducted in replicate design).
Does that sound about right? Can linking the results from the lower strength study to the higher strenth study be argued on scientific merit (assuming the science is correct)?
Activity
 Admin contact
20,145 posts in 4,248 threads, 1,385 registered users;
online 5 (0 registered, 5 guests [including 3 identified bots]).
Forum time (Europe/Vienna): 12:25 CET

A little Learning is a dang’rous Thing;
Drink deep, or taste not the Pierian Spring:
There shallow Draughts intoxicate the Brain,
And drinking largely sobers us again.    Alexander Pope

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5