Bioequivalence and Bioavailability Forum

Hi John,

❝ ❝ • Mean t_max of what?

❝

❝ Oh they did write "2 x mean Tmax of reference".

Doesn’t mean a thing unless one defines what the reference is. Please see again my #1/#2 above.

❝ I can see that if it's not a guidance requirement, but they should adhere to the guidance when the sponsor said it's for FDA ANDA…

Yep. I’m fed up with brainless protocols (and reports). I’m the last person taking guidances literally, but one has to understand them first if one intentionally [sic] wants to deviate from them.

• Oops, \(\overline{x}\) instead of \(\widetilde{x}\). Sorry.*
  
• Two-Stage rSABE study. Unknown patient’s risk; no statistical method published? Sorry.*
  
• Concentrations below the LLOQ used in the evaluation. Sorry.*
  
• λ_z from t_max to t_last, crossing three phases. Not terminal? Oops, sorry.*
  
• ≈ ∞

---

• We can change it.

Hi Luv,

❝ ❝ Absence of evidence ≠ evidence of absence. Indians claim a lot.

❝

❝ Ouch!!!

❝

❝ That Hurts!!

Point taken. Terribly sorry.

I’ll try to be more precise. Believe me, in my job I see an awful lot of protocols and reports. Some questions posted in the forum are also telling. Up to now I gave five workshops in India. My observations:

• Guidance-bondage intrinsically tied to copy-and-paste-disease.
  
• Lack of understanding even the most basic principles of GxP, PK, and good biostatistical practice (see the rather incomplete list in the post above).
  
• Worst: A strong tendency to bend the rules (“make” analytical batches valid, dealing with outliers, etc.).

❝ It's not only about India, we have done a couple of studies in EU and Canada and everywhere same word was used.

It’s not about the word we use. Mean and median are different things. If the FDA requires the median you can’t use the mean without a justification.

❝ Now coming to your point "The median of tmax-values after the reference in the study – without the subject who vomited" Do you mean that we need to get blood samples from all the volunteers, mesaure the plama concentration, run the stats and then exclude the volunteer.

It depends on what you stated in the protocol. If you know the reference’s median (!) t_max, write two times of this value in the protocol as a cut-off and stop sampling if a subject vomits ≤ 2×t_max,ref.
If you don’t know the cut-off it’s upon the PI’s discretion whether he excludes the subject right away or not. If a subject is already nauseous it may be unethical to continue blood sampling. Furthermore, the subject’s example may influence other volunteers.
However, if the PI decides to keep the subject in the study he may be excluded after calculating median t_max,ref of the rest of the group.

❝ Will it be acceptable by EMEA or FDA? (Remember In principle any reason for exclusion is valid provided it is specified in the protocol and the decision to exclude is made before bioanalysis)

That’s a gray area. No experience with the FDA. Luckily in Europe sometimes the median t_max is given in SmPCs (or more often in PARs). You are right about the statement in EMA’s GL:

Reasons for exclusion
In principle any reason for exclusion is valid provided it is specified in the protocol and the decision to exclude is made before bioanalysis.
Examples of reasons to exclude the results from a subject in a particular period are events such as vomiting and diarrhoea which could render the plasma concentration-time profile unreliable.
The permitted reasons for exclusion must be pre-specified in the protocol. If one of these events occurs it should be noted in the CRF as the study is being conducted. Exclusion of subjects based on these pre-specified criteria should be clearly described and listed in the study report.
Exclusion of data cannot be accepted on the basis of statistical analysis or for pharmacokinetic reasons alone, because it is impossible to distinguish the formulation effects from other effects influencing the pharmacokinetics.

That’s a sound text. For IR products 2×t_max is a reasonable cut-off based on PK (~98% of absorption completed). If I don’t have the median I use the mean instead (although teeth grinding). If we don’t know any value how can we state something in the protocol? See this post on what I use in my protocols. So far I never had a problem (which doesn’t mean a thing).

BTW, “Appendix IV of the Guideline on the Investigation on Bioequivalence (CPMP/EWP/QWP/1401/98 Rev.1): Presentation of Biopharmaceutical and Bioanalytical Data in Module 2.7.1” mandates applicants to state the medians of t_max of test and reference in Table 3.1. Therefore, in the future I expect to see more useful information in PARs.