ele2008 ☆ Korea, 2014-10-16 07:06 (3478 d 18:34 ago) (edited by ele2008 on 2014-10-16 15:46) Posting: # 13722 Views: 26,455 |
|
Dear Sir I am a clinical pharmacologist from China, thank you very much for keeping such a wonderful webiste for others, it has been extremely helpful. I have a question about the subject inclusion in a bioequivalence study: The Guidance from the EMA says: "The subject population for bioequivalence studies should be selected with the aim to minimise variability and permit detection of differences between pharmaceutical products." However the Guidance from the FDA says: "In general, unless otherwise recommended in a specific guidance: ... In vivo BE study subjects should be representative of the general population,taking into account age, sex, and race... If a drug product is intended for use in both sexes, the applicant should include similar proportions of males and females in the study..." I wonder if it is ok to maximize the homogeneity of subjects, especially for those highly variable drugs. Because recently a few studies found that genetics play an important role in determining the intra-individual variability. So we think if we recruit subjects with same genotype, we can have a smaller intra-indivdual variability therefore a smaller sample size. But we wonder if we are rational (or allowed) to do so. Thank you very much |
Helmut ★★★ Vienna, Austria, 2014-10-16 17:04 (3478 d 08:36 ago) @ ele2008 Posting: # 13724 Views: 25,630 |
|
❝ Dear Sir ↑↑↑↑ Not interested in opinions of female members of the Forum? ❝ […] thank you very much for keeping such a wonderful webiste for others, it has been extremely helpful. Welcome! ❝ I have a question about the subject inclusion in a bioequivalence study: […] ❝ I wonder if it is ok to maximize the homogeneity of subjects, especially for those highly variable drugs. Because recently a few studies found that genetics play an important role in determining the intra-individual variability. ❝ So we think if we recruit subjects with same genotype, we can have a smaller intra-indivdual variability therefore a smaller sample size. You are raising very interesting questions! Quoting from one of Les Benet’s presentations1: What is the justification for studying EMA followed the same rationale in their GL – minimize variability. This idea is even more clear for parallel designs, where geno-/phenotyping is mandatory (only a BE-study in one panel is required). For most drugs extensive/fast (E/FM) metabolizers are more common than poor/slow metabolizers (P/SM). Geno-/phenotyping is also important in cross-over studies, especially if steady state in PMs might lead to safety problems (example: paroxetine). Not sure were FDA’s requirement comes from. At times when Individual Bioequivalence (IBE) was discussed, one major topic was the subject-by-formulation interaction – which can be assessed only in a full replicate design. One idea was to explore subgroups (sex, age, ethicity, genotypes, …). It became immediately evident that a stratification for all possible combinations (keeping the power of comparisons at a reasonable level) is simply impossible unless one would opt for extremely high sample sizes. Therefore, the guidance states: We recommend that the total number of subjects in the study provide adequate power for BE demonstration, but it is not expected that there will be sufficient power to draw conclusions for each subgroup. Statistical analysis of subgroups is not recommended. We recommend that restrictions on admission into the study generally be based solely on safety considerations. (my emphasis)Generally we would expect that genotypes should be of no particular importance in BE, since the comparison between treatments is done within subjects. However, you are correct that there are some examples in the literature showing the opposite – if the intra-subject variability differs between genotypes.
❝ But we wonder if we are rational (or allowed) to do so. I don’t think so. At least the last sentence of FDA’s quotation speaks against selection of specific genotypes…
— 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 |
ele2008 ☆ Korea, 2014-10-16 18:18 (3478 d 07:22 ago) @ Helmut Posting: # 13727 Views: 25,187 |
|
Thank you very much, of course I appreciate whoever says something here. :D According to your reply, can I take it as: The EMA and the FDA follow different principles when it comes to selecting subjects. Therefore, if I want to submit a study to the EMA, I can try minimizing the subject variability (i.e., using young male healthy volunteers, maybe with same phenotype); but if I have to submit my study to the FDA, then I should not do that. The two studies you presented showed that people from two different subgroups can have nearly 2-folds differences in their CV%. The sample sizes calculated from these CV% can be nearly 2 fold different as well. Would not it be more ethically and financially sounding to both investigators and sponsors if they can use a much smaller sample size? But what I am talking about here is just an assumption because the relationship between the subjects' genotypes and intra-individual vriability is still largely uncertain (as other studies showed that there was no differences between CYP3A5 genotypes and tacrolimus intra-ondividual variability1,2)
Edit: References linked. [Helmut] |
nobody nothing 2014-10-16 19:43 (3478 d 05:56 ago) @ ele2008 Posting: # 13731 Views: 25,212 |
|
Ein Blick in's Gesetz erleichtert die Rechtsfindung.. http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2010/01/WC500070039.pdf states under 4.1.3 Phenotyping and/or genotyping of subjects may be considered for safety or pharmacokinetic reasons. ...supported here http://www.dovepress.com/getfile.php?fileID=18246 For the FDA I don't remember such information for BE... — Kindest regards, nobody |
Helmut ★★★ Vienna, Austria, 2014-10-16 20:13 (3478 d 05:27 ago) @ nobody Posting: # 13732 Views: 25,296 |
|
Ὀδυσσεύς ❝ http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2010/01/WC500070039.pdf states under 4.1.3 ❝ Phenotyping and/or genotyping of subjects may be considered for safety or pharmacokinetic reasons. Yessir. Again we have to start an exegesis on what the almighty oracle means by “pharmacokinetic reasons”. Sure, the different PK in PMs is the driving force leading to safety issues. So why not simply state “safety As shown in my references the CVw of PK metrics might be different between genotypes. Do they mean that as well? ❝ ...supported here ❝ http://www.dovepress.com/getfile.php?fileID=18246 Hey, that’s a goodie! At the end we are left out in the dark: However, global consensus is unavailable with regard to the inclusion or exclusion of poor versus extensive metabolizers (EMs) as a method to decrease variability. (my emphasis)— 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 |
nobody nothing 2014-10-16 20:24 (3478 d 05:15 ago) @ Helmut Posting: # 13733 Views: 25,088 |
|
❝ As shown in my references the CVw of PK metrics might be different between genotypes. Do they mean that as well? In my opinion: YES, see "pharmacokinetics". But i'm not the Pythia, just Οὐδείς — Kindest regards, nobody |
Helmut ★★★ Vienna, Austria, 2014-10-17 21:30 (3477 d 04:10 ago) @ nobody Posting: # 13749 Views: 25,050 |
|
Ὀδυσσεύς and all, ❝ ❝ As shown in my references the CVw of PK metrics might be different between genotypes. Do they mean that as well? ❝ ❝ In my opinion: YES, see "pharmacokinetics". OK, I’m convinced. Pharmacokinetics may be simply I never went that route so far (only for safety reasons). Now it gets interesting. Is this an academic issue or do our members have some experiences they could share, like:
— 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 |
nobody nothing 2014-10-17 22:04 (3477 d 03:35 ago) @ Helmut Posting: # 13750 Views: 25,132 |
|
Your proposal really pushes the approach to the edge I would never recommend something like that without a Sci Adv... Will be hard to present data to estimate intra CV for each subtype/subpopulation, (or?), but avoiding rare geno-/phenotypes might be worth going for. Cheers (*hicks*) Οὐδείς PS: Or you have read this http://www.ncbi.nlm.nih.gov/pubmed/20010459 http://www.ncbi.nlm.nih.gov/pubmed/23179471 ? — Kindest regards, nobody |
Helmut ★★★ Vienna, Austria, 2014-10-18 17:04 (3476 d 08:35 ago) @ nobody Posting: # 13751 Views: 25,067 |
|
Ὀδυσσεύς ❝ Your proposal really pushes the approach to the edge It’s far from being a proposal; I was just asking whether anybody has done that before. ❝ I would never recommend something like that without a Sci Adv... Absolutely. ❝ Will be hard to present data to estimate intra CV for each subtype/subpopulation, (or?), but avoiding rare geno-/phenotypes might be worth going for. Seems so. Actually I’m fine with adjusting for everything happening after absorption (variable clearance). Would not accept pre-systemic stuff (excipients influencing transporters!). I’m fine with performing a study in a major genotype showing lower variability. Don’t forget what the GL tells us: The subject population for bioequivalence studies should be selected with the aim of permitting detection of differences between pharmaceutical products. In order to reduce variability not related to differences between products, the studies should normally be performed in healthy volunteers unless the drug carries safety concerns that make this unethical. This model, in vivo healthy volunteers, is regarded as adequate in most instances to detect formulation differences and to allow extrapolation of the results to populations for which the reference medicinal product is approved (the elderly, children, patients with renal or liver impairment, etc.). EMA has no problem accepting studies in males only (which represent <50% of the population) and extrapolate to anybody…❝ PS: Or you have read this The first one was already quoted above; THX for the second! — 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 |
nobody nothing 2014-10-20 10:06 (3474 d 15:34 ago) @ Helmut Posting: # 13752 Views: 24,854 |
|
Transporters are one possible source of variability pre-absorption, but how about CYP-gutwall metabolism? Any knowledge on formulation aspects potentially influencing this? If not, selection of selected geno-/phenotypes might be a mistake, unless you pick the most sensitive subpopulation... Moreover, in case of a drug specifically indicated in e.g. hepatic impairment I would really doubt that (at least in special cases) healthy volunteers will be completely predictive, given the significant changes in physiology in this patient population... No data on hand, but a gutfeeling requiring investigation, in my opinion, instead of apodictic postulations. Οὐδείς — Kindest regards, nobody |
Helmut ★★★ Vienna, Austria, 2014-10-20 15:00 (3474 d 10:39 ago) @ nobody Posting: # 13755 Views: 24,874 |
|
Ὀδυσσεύς ❝ Transporters are one possible source of variability pre-absorption, but how about CYP-gutwall metabolism? Correct. That’s why I wrote ❝ ❝ I’m fine with adjusting for everything happening after absorption […]. Would not accept pre-systemic stuff […] ❝ Any knowledge on formulation aspects potentially influencing this? Nope. But should be possible as well. ❝ If not, selection of selected geno-/phenotypes might be a mistake, unless you pick the most sensitive subpopulation... Oh no, that’s tricky! ❝ Moreover, in case of a drug specifically indicated in e.g. hepatic impairment I would really doubt that (at least in special cases) healthy volunteers will be completely predictive, given the significant changes in physiology in this patient population... No data on hand, but a gutfeeling requiring investigation, in my opinion, instead of apodictic postulations. That’s the almighty oracle (see also the Les Benet’s quote over there). But in renal/hepatic impaired patients we are talking about clearance. That’s the drug, not the formulation. — 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 |
nobody nothing 2014-10-20 16:21 (3474 d 09:19 ago) @ Helmut Posting: # 13757 Views: 24,859 |
|
❝ But in renal/hepatic impaired patients we are talking about clearance. That’s the drug, not the formulation. Not necessarily, might also be on absorption level... I think pathophysiological alterations might include active as well as passive absorption mechanisms as well as presystemic metabolism (both phase I and II) Οὐδείς PS: Regarding LZB and "healthy volunteers": prospective trials is one piece of cake, clinical practice (including therapeutic drug monitoring) and e.g. epileptics/immunosuppressants a completely other... PPS: Finding BE-failures in clinical practice is almost like studying carcinogenicity of a drug from post-marketing surveillance data, i.e. virtually impossible. Maybe for an analgesic you would be able to detect a large difference in Cmax, but how about antihypertensives or even lipid lowering drugs in "primary prevention" (one of my favorite indications). — Kindest regards, nobody |
ele2008 ☆ Korea, 2014-10-23 07:25 (3471 d 18:14 ago) (edited by ele2008 on 2014-10-23 10:55) @ ele2008 Posting: # 13772 Views: 24,744 |
|
Thank you for the intriguing discussion. I posted this issue because my professor proposed we conduct such a study: to attest whether this genotype-dependancy of WSV exists or not, and if it does exist, is it same for different CYP enzymes? So his original intention was to reduce WSV (and hence the sample size) by recruiting genetically-homogeneous subjects. However, after reading the discussion, I wonder if it is still worth doing so (my first question). Currently, what I am doing is to search for references and to find out if WSV is genotype-dependent. Because there are not so many papers, I decided to find papers on drug-drug interaction (DDI) studies which included genetic information. Because I read your lecture that says WSV can also be estimated from DDI studies. But I wonder how different this estimation is from their real WSV that is determined from a replicate study or a BE study (my second question). I appreciate very much all opinions and comments here, also if you are interested in participating our study or have any better ideas, please feel free to share with me. Edit: I removed your e-mail address and activated the link to the contact form in your profile. [Helmut] |
ele2008 ☆ Korea, 2014-10-23 18:47 (3471 d 06:53 ago) @ ele2008 Posting: # 13776 Views: 24,744 |
|
Thank you Helmut for the editing. I just read this article titled "Gender Differences in Drug Bioequivalence: Time to Rethink Practices"1 Apparently, intra-individual variability is dependent on many factors (so far genetics and gender have been confirmed, I assume age and health condition are such factors as well). It seems impossible for any BE study to balance all the influencing factors in a study design, unless they use a huge sample size. therefore I am now more convinced by the viewpoint from the EMA. I read again what you posted: "We are trying to determine if two dosage forms of the same drug behave similarly. Therefore we want to keep any other variability not due to the dosage forms at a minimum. We choose the least variable “test tube”, that is, a healthy volunteer." In the end of article,the author says "Before the twentieth century, men used to perform women’s roles in theater, as a reflection of sexual puritanism. Until recently, women were not allowed to vote in many countries, and this reality continues today in some jurisdictions. It was assumed that men knew what was “good for women.” In the case of regulatory reality, the current approach is clearly not good for women, as the science shows." Really, is women not being included into BE studies a sign of sex discrimination? I did not know that. (Speaking as a woman myself) 1. Koren G, Nordeng H, MacLeod S Gender differences in drug bioequivalence: time to rethink practices. Clin Pharmacol Ther. 2013 Mar;93(3):260-2. doi 10.1038/clpt.2012.233. Epub 2012 Dec 3. |
ElMaestro ★★★ Denmark, 2014-10-23 19:25 (3471 d 06:15 ago) @ ele2008 Posting: # 13777 Views: 24,836 |
|
Hello ele2008 and all, ❝ I just read this article titled "Gender Differences in Drug Bioequivalence: Time to Rethink Practices"1 ❝ ❝ Apparently, intra-individual variability is dependent on many factors (so far genetics and gender have been confirmed, I assume age and health condition are such factors as well). I think the paper completely misses an extremely important point: In average BE we are just interested in having confidence in the GMR which is the similarity of two drugs. Yes, drugs sometimes/often/always behave differently in women and men, but that doesn't mean the GMR is different between women and men. That the variabilities differ between men and women is in that regard really not of much importance. It also differs between women and women, and between men and men. When we want to study the GMR -and this truly is what we wish to study- then we can do so very effectively (meaning our confidence in our conclusions drawn about it is increased) by cutting down the variability. Selecting the population so as to minimise variability is the tool to achieve it. Subgroups: The quote means that when you conduct a BE study you should draw the conclusions on basis of the population defined in the protocol. You should not take out a subgroup and try to prove BE on that fraction. There is a reason: When studies failed in the past companies wold embark on all sorts of subanalyses to check if they by chance or whatever could detect BE in a fraction of the studied subjects. Of course, on average the numbers would be against such findings but they would -will!- pop up by chance every now and then. I am all for the study of BE separately in men and women for drugs for which the GMR changes from sex to sex. And that is frankly speaking not a lot as far as I know. In fact, I find it very difficult to name a single one. So... which drugs or formulation pairs really have a GMR change between the sexes? — Pass or fail! ElMaestro |
nobody nothing 2014-10-23 19:53 (3471 d 05:47 ago) @ ElMaestro Posting: # 13778 Views: 24,690 |
|
Let's talk about sex, not gender (which is a sociolgical combat term, implying that all differences between men an women are made up by education. Biology and sex hormones are all secondary. Sounds strange, but...) So what you are looking for is examples of differences between formulations (not active compounds) in absorption between women and men. Will be hard to find, I guess. Maybe the reason why FDA started with women in BE trials, as there is no sound scientific basis to exclude them. But we started with genetic polymorphisms. If nothing is published on that: GREAT! Then let's start with studying intraindividual variability (repeated dosing) for different CYP geno/phenotypes. Should be worth going for and has not been done (published) so far! If you don't see differences on intraindividual variability it should not be worth going for selected geno/phenotypes for this CYP in the future, I guess. Otherwise... PS: Another term for "difference in geno/phenotype" might be "race". There are compounds with distinct differences in BA depending on race. In principle higher BA should correspond to lower variability (both intra- and interindividual, I guess). Might be worth having a look at in the literature — Kindest regards, nobody |
Helmut ★★★ Vienna, Austria, 2014-10-23 19:54 (3471 d 05:46 ago) @ ele2008 Posting: # 13779 Views: 24,738 |
|
Hi ele2008, I’m not sure whether something went wrong with the Forum’s software – your entire post disappeared; I restored it from electronic limbo. In case you want to edit your post, please see the Forum’s FAQ #3. — 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, 2014-10-23 21:50 (3471 d 03:50 ago) @ ele2008 Posting: # 13780 Views: 24,880 |
|
Hi ele2008, ❝ I just read this article titled "Gender Differences in Drug Bioequivalence: Time to Rethink Practices"1 I have some mixed feelings about this “Review” – with nine [sic] references; three of them being guidelines. I’m just writing a review about Two-Stage-Designs and digested fifty papers so far – and there is no light at the end of the tunnel yet. “To achieve BE, the AUC and peak concentrations of the generic drug need to be within 80–120% of the reference drug” referring to FDA’s guidance. 120%? Followed by “[…] typically BE studies are conducted in healthy, young adult male volunteers” referring to EMA’s BE-GL, where I read “Subjects could belong to either sex; however, the risk to women of childbearing potential should be considered”. ClinTrials, Completed BE Studies in adults: all: 1344, females: 1028, males: 1294. “Ashiru and colleagues have demonstrated that polyethylene glycol enhances the bioavailability of ranitidine in a dose-dependent manner by up to 63% among men, whereas it decreases absorption among women by up to 24%!” It’s funny to read a sentence ending with an exclamation mark in a scientific paper. Haven’t seen that before. I guess the authors are on a mission. They performed some cherry-picking in their tables 1&2 from Chen’s retrospective analysis. It is worthwhile to read their paper.* Five of the 26 studies had a sample size of twelve; the median sample size was ten / sex. The background of the analysis was individual bioequivalence (IBE), which was fashionable those days. Have a look at their table III, giving the 95% CI of the female/male standard deviation ratio. I’m not worried at all. From the conclusions: “[…] we found that in most instances, intrasubject variability in pharmacokinetic measures is similar between men and women, although it may indeed be higher for women in a few cases.” BTW, one of the main reasons why IBE never was implemented was the need for fully replicated designs with high sample sizes. Whereas μ is the first cumulant of the distribution, σ² is its second. With any given sample size a comparison of variances is less powerful than a comparison of means. FDA recently recommended a comparison of variances of NTIDs. The limit of σT/σR is 2.5… ❝ In the end of article,the author says "Before the twentieth century, men used to perform women’s roles in theater, as a reflection of sexual puritanism. Until recently, women were not allowed to vote in many countries, and this reality continues today in some jurisdictions. It was assumed that men knew what was “good for women.” In the case of regulatory reality, the current approach is clearly not good for women, as the science shows." Followed by “Women may become pregnant, which is not an issue for short BE studies.” What‽ ❝ Really, is women not being included into BE studies a sign of sex discrimination? I don’t think so. My bioanalytical team (♀:♂ ~ 1:1) was always happy about females in BE-studies because due to their lower body weight / smaller volume(s) of distribution concentrations where higher – less trouble with the LLOQ. Females/males in BE was our standard for two decades. As a welcomed side-effect the atmosphere in the recreation room of the CRO was much calmer in mixed-sex studies. See also this thread and this post.
— 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 |