Log in
Register|
Oiinkie ☆ The Netherlands, 2013-12-11 15:23 (4156 d 07:11 ago) Posting: # 12034 Views: 6,307 |
|
|
Dear all, We are developing a tablet formulation which we would like to take to a pilot PK study. The reference product is known to have a low intra-subject variability, but a very high inter-subject variability. We think we have found a concept to reduce the inter-subject variability and one of the aims of the pilot PK study would be to assess and compare inter-subject variability of test and reference product. Given the API's nature, we have to go with patients (healthy volunteers not allowed). What kind of setup would you recommend for such a study (e.g. standard 2x2 crossover)? What sample size would be appropriate? Many thanks for your help! — Regards, Oiinkie |
|
Helmut ★★★   Vienna, Austria, 2013-12-11 16:04 (4156 d 06:29 ago) @ Oiinkie Posting: # 12035 Views: 5,243 |
|
|
Hi Oiinkie, ❝ The reference product is known to have a low intra-subject variability, but a very high inter-subject variability. CVs of the product? Do you have data of a replicate study? ❝ […] one of the aims of the pilot PK study would be to assess and compare inter-subject variability of test and reference product. Hhm, why CVinter? Speaking cross-over only CVintra is relevant. I performed studies with CVinter of ~60%, where BE was shown in 12 subjects (due to CVintra ~10%). I have seen some anecdotal (!) reports that high standardization (i.e., inclusion criteria, sex, narrow BMI,…) reduced CVintra as well. I’m somewhat skeptical. IMHO that would only work for high correlation between the variance components. I don’t see a physiological basis for that. ❝ What kind of setup would you recommend for such a study (e.g. standard 2x2 crossover)? Still not clear to me whether and why you are interested in the CVs of the products. In order to get CVWR and CVWT you would need a fully replicated design (TRT|RTR or TRTR|RTRT). For CVBR and CVBT I guess you would have to perform the study in two groups. Don’t ask me for the model. Maybe some ideas by other members? ❝ What sample size would be appropriate? Define appropriate.  — Dif-tor heh smusma 🖖🏼 Довге життя Україна! ![[image]](https://static.bebac.at/pics/Blue_and_yellow_ribbon_UA.png) Helmut Schütz ![[image]](https://static.bebac.at/img/CC by.png) The quality of responses received is directly proportional to the quality of the question asked. 🚮 Science Quotes |
|
Oiinkie ☆ The Netherlands, 2013-12-11 18:47 (4156 d 03:46 ago) @ Helmut Posting: # 12040 Views: 5,200 |
|
|
Hi Helmut, Thanks for your reply! I am very familiar with setting up BE studies and thus taking into account CVintra, but struggling a bit with the "brilliant" idea of my boss for this product regarding CVinter. ❝ CVs of the product? Do you have data of a replicate study? We do not have data of a replicate study, but the CVintra reported in BE studies with IR tablets is about 10-15%. I should have said "CVintra of the compound when formulated in an IR tablet." ❝ Hhm, why CVinter? I should have provided a bit more background. I will try to explain the rationale  The IR tablets currently on the market show low CVintra, but high differences in terms of bioavailability (AUC) between subjects. Bioavailability after oral administration of IR tablets varies between about 30 to 95%. If a patient shows low BA (low-responder), he/she is switched to injections. We are developing MR tablets/capsules (eiter EC or CR) which we believe may increase bioavailability for those who show low bioavailability with IR tablets (however, this idea may be far-fetched). If we would be able to raise bioavailability to about 60 to 95% in general, we would have proof of our concept. In other words, more patients would be able to stay on tablets (instead of injections). ❝ ❝ What sample size would be appropriate? ❝ ❝ Define appropriate. Well, this is where I am having doubts. How large should the sample size be to show a potential difference in AUCs between subjects? In other words, how many patients do we need to show that bioavailability is increased from 30-95% with IR tablets to, say, about 50/60-95% with our formulation? As a consequence, we will obviously not aim for BE and probably end up in a phase III study... I know that this is kind of a long-shot, but we (read: my boss) are willing to give it a try... Thanks! — Regards, Oiinkie |
|
Dr_Dan ★★ Germany, 2013-12-12 10:25 (4155 d 12:09 ago) @ Oiinkie Posting: # 12043 Views: 5,116 |
|
|
Dear Oiinkie As far as I understood your drug administered as an IR formulation shows low intra-subject variability whereas the inter-subject variability in bioavailability is high. In other words patients will have consistently either high or low bioavailability. You aim to increase the low bioavailability with a MR formulation. Do you really know the reason for differences in bioavailability? Is it metabolism or just BMI? (my question in this context: the percentage you present is it absolute or relative bioavailability?). If you identify a patient population which tends to have low bioavailability then you do not have to aim for BE as you already suggested and consequently you do not need to perform a formal sample size estimation. You just have to show that bioavailability of the drug in this patient population is increased by the MR formulation in comparison with the IR formulation. However, this will also mean that you have to proof efficacy and safety of the MR formulation. I hope this helps Kind regards Dan — Kind regards and have a nice day Dr_Dan |
|
Oiinkie ☆ The Netherlands, 2013-12-17 11:02 (4150 d 11:31 ago) @ Dr_Dan Posting: # 12068 Views: 5,326 |
|
|
Dear Dan, Thanks for your input! ❝ As far as I understood your drug administered as an IR formulation shows low intra-subject variability whereas the inter-subject variability in bioavailability is high. In other words patients will have consistently either high or low bioavailability. You aim to increase the low bioavailability with a MR formulation. Correct! ❝ Do you really know the reason for differences in bioavailability? Is it metabolism or just BMI? (my question in this context: the percentage you present is it absolute or relative bioavailability?). If you identify a patient population which tends to have low bioavailability then you do not have to aim for BE as you already suggested and consequently you do not need to perform a formal sample size estimation. You just have to show that bioavailability of the drug in this patient population is increased by the MR formulation in comparison with the IR formulation. However, this will also mean that you have to proof efficacy and safety of the MR formulation. Besides characteristics like BMI, it also has to do with absorption/metabolism. The compound is actively absorbed via a saturable mechanism. With the pilot PK we will try to identify the patient population that shows low BA and, moreover, to see whether our MR formulation may show signs that the concept would work. It may be most informative to include patients who have been switched to injections (probably low-responders -> low BA), but this is regarded as unethical as we would then withdraw them from a therapy that works for them. To get some information from this pilot PK on the feasibility of our formulation, I think we will still go for a 2x2 crossover, probably with 16 patients. Please let me know if you think an other design or sample size would be more informative. Many thanks for your help! — Regards, Oiinkie |
Ing. Helmut Schütz