Log in
Register|
wligtenberg ☆ The Netherlands, 2017-01-03 16:48 (3029 d 16:53 ago) Posting: # 16923 Views: 7,580 |
|
|
As mentioned before, I am quite new to the bioequivalence and bioavailability field. I was asked to look into dissolution profiles of a few compounds. I have already implemented the F2 and mahalanobis distance in R, so the calculation is not an issue. However, I am not sure if the rules in the guidelines are sufficient to apply them in a correct manner. For instance, we have compounds for which the RLD and the in house product never reach 85% dissolution. (lipophilic compounds in membranes/vesicles) If I am not mistaken, the idea of having only one time point after dissolution goes above 85% is to make sure, that you don't just add arbitrary high time points to decrease the influence of the lower time points (if they would differ in the beginning). This would allow you to always get an F2 above 50, if you just measure enough points after dissolution is at its peak. So I guess, the reason is more that you don't want to compare time points where the dissolution has plateaued. That makes sense, in my opinion. However, when we have compounds that already plateau before 85% we can still have this issue. I have thought of 2 possible solutions: 1) Scale the dissolution, so that the maximum (of either product is 100%) and then calculate F2/mahalanobis as normal. 2) Define a way to detect the plateau and then use the first time point that reaches that plateau. For plateau detection we could use: have a 10% increase between each time point. Or fit a (Weibull) model and use that to define the plateau level. I personally like the second solution better, because it doesn't change the data. It behaves more like the normal rules, but it is also a bit more complicated. What do you think? Or are there some rules for instances like this, that I haven't found? |
|
nobody nothing 2017-01-03 20:19 (3029 d 13:22 ago) @ wligtenberg Posting: # 16924 Views: 6,441 |
|
|
Hi! As the experts are apparently still on holiday  I might start with some questions/remarks.Liposomes with doxorubicin (just wild guessing here  ) are no oral formulation afaik. Why would you do dissolution testing (simulation GI tract, normally)?The fraction of dose you get into solution pretty much depends on the media you use for your testing, in my opinion. So maybe some more details on the kind of products/projects your on might be helpful to get more response. — Kindest regards, nobody |
|
wligtenberg ☆ The Netherlands, 2017-01-04 11:46 (3028 d 21:55 ago) @ nobody Posting: # 16928 Views: 6,439 |
|
|
❝ As the experts are apparently still on holiday ❝ Liposomes with doxorubicin (just wild guessing here Wrong guess. :) ❝ The fraction of dose you get into solution pretty much depends on the media you use for your testing, in my opinion. ❝ So maybe some more details on the kind of products/projects your on might be helpful to get more response. In this case it is for eye drops. So some medium that resembles tears would make sense, right? So I do think that comparing dissolution makes sense in this case, even though it is not the GI tract. That still leaves the issue that we will not reach 85% dissolution in that medium. And I do think we should not keep many time points from the plateau phase. |
|
nobody nothing 2017-01-04 12:52 (3028 d 20:49 ago) @ wligtenberg Posting: # 16929 Views: 6,366 |
|
|
Dissolution testing for eye drops? WOW! Then I'm definitely out... Maybe more info from literature: https://www.ncbi.nlm.nih.gov/pubmed/27830515 https://www.ncbi.nlm.nih.gov/pubmed/26791934 https://www.ncbi.nlm.nih.gov/pubmed/24946794 — Kindest regards, nobody |
|
wligtenberg ☆ The Netherlands, 2017-01-16 18:27 (3016 d 15:14 ago) @ wligtenberg Posting: # 16956 Views: 6,189 |
|
|
Nobody on this forum has any experience/ideas about this? My own feeling goes more to identifying the plateau and limiting the data based on that. Because scaling messes with the F2 values. |
|
Dr_Dan ★★ Germany, 2017-01-16 19:10 (3016 d 14:31 ago) @ wligtenberg Posting: # 16959 Views: 6,273 |
|
|
Dear wligtenberg I am not experienced in this field, but if I look at EMA/CHMP/806058/2009/Rev. 02 see and take this analogously for products for ocular use I would ask you if it is really necessary to calculate f2 values in order to demonstrate similarity between the test liposomal product and the reference? Wouldn’t it be sufficient just to show the figures? According to this reflection paper IMHO you just need to address (among many others)
— Kind regards and have a nice day Dr_Dan |
|
wligtenberg ☆ The Netherlands, 2017-01-17 10:00 (3015 d 23:41 ago) @ Dr_Dan Posting: # 16962 Views: 6,167 |
|
|
Dear Dr_Dan, That was a very useful document, thank you. Although that document does not specifically state how to measure if the products are comparable, we would still need to show that they are. And I cannot imagine that eye-balling the graphs would cut it. Still, also during development, we would like to have some idea of how different batches compare as well, so I think we would still want some kind of distance measure and a generally accepted cut-off. But I do get from the document that you linked to, that for liposomes there are no fixed guidelines (yet). |
Ing. Helmut Schütz