Log in
Register|
Loky do ★ Egypt, 2024-11-10 12:52 (71 d 18:07 ago) Posting: # 24267 Views: 963 |
|
|
Dears I am working on fed and fasted studies for a bilayer drug containing empagliflozin/metformin XR. In the fed study, one volunteer showed no measurable concentration of empagliflozin for the test product, although a measurable concentration of metformin was observed in the same individual. The sponsor asserts that empagliflozin does not require a reliable concentration in the fed state since it is intended to be taken in the fasted state and is immediate-release, which would, in their view, negate the need for fed-state evaluation. We attempted to include this subject in the statistical calculations; however, SAS and WinNonLin automatically excluded the subject due to the missing concentration for empagliflozin, resulting in a passed study. As the CRO, we have raised concerns, viewing the lack of measurable concentration for empagliflozin in the fed state as a potential issue regarding product conformity. How can we configure the statistical software to include this subject in the analysis, potentially yielding a failed study outcome? Alternatively, what rationale could we provide in the study documentation to ensure regulatory acceptance of our approach and prevent potential questions for the CRO?. Thanks in advance Edit: Category changed; see also this post #1. [Helmut] |
|
Helmut ★★★   Vienna, Austria, 2024-11-10 13:33 (71 d 17:26 ago) @ Loky do Posting: # 24268 Views: 811 |
|
|
Hi Loky do, ❝ […] one volunteer showed no measurable concentration of empagliflozin for the test product, although a measurable concentration of metformin was observed in the same individual. […] ❝ We attempted to include this subject in the statistical calculations; however, SAS and WinNonLin automatically excluded the subject due to the missing concentration for empagliflozin, resulting in a passed study. ❝ As the CRO, we have raised concerns, viewing the lack of measurable concentration for empagliflozin in the fed state as a potential issue regarding product conformity. ❝ How can we configure the statistical software to include this subject in the analysis, potentially yielding a failed study outcome? ![[image]](https://static.bebac.at/img/Rlogo_15_12.svg) to calculate log(0) we will get -Inf as the – mathematically correct – answer. Nice, though not helpful if we want to use such a value in further calculations. Only by convention we say that \(\log(0)\) is ‘undefined’. Therefore, any \(\small{x\leq0}\) is either excluded in software or throws an error. There is no way to get around it.You could only perform a sensitivity analysis setting all BQLs to say, LLOQ/2. Unless the study is not very large and the CV very small, quite likely it would fail. ❝ Alternatively, what rationale could we provide in the study documentation to ensure regulatory acceptance of our approach and prevent potential questions for the CRO?.
— 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 |
|
Loky do ★ Egypt, 2024-11-10 13:48 (71 d 17:11 ago) @ Helmut Posting: # 24269 Views: 802 |
|
|
Hi Helmut Thanks for your reply ❝ Actually \(\lim_{x \to 0}\log(x)\small{=-\infty}\) and therefore, if we ask ❝ You could only perform a sensitivity analysis setting all BQLs to say, LLOQ/2. Unless the study is not very large and the CV very small, quite likely it would fail. as we use SAS im calculation, are there any way to calculate with SAS, in addition what if we assumed that lin(PK) equal zero, is this applicable ? another question could we consider these values as 5% of loq not LLOQ/2 as the observed results around 5-10 % of LLQ Thanks |
Ing. Helmut Schütz