Prednisone Lot O0C056

[thedissokid]

We received a new lot of Prednisone, Lot O0C056, this week. I was wondering if anyone had attempted a calibration with this new lot. Here's hoping that its better than the last Lot.

Thanks

[biomartin]

I've been exposed to many sets of results with this new calibrator and have seen no issues to date.

[ismithlj]

I had no problems with lot N for maybe the last 6 months and as soon as I switched to lot O0C056 I began failing on several units. I had one unit fail with both a high value and a low value. I stopped using the bottle that was giving me all the failing results and have not had any problems with other bottles.

[koufist]

Yes. We had failures for the last two months with one of our automated units. Both Salycylic runs passed, the prednisone paddles passed, but the prednisone basket test has failed in about 15 attempts. In every attempt we have tested different factors (new baskets, different ways of degassing medium, new bottles of tablets, different types of water source). We had one or two tablets failed in every one of those runs. Again we have used both lot N and lot O0C056.

[Rijkers]

I'm very glad that finaly a new lot has been released! With this lot I had no problems at all. I just finished calibrating 3 dissolution bath: not one value was out of range. While using lot N last year, it was some kind of disaster: every bath had to be calibrated for at least 2 times (paddels and baskets).

[LOLA]

We received a new lot of Prednisone, Lot O0C056, this week. I was wondering if anyone had attempted a calibration with this new lot. Here's hoping that its better than the last Lot.

Thanks
I have perform 3 apparatus with this lot and all of them inside of the limits. One of the apparatus is a full-automatic *****, USP1 and USP2, that detects any problem immediatly.

Have a nice calibration,
LOLA

;)

[syx]

I have tried this new lot last week and got results:
64.7 - 83.5% (apparatus 1) and 45.1 - 53.9% (apparatus 2).

[vloeto]

I have tried this new lot last week and got results:
64.7 - 83.5% (apparatus 1) and 45.1 - 53.9% (apparatus 2).
We have found the same rediculous high results for our apparatus 2 (48-52%). The solution porved to be very simple: it was just a calculation error.

As for the baskets, we have seen that the prednisone result is very sensitive for dirty vessel/baskets. Results may drop raddically after cleaning in a dish washer.
Furthermore, we also have had high results (just above the criterion) when the temperature of the medium is somewhat high (> 37.0 °C).

[qureshi]

syx and vloeto: I refer to some of my ealier posts e.g. http://www.dissolution.com/vbulletin/showthread.php?p=3913#post3913 and corresponding threads for infromation in this regard. Hope these will be helpful. Saeed

[Kronos]

It should be clarified that the goal of performing the calibration test is not to pass. It is to validate the suitability of the dissolution tester. However, if the calibrator tablets fail to perform as expected, and do not respond in predictive behaviors based on known physical parameters, and instead perform based on variables that cannot be controlled, then it is the calibrators tablets themselves that have failed. USP has failed to answer some important questions regarding the predenisone basket test. For example, was there some significant problem with the last lot (lot N)? Why are there so many concerns in the boards here that the test will fail high? Is it possible that the single most critical variable which determines the rate of release is whether the disintegrated sample clumps on one side, as opposed to forming a uniformed bed? Does the rate of release changed by the type of purified water used, or from what source the water is purified from. Is the release effected by surface tension based on the type of purified water used? Since the current lot (O0C056) is exactly the same formulation as the last lot, barring some compression force differences, they should behave equally (as has been the case for salicylic acid tablets from lot to lot). However, USP has changed the ranges on the new lot, and now claim that the ranges are set with 99% confidence limit. Hmmmm...

[Kronos]

This tablet is exactly the same formulation as the last lot (Lot N). I am experienceing signficant problem with this lot, as I have with the previous lot. Please contact USP if you have failures, or have concerns about the tablets. Their party line is that no one really has a problem with this lot, and any problem that has been raised with them has been "resolved".

[ganesh.s]

I am new comer for these group. Recently I have conducted calibration using the Lot O0C056. The calibration pass for apparatus I but failed for apparatus II, whereas the test was conducted under identical way. Only one tablet pass will the rest below the lower limit (32.59 - 46.19%). Could you please advice me on these issue.

[dpupovac]

We are observing re-occurance of USP Prednisone Tablets, Lot O0C056 exceeding the USP acceptance criteria for baskets. There is always one or two tablets slightly higher than 77%. Could you please advice me on these issue.

[qureshi]

Please keep USP informed about it. Out of spec. values for calibrator tablets are usually not reflection of a problem of dissolution testing (equipment, tablets or analysts), but a problem of tighter tolerance range (specifications). Tighter ranges do not truly reflect expected variability in dissolution results. Therefore, one should expect to observe out of range values. Sorry, could not be more helpful.

Saeed

[BanditVJ]

We to are having problems calibrating equipment using this batch, with our results being above the USP parameters even though calibration using Salicyclic acid was spot on.

We have used a fresh, unopened pot of standard, fresh unopened pot of prednisone tablets, new filters for extraction, media degassed with a vacuum degasser as well as using a Helium sparger. Two analysts have checked and rechecked the calculation. User error has been disqualified as the analyst doing the calibration was observed throughout by a second analyst.

Short of pulling ones hair out are there any other ideas as to what else can be done in an attempt to bring the results within the acceptance criteria?

Hayelp!!!

:confused:

[qureshi]

As I said earlier unfortunately there is no immediate solution to calibration problems. Until and unless USP is convinced about the problems (inherent deficiencies of the apparatuses), industry has to live with the frustration. Question is how to convey the message to the USP so that respective experts committee starts looking into the problem.

One option would be by conveying the message through USP Calibration Working Group (USP Project Team #5?). Would the Working Group be interested in pursuing it further. In my opinions, there are plenty of examples available in the literature to explain the issue and possibly to address it.

Any one from the working group likes to respond?

Saeed

Disclaimer: Views expressed here are for scientific discussion purposes only and not necessarily reflect opinions and policies of my employer.

[BanditVJ]

After trying all possible outlets to solve the problem we were having regarding calibration, we finally called in the service engineers to look at our apparatus. Despite the run-out at the bottom of the paddles being well with USP standards (max 2mm, ours were 0.75mm) we opted to change the paddles for a brand new set.

This appears to have resolved the problem, our Prednisone calibration results are in the middle of the range required by USP and we are happy as it means we can try and clear the backlog of work that built up having an apparatus out of use!

Many thanx to all who posted comments and offered assistance. :D

[qureshi]

Obviously, you were working with an instrument which was within the spec., however, changing spindles made the values fall within specs. One may argue that it may be considered as repeat testing. This has been our observation as well, based on published results, that just by repeating the test one can get values within spec., and vice versa.


Saeed

[kgauger]

Saeed,
If they attribute the failing test result to a specific set of paddles wouldn't all of the data generated with those paddles then be suspect? Yet another problem of having "calibrator" tablets which can fail with equipment that is within specifications.

[qureshi]

Kevin: I agree with you and have expressed this concern to peers a few times that in such situations results (in principle) may be considered invalid. However, peers prefer to keep quite. Otherwise it will open a huge can of worms. Big time (think of financial implications, and others)!

In one of our studies based on testing in 28 laboratories, it has been shown that results obtained for a pharmaceutical (commercial) product using apparatuses within specs., or out of specs (based on calibration ranges) do not show any significant differences. Perhaps, such results will help people to get of the hook, if issue arises.

Then the real question is, if one can obtain same or similar results, using apparatuses, which are within specifications or out of specifications, then what is the use of calibration. Exactly, that has been my view. Unfortunately, with calibrator tablets what we are measuring is variability (shortcomings/failures) due to the testing (technique) not of products, analysts etc. We need to convey this message to USP, FDA and in my view especially to industry, that there are problems with technique which needs to be addressed.

Saeed

Disclaimer: Views expressed here are for scientific discussion purposes and not necessarily reflect opinions and policies of my employer.

[Danielson]

I just finished up calibrating 7 instruments. We have three apparatuses that are Brand X and four that are Brand A. After having 9 failures with the Brand X baskets, I took the shafts and baskets from Brand A and put them on the Brand X instruments. I got three successful calibrations in a row!!

[qureshi]

Last week we also received the Technical Sheet for the USP Prednisone Tablets RS describing the revised ranges. The reason indicated for the change is the use of a revised approach of statistical analysis. It would be safe to assume that, their first statistical approach was not correct. Hopefully, it will be correct this time? Only time will tell.

This confirms our views that based on results from experimental studies, OOS values do not reflect lack of expected performance of equipment or analysts. It just reflects the high variability of the technique, which causes the OOS. Very recently, I posted on the board saying that often time problem is with setting of the ranges (tighter or not reflecting observed 95% CI), and thus ranges have to be adjusted to reflect the reality (expected high variability in results).

Obviously some, who believed in and advised others, must be feeling embarrassed by accepting and believing in the commonly stated opinions and assumptions as scientific facts regarding causes of OOS values such as related to de-aeration, vibration, and others. I hope it will be an alarming and learning experience for them. They should express their concerns to USP.

I do not think that problems of dissolution testing or calibration itself can be resolved by just adjusting (or widening) the ranges. Re-adjusting or widening of ranges establishes problems and instability (lack of repeatability/reproducibility) of technique, as others and we have clearly indicated.

Some corrective measures are clearly needed to address the deficiencies/problems of the technique, so that the testing properly reflects the quality of pharmaceutical products, as currently it is not.

Saeed

Disclaimer: opinions expressed here are for scientific discussion purposes only and may not necessarily reflect opinions and policies of my employer.

[jwb0071]

Hello All,

I have been performing bath calibrations on my totally automated MultiDose dissolution platform and I too have seen OOS results in accordance with Lot#O0C056. Since my method is totally automated it would bring into question the calibrator tablets as the MultiDose has a significantly less chance of variablity in sample analysis than a manual pull from the vessel. My calibration was OOS for three of the six vessels and I have checked everything on the system for any type of mechanical malfunctions and there are none. I am at a loss to explain why the tablets failed. This happened on another system and we simply repeated the test and while everything passed then they were all still on the high side. Just to note that all automated results were confirmed manually from the collected samples in the MultiFill units of the system.