Chromatography Forum

Please help:

Issue: Higher recoveries of up to 110% observed from dissolution run for a drug formulation

Compound: 5-methyl-2-[(4-methyl-1H-imidazol-5-yl)methyl]-2,3,4,5-tetrahydro-1H-pyrido[4,3-b]indol-1-one

Column: zorbax eclipse c18 50x4.6, 1.8u
Mobilephase: pH 3, potassium phosphate monobasic 80% with 20% acetonitrile
Column at 45c and wavelength at 220nm, flow 1ml
Sample concn 2ppm in water, injection volume 75 ul

Thanks

Two likely causes:

Std is degraded, giving low response.

Formulation has an impurity which coelutes with 5-methyl-2-[(4-methyl-1H-imidazol-5-yl)methyl]-2,3,4,5-tetrahydro-1H-pyrido[4,3-b]indol-1-one giving a higher response than expected.

Another non-chemistry cause is error in preparing solutions, but that could never happen, right?



Rod

standard solution stability verified and found to be stable for a week. Specificity also verified with individual components of placebo and found no interference. Filter study also fine. This compound has very high solubility in water.

An experiment done to match the dissolution conditions to verify if there is any temparature affect.

Experiment 1: Active and all placebo components mixed and diluted in water recovery found out to be about 90%

Experiment2: active and all placebo components mixed and diluted in water, kept at 37c for 30 min and bring it to RT. The recovery found out to be about 106%. It seems that there is some temparature affect on these.

Specificity: The most confusing issue is that if i do specificity with individual components of placebo with active recovery is fine. But if i do mix all placebo components together with active recovery found to be low. May be there are two issues, placebo interactions with active and temparature? I am confused if anybody has any ideas would be a great help.

Thanks

You have probably already looked at this, but are those recovery values (90% , 106%) repeatable and consistent? If not, it may be some other (uncontrolled) variable.

You did not address the possibility that the std may have been degraded. Not the std solution but the standard reference which was dissolved and put into solution.

Do you have a verification protocol in place that checks the purity of your reference materials? If you std has lost purity then you will be have a low response for your std solution and your test solutions will have a high recovery.

best wishes,

Rod

Thanks Tom. I have done only one time and did not confirmed the values of 90 and 106%. But when ever i do some dissolution analysis for the drug product recoveries found to be up to 110%. The high recovery for the drug product repeatable and consistent.

chromatographer1: thank you for your input. Right now i am using same API as a reference material which is used to formulate the drug product. Trying to compare the same thing (apples to apples).

Another interesting thing is that when i did the content uniformity which is 10ppm solution in water and injected 15ul using same conditions i have mentioned before recoveries found to be normal. The only difference between the CU and the dissolution methods are the sample concentration, temparature affect (dissos at 37 for 30min) and injection volume (15ul for CU and 75ul for dissolutions). Linearity studies done at both CU and disso methods and found to be normal. It is so confusing

What is your sample diluent.
What is the buffer strength in your mobile phase.

Sample diluent water
mobilephase: potassium phosphate monobasic 2.7gm to 2l water (i believe 20mM) add 6 ml of Triethyl amine and adjust pH to 3 with phosphoric acid.

Please check your calculation.

2.7 g in 1 liter (not 2 Liters) is 20 mM.

I suspect there is an inconsistency in the way you prepare your buffer for which you are getting varying recoveries.

Yes Mohan you are correct it is only 10mM. I have mentioned wrong. But the way we prepare buffer is consistent.

Is it possible to buffer the sample diluent (for both the dissolution and
content uniformity)?

Maybe 75uL of water is creating the injection precision issues.
If so, a buffered solution may help.

Yes for CUs i could change the diluent with buffer in it. But the CU are not the issue. The CU results i am getting with 10ul injs are OK. The dissolution method is a DBE and in order to change this we need to do lot of work. If there is any precision issue it would affect both stds and samples right?

Maybe evaluate different vials (including plastic) and/or septums.

Bryan,

thanks for your input. You brought up a good point. We have noticed low recoveries with clear vials but not with amber vials. But i still suspect something going wrong with amber vials especially with API/STD sample preparations coz if the std vial gives low response i would get the sample recoveries high. Is it possible the active sample solution can interact with glass? But i do not know how to prove or eliminate this. I appreciate any help

i like to try plastic vials could some one please give me some recommendations on plastic vials (like what kind/type etc)? I never used plastic vials in LC before.

Thanks

It looks like,

For CU : You are injecting 10 ppm at 15 uL inj volume.
For Disso: You are injecting 2 ppm at 75 uL inj volume.

Essentially the same sample load gets on the column in either case.

As per Bryan:

1. It could be an imprecision out of 75 uL inj volume. If your syringe volume is set at 100 uL (as with agilents), 75 uL inj is not a recommended value. You should aim at an inj volume no more than half, to obtain a good precision in your numbers.

2. Injection of pure water eventhough your compound is highly soluble is not a recommended choice, especially at the 2 ppm concentration. Due to surface tension, water will not be reproducibly introduced into the injection system - which eventually results in a lower or irregular compound recoveries.

I do not know if you can manipulate both the CU & Disso concentrations to 10 ppm, but if you can - go for it and inject 20 uL as the standard inj volume.

If not, make your diluent in a buffer (lowers surface tension) for injection purposes.