Bioequivalence and Bioavailability Forum • FDA Recommendation on Matrix Effect Determination in CC Assay

Obinoscopy
★

USA,
2018-10-01 16:38
(2427 d 17:14 ago)

Posting: # 19351
Views: 4,376

FDA Recommendation on Matrix Effect Determination in CC Assay [Bioanalytics]

Hi Guys,

I am going through the FDA BMV guideline and I noticed they didn't give any specific guidance on how to validate that a Chromatographic Assay method is free from Matrix Effect.

All I saw was this:

"Sponsor should ensure that there are no matrix effect throughout the application of the method"

No detail on how this could be achieved was provided.

For Ligand Binding Assays, they did give some hint on how to evaluate Matrix Effect but no Acceptance Criteria was provided.

Why is this so? Could it be that there are different school of thoughts on how to evaluate matrix effect and the FDA is trying not to take sides?

I personally feel they should have provided a method of evaluating matrix effect. Either they provide that which is in tandem with the EMA or they provide an alternative.

Or what do you think?

Regards,

—
Scopy

Ohlbe
★★★

France,
2018-10-01 17:30
(2427 d 16:22 ago)

@ Obinoscopy
Posting: # 19353
Views: 3,690

FDA Recommendation on Matrix Effect Determination in CC Assay

Post reply

Dear Scopy,

❝ I am going through the FDA BMV guideline and I noticed they didn't give any specific guidance on how to validate that a Chromatographic Assay method is free from Matrix Effect.

True. I find this is a real gap in their new Guidance. There was nothing about it in the 2001 Guidance, and not much more here, though matrix effects are a real pain in the quadrupole and can be dangerous.

❝ Why is this so? Could it be that there are different school of thoughts on how to evaluate matrix effect and the FDA is trying not to take sides?

No idea. I know that the FDA is trying not to be overly prescriptive, but they are giving much more details on other aspects of method validation.

We'll see what makes it into the ICH draft.

—
Regards
Ohlbe

ElMaestro
★★★

Denmark,
2018-10-01 18:30
(2427 d 15:22 ago)

@ Ohlbe
Posting: # 19355
Views: 3,719

FDA Recommendation on Matrix Effect Determination in CC Assay

Post reply

Hello both,

is there any method published anywhere which can be realistically thought to have no matrix effect?
Not just absence of mentioned effect, but truly verified absence of one and how would one go about proving it?

If you had a gun pointed to your head and someone asks you to develop an LC-MS/MS method without matrix effect for, say, Duloxetine or Simvastatin where would you start?

—
Pass or fail!
ElMaestro

Ohlbe
★★★

France,
2018-10-01 19:20
(2427 d 14:33 ago)

@ ElMaestro
Posting: # 19356
Views: 3,738

Reducing matrix effects

Post reply

Dear ElMaestro,

❝ is there any method published anywhere which can be realistically thought to have no matrix effect?

❝ Not just absence of mentioned effect, but truly verified absence of one and how would one go about proving it?

Define "no matrix effect". Same as concentration = 0: truly 0, or BLQ ? No matrix effect, or negligible ? Considering the analytical variability, I would consider it strictly impossible to demonstrate 0.000 matrix effect.

❝ If you had a gun pointed to your head and someone asks you to develop an LC-MS/MS method without matrix effect for, say, Duloxetine or Simvastatin where would you start?

Our Father who art in heaven...

I would start with:

using matrix obtained the same way as subject samples (vacutainers, not commercial plasma collected in blood collection bags full of plasticisers)
a real, proper sample cleaning (SPE or LLE, with a real development to really clean up, without too much loss in recovery. For instance, MTBE usually gives you a very good recovery of your drug... but also of quite a lot of things you'd like to get rid of)
a real chromatography (not a 90 seconds run on a 30 mm column)
comparing APCI and ESI. APCI is supposed to be less sensitive to matrix effects. Unfortunately, it is also less sensitive, full stop. May be problematic if a very low LLOQ is required (would not work for a number of the molecules you're working on).

And that's if I limit the definition of matrix effects to differences in detector response. You could expand the definition to include differences in recovery.

—
Regards
Ohlbe