Bioequivalence and Bioavailability Forum • FDA definition of non-accurately and reliably

jag009
★★★

NJ,
2013-01-09 23:20
(4522 d 00:24 ago)

Posting: # 9808
Views: 9,143

FDA definition of non-accurately and reliably [Bioanalytics]

Hi everyone,

Can someone clarify this for me... When a FDA draft guidance says bioequivalence is based on the prodrug unless "it is not possible to measure in plasma accurately and reliably", what does it really mean? Does it mean the LoQ has to be low enough? What about the overall variability of the concentration values?

The reason I ask is because I have a prodrug with highly variable concentration values. At the dose I studied in a pilot involving only the reference product, 8-9 out of 24 subjects had almost all BLQs in their entire PK profiles. For these 8-9 subjects, their detectable concentration values are not much higher than the LoQ value). The rest of the subjects are ok with only BLQ values in the last 6 timepoints (24 total). The prodrug is poorly water soluble.

The pivotal study will be dose at 2 x my pilot dose. Assuming linear kinetics and the resulting concentration will be doubled.

Thanks
John

Helmut
★★★

Vienna, Austria,
2013-01-10 15:03
(4521 d 08:42 ago)

@ jag009
Posting: # 9816
Views: 7,959

Experiences with EU submissions

Post reply

Hi John,

no experiences with the FDA so far. In the few similar cases for European submissions we had data demonstrating that we were pushing the analytical method to the state of the art’s limits. We also had pilot studies showing that the methods were not capable of delivering reliable PK-metrics. In all cases we went to scientific advisory meetings. The metabolite was accepted – in one case (December 2012) even a secondary one!

—
Dif-tor heh smusma 🖖🏼 Довге життя Україна!
Helmut Schütz

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes

jag009 ★★★ NJ, 2013-01-10 17:36 (4521 d 06:08 ago) @ Helmut Posting: # 9817 Views: 7,972	Experiences with EU submissions Post reply
	Thanks Helmut, When you wrote "We also had pilot studies showing that the methods were not capable of delivering reliable PK-metrics". What is your definition of reliable in that case? LoQ not low enough? Similar to my case where almost 1/2 the subjects have almost all BLQ values in their profiles? John

Helmut
★★★

Vienna, Austria,
2013-01-10 17:56
(4521 d 05:49 ago)

@ jag009
Posting: # 9819
Views: 7,980

Experiences with EU submissions

Post reply

Hi John,

❝ […] What is your definition of reliable in that case? LoQ not low enough? Similar to my case where almost 1/2 the subjects have almost all BLQ values in their profiles?

Mainly LOQ. In another case the LOQ was low enough but we had to deal with an endogenous compound. Sampling an entire blank profile showed that the values were (very, very) highly variable. Subtracting lead no nonsense.

—
Dif-tor heh smusma 🖖🏼 Довге життя Україна!
Helmut Schütz

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes

jag009
★★★

NJ,
2013-01-10 21:07
(4521 d 02:38 ago)

@ Helmut
Posting: # 9820
Views: 8,018

Experiences with EU submissions

Post reply

Thanks Helmut,

Do you consider 1ng/mL LLoQ sufficient with the current state of art equipments? I think it's too high (Sciex 5500+ can do better).

With the drug I am working with, based on pilot data, will be like 120 ng/mL for Cmax (the highest Cmax, the study was 200 mg so assuming 2x200mg). The range now is 1ng/mL - 200 ng/mL. I think I can push down the ULoQ to 100 ng/mL and do dilution for higher conc samples.

John

Helmut
★★★

Vienna, Austria,
2013-01-10 21:46
(4521 d 01:58 ago)

@ jag009
Posting: # 9821
Views: 8,055

LLOQ “low enough”?

Post reply

Hi John,

❝ Do you consider 1ng/mL LLoQ sufficient with the current state of art equipments? I think it's too high (Sciex 5500+ can do better).

Can’t tell. The lowest level one possibly can reach depends on the molecule, ionisation sites, clean-up/trace-enrichment, blah blah. I have seen a validated method for formoterol with an LLOQ of 400 fg/mL already eight years ago. That’s 2500times lower than your 1 ng/mL – but likely another drug, right? :-D

❝ With the drug I am working with, based on pilot data, will be like 120 ng/mL for Cmax (the highest Cmax, the study was 200 mg so assuming 2x200mg). The range now is 1ng/mL - 200 ng/mL. I think I can push down the ULoQ to 100 ng/mL and do dilution for higher conc samples.

Sounds OK, but didn’t you see no values >LLOQ in ~40% of the subjects?

—
Dif-tor heh smusma 🖖🏼 Довге життя Україна!
Helmut Schütz

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes

jag009
★★★

NJ,
2013-01-11 16:20
(4520 d 07:24 ago)

@ Helmut
Posting: # 9827
Views: 7,994

LLOQ “low enough”?

Post reply

Hi Helmut,

❝ Sounds OK, but didn’t you see no values >LLOQ in ~40% of the subjects?

Sorry typo :-P

... I meant to push down the ULoQ to 100ng/mL (or 150 ng/mL from current 200ng/mL) and lower the LLoQ to below 1ng/ml(Not sure how low yet. As you mentioned I will need to check the chemist first). I think with a 0.25-0.5ng/mL as LLoQ maybe doable.

If 40% of the subjects still show many BLQs with some detectable values not much higher than the LLoQ while the other 1/2 show normal concentration ranges then I don't see how I can justify that the parent compound measurement is deemed reliable for bioequivalence assessment. I don't think I will see this happening with future studies having 2x the dose though...

Thanks
John

Ohlbe
★★★

France,
2013-01-11 20:16
(4520 d 03:28 ago)

@ jag009
Posting: # 9830
Views: 7,912

LLOQ “low enough”?

Post reply

Dear John,

❝ I think with a 0.25-0.5ng/mL as LLoQ maybe doable.

❝

❝ If 40% of the subjects still show many BLQs with some detectable values not much higher than the LLoQ while the other 1/2 show normal concentration ranges then I don't see how I can justify that the parent compound measurement is deemed reliable for bioequivalence assessment.

In December 2009, CT Viswanathan participated in a meeting of the French Society for Quality Assurance (SoFAQ) in Paris. Somebody asked a question like "but if it is difficult to...". Vish's answer was:

what do you call difficult ? You'll have to show how hard you tried, not just say you did not succeed
if the lab you're using does not have the equipment needed to do the job: go to a better equipped lab
nobody said you had an easy life and an easy job...

Vish has left the FDA, but the spirit is probably still there. I don't know what molecule you're dealing with, maybe it is particularly difficult to ionise in MS/MS. Otherwise there are chances that going down to 0.25 ng/ml or lower will not sound such a challenge to the average FDA reviewer, with the latest LC-MS/MS systems and a proper sample preparation.

Regards
Ohlbe

—
Regards
Ohlbe

jag009 ★★★ NJ, 2013-01-11 21:14 (4520 d 02:30 ago) @ Ohlbe Posting: # 9831 Views: 7,808	LLOQ “low enough”? Post reply
	Thanks Ohlbe, The lab that I am working with said their MV with 1ng/mL - 200ng/mL is acceptable but they didn't elaborate. It kind of troubles me because I have existing data (they ran a reference vs reference study for me) to state my concerns to them. Granted that I did the study at 200 mg and the submission studies will be 400 mg. Anyhow I am waiting for their reply. John