Ladi
☆    

Thailand,
2016-02-23 10:22
(2956 d 00:55 ago)

Posting: # 16022
Views: 8,833
 

 calibration curve range [Bioanalytics]

Dear forum members,

I have questions regarding the calibration curve range considering the statement in "EMEA Guideline on bioanalytical method validation" that says "At least 2 QC sample levels should fall within the range of concentrations measured in study samples". My first question is whether this statement is considered acceptance criteria of each analytical run?

Supposedly, we validated a calibration curve range of 10-100 ng/mL, and LQC is 30, MQC is 50 and HQC is 80 ng/mL. Now please let me set up two scenarios as following.


Scenario 1--- After we analyzed our first batch of study samples, the Cmax are:

subject 1
period 1 is 48 ng/mL
period 2 is 35 ng/mL

subject 2
period 1 is 52 ng/mL
period 2 is 65 ng/mL

subject 3
period 1 is 70 ng/mL
period 2 is 30 ng/mL

Questions:

Since 2 QC levels do not fall within the range of drug concentrations measured from subject 1 (in both periods) and subject 3 (period 2),

1) Is this analytical run accepted?

2) Would you continue with the study or what would you do next?

3) If we continue the study with same curve and QC and the rest of data are somewhat in this trend, can authority reject this study?

Scenario 2---After we analyzed our first batch of study samples, the Cmax are:

subject 1
period 1 is 44 ng/mL
period 2 is 48 ng/mL

subject 2
period 1 is 47 ng/mL
period 2 is 45 ng/mL

subject 3
period 1 is 48 ng/mL
period 2 is 49.5 ng/mL

Similar questions:

Since 2 QC levels do not fall within the range of drug concentrations measured from every subjects,

1) Is this analytical run accepted?

2) Would you continue with the study or what would you do next?

3) If we continue the study with same curve and QC and the rest of data are somewhat in this trend, can authority reject this study?

I really appreciate your time in reading and answering my questions. So sorry for asking similar topic again in this forum but we still having this big issue.

Thank you,
Ladi
(Thailand)


Edit: Category changed. [Helmut]
ritesh.korat.pharma
☆    

India,
2016-02-26 20:35
(2952 d 14:42 ago)

@ Ladi
Posting: # 16032
Views: 7,234
 

 calibration curve range

Dear Ladi

To satisfy EMEA guideline which states that minimum two QC covered within Cmax of that particular period you can take following steps

Firstly monitor around 20% of first subject of the study. if Cmax of both periods of that subjects (at least 67% among them) will cross two QC than no need of additional QC and if its not cross than add one more QC between LQC and MQC. The all procedure must be incorporated in SOP prior to analysis

E.g. if you have to perform analysis of 100 subjects then you can monitor first twenty subjects among them

20% of 100 subjects is 20

if Cmax of both period in 14 (67% of 20) subjects out of that 20 subjects cross two QC level than no need of additional QC and if its not than add one more QC between LQC and MQC. so, in remaining subjects cmax cross two QC levels

Regards,
Ritesh Korat
Ladi
☆    

Thailand,
2016-03-03 07:02
(2947 d 04:15 ago)

@ ritesh.korat.pharma
Posting: # 16043
Views: 6,992
 

 calibration curve range

Dear Ritesh,

Thank you for your suggestion. I will take your suggestion to our team for further discussion.

Regards,
Ladi
Ohlbe
★★★

France,
2016-03-03 17:46
(2946 d 17:31 ago)

@ ritesh.korat.pharma
Posting: # 16046
Views: 7,045
 

 calibration curve range

Dear Ritesh Korat,

❝ Firstly monitor around 20% of first subject of the study.


Why wait till 20 %, if it is already visible after 2-3 runs that the concentrations are lower than expected ?

❝ if Cmax of both period in 14 (67% of 20) subjects out of that 20 subjects cross two QC level than no need of additional QC


Why 67 % ? Where did you take that figure from ?

IMHO what's important is not just how many subjects have a Cmax below the MQC sample - but also how much below it is.

❝ and if its not than add one more QC between LQC and MQC.


Or adjust the concentration of the existing MQC. The guidelines also gives this option.

Regards
Ohlbe
Ohlbe
★★★

France,
2016-03-03 17:41
(2946 d 17:36 ago)

@ Ladi
Posting: # 16045
Views: 7,048
 

 calibration curve range

Dear Ladi,

❝ My first question is whether this statement is considered acceptance criteria of each analytical run?


Clearly not:
  • this is not mentioned in section 5.2 (Acceptance criteria of an analytical run) but in section 5.3 (Calibration range);
  • section 5.3 states that It is not necessary to reanalyse samples analysed before optimising the standard curve range or QC concentrations.

❝ Supposedly, we validated a calibration curve range of 10-100 ng/mL, and LQC is 30, MQC is 50 and HQC is 80 ng/mL. Now please let me set up two scenarios as following.


With just a 10-fold difference between the LLOQ and the ULOQ you're in trouble...

❝ 2) Would you continue with the study or what would you do next?


I would stop and lower the LLOQ to be able to measure concentrations down to 5 % of Cmax... SCNR. I understand that you took this as a theoretical example.

To answer your questions:
  • Each run is acceptable. As mentioned above, this should not be considered as an acceptance criterion for each run.
  • Problems may not be seen right from the very fist run, but only in the 2nd or 3rd.
  • It can happen that some subjects will have a Cmax lower than the MQC sample - even after changing the MQC concentration or adding an extra QC level. No big deal. At least you can show that you did look at your data and that you did make the effort to adjust. If you make your MQC concentration too low in order to be sure it remains below the lowest Cmax, some regulators may say it is too low compared to the highest Cmax...
The idea is that the concentrations of the QC samples should not be totally disconnected from the concentations in the study samples. But the guideline leaves some flexibility.

Regards
Ohlbe
wienui
★    

Germany/Oman,
2019-02-20 19:13
(1862 d 16:04 ago)

@ Ohlbe
Posting: # 19943
Views: 3,863
 

 calibration curve range

Dear all,

In a BE study, the average Cmax of subjects samples was found out of the ULOQ of the calibration curve.

Please enlighten me, is this acceptable?

Best regards,
Osama

Cheers,
Osama
Helmut
★★★
avatar
Homepage
Vienna, Austria,
2019-02-20 19:19
(1862 d 15:58 ago)

@ wienui
Posting: # 19944
Views: 3,833
 

 calibration curve range

Salam Osama,

❝ In a BE study, the average Cmax of subjects samples was found out of the ULOQ of the calibration curve.


You’ve gotta be kidding! :-D

Dif-tor heh smusma 🖖🏼 Довге життя Україна! [image]
Helmut Schütz
[image]

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

Germany/Oman,
2019-02-20 19:32
(1862 d 15:45 ago)

(edited by wienui on 2019-02-21 05:30)
@ Helmut
Posting: # 19945
Views: 3,912
 

 calibration curve range

❝ Salam Osama,


❝ ❝ In a BE study, the average Cmax of subjects samples was found out of the ULOQ of the calibration curve.


❝ You’ve gotta be kidding! :-D


Salam Helmut,

No, my friend, I hoped I were kidding but unfortunately this is the reality. Moreover, it was given as a justification that samples were reanalyzed after dilution with blank human plasma to bring the analyte concentration into the validated calibration curve range which then multiplied by the dilution factor and reported in the final table of concentrations and that only 5% of subjects'samples have concentrations above ULOQ ( although we speak about the average Cmax!!!)

very nice!!!!
regards,
Osama

Cheers,
Osama
UA Flag
Activity
 Admin contact
22,957 posts in 4,819 threads, 1,639 registered users;
73 visitors (0 registered, 73 guests [including 5 identified bots]).
Forum time: 11:17 CET (Europe/Vienna)

Nothing shows a lack of mathematical education more
than an overly precise calculation.    Carl Friedrich Gauß

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5