Bioequivalence and Bioavailability Forum

Hi Ben and all,

❝ Well, another approach would be to fight for getting all measured values, including the ones where the lab tells us they are flagged as BLQ …

Join “El ingenioso hidalgo Don Quijote de la Mancha” in his tilting at windmills…

❝ … (yes, yes, I know they are "not reliable", whatever that means).

Easy. At the LLOQ: Accuracy ≤20% and precision ≤20% in chromatography, ≤30% in ligand binding assays. Can be higher, if justified (hard data demonstrating that it is impossible to comply with the rules).
<nitpick>

In chemistry we are bound to the IUPAC’s terminology: inaccuracy, imprecision.
A method with an accuracy of 20% would be useless.* 

</nitpick>

❝ But I have the feeling it is still better to use them than to ignore or set them to some fixed value.

Agree. Gut-feeling as well.

❝ I had a nice discussion with Helmut about this topic.

A summary: At the first Crystal City meeting about bioanalytical method validation (Arlington 1990) there were heated debates about the topic. Essentially there were two parties: Regulators wanted to have a ‘general rule’ in order to avoid discussions with applicants. Members of the PK modeling community strongly opposed that:

• Excluding concentrations based on an – arbitrary – cut off leads to truncated distributions and biased estimates.
  
• Give us all you have! We incorporate the error already in the model (most used a mixed approach: multiplicative+additive). The more data we have, the better. Yes, we are aware that at the end of the day something below the LOD is not possible.

At the archives of David Bourne’s PKPD-list you find a lot of related discussions (LLOQ, BQL). Experienced modelers like Roger Jeliffe, Nick Holford, and Hans Proost refused the idea of dropping BQLs. Quoting Nick Holford:

There is no reason not to use these values. It is just silliness that chemists fail to give you the measurements because of an arbitrary cut off that has no real meaning for pharmaco­kinetic analysis. Omitting these values will always cause bias.
One thing is sure about the true concentration – until sufficient time has passed for less than one molecule to be left in the body then the concentration is not 0. This is longer than most people live…

Hence, in my CRO we had an SOP:

• For BE we reported BQL to comply with the GLs.
  
• For PK modeling, we reported LOD ≤ measured < LLOQ with an asterisk and a footnote explaining that for these values A&P >20%. 

❝ … question is also how to simulate such unreliable data...

By (truncated?) lognormal distributions.

❝ is it just data with higher variability?

Nope. Higher (in)accuracy as well.

---

• There was not a single (‼) chemist in the group writing the EMA’s guideline on bioanalytical method validation… No offense but the fact that I have a heart does not qualify me to write a guideline for cardiology.