NK
☆

India,
2021-04-26 07:46
(56 d 00:11 ago)

Posting: # 22323
Views: 755

## Fully Replicate Study [Study As­sess­ment]

Dear All,

We did BE study (both fasting and fed) for ER formulation of a drug (HVD), with fully replicated design (RTRT/TRTR) for USFDA. The protocol says, plasma samples from subjects who complete all the four periods will be analyzed and will be included for PK and statistical analysis.

Where we observed following:
• Subject A - had BLQ (Below limit of Quantification) for all the time points for one of Test treatment and
• Subject B - had BLQ for all the time points in one of Reference treatment.
• Subject C – had very low T/R Ratio (outlier was not mentioned in the protocol)
Can we include Subject A & B for statistical analysis (protocol not specified about handling subject with three periods data. (ie. one T & two R and one R & Two T).

To handle this kind of situation in future, Can we mention in the protocol to exclude the subject (outlier) with above scenario.

Regards
NK

Helmut
★★★

Vienna, Austria,
2021-04-26 10:24
(55 d 21:32 ago)

@ NK
Posting: # 22324
Views: 647

## Outliers in study intended for RSABE

Hi NK,

» The protocol says, plasma samples from subjects who complete all the four periods will be analyzed and will be included for PK and statistical analysis.

Sounds strange to me. IMHO, all samples should be analyzed.* Excluding subjects with incomplete data from the statistical comparison is another cup of tea.

» Subject A - had BLQ (Below limit of Quantification) for all the time points for one of Test treatment and
» Subject B - had BLQ for all the time points in one of Reference treatment.
»
» Can we include Subject A & B for statistical analysis

Well, the statistical model is based on $$\small{\log_{e}}$$-transformed data. You have to exclude them, since $$\small{\log_{e}(0)}$$ is undefined (for nitpickers: $$\small{\lim \log_{e}(x\rightarrow 0) = -\infty}$$).

» (protocol not specified about handling subject with three periods data. (ie. one T & two R and one R & Two T).

AFAIK, the FDA requires analysis of complete data (see this post).

» Subject C – had very low T/R Ratio (outlier was not mentioned in the protocol)

When it comes to RSABE, the FDA is not interested in assessing ‘outliers’. See this article.

» To handle this kind of situation in future, Can we mention in the protocol to exclude the subject (outlier) with above scenario.

Of course, you can. But I recommend a controlled correspondence with the OGD beforehand. Otherwise, you risk an R-t-R.

• Sometimes volunteers are afraid of not being recruited in later studies and hence, don’t report AEs. Instead, they withdraw consent ‘for personal reasons’.
There is a ‘classical’ example of methylphenidate. One subject was an ultra-slow metabolizer and his concentrations were 100 (‼) times higher than usual. Since MPH is a CNS stimulant, the subject had just a great time and stayed in the study. If the subject would have dropped out and his samples not analyzed, we would never have learned anything from it.

Dif-tor heh smusma 🖖
Helmut Schütz

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

France,
2021-04-26 16:30
(55 d 15:27 ago)

@ Helmut
Posting: # 22325
Views: 606

## Outliers in study intended for RSABE

Dear Helmut,

» » Subject A - had BLQ (Below limit of Quantification) for all the time points for one of Test treatment and
» » Subject B - had BLQ for all the time points in one of Reference treatment.
» »
» » Can we include Subject A & B for statistical analysis
»
» Well, the statistical model is based on $$\small{\log_{e}}$$-transformed data. You have to exclude them, since $$\small{\log_{e}(0)}$$ is undefined (for nitpickers: $$\small{\lim \log_{e}(x\rightarrow 0) = -\infty}$$).

Well, indeed you can't analyse the study without excluding at least that period. NK's study is intended for the US FDA, I don't know how these guys react to such exclusions. In Europe, subject B could be excluded according to the EU BE guideline, but subject A could not as the period with all BLQ was after the test product. EU regulators will tell you that this might be due to your product and you can't exclude the data just because they don't suit you. In such a situation: you're screwed.

Regards
Ohlbe
mittyri
★★

Russia,
2021-04-26 17:01
(55 d 14:55 ago)

@ Ohlbe
Posting: # 22326
Views: 608

## failed study?

Dear Ohlbe,

» In Europe, subject B could be excluded according to the EU BE guideline, but subject A could not as the period with all BLQ was after the test product. EU regulators will tell you that this might be due to your product and you can't exclude the data just because they don't suit you. In such a situation: you're screwed.

what is the proposed solution for such situation?
say the subject did not swallow the IMP (just for fun) and investigators missed that.
There's no way to prove this without direct subject's confession. Game over? The study failed?

Kind regards,
Mittyri
Helmut
★★★

Vienna, Austria,
2021-04-26 22:33
(55 d 09:24 ago)

@ mittyri
Posting: # 22328
Views: 605

## Failed study

Hi mittyri,

» what is the proposed solution for such situation?
» say the subject did not swallow the IMP (just for fun) and investigators missed that.
» There's no way to prove this without direct subject's confession. Game over? The study failed?

Yep. To quote myself:

»[…] we cannot distinguish between lack of compliance (the subject didn’t take the drug), a product failure, and a subject-by-formulation interaction (the subject belongs to a subpopulation).«

Dif-tor heh smusma 🖖
Helmut Schütz

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

Vienna, Austria,
2021-04-26 22:28
(55 d 09:29 ago)

@ Ohlbe
Posting: # 22327
Views: 611

## Failed study

Dear Ohlbe,

» NK's study is intended for the US FDA, I don't know how these guys react to such exclusions.

IMHO, that would by a case for R-t-R, Section VI.A. because nothing about excluding ‘outliers’ is stated in any of the product-specific recommendations I’m aware of. That’s why I suggested a controlled correspondence for the next study to hope for Section VI.B.

» In Europe, subject B could be excluded according to the EU BE guideline, but subject A could not as the period with all BLQ was after the test product. EU regulators will tell you that this might be due to your product and you can't exclude the data just because they don't suit you. In such a situation: you're screwed.

Correct.

Dif-tor heh smusma 🖖
Helmut Schütz

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