Skip to main content
Home / Resources / Blog / Reference-scaled Average Bioequivalence

Reference-scaled Average Bioequivalence

The standard approach for approval of generic drugs is to run a bioequivalence study to demonstrate that a generic product is comparable to an  approved (ie, reference) drug in their rate and extent of absorption. The rate and extent of drug absorption are determined from the pharmacokinetic parameters: peak concentration (Cmax) and the area under the concentration-time curve (AUC), respectively. The approach is referred to as average bioequivalence (ABE) where the 90% confidence interval for the ratio of the average geometric means (test/reference)  for AUC and Cmax must fall between preset regulatory bioequivalence limits from 80% to 125%.

Reference-scaled average bioequivalence (RSABE) is a statistical methodology that is increasingly used to demonstrate bioequivalence for highly variable drugs (HVDs). A drug product is called highly variable if the intra-subject (ie, within-subject) variability is greater than 30% C.V. (coefficient of variation) in the pharmacokinetic measures of AUC and/or Cmax. In other words, if you take the same drug in two different occasions at similar conditions (eg, same dosage, administration route, fasted, same time a day, etc.) you would expect the measured AUC and Cmax to be very similar regardless of the time of administration. However, if instead the rate and extend of absorption differ by more than 30% between the occasions then the drug is considered highly variable. In those cases, running ABE with the standard sample size will likely fail to show bioequivalence due to the intrinsic variability even if the products were comparable. Indeed, some HVDs have failed to show bioequivalence to itself using standard ABE sample sizes. For HVDs, studies designed to show bioequivalence may need to enroll large numbers of subjects, even when the formulations themselves have no significant mean differences. This increases the expense of BE studies, places more subjects at risk, and ultimately, limits the availability of generics.

The RSABE method allows scientists to scale the acceptance bioequivalence window based on the within subject variability of the reference drug. Thus, the limits of the ABE can be scaled to the reference variability (ie, the permitted window increases as the variability increases). RSABE methods can be applied to show bioequivalence if the within subject variability for the reference drug has been shown to have at least 30% CV.

Specifics of RSABE methodology vary between regulatory agencies, but both the European Medicines Agency (EMA) and the United States Food and Drug Administration (FDA) require that subjects receive the reference drug more than once, eg, replicated 3-period (RRT/RTR/TRR) or 4-period (RTRT/TRTR) crossover designs, so that the BE analysis can account for within-subject variability. For both the EMA and the FDA, RSABE can be employed if the reference product within-subject variability, CVWR , is greater than 30%, which corresponds to a within-subject standard deviation SWR ≥ 0.294.  EMA also requires a sound justification that a wider difference in Cmax is clinically irrelevant and that the calculated intra-subject variability is a reliable estimate and that it is not the result of outliers. The FDA requires that a minimum of 24 subjects are enrolled in the replicated crossover design study.

  AUC Cmax
FDA If Swr >= 0.294then RSABE is permitted and acceptance criteria for 90% CI can be widened. The point estimate (or geometric mean ratio) must be within 80-125% regardless of the widened acceptance criteria.

If Swr < 0.294, conventional ABE methods should be used

If Swr >= 0.294then RSABE is permitted and acceptance criteria for 90% CI can be widened. The point estimate (or geometric mean ratio) must be within 80-125% regardless of the widened acceptance criteria.

If Swr < 0.294, conventional ABE methods should be used

EMA Always traditional ABE:  Acceptance criteria for 90% CI is 80-125% If Swr >= 0.294then RSABE is permitted and acceptance criteria for 90% CI can be widened to a maximum of 70-143%. The point estimate (or geometric mean ratio) must be within 80-125% regardless of the widened acceptance criteria.

If Swr < 0.294, conventional ABE methods should be used.

Let’s look at some equations that should help clarify why the BE methodology for HVDs is called ‘scaled’.

For the conventional ABE, the acceptance of BE is met if the difference between the logarithmic means (LSM(logCmax) is between preset regulatory limits (80-125%).  This can be expressed as:

-ln(0.80) leq (Mean (log  hspace{3pt} Cmax_{test}) - Mean (log hspace{3pt} Cmax_{ref})) leq ln(1.25)

OR

-ln(0.80) leq LSM (log  hspace{3pt} Cmax)  leq ln(1.25)

For RSABE both the difference and the limits are scaled.  The limits are scaled by a regulatory constant and the difference is scaled by the within-subject variability for the reference product.

EMA RSABE methodology first scales the acceptance range by the regulatory constant 0.294 (which is the Swr at 30% CV ) and the difference is scaled by the drug’s variability Swr :

frac{-ln(0.80)}{0.294} leq frac{LSM (log  hspace{3pt} Cmax)}{S_{wr}}  leq frac{ln(1.25)}{0.294}

OR

-0.760 leq frac{LSM (log  hspace{3pt} Cmax)}{S_{wr}}  leq 0.760

The FDA approach to RSABE is based on the same concept although the scaling of the acceptance limits is done by the regulatory constant 0.25.  In addition, to conclude BE for a HDV, a different set of criteria (not discussed here) derived from this equation needs to be shown (ref 1).

frac{-ln(0.80)}{0.25} leq frac{LSM (log  hspace{3pt} Cmax)}{S_{wr}}  leq frac{ln(1.25)}{0.25}

OR

-0.893 leq frac{LSM (log  hspace{3pt} Cmax)}{S_{wr}}  leq 0.893

Based on the intra-subject standard deviation of the reference formulation one can calculate the scaled acceptance range limits:

EMA FDA
CVwr Swr RSABE Limits CVwr Swr RSABE Limits
<30 ABE Methodology <30 ABE Methodology
30 0.294 80.00 – 125.00 30 0.294 76.94 – 129.97
35 0.340 77.23 – 129.48 35 0.340 73.82 – 135.47
40 0.385 74.62 – 134.02 40 0.385 70.89 – 141.06
45 0.429 72.15 – 138.59 45 0.429 68.15 – 146.74
≥ 50 0.472 69.84 – 143.19
  50 0.472 65.58 – 152.48
  60 0.555 60.95 – 164.08

Where S_{wr} = sqrt{ln(CV_{wr}^2+1)} and limits are determined by e^{pm S_{wr}*(0.760hspace{2pt} or  hspace{2pt} 0.893)}

The RSABE approach has supported several approvals of highly variable generic drug products and is becoming standard practice in the generic drug industry.


References

[1] EMA guideline: https://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2010/01/WC500070039.pdf

[2] FDA guideline: https://www.fda.gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidances/ucm209294.pdf

[3] Download Phoenix WinNonlin templates v 1.4 for RSABE analysis: https://s3.amazonaws.com/Pharsight-Phoenix_Platform/Documentation/Phoenix_1.4_RSABE_Templates.zip?AWSAccessKeyId=AKIAINNBR4JMIL4ZPEJQ&Expires=1609390800&Signature=fHll%2Fx%2FLrDgexTH2T3UZUx%2BsU5g%3D


To learn more about using Phoenix RSABE templates to to demonstrate bioequivalence for HVDs, please watch this webinar I gave on this topic.

About the author

By: Thomas Henry