Bioequivalence criterion 80-125% are the result of the history of the changes themselves criteria and options bioequivalence account statistically. These strange numbers are the result after log transformation for AUC and Cmax (drug concentration in the blood). If you are not a mathematician, the easiest way read material on Gary Buehler (in file).

I hope this helps :-)

the 20% in difference was a very subjective limit from physician when generics appeared. then, the transformation to logaritmic scale is 80 - 125% comparing with the original drug. 

90 - 110% is used in some countries as Australia or Canada to stablish bioequivalence for Critic dose drugs. 

For a clear, concise introduction, please read the one page 2011 article by Nathan Teuscher at http://learnpkpd.com/2011/01/24/where-did-the-80-125-bioequivalence-criteria-come-from/

As he notes, underlying the 80-120% rule was the assumption that a 20% difference was not significant.  Part of this was based on clinical intuition, part was based on CMC considerations. 

Regarding CMC considerations, at the time, in vitro drug content standards commonly allowed tolerances of 90-110% for oral dosage forms.  Thus, physiologically, the in vivo rule (80-120%) had to be wider than the in vitro rule (90-110%).

Gary Beuhler's slide deck from 2010 mentions the transition from 80-120% rule on the arithmetic scale to the 80-125% rule on the log scale and gives additional background on the entire effort to develop standards everyone could abide by:

http://www.fda.gov/downloads/%20AdvisoryCommittees/%E2%80%A6/UCM209319.pdf

Note that the historical development of the various BE rules was based on intense discussions between all relevant parties including government, academia, and the pharmaceutical industry.  Physicians, biostatisticians, pharmacologists and many other disciplines employed by the three key stakeholders mentioned in the first sentence of this paragraph were involved in intense and extensive discussions which extended over several decades.

Acceptance level is set at +-20%.

80/100 is 0.8, and 100/80=1.25. 

I agree with the comments above, but I want to add a different set of comments.  

As Monty Python would say, "And now for something dramatically new."

In the US, the FDA does not like to prevent new generic formulations from coming on the mkt. b/c the US is very commerce oriented.  So, IMHO, the +/- 20% limits provide for a reasonable range that will more likely than not be efficacious in pts. w/o being ineffective or more toxic than the pioneer product.

Variability among pts means that blood levels will vary from pt to pt regardless of how the new formulation relates to the std, and this can be said of most dosage forms by almost any route.  I think +/- 20% is reasonable to permit licensing, but it will take use in pts to confirm efficacy and R/O intolerance.  Of course, the TI of the drug would be an important factor.  An oncology drugs or anticoagulant would have to be monitored more carefully that a drug with a higher TI.  Also, certain antibiotics like aminoglycosides have to be monitored for trough levels to minimize renal and oto-toxicity.  Vancomycin needs to be monitored for "Red Man" syndrome due to excessive vasodilatation.

Back in 1968, Morelli and Melmon at UCSF wrote their book on Clin Pharmacology.  One of the things they stressed was that every time a doc administers a drug to a pt, he/she is conducting a new experiment.   I think these are wise words, and one must see how the pt responds on an individual basis, which is now heralded as pt-centered therapeutics, which stresses the need to evaluate host factors and treat pts accordingly.  Slow infusions w/close monitoring is essential until the pt's tolerance is determined.

Start low, go slow!

The link posted by Robert Shumaker is the appropriate document that will answer all queries of you. 

The mathematical reasoning has been nicely explained in several answers, i.e. why 80-125 seems more appropriate as e.g. 80-120.

The pharmacological rationale was well characterized by Robert Shumaker:  “…the historical development of the various BE rules was based on intense discussions between all relevant parties including government, academia, and the pharmaceutical industry. “

With other words, discussions rather than data are the basis or this assumption. In addition to pharmacological plausibility, technical feasibility and economic considerations may have contributed.

For most medicines, 80-125 seems to be a basis for the safe use of medicines.  Some reassurance may be found in the following:

If the same dose (e.g. 50 mg per tablet) is recommended for all patients, including both a five foot  45 kg female  and a 6.5 foot  125 kg male, the difference in the systemic exposure can be expected  to exceed the 80-125% range by far. Therefore, if that big an exposure range is considered acceptable (recommending the same dose for all patients) the 80-125% (accepted for generics) should not make a big difference.

There are, on the other hand, certainly medicines with a steeper dose response curve or so-called critical dose or narrow therapeutic index drugs. For these, other criteria apply.  

US-FDA classification of narrow therapeutic ratio:

less than a 2-fold difference in median lethal dose (LD50) and median effective dose values (ED50), -or- less than 2-fold difference in the minimum toxic concentrations (MTC) and minimum effective concentrations (MEC) in the blood

The FDA definition is quite clear and unambiguous – a weakness may be that it is based on the acute effects. This may be better with the more general definition by Health Canada (but of course this I s less precise)

Health Canada’s definition of critical dose drugs:

“Those drugs in which comparatively small differences in dose or concentration may lead to serious therapeutic failures and/or serious adverse drug reactions.

Interestingly, there is no published definition what the European Medicines Agency may consider critical dose or NTID.

Bottom Line: The generally accepted 80-125 % range worked reasonably well so far.

Here ends the politically correct part.

Xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Addendum, with more question than answers:

 The 80-125% range is generally accepted  - but is “generally accepted”  not just  a nicer expression for “based on discussions, not based on data”?

Consider medicines for which the dose finding studies during development did show dose dependence of  effects (efficacy or safety) for an 50% increase in dose, e.g. from 80 mg to 120 mg. And assume based on these dose finding studies  the final dose for the marketed product was then set to 100 mg. Should we not expect that different generic products – e.g. one with bioavailability at the lower acceptance limit (80%) and one with a the bioavailability at the upper acceptance limit (125%) will differ in efficacy and or safety?

Or to make it very simple - ask yourself: – would you not expect that e.g. 625 mg aspirin will be more likely to help your headache than 400 mg? And would you not expect that  400 mg aspirin will be less damaging to your gastric epithelium than 625mg?  

Bravo, Dr. Enzmann.  You have expressed the paradox correctly.  However, the FDA has to have some set of standards to use on a general basis, and 80 - 125% seems reasonable.  The FDA is always able to tell the manufacturer that they want a study with different upper and lower limits b/c the drug has a low TI.

Reasonable is the key word or objective here, and that appears to be what we have.

@ David M Benjamin.

Thank you. Using the 80-125% as default may be a more pragmatic than strictly scientific approach, but it has been working so far and there seems neither the need nor the will to change it.

But I would be interested in your view to what extent this approach can be extrapolated  to biosimilars. Is it justified to use the same criteria?

Kind regards

Harald

Well, Dr. Enzmann, I can see that you are a visionary, as well as a scientist.  I am in awe of all the biological agents on the market and under development.  I certainly would not apply "drug" regulations to a biological.  Different companies will use different methods to synthesize the product or fulfill patent filings and you cannot just attempt to measure blood levels and AUC,  etc.  Individual pt host factors and genetics will add to the variability of responses and biomarkers or other indicators of clinical response need to be monitored.  Bioavailability is not a concern to me, efficacy and tolerance are.

David

Dear Prof. Benjamin,

Thank you again, you are very kind – I am flattered, even more so as being visionary is not at all in my line of work. I am a regulator (Europe), so I have to reflect what data we need for the approval of medicines, generics or biosimilars.

Given my previous comments on the 80-125% limits for generics you will understand why I hesitate to apply the very same margins to biosimilars. Problem is, I may find that the 80-125% limits are not truly data-based; unfortunately, there are no other margins with a better data base or a more convincing rationale. I guess this is why US-FDA recommends the 80-125% margins for biosimilars the same way as for small molecules. See slide 29 of this presentation:

http://www.fda.gov/Drugs/DevelopmentApprovalProcess/HowDrugsareDevelopedandApproved/ApprovalApplications/TherapeuticBiologicApplications/Biosimilars/ucm428730.htm

 I really wish we (all regulators, no matter where) could come up with something better, with a more scientific rationale than just our basic “it worked so far”.

Kind regards

I agree, but there never will  be data to set a standard, but new data may emerge to make you "rethink" it.

I know a little about the German drug approval process b/c I had friends who practiced medicine in Germany and we discussed it.

Unfortunately, no system will pick up severe adverse drug effects on NDA data b/c the number of pts is too low, say 2500-3500, and some severe ADRs like ARF, or SJS or TEN may occur at 1 in 10,000 or less.

The thing to avoid is approving a drug with a worse safety profile than existing drugs, unless there is a great need for the drug.  As you know, the "safety" part of safety and efficacy is very subjective.  Where there are no clear cut rules, one must go by the seat of their pants.  That's all and hope for the best.  To do otherwise would inhibit progress.

David

I just found this report on biosimilar activity of Mylan's Myl-14010 to Herceptin in breast cancer and wanted to post it.

David

Study finds Mylan's biosimilar comparable to Roche cancer drug

Wall Street Journal (06/03/16) Winslow, Ron

The results of a new clinical trial show that a biosimilar drug being developed by Mylan Pharmaceuticals is comparable to trastuzumab (Herceptin—Roche Holdings). The study of 500 patients with metastatic disease found that Mylan's Myl-1401O was essentially equivalent in safety and effectiveness to trastuzumab. Participants received chemotherapy for 24 weeks plus Myl-1401O or trastuzumab, and they continued on the biotech drugs alone after 24 weeks until their disease progressed. The biosimilar drug "has the potential to meet the need for an affordable treatment option" for women diagnosed with Her2-positive breast cancer, said Hope Rugo, lead author of the report and professor of medicine at University of California San Francisco. The findings were presented at the annual meeting of the American Society of Clinical Oncology.

Read More | Return to Headlines

Most people involved in clinical pharmacokinetics are familiar with the 80-125% criterion. This criterion is used to compare two treatments with the purpose of evaluating if the treatments are bioequivalent. But, where did this come from? Why 80-125%? Why not 90-110%? or why not 80-120%?

Before we explain where 80-125% came from, let me explain the specifics of the criterion. When testing two treatments (e.g. 2 formulations, male vs. female, impaired vs non-impaired, etc.) often we would like to know if there is a difference in systemic exposure between the two treatments. The currently accepted test is often called “bioequivalence”.

The bioequivalence test states that we can conclude that two treatments are not different from one another if the 90% confidence interval of the ratio of a log-transformed exposure measure (AUC and/or Cmax) falls completely within the range 80-125%. It is important to note that we only conclude that the two treatments are “not different” from one another. We do not conclude that they are the “same”. However, if the 90% confidence interval falls outside the 80-125% range, we conclude that the two treatments are different from one another.

The basis for the 80-125% range is arbitrary … sort of. The FDA (and other regulatory bodies) “decided” that differences in systemic drug exposure up to 20% are not clinically significant. Now, that may lead you to believe that the appropriate range should be 80-120% (100% ± 20%) … but that isn’t the range. This is because the pharmacokinetic parameters for exposure (AUC and/or Cmax) are log-normally distributed. This means that if you transform these exposure parameters by taking the logarithm, you will get a normal distribution. Normal distributions are generally required for specific statistical tests. Thus, the symmetrical ± 20% has to be in the log-transformed space so that the statistical test of bioequivalence will be valid. The following table illustrates the different ratios, and the log-transformed difference.

Test    Reference Ratio Percentage ln(ratio)

0.8         1.0               0.8       80%         -0.223

0.9       1.0                 0.9       90%          -0.105

1.0       1.0                 1.0       100%           0

1.1       1.0                 1.1       110%        0.095

1.2         1.0               1.2        120%       0.182

1.25      1.0                1.25      125%        0.223

Starting at the lower limit (80%), we calculate the natural log of the ratio as -0.223. We can also see that the natural log of the ratio of 100% is 0. Therefore, a symmetrical distribution around 100% on the natural log transformed ratio would be ± 0.223. As shown in the table above, this corresponds to 125% at the upper limit. That’s how we get 80-125% as the target range that represents ± 20% systemic exposure.

This same principle can be used for other ranges that are commonly used for comparisons between 2 treatments where a wider range is acceptable.

Clinical Range ± ln(ratio) Acceptable Range

± 20%                   ± 0.223 80 – 125%

± 30%                   ± 0.357 70 – 143%

± 50%                   ± 0.693 50 – 200%

When conducting a study to compare two treatments, make sure you pick the correct range for the statistical test. All of these ranges are commonly accepted by regulatory agencies. In addition, if you need a custom range (e.g. ± 25%), you can calculate it by determining the ln(ratio) of the lower limit, then creating the symmetrical ln(ratio) for the upper limit and back-calculating the untransformed upper limit.