Non-Inferiority Clinical
`Trials to Establish
`Effectiveness
`
`Guidance for Industry
`
`U.S. Department of Health and Human Services
`Food and Drug Administration
`Center for Drug Evaluation and Research (CDER)
`Center for Biologics Evaluation and Research (CBER)
`
`November 2016
`Clinical/Medical
`
`Sawai (IPR2019-00789), Ex. 1049, p. 001
`
`

`

`Non-Inferiority Clinical
`Trials to Establish
`Effectiveness
`Guidance for Industry
`
`Additional copies are available from:
`Office of Communications, Division of Drug Information
`Center for Drug Evaluation and Research
`Food and Drug Administration
`10001 New Hampshire Ave., Hillandale Bldg., 4th Floor
`Silver Spring, MD 20993-0002
`Phone: 855-543-3784 or 301-796-3400; Fax: 301-431-6353
`Email: druginfo@fda.hhs.gov
`http://www.fda.gov/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/default.htm
`and/or
`Office of Communication, Outreach and Development
`Center for Biologics Evaluation and Research
` Food and Drug Administration
`10903 New Hampshire Ave., Bldg. 71, Room 3128
`Silver Spring, MD 20993-0002
`Phone: 800-835-4709 or 240-402-8010
`Email: ocod@fda.hhs.gov
`http://www.fda.gov/BiologicsBloodVaccines/GuidanceComplianceRegulatoryInformation/Guidances/default.htm
`
`U.S. Department of Health and Human Services
`Food and Drug Administration
`Center for Drug Evaluation and Research (CDER)
`Center for Biologics Evaluation and Research (CBER)
`
`November 2016
`Clinical/Medical
`
`Sawai (IPR2019-00789), Ex. 1049, p. 002
`
`

`

`Contains Nonbinding Recommendations
`
`TABLE OF CONTENTS
`
`INTRODUCTION............................................................................................................. 1
`I.
`BACKGROUND ............................................................................................................... 2
`II.
`GENERAL CONSIDERATIONS FOR NON-INFERIORITY STUDIES ................. 3
`III.
`A. The Non-Inferiority Hypothesis .................................................................................................... 3
`B. Reasons for Using a Non-Inferiority Design ................................................................................ 7
`C. The Non-Inferiority Margin .......................................................................................................... 8
`D. Assay Sensitivity ........................................................................................................................... 11
`E.
`Statistical Inference ..................................................................................................................... 14
`F. Regulatory Conclusions ............................................................................................................... 15
`G. Alternative Designs ...................................................................................................................... 16
`H. Number of Studies Needed .......................................................................................................... 17
`I.
`Choice of Active Control ............................................................................................................. 18
`IV.
`CHOOSING THE NON-INFERIORITY MARGIN AND TESTING THE NON-
`INFERIORITY HYPOTHESIS .................................................................................... 19
`Introduction .................................................................................................................................. 19
`A.
`Statistical Uncertainties and Quantification of the Active Control Effect .............................. 20
`B.
`C. Fixed Margin and Synthesis Methods ........................................................................................ 27
`D. Considerations for Selecting the Clinical Margin (M2) ............................................................ 30
`E. Estimating the Sample Size ......................................................................................................... 30
`F.
`Study Quality and Choice of Analysis Population .................................................................... 31
`G. Testing Non-Inferiority and Superiority in a Single Trial ....................................................... 31
`FREQUENTLY ASKED QUESTIONS AND GENERAL GUIDANCE .................. 32
`V.
`APPENDIX — EXAMPLES ...................................................................................................... 37
`REFERENCES FOR EXAMPLES ........................................................................................... 48
`REFERENCES ............................................................................................................................ 51
`
`Sawai (IPR2019-00789), Ex. 1049, p. 003
`
`

`

`Contains Nonbinding Recommendations
`
`Non-Inferiority Clinical Trials to Establish Effectiveness
`Guidance for Industry1
`
`This guidance represents the current thinking of the Food and Drug Administration (FDA or Agency) on
`this topic. It does not establish any rights for any person and is not binding on FDA or the public. You
`can use an alternative approach if it satisfies the requirements of the applicable statutes and regulations.
`To discuss an alternative approach, contact the FDA office responsible for this guidance as listed on the
`title page.
`
`I.
`
`INTRODUCTION
`
`This document provides guidance to sponsors and applicants submitting investigational drug
`applications (INDs), new drug applications (NDAs), biologics licensing applications (BLAs), or
`supplemental applications on the appropriate use of non-inferiority (NI) study designs to provide
`evidence of the effectiveness of a drug or biologic, usually because a superiority study design
`(drug versus placebo, dose response, or superiority to an active drug) cannot be used.2 The
`guidance gives advice on when NI studies intended to demonstrate effectiveness of an
`investigational drug can provide interpretable results, how to choose the NI margin, and how to
`test the NI hypothesis.
`
`This guidance does not provide recommendations for the use of NI study designs to evaluate the
`safety of a drug.
`
`In general, FDA’s guidance documents do not establish legally enforceable responsibilities.
`Instead, guidance documents describe the Agency’s current thinking on a topic and should be
`viewed only as recommendations, unless specific regulatory or statutory requirements are cited.
`The use of the word should in FDA guidance documents means suggested or recommended, but
`not required.
`
`This guidance finalizes the draft guidance for industry, Non-Inferiority Clinical Trials, published
`in 2010. In addition, it supersedes the guidance for industry, Antibacterial Drug Products: Use
`of Noninferiority Trials to Support Approval, also published in 2010, which will be withdrawn.
`
`1 This guidance has been prepared by the Office of Biostatistics and the Office of New Drugs in the Center for Drug
`Evaluation and Research (CDER) and the Center for Biologics Evaluation and Research (CBER) at the Food and
`Drug Administration.
`
`2 For the purposes of this guidance, all references to drugs include both human drugs and therapeutic biologic
`products unless otherwise specified. While most concepts discussed will be broadly applicable, certain issues related
`to vaccines, such as the choice of the NI margin when the study endpoint is the level of antibodies, would call for
`consultation from CBER.
`
`1
`
`Sawai (IPR2019-00789), Ex. 1049, p. 004
`
`

`

`II.
`
`BACKGROUND
`
`FDA’s regulations on adequate and well-controlled studies (21 CFR 314.126) describe four
`kinds of concurrently controlled trials that provide evidence of effectiveness. Three of them —
`placebo, no treatment, and dose-response controlled trials — are superiority trials that seek to
`show that a test drug is superior to the control (placebo, no treatment, or a lower dose of the test
`drug). The fourth kind, comparison with an active treatment (active control), can also be a
`superiority trial, if the intent is to show that the new drug is more effective than the control.
`More commonly, however, the goal of such studies is to show that the difference between the
`new and active control treatment is small — small enough to allow the known effectiveness of
`the active control, based on its performance in past studies and the assumed effectiveness of the
`active control in the current study, to support the conclusion that the new test drug is also
`effective. How to design and interpret the results of such studies so that they can support a
`conclusion about effectiveness of the new drug is challenging.
`
`Active controlled trials that are not intended to show superiority of the test drug but rather to
`show that the new treatment is not inferior to an unacceptable extent were once called clinical
`equivalence trials. The intent of an NI trial, however, is not to show that the new drug is
`equivalent, but rather that it is not materially worse than the control. Therefore, the interest is
`one-sided. The new drug could be better than the control, and therefore at a minimum non-
`inferior, but it would not be equivalent.
`
`The critical difference between superiority and NI trials is that a properly designed and
`conducted superiority trial, if successful in showing a difference, is entirely interpretable without
`further assumptions (other than lack of bias) — that is, the result speaks for itself and requires no
`extra-study information. In contrast, the NI study is dependent on knowing something that is not
`measured in the study, namely, that the active control had its expected effect in the NI study.
`When this occurs, the trial is said to have assay sensitivity, defined as the ability to have shown
`a difference from placebo of a specified size. A “successful” NI trial, one that shows what
`appears to be an acceptably small difference between treatments, may or may not have had assay
`sensitivity and therefore may or may not support a conclusion that the test drug was effective.
`Thus, if the active control had no effect at all in the NI trial (i.e., did not have any of its expected
`effect), then even ruling out a very small difference between control and test drug is meaningless
`and provides no evidence that the test drug is effective. (See Section III.D. for further discussion
`on assay sensitivity.) In the absence of a placebo arm, knowing whether the trial had assay
`sensitivity relies heavily on external (not within-study) information, giving NI studies some of
`the characteristics of a historically controlled trial.
`
`FDA regulations have recognized since 1985 the critical need to know, for an NI trial to be
`interpretable, that the active control had its expected effect in the trial. Thus, 21 CFR
`314.126(a)(2)(iv), unchanged since 1985, says:
`
`If the intent of the trial is to show similarity of the test and control drugs, the report of the
`study should assess the ability of the study to have detected a difference between
`treatments. Similarity of test drug and active control can mean either that both drugs were
`effective or that neither was effective. The analysis of the study should explain why the
`
`2
`
`Sawai (IPR2019-00789), Ex. 1049, p. 005
`
`

`

`drugs should be considered effective in the study, for example, by reference to results in
`previous placebo-controlled studies of the active control drug.
`
`This guidance consists of four parts:
`
`• A general discussion of regulatory, study design, scientific, and statistical issues
`associated with the use of NI studies to establish the effectiveness of a new drug
`• A focus on some of these issues in more detail, notably the statistical approaches used to
`determine the NI margin and to test for non-inferiority
`• Commonly asked questions about NI studies
`• Four examples of successful and unsuccessful efforts to define NI margins and test for
`non-inferiority3
`
`III.
`
`A.
`
`GENERAL CONSIDERATIONS FOR NON-INFERIORITY STUDIES
`
`The Non-Inferiority Hypothesis
`
`In a placebo-controlled trial, the null hypothesis (Ho) is that the beneficial response to the test
`drug (T) is less than or equal to the response to the placebo (P); the alternative hypothesis (Ha) is
`that the response to the test drug is greater than P. Thus:
`
`Ho: T ≤ P; T – P ≤ 0
`Ha: T > P; T – P > 0
`
`In most cases, the test for a treatment effect corresponds to showing that the lower bound of the
`two-sided 95% confidence interval (equivalent to the lower bound of a one-sided 97.5%
`confidence interval) for T-P is > 0. This result shows that the effect of the test drug is greater
`than 0 (see Figure 1).
`
`3 References: In this guidance, references to methods or studies are not included in the text; rather they are included
`in the general references section in the Appendix and are grouped by topic. A separate references section is also
`provided for the examples in the Appendix.
`
`3
`
`Sawai (IPR2019-00789), Ex. 1049, p. 006
`
`

`

`Figure 1. Possible Results of a Placebo-Controlled Superiority Study
`(Point Estimate and 95% Confidence Interval (CI))
`
`Test – Placebo (T-P)
`
`1. Point estimate of effect is 2; 95% CI lower bound is 1. Conclusion: Drug is effective
`and has an effect of at least 1.
`2. Point estimate of effect is 2; 95% CI lower bound is <0. Conclusion: Drug is not shown
`to be effective.
`3. Point estimate of effect is 0; 95% CI lower bound is well below 0. Conclusion: Drug is
`not shown to be effective.
`
`Although there is no difference in the conclusions of scenarios 2 and 3, the magnitude of the
`treatment difference and width of the confidence interval in scenario 2 may encourage the
`conduct of another study (possibly larger) before deciding that the test drug has no effect.
`
`In an NI study, the goal is to demonstrate that the test drug has an effect by showing that its
`effect is sufficiently close to the effect of an active control. There is no placebo arm in the study;
`therefore, the effect of the active control is not measured in the study but must be assumed. The
`goal of the study is to show that the effect of the test drug (T) is not inferior to the effect of the
`active control (C) by a specified amount, called the NI margin, or M.
`
`The null and alternative hypotheses correspond to a null hypothesis of inferiority and an
`alternative hypothesis of non-inferiority, as follows:
`
`Ho: C – T ≥ M (T is inferior to the control (C) by M or more)
` Ha: C – T < M (T is inferior to the control (C) by less than M)
`
`4
`
`Sawai (IPR2019-00789), Ex. 1049, p. 007
`
`

`

`A statistical test of the above hypothesis is provided by comparing the upper bound of the two-
`sided confidence interval for C-T with the NI margin, M, which is specified in advance. If the
`upper bound is less than M, non-inferiority of T relative to C is established.
`
`One choice for M (the largest possible value) is to set it equal to the entire known effect of the
`active control relative to placebo, based on past randomized controlled trials. With this choice
`for M, called M1, and assuming the control drug attains this level of efficacy in the NI study, a
`finding of non-inferiority means that the test drug has an effect greater than 0 (see Figure 2). A
`more usual choice is to set M equal to some clinically relevant portion of M1, namely, the portion
`of the control drug effect it is important to preserve with the test drug, based on clinical
`judgment.
`
`5
`
`Sawai (IPR2019-00789), Ex. 1049, p. 008
`
`

`

`Figure 2. Possible Results of an NI Study Showing Control Drug–Test Drug Differences
` (Point Estimate and 95% CI)
`
` Control – Test (C-T)
`
`1. Point estimate of C-T is 0, suggesting equal effect of C and T; upper bound of the 95%
`CI for C-T is 1, well below M1; NI is demonstrated.
`2. Point estimate of C-T favors C; the upper bound of the 95% CI for C-T is >2, above M1;
`NI is not demonstrated.
`3. Point estimate of C-T is zero, which suggests an equal effect; but the upper bound of the
`95% CI for C-T is >2 (i.e., above M1), so that NI is not demonstrated.
`4. Point estimate favors T; NI is demonstrated, but superiority is not demonstrated.
`5. Point estimate favors T; superiority and NI are demonstrated.
`6. Point estimate of C-T is 1, favoring the control. The upper bound of the 95% CI for C-T
`is <M1, demonstrating NI (the entire effect of C has not been lost) but at the same time
`the 95% CI for C-T is above zero, indicating that T is actually inferior to C, even while
`meeting the NI standard.
`
`The determination of M1 is a critical step in designing an NI trial and is often difficult; this
`determination is therefore a major focus of this guidance. M1 cannot be directly measured in the
`NI study, because there is no concurrent placebo group. It must be estimated based on the past
`performance of the active control, preferably in placebo-controlled trials, and then assumed to
`hold for the NI study based on a comparison of prior test conditions to the current test
`environment (see section III.D).
`
`6
`
`Sawai (IPR2019-00789), Ex. 1049, p. 009
`
`

`

`The choice of the margin M1 has important practical consequences. The smaller the margin, the
`lower the upper bound of the 95% two-sided confidence interval for C-T must be, and the larger
`the sample size needed to establish non-inferiority. Showing that the upper bound of the 95% CI
`of C-T is less than M1 demonstrates that the test drug has some effect (i.e., an effect > 0). The
`margin of interest, however, as noted above, is usually smaller than M1 (to show that an adequate
`portion of the clinical benefit of the control is preserved), in which case it is called M2. The
`basis for this expectation is described below in section III.C.4.
`
`B.
`
`Reasons for Using a Non-Inferiority Design
`
`The usual reason for using an NI active control study design instead of a superiority design is an
`ethical one. Specifically, this design is chosen when it would not be ethical to use a placebo, or a
`no-treatment control, or a very low dose of an active drug, because there is an effective treatment
`that provides an important benefit (e.g., life-saving or preventing irreversible injury) available to
`patients for the condition to be studied in the trial. Whether a placebo control can be used
`depends on the nature of the benefits provided by available therapy. The International
`Conference on Harmonisation guidance E10: Choice of Control Group and Related Issues in
`Clinical Trials (ICH E10) states:
`
`In cases where an available treatment is known to prevent serious harm, such as death or
`irreversible morbidity in the study population, it is generally inappropriate to use a
`placebo control. There are occasional exemptions, however, such as cases in which
`standard therapy has toxicity so severe that many patients have refused to receive it.
`
`In other situations, where there is no serious harm, it is generally considered ethical to
`ask patients to participate in a placebo-controlled trial, even if they may experience
`discomfort as a result, provided the setting is non-coercive and patients are fully informed
`about available therapies and the consequences of delaying treatment [ICH E10; pps.13-
`14].
`
`Aside from this ethical reason, there may be other reasons to include an active control, possibly
`in conjunction with a placebo control, either to compare treatments or to assess assay sensitivity
`(see section III.D). Caregivers, third party payers, and some regulatory authorities have
`increasingly placed an emphasis on the comparative effectiveness of treatments, leading to more
`studies that compare two treatments. Such studies can provide information about the clinical
`basis for comparative effectiveness claims, which may be helpful in assessing cost effectiveness
`of treatments. If a placebo group is included in addition to the active comparator, it becomes
`possible to judge whether the study could have distinguished treatments that differed
`substantially, e.g., active drug versus placebo. Such comparative effectiveness studies must be
`distinguished from NI studies, which are the main focus of this document. The word non-
`inferior is used here in a special sense. The methods described in this document are intended to
`show that a new treatment that demonstrates non-inferiority is effective, not that it is as effective
`as the active comparator. A new treatment may meet the standard of effectiveness (that it is
`superior to placebo) without justifying a conclusion that it is as effective or even nearly as
`effective as the active comparator.
`
`7
`
`Sawai (IPR2019-00789), Ex. 1049, p. 010
`
`

`

`C.
`
`The Non-Inferiority Margin
`
`As described above, the NI study seeks to show that the amount by which the test drug (T) is
`inferior to the active control (C), C-T, is less than some prespecified NI margin (M). M can be
`no larger than the presumed entire effect of the active control in the NI study, and the margin
`based on the entire active control effect is generally referred to as M1. It is critical to recognize
`that M1 is not measured in the NI trial (in the absence of a placebo arm), but rather is estimated
`based on past performance of the active control. The effect is assumed to be present in the
`current study based on a thorough comparison of the characteristics of the current NI study with
`those of prior studies and assessment of the quality of the NI study. The validity of any
`conclusion from the NI study depends on the choice of M1 and its relevance to the current NI
`study. If, for example, the NI margin is chosen as 10, and the study does indeed rule out a
`difference of 10 (but not a smaller difference), seeming to demonstrate effectiveness of T, but the
`true effect of C in this study was actually less than 10, then a conclusion that the study
`demonstrated non-inferiority would have been incorrect. The choice of M1, together with
`reasonable assurance that this effect occurred in the trial (i.e., the presence of assay sensitivity),
`is thus critical to obtaining a meaningful, correct answer in an NI study. Because assay
`sensitivity can never be proven in the absence of a placebo group, historical evidence of assay
`sensitivity is essential, but measures should be taken to make it likely that the active control will
`have the presumed effect in the NI study (see section III.A.5). This, together with careful
`selection of the NI margin, will increase the likelihood of valid and interpretable results.
`
`Because the consequence of choosing a margin greater than the actual treatment effect of the
`active control in the study may be a false conclusion that a new drug is effective, an undesirable
`public health outcome, there is a powerful tendency to be conservative both in the choice of
`margin and in the statistical analysis used to rule out a degree of inferiority of the test drug to the
`active control of more than that margin. This is generally done by ensuring that the upper bound
`of the 95% two-sided confidence interval for C-T in the NI trial is smaller than M1. Use of this
`interval corresponds to a one-sided test size (alpha level) of 0.025 in testing the NI hypothesis
`stated above. The upper bound of the confidence interval for C-T is not, however, the only
`measurement of interest, just as the lower bound of a 95% confidence interval for the effect of
`drug versus placebo in a superiority trial is not the only value of relevance. The point estimate of
`the treatment effect and the width of its confidence interval are also relevant. Nonetheless, the
`upper bound of the 95% confidence interval is typically used to judge the effectiveness of the test
`drug in the NI study, just as a two-sided test size (alpha level) of 0.05 is traditionally the standard
`used for defining success in a superiority trial. The 95% confidence interval upper bound for C-
`T is used, in conjunction with M1, to provide a reasonably high level of assurance that the test
`drug does, in fact, have an effect greater than zero (i.e., ruling out loss of all the effect of the
`active control).
`
`Although the NI margin used in a trial can be no larger than the entire assumed effect of the
`active control in the NI study (M1), it is usual and generally desirable to choose a smaller value,
`called M2, for the NI margin. Showing non-inferiority to M1 would provide assurance that the
`test drug had an effect greater than zero, but in many cases that would not be sufficient to
`conclude that the test drug had a clinically acceptable effect. Recall that the main reason an NI
`study is conducted is that it is not ethical to include a placebo arm. The active control has a
`
`8
`
`Sawai (IPR2019-00789), Ex. 1049, p. 011
`
`

`

`beneficial effect, and denying that benefit to subjects with a serious illness is not ethical. For the
`same reason, it would not usually be acceptable for a test drug to lose most of that active
`control’s effect. It is therefore usual in NI studies to choose a smaller margin (M2) that reflects
`the largest loss of effect that would be clinically acceptable. This can be described as an absolute
`difference in effect (typical of antibiotic trials) or as a fraction of the risk reduction provided by
`the control (typical in cardiovascular outcome trials). Note that a larger value for M2 may be
`justified clinically, if the test drug were shown to have some important advantage (e.g., on safety
`or on a secondary endpoint).
`
`Concern has been expressed that use of M2 represents an FDA “comparative effectiveness”
`standard that is not included in the Federal Food, Drug, and Cosmetic Act. In explaining the role
`of relative effectiveness under law in April 1995, President Clinton and Vice President Gore
`(“Reinventing Regulation of Drugs and Medical Devices,” part of the National Performance
`Review) stated the following:
`
`In certain circumstances, however, it may be important to consider
`whether a new product is less effective than available alternative therapies,
`when less effectiveness could present a danger to the patient or to
`the public. For example, it is essential for public health protection
`that a new therapy be as effective as alternatives that are already approved
`for marketing when:
`1. the disease to be treated is life-threatening or capable of causing
`irreversible morbidity (e.g., stroke or heart attack); or
`2. the disease to be treated is a contagious illness that poses serious
`consequences to the health of others (e.g., sexually transmitted disease).
`
`The reinvention statement was placed in the Federal Register of August 1, 1995 (60 FR 39180 at
`39181), as an FDA position by FDA’s Deputy Commissioner for Policy, William Schultz.
`
`The definitions used to describe these two versions of M are:
`
`M1 = the entire effect of the active control assumed to be present in the NI study
`M2 = the largest clinically acceptable difference (degree of inferiority) of the test drug
`compared to the active control
`
`M1 is estimated based on the historical experience with the active control drug. Its relevance to
`the current NI trial is based on:
`
`1. Assessment of the likelihood that the current effect of the active control is similar to that
`observed in the past studies used to estimate the active control effect (the constancy
`assumption)
`2. Assessment of the quality of the NI trial, particularly looking for deficiencies in study
`design and/or conduct that could reduce a difference between the active control and the
`new drug
`
`Note that in a trial seeking to show a difference (i.e., superiority), this diminution of the between-
`treatment difference in the second element is a “bias toward the null,” but in a non-inferiority
`
`9
`
`Sawai (IPR2019-00789), Ex. 1049, p. 012
`
`

`

`trial, it is a “bias toward the alternative” (i.e., non-inferiority). Because of this second element,
`in some situations M1, although prespecified, has to be “discounted” (i.e., a smaller value would
`be used), but the amount of discounting needed may not be known until the NI study has been
`completed.
`
`The choice of M2 is a matter of clinical judgment, but M2 can never be greater than M1, even for
`a situation in which the active control drug has a small effect, and clinical judgment might argue
`that a larger difference is not clinically important. Even if that clinical judgment were
`reasonable, choosing an M2 greater than M1 as the NI margin would not allow a conclusion that
`the test drug has any effect. As explained above, ruling out a difference between the active
`control and test drug larger than M1 is the critical finding that supports a conclusion of
`effectiveness. This analysis is approached with great rigor; that is, a difference (C-T) larger than
`M1 needs to be ruled out with a high degree of statistical assurance. As M2 represents a clinical
`judgment, there may be a greater flexibility in interpreting a 95% upper bound for C-T that is
`slightly greater than M2, as long as the upper bound is still well less than M1 (see Figure 3).
`
`10
`
`Sawai (IPR2019-00789), Ex. 1049, p. 013
`
`

`

`Figure 3. Possible Results of an NI Study Showing Active Control – Test Drug Differences
`(Point Estimate and 95% CI)
`
`Control – Test (C-T); (degree of inferiority of test drug)
`
`1. C-T point estimate = 0 and upper bound of 95% CI < M2, indicating test
`drug is effective and adequately rules out an unacceptable loss of the control
`effect (NI demonstrated).
`2. Point estimate of C-T favors C; upper bound of 95% CI < M1 but
`> M2, indicating test drug effect > 0 but an unacceptable loss of the control effect
`has not been ruled out.
`3. Point estimate of C-T is zero and upper bound of 95% CI < M1 but it is
`slightly greater than M2. Loss of the pre-specified M2 has thus not been ruled out,
`but whether the study has shown adequate preservation of the control effect would
`be a matter of clinical judgment.
`4. C-T point estimate favors C and upper bound of 95% CI > M1, indicating
`the study does not provide evidence of effectiveness for test drug.
`
`D.
`
`Assay Sensitivity
`
`Assay sensitivity is an essential property of an NI clinical trial. Assay sensitivity is the ability of
`the trial to have detected a difference between treatments of a specified size. Stated in another
`way, assay sensitivity means that had the study included a placebo, a control drug-placebo
`difference of at least M1 would have been present.
`
`11
`
`Sawai (IPR2019-00789), Ex. 1049, p. 014
`
`

`

`As explained above, the choice of M1, and the conclusion that a trial has assay sensitivity (i.e.,
`the active control would have had an effect of at least M1), is based on three considerations:
`
`1. Historical evidence of sensitivity to drug effects
`2. The similarity of the new NI trial to the historical trials (the constancy assumption)
`3. The quality of the new trial (ruling out defects that would tend to minimize differences
`between treatments)
`
`1. Historical Evidence of Sensitivity to Drug Effects (HESDE)
`
`HESDE means that prior studies, which were appropriately designed and conducted trials in the
`past and that used a specific active treatment (generally the one that is to be used in the new NI
`study or, in some cases, one or more pharmacologically closely related drugs), regularly showed
`this treatment to be superior to placebo (or some other treatment). Consistent findings in past
`studies allow for a reliable estimate of the drug’s effect compared to placebo. The estimate of
`the size of the effect must take the variability of past results into account; one should not
`presume that the largest effect seen in any trial, or even the point estimate of a meta-analysis of
`all relevant trials, is certain to be repeated. Analysis of historical data will be discussed further
`in section IV.
`
`HESDE cannot be determined for many symptomatic treatments (e.g., treatments for depression,
`anxiety, insomnia, angina, symptomatic heart failure, symptoms of irritable bowel disease, and
`pain), as even well-designed and conducted studies for such indications often fail to distinguish
`an effective drug from placebo. In those cases, it cannot be assumed that an active control would
`have shown superiority to a placebo (had there been one) in any given NI study, and NI studies
`of drugs for these indications may therefore be non-informative. These issues can also
`negatively affect the assessment of effectiveness in outcome studies. For example, in the case of
`aspirin, the largest placebo-controlled trial (AMIS, the Aspirin Myocardial Infarction Study; see
`Example 2) did not show an effect of aspirin even though other trials all favored aspirin.
`Similarly, of more than 30 post-infarction beta-blocker trials, only a small numbe