DEPARTMENT OF HEALTH AND HUMAN SERVICES
Food and Drug Administration
AGENCY: Food and Drug Administration, HHS.

21 CFR Parts 314 and 601
New Drug, Antibiotic, and Biological Drug Product
Regulations; Accelerated Approval

[Docket No. 91N-0278]
RIN 0905-AD66

57 FR 58942 

December 11, 1992 

 
 ACTION: Final rule.  

 SUMMARY: The Food and Drug Administration (FDA) is issuing final regulations 
under which the agency will accelerate approval of certain new drugs and 
biological products for serious or life-threatening illnesses, with provisions 
for any necessary continued study of the drugs' clinical benefits after approval 
or with restrictions on use, if necessary. These new procedures are intended to 
provide expedited marketing of drugs for patients suffering from such illnesses 
when the drugs provide meaningful therapeutic benefit compared to existing 
treatment. Accelerated approval will be considered in two situations: (1) When 
approval can be reliably based on evidence from adequate and well-controlled 
studies of the drug's effect on a surrogate endpoint that reasonably suggests 
clinical benefit or on evidence of the drug's effect on a clinical endpoint 
other than survival or irreversible morbidity, pending completion of studies to 
establish and define the degree of clinical benefits to patients; and (2) when 
FDA determines that a drug, effective for the treatment of a disease, can be 
used safely only if distribution or use is modified or restricted. Drugs or 
biological products approved under these procedures will have met the requisite 
standards for safety and effectiveness under the Federal Food, Drug, and 
Cosmetic Act (the act) or the Public Health Service Act (the PHS Act) and, thus, 
will have full approval for marketing.  

 EFFECTIVE DATE: January 11, 1993.  

 FOR FURTHER INFORMATION CONTACT: Marilyn L. Watson, Center for Drug Evaluation 
and Research (HFD-360), Food and Drug Administration, 7500 Standish Pl., 
Rockville, MD 20855, 301-295-8038.  

TEXT: SUPPLEMENTARY INFORMATION:  

   
I. Background  

   In the Federal Register  of April 15, 1992 (57 FR 13234), FDA published 
proposed procedures under which the agency would accelerate approval of  certain 
new drugs and biological products for serious or life-threatening illnesses, 
with provision for required continued study of the drugs' clinical benefits 
after approval or for restrictions on distribution or use, where those are 
necessary for safe use of the drugs. FDA provided 60 days for public comment, 
and, upon request, in the Federal Register  of June 18, 1992 (57 FR 27202), 
extended the comment period for an additional 30 days until July 15, 1992. The 
final rule incorporates all of the provisions of the proposed rule and provides 
additional clarification regarding both timing and content of the submissions of 
promotional materials and regarding the nature of required postmarketing 
studies. The agency has added a new provision clarifying when certain 
postmarketing requirements of the rule will be terminated.  

   Highlights of the final rule are summarized below, followed by a summary and 
discussion of the comments.  

   
II. Highlights of the Final Rule  

   This final rule establishes procedures under parts 314 and 601 (21 CFR parts 
314 and 601) under which FDA will accelerate approval of certain new drugs and 
biological products for serious or life-threatening illnesses, with provision 
for required continued study of the drugs' clinical benefits after approval or 
for restrictions on distribution or use, where those are necessary for safe use 
of the drugs. These procedures are intended to provide expedited marketing of 
drugs for patients suffering from such illnesses when the drugs provide 
meaningful therapeutic advantage over existing treatment. The preamble of the 
proposed rule (57 FR 13234) provides a description of other mechanisms available 
to facilitate access, speed development, and expedite review of therapeutic 
products (e.g., treatment investigational new drug applications (IND's), subpart 
E, parallel track). Where appropriate, these mechanisms can be utilized in 
concert with accelerated approval. The major provisions of the final rule are as 
follows:  

   
A. Scope  

   The new procedures apply to certain new drug, antibiotic, and biological 
products used in the treatment of serious or life-threatening diseases, where 
the products provide meaningful therapeutic advantage over existing treatment 
(21 CFR 314.500 and 601.40).  

   
B. Criteria for Approval  

   Accelerated approval will be considered in two situations: (1) When approval 
can be reliably based on evidence of the drug's effect on a surrogate endpoint 
that reasonably suggests clinical benefit or on evidence of the drug's effect on 
a clinical endpoint other than survival or irreversible morbidity, pending 
completion of studies to establish and define the degree of clinical benefits to 
patients; and (2) when FDA determines that a drug, effective for the treatment 
of a disease, can be used safely only if distribution or use is modified or 
restricted. Drugs or biological products approved under this final rule will 
have met the requisite standards for safety and effectiveness under the act or 
the PHS Act and, thus, will have full approval for marketing (21 CFR 314.510, 
314.520, 601.41, and 601.42). Ordinarily, products used to treat serious or  
life-threatening illnesses, for which approval is based on a surrogate endpoint 
that is recognized as validated by definitive studies, will be considered for 
approval under the traditional process rather than under accelerated approval.  

   
C. Postmarketing Studies  

   Where a drug's approval under these provisions is based on a surrogate 
endpoint or on an effect on a clinical endpoint other than survival or 
irreversible morbidity, the applicant will be required to conduct clinical 
studies necessary to verify and describe the drug's clinical benefit and to 
resolve remaining uncertainty as to the relation of the surrogate endpoint upon 
which approval was based to clinical benefit, or the observed clinical benefit 
to ultimate outcome. The requirement for any additional study to demonstrate 
actual clinical benefit will not be more stringent than those that would 
normally be required for marketing approval; it is expected that the studies 
will usually be underway at the time of approval. The proposed regulations have 
been revised to clarify that required postmarketing studies must also be 
adequate and well-controlled (21 CFR 314.510 and 601.41).  

   
D. Restrictions on Use After Marketing  

   FDA may grant marketing approval of a drug or biological product shown to be 
effective where safe use can only be assured if distribution or use is 
restricted. Under this final rule, FDA may: (1) Restrict distribution to certain 
facilities or to physicians with special training or experience, or (2) 
condition distribution on the performance of specified medical procedures. The 
restrictions on use will be tailored to the specific safety issue raised by the 
particular drug or biological product and agreed to by the applicant at the time 
of approval (21 CFR 314.520 and 601.42). FDA expects that the imposition of 
these restrictions on distribution will be rare.  

   
E. Promotional Materials  

   The final rule requires submission of planned promotional materials, 
including promotional labeling and advertisements, both prior to approval 
(reflecting the initial campaign), and following approval, unless informed by 
the agency that such submission is no longer necessary, at least 30 days before 
the intended time of initial dissemination of the promotional labeling or 
initial publication of the advertisement (21 CFR 314.550 and 601.45).  

   
F. Withdrawal of Approval  

   The final rule establishes an expedited procedure for the withdrawal of 
approval if: (1) Postmarketing clinical studies fail to verify clinical benefit; 
(2) the applicant fails to perform the required postmarketing study with due 
diligence; (3) use after marketing demonstrates that postmarketing restrictions 
are inadequate to ensure safe use of the drug or biological product; (4) the 
applicant fails to adhere to the postmarketing restrictions agreed upon; (5) the 
promotional materials are false or misleading; or (6) other evidence 
demonstrates that the drug or biological product is not shown to be safe or 
effective under its conditions of use (21 CFR 314.530 and 601.43).  

   
G. Termination of Requirements  

   In response to comments, the final rule provides that the requirements set 
forth in §§ 314.520, 314.530, and 314.550 for new drugs and antibiotics and §§ 
601.42, 601.43, and 601.45 for biological products ordinarily will terminate 
when FDA determines that the results of required postmarketing studies have 
demonstrated that the drug or biological product has clinical benefit, or, where 
restrictions on distribution or use have been imposed, when FDA determines that 
safe use of the drug or biological product can be ensured without such 
restrictions, e.g., through appropriate labeling. FDA will notify the applicant 
when these requirements no longer apply (21 CFR 314.560 and 601.46).  

   
III. Effective Date  

   This regulation will become effective on January 11, 1993.  

   
IV. Comments on the Proposed Rule  

   FDA received 54 comments on the proposed rule. The comments came from 
individuals, specific disease organizations, universities, pharmaceutical 
manufacturers, trade associations, health professionals, and professional 
societies. The comments reflect broad support and acceptance of the goal of 
expediting the approval of drugs intended for the treatment of serious and life-
threatening illnesses. A number of comments asked that the proposal be finalized 
expeditiously without change. Many comments posed specific questions and raised 
important concerns.  

   
A. General Comments  

   1. One comment suggested that the term "conditional approval" was less 
confusing and ambiguous than the term "accelerated approval." The comment also 
referred to the statement in the proposal that "Drugs * * * approved under this 
proposal will have met the requisite standards * * * under the (act)" and argued 
that because postmarketing conditions may be imposed, this statement can only be 
read to say that the requisite standards under the act can only be met by a 
lower standard of evidence in hand, combined with assurance that further 
evidence will be obtained.  

   Another comment expressed concern that the proposal appears to establish a 
standard for the evaluation of drug product effectiveness that is inconsistent 
with the substantial evidence requirement of section 505(d) of the act (21 
U.S.C. 355(d)), which means "evidence consisting of adequate and well-controlled 
investigations, including clinical investigations, by experts qualified by 
scientific training and experience to evaluate the effectiveness of the drug 
involved, on the basis of which it could fairly and responsibly be concluded by 
such experts that the drug will have the effect it purports or is represented to 
have under the conditions of use prescribed, recommended, or suggested in the 
labeling or proposed labeling * * *." The comment argued that, with few 
exceptions, the agency has consistently interpreted the "substantial evidence" 
requirement as an instruction that determinations of effectiveness be based on 
data unambiguously reflecting the clinical status of subjects evaluated under  
controlled conditions in bona fide clinical experiments. In the absence of 
compelling empirical evidence documenting that a drug-induced change in a 
surrogate measure reliably and consistently predicts improved clinical outcome, 
a surrogate indicator is no more than a hypothetical construct. The comment 
asserted that the proposed rule's endorsement of the use of unvalidated 
surrogate endpoints, therefore, appears to represent a significant departure 
from traditional agency interpretations of "substantial evidence" within the 
meaning of the act because it allows belief rather than evidence to serve as the 
basis for a conclusion about the effectiveness of a new drug.  

   Three comments asserted that the new regulations are not needed to approve 
drugs intended to treat serious or life-threatening illnesses. Two comments 
cited FDA's approval, without new regulations, of didanosine (formerly called 
ddi) and zalcitabine (formerly called ddc) in combination with zidovudine 
(formerly called AZT) based on a surrogate marker, i.e., an increase in CD4 cell 
counts and the "subpart E" procedures at 21 CFR part 312, which address the need 
for expediting the development, evaluation, and marketing of new therapies 
intended to treat life-threatening or severely debilitating illnesses as 
examples of existing mechanisms for the expedited approval of important new 
drugs. One comment argued that the act requires that drugs be shown to be "safe" 
and "effective," and proof of effectiveness is not limited by the act to 
demonstration of an effect on "survival or irreversible morbidity," as the 
proposed rule seems to assume. The comment further argued that FDA has 
considerable statutory discretion to define what type of data constitutes proof 
of effectiveness, and demonstration of an effect on a surrogate marker is one 
type of such proof.  

   The agency believes that what the procedures are called is much less 
important than what the procedures are. The shorthand term selected by the 
agency reflects the intent of the rule, especially that part related to use of 
surrogate markers, which is to make drugs that provide meaningful improvement 
over existing therapies for serious illnesses widely available (through 
marketing) at the earliest time consistent with the law. The essence of the 
proposal is thus acceleration, not the imposition of conditions. Approval under 
these procedures is dependent on compliance with certain additional 
requirements, such as timely completion of studies to document the expected 
clinical benefit. The evidence available at the time of approval under this rule 
will meet the statutory standard, in that there must be evidence from adequate 
and well-controlled studies showing that the drug will have the effect it is 
represented to have in its labeling. That effect will, in this case, be an 
effect on a surrogate endpoint that is reasonably likely to predict a clinical 
benefit and labeling will refer to the effect on the surrogate, not to effect on 
clinical outcome.  

   While the act does not refer to particular endpoints or state a preference 
for clinical, as opposed to surrogate, endpoints, it is well established that 
the effect shown in well-controlled studies, must, in the judgment of the 
agency, be clinically meaningful. Moreover, the safety standard in the act, that 
a drug must be shown to be safe for its intended use, implies a risk/benefit 
judgment. The effect shown must be such as to outweigh the risks of the 
treatment under the conditions of use. Approval under this rule requires, 
therefore, that the effect shown be, in the judgment of the agency, clinically 
meaningful, and of such importance as to outweigh the risks of treatment. This 
judgment does not represent either a "lower standard" or one inconsistent with 
section 505(d) of the act, but rather an assessment about whether different  
types of data show that the same statutory standard has been met.  

   Approval based on surrogate endpoints is not new, although the issue has not 
previously been considered in regulations. The agency has, in a number of 
instances, approved drugs based on surrogate endpoints. For example, drugs for 
hypertension have been approved based on their effects on blood pressure rather 
than on survival or stroke rate. Similarly, drugs for hypercholesterolemia have 
been approved based on effects on serum cholesterol rather than on coronary 
artery disease (angina, heart attacks). But, in those cases there was very good 
evidence from clinical trials (in the case of hypertension) and from 
epidemiologic and animal studies (in the case of hypercholesterolemia) that 
improving the surrogate would lead to or is associated with the desired effects 
on morbidity and mortality. Even so, there is still today considerable debate 
about who will benefit from cholesterol lowering. Controlled trials assessing 
effects on clinical endpoints of morbidity and mortality from use of 
cholesterol-lowering drugs have been, and are being, conducted.  

   Reliance on a surrogate endpoint almost always introduces some uncertainty 
into the risk/benefit assessment, because clinical benefit is not measured 
directly and the quantitative relation of the effect on the surrogate to the 
clinical effect is rarely known. The expected risk/benefit relationship may fail 
to emerge because: (1) The identified surrogate may not in fact be causally 
related to clinical outcome (even though it was thought to be) or (2) the drug 
may have a smaller than expected benefit and a larger than expected adverse 
effect that could not be recognized without large-scale clinical trials of long 
duration. Reliance on surrogate markers therefore requires an additional measure 
of judgment, not only weighing benefit versus risk, as always, but also deciding 
what the therapeutic benefit is based upon the drug effect on the surrogate.  

   The sections of the final rule that address approval based upon a drug effect 
on a surrogate endpoint specifically clarify the regulatory approval criteria 
when the agency relies on a surrogate endpoint that, while "reasonably likely" 
to predict clinical benefit, is not so well established as the surrogates 
ordinarily used as bases of approval in the past. Postmarketing studies required 
to verify and describe actual clinical benefits would also be required to be 
adequate and well-controlled studies. Sections 314.510 and 601.41 have been 
revised to clarify this point. If, on completion of required postmarketing 
studies, the effect on the surrogate is not shown to correspond to a favorable 
effect on clinical benefit, the rule provides an expedited means of removing the 
drug from the market.  

   Approval of didanosine and zalcitabine under current procedures does not show 
that the rule is of no value. Although approval did rely on a surrogate endpoint 
that is of the kind specifically addressed by the rule, the fact that studies to 
define clinical benefit were nearly complete and were being conducted under the 
auspices of the National Institute of Allergy and Infectious Diseases made it 
less crucial to have additional guarantees that such studies would be conducted 
promptly. Moreover, the sponsors of didanosine and zalcitabine agreed prior to 
approval to expedited withdrawal of the drug from the market if benefit were not 
shown. The provisions of the final rule will ensure that appropriate safeguards 
exist for timely generation of data on actual clinical benefit, for appropriate 
promotional information about labeled indications, and for prompt withdrawal of 
the drug from the market if clinical benefit is not confirmed.  

   2. Pointing to a statement in the preamble to the proposed rule that it is in 
the public interest to make promising new treatments available at the earliest 
possible point in time for use in life-threatening and serious illnesses, one 
comment expressed concern that the proposed rule may lead to the marketing of 
large numbers of clinically ineffective, but pharmacologically active, drugs and 
this may not be in the interest of the public health. The comment argued that 
early access to so-called "promising" drugs is not the same as early access to 
safe and effective drugs, and the number of potential markers that may be 
advanced as surrogates of clinical outcome is exceedingly large. The comment 
suggested that it may be more appropriate to seek adoption of the proposed 
requirements through an amendment to the act.  

   FDA agrees with the contention that providing people who have serious or 
life-threatening illnesses with numerous clinically ineffective drugs would not 
be helpful. However, the agency does not agree that the rule can be expected to 
have this result. Although studies using surrogate endpoints may provide less 
assurance of clinical benefit than studies using clinical endpoints, FDA 
believes compliance with all of the elements of the accelerated approval program 
will not result in the marketing of large numbers of clinically ineffective 
drugs. The new procedures apply to a limited group of circumstances, namely, to 
drugs intended for serious or life-threatening illnesses when the drugs provide 
a meaningful therapeutic benefit over existing therapy. Reliance on a surrogate 
endpoint is not equivalent to reliance on any evidence of pharmacologic 
activity. The endpoint must be reasonably likely, based on epidemiologic, 
therapeutic, pathophysiologic, or other evidence, to predict clinical benefit.  

   Whether a given endpoint is, in fact, reasonably likely to predict clinical 
benefit is inevitably a matter of judgment. FDA, using available internal and 
external expertise, will have to make informed judgments in each case presented, 
just as it does now. The agency acknowledges that there are well-recognized 
reasons for caution when surrogate endpoints are relied on. Certain putative 
surrogates have ultimately been shown not to correspond to clinical benefit. 
Perhaps the most noteworthy example is the failure of antiarrhythmic agents in 
the Cardiac Arrhythmia Suppression Trial (CAST) to improve survival by 
depressing ventricular ectopic beats; effective suppression of ectopic beats was 
associated with increased mortality.  

   A sponsor must persuasively support the reasonableness of the proposed 
surrogate as a predictor and show how the benefits of treatment will outweigh 
the risks. Such presentations are likely to be persuasive only when the disease 
to be treated is particularly severe (so that considerable risk is acceptable) 
and/or when the surrogate endpoint is well supported. In addition, it will be 
the sponsor's clear obligation to resolve any doubts as to clinical value by 
carrying out definitive studies.  

   FDA does not agree that it would be more appropriate to seek an amendment to 
the act than to adopt the proposed requirements. As discussed in the preamble to 
the proposed rule as well as elsewhere in this preamble to the final rule, 
existing provisions of the act and the PHS Act authorize promulgation of the 
requirements in the final regulations.  

   3. One comment expressed concern that because the proposed rule would 
establish conditions on a drug's approval, third-party payors may decline 
reimbursement because the so-called approval would have attributes of 
investigational status.  

   The agency expects that, because drugs approved under the accelerated 
approval process meet the statutory standards for safety and effectiveness, they 
would be eligible for reimbursement under State Medicaid programs or other 
third-party plans. Drug products granted accelerated approval will not be, under 
the law, investigational, as suggested by the comment.  

    4. One comment asked if all drugs considered for accelerated approval must 
be reviewed by an advisory committee. The comment stated that because advisory 
committees meet infrequently, waiting for the next meeting may slow down the 
approval process.  

   FDA is not required to consult with an advisory committee before approving an 
application under these accelerated approval regulations, or any other 
regulation. However, FDA intends to consult the appropriate committee in most 
instances. Advisory committee meetings can usually be scheduled to avoid 
significant delays in the review process. The agency will consider any request 
by an applicant for referral of the application to an advisory committee.  

   
B. Scope  

   5. Four comments asked for further clarification of what diseases are covered 
by the rule. One comment stated that the terms "serious," and "life-
threatening," are defined in the proposal by reference to 21 CFR 312.34, 
followed by a brief statement explaining the role of judgment and examples of 
diseases that are currently judged to be serious. The comment asked that FDA 
also describe: (1) Diseases that are not currently included in the category of 
"serious," (2) examples of diseases that are currently judged "life-
threatening," and (3) examples of diseases that are not currently included in 
the category "life-threatening."  

   One comment contended that the statement in the preamble that "seriousness of 
a disease is a matter of judgment, but generally is based on its impact on such 
factors as survival, day-to-day functioning, or the likelihood that the disease, 
if left untreated, will progress from a less severe condition to a more serious 
one" too narrowly limits diseases covered by the proposed rule (57 FR 13234 at 
13235). The comment argued that some "less severe" diseases, even if treated, 
may progress to a more serious state, and that these diseases should also be 
covered by the rule. On the other hand, two comments argued that the language in 
the preamble that classifies diseases as "serious" was overly broad and 
subjective and far too large a number of illnesses could be eligible as being 
"serious."  

   FDA discussed the meaning of the terms "serious" and "life-threatening" in 
its final rules on "treatment IND's" (52 FR 19466 at 19467, May 22, 1987) and 
"subpart E" procedures (54 FR 41516 at 41518-41519, October 21, 1988). The use 
of these terms in this rule is the same as FDA defined and used the terms in 
those rulemakings. It would be virtually impossible to name every "serious" and 
"life-threatening" disease that would be within the scope of this rule. In FDA's 
experience with "treatment IND's" and drugs covered by the "subpart E" 
procedures there have not been problems in determining which diseases fall 
within the meaning of the terms "serious" and "life-threatening," and FDA would 
expect no problems under this accelerated approval program. The likelihood of 
progression to a serious condition with available treatments would also be 
considered in assessing whether the disease is within the scope of the final  
rule. The preamble to the proposed rule (57 FR 13234 at 13235) referred to 
chronic illnesses that are generally well managed by available therapy, but can 
have serious outcomes for certain populations or in some or all of their phases. 
Applicants are encouraged to consult with FDA's reviewing divisions early in the 
drug development process if they have questions about whether their specific 
product is within the scope of this rule.  

    The concerns expressed in these and other comments about considering too 
many illnesses eligible for consideration under the accelerated approval 
procedures may arise from the underlying fear that reliance on surrogate 
endpoints will become routine, the "normal" way drugs are brought to the market. 
This fear is groundless. The vast majority of drugs are directed at symptomatic 
or short-term conditions (pain, heart failure, acute infections, 
gastrointestinal complaints) whose response to drugs, if it occurs, is readily 
measured and where there is no need to consider or accept surrogate endpoints. 
Surrogates, with few exceptions, are of interest in the following situations: 
(1) Where the clinical benefit, if there is one, is likely to be well in the 
future; and (2) where the implications of the effect on the surrogate are great 
because the disease has no treatment at all or the drug seems to treat people 
with no alternative (e.g., because they cannot tolerate the usual effective 
treatment). In the first case, great care is needed, and would be given, as 
there would generally be no experience linking an effect on the surrogate to 
clinical success, and there have been conspicuous examples of lack of linkage 
(CAST, referred to above; drugs that increase cardiac output in patients with 
heart failure but that decrease survival; imperfect agreement of effects on 
coronary artery patency and effects on survival in patients with myocardial 
infarction; lack of beneficial effect on bone fracture rate despite favorable 
effects on bone density in patients with osteoporosis). FDA and outside experts 
will be aware of these examples as proposed surrogates are considered. The 
implications are especially great when considering prophylactic therapy, i.e., 
treatments to prevent chronic illness (coronary artery disease, cancer), in an 
essentially well population. In the second case, there will generally have been 
experience (with the standard therapy) to evaluate in considering linkage of the 
surrogate to benefit; this was, for example, the case with didanosine, where 
evidence from zidovudine studies of the relationship of an effect on CD4 
lymphocytes and clinical outcome could be assessed. Similarly, there is 
considerable experience to show that durable complete responses in many cancers 
correspond to improved survival, so that an agent inducing them in refractory 
illness or in primary  

   
disease that had previously been poorly responsive would generally be seen as 
reasonably likely to provide a clinical benefit.  

   6. One comment stated that epilepsy is a serious and life-threatening 
condition and asked that it be included within the scope of the proposal. The 
preamble cited, among other illnesses, depression and psychoses as examples of 
chronic illnesses that can have serious outcomes even if they are generally well 
managed. One comment asserted that neither depression nor psychosis is a 
disease, nor is either one serious or life-threatening. The comment stated that 
depression and psychosis are diagnoses. The comment urged the agency to remove 
them from the definition of life-threatening "illnesses" or "diseases."  

   With respect to epilepsy, FDA notes that in the "treatment IND" final rule 
(52 FR 19466 at 19467, May 22, 1987), the agency listed "certain forms of  
epilepsy" as an example of a disease or stage of disease that would normally be 
considered "serious." Certain forms of epilepsy may also be considered "serious" 
under the accelerated approval program. It is unlikely, however, that a 
surrogate endpoint would be utilized in such a case, as seizure frequency, a 
clinical endpoint, is readily measured.  

   FDA's reference to depression and psychoses was intended to give examples of 
conditions or diseases that can be serious for certain populations or in some or 
all of their phases. While drugs for the treatment of depression and psychosis 
would be examples of those that could be covered by the accelerated approval 
program, it is not the use of surrogate endpoints that would be expected; the 
symptoms and signs of these diseases are readily studied. On the other hand, 
some of these drugs have been quite toxic (e.g., clozapine for refractory 
psychoses) and might be considered for approval with restrictions to ensure safe 
use.  

   7. Two comments asked how FDA will decide that a drug is eligible for 
accelerated approval. One comment asserted that the decision should be an option 
for the applicant to consider, not a decision for FDA to make unilaterally. 
Pointing to a statement in the preamble (57 FR 13234 at 13235) that FDA reserves 
the right not to apply accelerated approval procedures when it believes in good 
faith that the drug's foreseeable use is reasonably likely to be outside the 
scope of "life-threatening diseases without meaningful therapeutic benefit over 
existing therapy," the comments argued that, if there are patients with life-
threatening conditions that can benefit from expedited approval, the needs of 
the patients should determine the procedures used to approve the drug. One 
comment contended that applicants of products considered candidates for 
accelerated approval may have their drug or biological product "forced" into the 
accelerated approval process and be forced to conduct a program of studies to 
substantiate that surrogate endpoints actually predict significant clinical 
benefits.  

   The medical reviewing divisions within FDA's Center for Drug Evaluation and 
Research (CDER) and Center for Biologics Evaluation and Research (CBER) will 
determine the type of regulatory review that FDA may apply to an application. 
FDA encourages sponsors to meet with FDA early in the drug development process 
to discuss the applicability of the accelerated approval program to their 
product; however, FDA reserves the discretion to determine whether these 
procedures are applicable to a specific product.  

   With respect to the preamble statement cited by one comment, the comment 
misreads the preamble statement, which does not say that FDA will, in all cases, 
apply FDA's traditional approval mechanisms rather than this accelerated process 
for drugs where a majority of the drug's foreseeable uses are outside the scope 
of "life-threatening" diseases without meaningful therapeutic benefit over 
existing therapy. The statement merely informs applicants that FDA will consider 
the possible impact of widespread use of a drug for uses other than the one 
supporting accelerated approval; drugs approved under this program would often 
have only small safety data bases so that widespread off-label use might have 
serious implications. The agency does not believe that such a situation would 
regularly lead to exclusion from these provisions.  

   FDA does not agree that applicants seeking approval to market drug and 
biological products that would be candidates for accelerated approval will be 
forced to use the accelerated approval mechanism. It is true, however, that  
some proposed surrogate endpoints would not be considered acceptable bases for 
approval without assurance that the clinical studies to show clinical benefit 
will be conducted. A sponsor that wishes the application to be considered under 
the traditional approval process may request and receive such consideration.  

   The agency wishes to clarify the circumstances in which the accelerated 
approval regulations will apply. Sections 314.500 and 601.40 describe aspects of 
the scope of these regulations. Moreover, these regulations are intended to 
apply to applications based on surrogate endpoints whose validity is not fully 
established, to applications based on clinical endpoints that leave unanswered 
major questions about the product's effect on ultimate outcome, and to 
applications for products whose safe and effective use requires limitations on 
distribution or use. In all other situations, accelerated approval requirements 
will not apply.  

   Where approval is based on a surrogate endpoint that is accepted as validated 
to predict or correlate with clinical benefit, the product will be considered 
under the traditional process, and the postmarketing requirements under 
accelerated approval will not apply. Approvals of products for serious or life-
threatening illnesses based on clinical endpoints other than survival or 
irreversible morbidity will usually also be considered under traditional 
procedures. Approvals based on such clinical endpoints will be considered under 
the accelerated approval regulations only when it is essential to determine 
effects on survival or irreversible morbidity in order to confirm the favorable 
risk/benefit judgment that led to approval. Applications for products for 
serious or life-threatening illnesses that provide a meaningful therapeutic 
benefit over existing therapy will receive a priority rating and expedited 
review, even when not considered under the accelerated approval procedures.  

   The agency also wishes to clarify that whenever an application is approved 
under § 314.510 or § 601.41, postmarketing studies confirming the product's 
clinical benefit will thus be required. Therefore, in order to eliminate 
potential confusion, the agency has amended §§ 314.510 and 601.41 to clarify 
these points.  

   FDA also recognizes that over time a particular surrogate, once acceptable as 
a basis for approval only under the accelerated approval regulations, could 
become recognized as validated by definitive studies (just as high blood 
pressure, for example, over time became validated as a surrogate with clinical 
significance). In such cases, a future application relying on such a surrogate 
would not require postmarketing studies confirming the surrogate's clinical 
benefit and the application would be considered under traditional procedures.  

    8. Two comments asked for clarification of the phrase "meaningful 
therapeutic benefit over existing therapy" as used in the description of what 
drugs the accelerated approval program should apply to. Specifically, pointing 
to an example described in the preamble that a new therapy would be eligible for 
accelerated approval if there was "a clear improvement" over existing therapy in 
being more effective or better tolerated, one comment urged FDA to clarify the 
meaning of "clear improvement" to discourage applicants of "me-too" products 
from wasting the agency's time and resources by applying for accelerated 
approval of such products. The comment also asked that FDA specify that if a new 
drug is approved under the accelerated approval provisions because the drug 
exhibits a "clear improvement" over an existing drug that was also granted 
accelerated approval, then specific restrictions will be placed on the prior  
approved drug to limit its use only to patients who cannot tolerate the new 
drug, or whose physicians assess that a change to the new drug might involve 
significant risks to the patient that outweigh the benefits. One comment asked 
that the term "meaningful therapeutic benefit over existing therapy" be 
interpreted and consistently applied to both drugs and biological products.  

   FDA believes that the examples given to help clarify the phrase "meaningful 
therapeutic benefit over existing therapy" (ability to treat unresponsive or 
intolerant patients or improved response compared to available therapy) are 
readily understood illustrations of the intent of the requirement. A drug that 
is essentially the same as available treatment (what the comment refers to as a 
"me too" drug) will not have a credible claim to a meaningful therapeutic 
benefit over that existing treatment and this should be easily detected.  

   With respect to restricting use of a drug previously approved under 
accelerated approval procedures when a new drug granted accelerated approval is 
a clear improvement over the prior approved drug, this would rarely be 
appropriate. Although, in some instances, certain therapies are identified as 
"second-line," this requires essentially unequivocal evidence of an advantage of 
alternative therapy, not likely on the basis of a surrogate endpoint. Labeling 
for both drugs will be accurate, however, allowing physicians to prescribe both 
the newly approved drug and the prior drug properly.  

   9. One comment asked if a change in the route of administration would be 
considered as a meaningful benefit and within the scope of the proposal.  

   A change in the route of administration may be a candidate for accelerated 
approval depending upon the particular evidence presented.  

   10. One comment asked if subpart E drugs currently under investigation will 
be considered for accelerated approval. The comment assumed that new drug 
applications (NDA's) and supplemental NDA's considered for accelerated approval 
will have the highest priority for review.  

   Subpart E drugs will be considered for accelerated approval if they satisfy 
both eligibility criteria for accelerated approval, i.e., if they are being 
developed for the treatment of serious or life-threatening illnesses and the 
products will provide meaningful therapeutic benefits to patients over existing 
treatment. As discussed above, applicants should consult with FDA early in the 
development process to determine the nature of the regulatory review. Early 
consultations are a critical part of subpart E procedures. Drugs being reviewed 
under accelerated approval procedures will receive high priority review. 
However, applications for drugs for acquired immunodeficiency syndrome (AIDS) 
and human immunodeficiency virus (HIV)-related conditions will receive the 
highest priority review.  

   
C. Criteria for Approval  

   11. Two comments expressed concern that the proposal did not provide enough 
detail on what constitutes an appropriate surrogate endpoint. One comment 
recommended that FDA adopt specific criteria for what constitutes an appropriate 
surrogate endpoint. The comment suggested that such criteria should include: (1) 
The surrogate endpoint must be biologically plausible in that it must be 
consistent with what is known about the pathophysiology and pathogenesis of  the 
disease; (2) the surrogate endpoint must be present or abnormal in a large 
percentage of people who have the disease; (3) the surrogate endpoint must be a 
good predictor of the disease progression and should correlate closely with the 
significant clinical endpoint; (4) there should be a correlation between the 
quantitative aspect of the surrogate endpoint and the progression of the disease 
(e.g., the more severe the disease, the more deviant the surrogate endpoint from 
normal); (5) the regression of the surrogate endpoint should be significantly 
associated with clinical improvement (e.g., those with the greatest improvement 
in the surrogate endpoint should also show the greatest clinical effects); 
conversely, the lack of regression of the surrogate endpoint should be commonly 
associated with a lack of clinical improvement; and (6) the incidence of 
regression or improvement in the surrogate endpoint should be significantly 
greater in treated than untreated patients.  

   One comment asked if the use of microalbuminuria data is a surrogate for 
diabetic nephropathy and if all drugs relying on surrogate endpoints would be 
eligible for accelerated approval, e.g., an angiotensin receptor antagonist with 
potential utility for treatment of congestive heart failure. The comment also 
asked what would happen if postmarketing studies demonstrate beneficial changes 
of surrogate endpoints but not beneficial clinical endpoints. The comment also 
asked if FDA will consider publishing guidelines on which surrogate endpoints 
would be appropriate for the diseases that may be affected by the proposed rule. 
Another comment expressed the belief that there is no evidence that surrogate 
endpoints are necessarily good indicators of therapeutic benefit. The comment 
stated that a drug may have an effect on a surrogate endpoint, but will not make 
any clinical difference because the advanced stage of the patient's disease 
precludes any effective therapy or the surrogate marker is not synchronous with 
the patient's clinical condition.  

   Another comment asserted that the requirement to base an approval on a 
surrogate endpoint that is "reasonably likely, based on epidemiologic, 
therapeutic, pathophysiologic, or other evidence, to predict clinical benefit 
other than survival or irreversible morbidity" is not restrictive enough to 
assure adequate consumer protection. Terms like "reasonably likely" and "or 
other evidence" allow drug manufacturers too much latitude for claiming that 
there is a correlation between surrogate endpoints affected by their drugs and 
clinical endpoints. The comment argued that until a correlation between a 
surrogate endpoint and a clinical endpoint has been established, a particular 
surrogate endpoint should only be used to approve subsequent drugs, without 
adequate clinical evidence, if there is a very strong effect of the drug on the 
surrogate marker or, if the effect is not sufficiently strong, there is an 
additional surrogate marker which corroborates the results of the first.  

   FDA intends to publish informal guidance concerning surrogate endpoints, but 
does not believe specific requirements for an appropriate surrogate should be 
specified by regulation. Any given specifications may not be applicable to a 
particular case. For example, the thoughtful suggested criteria supplied by the 
comment would rarely, if ever, be applicable to the first effective drug for a 
disease, because criterion 5 requires that regression of the surrogate endpoint 
be associated quantitatively with clinical improvement. If there had never been 
effective treatment, this would never be known. Yet the surrogate could be 
persuasive on other grounds, such as a well-documented etiologic relation. In 
general, it is likely that one or another strongly supportive piece of evidence 
might outweigh gaps in other areas.  

   In developing informal guidance on surrogate endpoints, FDA will consider the 
suggestions in this comment. Interested persons will have an opportunity to 
comment on any guidance documents in this area developed by the agency. In some 
cases, new or revised drug class, or disease-specific, clinical guidelines may 
refer to surrogate endpoints. FDA is not prepared, at this time, to comment on 
the acceptability of an endpoint that it has not specifically considered, e.g., 
microalbuminuria.  

   The final regulations make it clear that not all drugs submitted for approval 
based on surrogate endpoint data are eligible for accelerated approval (§§ 
314.500 and 601.40). The drug in question must be for a serious or life-
threatening condition and must provide meaningful therapeutic benefit over 
existing therapy. In the case of an angiotensin receptor antagonist posed by the 
comment, there is existing documented life-prolonging treatment for congestive 
heart failure. An application for a new agent, to be eligible for accelerated 
approval, would have to show potential benefit over available therapy as well as 
identify a reasonable surrogate endpoint. This is problematic since no accepted 
surrogate endpoint for studies to treat congestive heart failure has been 
identified to date. For example, some drugs with favorable effects on 
hemodynamic measures in heart failure patients have been clinically ineffective.  

   The regulations are clear in requiring that, for drugs approved under these 
provisions based on surrogate endpoints, the postmarketing studies must show 
clinical benefit, not just the previously shown effect on the surrogate (§§ 
314.510, 314.530, 601.41, and 601.43).  

   Surrogates, or proposed surrogates, are not always good, nor necessarily bad, 
indicators of therapeutic benefit and must be judged on a case-by-case basis. 
Even very good surrogates may not be perfect: Blood pressure lowering has been a 
better predictor of effect on stroke than on coronary artery disease, 
cholesterol lowering has had a clearer effect on coronary artery disease than on 
survival. Moreover, a surrogate may be persuasive for a phase of disease with 
short expected survival but much less so in an earlier phase of the disease. 
Caution is always appropriate in evaluating surrogate endpoints and the 
particular therapeutic setting should always be considered. The agency believes 
that the evaluation of surrogate endpoint data and the safeguards built into 
these accelerated approval procedures will provide adequate consumer protection.  

    12. One comment expressed concern that if there is no accepted surrogate 
endpoint, an applicant's only option is to conduct a study using some clinical 
event as an endpoint, which may result in long, large studies that delay 
approval to the detriment of patients and sponsors. One comment suggested as an 
alternative that FDA permit approval of a drug based on a study using a clinical 
endpoint, but accept a less rigorous standard of statistical significance, e.g., 
0.20 or 0.15 instead of 0.05. The comment further suggested that the sponsor 
could then complete postmarketing studies to establish statistical significance 
at conventional levels. The comment argued that this alternative is totally 
consistent with FDA's willingness to accept greater uncertainty in approving 
drugs for serious and life-threatening illnesses.  

   The intent of the rule is to allow FDA to utilize a particular kind of 
evidence, an effect on a surrogate endpoint, as a basis for approval, and, where 
appropriate, to ensure that remaining doubts about the relationship of the 
effect on the surrogate to clinical benefit are resolved by additional adequate 
and well-controlled studies with clinical endpoints. The rule is not intended  
to place into the market drugs with little evidence of usefulness. Although 
there is no statutory requirement for significance testing of any particular 
value, there are well-established conventions for assessing statistical 
significance to support the statutorily required conclusion that the well-
controlled studies have demonstrated that a drug will have the effect it is 
represented to have. There is nothing about serious or life-threatening diseases 
that make them uniquely difficult to study. A meaningful effect on survival or 
morbidity where there is no effective therapy should be readily discerned. Such 
studies need be long and large only when the effect is small or difficult to 
detect. In that event, proper assessment of benefit, and valid weighing of its 
relation to risk, is especially critical.  

   13. One comment asked that FDA clarify that one study could be the basis of 
approval and that one postmarketing study should be all that is needed to 
establish the link between the endpoint used for approval and some relevant 
clinical benefit.  

   FDA interprets the statute, and good science, as requiring at least two 
adequate and well-controlled studies to establish effectiveness. In some 
instances, drugs have been approved on the basis of a single well-controlled 
study; this has been done where the study was of excellent design, showed a high 
degree of statistical significance, involved multiple study centers, and showed 
some evidence of internal replicability, e.g., similar effects in major study 
subsets. FDA encourages applicants to discuss with FDA early in a drug's 
development the basis for the applicant's choice of a specific endpoint and, 
where applicable, the basis for its belief that a single study would be a 
sufficient basis for approval. With respect to postmarketing studies, FDA 
anticipates that the requirement will usually be met by studies already underway 
at the time of approval. As stated in the proposed rule, the requirement for any 
additional study to demonstrate actual clinical benefit will not be more 
stringent than those that would normally be required for marketing approval of 
the same drug for the same claim.  

   14. One comment expressed concern that the preamble to the proposed rule 
implied that a sponsor of an AIDS drug might have to do a postmarketing study to 
establish an effect on survival after showing an effect on such endpoints as 
weight or incidence of opportunistic infection (57 FR 13234 at 13235-13236). The 
comment stated that FDA's own advisory committee indicated that it was pleased 
to see an effect from a nucleoside analogue on the incidence of opportunistic 
infections with AIDS patients but did not suggest that further work should be 
done to show an effect on mortality. The comment argued that in some cases 
direct correlation with clinical endpoints such as mortality is difficult to 
prove and urged FDA to be flexible on this issue to encourage sponsors to go 
through the accelerated approval process.  

   Ordinarily, an effect on a meaningful clinical endpoint, e.g., on rate of 
opportunistic infections in AIDS, is a sufficient basis for approval without 
need for followup studies. Other endpoints, however, might leave major questions 
unanswered. For example, a modest effect on weight gain in AIDS without other 
demonstrated benefit, if considered an adequate basis for approval, while a 
clinical endpoint, might leave sufficient doubt as to the ultimate value of the 
effect so that further studies would be necessary. FDA intends to interpret this 
provision of the regulations with flexibility. This provision should also serve 
as a reminder, however, that for life-threatening diseases, the ultimate aim of 
therapy is improved survival as well as improved symptoms.  

   15. One comment asked FDA to clarify what a sponsor's obligation is to 
continue supplying medication on a compassionate basis if clinical efficacy is 
not demonstrated to FDA's satisfaction in postmarketing studies but individual 
patients appear to be benefiting from use of the drug.  

   Sponsors are not obligated to supply drugs on a "compassionate basis." 
Whether, if clinical studies did not show effectiveness, further availability of 
the drug would be appropriate under any mechanism would be determined case-by-
case.  

   
D. Promotional Materials  

   16. Three comments asserted that requiring advance submissions of promotional 
materials is both beyond FDA's statutory authority and is unnecessary. Although 
FDA stated in the proposal that it does not intend specifically to approve 
promotional materials, two comments contended that is the likely effect of 
advance submission. The comment cited section 502(n) of the act (21 U.S.C. 
352(n)), which provides that no regulation promulgated under that provision 
shall require prior FDA approval of the content of any advertisement "except in 
extraordinary circumstances," and asserted that the "extraordinary 
circumstances" language would not apply to drugs approved under the accelerated 
approval program. One comment argued that submission of promotional material 
prior and subsequent to approval is unwarranted when dealing with treatments for 
serious or life-threatening illnesses where dissemination of the most current 
and timely information is important to the treating physician. One comment 
questioned why there would be any greater likelihood of misleading promotional 
claims for products approved under the proposed accelerated approval process 
than for drugs intended to treat serious or life-threatening diseases that are 
approved under the normal NDA procedures. The comment also expressed the hope 
that the proposed requirement for advance submission of promotional materials 
was not based upon an assumption that promotional materials for drugs intended 
to treat serious diseases are more likely to be misleading than promotional 
materials for other types of drugs because any such assumption would be 
unfounded. One comment argued that if an advertisement or labeling is 
inaccurate, the product is misbranded and FDA could then obtain injunctive 
relief, seize the product, and/or initiate criminal proceedings. Another comment 
considered requiring advance submission of promotional materials unreasonable 
because companies are not required to do so now. One comment questioned the 
legal authority for requiring presubmission of promotional material following 
approval of a drug product, and the reason for the requirement.  

   The agency believes that the requirements for submission of promotional 
materials in the context of accelerated approval are authorized by statute. 
Subsections 505(d)(4) and (d)(5) of the act provide that, in determining whether 
to approve a drug as safe and effective, the agency may consider not only 
information such as data from clinical studies but also "any other information" 
relevant to safety and effectiveness under the proposed conditions of use. Such 
information would include information about how the drug would be promoted. In 
determining whether the drug's proposed labeling would be "false or misleading" 
under section 505(d)(7) of the act, the agency is similarly authorized to 
evaluate "all material facts" during the approval process, including the facts 
about promotion.  

   FDA is also authorized by section 505(k) of the act to require reporting of 
information subsequent to approval necessary to enable the agency to determine 
whether there may be grounds for withdrawing the approval. Among the grounds for 
withdrawal specified in section 505(e) of the act are that the evidence reveals 
the drug is not shown to be safe and effective under its conditions of use. In 
addition, drug approval may be withdrawn if information shows the labeling to be 
false or misleading. Information on how the drug will be promoted is again 
relevant to whether the drug's marketing approval should be withdrawn. Section 
701(a) of the act (21 U.S.C. 371(a)) generally authorizes FDA to promulgate 
regulations for the efficient enforcement of the act.  

   For biological products, additional authority in section 351 of the PHS Act 
(42 U.S.C. 262) authorizes the promulgation of regulations designed to ensure 
the continued safety, purity, and potency of the products. The content of 
promotional materials is important to the continued safe and effective use of 
biologicals.  

   Therefore, the provisions of the final rule requiring submission of 
promotional materials prior to approval under the accelerated approval 
procedures and subsequent to such approval are authorized by statutory 
provisions. FDA might also invoke the authority of section 502(n) of the act (21 
U.S.C. 352(n)) to require prior approval of the content of any prescription drug 
advertisement in "extraordinary circumstances." Whether FDA could appropriately 
rely on section 502(n) of the act in promulgating §§ 314.550 and 601.45 need not 
be determined, however, because FDA is not relying upon section 502(n) of the 
act as legal authority for these (or any other) sections of the accelerated 
approval regulations.  

   The agency believes that advance submissions of promotional materials for 
accelerated approval products are warranted under the accelerated approval 
circumstances. The special circumstances under which drugs will be approved 
under these provisions and the possibility that promotional materials could 
adversely affect the sensitive risk/benefit balance justify review of 
promotional materials before and after approval. For example, if the promotional 
materials exaggerate the known benefits of the drug, wider and inappropriate use 
of the drug could be encouraged, with harmful results.  

   Similarly, high risk drugs that are approved based on postmarketing 
restrictions would not have been approved for use without those restrictions 
because the risk/benefit balance would not justify such approval. If promotional 
materials were to undermine the postmarketing restrictions, the health and 
safety of patients could be greatly jeopardized.  

   Although there is potential harm from any misleading promotion, and there is 
no reason to believe improper promotion is more likely in this setting than in 
others, the risk/benefit balance is especially sensitive in this setting. The 
relatively small data base available and the minimal published information 
available also can contribute to making the physician and patient populations 
particularly vulnerable under accelerated approval circumstances.  

   Reliance on court actions (such as seizures, injunctions, and criminal 
prosecutions) can be effective in ending false promotions, but can only be 
initiated after the fact, when harm has already occurred. Corrective efforts can 
be helpful but are always somewhat delayed. Under the circumstances of 
accelerated approval, FDA believes that it is far preferable to avoid problems  
by reviewing the promotional materials in advance of drug approval and of 
dissemination of the materials.  

   17. Two comments supported the provision about submission of promotional 
materials. One comment urged the agency to require that specific patient 
information be included in promotional materials to indicate the fact that the 
drug's clinical benefit has not yet been established. For drugs approved under 
the restricted use provision, the comment recommended that the labeling specify 
in detail the exact restrictions placed on the drug. In both cases, the comment 
recommended that this patient information appear as boxed warnings.  

   Section 502(n) of the act and regulations at § 202.1(e)(1) (21 CFR 
202.1(e)(1)) require prescription drug advertisements (promotional material) to 
contain, among other things, a true statement of information in brief summary 
relating to side effects, contraindications, and effectiveness, which would 
include warnings, precautions, and limitations on use. The information in brief 
summary relating to side effects, contraindications, and effectiveness is 
required to be based solely on the approved labeling. Therefore, to the extent 
that a drug's labeling reflects the extent of clinical exposure and includes 
appropriate warnings, a drug's promotional material would also include this 
information.  

   FDA regulations governing prescription drug labeling (21 CFR 201.56 and 
201.57) require that serious adverse reactions and potential safety hazards, as 
well as limitations in use imposed by them, be included in the "Warning" section 
of the labeling. In the case of approval based upon effect on a surrogate 
endpoint, the "Indications and Usage" section of the labeling would reflect the 
nature of the demonstrated effect. If the approval is based on use restrictions, 
the label would also specify the restrictions.  

   FDA may require boxed warnings if there are special problems associated with 
a drug, particularly those that may lead to death or serious injury (21 CFR 
201.57(e)). The agency does not agree that information related to clinical 
benefit or use restrictions for accelerated approval drugs would necessarily 
always require a boxed warning.  

   As indicated by §§ 314.550 and 601.45 of the final rule, applicants will be 
required to submit promotional materials prior to approval and in advance of 
dissemination subsequent to approval whether the product is a new drug, an 
antibiotic, or a biological product.  

   18. One comment contended that FDA review and approval of all promotional 
pieces before their use will indefinitely delay product marketing campaigns and 
other patient and physician educational activities, which are essential to 
market a product, thereby significantly diminishing the advantage of securing an 
early approval for the applicant. The comment further contended that the 
requirement to submit "all promotional materials * * * intended for 
dissemination or publication upon marketing approval" will be overly burdensome 
for FDA and will unnecessarily slow down the process for review of all 
materials, not just those for products subject to this proposed rule. The 
comment recommended that FDA only request for review the primary advertising 
pieces, such as the introductory letter to physicians, the main detail piece, 
and the main journal advertisement, but not the secondary materials, e.g., a 
letter to pharmacists, of the initial promotional campaign.  

   As previously discussed in this preamble, FDA will be reviewing an 
applicant's planned promotional materials both prior to approval of an 
application (reflecting the initial campaign) and subsequent to approval to 
ascertain whether the materials might adversely affect the drug's sensitive 
risk/benefit balance. Because all promotional materials, including those 
referred to by the comment as "secondary" materials, can have significant 
adverse effects if they are misleading, the agency does not agree that such 
materials should, as a matter of course, not be requested for review. Insofar as 
such materials may be directly derived from the introductory letter to 
physicians, or other materials characterized by the comment as "primary" 
materials, the additional time to review the derivative materials should not be 
extensive.  

   The agency does not agree with the comment's contention that the requirement 
to submit all promotional materials prior to and subsequent to approval will 
indefinitely delay marketing campaigns and educational activities or be overly 
burdensome to FDA reviewers. FDA is committed to rapid review and evaluation of 
all drugs considered for approval under this rule and will promptly review the 
promotional materials.  

   19. One comment suggested a passive, time-limited clearance system for review 
of advertising after the initial promotional campaign such as that used for 
review of IND's, which would allow the sponsor to proceed to use promotional 
materials after an allotted timeframe, such as 30 days, unless otherwise 
notified by FDA.  

   As indicated by this comment and others, additional clarification regarding 
both timing and content of the submissions of promotional materials seems 
useful. Therefore, the agency is revising proposed §§ 314.550 and 601.45 to make 
it clear that, unless otherwise informed by the agency, applicants must submit 
during the preapproval review period copies of all promotional materials 
intended for dissemination or publication within the first 120 days following 
marketing approval. The initial promotional campaign, sometimes referred to as 
the "launch campaign," often has a significant effect on the climate of use for 
a new product. As discussed elsewhere in this preamble, the risk/benefit balance 
of accelerated approval products is especially sensitive, and inappropriate 
promotion may adversely affect the balance with resulting harm.  

   There may be some instances in which promotional materials that had not been 
completed and submitted by the applicant prior to approval would be beneficial 
in fostering safe and effective use of the product during the first 120 days. 
Under revised §§ 314.550 and 601.45, FDA would have the discretion to consider 
such materials at a later time. An applicant who requested permission to include 
additional materials among those disseminated within the first 120 days 
following product approval would be notified of FDA's determination. If FDA 
agreed that dissemination of such materials was acceptable, the materials could 
then be disseminated or published upon notification.  

   For promotional materials intended for dissemination subsequent to the 
initial 120 days under §§ 314.550 and 601.45 FDA would review the submitted 
materials within 30 days of receipt. This 30-day period is meant to be time-
limited, so that the applicant will be assured of no unnecessary delay. It will 
be important for the applicant to identify the materials being submitted 
appropriately, so that it is clear that the materials are subject to the 30-day 
review period. The agency intends to review all such materials promptly, and  to 
notify the applicant of any identified problems as soon as possible. The agency 
expects that, if the agency notifies the applicant of significant objections to 
the proposed materials, no materials will be disseminated or published until the 
agency's objections are resolved. The applicant should plan to allow sufficient 
time after receiving FDA's comments for resolving differences and incorporating 
requested changes in the submitted materials prior to dissemination or 
publication.  

   When FDA removes the requirement for advance submission of promotional 
material, the agency will continue to offer a prompt review of all voluntarily 
submitted promotional material.  


   
E. Postmarketing Restrictions  

   FDA received many comments on the proposed requirement to limit distribution 
to certain facilities or physicians with special training or experience, or 
condition distribution on the performance of specified medical procedures if 
such restrictions are needed to counterbalance the drug's known safety concerns.  

   20. Several comments questioned FDA's authority to impose restrictions on 
distribution or use after an approved drug is marketed. Two comments disagreed 
with the statutory provisions cited by FDA in the proposed rule as its authority 
to impose restrictions on distribution or use stating that they refer only to 
FDA's general authority to ensure that drugs are not misbranded, which is an 
entirely separate issue. Another comment argued that section 503(b) of the act 
(21 U.S.C. 353(b)) contemplates that the issues warranting a restriction as to 
distribution are not factors in whether a drug product is "safe" for purposes of 
approval, but rather only whether the product must be limited to prescription 
status. Two comments said that, in the absence of specific statutory authority, 
the courts clearly have refused to permit FDA to impose restrictions on 
distribution and cited American Pharmaceutical Association (APhA) v. Weinberger, 
377 F. Supp. 824, 829 n. 9 (D.D.C. 1974), aff'd sub nom. APhA v. Mathews, 530 
F.2d 1054 (D.C. Cir 1976), a case concerning conditions placed on the approval 
of the drug methadone.  

   Some comments asserted that placing restrictions on the distribution of an 
approved drug to only certain facilities or physicians, or restricting use to 
certain medical procedures interferes with the practices of medicine and 
pharmacy, which the comments contended FDA does not have the authority to 
regulate.  

   The agency believes that the restrictions to ensure safe use contemplated for 
approvals under §§ 314.520 and 601.42 are authorized by statute. As discussed in 
the preamble to the proposed rule (57 FR 13234 at 13237), sections 501, 502, 
503, 505, and 701 of the act provide broad authority for FDA to issue 
regulations to help assure the safety and effectiveness of new drugs.  

   The agency does not agree with the comments' contention that the misbranding 
provisions of the act are irrelevant. Section 502(a) of the act prohibits false 
or misleading labeling of drugs, including (under section 201(n) of the act) 
failure to reveal material facts relating to potential consequences under 
customary conditions of use. Section 502(f) of the act requires drugs to have 
adequate directions for use and adequate warnings against unsafe use, such as 
methods of administration, that may be necessary to protect users. In  addition, 
section 502(j) of the act prohibits use of drugs that are dangerous to health 
when used in the manner suggested in their labeling. Each of these misbranding 
provisions is intended, at least in significant part, to protect consumers 
against the marketing of drugs that would not be safe under certain conditions 
of use. Section 701(a) of the act authorizes FDA to issue regulations for the 
efficient enforcement of the act. The restrictions on use contemplated by §§ 
314.520 and 601.42 help to ensure that products that would be misbranded under 
section 502 of the act are not marketed.  

   The restrictions on use imposed under section 503 of the act, which relate to 
prescription use limitations, primarily concern whether a drug is safe for use 
except under the supervision of a licensed practitioner. While the agency agrees 
that the restrictions imposed under §§ 314.520 and 601.42 concerning 
distribution to certain facilities or physicians with special training or 
experience would be in addition to ordinary prescription limitation, FDA 
believes these restrictions are consistent with the spirit of section 503 of the 
act, as well as the other provisions of the act referred to, in ensuring safe 
use.  

   New drugs may be approved under section 505(d) of the act only if they are 
safe for use under the conditions prescribed, recommended, or suggested in the 
proposed labeling. In addition, for approval, a drug's labeling must not be 
false or misleading based on a fair evaluation of all material facts, which 
would include details about the conditions of use. For biological products, 
section 351(d) of the PHS Act also authorizes the imposition of restrictions 
through regulations "designed to insure the continued safety, purity, and 
potency" of the products.  

   The agency disagrees with the comments' implication that the courts' rulings 
in American Pharmaceutical Association (APhA) v. Weinberger mean there is no 
statutory authority to impose restrictions on distribution for accelerated 
approval drugs. The situation considered in that case is readily distinguishable 
from the situation addressed in §§ 314.520 and 601.42 of the accelerated 
approval regulations. The APhA case concerned a regulation that withdrew 
approval of NDA's for methadone, but permitted distribution to certain 
maintenance treatment programs and certain hospital and community pharmacies. 
Because methadone is a controlled substance within the provisions of the 
Controlled Substances Act, which is implemented by the Drug Enforcement 
Administration with the Justice Department, the district court concluded that 
the question of permissible distribution of the drug was within the jurisdiction 
of the Justice Department, not FDA. The Court of Appeals determined that the 
type of misuse associated with methadone, i.e., misuse by persons who have no 
intent to try to use drugs for medical purposes, differed from safety issues 
contemplated for control under section 505 of the act. In contrast, the 
restrictions contemplated under §§ 314.520 and 601.42 are precisely those deemed 
necessary to ensure that section 505 criteria have been met, i.e., restrictions 
to ensure that the drug will be safe under its approved conditions of use. It is 
clearly FDA's responsibility to implement the statutory provisions regarding new 
drug approval.  

   Nor does FDA agree that the provisions placing restrictions on distribution 
to certain facilities or physicians, or conditioned on the performance of 
certain medical procedures, impermissibly interfere with the practice of 
medicine and pharmacy. There is no legal support for the theory that FDA may 
only approve sponsors' drugs without restriction because physicians or  
pharmacists may wish to prescribe or dispense drugs in a certain way. The 
restrictions under these provisions would be imposed on the sponsor only as 
necessary for safe use under the extraordinary circumstances of the particular 
drug and use. Without such restrictions, the drugs would not meet the statutory 
criteria, could not be approved for distribution, and would not be available for 
prescribing or dispensing. The agency, as a matter of longstanding policy, does 
not wish to interfere with the appropriate practice of medicine or pharmacy. In 
this instance, the agency believes that rather than interfering with physician 
or pharmacy practice, the regulations permit, in exceptional cases, approval of 
drugs with restrictions so that the drugs may be available for prescribing or 
dispensing.  

   21. One comment asserted that postmarketing restrictions on distribution to 
certain facilities or physicians with certain training or experience should be 
limited to rare occasions in cases of extreme hazard to patient safety in which 
toxicity of a particular drug may require it, but should not be applied because 
of insufficient efficacy data. Some comments argued that safety issues in the 
context of drug use should be addressed through patient management and effective 
product labeling, not through restricted distribution. In support of this 
argument, the comments cited the labeling of oncologic drugs, which provides 
physicians with adequate warnings and recommendations for their use without 
limiting distribution.  

   FDA agrees with these comments in part and intends to impose restrictions on 
distribution or use under this rule only in those rare instances in which the 
agency believes carefully worded labeling for a product granted accelerated 
approval will not assure the product's safe use. As stated in the preamble to 
the proposed rule (57 FR 13234 at 13237), FDA believes that the safe use of most 
prescription drugs will continue to be assured through traditional patient 
management by health professionals and through necessary safety warnings in the 
drug's labeling.  

   22. Two comments asked who will determine if restricted distribution should 
occur and what facilities or physicians with special training or experience will 
participate. Several comments expressed concern that restricted distribution 
and/or conditional use may not include all health care professionals who should 
participate in safe and effective patient care. Two organizations representing 
pharmacists asked that FDA develop functional and objective criteria that 
clearly establish the activities of pharmacists, physicians, and others in the 
care of patients receiving a drug under restricted distribution. The comments 
asserted that any health care professional that met these criteria should be 
allowed to participate in distribution of the drug and care of the patient. One 
comment recommended that any postmarketing restrictions on distribution or use 
of a drug approved under the accelerated approval process be developed by 
appropriate FDA advisory committees or panels expanded to include physicians and 
pharmacists with expertise in the therapeutic area being considered and in 
relevant drug distribution systems. Where appointment of pharmacists to these 
committees or panels is not feasible, the comment recommended that FDA use 
pharmacists in a consultant capacity. Another comment argued that current 
systems for drug distribution incorporate "checks and balances" such that 
prescribers and pharmacists work together to assure safe use of a drug by a 
patient. Two comments would oppose any restricted distribution system that 
allows manufacturers exclusively to deliver prescription drugs directly to 
patients. One comment asked whether FDA or the applicant would monitor the 
criteria for restricted distribution sites or physicians.  

   The medical reviewing divisions within FDA's CDER and CBER will determine if 
restricted distribution or use should be imposed. FDA will usually seek the 
advice of outside expert consultants or advisory committees before making this 
determination, and will, of course, consult with the applicant.  

   The agency does not agree that FDA should develop criteria that clearly 
establish the activities of health care professionals in the care of patients 
receiving a drug approved under this rule and for which restricted distribution 
has been imposed. Any postmarketing restrictions required under this rule will 
impose an obligation on the applicant to ensure that the drug or biological 
product is distributed only to the specified facilities or physicians. FDA will 
seek the advice of outside consultants with expertise in distribution systems or 
advisory committees when necessary in determining the need for or type of 
restricted distribution. The limitations on distribution or use imposed under 
this rule, including specific distribution systems to be used and the 
applicant's plan for monitoring compliance with the limitations, will have been 
agreed to by the applicant at the time of approval. The burden is on the 
applicant to ensure that the conditions of use under which the applicant's 
product was approved are being followed. As appropriate, FDA may monitor the 
sponsor's compliance with the specified terms of the approval and with the 
sponsor's obligations.  

   23. One comment recommended that proposed § 314.520 be modified to include 
therapeutic outcomes monitoring as a third example of a permissible 
postmarketing restriction. The comment defined therapeutic outcomes monitoring 
as the systematic and continual monitoring of the clinical and psychosocial 
effects of drug therapy on a patient which achieves the objective of preventing 
problems with drug therapy. Some comments argued that through therapeutic 
outcomes monitoring, a physician, a pharmacist, and a patient can work together 
to prevent problems with drug therapy by being constantly alert to signs of 
trouble. One comment said that indicator data can be routinely reported to a 
central collection point for utilization review by health care professionals, 
followed by educational programs to further improve the efficacy of drug 
therapy.  

   The postmarketing restrictions set forth in the proposal and in this final 
rule are intended to enhance the safety of a drug whose risks would outweigh its 
benefits in the absence of the restriction. Therapeutic outcomes monitoring does 
not contribute to that enhancement, and would not be required under this rule.  

   24. Some comments asked that FDA clarify how products will move from 
restrictive status to a regular prescription drug status. The comments asserted 
that all conditions associated with accelerated approval should automatically 
terminate following completion of confirmatory clinical trials; one comment 
urged FDA to explicitly state this in the final rule. One comment asserted that 
restrictions should automatically be removed 180 days after a supplemental 
application containing the data from the postmarketing study has been filed if 
FDA has not yet acted upon the supplemental application and the product should 
be deemed approved as if by "traditional" procedures and all other provisions of 
the act should apply, e.g., the applicant must have a formal hearing before 
removal of the product from the market.  

   FDA will notify the applicant when a particular restriction is no longer 
necessary for safe use of the product. In the case of drugs approved with a 
requirement for postapproval studies, FDA would expect that all of the  
postapproval requirements set forth in this rule, i.e., submission of 
promotional material and use of expedited withdrawal procedures, would no longer 
apply after postmarketing studies have verified and described the drug's 
clinical benefit. Concurrent with the review of the postmarketing studies, if 
requested, FDA will also review the need to continue any restrictions on 
distribution that have been imposed. In the case where restrictions on 
distribution or use have been imposed, such restrictions would be eliminated 
only if FDA determines that safe use of the product can be assured without them, 
through appropriate labeling. In some cases, however, that assurance could not 
be expected and the nature of the specific safety issue raised by the product 
might require continued restrictions. FDA has added new §§ 314.560 and 601.46 to 
state when postapproval requirements will no longer apply and state that the 
applicant may petition the agency, in accordance with 21 CFR 10.30, at any time 
to remove specific postapproval requirements.  

   With respect to the suggested time period for removing restrictions on 
distribution or use following submission of a supplemental application 
containing the data from a postmarketing study, FDA does not believe it should 
prescribe any specific time period. These applications will receive a priority 
rating and FDA is firmly committed to expedited review of an application 
considered for accelerated approval and all data submitted from a postmarketing 
study to verify clinical benefit and believes most reviews will be completed and 
action taken within 180 days.  

   25. One comment argued that, as proposed, it is not clear how accelerated 
approval would apply to drugs which fall under the conditions described in §§ 
314.520 and 601.42, which state the postmarketing restrictions on distribution 
or use that FDA may apply, because the language of these sections explicitly 
states that the sections apply to products "shown to be effective," which are 
already adequately covered by the act. To the comment, the language "shown to be 
effective" implies that full Phase 3 efficacy trials have been conducted, 
assessed, and deemed to demonstrate that the drug is effective for its proposed 
use. If the clinical data demonstrate that the product has an acceptable safety 
profile, the safe use of the drug should be addressed in the product labeling. 
Thus, the comment argued that §§ 314.520 and 601.42 should not be included in 
new subpart H of part 314 and subpart E of part 601, respectively, which deal 
with accelerated approval because these sections explicitly apply to products 
shown to be effective under a full drug development program.  

   Sections 314.520 and 601.42 apply not only to drugs and biological products 
approved on the basis of an effect on a surrogate endpoint but also to drugs and 
biological products that have been studied for their safety and effectiveness in 
treating serious or life-threatening illnesses using clinical endpoints and that 
have serious toxicity. In either case, if the products are so potentially 
harmful that their safe use cannot be assured through carefully worded labeling, 
FDA will approve the products for early marketing only if postmarketing 
restrictions on distribution or use are imposed. The phrase "shown to be 
effective" was not intended to distinguish drugs approved under new subpart H 
from drugs approved under any other subpart of the regulations. All drugs 
approved will have had effectiveness demonstrated on the basis of adequate and 
well-controlled studies, whether the endpoint of the studies is a surrogate 
endpoint or a clinical endpoint.  

   26. One comment expressed concern that the proposed restricted distribution 
or use provisions would restrict or eliminate the wholesale distribution of  
drugs approved through the accelerated approval process.  

   The limitations on distribution or use required under this rule are imposed 
on the applicant. Therefore, the burden is on the applicant to ensure that the 
conditions of use under which the applicant's product was approved are being 
followed. This rule does not specify how a manufacturer will distribute its 
product to those receiving the product under the approval terms. FDA will only 
determine which facilities or physicians may receive the drug, and the applicant 
will have agreed to this limitation on distribution or use.  

   27. One comment expressed concern that the proposed postmarketing restriction 
provision does not preclude a physician to whom restricted distribution applies 
from prescribing drugs approved under the accelerated approval process for 
unapproved (off-label) uses.  

   The comment is correct that this rule does not itself prevent a physician 
from prescribing a drug granted accelerated approval for an unapproved use. 
Under the act, a drug approved for marketing may be labeled, promoted, and 
advertised by the manufacturer only for those uses for which the drug's safety 
and effectiveness have been established and that FDA has approved. Physicians 
may choose to prescribe the drug for a condition not recommended in labeling. 
Such off-label use would, of course, be carried out under the restrictions 
imposed under this section. FDA also believes that physicians will be cognizant 
of the product's special risks and will use such drugs with particular care. The 
labeling of products approved under this rule will include all necessary 
warnings and full disclosure labeling would generally reflect the extent of 
clinical exposure to the drug.  

   
F. Postmarketing Studies  

   28. Three comments argued that FDA does not have the authority to require 
postmarketing studies to be performed as a condition of approval based on a 
"surrogate" endpoint. One comment stated that it is widely accepted that the act 
empowered the agency to define the type and extent of efficacy data necessary to 
approve a product application. If a surrogate marker can be shown to be 
sufficiently related to actual patient benefit, then, the comment asserted, data 
regarding the effect of a drug on a surrogate marker constitute acceptable proof 
of efficacy under the act. Two comments urged FDA to continue to ask applicants 
to agree voluntarily to perform postmarketing studies when medically warranted 
as is the current policy under the traditional approval process. One comment 
expressed concern that requiring postmarketing studies may become the norm 
rather than the exception.  

   The agency's response to comment 1. explained the circumstances in which FDA 
might conclude that a drug should be marketed on the basis of an effect on a 
surrogate endpoint reasonably likely to predict clinical benefit only if studies 
were carried out to confirm the presence of the likely benefit. As discussed in 
the preamble to the proposed rule (57 FR 13234 at 13236), FDA believes that it 
is authorized by law to require postmarketing studies for new drugs and 
biological products. Section 505(d) of the act provides for the approval of new 
drugs for marketing if they meet the safety and effectiveness criteria set forth 
in section 505(d) of the act and the implementing regulations (21 CFR part 314). 
As discussed in the proposed rule, to demonstrate effectiveness, the law 
requires evidence from adequate and well-controlled clinical studies on the  
basis of which qualified experts could fairly and responsibly conclude that the 
drug has the effect it is purported to have. Under section 505(e) of the act, 
approval of a new drug application is to be withdrawn if new information shows 
that the drug has not been demonstrated to be either safe or effective. Approval 
may also be withdrawn if new information shows that the drug's labeling is false 
or misleading.  

   Section 505(k) of the act authorizes the agency to promulgate regulations 
requiring applicants to make records and reports of data or other information 
that are necessary to enable the agency to determine whether there is reason to 
withdraw approval of an NDA. The agency believes that the referenced reports can 
include additional studies to evaluate the clinical effect of a drug approved on 
the basis of an effect on a surrogate endpoint. Section 701(a) of the act 
generally authorizes FDA to issue regulations for the "efficient enforcement" of 
the act.  

   With respect to biological products, section 351 of the PHS Act provides 
legal authority for the agency to require postmarketing studies for these 
products. Licenses for biological products are to be issued only upon a showing 
that they meet standards "designed to insure the continued safety, purity, and 
potency of such products" prescribed in regulations (42 U.S.C. 262(d)). The 
"potency" of a biological product includes its effectiveness (21 CFR 600.3(s)).  

   The agency notes that it has in the past required postmarketing studies as a 
prerequisite for approval for some drugs (see 37 FR 201, January 7, 1972; and 37 
FR 26790, December 15, 1972).  

   29. One comment recommended that FDA require that specific timelines for 
completion of the required postmarketing studies be included in the marketing 
application. The comment further suggested that, if the sponsor fails to meet 
its timelines, approval of its application be withdrawn, or in the event it is 
difficult to withdraw approval of drugs for serious or life-threatening 
diseases, FDA should establish substantial fines and penalties for sponsors that 
deliberately withhold information from FDA regarding the preliminary results and 
the progress of their postmarketing studies, or delay the completion of such 
studies. The comment also urged FDA to publish in the Federal Register  
identification of manufacturers who are not meeting their obligation to complete 
the required postmarketing studies on time. These recommendations were prompted 
by the comment's concern that once a manufacturer is granted approval for its 
product, the manufacturer will have little incentive to complete postmarketing 
studies in a timely manner, especially if the preliminary results of such 
studies indicate that the drug may not be safe and/or effective. Another comment 
urged FDA to include in the final rule language that requires the participation 
of pharmacists in postmarketing studies because pharmacists can serve as an 
additional source of information on therapeutic outcomes of patients taking 
drugs approved under this rule and monitoring for such drugs.  

   The agency expects that the requirement for postmarketing studies will 
usually be met by studies already underway at the time of approval and that 
there will be reasonable enthusiasm for resolving the questions posed by those 
studies. The plan for timely completion of the required postmarketing studies 
will be included in the applicant's marketing application. In addition, in 
accord with the annual reporting requirements at § 314.81(b)(2)(vii) (21 CFR 
314.81(b)(2)(vii), an NDA applicant is required to provide FDA with a statement 
of the current status of any postmarketing studies. FDA declines to impose the  
sanctions suggested by the comment for failure of an applicant to meet its plans 
for completion of a postmarketing study. FDA believes this rule applies 
appropriate regulatory sanctions. Under the proposed rule and this final rule, 
FDA may withdraw approval of an application if the applicant fails to perform 
the required postmarketing study with due diligence.  

   FDA believes that it is not within the scope of this rule to establish the 
role of pharmacists in postmarketing studies. That role should more properly be 
defined by the clinical investigator and each institution or facility at which a 
postmarketing study is conducted.  

   30. One comment asserted that the proposal sets forth an inherent 
contradiction between the way FDA evaluates the benefit and risk for drugs today 
and the way the proposal contemplates. The comment argued that now, if 
postmarketing data raise questions about the risk associated with a drug 
product, FDA considers that data along with the other data known about the 
product, and determines whether, based on the overall knowledge about the drug, 
there is a need to seek withdrawal of approval. Under this proposal, if the 
postmarketing study data raised questions about the risk of the product, FDA 
would seek withdrawal of approval, whether or not the new data really made a 
fundamental difference to what is known about the benefit and risk of the 
product.  

   FDA does not agree that the contradiction described by the comment exists. 
Under the circumstances of accelerated approval, approval would be based on a 
weighing of the benefit suggested by the effect on the surrogate endpoint 
against known and potential risks of the drug. Should well-designed postapproval 
studies fail to demonstrate the expected clinical benefit, the benefit expected 
at the time of approval (reasonably likely to exist) would no longer be expected 
and the totality of the data, showing no clinical benefit, would no longer 
support approval. This evaluation of the data is not different from 
considerations that would apply in evaluating data in the case of a drug 
approved under other provisions of the regulations.  

    31. Two comments expressed the view that the proposed requirement for 
postmarketing studies may raise important ethical questions because once a drug 
product is approved, it may be unethical, depending on the circumstances, for a 
physician to conduct a study using a placebo control. One comment also contended 
that a postmarketing study requirement could compromise the NDA holder's ability 
to enroll sufficient numbers of patients in the study when the new approved drug 
and possible alternative therapies are widely available to patients.  

   Usually, and preferably, because of problems suggested in the comment, the 
requirement for postmarketing studies will be met by studies already underway at 
the time of approval, e.g., by completion of studies that showed an effect on 
the surrogate. FDA recognizes that ethical considerations will play a central 
role in the type of study carried out, a choice that will depend upon the type 
and seriousness of the disease being treated, availability of alternative 
therapies, and the nature of the drug and the patient population. There often 
are alternatives to use of a placebo control, including active control designs 
and dose-response studies that can satisfy both the demands of ethics and 
adequacy of design.  

   32. One comment contended that the term "postmarketing study" is used 
inconsistently in the proposed rule. The comment argued that "postmarketing  
study" is an accepted regulatory term of art which, to this point, has referred 
to studies conducted to confirm safety (not efficacy), after an approval has 
been granted, whereas in this proposal, a "postmarketing study" refers to a 
study required to establish clinical efficacy (i.e., a Phase 3 study), but not 
necessarily safety, although safety data will be collected. To prevent confusion 
and to differentiate between these required postmarketing confirmatory efficacy 
studies and safety studies traditionally conducted after approval and to clarify 
that products granted accelerated approval have been approved on the basis of 
Phase 2 (surrogate endpoint) data, the comment suggested changing the term 
"postmarketing study" to "Phase 3 study" in this rule except where traditional 
postmarketing studies are intended. The comment also suggested that the term 
"Phase 3 study" be defined as a study required to confirm findings of efficacy 
based upon surrogate data collected in Phase 2, which will be conducted after an 
accelerated approval has been granted and will be required before restrictions 
set forth in § 314.520 are removed.  

   The agency does not believe that the comment has accurately described 
accepted meanings of various terms. The term postmarketing study does not refer 
to any particular kind of study, but to studies carried out after a drug is 
marketed, often as part of an agreement by a sponsor to do so. These have 
included pharmacokinetic, drug-drug interaction, and pediatric studies, studies 
of dose-response or of higher doses, and studies of new uses. The term is not 
limited to safety studies. Moreover, Phase 2 and 3 studies are not distinguished 
by the endpoints chosen. Phase 3 hypertension studies, for example, still 
measure blood pressure, not stroke rate. The agency believes that the use of the 
"postmarketing study" in the final rule is appropriate and consistent.  

   
G. Withdrawal of Approval  

   33. One comment supported the proposed withdrawal of approval procedure. 
Other comments asserted that the proposed procedure does not provide the 
applicant with the procedural safeguard of a formal evidentiary hearing 
guaranteed by section 505 of the act and the Administrative Procedure Act (APA). 
As an example, the comments said that based on a finding of a single study 
failing to show clinical benefit or misuse of any promotional material, an 
approved new drug would be subject to withdrawal from the market with only a 
minimal opportunity for the NDA holder to be heard. The comments argued that 
section 505(e) of the act guarantees applicants "due notice and opportunity for 
a hearing" on withdrawal of an NDA in compliance with APA hearing standards, 
thus FDA must conduct hearings on withdrawals of NDA's using the formal 
adjudicatory procedures of the APA. One comment asserted that, under the 
proposed procedure, there is the absence of a discernible legal standard, an 
inability to cross-examine, the prosecuting attorney and judge are one and the 
same person, and there is a lack of even minimal formal evidentiary procedures. 
The comment expressed doubt that the proposed procedure would be sufficient to 
create a record suitable for review by a Court of Appeals, which must be able, 
on the basis of such a record, to determine whether the approval is supported by 
"substantial evidence."  

   FDA believes the withdrawal procedures set forth in proposed §§ 314.530 and 
601.43 and in this final rule are consistent with relevant statutes and provide 
applicants adequate due process. As stated in the proposed rule, in issuing its 
general procedural regulations, FDA decided to afford NDA holders an opportunity 
for a formal evidentiary hearing even though the courts had not decided that  
such a hearing was necessarily legally required (see 40 FR 40682 at 40691, 
September 3, 1975). In promulgating its procedural regulations, FDA also 
determined that a formal evidentiary hearing is not required before withdrawing 
approval of biological products, but that it would be appropriate to apply the 
same procedures to biological products as to drug removal (see 40 FR 40682 at 
40691).  

   Through the hearing process in this final rule, as in the proposed rule, 
applicants will be afforded the opportunity to present any data and information 
they believe to be relevant to the continued marketing of their product. The 
proposed process also would have permitted the presiding officer, the advisory 
committee members, a representative of the applicant, and a representative of 
the Center that initiates the withdrawal proceedings to question any person 
during or at the conclusion of the person's presentation. As discussed below in 
response to a comment, FDA has decided to allow up to three representatives of 
the applicant and of the Center to question presenters. Participants could 
comment on or rebut information and views presented by others. As with ordinary 
21 CFR part 15 hearings, the hearing will be transcribed. Subsequent to the 
hearing, the Commissioner of Food and Drugs would render a final decision on the 
matter. The agency believes that the administrative record created through this 
process would be sufficient for judicial review.  

   The agency emphasizes that, as part of the approval process under this rule, 
applicants will have agreed that these withdrawal procedures apply to the drug 
for which they seek approval; applicants objecting to these procedures may 
forego approval under these regulations and seek approval under the traditional 
approval process. Under such circumstances, applicants would not have the 
benefit of accelerated approval; if the drug were subsequently approved, 
however, before withdrawal of the approval, the applicant would have an 
opportunity for a 21 CFR part 12 hearing.  

   34. One comment noted that the "imminent hazard" provision of section 505(e) 
of the act allows FDA to suspend approval of a product, immediately, if it is 
found to pose an imminent hazard to the public health. As an alternative to the 
proposed withdrawal procedure or in addition to the "imminent hazard" statutory 
provision, the comment suggested that, when confronted with a dangerous product 
on the market, FDA could request that the applicant voluntarily withdraw its 
product, and most applicants would comply if a legitimate hazard exists.  

   As noted in the proposed rule, FDA and applicants have often reached mutual 
agreement on the need to remove a drug from the market rapidly when significant 
safety problems have been discovered. However, applicants usually have been 
unwilling to enter into such agreements when doubts about effectiveness have 
arisen, such as following the review of effectiveness of pre-1962 approvals 
carried out under the Drug Efficacy Study Implementation (DESI) program. For 
drugs approved under the accelerated procedure regulations, the risk/benefit 
assessment is dependent upon the likelihood that the surrogate endpoint will 
correlate with clinical benefit or that postmarketing restrictions will enable 
safe use. If the effect on the surrogate does not translate into a clinical 
benefit, or if restrictions do not lead to safe use, the risk/benefit assessment 
for these drugs changes significantly. FDA believes that if that occurs, rapid 
withdrawal of approval as set forth in this rule is important to the public 
health.  

   35. Under the proposed withdrawal procedures, in addition to other persons, 
one representative of the Center that initiates the withdrawal proceedings may 
question participants at a withdrawal of approval hearing. One comment objected 
to limiting the Center to one representative because detailed knowledge about a 
drug product is likely to be available from several scientists.  

   The proposed limitation of questioning to single representatives of the 
initiating Center and the applicant was intended to make the proceedings 
manageable. On further consideration, the agency has determined that it would be 
appropriate and manageable to allow up to three persons to be designated as 
questioners for the applicant and for FDA. Sections 314.530(e)(2) and 
601.43(e)(2) have been revised accordingly.  

   36. Some comments questioned FDA's ability to withdraw approval under the 
proposed procedures efficiently or effectively because of: (1) The lack of 
assurance that the results of postmarketing studies will be promptly provided to 
FDA; (2) limited agency resources to review study results and act upon them 
promptly; (3) the difficulties associated with establishing that an approved 
drug is "ineffective;" and (4) political pressure not to rescind the approval of 
NDA's for drug products that may lack evidence of effectiveness, especially if 
no clearly effective alternative treatments are available. One comment offered 
the opinion that where a drug shows only modest evidence of benefit, perhaps on 
a surrogate endpoint, and only shows equivocal evidence of clinical efficacy in 
postmarketing studies it would be difficult and socially disruptive to withdraw 
approval and remove the drug from the market if the drug has become well 
established and accepted, and there is no issue of toxicity. Another comment 
believed it would be difficult to withdraw approval of a drug that may be 
beneficial in a subpopulation but which, in fact, has not been shown to be 
efficacious in broader patient population studies. The comments suggested the 
need for a lesser sanction.  

   Another comment suggested that expediting removal of a product from the 
market could be accomplished by using a procedure like the "imminent hazard" 
provision of the act, i.e., immediate removal of the drug from the market if any 
of the conditions listed in proposed § 314.530 were met followed by a hearing.  

   Although the potential difficulties cited by the comments are real, they are 
not fundamentally different from determinations FDA regularly must make in 
carrying out its responsibilities. The new regulations provide for an expedited 
procedure to withdraw approval; they do not guarantee that results of studies 
will be wholly unambiguous or that FDA will always be able to prevail in its 
view as to the need for withdrawal, any more than current withdrawal procedures 
do. The studies being carried out under these provisions will be conspicuous and 
important and their completion will be widely known. There is no reason to 
believe their results would or could be long hidden. A study that fails to show 
clinical effectiveness does not prove a drug has no clinical effect but it is a 
study that, under § 314.530, will lead to a withdrawal procedure because it has 
failed to show that the surrogate endpoint on which approval was based can be 
correlated with a favorable clinical effect. This may have occurred because the 
study was poorly designed or conducted; while FDA will make every effort to 
avoid this, the commercial sponsor has the responsibility for providing the 
needed evidence confirming clinical benefit. As previously discussed, §§ 314.510 
and 601.41 have been revised to clarify that required postmarketing studies must 
also be adequate and well-controlled. The possibility that an ineffective drug 
has become "accepted" is not a basis for continued marketing. FDA intends to  
implement the provisions of § 314.530 as appropriate; data that are ambiguous 
will inevitably lead to difficult judgments.  

   A drug with clear clinical effectiveness in a subset of the population, but 
not in the population described in labeling, would have its labeling revised to 
reflect the data. Withdrawal would be inappropriate under such circumstances.  

   If an imminent hazard to the public health exists, the Secretary of Health 
and Human Services may suspend approval of an application and then afford the 
applicant an opportunity for an expedited hearing. In the absence of a 
significant hazard requiring immediate withdrawal, FDA believes the expedited 
procedure described in the rule satisfies the need for prompt action while, at 
the same time, allowing opportunity for discussion and debate before withdrawal.  

   37. One comment noted that the proposed rule would allow FDA to withdraw 
approval for failure to perform the required postmarketing studies with due 
diligence. The comment asserted that the act does not permit FDA to withdraw 
approval on this ground. Another comment, however, suggested that because 
proposed §§ 314.530 and 601.43 cite grounds for withdrawal of approval that are 
not grounds under the act, the language of these proposed sections should be 
revised to use language that closer aligns to that used in the act, e.g., 
describe a "postmarketing study" in statutory language.  

   FDA reaffirms the position expressed in the preamble to the proposal (57 FR 
13234 at 13239) that there is adequate authority under the act to withdraw 
approval of an application for the reasons stated under proposed §§ 314.530 and 
601.43, which include failure of an applicant to perform the required 
postmarketing study with due diligence. Section 505(e) of the act authorizes the 
agency to withdraw approval of an NDA if new information shows that the drug has 
not been demonstrated to be either safe or effective. Approval may also be 
withdrawn if the applicant has failed to maintain required records or make 
required reports. In addition, approval may be withdrawn if new information, 
along with the information considered when the application was approved, shows 
the labeling to be false or misleading.  

   For biological products, section 351(d) of the PHS Act authorizes approval of 
license applications under standards designed to ensure continued safety, 
purity, and potency. "Potency" for biological products includes effectiveness 
(21 CFR 600.3(s)). The PHS Act does not specify license revocation procedures, 
except to state that licenses may be suspended and revoked "as prescribed by 
regulations."  

   For drugs approved under § 314.510, FDA will have determined that reports of 
postmarketing studies are critical to the risk/benefit balance needed for 
approval; if those reports are not forthcoming, then, under authority of section 
505(d) of the act, the drug cannot on an ongoing basis meet the standards of 
safety and efficacy required for marketing under the act. Therefore, it is 
important to ensure that the applicant make a good faith effort to complete any 
required postmarketing studies in a timely manner so that FDA can rapidly 
determine whether the surrogate endpoint upon which the drug was approved has 
been confirmed to correlate with clinical benefit. Failure to submit the study 
results in a timely fashion would also constitute failure to make a required 
report. Similarly, without submission of the information from required 
postmarketing studies on biological products approved under these procedures, 
the biological product is not assured of continued safety and effectiveness.  
The license application may, therefore, appropriately be revoked as described in 
§ 601.43.  

    FDA does not find the statements of the grounds for withdrawal of approval 
under §§ 314.530 and 601.43 of this rule inconsistent with statutory language or 
ambiguous. The agency notes that, in the event none of the grounds for 
withdrawal specifically listed in § 314.530 or § 601.43 applies, but another 
ground for withdrawal under section 505 of the act or section 351 of the PHS Act 
and implementing regulations at 21 CFR 314.150 or 601.5 does apply, the agency 
will proceed to withdraw approval under traditional procedures.  

   38. Two comments expressed concern that it may be difficult for the agency to 
enforce the requirement that postmarketing studies be pursued with due 
diligence. The comments asked what would happen if a sponsor using due diligence 
is unable to recruit enough patients, or if the sponsor questions the validity 
of the data from the required postmarketing study, and would clumsy data 
management be seen as sufficient reason to rescind approval for a marketed drug? 
Another comment stated that once a product is approved and, by definition, 
provides a "meaningful therapeutic benefit over existing therapies," study 
accrual may drop off dramatically as patients may refuse to receive the "old" 
therapy or placebo, or physicians may consider it unethical not to treat all 
patients with the approved indication with the new drug or biological product. 
Under these circumstances, the comment expressed the opinion that neither the 
sponsor nor the product should be penalized, nor should there be a threat to 
withdraw approval. Based on FDA's past history in postmarketing studies, which 
one comment characterized as resulting in poorly done studies, studies conducted 
much later than agreed upon, or not at all, the comment expressed the opinion 
that the "due diligence" with which applicants are expected to carry out 
postmarketing studies may be an overly great expectation. One comment asked FDA 
to give examples of when it may withdraw approval if "other evidence 
demonstrates that the drug product is not shown to be safe or effective under 
its conditions of use" (proposed §§ 314.530(a)(6) and 601.43(a)(6)).  

   FDA does not agree that it will be difficult to enforce the "due diligence" 
provision of this rule. The "due diligence" provision was designed to ensure 
that the applicant makes a good faith effort to conduct a required postmarketing 
study in a timely manner to confirm the predictive value of the surrogate marker 
or other indicator. Any requirement for postmarketing studies will have been 
agreed to by the applicant at the time of approval, and if the study is not 
conducted in a timely manner as agreed to by the applicant, approval of the 
applicant's application will be withdrawn. FDA will expect any required 
postmarketing study to be conducted in consultation with the agency. Therefore, 
should the applicant encounter problems with subject enrollment in a study or 
ethical difficulties about the type of study to conduct, FDA expects the 
applicant to discuss these problems with the agency and reach agreement on their 
resolution.  

   Examples of other evidence demonstrating the drug product is not shown to be 
safe and effective could include further studies of the effect of the drug and 
the surrogate endpoint that fail to show the effect seen in previous studies, 
new evidence casting doubt on the validity of the surrogate endpoint as a 
predictor of clinical benefit, or new evidence of significant toxicity.  

   39. Some comments objected to withdrawal of approval of a drug product 
approved under the accelerated approval process because of perceived  misconduct 
by the applicant, such as failure to perform a required postmarketing study with 
due diligence or use of promotional materials that are false or misleading. The 
comments argued that the primary purpose of the accelerated approval process is 
to provide improved treatments to desperately ill patients at the earliest 
possible time, and withdrawal of approval of the new treatments for reasons not 
directly related to safety or efficacy undermines the purpose of the proposed 
rule. Two comments suggested that correction of the promotional material without 
interruption of access to the drug would be a better approach. Another comment 
suggested that there may be circumstances where continued access to the drug, if 
accompanied by informed consent, would be appropriate even if substantial 
questions arise about a product's safety and effectiveness. One comment urged 
that anticipated withdrawal of approval be preceded by measures to ensure that 
patients and their physicians will have an uninterrupted supply until 
alternative treatment arrangements can be made.  

   The need for "due diligence" in conducting the agreed to postmarketing 
studies is discussed in paragraph 37. The reasons for concern about misleading 
promotional materials are discussed under paragraph 16. With respect to 
promotional materials, FDA expects that, in most cases, any disagreements 
between the applicant and FDA will be resolved through discussion and 
modification of the materials, so that the drug or biological product can 
continue to be marketed. If, however, FDA concludes that the promotional 
materials adversely affect the risk/benefit conclusion supporting the drug's 
marketing, the agency intends to minimize the risk to the public health by 
removing the product from the market through the withdrawal procedures in this 
rule.  

    40. One comment expressed concern that the proposed withdrawal procedure may 
give the appearance of bias or preconceived notions on the part of the agency 
because the final decision to withdraw approval of a drug would be made by the 
Commissioner of Food and Drugs and the intention to withdraw approval of the 
drug will already have been determined by the agency.  

   Under the withdrawal provisions of this rule, FDA's CDER or CBER, rather than 
the Commissioner, will initiate the withdrawal proceedings. The withdrawal 
process will begin with a letter from CDER or CBER notifying the applicant that 
the Center proposes to withdraw marketing approval and stating the reasons for 
the proposed action. Although separation of functions will not apply under the 
provisions of §§ 314.530 or 601.43, the Commissioner's decision regarding 
withdrawal would not occur until after the applicant had an opportunity for 
hearing as described in those sections. The Commissioner would then expect to 
review the issues with objectivity and fairness having had the benefit of the 
presentations and discussions at the hearing and of the advisory committee's 
recommendations.  

   
H. Safeguards for Patient Safety  

   41. One comment asked if drugs approved under the accelerated approval 
process will be held to the same standards concerning postmarketing safety as 
drugs approved by the traditional process.  

   As discussed in the preamble to the proposed rule, applicants gaining 
approval for new drugs through the accelerated approval procedures will also be 
expected to adhere to the agency's longstanding requirements for postmarketing  
recordkeeping and safety reporting (see 21 CFR 314.80 and 314.81). Information 
that comes to FDA from the applicant or elsewhere that raises potential safety 
concerns will be evaluated in the same manner that such information is evaluated 
for drugs approved under the agency's traditional procedures. If the 
postmarketing information shows that the risk/benefit assessment is no longer 
favorable, the agency will act accordingly to remove the drug from the market.  

   42. One comment urged FDA, if the proposed rule were adopted, to require 
written informed consent so that patients would know that the drugs with which 
they were being treated had risks and that the benefits had not been adequately 
established.  

   The agency does not agree that patients using drug products approved under 
the accelerated approval regulations should be asked to provide written informed 
consent. Drugs approved under these provisions are not considered experimental 
drugs for their approved uses. Like all approved drugs, drugs approved under 
these provisions will have both risks and benefits. As previously discussed in 
this preamble, for drugs approved based on studies showing an effect on a 
surrogate endpoint, the approved labeling will describe that effect. In 
addition, the labeling will contain information on known and potential safety 
hazards and precautionary information. As with all prescription drugs, the 
physician has the responsibility for appropriately advising the patient 
regarding the drug being prescribed.  

   43. One comment asked that FDA require manufacturers to maintain an updated 
list of names, addresses, and phone numbers of physicians prescribing their 
products approved under this rule, and in the case of recall or withdrawal of 
approval, require manufacturers to contact these physicians and encourage them 
to notify their patients.  

   FDA does not believe such a procedure is necessary. Furthermore, maintaining 
such a registry for drugs prescribed through pharmacies would be very difficult. 
Agency experience with recalls and product withdrawals indicates that the 
methods of notification that have been developed for such circumstances are 
adequate.  

   44. One comment recommended that FDA require patient package inserts (PPI's) 
for all drugs granted accelerated approval that would state the specific 
restrictions placed on a drug product and/or the reason for requiring 
postmarketing studies. In addition, the comment recommended that FDA require the 
manufacturer to include an adverse drug reaction "hotline" phone number in the 
PPI along with an FDA phone number. The PPI should inform the patient to report 
immediately any adverse drug reaction experienced to his or her doctor, the 
manufacturer, and FDA, and the manufacturer should be required to contact FDA 
immediately after receiving a report of a serious adverse reaction.  

   FDA concludes that patient package inserts are not routinely needed for drugs 
granted accelerated approval, although if circumstances made one appropriate, 
one would be developed for a particular drug. As with any prescription drug, the 
approved labeling for a product granted accelerated approval will contain 
information about the safe and effective use of the product, including all 
necessary warnings and the extent of clinical exposure. In addition, the 
conditions of use will be carefully worded to reflect the nature of the data 
supporting the product's approval. Physicians have the responsibility to inform 
patients about the safe and effective use of an approved product. Labeling  
includes suggestions to the physician concerning information to be provided to 
patients.  

   The agency notes that in this final rule limited editorial changes have been 
made to the wording of the proposed rule. The agency has determined that these 
changes do not affect the intent of the proposed rule.  

   
V. Economic Impact  

   In accordance with Executive Order 12291, FDA has carefully analyzed the 
economic effects of this final rule and has determined that it is not a major 
rule as defined by the Order. Indeed, because firms will not be forced to use 
the accelerated approval mechanism, applicants will most probably choose to take 
advantage of the program only where its use is expected to reduce net costs, 
Similarly, the final rule does not impose a significant economic impact on a 
substantial number of small entities so as to require a regulatory flexibility 
analysis under the requirements of the Regulatory Flexibility Act of 1980.  

   
VI. Environmental Impact  

   The agency has determined under 21 CFR 25.24(a)(8) that this action is of a 
type that does not individually or cumulatively have a significant effect on the 
human environment. Therefore, neither an environmental assessment nor an 
environmental impact statement is required.  

   
VII. Paperwork Reduction Act of 1980  

   This rule does not contain new collection of information requirements. 
Section 314.540 does refer to regulations that contain collection of information 
requirements that were previously submitted for review to the Director of the 
Office of Management and Budget (OMB) under section 3504 of the Paperwork 
Reduction Act of 1980 (Adverse Drug Experience Reporting, OMB No. 0190-0230).  

   
List of Subjects  

   
21 CFR Part 314  

   Administrative practice and procedure, Confidential business information, 
Drugs, Reporting and recordkeeping requirements.  

   
21 CFR Part 601  

   Biologics, Confidential business information.  

   Therefore, under the Federal Food, Drug, and Cosmetic Act, the Public Health 
Service Act, and under authority delegated to the Commissioner of Food and 
Drugs, 21 CFR parts 314 and 601 are amended as follows:  

   PART 314 -- APPLICATIONS FOR FDA APPROVAL TO MARKET A NEW DRUG OR AN 
ANTIBIOTIC DRUG  

   1. The authority citation for 21 CFR part 314 continues to read as follows:  

   Authority: Secs. 201, 301, 501, 502, 503, 505, 506, 507, 701, 706 of the 
Federal Food, Drug, and Cosmetic Act (21 U.S.C. 321, 331, 351, 352, 353, 355, 
356, 357, 371, 376).  

   2. Subpart H consisting of §§ 314.500 through 314.560 is added to read as 
follows:  

   Subpart H -- Accelerated Approval of New Drugs for Serious or Life-
Threatening Illnesses  

   Sec.  

   
314.500 Scope.  

   
314.510 Approval based on a surrogate endpoint or on an effect on a clinical 
endpoint other than survival or irreversible morbidity.  

   
314.520 Approval with restrictions to assure safe use.  

   
314.530 Withdrawal procedures.  

   
314.540 Postmarketing safety reporting.  

   
314.550 Promotional materials.  

   
314.560 Termination of requirements  

   Subpart H -- Accelerated Approval of New Drugs for Serious or Life-
Threatening Illnesses  

   § 314.500 Scope.  

   This subpart applies to certain new drug and antibiotic products that have 
been studied for their safety and effectiveness in treating serious or life-
threatening illnesses and that provide meaningful therapeutic benefit to 
patients over existing treatments (e.g., ability to treat patients unresponsive 
to, or intolerant of, available therapy, or improved patient response over 
available therapy).  

   § 314.510 Approval based on a surrogate endpoint or on an effect on a 
clinical endpoint other than survival or irreversible morbidity.  

   FDA may grant marketing approval for a new drug product on the basis of 
adequate and well-controlled clinical trials establishing that the drug product 
has an effect on a surrogate endpoint that is reasonably likely, based on 
epidemiologic, therapeutic, pathophysiologic, or other evidence, to predict 
clinical benefit or on the basis of an effect on a clinical endpoint other than 
survival or irreversible morbidity. Approval under this section will be subject 
to the requirement that the applicant study the drug further, to verify and 
describe its clinical benefit, where there is uncertainty as to the relation of 
the surrogate endpoint to clinical benefit, or of the observed clinical benefit 
to ultimate outcome. Postmarketing studies would usually be studies already 
underway. When required to be conducted, such studies must also be adequate and 
well-controlled. The applicant shall carry out any such studies with due 
diligence.  

   § 314.520 Approval with restrictions to assure safe use.  

   (a) If FDA