Petitioners challenge an order of the Commissioner of the Food and Drug Administration (FDA), entered April 27, 1973, which withdrew approval of their New Animal Drug Applications (NADA's) for pellets of diethylstilbestrol (DES). They also challenge the Commissioenr's order of May 3, 1973, issued ancillary to the withdrawal of the NADA's, and revoking certain use regulations. The Commissioner acted in the exercise of the authority conferred on the Secretary of Health, Education and Welfare by the Food, Drug and Cosmetic Act of 1938 (the Act) and delegated by the Secretary to the Commissioner.

We set aside the Commissioner's orders and remand the case to the FDA for further proceedings.

Pursuant to FDA approval of their NADA's, petitioners began manufacturing and selling DES in the 1950's.1 DES is a synthetic estrogen which is fed to animals to increase their feeding efficiency; when given DES, animals gain weight more quickly with less feed. The drug is available in two dosage forms, a feed premix which the animals eat, and pellets which are implainted in the animal's ear.2 Petitioners manufacture DES pellets.

DES has been recognized as a carcinogen for some years.3 As such the Delaney Clause4 would ordinarily have kept it from the market; the Clause automatically prohibits FDA approval of any drug that has been shown to be carcinogenic. But DES apparently passes out of the system of an animal in a very short time. An exception to the Delaney Clause is applicable if

the Secretary finds that, under the conditions of use specified in proposed labeling and reasonably certain to be followed in practice (i) such drug will not adversely affect the animals for which it is intended, and (ii) no residue of such drug will be found (by methods of examination prescribed or approved by the Secretary by regulations . . .) in any edible portion of such animals after slaughter or in any food yielded by or derived from the living animals . . .5

This clause, carving an exception out of the Delaney Clause, was enacted in 1962 with the DES situation in mind, and it is apparently sometimes referred to as the 'DES clause.'

Since 1962 the FDA has examined carcasses of slaughtered animals using its approved method of examination, the 'mouse-uterine' test. To date this testing has revealed no DES residues in edible tissues under conditions of use of 10 milligrams of DES per head per day, combined with a minimum 48 hour withdrawal prior to slaughter. FDA accordingly continued approval of all NADA's for DES until 1972.

During 1971, however, the United States Department of Agriculture (USDA) began to test carcasses using a new method, Gas-Liquid Chromatography (GLC). The FDA represents that this method is quicker than the 'mouseuterine' test, and apparently more sensitive.6 Using the GLC test, the USDA detected small amounts of residue in beef livers, a fact which it reported to the FDA.

Acting on this information the FDA, on June 21, 1972, issued a Notice which was in form a notice to withdraw approval of NADA's for DES, but in substance a notice triggering a public hearing to develop pertinent information.7 The Notice cited increased public concern about the use of DES, and stated that

The Commissioner has concluded that the most appropriate forum for public consideration of this matter is a public hearing to develop on the public record the information necessary for a conclusion as to the proper handling of this matter . . .. Accordingly, the Commissioner has concluded that it would be appropriate to propose withdrawal of the approval of the new animal drug applications for (DES) in order to utilize the hearing mechanism provided in the statute for this purpose.8

The agency stated that it had not yet decided to withdraw approval for DES, and that it contemplated consideration of proposals that would provide both a check on misuse and more effective restrictions on use of DES to reduce illegal residues. It requested that the manufacturers' responses suggest protocols that would check such misuse and establish more effective restrictions on use of DES.9 The Notice then listed several types of information that the FDA wished to obtain from interested persons in the event a hearing were held.10

In accordance with agency regulations,11 the Notice required interested parties requesting a pre-withdrawal hearing to file responses within thirty days, requesting such hearing and giving reasons why the NADA's should not be withdrawn. The petitioners responded within the thirty-day period.

Thereafter, on August 4, 1972, the FDA issued an order denying a hearing to the manufacturers of DES feed premixes, and withdrawing FDA approval of NADA's for premixes.12 The order specified that the withdrawal of approval resulted from the failure of the manufacturers to come forward with practicable guidelines for curtailing abuses in the administration of DES by premix.13 But finding that the matter 'is a regulatory, not a public health problem,' with discontinuance of DES premix resulting from the inability to develop controls on the use of DES in animal feed that are reasonably certain to be followed in practice, and 'because there is no evidence of a public health hazard,' the FDA allowed the manufacturers to continue to sell their stocks of DES until December, 1972, and permitted the continued slaughter and marketing of cattle that had been fed DES.14

In denying a hearing, the Commissioner stated that the objections which he had received from premix manufacturers had failed to 'set forth specific facts showing that there is a genuine and substantial issue of fact that requires a hearing.'15 The FDA cited the USDA's particularized test results, which it found to be uncontroverted by any specific data in the responses.

With regard to implants, the FDA stated that the latest USDA study did not include implants, that earlier study had failed to detect any implant-related residues, and that therefore the implants could be continued on the market pending further study by the USDA.16 The order particularly recognized that the implants normally given to a cow during feeding placed into the cow's system one-thirtieth of the amount of DES which feed premix would place. The order advised that testing of implants had recently begun, utilizing the most recently evolved technique, radioactive tracer studies, with results expected within several weeks. For that reason, the Commissioner stated:

The Commissioner has therefore concluded that it is premature to rule at this time on the objections and requests for a hearing with respect to DES implants. A ruling on this matter will await the results of the USDA study now underway.

At the present time, the Commissioner has no reason to believe that DES implants raise a public health hazard. Thus, while it is prudent to pursue and to resolve existing scientific questions about DES implants, it is unnecessary to remove existing implants or to be concerned about the safety of meat from animals implanted with DES.17

Thus the FDA stated its intent 'to pursue and resolve existing scientific questions about DES implants' through 'the USDA study underway.' It invited no further submissions from implant manufacturers.

Almost nine months passed. Then on April 27, 1973, the FDA issued an order withdrawing its approval of the NADA's for DES pellets, and denying a hearing to the petitioners.18 The order contained these findings:

The data obtained from the USDA radioactive-tagged DES implant study clearly establish the presence of residues of DES and/or its conjugates 120 days after implantation. No known data utilizing comparably sensitive methodology shows the absence of DES residues. The Commissioner and USDA are unaware of any valid basis for rejecting the results of this study as erroneous. The data now available in the NADA's demonstrate no residues only when tested by the mouse uterine and GLC methods, neither of which is sufficiently sensitive to permit a finding that no residue of DES exists 120 days after implantation.19

The order stated that the FDA had notified the petitioner of its intention to withdraw approval on June 21, 1972, that the petitioners' response to that notice had not raised any issues that required a hearing, and that the results of the latest USDA study made it impossible to allow the marketing of DES Pellets under any controls. Once again, the FDA stressed that the withdrawal was a regulatory matter and that:

. . . The Commissioner has no reason to believe that the use of DES implants has presented a public health hazard . . .20

Accordingly, the order allowed the continued marketing of cattle which had already received DES by implant.21

Both petitioners challenge the validity of the April 27, 1973, Order on the ground, inter alia, that the Commissioner erred in failing to grant a pre-with-drawal hearing. Because we agree with this contention, we do not reach their other assertions.22

II. INADEQUACY OF FDA's NOTICE AS A BASIS FOR SUMMARY DISPOSITION WITHOUT OPPORTUNITY FOR A HEARING

The Commissioner bases his withdrawal of approval of the NADA for DES pellets on 21 U.S.C. 360b(e)(1)(B). The Act provides that the Secretary may issue such an order only after 'due notice and opportunity for hearing.'23 The only bases for permitting withdrawal of approval for an NADA without a prior hearing are (1) the statutory proviso relating to emergency action in the face of an imminent hazard to health, and (2) the provisions in the regulations for 'summary judgment.' We discuss these in turn.

The Act provides that if the Secretary determines that 'there is an imminent hazard to the health of man or of the animals for which such drug is intended, he may suspend the approval of such application immediately . . . and afford the applicant the opportunity for an expedited hearing under this subsection . . .'24

In the instant case, there is no determination by the Secretary that there was an imminent hazard from DES implants to the health of man or animals. We pause to take note of the provision of the Act that such a determination must be made by the Secretary, and cannot be delegated to the Commissioner. This fortifies the Congressional presumption that a hearing is ordinarily required. It cannot be dispensed with on emergency health grounds by the FDA acting alone.

However, the present case does not even present us with the Commissioner finding a health hazard and the Secretary declining to make such a determination. Here the Commissioner himself expressly stated in his order that he had 'no reason to believe that the use of DES presented a public health hazard.'25

The FDA's regulations establishing a 'summary judgment' procedure provide:

When it clearly appears from the data in the application and from the reasons and factual analysis in the request for the hearing that there is no genuine and substantial issue of fact which precludes . . . the withdrawal of approval of the application . . . the Commissioner will enter an order on this data, making findings and conclusions on such data.26 There is no doubt of the general validity of the agency's 'summary judgment' regulations and procedure. But they must be applied consistently with basic fairness, and with the statutory requirement (see note 23) that the order may be issued only after 'due notice and opportunity for hearing.' Thus, the issue is whether the Commissioner gave adequate notice to the petitioners and then received responses insufficient to raise material issues of fact.

The agency contends that the June, 1972, Notice satisfied the statutory requirement, especially when read with the specific advisory to implant manufacturers in the August, 1972, announcement. The agency asserts that in June and in August, 1972, it alerted petitioners to the new tests being used. Moreover, the FDA contends that the June Notice indicated to petitioners that the agency intended to withdraw its approval of DES implants should the testing reveal DES residues attributable to implants. The FDA concludes from this that the petitioners had adequate notice of their 'theory' of withdrawal, and that this notice satisfied the statutory requirement.

In response, petitioners contend that adequate notice within the statute means that the FDA should have provided them with the test results, and allowed them a chance to respond to those results. We agree with petitioners.

3. General requirements for notice intended to initiate an agency summary disposition.

An agency may not validly take action against an individual without a hearing unless its notice to the individual of the adverse action proposed to be taken against him specifies the nature of the facts and evidence on which the agency proposes to take action.27 Such notice enables the affected party to prepare an informed response which places all the relevant data before the agency.

This principle has vitality in a case involving the FDA's powers of summary judgment. When the FDA issues a Notice of Opportunity for Hearing, its summary judgment procedures are available if the requesting party fails to raise material issues of fact. For that reason, the contents of the response are of critical importance, and the need for and importance of the response in turn enhances the significance of the notice given the adverse party. In order to be adequate, such notice given by the agency to an adverse party must contain enough information to provide the respondent a genuine opportunity to identify material issues of fact. This is needful to provide the 'due notice and opportunity for hearing' required by the Act.

The requirements of due notice and opportunity for hearing were explored, though not in terms, in USV Pharmaceutical Corporation v. Secretary of HEW.28 We examined the FDA's responsibilities when it seeks to invoke its summary judgment procedure. The FDA had undertaken to determine whether the petitioner's approved drug was efficacious. This determination was part of a larger undertaking to review all approved drugs, an undertaking initiated by 1962 amendments to the Food and Drug Act of 1938.29 On the basis of review by the National Academy of Sciences-National Research Council (NAS-NRC), and of his own findings, the Commissioner decided that USV had failed to prove efficacy by 'substantial evidence.' Despite additional submissions from USV, the FDA withdrew approval of USV's NDA, simultaneously denying petitioners a hearing.

We vacated the FDA's order of withdrawal because we found that the Commissioner had failed to make the necessary prima facie case for denial of a hearing. We found the FDA's notice to USV to have been 'cryptic and conclusory, without any statement of supporting facts.'30 As we stated:

At no time did the Commissioner set forth the facts and reasons upon which he relied in reaching his conclusion that no material issue of fact existed. We think that such an application of the Commissioner's summary judgment rule is not in harmony with the principle of the rule and is fundamentally unsound and unfair. Before calling upon the petitioner to answer, the Commissioner, as the moving party, had an obligation to present at least a prima facie case for denial of a hearing. Only after such a presentation could the Commissioner fairly impose upon the petitioner the heavy sanctions of 21 CFR 130.14.31

Because of the context, we did not explicitly reach in USV the relationship between 'adequate notice' and the Commissioner's pre-summary judgment burden. But the relationship is evident. Although USV did not specifically discuss the legal issue in terms of 'due notice and opportunity for hearing,' that concept is fairly implicit.

A notice that may be 'adequate' for the purpose of scheduling a hearing is not necessarily adequate for the purpose of beginning a summary judgment procedure. The difference is well known to the law. While a broad complaint may be legally adequate for the purpose of initiating a lawsuit and trial, the Federal Rules of Civil Procedure do not permit a summary judgment procedure to be used unless a motion is made which specifically sets forth the uncontested facts that warrant summary disposition. See Rule 56.

Of course, administrative agencies are not bound by the same details of procedure as the courts. But the agencies are governed by the same basic requirements of fairness and notice, and these include specificity of notice and opportunity to respond if what is instituted is intended to be a procedure for summary disposition without hearing. If the Commissioner of FDA is relying on his Notice as a device for invoking a summary judgment procedure that avoids the statute's general requirement of a hearing, he must include in such notice references to the 'facts' that he deems to be established in order that there may be meaningful opportunity to controvert the alleged facts and present a material issue for hearing. This includes, at a minimum, presentation of the prima facie case required in USV as a predicate for withholding the hearing required in general for revocation of an approved application.

Each time the Commissioner offers an NADA holder an opportunity for a hearing prior to withdrawing approval of the NADA, he begins a procedure that may, in the absence of a controverting response, result in a summary judgment. If the FDA wishes to retain a procedure in which the Notice may serve as the basis for a summary judgment (in the absence of controverting material) it is critical that with or in the Notice to the adverse party the Commissioner must come forward and present to the NADA holder the information which leads him to conclude, of course tentatively and subject to dispute and hearing, that approval of the NADA should be withdraw.

Our reference to the USV decision is intended to relate to the general principles of notice and summary disposition. The validity of those general principles is confirmed by Weinberger v. Hynson, Wescott & Dunning, 412 U.S. 609, 93 S. Ct. 2469, 37 L. Ed. 2d 207 (1973), although it may be that in some particulars the application of USV must be refined in the light of Hynson. In Hynson the Supreme Court approved the FDA's summary judgment procedure permitting withdrawal of an NDA without a hearing if the manufacturer failed to produce 'substantial evidence' of efficacy. After noting that the statute defined 'substantial evidence' to include 'evidence consisting of adequate and well-controlled investigations,' and that the FDA regulations detailed the 'principles . . . recognized by the scientific community as the essentials of adequate and well-controlled clinical investigations,' the Court upheld, 'as to those regulations that are precise,' an FDA denial of a hearing upon a finding that the study adduced by the manufacture had deficiencies that 'conclusively render the study inadequate or uncontrolled in light of the pertinent regulations.' (412 U.S. 622, 93 S. Ct. 2479). The Court held by a divided vote that the submission of the manufacturer in that case, though rejected by the FDA, 'was sufficient to warrant a hearing.' (412 U.S. at 623, 93 S. Ct. at 2480).

Hynson in effect reaffirms the propriety of administrative summary judgment, if taken in a context where the pleadings on their face 'conclusively' show that the hearing can serve no useful purpose. It did not overturn USV's requirement that the agency make some showing as a predicate for summary adjudication. It rather found that such a showing and predicate was supplied by particularized regulations setting forth precisely what the manufacturer was required to supply and by findings that he study adduced was conclusively deficient.

A further caution. The USV and Hynson cases both involved efficacy. The present case involves safety. We wish to make it clear that we are in no way suggesting that the FDA's course must or should be the same regardless of whether the ultimate issue is efficacy or safety. These different ultimate issues are governed by different sections of the law, different legislative histories and different regulations. But the ultimate principle is the same for both issues-- that where the case is governed by a statutory requirement for hearing (there being no imminent hazard to health), that hearing is not to be denied in the absence of a fair opportunity to identify material issues that require a hearing, an opportunity that embraces a suitable notice of the basis on which the agency proposes to act summarily.

4. Inadequacy of FDA notice(s) to petitioners as foundation for summary disposition.

Examining the Notice published on June 21, 1972, in the light of the foregoing principles, we find it inadequate as a foundation for summary disposition because it failed to establish a prima facie case for withdrawal without hearing. In that Notice, the FDA expressed its generalized concern about the use of DES in cattle, particularly about abuses in the premix type of administration. Moreover, while the FDA alluded in the June, 1972 Notice to a new type of test being used to detect residues, no particular test results were given.32 Rather, the record indicates that the testing program which led to FDA's removal of pellets from the market-- using a third method, radioactive tracing-- did not begin until October, 1972, and was not completed until April, 1973, with the Commissioner receiving the results on April 16, 1973.33

The June, 1972, Notice did not and could not contain any of the data on which the FDA relied in withdrawing its approval of the petitioner's NADA's. Other difficulties with the Notice, if treateld as a serious statement of present intention to withdraw approval of petitioner's NADA's, include: (1) that it took no particular account of the difference between normal label-specified dosage of DES, whether by premix or by implant, and excessive or abusive dosage; (2) that it failed to specify the nature of the residues detected, but implied only that they consisted of DES; and (3) that it referred to no specific testing program, but only to the use of new test methods in normal surveillance of commercially slaughtered animals.

Indeed, while this point is not essential to our ruling, we think it fair to observe that the June, 1972, Notice, looking to its life and without over-dwelling on its letter, does not communicate that the FDA was contemplating withdrawal, and certainly does not indicate that FDA was contemplating withdrawal without a hearing. On the contrary, the Notice specifically states that withdrawal was being noticed 'in order to use the hearing mechanism provided in the statute,' and to project a broad hearing in which various interested individuals and organizations would participate in addition The Commissioner set forth the subjects on which he wished data at the hearing (see note 10), including 'the accuracy and reliability of present detection methods for diethylstilbestrol, the potential availability of more sensitive detection methods for diethylstilbestrol and the likely result of their use.' Manifestly, this Notice could not, fairly construed, begin to serve as an alert that manufacturers were required to analyze characteristics of a potential, as yet undisclosed, testing method, under pain of forfeiting their statutory opportunity for hearing.

The Commissioner's brief tries to offset deficiencies in the June, 1972, Notice by contending that the August, 1972, Order that withdrew approval of the premixes should have alerted suppliers of pellets, like petitioners, to the test method involved, and to the possibility of actual withdrawal. This argument cannot stand. What we have said about the June, 1972, Notice is fully applicable to the August, 1972, Order if treated as a constructive notice: at this point the USDA had not begun the pellet tests using the radioactive tracing method ultimately relied upon, testing whose results did not emerge until April, 1973. Thus, even a back-dated synthetic 'combination' notice derived from the June and August publications would fail to contain the information necessary to meet the requirement of due notice that we have set forth above in section 3. Nor did the August, 1972, Order abandon or modify the character of the June Notice as the precursor of a hearing.

The FDA also argues that petitioners had notice of the nature of the tests as early as February, 1972. It does appear that one of the petitioners, Hess & Clark, manufactured special implants forl use in the testing. But this is not conclusive, even if we assume that both petitioners knew tests were underway, knew the nature of the tests, and indeed supplied some testing materials. The critical facts were the exisstence of residues and the nature of those residues. The tests have importance only in relationship to those facts. So until petitioners had notice of the test results, and of the FDA interpretation of those results, they cannot be said to have had due notice and opportunity for hearing.

5. FDA's claim of disclosure of 'theory of withdrawal.'

The Commissioner also argues that petitioners knew his theory of withdrawal in June, 1972, and that this knowledge gave petitioners adequate notice. The Commissioner cites Rodale Press,34 but that precedent is inappropriate. Rodale involved a Federal Trade Commission (FTC) attempt to limit certain advertising practices. In its notice to Rodale, the FTC stated one theory supporting the illegality of the challenged practices. In taking final action the agency employed a different theory. This court reversed the FTC, holding that the agency changing its theories in midstream was required to give reasonable notice of the change in order to avoid a deprivation of required notice and opportunity for hearing.

The Commissioner argues that because he retained the same 'theory,' Rodale supports his case for 'notice.' Of course, that the present case did not involve exactly the same vice as Rodale, if true, hardly establishes that it is without blemish. But there are manifest differences between the FTC action in Rodale and the FDA action in the case at bar. First, in the FTC context, the theory of illegality is the basis for adverse action. The advertisement exists on the public record. The FTC's characterization of that advertisement, i.e. its presentation of its theory, is therefore an adequate presentation of the prima facie case for adverse action, and for triggering the advertiser's response. The FDA context is very different. A theory that the presence of DES residues will lead to withdrawl of approval is not a prima facie case for such withdrawal. Rather, the agency may withdraw approval only when it actually finds such residues. The date about residues comprise the substance of the prima facie case. In short, the Commissioner's citation of Rodale adds nothing to his argument about the June, 1972 Notice.

6. FDA's change of theory in this case Moreover, the Commissioner, like the FTC in Rodale, did change his theory. The June, 1972, Notice indicated that the Commissioner planned to withdraw his approval of the NADA's pursuant to 21 U.S.C. 306b(e).35 This section admits of two theories, as discussed in Part III.36

However, in the August, 1972, Order to the premix suppliers-- which FDA now wishes to be reconstituted as a 'notice' to the implant suppliers-- the Commissioner stated that his action withdrawing approval of the premix NADA's was 'required under the strict terms of section 512(d)(1)(H) and 512(e) (1)(B) of the Act.'37 (Codified as 21 U.S.C. 360b(d)(1)(H) and (e)(1)(B)).

Section 512(d)(1)(H) is the Delaney Clause, and the statute is written to incorporate the Delaney Clause into the general section for withdrawal of approval, section 512(e)(1)(B). Hence the Commissioner's statement, fairly read, identifies a reliance on the Delaney Clause theory.38

Thereafter, the Commissioner affirmed his reliance on the Delaney Clause theory in the April 27, 1973, Order:

. . . The NADA's lack data on the basis of which it can be concluded that section 512(d)(1)(H) (The Delaney Clause) does not apply.39

In this court, however, counsel for the Commissioner argue that the Commissioner's action was never based on the Delaney Clause (Resp.Br. at 10):

. . . (The statute) requires withdrawal of a drug if it is no longer shown to be safe or if 360b(d)(1)(H) applies. 21 U.S.C. 360b(e)(1)(B).

. . . This Order (of April 27, 1973) was based upon the first clause . . .

As a matter of legal tactics, counsel representing FDA may have been well-advised to forsake the Delaney Clause contention, as appears from the discussion in Part III. But it is by now elemental doctrine that an agency's action must be defended on its own findings and determinations, and not on subsequent legal arguments of counsel.40 In short, we do not see Rodale as supporting FDA, but rather have the over-all perception that this is, like Rodale, a case of changes in theory.

7. Concluding analysis of inadequacy of notice as basis for summary disposition

And so it appears that by April, 1973, in deciding whether to accord the petitioners a hearing, the FDA had before it the substantial data from the USDA testing program, the information contained in the NADA's that the petitioners had filed and that the FDA had approved in the 1950's, and the responses which the petitioners had made to the June 21, 1972, Notice. At that time, the specific information received from the USDA should have been transmitted by FDA to the petitioners with provision of opportunity to answer. Without such a submission of the basis on which he proposed to act, in a statement transmitted to petitioners prior to the cut-off date for a response, the Commissioner's use of summary judgment was invalid.

Instead, on April 27, 1973, the FDA Commissioner at one and the same time disclosed the radioactive study of the Department of Agriculture, and announced that because its results had not been suitably contested in responses previously filed there was no need for a hearing. This is palpably impermissible procedure. The Commissioner must be held not to have given the petitioners a proper opportunity for a hearing.

The Commissioner's decision that a hearing was unnecessary was improper, based as it was upon information in the Commissioner's hands to which the adversely affected parties had not been given opportunity to respond. This procedural failure invalidates the Commissioner's order withdrawing approval of the NADA's for DES pellets.

III. CONSIDERATION OF DEVELOPMENTS SUBSEQUENT TO THE FDA's WITHDRAWAL ORDER

Even if the April 27, 1973, Order is procedurally defective, says the FDA, it should not be reversed. FDA argues: (1) there is no need to do an unnecessary act;41 (2) petitioners cannot raise any material issues of fact; and (3) therefore no hearing would have followed proper notice to petitioners.

The agency's request, in essence, is that we examine the April 27, 1973, Order as if that were the agency's prima facie case. The difficulty with this request is manifest. The purpose of requiring the FDA to present a prima facie case is to allow adverse parties a genuine opportunity to respond. By acting to withdraw their approval of petitioners' NADA's at the same time as they presented their prima facie case, the FDA frustrated the correct operation of this procedural mechanism.

In the instant case, however, certain actions of the court prior to our consideration of the merits of this appeal prompted extensive submissions by petitioners to the FDA and to this court.42

For that reason we have undertaken to explore how the case would stand if we were to evaluate the FDA's action as if the Commissioner had followed the proper procedures. We have undertaken this, grudgingly, in order to assure sustained and reflective on-going administrative consideration, particularly in the light of FDA's tenacity and reiterated resolution to refuse a hearing on this matter.43 But if this exploration is prompted by respect for the role of this court, as in a kind of partnership with the agency under review in furtherance of the public interest, our simultaneous function of surveillance calls on us to underline our concern about the manner in which the FDA has proceeded in this case.

The reasons for our concern begin with the fact that FDA cast its June 21, 1972, Notice as a mere precursor of a broad hearing. In August, 1972, it withdrew feed premixes from the market, but specifically stated (see note 18) that it was 'prudent to pursue and resolve existing scientific questions about implants.' Months went by and then, in April, 1973, the FDA withdrew implants on the basis of results received earlier in 1973 of tests not begun until months after the Notice of Opportunity for Hearing. The FDA accompanied its revocation with reliance on the Delaney Clause-- possibly only a 'scare tactic,' for it abandoned that reliance when called upon to make a considered submission to this court. At the very least, it refrained from stating the precise basis of its actions, which should be the first business of any agency, until it filed its briefs in this Court, tardily, a few days before oral argument.

The FDA has an important responsibility for the public health. In this case, however, as we have noted, neither the Secretary nor the Commissioner has found a hazard to the public health-- the Commissioner has consistently acted for 'regulatory' reasons alone. Even in that capacity, of course, the Commissioner's job is a difficult and delicate one. Nonetheless, where, as here, action is taken for regulatory reasons, the statute does not permit us to ignore procedural deficiencies on the basis of a public health hazard that has not been declared.

As a result of the foregoing, this case has become unnecessarily confused. We have been thrust into a review on a piecemeal record, with material gleaned, variously, from the original proceedings at the agency; from the motion for stay pending appeal; from the response to our order granting the stay on specified conditions. Still other material-- including substantial additional briefing of legal issues-- has arrived as part of a 'Motion for Clarification' of our September 14, 1973, Order.44 We are loath to proceed in this fashion, to assist the Commissioner in a repair carpentry to substitute for a sound basic structure. In his latest submission, the Reply to Opposition to Motion for Clarification, the Commissioner in effect puts the case to the court that its sound course lies in deference to his expertise and judgment as the agent charged with responsibility for the public health.

What is at stake in this case is the integrity of the Commissioner's expertise in a broad sense, exercised, in the manner ordained by Congress, after genuine opportunity for hearing and consideration of controverted issues. We cannot allow important issues of law and public policy to be decided in a patchwork of legal theory that is sewn in a confusion inconsistent with responsible review.45 Thus, we have attempted to sort out the strands of approach possibly available in this case, and to consider whether the Commissioner's denial to petitioners of a hearing can be justified.

As we noted in Part II, there are two justifications for withdrawal of approval of NADA's without prior hearing. First, the Secretary may act without a hearing when he finds that an approved drug constitutes an imminent hazard to public health.46 Here, the Secretary has not found an imminent hazard to public health and indeed, as we have already noted, the Commissioner's order stated that he had 'no reason to believe that the use of DES presented a public health hazard.'47 Second, the FDA may invoke its summary judgment procedures if, after proper notice, and adverse party fails to raise material issues of fact in its request for a hearing.48 The issue is whether, in this case, the FDA may properly invoke its summary judgment procedures to deny petitioners a hearing.

Possible Theories of FDA Withdrawal of Approval of Petitioner's NADA's.

The statute provides that the FDA may withdraw its approval of an NADA when

new evidence . . . together with the evidence available to the Secretary when the application was approved, shows that such drug is not shown to be safe for use under the conditions for use upon the basis of which the application was approved or that subparagraph (H) of paragraph (1) of subsection (d) of this section applies to such drug; . . .49

Relying on this section, there are two theories by which the FDA could proceed to withdraw approval of petitioners' NADA's: (1) the Delaney Clause theory; or (2) the general safety clause theory.

1. The Delaney Clause.

Under the above quoted statute, the FDA can remove a drug from the mark when new evidence indicates a violation of the Delaney Clause, 21 U.S.C. 360b(d)(1) (H). If the FDA, using an approved test method, detected residues of DES in edible portions of slaughtered animals, then it could show a violation of the Delaney Clause.

In the instant case, however, despite references to the Delaney Clause, and despite continual reference in the briefs and at oral argument to the fact that DES is a carcinogen, the FDA has plainly not used the Delaney Clause theory. One possible reason for this election is that the detected residues may not be DES residues and hence the Delaney Clause may be inoperative. In any event, the USDA did not detect the residues while using an 'approved' test method as required by the Delaney Clause.50 In its regulations, the FDA has approved only the 'mouse uterine' test.51 Using this test, no residues have been found in the tissues of slaughtered animals.52 Rather, the only method by which residues have been detected is the radioisotope tracer test,53 but that method has not been approved. For that reason, the Delaney Clause is plainly inapplicable, without regard to the composition of the residues.

The Commissioner relies on the alternative theory of the 'general safety' clause of section 360b(e)(1)(B), contending that the new evidence from the USDA tests 'shows that (DES) is not shown to be safe.' More precisely, the Commissioner asserts that nothing in petitioners' responses raises any material issues of fact about his conclusion that DES is no longer shown to be safe. Thus, the Commissioner concludes that he need hold no hearing. In response, petitioners claim that a material issue exists whether the Commissioner has made the proper showing.

a) Operation of the General Safety Clause

The Commissioner urges that once he has adduced evidence that shows a drug is not shown to be safe, then the drug manufacturer must come forward and prove his drug is safe. Moreover, the Commissioner argues that such a showing can only be made by presentation of adequate tests showing safety.54 Because he finds no such 'adequate test' results in petitioners' responses, the Commissioner has concluded that these responses fail to place his factual showing in issue. Therefore he has refused to grant petitioners a hearing.

This argument reflects an imprecise reading of the statute. The pertinent sections permit withdrawal of an outstanding approval on the basis of 'new evidence' (a term used broadly here to include 'tests by new methods, or tests by methods not deemed reasonably applicable when such application was approved'). The Secretary or Commissioner, may withdraw the approval if this 'new evidence . . . evaluated together with the evidence available . . . when the application was approved shows that such drug is not shown to be safe for use . . .' 21 U.S.C. 360b(e)(1)(B).

The statute plainly places on the FDA an initial burden to adduce the 'new evidence' and what that new evidence 'shows.' Only when the FDA has met this initial burden of coming forward with the new evidence is there a burden on the manufacturer to show that the drug is safe. Only at this later stage must the manufacturer produce 'adequate tests' of safety.

In the instant case, we have not yet reached the second stage in this process. Rather, the issue is whether the FDA has met the initial burden of adducing the 'new evidence' and what it shows in terms of undermining the previous conclusions as to safety. The manufacturers have introduced affidavits and other evidence which challenge the FDA's showing in every particular. Because this evidence has been directed to the adequacy of the FDA's initial showing, it need not consist of the results of adequate tests in order to raise material issues of fact. We therefore reject the FDA's glib assertion that petitioners cannot possibly raise material issues of fact because they lack adequate test results. Instead, we must determine if petitioners' submissions in fact do raise material issues concerning the FDA burden of making a showing concerning the new evidence.

b) Material Issues of Fact

Because he is not using the Delaney Clause, it is not enough for the Commissioner merely to show that animal carcasses contain residues and that DES is a carcinogen. Instead, the FDA must show that two different issues are resolved in its favor before it can shift to petitioners the burden of showing safety: (1) whether the detected residues are related to the use of DES implants; (2) if so, whether the residues, because of their composition, and in the amounts present in the tissue, present some potential hazard to the public health.

(1) Relationship of commercial DES implants and the detected substances.

Petitioners have submitted material supporting the assertion that the testing method itself produced the detected residues. The thrust of these submissions is that the residues resulted from the presence of impurities in the specially made 'tagged' implants. Hess & Clark speculates that some of these impurity-caused residues may be so-called 'pseudo-DES,' while the composition of the remainder is unknown.

In response, the FDA contends that Hess & Clark raises no material issue of fact because it has failed to identify the precise nature of the detected residue. But this failure by Hess & Clark is no automatic validation of the FDA's characterization of the residues. Rather, the submission raises an issue about the composition of the residue, and thus about the relationship of that residue to the tests by which it was detected. This issue is quite material: if the residues are caused by the test, then the Commissioner could hardly maintain that their presence shows that commercial use of DES implants is not shown to be safe. Accordingly, we conclude that petitioners' submissions raise substantial and material issues of fact about the very premise of the FDA action, i.e. that commercial use of DES implants has caused harmful residues in human food.

(2) Relationship between the detected residues and safety

Even if the commercial use of DES implants causes the detected residues, petitioners further assert that a material issue exists whether the implants are safe. The FDA argues that if the residues exist, and if the implants caused them, then the manufacturers have the burden of proving the safety of their products.

The FDA argument is as follows. The statute states that, prior to approval of an NADA,

in determining whether such drug is safe for use under the conditions prescribed . . . the Secretary shall consider, among other relevant factors, (A) the probable consumption . . . of any (other) substance formed in . . . food because of the use of such drug.55

The FDA asserts that the residues are 'substances formed in food' because of the use of DES implants. That being the case, the FDA argues that a showing that such substances exist is a showing that DES implants are no longer shown to be safe.

This argument skimps FDA's responsibility. The quoted statute requires the FDA to consider the relationship between a drug's safety and residues left in food because of use of the drug. The statute does not say that because a drug leaves residues, it is unsafe per se. We think it implicit in the statute that when the FDA proposes to withdraw an approval because new evidence shows the drug leaves residues, it has an initial burden of coming forward with some evidence of the relationship between the residue and safety to warrant shifting to the manufacturer the burden of showing safety. This is at least the case where, as here, the residues are of unknown composition.

The withdrawal of the approval in the case before us cannot be sustained in the absence of a hearing, in view of the substantial material issue as to whether the detected residues are unsafe. Petitioners have submitted affidavits and other information which indicate that the detected residues may be wholly innocuous in composition and effect. For instance, petitioners assert that the studies on which the FDA relies found no 'free DES,' but only DES conjugate. Petitioners urge that DES conjugate is estrogenic and they offer expert opinion that such estrogenic residues are safe, even if they are caused by DES implants. These assertions create a material issue of fact which remains unresolved.

Even if the residues were 'free DES' rather than DES conjugate, petitioner Vineland has submitted data indicating that such small amounts of DES are safe. So long as the Commissioner does not invoke the Delaney Clause, that issue is not foreclosed by a showing that larger amounts of DES are carcinogenic. The FDA routinely allows sale in limited quantities of drugs that would be lethal in substantial amounts,56 and it has acknowledged that small amounts of DES may present no danger to the consumer.57

Outside of the per se rule of the Delaney Clause, the typical issue for the FDA is not the absolute safety of a drug. Most drugs are unsafe in some degree.58 Rather, the issue for the FDA is whether to allow sale of the drug, usually under specific restrictions. Resolution of this issue inevitably means calculating whether the benefits which the drug produces outweigh the costs of its restricted use.59 In the present case, DES is asserted to be of substantial benefit in enhancing meat production, and this is not gainsaid by FDA. The FDA must consider, after hearing, whether DES pellets would be safe in terms of the amounts of residue consumed. Or the FDA might restrict such consumption by a ban on sale of liver, the only food material in which any residues have even been detected.60

This situation presents a substantial and complex problem consisting of issues of both fact and policy. The issues of fact mandate a hearing, and the hearing when held may soundly range into the inter-related policy issues. As we stated in Citizens for Allegan County v. Federal Power Commission,

. . . questions of public interest confronting an administrative agency will often be illuminated by an exploration in greater depth than can be provided simply by pleadings and documents.61

Thus, as to the relationship of the residues and safety, there exist in this case at least three issues: (1) whether the detected residues are composed solely of DES conjugates, and whether that substance is harmful; (2) whether, if they consist of DES, the residues are nonetheless harmless; (3) whether, if the residues are harmful, the FDA should keep DES implants from the lmarket, in light of their acknowledged value and of the possibility for meaningful restrictions on consumption of the residues.

We hold, as a matter of law, that the FDA has failed to carry its statutory burden, under the general safety clause theory of section 360b(e)(1)(B), as a justification for refusal to grant petitioners a hearing. To meet its burden in the context of the instant case, the FDA would have had to show that the procedures 'conclusively' establish that there is no need for a hearing, and that no material issues of fact existed concerning: (1) the relationship between DES implants as used in commercial applications and the residues detected in a special testing program; and (2) the relationship between the detected residues and the safety of the DES implants. This record leaves open substantial and material issues of fact about the nature of both relationships. The FDA must resolve these issues through the hearing mechanism. In the absence of the restraints available in the emergency situation, i.e., on a finding of imminent hazard to safety, the FDA cannot assert, as a matter of paternalistic sagacity, that it can dispose of these matters without opportunity for hearing.

Our holding today invalidates the FDA Order of April 27, 1973. Therefore, the petitioners' NADA's remain approved until the FDA has held a hearing to resolve the material issues of fact raised by petitioners submissions. Because their NADA's have not been lawfully withdrawn, petitioners may continue to market their products. Of course, should the Secretary conclude that such marketing constitutes an imminent hazard to public health, sale of the products may be suspended forthwith.

The Commissioner argues in his Motion for Clarification that this court is without power to allow continued marketing of DES implants. He contends that he revoked the regulations governing the use of DES implants when he withdrew his approval of petitioners' NADA's, and that this court cannot review his revocation.

We reject this contention. The statute, 21 U.S.C. 360b(i), requires the Commissioner to revoke use regulations when he revokes approval of an NADA. Likewise the agency is expressly required to issue regulations to govern use of approved drugs. The corollary of the FDA's revocation of the approval of the NADA's was its revocation of the use regulations governing DES implants. Our reversal reinstates the status quo ante, and the corollary of our reinstatement of the approval of the NADA's is the reinstatement of the use regulations governing DES implants that accompanied the former approvval of the NADA's.

When, as here, the Commissioner's revocation of use regulations is inextricably unied to his withdrawal of approval of petitioner's NADA's, and that withdrawal is terminated as invalid, the revocation of the use regulations is also terminated unless the Commissioner advances some new, untainted basis for such revocation.

One last procedural point remains. The Commissioner chose to publish his continued refusal to hold a hearing in the Federal Register. He denominated his refusal an 'Order.'62 This 'order' has no independent validity; issued after the record was certified to this court, it is no more than a renewed expression of the position occupied by the Commissioner since April 27, 1973. And it is rejected by our judgment vacating the April 27, 1973 order.

The Commissioner's Orders of April 27, 1973, and May 3, 1973, are vacated, and the case is remanded for further proceedings consistent with this opinion.

So ordered.