United States v. 225 Cartons, More or Less, of an Article or Drug

OPINION OF THE COURT

Appellant Sandoz Pharmaceuticals Corporation has been unsuccessful in its attempt to obtain Food and Drug Administration (FDA) approval of its new drug applications for the products in question under the usual premarket clearance procedures required by 21 U.S.C. § 355(a) & (b). Its contention that the products can nonetheless be distributed in commerce rests on its claim that they are “generally recognized” to be safe and effective, a claim the district court rejected. The issue presented on appeal requires this court to interpret for the first time the regulation promulgated by the FDA governing the effectiveness of “combination” prescription drug products, i.e., products containing fixed combinations of ingredients, some or all of which may have been granted prior approval by the FDA. See 21 C.F.R. § 300.50 (1988).

I.

Facts

The facts are fully set forth in the comprehensive opinion of Judge Debevoise reported at United States v. 225 Cartons More or Less, of an Article of Drug, 687 F.Supp. 946, 948-53 (D.N.J. 1988), and we set forth only those essential to our discussion.

Sandoz has been marketing products containing Fiorinal and codeine since 1963. In 1968, the FDA revoked all prior advice that such products were not regarded as new drugs, see 21 C.F.R. § 310.100(d) (1988), because the Food, Drug and Cosmetic Act, 21 U.S.C. § 301 et seq. (FDC Act), was amended in 1962 to require the FDA to evaluate the effectiveness as well as the safety of all drug products that had been placed in commerce since 1938, with certain exceptions for “grandfathered” products not applicable here. See generally Weinberger v. Hynson, Westcott and Dunning, Inc., 412 U.S. 609, 613-16, 93 S.Ct. 2469, 2475-76, 37 L.Ed.2d 207 (1973) (discussing 1962 amendments); Warner-Lambert Co. v. Heckler, 787 F.2d 147, 156 (3d Cir. 1986) (same).

The effect of this revocation was to require that manufacturers file a new drug application (NDA) as required under 21 U.S.C. § 355(b) for all such products, except that an abbreviated new drug application (ANDA) could be used, if approved by the FDA, for duplicates of approved drugs or drugs “very closely related” thereto. See 21 C.F.R. § 314.56 (1988). In 1973, Sandoz submitted an NDA for Fiorinal, an analgesic containing aspirin, caffeine, phenacetin and the barbiturate butalbital. In 1977, the FDA approved Fiorinal as a “new drug” effective for treating tension headaches following its evaluation under the specially established procedure for “drug efficacy study implementation” (DESI) with the assistance of panels of the National Academy of Sciences. See 42 Fed.Reg. 59,115 (1977).

*412Thereafter, Sandoz sought to get FDA’s approval for its Fiorinal with Codeine (FWC) products through the abbreviated procedure applicable to products closely related to products previously approved. See 21 C.F.R. § 314.55(a) (1988). The products for which approval was sought through the abbreviated procedure were the prescription drug products Fiorinal with 7.5 mg. of codeine phosphate (FWC No. 1) and Fiori-nal with 15 mg. of codeine phosphate (FWC No. 2).1 The FDA twice rejected these abbreviated applications for the FWC products on the ground that the approved products on which Sandoz sought to “piggy back” the FWC applications, plain Fiorinal and Trigesic with Codeine, contained different ingredients. Sandoz did not appeal these decisions.2

The FDA advised Sandoz at a meeting in 1978 that Sandoz’s FWC products were considered new drugs and would require the filing of full new drug applications, that Fiorinal with Codeine is not related to any DESI drug indicated for pain states because of the presence of the barbiturate butalbital which has not been proven effective in analgesia, and that the process permitting ANDAs for similar, identical, or related drugs was not intended to permit drug sponsors to avoid the required clinical testing. At the same meeting the FDA advised Sandoz that it believed that FWC No. 1, containing the lowest dosage of codeine, would probably not meet the requisite test of effectiveness.

At the time Fiorinal had been approved as efficacious it had contained phenacetin as one of its ingredients. In 1983, the FDA required removal of phenacetin and Sandoz substituted additional aspirin as permitted by the FDA. See 47 Fed.Reg. 34,636, 34,-640-41 (1982).

Thereafter, in 1984 and 1985, the FDA sent Sandoz Regulatory Letters that the FWC products were unapproved new drugs covered by a 1980 notice finding other barbiturate analgesic combination drugs to be lacking substantial evidence of effectiveness. The letters warned Sandoz that it could not continue to market the FWC products without submission of the necessary NDAs. Sandoz submitted both a full new drug application and another request that the FDA accept the abbreviated applications for the FWC products based on the earlier approval of Fiorinal. The FDA’s preliminary review of the new drug application was unfavorable and it rejected the request with respect to the abbreviated application. In letters dated April 26 and May 16, 1985, Sandoz submitted to the FDA an unpublished clinical trial study and two published studies along with declarations from three physicians who had examined the published studies and who gave their opinions that the FWC products were not new drugs because they were generally recognized as safe and effective for their intended uses.

II.

Procedural History

The FDA filed a complaint for forfeiture of the FWC products in October 1986, alleging that they were “new drugs” within the meaning of 21 U.S.C. § 321(p) and could not be marketed without obtaining FDA approval through the filing of a new drug application. The complaint also alleged that the drugs were misbranded because they were being sold without labels containing adequate directions for their use, a requirement from which they were not exempt in light of their status as unapproved new drugs. In January 1987, a United States Marshal seized the FWC products. Sandoz intervened as claimant.

*413The FDA moved for summary judgment contending that there was no genuine issue of material fact regarding Sandoz’s failure to meet the legal standard for exemption from the requirements for new drug approval. After evaluating documentary evidence submitted by both parties, the district court granted the government’s summary judgment motion, holding that it had sustained its burden of demonstrating the absence of a genuine issue of material fact. 687 F.Supp. at 956.

Sandoz appeals, arguing that the district court incorrectly interpreted the legal standards applicable to this case and improperly granted summary judgment when issues of material fact remained in dispute. Our review of the legal standard formulated by the district court is plenary. See Dent v. Cunningham, 786 F.2d 173, 175 (3d Cir. 1986). For the reasons set forth below, we will affirm the district court’s judgment.

III.

Discussion

A.

The “Generally Recognized” Exclusion

Sandoz’s principal argument on appeal is that it was not obliged to follow the statutory and regulatory requirements applicable to new drugs because the FWC products seized are “generally recognized” as effective. The FDC Act mandates that no “new drug” may be introduced into interstate commerce until it has been approved by the FDA as safe and effective. 21 U.S.C. § 355(a) & (b). In this case, the FDA’s challenge is directed only to efficacy, not safety.

Under the relevant provision of the statute, the term “new drug” means,

Any drug ... the composition of which is such that such drug is not generally recognized, among experts qualified by scientific training and experience to evaluate the safety and effectiveness of drugs, as safe and effective for use under the conditions prescribed, recommended, or suggested in the labeling thereof.

21 U.S.C. § 321(p) (emphasis added). As we explained in Tri-Bio Laboratories, Inc. v. United States, 836 F.2d 135, 141 (3d Cir. 1987), cert. denied, — U.S. -, 109 S.Ct. 57, 102 L.Ed.2d 35 (1988), the exclusion for “generally recognized” drugs is a narrow one.

The scope of the exclusion for general recognition among experts was considered by the Supreme Court in Weinberger v. Hynson, Westcott and Dunning, Inc., 412 U.S. 609, 93 S.Ct. 2469, 37 L.Ed.2d 207 (1973), where the Court held that “the statutory scheme and overriding purpose of the 1962 amendments compel the conclusion that the hurdle of ‘general recognition’ of effectiveness requires at least ‘substantial evidence’ of effectiveness for approval of an NDA.” Id. at 629, 93 S.Ct. at 2483. Since “substantial evidence” is defined by the statute as “evidence consisting of adequate and well-controlled investigations, including clinical investigations,” 21 U.S.C. § 355(d), experts attesting to a drug’s “general recognition” of efficacy must base their opinions on evidence that meets the FDA’s extensive regulations governing whether a study is “adequate and well-controlled” for purposes of the FDC Act, see 21 C.F.R. § 314.126 (1988). Thus, “general recognition” cannot be satisfied by “clinical impressions of practicing physicians, ... poorly controlled experiments, ,.. [or] reports lacking the details which permit scientific evaluation.” Hynson, 412 U.S. at 630, 93 S.Ct. at 2483. General recognition in this context is a term of art. It does not involve establishing that a consensus of opinion by experts exists, but requires the submission of clinical studies meeting scientific standards at least as stringent as those applied under the NDA approval procedures.

B.

The District Court’s Decision

In support of its motion for summary judgment, the FDA submitted declarations from six experts who had been asked by the FDA to provide their professional opinion regarding the general recognition, or lack thereof, of the safety and effective*414ness of FWC No. 1 and FWC No. 2. Five of these experts stated that they were unable to conclude that FWC No. 1 and FWC No. 2 are generally recognized as effective for any of their intended uses; the sixth gave no opinion on this. Five of the experts stated that the published studies which Sandoz submitted to the FDA did not establish that each ingredient in the FWC products contributes to their effectiveness, noting that none of the studies contained enough control groups to allow an evaluation of whether the addition of each ingredient in the FWC products provided an additional benefit to the patient as required under 21 C.F.R. § 300.50. The sixth FDA expert stated that the studies failed to establish that the addition of codeine contributed to the claimed effects.

With respect to Sandoz’s attempt to rely on other approved drugs for the claim of efficacy, all of the FDA experts noted that none of the three studies initially submitted by Sandoz analyzed an FWC product containing a dose of codeine as low as the 7.5 mg. present in FWC No. 1, therefore leaving them unable to conclude that the addition of this relatively low dose of codeine contributed to the claimed effects for FWC No. 1. In addition, all of the experts concluded that there were various deficiencies in some of the studies which rendered them inadequate under the criteria for well-controlled studies specified in 21 C.F.R. § 314.126. See 687 F.Supp. at 955-56 (summarizing deficiencies in control procedures used in certain of the studies). Finally, Sandoz had not submitted dissolution and/or bioavailability3 studies showing that the removal of phenacetin and the substitution of additional amounts of aspirin did not affect the product, rendering the studies conducted on the old formula of FWC an inadequate basis for drawing conclusions about FWC as currently formulated.

In opposition to summary judgment, San-doz filed the Declarations4 of eight of its own experts and a total of six published studies, which included those it had originally submitted to the FDA. The district court discussed the experts’ submissions in detail, see 687 F.Supp. at 957-58, and no repetition of that material is necessary. The Sandoz experts opined that the FWC products are safe and effective for their intended uses and generally recognized as such among experts qualified to evaluate the drugs’ safety and effectiveness, and that the “pharmacological rationale” for the products suggests that each component of Fiorinal with Codeine contributes to the claimed effect. What is significant for our purposes is that none of the experts referred to clinical studies establishing the effectiveness of each ingredient of FWC No. 1 and FWC No. 2 in their present formulations. Instead, they stated that their clinical experience and the extrapolation of data from the published literature provided sufficient evidence to conclude that each ingredient in FWC products contributes to their therapeutic effect and that studies designed with controls to demonstrate the contribution of each ingredient were therefore not necessary. Each stated that the substitution of additional aspirin for phenacetin could be expected not to alter the effectiveness of the drug, and that the contribution made by the addition of 7.5 mg. of codeine in FWC No. 1 could be extrapolated from the results obtained in tests conducted on formulas containing higher doses of codeine.

In granting summary judgment, the district court assumed that the studies submitted by Sandoz met the FDA’s standards for adequate and well-controlled investigations even though that was contested by the FDA’s experts. The district court also stated that if the relevant issue were whether FWC No. 1 and FWC No. 2 were in fact safe and effective for their intended uses and whether each ingredient made a contribution, the affidavits and material submitted by Sandoz would be sufficient to *415create a genuine issue of fact. 687 F.Supp. at 959. However, the district court concluded that the studies submitted by San-doz did not establish “general recognition” by qualified experts as required to obtain an exemption from the new drug approval procedures because they did not show that each of the components of the FWC products is effective in producing the claimed results. 687 F.Supp. at 957. The court construed the applicable FDA regulation, 21 C.F.R. § 300.50, as requiring a showing that each active ingredient in a combination drug makes a contribution to the claimed effect for the drug, and held that this requirement had a rational basis.

The district court reviewed the three studies the FDA experts evaluated as well as three additional studies submitted by Sandoz, and found that the six studies relied on by Sandoz “do not establish general recognition by qualified experts that the particular drugs in this case and each of the components of those drugs are effective to produce the claimed results.” 687 F.Supp. at 957. The studies are deficient because “they did not adequately address the two drugs at issue” and because “they did not adequately address the ingredients of those two drugs as used in these two drugs.” Id. at 959. The court held that these deficiencies cannot be remedied by the extensive clinical experience of San-doz’s experts or by the “extensive extrapolation of data to which these experts resorted.” Id.

To summarize the deficiencies, “there has been no study of FWC No. 1 either with respect to its overall efficacy or with respect to its components.” Id. at 961. “[T]here are no published, well-controlled clinical studies showing ... what contribution caffeine in any amount makes to FWC No. 1 and FWC No. 2.” Id. The only study purporting to establish the contribution of butalbital was conducted on a different formulation of FWC No. 2 that had not been shown to be the bioequivalent of FWC. Id. Finally, even though Sandoz’s evidence could show the pain-relieving quality of codeine, “it has not produced evidence of published, well-controlled clinical studies that codeine or the amount of codeine in FWC No. 1 (7.5 mg. codeine) and FWC No. 2 (15 mg. codeine) contribute to the effectiveness of the product.” Id.

The district court, framing the rule of law which it believed was proper and which this circuit would follow, concluded that “the combination policy of 21 C.F.R. § 300.50 can only be met by studies of the particular drug product at issue or a drug having the same active ingredients which is demonstrably its bioequivalent.” Id. at 960-61.5 Sandoz had failed to produce such studies for its FWC combination products, and thus the court rejected its claim of general recognition. Moreover, since Sandoz did not argue that it would be able to produce such studies, its request under Fed.R.Civ.P. 56(f) for a continuance to permit further discovery was denied. Responses to the requested discovery “could not fill the void in necessary published clinical investigations.” 687 F.Supp. at 962.

C.

The Applicable Legal Standard for Evaluating the Effectiveness of Combination Products

The FDA’s current policy with respect to the safety and effectiveness of fixed combination dosage form prescription drugs for humans is set forth in a regulation that it promulgated in 1975, see 21 C.F.R. § 300.50, which provides, in relevant part, that:

Two or more drugs may be combined in a single dosage form when each component makes a contribution to the claimed effects and the dosage of each component (amount, frequency, duration) is such that the combination is safe and effective for a significant patient population requiring such concurrent therapy as defined in the labeling for the drug.

*41621 C.F.R. § 300.50(a) (1988). In other words, a product with an additional active ingredient must be more effective for its intended purpose than that product would be without the extra ingredient.

Sandoz does not challenge the validity of the FDA’s combination drug regulation. Its challenge, however, is to the FDA’s insistence in this case that the contribution of each of the active ingredients in the two FWC products must be proven either (1) by studies of the FWC products showing that each ingredient is making a real clinical contribution of statistical significance or (2) by studies of another product with identical active ingredients together with studies showing that the other product is bioequi-valent to the FWC product.

Sandoz does not argue that it has satisfied the first alternative. It argues that it does not have to meet the second alternative which requires bioequivalency data in order to show general recognition of its combination drug products. It argues that it can show general recognition instead through extrapolation from data derived from tests of other products when justified as a matter of scientific fact.

1. The Ingredient Contribution Requirement

We turn first to the district court’s determination that when the drug at issue is a combination drug FDA regulation § 300.50 requires investigation showing the contribution of each ingredient. The court accepted the FDA’s position that a fixed-combination drug product may not be excluded from the statutory premarketing clearance procedures unless there are adequate and well-controlled clinical investigations that are published in the scientific literature and show that the dosage of each ingredient in the product contributes to the effects claimed for the product. The district court found the two studies submitted by Sandoz which involved tests of the current FWC formula inadequate because neither study purported to measure the relative contribution of the constituents of the combinations studied. 687 F.Supp. at 961.

Although this court has not previously spoken on the issue, other circuits have upheld the FDA’s combination drug policy. For example, in United States v. Articles of Drug ... Promise Toothpaste, 826 F.2d 564, 570-71 (7th Cir. 1987), the court rejected the arguments advanced by experts for a drug manufacturer that the effectiveness of its combination product could be surmised by extrapolating from the results of studies conducted on the product’s individual ingredients. The court held that the manufacturer should be required to submit studies which tested the contribution of each ingredient to the claimed effect of the end product. Id. at 570-71. “[T]he regulations,” the court wrote, “require more than the opinions of experts.” Id. at 572. Accord United States v. An Article of Drug Consisting of 4,680 Pails, 725 F.2d 976, 986-87 (5th Cir. 1984) (“[ajdequate and well-controlled investigations required by the combination policy necessitate studies that compare the effect of the individual ingredients with the effect of the combination, and that include a control group.” Id. at 987.)6 See also Masti-Kure Prods. Co., Inc. v. Califano, 587 F.2d 1099, 1103-04 (D.C.Cir. 1978) (applying same test).

We have previously found persuasive the Commissioner’s interpretation of the statutory requirement of effectiveness as requiring a showing that patients will receive a clinically, i.e., therapeutically, significant benefit from the drug. See Warner-Lambert, 787 F.2d at 155. For the same reasons that drug companies may not market drugs that are not clinically significant, they may also not include components in combination drugs that are not shown to be necessary for the effect claimed.

Judge Debevoise correctly anticipated that this court would adopt the same ingredient contribution rule that other circuits *417have. We thus expressly affirm the district court’s reasoning and its conclusion that 21 C.F.R. § 300.50 requires manufacturers to submit clinical evidence establishing the contribution of each ingredient to the effectiveness of a product for its indicated applications.

2. The Bioequivalency Requirement

Sandoz also argues that the district court applied too stringent a standard in evaluating the evidence it submitted. It contends that it did submit studies that were both adequate and well-controlled and sufficient to satisfy the combination policy, as well as studies on the products in question which established that the products were effective. The FDA, however, found the studies deficient because they were made on drugs which were not shown to be bioequivalent to FWC No. 1 and FWC No. 2.

The studies on which Sandoz relied involving tests of FWC under its prior formula tested a different product because FWC then contained phenacetin, the ingredient which was removed from Fiorinal and FWC products in 1983 pursuant to the FDA’s directive. See 47 Fed.Reg. 34,636 (1982). The FDA directive did not require those manufacturers who replaced phenacetin with aspirin to conduct new clinical safety and effectiveness studies, but they were required to conduct in vitro dissolution rate or bioavailability studies to establish that any new formulation was the bioe-quivalent of the old. Id. at 34,641. Sandoz submitted no studies to the district court demonstrating the bioequivalency of the new and old Fiorinal formulas and does not argue that such studies are available. Furthermore, as noted by the district court, none of Sandoz’s experts mention bioequi-valence and none state that they have any knowledge of the inactive ingredients in the new or old formulation. 687 F.Supp. at 960.

“Bioequivalent drugs” have similar “bioavailability,” i.e., “the active ingredients in the drugs are absorbed from the drug product at a similar rate and to a similar extent.” United States v. Undetermined Quantities of an Article of Drug ... Anucort Hemorrhoidal Suppositories, 709 F.Supp. 511, 515 (D.N.J. 1987), aff'd mem., 857 F.2d 1466 (3d Cir.), cert. denied sub nom. G & W Laboratories, Inc. v. United States, _ U.S. _, 109 S.Ct. 864, 102 L.Ed.2d 988 (1988); see 21 C.F.R. § 320.1(a) (1988). As we explained in TriBio Laboratories, 836 F.2d at 137 n. 3, “Bioequivalence compares the action of two drugs that, ‘when administered to the same individual in the same dosage regimen, result in equivalent concentrations of drug in blood and tissue,’ Merck Manual 2241 (R. Berkow 14th ed. 1982). See also 21 C.F.R. § 320.1(e) (1987).”

The scientific rationale for the bioequiva-lency requirement is well-recognized since studies on bioequivalency are necessary to establish whether products with different ingredients deliver the same dose of active ingredients. See United States v. Undetermined Quantities of Various Articles of Drug ... Equidantin Nitrofurantoin Suspension, 675 F.2d 994, 996 (8th Cir. 1982), cert. denied, 460 U.S. 1051, 103 S.Ct. 1497, 75 L.Ed.2d 929 (1983) (drug that is less bioavailable delivers less of active ingredient; drug that is more bioavailable presents danger of overdose) (citing United States v. Premo Pharmaceutical Laboratories, Inc., 511 F.Supp. 958, 962-63 (D.N.J. 1981)). See also United States v. Articles of Drug ... 5,906 Boxes, 745 F.2d 105, 117-18 (1st Cir. 1984), cert. denied, 470 U.S. 1004, 105 S.Ct. 1358, 84 L.Ed.2d 379 (1985) (“substantial evidence” standard not met where no showing of bioequivalency between drug in question and drug with studies on which manufacturer sought to rely).

The Supreme Court’s decision in United States v. Generix Drug Corp., 460 U.S. 453, 454-55, 103 S.Ct. 1298, 1299-1300, 75 L,Ed.2d 198 (1983), defeats Sandoz’s argument that it need not show the bioequiva-lenee of the new and old FWC formulas before relying on studies conducted on the old formula. In Generix, a drug manufacturer of generic drugs sought to avoid the necessity of filing an NDA on the ground *418that the active ingredients of its products were the same as those in previously approved pioneer drugs. The district court enjoined the distribution of the generic drug products, finding that there was a reasonable possibility that the differences in the inactive ingredients between the generic drugs at issue and the pioneer drugs which they copied might affect their bioe-quivalency, and hence their safety and effectiveness. Id. at 461, 103 S.Ct. at 1302. The Supreme Court agreed, holding that manufacturers of the generic drugs containing the same active ingredients as an approved product but different inactive ingredients must file an NDA. In light of the holding in Generix that a showing of bioequivalency is necessary to avoid filing an NDA even when only the inactive ingredients are different, Sandoz cannot reasonably contend that it need not submit bioe-quivalency data when the drugs it seeks to compare contain different active ingredients. It follows that a manufacturer of a drug is not excused from filing an NDA on the basis of the approval of another drug without also showing the bioequivalence of the two products.

We therefore hold that the district court did not err in concluding that in light of the absence of bioequivalency data the “studies conducted using the old formulation are not well-controlled clinical investigations of products using the new formulation and extrapolation of the data derived from them is not justified for the purpose of a ‘new drug’ determination.” 687 F.Supp. at 960.

3. Sandoz’s Claim for a “Lesser” Standard of Supporting Evidence

Sandoz, taking a statement of the Supreme Court out of context, argues that “adequate and well-controlled studies are not always legally necessary for a conclusion of general recognition.” Appellant’s Brief at 33. It points to the “long-standing acceptance of FWC” and “the substantial body of published studies” as sufficient to raise factual issues of the general recognition of FWC Nos. 1 and 2. Id.

In Weinberger v. Bentex Pharmaceuticals, Inc., 412 U.S. 645, 93 S.Ct. 2488, 37 L.Ed.2d 235 (1973), on which it relies, the issue before the Supreme Court was whether the FDA had jurisdiction to decide in the first instance the validity of the claim of drug manufacturers that their drugs were generally recognized as safe and effective under the new effectiveness requirements imposed by the 1962 amendments. The Court held that the FDA did have such jurisdiction, and that it would be commonplace for a court to await the FDA’s declaration before it acted in an enforcement proceeding. In the course of its discussion, the Supreme Court stated, “It may, of course, be true that in some cases general recognition that a drug is efficacious might be made without the kind of scientific support necessary to obtain approval of an NDA.” Id. at 652-53, 93 S.Ct. at 2493.

Nothing in the opinion suggests that by this sentence the Court was establishing a “flexible” standard as to the proof required to show general recognition. In fact, in the very next sentence, the Court reiterated its statement in Hynson, also decided that day, that “the reach of scientific inquiry under both § 505(d) [covering the FDA’s premarket clearance for new drugs] and § 201(p) [under which drugs “generally recognized” are excluded from the definition of “new drugs”] is precisely the same.” Id. at 653, 93 S.Ct. at 2494; see also United States v. Articles of Food and Drug ... Coli-trol 80, 518 F.2d 743, 746 (5th Cir. 1975).

Sandoz also bases its claim for a “flexible” approach to the proof required to show general recognition on United States v. An Article of Drug ... “Entrol-C Medicated”, 513 F.2d 1127 (9th Cir. 1975). The holding in that case does not support the proposition for which Sandoz seeks to use it. On the contrary, the court there expressly rejected the drug manufacturer’s contention that general recognition of a combination drug could be established by showing the general recognition of its component parts. Id. at 1129-30. Thus, in Entrol-C, the court held the manufacturer was required to show “general recognition accorded to the end product of this combi*419nation of component parts,” id. at 1130, a requirement similar to that also adopted in United States v. An Article of Drug ... Mykocert, 345 F.Supp. 571, 575-76 (N.D.Ill. 1972), where the court pointed out that even if each of the component parts were generally recognized, the combination of those parts must not in any way create a new drug.

Sandoz has cited no case which has adopted a test that is “lesser” or more “flexible” than that required by the FDA’s regulation. As we recently reiterated, “the FDA possesses an expertise entitled to respectful consideration by this court.” TriBio Laboratories, 836 F.2d at 142 (citation omitted); see Warner-Lambert, 787 F.2d at 152, 160; see also SmithKline Corp. v. FDA, 587 F.2d 1107, 1119 (D.C.Cir. 1978) (“ ‘technically illiterate judges’ ” should confine review to legal and not scientific questions presented by drug regulation). Therefore, in the absence of any sound authority to the contrary, we accept the FDA’s position that a combination drug cannot be shown to be generally recognized as effective without evidence of published clinical investigations that show that each component makes a statistically significant clinical contribution to the effects claimed for the products and that the results of tests on other products cannot be extrapolated without published dissolution, bioa-vailability or bioequivalence studies on the products compared. Neither Bentex nor Entrol-C can be read as supporting San-doz’s contention that the body of experience from long-standing market acceptance, without the controlled research required under the FDA regulation, can support a claim of general recognition.

Sandoz claims that the FDA has in the past permitted extrapolations from one strength of codeine to another in assessing the effectiveness of other drugs. We need not evaluate the accuracy of this claim or its implications if true because FDA decisions in other cases “do not bear on the decision in this case.” Warner-Lambert, 787 F.2d at 159; see also United States v. An Article of Drug ... Labeled ... (Box) “Furestrol Vaginal Suppositories”, 415 F.2d 390, 393 (5th Cir. 1969) (district court did not err in excluding as irrelevant to general recognition issue evidence that FDA had taken inconsistent positions regarding drug at issue); Anucort, 709 F.Supp. at 517-18 (D.N.J. 1987).

In summary, applying a plenary standard of review, we hold, as did the district court, that the FDA’s decision that Sandoz had not made the necessary showing of general recognition for FWC No. 1 and FWC No. 2 was consistent with the FDC Act and the FDA’s own regulations.

D.

The Propriety of Summary Judgment

Sandoz argues that the district court improperly weighed the evidence because it resolved a disagreement between the FDA’s expert witnesses and its own witnesses as to the effectiveness of the FWC products, a resolution impermissible on summary judgment.

This argument misunderstands the nature of the summary judgment test. Summary judgment requires that there be no dispute as to material facts; the materiality of facts is determined by the applicable legal standard. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248, 106 S.Ct. 2505, 2510, 91 L.Ed.2d 202 (1986). It is irrelevant that some facts remain in dispute if these facts could not affect the outcome of the case. Here, because it is undisputed that Sandoz did not produce evidence of general recognition of the products under the applicable legal standard discussed above, there was no impediment to the grant of summary judgment.

Sandoz’s suggestion that such a grant of summary judgment in such a new drug case is “unprecedented”, Brief of Appellant at 17, is belied by the case law, in which grants of summary judgment in favor of the government on the basis of documentary evidence in new drug cases have been routinely upheld. See, e.g., United States v. Atropine Sulfate 1.0 Mg., 843 F.2d 860, 861 (5th Cir. 1988) (affirming grant of summary judgment where lower court found that “as a matter of law, ‘any consensus *420which may exist regarding the safety or efficacy of [the drug] is not based upon adequate or well controlled studies’ ”); United States v. Articles of Drug ... Promise Toothpaste, 826 F.2d 564, 568-69 (7th Cir. 1987) (summary judgment appropriate where evidence did not contain the data necessary to assess effect of mixing ingredients to produce combination product); United States v. An Article of Drug Consisting of 4,680 Pails, 725 F.2d 976, 985-86 (5th Cir. 1984) (affirming grant of summary judgment based on government experts’ explanation of deficiencies of manufacturer’s studies in spite of contrary opinions of manufacturer’s experts); United States v. An Article of Drug ... “Entrol-C Medicated", 513 F.2d 1127, 1130 (9th Cir. 1975) (summary judgment appropriate where affidavits submitted by manufacturer failed to raise issue of “general recognition” accorded end product); see also Upjohn Co. v. Finch, 422 F.2d 944, 955 (6th Cir. 1970) (FDA Commissioner did not err in refusing to grant evidentiary hearing where evidence submitted failed to meet “substantial evidence” standard; “ ‘No amount of examination and cross-examination can change the scientific studies and the data reported into something they are not.’ ”).7

Sandoz also argues that the district court abused its discretion in denying its Rule 56(f) motion. Sandoz must establish that the district court’s action “ ‘made it impossible to obtain crucial evidence.’ ” Wisniewski v. Johns-Manville Corp., 812 F.2d 81, 90 (3d Cir. 1987) (citation omitted).

The evidence Sandoz sought to elicit on further discovery was, as the district court concluded, immaterial as a matter of law. Sandoz expressed its interest in “deposing] the Government’s expert witnesses to ascertain the depth and validity of their understanding of the issues on which they opined.” Brief of Appellant at 38. This would not have shed any light on the crucial issue, Sandoz’s failure to submit the clinical evidence necessary to meet the “substantial evidence” standard for finding “general recognition” of the safety and effectiveness of the FWC products in order to be excluded from the new drug application requirements of 21 U.S.C. § 355.

Similarly, Sandoz’s other proffered reason for such proposed discovery, obtaining evidence that would allow Sandoz to show that the FDA has applied inconsistent standards to evaluate studies submitted with other applications for exemptions from the new drug approval requirements for combination drugs, is, as we have held, not material in this ease. It follows that the district court neither abused its discretion nor erred as a matter of law in denying San-doz’s Rule 56(f) motion.

IV.

Conclusion

For the reasons stated above, we adopt as the rule for this circuit a legal standard for interpreting the FDA’s regulation for combination products, 21 C.F.R. § 300.50, under which manufacturers may be required to submit data demonstrating the contribution made by each ingredient to the claimed effect of a combination product. We also affirm the FDA’s position that manufacturers may be required to submit data establishing the bioequivalency between their product and another product before relying on studies conducted on the other product.

We will therefore affirm the judgment of the district court.