Serono Laboratories, Inc. v. Shalala, 974 F. Supp. 29 (D.D.C. 1997)

*30 Bruce S. Manheim, Fox, Bennett & Turner, Washington, DC, for Serono Lab., Inc.

Jeffrey Bruce Chasnow, U.S. Dept. of Justice, Washington, DC, for Donna Shalala and Michael Friedman.

David F. Weeda, Arthur YaShih Tsien, Olsson, Frank & Weeda, P.C., Washington, DC, for Ferring Pharmaceuticals, Inc.

SPORKIN, District Judge.

This matter is before the Court on Plaintiff's motion for a preliminary injunction. The Court has considered the motion and the opposition thereto and heard argument on July 1, 1997.

The Plaintiff, Serono Laboratories, Inc. ("Serono"), manufacturer of the infertility drug Pergonal, contends that the FDA erred in approving an abbreviated new drug application ("ANDA") which permits Ferring Laboratories to market Repronex, a generic alternative to Pergonal. The Court finds the FDA failed to comply with its governing statute and its own regulations in approving the ANDA. The Court also has certain questions related to the safety of Repronex. Based on this finding, the Court grants Plaintiffs motion for a preliminary injunction, barring the marketing of Repronex in the United States.

The FDA has the authority to regulate the manufacture, distribution, and sale of drugs within the United States pursuant to the Food, Drug, and Cosmetics Act ("FFDCA"). 21 U.S.C. § 355(a). Pharmaceutical companies seeking approval to market a new drug (also known as a "pioneer" drug) initiate this process by filing a new drug application ("NDA") with the FDA. A successful application will detail the numerous tests undertaken by the pharmaceutical company in its effort to demonstrate the safety and effectiveness of the candidate drug. 21 U.S.C. § 355(d). Only after the FDA approves an NDA can the applicant market the product in the United States. 21 U.S.C. § 355(a).

*31 In an attempt to encourage the marketing of generic alternatives to existing drugs, Congress enacted the Drug Price Competition and Patent Term Restoration Act of 1984 (also known as the "Hatch-Waxman Amendments" to the FFDCA). The Hatch-Waxman Amendments permit the submission of an ANDA when an applicant can demonstrate that a generic drug meets stringent requirements designed to ensure that it is as safe and effective as the pioneer drug. 21 U.S.C. § 355(j) (2) (A); 21 C.F.R. § 314.94.

Specifically, in order for an ANDA involving a drug product administered by injection to be approved, an applicant must demonstrate and the FDA must conclude that "the active ingredients [of the generic product] are the same as those of the [pioneer] drug." 21 U.S.C. § 355(j) (3) (C) (ii). The agency's implementing regulations state that "for determining the suitability of an abbreviated new drug application, the term `same as' means identical in active ingredients." 21 C.F.R. § 314.92(a) (1).

The Hatch-Waxman Amendments also require that an ANDA applicant demonstrate that a generic drug, is not unsafe "because of the type or quantity of inactive ingredients included ..." 21 U.S.C. § 355(j) (3) (H) (ii). In order to meet these stringent requirements, regulations mandate that the "FDA ... refuse to approve ... [an] abbreviated new drug application unless [the applicant drug] contains the same inactive ingredients ... in the same concentrations as the listed drug ..." 21 C.F.R. § 314.127(a) (8) (ii) (B). The agency's rationale for requiring equivalence with regard to inactive ingredients is based on its conclusion "that changing the inactive ingredients in a drug can adversely affect the drug's safety or effectiveness." 54 Fed.Reg. 2872, 28902 (July 11), 1989).

In 1969 the FDA approved an NDA for Pergonal, a drug marketed by the plaintiff, Serono. Pergonal is an injectable, intramuscular medication which treats female and male infertility. Its active ingredients,^[1] follicle stimulating hormone ("FSH") and leteunizing hormone ("LH"), are derived from the urine of post-menopausal women and make up five percent of the drug's overall chemical composition. Because the drug is crafted from mammalian urine, it is labeled a menotropins product. The remaining ninety-five percent of the drug contains lactose and uncharacterized, urinary protein ("UUP") impurities, both inactive ingredients.^[2] Certain of the UUP are biologically active and may have effects on individuals who undergo fertility therapy.

On June 21, 1990, Lederle Laboratories filed an ANDA with the FDA for approval of Repronex, a generic fertility drug based on Serono's Pergonal. This application was transferred to Ferring Laboratories ("Ferring") on July 3, 1996, and was subsequently approved by the FDA on January 30, 1997. In connection with its approval of the ANDA for the generic product, the FDA relied or Serono's Pergonal as the pioneer drug (or reference listed drug) for which pre-clinical and clinical studies had been undertaken, since it concluded that Repronex was "therapeutically similar" to Pergonal. On the basis of this approval, the FDA assigned an "AB" rating to Repronex in its publication entitled "Approved Drug Products with Therapeutic Equivalence Evaluations," commonly known as the "Orange Book." Physicians and pharmacists who rely upon this document in writing and filling prescriptions, consider the "AB" rating to mean that the generic product is fully substitutable for the pioneer drug.

On December 8, 1992, after learning that at least one ANDA for menotropins was being considered by the FDA's Office of Generic Drugs, Serono filed a citizen petition with *32 the FDA pursuant to the agency's procedures published at 21 C.F.R. § 10.30. The FDA regulations require the agency to respond to a citizen petition within 180 days of its submission. 21 C.F.R. § 10.30(e) (2). When no action was taken on Serono's petition for over four years, the Plaintiff on May 30, 1997 filed its complaint for declaratory and injunctive relief.^[3] Plaintiff's complaint challenging the approval of an ANDA for Repronex has four counts: (1) failure to establish that Repronex had the same active ingredients; (2) failure to establish that Repronex had the same inactive ingredients and composition; (3) inconsistent agency action; and (4) agency action unreasonably delayed and unlawfully withheld.^[4] Serono now moves for a preliminary injunction.

The agency argues that its approval of the ANDA for Repronex was consistent with the Hatch-Waxman Amendments and the applicable regulations. With respect to the active ingredients, the FDA argues that the active ingredients of Repronex are the "same" as those of Pergonal for the purposes of the FFDCA. With respect to the inactive ingredients, the agency believes that the Repronex ANDA should be judged by the standards in place at the time the application was submitted in 1990, and not the more stringent implementing regulations established in 1992.

Four factors must be considered in deciding whether to issue a preliminary injunction: (1) the likelihood of success on the merits; (2) irreparable harm to the plaintiff; (3) whether the issuance of a stay would substantially harm other interested parties; and (4) a determination of where the public interest lies. Washington Metro. Area Transit Comm'n v. Holiday Tours, Inc., 559 F.2d 841 843 (D.C.Cir.1977). A party need not establish all four factors in order to prevail. Rather "[i]f the arguments for one factor are particularly strong, an injunction may issue even if the arguments in other areas are rather weak." CityFed Fin. Corp. v. Office of Thrift Supervision, 58 F.3d 738, 747 (D.C.Cir.1995).

In evaluating the likelihood of Plaintiff's success on the merits, the Court must look to see whether the FDA's approval of the Ferring ANDA meets the specific criteria set out in the FFDCA. The Court's resolution of this case is governed by the principles of statutory construction enunciated in Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 841-46, 104 S. Ct. 2778, 2781-83, 81 L. Ed. 2d 694 (1984).^[5] As Chevron sets forth, "if the statute is silent or ambiguous with respect to [a] specific issue, the question for the court is whether the agency's answer is based on a permissible construction of the statute." Chevron, 467 U.S. at 843, 104 S. Ct. at 2782.

In this case, the Court finds that the FDA violated Section 505(j) of the FFDCA and its own implementing regulations when it approved Ferring's ANDA on January 30, 1997.

As described above, Section 505(j) (3) of the FFDCA states that the FDA may only approve *33 an ANDA when the information submitted is sufficient to show that "the active ingredients of the proposed generic product are the same as the active ingredients of the listed drug." 21 U.S.C. § 355(j) (3) (C) (ii). In its regulations implementing the Hatch-Waxman Amendments, the agency indicated that ANDAs must include information to show that the active ingredients in a proposed generic product are "the same as" those of the reference listed drug and that the agency will "refuse to approve an ANDA" otherwise. 21 C.F.R. § 314.127(a) (3) (ii). The regulations state further that "[f]or determining the suitability of an abbreviated new drug application, the term `same as' means identical in active ingredient(s)." 21 C.F.R. § 314.92(a) (1).

In applying these requirements to the immediate case, the Court must conclude that the FDA erred in approving the Ferring ANDA. The FDA scientists responsible for evaluating the chemical composition of a drug's active ingredients have, in two separate reviews, concluded that Repronex does not contain the same active ingredients as those in Pergonal.^[6] Instead of requiring that the active ingredients be identical, the FDA substituted a different criteria, requiring that the Ferring product pass the animal bioassay for potency developed by the U.S. Pharmacopoeia (USP) for determining whether menotropins products may be labeled "menotropins USP." Nothing in Section 505(j) permits an ANDA applicant to substitute USP animal assays for information demonstrating that the active ingredients of the proposed generic product are identical to those in the innovator product. What is more, the FDA had rejected the use of such a substitute test in 1993.^[7]

The FDA argues that the active ingredients of Repronex are the "same" as those of Pergonal for the purposes of the FFDCA. FSH, one of the active ingredients, is a polypeptide hormone consisting of two protein chains (forming a polypeptide "backbone") and a configuration of carbohydrate side-chains. A.R. at 402, 412. Natural variation in the carbohydrate elements of FSH a phenomenon known as "microheterogeneity" creates different "isoforms" of the hormone. A.R. at 412, 631, 724-43. While the government concedes that batches of Repronex exhibit variable isoform patterns compared with batches of Pergonal, it argues that any one batch of Pergonal would also contain different isoforms from any other batch of Pergonal. In essence, the FDA argues that the existence of varying isoforms does not affect the "sameness" of the product's active ingredients since "no evidence exists to suggest that isoform pattern impacts on clinical safety and efficacy." A.R. at 393.

It may well be that variations in isoforms are unavoidable and that there are no reasonable safety concerns with the generic product, as the government claims. But the FDA cannot selectively choose to reinterpret the FFDCA and its own implementing regulations in such an arbitrary manner. "Same" *34 means "identical," just as the agency's own regulations say and an agency must follow its own regulations and not arbitrarily reinterpret those regulations. The FDA violated Section 505(j) (2) (A) (ii) when it approved the Ferring ANDA because the active ingredients were not the "same."

Section 505(j) (3) of the FFDCA states that the FDA may not approve an ANDA if information in the application or any other information available to the agency shows that "(i) the inactive ingredients of the drug are unsafe for use ..., or (ii) the composition of the drug is unsafe ... because of the type or quantity of inactive ingredients included or the manner in which the inactive ingredients are included ..." 21 U.S.C. § 355(j) (3) (H). FDA's implementing regulations expressly prohibit approval of an ANDA unless the inactive ingredients in an injectable generic product are identical to and occur in the same concentrations as those in the innovator drug. In pertinent part, the regulations provide that:

"FDA will consider an inactive ingredient in, or the composition of, a drug product intended for parenteral use to be unsafe and will refuse to approve the abbreviated new drug application unless it contains the same inactive ingredients, other than preservatives, buffers, and antioxidants, in the same concentrations as the listed drug ..." 21 C.F.R. § 314.127(a) (8) (ii) (B).

As the FDA made clear in describing this requirement, the agency requires the same inactive ingredients because "[i]t is well-established that changing the inactive ingredients in a drug can adversely affect the drug's safety or effectiveness." 54 Fed.Reg. 28872, 28902 (July 10, 1989).

In this case, the FDA has consistently recognized that differences exist between the UUP in Pergonal and those in Repronex.^[8] But, rather than insist that the inactive ingredients in Repronex be the "same" as Pergonal, the FDA relied on certain animal studies to assure itself of the safety of the inactive ingredients in Ferring's generic product. There has been some suggestion that the animal tests raise a safety issue with respect to Repronex, a suggestion that the agency disputes.^[9] The agency has also sought to redefine the UUP as "impurities" rather than as inactive ingredients. The Court is perplexed by the agency's efforts to skirt the application of its own regulations.

Pergonal and Repronex also differ in concentration with respect to the other inactive ingredient, lactose. Repronex contains 20 milligrams of lactose per ampule, while Pergonal only contains 10 milligrams of lactose per ampule.

The FDA contends that the regulation that a general parenteral product contain the same inactive ingredients in the same concentration as the pioneer drug is inapplicable to the granting of an ANDA for Repronex. The Repronex ANDA was filed with the FDA in June 1990 and the regulations addressing inactive ingredients were not adopted until June 1992. This contention dumbfounds the court. It was the court's impression that the FDA's role was to protect the public. How is the public protected by allowing a new drug to come to market on *35 a more lenient basis than required by existing law? While the court understands Grandfather clauses, if one does exist in this case, they have no place where the public safety is involved.

Even assuming arguendo that the FDA need not apply its regulations on inactive ingredients to its consideration of the Repronex ANDA, it is uncertain whether the agency's decision meets the requirements of the Hatch-Waxman Amendments. The Court refers back to the FFDCA and its legislative history. See, generally, H.R.Rep. No. 857, 98th Cong. 1st Sess., pt. 1. at 36 (1984), reprinted in 1984 U.S.C.C.A.N. 2647, 2669 (explaining that ANDAs submitted before the effective date of the generic drug approval regulations "may be submitted in accordance with current regulations."). While the Hatch-Waxman Amendments of 1984 do not specifically require that the inactive ingredients of a generic drug be identical to those of the pioneer, they state that an ANDA must be denied if the drug's inactive ingredients make it unsafe. 21 U.S.C. § 355(j) (3) (H) (i).

The tentativeness of the agency's conclusion that "the difference in the amount of lactose present in [Repronex] does not raise serious questions of safety" is not reassuring. A.R. at 434 (emphasis added). The record makes clear that there has been debate within the FDA as to whether a full NDA should be required of Repronex.^[10] This is not an area where the FDA can cut comers. The FDA shall either require a NDA of Repronex or establish a fuller basis on the record as to why a NDA is unnecessary.

Serono will suffer irreparable injury in the absence of a preliminary injunction preventing Ferring from marketing Repronex in the United States. The analysis here is one of simple economics. A change in the number of manufacturers of a given drug product will change each manufacturer's market share, and those who lose market share are irreparably harmed. Allegran v. Shalala. 6 Food and Drug Rep. 389, 391, No. 94-1223 (D.D.C. Nov. 10, 1994). The FDA's approval of the Repronex ANDA would change the number of menotropins products in the market from two to three, with a corresponding loss of market share and revenue for Serono's Pergonal. In sum, Plaintiff represents that it expects "an annual loss of up to $50 million nearly a 25% drop from the company's total 1996 revenues ($207 million) for all products, causing irreparable and severe economic harm to Serono." Memorandum of Points and Authorities in Support of Plaintiff's Motion for a Preliminary Injunction, p. 38. Plaintiff also represents that it is in imminent danger of sustaining irreparable injury to the goodwill earned by its product during the past twenty-eight years.

The government argues that Plaintiff's fear of lost revenue and reputation is too speculative to constitute "irreparable harm" and falls short of the harm necessary to justify emergency injunctive relief. The government also claims that Plaintiff, a part of the multinational "Ares-Serono Group," can afford to endure economic loss while this matter is litigated and that such loss is not then "irreparable."

This Circuit has long used a flexible approach in weighing the preliminary injunction factors. See, e.g., Bracco Diagnostics, Inc. v. Shalala, 963 F. Supp. 20, 27 (D.D.C.1997) (preliminary injunction factors "must be viewed as a continuum") (citing Washington Metropolitan Area, Etc. v. Holiday Tours, Inc., 559 F.2d 841, 843 n. 3 (D.C.Cir.1977)). While Plaintiffs actual and reputational losses can only be speculative, it is clear to this Court that Plaintiff's economic losses, at the very least, will be substantial in *36 absolute terms.^[11] What is more any loss by Serono is not recoverable due to governmental immunity.^[12] Plaintiff will be irreparably harmed if a preliminary injunction is not granted in this case.

On balance, the public interest will be served in this case by granting Plaintiff's motion for a preliminary injunction. It is true that this injunction will delay the introduction into the market of a lower-priced generic drug. The Court recognizes that "[t]he purpose of [the Hatch-Waxman Amendments] was to increase competition in the drug industry by facilitating the approval of generic copies of drugs." Mead Johnson Pharmaceutical Group v. Bowen, 838 F.2d 1332, 1333 (D.C.Cir.1988).

"[T]here is a strong public interest in meticulous compliance with law by public officials." Fund for Animals v. Espy, 814 F. Supp. 142, 152 (D.D.C.1993). Here the FDA violated both the FFDCA and its own implementing regulations by approving a generic infertility product that, by the FDA's own admission, is not the same as the pioneer drug it seeks to copy. The FDA took such action in the face of information, studies and internal debate indicating that such differences between Pergonal and Repronex could present problems of safety and efficacy.

The public has a tremendous interest in ensuring that the FDA only approve generic drugs that are the same as the pioneer drugs they purport to copy. That is certainly true for patients undergoing fertility treatment with menotropins products. Because the reproductive potential of each woman declines with age, and this occurs quite rapidly, any lost cycle (usually of a month's duration) will contribute to the loss of reproductive potential if the fertility treatment is ineffective. FDA's finding of "therapeutic equivalence" and "AB" rating for Repronex could cause physicians to switch patients from Pergonal to Repronex in the middle of a fertility cycle. Yet, because the ingredients in Repronex are different from Pergonal, patients may lose the opportunity for conception and could be exposed to side effects resulting from administration of different menotropins products. See Plaintiff's Exh. Nos. 1 and 2.

While generic drugs provide the public with immediate cost savings, more is at stake. The U.S. has made tremendous strides in developing new medicines to cure the sick. To develop a new medicine requires the expenditure of a great deal of money and many years of testing to establish its safety and efficacy. The reasonable costs of developing new curative drugs must be able to be recouped by our research and pioneering pharmaceutical companies. The Congress certainly has every right to find that at times those companies may have gouged the public. To make certain allowances for the production of generic copies of "the real thing" without unduly cutting back on new drug research certainly will benefit the public. It is clear, however, for this policy to be effective there can be no short cuts. For the public safety, the generic drug must have the same therapeutic effect as the original. If this cannot be appropriately documented, the generic drug must not receive FDA approval without full testing as to the its safety and effectiveness. The record does not show that Repronex is a safe and effective equivalent of Pergonal. It will serve the *37 public interest to grant a preliminary injunction in this case.

Plaintiffs motion for a preliminary injunction will be granted. The Court will preliminarily enjoin the FDA from approving the Ferring ANDA and will compel the FDA to rescind immediately its designation of an "AB" rating in the Orange Book. An appropriate order is attached hereto.

This matter is before the Court on Plaintiffs motion for a preliminary injunction. The Court has considered the motion and the opposition thereto and heard argument on July 1, 1997. For the reasons cited in the accompanying Memorandum Opinion, it is hereby

ORDERED that Plaintiffs motion is granted; and it is

FURTHER ORDERED that, pending a trial of the permanent injunction, the FDA is preliminarily enjoined from approving the Ferring ANDA; and it is

FURTHER ORDERED that the FDA rescind immediately its designation of an "AB" rating for Repronex in the Orange Book.

[1] The FDA has defined "active ingredient" to mean "any component that is intended to furnish pharmacological activity or other direct effect in the diagnosis, cure, mitigation, treatment or prevention of disease, or to affect the structure or any function of the body of man or other animals The term includes those components that may undergo chemical change in the manufacture of the drug product and be present in the drug product in a modified form intended to furnish the specified activity or effect." 21 C.F.R. § 210.3(b) (7).

[2] The FDA has defined "inactive ingredient" to mean "any component of a drug other than an active ingredient." 21 C.F.R. § 210.3(b) (8).

[3] The FDA did answer Serono's citizen petition after this law suit was filed.

[4] The FDA challenges Plaintiffs standing to bring claims 3 and 4. The Court does not reach that issue, but grants Plaintiff's motion for preliminary injunction solely on the basis of counts 1 and 2.

[5] In Arent v. Shalala, 70 F.3d 610 (D.C.Cir.1995), the Court of Appeals discussed the distinction between Chevron analysis (under which no deference is owed to agency decisions that conflict with clear statutory provisions) and "arbitrary and capricious" review. Arent establishes that Chevron analysis applies here, just like other cases involving interpretations of the Hatch-Waxman Amendments. See, e.g., Abbott Laboratories v. Young, 920 F.2d 984, 987 (D.C.Cir.1990) (substituting Chevron analysis for the District Court's "arbitrary and capricious" review); Mova Pharmaceutical Corp. v. Shalala, 955 F. Supp. 128 (D.D.C.1997); and Inwood Laboratories, Inc. v. Young, 723 F. Supp. 1523 (D.D.C.1989).

[6] In March 1994, the FDA's scientists declared that two analyses have "clearly demonstrated that Serono Pergonal (reference) and Lederle [Ferring] menotropins are chemically different. This is a result of differences in their purification process of the active drug substance." Memorandum from Chien-Hua Niu, Ph.D., to Dir., Office of Generic Drugs, March 3, 1994, at 2. (Plaintiff's Exh. No. 8.). At the same time, the FDA scientists noted that "[p]lasma disappearance (i.e. biologic half-life of FSH and LH) may be affected by different FSH and LH isoform/glycosylation" and that no information on isoforms of FSH and LH was provided by the drug applicant." In light of the importance of the latter issue, FDA requested the applicant to provide the agency with a "comparative isoform study" and eventually concluded that the information submitted in response to that request was not sufficient for the purposes of compliance with the FFDCA. See Letter, Division of Metabolism and Endocrine Drug Products, Consult Review of Chemistry, Manufacturing, and Controls for Lederle Laboratory' Repronal, Oct. 21, 1996. (Plaintiff's Exh. No. 11).

[7] On May 27, 1993, FDA officials met with Lederle representatives to "establish a methodology for characterization of the bulk active drug substance which demonstrates that the Lederle and Serono Menotropins for Injection products are pharmaceutical equivalents." (5/27/93 Mtg. Minutes, Plaintiff's Exh. No. 10 at 2.) At that meeting, Lederle sought to persuade the FDA to evaluate the "pharmaceutical equivalence" of the two products based upon the USP bioassay for menotropins." Id., at 4. However, FDA implicitly rejected that proposal by requiring additional chemical testing.

[8] For example, in April 1994 the FDA's Division of Bioequivalence concluded that, in addition to differences between the active ingredients of the Serono and Ferring menotropins products, "differences are also observed in the inactive protein components and composition of the test product." Review of Analytical Data and Three Protocols, Addendum to Review of March 31, 1994, April 14, 1994 ("Protocol Review"), at 2. (Plaintiff's Exh. No. 15).

[9] In the animal study, the FDA required Ferring to conduct a study involving guinea pigs to address the issue of immunogenicity, or whether the product would cause allergic reactions. Lederle had earlier proposed a similar test for its product. The study of the Ferring menotropins product demonstrated that the menotropins "were consistently more antigenic" than the menotropins of Serono. Memorandum from Alex Jordan, Ph.D., to Office of Generic Drug Products, July 26, 1996 ("Jordan Mem.," Plaintiff's Exh. No. 18, at Summary). However, the FDA dismissed the results of the study, stating that the "clinical significance of human proteins being more or less antigenic in guinea pigs is difficult to ascertain ... The consensus was that the results from these tests have little clinical relevance." (Jordan Mem., Plaintiff's Exh. No. 18, at Summary).

[10] In an October 1993 memorandum to the Director of FDA's Division of Metabolism and Endocrine Drug Products, FDA's Associate Director for Science and Medical Affairs wrote that "[p]otential differences in the urinary proteins that comprise 95% of the drug substance may result from different extraction processes. In-house discussions suggest that these pharmaceutical equivalence differences raise a safety issue." See Plaintiff's Exh. No. 14.

[11] Indeed, Ferring itself essentially concedes that it would take away at least $10 million in Pergonal sales in the coming year if the granting of the ANDA is upheld. Ferring tells the Court that it will have $13.5 million in U.S. revenues during the first year of Repronex marketing, and that half of this ($6.75 million) will come from taking market share from Pergonal. (Ferring Mem. at 28.) Since the $6.75 million will come from pricing 40 per cent below Pergonal, id., according to Ferring Pergonal will lose a total of the $6.75 million plus the 40 per cent markdown, or a little over $10 million.

[12] See, e.g. Woerner v. United States Small Bus. Admin., 739 F. Supp. 641, 650 (D.D.C.1990) ("[Plaintiffs] claim, persuasively, irreparable injury because the government is immune from damage suits"); Independent Bankers Ass'n of America v. Conover [1984-85] Fed. Banking L. Rep. (CCH) ¶ 86,178, at 90,538 (M.D.Fla. 1985) ("where a defendant has sovereign immunity, monetary loss constitutes irreparable injury").