Amoco Production Co. v. Village of Gambell, 480 U.S. 531 (1987)	10, 27
Animal Legal Defense Fund, Inc. v. Glickman, 154 F. 3d 426 (D.C. Cir.1998)	7
Arrow Air, Inc. v. United States, 649 F. Supp. 993 (D.D.C. 1986)	30
*1 Bristol-Meyers Squibb Co. v. Shalala, 923 F. Supp. 212 (D.D.C.1996)	27,32,33
Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984)	12
Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402(1971)	10,23,26
CityFed Fin. Corp. v. OTS, 58 F. 3d 738 (D.C. Cir. 1995)	29
Dow Chemical, USA v. Consumer Product Safety Comm’n, 459 F. Supp. 378 (W.D.La.1978)	31
Federal Express Corp. v. Air Line Pilots Ass’n, 67 F. 3d 961 (D.C. Cir.1995)	4
FW/PBS, Inc. v. Dallas, 493 U.S. 215 (1990)	4
GTE Sylvania, Inc. v. Consumer Product Safety Comm’n, 404 F. Supp. 352 (D.Del. 1975)	31
Honeywell, Inc., v. Consumer Product Safety Comm’n, 582 F. Supp. 1072 (D.D.C.1984)	31
Lemon v. Kurtzman, 411 U.S. 192, 200-01(1973)	9
Lujan v. Defenders of Wildlife, 504 U.S.555(1992)	4,31
Malarkey-Taylor Assoc., Inc. v. Cellular Telecommunications Indus. Ass’n, 929 F. Supp. 473. 473, 478 (D.D.C. 1996)	31
Martin v. Occupational Safety and Health Review Comm’n, 499 U.S. 144(1991)	12
Mead Johnson Pharm. Group v. Brown, 655 F. Supp. 53 (D.D.C. 1986)	30
*Microwave Acquisition Corp. v. Federal Communications Commission, 145 F. 3d 1410 (D.C. Cir. 1998)	8
National Medical Center. v. Shalala, 43 F. 3d 691 (D.C.Cir.1997)	12
National Trust for Historic Preservation v. Dole, 828 F. 2d 776 (D.C. Cir.1987)	12
*Rollins v. U.S. Environmental Protection Agency, 937 F.2d 649 (D.C.Cir.1991)	13,14
San Antonio General Maintenance v. Abnor, 691 F. Supp. 1462(D.D.C.1987)	9,10,27
San Diego Gun Rights Committee v. Reno, 98 F. 3d 1121 (9th Cir. 1996)	64
Steel Co. v. Citizens for a Better Environment, 118 S.Ct. 1003 (1998)	4,5
Synthetic Organic Chemical Manufacturers Assoc. v. Secretary, Dep’t of Health & Human Services, 720 F. Supp. 1244 (W.D.La. 1989)	23
Telephone and Data Systems, Inc. v FCC, 19 F.3d 42 (D.C. Cir. 1994)	6
*Thomas Jefferson University v. Shalala, 512 U.S. 504(1994)	12,13,15
*Troy Corp v. Browner, 120 F. 3d 277 (D.C. Cir.1997)	10,13,23,26
University Medical Center of Southern Nevada v. Shalala, 173 F.3d 438 (D.C. Cir. 1999)	8
Valley Forger Christian College v. American United of Separation of Church and State, Inc., 454 U.S. 464, 472 (1975)	4
Varicon Int. Mgmt. v. OPM, 934 F. Supp. 440, 448 (D.D.C.1996)	30
Wisconsin Gas v. FERC, 758 F.2d 669 (D.C. Cir.1985)	27
Wyoming Outdoor Council v. United States Forest Service, 165 F. 3d 43, 52 (D.C. Cir.1999)	12,13,15

¹ Cases marked with an asterisk are cases chiefly relied upon. In addition to the cases marked herein, defendants also chiefly rely upon Pfizer, Inc. v. Shalala, 182 F.3d 975 (D.C. Cir. 1999), which was cited in the Defendant's previously filed Memorandum in Support of its Motion To Dismiss, and which memorandum is incorporated herein by reference.

5 U.S.C. § 706(2) (A)

10,30

REGULATIONS

61 Fed. Reg. 50449	9
63 Fed. Reg. 13418	29
60 Fed. Reg. 30435	15

______________________________________

UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF COLUMBIA

______________________________________

______________________________________

JIM J. TOZZI, et al.,
Plaintiffs,

v.

UNITED STATES DEPARTMENT OF HEALTH AND HUMAN SERVICES, et al.,
Defendants

______________________________________

Plaintiffs² have filed this lawsuit seeking to enjoin the defendants³ from including any listing in the Biennial Report on Carcinogens ("RoC"), "that identifies the chemical tetrachlorodibenzo-p-dioxin ("dioxin" or "TCDD") as a ‘known human carcinogen’ or ‘known to be carcinogenic to humans.’" See Plaintiff’s Proposed Order, attached to the Complaint. Plaintiff’s action is fatally flawed in a number of respects. First, none of the plaintiffs have standing to bring this action. Second, the RoC is not reviewable for two reasons. First, it is still, at present, unripe, non-final agency action over which this Court has no jurisdiction. Second, even once published it remains an merely an informational document that has no regulatory or other practical effect sufficient to render it reviewable under the APA. Defendant’s arguments as to why the plaintiffs listed in the original Complaint do not have standing, and as to the unripeness and unreviewability of the RoC are fully briefed in the Defendant’s Memorandum in Support of its Motion to Dismiss, filed on July 20, 1999, as well as in its Reply Brief, filed on August 30, 1999. The defendants respectfully refer the Court to those submissions and the exhibits attached thereto, and incorporate them fully herein by reference as support for their renewed Motion to Dismiss.

Because, as of November 29, 1999, the Court has permitted the addition of Brevet Industries and Brevet, Inc. as new plaintiffs, defendants will supplement herein their earlier arguments to specifically address the lack of standing of these new plaintiffs.

Plaintiffs’ action fails completely on the merits as well, and the Court should grant summary judgment in the defendants favors. Plaintiffs’ claim that a listing in the "Known To Be A Human Carcinogen" category cannot be supported by so-called mechanistic data³ or animal data is meritless. The striking premise of this claim is that the defendants do not know how to interpret and apply their own criteria by which substances are listed in the RoC as either "Known To Be A Human Carcinogen" or as "Reasonably Anticipated To Be A Human Carcinogen" or as Reasonably Anticipated To Be A Human Carcinogen." The defendants themselves have promulgated the criteria at issue, and their interpretation is in no way arbitrary or capricious.

Plaintiffs’ other claims, that the evidence in support of the dioxin listing was misleadingly presented, and was based on irrelevant and otherwise insufficient studies, are easily disposed of given the depth and breadth of the scientific review at issue here and the deferential standard of review used by this Court.

Moreover, the various plaintiffs’ are, for various different reasons, unable to show any irreparable injury sufficient to justify an injunction. Nor does a balancing of the various interests weigh in plaintiffs’ favor. Thus, plaintiffs’ claims must fail and the defendants are entitled to summary judgment.

As defendants have previously shown, the well-known elements of standing are threefold. First, plaintiffs must allege (and prove) "injury in fact," which must be "concrete and actual or imminent, not conjectural or hypothetical... Second, there must be causation - a fairly traceable connection between the plaintiff’s injury and the complained-of conduct of the defendant . . . And third, there must be redressability - a likelihood that the requested relief will redress the alleged injury." Steel Co. v. Citizens for a Better Environment, 118 S.Ct. 1003, 1016 (1998); see also Lujan v. Defenders of Wildlife, 504 U.S. 555, 560-61 (1992); Valley Forge Christian College v. American United of Separation of Church and State, Inc., 454 U.S. 464, 472 (1975).

It is the plaintiff’s burden to allege and prove standing. See FW/PBS, Inc,. V. Dallas, 493 U.S. 215, 132 (1990). Moreover, where, as here, plaintiff seeks only declaratory and injunctive relief, "there is a further requirement that [plaintiff] show a very significant possibility of future harm." San Diego Gun Rights Committee v. Reno, 98 F.3d 1121, 1126 (9^th Cir. 1996); see also Federal Express Corp. v. Air Line Pilots Ass’n, 67 F.3d 961, 963 (D.C. Cir. 1995) (standing requirements are "no less strict" when a party seeks declaratory judgment). Here, Brevet Industries and Brevet, Inc. ("Brevet") fail to meet any of the three necessary prongs of the standing analysis.

Brevet makes medical tubing connectors that use polyvinyl chloride ("PVC") plastic. These connectors are designed exclusively to connect with medical tubing made of PVC plastic. See Affidavit of Charles Brewer, at ¶ 4, 5. PVC-based plastic, when burned, apparently produces dioxin as an unwanted by-product. See Cmplt. Ex. 6A at ¶¶ 17, 19-20; see also R., Ex. 1, DBD at 2-1.

Brevet is concerned because the cities of San Francisco, Oakland and Berkeley, California have all, within the past 18 months, passed resolutions setting as a goal the elimination of PVC incineration and the reduction and/or elimination of PVC plastic use in the Bay area. See Amen. Cmplt. Ex. 5, 6, 6A. These resolutions, which are very similar to each other, have attached to them an appendix listing scientific publications or pronouncements supporting the carcinogenicity of dioxin. See id. Listed there, among some twenty other listings and publications or pronouncements supporting the carcinogenicity of dioxin. See id. Listed there, among some twenty other listings and publications, including the IARC Monograph, is a reference to the NTP Board of Scientific Counselors of NIEHS, dated 1997. Id.

These city resolutions may be harbingers of public opinion against dioxin and PVC plastic in the Bay area. Brevet may thus conceivably be engaged in an industry that has a limited future. Buy any conclusion as to whether, when or how much business Brevet will or may lose is wholly speculative. Brevet has not shown that it has or will suffer an "injury in fact [that is] . . . actual or imminent, not conjectural or hypothetical." Steel Co. v. Citizens for a Better Environment, 118 S.Ct. at 1016. None of the cited city resolutions have any binding effect. They are simply statements of intent and future goals. Brevet has not alleged that it has lost any business at present. Nor has it shown that it will suffer any "imminent" loss of business.

Exactly when local government resolutions, that currently have no regulatory "bite," will translate into actual lost business for Brevet is totally speculative. Indeed, the formal effect of the three city resolutions was to pass on their dioxin reduction goals to a regional task force on dioxin, the political authority and power of which is unclear, to say nothing of whether the stated dioxin reduction goals will ever become binding law. See Amen. Cmplt. Ex. 5 at 3, Ex, 6 at 3, Ex. 6A at 5. Which of Brevet’s customers will ever stop buying PVC plastic, when and how much is c9ompletely speculative. Thus, Brevet’s alleged injuries are conjectural and hypothetical.

Brevet also fails the causation prong of the standing analysis. Brevet has not shown that publication of the RoC will cause it any harm. First, publication of the RoC itself has no direct impact on Brevet; the RoC obviously does not prevent anyone from buying or incinerating Brevet’s products. Rather, Brevet’s argument is that publication of the RoC will cause local or other governmental bodies to pass laws reducing or eliminating the use of PVC plastics, or plastics buyers to stop buying its product. In the affidavit submitted along with plaintiffs’ amended Complaint, Charles Brewer states that Brevet will lose business if there are "regulatory initiatives" or "product deselection decisions" reducing the desire for PVC plastics. Brewer Aff. at p. 10.

While standing may be predicated on third party causation, there are no cases permitting the type of attenuated link proposed here by Brevet. In Telephone and Data Systems, Inc. v. FCC, 19 F.3d 42, 47 (D.C. Cir. 1994), the Circuit held that "injurious private conduct is fairly traceable to the administrative action contested in the suit if that action authorized the contested conduct or established its legality." See also Animal Legal Defense Fund, Inc. v. Glickman, 154 F.3d 426, 442 (D.C. Cir. 1998). Thus, government conduct that allows or permits third party conduct that injures the plaintiff may be said to have "caused" plaintiff’s harm, and therefore supports standing. Animal Legal Defense Fund, 154 F.3d at 443.

However, here, the question of "permitting" or "allowing" governmental bodies to pass laws is inapposite. It is absurd to think that any governmental body is waiting for permission or authority as a result of RoC publication to pass any law or ordinance whatsoever. Indeed, as plaintiffs have shown, local governments clearly have not waited for RoC publication to pass any law or ordinance whatsoever. Indeed, as plaintiffs have shown, local governments clearly have not waited for RoC publication to take action. Moreover, as to the three existing city resolutions cited, the simple fact that the NTP Board of Scientific Counselors was listed as one of over 29 references on a list appended to the end of the resolution hardly shows that NTP discussions or RoC publication did, will or could "cause" the ordinance to come into existence.

The same goes for any third party conduct by PVC plastic buyers as well. Marketplace decisions to buy PVC plastic are governed by a wide variety of factors beyond the purview of RoC publication. Indeed, attached to the Charles Brewer affidavit is an article indicating that Tenet Healthcare, a corporate operator of over 120 hospitals nationwide, announced recently that is will, on its own accord, seek to buy and use medical supplies that are not made from PVC plastic. This corporate decision was made voluntary and wholly independent of any action by the defendants or the RoC.⁴ Moreover, it was obviously made before the proposed dioxin listing has even been finalized or published in the RoC.

The decision by Tenet Healthcare therefore supports the defendants’ argument that marketplace events (like government regulations), are not "authorized," "permitted" or in any way caused by publication of the RoC. There is simply no way to link these occurrences to the RoC with a straight face. Brevet’s suggestion, that future marketplace decisions not to buy PVC plastics (if any) or future government action limiting PVC use (if any) can be linked to the RoC, is even more speculative.

This cause is similar to Microwave Acquisition Corp. v. Federal Communications Comm’n, 145 F.3d 1410, 1412 (D.C. Cir. 1998), wherein the Circuit Court held that plaintiff lacked standing because the alleged third party conduct was too attenuated to support standing. The plaintiff’s alleged loss in Microwave, as here, was attributable to a third party whose conduct and was not closely enough linked to the government action at issue.

Finally, Brevet’s alleged future harm could not be redressed by a decision of this Court in this case in any event. See University Medical Center of Southern Nevada v. Shalala, 173 F.3d 438, 441 (D.C. Cir. 1999). Even if the plaintiffs were to successfully convince this Court that the defendants have been arbitrary and capricious in the steps leading up to publication of the RoC, a declaratory judgment or an injunction against the RoC publication pending further review of dioxin would not solve Brevet’s problem.

It is apparent just from the list of citations appended to the three city resolutions referred to above that a great number of scientific and political groups have already taken positions supporting the containment and regulation of dioxin exposure. Even if a RoC re-review of dioxin were ordered by the Court, it is most unlikely that such a decision would alter the underlying groundswell of public opinion against dioxin. Thus, for all the above reasons, Brevet lacks standing to bring this action.

II. The Court Should Grant Summary Judgment In The Defendants’ Favor Because There are No Material Facts in Dispute And Defendants Are Entitled To Judgment As A Matter Of Law.

Before addressing in detail the particular claims plaintiffs advance, it is helpful to review the legal standards for consideration of injunctive relief. The criteria that govern the issuance of permanent injunctive relief. The criteria that govern the issuance of permanent injunctive relief in the District of Columbia are well-settled. Plaintiff’s burden is a high one. Injunctive relief is "an extraordinary and drastic remedy, not to be routinely granted and only to lie when the right to relief is clearly established." San Antonio General Maintenance v. Abnor, 691 F. Supp. 1462, 1467 (D.D.C. 1987). The decision to grant or deny injunctive relief is "quintessentially a discretionary act for the trial court." Id.; see also Lemon v. Kurtzman, 411 U.S. 192, 200-01 (1973).

Plaintiff must: (1) establish his right to relief by succeeding on the merits; (2) show that without injunctive relief it will suffer irreparable harm and that remedies available to it at law are inadequate, (3) show that in balancing hardships, the issuance of an injunction will not substantially harm other interested parties; and (4) show that the public interest favors the requested injunction. E.g., San Antonio General Maintenance, 691 F. Supp at 1467; Amoco Production Co. v. Village of Gambell, 480 U.S. 531, 547 n.12 (1987) (standards for a preliminary and a permanent injunction are essentially the same except that for a permanent injunction plaintiff must show "actual success," rather than a mere likelihood of success on the merits). This is an exacting standard that, as we explain below, plaintiff has not met.

Under the Administrative Procedure Act ("APA"), agency action will be upheld unless it is "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." 5 U.S.C. § 706(2) (A). The court will not "substitute its own judgment for that of the agency." Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 416 (1971); see also Troy Corp. v. Browner, 120 F.3d 277, 283 (D.C. Cir. 1997).

Plaintiffs four putative causes of action may be summarized as follows: First, plaintiffs’ primary allegation is that the defendants have misapplied their own criteria for listing dioxin in the RoC as a known human carcinogen, by relying on mechanistic and animal data. Amen. Cmplt. at ¶ ¶ 103-106. Second, plaintiffs allege that, during the ongoing review process for the dioxin listing in the RoC, defendants have presented evidence of dioxin’s carcinogenicity in a misleading manner to the various scientific review committees. Id. at ¶ ¶ 107-108. Third and fourth, plaintiffs allege that defendants have relied on otherwise improper and insufficient evidence upon which to list dioxin as a known human carcinogen. Id. at ¶ ¶ 109-114.

The September 30, 1997 Draft Background Document for dioxin outlines and summarizes the scientific findings that support the proposed dioxin listing. In summary, these findings indicate that the dioxin listing as a "known" human carcinogen is supported by (1) human studies that found "an association between dioxin exposure and cancer mortality with respect to all cancers combined, non-Hodgkin’s lymphoma, and lung cancer;" (2) "studies in experimental animals that have shown that TCDD induces benign and malignant neoplasms at multiple tissue sites in multiple species;" and (3) a "compelling body of evidence [that] indicates a basic similarity in the mechanism of induction of animal and human biochemical and toxicological responses to TCDD at comparable doses and tissue levels." See R., Ex. 1, DBD at RC-1.

Plaintiffs assert that the Draft Background Document’s summary of three sources that support the listing of dioxin in the "known To Be A Human Carcinogen" category, ss opposed to simply and solely basing that listing on human epidemiological evidence, violates the plain meaning of, and is a misapplication of, the listing criteria. Plaintiffs’ interpretation of the defendants’ listing criteria is simply erroneous.

Before explicitly discussing the listing criteria, it is helpful to point out the applicable standard by which the Court evaluates an agency’s interpretation of its own rules and regulations. This standard of review applies to the listing criteria here, and it is particularly deferential.

The Supreme Court has made it clear that a court reviews with "substantial deference" an agency’s interpretation of its own regulations. Thomas Jefferson University v. Shalala, 512 U.S. 504, 512 (1994); see also Martin v. Occupational Safety and Health Review Comm’n, 499 U.S. 144, 150-51 (1991). Indeed, the defendants’ interpretation of their own listing criteria here "is controlling unless it is plainly erroneous or inconsistent with the [criteria]." Wyoming Outdoor Council v, United States Forest Service, 165 F.3d 43, 52 (D.C. Cir. 1999) (internal quotations omitted and emphasis added); see also National Trust for Historic Preservation v. Dole, 828 F.2d 776, 782 (D.C. Cir. 1987). Here, where the agency itself has drafted the rule or regulation at issue, not only is the Court’s review deferential, it is "more deferential . . . than that afforded under Chevron."⁵ National Medical Center v. Shalala, 43 F.3d 691, 697 (D.C. Cir. 1997). Indeed, the Supreme Court has made it clear that "broad deference is all the more warranted when...the regulation concerns ‘a complex and highly technical regulatory program.’" Thomas Jefferson, supra, 512 U.S. at 512. The scientific review of toxic substances is such a program. Cf. Troy, 120 F.3d at 283 (the Court reviews scientific judgements of an agency "‘not as the chemist, biologist or statistician that (it is) qualified neither by training nor experience to be, but as a reviewing court exercising (its) narrowly defined duty of holding agencies to certain minimal standards of rationality.’").

Indeed, a court "need not find that the agency’s construction is the only possible one, or even the one that the court would have adopted in the first instance." Wyoming Outdoor Council, 165 F.3d at 52. In Rollins v. U.S. Environmental Protection Agency, 937 F.2d 649, 651 (D.C. Cir. 1991), our Court of Appeals sustained an EPA interpretation of its own regulation that the court believed "would not exactly leap out at even the most astute reader," and was "rather more restrained" than the plaintiff’s reading of the regulation. The court upheld the interpretation because it was consistent with the regulation and, "in a competition between possible meanings of regulation, the agency’s choice receives substantial deference." Id. Here, in any event, the defendant’s interpretation of their own listing criteria is eminently reasonable.

First, it is clear that the plain meaning of the listing criteria permits evaluation of both mechanistic and animal data in considering the carcinogenic status of any given substance. The final paragraph of the criteria state that "Conclusions regarding carcinogenicity in humans or experimental animals are based on scientific judgment, with consideration given to all relevant information." R., Ex. 2 at 2 (emphasis added). Clearly both animal and mechanistic data concerning the listed substance are "relevant data."

Plaintiff’s argument is centered on a superficial and mistaken understanding of the specific language of the "Known To Be A Human Carcinogen" category. That language states that, "There is sufficient evidence of carcinogenicity from studies in humans which indicates a causal relationship between exposure to the agent, substance or mixture and human cancer." Plaintiff is apparently complaining that the listing summary in the Draft Background Document mentions animal data as supporting the "known" listing for dioxin, and that animal data is not mentioned as a data source in the language for the "known" listing criteria.

But the fact that animal data is referred to in the Draft Background Document as support for any listing of a "known" carcinogen, including dioxin, does not mean that the listing is based upon animal data, as opposed to "studies in humans." Rather, the animal data provides, in the case of dioxin as well as in all other "known" listings, additional support for the listing. This point is made perfectly clear by the fact that virtually all listings in the "known" category in the published Eight RoC make reference to animal data in the Summary section. See R., Ex. 2, Eight RoC, at 15-49. Indeed, most "known" listing summaries in the Eight RoC actually begin by referencing the animal data. This makes eminent sense, since most experimental studies on toxic substances begin with animals before progressing to humans or mechanistic data. The fact that further studies in humans have supported a listing in the "known" category does nothing to invalidate the animal studies. It is therefore wholly reasonable to refer to the animal studies in the listing summaries. See Thomas Jefferson v. Shalala, 512 U.S. at 512.

Plaintiffs also assert that the defendants have violated their own listing criteria by evaluating mechanistic data in connection with the dioxin listing. Mechanistic data in humans not only provides additional support for the dioxin listing, the listing is based in part on human mechanistic data. This is so because the "known" listing language "from studies in humans," ma reasonably be interpreted to mean "from human studies." Human studies may reasonably be interpreted to mean "from human studies." Human studies may reasonably be interpreted to include human mechanistic studies, that is, studies that show the mechanism by which dioxin works on and in human cells, and the connection such a mechanism has to cancer. Such an interpretation is eminently reasonable. E.g., Wyoming Outdoor Council, 165 F.3d at 52 (agency’s reasonable interpretation of its regulations is controlling); see also Rollins, 937 F.2d at 651.

In order to avoid the clear language in the criteria’s final descriptive paragraph permitting "scientific" judgment" on the carcinogenicity of any substance to be based on "all relevant information . . . includ[ing] . . . data relating to mechanism of action," plaintiffs are forced to argue that the final descriptive paragraph applies only to the "Reasonably Anticipated To be A Human Carcinogen" listing. But this argument contravenes all the administrative history leading up to the promulgation of the revised criteria in 1996.

Revisions to the previous 1982 listing criteria were first proposed at an April 24-25, 1995 meeting of the NTP Board of Scientific Counselors Ad Hoc Working Group. See Ex. 5, at 2; see also R., Ex. 8 at 3 (Summary Minutes referring to the April 24, 1995 meeting as the "first step in the criteria review process). At that first meeting, one of the primary topics of discussion was the possible "incorporation of mechanistic data as part of the criteria for listing substances in future Reports . . .". Ex. 5, at 2 (emphasis added). Individual "break-out" groups were assembled and each group commented on the topic. See generally Ex. 5.

Following the publication of the proposed revisions in the Federal Register on June 8, 1995, (R., Ex. 6), these comments were then presented and discussed at a June 29, 1995 NTP Board meeting. See R., Ex. 7 at 3-9. Following discussion and public comment, it was "moved that mechanistic information should be included in the selection process for agents to be listed in the [RoC]." Id at 7 (emphasis added). The motion passed unanimously. Id. Nowhere during the discussions of mechanistic data was it agreed or decided that such data should be used only for the "Reasonably Anticipated" and not for the "Known category.

The references to mechanistic data were placed in an explanatory paragraph that clearly applied to both listing categories. This is made plain by the fact that the descriptive or explanatory paragraph was, in some interim drafts, placed above the specific language for the two listing categories. See Ex. 7, last page; see also Ex. 8, at 4, the Minutes of the NTP Executive Committee Meeting on July 27, 1995, summarizing what had been done up to that point, show that the NTP Board had approved a resolution to place the "explanatory paragraph" at the beginning of the criteria. Ultimately the final paragraph was placed below the two listing categories. See Ex. 5, at 6; and see final criteria in R., Ex. 2, Eighth RoC at 2. If the explanatory paragraph had only been intended to refer to the second, "reasonably anticipated" category, it would never have been suggested to lace it above both listing categories. That simply would have made no sense.

After the explanatory paragraph had been moved back to the bottom again, the minutes of the January 26, 1996 NTP Executive Committee Meeting clearly state that "Dr. Jameson then read a final paragraph which applies to all the criteria and discusses the role of scientific judgment, and other relevant information . . .". R., Ex. 11 at 8 (first page); see also R., Ex. 12 at 3-5.

Moreover, the original proposed criteria show the explanatory paragraph with wider margins than either of the two specific listing categories, indicating by that format that they apply to both. See Ex. 5, at 6.

Not only does the administrative history, and logic,⁶ dictate that mechanistic data be used for the "Known" category, but the defendants’ own statements about the revised criteria following their publication show, with one minor exception, that defendants understood that mechanistic data was to be considered under the "Known" category. When the final, revised criteria were published in the Federal Register on September 26, 1996, the accompanying text there provided a brief summary of the review process. It stated that at "each step of the review process there was concurrence with the following points. . .(2) mechanistic information should be used as part of the listing criteria." See R., Ex. 13 at 2, 61 Fed. Reg. 50499. The Federal Register text also notes just above a printing of the descriptive paragraph containing the reference to mechanistic data that "the following descriptive paragraph has been added to the criteria." Id. Nowhere does the Federal Register text differentiate between the criteria for the "Known" category and the "Reasonably Anticipated" category. Thus, the descriptive paragraph reasonably applies to both categories.

Plaintiff points, however, not to the Federal Register Notice but to a staff authored press release in the "HHS News," entitled "Updated Criteria Approved For Anticipated Human Carcinogens." See Amen. Cmplt. Ex. 2. This press release did state that the criteria had been updated and mechanistic data was included for use in the "Reasonably Anticipated" category. It also stated, confusingly, that "the original criteria for listing a substance as a known human carcinogen remain unchanged . . ." Id. at 2. This second statement, which was not a quote from, or attributed to, anyone connected to the program, is both factually incorrect and misleading. It is factually incorrect because the actual literal text of the "Known" category did change from the previous version. See R., Ex. 13 at 2. It is also clearly misleading to the extent that it implies that the final descriptive paragraph does not apply to the "Known" category.

While the statement is misleading, it may be understood as referring solely to the text of the "Known" category itself. Even administrative references prior to the adoption of the revised criteria indicate that there were few or modest changes to the text of the "Known" category. E.g., R., Ex. 12 at 4. An NIEHS newsletter put out at the same time as the revised criteria makes the same point. See Amen. Cmplt. Ex. 3. On one hand the newsletter states that the "major change in the BRC is that the . . . NTP . . . will consider all relevant information - including mechanistic data - in determining whether to list a substance." See Amen. Cmplt. Ex. 3 (emphasis added). The newsletter did not state that mechanistic data could be used only to list a substance in the "Reasonably Anticipated" category. On the other hand the newsletter states that, with regard to the "Known" category, the revised criteria "are substantively unchanged," without referring to the final descriptive paragraph, and referring simply to the language under that heading.

The emphasis on the "Reasonably Anticipated" category makes sense in light of the opinion at the time that consideration of mechanistic data would have the biggest impact in connection with the "Reasonably Anticipated" category. Indeed, the NIEHS newsletter goes on to say that "[m]ost substances listed in the BRC are contained in the reasonably anticipated to be a human carcinogen’. . ." Id.

It is in this context that one should view the press release reference in the HHS News as well, i.e., it refers, albeit misleadingly, simply to the language of the category and does not make reference to the final descriptive paragraph. Dr. George Lucier, Director of the Environmental Toxicology Program at NIEHS and Associate Director of NTP, states that such statements "indicating that there was little or no change in the known human carcinogen’ category after publication of the criteria in September, 1996 related to the precise wording of the category only." See R., Ex. 30 at 2.

Following adoption of the revised criteria, the defendants have consistently applied its intended meaning, i.e., the final descriptive paragraph, with is reference to mechanistic data, applies to both categories. Shortly after the revised criteria were published, at the October 30, 1997 NTP Board Meeting, Dr. Lucier clearly stated that mechanistic data applied to both categories. See R., Ex. 16 at 7. He made similar comments in December, 1998, one year later, and noted specifically that mechanistic data "impacts both categories." See R., Ex. 25 at 10. Thus, it is clear that mechanistic data is properly used during considerations for listing in the "Known" category.

Plaintiffs also argue, however, that mechanistic data cannot support a finding or "sufficient evidence" of a "causal relationship," which is necessary in order to list a substance under the "Known" category, if the epidemiological evidence is less than sufficient by itself. This is belied by the very words of the explanatory paragraph, which states that "Conclusions regarding carcinogenicity in humans . . . are based on scientific judgment, with consideration given to all relevant information . . . includ[ing] . . . data relating to mechanism of action." R., Ex. 1 at LC-1. The ultimate decision is, obviously, left to "scientific judgment," and there is nothing to preclude NTP’s scientific judgment from finding that epidemiological evidence in conjunction with human mechanistic data may together rise to the necessary level of "sufficient evidence" indicating a "causal relationship."⁷

Moreover, although the explicit words of the final, descriptive paragraph by themselves preclude plaintiffs’ argument on this point, administrative history also explicitly supports the defendants’ interpretation. The Minutes of the June 29, 1995 NTP Board Meeting note that the explicit issue that plaintiffs now raise was raised then. The Minutes show that,

Dr. Lucier stated that a point of discussion might be whether there could be compelling mechanistic data that would sallow a chemical to be classified a human carcinogen even though it may lack good epidemiological data. Dr. Carl Barrett, NIEHS Scientific Director, contended that a chemical or agent could be placed in Category 1, lacking convincing epidemiological evidence, if there was as consensus of experts that available mechanistic data strongly supported the R., Ex. 7 at 5.

This exchange clearly shows that drafters of the revised criteria intended that mechanistic data could be used to support a finding in the "Known" category even where the traditional epidemiological evidence was less than "sufficient"."⁸ Thus, it is clear both that (a) human mechanistic data may be properly used to consider a listing under the "Known" category, and (b) that such data may, if the evidence supports it, be combined with limited human epidemiological data to jointly support a finding a "sufficient evidence" of as "causal relationship" between the agent and cancer in humans.

That is exactly what the proposed listing for dioxin does. The epidemiological evidence is stated as showing an "association" between cancer and dioxin. R., Ex. 1 at RC-1. The proposed listing goes on to state that this epidemiological evidence showed that industrial exposure to dioxin in the cohorts studied indicated an increased risk for all cancers in general, rather than for any one particular type of cancer. Id. at ¶3.0, p. 3-1, 3-2. The Draft Background Document goes on to show, however, that there is "scientific consensus" relating to the mechanism of action data for dioxin. In particular, the mechanistic studies indicate "that Ah receptor activation is required for the carcinogenicity of [dioxin]." Id. at RC-2 and ¶ 7.1. Thus, in the defendants’ scientific judgment, whose job it is to make exactly these types of toxicological judgments, this information together properly supports a listing of dioxin in the "Known" category.

Plaintiffs assert that the defendants did not present the evidence supporting the dioxin listing accurately and that the evidence does not support the listing. Unlike plaintiffs’ first claim, which was largely procedural, these claims assert infirmities with the actual scientific support for the dioxin listing. The claims fail.

As defendants have already noted, when reviewing scientific judgments made by an agency charged with making such judgments, this court will not "substitute its own judgment for that of the agency." Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. at 416; Troy Corp. v. Browner, 120 F.3d at 283; see also Synthetic Organic Chemical Manufacturers Assoc. v. Secretary, Dept. of Health & Human Services, 720 F. Supp. 1244, 1250 (W.D.La. 1989). Indeed, the Court reviews scientific judgments of an agency "‘not as the chemist, biologist or statistician that [it is] qualified neither by training nor experience to be, but as a reviewing court exercising [its] narrowly defined duty of holding agencies to certain minimal standards of rationality.’". Troy, 120 F3d at 283.

Here, plaintiffs have argued that the following DBD statement is inaccurate: "In the highly exposed industrial sub-cohorts, a causal relationship between TCDD exposure and mortality from all cancers combined was noted, but there was less strong evidence for cancers of any particular site." R., Ex. 1 at RC-1. That reference is attributed to the IARC monograph at pages 192-93. See id. at 3-2. The IARC reference clearly states that two of the studies reviewed by IARC provided evidence "supporting a causal relationship between [dioxin] exposure and mortality from all cancers." Id. at 193. Thus, the reference is not inaccurate.

Plaintiffs argue that the DBD improperly presented this statement as IARC’s overall finding of the value of the epidemiological evidence. But this simply isn’t so. The DBD plainly states in the first sentence in section 3.0, dealing with epidemiological evidence that, "IARC (1997) stated that there is limited evidence from epidemiology studies of humans . . ." Id. at 3-1. Thus, plaintiffs’ argument fails.

Plaintiffs also assert that the defendants should not have relied on two studies authored in 1997 and 1998 by Bertazzi et al. These studies follow, and periodically update, the toxicological effects on a population exposed to dioxin as a result of a factory explosion in Seveso, Italy. See id. Supplement at 2. Plaintiffs have argued that because IARC did consider an earlier, 1996 version of the Bertazzi studies, that IARC therefore somehow "considered" all the subsequent Bertazzi follow-up studies. This is simply not true. The bibliography to the IARC monograph shows that the 1997 and 1998 Bertazzi studies were not considered. See id. at 531.

Moreover, plaintiffs’ argument that the Bertazzi studies, both new and old, do not support the dioxin "Known" listing must also fail. Plaintiffs cite to various submissions by, as well as an oral presentation by, Dr. Raymond Greenberg at the October 31, 1997 NTP peer review meeting, all of which criticize the Bertazzi studies.⁹ However, at the October, 1997 meeting at which Dr. Greenberg spoke, the assembled scientists voted to upgrade dioxin, despite Dr. Greenberg’s comments. See R., Ex. 16 at 339-40.¹⁰ Thus, Dr. Greenberg’s views simply did not carry the day.

Plaintiffs also assert that the proposed dioxin listing inappropriately relies on studies of exposure to dioxin mixtures, as opposed to 2, 3, 7, 8-TCDD itself, i.e., "pure" dioxin. But the defendants have made it clear that they "placed greater weight on those studies which have used TCDD blood levels as the exposure index rather than those which use the workplace situation." See R., Ex. 31. Thus, the proposed dioxin listing has a valid scientific basis, and plaintiffs’ efforts to criticize it must fail.

As a general matter, the transcripts of the both the October 31, 1997 meeting, at which the sub-committee voted 4-3 to upgrade dioxin, ss well as the December 2, 1998 re-review meeting, at which the sub-committee voted 7-5 to keep dioxin in the reasonably anticipated category, are enlightening. See R., Exs. 16 & 25. It is clear from reading these transcripts that the vote on dioxin was a close one and was hotly contested.

But the fact that the scientific community had an intense debate over the strength of the evidence supporting dioxin’s listing in the "known" category is precisely the point. It is not for the Court to jump in sand second guess the decision of reasonable scientists who had a fierce debate over the merits of the scientific issues. Following the public meetings and debate, the NTP Executive Committee voted 8 to 1 to list dioxin as s "known" carcinogen, and that recommendation, as well as the recommendations of the other review groups, is presently pending before Dr. Olden, Director, NTP. There is no basis for second guessing this decision. Citizens to Preserve Overton Park, Inc., 401 U.S. at 416; Troy Corp. v. Browner, 120 F.3d at 283.

Quite separate and apart from plaintiff’s lack of likelihood of success on the merits, is the plain fact that plaintiffs have proffered only the most speculative type of economic harm. For this reason alone plaintiffs’ motion for a preliminary injunction should fail.¹¹

In order to qualify for either preliminary or permanent injunctive relief, plaintiffs must show that they will suffer irreparable injury in the absence of such relief. E.g., San Antonio General Maintenance, 691 F.Supp. at 1467; Amoco Production Co., 480 U.S. at 547 n.12. But "[i]rreparability of injury is a very high standard." Bristol-Meyers Squibb Co. v. Shalala, 923 F.Supp. 212, 220 (D.D.C. 1996) (preliminary injunction case, internal citation omitted).

Here, the original plaintiffs have asserted that the proposed agency action they seek to enjoin will create public confusion over the safety of the food that plaintiffs’ serve, leading to a food scare and loss of business. This claim of harm fails for two basic reasons; it is speculative, and even if it did occur, it would be insufficient to constitute "irreparable" harm and could not justify injunctive relief.

First, the claimed harm is speculative. As our Court of Appeals explained, to constitute irreparable harm, the injury must be certain and great; it must be actual and not theoretical. Wisconsin Gas Co. v. FERC, 758 F.2d 669, 674 (D.C. Cir. 1985). Injunctive relief will not be granted against something merely feared as liable to occur at some indefinite time; rather, the movant must substantiate the claim that the irreparable injury is likely to occur:

Bare allegations of what is likely to occur are of no value since the court must decide whether the harm will in fact occur. The movant must provide proof that the harm has occurred in the past and is likely to occur again, or proof indicating the harm is certain to occur in the near future. Further, the movant must show that the alleged harm will directly result from the action which the movant seeks to enjoin.

Wisconsin Gas, 758 F.2d at 674.

As defendants have previously argued at some length,¹² there simply is no support for the claim that, were dioxin to be listed and published in the RoC under the "Known To be A Human Carcinogen" category, the general public will swear off meats, fish, and dairy products as part of a "food scare." The carcinogenicity of dioxin has already been well publicized and plaintiffs have suffered no harm as of yet. Sas plaintiffs well know, IARC, a highly respected subgroup of the World Health Organization, in a well publicized monograph, has already listed dioxin in its highest category as "carcinogenic to humans." IARC monograph, Ex. 3, at 343. Similarly, large metropolitan areas, such as Berkeley, Oakland and San Francisco, California, have passed ordinances declaring dioxin to be the "single most carcinogenic substance known to science." Amen. Cmplt. Exs. 5, 6 & 6A. Plaintiffs have suffered no injury yet, and provide no basis for believing that anything different will happen following publication of the RoC, should it ultimately list dioxin under the "Known" category.

Moreover, plaintiffs’ comparisons to reports concerning dioxin-contaminated sources for Belgian livestock and poultry are inapposite. Plaintiffs’ Memorandum in Support of Preliminary Injunction at 34. Such reports relate to specific high level contaminations and specific foods. The proposed RoC listing for dioxin relates to no particular use for or source of dioxin. Nor could consumers really avoid dioxin altogether in any event, given its ubiquity. See R., Ex. 1, ¶2.3, p. 2-2. (Stating dioxin most likely spread to soil through "atmospheric fallout"). Thus, plaintiffs’ asserted "food scare" is totally speculative.

Second, even if there were, as plaintiffs argue, "some degree of consumer aversion" to meat, fish and poultry as a result of the RoC, such aversion is completely inadequate to show irreparable injury. "It is well settled that economic loss does not, in and of itself, constitute irreparable harm . . . ‘The key word in this consideration is irreparable. Mere injuries, however substantial in terms of money, time and energy necessarily expended in the absence of a stay are not enough.’" Bristol-Myers, 923 F.Supp. at 220 (quoting Wisconsin Gas, 758 F.2d at 774).

Courts apply "[d]emanding scrutiny" to claims of irreparable injury. Bristol-Myers, 923 F.Supp. at 220. In CityFed Fin. Corp. v. OTS, 58 F.3d 738, 747 (D.C. Cir. 1995), our Court of Appeals affirmed the denial of injunctive relief even when the defendant Office of Thrift Supervision had completely frozen plaintiff’s assets, and in the face of plaintiffs’ claim of bankruptcy, because the court found that plaintiff could apply for "hardship" relief.

Plaintiffs argue that they have shown irreparable injury because they will be unable to recover money damages from the government. But this fact alone does not obviate plaintiffs’ need to show "severe economic impact." Bristol-Myers, 923 F.Supp. at 221 (quoting WMATC v. Holiday Tours, Inc., 559 F.2d 841, (D.C. Cir. 1977)). Plaintiffs have failed to show anything close to sever economic impact. E.g., Varicon Int. Mgmt. V. OPM, 934 F.Supp. 440, 448 (D.D.C. 1996) (ten percent loss of business insufficient to show irreparable harm); Arrow Air, Inc v. United States, 649 F. Supp. 993 (D.D.C. 1986) (twenty-five percent of gross revenue insufficient to show irreparable harm); Mead Johnson Pharm. Group v. Bowen, 655 F. Supp. 53 (D.D.C. 1986) (twenty to thirty percent loss of market share not show irreparable harm). Rather, plaintiffs’ claim of a "food scare" is speculative, and there is nothing to suggest that plaintiffs would lose more than a de minimis amount of business as a result of the dioxin listing in the RoC. See Bristol-Myers, 923 F.Supp. at 221.

Plaintiffs understandably seek to avoid the requirement that they show significant economic impact by asserting that publication of the RoC would damage their business reputation. Such assertions are completely inapposite here, however, because any link between the plaintiffs’ restaurants and publication of the RoC is totally speculative. The cases plaintiffs cite in their Memorandum in Support of a Preliminary Injunction relate to release of information that would be specifically damaging to the plaintiff’s business. E.g., Honeywell, Inc. v. Consumer Product Safety Comm’n, 582 F.Supp. 1072, 1073 (D.D.C. 1984) (manufacturer of gas control products sought to prevent document describing its product as a hazard); GTE Sylvania, Inc. v. Consumer Product Safety Comm’n, 404 F. Supp. 352, 378 (D.Del. 1975) (TV manufacturers sought to prevent release of accident, safety and trade secret information about TVs); Dow Chemical, USA v. Consumer Product Safety Comm’n, 459 F. Supp. 378 294 (W.D.La.1978) (chemical manufacturers sought to prevent certain classification of their own chemical products). Here, however, plaintiffs have failed to show why their reputations would in any way be damaged by the RoC, which neither refers to any of the plaintiffs of their businesses or to any product that they sell.

Moreover, plaintiffs reference to Malarkey-Taylor Assoc., Inc. v. Cellular Telecommunications Indus. Ass’n, 929 F. Supp. 473, 478 (D.D.C. 1996), for the proposition that business reputation may be damaged by "consumer confusion" is equally inapposite. Malarkey-Taylor was a trademark infringement case where the issue of confusion between products was the central issue in the case. No such issue is presented here.

There is no support in any of the cases cited by plaintiffs that would serve to obviate the requirement, which plaintiffs plainly cannot meet, that they show significant economic damage prior to issuance of an injunction. Moreover, plaintiffs have failed to show that publication of the dioxin listing in the RoC would result in any damage to their business reputation.

Plaintiff Brevet has also failed to show irreparable injury. While it has alleged that it stands to lose a large percentage of its business if its customers stop buying its products, it has not shown, in anything other than a completely speculative way, that anything the defendants may do will cause this to happen. The fact that Brevet fears a substantial loss of business cannot simply be laid at the feet of RoC publication. Such allegations of injury are in no way certain; they are merely theoretical. In Wisconsin Gas Co. v. FERC, 758 F.2d at 674, the Circuit Court made it clear that injunctive relief will not be granted against something merely feared as liable to occur at some indefinite time. These are precisely the type of fears, unconnected to RoC publication, that Brevet has alleged. Thus, none of the plaintiffs have shown irreparable injury.

As part of its injunctive relief analysis, the Court must balance the relative harms to the parties of issuing an injunction. See Bristol-Myers, 923 F.Supp. at 221. Here, were an injunction to issue, likely the entire RoC might have to be delayed, given the uncertain space allocation, pagination and other printing requirements relating to the dioxin portion of the report. Withholding the entire RoC based on the speculative claims proffered herein presents a greater harm to the defendants and the public at large than publication would case to the plaintiffs, given the large public expenditure of money thus far and the speculative nature of plaintiffs’ claims.

Plaintiff has failed to establish that the public interest would be better served by issuing an injunction. The RoC is Congressionally mandated, and provides a necessary and important public service. It contains a list of substances that may pose a potential hazard to human health. The RoC is an informational, public health document. Its publication in a timely fashion, after full and complete scientific deliberation should not be delayed based upon speculative arguments claiming that it will result in a "food scare" or decreased PVC use.

The instant case should be dismissed because none of the plaintiffs have standing and the RoC is not ripe for review. Moreover, defendants are entitled t summary judgment as a matter of law because they have acted reasonably, not arbitrarily or capriciously, in interpreting their own listing criteria and in applying the evidence in support of the dioxin listing. Plaintiffs have not and cannot show that they are entitled to the drastic, injunctive and declaratory relief they seek. Thus, the case should be dismissed.

Respectfully Submitted,

____________________________________
WILMA A. LEWIS, DC BAR #358637
United States Attorney

____________________________________
MARK E. NAGLE, DC Bar #416364
Chief, Civil Division

____________________________________
DAVID L. SMITH, DC Bar #453709
Assistant United States Attorney
555 4^th Street N.W. Room 10-452
Washington, D.C. 20001
202/353-9895

Of Counsel:
Patricia A. Kvochak, Esq.
Office of General Counsel
National Institutes of Health

² Plaintiffs in the original action are Beduci, a Washington, D.C. restaurant, Greenbaum & Gilhooley's, a New York restaurant, the Empire State Restaurant & Tavern Association, and Jim Tozzi, an individual business man and investor in one of the restaurants. On November 29, 1999, the Court granted plaintiff's motion to amend their Complaint. The amended Complaint added Brevet Industries and Brevet, Inc. as plaintiffs in the action, and was otherwise identical to the original complaint.

³ Defendants in this action are the United States Department of Health and Human Services("HHS"); Donna E. Shalala, Secretary of HHS; Kenneth Olden, Director, National Institute of Environmental Health Sciences ("NIEHS") and Director, National Toxicology Program ("NTP"); and George W. Lucier, Director, Environmental Toxicology Program ("ETP").

³ Mechanistic data includes data on the effect of toxic substances introduced to cells in a laboratory and studied at the cellular level, as opposed to epidemiological data, which studies the effect of toxic substances on living people over time.

⁴ Plaintiff has not, in any event, alleged that it sold PVC plastic products to Tenet or that it will lose business as a result of Tenet's decision.

⁵ See Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984) outlining a two-step analysis for reviewing with deference an agency's interpretation of a statute that Congress has entrusted to the agency for implementation.)

⁶ It simply makes no sense that, after lengthy debate and discussion, the defendants would preclude the use of mechanistic data, a powerful and emerging tool in toxicologic analysis, for consideration under the "Known" category, which is the highest listing category possible.

⁷ That is, of course, if the evidence may scientifically support such a conclusion. Plaintiffs assert in their third argument, discussed below, that the evidence for dioxin falls short. They are incorrect.

⁸ Although the reference was to "Category 1," rather than to the "Known" category, it is clear from later references in the same paragraph to "Category 2" that these statements were intended to refer to the "Known" and "Reasonably Anticipated" categories.

⁹ Dr. Greenberg was representing the American Forestry and Paper Association in October, 1997, (see R., Ex. 17 at 15), and was representing the Chlorine Chemistry Council in December, 1998. R., Ex. 26 at 4.

¹⁰ In any event, nowhere does the DBD state that the epidemiological evidence, such as the Bertsazzi studies, are "sufficient" by themselves. Clearly, the dioxin listing is based on many epidemiological studies, and on the mechanistic data discussed above. See R., Ex. 24, Supplemental DBD, listing additional sources not reviewed by IARC.

¹¹ Thus, even if the plaintiffs were likely to succeed on the merits, which they are not, their failure to make the separate and independent showing of irreparable harm necessarily scuttles their claim for injunctive relief. See Sea Containers Ltd. v. Stena AB, 890 F.2d 1205, 1210-11 (D.C. Cir. 1980) (affirming denial of injunctive relief where plaintiff may have been likely to succeed but failed to show irreparable injury).

¹² This argument is separate from, although factually relevant to, defendants' jurisdictional and constitutional arguments that plaintiffs lack standing. See Defendants' Memorandum In Support of Motion To Dismiss at 21-25.