dissenting.
It has been confirmed that the remedy of injunction in patent cases is subject to *1349the sound discretion of the district court, upon application of the traditional criteria by which injunctive relief is evaluated and applied. eBay, Inc. v. MercExchange, L.L.C., — U.S. -, -, 126 S.Ct. 1837, 1841, 164 L.Ed.2d 641 (2006) (“We hold only that the decision whether to grant or deny injunctive relief rests within the equitable discretion of the district courts, and that such discretion must be exercised consistent with traditional principles of equity, in patent disputes no less than in other cases governed by such standards.”)
These traditional principles are no less applicable when a preliminary injunction is at issue, particularly when the purpose is to preserve — not to change — the relationship of the litigants during the litigation. “The purpose of a preliminary injunction is to preserve the relative positions of the parties until a trial on the merits can be held.” Univ. of Texas v. Camenisch, 451 U.S. 390, 395, 101 S.Ct. 1830, 68 L.Ed.2d 175 (1981); Smith Int’l, Inc. v. Hughes Tool Co., 718 F.2d 1573, 1578 (Fed.Cir.1983) (“A preliminary injunction will normally issue only for the purpose of preserving the status quo and protecting the respective rights of the parties pending final disposition of the litigation.”) Precedent counsels against making an important change in the relationship of the parties while their dispute is being litigated, while recognizing that there may be circumstances warranting such change, when all of the legal and equitable aspects relevant to a particular case are considered. See, e.g., Mikohn Gaming Corp. v. Acres Gaming, Inc., 165 F.3d 891, 895 (Fed.Cir.1998) (the preliminary injunction serves to preserve the status quo “lest one side prevent resolution of the questions or execution of any judgment by altering the status quo”); Globetrotter Software, Inc. v. Elan Computer Group, Inc., 236 F.3d 1363, 1367 (Fed.Cir.2001).
The trial court’s decision with respect to the discretionary grant of a preliminary injunction warrants significant deference, for equitable considerations weigh heavily in matters of change or stability pendente lite. See Deckert v. Independence Shares Corp., 311 U.S. 282, 290, 61 S.Ct. 229, 85 L.Ed. 189 (1940) (“ ‘It is well settled that the granting of a temporary injunction, pending final hearing, is within the sound discretion of the trial court; and that, upon appeal, an order granting such an injunction will not be disturbed unless contrary to some rule of equity, or the result of improvident exercise of judicial discretion.’ ”); Meccano, Ltd. v. John Wanamaker, New York, 253 U.S. 136, 141, 40 S.Ct. 463, 64 L.Ed. 822 (1920) (“The correct general doctrine is that whether a preliminary injunction shall be awarded rests in sound discretion of the trial court.”). It is particularly irregular for an appellate court to reverse this discretionary decision and thereby to make a significant change in the relationship of the parties, while presenting no explanation of how the district court abused its discretion.
Reversal of a preliminary injunction that preserves the status quo requires a clear showing that the district court exceeded its discretionary authority. See We Care, Inc. v. Ultra-Mark Int’l Corp., 930 F.2d 1567, 1570 (Fed.Cir.1991) (“The court’s determination can be overturned only on a showing that it abused its discretion, committed an error of law, or seriously misjudged the evidence.”) My colleagues do not discuss the trial judge’s careful explanations, but, upon finding that Teva has raised a “substantial question” about patent validity, they hold that Teva should be permitted to practice the Abbott invention before patent validity is decided. With all respect to my colleagues’ concerns, they misapply not only the criteria of the preliminary injunc*1350tion but also the standard of appellate review:
First, as to patent validity, the panel majority rejects the requirement that in determining the likelihood that the patent will be proved invalid it is necessary to consider the burdens of proof that would inhere at trial. See Canon Computer Sys., Inc. v. Nu-Kote Int’l., Inc., 134 F.3d 1085, 1088 (Fed.Cir.1998) (“However, a patent is presumed valid, and this presumption exists at every stage of the litigation.”); Genentech, Inc. v. Novo Nordisk, A/S, 108 F.3d 1361, 1364 (Fed.Cir.1997) (criterion “substantial questions of validity” means that “in light of the presumptions and burdens that will inhere at trial on the merits” the attacker has “a likelihood of success” in invalidating the patent); PPG Indus., Inc. v. Guardian Indus., Inc., 75 F.3d 1558 (Fed.Cir.1996) (“The ultimate question, however, is whether the challenger’s evidence of invalidity is sufficiently persuasive that it is likely to overcome the presumption of patent validity.”) In the case now before us the district court analyzed the evidence, in which technologically complex questions are presented, and concluded that Teva was not likely to prove the patent invalid by clear and convincing evidence. In contrast, the panel majority holds that if the attacker raises no more than a “substantial question” of invalidity, that suffices to establish the likelihood that the attacker will succeed on the merits. That is incorrect in law and in procedure.
Next, even as the panel majority states its agreement with the district court’s finding that “the balance of hardships favors Abbott,” maj. op. at 30, the majority declines to weigh this factor in its decision. See Chrysler Motors Corp. v. Auto Body Panels of Ohio, Inc., 908 F.2d 951 (Fed.Cir.1990) (“Our rule regarding whether a preliminary injunction should be granted or denied is that the trial court should weigh and measure each of the four factors against the other factors and against the magnitude of the relief requested.”). Abbott points out that the status quo ante will not easily be recoverable if interim infringement is authorized; Abbott also points out that there is no patent barrier to Teva’s entry into commerce with its own extended release formulation instead of that of Abbott, for the basic patent on clarithromycin has expired. Thus the panel majority again applies a flawed methodology, for “Where it is clear that the moving party will suffer substantially greater harm by the denial of the preliminary injunction than the non-moving party would by its grant, it will ordinarily be sufficient that the movant has raised ‘serious, substantial, difficult and doubtful’ questions that are the proper subject of litigation.” Ugine-Savoie Imphy v. United States, 121 F.Supp.2d 684, 689 (Ct. Int’l Trade 2000). The district court’s consideration of this aspect was proper, and warrants appellate deference.
Next, the panel majority states that the question of the sufficiency of money damages is “arguable” — ignoring the district court’s finding that this aspect may also favor Abbott. Instead, the majority opinion announces that this aspect will not be considered at all. On this accumulation of flaws, and with no reference to the district court’s well reasoned opinion, my colleagues reverse the preliminary injunction, change the status quo, and authorize infringement before validity is decided. I must, respectfully, dissent.
The Considerations Pendente Lite
At issue are claims 2, 4, and 6 of U.S. Patent No. 6,010,718 (the ’718 patent) and claim 2 of U.S. Patent No. 6,551,616 *1351(the ’616 patent). The claims are directed to an extended-release formulation of er-ythromycin in a polymer matrix, and require that the minimum plasma concentration for the extended release formulation is substantially equivalent to that of the immediate release formulation; that is, the drug is released so as to be present in the plasma at the same minimum level for both the immediate release and extended release formulations, and with less fluctuation for the extended release product.
In the challenge to validity, Teva and Abbott both presented extensive argument and briefing, citing various references. The district court, explaining its decision on the question of Teva’s likelihood of success in proving the patents invalid, analyzed the evidence and concluded:
This court finds that Teva has failed to raise a substantial question as to the validity of Abbott’s claims 2 and 4. The prior art cited by Teva discloses discrete portions of the asserted claims, but Teva fails to demonstrate that this would be sufficient to give a person of ordinary skill in the art a reasonable expectation of success. Teva’s prior art references reveal that using HPMC was a logical line of inquiry but the dissimilarities between the drugs with which HPMC had been successfully combined and clari-thromycin defeat Teva’s claim of obviousness.
The district court included discussion of the issues and arguments presented by the parties, remarked on the uses of various known release agents, and the unpredictability of achieving successful extended release as to any particular product. The record shows discussion at the district court that the metabolic pathway of the active ingredient must be determined, as well as the physical and chemical properties and physiologic behavior and effectiveness of the metabolites and their interaction with the materials in the extended release formulation. The record shows discussion at the district court of the need for extended release performance that will produce an effective drug exposure in the bloodstream over the entire release period, and the unpredictability of this performance. The record shows discussion of the non-linear pharmacokinetics 1 exhibited by clarithromycin. The district court explained its decision that Teva had not shown that it was likely to prove invalidity of the claimed formulations:
This court is mindful of the Federal Circuit’s warning about the risk of the “hindsight trap,” or the post facto belief that an invention, which seems obvious once created, would have been obvious to people skilled in the art at the time. Abbott has provided ample evidence that its invention was not obvious and that there were many other extended release formulation methods known in the prior art. In fact, the existence of alternate methods and the attempted exploitation of some of those methods provide secondary considerations of nonobviousness. These factors suggest that there was a long-felt need for the invention, that others, including Abbott, initially failed to develop the invention, and go a long way to account for the commercial success that Abbott has unquestionably enjoyed with its BIAXIN XL product.
The panel majority does not discuss, and assigns no flaw, to the district court’s re*1352fusal to apply judicial hindsight; nonetheless, the majority applies such hindsight for itself, starting with the template of the Abbott invention and then selecting portions of references to reconstruct the invention within that template. To guard against such incorrect analysis, precedent teaches that references cannot be selected, and selected elements from selected references cannot be combined, without some suggestion, motivation, or teaching that would make obvious that selection and that combination. See, e.g., Karsten Mfg. Corp. v. Cleveland Golf Co., 242 F.3d 1376, 1385, 58 USPQ2d 1286, 1293 (Fed.Cir.2001) (“In holding an invention obvious in view of a combination of references, there must be some suggestion, motivation, or teaching in the prior art that would have led a person of ordinary skill in the art to select the references and combine them in the way that would produce the claimed invention.”); Brown & Williamson Tobacco Corp. v. Philip Morris Inc., 229 F.3d 1120, 1124-25 (Fed.Cir.2000) (“a showing of a suggestion, teaching, or motivation to combine the prior art references is an ‘essential component of an obviousness holding’”).
The panel majority acknowledges the law, but finds a motivation to make the claimed formulation by combining the information in Abbott’s prior art Patent No. 5,705,190, which shows extended release formulations of clarithromycin and azithro-mycin in “alginate,” a known release agent derived from seaweed, with a Pfizer publication designated WO 95/30422, which shows the HPMC (hydroxypropyl methyl cellulose) of the ’718 patent used with azi-thromycin. Abbott stated at the preliminary injunction hearing that what works for a product in an alginate matrix is not predictably applicable to other products; this statement was not contradicted. The district court analyzed the interchangeability of clarithromycin and azithromycin, stating:
The questions are: how similar and dissimilar are the two molecules; and what are the implications of these similarities and dissimilarities to a person of ordinary skill in the art in light of prior art at the time of the invention. Specifically, would a person of ordinary skill in the art have had a reasonable expectation of success in creating an extended release formulation of clarithromycin using a hydrophilic water-soluble polymer based on the prior art, including the ’422 patent for an extended release formulation of azithromycin with such a polymer?
The court concluded that they were not so similar as to be interchangeable in the context of polymers like HPMC, correctly rejecting the argument that “obvious to try” can establish obviousness. The court stated:
Teva’s prior art references reveal that using HPMC was a logical line of inquiry but the dissimilarities between the drugs with which HPMC had been successfully combined and clarithromycin defeat Teva’s claim of obviousness.
My colleagues ignore the district court’s analysis, offering neither deference nor acknowledgment. Instead, the panel majority explains that its finding of likelihood of success in proving obviousness is supported “not by what the 190 patent discloses but what it does not disclose” (emphasis in maj. op.), proposing that: “Abbott has represented to the U.S. Patent and Trademark Office that the differences between clarithromycin and azithro-mycin were such that azithromycin could be substituted into a controlled release clarithromycin composition by a person of *1353ordinary skill in the art without undue experimentation.” Maj. op. at 17-18. Thus my colleagues conclude that claim 4 of the ’718 patent is “vulnerable to allegations of invalidity,” and find “a substantial argument” as to other claims. These are not the criteria of likelihood of success.
Reversible error has not been shown in the district court’s analysis, and no basis whatsoever has been shown for overturning the court’s discretionary decision to preserve the status quo while the matter is litigated. Even if Teva had raised a substantial argument, as my colleagues find, the criteria of abuse of discretion have not been met. To support a change in the status quo before the merits are decided, it must be shown to be likely that the patent will be held invalid under the presumptions and burdens in effect at trial. The panel majority is incorrect in holding that it “requirefs] less proof’ to authorize infringement before the merits are decided; such a rule, whereby a patent is deprived of exclusivity during litigation, is not readily invoked, for it is excessively disruptive of the processes of law. As the Court said in eBay v. MercExchange: “As this Court has long recognized, ‘a major departure from the long tradition of equity practice should not be lightly implied.’ ” 126 S.Ct. at 1839 (quoting Weinberger v. Romero-Barcelo, 456 U.S. 305, 320, 102 S.Ct. 1798, 72 L.Ed.2d 91 (1982)).
Conclusion
The district court’s conclusion as to the challenger’s likelihood of success in invalidating all of the claims in suit, and the district court’s view of the balance of harms, are well reasoned and fully supported by precedent. The district court’s ruling, preserving the status quo during litigation, warrants, and requires, our deference. From my colleagues de novo and incorrect contrary ruling, I must, respectfully, dissent.
. Abbott explains that "non-linear” here means that the amount of drug in the blood is not directly proportional to the dosage amount, but increases disproportionately with higher doses. This is not a characteristic of azithromycin, the product whose formulation is relied on by the panel majority, as discussed infra.