dissenting from denial of the petition for rehearing en banc.
The issue for rehearing en banc is the Federal Circuit’s rejection of the district court’s construction of the claim term “a therapeutically effective amount.” My concern is with this court’s methodology and rationale of claim construction, and with the sources on which the panel majority relied (or failed to rely) in arriving at its changed claim construction. The district court had correctly applied this court’s precedent, requiring affirmance.
I
In brief, the panel majority construed the claims more broadly than the invention that was patented; thus the court ignored the patentee’s purpose of writing claims that avoid the prior art — a purpose essential to every claim in every patent — by now construing the claims so that, as the panel majority recognized, they may read on the prior art. The patentee’s purpose of including the critical limitation to “a therapeutically effective amount” of the genetically engineered erythropoietin (EPO) was to distinguish the prior art EPO isolates, which were not therapeutically effective to “heal or cure,” in the district court’s words. Thus this court’s claim construction diverges from the specification and the prosecution history, and presents a claim construction that impinges on the prior art and thereby fosters invalidity. From this analytic method of claim construction and the denial of rehearing en banc I must, respectfully, dissent.
The methodology of claim construction that the court has here adopted raises issues that were laid to rest in Phillips v. AWH Corp., 415 F.3d 1303 (Fed.Cir.2005), where the en banc court disavowed the view that a patent claim can be construed more broadly than the invention to which it gives legal effect. In Phillips that unwarranted breadth was found in claim construction based on general dictionary definitions of claim terms, rather than relying on the specific technical usage in the specification; in this case the unwarranted breadth is found in the court’s claim construction that may embrace the prior art, despite the exclusion of that subject matter by the prosecution history.
The district court, construing the claims in light of the specification and the prosecution history, construed “a therapeutically effective amount” as “a quantity that produces a result that in and of itself helps to heal or cure.” That construction was the basis on which the patentee distinguished its genetically engineered EPO from the chemically identical EPO in known isolates of naturally occurring EPO, for the prior art product contained insufficient EPO to heal or cure. Thus the claim limitation to “a therapeutically effective amount” of *1042EPO was critical to allowance. During an extensive prosecution, the applicant stated that
naturally occurring human erythropoiet-in is not a viable human therapeutic product; human recombinant erythro-poietin, on the other hand, has been proven to be clinically effective, and is the first therapeutic product which can be used to effectively treat the hundreds of thousands of patients who suffer from anemia and other disorders involving low blood counts.
App. No. 113,178, Amendment under Rule 16 at 4, June 5,1989 (emphasis in original).
The record in the district court contains evidence, presented by persons experienced in this field of science, that the limitation to “a therapeutically effective amount” distinguishes the genetically engineered EPO from known EPO isolates. The district court construed the claims from the viewpoint of the skilled artisan, upon reading and understanding the specification and the restrictions flowing from the prosecution history. My concern about this court’s methodology is that the panel majority does not apply these standard tools of claim construction. The court’s holding that “a therapeutically effective amount” includes any amount of EPO that exhibits a physiologic effect, whether or not the amount of EPO is adequate to provide therapy for any disorder, imparts to the claims the scope that was excluded during prosecution.
Although this court has urged caution in construing claims in order to preserve their validity, no precedent or logic requires that when more than one claim construction is available, the court must choose the broader one although it may invalidate the claim. See Phillips, 415 F.3d at 1327 (court may not construe claims as would be appropriate to preserve their validity unless “after applying all the available tools of claim construction ... the claim is still ambiguous”); see also Smith v. Snow, 294 U.S. 1, 14, 55 S.Ct. 279, 79 L.Ed. 721 (1935) (“if the claim were fairly susceptible to two constructions, that should be adopted which will secure to the patentee his actual invention”); Modine Mfg. Co. v. United States Int’l Trade Comm’n, 75 F.3d 1545, 1556 (Fed.Cir.1996) (“When claims are amenable to more than one construction, they should when reasonably possible be interpreted so as to preserve their validity.”).
Here, the specification and prosecution history make clear that the claimed invention is the “therapeutically effective amount” of engineered recombinant EPO. As was explained to the patent examiner and again in the district court, the prior art isolates of EPO were ineffective to heal or cure. Just as the prosecution record cannot enlarge the claims beyond what the inventor has presented as his invention, so the court cannot enlarge the claims beyond the limitations imposed by the patentee. See Phillips, 415 F.3d at 1319 (rejecting definitions broader than the technical usage specific to the invention). The court in this case has departed from these fundamental principles. We should speak en banc to clarify that it is appropriate, and necessary, to look at what has in fact been invented, prosecuted, and patented, and construe the claims accordingly.
II
I do not share the view, expressed here by some colleagues, that this court should not intrude upon panel decisions when major errors of claim construction are pointed out on petition for rehearing en banc. The Federal Circuit has a special obligation to provide predictability and consistency in patent adjudication, for our panel decisions are of nationwide effect; indeed, this obligation was a justification for the court’s *1043holding, a decade ago, that the district court’s claim construction receives non-deferential review on appeal. This appellate position imposes on us the obligation to state the correct law, even on rehearing en banc. If the meaning of “therapeutically effective amount” is treated as a question of law, its correct definition as well as the methodology by which it is defined are squarely within the criteria for rehearing en banc. And if the meaning is recognized as a case-specific finding of fact, appellate review warrants deference to the trier of fact, a deference here lacking.
Ill
I continue to believe that findings of science/technology-based facts in patent cases should receive appellate review on the same basis as other science-based findings, guided by Daubert and the Court’s ensuing elaborations. The Court’s and our own precedent require the trial judge to evaluate scientific evidence and expertise from the viewpoint of a person experienced in the field of science, a framework that aptly fits evaluation of the technologic content and scope of patents, an analysis whose intermingling of fact and law is well served by the procedures and the adjudicatory skill of the district courts. The Federal Circuit’s position that patent interpretation requires more rigorous appellate review than other fact/law issues has not well withstood the test of experience. It is time to reopen the question and to rethink, en banc, the optimum approach to accuracy, consistency, and predictability in the resolution of patent disputes, with due attention to judicial structure, litigants’ needs, and the national interest in invention and innovation.