Opinion
Civ. A. No. 88 C 4997.
August 27, 1990.
Roy H. Wepner, Joseph S. Littenberg, John R. Nelson, Lerner, David, Littenberg, Krumholz Mentlik, Westfield, N.J., Daniel W. Vittum, Jr., Russell E. Levine, Kirkland Ellis, Chicago, Ill., for Vas-Cath Inc., Gambro, Inc. and Omni Medical Products, Inc.
Arthur Sternberg, Pedersen Houpt P.C., Chicago, Ill., Local counsel, for Gambro, Inc., for notice purposes only.
Michael W. Coffield, Michael J. Philippi, Coffield, Ungaretti, Harris Slavin, Chicago, Ill., for Geoffrey Martin and M. Jane Martin.
Samuel Fifer, Sonnenschein Nath Rosenthal, Chicago, Ill., for Quinton Instruments Co.
Joseph N. Hosteny, Raymond P. Niro, John C. Janka, Niro, Scavone, Haller Niro, Ltd., Chicago, Ill., for Sakharam D. Mahurkar and Quinton Instruments Co.
OPINION
Sakharam D. Mahurkar, a physician, holds several patents on dual-lumen hemodialysis catheters. A dual-lumen catheter is a pair of tubes (lumens) designed to allow blood to be removed from an artery, processed in a machine that removes impurities, and returned close to the place of removal.
Hemodialysis catheters are used to palliate kidney failure. In the event of chronic failure, physicians construct a fistula, a permanent internal connection between vein and artery that provides ready, long-term access to the circulatory system. During the several weeks required for the fistula to mature and heal, physicians need another entry point, which the catheter provides. In the event of acute failure, immediate access is essential, again via catheter.
Catheters cause trauma on insertion and while they remain. Every entry of a needle injures the blood vessel, the walls of which eventually collapse. The fewer entries, the better. The more comfort, the better. (Sometimes patients try to rip out catheters that aggravate them.) A catheter also should allow a high rate of flow without injury to the blood. Whirlpools, eddies, sharp edges, and collisions with rigid walls can rupture the red cells, which may lead to "hemolysis" (inability of the blood to carry sufficient oxygen) and to clotting. Either can cause brain damage or the death of the patient. Knowledge of fluid dynamics and the structure of the blood and vascular system is essential to design a catheter that can handle the necessary rate of flow at an acceptable rate of injury to the red cells.
Dual-lumen catheters allow the return and removal of blood with a single insertion. For years physicians used coaxial dual-lumen catheters. Blood leaves the body through the large outer tube and returns through the inner tube. In the cross-sectional diagram (Figure 1) the shaded area is the inner (return) tube, and the unshaded area the annulus through which blood is withdrawn.
Figure 1
The tip of the inner tube extends beyond the inlet, so that the cleansed blood is returned "downstream". Coaxial catheters usually were inserted surgically. Even so, inaccurate centering in the blood vessel could produce substantial damage to the blood, as could eddies that developed in the fluid motion.
The dual-lumen catheters covered by Dr. Mahurkar's patents take a different approach. Instead of concentric circles, they use joined semi-circular tubes, as the following figure shows. The shaded areas in this figure are the walls.
Figure 2
These tubes come to a single tip, a conic section with two openings. The opening at the tip (downstream) releases the cleansed blood; the other opening takes it in. Figure 3 shows illustrations of such a tip, from one of Dr. Mahurkar's patents.
Figure 3
Dual-lumen catheters of this construction may be inserted without surgery. Their small tip, coupled with the holes near the tip, allows physicians to use the "Seldinger technique": a wire guides the catheter into place. It may remain in place for weeks, so that one insertion is enough to complete the transition to the use of a fistula. The puncture area of a coaxial catheter is 42% greater than that of a semi-circular catheter carrying the same quantity of blood. And semi-circular designs with conical tips yield low rates of injury to the blood.
Catheters of this general construction are the medical community's first choice today for temporary access. Coaxial catheters are obsolete. Dr. Mahurkar's particular catheters have been successful, and they appear to represent more than half of the world's sales.
Vas-Cath Incorporated and its licensee Gambro, Inc. (the American subsidiary of Sopamed SA, the Swiss subsidiary of Gambro AB, a Swedish corporation), filed this suit seeking a declaratory judgment that their dual-lumen hemodialysis catheters do not infringe Dr. Mahurkar's U.S. patents. The complaint contends that (i) those patents are invalid, (ii) Geoffrey Martin, Vas-Cath's principal, is a co-inventor entitled to practice the patents, and (iii) the claims of the patents do not cover Vas-Cath's catheters. This suit is complex not only because of the many legal issues but also because Dr. Mahurkar holds numerous patents, and Vas-Cath's catheters have gone through two generations. Each generation of catheters must be assessed under each patent. As the case has proceeded, the Patent Office has issued some additional patents, which have been added to the litigation.
Mahurkar and his licensee Quinton Instruments Co. (a subsidiary of American Home Products Corp.) filed a counterclaim against Vas-Cath, contending that Vas-Cath's products infringe his patents. Mahurkar seeks an injunction and damages. Mahurkar brought into the suit Geoffrey Martin and his wife M. Jane Martin, as principals and stockholders of Vas-Cath, and Omni Medical Products Incorporated. The nature of the claims against these defendants is not important for current purposes. For simplicity, I call Mahurkar and Quinton "Mahurkar", and everyone else "Vas-Cath", except in Part III of this opinion when it becomes necessary to distinguish them.
Each side requests summary judgment. More precisely, each side requests summary judgment on particular counts and issues. All agree that summary judgment is not yet (and may never be) appropriate on all claims. The Rules of Civil Procedure do not expressly authorize motions for partial summary judgment. Nonetheless, the Manual for Complex Litigation § 21.34 (2d ed. 1985), recommends that judges pare off issues that may be resolved as a matter of law. The Advisory Committee on Civil Rules has recommended an amendment to Fed.R.Civ.P. 56 that if adopted would make official the practice of issuing legal rulings that dispose of issues but not the whole suit (or even entire claims). 127 F.R.D. 258, 370-85 (1989). The new caption of the rule conveys its scope: "Summary Establishment of Fact and Law; Summary Judgment". Both sides ask me to cut down the scope of this dispute, as if the proposal were, already in force. I shall do so without attempting to resolve the suit, or particular counts, in full. After considering this opinion, the parties may propose an order that will put these rulings into effect.
I
Dr. Mahurkar's patents include Nos. 4,568,329 and 4,692,141. Vas-Cath asks me to declare these invalid under 35 U.S.C. § 102(b), contending that they were anticipated by Canadian Industrial Design No. 50089, which issued on August 9, 1982. The parties agree that this patent (Canadian '089) contains drawings identical, except for labels, to those underlying the U.S. '329 and '141 patents. No surprise, for Canadian '089 is also Dr. Mahurkar's patent.
An invention is unpatentable if it was "described in a printed publication in this or a foreign country . . . more than a year prior to the date of the application for patent in the United States". 35 U.S.C. § 102(b). Mahurkar applied for the '329 and '141 patents more than a year after August 9, 1982. Although it may seem odd that an inventor can be "anticipated" by his own work, the parties agree that Canadian '089 qualifies as a prior publication under § 102(b). Section 102(b) requires inventors to apply for U.S. patents with dispatch. They cannot lollygag while other persons start making an item and claim their patent at convenience, a maneuver that could both mousetrap business rivals and extend the effective life of the patent beyond 17 years.
Mahurkar does not deny that if the '329 and '141 patents take their own filing date, they are invalid. This outcome follows from his concession (for purposes of this motion) that Canadian '089 is enough to allow practice of the invention and therefore to anticipate the '329 and '141 patents. Ralston Purina Co. v. Far-Mar-Co, Inc., 772 F.2d 1570, 1574-75 (Fed. Cir. 1985). He submits, however, that they should be treated as if filed on March 8, 1982, when he applied for a U.S. design patent, Serial No. 356,081. Mahurkar abandoned serial '081 in 1984, after he applied for the utility patents that became '329 and '141. Because the drawing underlying serial '081 is functionally identical to Canadian '089 and illustrates the '329 and '141 patents, Mahurkar invokes 35 U.S.C. § 120, which gives an application the same filing date as "an application for patent for an invention disclosed in the manner provided by the first paragraph of section 112 in an application previously filed in the United States". Serial '081 qualifies under this language, Mahurkar submits, and so the '329 and '141 patents have filing dates before Canadian '089.
Vas-Cath meets this by contending that serial '081 does not satisfy the requirements of the first paragraph of § 112, which are:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Conceding for purposes of this motion that the diagrams in serial '081 enable one skilled in the art to practice the invention, Vas-Cath nonetheless maintains that serial '081 does not contain a "written description of the invention". Indeed, it does not contain a writing, period. For it is a design patent, and under the Patent Office's rules applications for design patents need not contain written descriptions "other than a reference to the drawing". 37 C.F.R. § 1.153(a).
"Written" does not necessarily mean "using words". It could mean "committed to paper" or something similar. Writings are symbols conveying thoughts. Pictures and diagrams also are symbols conveying thoughts. So there is no sharp line between language and pictures. Recall that "a picture is worth 1,000 words." In re Berkman, 642 F.2d 427, 429 (CCPA 1981), holds that drawings may satisfy the disclosure requirement.
Vas-Cath's objection to the use of serial '081 is not so much the missing language as it is a contention that the diagrams in the design application do not describe the invention — that is, the things that make the '329 and '141 patents distinctive. What matters, Vas-Cath insists, is that no words of limitation appear in serial '081 and no way to derive them. It shows a dual-lumen catheter, with D-shaped lumens and a conical tip. The tip has a distinctive cut, taper, and bevel, and holes obviously designed for use with the Seldinger technique. But it does not say what the invention consists in. Without words of limit, Vas-Cath insists, it is impossible to satisfy the first paragraph of § 112.
There are two ways out of this for Mahurkar. One is to establish that anything sufficient to allow reduction to practice is necessarily "a written description of the invention". If this is so, then Vas-Cath's concession that serial '081 allows practice — essential to Vas-Cath's position that Canadian '089, with the same basic diagrams, anticipates the '329 and '141 patents — is fatal. When Vas-Cath argues that Canadian '089 anticipates the '329 and '141 patents, it necessarily concedes that serial '081 discloses the invention, the argument concludes. The second way out is to demonstrate that the drawings in serial '081 do describe "the invention", as the ceramic described in Kennecott Corp. v. Kyocera International, Inc., 835 F.2d 1419 (Fed. Cir. 1987), revealed a crystal lattice by operation of the laws of physics. I discuss these in turn.
Decisions of the Federal Circuit control on the question whether something may allow the practice of an invention without automatically providing "a written description of the invention". Unfortunately, it is not so easy to tell what the law of the Federal Circuit is. In re Lukach, 442 F.2d 967, 969-70, 58 CCPA 1233 (1971), In re Barker, 559 F.2d 588, 591 (CCPA 1977), and In re Scheiber, 587 F.2d 59 (CCPA 1978), hold that the two are distinct. Paperless Accounting, Inc. v. Bay Area Rapid Transit System, 804 F.2d 659, 664-65 (Fed. Cir. 1986), pretty much merges the two. Then In re Gosteli, 872 F.2d 1008, 1012 (Fed. Cir. 1989), reiterates that they differ — citing Lukach and not Paperless Accounting. You can draw from these cases almost anything you please, and the parties have done so. I do not think that I can resolve the question on the basis of precedent, or by claiming that the most recent word must be dispositive. It is better to start from scratch.
Section 112 says that the application must contain "a written description of the invention". It also says that the application must contain enough to allow practice. Separate statement of logically distinct requirements implies that there are indeed two elements. Sometimes a single disclosure may satisfy both; sometimes not. Words and pictures in conjunction (the stuff of patent applications) frequently will show the nature (and extent) of the invention as well as permit practice. Sometimes, as in Kennecott, a description of one feature of the invention will imply (and limit) others because the laws of physics allow only one possibility.
More commonly, however, a single diagram allows practice without showing "the invention". A diagram may show a dual-lumen catheter, with a given tip. Any competent professional could look at the diagram and the light would come on. The "Aha!" reaction would be followed by some simple steps to produce a workable catheter, just as blueprints enable the construction of buildings without verbal descriptions. Yet the information allowing practice would not necessarily show what the invention is, when "the invention" could be a subset or a superset of the features shown. Is the invention the semi-circular lumens? The conical tip? The ratio at which the tip tapers? The shape, size, and placement of the inlets and outlets? You can measure all of these things from the diagrams in serial '081 and so can practice the device, but you cannot tell, because serial '081 does not say, what combination of these things is "the invention", and what range of variation is allowed without exceeding the scope of the claims. To show one example of an invention, even a working model, is not to describe what is novel or important.
One could say — perhaps the Federal Circuit eventually will say, picking up on Paperless Accounting — that any filing establishing that the inventor's job was complete, and that only the scrivener's job remained, satisfies § 112. Viewed in retrospect, serial '081 shows this, and if this demonstration is enough then Mahurkar must prevail. Yet, in patent law, the translation from invention in a scientific sense to the legal description of the invention is important, because the law establishes a system of deadlines that could be subverted by allowing the scientific sense to control. Unless it conveys a sense of which features matter, a diagram could enable later filings on the entire genus of which it is a species, a process that would substantially enlarge the time for filing the real patent application. If a picture in Year 1 is to be the priority date for a dozen inventions more or less different years later, the timing rules would fall to pieces. These considerations lead me to conclude that Lukach and Gosteli state the law, that disclosure allowing practice does not necessarily describe "the invention".
This brings us to Mahurkar's second argument, that the diagrams in serial '081 satisfy the requirement that the application describe the invention. Whether it does is a question of fact, and Mahurkar offers the affidavit of Mark Newman to the effect that the diagrams disclosed what was important. But the expert leaves out his process of reasoning, so the affidavit is entitled to little if any weight. Cable Electric Products, Inc. v. Genmark, Inc., 770 F.2d 1015, 1025-26 (Fed. Cir. 1985); Union Carbide Corp. v. American Can Co., 724 F.2d 1567, 1572 (Fed. Cir. 1984). See also Mid-State Fertilizer Co. v. Exchange National Bank, 877 F.2d 1333, 1338-40 (7th Cir. 1989); Kenosha Liquor Co. v. Heublein, Inc., 895 F.2d 418, 420 (7th Cir. 1990). With or without the affidavit, the record contains no disputed issue of material fact, so that summary judgment is appropriate. Although it has hinted that a diagram could satisfy the requirement that the application describe "the invention" KangaROOS U.S.A., Inc. v. Caldor, Inc., 778 F.2d 1571, 1574 (Fed. Cir. 1985), the Federal Circuit has never found one to be sufficient. This case would not be an auspicious beginning. Even giving the '329 and '141 patents the presumption of validity, 35 U.S.C. § 282, and applying the rule that a patent stands in the absence of "clear and convincing" absence of invalidity, as I have done, does not produce enough to support the conclusion that serial '081 satisfies the first paragraph of § 112. (There is, by the way, no evidence that the examiners on the the '329 and '141 patents considered this question.)
The drawings in serial '081 show cross-sections of a certain size, and openings a certain distance from the tip. But they do not require these features. Nothing in the laws of physics implies them. It may be that alterations would not work well, but this is a "utility" aspect that was not mentioned in the design patent. There were no clues in the design patent on how far if at all one could alter structure and still have the gist of this invention.
For example, the utility patents claim a return lumen that is "substantially greater than one-half but substantially less than a full diameter" after it makes the transition from semi-circular to circular cross-section, and the drawings of serial '081 fall in this range. But until the utility application was filed, nothing established that they had to — for that matter that the utility patent would claim anything other than the precise ratio in the diagrams, or even that the utility application would claim that the ratio mattered (that it was part of "the invention"). The diagrams therefore could have been used to support a sequence of modifications, tearing up the time rules. The wide scope for elaboration on the theme of the diagrams in serial '081 prevents them from satisfying the initial requirement of § 112. When a description, written or otherwise, could imply one of many forms, disclosures have been held insufficient. E.g., In re Wright, 866 F.2d 422, 424 (Fed. Cir. 1989); Titanium Metals Corp. v. Banner, 778 F.2d 775, 782 (Fed. Cir. 1985); In re Wilder, 736 F.2d 1516, 1520 (Fed. Cir. 1984).
Mahurkar replies with an affidavit by Dr. Stephen Ash concluding that someone expert in the field of hemodialysis catheters could have worked it out from the diagrams that the return (longer) lumen had to be less than 1.0 and greater than 0.5 the diameter of the tubing. This is logical, for a return lumen with a diameter as large as the entire catheter would produce a drop in pressure, with unhappy consequences, and unless the ratio exceeds 0.5 there is an increase in pressure. (Recall that this ratio applies after the cross-section has been converted to circular, below the intake.) Still, it is impossible to understand that Mahurkar believed that he had invented this ratio, or was claiming it, or if he was claiming it, within what range (" substantially greater than one-half but substantially less than a full diameter", as things turned out). Even this may be a mite fast. Later patents issued to Mahurkar disclose ratios closer to 1.0 (No. 4,583,968) and exactly 0.5 (No. DES 272,651). If these are do-able, even desirable, ratios — at least when some other variations in the geometry accompany them — how does serial '081 necessarily exclude then?
All in all, the best one can say is that what Mahurkar eventually patented is exactly what the pictures in serial '081 show. The utility patents contain the same diagrams. That cannot be sufficient, however, if the design application would have allowed variations — that is, if the question be asked ex ante, as it should be. Moreover, although Mahurkar's patents use the same diagrams, they contain limitations that did not follow ineluctably from the diagrams.
Priority depends on a demonstration that the party has "possession of every feature" of the invention. Coleman v. Dines, 754 F.2d 353, 359 (Fed. Cir. 1985). We know in retrospect that Mahurkar was in possession of all the features, and that serial '081 shows an embodiment of them; but it does not show which of the attributes are "the features", and how much variation in what is shown is allowable. Mahurkar's utility applications claimed the ratio of the return lumen's diameter to the catheter's. They might have claimed the distance between intake and outlet, or the shape of the inlet. They claimed neither. Nothing in serial '081 showed that the ratio of diameters was important, and the distance between intake and outlet irrelevant. If the applications for the '329 and '141 patents had contained the diagrams from serial '081 together with the single claim: "A dual-lumen catheter constructed like this", the examiner would have bounced the applications without much ado. For essentially the same reasons, I conclude that serial '081 does not satisfy the first paragraph of § 112.
Mahurkar emphasizes that in 1982 he was a practicing nephrologist, doing research on his own time and money. He filed a design application because it is a lot cheaper to prosecute than is a utility application. Since 1982 the value of his inventions has been recognized around the world, and he has acquired financial backing allowing him to perfect the utility applications and carry on with this costly litigation. His first filings now haunt him; his own invention is the "prior art" cited against him. In an ideal world, these facts might count for something. Still, one could reply that Mahurkar's first catheter went on the market in June 1983, less than a year after Canadian '089, and if the manufacturer (Quinton) thought it safe to rest with the design application, there should be no crocodile tears seven years later. At all events, the patent laws of the United States do not inquire into such matters. The '329 and '141 patents are invalid under § 102(b) because the applications were filed more than a year after the issuance of Canadian '089, and are not entitled to the filing date of serial '081.
II
Vas-Cath and Mahurkar are not strangers. They have been friends before, and adversaries before. In April 1981 Mahurkar sent Vas-Cath some double-D tubing with a request for help in "forming a smooth tip and bevel satisfactory for penetration test". Vas-Cath crafted some tips along lines Mahurkar suggested, and sent along at least two tips of its own design. Mahurkar contends that Vas-Cath stole his ideas when it began to market its own dual-lumen catheters. Vas-Cath contends that Mahurkar stole its ideas (demonstrated in the additional sample tips) when obtaining his patents.
This dispute led to patent litigation in Canada, Vas-Cath's base. Mahurkar sued Vas-Cath for infringing his Canadian patent No. 1,193,508. Vas-Cath defended on the usual grounds (obviousness, anticipation, non-infringement, and the like), adding the contention that Geoffrey Martin is the real inventor of the tip. After a trial on the merits, Justice Strayer of the Federal Court of Canada resolved almost every issue of fact and law in favor of Mahurkar, throwing Vas-Cath a bone by holding that several of the claims of Canadian '508 were too broad. Mahurkar v. Vas-Cath of Canada Ltd., 18 C.P.R.3d 417, affirmed, No. A-384-88 (Fed.Ct.App. Jan. 11, 1990).
Mahurkar's U.S. patent No. 4,583,968 covers the same invention as Canadian '508. He seeks summary judgment that Vas-Cath's first generation of catheters (the Vaccess 3500 and 4000 catheters) infringe the '968 patent, on the theory that the Canadian judgment precludes Vas-Cath from denying either the validity of the '968 patent or infringement by its catheters. Vas-Cath replies that the judgment concerning Canadian '508 has no force with respect to U.S. '968, but that, if it does, Vas-Cath is at least entitled to a declaration invalidating the American claims corresponding to the Canadian claims that Strayer J. found overbroad.
First in line comes the question whether principles of preclusion apply to foreign judgments concerning patents. Vas-Cath insists that they do not, on the ground that U.S. law differs from foreign law: both the substantive and procedural rules about patents vary from nation to nation, so that what is patented in Canada may be in the public domain here, and vice versa. To give preclusive effect to the Canadian judgment, Vas-Cath believes, would allow Canada to export its patent law to the United States. It musters an impressive array of precedent for the proposition that principles of claim preclusion (res judicata) do not carry across national borders in patent cases. E.g., Medtronic, Inc. v. Daig Corp., 789 F.2d 903, 907-08 (Fed. Cir. 1986); In re Dulberg, 472 F.2d 1394 (CCPA 1973); Skil Corp. v. Lucerne Products, Inc., 489 F. Supp. 1129, 1158 (N.D.Ohio 1980), affirmed, 684 F.2d 346, 351 n. 3 (6th Cir. 1982); Cuno Inc. v. Pall Corp., 729 F. Supp. 234 (E.D.N.Y. 1989).
To this Mahurkar replies that differences in law do not prevent enforcement of judgments from other jurisdictions. Although this proposition has overwhelming support, Restatement 3d of Foreign Relations § 481, § 482(2) (1987), cf. Restatement 2d of Judgments § 27 (1982), it is beside the point. No one thinks twice about enforcing in the United States a German court's judgment that a particular contract is valid and that the offender must pay damages, even though the Uniform Commercial Code is substantially different from the commercial law of Germany. We are not dealing with enforcement of a judgment. We are not even dealing with the judgment of a Canadian court that U.S. '968 is valid and infringed, so that claim preclusion might apply. We have instead a judgment that Canadian '508 is valid and infringed. The claims differ between the suits, and claim preclusion applies only to a claim that was or could have been prosecuted in the first action. Mahurkar did not contend, and could not have contended in the Canadian action, that Vas-Cath's products infringed the U.S. '968 patent.
Although a court may not say "Canadian '508 is valid and infringed, therefore U.S. '968 is valid and infringed", it is commonplace for courts in the United States to employ issue preclusion (collateral estoppel) even when claim preclusion is unavailable. E.g., Harris Trust Savings Bank v. Ellis, 810 F.2d 700, 705-06 (7th Cir. 1987); Ackermann v. Levine, 788 F.2d 830, 844 (2d Cir. 1986). Patent litigation is costly, and the Canadian case was hard fought. Why begin from scratch? Conservation of resources is the principal objective of the law of preclusion, and that is a vital objective when costs are high, the more so when similar patents have been secured in many of the industrial countries. Patent litigation should not be allowed to become a war of attrition, in which after the conclusion of one battle parties move on to another and duplicate the engagement. If litigation squanders the returns to invention, we will have less innovation — a depressing thought, given the importance of invention to economic growth. Whenever there is a choice, a court ought to opt for cost-saving and decision-expediting devices, of which preclusion is an attractive one. See generally Blonder-Tongue Laboratories, Inc. v. University of Illinois Foundation, 402 U.S. 313, 91 S.Ct. 1434, 28 L.Ed.2d 788 (1971) (enlarging the scope of preclusion in patent cases).
Is there a choice? Judge Weinstein believes that the Federal Circuit is so hostile to preclusion in patent cases that a district court must decline to give effect to a foreign judgment even on questions of fact extensively litigated. Cuno, 729 F. Supp. at 238-39. This did not please Judge Weinstein, who remarked on "[t]he law's absurdity as revealed by this case" — not only compounding costs but also substituting the judgment of "American jurors with no background in science or patents, whose average formal education will be no more than high school", id. at 239, for the considered decisions of experts in the United Kingdom. I do not read the Federal Circuit's cases as compelling courts of the United States to ignore informed decisions rendered abroad; the judges of that distinguished court are not xenophobes. Decisions from the Federal Circuit instead reflect the legitimate concern that, when considering what the first case settled, courts not disregard differences in law. If a foreign court renders judgment on a question of fact with significance in each system of law, there is no reason not to take over that decision. Despite Judge Weinstein's omens, I propose to do just that: to examine the Canadian judgments, to learn what has been decided, and to apply those decisions to this litigation to the extent — and only to the extent — they are legally relevant, and the findings are free of the influence of legal differences.
Before I turn to this task, however, three more possibilities need attention. One is that Canadian law supplies the rule of preclusion, the second that the Canadian proceedings are unreliable, and the third that the presumption of validity accorded to U.S. patents colors the appreciation of foreign decisions. Start with the first. Federal courts in the United States must use the rules of the rendering court when deciding the preclusive effect of state judgments, even when the judgment affects a subject within the exclusive jurisdiction of the federal courts. 28 U.S.C. § 1738; Marrese v. American Academy of Orthopoedic Surgeons, 470 U.S. 373, 105 S.Ct. 1327, 84 L.Ed.2d 274 (1985). Section 1738 refers only to state judgments, however, which means that federal common law supplies the rules of preclusion.
Vas-Cath asks me to apply Canadian law as a matter of comity, and it argues that Canada would allow it to bring a sequential "impeachment" action against Mahurkar's patent. Mahurkar maintains that Vas-Cath had its opportunity at impeachment as part of its defense against Mahurkar's suit. I do not think it necessary to decide who is right. When deciding the force to be given to foreign judgments, the Federal Circuit has crafted its own common law, without reference to the rules of preclusion applied in the rendering nations. Cf. Robert C. Casad, Issue Preclusion Foreign Country Judgments: Whose Law?, 70 Iowa L.Rev. 53 (1984). This means that I need not settle this debate between the parties. At all events, Canada applies the broad, British version of issue preclusion, see Restatement 3d of Foreign Relations § 481 note 6(b) (summary of approach of foreign nations). As I do not venture beyond issue preclusion, arguments about the extent of claim preclusion in Canada and the obligations of comity need not cause concern.
Similarly, I do not think it important that Canada does not afford parties the same discovery rights as the Federal Rules of Civil Procedure. Preclusion survives substantial differences in the rules of procedure (provided the differences are not so great that the original judgment was obtained without due process of law, as we understand that phrase). Even within the United States, some states are more stingy with discovery than are federal courts, yet their judgments are honored. Vas-Cath does not contend that the system of litigation in Canada is uncivilized or unlikely to produce sound judgments. Most nations of the world view our system of discovery as uncivilized. No case I could find declined to give force to a Canadian judgment on the ground that Canadian procedural rules differ from ours. Vas-Cath has not pointed to any fact that it wanted to bring to the Canadian court's attention but could not because of procedural rules. Even after extensive discovery in this case, Vas-Cath has not identified any fact that could have made a difference, if only it had been found in time for presentation to the Canadian court.
As for the presumption of validity: this is a rule of law, not a filter through which to understand findings of fact. If foreign findings are colored by a foreign presumption against patents, or if the findings were rendered under a different burden of proof, these might prevent application in U.S. litigation. In the main, however, the Canadian judgment supports the validity of the patents; the presumption of validity reinforces rather than undercuts their force. The judgment of the Federal Court of Canada accordingly receives the same weight it would have if it had been rendered by a federal court in the United States.
A
Strayer J. concluded that Mahurkar and not Geoffrey Martin invented the tip covered by Canadian '508. 18 C.P.R.3d 425-30. Strayer J. found that Mahurkar devised the basic idea for his catheters during summer 1979 and developed a tapered tip by January 1981, before sending anything to Vas-Cath. He also concluded, id. at 426, that Quinton and not Vas-Cath developed Mahurkar's designs into a commercially acceptable catheter. In reaching these conclusions, Strayer J. relied on the credibility of the witnesses and their ability to do what they said they did:
I have preferred the evidence of the plaintiff to that of Mr. Martin on behalf of Vas-Cath. . . . [No evidence support's Martin's claims concerning the dates of his work.] Nor was there any evidence that he developed this design into a workable product until after the Mahurkar catheter was actually on the market. I am further inclined to believe that Dr. Mahurkar was the inventor rather than Mr. Martin because Dr. Mahurkar had the requisite skills to create an appropriate design and Mr. Martin did not. In expert evidence given by both Dr. Mahurkar and Dr. Uldall it was said that the design of such a catheter requires extensive medical knowledge, particularly that related to the behaviour of blood, as well as knowledge of the technology of hypodermic needle and catheter manufacture. Dr. Mahurkar's training and experiences equipped him for this task. By contrast, Mr. Martin's academic training is in music. . . . While he no doubt has developed a certain amount of technical skill in the manufacture of catheters I have concluded, from the foregoing and other evidence, that with respect to this catheter he was more opportunist than inventor.
18 C.P.R.3d at 428-29. Vas-Cath did not pursue this issue on appeal in Canada.
This factual finding is as relevant to the inventorship of U.S. '968 as it is to Canadian '508. Who is the inventor and who the copier is a staple question in any patent system, and the Canadian decision did not depend on any features of Canadian law not shared by U.S. law. It is a decision of fact, reached after full and fair litigation.
Mahurkar is not satisfied with victory on the '968 patent; he wants a decision that Martin contributed nothing. Vas-Cath insists that even if Mahurkar was the sole inventor of the '508/'968 subject matter, it is still entitled to credit for (or shop rights in) the U.S. '329 and '141 patents and DES '651, which Vas-Cath insists derive from the sample tips Martin devised in 1981. The Canadian judgment did not mention these tips; Mahurkar believes that the omission was natural, because Strayer J. thought Martin incapable of making such a contribution. There is, however, another reason for the omission: Vas-Cath sent Mahurkar these tips after the date of invention relevant for Canadian purposes. Issue preclusion applies only to issues actually and necessarily decided. Strayer J. did not decide the provenance of Mahurkar's other tips. Consequently, Vas-Cath is not precluded from arguing that tips Mahurkar devised after receiving the samples from Vas-Cath were derived from those samples. Mahurkar may contend that because he had hired Vas-Cath to work for him, he is entitled to the benefit of those tips. Nonetheless, there is no need to proceed along these lines. As I have declared the '329 and '141 patents invalid, the origin of their tips is immaterial. Vas-Cath may still wish to argue that the tip in DES '651 derives from these samples. From what I have seen of the tips, it is not apt to get far (the designs are quite different), and a motion for summary judgment might put any remaining issue of inventorship to rest. Nonetheless, it should be clear that the only question covered by issue preclusion is who invented the subject matter of the U.S. '968 patent.
B
Vas-Cath argues that Canadian '508 was anticipated by prior art. Strayer J. fully considered this contention, particularly the reference to Pfarre 1882, and concluded that the design is original with Mahurkar. 18 C.P.R.3d at 430-32. Again this seems to be a question on which differences in the legal system do not matter.
Vas-Cath identifies two potential differences. One need not detain us. In considering the Pfarre reference, the Canadian judge remarked that Pfarre's designs had not been reduced to practice and that a person skilled in the art would not have been aware of Pfarre's invention. Both of these are relevant in Canadian law. 18 C.P.R.3d at 430. The former is irrelevant in the United States, which also imputes to experts awareness of prior art. All the same, a fair reading of Strayer J.'s findings shows that neither of these was important. Pfarre invented an enema tube for use in cattle. As my Canadian colleague writes, Pfarre "describes a catheter of soft rubber rounded, not tapered, at the end. It has within it two channels or lumens which are not of equal size. . . . [T]he Pfarre patent says nothing of the importance of having two D-shaped lumens of similar size." 18 C.P.R.3d at 430. Enema tubes for washing parts of the body are not designed to minimize the injury to blood cells. The inventions are sufficiently different that I would be willing to declare as a matter of law that Pfarre does not anticipate the '968 patent. I need not, because after listening to expert opinion Strayer J. reached exactly that conclusion, which I deem preclusive against Vas-Cath.
The second, and most important, difference is that the timing rules of the patent laws set different cutoff dates for looking at prior art. Canada cuts off references to the prior art as of the date of the invention, which Strayer J. found to be January 1981. American law takes note of prior references up to one year before the filing date of the application — a difference that dooms Mahurkar's '329 and '141 patents. Mahurkar filed the application that matured into the '968 patent on August 15, 1984. He claimed the benefit of the filing date of serial No. 538,671, an application for a design patent filed on October 3, 1983. The parties have not (yet) asked me to decide whether Mahurkar obtains the benefit of that filing date. If he does, then prior art references between January 1981 and October 3, 1982, remain to be considered; if he does not, then references between January 1981 and August 15, 1983, must be considered.
At oral argument I asked whether there are significant references in these periods. Vas-Cath replied that Canadian '089 (Mahurkar's own design again!) falls in the gap and suggested that two U.S. patent applications filed in 1981 are significant. The priority date of the application for the '968 patent also may matter because of the "on sale" bar in 35 U.S.C. § 102(b): no patent may issue if the application was filed more than a year after the invention was "in public use or on sale in this country". Quinton started selling one of Mahurkar's catheters in June 1983. If the '968 patent takes August 15, 1984, as its filing date, these sales may be significant. Moreover, Quinton furnished samples of the catheter to the Northwest Kidney Center in Seattle in August 1982, which poses a difficulty for the '968 patent no matter which filing date obtains.
So far neither side has asked me to determine (a) which filing date applies to the '968 patent, (b) whether references between January 1981 and the year before the filing date of the '968 patent (whatever that turns out to be) anticipate it, or (c) whether either the sales in June 1983 or the samples in August 1982 activate the "on sale" bar. Envirotech Corp. v. Westech Engineering Inc., 904 F.2d 1571 (Fed. Cir. 1990), doubtless sheds some light on the latter question, holding that a product is not "on sale" unless made widely available. I mention these potential difficulties only to show that they lie ahead.
In a case this complex, it pays to reduce the number of open issues even if some remain for future decision. One issue that can be sliced off is whether references in the prior art before January 1981 anticipate the '968 patent. I hold, on the basis of the Canadian decision, that they do not.
C
According to Vas-Cath, Canadian '508 is obvious to one skilled in the art. This became a principal subject of the trial, and the Canadian judge made extensive findings. 18 C.P.R.3d at 432-46. Mahurkar won a smashing victory. Strayer J. reached this decision because (a) there was a recognized problem and unmet need, suggesting that those skilled in the art did not find the solution obvious; (b) Dr. Robert Uldall, a witness taken as a paradigm skilled practitioner, not only did not know how to solve the problem but also thought initially that the Mahurkar solution would not work; and (c) Mahurkar's solution swept the market, which implies an advance beyond others skilled in the art. If the rewards were as large as Mahurkar's success shows, then someone else would have had ample motive to find an "obvious" answer. The Federal Court of Appeal considered and rejected Vas-Cath's objections to these conclusions. All of the judge's reasons are as appropriate to decision under U.S. law as they were under Canadian law.Canadian and American standards on obviousness are identical in all significant respects but one: the United States presumes that practitioners know the prior art, Custom Accessories, Inc. v. Jeffrey-Allan Industries, Inc., 807 F.2d 955, 962 (Fed. Cir. 1986); In re Nilssen, 851 F.2d 1401, 1403 (Fed. Cir. 1988), and Canada apparently does not. Beloit Canada Ltd. v. Valmet Oy, 8 C.P.R.3d 289, 297 (Fed.Ct.App.). This distinction does not make a difference, because Strayer J. relied on the fact that even the most skilled practitioners at the time, witnesses Uldall and Dr. Stanley Shaldon, who were acutely aware of the prior art, did not find Mahurkar's solution obvious. Vas-Cath offers no reason to disregard that decision. Accordingly, I conclude that as between Mahurkar and Vas-Cath, the '968 patent is not obvious.
Questions remain about whether something coming to light between January 1981 and the priority date of the '968 patent could establish obviousness. Strayer J. disregarded all references after January 1981, see 18 C.P.R.3d at 435. So far as I can see, however, nothing cited in that period would have an effect on obviousness distinct from its effect on anticipation. It therefore makes sense to remove obviousness from the case as a separate issue with respect to the '968 patent, and focus attention on any intervening references under the rubric of anticipation.
D
Next comes the question whether the specifications in the patent are sufficient to allow one skilled in the art to practice the invention. Justice Strayer concluded that the specifications in Canadian '508 meet that test. 18 C.P.R.3d 436-42.
Canada asks whether the specifications in the patent as issued enable reduction to practice. See § 36(1) of Canada's Patent Act. The United States asks whether the information in the application as filed satisfies § 112. W.L. Gore Associates v. Garlock Inc., 721 F.2d 1540, 1556 (Fed. Cir. 1983). We do not yet know what disclosures Mahurkar made as of the date of filing the U.S. patent application, because we do not yet know which date Mahurkar will obtain — that of the design patent application in 1983, or the utility patent application in 1984.
As a practical matter, this means that the Canadian judgment is not preclusive on the sufficiency of disclosure if Mahurkar needs and obtains priority from the design application in 1983. If on the other hand Mahurkar accepts the 1984 priority date of the utility patent application, the Canadian judgment is preclusive. The only differences between the Canadian patent issued in 1985 and the U.S. application filed in 1984 have to do with four elements of Claims 1 and 7 (parallel to but slightly narrower than the Claims I and 5 in the Canadian patent). My reading of the documents and examination of the diagrams leads to the conclusion that the differences make explicit what was implicit in the Canadian papers. Because claims must be read in light of specifications, D.M.I., Inc. v. Deere Co., 755 F.2d 1570, 1574 n. 2 (Fed. Cir. 1985), I conclude that the changes do not affect the judgment on sufficiency of disclosure.
In sum, if Mahurkar obtains the 1983 date of the design application, the Canadian judgment is not preclusive; if Mahurkar takes the 1984 date of the utility application, the Canadian judgment is preclusive on sufficiency. Which date is the right one remains an open question.
E
Claims 1-6, 8-10, 12-13, 17, and 21-25 of Canadian '508 were held invalid because they claimed more than the invention actually described. 18 C.P.R.3d at 442-44. Strayer J. concluded that the "invention" in Canadian '508 has four essential features: (1) twin lumens, each semi-circular (D-shaped); (2) one lumen shorter; (3) a conical tapered tip at the distal end, with extra material for rigidity; and (4) a tip guidance point at the center of the conical tip. Claims asserting more were held overbroad because they do not disclose anything (as opposed to lay claim to it). Mahurkar says that Canadian '508 is essentially identical to U.S. '968. It ought to follow that the parallel claims of U.S. '968 are invalid too, if the law is the same. And it is. Although the United States does not use the word "overbreadth", the fundamental concept applies under § 112. In re Borkowski, 422 F.2d 904, 909, 57 CCPA 946 (1970); In re Hyatt, 708 F.2d 712, 714 (Fed. Cir. 1983).
Vas-Cath demands the benefit of the favorable portion of the Canadian judgment if Mahurkar has the benefit of the portions favorable to him. Mahurkar tries to avoid this on three grounds: the presumption of validity for U.S. patents; a difference in the burden of proof imposed on the party attacking the patent (preponderance in Canada, "clear and convincing evidence" in the United States); and a difference in Claim 1 of the U.S. '968 patent.
Burdens and presumptions fall away because interpretation and scope of a claim are legal issues in the United States. E.g., Corning Glass Works v. Sumitomo Electric U.S.A., Inc., 868 F.2d 1251, 1256 (Fed. Cir. 1989); Moleculon Research Corp. v. CBS Inc., 793 F.2d 1261, 1270 (Fed. Cir. 1986). "Clear and convincing" is a standard by which to assess the sufficiency of factual evidence, Raytheon Co. v. Roper Corp., 724 F.2d 951, 956 (Fed. Cir. 1983); SSIH Equipment S.A. v. ITC, 718 F.2d 365, 375 (Fed. Cir. 1983), and the question at hand does not depend on resolving factual conflicts or weighing evidence. The presumption of validity is a tie-breaker, a solution if all is in equipoise, and again the concept does not apply to a legal argument.
I am inclined to believe that he who lives by the sword must be prepared to die by the sword. Mahurkar has convinced me to apply rules of preclusion. Sometimes he gains, sometimes he loses. Despite this inclination, however, I am uncomfortable about treating the decision as preclusive because it is one of "law". Vas-Cath escaped the effect of the different burdens of persuasion — which usually defeat issue preclusion — by showing that the interpretation and validity of claims is a question of law. Usually that also means that claim preclusion does not apply. I therefore do not come to rest on this subject. As part of their next wave of motions, the parties should address the question whether preclusion applies to this "legal" aspect of the Canadian judgment.
F
Where does all this leave us? These issues remain to be decided concerning the '968 patent:
1. First, because it affects almost everything else, to what filing date is the '968 patent entitled?
2. Do samples or sales before that filing date (to the Northwest Kidney Center in August 1982 and more widely in June 1983) activate the "on sale" bar of § 102(b)?
3. Do disclosures after January 1981 but more than a year before the filing date anticipate the '968 patent?
4. Are the claims of the '968 patent parallel to Claims 1-6, 8-10, 12-13, 17, and 21-25 of Canadian '508 excessive in relation to the invention disclosed?
5. Do Vas-Cath's products infringe the '968 patent?
At least the first three of these questions must be decided without assistance from the Canadian judgment. The effect of that judgment on questions 4 and 5 remains in limbo. It is not clear that No. 5 is even disputed for the first generation of Vas-Cath's catheters, if the '968 patent is valid.
Patent DES '651 presents a sixth issue: Martin's role in the design of the tip. On this subject, for reasons I have explained, there is no preclusion, but I do not rule out future decision as a matter of law on any questions concerning this patent. Some briefs on shop rights have been tendered, but I defer any decision until the parties have had an opportunity to make arguments based on this decision.
III
Gambro, Inc. (Gambro US), one of the plaintiffs in this case, was not a party to the Canadian litigation. It insists that it is free to litigate from scratch even if Vas-Cath is precluded. Mahurkar asks me to hold that Gambro US is bound by the Canadian judgment to the same extent as Vas-Cath. Gambro Canada Ltd. was aligned with Vas-Cath in the Canadian action. Gambro Canada is a subsidiary of Gambro Lundia AB, a subsidiary of Gambro AB (Gambro Sweden). Gambro US likewise is an indirectly held subsidiary of Gambro Sweden. Mahurkar believes that Gambro US received "virtual representation" in Canada through the offices of Gambro Canada; Gambro US submits that it is independent of Gambro Canada in management, operation, and finance, and therefore is entitled to litigate the question for itself.
"Virtual representation" is a doctrine allowing the preclusion of parties not present in the first case, provided that someone who was present had the same interests as the absent party and so had every reason to prosecute or defend the case as vigorously. E.g., Benson Ford, Inc. v. Wanda Petroleum Co., 833 F.2d 1172 (5th Cir. 1987); Mother's Restaurant, Inc. v. Mama's Pizza, Inc., 723 F.2d 1566, 1572-73 (Fed. Cir. 1983). Judicial resources are scarce, and if the first litigation was as complete as the extra party could have made it, had it been present, why not accept the outcome? This is what courts do all the time in class actions, in which the representative member of the class supplies virtual representation to the missing members; as a practical matter, it is what happens in many cases even when the party nominally appears — a lawyer makes all the important decisions.
Notwithstanding the attractions of the doctrine, the idea that someone may be bound by the results of a case even though he did not participate is in tension with the autonomy recognized in our legal system. It amounts to saying that intervention is compulsory, that if a pending case affects your interests you must intervene in it, because you will be bound whether you intervene or not — at least if someone already there is "on your side" (that is, offers "virtual representation"). Yet the rules of intervention under Fed.R.Civ.P. 24 run the opposite way: if a party in the case represents your interests, the court is unlikely to allow intervention. The upshot is that when representation is "virtual" it may also be exclusive: the existence of a "virtual representative" will prevent you from intervening to protect your own interests.
All of this is in sharp conflict with the principles that an entity is not bound by the decision of a case in which it could not participate, and that no one is obliged to intervene. The onus, rather, is on the adverse party to join anyone it would like to have bound. If "virtual representation" is enough to bind a party, then Martin v. Wilks, 490 U.S. 755, 109 S.Ct. 2180, 104 L.Ed.2d 835 (1989), should have come out the other way: the court should have held that a consent decree binds anyone who could have intervened but did not. Martin concluded, however, that persons affected by a decree to which they were not parties may attack it collaterally, because judicial decrees do not bind non-parties.
Class actions do not offend accepted principles in the same way "virtual representation" does. A member of a class under Fed.R.Civ.P. 23(b)(3) may opt out. (Classes under Rule 23(b)(1) and (2) do not allow opting out, but in such injunctive cases there will be only one outcome without regard to concepts of "representation.") A member of the class who opts out may litigate independently, and is not bound by an adverse judgment in the main action — and correspondingly may not take the benefit of a favorable one, "virtual representation" by the class representative notwithstanding. Premier Electrical Construction Co. v. National Electrical Contractors Ass'n, Inc., 814 F.2d 358, 361-67 (7th Cir. 1987). See also Colby v. J.C. Penney Co., 811 F.2d 1119, 1125 (7th Cir. 1987), and Secretary of Labor v. Fitzsimmons, 805 F.2d 682, 692-94 (7th Cir. 1986) (in banc), both expressing doubts about "virtual representation".
A venerable rule makes it unnecessary to decide whether to embrace the doctrine of "virtual representation". Preclusion applies to the parties and those in privity with them. Legal rulings may affect a subject such as a contract or patent, and the holder of the patent may not have another go at litigation by issuing a new license. The licensee takes the patent encumbered, as it were, with the decision affecting its validity. So the right question is not whether Gambro US is the an alter ego of (or was "virtually represented" by) Gambro Canada. It is whether Gambro US, as Vas-Cath's licensee, is in privity of contract with Vas-Cath. If it is, the jig is up. Montana v. United States, 440 U.S. 147, 99 S.Ct. 970, 59 L.Ed.2d 210 (1979).
Technically Gambro US is not in privity with Vas-Cath, which licensed Gambro Sweden, not Gambro US or Gambro Canada. Gambro Sweden decided (through Gambro Lundia and Gambro Switzerland) that Gambro US would hold rights to Vas-Cath's products in the United States, and Gambro Canada would hold rights to them in the rest of the world. Gambro Sweden is in privity of contract with Vas-Cath, and I cannot see why Gambro Sweden should be entitled to receive two or more adjudications on the validity of Mahurkar's patents and whether Vas-Cath's products infringe them, just because Gambro Sweden carved up the world into two or more territories with two or more subsidiaries.
Is a parent corporation bound by adjudications in which its subsidiaries participate? Again in a technical sense the answer is no, for the subsidiary has a separate legal personality. Gambro Sweden cannot be called on to pay the debts of Gambro US or Gambro Canada. All the same, the parent is in a position to control the litigation if it chooses. Franchise Tax Board of California v. Alcan Aluminum Ltd., ___ U.S. ___, 110 S.Ct. 661, 666, 107 L.Ed.2d 696 (1990), holds that a foreign parent's ability to direct the litigation of its subsidiary in the United States is enough to require it to abide the results. Powers to control justified preclusion in Alcan and Montana; they do so here as well. Because Gambro Sweden, as licensee of Vas-Cath with the power to control Gambro Canada, is bound by the Canadian judgment, its sub-licensee Gambro US is bound. I hold that Gambro US is precluded by the Canadian judgment to the same extent as Vas-Cath.
IV
The papers filed in this case already threaten to overrun my chambers. Today's opinion makes a dent in the issues, but as Part II.F demonstrates many could not be resolved. Other patents and new Vas-Cath products have not even been mentioned. The Martins have filed two motions that are being briefed. Three more suits concerning Mahurkar's patents are pending. One has been transferred to the Northern District of Illinois under 28 U.S.C. § 1404(a), and the Judicial Panel on Multidistrict Litigation is contemplating whether to consolidate the others for pre-trial proceedings. I hope that this opinion will narrow the differences between the parties and facilitate more concise motions and briefs directed to the remaining disputes.