PATENTS

www.nature.com/naturebiotechnology • SEPTEMBER 2002 • VOLUME 20 • nature biotechnology

Research tool patents typically claimmethods or compositions used to dis-

cover novel, biologically important com-pounds and therapeutics. Patentees often tryto maximize the rewards of their innovationby conditioning licenses to research toolpatents on royalties from sales of commer-cial products that may ultimately be discov-ered and brought to market using thepatented tools. Such royalties are called“reach-through royalties,” and arrangementsfor them have become increasingly commonin the biotechnology and pharmaceuticalindustries. However, they may be anticom-petitive and unenforceable if the patentee iseffectively seeking royalties on more than itinvented, disclosed, and taught in its patent.

Entitlement to reach-through royalties onresearch tool patents has become a matter ofincreasing concern for biotechnology andpharmaceutical companies alike. Hundredsof millions of dollars in potential royaltiesand profits on blockbuster drugs and othertherapies are at stake, as well as traditionalnotions of scientific progress, which may beimpeded by an increasingly complex andrestrictive intellectual property regime.

To date, the courts have provided surpris-ingly little guidance to researchers and indus-try concerning these important issues, andmost pharmaceutical companies have takenthe cautious approach of securing licenses toresearch tool patents used in the developmentof products in their commercial pipelines. Thevalidity and enforceability of these licenses(and the royalty streams they represent) haveyet to be determined. In this article, we reviewthe legal framework relevant to this issue, andthe traditional bounds of patent scope,enforceability, and licensing. We also offersome recommendations to potential licensorsand licensees of research tool patents.

Claim scope and reach-throughroyaltiesOwners of patents directed to researchtools naturally seek the highest possible

financial return on their research anddevelopment efforts. To maximize thatreturn, patentees typically negotiate licens-es based on flat fees or milestone paymentsdue at various stages of drug developmentthat rely on their patented research tools.More recently, however, many research toolpatentees have been seeking reach-throughroyalties based on sales of drugs and thera-peutics ultimately developed using theirresearch tools. Such licensing regimensbring up questions of how broadly claimsto research tools ought to be applied andwhat royalty base they ought to command.The narrow application of such claims maylimit patentees to reach only those meth-ods and compositions specifically disclosedand claimed in a patent. In contrast, thebroader application of such claims mightalso reach compositions unidentified at thetime of the filing of the patent application,but discoverable through the use of theclaimed research tool. Such broad applica-tion would allow patentees to commandreach-through royalties1 based on sales ofundisclosed compositions discovered usingtheir claimed methods.

Reach-through royalty licenses areincreasingly common because (i) they aremore profitable than licenses based solely onsales or use of the research tool, (ii) they maybe easier to enforce than licenses based solelyon sales or uses of the research tool, and (iii)they potentially maximize the patentee’sreturn on investment in otherwise limitedmarkets (for example, if only one or a fewpotential licensees conduct experimentsusing their patented tools).

Some reach-through royalty arrange-ments, however, may be improper. Thepatent statute requires that patent applica-tions (also known as “specifications”) con-tain a written description of the invention,essentially to show that an inventor was inpossession of the invention as claimed at thetime the application was filed (35 USC §112)2. Arguably, reach-through royaltyarrangements should not award patenteesroyalties based on sales of products that lackdescriptive support in the specification.

The specification must also enable a per-son skilled in the art to make and use theinvention without undue experimentation.

This means that the teachings of the specifi-cation must be “commensurate in scope”with the invention as claimed. If a person ofskill in the art could not practice a claimedinvention by following the teachings of thespecification combined with the knowledgein the art at the time the application wasfiled, the claim would fail for lack of enable-ment3. Arguably, reach-through royaltyarrangements should not award patenteesroyalties based on the sale of products thatwere not enabled by the disclosure of thespecification at the time of the invention.

In sum, patents that adequately describeand enable a method or tool for discoveringbioactive compounds or using specificscreening systems under certain experimen-tal conditions may not describe and enablethe commercial products that directly resultfrom the licensee’s use of those methods andtools. The tension between the indispens-ability of licensed research tools to drugdevelopment and the limitations of theunderlying research tool patents will likelybe at the center of the legal debate over thevalidity and enforceability of reach-throughlicenses.

Conflicting public policies: patentprotection versus patent misuseResearch tool developers are left to deter-mine the proper scope of their research toolpatents and what the proper scope of theirlicenses ought to be. It is well established thatinnovation should be rewarded under thepatent statute. However, the licensing andenforcement of patents on research toolsraise unique issues in the biotechnology andpharmaceutical fields, where relatively fewproducts reach the market annually and thecost of getting them there is often extraordi-narily high4.

As a matter of public policy, patentsshould not be construed to cover subjectmatter beyond the scope of what they dis-close and claim. However, some in thebiotechnology and pharmaceutical indus-tries have attempted to extend the reach oftheir patents to do just that, by seekinglicenses based on sales of products potential-ly discoverable using the claimed methods,but neither claimed nor supported by thedisclosure of their patents. Overbroad

945

Licensing research tool patentsPatent holders often seek royalties on the sales of products developed using their tools, but how broadly should "reach-through" claims be applied?

Gerald J. Flattmann and Jonathan M. Kaplan

Gerald J. Flattmann is a partner and JonathanM. Kaplan is an associate at Fish & Neave,1251 Avenue of the Americas, New York, NY10020 (gflattmann@fishneave.com).

©20

02 N

atu

re P

ub

lish

ing

Gro

up

h

ttp

://w

ww

.nat

ure

.co

m/n

atu

reb

iote

chn

olo

gy

nature biotechnology • VOLUME 20 • SEPTEMBER 2002 • www.nature.com/naturebiotechnology

PATENTS

946

enforcement and licensing of such researchtool patents may be anticompetitive; insteadof rewarding innovation by the grant of alimited monopoly, it may stifle subsequentinnovation and discovery by making the useof such tools prohibitive or unduly expen-sive. Potentially anticompetitive behaviorarising from such overbroad application ofresearch tool patents can manifest itself inthe form of (i) overly expensive licensingschemes; (ii) coercive licensing practices5;(iii) attempted foreclosure of subsequentresearch6; or (iv) creation of barriers to theentry of other firms into a field of scientificinvestigation or research6.

Restraints in trade based on the use ofintellectual property come in the familiarflavors embodied by the antitrust laws. Onetype of restraint relating specifically topatents, known as “patent misuse,” is a flexi-ble doctrine that broadly condemns “everyuse of a patent as a means of obtaining a lim-ited monopoly of unpatented material.”7 Amisused patent is unenforceable againstalleged infringers until affirmative steps aretaken to counteract the effects of the misuse.A showing of misuse, typically an affirma-tive defense to a suit for patent infringement,requires that the alleged infringer prove thatthe patentee has impermissibly broadenedthe “physical or temporal scope” of thepatent grant with anticompetitive effect.Arguably, licensing arrangements thatemploy reach-through royalties may imper-missibly broaden the scope of research toolpatents if they reach subject matter neithercontemplated by the inventor at the time offiling of the patent application nor ade-quately disclosed under 35 USC § 112. Insum, because reach-through royaltyarrangements tie royalties from a commer-cial product to a research tool patent thatdoes not itself cover that product, they maybe subject to scrutiny under the patent mis-use doctrine and the antitrust laws.

Not all reach-through royalty arrange-ments are alike. The question of whether aparticular arrangement constitutes animpermissible broadening of the patentgrant can only be answered in the context ofthe specific disclosure and claims of thepatent, the nature of the commercial productat issue, and the specific terms of the license.If a patent disclosure contains an enablingdescription of the commercial product (or agenus of similar products) or if the product isexplicitly claimed, then a licensing arrange-ment directed to that product is likely validand enforceable. On the other hand, reach-through licensing arrangements may beunenforceable in the absence of such teach-ings and claims. Notably, the US SupremeCourt has traditionally rejected licenses con-ditioned on the payment of royalties based

on subject matter the patent does not teachor claim8. Thus, patentees may potentiallyengage in misuse if they coerce prospectivelicensees into accepting licenses conditionedon the payment of royalties based on sales ofcommercial products that are not claimed orlack support in their patent.

Despite the increasing number of reach-through royalty arrangements being negoti-ated today, there is surprisingly little case lawregarding their scope and enforceability.Two relevant cases are discussed below.

Sibia v. CadusThe first of two recent cases that providesome useful insights into how courts maytreat reach-through licenses in the future isSibia Neurosciences v. CadusPharmaceutical9. After a six-week trial, aCalifornia jury returned a verdict awardingSibia a royalty of $18 million based on itsfinding that Cadus’s products infringedSibia’s research tool patent10 directed to cer-tain methods of drug discovery.

The ‘629 patent, issued on March 28, 1995to Harpold et al., broadly disclosed, amongother things: “[N]ovel recombinant cellswhich are useful for assaying compounds fortheir agonist or antagonist activity withrespect to specific ion channels and/or spe-cific cell surface localized receptors ... rapid,reliable methods to identify compoundswhich interact with, and thereby affect thefunction of, specific ion channels and/orspecific cell surface-localized receptors;[and] rapid reliable methods to determine ifcells are producing specific functional ionchannels and/or cell specific functional sur-face-localized receptors.”

The specification listed as propheticexamples scores of ion channels and cell sur-face receptors that could be used in the dis-closed methods. The ‘629 patent also dis-closed as actual examples (i) the activationof the M1 muscarinic receptor by its agonistcarbamylcholine in the presence or absenceof its antagonist atropine; and (ii) the activa-tion of gene expression by carbachol, bovineserum, or atropine.

Notably, the ‘629 patent claimed researchtools rather than pharmaceutical composi-tions discoverable through the use of thosetools. These research tools were recombinantcells and assays useful for identification ofone or more pharmaceutical compositionsdirected to specific, clinically relevant dis-eases. The claims of the ‘629 patent broadlyrecited a “method for identifying com-pounds that modulate cell surface protein-mediated activity...” The patent containedno claims to products or compositions iden-tifiable through that claimed method.

Sibia’s damages award was based onCadus’s actual and potential profits on its

commercialized products. The propriety ofbasing Sibia’s damages on Cadus’s sales of itscommercialized products was resolved inSibia’s favor at the trial court level. However,it was never ultimately resolved because thejury verdict was overturned by the Court ofAppeals for the Federal Circuit on differentgrounds—the Federal Circuit found thatSibia’s claims were invalid because they wereobvious over the prior art11.

Bayer v. HouseyThe second recent case that may provideguidance to the biotechnology and pharma-ceutical industries in negotiating researchtool patent licenses is Bayer AG v. HouseyPharmaceuticals. Bayer filed suit seeking adeclaratory judgment that defendantHousey Pharmaceuticals’ patents wereinvalid, unenforceable, and not infringed12.Housey’s US patent 4,980,281 (“the ‘281patent”) relates to a general screeningmethod for the discovery and identificationof both inhibitors and activators of enzymes,receptors, and other proteins. Specifically,one of its claims is directed to: “A method ofdetermining whether a substance is aninhibitor or activator of a protein whoseproduction by a cell evokes a responsivechange in a phenotypic characteristic otherthan the level of said protein in said cell perse, comprising…” The specification of the‘281 patent disclosed few substances thatmay act as inhibitors or activators and only ahandful of target proteins.

Bayer alleged that Housey engaged inpatent misuse by demanding a license withterms that would require substantial royaltiesbased on Bayer’s total research and develop-ment budget and total sales revenues for anycommercialized products that Housey con-tended were developed using its patentedmethod13. Notably, the court declined to dis-miss Bayer’s patent misuse defense: “Certainpractices … may constitute [patent] misuse ifa court determines that such practices do notreasonably relate to the subject matter withinthe scope of the patent claims.”14

Because Bayer’s patent misuse claim hasnot been decided on the merits, the court hasyet to provide guidance on which elementsof Housey’s licensing practices may consti-tute patent misuse. However, the Court’srefusal to dismiss the claim indicates thatunder certain circumstances attempts tolicense research tool patents with reach-through royalty provisions could constitutepatent misuse.

Reach-through royalties versusestablished licensing principlesTraditionally, royalties are calculated basedon the sales or uses of a patented product orprocess. However, parties may choose other

©20

02 N

atu

re P

ub

lish

ing

Gro

up

h

ttp

://w

ww

.nat

ure

.co

m/n

atu

reb

iote

chn

olo

gy

PATENTS

www.nature.com/naturebiotechnology • SEPTEMBER 2002 • VOLUME 20 • nature biotechnology

methods of licensing a patented technology,such as flat fees or milestone payments, inthe case of pre-commercialization licenses.Licenses to research tool patents presentunique problems regarding the calculationof appropriate royalties. One complicatingfactor is that the ultimate commercial valueof a patented research tool is as unpre-dictable as its prospects for successful appli-cation. As commentator Rebecca S.Eisenberg notes: “Licensing of patentedinventions may pose special problems wherethe inventions are useful primarily for subse-quent research rather than for commercialapplications. The serendipitous nature ofresearch discoveries may make it difficult toplace a value on the right to use a patentedinvention before the outcome of a researchproject is known.”15

Of course, despite unpredictability, licens-es will still be negotiated based on the expec-tations of the parties and their relative bar-gaining positions. The landmark Georgia-Pacific case16 provides a summary of the thencurrent case law creating a framework fordetermining a reasonable royalty in a patentinfringement action. Although Georgia-Pacific involved a method of determining areasonable royalty by constructing a hypo-thetical licensing negotiation that mighthave occurred between parties before litiga-tion, it sets out guidelines that are instructivein discussing how parties negotiate patentlicenses in the real world. Of the 15 factorsidentified by the Georgia-Pacific court as rel-evant in determining a reasonable royalty,factors 8–10 are most pertinent here.

Factor eight refers to the established prof-itability of the product made under a patent.Profitability is difficult to assess when suc-cessful marketing of a patented technology,or a potential commercial derivative thereof,is uncertain and unpredictable. Researchtool patentees often attempt to minimize theimpact of this uncertainty by negotiatinglicenses based on a combination of bothfixed fee payments geared to specific mile-stones in product development and reach-through royalties on sales of the ultimateproduct. Prospective licensees, however,often understandably object to agreeing topay royalties based on commercialized prod-ucts that the patentee will play no role indeveloping17,18.

Factor nine concerns the utility andadvantages of the patented technology overthe prior art19. It involves considerationssuch as the superiority, efficiency, andexpense of using the patented technologyand the ability of the prospective licenseesimply to design around the patent (usingdifferent research tools).

Factor ten is concerned with the nature ofthe patented invention, the character of its

commercial embodiments, and its benefitsto those who have used it19. These considera-tions are complicated by the unpredictabilityof research and development. In the contextof research tool patents, the nature of thepatented invention can be quite differentfrom the character of many of its resultingcommercial embodiments.

In sum, the unpredictability of the valuelicensed companies will derive from usingpatented research tools complicates thetraditional framework for negotiatinglicenses. Creative solutions incorporating amix of flat fees, milestone payments, and,perhaps, noncoercive reach-through royal-ties, may provide negotiators with somebasis for hammering out a license benefi-cial to both parties.

Recommendations for licensingresearch tool patentsThose who develop novel and usefulresearch tools should seek patent protectionfor their inventions by drafting applicationscontaining detailed disclosures. The specifi-cations of these patent applications shouldinclude examples of the various types ofcompounds and biomolecules that thepatentee believes may be used in, or discov-ered by, the research tool, as well as repre-sentative species of those compounds orbiomolecules. The specifications shouldalso include as many working examples aspossible to increase the likelihood that theapplication describes and enables commer-cial products on which the patentee can baseits royalties.

Once patents are obtained, patenteesshould pursue licenses with those in thebiotechnology or pharmaceutical industrywhom they reasonably believe may be usingor could use their technology to developcommercial products. In offering licenses,patentees should consider proposing royal-ty arrangements that include flat fees andmilestone payments geared to specificstages of product development. Patenteesmay also consider proposing reach-throughroyalties based on sales of the ultimatecommercial products. In doing so, paten-tees must be careful not to condition thelicensing of the research tool claims onreach-through royalties based on productsneither described nor enabled by theirpatent specifications. Such a “take it or leaveit” approach in the absence of a reasonableargument that the patent claims shouldreach those commercial products may con-stitute patent misuse, rendering the patentsunenforceable until the effects of that mis-use have been mitigated.

Potential licensees may often have littlechoice but to take a license to patentedresearch tools if (i) they intend to use those

tools to discover or create commercial prod-ucts, and (ii) after investigation, they believethat the patents are valid, enforceable, andcover the potential licensee’s activities.Ignoring the potential need for such licensestoday is tantamount to risking a patentinfringement lawsuit, a finding of willfulinfringement, and substantial damages laterif the products ultimately developed arecommercially successful. However, in viewof the current uncertainty concerningenforcement of research tool patents andthe unpredictability of the value that theymay bring to the prospective licensee, it maybe possible to negotiate low royalty rates. Asdiscussed above, the patent misuse doctrineprovides prospective licensees with a basisfor rejecting coercive reach-through royaltyregimens that are not reasonably supportedby the claims and specifications of thepatents in question. Until the courts providemore guidance, we offer these recommenda-tions to aid both sides in negotiations ofresearch tool patent licenses to achievemutually beneficial and enforceablearrangements.

1. Cullem, J.G. Panning for biotechnology gold:reach-through royalty damage awards for infring-ing uses of patented molecular sieves. Idea 39,553 (1999).

2. 35 USC § 112, first paragraph.3. Genentech, Inc. v. Novo Nordisk et al., 108 F.3d

1361, 1366 (Fed. Cir. 1997).4. Carroll, A.E. A review of recent decisions of the

United States Court of Appeals for the FederalCircuit: comment: not always the best medicine:biotechnology and the global impact of US patentlaw. Am. U.L. Rev. 44, 2433, 2482–2483 (1995).

5. Zenith Radio Corp. v. Hazeltine Research, Inc., 395US 100, 135–137 (1969).

6. Kobak, J.B. Intellectual property, competition lawand hidden choices between original and sequentialinnovation. Va. J.L. & Tech. 3, 6 at ¶23 (citingBarton, J.H. Patents and antitrust: a rethinking inlight of patent breadth and sequential innovation.Antitrust L.J. 65, 449, 454 (1997)).

7. Leitch Mfg. Co. v. Barber Co., 302 US 458, 463(1938); See also B. Braun Medical Inc. v. AbbottLab., 124 F.3d 1419, 1426 (Fed. Cir. 1997).

8. Zenith Radio Corp. v. Hazeltine Research, Inc., 395US at 135 (1969).

9. Sibia Neurosciences, Inc. v. Cadus PharmaceuticalCorp., 225 F.3d 1349, 1354 (Fed. Cir. 2000).

10. Harpold, M.M. & Brust, P. Assay methods and com-positions useful for measuring the transduction ofan intracellular signal. US patent 5,401,629 (“the‘629 patent”).

11. Sibia Neurosciences, Inc. v. Cadus PharmaceuticalCorp., 225 F.3d at 1357–1359.

12. Bayer AG and Bayer Corp. v. HouseyPharmaceuticals, Inc., 169 F.Supp 2d 328 (D. Del.2001).

13. Idem at 329.14. Idem at 331.15. Eisenberg, R.S. Proprietary rights and the norms of

science in biotechnology research. Yale L.J. 97,177, 217 (1987).

16. Georgia-Pacific Corp. v. U.S. Plywood Corp., 318 F.Supp. 1116 (S.D.N.Y. 1970).

17. Eisenberg, R.S. Patents and the progress of sci-ence: exclusive rights and experimental use. U. Chi.L. Rev. 56, 1017, 1078 (1989)

18. Barton, J.H. Patents and antitrust: a rethinking inlight of patent breadth and sequential innovation.Antitrust L.J. 65, 449 (1997).

19. Georgia-Pacific Corp. v. U.S. Plywood Corp., 318 F.Supp. at 1120.

947

©20

02 N

atu

re P

ub

lish

ing

Gro

up

h

ttp

://w

ww

.nat

ure

.co

m/n

atu

reb

iote

chn

olo

gy