Congress enacted the Hatch-Waxman Act in 1984 where a limited industry-specific experimental use exception was incorporated for testing drugs and medical devices for purposes reasonably related to regulatory data gathering. Courts had given a broad interpretation of the wording 'reasonably related' to include food additives, colour additives, new drugs, antibiotic drugs and human biological products 5 . However, the recent case Integra v Merck 6 , the Federal Circuit Court seems to have limited the scope of this statutory exception to experiments carried out for the purposes of facilitating expedited marketing approvals for generic drugs only. The Court, furthermore, held that extending such an exception to new drug development would be contrary to the purpose and language of Hatch-Waxman Act.

This exception was first laid down in US patent law by Justice Story in the early nineteenth century, holding that where a patented product is made or used as an experiment, whether for gratification of scientific tastes, curiosity, or to ascertain the verity and exactness of the specification or for amusement, without an intent to use for profit, would not amount to patent infringement 7 . This exception was held to be 'truly narrow' in Roche v Bolar 8 and the slightest commercial purpose or intention for carrying out an experiment has been held to be patent infringement 9 . Such narrow interpretation of experimental use exception has also taken its toll on non- state university research. Regarding state universities, they are apparently immune from infringement liability of federal patent laws as held by the US Supreme Court in Florida v College Savings Bank 10 on grounds of the Eleventh Amendment immunity from patent infringement liability for states, instrumentalities of states, and state employees acting in their official capacity. According to Duffy 11 , this judgement may lead patentees to bring patent infringement suits before the state courts under the guise of unfair competition actions and secondly, the effect of state universities being granted immunity could lead to private sector industry utilising state universities for carrying out research, which would otherwise require a license from patentee. Also, it is arguable that such immunity is contrary to Art. 30 of the Trips Agreement, but this debate is outside the scope of this paper. However, the position of non-state universities appears to be different. In Madey v Duke University 12 Duke University employed Madey as a laboratory director. He owned two patents, before his appointment at the University. After his services were terminated, he sued Duke University for infringement of his patents. Duke University relied on, inter alia, the experimental use exception defence. The Federal Circuit refused to apply the experimental use exception to exempt university research activities from infringing a patent, as it held that these research activities "unmistakably further the institution's legitimate business objectives, including educating and enlightening students and faculty participating in these projects." It also disregarded the non-profit status of Duke University. Bee cited this view as early as in 1958 whilst criticising Ruth v Stearns-Roger Mfg. Co. 13 , where academic institutions were held to be not infringing on the ground of common law experimental use exception, he commented that "it is not clear that the use by the School of Mines was "for the sole purpose of gratifying a philosophical taste, or curiosity or for mere amusement." Was the School not in the business of educating students and did it not charge a tuition fee for such education? Also, schools such as this, frequently perform services for outside parties for which a fee is charged." 14

The rationale of the Duke University judgement could perhaps be due to the 1980 enactment of the Bayh-Dole Act whose explicit objective is to promote collaboration between commercial concerns and non-profit organisations, including universities. It encourages universities to patent 'subject inventions' (means any invention of the contractor conceived or first actually reduced to practice in the performance of work under a government contract, as defined in Federal Acquisition Regulations-FAR, Sec. 27.301), made with federal government funds and contemplates them to grant licenses of patents to the private sector. The judgement also emphasises on the importance of classification of a university as a pure academic institution or a commercial enterprise/entrepreneur, as debated by Monotti and Ricketson 15 . According to Mueller 16 , legal, technological and economic factors have removed the strict barriers between academic research and commercial research, thus pure 'philosophical' experimentation is no longer possible, calling for a broader interpretation of the common law experimental use exception to ensure that university research does not get stifle out. Furthermore, Straus 17 has pointed that narrow experimental use defence would be problematic particularly in plant biotechnology sector, where access to patented germplasm is crucial for further [university based research] innovation in plants. Since the US is the most advanced user of transgenic patented crops, one may not lose sight of the special importance which US developments may have in general on technical and scientific progress in plant biotechnology area. However, the AIPPI report of the American group on 'Patentability Requirements and Scope of Protection of Expressed Sequence Tags (ESTs), single Nucleotide Polymorphisms (SNPs) and Entire Genomes' 18 has suggested no change to the experimental use exception for biotechnology law. Furthermore, a recent National Science Foundation analysis has demonstrated that the proportion of university patents amongst the total utility patents have increased to almost five percent of all new US patents, which shows an increment of more than one hundred percent in last twenty years 19 . On the strength of these statistics, Johnson has expressed the view that a narrow interpretation of experimental use exception is unlikely to restrict university research 20 . Also, universities in US are very active in enforcing their patents. For instance, recently, University of California and Eolas Technologies sued Microsoft for patent infringement, whereby the federal jury awarded damages worth USD 520 million to University of California and Eolas Technologies for patent infringement 21 . Though, the debate about the nature and scope of experimental use exception has yet not concluded in the US as the Duke University judgement is most likely to be appealed to the US Supreme Court, the present position is that experimental use exception to patent rights in the US is 'truly narrow' and non-state academic institutions are not exempt from patent infringement liability as such. In light of Madey v Duke University, it is pertinent to examine the European, German and UK position as regards the experimental use exception to patent rights.

Under the German Patent Act of 1968, in force until the end of 1980, there was no specific provision excepting experimental use. The German Federal Supreme Court (BGH) in "Ethofumesat", decision of February, 1989 stated that experiments or trials with a protected pharmaceutical, would only be permitted insofar as such experiments were directed to the substance itself 25 , for example, to determine the substance's inherent properties. Clinical trials, however, were considered as being of a different nature and were not permitted.

Under the German Patent Act of 1981, which replaced the 1968 Act, Art.11.2 uses the identical language of CPC in providing the experimental use exception, under which two decisions are noteworthy. In "Clinical Trials I" 26 , the Court held that experiments including clinical trials for obtaining second usage of a patented product is well within Art. 11.2 of the 1981 Act and hence does not amount to infringement of patent rights. The BGH did not address the legal consequences if clinical trials were not directed toward a new indication, but were instead conducted to obtain government approval for a pharmaceutical having an indication identical to that of the protected substance. In "Clinical Trials II" 27 , the Court held that experimental use exception encompassed any trials including those to obtain data for clinical approval, even if such clinical trials were conducted for the same indication as that of the protected product. The BGH excluded from this general permission, clinical trials which would not be justified for the purpose of the experiments or which were conducted for the purpose of interfering with the marketing efforts of the patentee.

"Clinical Trials II" generated considerable criticism, particularly from "Big Pharma," but it was affirmed by the German Constitutional Court, in a decision of May 10, 2000. Based on this case law, Goddar 28 rates Germany as being one of the most liberal countries in Europe concerning the experimental use exception. It is submitted that German universities would not face a Madey v Duke University like outcome, at least in the pharmaceutical and medical fields.

Section 60(5)(b) of the Patents Act, 1977 incorporates the experimental use defence using the same words as that of the CPC. According to Judge Michael Fysh, of the Patents County Court 29 , Sec. 60(5)(b) contains two elements: (a) the acts must be done for experimental purposes; and (b) those purposes must relate to the subject matter of the invention. This demands enquiry of the nature of the invention and more significantly, the nature of the 'relationship'. It does not specify that experiments carried out should be non-commercial. It may also be noted that Sec. 60 (5) (a) dealing with 'private use' exception specifies that such acts should be non commercial. Thus, Sec. 60(5) (b) does not preclude experiments, which are done for commercial purposes.

Act must be done for experimental purposes

In Frearson v Loe 30 , the Court held that if a patented product was made only for bona fide experiment, without any intention to sell or use it, but with the view of improving upon the invention or with a view of seeing whether an improvement can be made or not, would not amount to patent rights infringement. Subsequently however, judicial attitudes have changed. In Monsanto v Stauffer 31 , Stauffer had developed a market variant 'Touchdown' of Monsanto's successful patented weed-killer 'Roundup' for which they had obtained provisional clearance from relevant authorities. In order to obtain final clearances, Stauffer had established tests at their own research farm and also organised a series of tests outside their research farm where interested parties could observe the results. Monsanto moved for an interlocutory injunction the grounds of patent infringement, which was granted by the Patents court, negativing the ground that tests done outside the research farm to check their product in different soil and climatic conditions, amounts to an experimental use. The Court of Appeal, although it agreed that tests done outside could not qualify for an experimental use exception, it exempted all trials carried out at Stauffer's research farm and at laboratories and greenhouses in the UK. The Court limited the interpretation of the word 'experimental' in accordance to its size, scale, recipient and methodology. The Court stated

...trials carried out in order to discover something unknown or to test an hypothesis or even in order to find out something which is known to work in specific conditions would work in different conditions or even perhaps to see if the experimenter could manufacture commercially in accordance with the patent. These can fairly be regarded as experiments but trials carried out in order to demonstrate to a third party [such as a regulatory body] that a product works or, in order to amass information to satisfy a third party, whether a customer or body, that the product works as its maker claims is not be regarded as acts done for 'experimental purposes'....

Loughan has remarked that in light of Monsanto v Stauffer, it is uncertain how far the university researchers could apply experimental use exception as some types of testing and field trials have been held to be infringing 32 .

The Court, in Smith Kline & French Laboratories Ltd v Evans Medical Ltd 33 observed that "what is or is not an experiment must depend upon the facts of each case but can include experiments designed with a commercial end in view."

Although it appears that the UK courts do not preclude all tests done for commercial purposes, as in the USA, the Patents County Court in its judgement in Inhale Therapeutic Systems Inc v Quadrant Healthcare Plc 34 , observed: "...it is said that there is a defence of experimental use. This fails on the facts. None of the exploitation of the products or technology complained of was conducted by Quadrant for its own experimental purposes. In all cases Quadrant was trying to exploit and sell its technology to third parties. This is not experimental use."

Cordery has commented that any commercial activity would now fall outside the scope of the 'experimental use' exception 35 . From this viewpoint, it appears that the Court has narrowed the scope of the experimental use exception, compared with the Monsanto and Smith Kline & French decisions. If this is true, English law seems to have moved closer towards common law experimental use exception of the US and drifted away from the German law.

The judgement, however, does not explain exhaustively why experimental use exception fails; it merely states that it fails on the basis of facts, but does not consider which facts. This could be because the Court had already held the patent invalid, which would have diminished the significance of this defence. This is further corroborated, by the brevity in which the Court dealt with the exception. Furthermore, the meaning of the sentence 'None of the exploitation of the products or technology complained of was conducted by Quadrant for its own experimental purposes.' does not appear to be clear. Some commercial activities which fall within the users own experimental purposes could still qualify for the experimental use defence. However, what constitutes 'for its own experimental purposes' is left to be interpreted by future judgements. The other alternative is to read the phrase: 'In all cases Quadrant was trying to exploit and sell its technology to third parties.' As meaning that any commercial activity would make the experimental use defence inapplicable. It would be difficult for a commercial organisation to carry out experiments for 'non-commercial purposes' as their ultimate aim and objective is to exploit and sell their technology commercially.

If Cordery's interpretation is true, then the question which has to be asked is whether universities' research activities in the UK would be considered in the same way as in Duke University v Madey as synonymous to commercial activities, but if so, could they be granted a special exemption from patent infringement.

Purposes must relate to the subject matter of the invention

In Auchinloss v Agriculture & Veterinary Supplies 36 , the Court of Appeals had to consider whether a given experiment fell within the 'subject matter of the invention' or not. The Court, relying on the Smith Kline & French case, held that a broader interpretation must be given to the term 'subject matter of the invention' which must be ascertained from the patent as a whole and not only from the claims.