UBMITTED FOR THE
TANDARDS OF THE
Carlos María Correa
University of Buenos Aires
In August 1995, the South Centre became a permanent intergov-
ernmental organization of developing countries. In pursuing its
objectives of promoting South solidarity, South-South co-operation,
and coordinated participation by developing countries in interna-
tional forums, the South Centre prepares, publishes and distributes
information, strategic analyses and recommendations on interna-
tional economic, social and political matters of concern to the South.
For detailed information about the South Centre see its website
www.southcentre.org and in particular the publication A Guide to the South
The South Centre enjoys support from the governments of its
member countries and of other countries of the South and is in
regular working contact with the Group of 77 and the Non-Aligned
Movement. Its studies and publications benefit from technical and
intellectual capacities existing within South governments and institu-
tions and among individuals of the South. Through working group
sessions and consultations that involve experts from different parts
of the South, and also from the North, common challenges faced by
the South are studied and experience and knowledge are shared.
This “South Perspectives series” comprises authored policy
papers and analyses on key issues facing developing countries in
multilateral discussions and negotiations and on which they need to
develop appropriate joint policy responses. It is hoped that the pub-
lications will also assist developing country governments in formu-
lating the associated domestic policies which would further their
Protection of Data Submitted for the Registration of Pharmaceuticals: Implementing the
Standards of the Trips Agreement, by Carlos Correa was first published in June
2002 by the South Centre in collaboration with the Department of Essential
Drugs and Medicines Policy of the World Health Organization. Reproduc-
tion of all or part of this publication for educational or other non-
commercial purposes is authorized without prior written permission from
the copyright holder provided that the source is fully acknowledged and any
alterations to its integrity are indicated. Reproduction of this publication for
resale or other commercial purposes is prohibited without prior consent of
the copyright holder.
South Centre, POB 228, Chemin du Champ-d'Anier 17, 1211 Geneva 19,
© South Centre, 2002
Printed and bound by SADAG.
ISBN 92-9162-011-6 Paperback
ISSN 1607-5323 Paperback
Executive Summary.............................................................................................. ix
Introduction ........................................................................................................ xi
EQUIRED FOR THE
A. Approaches to Data Protection ....................................................... 5
B. National Practices before TRIPS..................................................... 7
A. Protection of test data under the TRIPS Agreement................... 13
B. The Article 39.3 conditions of protection...................................... 14
1. Data necessary for marketing approval.............................. 14
2. Protected data ......................................................................... 15
3. Undisclosed data..................................................................... 15
4. New chemical entities............................................................ 16
5. Considerable effort (investment)......................................... 18
A. The TRIPS Agreement text.............................................................. 25
1. “Unfair”.................................................................................... 25
2. “Commercial” ......................................................................... 28
3. “Use” ........................................................................................ 30
4. Analysing “Unfair Commercial Use”.................................. 31
B. National case law................................................................................. 33
ISTORY OF THE
Annex I. Exclusive Use of Data and Compensation under the U.S.
Federal Insecticide, Fungicide and Rodenticide Act (FIFRA).... 59
The South Centre published in October 2000 a study by Carlos
Correa entitled Integrating Public Health Concerns into Patent Legislation in
Developing Countries. Second edition of the study was issued in Sep-
tember 2001, and the French and Spanish translations were pub-
lished, also in 2001. The study, which was prepared with the collabo-
ration of the World Health Organization, has been given wide circu-
lation in the developing countries and has been found very useful by
them in dealing with public health and patent legislation in the con-
text of TRIPS implementation.
The current study by Carlos Correa entitled Protection of Data
Submitted for the Registration of Pharmaceuticals: Implementing the Standards
of the TRIPS Agreement is continuation of the same work, focusing in
depth on the specific issue of marketing approval protection under
the TRIPS Agreement and its interpretation, an issue of major prac-
tical importance for the developing countries. The World Health
Organization has agreed to co-publish this study.
As the earlier study, the current study was produced with the
support of the Rockefeller Foundation. The author benefited from
an extensive discussion on data protection issues with Prof. Jerome
Reichman, whose valuable inputs the author wishes to acknowledge.
The author also wishes to acknowledge the support from WHO,
which organized a consultation meeting in New York, on 22 Octo-
ber, 2001 to review the draft of the study, at which participated Al-
fred Engelberg, Trevor Cook, Jim Keon, James Love, Jerome
Reichman, Robert Weissman and German Velasquez. The author is
thankful for the comments received during this meeting. He is also
grateful for the comments by Octavio Espinosa (WIPO), in his per-
sonal capacity, Adrian Otten (WTO) and Jayashree Watal (WTO).
The text was edited by Robert Weissman.
The author, Dr. Carlos Correa, is the Director of the Masters
Programme on Science and Technology Policy and Management at
the University of Buenos Aires, Argentina. He participated in the
negotiations of TRIPs during the Uruguay Round and has actively
pursued the issue of the global intellectual property regime and its
implications for development and for the countries of the South.
Several of his studies and working papers have been published by the
The author is grateful for the support from the Rockefeller
Foundation for the preparation of this document.
Any views expressed are the views of the author and do not
necessarily reflect the views of the Rockefeller Foundation or of the
World Health Organization. The author is solely responsible for the
1. As a condition for registering pharmaceutical products, national
authorities normally require registrants to submit data relating
to drugs’ quality, safety and efficacy (“test data”), as well as in-
formation on the composition and physical and chemical char-
acteristics of the product. A particularly important issue is the
direct or indirect use of the data for subsequent registration of
products similar to those originally registered.
2. The World Trade Organization’s Trade Related Aspects of
Intellectual Property Agreement (TRIPS), Article 39.3, requires
member countries to establish protections for submitted test
data. But this requirement is in fact narrowly drawn, and coun-
tries maintain substantial flexibility in implementation. The
public interest in limiting protections for data is to promote
competition, and to ensure that data protections do not become
the means to block the timely entrance of generic competitors
to off-patent drugs. Generic competitors drive down price,
thereby promoting greater accessibility of medicines.
3. Article 39.3 requires governments to provide protection to
marketing approval data only under certain conditions. Test
data must be protected if national authorities require its submis-
sion. Article 39.3 does not require protection be given to al-
ready public data. Protection is required only for new chemical
entities. Members have considerable discretion in defining
“new,” and may exclude applications for second indications,
formulations and dosage forms. And, prior to granting protec-
tion, national regulatory authorities may request the applicant to
prove that the information for which protection is sought is the
result of significant investment.
4. Article 39.3 requires countries to protect against “unfair com-
mercial use” of marketing approval data. Countries have con-
siderable discretion to define “unfair” in the context of their
own national laws and culture. Use by the government to assess
the efficacy and toxicity of a pharmaceutical or agrochemical
product is not a commercial use subject to Article 39.3. Grant-
ing marketing approval to a second entrant, based on the sec-
ond product’s similarity to a previously approved first product,
is not a proscribed “use” under Article 39.3. These interpreta-
tions are supported by United States and Canadian Supreme
Court decisions interpreting national laws.
5. Countries can meet their obligations to protect against “unfair
commercial use” under Article 39.3 by barring “dishonest” uses
of test data. This would require, for example, proscribing situa-
tions in which a competitor obtains the results of testing data
through fraud, breach of confidence or other “dishonest” prac-
tices, and uses them to submit an application for marketing ap-
proval for its own benefit. It would also apply in cases where
the government provides access to undisclosed testing data in
order to provide an advantage to a firm which did not produce
them or share their cost.
6. Countries are not obligated under Article 39.3 to confer exclu-
sive rights on the originator of marketing approval data.
7. The pharmaceutical industry and some countries have argued
for much broader coverage of Article 39.3, and for a require-
ment that countries confer exclusive rights on originators of
marketing approval data. But these positions are not well
grounded in either the text or negotiating history of TRIPS.
TRIPS negotiators specifically considered and rejected language
requiring grants of exclusive rights to test data.
As a condition for registering pharmaceutical products, national
authorities normally require registrants to submit data relating to a
drug’s quality, safety and efficacy as well as to its physical and chemi-
cal characteristics. A particularly important issue is third parties’ use
of the data for subsequent registration of products similar to those
In some jurisdictions, the data submitted for the registration
of pharmaceutical (and agrochemical products), are subject to a sui
generis system of protection, based on a temporary right to the exclu-
sive use of such data by the first applicant (generally the company
that developed a new product). In such a system, other companies
(often “generics” manufacturers) cannot rely on the data submitted
by the first applicant for the purpose of registering a similar product
for commercial use. The rationale for this exclusivity model is to
permit the originator of data to recover the investments made for
their development. The underlying assumption is that, without such
protection, private firms would have no incentive to bear the consid-
erable costs of producing the required data.
In other countries, however, health authorities are permitted
to rely on data submitted by the first applicant to process and ap-
prove third parties’ subsequent applications for a similar product,
subject to evidence that its physico-chemical attributes are equivalent
to those of the first applicant’s product. This approach emphasizes
that the registration of products should not erect barriers to other-
wise legitimate competition. It holds, instead, that the registration
system should promote price competition and access to more af-
In some cases, national authorities are allowed to rely on the registration made
in a foreign country to approve subsequent applications.
The issue of data protection is especially relevant for off-
patent products as well as for products, such as biologicals, that are
often difficult to patent. In cases where the product is patented, the
patent holder can, in principle, exclude any competition during the
lifetime of the patent -- a period of exclusion which will generally run
longer than that afforded by data protections. Data protection rules
are of particular importance to many developing countries that until
recently did not provide patent protection for pharmaceuticals (and
to those under the transitional periods of the WTO’s TRIPS Agree-
ment, which still do not provide pharmaceutical patent protection).
In these countries, there is a large pool of unpatented pharmaceutical
products. Data protection systems could, if they provided exclusivity,
become a partial substitute for patent protection in these cases and
nullify, in practice, the transitional periods granted to developing
Before the entry into force of the TRIPS Agreement, coun-
tries had considerable latitude to determine rules for the protection
of test data. The Agreement introduced the first international standard
on the subject, as contained in its Article 39.3.
But the Agreement is
not a “uniform law” -- it only establishes broad parameters for na-
tional rules. An important question is the extent to which the
Agreement allows WTO Member countries freedom to apply differ-
ent approaches for the protection of test data protection and, in
particular, the extent to which a competitive model -- i.e., protection
without exclusivity -- is compatible with the minimum standards set
forth in Article 39.3.
To properly interpret Article 39.3, the Vienna Convention on
the Law of the Treaties instructs that the ordinary meaning and
context of the terms used, and the object and purpose of the treaty
The full text of Article 39.3 reads: “Members, when requiring as a condition of
approving the marketing of pharmaceutical or of agricultural chemical products
which utilize new chemical entities, the submission of undisclosed test or other
data, the origination of which involves a considerable effort, shall protect such
data against unfair commercial use. In addition, Members shall protect such
data against disclosure, except where necessary to protect the public, or unless
steps are taken to ensure that the data are protected against unfair commercial
must be carefully considered. The history of the negotiation is also
an important complementary element for interpretation (Article 31
(2) of the Vienna Convention).
The first section of this paper describes the different stages of
drug development and the testing required for marketing approval of
new pharmaceutical products. The second section discusses the ra-
tionale for test data protection. The third section examines the con-
ditions, established by Article 39.3 of TRIPS, under which protection
must be given to marketing approval data. The fourth section ex-
amines the concept of “unfair commercial use” of data -- the con-
duct proscribed by Article 39.3. The fifth section examines the legal
means that States may adopt to provide protection against commer-
cial use. The sixth section offers a brief analysis of the negotiating
history of Article 39.3, which provides the backdrop for interpreta-
tion of the TRIPS Agreement’s data protection obligations. A final,
concluding section assesses the obligations on countries to provide
marketing approval protection under the TRIPS Agreement, and
reviews the flexibilities available to Member countries.
EQUIRED FOR THE
The development of a new drug involves different stages, during
which a variety of data are produced.
The “discovery” stage involves the synthesis or isolation of
new chemicals. Initial screening tests determine whether the new
chemicals possess sufficient biological activity to be worthy of fur-
ther investigation. The nature of pharmaceutical research has
changed dramatically in the last twenty years with the application of
the “rational drug design” method and the use of combinatorial
chemistry. With discovery by design, scientists use knowledge about
the causes of human disorders, the properties of drug compounds,
and their action in the human organism to conceptualize the struc-
ture of an “ideal” molecule that is expected to restore the altered
equilibrium. Laboratory chemists then search for substances whose
molecular structures match as closely as possible the theoretical
model (Gambardella, 1995, p. 23). This methodology reduces the
cost of the “discovery” stage, but does not eliminate the need for
bioassay, animal and other tests of the new drugs.
Once a promising new chemical is identified, its non-toxicity
and efficacy must be confirmed. The testing procedures involve
different stages and phases (see Box 1).
On the basis of the results of these tests, national authorities
can assess whether to grant marketing authorization for a new
chemical entity. All the safety and efficacy tests must normally be
completed before the authority approves the product. The authority
may also require additional clinical tests. In 1980, the duration of
these studies varied from about 1 to 7 years and averaged slightly less
than 3 years. This period has been significantly reduced since then
(Raggett, 1996, p. 26).
2 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
Testing new drugs
In the preclinical stage, the new chemical entity (NCE) is tested in
animals to assess its pharmacodynamic, phamacokinetic and toxico-
logical profile. Results of these tests are studied carefully before tests
in human beings are carried out.
Safety and Efficacy Testing
The types of tests, the procedures to be used, and the standards to be
met to demonstrate safety and efficacy may vary among therapeutic
classes and even among drugs for use within a therapeutic class. This
stage includes different phases.
In Phase I chemical testing, a small group of healthy volunteers receive
dosages of the investigational drug for a short period of time. The
primary purpose is to look for evidence of toxicity or unexpected
undesirable reactions, and to study the bioavailability and pharmaco-
kinetics of the NCE/drug applied to patients.
Phase II of clinical testing has a similar purpose to phase I, but consid-
ering the therapeutic context. Its primary objective is to ascertain the
effectiveness of the investigational drug.
Phase III clinical trials are conducted on a large member of patients;
they often involve several hundred human subjects and are conducted
for substantial periods. These tests are designed to determine the
efficacy of the investigational drug and to uncover any unanticipated
side effects that the drug may have, considering age and gender influ-
ence, drug interactions and specific dosage for different indications.
While the phase III trials are under way, long-term animal toxicity
studies are undertaken to determine the effects of prolonged exposure
and the effects on subsequent generations. The duration of the studies
vary widely among therapeutic classes. For drugs that affect the
reproductive system or that will be used over long periods of time,
animal toxicity studies are typically expensive and lengthy.
Data Required for the Registration of Pharmaceuticals 3
In addition to test data, national authorities require informa-
tion on the quantitative and qualitative composition and other attrib-
utes of the product, as well as on manufacturing methods.
Marketing approval is generally granted for a specific drug
used for a specific therapy. Changing the composition of the drug,
combining it with other drugs in a single product or selling the drug
for a different therapeutic purpose requires new approval.
A. Approaches to data protection
A basic element of data protection is the obligation imposed on third
parties not to disclose the data; that is, to keep them confidential.
Some health specialists have argued against any concealment
of data submitted for the approval of pharmaceuticals (Olilla and
Hamminki, 1996, p. 169). In their view, non-disclosure contradicts
the right of the public to be informed about the efficacy and safety
of approved pharmaceuticals. According to this opinion, the con-
cealment of data on clinical, pharmacological and toxicological ex-
periments retards the development of knowledge, and poses risks
that consumers of a drug may be injured unnecessarily. Since confi-
dentiality prevents the scientific community from scrutinizing the
scientific basis of a licensing decision, it is not possible to determine
whether there is commercial bias in the information, or whether it
meets high standards. Drug companies have an interest in not pub-
lishing research that is not favourable to their products, and may
even try to hinder the publication of such studies (Dukes, 1996, p.
Other experts emphasize that health authorities should be able
to use and rely on registration data submitted for similar products, or
on the existence of a prior registration elsewhere.
If the regulatory
body is not free, when assessing a file, to use all the knowledge avail-
able to it, including data from other files and published information,
a great deal of repetitive toxicological and clinical investigation will be
In this case, the authority bases its decision on the fact that a foreign country
has granted registration, and on the proof of equivalence in terms of the physi-
cal and chemical characteristics and other relevant attributes of the product.
6 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
required, which will be wasteful and in the case of animal testing,
ethically questionable (Dukes, 1996, p. 146).
According to this position, when the authorities already know
the characteristics and effects of the product (due to the first regis-
tration), it is not rational from the society’s point of view to duplicate
tests to recreate existing information. All that the authorities need for
the second application is confirmation that the second product is
similar to the first product. How to prove similarity is a matter for
national regulation; some countries require bio-equivalence and bio-
availability tests, while others are satisfied with the proof of chemical
similarity and prior registration.
This position is also grounded in the pro-competitive effects
of low entry barriers for pharmaceutical products. If producers (par-
ticularly generics manufacturers) are obliged to repeat long and costly
testing, competition will be reduced because of time delays and,
more importantly, because some small and medium firms -- espe-
cially local firms in developing countries -- will lack the resources to
undertake such testing. This reduces competition and the afforda-
bility of medicines that -- by definition -- are off-patent and, there-
fore, should be broadly available at the lowest possible price.
The research-based industry has, however, argued for stronger
test data protection, using both equity and health policy arguments.
The industry position argues that the manufacturer has invested,
often heavily, in conducting tests and deserves a return on invest-
ment. Where patent law fails to provide protection (for example,
because the patent on an active component is shortly to expire, or
because the drug is based on a combination of known substances
used in novel manner) data exclusivity is a necessary barrier to com-
petitors rapidly producing and registering an exact copy of the drug.
In accordance with this view,
“equity demands that protection be provided for data, which can
cost the original submitter several million dollars to produce. Dis-
closing this data to the public or allowing its use by another appli-
cant unfairly denies the compiler of the data the value of its efforts
The Rationale for Data Protection 7
and grants an economic advantage to later applicants for market-
ing approval, enabling them to avoid the cost of developing test
data for their own products. Countries that allow such unfair ad-
vantages to later applicants discourage developers of new pharma-
ceuticals and agricultural chemicals from seeking to introduce their
state-of-the-art products in the country´s market. So, not only is
such protection required by the TRIPS Agreement, it is both eq-
uitable and wise from a public and health policy standpoint”
(Priapantja, 2000, p. 4).
Finally, consumer groups such as the Trans-Atlantic Consumer
Dialogue have proposed that, since data exclusivity is intended to
protect investment, companies seeking data exclusivity should be
required to disclose the amount actually invested. This would en-
hance transparency and allow the establishment of a relation be-
tween the actual investment and the protection provided (WHO,
2000, p. 40).
In the light of these contrasting approaches, a key issue is the
extent to which, under the TRIPS Agreement, Member countries are
obliged to provide exclusivity, and whether authorities can rely on
the data from a prior registration or on a registration made in a for-
B. National practices before TRIPS
Companies originating data for the registration of new products have
requested from national health authorities and generally obtained
protection of submitted data against disclosure. Confidentiality is
essentially intended to protect secret information from misappro-
priation by third parties. However, problems with secrecy in drug
regulation have historically raised public concern in several countries,
including Great Britain, New Zealand, Germany, Sweden and the
USA (Ollila and Hamminki, 1996, p. 168).
8 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
Historically, some health authorities relied on the first appli-
cation data for the evaluation of second-entrant applications for
similar products. Some companies brought legal action against the
authorities arguing that reliance on the knowledge derived from one
file to evaluate another one (e.g., a generic equivalent) caused them
In a number of court cases relating to Cimetidine decided in the
United Kingdom, Australia and New Zealand, first entrants origi-
nating registration data invoked the ordinary law of confidential
information to prevent regulatory authorities from relying on the
originator’s file when assessing an application for the approval of an
equivalent drug by a generic competitor. Courts were, however, re-
luctant to apply such law (Cook, 2000, p. 5).
As a result of industry lobbying, some developed countries
established sui generis protections for test data submitted for the ap-
proval of pharmaceuticals (and agrochemicals). Under different mo-
dalities, they adopted the concept of exclusive use of the test data by
the originator company. The U.S. adopted a regulatory data protec-
tion regime for pesticides,
and in 1984 regulatory exclusivity provi-
sions for medicines. The U.S. health registration regulations provide
for five years of exclusivity for new chemical entities, and three years
for data filed in support of authorizations based on new clinical re-
search relating to chemical entities which had already been approved
for therapeutic use.
This regime limits exclusivity by allowing third parties to use originator’s test
data if compensation is paid. In case of disagreement, the amount is determined
through arbitration. See in Annex I a summary of the relevant legislation.
In October 1997, the U.S. Senate held hearings on “Health Registration Data
Exclusivity, Biomedical Research, and Restrictions on the Introduction of Ge-
neric Drugs” (Subcommittee on Labor, Health and Human Services and Edu-
cation and Related Agencies, Committee on Appropriations). These hearings
considered a proposal for a voluntary five- year extension of the U.S. data ex-
clusivity period, coupled with a 6 percent R&D commitment from the com-
pany electing to take the extension. The U.S. Congress did not adopt this pro-
The Rationale for Data Protection 9
In the European Union (EU), the Member States have pro-
vided exclusivity protection for the data filed in support of marketing
authorizations for pharmaceuticals since 1987. One of the original
objectives of this regime was to compensate for the lack of patent
protection for pharmaceuticals in some Members States (Portugal,
Spain), but it was maintained after those countries introduced such
protection (Watal, 2001, p. 201). During the exclusivity period,
health authorities cannot rely on an originator’s test to approve other
applications without the originator’s consent.
The minimum period
Article 8 of Directive 65/65 , as amended by Directive 87/21/EEC, estab-
lishes that “without prejudice to the law relating to the protection of industrial
and commercial property:
(a) The applicant shall not be required to provide the results of pharmaceutical
and toxicological tests or the results of clinical trials if he can demonstrate:
i. either that the proprietary medicinal product is essentially similar to a
product authorized in the country concerned by the application and
that the person responsible for the marketing of the original proprie-
tary medicinal product has consented to the pharmacological, toxico-
logical or clinical references contained in the file on the original pro-
prietary medicinal product being used for the purpose of examining
the application in question;
ii. or by detailed references to published scientific literature presented in
accordance with the second paragraph of Article 1 of Directive
75/318/EEC that the constituent or constituents of the proprietary
medicinal product have a well established medicinal use, with recog-
nized efficacy and an acceptable level of safety;
iii. or that the proprietary medicinal product is essentially similar to a
product which has been authorized within the Community, in accor-
dance with Community provisions in force, for not less than six years
and is marketed in the Member State for which the application is
made; this period shall be extended to 10 years in the case of high-
technology medicinal products within the meaning of Part A in the
Annex to Directive 87/22/EEC or of a medicinal product within the
meaning of Part B in the Annex to that Directive for which the pro-
cedure laid down in Article 2 thereof has been followed; furthermore,
a Member State may also extend this period to 10 years by a single
Decision covering all the products marketed on its territory where it
considers this necessary in the interest of public health. Member states
are at liberty not to apply the abovementioned six-year period beyond
the data of expiry of a patent protecting the original product. How-
ever, where the proprietary medicinal product is intended for a differ-
ent therapeutic use from that of the other proprietary medicinal prod-
10 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
of such protection is six years, but 10 years is obligatory for “high
technology products” (most biotechnology products), and also for
new chemical entity authorizations granted by the European Medi-
cines Evaluation Agency (EMEA). EMEA may also grant 10 years
exclusive protection for test data related to medicines administered
by means of “new delivery systems which … constitute a significant
innovation”, and “medicinal products containing a new substance or
an entirely new indication which…is of significant therapeutic inter-
est” (Cook, 2000, p.18).
Most Member States (Belgium, France, Germany, Italy, the
United Kingdom, the Netherlands and Sweden) have applied the 10-
year period to all medicinal products (Dodds Smith, 2000, p. 113).
Moreover, the “data exclusivity that this affords can, if a marketing
authorization is obtained only late in the life of a patent, extend be-
yond patent expiry. The only qualification to this is an option avail-
able to those few Member States which have not availed themselves
of the 10-year period for all medicinal products, and which can also
elect for such data exclusivity ‘not to extend beyond patent expiry’”
(Cook, 2000, p.18).
Article 1711 of the North American Free Trade Agreement
(NAFTA) of 1992 also establishes an exclusivity standard, requiring
signatory countries to provide a minimum five years exclusivity pe-
riod counted from the date of marketing approval. This model was
followed in 1993 by the Andean Group countries under Decision
344 (“Common Regime on Industrial Property”).
At the time of conclusion of the TRIPS Agreement, few
countries had adopted the exclusivity approach developed in the
ucts marketed or is to be administered by different routes or in differ-
ent doses, the results of appropriate pharmacological and toxicological
tests and/or of appropriate clinical trials must be provided.
(b) In the case of new proprietary medicinal products containing known con-
stituents not hitherto used in combination for therapeutic purposes, the re-
sults of pharmacological and toxicological tests and of clinical trials relating
to that combination must be provided, but it shall not be necessary to pro-
vide references relating to each individual constituent”.
The Rationale for Data Protection 11
United States and Europe. At the time, most countries in the world
did not provide for exclusivity and most allowed the national health
authorities to rely on test data submitted by the first applicant to ap-
prove subsequent applications on “similar” products.
countries (e.g.. Argentina, Singapore, Taiwan, and the territory of
Hong Kong) it was sufficient to prove that a similar product had
been approved or commercialized in a foreign country.
Though the time of the adoption of the Agreement is to be taken into ac-
count, according to general principles of international law, for the interpretation
of its obligations, it should be noted that even today, after the expiration of all
except the transitional period for LDCs, only a minority of the WTO Members
apparently confer data exclusivity (see, e.g. the February 2000 Pharma submis-
sion to the USTR on Section 301, at www.pharma.org). New Zealand introduced
an exclusivity period in 1994, as part of implementing legislation of the TRIPS
Agreement, and Australia did it in 1998 as a result of U.S. action under “Special
301” of U.S. Trade Act. The Andean Group countries, instead, revised Decision
344 in 2000 and eliminated the exclusivity period. A special exclusivity granted
under the “Safety Monitoring Program” in Thailand was also abolished in Janu-
The TRIPS Agreement establishes a minimum international standard
for the protection of marketing approval data. WTO Member coun-
tries need to determine what is actually needed to fulfil their obliga-
tions under the Agreement. Understanding the obligations imposed
by Article 39.3 requires a close reading of the text, an assessment of
each of its components, as well as a review of the negotiating history
and national practice. The remainder of this paper turns to these
A. Protection of test data under the TRIPS Agreement
The inclusion of test data as a category of intellectual property in
TRIPS does not mean countries must provide exclusivity protections
for such data.
According to Article 1.2 of the TRIPS Agreement, the pro-
tection of test data is a category of “intellectual property” like pat-
ents, copyrights and trademarks. The structure of Article 39 suggests
that the regime for test data has been conceived by the negotiating
parties as a particular case in the framework of the protection of “un-
disclosed” information. In this sense, the protection conferred can-
not be properly deemed a sui generis system.
The categorization of test data as a subject matter of “intel-
lectual property” does not mean that Article 39.3 puts their protec-
tion on the same footing as other intellectual property rights. In
particular, it cannot be inferred that such protection requires exclusive
rights. Though in most instances intellectual property rights confer a
ius excluendi, this is far from being an absolute rule. It is well accepted,
for example, that trade secrets protection in the framework of unfair
14 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
competition does not give rise to a right to exclude. Nor does the
protection of geographical indications under the TRIPS Agreement
entail the granting of such faculty.
Likewise, there are many situa-
tions in which copyright protection only allows the title-holder to
claim remuneration, but not to prohibit unauthorized acts.
As Article 39.3 itself indicates (see below), test data protection
is a reward for the investment in data production, rather than for the
creativity or inventiveness involved in generating the data. Test data
are developed in accordance with standard protocols and proce-
dures, involving a systematic compilation of factual information.
Though the testing may refer to a novel drug, the test results them-
selves are merely the outcome of routine scientific practices.
Thus, the inclusion of test data in the TRIPS Agreement as a
category of “intellectual property” does not determine the nature of
the protection conferred. In particular, it does not indicate that such
data should be protected through grant of exclusive rights.
B. The Article 39.3 conditions of protection
1. Data necessary for marketing approval
A basic premise for the application of Article 39.3 is that test data
must only be protected if national authorities require their
submission for obtaining marketing approval of pharmaceuticals or
agrochemical products. The first sentence of this article states:
“Members, when requiring , as a condition of approving the mar-
Given the territoriality of the intellectual property system -- a feature
that the TRIPS Agreement has not altered -- the obligation to pro-
tect test data only arises in the Member countries where national
See article 22.2 of the TRIPS Agreement.
Conditions of Protection Under TRIPS 15
regulations require the submission of such data. If a Member coun-
try opts not to require those data, Article 39.3 will be not apply.
In addition, the submission of data must be necessary to obtain
approval. Data voluntarily submitted by an applicant, or in excess of
what is required for approval, are not subject to protection under
2. Protected data
The subject matter of the protection under this article is written
material which details the results of scientific health and safety test-
ing of drugs and agrochemicals, in relation to human, animal and
plant health, impact on the environment and efficacy of use. The
provision covers tests and other data that may be required by the
authorities. These “other” data may include, for instance, manufac-
turing, conservation and packaging methods and conditions, but only
to the extent that submission of this information is necessary to obtain
3. Undisclosed data
Article 39.3 does not require protection be given to public data sub-
mitted for marketing approval. To qualify for protection under Arti-
cle 39.3, the pertinent information must be “undisclosed”. This
means that information that is already public does not fall within the
scope of this article. Any requirement for the submission of pub-
lished or otherwise disclosed information to national regulators shall
not generate any private right limiting the use of such information by
the government or third parties, since the information would be in
the public domain.
While a substantial part of the information on tests relating to
safety and efficacy of approved drugs becomes publicly available --
16 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
because the information is published in scientific journals,
public by the health authority,
-- many data remain confidential
such as data relating to some of the product’s physical and chemical
attributes and manufacturing processes.
Given that under Article 39.3 protection is only conferred on
undisclosed information, it will be necessary to determine in cases of
controversy which of the information accompanying an application
for marketing approval is confidential and need to be protected, and
which is not. The undisclosed or disclosed nature of information is
an objective feature, and it is not dependent on the qualification given
by the applicant to the information that it is submitted. Hence, any
applicant’s declaration that all or certain information is “confiden-
tial” or “undisclosed” should be subject to scrutiny.
4. New chemical entities
Another important condition for the application of Article 39.3 is
that the data must refer to a “new chemical entity”. The Agreement
does not define the term “new”. While the term presumably does
not impose a patent standard of novelty, Member countries may
choose under the Agreement to apply such a standard.
It may be also held that the test for newness under Article
39.3 refers to the date of application for approval. Thus, a chemical
entity may be deemed “new” if there were no prior application for
In the case of the EU regulations (Directive 65/65, as amended) the possibility
of obtaining market approval on the basis of published literature has been
interpreted very restrictively. It only applies where a product has a well estab-
lished medicinal use and the documentation submitted by the applicant covers
all aspects of the safety and efficacy assessment (Dodds-Smith, 2000, p. 111).
For instance, the European Medicines Evaluation Agency (EMEA) publishes
summaries of clinical trials in the “European Public Assessment Report”
(EPAR) However, no detailed information on toxicological/pharmaceutical
tests or clinical trials is published which could be used for registration by an-
other company. The manufacturing process is not published either.
Conditions of Protection Under TRIPS 17
approval of the same drug, or where the same drug was not previ-
ously known in commerce.
Article 39.3 does not clarify either whether newness should be
absolute (universal) or relative (local), that is, whether “new” would
mean the first application in the world or in the Member country
where it was filed (Cook, 2000, p.6).
Occasionally, a product which is known and used in a certain
field (e.g. chemical industry), may find a new application in the
pharmaceutical sector. Such a new therapeutic product (generally
known as “first indication”) may be deemed not to constitute a “new
chemical entity”, since the chemical was already known. Alterna-
tively, the newness may be assessed within a particular regulatory
framework, and without regard to the fact that the same chemical
may have been used in the context of another regulatory framework
(Cook, 2000, p. 6).
All the above interpretations are equally permissible. The
TRIPS Agreement deliberately avoids defining the concept of “new
chemical entity”. This is one of the clear areas in which Member
countries enjoy room for manoeuvre to implement the Agreement’s
Based on the ordinary meaning of the terms used, it may be
also interpreted that there is no obligation to provide for protection
when the test data were developed for a new use of a pharmaceutical
product (generally called a “second indication”). In this case, it is the
application or method of use of a known chemical entity that is new,
but not the entity as such.
Similarly, Article 39.3 would not apply in cases where approval
is sought for new indications, dosage forms, combinations, new
forms of administration, crystalline forms, isomers, etc. of existing
drugs, since there would be no novel chemical entity involved. The
European Court of Justice indirectly addressed this issue in the
18 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
The Court held that a (second) product is “essentially
similar” to an earlier approved product if the second product has
“the same qualitative and quantitative composition in terms of active
principles”, “the same pharmaceutical form” and is bio-equivalent
to the first product, “unless it is apparent in the light of scientific
knowledge that it differs significantly from the original product as
regards safety or efficacy”. In these cases, the original applicant does
not receive new periods of so-called “marketing exclusivity” for each
new indication, dosage form or dosage schedule (Jones and Nitten-
berg, 1998/1999, p. 152).
5. Considerable effort (investment)
The subject matter of the protection under Article 39.3 is test data
which cover matters such as toxicology, clinical trials for the phar-
maceuticals and field trials for agrochemicals. Because this informa-
tion is not “invented” or “created”, the TRIPS Agreement does not
define any substantive standard for granting protection (like inven-
tive step or novelty). It simply mandates protection when the process
of obtaining the data involved “a considerable effort”.
The text is vague about the type of effort involved (technical,
economic?) and also with respect to its magnitude (when would it be
deemed “considerable”?). As mentioned, the proponents of this
formulation intended to protect the investment
made in producing test
The ECJ decision was given in response to questions referred to it from the
English High Court in relation to three cases. In all of them, the research-based
pharmaceutical companies had made changes to certain aspects of their prod-
ucts and obtained marketing approval for each change. Subsequently, generic
companies sought to rely not only on the original versions of the products but
also on the products which had been approved more recently. The Medicines
Control Agency acceded to certain of the generic companies requests, but not
all of them (Jones and Nittenberg, 1998/1999, p. 152). See also Dodds-Smith,
2000, p. 112.
Conditions of Protection Under TRIPS 19
The extension of intellectual property beyond its boundaries
so as to protect investment, and not intellectual contributions
rupts the essence of a system conceived to reward the creators of
original ideas and new inventions.
Even if it may be argued that
“free riding” or “unfair use” of such data by third parties may create
unfair advantages or unjust enrichment, it is not the role of the in-
tellectual property system to solve competition problems that do not
relate to the creation or use of ideas. Nonetheless, Article 39.3 ex-
ists. And it includes the considerable effort standard. Inclusion of
this standard suggests national regulatory authorities may request the
applicant prove that the information for which protection is sought
is the result of considerable effort.
An investment-based system was adopted by the European Community in
the form of a sui generis regime for the protection of data bases. Despite the
efforts of WIPO, however, no agreement has been reached so far to adopt an
international convention modeled on the European approach. A bill on the
matter proposed in the United States has also found strong opposition, par-
ticularly from the scientific and librarian communities (Reichman and Uhlir,
According to the Trans Atlantic Consumer Dialogue (TACD), “data exclu-
sivity provisions are part of a growing class of sui generis forms of protection that
are designed to protect investment, rather than innovation. Because data exclu-
sivity is not a reward for invention (which is already rewarded by patents) but
rather a protection of investment, there should be greater transparency of the
basis for the protection and a reasonable relationship between the investment
and the protection” (available at www.tacd.org).
Since the TRIPS Agreement’s obligations with regard to test data
protection relates exclusively to undisclosed information, it seems
clear that WTO Members’ obligations are limited to information,
effectively requested by and submitted to the government, which
was at the time of submission, and later remains, “undisclosed”.
The non-disclosure obligation requires that the test data be
protected against “disclosure” unless:
a) it is necessary to protect the public; or
b) steps are taken to ensure that the data are protected
against unfair commercial use.
The application of the first exception is subject to a “necessity test”.
In determining necessity, GATT/WTO rules and jurisprudence gen-
erally provide deference to Member countries to determine when a
necessity arises, but impose an often heavy burden of proof on the
Member invoking it (Trebilcock and Howse, 1999, p. 140; Correa,
The second exception would permit a Member to disclose any
information, if its unfair commercial use can be prevented. The key
questions are what constitutes unfair use and how that protection
can be guaranteed. This issue is discussed below.
Article 39.3 aims at preserving the confidentiality of the in-
formation submitted for marketing approval without any time limit.
There is no indication in the provision about the duration of the
obligation, certainly a weak point in the text. In principle, the confi-
dentiality obligation continues until the information becomes known.
It may also be possible, however, for a Member to establish a maxi-
mum period of confidentiality.
22 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
In any case, as mentioned above, because of the public health
implications of the release into the market of a new drug, a substan-
tial part of, but not all, the results of safety and efficacy tests and
other data become available to the public. Some public health spe-
cialists have strongly opposed the possibility of keeping confidential
pharmaceutical data, such as information obtained during pre-clinical
tests. It has been argued that
“The earliest point in the career of the drug when one obtains a
glimpse as to which its adverse effects might be is, without doubt,
the phase of pharmacological and toxicological studies in animals.
Very properly, the community requires of the pharmaceutical in-
dustry that the work performed at this stage be conscientiously car-
ried out and painstakingly reported when the drug is submitted to
Drug Control Authorities…Very improperly, the community
then goes on to tolerate a situation whereby these reports, having
been used for this purpose, are then commonly deposited in confi-
dential archives where they are inaccessible to the medical world at
large…It follows that when the first clinical evidence of a particu-
lar and unexpected side effect reaches us there is often no simple
and direct means of comparing it with what has been reported in
dogs, rabbits and mice. If these data were public property, it might
be simpler to identify at an early stage those adverse reaction re-
ports from the clinics which, because they run parallel to animal
findings, deserve particular attention….” (Dukes, 1977).
Public health concerns were only marginally present in the negotia-
tion of the TRIPS Agreement.
The non-disclosure obligation was
established on the basis of commercial considerations, without a
proper weighing of public health interests in the openness of drug
information (see Box 2).
See, e.g., Article 8.1.
Non-Disclsoure Obligations 23
The importance of access to information
Full availability of information is essential if all parties involved in health
care are to participate effectively. Openness facilitates adequate feed-
back, proper setting of priorities and development of trust. A culture of
openness protects conscientious individuals working in organizations of
Knowledge relating to all drugs evolves constantly, as do standards
and expectations relating to them, their producers and health care pro-
viders. However thorough the investigations made before a drug is
licensed and marketed, much more will be learned about its efficacy,
proper use and risks once it is marketed and used on a much larger
Almost no new element of knowledge emerges suddenly; as a rule
it begins with impressions and hypotheses. Where these arise – for ex-
ample, in reports of possible serious side effects in the journals -- all
existing relevant information will need to be mobilized to verify or dis-
count this evidence so that the trust can be established as quickly as
possible. Much of the information needed for that purpose, including
data on both animal and human experience, is unpublished and lies only
within the files of agencies. By using it, the truth can be established
much more quickly than if one is reliant purely on published evidence.
Extracts from the Statement of the “International Working Group on trans-
parency and accountability in drug regulation “(Uppsala, 11-14 September,
24 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
Box 2 (continued)
Consequences of excessive secrecy in drug regulation
If a substantial part of the information existing on drugs remains hidden
within regulatory agencies, and sometimes fragmented between them,
the development of knowledge will be impeded. This is particularly
dangerous where suspicion arises of a hitherto unknown risk.
Malpractice can be hidden from view; legal discovery in the course
of litigation has for example revealed cases of falsification or suppres-
sion of unfavorable data by certain companies, or submission of incon-
sistent files on the same drug to different agencies. Secrecy facilitates the
circulation and use of sub-standard drugs.
Where a drug is subject to negative findings, the failure of a drug
agency to explain its conclusions or provide background data, can leave
the way clear for the sometimes very different and emphatic account
given from the manufacturer. In a climate of secrecy and mistrust, the
public is unlikely to believe even accurate and meticulously prepared
official statements – assuming that they cannot be taken at face value
and that some relevant information has probably been withheld.
The incomplete availability and irregular release of information
promotes a climate in which suspicion is generated and in which sensa-
tional and poorly founded stories on drugs break in the popular press,
their reliability cannot be checked and unnecessary panic can be caused.
Secrecy has consequences which can be wasteful and even inhu-
mane; scientific work, e.g., in humans or animals which has already been
performed by one company but hidden within regulatory files, may be
If drug utilization data are not available irrational drug use may
continue unrecognized and unchecked.
If research is sponsored by companies, unfavourable or unclear re-
sults may be withheld or the research itself may be stopped.
A. The TRIPS Agreement text
One of the crucial interpretative issues in Article 39.3 is whether the
reliance by a national authority on data submitted by one company
(the “originator”) to evaluate a subsequent application by another
company (a “follower”), constitutes an “unfair commercial use” of
The expression “unfair commercial use” is not defined in Article
39. Pursuant to Article 31 (1) of the Vienna Convention, its inter-
pretation should be based on the ordinary meaning of the terms of
the treaty in their context and in the light of the agreement’s object
The ordinary meaning of “unfair” is “not equitable or honest or impartial
or according to rules”
. In the case of Article 39.3, this concept must be
understood in the light of Article 10bis of the Paris Convention.
The concept of “unfair” is relative to the values of a particular
society at a given point in time. It varies among Members, and this
variation is in fact one of the premises on which the discipline of
unfair competition is grounded. There is no absolute, universal rule
to determine when certain practices should be deemed “unfair”:
“Morality, which is the source of the law of unfair competition, is
a simple notion in theory only. In fact it reflects customs and hab-
its anchored in the spirit of a particular community. There is no
The Concise Oxford Dictionary, Seventh Edition, Oxford University Press, Ox-
26 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
clearly objective standard of feeling, instincts, or attitudes toward a
certain conduct. Therefore, specific prescriptions involving uniform
evaluation of certain acts are extremely difficult.
The pressures existing in the various countries for the suppression
of acts of unfair competition differ greatly. Generally, the develop-
ment of law of unfair competition depends on active and intense
competition in the marketplace by competing enterprises. It is the
pressure of conflicting interests which leads to the establishment of
clear rules of law. This pressure is not uniform in all countries and
indeed it is evolving continuously” (Ladas, 1975, p. 1685-
Ladas concludes his treatise’s discussion of the issue by indicating
We look for a standard by which we may judge the act complained
of. This is an objective standard: the honest practices in the course
of trade in the particular community and at the particular time”
(Ladas, 1975, p. 1689).
Given this diversity, it is likely that different countries will judge
certain situations differently, depending on their values and competi-
tive advantages. Some countries may consider it an “unfair practice”
for a “follower” company to commercially benefit from the data
produced by the originator, via a marketing approval system based
on “similarity”; or hold that such commercial benefit gives rise to
claims of “unjust enrichment” leading to a compensation for the use
of the data. In others, it may be regarded as the legitimate exploita-
tion of an externality created during legitimate competition in the
market. As noted by Kamperman Sanders,
“Where exploitation of another’s achievements becomes inequita-
ble, unfair competition law acts provides a remedy. This means
that the mere fact that another’s achievement is being exploited
does not call for any impediment on the basis of unfair competition
provisions. On the contrary, appropriating and building on others’
Proscribed Acts of Unfair Comemrcial Use 27
achievements is the cornerstone of cultural and economic develop-
ment. The axiom of freedom to copy epitomizes the principles of
the free market system”.
Certainly, specific regulations could be adopted at the international
level in order to harmonize the treatment of these cases. The United
States made such a proposal in the TRIPS negotiations,
but it was
not incorporated into the final text of the TRIPS Agreement. The
U.S. proposal would have obliged countries to prevent any use of
test data, without the consent of the right holder or on payment of
“the reasonable value of the use”, if that use led to the “commercial
or competitive benefit of the government or of any person”. This
provision would have obliged countries to prevent any practice that
would create such benefit. The final proposal, by contrast, used the
term “unfair commercial practices”. The rejection of the US pro-
posal indicates that the negotiating parties deliberately opted under
Article 39.3 to mandate regulation of certain types of practices (those
that are commercially unfair) and not to prevent any practice based
on its possible effects on benefits allocation.
In other words, Article 39.3 only applies when a competitor
obtains a benefit or advantage from the use of the originator’s testing
data as the result of unfair commercial practices. It is the qualification of
the practice that counts, not the mere existence of an advantage or
benefit. Such qualification is left to Members’ discretion; it is part of
the room for manoeuvre that they retained when signing the Agree-
There are many instances in which the production of goods,
notably intangibles, in a competitive environment generate external-
ities that benefit competitors. In describing the nature of competi-
tion, Ladas has noted that:
“it is an undeniable fact of modern business life that successful
manufacturers or traders have to cope with the danger of having the
goodwill of their business, their connection with the purchasing
See below the history of the negotiation of article 39.3.
28 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
public, interfered with by competitors...In a competitive economy is
it to be expected that each manufacturer or trader necessarily seeks
to maintain and improve his market position by obtaining the
benefit of a public demand, even though this demand be created by
other manufacturers or traders...
“…where does lawful competition end and unlawful competition
begin? The fact that a competitor may derive a profit from his act
of competition or cause monetary loss to another is not, in itself,
unlawful. The dictum “no one should reap where he has not sown”
requires delicate application. Progress would be paralyzed and mo-
nopoly would become general if we should attempt to prevent per-
sons from using the work or experience of others. We must encour-
age people in the same trade or industry to compete for the custom
of the public on the most favorable terms. The issue is whether the
means employed in such competition are fair and lawful. An act
may lack tact or taste but not be dishonest” (Ladas, 1975, pp.
1676, 1677 and 1689).
Many countries do not treat commercialization of a “similar” prod-
uct approved by reference to a previous registration, or by reliance
on data submitted by the originator company, as an unfair commer-
cial practice, but some do. Under Article 39.3, each approach is valid.
Article 39.3 mandates protection against “unfair commercial prac-
tices”, but permits Member countries to determine which practices
will be deemed commercially unfair. As mentioned, differences
among countries are likely to exist, consistent with Article 10bis of
the Paris Convention.
Article 39.3 only covers “commercial” uses. This requirement clearly
excludes use by the government, notably by the national health
authority to assess the efficacy and toxicity of a pharmaceutical or
Proscribed Acts of Unfair Comemrcial Use 29
In the view of the European Union, however, there is a sub-
stantial difference between the underlying principle in Article 39.1,
which refers to relationships between competitors and Article 39.3,
which includes governmental acts:
“The main question of interpretation is what is meant by “unfair
commercial use”. Clearly, this concept is different from the concept
of “unfair competition”, as used in Article 39.1 with a reference
to Article 10bis of the Paris Convention on the protection of In-
dustrial Property, and which relates to behaviour among competi-
tors. Protection of registration data is a government function. Arti-
cle 39.3 does not indicate whether the notion of “unfair commercial
use” refers to unfair commercial use by generic manufacturers to
those who have submitted the data (usually research-based phar-
maceutical industry) or to use by regulatory authorities of these
data to the benefit of competitors. Protecting data against “unfair
commercial use” is also different from protecting them from disclo-
sure, since the latter is a separate and distinct obligation under
Article 39.3” (EU, 2001, p.3).
The EU argument, however, disregards that Article 39 develops and
does not add to Article 10bis of the Paris Convention. It only incor-
porates examples of the general principle contained in paragraph (2) of Article
In addition, though the use by the governments will indirectly
have commercial consequences (the entry of a competitor in the
market), it does not represent a commercial activity as such, but a le-
gitimate State practice. In order to be “commercial”, the use of the
information should be made by an entity which is actually in com-
merce. As also noted by Ladas,
“The general clause of Article 10bis, in establishing as its foun-
dation “honest usages,” looks to the relations between competitors
and to the interests of customers, and these provide an objective test
which reflects an evolving pattern of competition in most of the pre-
sent world...By definition, competition in commerce refers to the ef-
forts of two or more persons, acting independently, to secure the
30 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
custom of third parties, with the results that one may increase the
sale of his goods and reduce the sale of the goods of the other”
(Ladas. 1975, p. 1688).
The same concept underlies the WIPO “Model Provisions on Pro-
tection Against Unfair Competition” which, in relation to data pro-
tection, suggests the adoption by national laws of the following pro-
“Use or Disclosure of Secret Information Submitted for
Procedure of Approval of Marketing:
Any act or practice,
in the course of industrial or commercial activities, shall be con-
sidered an act of unfair competition if it consists or results in an
unfair commercial use of secret test or other data, the origination
of which have been submitted to a competent authority for the
purposes of obtaining approval of the marketing of pharmaceuti-
cal or agricultural chemical products which utilize new chemical
entities” (emphasis added) (WIPO, 1996).
Finally, for Article 39.3 to apply, there must be “use” of the infor-
mation submitted by the originator.
In one of the texts under consideration by the negotiating parries in July
1990, the broader concept of “exploitation” was proposed (but not finally
adopted). The text read:
“3Aa. Parties, when requiring the publication or submission of undisclosed informa-
tion consisting of test [or other] data, the origination of which involves a considerable
effort, shall protect such data against unfair exploitation by competitors. The protec-
tion shall last for a reasonable time commensurate with the efforts involved in the
origination of the data, the nature of the data, and the expenditure involved in their
preparation, and shall take no account of the availability of other forms of protec-
Proscribed Acts of Unfair Comemrcial Use 31
4. Analysing “Unfair Commercial Use”
Thus, given the flexibility inherent in Article 39.3, and depending on
the applicable legal system, national laws can follow different ap-
proaches for the approval of a second-entry marketing application.
a) require the second-entrant to produce its own testing and
other data or to obtain an authorization of use from the
“originator” of the data;
b) allow the second-entrant to rely on the “originator’s”
data against payment of a compensation to the “origina-
tor” (when the “originator” has not given his consent for
the use of the data);
c) examine and rely upon the data submitted by the “origi-
nator” to evaluate the second-entrant application;
d) approve a second entry marketing application without
examining or otherwise relying upon confidential infor-
mation submitted by the originator.
In all cases, the authorities will normally require that the second-
entrant prove that his product is similar or “essentially similar” to the
already registered product (in terms of its physical and chemical
characteristics and attributes).
Different types of bioequivalence
studies are generally required for this purpose.
In cases a) and b) the data receive specific protection, either
on the basis of exclusivity or compensation. In case c) the second-
entrant does not use the data; it is the authority who examines and
This compulsory licence approach is the one applicable, under certain cir-
cumstances, in accordance with the U.S. FIFRA. See Annex I.
See, e.g., article 4.8 (a)(ii) of the EC Directive 65/65/EEC.
In some countries, bio-availability studies are also required for the approval of
generic versions of existing products.
32 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
relies on the data in its possession. In case d), finally, there is no
“use” at all, since the authority does not use the testing and other
data(which it may not even possess); it merely relies on public infor-
mation and/or on the existence of a prior (domestic or foreign)
Neither in cases c) or d) is there a “commercial use” of the
data. A contrary interpretation holds that even indirect reliance on
data by a national authority constitutes a form of commercial use.
Under this interpretation, the competent authority must be pro-
scribed from “using” the data to support, clear or otherwise review
second entrant applications for marketing approval for a set amount
of time unless authorized by the “originator” (WHO, 2000, p. 39)
According to this interpretation, national authority reliance on
the data submitted by the originator in order to assess a subsequent
application constitutes “unfair commercial use”, even when neither
the authority nor the competitor actually “use” the data without the
originator’s authorization (for instance, when approval is given with-
out any re-examination of the data). In the U.S. complaint against
Australia, for instance, the USA argued that relying on the innova-
tor’s data allowed free-riding by generic drug companies on
“the innovator company´s investment in developing the test data
and thus puts the innovator company at a competitive disadvan-
tage...The U.S. claims that Article 39 para.(3) means that ge-
neric companies are not allowed to derive commercial benefit from
the innovator´s test data” (Priapantja, 2000, p.6).
Under this view, the fact that a competitor obtains a commercial
benefit or advantage constitutes an “unfair commercial use” of the
data, notwithstanding that actual use may not occur and that the
practice as such may not be “dishonest” or contrary to a country’s
prevailing values of morality or fairness in commercial activities.
This latter interpretation, however, clearly goes beyond what
the provision mandates. It does introduce an obligation not negoti-
ated during the Uruguay Round that, in practice, would limit legiti-
Proscribed Acts of Unfair Comemrcial Use 33
mate competition and thereby erect barriers to the access to medi-
B. National case law
Available national case law supports the view that granting marketing
approval to a second entrant, based on the second product’s similar-
ity to a previously approved first product, is not a proscribed “use”
under Article 39.3.
The nature and extent of data exclusivity rights were examined
in two important decisions by the U.S. Supreme Court (Ruckelshaus v.
Monsanto Co., 467 US 986, 104 S.Ct.2862, June 26, 1984) and the
Canadian Federal Court of Appeal (Bayer Inc. v. The General Attorney of
Canada, the Minister of Health, Apotex Inc. and Novopharm Ltd., May 19,
1999). The second decision, in particular, examined the extent to
which a national health authority can rely on the originator’s data, even
when an exclusivity period applies.
The Ruckelshaus v. Monsanto Co. case relates to the protection
of data submitted for the registration of an agrochemical product.
Though a subsequent applicant was obliged to compensate for the
use of Monsanto’s original data, Monsanto argued that such use
undermined its reasonable “investment backed expectations” and
was unconstitutional. A basic argument of the plaintiff was that the
possibility given to a competitor of using the data against payment of
a compensation nullified its “reasonable investment-backed expecta-
tion”. However, the Supreme Court described the extensive practice
of relying on data submitted by the first applicant in the United
States, and rejected Monsanto’s complaint (see Box 3).
34 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
Relying on data: the U.S. Supreme Court decision
in Ruckelshaus v. Monsanto Co.
The Supreme Court considered that Monsanto could not have had a
reasonable, investment-backed expectation that the Environmental
Protection Agency (EPA) would keep the data confidential beyond the
limits prescribed in the amended statute itself. Monsanto was on notice
of the manner in which EPA was authorized to use and disclose any
data turned over to it by an applicant for registration.
Excerpts from the Court’s decision:
“In addition, Monsanto was aware that information relating to
formulae of products could be revealed by EPA to “any Federal
agency consulted and [could] be revealed at a public hearing or in
findings of fact” issued by EPA “when necessary to carry out”
EPA´s duties under the Federal Insecticide, Fungicide and Rodenti-
cide Act (FIFRA) § 10(B)
. The statute also gave Monsanto notice
that much of the health, safety, and efficacy data provided by it
could be disclosed according to the data-consideration and data-
disclosure provisions in the statute. Monsanto chose to submit the
requisite data in order to receive a registration, it can hardly argue
that its reasonable investment-backed expectations are disturbed
when EPA acts to use or disclose the data in a manner that was
authorized by law at the time of the submission.”
“Because the market for Monsanto’s pesticide products is an inter-
national one, Monsanto could decide to forego registration in the
United States and sell a pesticide only in foreign markets. Presuma-
bly, it will do so in those situations where it deems the data to be
protected from disclosure more valuable than the right to sell in the
See a summary of FIFRA in Annex I.
Proscribed Acts of Unfair Comemrcial Use 35
Box 3 (continued)
“A fortiori, the Trade Secrets Act cannot be construed as any sort of
assurance against internal agency use of submitted data during con-
sideration of the application of a subsequent applicant for registra-
tion. Indeed, there is some evidence that the practice of using data
submitted by one company during consideration of the application
of a subsequent applicant was widespread and well known. Thus,
with respect to any data that Monsanto submitted to EPA prior to
the effective date of the 1972 amendments to FIFRA, we hold that
Monsanto could not have had a “reasonable investment-backed ex-
pectation” that EPA would maintain those data in strictest confi-
dence and would use them exclusively for the purpose of consider-
ing the Monsanto application in connection with which the data
“When Monsanto provided data to EPA during this period, it was
with the understanding, embodied in the FIFRA, that EPA was free
to use any of submitted data that were not trade secrets in consid-
ering the application of another, provided that EPA required the
subsequent applicant to pay “reasonable compensation” to the
original submitter. § 3(c)(1)(D), 86 Stat. 979. But the statute also
gave Monsanto explicit assurance that EPA was prohibited in con-
nection with the application of another, to use any data submitted
by an applicant if both the applicant and EPA determined the data
to constitute trade secrets.”
The U.S. Supreme Court in this case recognized that the authority
could use the data submitted by the originator to assess second-
entrant applications. According to the law applicable at the time of
the complaint, Monsanto was entitled to compensation, but not to
exclusive use of the data. The solution has probably not substantially
changed in the United States despite the adoption of the Second
Restatement of Unfair Competition Law (1997). In the absence of a
36 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
specific provision granting an exclusivity period as currently pro-
vided for medicines by U.S. law, relying on data to approve subse-
quent applications would not be considered an illegitimate misap-
propriation of trade secrets.
The General Court Appeal of Canada decided a second and
more significant case on issues related to data exclusivity. Despite the
fact that NAFTA provisions, as mentioned before, provide for a
minimum term of exclusivity, the Court found legitimate the ap-
proval of a subsequent application on the basis of a prior registra-
tion. The court argued that the health authority neither requested
undisclosed information a second time nor examined it; the authority
just checked whether the original and subsequent products were
indeed the same (see Box 4). The issue was decided under Canadian
law and NAFTA Article 1711 on “Trade Secrets”, which establishes
“5. If a Party requires, as a condition for approving the
marketing of pharmaceutical or agricultural chemical
products that utilize new chemical entities, the submis-
sion of undisclosed test or other data necessary to de-
termine whether the use of such data involves consider-
able effort, the Party shall protect against disclosure of
the data of persons making such submission, where the
origination of such data involve considerable efforts,
except where the disclosure is necessary to protect the
public or unless steps are taken to ensure that the data is
protected against unfair commercial use.
6. Each Party shall provide that for data subject to para-
graph 5 that are submitted to the Party after the date of
entry into force of this Agreement, no person other
than the person that submitted them may, without the
latter’s permission, rely on such data in support of an
application for the product approval during a reasonable
period of time after their submission. For this purpose,
Personal communication by Prof. J. Reichman (Duke University), October
Proscribed Acts of Unfair Comemrcial Use 37
a reasonable period shall normally mean not less than
five years from the date on which the Party granted ap-
proval to the person that produced the data for ap-
proval to market its product, taking account of the na-
ture of the data and the person’s efforts and expendi-
tures in producing them. Subject to this provision, there
shall be no limitation on any Party to implement abbre-
viated approval procedures for such products on the ba-
sis of bioequivalence and bioavailability studies.
7. Where a Party relies on a marketing approval granted
by another Party, the reasonable period of exclusive use
of the data submitted in connection with obtaining the
approval relied on shall begin with the date of the first
marketing approval relied on”.
Canadian Federal Court of Appeal: the Bayer case
The Federal Court of Appeal held, inter alia, the following:
“When a generic manufacturer files an Abbreviated New Drug
Submission (ANDS), the safety and effectiveness of the generic product
may be demonstrated by showing that the product is the pharmaceutical
and bioequivalent of the innovator´s product. If the generic manufac-
turer is able to do so solely by comparing its product with the innova-
tor´s product which is being publicly marketed, the Minister will not
have to examine or rely upon confidential information filed as part of
the innovator´s New Drug Submission (NDS). In such case, the mini-
mum five year market protection referred to in the regulation will not
38 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
Box 4 (continued)
“On the other hand, if in order to be satisfied of the safety and ef-
fectiveness of the generic product, the Minister examines and relies
upon information filed by the innovator in its NDS, the minimum five
years market protection for the innovator will apply. This is because the
safety and effectiveness of the generic product will only be established
by reference to confidential information provided to the Minister by the
innovator. It is only this use of that confidential information by the
Minister on behalf of the generic manufacturer that gives rise to the
minimum five years protection form competition for the innovator.
“The appellant says that whenever an ANDS is filed by a generic
manufacturer comparing the generic product with the innovator´s
product, the Minister must implicitly be examining and relying upon the
confidential information filed by the innovator in its NDS. We do not
read subsection C.08.004.1(1) in this way. To do so would be to inter-
pret it as invariably providing a minimum five years of market protec-
tion to an innovator when an ANDS is filed by a generic manufacturer.
Rather, the regulation contemplates that the Minister may or may not
examine and rely upon confidential information filed by the innovator.
The appellant’s argument reads out of the regulation the option given to
the Minister as to whether or not to examine and rely on the confiden-
tial information filed by the innovator.
“The NAFTA provisions are intended to protect trade secrets. If the
generic manufacturer exercises the option of having the Minister exam-
ine the confidential information filed by the innovator in support of its
application for a Notice of Compliance, it is, in effect, relying on that
information within the meaning of section 6 of Article 1711. It is appar-
ent that if confidential data is not relied upon, the trade secrets provi-
sions of the NAFTA are not applicable. Specifically, if a generic manu-
facturer is able to establish the safety and effectiveness of its product on
the basis of bioequivalence or bioavailability studies without the Minis-
ter having to examine and rely upon confidential data filed by the inno-
vator, there is no reason or justification for the minimum five years
protection from competition. This interpretation of subsection
C.08.004.01(1) is consonant with section 5 and 6 of Article 1711 of the
Proscribed Acts of Unfair Comemrcial Use 39
Box 4 (concluded)
“If a generic manufacturer compares its product to an innovator´s
product solely on the basis of public information, providing the inno-
vator with protection from competition for a minimum of five years is
tantamount to granting it the protection a patent would provide. Put
another way, even if the Minister did not examine and rely on the inno-
vator´s confidential information, the innovator would be entitled to the
minimum of five years protection form competition. The words of
subsection C.08.004.1(1) cannot be construed to yield such a result.”
The Court, in sum, concluded that, under Canadian law and
NAFTA, if the health authority actually uses the data submitted by
the originator on behalf of the generic manufacturer in order to as-
sess the latter’s application, the minimum five years protection from
the competition for the innovator applies. But if the authority does
not examine and rely on that confidential or trade secret information
on behalf of the generic manufacturer, there is no use of data and
the exclusivity provision is not applicable.
If despite the express provision of exclusivity, the mere reli-
ance on a prior registration without use of the data does not allow to
claim exclusivity, a fortiori the same conclusion should be reached
when the exclusivity is not specifically established, as in the case of
In sum, whatever the desire of some of the TRIPS negotiating
parties might have been, the expression “unfair commercial use”,
reasonable interpreted, does not sustain a reading that Article 39.3
requires the provision of exclusivity, or of a compensation. It has left
wide room for manoeuvre for Member countries to determine:
a) when such a use exists, and
b) the means of protection (see next section).
40 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
An “unfair commercial use” may be determined to exist, for in-
stance, in situations in which a competitor obtains through fraud,
breach of confidence or other “dishonest” practices, the results of
testing data and uses them to submit an application for marketing
approval in its own benefit. It would also apply in cases where the
government provides access to undisclosed testing data in order to
provide an advantage to a firm which did not produce them or share
This would represent a violation of the non-disclosure obligation as well as an
“unfair commercial use”.
A key issue for the application of Article 39.3 is to determine the
nature and extent of the obligation to protect “against unfair com-
mercial use”. As noted, the interpretation of this rule has created
The TRIPS Agreement mandates the protection of “undis-
closed information” in the framework of the discipline of “unfair
competition”. Article 39.1 of Agreement stipulates that
“in the course of ensuring effective protection against unfair compe-
tition as provided in Article 10bis of the Paris Convention
(1967) Members shall protect ...the data submitted to govern-
ments or governmental agencies in accordance with paragraph 3”.
Article 10bis of the Paris Convention requires protection against
“unfair competition”, defined as
“any act of competition contrary to honest commercial practices in
industrial or commercial matters”.
The discipline of unfair competition protects fairness in commercial
activities. As mentioned, there are no universal moral values or a
unique concept of what is “honest” in commercial behaviour. The
definition of what constitutes “fair” or “honest” practices varies
among countries. They may include competitor’s misrepresentation,
fraud threats, defamation, disparagement, enticement of employees,
betrayal of confidential information commercial bribery, among
others. In many but not all jurisdictions, the misappropriation of
trade secrets is regulated under unfair competition law, as is the case
with the TRIPS Agreement.
42 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
The present status of international competition law in the
Paris Convention is outlined in Box 5.
International protection against unfair competition under the
Pursuant to Art.10 bis (1), the contracting states are obliged to en-
sure citizens of other contracting states “effective protection against
In Art.10 bis (2) (general clause), unfair competition is defined as
any act of competition which is “contrary to honest practices in in-
dustrial or commercial matters”.
Three cases are named in Art.10 bis (3) which “in particular shall be
prohibited”, namely creating the risk of confusion, discrediting
competitors through false allegations and making misleading indi-
cations or allegations about one’s own goods. This list is not enu-
merative, so that other competitive acts can also be covered by the
Pursuant to Art.10 bis (1) of the Paris Convention, appropriate legal
remedies must be made available to the citizens of other contracting
states, in order to ensure the effective repression of acts contra-
vening Art.10 bis; these must also include the power of federations
and associations to take legal action (Art.10 ter (2)).
Source: Henning-Bodewig, 1999, p. 173.
Under the discipline of unfair competition, protection is not based
on the existence of “property” rights. Hence, the provision of pro-
tection under such discipline does not give rise to claims of property
rights, including in respect of trade secrets and data submitted for
marketing approval. There is only “possession” of this information.
Means of Protection Against Unfair Commercial Use 43
The TRIPS Agreement itself, in Article 39.3 refers to undisclosed
information “under the control” of a person, in clear contrast to the
concept used in the sections relating to other categories of intellec-
tual property rights.
A comparison between patents and trade se-
crets protection illustrates this important difference (see Box 6):
Patents vs. undisclosed information
A patent confers property rights.
A patent owner obtains the exclusive use of his/her rights. This
means he/she is the only one who can use the invention, commer-
cialize the product, etc. A patent owner can prevent any other per-
son from using that invention. Even if a third party has developed
the same product in an independent manner, without knowing or
relying on the technology of the patent owner, the former is not
allowed to use it, since the exclusive rights conferred are absolute.
In the case of undisclosed information, under most legal systems,
there is only “possession” of certain information.
The value of undisclosed information does not lie in the inventive
step or novelty -- even a list of clients can be protected, though ob-
viously this is not an invention -- but in the fact that the undis-
closed information has commercial value and in the fact that it is
Unlike patents, which in general last for 20 years from the filing
date, in the case of undisclosed information there is no defined time
limit. Undisclosed information is protected as long as it is kept un-
disclosed. The duration of the protection, therefore, depends on the
factual situation, not on any legal provision.
See, e.g., articles 16.1 and 28.1 which refer to the “owner” of a trademark
and of a patent, respectively.
44 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
Though during the TRIPS negotiations the United States sug-
gested the consideration of undisclosed information as “property” --
in accordance with the concepts developed in its own legal system --
that approach did not find support, particularly from European and
The TRIPS Agreement clearly does not treat undisclosed in-
formation as “property”.
The fact that TRIPS deems “undisclosed
information” to be a “category” of intellectual property does not
imply, as mentioned before, the existence of a property right.
Because the TRIPS agreement embraces an unfair competi-
tion approach to undisclosed information, a logic consequence of
the Agreement is that Article 39 does not obligate countries to con-
fer exclusive rights.
Exclusive rights are merely one “TRIPS-plus” option to deal
with issues covered by Article 39.3. There are heavy costs and ethical
concerns associated with such an approach, however. In the absence
of mechanisms that permit the use of the data, an exclusive rights
system leads to the need for competitors to duplicate tests (often
involving suffering of animals) in order to reach results that are al-
The Article 39.3 obligation may be implemented through less
onerous means, such as through the legal faculty to impede the use
of information acquired through dishonest practices (e.g. espionage,
On the different approaches in continental and common law with regard to
trade secrets, see Coleman, 1992; Font Segura, 1999.
According to Engelberg, U.S. law does not recognize “any property rights in
the data submitted to support an application for approval of a new drug… The
non-patent, market exclusivity provisions of the Drug Competition Act of 1984
were created as an arbitrary means of providing investment incentive for the
development of drug products that had little or no patent protection and not as
a purposeful determination to create a new form of intellectual property based
on undisclosed data”.
It is generally accepted, particularly under European law, that unfair competi-
tion is one of the disciplines of industrial property, and it is in this sense that
article 1.2 should be interpreted.
Means of Protection Against Unfair Commercial Use 45
breach of confidence), as background for an independent submis-
sion for marketing approval.
Implementing legislation may also require the subsequent user
to pay compensation, without providing for exclusive rights. The
U.S. FIFRA, for instance, recognizes the possibility of using the
originator’s test data for the approval of a subsequent application,
without the originator’s consent but with payment of compensation.
The law thus establishes a form of compulsory licensing for such
data. The United States required such a compulsory licence -- with-
out payment of compensation -- in approving Dow Chemical’s ac-
quisition of Rugby-Darby Group Companies. Approval of the
merger was contingent on the issuance of a licence for registration
data to all potential competitors (see Box 7).
Compulsory licensing in the U.S. involving test data
Acquisition of shares of Rugby-Darby Group Companies by Dow
The Federal Trade Commission required Dow to license to potential en-
trants, intangible dicyclomine assets, including all formulations, patents,
trade secrets, technology, know-how, specifications, designs, drawings,
processes, quality control data, research materials, technical information,
management information systems, software, the Drug Master File, and all
information relating to the United States Food and Drug Administration Approvals
that are not part of the acquired company’s physical facilities or other tan-
In sum, Article 39.3 -- interpreted according to the ordinary
meaning of the words used, in their context (notably Article 39.1)
46 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
and taking into account the object and purpose of the Agreement as
expressed in Articles 7 and 8 -- does not require the granting of ex-
The obligation that it imposes may be satisfied by
other means, not specified in the Agreement. As stated by
UNCTAD in relation to data covered by Article 39.3,
“authorities are not prevented…from using knowledge of such data,
for instance, to assess subsequent applications by third parties for the
registration of similar products” (UNCTAD, 1996, p.48).
See also Watal (2001) who concludes that “in the end in the TRIPS text there
is no clear obligation not to rely on the test data for the second or subsequent
applicants nor a fixed duration of market exclusivity, failing which the first
registrant is assured reasonable compensation. This is a clear contrast to the
corresponding provisions in NAFTA” (p. 199).
The pharmaceutical industry, the United States and the European
Union disagree with the contention that Article 39.3 does not require
the granting of exclusive rights. According to the industry, the only
way to effectively protect test data against unfair commercial use is
to provide an exclusivity period for the use of the data:
“To have a meaningful purpose this provision (Article 39.3) must
be interpreted to require the protection of data against use by the
competitors. Even if there is some concern about government use of
such data in a commercial manner, it is minuscule in comparison
to the problem of competitors´ use of the data. Consequently, in
light of the maxim of statutory construction that a provision will
be interpreted so that no part will be inoperative or superfluous,
void or insignificant, Article 39 para.(3) must be interpreted to
provide protection against the use of data by competitors for some
period of time” (Priapantja, 2000, p. 4).
The Office of the U.S. Trade Representative (USTR) has interpreted
Article 39.3 of the TRIPS Agreement to mean that
“the data will not be used to support, clear or otherwise review
other applications for marketing approval for a set amount of
time unless authorised by the original submitter of the data. Any
other definition of this term would be inconsistent with logic and
the negotiating history of the provision”.
Office of the General Counsel, U.S. Trade Representative, “The protection
of Undisclosed Test Data in Accordance with TRIPS Article 39.3”, unattributed
paper for submission in bilateral discussions with Australia (May 1995).
48 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
The United States maintained this position, for instance, in its
complaint (initiated in April 1996), under Special 301 Section of U.S.
Trade Act, against Australia.
Australia did not grant exclusivity,
and generic companies only had to demonstrate bio-equivalence in
order to obtain marketing approval of a similar product. In addition,
Australian authorities granted certificates of free sale which permit-
ted generic companies to export to other countries where marketing
approval was automatically granted on the basis of the Australian
The U.S. argued that the Australian regime violated Article
39.3. The U.S. pressure forced an amendment to the Australian law.
Under the Therapeutic Goods Legislation Amendment Act 1998
(No.34, 1998) test data in Australia now have five years of “exclusiv-
ity”. During this time, another company wishing to register a generic
copy of an originator’s product will be required to seek the agree-
ment of the originator company to use its data, or to develop its own
data package (Priapantja, 2000, p. 6).
The exclusivity approach was also incorporated, as a result of
U.S. demands, in the USA-Jordan Agreement on the Establishment
of a Free Trade Area (Washington D.C., 24 October, 2000), accord-
ing to which “in situations where there is reliance on evidence of
approval in another country, Jordan shall at a minimum protect such
information against unfair commercial use for the same period of
time the other country is protecting such information against unfair
commercial use” (Article 22, fn. 11)
This case was not brought to a panel resolution under the WTO’s Dispute
Settlement Understanding (DSU) rules. The U.S. instead threatened to impose
unilateral trade sanctions on Australia, even though TRIPS had already entered
into force in both countries. The U.S. also applied economic sanctions to Ar-
gentina in 1997, arguing Argentina maintained insufficient protection of confi-
dential information. More recently, the U.S. has started consultations under the
DSU on, inter alia, Argentina’s compliance with Article 39.3.
In addition, this Agreement establishes a TRIPS-plus standard in relation to
the concept of “new chemical entities”: it is understood that such concept
“shall also include protection for new uses for old chemical entities for a period
of three years” (article 22, fn. 10). The U.S. has also criticized the amendment
of the Thai “Safety Monitoring Programme” (SMP) established in 1993 as a
The Exclusivity Approach 49
The EU argues, similarly, that Article 39.3 established an ex-
clusivity obligation. All that is left to Member countries, according to
the EU, is the determination of the duration thereof.
“the only way to guarantee that no “unfair commercial use” within
the meaning of Article 39.3 shall be made is to provide that
regulatory authorities should not rely on these data for a reason-
able period of time, the determination of what is a reasonable pe-
riod of time being left to the discretion of the Members.
In theory, Article 39.3 appears to give Members the discretion to
provide for different means of data protection, although it is very
difficult to imagine other ways than non-reliance over a certain pe-
riod of time, except for a (temporary) refusal to grant any second
market approval to similar products (even if the second applicant
submits its own data), as is the case in at least one WTO Mem-
ber and maybe for an obligation to pay as a compensation for reli-
ance on proprietary data without having to obtain consent from the
first applicant. The question remains whether such payment would
indeed be sufficient to guarantee that any “unfair commercial use”
of test data takes place. For instance, it would be essential that
such payment reflects the investments made by the original appli-
cant -- which may not always be easy to establish.
In theory, any country maintaining an effective system to imple-
ment obligations under 39.3 even if different from non-reliance
result of USTR demands aimed at ensuring a two years minimum exclusivity
period for drugs patented abroad between 1986 and September 1991. In Janu-
ary 2001 the SMP was amended and generics companies were allowed to con-
duct bioequivalence studies at any time regardless of whether or not the original
products are under the SMP. If the original products are under the SMP, how-
ever, those generic products must also be under the SMP. According to the
Thai authorities, the SMP had led to unaffordable high prices for new drugs. In
the USTR view, this reform would mean the relinquishment of the benefits
affording data protection in accordance with the U.S.-Thai 1993 Bilateral
Agreement, and would create – at least until a “satisfactorily TRIPS-consistent
Trade Secrets Law is enacted and implemented in Thailand -- an unacceptable
gap in data protection coverage (Kwa, 2001, p. 49-50).
See also Lobato García-Miján, 2000.
50 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
over time, would not be in breach of its TRIPS obligations, but we
are not aware of many alternatives and it is clear that what the
TRIPS-negotiations had in mind was data exclusivity over a cer-
tain period of time. On the other hand, as it does not set any time
limit, Article 39.3 would not prevent a country from providing for
data exclusivity for an unlimited period of time” (EU, 2001, p.
The EU position suffers from several shortcomings, however. First,
had the negotiating parties agreed to embrace the concept of exclu-
sivity, they simply could have done so explicitly. The TRIPS Agree-
ment´s obligations in relation to copyrights, trademarks, industrial
designs, patents and integrated circuits (via incorporation of the
Washington Treaty), all explicitly provide for exclusivity.
The EU admits that there was substantial disagreement during
“It must be admitted that the following of Article 39.3 does not,
from a prima facie reading, appear to impose data exclusivity
during a certain period of time. This lack of clarity is the obvious
result of a difficult negotiation process where divergences of views
arose between developing and industrialized countries as to the ne-
cessity of EC/U.S. like type of data protection as well as among
industrialized countries on the length of the data exclusivity pe-
riod” (EU, 2001, p. 3).
The disagreement among the parties was, however, more substantial
than that argued by the EU, and there was no international estab-
lished practice on which to rely. The negotiating history of Article
39.3 reveals that the parties considered at length, but did not adopt,
text which clearly required exclusivity for test data.
Second, if the negotiating parties only left the Members the
freedom to determine the duration of the exclusivity period, on what
basis could a panel or the Appellate Body establish an “adequate”
duration? The basic rule of Article 3.2 of the WTO’s Dispute Settle-
The Exclusivity Approach 51
ment Understanding prohibits dispute settlement bodies from add-
ing to or subtracting from WTO agreement rights.
The EU itself
admits that there was disagreement among the developed countries
even about the duration of such period:
“Would this be 5 years (as in the case inter alia in US), 4 years
or 3 years? This remains an open question”.
As noted by Watal,
“It can be argued that if the intention had been to have such ex-
clusive marketing rights, this term, which is used in Article 70.9
of TRIPS, would have been used here too. Further, given the dif-
ferences in the TRIPS and NAFTA texts of this provision, it is
clear that the scope and purpose in TRIPS is intended to be more
limited as otherwise the text would have been as specific. No addi-
tional obligations, which are not present in the text, can be im-
ported through interpretation. Therefore, a reasonable interpreta-
tion would be that the obligation on the authorities would be to
keep the test data secret and to prohibit others from accessing this
test data for unfair commercial use, such as sale to rival firms”
(Watal, 2001, p. 204).
In sum, Article 39.3 clearly requires some form of protection for test
data. Its main purpose is not to prevent the commercial use of such
data by governments, but the use by competitors. The wording,
context and purpose of the article does not support an interpretation
that the required protection can be implemented only on the basis of
an exclusivity protection. This interpretation is confirmed by the his-
Article 3.2: “Recommendations and rulings by the Dispute Settlement Body
cannot add to or diminish the rights and obligations provided in the covered
52 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
tory of the negotiation of the TRIPS Agreement, as reviewed be-
The suggested interpretation has also been held by the Africa Group, Barba-
dos, Bolivia, Brazil, Dominican Republic, Ecuador, Honduras, India, Indonesia,
Jamaica, Pakistan, Paraguay, Philippines, Peru, Sri Lanka, Thailand and Vene-
zuela in a recent submission to the Council of TRIPS on “TRIPS and Public
Health”: “Protection of Test Data: Article 39.3 of the TRIPS Agreement leaves
considerable room for Member countries to implement the obligation to pro-
tect test data against unfair competition practices. The Agreement provides that
“undisclosed information” is regulated under the discipline of unfair competi-
tion, as contained in article 10 bis of the Paris Convention. With this provision,
the Agreement clearly avoids the treatment of undisclosed information as a
“property” and does not require granting “exclusive” rights to the owner of the
data” (para. 39) (IP/C/W/296, 19 June, 2001).
ISTORY OF THE
The history of the TRIPS Agreement negotiations also provides
important evidence for interpreting Article 39.3. Such history has
been accepted in recent WTO jursiprudence as an interpretative
source under Article 31 (2) of the Vienna Convention on the Law of
the treaties. It has been used to confirm the interpretation reached by
the application of the principles of Article 31 (1) of the Conven-
An early precedent of Article 39.3 can be found in the
“Statement of Views of the European, Japanese and United States
which also influenced the drafting of
other articles of the TRIPS Agreement. In their submission, the
business communities advocated for the protection of test data as
“1. Information required by a government to be disclosed by any
party shall not be used commercially or further disclosed without
the consent of the owner.
2 Information disclosed to a government as a condition for regis-
tration of a product shall be reserved for the exclusive use of the
registrant for a reasonable period from the day when government
approval based on the information was given. The reasonable pe-
riod shall be adequate to protect the commercial interests of the
This proposal clearly specified the obligation to establish a
data exclusivity period. The same approach was reflected in the U.S.
See, e.g. the panel report in United States-Section 110(5) of the Copyright U.S.
Copyright Act, WT/DS160/R, 15 June 2000, para. 6.50; United States-Section 211
Omnibus Appropriations Act of 1998, WT/DS176/R, 6 August, 2001, para.8.378.
See Annex III of Beier and Schricker, 1989.
54 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
“Contracting parties which require that trade secrets be submitted
to carry out governmental functions, shall not use the trade secrets
for the commercial or competitive benefit of the government or of
any person other than the right-holder except with the right
holder’s consent, on payment of the reasonable value of the use, or
if a reasonable period of exclusive use is given to the right-
It is interesting to note that this proposal referred to the “commer-
cial or competitive benefit” obtained by a third party, rather than to
“unfair commercial use” as is the Agreement’s text. The proposal is
based on the effects of the use (the creation of a benefit), while Arti-
cle 39.3 is based on an ethical qualification of the use as “unfair”.
The U.S. proposal turned on whether a commercial or competitive
benefit (independently of the qualification of the use that generated
it) was obtained; under Article 39.3, the key issue is whether there is
unfairness in the use (as provided by Article 10bis of the Paris Con-
vention) and not whether a third party obtains a benefit.
The negotiating parties considered requiring test data exclu-
sivity, but rejected this approach. Bracketed text under consideration
at the Brussels Ministerial Meeting (December 1990) would have
required not less than five years of data exclusivity. The draft read as
“Parties, when requiring, as a condition of approving the market-
ing of new pharmaceutical products or of a new agricultural chemi-
cal product, the submission of undisclosed test or other data, the
originator of which involves a considerable effort, shall [protect such
data against unfair commercial use.
Unless the person submitting the information agrees, the data may
not be relied upon for the approval of competing products for a rea-
sonable time, generally no less than five years, commensurate with
the efforts involved in the origination of the data, their nature, and
the expenditure involved in their preparation. In addition, Parties
See MTN.GNG/NG11/W/70, reproduced in Correa and Yusuf, 1998..
The History of the TRIPS Negotiations 55
shall] protect such data against disclosure, except where necessary
to protect the public]”.
Notably, this text also explicitly included a prohibition on reliance on the
data submitted by the originator. But this concept disappeared from the
The negotiating history of Article 39.3, in sum, does not
support the thesis that it was intended to provide exclusive rights.
On the contrary, it shows that such concept was rejected. It is also
suggestive in this sense that the most active proponents of such ap-
proach are currently proposing to review the TRIPS Agreement in
order to include an exclusivity period
See Gervais, 1998, pp. 182-183.
The EU has pointed out that “according to one commentator, the U.S. nego-
tiations finally decided to drop the more explicit language of above drafts be-
cause they did not view such wording as essential because, in any event, “the
accepted definition at the time of protection against unfair commercial use
included non-reliance for a fixed period of time for new chemical entities” (EU,
2001, p. 4).
In its position paper on the WTO Millennium Round, the International Fed-
eration of Pharmaceutical Manufacturers’ Association (IFPMA) has called, inter
alia, for the adoption of a 10-year data exclusivity period (see “What is at stake
in Seattle”, in www.pharma.org). See also IIPI (2000), where it is noted that
Article 39.3 “requires WTO Members to protect health registration data from
disclosure or unfair commercial use, but its exact boundaries of “unfair com-
mercial use” are not entirely clear (p. 26).
The use by health authorities and competitors of test data which
must be submitted to obtain marketing approval of pharmaceutical
(and agrochemical) products has been subject to specific regulations
in several jurisdictions. Some developed countries, notably the U.S.
and EU, have established data protection regulations based on the
exclusive use of such data by the originator company. In other
countries, however, off-patent generic products can be approved by
relying on the data available to health authorities or by reference to a
prior registration either domestic or in third countries. In all cases,
the physical and chemical similarity (or essential similarity) with the
registered product must be demonstrated.
The TRIPS Agreement has obliged WTO Member countries
to treat test data as a component of “intellectual property”. How-
ever, the rationale for test data protection is the investment made in
data production, rather than their creative or inventive content.
Article 39.3 of the TRIPS Agreement requires data protection
against disclosure and “unfair commercial use”. Article 39.3 develops
Article 10bis of the Paris Convention; that is, it requires the protec-
tion of data against dishonest commercial practices.
The non-disclosure obligation admits exceptions where neces-
sary to protect the public, and in other cases where measures are
adopted to ensure that the information is not used in an unfair
commercial manner. Considerable room has been left to Members
for defining the grounds for the application of these exceptions.
In implementing the obligation to protect against unfair
commercial use, the Member States can determine, in accordance
with their own values and practices, the standards demarcating dis-
honest commercial practices. Further, the TRIPS Agreement has
58 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
deferred to Members the determination of the legal means to be
used in order to make such protection effective. Hence, Members
may opt for means of protection against unfair commercial use
which allow for the approval of “similar” products without the use
of the data or relying on them. Members may also opt, but are not
obliged to, grant “TRIPS-plus” protection on the basis of data exclu-
sivity, as some countries currently do.
In making such choices, policymakers will have to weigh the
protection of the interests of originator companies against the im-
portance of creating a competitive environment in order to increase
access to medicines that are outside patent protection. From a public
health perspective, the introduction of TRIPS-plus standards does
not seem the best approach for developing countries.
In sum, developing countries should carefully consider the
scope of regulations on approval of pharmaceutical products. Such
regulations should be enacted with a pro-competitive intent, in a
manner that maximizes legitimate competition and access to drugs,
while respecting the legitimate interests of the originators of data in
accordance with the standards of protection established by the
1. Under the “exclusive use” provision, some data are temporarily
protected from use by a data submitter’s competitors. FIFRA §
3(c) (1) (D) (i). Registrants are granted a ten-year period of ex-
clusive use for data on new active ingredients first registered af-
ter September 30, 1978. FIFRA § 3(c) (1) (D) (i). During that pe-
riod, no other applicant may use the data to support an applica-
tion for registration. To be eligible for exclusive use, data must
pertain to a “new” complete data package. The second regis-
trant, however, is not necessarily required to duplicate exactly
the original submitter’s data package.
2. Under the “data compensation” provision, most data can be
used by any company willing to pay compensation to the data
submitter. FIFRA § 3(c) (1) (D) (ii). Compensation is required
whenever data submitted after December 31, 1969 is considered
by EPA in support of another company’s registration. § 3(c) (1)
(D) (ii). The duty to pay compensation, however, ends fifteen
years after the data are submitted, after which no further pay-
ment is required for use of the data. § 3(c) (1) (D) (iii). In the
case of an active ingredient subject to exclusive use protection,
the data are subject to compensation for five years after the ten-
year period of exclusivity expires.
3. Under the “joint data development” provision, two or more
registrants can agree to develop jointly, or to share the cost of,
new data needed for re-registration or to respond to data call-ins.
FIFRA § 3(c) (2) (B) (ii). Registrants may agree to develop jointly
new data needed by EPA for re-registration. FIFRA § 3(e) (2)
(B) (ii). Applicants for re-registration which are not developing
re-registration data either alone or jointly must offer to share in
60 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
the cost of the data being developed by other registrants. See
FIFRA §§ 4(d) (3) (B) (ii), 4(e) (1) (H) (ii). FIFRA does not es-
tablish a formula or standard for determining the amount of
compensation under § 3(c) (1) (D) or the manner in which costs
should be shared under § 3(c) (1) (D). Instead, the statute leaves
it to the parties themselves to work out compensation or cost
sharing arrangements, or to agree upon a dispute resolution pro-
cedure. Any party, however, has the right to initiate binding ar-
bitration proceedings in order to resolve a data compensation or
cost sharing dispute. FIFRA § 3(d) (1) (D) (ii), 3(c) (2) (B) (iii).
4. Under FIFRA´s binding arbitration provisions, data compensa-
tion or cost sharing disputes can be resolved by a neutral arbi-
trator. FIFRA § 3(c) (1) (D) (ii), 3(c) (2) (B) (iii). In the Thomas v.
Union Carbide Agricultural Products Co., 473 U.S. 568 (1985). case,
several pesticide manufacturers challenged the FIFRA arbitra-
tion system on the grounds that it delegated too much power to
an arbitrator to determine compensation, without review of the
soundness of arbitration awards by the federal courts. The Su-
preme Court concluded that this delegation of adjudicatory
power to arbitrators, rather than the courts, does not violate the
“separation of powers” required by the Constitution. 473 U.S. at
592-93. A federal district court subsequently held that the lack of
a standard in FIFRA for measuring compensation is not uncon-
stitutional. PPG Industries v. Stauffer Chemical Co., 637 F. Supp. 85
(D.D.C.1986), appeal dismissed, No. 86-5502 (D.C. Cir.Nov.4,
" Beier, Friedrich-Karl and Schricker, Gerhard (1989), GATT or
WIPO –New ways in the international protection of intellectual prop-
erty, VCH, Winheim.
" Coleman, Allison, (1992), The legal protection of trade secrets, Sweet
& Maxwell, London.
" Cook, Trevor, (2000). Special Report: The protection of regulatory
data in the pharmaceutical and other sectors, Sweet & Maxwell, Lon-
" Correa, Carlos, (2000) “Implementing national public health
policies in the framework of the WTO Agreements” (2000),
Journal of World Trade, vol. 34, No. 5, September.
" Correa, Carlos and Yusuf, Abdulqawi (1998), Intellectual property
and international trade. The TRIPS Agreement, Kluwer Law Inter-
" Dodds-Smith, Ian, (2000), “Data protection and abridged
applications for marketing authorisation in the pharmaceutical
industry”, Goldberg, Richard and Lonbay, Julian, (Eds.),
Pharmaceutical Medicine. Biotechnology, and European Law, Cam-
bridge University Press, Cambridge.
" Dukes, M.N. Graham, (1996), “Drug regulation and the tra-
dition of secrecy”, International Journal of Risk & Safety in Medi-
" Dukes, M.N. Graham, (1977), “The moments of truth”, Side
Effects of Drugs Annual, 1, Wxcerpta Medica, Amsterdam and
62 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
" Engelberg, Alfred (2001) “Data Exclusivity Under Article 39.3
of TRIPS: Does Current US Law Comply?”, Working Paper,
" European Union (EU), (2001), Questions on TRIPs and data
exclusivity. An EU contribution, Brussels.
" Font Segura, Albert, (1999), La protección internacional del secreto
empresarial, Colección Estudios Internacionales, Ed.
EUROLEX, No.XXXI, Madrid.
" Gambardella, Alfonso, (1995), Science and innovation. The Us
pharmaceutical industry during the 1980s, Cambridge University
" Gervais, Daniel (1998), The TRIPS Agreement. Drafting history
and analysis, Sweet & Maxwell, London.
" Henning-Bodewig, Frauke, (1999), “International protection
against unfair competition –Art. 10bis Paris Convention,
TRIPS and WIPO Model Provisions”, IIC, vol.30, No.2.
" International Intellectual Property Institute (IIPI) (2000), Pat-
ent protection and access to HIV/AIDS pharmaceuticals in Sub-
Saharan Africa, Washington D.C.
" Jones, Nigel and Nittenberg, Ronald, (1998/|1999), “Essen-
tially Similar´ Despite Being Different: The Squibb Case”,
" Kamperman Sanders, Anselm (1997), Unfair Competition
Law,1997, Clarendon Press, Oxford.
" Kwa, Aileen (2001), “Dying for free trade”, Resurgence, No.
" Ladas, Stephen, (1975), Patents, Trademarks, and Related Rights.
National and International Protection, vol. III, Cambridge.
" Lobato García-Miján, Manuel, (2000), Estudio sobre el funda-
mento jurídico de la protección de los secretos empresariales regulados en el
artículo 266 de la Decisión Andina 486, FIFARMA.
" Ollila, E. and Hamminki, E., (1996), “Secrecy in drug regula-
tion”, International Journal of Risk & Safety in Medicine, No. 9.
" Priapantja, Priapantja, (2000), Trade Secret: How does this apply to
drug registration data?, paper presented at “ASEAN Workshop
on the TRIPS Agreement and its Impact on Pharmaceuticals”,
Department of Health and World Health Organization, Ja-
karta, 24 May 2000.
" Raggett, Tom, (1996), GATT and patent reform. The global
strengthening of patent protection and the implications for the pharma-
ceutical industry, Financial Times Management Reports, Lon-
" Reichman, J. And Uhlir, Paul, (1999), “Database protection at
the crossroads: recent developments and their impact on science
and technology”, Berkeley Technology Law Journal, vol. 14, No.2.
" Trebilcock, Michael and Howse, Robert, (1999), The Regulation
of International Trade, Routledge, London and New York.
" UNCTAD, (1996), The TRIPS Agreement and developing Countries,
UNCTAD/ITE/1, New York and Geneva.
" Watal, Jayashree, (2001), Intellectual Property Rights in the WTO
and Developing Countries, Kluwer Law International, The
" WHO, (2000), The TRIPS Agreement and Pharmaceuticals. Report
of an ASEAN Workshop on the TRIPS Agreement and its Impact on
Pharmaceuticals, Jakarta, 2-4 May 2000.
64 Protection of Data for the Registration of Pharmaceuticals: Implementing Standards …
" WIPO, (1996), Model Provisions on Protection Against Unfair Com-