FINAL REPORT FINAL REPORT

FINAL REPORT STEM CELLS Analysis of an Emerging Domain of Scientific and Technological Endeavour FINAL REPORT by Wolfgang GLÄNZEL Arnold VERBEEK Mar...
Author: Sybil Lamb
8 downloads 1 Views 1MB Size
FINAL REPORT

STEM CELLS Analysis of an Emerging Domain of Scientific and Technological Endeavour

FINAL REPORT by Wolfgang GLÄNZEL Arnold VERBEEK Mariëtte DU PLESSIS Bart VAN LOOY Tom MAGERMAN Bart THIJS Balázs SCHLEMMER Koenraad DEBACKERE Reinhilde VEUGELERS IN COLLABORATION WITH PROF. BALÁZS SARKADI (HEAD OF THE DEPARTMENT OF CELL METABOLISM AT THE NATIONAL INSTITUTE OF HAEMATOLOGY AND IMMUNOLOGY (OHII), BUDAPEST, HUNGARY)

Steunpunt O&O Statistieken K.U. Leuven Dekenstraat 2 B-3000 Leuven http://www.steunpuntoos.be/ DECEMBER 2004

FINAL REPORT

TABLE OF CONTENTS PREFACE____________________________________________________________3 PART I STEM-CELL RESEARCH – A BIBLIOMETRIC APPROACH _______________________6 I. The coverage of stem-cell research in the bibliographic database Science citation Index Expanded _______________________________________________________________ 6

The field of stem-cell research in the mirror of bibliometric analyses: A concise overview _____________6 Data sources and data processing ___________________________________________________6 The delineation of the research field stem-cell research (SCR) _______________________________7 Questions of national stem cell policies in embryonic-stem-cell research ________________________8

II. The evolution of publication output and citation impact in stem-cell research in the period 1992-2001 _______________________________________________________________ 8

Evolution of publication output and citation impact of the field ________________________________8 Publication output and citation impact of Flanders and the 35 most active countries (1994-2003) ______13

III. Flemish stem-cell research in the mirror of bibliometric indicators (1994-2003) _______ 17 IV. International co-authorship patterns in Stem-Cell Research ______________________ 19

The global collaboration network of research in SCR _____________________________________19 Mapping mutual co-authorship links _________________________________________________23 International co-publication patterns of Flemish research in Stem-Cell Reasecrh__________________26

REFERENCES _______________________________________________________30 APPENDIX - Definition of country abbreviations______________________________32 PART II STEM-CELL TECHNOLOGY – A TECHNOMETRIC APPROACH___________________33 I. Introduction ___________________________________________________________ 33 II. Search Strategy ________________________________________________________ 33

The search procedure and it’s effectiveness ___________________________________________33 A closer look at the evolution in stem cell patenting ______________________________________35

III. Detailed analysis of stem cell patenting _____________________________________ 39

Development of stem cells worldwide ________________________________________________39 Technological specialisation_______________________________________________________41 International Collaboration ________________________________________________________44 Organisational Analysis __________________________________________________________48

IV. Conclusions __________________________________________________________ 52

REFERENCES _______________________________________________________53 APPENDIX - Definition of country abbreviations _____________________________54 PART III OVERALL CONCLUSIONS ______________________________________________55

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

PREFACE BY PROFESSOR BALÁZS SARKADI Head of the Department of Cell Metabolism at the National Institute of Haematology and Immunology (OHII), Budapest, Hungary International Research Scholar of the Howard Hughes Medical Institute (HHMI), Chevy Chase, MD, USA

Background on Stem Cell Research Embryonic or somatic stem cells are seen as promising therapeutic tools for the treatment of number of several severe human diseases such as leukemia, diabetes, Parkinson disease, multiple sclerosis and other degenerative diseases. Embryonic stem (ES) cells have been isolated from the mouse more than twenty years ago, and it is only during the last five years that human ES cells have successfully been isolated and propagated in a very limited number of laboratories mostly in United States, Australia, Israel and Sweden. Somatic stem cells became also highly promising reagents in the past few years when a number of data suggest their potential for efficient differentiation into various cell types. In the hematopoietic system, somatic stem cells (hematopoietic stem cells) have been used for transplantation therapy for a long time. There is also number of studies which indicate that cancer takes place in somatic stem cells. This is particularly true in tissues with high level turnover such as skin, intestine, blood and human breast gland. Striking parallels can be found between stem cells and cancer cells and similar mechanisms may regulate self-renewal in those two cell types. Because of the expected demand for stem cells for human medical applications, there is a real need for supporting research aimed at developing human stem cell lines and their applications. This aim requires that we rapidly increase our knowledge of the basic features and properties of stem cells either from embryonic or somatic origin, human as well as from animal models. In a recent (2004) European ESF program called “EUROCORES on Development of a stem cell tool box” the following strategic purposes have been declared: o o o o o o

Fill the immediate need for tools, biological materials and protocols in stem cell technology, Create a critical mass of expertise in Europe in the field, Harmonize definitions, tools, reagents and protocols in stem cell biology, Promote and support training and access to European laboratory facilities, Set up the bases for comparative analyses between stem cells of different origins, i.e. embryonic vs. foetal and somatic, and between somatic stem cells from various tissues, Prepare application of the stem cell technology to therapeutic developments in human

The ultimate goal of research on stem cells will be to develop new therapeutic approaches in human, while this ESF programme aims at developing fundamental knowledge on stem cells and conceptual and technological expertise in the field. For the moment most of the experimental investigations on stem cells have to be performed on animal models of different origins. In the future, the diversity of the approaches should favour emergence of new concepts and technologies.

3

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Scientific priorities Propagation and expansion of stem cell cultures Ability to cultivate and manipulate stem cells ex vivo is a critical step towards elucidating their biological properties and developing their biotechnological and therapeutic potential. At present only a handful of stem cell types can be maintained in vitro, and of these only Embryonic Stem (ES) cells, certain neural stem cells, mesenchymal stem cells and the recently described Multipotent Adult Progenitor Cells (MAPCs) undergo significant multiplication. Most commonly in stem cell cultures proliferation is associated with differentiation. There is a pressing need therefore to •

acquire understanding of self-renewal mechanisms



develop procedures for expanding stem cells in the laboratory



optimise and standardise culture protocols

These goals are critical for future applications of human stem cells but are also important for fundamental investigations in mice and other model organisms. The research programme should therefore include a range of animal stem cells and encourage comparative approaches. Investigations are also necessary into the genetic, epigenetic and phenotypic fidelity of stem cells during long-term culture.

Optimization of stem cells cultures Non-human materials in cultures bear a risk for inter-species infections, and optimally no such materials should be used. To keep human ES cells undifferentiated, feeder cells have been necessary. When the first cell lines were established, foetal mouse feeder cells have been used and, only recently, establishing and culturing these cell lines on human feeder cells have been successful. Optimally no feeder cells should be used, but much research is still needed to identify the factors which are necessary for promoting the growth of stem cells as undifferentiated cells. If cells are used in human cell transplantation, Good Manufacturing Practice (GMP) quality is required. Optimally, the culture techniques have to allow large-scale production of cells which maintain their stem cell characteristics during the process.

Phenotypic and genotypic characterisation A given candidate stem cell population is as good as the model in which it was characterised. Hematopoietic stem cells have been formerly characterised and purified to homogeneity because reliable, sensitive and quantitative assay systems were available, in vitro and especially in vivo. Conversely, a major limitation in many current projects related to stem cell research is the lack of appropriate assays. Hence special consideration should be given to projects in which novel stem cell assays will be developed and validated:



In vitro assays

These include the classical two-dimensional culture of dissociated cells but also the development of three-dimensional cultures of either intact or reconstructed tissues. Conditions required to maintain “stemness” or induce differentiation may include cell-cell and cell-substrate interactions, substrate nature, oxygen pressure, medium composition, presence of growth/differentiation factors, optimal cell density.

4

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken



In vivo assays

Isogenic assays should be conducted mainly in small laboratory animal models. Host conditioning should be determined depending on the stem cell type analysed: irradiation, chemo-, immuno- or surgical ablation; cell lineage ablation or damage in transgenic mice. Allogenic assays should be conducted mainly in small laboratory animal models. Activities should include at least: a/ definition of the immuno-phenotype b/ characterisation of the host immune response Xenogenic assays should be developed primarily for human stem cell characterisation taking into account the ethical guidelines concerning human stem cells. Congenitally immuno-deficient mice or rats, early blastocysts or pre-immune foetuses in rodents or larger animals can be used as hosts. Human stem cells to be assayed in tolerant animal hosts can be administered directly to the host target tissue, following appropriate conditioning (see above), or engrafted into human target tissue previously implanted in the animal host. An important point to be taken into account is the marking system used in these experiments to assess chimerism, be it intrinsic (HLA or other markers) or secondarily brought to the cell under study (marker transgene).

Genetic modification of stem cells Genetic modification of stem cells could pave the way for their successful medical application. In current gene therapy efforts, the most promising methods involve ex vivo modification of stem cells, and then auto-transplantation of the modified cells into the patient. Based on this strategy, non-functional or malfunctioning genes in the stem cells of a patient may be replaced by genes properly regulated and producing normally functioning proteins. Thereafter the genetically modified stem cells could be inserted into the patient’s body. A successful insertion and propagation of genetically modified stem cells should provide the basis of a new era of curing e.g. heritable, autoimmune or malignant diseases. Before reaching this stage, there are numerous problems to be solved by extensive basic research. The key issues in stem-cell based gene therapy research should include the establishment of efficient and safe methodologies in order to: o

genetically modify stem cells, including the development of efficient and safe gene insertion systems,

o

avoid unwanted stem cell transformation or differentiation during this procedure,

o

control differentiation of stem cells towards a desired direction,

o

allow an efficient re-insertion and long-term survival of genetically modified stem cells in the body and avoid immunological rejection,

o

avoid any possible malignant transformation of the modified and re-inserted stem cells,

o

provide a selective advantage of the genetically modified stem cells after their re-insertion,

o

use stem cells as delivery vehicles and development of appropriate homing assays

5

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

PART I STEM-CELL RESEARCH – A BIBLIOMETRIC APPROACH I. The coverage of stem-cell research in the bibliographic database Science citation Index Expanded

The field of stem-cell research in the mirror of bibliometric analyses: A concise overview During the last years, very few studies on the field stem-cell research (SCR) and its evolution in the mirror of scientific papers and granted patents have been published. A compilation of bibliometric indicators on Embryonic Stem Cells has recently been published by the Institute for Scientific Information (Thomson – ISI, Philadelphia, PA, USA). This compilation is part of ISI’s Essential Science Indicators TM (2004). The set of micro, meso and macro indicators is based on the publication period 1991–2001. Recently two studies have been published, a small study Ho et al., 2003 in which a derivate of the ISI Impact Factor was used to study the scientific activity in the Asian Four Dragons (Hong Kong, Singapore, South Korea and Taiwan) in SCR in the period 1981–2001. This study was based on the Science Citation Index Expanded TM (SCIE) of Thomson – ISI. The so far largest study on stem-cell research and technology has been published by Campbell et al. in 2004. The bibliometric part of this study was based on publications retrieved from Medline database using a set of keywords defining the SCR field. The underlying time period was 1991–2002. The technometric part was based on patents retrieved from the United States Patents and Trademark Office (USPTO) database over the period 1991-2003. Three scientometric and two technometric indicators have been used to outline the scientific and technological output in stem cells at the international level for the G7 countries and at the national level for Canada. The following bibliometric analysis will be based on publication and citation indicators. The Science Citation Index Expanded TM (SCIE) is, therefore, an appropriate source to depict publication activity and citation impact of scientific research in the field of stem-cell research in an adequate manner.

Data sources and data processing All results are based on raw bibliographic data extracted from the 12-year annual cumulations (19922003) of the Web of Science edition of the Science Citation Index ExpandedTM (SCIE) of the Institute for Scientific Information (ISI, Philadelphia, PA, USA). Only “citable” publications, that is, papers recorded as article, letter, note or review were taken into consideration. The extracted data have undergone a very detailed cleaning and then processed to bibliometric indicators according to the fundamental principles underlying the construction of basic indicators and the methodology of data processing published in earlier studies by the author (Glänzel, 2001). The papers were assigned to countries and Flemish institutions based on the corporate address given in the by-line of the publication. All countries and institutions indicated in the address field were thus taken into account. Co-authorship was counted for the corresponding country pairs if the names of the concerning countries occurred simultaneously. It has to be stressed here that publication counts and citation frequencies cannot be summed up over co-authorship links to the total. For the meso study, Flemish addresses were cleaned-up, unified and according de-duplicated at the level of main institution.

6

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Citation counts have been determined on the basis of an item-by-item procedure using special identification keys. Citations were counted in a three-year period: in the year of publication and the two subsequent years, that is, for instance, if papers published in 1998 were considered, all citations received by them in the period 1998-2000 have been counted. The choice of the citation window is in keeping with recent methodological considerations and practical experience (see, for instance, Glänzel et al., 1999). As a consequence of the use of 3-year citation windows, citations could be counted for papers published in the years 1992 (citations received in 1992-1994) up to 2001 (citations received in 2001-2003).

The delineation of the research field stem-cell research (SCR) Earlier bibliometric studies were based on the search key ‘stem cell*’ (as for instance in ISI’s Essential Science Indicators on embryonic stem cells). Here we have to stress that literature indexed by the SCIE has undergone a careful reviewing process as a result of which ISI has added the term ‘stem-cell’ as Keyword Plus whenever ISI considered a publication relevant in the context of SCR. In the present study we apply a strategy that is based on a combination of keywords, core journals and bibliometric analysis of reference literature. The strategy consists of two parts which, in turn, have three components each. The first part comprises three unconditional criteria, particularly, UC1: UC2: UC3:

Journal = STEM CELLS Address word = STEM CELL* Keywords = (STEM CELL* OR STEM (ES) CELL* OR PROGENITOR* CELL* OR HEMATOPOI* CELL*)

Explanation: All papers meeting at least one of the criteria UC1–UC3 are considered relevant. UC1: UC2: UC3:

Papers published in the core journal STEM CELLS; Papers published by authors with affiliation in the name of which the stem cell* occurs (e.g., ‘Inst Stem Cell Res’ at the University Edinburgh); Papers with title, abstract or keywords/keyword plus in which the above terms occur.

The second part comprises conditional criteria, particularly, CC1: CC2: CC3:

Journal = JOURNAL OF HEMATOTHERAPY & STEM CELL RESEARCH Keywords = (BONE-MARROW OR UMBILICAL-CORD-BLOOD OR UCB OR HUCB OR CYTOPOI* OR MEGAKARYOPOI* OR ERYTHROPOI* OR MYELOPOI* OR THROMBOPOI* OR STROMAL CELL* OR PRECURSOR CELL*) Cited source = UC1 OR UC2 OR UC3

Explanation: All papers meeting at least one of the criteria CC1–CC3 are considered potentially relevant. CC1: Papers published in the journal JOURNAL OF HEMATOTHERAPY & STEM CELL RESEARCH; CC2: Papers with title, abstract or keywords/keyword plus in which the above terms occur. CC3: Papers citing 3–5 other papers classified as unconditionally relevant making up at least 40% of all SCIE references, or 6–10 UC papers making up at least 30% of all SCIE references, or citing more than 10 UC papers. Papers meeting one of the first two conditional criteria and citing relevant papers are also considered relevant. The above restriction to the percentage shares of 40% and 30% proved necessary to avoid noise. The resulting final formula for identifying relevant papers thus reads

7

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

UC1 OR UC2 OR UC3 OR ((CC1 OR CC2) AND CC3) The retrieval of papers meeting CC3 is an iterative process. In order to identify CC3 papers published in the year x all UC1–UC3 papers published in the period [x–3, x] have to be retrieved first. A larger citation window is not practically necessary because older references tend to become less specific (cf. Glänzel and Czerwon, 1996). The results of the search strategy have been validated by Professor Balázs Sarkadi who is Head of the Department of Cell Metabolism at the National Institute of Haematology and Immunology (Budapest, Hungary) and International Research Scholar at the Howard Hughes Medical Institute (Chevy Chase, MD, USA). A set of 46964 publications was retrieved on the basis of the combination of conditional and unconditional criteria. 1606 papers were added on the basis of the conditional criteria. This corresponds to a share of 3.4% in the total. Only 2.6% of the “conditional” papers and less then 0.1% of all retrieved publications was considered being at the “low end” of medium relevance. Practically all papers could be considered being of high or at least medium relevance. These results also substantiate that the abovementioned reviewing process at ISI guarantees a very good coverage of SCR.

Questions of national stem cell policies in embryonic-stem-cell research Preparing a descriptive or even evaluative bibliometric study on stem-cell research is quite impossible without addressing the question of national restrictions concerning research that uses embryonic stem cells. Although the share of publication explicitly reporting on research results basing on human embryonic stem (ES) cells amounts according to the above-mentioned indicator compilation (ISI Essential Science Indicators TM , 2004) only roughly 10% of all stem-cell related papers, the question of national restrictions might heavily influence both scientific migration and publication output in the domain. As a consequence of the decision made in 2001 by the US government to deny funding for research on embryonic stem cells, a brain drain of US researchers to Europe was predicted. However, most of the developed European countries have their own restrictions that are in part more severe than those in the USA. Germany, France, Austria, Denmark, Norway, Ireland, Slovakia and Lithuania have the most severe restrictions, and do (cf., Jonietz, 2003) permit embryonic-stem-cell research only in exceptional cases. In Germany, for instance, import and use of embryonic stem cells are prohibited since 2002 – with some exceptions, namely, if the research uses ES cells acquired before 2002. Other European countries have less severe restrictions; the same applies to Japan (Cheng, 2004). Great Britain is the only country without restrictions; moreover UK was the first country that legalised human cloning, in particular for the obtainment of stem cells as long as the embryos used are destroyed after 14 days. These examples might just serve to outline national policies and the diversity of legal conditions of use of embryonic stem cells in biomedical research.

II. The evolution of publication output and citation impact in stem-cell research in the period 1992-2001 Evolution of publication output and citation impact of the field Figure 1 presents the number of papers retrieved. The chart at the top shows the annual growth, that at the bottom the cumulated number of SCR publications. A certain increase of publications can be observed (cf., Figure 1, top); the share of SCR papers in the SCIE database has grown by 0.4 percent points, namely from 0.42% in 1994 to 0.82% in 2002. Nevertheless, this corresponds to an increase by 96.5%, that is, the share of SCR paper has almost doubled during the last ten years. The growth of cumulated number of publications (cf. Figure 1, bottom) lies in between the linear (as for instance observed in biotechnology, Glänzel et al. 2003a) and the exponential model (as for instance observed in nanoscience and technology, Glänzel et al. 2003b).

8

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

8,000

Number of publications

7,000 6,000 5,000 4,000 3,000 2,000 1,000 0 1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

1999

2000

2001

2002

2003

50,000 45,000

Number of publications

40,000 35,000 30,000 25,000 20,000 15,000 10,000 5,000 0 1994

1995

1996

1997

1998

Figure 1 Evolution of publication output in the period 1994-2003 (world total) Top: Annual publication output; Bottom: Cumulated number of publications

The powerful growth of literature in SCR is contrasted by a less pronounced increase of citations. Although the deviation looks at the first sight quite dramatic, relative increments of citations roughly follow those of papers except for a period of almost stagnation in 1995-1997. The patterns are shown in Figure 2. Citations have, as already mentioned in the outset, determined on a basis of three-year citation windows. Before we have a closer look at citation patterns, we will introduce the indicators used for the analysis.

9

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

200% Publications

190%

Citations

180% 170% 160% 150% 140% 130% 120% 110% 100% 1994

1995

1996

1997

1998

1999

2000

2001

Figure 2 Annual change of citations compared with that of publications in biotechnology for 1994-2001 (1992 = 100%) The following citation-based indicators are used in this study. i)

Mean Observed Citation Rate (MOCR). MOCR is defined as the ratio of citation count to publication count. It reflects the factual citation impact of a country, region, institution, research group etc. Jointly with the share of uncited papers (or cited papers, respectively) this indicator characterises the citation distribution of the unit under study.

ii)

Mean Expected Citation Rate (MECR). The expected citation rate of a single paper is defined as the average citation rate of all papers published in the same journal in the same year. Instead of the one-year citation window to publications of the two preceding years as used in the Journal Citation Report (JCR), a three-year citation window to one source year is used, as explained above. For a set of papers assigned to a given country, region or institution in a given field or subfield, the indicator is the average of the individual expected citation rates over the whole set.

iii) Relative Citation Rate (RCR). RCR is defined as the ratio of the Citation Rate per Publication to the Expected Citation Rate per Publication, that is,

RCR =

MOCR . MECR

This indicator measures whether the publications of a country or institution attract more or less citations than expected on the basis of the impact measures, i.e., the average citation rates of the journals in which they appeared. Since the citation rates of the papers are gauged against the standards set by the specific journals, it is largely insensitive to the big differences between the citation practices of the different science fields and subfields. It should be stressed that in this study, a 3-year citation window to one source year is used for the calculation of both the enumerator and denominator of RCR. RCR = 0 corresponds to uncitedness, RCR < 1 means lower-than-average, RCR > 1 higherthan-average citation rate, RCR = 1 if the set of papers in question attracts just the number of citations expected on the basis of the average citation rate of the publishing journals.

10

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

This indicators has been introduced by Schubert et al. (1983, 1986), and largely been applied to comparative macro and meso studies since. It should be mentioned that a version of this relative measure, namely, CPP/JCSm is used at CWTS in Leiden (see Moed et al, 1995). If subjects are delineated by subject categories, for instance, by those defined by ISI and journals as entities are assigned to such categories, then the RCR values of the world total in fields or subfields defined on the basis of these categories equal to the neutral value one. In the SCR database, most journals are only partially covered, that is, the RCR of the world total in the total field is not necessarily close to the neutral value of 1.0. RCR ≠ 1 for the world total is due to the fact that because of the partial coverage of most journals, the citation sets, compared in numerator and denominator of the RCRindicator for the total, are different. However, if SCR would have its own subject category or categories in the ISI-SCIE database, then this discrepancy would disappear and the RCR at the level defined by the ISI-SCIE SCR subject categor(y)/(ies) would then be equal to one, just as with all other subject categories. As a consequence of the observed discrepancy between the growth of publications and citations described above, the Mean Observed Citations Rate of publications in stem-cell research decreases slightly (see Figure 3, top). The decrease is, however, not remarkable. The linear trend-line indicates a gradient of -0.22. It amounts to 10% with respect to 1994. The Relative Citation Rate shows a slight decline; its decrease just amounts to 2% and the gradient of the trend-line is less than -0.01.Moreover, RCR is distinctly higher then 1.0 for the whole period. Although it is slightly decreasing, it does not much deviate from the median value of 1.14 at any time of the period. The SCR literature represents – on an average – the high-end of journal publications as measured through citations since the Mean Observed Citation Rate is pronouncedly higher that their expectation on the basis of the underlying set of journals. The RCR value of the world total significantly higher than 1.0 may be interpreted as a sign that SCR still is an emerging discipline.

11

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

14 12

MOCR

10 8 6 4 2 0 1994

1995

1996

1997

1998

1999

2000

2001

1994

1995

1996

1997

1998

1999

2000

2001

1.4 1.2 1.0

RCR

0.8 0.6 0.4 0.2 0.0

Figure 3 Evolution of citation impact in the period 1994-2001 (world total) Top: MOCR; Bottom: RCR

In order to gain detailed information about the increase of citation means, we analyse the distributions of citations over individual papers, one each for the beginning and the end of the period under study. The diagram is presented in Figure 4. The distributions for 1994 and 2001 are quite similar except for the shares of poorly and frequently cited papers. The distribution has evolved into a slightly less skewed one. The moments of the two distributions are very similar: The mean in 1994 amounts to 11.6, that in 2001 to 10.4. The harmonic means are with 2.42 and 2.41, respectively, almost identical. In verbal terms, the frequency distributions of citations over publications characterise the field as a specialty with high citation impact the citation patterns of which are, however, quite polarised.

12

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

16% 1994

2001

14%

Relative frequency

12% 10% 8% 6% 4% 2% 0% 0

1

2

3

4

5

6

7

8

9

10 11 12 13 14 15 16 17 18 19 20 >20

Citations

Figure 4 Distribution of citations over papers published in 1994 and 2001 Publication output and citation impact of Flanders and the 35 most active countries (19942003) For the analysis of national publication activity and citation impact, the 35 most active countries in the period 1994-2003 have been selected. Countries with less than 60 papers in the ten-year period have by reasons of statistical reliability not been selected. Flemish publications have been selected according to the following rule. A paper has been considered Flemish whenever an author contributing to the paper had a Flemish address. In addition, all relevant publications of the Vrije Universiteit Brussel (VUB) and all articles that could be assigned to the Flemish community have been taken into account. 20% of all other papers with a corporate address in Brussels but with no address in Flanders have been assigned to Flanders for the analysis of publication profiles. The publication output of the 35 most active countries in SCR and their share in the world total in stemcell research are presented in Table 1. In order to provide information about the evolution of national publication activity in the field, the period 1994-2003 has been split into two 4-year sub-periods, particularly, 1994-1997 and 2000-2003. National data in Table 1 are ranked in descending order by their publication output in the whole 10-year period. If we compare the list with similar lists on national publication output in all fields combined (cf., Glänzel et al., 2002), we can conclude that those countries that are most active in scientific research in all fields combined have top activity in stem-cell research, too. However, the UK ranks distinctly lower in SCR than in all fields combined. The contribution of the USA amounts to almost 50% of all SCR papers in the world. The corresponding value in all science fields combined is about one third. The share of the USA in SCR, however, decreased by about 10% if one compares the two sub-periods. The shares of Japan and Germany, in turn, grew considerably. The low share of Russia – otherwise holding rank seven in all fields combined – is worth mentioning. The share of Flanders reflects an “average” growth and the share of Flemish SCR papers in the Belgian national output amounts to 60%.

13

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Table 1

Publication output of Flanders and the 35 most active countries in the two sub-periods 1994-1997 and 2000-2003

Country/Region

USA Japan Germany UK France Italy Canada Netherlands Australia Spain Sweden Switzerland Austria Israel Belgium China PR South Korea Flanders Finland Russia Norway Denmark Czech Republic Taiwan Poland Turkey Brazil India Greece Mexico Argentina New Zealand Hungary Ireland Singapore Portugal WORLD TOTAL

1994-2003 Papers

21780 5468 4750 3995 3324 2611 2201 1479 1297 1053 965 938 701 670 569 426 410 338 332 289 288 277 213 204 176 168 166 148 147 105 102 102 90 84 84 65 46964

1994-1997

2000-2003

Share

Papers

Share

Papers

Share

46.4% 11.6% 10.1% 8.5% 7.1% 5.6% 4.7% 3.1% 2.8% 2.2% 2.1% 2.0% 1.5% 1.4% 1.2% 0.9% 0.9% 0.7% 0.7% 0.6% 0.6% 0.6% 0.5% 0.4% 0.4% 0.4% 0.4% 0.3% 0.3% 0.2% 0.2% 0.2% 0.2% 0.2% 0.2% 0.1% 100.0%

6921 1508 1238 1167 1010 709 676 444 444 226 221 268 207 185 141 39 46 89 90 106 88 61 55 49 28 25 26 28 29 26 21 40 19 20 9 8 14020

49.4% 10.8% 8.8% 8.3% 7.2% 5.1% 4.8% 3.2% 3.2% 1.6% 1.6% 1.9% 1.5% 1.3% 1.0% 0.3% 0.3% 0.6% 0.6% 0.8% 0.6% 0.4% 0.4% 0.3% 0.2% 0.2% 0.2% 0.2% 0.2% 0.2% 0.1% 0.3% 0.1% 0.1% 0.1% 0.1% 100.0%

10774 2925 2555 2036 1630 1359 1066 732 583 602 571 480 361 364 302 338 314 183 185 136 147 155 126 120 118 117 111 90 100 60 65 45 55 50 61 45 23984

44.9% 12.2% 10.7% 8.5% 6.8% 5.7% 4.4% 3.1% 2.4% 2.5% 2.4% 2.0% 1.5% 1.5% 1.3% 1.4% 1.3% 0.8% 0.8% 0.6% 0.6% 0.6% 0.5% 0.5% 0.5% 0.5% 0.5% 0.4% 0.4% 0.3% 0.3% 0.2% 0.2% 0.2% 0.3% 0.2% 100.0%

14

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Table 2 Country/Region USA Japan Germany UK France Canada Italy Netherlands Australia Sweden Switzerland Spain Israel Austria Belgium Flanders Finland Denmark Norway South Korea China PR Taiwan Brazil New Zealand Greece Russia Poland Portugal Hungary Czech Republic Singapore Mexico Ireland Argentina Turkey India WORLD TOTAL

Citation impact of Flanders and the 35 most active countries in 1994-2001 in the two subperiods 1994-1996 and 1999-2001 1994-2001 Share MOCR 62.1% 13.42 9.3% 8.35 8.8% 9.22 8.8% 10.58 7.0% 9.64 6.6% 13.94 4.4% 8.16 3.0% 10.00 2.9% 10.27 2.5% 13.28 2.3% 11.88 2.0% 9.17 1.5% 11.27 1.3% 8.90 1.2% 10.50 0.8% 11.86 0.6% 9.72 0.5% 8.64 0.4% 6.90 0.3% 4.48 0.2% 3.28 0.2% 4.80 0.2% 6.54 0.2% 7.33 0.2% 5.88 0.1% 2.36 0.1% 4.72 0.1% 10.49 0.1% 7.41 0.1% 3.21 0.1% 9.70 0.1% 3.72 0.1% 5.33 0.1% 3.84 0.1% 2.09 0.0% 1.79 100.0% 10.20

RCR 1.21 1.01 1.18 1.14 1.12 1.25 1.17 1.07 1.14 1.45 1.17 1.31 1.25 1.12 1.40 1.41 1.15 1.13 0.90 0.89 0.73 0.78 1.26 1.06 1.13 0.65 0.85 1.31 1.12 0.67 1.12 0.97 0.84 1.15 0.50 0.58 1.14

1994-1996 Share MOCR 65.1% 14.37 9.2% 9.43 8.3% 10.60 8.0% 10.65 6.5% 9.89 6.4% 14.30 3.3% 7.46 2.7% 9.78 3.1% 11.34 1.6% 12.43 2.3% 13.32 0.7% 5.74 1.1% 8.63 1.2% 10.11 1.2% 12.56 0.8% 12.68 0.7% 11.40 0.3% 9.32 0.5% 7.83 0.1% 2.70 0.1% 4.00 0.1% 1.86 0.1% 7.06 0.2% 6.79 0.0% 1.71 0.2% 2.89 0.0% 2.89 0.1% 13.50 0.0% 1.75 0.1% 4.35 0.0% 3.00 0.1% 4.40 0.1% 5.69 0.0% 3.60 0.0% 2.12 0.0% 0.87 100.0% 11.02

RCR 1.26 1.09 1.28 1.15 1.12 1.26 1.08 1.02 1.13 1.35 1.14 0.95 0.96 1.31 1.74 1.58 1.29 1.12 0.75 0.63 0.93 0.43 1.35 0.95 0.57 0.65 0.57 1.46 0.40 0.70 0.31 1.01 1.01 1.07 0.54 0.34 1.17

1999-2001 Share MOCR 60.6% 13.25 9.6% 8.00 9.3% 8.97 9.1% 10.70 7.1% 9.27 6.3% 12.86 5.1% 8.52 3.1% 10.47 2.3% 8.65 3.2% 14.09 2.3% 11.26 2.7% 10.27 1.7% 11.68 1.4% 8.76 1.4% 10.24 0.9% 11.77 0.7% 9.79 0.6% 8.24 0.4% 6.86 0.4% 4.89 0.3% 2.89 0.3% 6.14 0.2% 4.41 0.1% 8.08 0.3% 7.21 0.1% 2.61 0.2% 5.50 0.2% 12.11 0.2% 11.39 0.1% 2.95 0.2% 12.16 0.1% 3.68 0.1% 3.96 0.1% 3.95 0.1% 2.68 0.1% 2.08 100.0% 9.91

RCR 1.21 0.99 1.16 1.12 1.07 1.20 1.20 1.15 1.00 1.50 1.22 1.39 1.34 1.10 1.26 1.29 1.09 1.00 1.11 0.94 0.68 0.86 0.90 1.01 1.17 0.79 0.85 1.39 1.32 0.69 1.84 1.04 0.68 1.09 0.58 0.61 1.13

15

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

The citation impact of Flanders and the 35 selected countries in the period 1994-2001 is presented in Table 2. Analogously to the previous table, the period has been split into two sub-periods; taking into account that because of the 3-year citation windows, the upper limit for the publication year is 2001, two 3-year sub-periods have been chosen, particularly, 1994-1996 and 1999-2001. Countries have been ranked by the share of citations their papers published between 1994 and 2001 have received. Here again, the highly developed countries are ranking highest. The USA attracted the highest share of citations. The USA contributes with about 46% of all publications to the world total, but attract about 62% of all citations. Switzerland, USA, Canada, Sweden, Switzerland, Flanders and Israel have the highest MOCR values being in part far above the world standard. UK, Belgium and Portugal are also above the reference standard. Portugal is, however, subject to fluctuations because of its small publication output (cf. Table 1). This citation standard is, of course, set by “largest” countries, so that several developed countries – even big and medium-sized countries, for instance, France and Italy, have lower-than-average MOCR values. By contrast, the RCR values of most highly developed countries are significantly above the standard value of 1.0. Among these countries are, for instance, Sweden, Israel, Canada, Flanders, Belgium, Portugal, and in the second sub-period also Spain with RCR values distinctly above 1.10. Table 3 presents publication activity and citation impact of those 20 institutions that have been identified as most active in the field in the period 1994-2003. Institutions are ranked by their share in the world total. The shares range between about 1% and almost ten percent. The world total publication output in 1994-2003 amounted to 46964 publications, thus all institutions with more than 450 publications have been listed in Table 3. Institutional addresses have been cleaned up as far as possible, dis-aggregation of the big national institutions CNRS, INSERM, Max Planck Society and INCRA was not possible since the research institute or unit was in many cases not indicated in the address field. Also the National Institutes of Health (NIH) have not been disaggregated. Consequently, these institutions are ranking high. Besides large public research institutions, above all, universities proved to be most active. The field of most active research institutions is clearly dominated by American institutions. In particular, Harvard University and the National Institutes of Health – with more than 6% each – have by far the highest share in the world’s institutional publication output. Only two Japanese universities can be found among the world’s 20 “leading” institutions. European institutions with more than 300 publications each can be found at rank 23-35. These institutions are not listed in Table 3, and comprise the Karolinska Institute, the University of Tübingen, the University of Heidelberg, the University of Cambridge, the Max Planck Society, INRCA and the University of Vienna. The share of citations in all citations attracted in the field of SCR and the citation impact of the most active institutions is widely ranges. No doubt, American universities are the most active and leading institutes in SCR. All institutions listed in Table 3 have high (both, observed and relative) citation impact. Nevertheless, the HHMI and Stanford University have outstanding MOCR values that are even beyond the value of 20 citations per paper. At the first sight, non-American institutions seem to have distinctly lower citation-indicator values than the American ones. However, citation impact of Karolinska Institute (MOCR = 15.45; RCR = 1.70), Max Planck Gesellschaft (MOCR = 17.01; RCR = 1.16) and the Medical Research Council (MOCR = 15.45; RCR = 1.70) are at the same high level as that of the leading American institutes.

16

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Table 3

Publication output of the 20 most active institutions in 1992-2001 and their citation impact in 1992-1999

Institution HARVARD UNIVERSITY NATIONAL INSTITUTES OF HEALTH (ALL INSTITUTES) UNIVERSITY OF WASHINGTON INSERM UNIVERSITY OF TEXAS CNRS HOWARD HUGHES MEDICAL INSTITUTE UNIVERSITY OF TOKYO FRED HUTCHINSON CANCER RESEARCH CENTER UNIVERSITY OF PENNSYLVANIA UNIVERSITY OF TORONTO OSAKA UNIVERSITY INDIANA UNIVERSITY UCSF UNIVERSITY OF CALIFORNIA SAN FRANCISCO UCLA UNIVERSITY OF CALIFORNIA LOS ANGELES STANFORD UNIVERSITY UNIVERSITY OF MINNESOTA JOHNS HOPKINS UNIVERSITY MEMORIAL SLOAN KETTERING CANCER CENTER KYOTO UNIVERSITY 1

Share of papers1

Share of citations2

4.81% 4.30% 3.79% 3.01% 2.55% 2.34% 1.89% 1.79% 1.65% 1.49% 1.31% 1.23% 1.18% 1.08% 1.06% 1.05% 1.05% 1.02% 0.96% 0.96%

7.78% 6.04% 4.61% 3.08% 2.97% 2.31% 4.96% 1.77% 2.12% 1.91% 2.56% 1.74% 2.56% 1.79% 1.55% 2.09% 1.15% 1.59% 1.50% 1.07%

MOCR2

18.63 15.80 15.58 11.85 12.77 11.83 28.25 11.51 14.41 13.81 18.57 15.46 18.57 16.93 15.25 21.08 11.24 17.26 15.56 12.68

RCR2

1.24 1.30 1.26 1.09 1.39 1.02 1.17 1.05 1.31 1.25 1.33 1.26 1.33 1.18 1.21 1.28 1.17 1.53 1.27 1.05

Share in the world total is based on the sub-period 1994-2003.

2 Share in the world total as well as MOCR and RCR are based on the sub-period 1994-2001.

III. Flemish stem-cell research in the mirror of bibliometric indicators (19942003) Figure 5 presents the distribution of Flemish publications over sectors. The chart at the top show the Flemish publication profile in SCR literature, that at the bottom gives the profile in all fields combined for comparison. The predominance of university research in both charts is obvious. Nevertheless, the shares of both Sector I and Sector IV are somewhat greater in SCR than in all fields combined. In this context, it should be stressed that because of scientific collaboration between sectors percentage shares cannot be summed up to 100%.

17

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

V

0.3%

IV

5.6%

III

4.1%

II

7.1%

91.1%

I

0%

V

0.4%

IV

4.1%

III

6.0%

II

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

60%

70%

80%

90%

100%

8.3%

88.3%

I

0%

10%

20%

30%

40%

50%

Figure 5 Flanders’ publication profile of SCR by sectors in 1994-2003 (top) compared with that in all science fields combined (bottom) (I Higher education, II Public research institution or government, III Private institution, IV Hospital, V Others)

Table 4 presents indicators of publication activity and citation impact of the Flemish institutions by sectors in 1994-2001. The shares of publications in the Flemish total are in line with the corresponding data shown in Figure 5. The analysis of citation indicators, however, reveals interesting details. The high relative citation rate of about 1.4 in the educational sector as well as the high MOCR value of 14.8 in the private sector is remarkable. The high citation impact in the private sector cannot be explained with subject specific publication profiles alone since the corresponding RCR values also lies distinctly above the neutral value of 1. The low impact of the output of public institutions is somewhat striking. More than 50% of all papers in this sector have an author at VITO, the Flemish Institute for Technological Research. These papers are mostly concerned with stem-cell research in the context of (environmental) toxicology and mutagenesis. This might in part explain their low impact.

18

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Table 4

Publication activity and citation impact of the Flemish institutions by sectors in 1994-2001 Sector I II III IV V Total

Papers 211 22 10 13 0 236

1994-2001 Share of MOCR papers 89.5% 12.73 9.2% 2.82 4.2% 14.80 5.5% 10.62 0.0% 100.0% 11.86

RCR 1.42 0.71 1.28 1.26 1.41

IV. International co-authorship patterns in Stem-Cell Research The global collaboration network of research in SCR International collaboration may reflect individual interests and motivation of individual scientists. Some of the factors influencing co-publication patterns have already been discussed in the basic papers on scientific collaboration by deB. Beaver and Rosen (1979) and Luukkonen et al. (1992). In a recent paper, deB. Beaver (2001) has summarised eighteen main criteria for which authors collaborate. When one considers international collaboration, the economic and/or political dependence of a country or geopolitical region (such as the different forms and degrees of neo-colonial ties) or large or special equipment (such as CERN in Switzerland and the observatories in Spain or Chile), which are often shared in large multinational projects, also condition co-operation, apart from any individual motivation. And scientific collaboration between member countries of the EU promotes European integration into one of the world's most advanced systems of science and technology. On the other hand, co-publications might simply result as mandatory exercises within the framework of bilateral agreements between institutions, science administrations or governments. It is clear that a variety of different purposes and motivations, the manifold of factors influencing (international) collaboration must have at least in part a measurable impact on the published results of joint research work. In a recent study by Glänzel and Schubert (2001), the relation between international co-authorship and citation impact in the field of chemistry has been studied. Their results often confirmed but sometimes contradicted widespread notions on the efficiency of international collaboration. Using a more complex scheme, national characteristics in international scientific co-authorship patterns have been studied by Glänzel (2001). This study has confirmed again that international collaboration has strongly intensified in the 90s. An interesting observation has been made concerning the re-integration of EIT countries into the scientific collaboration structures of Europe and the Western world. The strong co-authorship link between Hong Kong and China might indicate the beginning coalescence of the two scientific systems. The absolute number of international papers and their share in the total national publication output serve as basic indicators of international co-authorship relations and scientific collaboration. International collaboration depends on the country's ‘size’ (cf., for instance, Schubert and Braun, 1990 and Katz, 2000). At the national level, the share of international collaboration in large countries is necessarily lower than that of medium-sized or even small countries. The share of all international papers in the world can, in principle, be determined as the complementary share of the ratio of all countries’ domestic papers and the total world publication output. However, such ‘world average’ has not proved an appropriate reference standard for international collaboration activity (cf. Schubert and Braun, 1990), and will, therefore, not be used here. Unlike in the case of basic statistics on publication activity and citation impact, where according to the 20%/80% rule, one fifth of all papers with an address in Brussels and all citations they have received

19

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

could be assigned to Flanders, the international co-publication patterns of Flemish and the citation impact of internationally co-authored papers must be analysed on individual links that are split into country pairs. Such procedure is not applicable here. The analysis is, therefore, restricted to papers with at least one address in Flanders or an address at VUB. Since this set forms the overwhelming share of all Flemish papers, this data collection can be considered representative. Table 5 presents number and share of internationally co-authored publication of Flanders and the 35 most active countries in the period 1994-2003 and the two sub-periods 1994-1997 and 2000-2003, respectively. Countries have been ranked by the share of international co-publications in the total national publication output. In order to allow a fair comparison, SCR papers published in multidisciplinary journals have not been taken into account for Flanders. Hungary, Switzerland, Finland, Portugal and Ireland have the highest share of international copublications. More than half their papers have been published in international collaboration. Among the countries with high share of international co-publications, we also find the other Nordic countries, Canada, Flanders, Belgium, Poland, Greece and Singapore. Although there is in overall trend towards increasing the share of internationally co-authored papers followed by most countries, several countries show decreasing collaboration patterns in so far the low number of publications allows such conclusions. Among these countries, we find the Czech Republic and Brazil. Also in the Nordic countries certain stagnation can be observed. The share in Sweden even decreased. The question arises of whether this trend could be interpreted in the context of general trends in Swedish life-science research (cf. Glänzel, 2000, Glänzel et al., 2003).

20

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Table 5

International co-publications of Flanders and the 35 most active countries in the two subperiods 1994-1997 and 2000-2003 (ranked by share of international co-publications in 2000-2003) Country/Region Hungary Switzerland Finland Ireland Portugal Denmark Norway Singapore Greece Poland Flanders Canada Belgium Sweden New Zealand Austria Netherlands Argentina Israel Germany Australia UK France Czech Republic Italy Spain Brazil Turkey Taiwan China PR South Korea Russia USA Japan India Mexico

1994-1997 Papers Share 7 36.8% 133 49.6% 48 53.3% 11 55.0% 5 62.5% 29 47.5% 44 50.0% 4 44.4% 11 37.9% 13 46.4% 30.3% 27 268 39.6% 48 34.0% 110 49.8% 11 27.5% 94 45.4% 174 39.2% 8 38.1% 75 40.5% 388 31.3% 147 33.1% 337 28.9% 291 28.8% 30 54.5% 233 32.9% 65 28.8% 13 50.0% 1 4.0% 12 24.5% 18 46.2% 21 45.7% 35 33.0% 1326 19.2% 299 19.8% 2 7.1% 5 19.2%

2000-2003 Papers Share 33 60.0% 284 59.2% 107 57.8% 27 54.0% 23 51.1% 77 49.7% 71 48.3% 29 47.5% 47 47.0% 55 46.6% 85 46.4% 492 46.2% 138 45.4% 254 44.5% 20 44.4% 155 42.9% 310 42.3% 27 41.5% 151 41.5% 1013 39.6% 228 39.1% 787 38.7% 585 35.9% 44 34.9% 452 33.3% 192 31.9% 33 29.7% 34 29.1% 34 28.3% 93 27.5% 86 27.4% 37 27.2% 2491 23.1% 619 21.2% 19 21.1% 11 18.3%

21

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Table 6

Citation impact of international co-publication of Flanders and the 35 most active countries in the two sub-periods 1994-1996 and 1999-2001 (ranked by share of citations through international co-publications in 1999-2001) Country/Region Singapore Hungary Argentina Portugal Poland Russia New Zealand Ireland Turkey Switzerland Greece Austria Finland Norway Spain Denmark Belgium Canada Sweden Italy Israel Germany France Flanders Netherlands Brazil Australia UK Taiwan Czech Republic India South Korea Japan China PR USA Mexico

Share 72.2% 71.4% 87.0% 92.6% 75.0% 81.1% 71.1% 83.8% 8.3% 54.6% 44.8% 60.4% 87.9% 65.5% 57.5% 77.2% 54.7% 50.4% 66.3% 49.4% 65.2% 56.4% 51.1% 62.1% 53.0% 65.8% 43.5% 43.8% 32.3% 89.2% 15.4% 74.1% 33.7% 80.2% 22.5% 50.0%

1994-1996 MOCR – – – – 3.55 7.04 – – – 15.56 – 13.97 17.59 9.84 11.24 13.41 19.32 19.49 16.26 11.71 13.73 18.48 17.73 24.32 14.10 – 16.13 15.81 – 6.29 – 4.00 16.65 7.75 17.10 –

RCR – – – – 0.51 0.76 – – – 1.27 – 1.27 1.46 0.80 1.21 1.33 2.12 1.44 1.61 1.19 1.10 1.45 1.47 2.39 1.16 – 1.42 1.36 – 0.85 – 0.64 1.31 1.14 1.31 –

Share 97.0% 89.5% 88.4% 86.1% 79.8% 78.1% 73.3% 70.7% 70.4% 69.6% 68.6% 68.1% 66.8% 65.5% 65.2% 62.3% 60.4% 60.3% 57.9% 56.9% 56.0% 55.0% 53.9% 53.5% 52.5% 51.9% 51.4% 47.7% 47.6% 46.3% 44.0% 40.6% 39.0% 38.2% 26.4% 25.2%

1999-2001 MOCR 20.36 22.57 10.75 17.18 8.46 5.28 12.33 4.67 6.25 13.42 10.92 12.90 11.21 10.35 22.19 10.67 12.54 16.69 17.50 15.52 16.08 13.09 14.42 12.75 13.42 6.75 11.71 13.60 9.62 4.82 3.67 8.19 13.79 3.57 15.25 –

RCR 2.31 1.63 1.22 1.57 0.90 0.86 1.07 0.73 0.85 1.27 1.30 1.23 1.02 1.31 1.74 0.95 1.28 1.30 1.51 1.56 1.42 1.28 1.19 1.24 1.34 1.03 1.10 1.24 0.94 0.79 0.58 1.12 1.10 0.58 1.26 –

The evolution of co-publication patterns is several countries such as Portugal and Singapore must be interpreted with utmost care because of their minute publication output in this field in the first sub-period. Table 6 presents the share of citations received by internationally co-authored publications in all citations received by SCR papers in the corresponding country. Citations means are only given if the underlying paper set contains at least 10 publications. Many medium-sized countries did not meet this condition in the sub-period 1994-1996. Internationally co-authored publications are often the main source of citations for EIT countries and less developed countries. The increase of the citation means in Hungary and Argentina are quite dramatic. In other countries, e.g., Brazil, China, Poland, Czech Republic and Turkey changes are less spectacular. The share of citations received by international copublications in these countries is higher than the share of international co-publications itself (cf., Table

22

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

6). In developed European, North American and Asian countries, the deviation between the share of citations and publications is somewhat lower. In most countries under study, international co-authorship, on average, results in publications with higher citation rates than purely domestic papers. This effect can be observed in stem-cell research, too. Above all, Hungary, Singapore, Portugal, Italy, France and Spain benefited most from international collaboration.

Mapping mutual co-authorship links In order to measure the strength of mutual collaboration links, an appropriate similarity measure has to be used. Collaboration networks will be analysed on the basis of such measure determined for country pairs. Multinational collaboration will therefore be split up to a group of bilateral relations. In order to give an example for this procedure, we proceed from a two fictitious Flemish publication where one paper has two other addresses, one in the Netherlands and one in Germany; the second Flemish paper has a co-author with affiliation in the Netherlands. These papers thus define exactly four links, two between Flanders and Netherlands, and one each between Flanders and Germany, and between Germany and the Netherlands, respectively. A frequently used measure for the strength of co-publication links is the cosine measure according to Salton. It is defined as the number of joint publications divided by the square root of the product of the number (i.e., the geometric mean) of total publication outputs of the two countries, that is,

r=

pij pi ⋅ p j

,

where pij the number of links between the countries i and j, and pi (pj) the total number of publications of the country i (j) is. Following the practice of earlier studies, the ‘natural topology’ is used to illustrate the structure defined by scientific co-authorship on the basis of Salton’s measure calculated for country pairs. It should be mentioned here that as a consequence of treating collaboration links of each country pair separately, co-publication counts and shares are not additive, and thus cannot be summed up to the total over any part of the world. One has, consequently, to distinguish between the number of co-publications and of co-authorship links.

23

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

FIN N OR SW E DN K

GB R

NLD

POL DEU

B EL

CZE

FR A

AUT

CHE

ITA ESP TUR

ISR

RUS

CHN

CAN

KOR

USA

JPN

TWN

AUS

Map source: University of Alabama, Cartographic Research Lab

Figure 6 International co-publication links in the world (1994-1997) (dotted line ≥ 2.5%, solid line ≥ 5%, thick line ≥ 7.5%) 24

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

FIN NO R SW E DNK

GBR

N LD

POL DEU

BEL

CZE

FR A

AUT

CHE

ITA ESP TUR

ISR

RUS

CH N

CAN

KOR

USA

JPN

TW N

AUS

Map source: University of Alabama, Cartographic Research Lab

Figure 7 International co-publication links in the world (2000-2003) (dotted line ≥ 2.5%, solid line ≥ 5%, thick line ≥ 7.5%)

25

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

In the light of the results of this study, the co-publication network formed, on one hand, by the 25 most active countries and by Flanders and its most important collaborations partners, on the other hand, will be studied. The restriction to 25 countries for the “global” part had to be made to guarantee the clarity of the presentation of collaboration structures. Strong links beyond this level will be discussed separately. The following three thresholds have been chosen. rij = 2.5% is the lower threshold indicating medium strength; the second threshold of rij = 5.0% indicates strong links, and rij = 10%, finally, indicates very strong links between the countries i and j. The co-publication networks of the 25 most active countries for the two sub-periods 1994-1997 and 2000-2003 according to Salton’s measure are presented in Figures 6 and 7, provided the links are based of at least 25 joint papers. The changes are not striking; they are in line with the evolution of co-publication networks already described in other fields (e.g., Glänzel and Schubert, 2001, Glänzel, 2001). The general trend towards intensifying collaboration is reflected by the increasing number of solid and thick lines. Four important nodes could be found in this network: USA in America and Germany, UK, France – and to a lesser extent, Italy – in Europe. Several distinct clusters could be identified although the selection of most active countries results in some restrictions. There is a small Scandinavian cluster, a big European one, and a small North American cluster. The European and American Clusters are connected through the USA, particularly through medium links with Germany, France, UK and Italy. The German-US link has evolved to a strong one in the second sub-period. The “density” of the network has considerably increased; however, traditional patterns of geopolitical and economic affinities are still present. The links among Scandinavian countries and the link between Germany and Austria might be interpreted in this context. Finally, we will also mention several strong links not presented in the two maps because at least one of the partner countries was not in the group of the 25 most active countries. In the first sub-period (1994-1997), there were strong links with rij > 7.5% each between Germany and Austria, between Finland and Sweden, and between USA and Canada. In the second sub-period (1998-2001), seven strong links with rij > 7.5% could be observed. In Far East, Japan is strongly linked with the North American through the USA. The strongest link in the second sub-period could be observed between the Sweden and Finland; Salton’s measure lies above 11% here. This was followed by Germany–Switzerland and USA–Canada (both >9% each).

International co-publication patterns of Flemish research in Stem-Cell Reasecrh In the following, the international co-publications in Flanders and their citation impact will be analysed. Table 7 presents the most important partners, the number of joint publications, provided this number was not less than 10, and their citation impact in the 8-year period. Similarly to the macro analysis of internationally co-authored papers in SCR, the following analysis had to be restricted to papers with at least one address in Flanders or an address at VUB. This data collection can, however, be considered representative. The most important partners are USA, Netherlands, Germany, France, UK and Italy. Co-publications with all countries shown in Table 7 have, on an average, been frequently cited; the MOCR values of Flemish co-publications with these countries can be considered very high. Collaboration with these countries seems to pay. Above all, joint papers with Germany and UK have an almost extremely high impact since not only the MOCR values are far above the standard, but also the corresponding RCR values of 2.3 and 3.0, respectively, are far beyond the reference standard. However, several publications are result of multinational collaboration, that is, the same papers are then assigned to all individual countries involved. These papers have usually attracted many citations, indeed.

26

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Table 7

Citation impact of international co-publications of Flanders with individual countries (at least 10 joint publications in 1994-2001) Country USA Netherlands Germany France UK Italy

Papers 29 24 23 19 17 13

1994-2001 MOCR 18.17 11.04 23.78 17.74 26.94 17.08

RCR 1.50 1.12 3.00 1.67 2.31 1.74

Figures 8 and 9 present the international co-publication network of Flanders in the field in the two sub-periods. For these maps, three thresholds different of those used in Figures 6 and 7 have been chosen. A second condition was that collaboration links are based on at least 5 joint papers. The intensification of collaboration at this level is obvious. Flanders has traditionally strong links with the Netherlands. Collaboration with the Netherlands did not result in a sufficiently large paper set in the first sub-period. By the end of the 90’s, Flanders established a strong link with the Netherlands, the strength of which now exceeds 5% according to Salton’s measure, and we have also found medium links with France, Germany, UK and Italy.

27

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

USA

VL

DEU

ITA

Map source: University of Alabama, Cartographic Research Lab

Figure 8 International co-publication links of Flanders (1994-1997) (dotted line ≥ 1%, solid line ≥ 2.5%, thick line ≥ 5%)

28

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

CAN USA GBR

NLD VL

FRA

ESP

DEU

CHE

ITA

Map source: University of Alabama, Cartographic Research Lab

Figure 9 International co-publication links of Flanders (2000-2003) (dotted line ≥ 1%, solid line ≥ 2.5%, thick line ≥ 5%)

29

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

REFERENCES D. DEB. BEAVER, R. ROSEN, Studies in scientific collaboration. Part I. The Professional Origins of Scientific Co-authorship, Scientometrics, 1, 1978, 65-84, Part II. Scientific Co-authorship, Research Productivity and Visibility in the French Elite, Scientometrics, 1, 1979, 133-149. D. DEB., BEAVER, Reflections on scientific collaboration (and its study): Past, present, and future, Scientometrics, 52 (3), 2001, 365-377. D. CAMPBELL, M. NOISEUX, G.CÔTÉ, Potential for Stem Cells Science and Technology in Canada: Great Promises and Challenges, Science-Metrix report, April, 2004, pp. 63. N. CHENG, Paradox in a Petri Dish, Ethical Issues in Human Embryonic Stem Research, Journal of the Hippocratic Society, 2004, 56-60. W. GLÄNZEL, H. J. CZERWON, A New Methodological Approach to Bibliographic Coupling and Its Application to the National, Regional and Institutional Level, Scientometrics, 37 (2), 1996, 195-221 W. GLÄNZEL, A. SCHUBERT, H.-J. CZERWON, A Bibliometric Analysis of International Scientific Co-operation of the European Union (1985-1995), Scientometrics, 45, 1999, 185-202.s W. GLÄNZEL, Science in Scandinavia: A Bibliometric Approach, Scientometrics, 48 (2), 2000, 121-150. W. GLÄNZEL, R. DANELL, O. PERSSON, The decline of Swedish neuroscience – decomposing a bibliometric national science indicator, Scientometrics, 57 (2), 2003, 197-213. W. GLÄNZEL, National Characteristics in International Scientific Co-authorship, Scientometrics, 51 (1), 2001, 69-115. W. GLÄNZEL, A. SCHUBERT, T. BRAUN, A relational charting approach to the world of basic research in twelve science fields at the end of the second millennium, Scientometrics, 55 (3), 2002, 335-348. W. GLÄNZEL, M. MEYER, B. SCHLEMMER, M. DU PLESSIS, B. THIJS, T. MAGERMAN, K. DEBACKERE, R. VEUGELERS, Nanotechnology – Analysis of an emerging domain of scientific and technologic endeavour, 2003a, http://www.steunpuntoos.be/nanotech_domain_study.pdf W. GLÄNZEL, M. MEYER, B. SCHLEMMER, M. DU PLESSIS, B. THIJS, T. MAGERMAN, K. DEBACKERE, R. VEUGELERS, Biotechnology” - An Analysis based on Publications and Patents, 2003b, http://www.steunpuntoos.be/Biotech.Report.November2003.pdf Y.S. Ho, C.H. Chiu, T.M. Tseng, W.T. Chiu, Assessing stem cell research productivity, Scientometrics, 57, 2003, 369-376. ISI Essential Science Indicators, Embryonic Stem Cell Research, Retrieved 22 November, 2004 from: http://esi-topics.com/stemcells/ E. JONIETZ, Mapping a stem cell policy, Technology Review 106 (8), 24 October 2003. J. S. KATZ, Scale-independent indicators and research evaluation, Science and Public Policy, 27 (1), 2000, 23-36. T. LUUKKONEN, O. PERSSON, G. SILVERTSEN, Understanding Patterns of International Scientific Collaboration, Science, Technology & Human Values, 17, 1992, 101-126. H. F. MOED, R.E. DE BRUIN, TH.N. VAN LEEUWEN, New bibliometric tools for the assessment of national research performance: database description, overview of indicators and first applications, Scientometrics, 33, 1995, 381-422. A. SCHUBERT, T. BRAUN, Relative indicators and relational charts for comparative assessment of publication output and citation impact, Scientometrics, 9, 1986, 281-291. A. SCHUBERT, T. BRAUN, International Collaboration in the Sciences, 1981-1985, Scientometrics, 19, 1990, 3-10.

30

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

A. SCHUBERT, W. GLÄNZEL, T. BRAUN, Relative Citation Rate: A New Indicator for Measuring the Impact of Publications. In: D. Tomov, L. Dimitrova (Eds.), Proceedings of the 1st National Conference with International Participation on Scientometrics and Linguistic of the Scientific Text, Varna 1983, 80-81. A. SCHUBERT, W. GLÄNZEL, T. BRAUN, World flash on basic research: Scientometric datafiles. A Comprehensive set of indicators on 2649 journals and 96 countries in all major science fields and subfields, 1981-1985, Scientometrics, 16 (1-6) (1989) 3-478.

31

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

APPENDIX - Definition of country abbreviations Standard ISO codes of countries FLANDERS ARGENTINA AUSTRALIA AUSTRIA BELARUS BELGIUM BRAZIL BULGARIA CANADA CHINA CZECH REPUBLIC DENMARK FINLAND FRANCE GERMANY GREECE HONG KONG HUNGARY INDIA IRELAND ISRAEL ITALY JAPAN KOREA LITHUANIA MEXICO NETHERLANDS NEW ZEALAND NORWAY POLAND PORTUGAL ROMANIA RUSSIAN FEDERATION SINGAPORE SLOVAKIA SLOVENIA SOUTH AFRICA SPAIN SWEDEN SWITZERLAND TAIWAN TURKEY UKRAINE UNITED KINGDOM USA YUGOSLAVIA



VL* ARG AUS AUT BLR BEL BRA BGR CAN CHN CZE DNK FIN FRA DEU GRC HKG HUN IND IRL ISR ITA JPN KOR LTU MEX NLD NZL NOR POL PRT ROU RUS SGP SVK SVN ZAF ESP SWE CHE TWN TUR UKR GBR USA YUG

Flanders has no standard ISO code

32

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

PART II STEM-CELL TECHNOLOGY – A TECHNOMETRIC APPROACH I. Introduction The structure of this chapter is as follows. In section II the approach used to identify and select the stem cell related patents is being described in detail. Subsequently, the results of the search strategy are presented and discussed. A total number of 780 USPTO patents and 736 EPO patents have been identified in the respective window of analysis (being: 1994-2003). Section III presents a detailed analysis of stem cell patents, first on a worldwide level, then at the specialization degree of different countries in stem cells and the different collaboration patterns, and finally followed by an analysis of the main organisations behind stem cell patenting. Section IV briefly elaborates on a number of ethical/legal issues possibly affecting technology development in stem cells. Section V summarizes the main findings and concludes the elaboration.

II. Search Strategy The search procedure and it’s effectiveness Any subject-specific research based on patent data requires a specific search strategy for retrieving the adequate patent subset. The reason for this is that patent classification systems like the International Patent Classification (IPC) have not been designed in the first place for research purposes, but instead they have been designed to assist patent examiners in their search for prior-art. As a consequence, search procedures need to be customized based on the objective of the study in question and mainly based on the characteristics of the technology (sub-) area that is under investigation. Usually two search strategies can be identified: 1) a search based on a combination of IPC-codes, and 2) a search based on a combination of keywords best describing the topic under analysis. In many cases a combination of the two can also be quite satisfactory. In the case of ‘stem cells’ which is a rather novel field of investigation, the traditional IPC-classification does not foresee in one specific class where stem cell related patents could be classified in. However, in the IPC system there are several subclasses containing patents related to a very specific aspect of stem cell related research, namely ‘cultivation of cells’ (C12M, C12N, and C12Q) - not all patents in these subclasses are related to stem cell research and/or technology. These subclasses are, according to the OST/INPI/ISI Technology Classification scheme, part of the broader field of Biotechnology. In the light of the previous, the search strategy employed is based on a selection of keywords best representing the field. Two levels have been distinguished in the search strategy: Level 1: Identification of core stem cell patents On this level, the identification of patents forming the core of stem cell research and technology was aimed for. To identify the core stem cell patents we focussed on four keywords, which according to the internal document analysis and the external expert validation best represented ‘stem cells’. Subsequently, we have identified these keywords in the title and/or abstract of a selection of patents. The core keywords could co-occur among each other or with the related keywords (see below). If a patent had one or more of these keywords in the title or abstract it is considered a ‘core’ stem cell patent.

33

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

The four keywords that have been used are: a) *stem cell* b) * ES cell* or *.ES cell* or -ES cell* - (meaning: Embryonic Stem) c) *progenitor cell* d) *hematopoi* cell* Level 2: Identification of related stem cell patents The second level of identification made it possible to identify patents which are ‘closely’ related to the ‘core’ stem cell research and technology. Although they do not have one of the above mentioned keywords in the title or abstract, the occurrence of at least 2 stem cell related keywords made it possible to identify strongly related patents. The list of keywords is also based on the analysis of literature reviews and articles about stem cell science. The ‘related’ keywords include: 1. *blastocyst* 2. *pluripotent* 3. *multipotent* 4. *totopotent* 5. *Oct 4* 6. *leukaemia inhibitory factor* or *leukemia inhibitory factor* 7. *fibroblast* 8. *quiescent cell* 9. *mammalian telomerase* 10. *human telomerase* 11. *germ cell* 12. *nuclear transfer* 13. *donor cell* 14. *allogeneic* 15. *clonogenic* 16. *inner cell mass* 17. *stromal cell* 18. *feeder layer* or *feeder cell* 19. *embryoid bod* 20. *transdifferentiation* 21. *initiating cell* 22. *immortalized* 23. *precursor cell* 24. *bone marrow* 25. *umbilical cord blood* or *umbilical-cord-blood* 26. *cytopoi* 27. *megakaryopoi* 28. *erytropoi* 29. *myelopoi* 30. *trombopoi* 31. *CD34* 32. *myogenic cell* 33. *neurosphere* Patents related to stem cells had to have at least 2 of these related keywords in the title or abstract. After this selection a final step in identifying the closely related patents was done. By reading the abstract of the related patents, the conclusion was drawn that only the ‘related’ patents with at least one of the following keywords in the title or abstract should be included in the closely related stem cell group; these keywords are: *pluripotent*, *progenit*, *precursor*, *differentiat*, immortal* and *totipotent*. When two of the previously mentioned ‘related’ keywords were used in combination with at least one of the additional keywords in the patent abstract or title, then the patents were identified as ‘closely’ related. The results of the search strategy are presented in table 1.

34

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Table 1 - Results of the keyword based patent retrieval in ‘stem cells’ U.S. Patent data 1991 1994-20031 onwards 806 66 872

Level 1: ‘Core’ Level 2: ‘Related’ Total

European Patent data 1981 onwards 1994-20031

718 62 780

849 39 888

705 31 736

As the table shows, the sample of ‘stem cell’ patents between 1994-2003, the period under analysis, is quite representative for the overall patenting activity in this field (89% of the EPO stem cell patents fall in the period 1994-2003; 82,8% in the case of U.S. patents). In general, the evaluation of the retrieved patents by Dr. Szarkadi resulted in a relevance analysis to be varying between ‘medium’ to ‘high’, very similar to what was found in the publication analysis. A set of 9 UPSTO patents (or +/- 1,0% of the sample) could be positioned in the periphery of stem cell technology. This is mainly due to the inclusion of the keyword ‘immortalized’, which appeared not selective enough to filter less relevant items. As these 9 patents are related, even though to a less extent, they will be included in the further analysis. The two-level search strategy proved to be a feasible and efficient method in retrieving the ‘core’ and ‘related’ stem cell patents from the EPO and USPTO patent databases.

A closer look at the evolution in stem cell patenting The application of the above mentioned search strategy lead to the retrieval of respectively 780 USPTO patents and 736 EPO patents for the period under review. Figure 1 illustrates the evolution in stem cell patenting over time. Figure 1 - Evolution of stem cell related patenting over time 160

Number of patents

140 120 100 80 60 40 20 0 1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

Year EPO applications

EPO grants

USPTO grants

Even though the main period under analysis comprises 10 years (1994-2003) the evolution of patenting in stem cells starts already in the early 1980s. During the 1980s, specifically the mid 1980s, we have found a strong increase in the number of patent applications in Europe, followed by an almost parallel increase in the number of granted patents in the U.S.. 1

The period mentioned refers to the ‘Application Year’ of the patent. In case a patent has been applied for between 1994 and 2003 the patent is included in the sample.

35

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Noteworthy is the low number of granted stem cell related patents by the European patent office (EPO) over time (in total 122 granted patents compared to 878 grants by the United States Patent and Trademark Office (USPTO) over the full time period). By speculating, one might assume that this could be the result of a different attitude towards the ethical and legal issues around stem cell research and technology in the U.S. and Europe and thus an increased prudence in patent granting in Europe. Moreover, in the U.S. it seems that at a very early stage stem cell research was being transferred to a private setting in order to avoid the legal barriers (Campbell et al., 2004). During the beginning of the 1990s we can observe a strong increase in patenting (applications in Europe and grants in the U.S.) reaching a high point one decade later, at the beginning of 2000 (EPO applications: 134 in 2001; USPTO grants 127 in 1999). The sharp decrease in European patent applications is of specific interest. On the one hand this could be due to a data artefact, the completeness of the data available in 2002 and 2003 in view of publication practices, on the other hand there might be other explanations perhaps explaining a ‘real’ decrease in stem cell related patent applications at the EPO, a suggestion that is also posed in a Canadian study on stem cells research and technology. This study (the only available to our knowledge) carried out for the Canadian Biotechnology Secretariat (Campbell et al., 2004) confirms the overall pattern presented above (also the levels of absolute number of retrieved stem cell related patents). Currently, stem cell patents only represent a very small proportion of total EPO and USPTO patents, mainly due to the fact that research in this field was and still is at a very early stage of development while the more fundamental (also ethical) issues in relation to the biology of stem cells are still being dealt with. As more of these fundamental issues are being answered it is likely that the patenting activity will increase the coming years (Campbell et al., 2004). As reported in the previous section, the identified ‘stem cells’ set has been judged to be representative for patented stem cell technology. In this section we will look more closely at the different technology classes and fields to which the identified ‘stem cell’ patents have been classified in. It should be mentioned that the IPC-classification system, on which the classification in broader technology fields is based, is specifically designed as a classification system to support the work of the patent examiner and is not ‘directly’ related to the more explicit notions of technological and/or economic classes. However, the classification designed by OST/INPI/ISI offers the possibility to approximate these notions. Figure 2 shows the technology fields related to stem cell research and technology.

36

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Figure 2 - Distribution of stem cell patents (source: EPO) over the different technology subclasses

Percentage stem cell patents

60,00%

50,00%

40,00% 30,00%

20,00%

10,00%

Ag ric

ul tu

ra la nd

fo od

pr oc es si

ng ,m

ac hi ne ry Ag an An r ic Ch d al ul ap ys em t u is, pa re ica ,f ra m la oo tu ea nd s d su c pe re he m m tr o en is l in try t, du te ch st ry no ,b lo Bi as gy ot ic ec m hn at ol er M og ia ac ls y ro Ch c h m em em ol ec is ica tr y ul le ar n ch gi ne em er ist in ry g ,p M ol ym ed ic er al s O te rg ch a Ph ni no c lo ar f in gy m e ac c h eu e m tic ist al s, ry co sm et ic s

0,00%

Technology subfield

As expected, based on the analysis of the EPO data, the majority of stem cell patents (over 50%) are related to the broader area of biotechnology, in second instance we find around 25% of patents to be related to the area of pharmaceuticals. Around 10% of stem cell patents are traced back to organic fine chemistry, followed by medical technology and other chemistry and agriculture related fields to which smaller proportions of stem cell patents can be attributed. A similar field pattern occurs when looking at the USPTO patent data (see figure 3). Here also we find most of the stem cell patents to fall into the areas of biotechnology and pharmaceuticals. Analysis and measurement technology accounts for a slightly higher share of stem cell patents in comparison to the EPO patents (see figure 2).

37

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Figure 3 - Distribution of stem cell patents (source: USPTO) over the different technology classes

Percentage stem cell patents

60,00%

50,00%

40,00%

30,00%

20,00%

10,00%

Ag ric

ul tu

ra la nd

fo od

pr oc es si

ng ,m

ac hi ne ry Ag an An ric Ch d al u ap ys ltu em i pa r s e ica ,m ,f ra la oo tu ea nd s d su c pe re he m m tr o en is l in tr y t, du te ch st ry no ,b lo Bi as gy o ic te m ch at no er lo ia gy ls Ch Co ch e em ns m um is ica try er le go ng od in ee s an rin d g eq M u ed ip m ic en al O t te rg ch a Ph ni no c lo ar f in gy m e ac c he eu m tic ist al s, ry co sm et ic s

0,00%

Technology subfield

In the following sections the results of a more detailed analysis of patented stem cell research and technology will be presented.

38

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

III. Detailed analysis of stem cell patenting Development of stem cells worldwide In this section we will discuss the evolution in stem cell patenting from a worldwide perspective, after which we will zoom into the Belgian and Flemish performance. In table 2 the country rankings are presented based on USPTO patent data. Table 3 presents the results based on the EPO patent data. In line with the stem cell study carried out for the Canadian Biotechnology Secretariat (Campbell et al., 2004) we find the U.S. to be the leading nation in stem cell patenting (74%), followed far behind by Canada (4,51%), Japan (4,07%), Germany (3,19%) and France (2,75%), all of these countries accounting for an absolute number of 20 or more patents. Among the countries accounting for less than 20 patents we find Italy, Switzerland, The Netherlands, Australia and Belgium that accounts for only two patents in the period under review. Table 2 - International comparison of the number of stem cell patents granted by the USPTO (Based on inventor country and/or assignee country) Rank

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 15 15 15 15

Country

United States Canada Japan Germany France Great Britain Israel Italy Switzerland Netherlands Australia Russia India Austria Belgium China Korea Norway Sweden

1994-2003 Patents 672 41 37 29 25 19 18 16 10 9 6 5 4 3 2 2 2 2 2

Share 73,93% 4,51% 4,07% 3,19% 2,75% 2,09% 1,98% 1,76% 1,10% 0,99% 0,66% 0,55% 0,44% 0,33% 0,22% 0,22% 0,22% 0,22% 0,22%

1994-1998 Patents 469 24 23 24 16 17 13 11 5 6 5 3 1 1 2 1 0 1 1

Share 74,92% 3,83% 3,67% 3,83% 2,56% 2,72% 2,08% 1,76% 0,80% 0,96% 0,80% 0,48% 0,16% 0,16% 0,32% 0,16% 0,00% 0,16% 0,16%

1999-2003 Patents 203 17 14 5 9 2 5 5 5 3 1 2 3 2 0 1 2 1 1

Share 71,73% 6,01% 4,95% 1,77% 3,18% 0,71% 1,77% 1,77% 1,77% 1,06% 0,35% 0,71% 1,06% 0,71% 0,00% 0,35% 0,71% 0,35% 0,35%

Based on the EPO-data a slightly different picture occurs. Here also we find the U.S. to be dominant in terms of patents, but to a lesser degree than at the USPTO patent system (53% EPO instead of 74% USPTO). The second country in the ranking is Japan (8,6%) followed by Germany (6,33%), France (4,41%) and Great Britain (4,07%). Canadian patenting in stem cells is less prominent here (3,62%) than in the case of the USPTO patents (possibly due to proximity effects). The Netherlands account for 21 patents (or 2,38%) of all EPO stem cell patents.

39

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Table 3 - International comparison of the number of stem cell patent applications at the EPO (Based on inventor country and/or assignee country) Rank

1 2 3 4 5 6 7 8 9 10 10 11 11 12 12 13 13 14 15

Country

United States Japan Germany France Great Britain Canada Israel Netherlands Australia Italy Switzerland Belgium Sweden Austria Korea Singapore Russia Spain Netherlands Antilles

1994-2003 Patents 472 76 56 39 36 32 24 21 18 17 17 10 10 7 7 6 6 5 4

Share 53,39% 8,60% 6,33% 4,41% 4,07% 3,62% 2,71% 2,38% 2,04% 1,92% 1,92% 1,13% 1,13% 0,79% 0,79% 0,68% 0,68% 0,57% 0,45%

1994-1998 Patents 237 25 17 16 17 12 10 9 3 8 8 3 1 4 2 1 4 0 1

Share 61,88% 10,12% 6,88% 6,48% 6,88% 4,86% 4,05% 3,64% 1,21% 3,24% 2,83% 1,21% 0,40% 1,62% 0,81% 0,40% 1,62% 0,00% 0,40%

1999-2003 Patents 235 51 39 23 19 20 14 12 15 9 9 7 9 3 5 5 2 5 3

Share 46,71% 9,98% 7,58% 4,59% 3,79% 3,99% 2,79% 2,40% 2,99% 1,80% 1,80% 1,40% 1,80% 0,60% 1,00% 1,00% 0,40% 1,00% 0,60%

Belgium accounts for 10 patents here (or 1,13%) and as a result is 11th in the ranking of most patenting countries in stem cell research and technology. Table 4 shows several characteristics of the Belgian/Flemish patent portfolio in stem cells.

40

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Table 4 - Characteristics of the Belgian/Flemish patent portfolio in stem cells Patent(s)

Assignee(s)National level

EP 1109923 EP 1257633

BE (Flemish) BE (Flemish)

EP 1270732

BE (Flemish)

EP 1305401 EP 0696639

BE (Flemish) DE

EP 08426682

LU

EP 1307205 EP 1336658

GB FR

EP 1320622 EP 0907722

BE (Flemish) BE (Flemish)

US 6103523 US 5807744

BE (Flemish) DE

Assignee names Aventis CropScience N.V. Vlaams Interuniversitair Instituut voor Biotechnologie vzw. N.V. Antwerpes Innovatiecentrum Thromb-X N.V. Roche Diagnostics GmbH

Inventor(s) -

Prof. Berneman, Z., Dr. V. Van Tendeloo, Dr. H.W. Snoeck Dezeure, F., and Vanstraelen, D. Wulfert, E.A. Schoonjans, R.

European Atomic Energy Community (Euratom) BTG Inernational Ltd. Centre Nationale de la recherché scientifique DAE Galapagos Genomics N.V. Leuven Research and Development vzw. Thromb-X N.V. Boehringen Mannheim Berneman Z., Van GmbH Bockstaele D., Snoeck H.W.

In total (EPO and USPTO patents) Belgium (Flanders) accounts for 12 patents; 7 patents are assigned directly to Flemish organisations and 5 are the result of the contribution of mainly Flemish inventors, but are nevertheless assigned to foreign organisations. For these patents the table shows the inventors based upon which these patents have been included in the Belgian/Flemish subset.

Technological specialisation To what extent are there, even in the still developing phase of stem cell research and technology, signs of technological specialisation of countries? There are several ways to track technological specialisation of countries in technological areas. The so-called ‘revealed technology advantage’ is the most popular measure of technological specialisation. The revealed technological advantage (RTA) is defined as the share of patents in class r by entity j divided by the share of patents in class r by all entities considered in the benchmark. In other words, the RTA-index compares the share of a particular country’s patents for a particular technological sub-domain with the share of other countries in the same domain. As such, it is a relative indicator of technological specialisation (strength). If country X has a share that appears to be bigger than that of other countries, we can say that country X has a ‘revealed technological advantage’ for that specific technological domain, stem cell research and technology in the present case. As a result, the RTA was applied to the country level, the share of a country in stem cell patents divided by the country’s share in all patents. However, in view of the sensitivity of the indicator for a low number of patents (as in the case of stem cells) one should interprete the outcome of the RTA-analysis with great care.

Although the patent is assigned to the Euroepan Atomic Energy Community it is related to stem cells. The title of the patent is “Ex-corpore method for treating human blood cells”. One of the claims of the invention states: “stem cells extracted from said blood and/or bone marrow, extracted from a patient for autologous transplant, are mixed ex-corpore with said conjugated radionuclides”.

2

41

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

We have calculated the measure for both databases (see figures 4 and 5). See annex 1 for an explanation of the used country abbreviations. Figure 4 - Revealed technological advantage in stem cells: a country comparison based on USPTO data

AU 10,00 US

BE

SG

CA 1,00

SE NL

CH CN

0,10

KR

DE

JP

FR IT

IN

IL

GB

The value of the RTA-index varies from 0 to + ∞. A value lower than 1 reflects that country i has a relative disadvantage in category j. A value of 1 corresponds to a neutral position, whereas a value exceeding 1 signifies a relative advantage. Based on the USPTO stem cell data we find Canada (CA) together with India (IN) and Israel (IL) to have a strong position in patented stem cell research and technology, the RTA points towards a relative advantage in this area. As far as Belgium is concerned, we find an RTA index below 1 suggesting a relative disadvantage in stem cells; Belgium accounts for only 2 patents (see also table 2) in the period under consideration and clearly has to catch up the coming years. As we will see further on in this report, one of these patents is assigned to a Flemish company (ThrombX N.V. in Leuven), whereas the second patent is assigned to Boehringer Mannheim GmbH in Germany, but developed in collaboration with 3 Flemish inventors. A relative disadvantage is also found for countries like Germany (although having a high absolute number of stem cell patents), Switzerland, Japan, Korea and Sweden. However, the interpretation of these findings should be done with care. The RTA-index3 relates the patenting output of country in a certain field, stem cells in this case, to the overall output of that country in all fields together. Because of the fact that the total patent portfolio of India (IN) amounts 95 in the period under consideration, and out of these 95 there are 4 patents that are stem cell related (1 due to IN inventor involvement, 3 due to IN assignee involvement), the RTA index for India suggests a certain degree of specialisation of India in stem cell research and technology (see also the results presented by the Canadian stem cell study Campbell et al., 2004).

3

A brief methodological note: the RTA values were calculated using the number of patents per country where the assignee is of that respective country. When the number of patents for a country is low (threshold: 2 or less patents), these countries were not included in the analysis of the relative technological advantage.

42

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

One has to be aware that, when making international comparisons with respect to patenting activity of various actors and/or countries, such comparisons are hampered by international differences in the legal conditions surrounding the granting of a patent. In practice, this results in U.S. companies having a comparable advantage with respect to patent grants in the USPTO system relative to foreign companies, because of the acquaintance with and the thorough understanding of the patenting system on the one hand, but also the fundamental differences in patenting procedures on the other hand (think of the obligation in the U.S. to mention any prior art know to the inventor, the “duty of disclosure”, v.s. in Europe where no such obligation exists). In the EPO system, however, such distortions seem to influence the granting of a patent to a lesser extent. This is mainly due to the fact that the European Patent Office is a supranational agency. European applicants therefore will have little, if any "home advantage" in the EPO system, when compared to the U.S. situation (think only of the complexity and the costs related to translating the patent in the different languages). In any case, the same analysis should be performed but now based on EPO-data (see figure 5). Figure 5 - Revealed technological advantage in Stem cells: a country comparison based on EPO data

US

AN 10

AT

SG

AU

SE

BE

1

NL

CA 0,1

KR

CH

JP

CN IT

DE IL

FR GB

Based on the EPO stem cell patent data, we see that Canada (CA), Australia (AU), the United States (US), Israel (IL), China (CN), and Singapore (SG) have RTA-indices above 1 indicating a certain degree of specialization in stem cell research and technology. A rather high RTA-index is found for the Netherlands Antilles (AN), which is the result of the activity of Applied Research Systems ARS Holding N.V., a Curacao based company that is a wholly owned subsidiary of Serono International, S.A. Serono is a Swiss-based pharmaceutical company listed on the NYSE. Looking specifically at Belgium in the case of EPO patents, we find a rather neutral position (RTA around 1) in regard of stem cells. Belgium accounts for 10 patents in stem cell related research and technology, 6 of these patents are assigned to Belgian organisations, whereas 4 of these patents have been developed with the involvement of Belgian inventors, perhaps working for foreign organisations. All 6 patents, as will be discussed in more detail in the section about the organisations active in stem cell patenting, are in fact patents assigned to Flemish organisations (Aventis CropScience N.V., Vlaams Interuniversitair Instituut voor Biotechnologie vzw., N.V. Antwerps Innovatiecentrum, Thromb-X N.V., Galapagos Genomics N.V., Leuven Research and Development vzw.).

43

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Taking a closer look at the 4 patents that have been co-invented by Belgian inventors but assigned to foreign organisations (Roche Diagnostics GmbH, European Atomic Energy Community, BTG International Ltd., Centre National de la recherché scientifique DAE) reveals that out of the 7 Belgian inventors involved, there are 6 Flemish inventors and 1 from Brussels. Based on this, it can be stated that in Belgian stem cell research, we find mainly Flemish organisations and Flemish inventors, to be the driving force behind the still limited patenting activity. In the following section we shall focus on the (international) collaboration patterns in stem cell technology development.

International Collaboration The advantage of international and national collaboration for advancing science and technology in general is widely acknowledged in literature (see for instance the work of Hagedoorn, Schakenraad, Duysters, Carayol, Murray, Branscomb and others). International collaboration in stem cell research and technology seems to be multifaceted and complex in view of the different national regulations and ethical viewpoints (see also section IV on the impact of ethical issues and regulations on the course of science and technology development in stem cells). Most countries have adopted different regulations on the use of human stem cells in research. However, several common features can be identified (Campbell et al., 2004): “Human reproductive cloning is usually forbidden, creation of embryos exclusively for research purposes is usually prohibited, derivation of ES cell lines from excess IVF (in vitro fertilization) embryos is usually permitted, the utilization of adult stem cells for research purposes does not pose any ethical problem and is generally not addressed in the different regulations. The position on therapeutic cloning is more divided, with about 30 countries in favour and 60 countries against.” In this section we will look at the extend to which international collaboration in stem cell research and technology occurs, perhaps in many cases driven by ‘local’ legal and/or ethical limitations. There are two approaches to trace collaborative activity in patent documents (see Hicks and Narin, 2001 for a conceptual discussion on co-invention and co-assignee analysis): ƒ

Co-inventions: This measure is most closely related to co-authorship in publications. A coinvention links points to individuals who generated technology in a common endeavour.

ƒ

Co-assignations: This link connects actors that share the ownership of a patent. Contrary to coinventions, co-assignations op patents point to a shared interest in utilising a patented invention rather than co-operation in the creation of a technology. Co-assignations occur usually at the organisational level and not the individual level.

For the analysis of collaborative activity the Salton´s measure is used, which is defined as the number of joint inventions (assignations) of the two countries divided by the geometric mean of the total number of inventions (assignations) attributed to the two respective countries, that is, r=

pij pi . p j

where pij = the number of links between the countries i and j and pi (pj) the total of inventions (assignations) of the county i (j). The interpretation is rather straightforward, the higher the measure the more intense the collaboration interaction between two countries.

44

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Tables 5 and 6 present the findings of the co-inventor analysis based on respectively USPTO and EPO data. Collaboration on the inventor level is a way to capture a more direct form of potential knowledge transfer between individuals. As table 4 shows, the United States inventors are by far most frequently involved in collaborative settings with other countries. This can of course be explained when we look at the overall position of the U.S. in patented stem cell research and technology. If one can speak in terms of ‘partnering up’, we find the U.S. inventors to partner up most frequently with their Canadian colleagues, also a country with a strong position in stem cells, followed by collaboration with Australian, Italian, Japanese, Russian, German and French inventors. Other notable collaborative relations (cells are in grey) on the inventor level are the relations between Australian and Italian inventors, the relation between Australian and UK inventors, and finally, the relations between Japan and China and between France and Austria. The number of Belgian patents is too low to expect visible intense collaboration with other country’s inventors. Table 5 - International co-inventions based on USPTO data (for the application years 1993-2004), values given in Salton´s measure (all collaborations) US US

CA

AU

JP

6,15% 7,54%

AU

7,78% 6,15%

IT

6,67% 7,54% 10,21%

JP

8,03%

DE

6,72% 2,96%

FR

6,22% 9,23%

CH

4,04%

RU

8,52%

GB 4,49%

FR

CH

RU

GB

IL

AT

1,75%

BE

CN

2,69%

EE

IN

KR

NL

10,21%

9,62% 4,90%

5,74% 11,47%

4,90%

6,93% 7,22%

4,50% 11,79%

6,93% 7,22% 9,62% 5,74%

4,50% 11,79%

2,69%

CN

11,47%

EE

3,81%

IN

1,90%

KR

2,69%

NL

1,44%

ES

5,39%

SE

5,39%

ES

SE

3,81% 1,90% 2,69% 1,44% 5,39% 5,39%

2,96% 9,23%

AT BE

DE

13,21% 7,78% 6,67% 8,03% 6,72% 6,22% 4,04% 8,52% 4,49% 1,75%

CA 13,21%

IL

IT

45

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Based on the EPO data (see table 6), which should give a better picture of the collaboration patterns of European actors, we see that here too US-inventors are almost always involved in collaborative settings with other stem cell active countries. Keeping in mind the small number of patents involved, we find Belgian inventors to collaborate mostly with German inventors, followed by UK, French and U.S. inventors. Other interesting collaborative relations are the ones between Germany and Switzerland, U.S. and Russian inventors, and Spanish and Swiss inventors (cells are in grey). Table 6 - International co-inventions based on EPO data (for the application years 1993-2004), values given in Salton´s measure (all collaborations) US US FR DE CA JP GB ES NL BE IL SE IT RU CH AU

FR

DE 3,62%

3,62% 4,57% 8,54% 5,67% 5,43% 6,07% 1,07% 3,20% 0,91% 5,73% 4,39% 12,94% 2,86% 6,40%

2,31% 2,75% 1,82% 5,49%

CA 4,57% 2,31%

JP 8,54% 2,75%

1,95% 4,95%

5,66% 10,21% 5,77% 4,56% 7,00%

GB ES NL BE IL SE IT RU CH AU 5,43% 6,07% 1,07% 3,20% 0,91% 5,73% 4,39% 12,94% 2,86% 6,40% 5,49% 5,66% 5,06% 4,95% 10,21% 5,77% 4,56% 7,00% 18,26% 1,95% 4,16% 8,08% 2,28% 8,08% 6,06% 5,42% 4,04% 14,14% 8,08% 6,06% 2,28% 9,43% 5,42% 5,67% 1,82%

4,16%

5,06% 18,26%

14,14% 8,08%

4,04%

9,43%

* The table presents the collaborative relations between the main actors active in stem cell research and technology; other countries involved in minor collaborations are: Singapore, Taiwan, Korea, Greece and others.

In tables 7 and 8 the results of a similar analysis (respectively based on USPTO and EPO data) are presented, but now instead of looking at co-inventor relations we look at co-assignee relations: the cases when a patent is assigned to more than 1 organisation. According to Hicks and Narin (2004), the rate of co-assignment varies across technologies. In chemicals, for example, almost no patents are coassigned (0.08%), while in biotechnology about 7% are co-assigned (based on USPTO data). The growth in co-assignment is also uneven. In some areas, co-assignment has taken off; in others, the rate has not changed in 20 years. In 1980, excluding biotechnology which always had a higher rate of coassignment, the maximum share of patents co-assigned in any one of CHI's 30 technology classification was 1.43% and the minimum was 0. By 1999, the maximum had risen to 7.1% while the minimum remained at 0%. To a large extent, this explains the reason for the low number of co-assignments found, even though stem cells is a part of the broader area of biotechnology and as such shows already a larger number of co-assignments.

46

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Table 7 - International co-assignations based on USPTO data (for the application years 1993-2004), values given in Salton´s measure (all collaborations) US US IT CH DE GB IL CA JP FR

IT

CH

DE

1,41% 1,41%

GB 0,94%

IL 2,21%

CA 2,90%

JP 2,91%

FR 3,01%

4,45%

17,68% 17,68%

0,94% 2,21% 2,90% 2,91% 3,01% 4,45%

Based on the USPTO data (table 7) we see that intense co-assignment relations exist between Italy and Switzerland (17,68%) - reflecting collaboration between neighbouring countries and economies. Another collaboration relation worth mentioning is the one between France and the U.S. (4,45%). Belgium is not included in the table because of the fact that Belgium has only 1 patent owned solely by a Belgian (Flemish) company. Table 8 - International co-assignations based on EPO data (for the application years 1993-2004), values given in Salton´s measure (Threshold: 4 collaborative patents) US US CH DE AU SG IL LU NL PL BE FR TW CA GB IT JP AT

CH DE AU SG IL LU NL 2,76% 1,31% 4,89% 3,24% 2,76% 4,31% 9,09% 1,31% 4,31% 9,09% 13,48% 6,43% 13,48% 9,53% 4,89% 6,43% 9,53% 3,24%

PL BE 4,58%

FR TW CA GB IT JP AT 1,70% 3,24% 1,67% 2,55% 1,45% 1,61% 73,85% 5,83% 9,04% 21,32%

9,37% 4,58% 5,83%

9,37%

1,70% 3,24% 1,67% 2,55% 1,45% 1,61%

21,32% 2,14% 9,04% 2,14% 73,85%

47

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

In table 8 we find the results of the collaboration analysis based on the EPO data. For Belgium (and thus for Flanders) the main collaborating nations are Germany (5,83%) and The Netherlands (9,37%). Other important transnational collaborative relations between organisations are be found between Switzerland and Australia, between Germany and Italy, between Australia and Singapore, and finally, between Singapore and Israel. It may not be surprising that to a large extent the collaboration patterns on the inventor level coincide with the collaboration pattern on the organisational level. In the next section we will look at the actors behind the described collaborative relations.

Organisational Analysis Tracing organisations that own the patents is an additional way of exploring a technological area. This way the analyst can get an idea whom the important industrial actors are in this field. An organisational analysis based on assignee/applicant addresses also allows us to identify the extent to which other actors, than solely companies, in the innovation system are active in technological development in this area. To this end, this section presents two types of analysis. First, stem cell patents will be examined by organisational type, i.e. whether a company or educational or government institution or an individual person owns them. In another step, assignee rankings will be introduced for both European stem cell patent data and U.S. stem cell patent data. The results of the first type of analysis are presented in tables 9 and 10. Table 9 - Breakdown by organisational type source EPO data 1994-2003

1994-1998

1999-2003

Type of Assignee Company University/University hospital/Higher Education Admin/Public Institute Hospital Person/Other/Unknown

Patents Share

Patents Share

Patents Share

465

51,72%

223

56,31%

242

48,11%

190 108 38 98

21,13% 12,01% 4,23% 10,90%

83 38 12 40

20,96% 9,60% 3,03% 10,10%

107 70 26 58

21,27% 13,92% 5,17% 11,53%

Table 10 - Breakdown by organisational type source USPTO data 1994-2003

1994-1998

1999-2003

Type of Assignee Company University/University hospital/Higher Education Admin/Public Institute Hospital Person/Other/Unknown

Patents Share

Patents Share

Patents Share

421

47,62%

294

48,04%

127

46,69%

233 108 49 73

26,36% 12,22% 5,54% 8,26%

162 82 28 46

26,47% 13,40% 4,58% 7,52%

71 26 21 27

26,10% 9,56% 7,72% 9,93%

The majority of stem cell patents, over the full period, are in hands of companies (based on EPO and USPTO respectively 52% and 48%). In second place we find an important role for universities and university hospitals, followed by government organisations like public research institutes. A significant number of patents is also owned by individuals or other type of actors.

48

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

A plausible explanation for the strong presence of companies in stem cell research and technology, despite the highly scientific nature of stem cell research, may be that the constraining policies regarding funding of stem cell S&T stimulate the transfer of research from public facilities to private facilities where no policies constrain stem cell research (Campbell et al., 2004). In the U.S., concern has been expressed about this transfer, specifically in relation to protection of human subjects, obtaining the right approvals etc. Following this, tables 11 and 12 present the top-10 patenting organisation in stem cells research and technology, respectively based on USPTO and EPO data. Heading the list with respectively 3% and 2,66% of all stem cell patents, we find Osiris Therapeutics, a U.S. based company. Osiris was founded in 1992 to commercialize stem cell products that are isolated from a readily available source – adult bone marrow. The technology is based on the pioneering work of Dr. Arnold Caplan and his colleagues at Case Western Reserve University. They showed that Mesenchymal Stem Cells (MSCs) can engraft and selectively differentiate, based on the tissue environment, to such lineages as muscle, bone, cartilage, marrow stroma, tendon and fat. Due to their cell-surface characteristics, MSCs are relatively non-immunogenic, allowing for the development of therapies that rely on the transplantation of MSCs derived from unrelated human donors (see http://www.osiristx.com). Table 11 - Ranking of the Top-10 organisations in stem cell patenting (application year 1994-2003; source: USPTO data) Rank 1 2 3 4 4 5 5 5 6 6 6 7 7 8 8 8 8 9 9 9 9 9 10 10 10 10 10 10 10 10

Organisation

Organisational type

OSIRIS THERAPEUTICS, INC. Company UNIVERSITY OF MICHIGAN University UNIVERSITY OF CALIFORNIA University THE UNITED STATES OF AMERICA Administration SYSTEMIX, INC. Company AMGEN INC. Company SLOAN-KETTERING INSTITUTE FOR CANCER RESEARCH Hospital UNIVERSITY OF MINNESOTA University CALIFORNIA INSTITUTE OF TECHNOLOGY University HUMAN GENOME SCIENCES, INC. Company ZYMOGENETICS, INC. Company CELL GENESYS, INC. Company CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE Public Research Institute GENENTECH, INC. Company GENETICS INSTITUTE, INC. Company NEUROSPHERES HOLDINGS LTD. Company THE GENERAL HOSPITAL CORPORATION Hospital IMMUNEX CORPORATION Company INSTITUT NATIONAL DE LA SANTE ET DE LA RECHERCHE MEDICALE Public Research Institute (INSERM) MICHIGAN STATE UNIVERSITY University UNIVERSITY OF WASHINGTON University YEDA RESEARCH AND DEVELOPMENT CO. LTD. University CASE WESTERN RESERVE UNIVERSITY University HADASIT MEDICAL RESEARCH SERVICES AND DEVELOPMENT LTD. Company JOHNS HOPKINS UNIVERSITY University NEXELL THERAPEUTICS INC. Company STEMCELL TECHNOLOGIES INC. Company UNIVERSITY OF PENNSYLVANIA University UNIVERSITY OF PITTSBURGH University UNIVERSITY OF TEXAS SYSTEM University

Patent Share 3,00% 2,58% 2,15% 2,04% 2,04% 1,50% 1,50% 1,50% 1,29% 1,29% 1,29% 1,07% 1,07% 0,97% 0,97% 0,97% 0,97% 0,86% 0,86% 0,86% 0,86% 0,86% 0,75% 0,75% 0,75% 0,75% 0,75% 0,75% 0,75% 0,75%

49

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Based on the USPTO data, Osiris is directly followed by a university, the University of Michigan (2,58% of all stem cell patents); based on the EPO data one has to go down to the fourth position to find the first university accounting for a substantial proportion of European stem cell patents, the University of California. In general, universities account for approximately 20-25% of all stem cell patents, shares that stay rather constant over time (see tables 9 and 10). As already mentioned, the only patent held by a Flemish (and Belgian) organisation in the USPTO, is held by Thromb-X. Thromb-X operates alongside but separate from the Centre for Molecular and Vascular Biology (CMVB) and the Centre for Transgene technology and Gene Therapy (CTG) of the Flanders Interuniversity Institute for Biotechnology (VIB). Thromb-X is originally a spin-off from the University of Leuven. Table 12a - Ranking of the Top-10 organisations in stem cell patenting (application year 1994-2003; source: EPO data) Rank 1 2 3 4 4 4 5 6 6 6 6 6 7 7 7 8 8 8 8 8 8 9 9 9 9 10 10 10 10 10 10 10 10 10 10

Organisation OSIRIS THERAPEUTICS, INC. ZYMOGENETICS, INC. THE GENERAL HOSPITAL CORPORATION CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE (CNRS) UNIVERSITY OF CALIFORNIA THE GOVERNMENT OF THE UNITED STATES OF AMERICA SYSTEMIX, INC. HADASIT MEDICAL RESEARCH SERVICES & DEVELOPMENT COMPANY, LTD UNIVERSITY OF MASSACHUSETTS INSTITUT NATIONAL DE LA SANTE ET DE LA RECHERCHE MEDICALE (INSERM) NEUROSPHERES HOLDINGS LTD. NOVARTIS AG BAXTER INTERNATIONAL INC. CASE WESTERN RESERVE UNIVERSITY GENENTECH, INC. AVENTIS CROPSCIENCE CHUGAI SEIYAKU KABUSHIKI KAISHA GERON CORPORATION UNIVERSITY OF EDINBURGH UNIVERSITY OF PITTSBURGH MONASH UNIVERSITY CHIRON CORPORATION HUMAN GENOME SCIENCES, INC. UNIVERSITY OF SOUTHERN CALIFORNIA UNIVERSITY OF MICHIGAN ECOLE NORMALE SUPERIEURE DE LYON IMCLONE SYSTEMS, INC. INSTITUT NATIONAL DE LA RECHERCHE AGRONOMIQUE (INRA) KIRIN BEER KABUSHIKI KAISHA ONO PHARMACEUTICAL CO., LTD. THE HEBREW UNIVERSITY OF JERUSALEM WISCONSIN ALUMNI RESEARCH FOUNDATION CELL GENESYS, INC. LELAND STANFORD JUNIOR UNIVERSITY YEDA RESEARCH AND DEVELOPMENT CO. LTD.

Organisational type

Patent Share

Company Company Hospital Public Research Institute University Public Research Institute Company Company

2,66% 1,70% 1,49% 1,38% 1,38% 1,38% 1,28% 1,06%

University Public Research Institute

1,06% 1,06%

Company Company Company University Company Company Company Company University University University Company Company University University University Company Public Research Institute Company Company University University Company University University

1,06% 1,06% 0,85% 0,85% 0,85% 0,74% 0,74% 0,74% 0,74% 0,74% 0,74% 0,64% 0,64% 0,64% 0,64% 0,53% 0,53% 0,53% 0,53% 0,53% 0,53% 0,53% 0,53% 0,53% 0,53%

Based on the EPO-patent data, we have constructed a sub-table with the Flemish (also being the Belgian) organisations active in stem cell patenting (see table 12b).

50

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

Table 12b - Ranking of the Top-patenting Flemish organisations in stem cell patenting (application year 1994-2003; source: EPO data) Rank 8 13 13 13 13 13

Organisation

Organisational type

Patent share

AVENTIS CROPSCIENCE N.V. VLAAMS INTERUNIVERSITAIR INSTITUUT VOOR BIOTECHNOLOGIE VZW. N.V. ANTWERPES INNOVATIECENTRUM THROMB-X N.V. GALAPAGOS GENOMICS N.V. LEUVEN RESEARCH AND DEVELOPMENT VZW.

Company Public Research Institute

0,74% (7 patents) 0,11% (1 patent)

University Company Company University

0,11% (1 patent) 0,11% (1 patent) 0,11% (1 patent) 0,11% (1 patent)

The most active Flemish company in stem cell patenting is Aventis (the Belgian branch), accounting for 7 EPO patents in total. The Flemish Interuniversity Institute for Biotechnology, just as the Antwerp Innovation Centre (University of Antwerp), Thromb-X, Galapagos Genomics, and Leuven Research and Development (University of Leuven) each account for 1 stem cell patent. In the following section we will briefly elaborate on a number of ethical issues potentially affecting the course of scientific and technological development in stem cell related research and technology.

51

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

IV. Conclusions The developed search strategy for identifying stem cell patents has proven to be highly effective in retrieving a representative and relevant set of patents. In total over the period 1994-2003, 780 USPTO and 736 EPO patents have been retrieved. Their relevance was moreover confirmed by an expert in the field of stem cell research and technology. Patenting in stem cell research and technology started already in the early 1980s. A strong increase is noted beginning of the 1990s followed by stabilization (USPTO) and even decline (EPO) at the beginning of the years 2000 and 2001. When looking at the technology areas stem cell patents are related to, we find biotechnology and pharmaceuticals to be the largest areas where stem cell patents relate to. As also indicated by the, to our knowledge, only study on stem cell technology commissioned by the Canadian Biotechnology Secretariat, our analysis also point out the U.S. as the most important nation in (developing) stem cell technology (measured by the number of patents). The U.S. by far accounts for the largest share in stem cell patents, followed by Japan, Germany, the UK and France. Belgian and Flemish patenting activity in stem cells is limited and low with 12 patents (10 in the EPO and 2 in the USPTO). Flanders accounts for almost all of these patents when looking more in detail at the involved organisations/inventors. A number of countries is strongly specializing in stem cell research and technology. Among them is Canada, the U.S., China, the UK, Israel, Singapore and India. The Belgian/Flemish focus on stem cell research and technology can be typified as rather ‘neutral’. The investigation of the collaborative relations between countries (based on co-inventor and coassignee analysis) reveals that the U.S. are most frequently collaborating with other nations when developing stem cell technology. Furthermore, we find regional elements, and perhaps preferences, in the collaboration relations, e.g. intensive collaboration between Italy and Switzerland, between China and Japan, and also between France and Austria. Belgian/Flemish patents reflect collaboration with the U.S., Germany, France and the U.K. Patenting in stem cells is dominated by companies that is, as suggested, a possible result of the transfer of research to private settings in order to surpass legal barriers. Companies are followed by universities and public research facilities. Both analysed data sources point towards Osiris Therapeutics as the largest holder of stem cell patents (ranging between 2,7% and 3,0%). Based on the USPTO data we find the universities of Michigan and California to hold large patent shares. Even within the European patent system we find the university of California to have a strong position (1,38%). Belgian patenting is almost entirely attributable to Flemish organisations and inventors. Among the Flemish organisations we find Aventis Cropscience to be the largest patent holder in stem cells (7 patents), followed by the Flemish Interuniversity Institute for Biotechnology, the Antwerp Innovation Centre (University of Antwerp), Thromb-X, Galapagos Genomics, and Leuven Research and Development (University of Leuven), which account for 1 stem cell patent each.

52

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

REFERENCES D. CAMPBELL, M. NOISEUX, G.CÔTÉ, Potential for Stem Cells Science and Technology in Canada: Great Promises and Challenges, Science-Metrix report, April, 2004, pp. 63. HICKS, D. NARIN F., Strategic Research Alliances and 360 Degree Bibliometric Indicators, CHI Research Inc., Strategic research partnerships: proceedings from an NSF workshop, 2004

53

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

APPENDIX - Definition of country abbreviations

Abbreviation AU AT AN BE CA CH CN DE EE ES FR GB IL IN IT JP KR NL RU SE SG US

Country Australia Austria Netherlands Antilles Belgium Canada Switzerland China Germany Estonia Spain France Great Britain Israel India Italy Japan Korea The Netherlands Russian Federation Sweden Singapore United States

54

Analysis of “Stem-Cell” research and technology Steunpunt O&O Statistieken

PART III OVERALL CONCLUSIONS Embryonic or somatic stem cells are seen as promising therapeutic tools for the treatment of number of several severe human diseases such as leukemia, diabetes, Parkinson disease, multiple sclerosis and other degenerative diseases. Embryonic stem (ES) cells have been isolated from the mouse more than twenty years ago, and it is only during the last five years that human ES cells have successfully been isolated and propagated in a very limited number of laboratories mostly in United States, Australia, Israel and Sweden. Somatic stem cells became also highly promising reagents in the past few years when a number of data suggest their potential for efficient differentiation into various cell types. In the hematopoietic system, somatic stem cells (hematopoietic stem cells) have been used for transplantation therapy for a long time. There is also number of studies that indicate that cancer takes place in somatic stem cells. This is particularly true in tissues with high level turnover such as skin, intestine, blood and human breast gland. Striking parallels can be found between stem cells and cancer cells and similar mechanisms may regulate self-renewal in those two cell types. Because of the expected demand for stem cells for human medical applications, there is a real need for supporting research aimed at developing human stem cell lines and their applications. This aim requires that we rapidly increase our knowledge of the basic features and properties of stem cells either from embryonic or somatic origin, human as well as from animal models. In this report, we provide an overview of stem cell research using both bibliometric and technometric indicators. As far as bibliometric indicators are concerned, we can conclude that those countries that are most active in scientific research in all fields combined have top activity in stem-cell research, too. The contribution of the USA amounts to almost 50% of all stem cell related papers in the world. The corresponding value in all science fields combined is about one third. The share of the USA in stem cell research, however, decreased by about 10% if one compares the two sub-periods. The shares of Japan and Germany, in turn, grew considerably. The low share of Russia – otherwise holding rank seven in all fields combined – is worth mentioning. The share of Flanders reflects an “average” growth and the share of Flemish stem cell research papers in the Belgian national output amounts to 60%. Patenting in stem cell research and technology started already in the early 1980s. A strong increase is noted beginning of the 1990s followed by stabilisation (USPTO) and even decline (EPO) at the beginning of the years 2000 and 2001. When looking at the technology areas stem cell patents are related to, we find biotechnology and pharmaceuticals to be the largest areas where stem cell patents relate to. As also indicated by the, to our knowledge, only study on stem cell technology commissioned by the Canadian Biotechnology Secretariat, our analysis also point out the U.S. as the most important nation in (developing) stem cell technology (measured by the number of patents). The U.S. by far accounts for the largest share in stem cell patents, followed by Japan, Germany, the UK and France. Belgian and Flemish patenting activity in stem cells is limited and low with 12 patents (10 in the EPO and 2 in the USPTO). Flanders accounts for almost all of these patents when looking more in detail at the involved organisations/inventors. A number of countries is strongly specialising in stem cell research and technology. Among them is Canada, the U.S., China, the UK, Israel, Singapore and India. The Belgian/Flemish focus on stem cell research and technology can be typified as rather ‘neutral’.

55