WO2009125219A2 - Markers of oocyte viability - Google Patents

Abstract

There is disclosed a method of determining the potential of individual oocyte to develop into a viable embryo, which method comprises establishing the concentration of any one or more of the markers identified in Table 2 in the follicular fluid of a single follicle from which said oocyte is derived and comparing the concentrations obtained to a reference concentration range that is established by comparing the concentration of said one or more markers from one or more follicles containing a fertilisation competent oocyte(s), wherein differential levels of said markers from said follicle to be tested compared to the reference range at least at the 5% significance level is predictive of a reduced potential of said oocyte to develop into a viable embryo following fertilisation.

Description

Markers of Oocyte Viability

The present invention is concerned with identifying markers that are predictive of oocyte viability and their use in methods of predicting the outcome of assisted reproduction techniques.

Background

The progressive increase in demand for assisted conception by the 1 in 6 infertile couples in the UK has not been matched by improvements in pregnancy rates following fertility treatment and hence multiple embryos are regularly transferred to maximise a couple's chances of achieving a pregnancy. Thus pregnancies are frequently multi-fetal, prompting the Human Fertilisation and Embryology Authority (HFEA) to limit the number of embryos transferred to 2 in women aged under 40, and to 3 for older women in 2003. Even then approximately 1 in 4 of all IVF/ICSI pregnancies are twins, an incidence which is 20-fold higher than that after a spontaneous pregnancy (1 in 80). Consequently, the proportion of twins born following assisted conception is significantly elevated, accounting for 40% of all IVF babies. Overall numbers have also increased: in 1995, around 2,600 multiple pregnancy IVF babies were born and this figure had risen by >41% to more than 3,700 by 2003 (HFEA, www.hfea.gov.uk). In addition to the obvious financial, physical and emotional burden of multi- fetal gestations, twin or higher order multiple pregnancies feature an increased likelihood of pregnancy/perinatal complications (Table 1).

Table 1: Infant/maternal risks of multiple pregnancies; *compared to singleton births, **with preterm delivery, ***with delivery of low birth weight baby (HFEA, www.bt^'ea.gov.uk; Sattar, N., et al., 2003).

Several groups have published significant improvements in pregnancy rates with Elective Single Embryo Transfer (eSET) by prolonging in-vitro culture of embryos and transferring on day 3 or day 5-6 after oocyte retrieval. European Society of Human Reproduction and Embryology (ESHRE) working groups have recommended the wider use of Elective Single Embryo Transfer (eSET) cycles to reduce the neonatal mortality and morbidity associated with prematurity and low birth weight in multi- fetal pregnancies after fertility treatment (ESHRE Task Force on Ethics & Law, 2003; Land, J. A., et al., 2003). Indeed, the HFEA has been convinced by the argument made by the Independent Expert Group report that most of these risks can be avoided if only one embryo is transferred to those patients who have the highest chance of conceiving (and therefore the highest risk of twinning) (HFEA, www.hfea.gov.uk). However, the Scandinavian experience is difficult to replicate in the UK because of significant differences in the characteristics of the treatment population and state sponsored health service provision. In contrast to those countries, approximately 70% of treatment cycles are self- funded in the UK and where there is NHS funding, there are wide geographical variations (the post-code lottery) with variable eligibility criteria posed by many Health authorities and Primary Care Trusts (PCT) for NHS funded treatment. As the cost of self-funded treatment is substantial, sub-fertile couples are reluctant to accept eSET because of the fear of the lowering of their chances of success. Instead they readily accept the risk of a multiple pregnancy in favour of a higher pregnancy rate. The HFEA publishes annual results for all centres that are commonly seen as 'league tables' by the public and media alike. This makes the centres also less eager to adopt eSET in case the reduction in their success rates lowers their standing and popularity. Hence, the difficulty lies with finding a balanced way of making eSET work to reduce the risk of twining without taking the risk of compromising the success rate which would be at the patient's expense.

At present, embryo viability is assessed simply by assessing the rate of cell division and gross morphology of the blastomeres on the day of embryo transfer (Terriou, P. et al., 2007). Although this approach allows embryo logists/clinicians to remove embryos that are of overtly poor quality (e.g. fragmented, delayed, cleaving abnormally, arrested), its principal limitation lies in failing to identify the most "viable embryo/s" from a morphologically and developmentally comparable cohort. Therefore, there has always been a need to develop novel non-invasive diagnostic approaches to determine which embryos within such identical cohorts are the most viable in order to improve pregnancy rates. In turn, this would also eliminate the need for multiple embryo transfer and consequently reduce the number of ensuing multiple pregnancies and their associated obstetric and paediatric complications.

There is extensive evidence suggesting that embryo viability is very heavily influenced by oocyte quality or viability (Wilding, M. et a\. 2007). Many parameters influencing oocyte viability are determined during oogenesis and, more specifically, during the final stages of development/maturation in the ovarian follicle. The follicular fluid surrounding the oocyte contains a variety of autocrine and paracrine factors responsible for the regulation of oocyte development, folliculogenesis and ovulation. While many such agents are largely unidentified, it is accepted that they are produced locally within each follicle in response to stimulation by the gonadotrophins and the steroidal milieu (Schams, D. et al., 2001). Although these factors may putatively affect the oocyte directly, they may also do so indirectly, through the induction of effects on cumulus and mural granulosa cells and consequently may be key determinants of oocyte viability, both pre- and post-fertilization (Gilchrist, R.B. et al., 2004). Some of these factors are understood to include growth factors/cytokines, which, due to their secretory nature and paracrine activity, are detectable within the follicular fluid of antral/preovulatory follicles, and are believed to have a bearing on cycle outcome (Bedaiwy, M. et al., 2007).

Cytokines are an extensive array of pleiotropic glycoproteins involved in mediating a wide range of physiological responses, including immunity, inflammation, and haematopoiesis. Although their 'traditional' role is commonly perceived in relation to their immunoregulatory properties, cytokines also have a range of mitogenic and proapoptotic functions on non-immune cell targets based on tightly-regulated changes in their respective ratios (Chaouat et al., 2002; Kawamura, K. et al., 2007) . They operate as parts of highly complex integrated networks that exhibit multiple stimulatory/antagonistic interactions, synergism and functional redundancy. Although cytokines are ideal biomarkers investigated in a number of disorders, the complexity of their network regulation makes their study in isolation (a feature of most studies) largely meaningless. In view of their biological properties, cytokines should be investigated in relation to each other in order to attain a meaningful view of their effects and merit as biomakers. These include pleiotropism, where each cytokine has multiple target cells in various organs, and responses may differ according to cell type. Cytokines also synergise to potentiate/modulate each other's actions so as to induce a specific effect (Weiser, S. et al., 2007). They are also capable of mutual antagonism, wherein different cytokines have opposing actions (Numerof, R.P. et al., 2006). Finally, cytokine networks exhibit a marked degree of functional redundancy: different cytokines may act on a cell type individually to induce the same response, often through their use of a common receptor complex (Andrews, A.L. et al., 2006). Cells are also typically differentially sensitive to both exposure time and the absolute concentration of specific cytokines. For example, tumour necrosis factor (TNF)-α can trigger apoptotic pathways while paradoxically also being able to induce opposing bioregulatory effects, such as mitogenesis and differentiation depending on concentration (Baud, V. et al., 2001). The correct operation of cytokine networks is essential for normal physiology, a central role underlined by the fact that inflammatory/immune deregulations underpin many pathological reproductive conditions, which typically exhibit changes in both local and systemic cytokine profiles (Hagberg, H. et al., 2005; Christiansen, O.B. et al., 2006; Agic, A. et al., 2006).

The role of cytokines in infertility has already been the focus of investigation for a number of years, given that they are associated with the majority of reproductive disorders linked to conception and early pregnancy, including pelvic inflammatory disease (Guerra-

Infante et al., 1999), endometriosis (Bersinger et al., 2006), polycystic ovary syndrome

(Wu et al., 2006) and recurrent miscarriage (Wilczynski, 2006). As such, studies have focussed on the role of cytokines in specific patients rather than their individual follicles, and commonly operated on pooled follicular fluid samples from one or both ovaries

(Gutman et al., 2004). While this helps the understanding of the inflammatory pathophysiology associated with specific disease states and the overall fertility of individual patients, it provides little information regarding individual follicle/s or of the viability of individual oocytes retrieved from these patients. It is known that follicles within each ovary, after ovarian stimulation with gonadotrophins, generate multiple cohorts of oocytes all which act as functionally independent entities and vary in their developmental potential within the same treatment cycle (REF). Identifying markers within these individual follicles could make it possible to determine which factors are likely to be important in oocyte development, maturation, fertilisation and viability of their resulting embryos following fertilisation. Identification of embryos most likely to implant and become successful pregnancies will improve success rates such that it would becomes easier to embrace eSET and eliminate the need to transfer multiple embryos. This act alone will reduce the risk of multiple pregnancies and thus remove a cause of serious neonatal and maternal morbidity as well as mortality.

Currently there are no reliable non- invasive approaches that do not require changes to clinical practice in assisted conception. The only attempt to date has been disclosed in

WO 0153518 which bases its embryo viability assessment on assessing individual embryo metabolism as a predictive marker of viability. However, in order to achieve this, the embryos need to undergo further manipulations which are not part of routine clinical practice and which may potentially reduce their viability due to additional changes in gas tension, temperature and pH. In addition, pre-implantation embryo amino acid metabolism is acutely sensitive to changes in amino acid, carbohydrate and protein profiles in the culture medium (Orsi & Leese, 2004a,b), and meaning that whatever markers are chosen can only be applied using one type of culture medium. Given the frequent changes in medium usage following improvements in composition, this approach would have to be periodically reviewed. Furthermore, it would require standardisation on the number of cells of each embryo, which will vary between the start and the end of culture (cell number around the time of transfer is critical because of major embryonic changes allied to genome activation).

Summary of the Invention

Accordingly, it would be particularly beneficial to identify improved methods that could provide an indication of whether an oocyte retrieved from any given follicle would result in fertilisation and develop a 'viable embryo ' which would offer the greatest chance of proceeding to a successful implantation, pregnancy and normal birth. The present inventors have now advantageously identified a number of proteins in the follicular fluid of individual follicles that can function as markers in predicting the likelihood of the resulting fertilised embryo successfully implanting into the wall of the uterus and hence the outcome of assisted reproductive therapy, such as in vitro fertilisation,.

Therefore, according to a first aspect of the present invention, there is provided a method of determining the potential of an individual oocyte to develop into a viable embryo, which method comprises establishing the concentration of any one or more of the markers identified in Table 2 in the follicular fluid of a single individual follicle from which said oocyte is derived and comparing the concentrations obtained to a reference concentration range that is established by determining the concentration of said one or more markers from one or more individual follicles containing a fertilisation competent oocyte(s), wherein differential levels of said markers from said follicular fluid to be tested compared to the reference range, for example at least at the 5% significance level, is predictive of a reduced potential of said oocyte to develop into a viable embryo following fertilisation.

According to a further aspect of the present invention, there is provided a method of determining the potential of an individual oocyte to develop into a viable embryo, which method comprises establishing the concentration of any one or more of the cytokine markers identified in Table 2 in the follicular fluid of a single individual follicle from which said oocyte is derived and comparing the concentrations obtained to a reference concentration range that is established by determining the concentration of said one or more markers from one or more follicles containing a fertilisation incompetent oocyte(s), wherein differential levels of said markers from said follicular fluid to be tested compared to the reference range, for example at least at the 5% significance level, is predictive of an increased potential of said oocyte to develop into a viable embryo following fertilisation.

Also provided by the present invention is a method of evaluating the fertilisation competence or fertilisability of an oocyte derived from a single follicle in female subject, which method comprises establishing the concentration of any one or more of the cytokine markers identified in Table 2 in the follicular fluid of a single follicle from which said oocyte is derived and comparing the concentrations obtained to a reference concentration range that is established by determining the concentration of said one or more markers from one or more fertilisation competent oocytes from one or more females, wherein differential levels of said markers from said follicle to be tested compared to the reference range is indicative of the fertilisation incompetence or reduced fertilisability of said oocyte.

According to a further aspect of the present invention, there is provided a method of determining the potential of an individual oocyte to develop into a viable embryo, which method comprises establishing the concentration of any one or more of the cytokine markers identified in Table 2 in the follicular fluid of a single follicle from which said oocyte is derived and comparing the concentrations obtained to a reference concentration range that is established by determining the concentration of said one or more markers from one or more follicles containing a fertilisation incompetent oocyte(s) and a fertilisation competent oocyte(s), and wherein the proximity of the concentration ranges obtained from said oocyte to be tested to said marker levels from either said follicles containing either said fertilisation incompetent or competent oocytes is predictive of the potential of said oocyte to develop into a viable embryo.

Cytokine

IL-lα Inter leukin - lα

IL-lβ Interleukin - 1 β

IL-lra Interleukin - 1 receptor antagonist

IL-2ra Interleukin - 2 receptor antagonist

IL-2 Interleukin - 2

IL-3 Interleukin - 3

IL-4 Interleukin - 4

IL-5 Interleukin - 5

IL-6 Interleukin - 6

IL-7 Interleukin - 7

IL-8 Interleukin - 8

IL-9 Interleukin - 9 IL-IO Inter leukin - 10

IL12 (p40) Interleukin - 12(p40)

IL-12(p70) Interleukin - 12(p70)

IL-13 Interleukin - 13

IL-15 Interleukin - 15

IL-16 Interleukin - 16

IL-17 Interleukin - 17

IL-18 Interleukin - 18

Eotaxin

FGF Basic Fibroblast growth factor

G-CSF Granulocyte- colony stimulating factor

Granulocyte macrophage- colony

GM-CSF stimulating factor

IFN-α 2 Interferon - α2

IFN-γ Interferon - γ

IP-IO IFN-γ inducible protein- 10

MCP-I Macrophage chemotactic protein- 1

MIP-Ia Macrophage inflammatory protein -lα

MIP-lβ Macrophage inflammatory protein -lβ

PDGF Platelet derived growth factor

Regulated upon activation normal T-

RANTES cell expressed and secreted

TNF-α Tumour necrosis factor

VEGF Vascular endothelial growth factor

CTACK Cutaneous T-cell attracting chemokine

GROα Growth regulated ongogene-α

LIF Leukeamia inhibitory factor

MCP3 Monocyte chemoattractant protein-3

Macrophage- colony stimulating

M-CSF factor

Macrophage migration inhibitory

MIF factor

MIG Monokine induced by IFN-γ b-NGF basic-Nerve growth factor

SCF Stem cell factor

SCGF-β Stem cell growth factor- β

SDF-lα Stromal cell derived factor- lα

TNF-β Tumour necrosis factor-beta

Tumour necrosis factor related

TRAIL apoptosis inducing ligand CRP C-reactive protein

TGF-β

Table 2. List of markers tested in follicular fluid

In accordance with the present invention fertilisation competence in the context of an oocyte is defined as the measure of fertilisability of an oocyte from a female. It may also be referred to as oocyte viability. A fertilisation competent oocyte is defined as an oocyte that is capable of being fertilised naturally or by virtue of assisted reproductive therapy including intracytoplasmic sperm injection (ICSI). A fertilisation incompetent oocyte is defined as an oocyte that is not capable of being fertilised naturally or by virtue of assisted reproductive therapy including intracytoplasmic sperm injection (ICSI).

Some fertilised oocytes may develop into a viable embryo and successfully implant into the uterus wall. In accordance with the present invention, reference to embryo viability is defined as the potential of an embryo, for example, that was generated by assisted reproduction techniques, such as IVF/ICSI or the like, to develop and successfully implant into the lining of the uterus and to be maintained successfully for a period of at least six weeks. In comparing those cytokines/marker levels that are present in the follicular fluid of oocytes that successfully fertilised and led to viable embryos compared to those that fertilised and did not lead to viable embryos, it is also possible to determine the optimum cytokine concentrations that provide for both successful fertilisation and embryo viability in order to maximise the chances of a successful pregnancy.

Therefore, the methods of the invention may advantageously be utilised to determine the most viable embryos to be transplanted into the lining of the uterus of a female undergoing assisted reproduction therapy by establishing, using the methods defined herein, the concentrations of the markers identified in table 2 in the follicular fluid in the follicle from which the oocyte that developed into the embryo originally derived.

Advantageously, the method of the invention may be particularly useful to establish markers that are targets for therapeutic intervention. Therefore, also provided by the present invention is a method of treating infertility in a female subject which method comprises determining the potential of an oocyte from a female subject to develop into a viable embryo in accordance with the methods of the invention as defined herein, reducing or increasing activity or expression of any one or more of said markers that are present at differential levels compared to said reference range so that they are within said reference range for optimum viability. Similarly, the markers identified may also be useful in reducing fertility thus acting as targets for use in a contraceptive context.

Furthermore, the methods of the invention may be utilised to monitor the fertility of a female subject and which may be utilised in the assessment of a female's fertility status to assess, for example, an appropriate time to undergo assisted reproduction therapy.

By "reference concentration range or reference range" is meant any range of values, including a single value that may be used for the purposes of comparison. A reference range can reflect the outcome of a single assay from a single follicle or from a plurality of follicles in an ovary or from assays from multiple ovaries in respect of multiple subjects. The reference range may also be a statistical function of the results of multiple assays. By comparing cytokines/marker concentrations in the follicular fluid of oocytes that successfully fertilised and led to 'viable embryos ' with those that fertilised but did not lead to 'viable embryos ' it is also possible to determine the optimum cytokine concentrations that are associated with both successful fertilisation and embryo viability. Exemplary reference ranges are described in Tables 3 to 5 herein.

Thus, for example, the reference range may be established by identifying the concentration levels of the markers identified in Table 2 in the follicular fluid of individual follicles of females having successful and unsuccessful outcomes in assisted reproduction therapy for the oocytes/embryos derived from said follicles, correlating the concentration levels of one or more of said markers with the successful and unsuccessful outcomes to identify the optimum concentration ranges of said cytokines in the individual follicles to achieve a positive outcome in assisted reproduction therapy. In one embodiment of the invention, the reference range of said cytokine markers may be established by measuring the cytokine levels in the follicular fluid of individual follicles containing fertilisation competent or incompetent oocytes from a female that is not suffering from any medical or inflammatory condition that may alter the cytokine profile in said follicular fluid. These females may or may not have been undergoing assisted reproductive therapy. Therefore, in one embodiment the reference ranges may be established using information on the marker levels for oocytes derived from individual follicles in females undergoing assisted reproductive therapy because of male factors.

By differential levels of said markers it is meant that a protein in said follicular fluid is present at different levels compared to the same protein taken from the follicular fluid of a follicle containing fertilisation competent or incompetent oocytes depending on the reference range utilised. Recognition of such differential expression may be utilised in the diagnosis and/or development of therapeutic interventions for the treatment of various inflammatory conditions that may cause infertility in a female subject. In accordance with the present invention, reference to "differential level(s)" of said markers may be defined as being significantly different at the 5% significance level.

The present inventors have, therefore, surprisingly identified that levels of one or more markers, particularly cytokine markers, in the follicular fluid derived from single or individual follicles compared to said reference concentration of said cytokine markers, provides a correlation to the potential of an oocyte to be successfully fertilised and/or for the successfully fertilised oocyte to develop into a viable embryo. Thus, advantageously, the methods of the invention can be used to identify those oocytes from individual follicles that when used in assisted reproductive therapy will be more likely to result in a successful pregnancy. Alternatively, the method may be particularly useful in facilitating the assessment of the Oocyte and embryo viability' over and above that which is dependent on embryonic growth, morphology and age-related factors. For example, if a couple fail over a prolonged period to conceive naturally there may be a number of underlying causes such as endometriosis, chronic pelvic inflammatory disease and polycystic ovaries in the female. In these patients there may be additional issues with respect to Oocyte and embryo viability '. The present invention, may permit or facilitate the prospective assessment of prognosis with assisted conception (IVF/ICSI) by allowing a medical practitioner to assess the quality of oocytes in addition to the other medical impediments to conception. Thus, advantageously, the methods may be used both in a diagnostic context and as a tool to further enhance the chances of a positive outcome in assisted reproduction therapy.

In determining the viability of embryos for their selection prior to transfer in assisted reproduction therapy, the methods of the invention may be used alone or in combination with traditional morphological or metabolic assessment of the embryo. Furthermore, in determining the viability of oocytes for their selection for fertilisation

(where only a finite number of oocytes may be fertilised e.g. in Italy, Ireland & Germany) the methods of this invention may be used alone or in combination with traditional morphological assessment of the oocyte.

In view of the increased chances of a successful pregnancy being associated with the performance of the method of the invention, it is possible to reduce the number of embryos that may be implanted thereby obviating the risks associated with multiple pregnancies and the obstetric and perinatal complications these bring. This may also render it possible to determine the current putative understanding that selecting two good embryos for transfer (the determination of "good" and "bad" being based on the previously used morphological assessment methods) will give rise to a twin pregnancy, whereas transferring one "good" embryo and one "bad" embryo, leading to a singleton pregnancy, ultimately affects the surviving foetus, as is evident from the increased incidence of low birth weight in singleton IVF pregnancies (De Geyter et al., 2006). Ideally, selecting one good embryo alone will maximise benefits for both mother and baby (De Neubourg et al., 2006).

Identifying the fertilisation competence/viability of an oocyte and the viability of an embryo is also fundamental for oocyte or embryo cryopreservation. On one hand, it is important to determine which embryos are likely to survive cryopreservation, which will have an impact on determining whether it is worthwhile freezing the embryos from a single cohort after selection as per normal routine by growth rate and morphology. However, another meritorious application of the performance of the present invention lies with women who may attend assisted conception units for oocyte freezing. This may be because of the risk of premature menopause, gonadal toxicity from chemotherapy or exceptionally for social reasons such as career choices to postpone having children and thereby obviating age-related detrimental changes to their oocytes. Accordingly, the methods of the invention may also be utilised as a pre-cycle screen to ensure oocytes of optimum viability are obtained for any subsequent IVF cycle whether or not the oocytes are to be cryopreserved.

The methods of the invention may also, advantageously, be utilised to develop a medium for use in in vitro culture of mature and immature oocytes. The medium may be enriched with one or more of the cytokines/proteins identified in the follicular fluid and in proportion such that the culture medium provides the optimum profile for in-vitro maturation of the oocytes that will develop the most viable embryos. Therefore, not only will it be possible to select those oocytes from the individual follicles that are likely to produce the most viable embryos but the media used to culture such oocytes and embryos in the laboratory can also be enriched to include the proteins from the follicular fluid that are likely to provide the best in vitro culture medium. This particularly useful aspect of the invention may enable a rescue of the borderline oocytes or may act to nullify the attretic effects of age related and other inflammatory, pro-apoptotic changes.

The methods of invention may also provide for a comparative assessment of the different treatment regimens, using various GnRH analogues and antagonists, urinary versus recombinant follicle stimulating and luteinising hormones as cytokine production in the follicle is a reflection of gonadotrophic stimulation and steroid milieu of the follicle. Identification of methods that promote optimum Oocyte and embryo viability ' could thus be developed.

There may be a potential for the study of local effects of inflammatory cytokines and therapeutic modalities may be developed to arrest progression of the disease or to prevent the detrimental effects on the gametes. Advantageously, identification of individual markers in follicular fluid derived from individual follicles has been found to be predictive of oocyte and embryo viability. However, in view of the complex nature of the interactions of cytokines, it is preferred that a plurality of the markers may be utilised and upon which a more accurate predictive assessment may be made. Therefore, the plurality of markers may comprise any of two, three, four....up to all forty eight markers listed in table 2 herein. Since some markers are more strongly predictive of positive outcomes in assisted reproduction therapy, these markers may advantageously be utilised either individually or alternatively in combination as predictors of fertilisation competence or embryo viability. The markers may also be ranked individually and in combination in terms of their predictive value.

In one embodiment of the invention, the method involves identifying a plurality of markers in the follicular fluid from said individual follicles. In a further embodiment of the invention, a reference profile of either the critical cytokine or protein markers, or alternatively a more extensive complement of markers as set out in Table 2, (ranging from two to all forty eight markers) may be established against which a follicle to be tested may be compared. In such a profile, those markers that are critical to the viability of an oocyte and/or embryo and their respective concentrations may be established. The concentrations of these critical markers in a follicle to be tested may, therefore, be identified and therefore the suitability or viability of the oocyte obtained therefrom may be ranked or graded for its fertilisation competence or viability. In accordance with the methods of the present invention, it is also possible to rank embryos using the marker concentration data in terms of their suitability for assisted reproduction techniques. The embryos obtained from a single IVF cycle, for example, may be ranked with respect to one another and in the context of their overall viability compared to a reference profile. A number of cytokines have been identified which either alone or in combination provide a particularly accurate prediction of the likelihood that an oocyte from a particular individual follicle will result in a successful pregnancy in assisted reproduction therapy. Therefore, the markers may be used either alone or in combination, for example, to provide a more extensive cytokine profile reflecting the spectrum of cytokines that are functioning in the follicular fluid of individual follicles. In accordance with the present invention a successful pregnancy is defined as an embryo that has successfully implanted in the lining of the uterus and continues to develop for a period of at least six weeks thereafter.

Measurement of the markers in the reference range established using viable embryos or fertilisation competent oocytes indicates an increased likelihood of embryo viability and therefore its chances of developing into a successful pregnancy. Furthermore, a measurement of at least 0.01, 0.1, 0.5, 1, 5, 10, 15, 20% outside the reference range indicates a correspondingly decreased likelihood of embryo viability and the oocyte being able to develop in to a successful pregnancy.

This invention may also be very important in establishing markers that could be targets for therapeutic intervention. The present invention provides a method of assessing the potential of an oocyte or cohorts of oocytes from a female subject to develop into 'a viable embryo/s ' as defined herein. Therapeutic interventions may be developed to reduce or increase the activity or expression of any one or more of the said markers which are present at differential levels so that they are within the said 'reference range ' to induce/improve optimum viability.

Similarly, therapeutic interventions could be developed to reduce or increase the activity or expression of any one or more of the said markers which are present at differential levels so that they are NOT within the said 'reference range ' to reduce viability or induce attresia / apoptosis, thus reducing fertility and acting as contraceptives.

In one embodiment of the invention, the markers to be identified comprise any of

IL-2, IL-4, IL-6, IL-7, IL-8, IL-9, IL-10, IL-12 (p40), IL-12 (p70), IL-13, IL-15, IL-17, IL- 18, Eotaxin, FGF, G-CSF, GM-CSF, IFN-α2, IFN-γ, MCP-I, MIP- lα MIP- lβ,

RANTES, TNF-α, VEGF, CTACK, M-CSF, MIF, b-NGF, SCF, SDF-lα, TNF-β, TRAIL and CRP. When the cytokine to be assessed is any of IL-2, IL-4, IL-6, IL-9, IL-12 (p40), IL- 12 (p70), IL-15, Eotaxin, IFN-α2, MIP-I β, RANTES, TNF-α, M-CSF, MIF, SDF-lα, elevated levels of any one of said cytokines are individually predictive of better quality embryos and greater pregnancy success rates and lower levels are predictive of reduced embryo viability. On the other hand, reduced levels of any one of IL-7, IL-8, IL-IO, IL-13, IL- 17, IL- 18, FGF, G-CSF, GM-CSF, IFN-γ, MCP-I, MIP- lα, VEGF, CTACK and CRP are individually predictive of better quality embryos and increased pregnancy rates.

In one embodiment at least three cytokines may be tested which comprise any of either of the following groupings i) VEGF, IFNα and MIF, or ii) IL-9, G-CSF or RANTES. Alternatively, when at least four cytokines are to be tested the cytokines tested may include VEGF, IFNα, CTACK and MIF. Alternatively, when at least five cytokines are to be tested the cytokines may include IL-7, VEGF, IFNα, CTACK and MIF. When at least six cytokines are to be included in the profile, they may include any of the following combinations, i) IL- lβ, IL-7 VEGF, IFNα CTACK, and MIF; ii) IL-5, IL-7, VEGF, IFNα, CTACK and MIF; iii) IL-7 VEGF, IFNα CTACK GROα and MIF; or iv) IL-7, IL-9, G- CSF, RANTES, CTACK and SDF. lα. When at least 7 cytokines are utilized the profile preferably includes IL-5, IL-7, IL-9, G-CSF, RANTES, CTACK and SDF.lα or IL.7 IL.9 G.CSF RANTES VEGF CTACK SDF.1. alpha

When at least eight cytokines are included in the profile, the profile may include any of the following groupings; i) IL-I β, IL-7, MCP.l, VEGF, IFNα, IL2rα, CTACK and MIF; or ii) IL- lβ, IL-7, VEGF, IFNα, IL2rα, CTACK, MIF and SDF-lα; or iii) IL.5 IL.7 IL.9 G.CSF RANTES VEGF CTACK SDF.1. alpha; or iv) IL.7 IL.9 G.CSF MCP.l RANTES VEGF CTACK SDF.1. alpha; or v) IL.7 IL.9 G.CSF MIP. Ia RANTES VEGF CTACK SDF.1.alpha; or vi) IL.7 IL.9 G.CSF RANTES VEGF CTACK LIF SDF.l.alpha; or vii) IL.13 FGF G.CSF VEGF CRP IL18 CTACK MIF.

When at least nine cytokines are to be utilized the cytokines include any of the following groupings; i)IL-lβ, IL-7, MCP-I, VEGF, IFNα, IL2rα, CTACK, MIF and SDF- lα, or ii) IL.5 IL.7 IL.9 G.CSF MCP.l RANTES VEGF CTACK SDF.l.alpha; or iii) IL.13 FGF G.CSF GM.CSF VEGF CRP IL18 CTACK MIF. When at least 10 cytokines are to be ulitised, the cytokines may include any of the following groupings; i) IL- lβ, IL-7, MCP-I, VEGF, IFNα, IL2rα, CTACK, MIF, SDF- lα and TRAIL; or ii) IL.5 IL.7 IL.9 G.CSF GM.CSF MCP.1 RANTES VEGF CTACK SDF.l.alpha; or iii) IL.13 IL.15 FGF G.CSF GM.CSF VEGF CRP IL18 CTACK MIF; or iv) IL.13 FGF G.CSF GM.CSF MCP.1 VEGF CRP IL 18 CTACK MIF; or v) IL.13 FGF G.CSF GM.CSF VEGF CRP IL16 IL18 CTACK MIF.

When at least eleven cytokines are to be included they may include any of the following grouped combinations of cytokines i) IL- lβ, IL-7, IL-8, MCP-I, VEGF, IFNα, IL2rα, CTACK, MIF, SDF- 1 α and TRAIL; ii) IL- 1 β, IL-7, MCP- 1 , VEGF, CRP, IFNα, IL2rα, CTACK, MIF, SDF- lα and TRAIL; iii) IL- lβ, IL-7, MCP-I, VEGF, IFNα, IL2rα, IL- 16, CTACK, MIF, SDF- lα and TRAIL; iv) IL- lβ, IL-7, MCP-I, VEGF, IFNα, IL2rα, IL- 18, CTACK, MIF, SDF- lα and TRAIL; v) IL- lβ, IL-7, MCP-I, VEGF, IFNα, IL2rα, CTACK, GROα, MIF, SDF- lα and TRAIL; or vi) IL. Ib IL.7 MCP.1 VEGF IFNa IL12..p40. IL16 IL18 CTACK LIF MIF

When at least twelve cytokines are to be tested the cytokines may include any of the following groups i) IL-lβ, IL-7, MCP-I, MlP-lβ, VEGF, CRP, IFNα, IL2rα, CTACK, MIF, SDF- lα and TRAIL; ii) IL-lβ, IL-7, IL-8, MCP-I, VEGF, CRP, IFNα, IL2rα, CTACK, MIF, SDF- 1 α and TRAIL; iii) IL- 1 β, IL-7, MCP- 1 , VEGF, CRP, IFNα, IL- 16, IL2rα, CTACK, MIF, SDF- lα and TRAIL, iv) IL-lβ, IL-7, MCP-I, VEGF, CRP, IFNα, IL2rα, CTACK, GROα, MIF, SDF- lα and TRAIL; or v) IL-5, IL-7, MIP- lβ, RANTES, VEGF, CRP, IFNα, CTACK, M-CSF, MIF, SDF- lα and TRAIL ; vi) IL. Ib IL.7 IL.9 G.CSF MCP.l MlP.la RANTES VEGF CRP IL12..p40. CTACK SDF.l.alpha; or vii) IL.7 IL.9 G.CSF MCP.l MlP.la PDGF RANTES VEGF IL12..p40. IL18 CTACK SDF.l.alpha; or viii) IL.7 IL.9 G.CSF MCP.l MlP.la RANTES VEGF IL3 IL12..p40. IL18 CTACK SDF.l.alpha; or ix) IL.7 IL.9 G.CSF MCP.l MlP.la RANTES VEGF IL12..p40. IL18 CTACK GROalpha SDF.l.alpha.

When at least thirteen cytokines are to be tested the cytokines may include IL.7

IL.9 G.CSF MCP.l MlP.la PDGF RANTES VEGF IL12..p40. IL18 CTACK MCP3

SDF.l.alpha. When at least fourteen cytokines are tested the cytokines may include any of the following groups i) IL.5 IL.7 IP.10 MlP.lb RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SDF.1. alpha TRAIL; ii) IL.5 IL.7 MIP. Ib RANTES VEGF CRP IFNa IL12..p40. CTACK MCP3 M.CSF MIF SDF.l.alpha TRAIL; iii) IL.5 IL.7 IL.8 MlP.lb RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SDF.l.alpha TRAIL; iv) IL.5 IL.7 IL.12.p70. MlP.lb RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SDF.l.alpha TRAIL; v) IL.5 IL.7 IL.15 MlP.lb RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SDF.l.alpha TRAIL; vi) IL.5 IL.7 MlP.lb PDGF RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SDF.l.alpha TRAIL; vii) IL.5 IL.7 MlP.lb RANTES TNF.a VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SDF.l.alpha TRAIL; viii) IL.5 IL.7 MlP.lb RANTES VEGF CRP IFNa IL12..p40. CTACK MCP3 M.CSF MIF SDF.l.alpha TRAIL; ix) IL.5 IL.7 MlP.lb RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SCF SDF.l.alpha TRAIL; x) IL.7 IL.9 G.CSF MCP.l MlP.la PDGF RANTES VEGF IL12..p40. IL18 CTACK GROalpha MCP3

SDF.l.alpha; xi) IL.7 IL.9 G.CSF MCP.l MlP.la PDGF RANTES VEGF IL12..p40. IL18 CTACK MCP3 SCF SDF.l.alpha.

When at least 15 cytokines are to be utilised the cytokines may include any of the following groups i) IL. Ib IL.7 IL.8 FGF MCP.1 VEGF CRP IFNa IL2ra CTACK MCP3 MIF b.NGF SDF.l.alpha TRAIL; or ii) IL.5 IL.7 MlP.lb PDGF RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SCF SDF.l.alpha TRAIL.

In a preferred embodiment of the invention, the measured values for the level of each of the markers in said follicular fluid may be subjected to a standardisation protocol to account for contamination of said follicular fluid sample with non- follicular fluid as a result of, for example the oocyte collection procedure. In one embodiment, the measured values for the levels of each of the markers is standardised against a protein not found in follicular fluid and which preferably has a molecular weight above approximately 7OkD. In a preferred embodiment, said standardisation protocol comprises applying the measured concentration levels of said markers in the follicular fluid to the following equation, FF = (Zy-Bx)/(y-x) wherein,

Z = Follicular fluid cytokine concentration measured FF =The corrected concentration of cytokines in follicular fluid B = Blood cytokine concentration x = concentration in follicular fluid sample of standardising protein y = concentration in blood sample of standardising protein all values being expressed as pg per mg of protein

The use of a standardisation protocol is particularly advantageous because it enables all of the follicles in the ovaries to be used in the method of the invention. Typically, the oocyte collection procedure involves puncturing the follicle containing the oocyte with an aspiration needle and drawing out the follicular fluid contained within the oocyte. While the fluid from the first follicle may be relatively free of contaminating blood and plasma, the remainder of the follicles will invariably be contaminated thus rendering the predictive reliability of any technique involving cytokine profiles obtained from the second and subsequent follicles sampled highly questionable. Substantially all of the follicular fluid sampled from the second and subsequent follicles will contain some contamination by virtue of the invasive nature of the oocyte collection procedure. Accordingly, any methodology that can substantially mitigate the effect of such contamination of the follicular fluid sample with proteins, including cytokines, from blood will be highly advantageous as it permits an accurate assessment of the follicular fluid contents and subsequent comparison of markers contained therein with those of reference marker levels obtained, for example, from individuals that had a successful pregnancy as a result of IVF and whose follicles would also be subjected to contamination. It is believed that it has never previously been considered to apply the measured levels of any proteins obtained from follicular fluids of single follicles to such a standardisation protocol.

Thus advantageously, there is further provided by the present invention, a method of preparing a follicular fluid sample(s) for analysis to assess viability of an oocyte therein, which method comprises providing a sample of follicular fluid, preferably from a single ovarian follicle, subjecting said sample to a standardisation protocol wherein measurement of the levels of proteins to be tested in said sample reflects the level of said protein in said follicular fluid and not from cross contamination. Thus, advantageously, the measurements are more reflective of the levels of said proteins in said follicular fluid and substantially reduce the possibility of the measured marker levels being contaminated by the presence of similar markers, for example, from blood. In one embodiment, the protein markers identified in the follicular fluid samples are standardised against a protein that is not found in the follicular fluid and which may, for example, have a molecular weight not exceeding approximately 70 kD. In a preferred embodiment the standardising protein is von Willebrand factor (vWF). This is found in blood but because the ovarian follicle is poorly vascularised the only source of vWF in ovarian follicles is via contamination as a result of the oocyte collection process.

As aforementioned, the cytokine markers utilised in the performance of the present invention may themselves be graded or ranked in terms of their predictive value. This may be carried out by analysing the concentration ranges found in females undergoing assisted reproduction therapy and identifying those cytokines that are present at significantly different levels between preganant and non-pregnant females.

Furthermore, various statistical techniques can also be applied to the measured cytokine values to rank a collection of embryos in the order of their predicted chances of resulting in a successful pregnancy. It is also possible to generate an algorithm that will facilitate predictive assessment of the likelihood of achieving an implantation and ongoing pregnancy. This is particularly important where only a limited number of embryos can be legally transferred, (for example, in the UK, it is only possible under current legislation to implant two embryos in women under 40) and the spare embryos need to be assessed for their implantation potential and ability to survive the cryo-preservation and thawing procedure.

A number of statistical techniques are available and which are described more fully in the examples below and which can be used to establish a ranking of embryos that are suitable for IVF, for example. Logistic regression may be used to predict the probability of a successful outcome, which in the context of the present invention is a successful pregnancy. Using logistic regression, the cytokine concentrations in the follicular fluid of a single follicle are attributed a value, as set out more clearly in the examples. A value called Akaike's Information Criterion (AIC) can be calculated and used to choose which cytokines to include in the model. Using this procedure, 27 cytokines were identified as being relevant to model the success of the embryos. The model can, advantageously, therefore be used to predict the probability of success for any embryo given the cytokine concentration measurements from the follicular fluid from which the oocyte was derived. Accordingly, the predicted probabilities can be used as a basis to rank the oocytes in terms of their likelihood of success in any assisted reproduction techniques, with higher predicted probability of success corresponding to a better oocyte.

Logistic regression is only one statistical method which can be used to rank embryos in this manner. Other possible statistical methods (the Mahalanobis distance and discriminant analysis) are described below.

The cytokine concentration profiles obtained from individuals may also reflect any underlying inflammatory conditions that are affecting the individual at the time the sample of follicular fluid is collected and which may have had an additional impact on the oocyte and embryo viability. Therefore, in establishing the cytokine profiles, it is possible to correlate the cytokine levels with clinical conditions and study their impact on folliculogenesis and oogenesis at the ovarian and oocyte level so that women with an underlying inflammatory condition or a condition that has the indirect effect of raising or otherwise altering cytokine profiles, can have their cytokine profiles compared with those that have resulted in a successful pregnancy and which were also affected by the same condition.

Therefore, advantageously, the method of the invention facilitates the generation of specific subpopulation profiles to provide a specific background against which a particular individual's cytokine profile may be appropriately compared. In this way an individual's clinical profile may be compared to that of a specific subpopulation to make the method even more accurate. The generation of such cytokine profiles may also in some circumstances permit the diagnosis of any such conditions in an individual that have heretofore gone undetected and may, for example, provide a clinical explanation for the perceived infertility of a female individual.

The method of the invention has, advantageously, been achieved by virtue of a multiplex immunoassay for screening for any one or more of the highlighted cytokines. The merit of the cytokine multiplexing approach on a per follicle basis is the fact that it is entirely non-invasive, accounts for variability within patients and is not influenced by culture conditions (e.g. use of different media) in different assisted conception units. It thus has a broad array of applications to determine oocyte fertilising ability and embryo viability to establish pregnancy following cryopreservation. This method, can therefore, readily be used to select sub-cohorts of high-grade embryos that have been pre-selected based on their morphology and in-vitro development. While this has immediate and obvious benefits as a supplementary approach to conventional morphometric embryo grading in, for example the UK, there are also non Human Fertilisation and Embryology Authority legislated systems, such as those seen under applications in Germany and Italy. In these countries, only a limited number of oocytes can be fertilised by law, so that only one or two embryos for transfer can be created at any one time. This in return poses a significant restriction on the success rates that can be achieved by assisted conception in those countries. Hence it is even more advantageous in this scenario to have the ability to predict the fertilising ability and viability of the oocyte and hence also the embryo. In one embodiment, the multiplex immunoassay comprises xMAP detection technology However, other suitable multiplex assays may be used and which are known to those of skill in the art.

Also provided by the present invention is a kit for the performance of the method of the invention and which kit comprises one or more binding agents, e.g, antibodies or the like that are capable of binding to and recognising one or more of the follicular fluid markers identified herein, for contacting with a sample of follicular fluid obtained from a single ovarian follicle, means for contacting said sample and said one or more binding agents, the binding agent including an appropriate label or reporter molecule to confirm the presence of the marker and means for determining the presence of said label or reporter molecule(s) and it's concentration. The means for determining the presence of said label may utilise said xMAP detection technology or other multiplex immunoassay techniques. This may also be coupled to a computer processer that includes details of said optimum reference ranges against which the follicular fluid profiles from said single ovarian follicles may be compared and ranked in accordance with the viability.

In accordance with this embodiment, an analyzer for identifying the cytokines present in a sample may provided. The data obtained by the analyzer 1 may then be transmitted to a computer which is configured by software either provided on a disk or by receiving an electrical signal by a communications network, to be configured into a number of functional modules which cause the computer to process the image data received from the analyser to generate either an output image or a readout of cytokine levels which is shown on a display . The analyzer 1 may comprises a multiplex assay kit such as a Bio-Plex assay system. A processing module including processing software to process the information received from said analyzer is provided and which may produce said output image/readout generated as a result of the cytokine/marker measurements identified in said sample. Another module, such as a viability model, may also be provided through which the data from the processing module may be processed and against which the data from said sample may be compared to assess the viability of the embryo in the follicle from where the follicular fluid of said sample is derived.

Although the embodiments of the invention described with reference to the drawings comprise computer apparatus and processes performed in computer apparatus, the invention also extends to computer programs, particularly computer programs on or in a carrier, adapted for putting the invention into practice. The program may be in the form of source or object code or in any other form suitable for use in the implementation of the processes according to the invention. The carrier can be any entity or device capable of carrying the program.

The invention may be more clearly understood by reference to the following examples and accompanying drawings. Description of Drawings

Figure 1 is an illustration of the development of an ovarian follicle

Figure 2 is an illustration of the principles of assisted reproduction therapy

Figure 3 is a diagrammatic representation of the principle underlying fluid-phase multiplex immunoassays. Blue lines represent the excitation light beam, coloured lines the emission beams

Figure 4 is a diagrammatic illustration of the results obtained for simulated cytokines Ci and C₂.

Figure 5 is a diagrammatic representation of the results obtained utilising the Mahalanobis distance.

Figure 6 is a diagrammatic representation of the results obtained using linear and quadratic discriminant analysis

Figure 7 is an illustration of the results obtained from embryos ranked according to the method of the invention for a number of patients of known outcome. The ordinate represents the arbitrary ranking score given by the algorithm. Each column of circles represents one patient, with filled circles being those embryos chosen and transferred by the embryologist. Where the uppermost circle in every column is filled, both the embryologist and the model agree.

Figure 8 is a schematic block diagram of an embodiment of the apparatus of the present invention Examples

Methodology

The follicular fluid from patients undergoing in-vitro fertilisation was collected. The stimulation protocol utilised was standard and includes pituitary down regulation with either a GnRH agonist or antagonist, ovarian hyperstimulation with urinary or recombinant FSH (or both), hCG as an LH surrogate to trigger ovulation and hCG, intramuscular or intravaginal progesterone for luteal support. Oocytes are collected, approximately 35 to 37 hours after the hCG injection, transvaginaly under ultrasound guidance. Each follicular fluid sample is examined for the presence of an oocyte. The oocyte is removed for fertilisation whilst the follicular fluid is saved for processing in the laboratory The oocytes are then fertilised by insemination of the eggs using a sperm separation technique as in IVF or injected with the sperm as in the ICSI procedure. The resulting embryo is then replaced in the uterus and the woman receives further medication afterwards to provide hormonal support in the luteal phase of the cycle the nature of which is dependent on the assessment of the risk of ovarian hyper-stimulation syndrome.

Follicular fluid/blood and standardisation protocols

Sample collection

Follicular fluid (aspirate) samples are collected from individual follicles and examined for the presence of an oocyte. If an oocyte is present it is removed prior and the sample transferred into a 15ml centrifuge tube, labeled contemporaneously with the corresponding follicle number. Each follicle aspirated and each oocyte retrieved is assigned a chronological number in the order of aspiration and retrieval. The oocyte number is carried forward by the embryo to its final destiny and hence each follicle aspirated, its oocyte when retrieved and the fate of the oocyte/embryo can be followed to the end. When an oocyte is retrieved, the oocyte number is also recorded on the corresponding sample tube. Sample processing and handling

Using a benchtop centrifuge, samples are spun at 2,00Og for 5 minutes to deposit the contaminating cellular debris into a pellet.

The supernatants of the centrifuged samples are pipetted into autoclaved 1.5ml eppendorf tubes in ~lml aliquots. These eppendorf tubes identify the pertinent information:

• Patient Study Number

• Follicle Number • Oocyte Number (or "NE" for No Egg retrieved from this follicle)

• Patient's HFEA Number

When less than ImI of follicular fluid is retrieved from a follicle, the sample is divided between several aliquots to accommodate the labile nature of the proteins in the sample. Samples that cannot be processed immediately are stored in a refrigerator (4-8⁰C) or on ice. All samples must be processed within 2hs of collection.

Sample storage and preservation

The aliquots of follicular fluid are sequentially ordered into freezer storage boxes. Records are maintained on the number of eppendorf tubes corresponding to a follicle, the participants and follicles located in a storage box, and the location of the storage box within the freezer. Samples are processed and then immediately frozen at -8O⁰C after sample processing. Sample preservation is most adequately achieved by minimizing the freeze/thaw insult. Samples should only be removed from the freezer with the intention of performing an assay.

Preparation of Serum / Plasma

Serum / plasma was centrifuged at 2000 rpm for 10 minutes and the supernatant placed in aliquots and stored at -8O⁰C. Cytokine/Protein Marker Assay

Sample Preparation for analysis

Follicular fluid and plasma samples are all kept on ice until ready for use. Thawed samples are prepared for analysis by diluting 1 volume of sample with 1 volume of plasma/serum diluent prior to analysis.

Cytokine sample analysis

The cytokines and other markers are typically measured either by bioassay or immunoassay. Both techniques are time consuming and can facilitate the analysis of only a single cytokine at a time. The Bio-Plex suspension array system, which utilizes xMAP

(Luminex) detection technology, incorporates novel technology using colour-coded beads, permits the simultaneous detection of up to 100 cytokines in a single well of a 96-well microplate. Bio-Plex cytokine assays are multiplex bead-based assays designed to quantify multiple cytokines in diverse matrices; including serum samples, plasma samples, and tissue culture supernatants, and therefore also lends itself well to the analysis of follicular fluid. By multiplexing, it is possible to quantify the level of multiple cytokines in a single well in just 3h, using as little as 12.5-25μl of sample. The advantages over traditional immunoassays that analyze only a single cytokine at a time include the ability to create a complete cytokine profile from limited sample, reduce sample preparation time, and increase throughput. Although the Luminex platform is optimally used, there are others alternative available which would allow us to achieve the same end.

This suspension array system is built around three core technologies. The first is the family of fluorescent Iy dyed microspheres (beads) to which bio molecules are bound. The second is a flow cytometer with two lasers and associated optics to measure biochemical reactions that occur on the surface of the microspheres. The third is a highspeed digital signal processor that efficiently manages the fluorescent output. The Bio- Plex suspension array system employs patented multiplexing technology that uses up to 100 colour-coded bead sets, each of which can be conjugated with a specific reactant. Each reactant is specific for a different target molecule. Bio-Plex cytokine assays are designed in a capture sandwich immunoassay format. Antibody specifically directed against the cytokine of interest is covalently coupled to colour-coded 5.6μm polystyrene beads. The antibody-coupled beads are allowed to react with a sample containing an unknown amount of cytokine, or with a standard solution containing a known amount of cytokine. After performing a series of washes to remove unbound protein, a biotinylated detection antibody specific for a different epitope on the cytokine is added to the beads. The result is the formation of a sandwich of antibodies around the cytokine. The reaction mixture is detected by the addition of streptavidin-phycoerythrin (streptavidin-phycoerythrin), which binds to the biotinylated detection antibodies. The constituents of each well are drawn up into the flow-based suspension array system, which identifies and quantifies each specific reaction based on bead colour and fluorescence. The magnitude of the reaction is measured using fluorescent Iy labelled reporter molecules associated with each target protein. Unknown cytokine concentrations are automatically calculated by Bio -P lex Manager software using a standard curve derived from a recombinant cytokine standard. By using coloured beads as the solid phase instead of a coated well, up to 100 differently coloured beads can be mixed and used for quantifying <100 different analytes simultaneously.

The Bio-Plex cytokine assays and panels contain anti-cytokine conjugated beads, cytokine detection antibody and cytokine standard. In addition, the following materials are also required: serum diluent (to optimise recovery), a Luminex system (cytometer and sheath fluid pump), Bio-Plex validation kit (includes optics/classification/reporter/fluidics validations), Bio-Plex Manager 4.0 software, MCV plate, Bio-Plex calibration kit, microplate shakers, vacuum manifold, vortexer and general consumables (pipettes and tips, sterile distilled water, aluminium foil).

All samples have been analysed for 50 cytokines (outlined in table 2) using a commercially available kit (Bio-rad: Human 27- and 23-plex cytokine assay). Assays are otherwise prepared according to manufacturer's instructions, available from Bio-Rad Laboratories, Inc.; Web site ww w. bio -rad . com; United Kingdom Tel: 020 8328 2000. C-reactive protein (CRP)

CRP is one of the most sensitive acute phase proteins. The analysis of CRP utilises an enhanced latex-turbimetric immunoassay with samples being processed using an clinical chemical autoanalyser (COBAS MIRA) and a commercially available kit (Thermo Electron Corporation, Alpha Labs, UK).

Transforming growth factor (TGF)-β

TGF-βi is one of several TGF-β isoforms with similar biological functions. It is recognised as inducing growth inhibition of various normal and transformed cell types, as well as stimulating the synthesis of/inhibiting the degradation of extracellular matrix proteins. More importantly, TGF-βi is recognised for its immunosuppressive effects: it inhibits IL2-dependent proliferation of both T and B-lymphocytes. It also suppresses interferon-induced natural killer cell cytotoxicity of natural killer cells, the activity of cytotoxic T-lymphocytes, and the proliferation of the precursors of lymphokine-activated killer cells. In particular, follicular fluid TGF-βi levels have been associated with the establishment of pregnancy in man and embryo viability in the mouse (Osterlund and

Fried, 2000; Robertson et al., 1997). Accordingly, clarifying its role in predicting human embryo viability was deemed of interest.

TGF-βi was measured using a commercially available ELISA kit (R&D systems, Abingdon, UK), as the specific chemistry of the reaction did not allow its analysis in a multiplex immunoassay format.

Osterlund C, Fried G (2000) TGFbeta receptor types I and II and the substrate proteins Smad 2 and 3 are present in human oocytes MoI Hum Reprod 6(6):498-503.

Robertson SA, Mau VJ, Hudson SN, Tremellen KP (1997) Cytokine-leukocyte networks and the establishment of pregnancy Am J Reprod Immunol. 37(6):438-42. As indicated in the results below, follicular fluid levels of TGF-beta were significantly higher in the twin pregnancy group compared to the non-pregnant group (P<0.05), suggesting that this cytokine was associated with improved embryo viability.

Not pregnant Singleton Twins

Average 44.85034 Average 49.09422 Average 54.48646

SEM 6.011347 SEM 8.945225 SEM 35.201

Range low- Range Range high 0-1516.45 high 0-961.18 high 0-240.27

von Willebrand Factor von Willebrand Factor (vWF) is a large multimeric glycoprotein present in blood plasma and as such does not pass through the basement membrane which surround the follicle and is not found in follicular fluid, therefore its presence is an indicator of contamination. The concentration of vWF in plasma and follicular fluid samples was measured using an commercially available ELISA kit (Stago, Diagnostica Satgo UK LTD, Reading, UK).

Protein Assay

The concentration of protein was measured using the Bradford assay in all follicular fluid and plasma samples in order to standardise for the flush component within the follicles. This was performed using a commercially available kit (Bio-rad Protein assay, Bio-rad, UK).

Data analysis for cytokine concentrations

If the concentrations are expected to be in the range 0-1,000 pg/ml, such as in serum, then the PMT setting is set to 'high' for a narrow range standard curve using conventional 4-fold standard dilutions optimised for 5-parameter logistic curves. Bio-Plex Manager software contains features that simplify the process of multiplex cytokine assay data analysis including determination of assay precision, selection of an appropriate curve fitting routine, and determination of the goodness of fit of the regression algorithm Standardisation protocol

Both plasma and follicular fluid profiles have been assessed for cytokine profile since some follicular fluid samples are contaminated with blood. This is a normal consequence of oocyte retrieval but means that follicular levels may be slightly altered by blood (plasma) contamination. The extent of this contamination and its consequent effect on cytokine profile has been rectified by relating it to the amount of detectable factor von Willebrand factor present in follicular fluid samples. This is determined using the following equation;

Z = FF(I -x/y) + B(x/y) therefore,

FF = (Zy - Bx)/(y-x) where, Z = Follicular fluid cytokine concentration measured by the Luminex

FF =The corrected concentration of cytokines in follicular fluid

B = Blood cytokine concentration measured by the Luminex x = vWF concentration in follicular fluid sample y = vWF concentration in blood sample All values are then expressed as pg per mg of protein.

The standardised and non-standardised values for the measured concentration cytokine levels in pregnant and non-pregnant females are set out in Tables 3 and 4 below.

Standardised Results

CVtokint* Average Average Range Twin Range NPT Fredimve Predictive

Twin NPT range range

0.00097 0.0014 -0.0002-0.031 -0.0018-0.083 0.0004-0.0016 0.001 -0.0017

IL-lα

IL-lβ 0.0086 0.0043 -0.020-0.178 -0.11-0.17 0.0045-0.0126 0.003-0.005

IL-lra 0.669 4.758 -3.45-75.55 -4.05-216 0.348-0.989 3.68-5.82

IL-2ra 2.719 2.454 -0.003-11.83 -99.36-41.87 2.401-3.038 2.07-2.83

IL-2 0.0095 0.004 -0.396-7.859 -0.75-1.93 -0.016-0.035 -0.0009 - 0.009

IL-3 1.51 1.76 -0.747-12.410 -6.15 -85.13 1.186-1.828 1.43-2.09

IL-4 -0.008 0.006 -0.0014-1.074 -0.11-1.13 -0.034-0.0177 0.003-0.007

IL-5 0.0037 0.009 -0.0067-2.156 -0.091-3.38 0.0025 - 0.0049 0.006-0.011

IL-6 0.123 0.07 -0.043-3.606 -0.414-5.003 0.0826-0.164 0.062 - 0.079

IL-7 0.03 0.08 -0.074-2.990 -0.20-5.61 0.022-0.038 0.07-0.091

IL-8 0.465 0.844 -0.016-13.63 -1.23-14.97 0.397-0.534 0.789 - 0.90

IL-9 0.017 -0.03 -0.259-2.813 -5.29 - 8.93 -0.008 - 0.041 -0.048 - -0.01

IL-IO 0.099 0.035 -0.272-10.284 -0.522-15.28 0.056-0.142 0.029 - 0.042

IL12 13.27 12.95 -1.527-80.041 -13.15-320.31 11.252-15.278 11.42-14.47

(p40) IL- 0.038 0.027 -0.214-7.082 -1.23-8.42 0.0188-0.0582 0.021-0.033

12(p70) IL-13 -0.006 0.016 -0.0019-1.216 -0.063-1.96 -0.0245-0.0112 0.013-0.018

IL-15 0.062 0.080 -0.07-2.218 -0.08 - 10.4 0.019-0.011 0.044-0.117

IL-16 2.383 1.808 -0.001-37.55 -28.16-22.21 1.737-3.029 1.655-1.962

IL-17 -0.012 -0.003 -0.315-0 -0.828 - 0.402 -0.018 - -0.005 -0.006-0.0001

IL-18 0.581 0.793 -0.010-1.667 -4.084-15.418 0.53-0.631 0.729-0.856

Eotnxin 1.20 0.74 0-11.77 -0.32-10.97 1.02-1.37 0.69 - 0.79

FGF 0.0995 0.080 -0.149-4.822 -0.386-91.495 0.025-0.174 0.789 - 0.900

G-CSF 0.127 0.268 -0.279 - 5.965 -0.212 -14.234 0.032-0.223 0.249 - 0.287

GM-CSF 0.274 0.55 -0.142-7.542 -0.727-47.876 0.128-0.422 0.354-0.747

IFN-α 2.188 2.761 -0.002-7.352 -11.11-44.18 2.007-2.368 2.594-2.992

IFN-γ 0.111 0.181 -0.196-5.398 -0.934-11.272 -0.025 - 0.248 0.151-0.212

IP-IO 5.014 6.410 -0.104-25.183 -3.48-64.2 4.187-5.842 6.020-6.801

MCP-I 0.761 0.953 -0.003-8.839 -0.45-12.165 0.634-0.887 0.907 - 0.999

MIP-Ia 0.0031 0.011 -0.019-0.039 -0.032 - 0.248 0.00007 - 0.005 0.0099-0.0128

MlP-lβ 0.283 0.399 -0.033-2.461 -0.429 - 6.667 0.245 - 0.32 0.374 - 0.424

PDGF 5.353 7.393 -3.049-83.215 -4.711-790 3.178-7.528 4.617-10.168

RANTES 7.810 8.401 0-141.46 -0.425-275.61 4.773 - 10.847 7.101-9.01

TNF-α 0.109 0135 -0.712-4.575 -0.939-13.26 0.014-0.203 0.089-0.181

VEGF 5.45 10.34 0.87-29.85 1.28-79.11 4.77-6.13 9.8- 10.88

CTACK 6.232 7.765 -0.003-22.620 -0.282-141.27 5.625 - 6.84 7.196-8.333

GROα 2.216 2.698 -3.032-23.08 -10.85-48.474 1.753-2.779 2.453-2.943

LIF 0.025 0.160 -0.25-0.736 -0.048-18.43 0.008-0.041 0.099 - 0.220

MCP3 0.1274 0.112 -0.186-1.240 -0.351-3.736 0.097 -0.158 0.097-0.127

M-CSF 0.456 0.616 -0.0686-3.532 -0.166-7.457 0.378-0.535 0.572-0.689 27.91 94.64 -0.004 - 707.13 -6.074 - 14.84-40.99 21.85-167.43

MIF 1963.4

MIG 3.427 4.137 -0.0035 - 14.58 -31.72-66.39 3.045-3.809 3.788-4.486 b-NGF 0.056 0.051 -0.0001 -0.307 -0.042-0.901 0.049 - 0.062 0.047-0.055

SCF 1.462 1.457 -0.001 - -9.252 -1.373-18.42 1.244-1.68 1.372-1.187

5.062 6.271 -0.006- 15.488 -4.777 - 4.676 - 5.448 1.372-1.543

SDF-lα 76.379

TNF-β 0.010 0.020 -0.005 - - 0.478 -0.024-5.001 0.019-0.19 0.0046-0.035

6.23 7.34 -0.004 - 17.899 -33.22- 5.77-6.67 6.80-7.87

TRAIL 102.18

CRP 0.039 0.055 -0.037- -0.307 -0.021-0.723 0.03 - 0.047 0.050-0.061

Table 3

Non-standardised results

Not pregnant Twins

Range Predictive Range Predictive

Average SEM Min Max Low High Average SEM Min Max Low Hig

IL-Ib 0.187 0.017 0.000 2.400 0.169 0.204 0.226 0.059 0 .000 2.560 0.167 0.28

IL-lra 384.214 82.910 0.000 17243.440 301.304 467.124 119.202 54.751 0 .000 2529.010 64.450 173.95

11-2 1.350 0.426 0.000 98.480 0.924 1.776 1.534 1.210 0 .000 78.620 0.325 2.7

IL-4 0.863 0.307 0.000 68.040 0.557 1.170 1.482 0.761 0 .000 45.440 0.721 2.24

IL-5 1.372 0.424 0.000 82.970 0.948 1.796 2.651 1.763 0 .000 91.200 0.888 4.4

IL-6 5.813 1.300 0.000 276.920 4.513 7.113 5.619 1.799 0 .000 105.080 3.820 7.41

IL-7 5.127 0.863 0.000 151.420 4.264 5.990 7.354 3.437 0 .000 170.750 3.917 10.79

W IL-8 47.140 3.653 0.000 729.020 43.487 50.793 30.441 5.013 0 .000 232.650 25.428 35.45

IL-9 3.370 1.511 0.000 437.690 1.858 4.881 4.433 2.260 0 .000 118.990 2.172 6.69

IL-IO 7.801 2.690 0.000 532.450 5.111 10.492 14.356 8.018 0 .020 435.020 6.338 22.37

IL-12(p70) 5.900 1.756 0.000 327.080 4.144 7.657 7.698 4.749 0 .000 299.560 2.950 12.44

IL-13 1.238 0.286 0.000 52.920 0.953 1.524 1.838 0.919 0 .000 51.440 0.919 2.75

IL-15 2.207 1.387 0.000 507.880 0.820 3.594 1.025 0.387 0 .000 17.270 0.638 1.41

IL-17 0.256 0.131 0.000 40.960 0.125 0.387 0.000 0.000 0 .000 0.000 0.000 0.00

Eotaxin 40.011 3.204 0.000 426.460 36.807 43.215 66.265 10.534 0 .000 345.550 55.731 76.79

FGF 5.465 1.755 0.000 342.040 3.710 7.219 4.603 3.216 0 .000 203.980 1.387 7.81

G-CSF 16.011 1.760 0.000 297.490 14.251 17.772 21.835 7.725 0 .000 399.260 14.110 29.56

GM-CSF 9.028 2.640 0.000 958.760 6.388 11.669 10.945 3.593 0 .000 226.690 7.352 14.53

IFN-g 9.587 1.451 0.000 291.590 8.136 11.038 13.344 4.647 0 .000 221.250 8.697 17.99

IP-IO 293.583 14.317 0.000 1863.970 279.266 307.900 244.782 30.668 0 .000 1035.840 214.114 275.45

MCP-I 50.650 3.002 0.000 550.050 47.649 53.652 43.522 7.306 0 .000 331.990 36.216 50.82

MIP-Ia 0.478 0.038 0.000 4.410 0.440 0.517 0.294 0.055 0 .000 1.660 0.239 0.35

MIP-Ib 20.751 1.501 0.000 348.140 19.250 22.252 13.813 1.375 0 .000 47.730 12.437 15.18

PDGF 130.670 20.657 0.000 3801.360 110.013 151.326 198.635 55.170 0.000 2351.080 143.465 253 •80

RANTES 243.316 17 .298 0 .000 2478.020 226.019 260.614 290.007 38.443 0 .000 1306 .860 251 .564 328 •45

TNF-a 6.656 2. 661 0 .000 756.920 3.995 9.318 4.833 1.399 0 .000 63 .270 3 .434 6 23

VEGF 503.989 23 .729 0 .000 4266.520 480.260 527.718 289.604 29.608 1 .380 1262 .660 259 .996 319 .21

CRP 2.693 0. 195 0 .000 30.000 2.499 2.888 2.095 0.330 0 .000 11 .290 1 .765 2 .42

IFNa 106.910 2. 173 0 .000 226.750 104.737 109.084 112.457 4.910 22 .470 201 .480 107 .547 117 .36

IL-Ia 0.105 0. 021 0 .000 3.190 0.084 0.126 0.071 0.045 0 .000 2 .100 0 .026 0 .1 I

IL2ra 120.135 4. 789 0 .000 697.020 115.347 124.924 137.600 11.562 2 .870 459 .970 126 .038 149 .16

IL3 60.789 4. 615 0 .000 605.000 56.174 65.404 82.409 14.823 0 .000 559 .480 67 .586 97 .23

IL12 (p40) 518.077 24 .609 0 .000 2865.600 493.468 542.686 649.125 75.854 0 .000 2826 .760 573 .271 724 .97

IL16 83.505 2. 636 0 .000 298.050 80.869 86.141 107.246 15.087 2 .470 944 .090 92 .159 122 .33

IL18 35.598 1. 260 0 .000 205.170 34.338 36.859 32.624 2.345 1 .180 85 .290 30 .279 34 .96

CTACK 326.367 9. 930 8 .400 1314.840 316.437 336.297 323.868 24.637 29 .850 991 .790 299 .231 348 .5O _w GROalpha 140.126 9. 404 0 .000 1348.980 130.722 149.530 146.767 30.875 0 .000 1539 .450 115 .891 177 .64

^ LIF 4.314 0. 692 0 .000 67.180 3.622 5.006 1.972 0.970 0 .000 38 .480 1 .002 2 .94

MCP3 5.090 0. 369 0 .000 69.570 4.721 5.459 8.091 2.080 0 .000 100 .550 6 .011 10 .17

M-CSF 30.091 2. 319 0 .000 727.730 27.772 32.409 26.050 3.115 0 .000 95 .690 22 .934 29 .16

MIF 599.850 90 .876 0 .000 25300.390 508.974 690.725 951.283 401.990 46 .400 18828 .670 549 .294 1353 .27

MIG 184.429 9. 680 0 .000 1807.180 174.749 194.109 173.887 16.651 6 .860 773 .010 157 .236 190 .53 b-NGF 2.235 0. 071 0 .000 7.530 2.164 2.306 2.805 0.248 0 .230 8 .360 2 .557 3 .05

SCF 67.432 2. 199 0 .000 234.300 65.233 69.631 74.347 6.587 0 .200 248 .950 67 .760 80 .93

SDF-I alpha 267.517 5. 906 0 .000 736.150 261.611 273.423 267.730 13.437 51 .060 672 .230 254 .293 281 .16

TNF-b 0.496 0. 101 0 .000 13.350 0.395 0.597 0.404 0.268 0 .000 12 .900 0 .135 0 .67

TRAIL 309.256 7. 340 0 .960 878.700 301.917 316.596 334.958 20.959 33 .470 977 .440 313 .998 355 .9I

Table 4

Ranges

Male factor SEM Min Max

IL-I b 0.001704608 0.001391 -0.00197 0.012479

IL-1 ra 2.512182109 2.424476 -0.06288 26.74972

II-2 0.005926138 0.003894 -0.03911 0

IL-4 0.002544641 0.001368 -0.00092 0.013728

IL-5 0.003779201 0.001131 -0.00252 0.01034

IL-6 0.054283621 0.019487 0 0.249267

IL-7 0.034287564 0.011059 -0.00437 0.097002

IL-8 0.907690084 0.207068 0.203061 2.26341

IL-9 0.002680167 0.001224 -0.01031 0

IL-10 0.010484237 0.002671 0.001985 0.035788

IL-12(p70) 0.016348166 0.007064 -0.011 0.055732

IL-13 0.010222278 0.003094 -0.00033 0.030611

IL-15 0.008751815 0.009071 -0.00395 0.108466

IL-17 -0.00053864 0.000444 -0.00537 0

Eotaxin 1.116188 0.402534 0 4.731213

FGF 0 0 0 0

G-CSF 0.183740801 0.073776 -0.00837 0.850444

GM-CSF 0.100776743 0.045957 -0.01336 0.441093

IFN-g 0.089628973 0.050072 -0.03879 0.583713

IP-10 4.200326276 0.954718 0.49314 9.780776

MCP-1 0.645444379 0.109906 0.037642 1.429814

MIP-Ia 0.006289872 0.002882 -0.00124 0.02638

MIP-I b 0.295651493 0.064623 0.007163 0.89006

PDGF 1.029136578 0.491714 0 4.814144

RANTES 3.234041022 0.872003 0 10.52943

TNF-a 0.023520286 0.014175 -0.02022 0.134215

VEGF 10.98468579 2.39434 1.040357 31.75414

CRP 0.036714862 0.009623 -0.00051 0.088972

IFNa 2.1 1738801 0.308289 1.240781 5.194127

IL-Ia 0.00303434 0.003034 0 0.036412

IL2ra 2.1 19052187 0.57307 0.791111 7.880256

IL3 1.033200851 0.566091 -1.10083 6.503225

IL12 (p40) 11.09490799 4.226506 0.385314 54.08839

IL16 1.95068705 0.608796 0.565272 8.104373

IL18 0.75571225 0.188278 0.092328 2.312256

CTACK 6.269624494 1.09303 1.149008 13.67318

GROalpha 3.296121064 1.500721 -0.03394 17.44192

HGF 143.7455501 28.62453 61.01329 384.2054

ICAM 1 111.2546001 43.70768 -105.75 360.3401

LIF 0.000821087 0.000644 -0.0076 0

MCP3 0.066265695 0.024091 -0.01964 0.26437

M-CSF 0.334165784 0.055647 -0.00436 0.611845

MIF 10.82816609 4.948084 0.960386 59.39506

MIG 2.739457911 0.511725 1.037089 6.508703 b-NGF 0.040541825 0.010984 0.008657 0.155619 SCF 1.472448446 0.513325 0.34725 6.708201

SCGF-b 204.9474085 30.05441 36.52863 412.1295

SDF-1 alpha 5.608556754 0.719692 2.88926 11.48467

TNF-b 1.62443E-05 1.62E-05 0 0.000195

TRAIL 5.139086008 0.611446 2.057232 10.84172

VCAM- 1 134.7076592 34.60354 -38.7014 368.1143

Table 5: Cytokine levels for women having assisted conception due to male factor only

Ranking embryos using cytokine data

In the present study, embryos which led to no pregnancy are referred to as "unsuccessful" embryos, while those which led to a twin pregnancy (and hence were known to both be successful) are referred to as 'successful" embryos. At this stage, embryos where the outcome was unknown are not used in the analysis.

Preprocessing of data

Some additional pre-processing of the cytokine/marker data was appropriate before further analysis. These data were subject to a number of missing values and the presence of outliers.

Missing values were imputed by replacing them by a suitable value.

• For unsuccessful embryos, missing values were replaced by the median of all measurements of that cytokine on unsuccessful embryos.

• Similarly, for successful embryos, missing values were replaced by the median of all measurements of that cytokine on successful embryos.

To reduce the influence of outliers, all cytokine values were "shrunk" towards the median value of that cytokine. Data on successful and unsuccessful embryos were shrunk towards the medians of their respective groups. This was done by taking logarithms of the amount by which the cytokine values were above or below the median value for each cytokine. As a result, cytokine values close to the median were changed little, while extreme outliers were brought substantially closer to the median values. This allows us to use the pertinent information - that those particular embryos had particularly high / low cytokine measurements - without the extreme values having an undue influence on later results.

Using cytokine data to rank embryos

Three methods have been considered to rank a collection of embryos in order of their predicted chances of resulting in a successful pregnancy, based on their cytokine profiles. These methods are described below. The descriptions in this section give an overview of each method; more detailed descriptions are set out more fully below, where some mathematical details and an example of applying these methods to an illustrative artificial data set are given.

Logistic regression

Logistic regression can be used to predict the probability of a successful outcome (pregnancy in our case) given values of a collection of predictors (the cytokine values). Once a logistic regression model is fitted, a value called Akaike's Information Criterion (AIC) can be calculated and used to choose which cytokines to include in the model. Using this procedure identified 27 cytokines as being relevant to model the success of the embryos.

This model can be used to predict the probability of success for any embryo given cytokine measurements on that embryo. Hence we can use these predicted values to rank embryos - higher predicted probability of success would indicate a "better" embryo.

Discrepancies from known baselines

An alternative approach is to assess how far each embryo is from the established baseline profile of cytokines in the successful and unsuccessful groups. These baseline profiles were formed as the median value of each cytokine i.e. a basal unsuccessful embryo would be one which had the median value of each cytokine from the "no pregnancy" embryos. For any particular embryo, the distance from both of these baselines was calculated using a measure called the Mahalanobis distance, which allows for correlations between cytokines. A score was calculated from these distances as

distance from unsuccessful baseline distance from unsuccessful baseline + distance from successful baseline,

which will be zero for an embryo which has cytokines matching the unsuccessful baseline and one for cytokines matching the successful baseline. New embryos can be ranked in terms of this score.

Classification-based rankings

The statistical technique of discriminant analysis is used to build a rule to separate observations into two or more classes based on values of predictor variables. Different versions of discriminant analysis are available for data sets satisfying different assumptions. For example, linear discriminant analysis assumes that the relationships between cytokines will be the same among successful embryos as among unsuccessful embryos; quadratic discriminant analysis is more complicated to use but avoids this restriction.

In the present case, it is possible to build a rule to separate successful and unsuccessful embryos. When applied to new embryos, the rule can be used to give the probability of a new embryo being in each class. Thus it is possible to build a rule on the available data and use this rule to predict the probability of success for new embryos, which can then be ranked in order of this predicted probability.

Validation of methods

Clearly, the performance of any model should be assessed to validate that model before using it in any trial. The models described above were assessed in terms of their ability to discriminate between successful and unsuccessful embryos. Each embryo was predicted to be either successful or unsuccessful if their predicted probability / score exceeded 0.5. Then the predicted and actual status of the embryos were tabulated to give a 2 x 2 table of the following form.

True status Unsuccessful Successful

Predicted Unsuccessful a b status Successful c d

The percentage of embryos which were correctly classified is then

a + b

-X lOO a + b + c + d

To ensure that the methods were fairly assessed, the technique of cross-validation was used. This is an extension of dividing the data into a training set (to develop the model) and a test set (to assess the model's performance). With cross-validation we form a test set of a single observation and train a model on the remaining data. The model is used to predict the single test observation and we evaluate whether this was a correct prediction. Then the procedure is repeated with a different test observation. We continue to apply this procedure until each observation has been predicted from the other observations in the data set, giving us a percentage of correct classifications as above.

The techniques described above were assessed using the technique of cross-validation and the following percentage rates of correct classification obtained.

Method % correct classification Logistic regression 90.3%

Distance from baseline 99.3%

Linear discriminant analysis 88.4% Quadratic discriminant analysis 93.1%

Clearly, for these data, using the Mahalanobis distance from the baselines gives a highly accurate classification of successful and unsuccessful embryos. Since these percentages are based on a cross-validation procedure, the models have been tested on data not used for the model building process, giving a more accurate assessment of the models' performance.

Illustrative examples

The present example illustrates the use of the methods described above on some artificial data. For the purposes of this example, we use simulated data where we have only measurements on two cytokines (labelled Ci and C₂) for 100 successful and IOU unsuccessful embryos. The simulated data are shown in figure 4, where open circles denote unsuccessful embryos and triangles successful ones. Filled circles show imaginary "new" embryos which we shall rank by the methods discussed below.

Logistic regression

Logistic regression estimates the probability that all embryos will be successful, denoted p, by the log odds value

log odds = log -

¹ I - P The log odds are then modelled in the same fashion as a simple linear regression, resulting in an equation of the form log odds = α + β\ x Ci + β₂ x C₂ + β₃ ^x Ci x C₂

Here, the β values represent the following: α the base chance of success β\ the effect of cytokine Ci βi the effect of cytokine C₂ /?3 the effect of any statistical interactions between cytokines Ci and C₂.

The simplest possible model here has all β values set to zero, implying that the cytokine values have no influence. Successively more complicated models include more of the β parameters at non-zero values. As outlined above, we use the AIC statistic to select the appropriate model. The possible models and associated AIC values are

Model AIC log odds = α 279

log odds = α+/?₂C₂ 72

log odds = α +

+ AsCiC₂ 38

Since we use the model with the lowest AIC value, the appropriate model here is the most complicated model. Fitting this model gives us the equation

log odds = -21.9 + 7.5Ci + 3.0C₂ - 1.2CiC₂

We convert the log odds values, denoted I, to probabilities of success, denoted p via the equation e¹

^{P =} \ ₊ e^>

so that the rankings for this method will be given on a 0-1 scale as with the other methods.

Distance-based classification

This method is based on the use of the Mahalanobis distance, which incorporates correlations between variables while a normal distance does not. This is illustrated in figure 5, showing the artificial unsuccessful data. In this figure, the two squares are further apart than the two triangles using our normal concept of "distance". However, the data are more spread out in the bottom left-top right direction than in the top left-bottom right direction, and this should be taken into account. In fact, the Mahalanobis distance between the squares is equal to that between the triangles.

Discriminant analysis

These methods work by finding a line on figure 4 separating the successful and unsuccessful embryos. The lines found by linear and quadratic discriminant analysis are shown in figure 6. As the names suggest, "linear" DA is restricted to using a straight line, while "quadratic" DA can use a curve.

Ranking new embryos

For the four "new" embryos represented by filled circles on figure 4, the rankings implied by the methods above are shown in the table below. In this table, embryo el is the top-left point on figure 4, close to the centre of the unsuccessful embryos, grading to embryo e4, which is close to the centre of the successful embryos. The tabulated values are in all case on a 0-1 scale with zero corresponding to unsuccessful and one to successful.

Embryo Logistic Mahalanobis Linear DA Quadratic DA regression distance el 0.033 0.380 0.004 0.023 e2 0.325 0.497 0.131 0.410 e3 0.998 0.668 0.858 0.982 e4 1.000 0.897 0.996 1.000

In this case, all four methods have correctly ranked the four embryos.

Results

The results of the standardised and non-standardised results obtained from the Student T test are set out in Tables 3 and 4. Cytokines concentrations in ovarian follicles sampled indicated a number of cytokines that were present at significantly different concentrations in the successful and unsuccessful embryos. Cytokines that were identified as predictors of oocyte viability using this method were, IL-6, IL-7, IL-8, IL-IO, IL-13, Eotaxin, G-CSF, MIP- lα and VEGF. For these cytokines, the ranges identified were those that were significantly different at the 5% significance level.

As a result of the standardisation process, to take into account blood contamination and any protein in the sample, up to sensitivity level of kit, results close to zero that are then standardised may become negative due to this process.

In assessing the results obtained for the Student T-test comparison of pregnant versus non pregnant data, a positive result was considered to be the case if the measured concentrations were significantly different at the 5% significance level.

Logistic regression can be used to predict the probability of a successful outcome (pregnancy in our case) given values of a collection of predictors (the cytokine values). Once a logistic regression model is fitted, a value called Akaike's Information Criterion (AIC) can be calculated and used to choose which cytokines to include in the model. Using this procedure identified 27 cytokines as being relevant to model the success of the embryos.

This model can be used to predict the probability of success for any embryo given cytokine measurements on that embryo. Hence we can use these predicted values to rank embryos - higher predicted probability of success would indicate a "better" embryo.

Of the cytokines adopted utilising the T test above, logistic regression also identified IL-6, IL-IO, Eotaxin, G-CSF, MIP- lα and VEGF as being predictive of oocyte viability and thus pregnancy success rates. In particular, a higher concentration of IL-6 and Eotaxin were identified as being predictive of better quality embryos and greater pregnancy success rates, whereas lower concentrations of IL-10, G-CSF, MIP- lα and VEGF were predictive of better quality embryos. Logistical regression also additionally identified that the presence of IL-2, IL-4, 11-9, IL- 12 (p40), IL- 12 (p70), IL- 15, IFN-α, MIP- lβ, RANTES, TNF-α, M-CSF, MIF and SDF lα in higher concentrations is predictive of embryo quality and increased success rates in pregnancy. Lower concentrations of IL- 17, IL- 18, FGF, GM-CSF, IFN-γ, MCP-I, CTACK and CRP are predictive of embryo quality.

Identification of Combinations of Cytokines

The search method to identify combinations of cytokines that together would be predictive of oocyte viability was as follows:

1. Start with the two good triples identified by checking all possible triples. These triples were the ones which gave the highest percentage "score" of correctly classifying the embryos as successful or unsuccessful.

2. For each triple, try adding each other cytokine in turn. Whichever one(s) give the best score become identified as candidate "sets of four"

3. For each set of four, find which are the best cytokines to add to form a "set of five".

Continue to increase the number of cytokines used, in each case only adding those that give the best scores. For any given number of cytokines, some sets did better than others. The primary interest is in the sets which gave the best possible score for a given number of cytokines.

The results obtained for a given number of cytokines are:

number of cytokines number of sets best % correct nnuumber of "best" sets considered classification

3 16215 87.7% 2

4 2 91.4% 1

5 2 94.3% 1

6 4 95.5% 4

7 8 97.2% 1

8 12 97.9% 2

9 34 98.6% 1

10 70 98.8% 1

11 199 99.1% 5

12 493 99.5% 5

13 1664 99.5% 55

14 6317 100% 2

Thus, for example, there were four sets of six cytokines considered and these all led to a score of 95.5% correct classification. On the other hand, over 6,000 sets of 14 cytokines were considered and of these only 2 gave the best score of 100% correct classification.

The details for up to 15 cytokines are

Best selections of 3 cytokines found are i)VEGF IFNα MIF ii) IL.9 G.CSF RANTES

Best selections of 4 cytokines found are VEGF IFNα CTACK MIF

Best selection of 5 cytokines found IL.7 VEGF IFNα CTACK MIF 4 best selections of 6 cytokines found i) IL. Ib IL.7 VEGF IFNa CTACK MIF ii) IL.5 IL.7 VEGF IFNa CTACK MIF iii)IL.7 VEGF IFNa CTACK GROalpha MIF iv) IL.7 IL.9 G.CSF RANTES CTACK SDF.1.alpha

Best selection of 7 cytokines found are IL.5 IL.7 IL.9 G.CSF RANTES CTACK SDF.1. alpha

IL.7 IL.9 G.CSF RANTES VEGF CTACK SDF.1. alpha

Best selections of 8 cytokines found are i) IL. Ib IL.7 MCP.1 VEGF IFNa IL2ra CTACK MIF ii) IL. Ib IL.7 VEGF IFNa IL2ra CTACK MIF SDF.1. alpha iii) IL.5 IL.7 IL.9 G.CSF RANTES VEGF CTACK SDF.1. alpha iv) IL.7 IL.9 G.CSF MCP.1 RANTES VEGF CTACK SDF.1. alpha v) IL.7 IL.9 G.CSF MIP. Ia RANTES VEGF CTACK SDF.1. alpha vi) IL.7 IL.9 G.CSF RANTES VEGF CTACK LIF SDF.1. alpha vii) IL.13 FGF G.CSF VEGF CRP IL 18 CTACK MIF

Best selections of 9 cytokines found are i) IL. Ib IL.7 MCP.1 VEGF IFNa IL2ra CTACK MIF SDF.1. alpha ii)IL.5 IL.7 IL.9 G.CSF MCP.1 RANTES VEGF CTACK SDF.1. alpha; iii)IL.13 FGF G.CSF GM.CSF VEGF CRP IL 18 CTACK MIF

Best selections of 10 cytokines found are i) IL. Ib IL.7 MCP.1 VEGF IFNa IL2ra CTACK MIF SDF.1. alpha

TRAIL; ii) IL.5 IL.7 IL.9 G.CSF GM.CSF MCP.1 RANTES VEGF CTACK SDF.1. alpha iii) IL.13 IL.15 FGF G.CSF GM.CSF VEGF CRP IL18 CTACK MIF iv) IL.13 FGF G.CSF GM.CSF MCP.l VEGF CRP IL18 CTACK MIF v) IL.13 FGF G.CSF GM.CSF VEGF CRP IL 16 IL 18 CTACK MIF

Best selections of 11 cytokines found are i) IL. Ib IL.7 IL.8 MCP.l VEGF IFNa IL2ra CTACK MIF SDF.1. alpha TRAIL ii) IL. Ib IL.7 MCP.1 VEGF CRP IFNa IL2ra CTACK MIF SDF.1. alpha TRAIL iii)IL. Ib IL.7 MCP.1 VEGF IFNa IL2ra IL 16 CTACK MIF SDF.1. alpha TRAIL iv)IL.lb IL.7 MCP.1 VEGF IFNa IL2ra IL 18 CTACK MIF SDF.1. alpha TRAIL v) IL. Ib IL.7 MCP.1 VEGF IFNa IL2ra CTACK GROalpha MIF SDF.1. alpha TRAIL vi) IL.lb IL.7 MCP.l VEGF IFNa IL12..p40. IL16 IL18 CTACK LIF MIF

Best selections of 12 cytokines found are i)IL.lb IL.7 MCP.l MIP. Ib VEGF CRP IFNa IL2ra CTACK MIF SDF.1. alpha TRAIL ii)IL. Ib IL.7 IL.8 MCP.l VEGF CRP IFNa IL2ra CTACK MIF SDF.1. alpha TRAIL iϋ)IL.lb IL.7 MCP.l VEGF CRP IFNa IL2ra IL16 CTACK MIF SDF.l.alpha TRAIL iv)IL.lb IL.7 MCP.l VEGF CRP IFNa IL2ra CTACK GROalpha MIF SDF.l.alpha

TRAIL v)IL.5 IL.7 MIP. Ib RANTES VEGF CRP IFNa CTACK M.CSF MIF SDF.l.alpha

TRAIL vi)IL.lb IL.7 IL.9 G.CSF MCP.l MlP.la RANTES VEGF CRP IL12..p40. CTACK

SDF.l.alpha; vϋ)IL.7 IL.9 G.CSF MCP.l MlP.la PDGF RANTES VEGF IL12..p40. IL18 CTACK

SDF.l.alpha; viϋ)IL.7 IL.9 G.CSF MCP.l MlP.la RANTES VEGF IL3 IL12..p40. IL18 CTACK SDF.l.alpha; ix)IL.7 IL.9 G.CSF MCP.l MlP.la RANTES VEGF IL12..p40. IL18 CTACK GROalpha

SDF.l.alpha;

Best selection of 13 cytokines found i)IL.7 IL.9 G.CSF MCP.l MlP.la PDGF RANTES VEGF IL12..p40. IL18 CTACK MCP3 SDF.l.alpha

Best selections of 14 cytokines found are i)IL.5 IL.7 IP.10 MlP.lb RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF

SDF.l.alpha TRAIL

ii)IL.5 IL.7 MIP. Ib RANTES VEGF CRP IFNa IL12..p40. CTACK MCP3 M.CSF MIF SDF.l.alpha TRAIL, ϋi)IL.5 IL.7 IL.8 MlP.lb RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SDF.1. alpha TRAIL iv)IL.5 IL.7 IL.12.p70. MlP.lb RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SDF.1. alpha TRAIL v)IL.5 IL.7 IL.15 MlP.lb RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SDF.1. alpha TRAIL vi)IL.5 IL.7 MlP.lb PDGF RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SDF.1. alpha TRAIL vii)IL.5 IL.7 MlP.lb RANTES TNF.a VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SDF.1. alpha TRAIL viii)IL.5 IL.7 MlP.lb RANTES VEGF CRP IFNa IL12..p40. CTACK MCP3 M.CSF MIF SDF.1. alpha TRAIL ix)IL.5 IL.7 MlP.lb RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SCF SDF.1. alpha TRAIL x)IL.7 IL.9 G.CSF MCP.l MIP. Ia PDGF RANTES VEGF IL12..p40. IL18 CTACK GROalpha MCP3 SDF.1. alpha xi)IL.7 IL.9 G.CSF MCP.l MlP.la PDGF RANTES VEGF IL12..p40. IL18 CTACK MCP3 SCF SDF.1. alpha

Best selections of 15 cytokines found are i)IL.lb IL.7 IL.8 FGF MCP.l VEGF CRP IFNa IL2ra CTACK MCP3 MIF b.NGF SDF.1. alpha TRAIL ii)IL.5 IL.7 MlP.lb PDGF RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SCF SDF.1. alpha TRAIL

Based on the initial data and the successful development of a ranking model, a further cohort of follicles was analysed retrospectively in order to determine whether there was agreement between morphometric grading and the LRB model. It also assessed whether the preliminary mathematical model could rank embryos from an individual woman in order of their putative developmental potential. Women were selected if they had a number of good morphological grade embryos that were both transferred and frozen. This subset of women from the twin (n=l2) and non-pregnant (n=20) group allowed the comparison of a number of follicular profiles using our preliminary mathematical model based on VEGF, IFN-α, CTACK and MIF alone. The model was used to rank their embryos. In instances where the embryologist agreed with the model's choice, there was a 64% success (pregnancy) rate. By contrast, when the embryologist and the model disagreed, success rate was only 27%. This suggests that the model picks viable embryos with good morphological grading. However, the model can also identify viable embryos which may be erroneously disregarded by the embryologist.

Based on these data, the model would have chosen a different embryo to the embryologist in circa 2/3 of these patients. Using these assumptions on how many of the non-pregnant patients may have become pregnant as a result of transferring embryos chosen by the model, we can hypothesise possible overall success rates:

• if 50% of all the model's selections were successful, the success rate would be 42%

• if 75% of all of the model's selections were successful, the success rate would be 50% (a doubling of current success rates)

• if all of the model's selections were successful, the success rate would be 66% (almost a trebling of success rates)

Any of these improvements in success rates would ensure uptake of the technology by units worldwide. The caveat to this interpretation is that these predictions depend on the assumption similar numbers of changes in the general population would be made as in this study population, and how many implemented changes truly lead to successful outcomes.

Figure 8 is a schematic block diagram of an embodiment of the apparatus of one aspect of the present invention. In accordance with this embodiment, an analyzer 1 for identifying the cytokines present in a sample 2 placed therein is provided. The data obtained by the analyzer 1 is then transmitted to a computer 3 which is configured by software either provided on a disk 4 or by receiving an electrical signal 5 by a communications network to be configured into a number of functional modules 6-7 which cause the computer 3 to process the image data received from the analyser 1 to generate either an output image 8 or a readout of cytokine levels which is shown on a display 9. The analyzer 1 comprises a multiplex assay kit such as a Bio-Plex assay system.

Processing module 6 includes processing software to process the information received from said analyzer 1 and to produce said output image/readout and which is generated as a result of the cytokine/marker measurements identified in said sample 2. Module 7 comprises a viability model through which the data from module 6 may be processed and against which the data from said sample may be compared to assess the viability of the embryo in the follicle from where the follicular fluid of said sample is derived.

Although the embodiments of the invention described with reference to the drawings comprise computer apparatus and processes performed in computer apparatus, the invention also extends to computer programs, particularly computer programs on or in a carrier, adapted for putting the invention into practice. The program may be in the form of source or object code or in any other form suitable for use in the implementation of the processes according to the invention. The carrier can be any entity or device capable of carrying the program.

For example, the carrier may comprise a storage medium, such as a ROM, for example a CD ROM or a semiconductor ROM, or a magnetic recording medium, for example a floppy disc or hard disk. Further, the carrier may be a transmissible carrier such as an electrical or optical signal which may be conveyed via electrical or optical cable or by radio or other means.

REFERENCES

1. ESHRE Task force on Ethics and Law. Ethical issues related to multiple pregnancies in medically assisted procreation. Hum Reprod 2003; 18:1976-9.

2. Land JA, Evers JLH. Risks and complications in techniques: Report of an ESHRE consensus meeting. Hum Reprod 2003;18:455-7. 3. HFEA figures accessible online at http://www.hfca.gov.uk.

4. Sattar N, Greer IA. Pregnancy complications and maternal cardiovascular risk: opportunities for intervention and screening? BMJ 2002;325: 157-60.

5. Terriou P, Giorgetti C, Hans E, Salzmann J, Charles O, Cignetti L, Avon C, Roulier R. Relationship between even early cleavage and day 2 embryo score and assessment of their predictive value for pregnancy. Reprod Biomed Online 2007;14:294-9.

6. Wilding M, Di Matteo L, DAndretti S, Montanaro N, Capobianco C, Dale B. An oocyte score for use in assisted reproduction. J Assist Reprod Genet 2007;24:350-8.

7. Schams D, Kosmann M, Berisha B, Amselgruber WM, Miyamoto A. Stimulatory and synergistic effects of luteinising hormone and insulin like growth factor 1 on the secretion of vascular endothelial growth factor and progesterone of cultured bovine granulosa cells. Exp Clin Endocrinol Diabetes 2001; 109: 155-62.

8. Gilchrist RB, Ritter LJ, Armstrong DT. Oocyte-somatic cell interactions during follicle development in mammals. Anim Reprod Sci 2004;82-83:431-46.

9. Bedaiwy M, Shahin AY, AbulHassan AM, Goldberg JM, Sharma RK, Agarwal A, Falcone T. Differential expression of follicular fluid cytokines: relationship to subsequent pregnancy in IVF cycles. Reprod Biomed Online 2007; 15:321-5.

10. Kawamura K, Kawamura N, Kumagai J, Fukuda J, Tanaka T. Tumor necrosis factor regulation of apoptosis in mouse preimplantation embryos and its antagonism by transforming growth factor alpha/phosphatidylionsitol 3-kinase signaling system. Biol Reprod 2007;76:611-8. 11. Weiser S, Miu J, Ball HJ, Hunt NH. Interferon-gamma synergises with tumour necrosis factor and lymphotoxin-alpha to enhance the mRNA and protein expression of adhesion molecules in mouse brain endothelial cells. Cytokine 2007;37:84-91.

12. Numerof RP, Asadullah K. Cytokine and anti-cytokine therapies for psoriasis and atopic dermatitis. BioDrugs 2006;20:93-103.

13. Andrews AL, Nasir T, Bucchieri F, Holloway JW, Holgate ST, Davies DE. IL- 13 receptor α 2: a regulator of IL- 13 and IL-4 signal transduction in primary human fibroblasts. J Allergy Clin Immunol 2006;l 18:858-65.

14. Baud V, Karin M. Signal transduction by tumor necrosis factor and its relatives. Trends Cell Biol 2001 ;11 :372-7.

15. Hagberg H, Mallard C, Jacobsson B. Role of cytokines in preterm labour and brain injury. BJOG 2005; 112(Suppl. l):16-8.

16. Christiansen OB, Nielsen HS, Kolte AM. Inflammation and miscarriage. Semin Fetal Neonatal Med 2006; 11:302-8.

17. Agic A, Xu H, Finas D, Banz C, Diedrich K, Hornung D. Is endometriosis associated with systemic subclinical inflammation? Gynecol Obstet Invest 2006;62: 139-47.

18. Machelon V, Nome F, Durand-Gasselin I, Emilie D. Macrophage and granulosa interleukin-1 beta mRNA in human ovulatory follicles. Hum Reprod. 1995 Aug;10(8):2198-203.

19. Arici A, Oral E, Bukulmez O, Buradagunta S, Bahtiyar O, Jones EE. Monocyte chemotactic protein- 1 expression in human preovulatory follicles and ovarian cells. J Reprod Immunol. 1997b Feb;32(3):201-19. 20. Arici A, Oral E, Bahtiyar O, Engin O, SeIi E, Jones EE. Leukaemia inhibitory factor expression in human follicular fluid and ovarian cells. Hum Reprod. 1997a Jun;12(6): 1233-9.

21. Macciό A, Mantovani G, Turnu E, Artini P, Contu G, Serri FG, Volpe A. Evidence that granulosa cells inhibit interleukin-1 alpha and interleukin-2 production from follicular lymphomonocytes. J Assist Reprod Genet. 1993 Nov;10(8):517-22.

22. BiIi H, Tarlatzis BC, Daniilidis M, Fleva A, Bontis J, Tourkantonis A, Mantalenakis S. Cytokines in the human ovary: presence in follicular fluid and correlation with leukotriene B4. J Assist Reprod Genet. 1998 Feb;15(2):93-8.

23. Bϋscher U, Chen FC, Kentenich H, Schmiady H. Cytokines in the follicular fluid of stimulated and non-stimulated human ovaries; is ovulation a suppressed inflammatory reaction? Hum Reprod. 1999 Jan; 14(1): 162-6.

24. Barak V, Mordel N, Zajicek G, Kalichman I, Treves AJ, Laufer N. The correlation between interleukin 2 and soluble interleukin 2 receptors to oestradiol, progesterone and testosterone levels in periovulatory follicles of in- vitro fertilization patients. Hum Reprod. 1992 Aug;7(7):926-9.

25. Qstanm AA, Aizikosdch BL Abnkovieb IV, Kozhin AY, Chernykh ER, Role of cytokines in the regulation of reproductive function. Bull Exp Biol Med. 2007 Jan;143(l):75-9.

26. M.QI!.^.oJ;!E_a..Pc^ endothelial growth factor in serum and in the follicular fluid of patients undergoing hormonal stimulation for in- vitro fertilization. Hum Reprod. 1998 Dec;13(12):3310-4.

27. Vujisic S, Lepej SZ, Emedi (, Baurrsan R, Retnenar A, TiJjak M K., Ovarian follicular concentration of IL-12, IL-15, IL-18 and p40 subunit of IL-12 and IL-23. Hum Reprod. 2006 Oct;21(10):2650-5. Epub 2006 Jun 13. 28. Hammadeh ME, Fischer-Hammadeh C Hoffmeister H, Herrmann W, Rosenhaum Ej_..S_.cbniidt___W. Relationship between cytokine concentrations (FGF, si CAM-I and SCF) in serum, follicular fluid and ICSI outcome. Am J Reprod Immunol. 2004 Jan;51(l):81-5.

29. Salmassi A, Schmutzler AG, Schaefcr S, Koch K, Hcddcrich J, Jonat W, Mcttlcr L_. Is granulocyte colony- stimulating factor level predictive for human IVF outcome? Hum Reprod. 2005 Sep;20(9):2434-40. Epub 2005 May 12.

30. Hammadeh ME, Fi seller- Hammadeh C, Gcorg T. Roscnbaum P, Schmidt VV. Comparison between cytokine concentration in follicular fluid of poor and high responder patients and their influence of ICSI-outcome. Am J Reprod Immunol. 2003 Aug;50(2):131-6.

31. Dahm-Kahlcr P, Rxmesson E, Lind AK, Brannstrόm M. Monocyte chemotactic protein- 1 in the follicle of the menstrual and IVF cycle. MoI Hum Reprod. 2006 Jan;12(l):l-6. Epub 2006 Jan 26.

32. Xu H, Schultzc-Mosgau A. Agic A, Dicdrich K, Taylor RN, Homung D. Regulated upon activation, normal T cell expressed and secreted (RANTES) and monocyte chemotactic protein 1 in follicular fluid accumulate differentially in patients with and without endometriosis undergoing in vitro fertilization. Fertil Steril. 2006 Dec;86(6): 1616-20. Epub 2006 Sep 25.

33. Asimakopoulos B, Nikolettos N, Papachristou DN, Simopoulou M, Al-Hasani S, Diedrich K. Follicular fluid levels of vascular endothelial growth factor and leptin are associated with pregnancy outcome of normal women participating in intracytoplasmic sperm injection cycles. Physiol Res. 2005;54(3):263-70. Epub 2004 Dec 9. 34. Kawano Y, Zeineh Hasan K, Fukuda J, Mine S, Miyakawa I. Production of vascular endothelial growth factor and angiogenic factor in human follicular fluid. MoI Cell Endocrinol. 2003 Apr 28;202(l-2): 19-23.

35. Viganό P, Fusi F, Gaffuri B, Bonzi V, Ferrari A, Vignali M.X. Soluble intercellular adhesion molecule-1 in ovarian follicles: production by granulosa luteal cells and levels in follicular fluid. Fertil Steril. 1998 Apr;69(4): 774-9.

36. Kawano Y, Kawasaki F, Nakamura S, Matsui N, Narahara H, Miyakawa I. The production and clinical evaluation of macrophage colony- stimulating factor and macrophage chemoattractant protein- 1 in human follicular fluids. Am J Reprod Immunol. 2001 Jan;45(l):l-5

37. Wada S, Kudo T, Kudo M, Sakuragi N, Hareyama H, Nishihira J, Fujimoto S. Induction of macrophage migration inhibitory factor in human ovary by human chorionic gonadotrophin. Hum Reprod. 1999 Feb;14(2):395-9.

38. Seifer DB, Lambert-Messerlian G, Schneyer AL. Ovarian brain-derived neurotrophic factor is present in follicular fluid from normally cycling women. Fertil Steril. 2003 Feb;79(2):451-2.

39. Srivastava MD, Lippes J, Srivastava BI. Cytokines of the human reproductive tract. Am J Reprod Immunol. 1996 Sep;36(3): 157-66.

40. Kryczek I, Frydman N, Gaudin F, Rrzysiek R, Fanchin R, Emilie D, Chouaib S, Zou W, Machelon V The chemokine SDF-I /CXCL 12 contributes to T lymphocyte recruitment in human pre-ovulatory follicles and coordinates with lymphocytes to increase granulosa cell survival. Am J Reprod Immunol. 2005 Nov;54(5):270-83.

41. Karstrδm-Encrantz L, Runesson E, Bostrόm EK, Brannstrόm M Selective presence of the chemokine growth-regulated oncogene alpha (GROalpha) in the human follicle and secretion from cultured granulosa- lutein cells at ovulation. MoI Hum Reprod. 1998 Nov;4(ll): 1077-83 42. Benifla JL, Bringuier AF, Sifer C, Porcher R, Madelenat P, Feldmann G.Vascular endothelial growth factor, platelet endothelial cell adhesion molecule-1 and vascular cell adhesion molecule-1 in the follicular fluid of patients undergoing IVF. Hum Reprod. 2001 Jul;16(7):1376-81

43. Fallon PG, John HE, Smith P, Emson CL, Townsend MJ, Fallon R, Smith P, McKenzie AN. IL-4 induces characteristic Th2 responses even in the combined absence of IL-5, IL-9, and IL-13. Immunity. 2002 Jul;17(l):7-17.

44. Gazvani MR, Bates M, Vince G, Christmas S, Lewis- Jones DI, Kingsland C. Follicular fluid concentrations of interleukin-12 and interleukin-8 in IVF cycles. Fertil Steril. 2000 Nov;74(5):953-8.

45. Koga K, Osuga Y, Yoshino O, Hirota Y, Yano T, Tsutsumi O, Taketani Y. Elevated interleukin-16 levels in the peritoneal fluid of women with endometriosis may be a mechanism for inflammatory reactions associated with endometriosis. Fertil Steril. 2005 Apr;83(4):878-82.

46. Zhang X, Xu H, Lin J, Qian Y, Deng L. X Peritoneal fluid concentrations of interleukin-17 correlate with the severity of endometriosis and infertility of this disorder. BJOG. 2005 Aug;l 12(8): 1153-5.

47. Aust G, Simchen C, Heider U, Hmeidan FA, Blumenauer V, Spanel-Borowski K.Eosinophils in the human corpus luteum: the role of RANTES and eotaxin in eosinophil attraction into periovulatory structures. MoI Hum Reprod. 2000

Dec;6(12):1085-91.

48. Roberts RM, EaIy AD, Alexenko AP, Han CS, Ezashi T. Trophoblast interferons. Placenta. 1999 May;20(4):259-64.

49. Imakawa K, Nagaoka K, Nojima H, Hara Y, Christenson RK. Changes in immune cell distribution and IL-IO production are regulated through endometrial IP-IO expression in the goat uterus. Am J Reprod Immunol. 2005 Jan;53(l):54-64 50. Liao F, Rabin RL, Yannelli JR, Koniaris LG, Vanguri P, Farber JM. Human Mig chemokine: biochemical and functional characterization. J Exp Med. 1995 Nov l;182(5):1301-14.

51. Chang AS, Dale AN, Moley KH. Maternal diabetes adversely affects preovulatory oocyte maturation, development, and granulosa cell apoptosis. Endocrinology. 2005 May;146(5):2445-53. Epub 2005 Feb 17.

52. Orvieto R, Chen R, Ashkersazi J, Ren-Harou&b A, Bar J, Finch B. C-reactive protein levels in patients undergoing controlled ovarian hyperstimulation for IVF cycle. Hum Reprod. 2004 Feb;19(2):357-9.

Claims

Claims

1. A method of determining the potential of individual oocyte to develop into a viable embryo, which method comprises establishing the concentration of any one or more of the markers identified in Table 2 in the follicular fluid of a single follicle from which said oocyte is derived and comparing the concentrations obtained to a reference concentration range that is established by comparing the concentration of said one or more markers from one or more follicles containing a fertilisation competent oocyte(s), wherein differential levels of said markers from said follicle to be tested compared to the reference range, for example at least at the 5% significance level, is predictive of a reduced potential of said oocyte to develop into a viable embryo following fertilisation.

2. A method of determining the potential of an individual oocyte to develop into a viable embryo, which method comprises establishing the concentration of any one or more of the cytokine markers identified in Table 2 in the follicular fluid of a single follicle from which said oocyte is derived and comparing the concentrations obtained to a reference concentration range that is established by determining the concentration of said one or more markers from one or more follicles containing a fertilisation incompetent oocyte(s), wherein differential levels of said markers from said follicle to be tested compared to the reference range, for example, at least at the 5% significance level, is predictive of an increased potential of said oocyte to develop into a viable embryo following fertilisation.

3. A method of evaluating the fertilisation incompetence of an oocyte derived from a single ovarian follicle in female subject, which method comprises establishing the concentration of any one or more of the cytokine markers identified in Table 2 in the follicular fluid of a single follicle from which said oocyte is derived and comparing the concentrations obtained to a reference concentration range that is established by determining the concentration of said one or more markers from one or more follicles containing a fertilisation competent oocyte(s), wherein differential levels of said markers from said follicle to be tested compared to the reference range, for example, at least at the 5% significance level, is indicative of the fertilisation incompetence of said oocyte(s).

4. A method of evaluating the fertilisation competence of an oocyte derived from a single ovarian follicle in female subject, which method comprises establishing the concentration of any one or more of the cytokine markers identified in Table 2 in the follicular fluid of a single follicle from which said oocyte is derived and comparing the concentrations obtained to a reference concentration range that is established by determining the concentration of said one or more markers from one or more follicles of a fertilisation incompetent female, wherein differential levels of said markers from said follicle to be tested compared to the reference range, for example, at least at the 5% significance level is indicative of the fertilisation competence of said oocyte(s).

5. A method according to any of claims 1 to 4 wherein said differential concentrations of said follicular fluid markers are either elevated or reduced levels compared to those of the reference range.

6. A method of determining the potential of an individual oocyte to develop into a viable embryo, which method comprises establishing the concentration of any one or more of the markers identified in Table 2 in the follicular fluid of a single follicle from which said oocyte is derived and comparing the concentrations obtained to a reference concentration range that is established by determining the concentration of said one or more markers from one or more follicles containing a fertilisation incompetent oocyte and a fertilisation competent oocyte, and wherein the proximity of the concentration ranges obtained from said oocyte to be tested to said marker levels of either said fertilisation incompetent or competent oocyte is predictive of the potential of said oocyte to be tested to develop into a viable embryo.

7. A method according to any of claims 1 to 6 wherein said method comprises determining the concentration of any one of IL-2, IL-4, IL-6, IL-9, IL- 12 (p40), IL- 12 (p70), IL-15, Eotaxin, IFN-α, MIP-I β, RANTES, TNF-α, M-CSF, MIF and SDFlα and wherein elevated levels of any one or more of said markers is predictive of better embryo quality and increased success rates in pregnancy.

8. A method according to any of claims 1 to 6 wherein said method comprises determining the concentration of any one of IL-IO, IL- 17, G-CSF, IL- 18, FGF, GM-CSF, MCP-I, IFN-γ, CTACK, CRP, MIP- lα and VEGF wherein lowered levels of any one or more of said markers is predictive of better embryo quality and increased success rates in pregnancy.

9. A method according to any of claims 1 to 8 wherein a plurality of said markers are utilised to provide an indication of the potential of an oocyte to develop into a viable embryo.

10. A method according to claim 9 wherein said plurality of markers comprises at least three of said markers.

11. A method according to claim 9, wherein said three markers comprise any of i) VEGF, IFNα and MIF, or ii) IL-9, G-CSF or RANTES, or iii) IL2ra, SDF.1.alpha TGF .b

12. A method according to claim 9, wherein said plurality of markers comprises at least four of said markers.

13. A method according to claim 12, wherein said plurality of markers comprise VEGF, IFNα, CTACK and MIF, or IL2ra IL12..p40. SDF.1. alpha TGF.b.

14. A method according to claim 9, wherein said plurality of markers comprises at least five of said markers.

15. A method according to claim 14, wherein said five markers comprise i) IL-7, VEGF, IFNα, CTACK and MIF or ii) CRP IL2ra IL12..p40. SDF.l. alpha TGF.b

16. A method according to claim 9, wherein said plurality of markers comprises at least six of said markers.

17. A method according to claim 16, wherein said six markers comprise any of i) IL- lβ, IL-7 VEGF, IFNα CTACK, and MIF; ii) IL-5, IL-7, VEGF, IFNα, CTACK and MIF; iii) IL-7 VEGF, IFNα CTACK GROα and MIF; or iv) IL-7, IL-9, G-CSF, RANTES, CTACK and SDF.lα, or v) CRP IL2ra IL12..p40. SDF.1. alpha TGF .b

18. A method according to claim 9, wherein said plurality of markers comprises at least seven of said markers.

19. A method according to claim 18, wherein said seven markers comprise any of i) IL-5, IL-7, IL-9, G-CSF, RANTES, CTACK and SDF.lα, ii) IL.7 IL.9 G.CSF RANTES

VEGF CTACK SDF.l. alpha, iii) CRP IFNa IL2ra IL12..p40. GROalpha SDF.1. alpha TGF .b , iv) CRP IL. Ia IL2ra IL12..p40. GROalpha SDF.l. alpha TGF.b, v) CRP IL2ra IL12..p40. IL16 GROalpha SDF.l.alpha TGF.b, or vi) CRP IL2ra IL12..p40. IL18 GROalpha SDF.l.alpha TGF.b.

20. A method according to claim 9, wherein said plurality of markers comprises at least eight of said markers.

21. A method according to claim 20, wherein said eight markers comprise any of i) IL- 1 β, IL-7, MCP.1 , VEGF, IFNα, IL2rα, CTACK and MIF; ii) IL- 1 β, IL-7, VEGF, IFNα,

IL2rα, CTACK, MIF and SDF- lα; iii) IL.5 IL.7 IL.9 G.CSF RANTES VEGF CTACK SDF.l.alpha; iv) IL.7 IL.9 G.CSF MCP.1 RANTES VEGF CTACK SDF.l.alpha; v) IL.7 IL.9 G.CSF MIP. Ia RANTES VEGF CTACK SDF.l.alpha; vi) IL.7 IL.9 G.CSF RANTES VEGF CTACK LIF SDF.l.alpha; or vii) IL.13 FGF G.CSF VEGF CRP IL18 CTACK MIF, viii) CRP IFNa IL2ra IL12..p40. GROalpha M.CSF SDF.1.alpha TGF.b, or ix) CRP IL. Ia IL2ra IL12..p40. IL16 GROalpha SDF.l.alpha TGF.b.

22. A method according to claim 9, wherein said plurality of markers comprises at least nine of said markers.

23. A method according to claim 22, wherein said nine markers comprise any of i) IL- lβ, IL-7, MCP-I, VEGF, IFNα, IL2rα, CTACK, MIF and SDF- lα; ii) IL.5 IL.7 IL.9 G.CSF MCP.1 RANTES VEGF CTACK SDF.1. alpha; or iii) IL.13 FGF G.CSF GM.CSF VEGF CRP IL18 CTACK MIF, or iv) CRP IFNa IL.la IL2ra IL12..p40. GROalpha M.CSF SDF.1. alpha TGF.b.

24. A method according to claim 9, wherein said plurality of markers comprises at least ten of said markers.

25. A method according to claim 24, wherein said ten markers comprise any of i) IL- lβ, IL-7, MCP-I, VEGF, IFNα, IL2rα, CTACK, MIF, SDF- lα TRAIL; or ii) IL.5 IL.7 IL.9 G.CSF GM.CSF MCP.1 RANTES VEGF CTACK SDF.1. alpha; or iii) IL.13 IL.15 FGF G.CSF GM.CSF VEGF CRP IL18 CTACK MIF; or iv) IL.13 FGF G.CSF GM.CSF MCP.l VEGF CRP IL18 CTACK MIF; v) IL.13 FGF G.CSF GM.CSF VEGF CRP IL16 IL18 CTACK MIF, vi) CRP IFNa IL.la IL2ra IL12..p40. IL18 GROalpha M.CSF SDF.l.alpha TGF.b, vii) CRP IFNa IL.la IL2ra IL12..p40. GROalpha M.CSF SDF.l.alpha TNF .b TGF.b, or viii) CRP IL2ra IL12..p40. IL16 IL18 GROalpha MCP3 SDF.l.alpha TRAIL TGF.b .

26. A method according to claim 9, wherein said plurality of markers comprises at least eleven of said markers.

27. A method according to claim 26, wherein said eleven markers comprise any of the following combinations i) IL-lβ, IL-7, IL-8, MCP-I, VEGF, IFNα, IL2rα, CTACK, MIF, SDF- lα and TRAIL; ii) IL-lβ, IL-7, MCP-I, VEGF, CRP, IFNα, IL2rα, CTACK, MIF, SDF- lα and TRAIL; iii) IL-lβ, IL-7, MCP-I, VEGF, IFNα, IL2rα, IL- 16, CTACK, MIF, SDF- lα and TRAIL; iv) IL-lβ, IL-7, MCP-I, VEGF, IFNα, IL2rα, IL- 18, CTACK, MIF, SDF- lα and TRAIL; v) IL-lβ, IL-7, MCP-I, VEGF, IFNα, IL2rα, CTACK, GROα, MIF, SDF-lα and TRAIL; or vi) IL.lb IL.7 MCP.l VEGF IFNa IL12..p40. IL16 IL18 CTACK LIF MIF, or vii) CRP IL2ra IL12..p40. IL16 IL18 GROalpha MCP3 MIF SDF.l.alpha TRAIL TGF.b.

28. A method according to claim 9, wherein said plurality of markers comprises at least twelve of said markers.

29. A method according to claim 28, wherein said twelve markers comprise any of the following combinations i) IL- lβ, IL-7, MCP-I, MIP- lβ, VEGF, CRP, IFNα, IL2rα, CTACK, MIF, SDF- lα and TRAIL; ii) IL- lβ, IL-7, IL-8, MCP-I, VEGF, CRP, IFNα, IL2rα, CTACK, MIF, SDF- 1 α and TRAIL; iii) IL- 1 β, IL-7, MCP- 1 , VEGF, CRP, IFNα, IL- 16, IL2rα, CTACK, MIF, SDF- lα and TRAIL, iv) IL- lβ, IL-7, MCP-I, VEGF, CRP, IFNα, IL2rα, CTACK, GROα, MIF, SDF- lα and TRAIL; or v) IL-5, IL-7, MIP- lβ, RANTES, VEGF, CRP, IFNα, CTACK, M-CSF, MIF, SDF-lα and TRAIL; vi) IL. Ib IL.7 IL.9 G.CSF MCP.l MlP.la RANTES VEGF CRP IL12..p40. CTACK SDF.l.alpha; or vii) IL.7 IL.9 G.CSF MCP.l MIP. Ia PDGF RANTES VEGF IL12..p40. IL18 CTACK SDF.l.alpha; or viii) IL.7 IL.9 G.CSF MCP.l MIP. Ia RANTES VEGF IL3 IL12..p40. IL18 CTACK SDF.l.alpha; ix) IL.7 IL.9 G.CSF MCP.l MlP.la RANTES VEGF IL12..p40. IL18 CTACK GROalpha SDF.l.alpha, x) CRP IFNa IL.la IL2ra IL12..p40. IL18 CTACK GROalpha MCP3 M.CSF SDF.l.alpha TGF.b, xi) CRP IFNa IL.la IL2ra IL12..p40. IL18 CTACK GROalpha M.CSF MIF SDF.l.alpha TGF.b, xii) CRP IFNa IL.la IL2ra IL12..p40. IL18 CTACK GROalpha M.CSF MIG SDF.l.alpha TGF.b, xiii) CRP IL.la IL2ra IL12..p40. IL16 IL18 GROalpha MCP3 MIF SDF.l.alpha TRAIL TGF.b, xiv) CRP IFNa IL2ra IL12..p40. IL16 IL18 GROalpha MCP3 MIF SDF.l.alpha TRAIL TGF.b, or xv) CRP IL2ra IL12..p40. IL16 IL18 GROalpha MCP3 M.CSF MIF SDF.1.alpha TRAIL TGF.b.

30. A method according to claim 9, wherein said plurality of markers comprises at least thirteen of said markers.

31. A method according to claim 30, wherein said thirteen markers comprise any of the following combinations: i) IL.7 IL.9 G.CSF MCP.l MlP.la PDGF RANTES VEGF IL12..p40. IL18 CTACK MCP3 SDF.l.alpha, CRP IFNa IL.la IL2ra ILl 2.. p40. IL16 IL18 CTACK GROalpha M.CSF MIF SDF.l.alpha TGF.b, ii) CRP IFNa IL.la IL2ra IL12..p40. IL18 CTACK GROalpha M.CSF MIF MIG SDF.l.alpha TGF.b, iii) CRP IFNa IL.la IL2ra IL12..p40. CTACK GROalpha LIF M.CSF b.NGF SDF.l.alpha TNF .b TGF.b, iv) CRP IFNa IL2ra IL12..p40. IL16 IL18 CTACK GROalpha MCP3 MIF SDF.l.alpha TRAIL TGF.b.

32. A method according to claim 9, wherein said plurality of markers comprises at least fourteen of said markers.

33. A method according to claim 32, wherein said fourteen markers comprises any of i) IL.5 IL.7 IP.10 MIP. Ib RANTES VEGF CRP IFNa IL 12 p40 CTACK M.CSF MIF SDF.1. alpha TRAIL; ii) IL.5 IL.7 MIP. Ib RANTES VEGF CRP IFNa IL12..p40. CTACK MCP3 M.CSF MIF SDF.1. alpha TRAIL; iii) IL.5 IL.7 IL.8 MIP. Ib RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SDF.1. alpha TRAIL; iv) IL.5 IL.7 IL.12.p70. MlP.lb RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SDF.1. alpha TRAIL; v) IL.5 IL.7 IL.15 MIP. Ib RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SDF.1. alpha TRAIL; vi) IL.5 IL.7 MIP. Ib PDGF RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SDF.1. alpha TRAIL; vii) IL.5 IL.7 MlP.lb RANTES TNF.a VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SDF.1. alpha TRAIL; viii) IL.5 IL.7 MlP.lb RANTES VEGF CRP IFNa IL12..p40. CTACK MCP3 M.CSF MIF SDF.1. alpha TRAIL; ix) IL.5 IL.7 MlP.lb RANTES VEGF CRP IFNa IL12..p40. CTACK M.CSF MIF SCF SDF.1. alpha TRAIL; x) IL.7 IL.9 G.CSF MCP.1 MIP. Ia PDGF RANTES VEGF IL12..p40. IL18 CTACK GROalpha MCP3 SDF.l. alpha; xi) IL.7 IL.9 G.CSF MCP.l MIP. Ia PDGF RANTES VEGF IL12..p40. IL18 CTACK MCP3 SCF SDF.l.alpha, xii) CRP IFNa IL.la IL2ra IL12..p40. IL18 CTACK GROalpha MCP3 M.CSF SCF SDF.l.alpha TRAIL TGF, xiii) CRP IFNa IL.la IL2ra IL12..p40. IL18 CTACK GROalpha LIF M.CSF SDF.l.alpha TNF.b TRAIL TGF .b, or xiv) CRP IFNa IL2ra IL12..p40. IL16 IL18 CTACK GROalpha MCP3 MIF MIG SDF.l.alpha TRAIL TGF .b.

34. A method according to claim 9, wherein said plurality of markers comprises at least fifteen of said markers.

35. A method according to claim 34, wherein said fifteen markers comprise any of i) IL.5 IL.7 MlP.lb RANTES VEGF CRP IFNa IL12 p40 CTACK MCP3 M.CSF MIF SDF.l.alpha TRAIL; ii) IL. Ib IL.7 IL.8 FGF MCP.l VEGF CRP IFNa IL2ra CTACK MCP3 MIF b.NGF SDF.1. alpha TRAIL; iii) CRP IFNa IL.la IL2ra IL12..p40. IL18 CTACK GROalpha MCP3 M.CSF MIF SCF SDF.1. alpha TRAIL TGF.b, iv) CRP IFNa IL.la IL2ra IL12..p40. IL16 IL18 CTACK GROalpha MCP3 M.CSF SCF SDF.l.alpha TRAIL TGF.b; v) CRP IFNa IL.la IL2ra IL12..p40. IL18 CTACK GROalpha MCP3 M.CSF b.NGF SCF SDF.l.alpha TRAIL TGF.b; vi) CRP IFNa IL.la IL2ra IL12..p40. IL16 IL18 CTACK GROalpha M.CSF MIF SCF SDF.l.alpha TNF .b TGF.b, vii) CRP IFNa IL.la IL2ra IL12..p40. IL18 CTACK GROalpha LIF MCP3 M.CSF SDF.l.alpha TNF .b TRAIL TGF.b; viii) CRP IFNa IL.la IL2ra IL12..p40. CTACK GROalpha LIF M.CSF b.NGF SCF SDF.l.alpha TNF .b TRAIL TGF.b; or ix) CRP IL.la IL2ra IL12..p40. IL16 IL18 CTACK GROalpha M.CSF MIG b.NGF SDF.l.alpha TNF.b TRAIL TGF.b.

36. A method according to claim 9, wherein said plurality of markers comprises at least sixteen of said markers.

37. A method according to claim 36 wherein said markers comprise CRP IFNa IL2ra IL12..p40. IL16 IL18 CTACK GROalpha MCP3 M.CSF MIF b.NGF SCF SDF.l.alpha TRAIL TGF.b

38. A method according to claim 9, wherein said plurality of markers comprises at least seventeen of said markers.

39. A method according to claim 38 wherein said markers comprise CRP IFNa IL2ra IL3 IL12..p40. IL16 IL18 CTACK GROalpha MCP3 M.CSF MIF b.NGF SCF SDF.1.alpha TRAIL TGF.b.

40. A method according to claim 9, wherein said plurality of markers comprises at least eighteen of said markers.

41. A method according to claim 40 wherein said markers comprise CRP IFNa IL2ra IL3 IL12..p40. IL16 IL18 CTACK GROalpha LIF MCP3 M.CSF MIF b.NGF SCF SDF.1. alpha TRAIL TGF .b.

42. A method according to claim 9 wherein said plurality of markers comprises at least nineteen of said markers.

43. A method according to claim 42 wherein said markers comprise any of i) CRP IFNa IL.la IL2ra IL3 IL12..p40. IL16 IL18 CTACK GROalpha LIF MCP3 M.CSF MIF b.NGF SCF SDF.l.alpha TRAIL TGF.b, or ii) CRP IFNa IL.la IL2ra IL12..p40. IL16 ILl 8 CTACK GROalpha MCP3 M.CSF MIF MIG b.NGF SCF SDF.l.alpha TNF .b TRAIL TGF.b.

44. A method according to claim 9 wherein said plurality of markers comprises at least twenty of said markers.

45. A method according to claim 44 wherein said markers comprise any of i) CRP IFNa IL.la IL2ra IL3 IL12..p40. IL16 IL18 CTACK GROalpha LIF MCP3 M.CSF MIF MIG b.NGF SCF SDF.l.alpha TRAIL TGF.b; or ii) CRP IFNa IL.la IL2ra IL3 IL12..p40. IL16 IL18 CTACK GROalpha LIF MCP3 M.CSF MIF MIG SCF SDF.l.alpha TNF .b TRAIL TGF.b.

46. A method according to claim 9 wherein said plurality of markers comprises at least twenty one of said markers.

47. A method according to claim 46 wherein said markers comprise CRP IFNa IL.la IL2ra IL3 IL12..p40. IL16 IL18 CTACK GROalpha LIF MCP3 M.CSF MIF MIG b.NGF

SCF SDF.l.alpha TNF.b TRAIL TGF.b

48. A method according to any preceding claim wherein the measured values for the level of the markers in said follicular fluid is subjected to a standardisation protocol to account for contamination of said follicular fluid sample with non- follicular fluid as a result of the harvesting/collection process.

49. A method according to claim 48, wherein markers in said follicular fluid are standardised to a standardising protein that has a molecular weight not exceeding approximately 7OkD.

50 A method according to claim 48 or 49, wherein said standardisation protocol comprises applying the values for the levels of said markers in the follicular fluid to the following equation,

FF = (Zy-Bx)/(y-x) wherein,

Z = Follicular fluid cytokine concentration measured

FF =The corrected concentration of cytokines in follicular fluid

B = Blood cytokine concentration x = concentration in follicular fluid sample of standardising protein y = concentration in blood sample of standardising protein all values being expressed as pg per mg of protein.

51. A method according to any of claims 48 to 50, wherein said standardising protein is von Willebrand factor.

52. A method according to any preceding claim, wherein said concentration of said cytokine is measured by a multiplex immunoassay procedure.

53. A kit for determining the potential of an individual oocyte to develop into a viable embryo following fertilisation, which kit comprises one or more binding agents capable of binding to and recognising one or more follicular fluid markers as set out in Table 2, for contacting a sample of follicular fluid obtained from a single ovarian follicle, means for contacting said sample and said one or more binding agents, the binding agent including an appropriate label or reporter molecule, and means to determine the concentration of said marker.

54. A kit according to claim 53, wherein a plurality of said binding agents capable of recognising the markers according to table 2 are provided.

55. A method of treating infertility in a female subject which method comprises determining the potential of an oocyte from a female subject to develop into a viable embryo in accordance with the method of any one of claims 1 to 52, reducing or increasing activity or expression of any one or more of said markers that are present at differential levels compared to said reference range.

56. A method of preparing a follicular fluid sample(s) for analysis, which method comprises providing a sample of follicular fluid, preferably from a single ovarian follicle, subjecting said sample to a standardisation protocol wherein measurement of the levels of proteins in said sample reflects the level of said protein in said follicular fluid and not from cross contamination.

57. A method according to claim 56, wherein said standardisation protocol comprises standardising said proteins in said sample against a standardising protein.

58. A method according to claim 56 or 57 wherein said standardising protein has a molecular weight not exceeding approximately 7OkD.

59. A method according to any of claims 55 to 58 wherein said standardisation protocol comprises subjecting the measured levels of said proteins to the following protocol

FF = (Zy-Bx)/(y-x) wherein,

Z = Follicular fluid cytokine concentration measured

FF =The corrected concentration of cytokines in follicular fluid B = Blood cytokine concentration x = concentration in follicular fluid sample of standardising protein y = concentration in blood sample of standardising protein all values being expressed as pg per mg of protein

60. A method according to any of claims 57 to 59, wherein said standardisation protein is von Willebrand factor.

61. Apparatus for determining the viability of an embryo, said apparatus comprising an analyzer for identifying the cytokines/markers present in a follicular fluid sample placed therein and their concentration, a computer for receiving the data from said analyser and which is configured by software, either provided on a disk or by receiving an electrical signal by a communications network, into a number of functional modules which cause the computer to process the data received from the analyser to generate an output of cytokine levels, said computer comprising a viability model through which the data from said sample may be processed and against which the data from said sample may be compared to assess the viability of the embryo derived from an oocyte from the follicle from where the follicular fluid of said sample is derived.

62. A method according to any of claims 1 to 52 and which further comprises ranking any viable embryos in terms of their potential to result in a successful pregnancy.