Cortisol and cortisone in human follicular fluid and serum and the outcome of IVF treatment

Sabina Lewicka1,3, Cornelia von Hagens2, Ute Hettinger1, Klaus Grunwald2, Paul Vecsei*,1, Benno Runnebaum2 and Thomas Rabe2

Departments of 1 Pharmacology and 2 Gynecology, University of Heidelberg, 69120 Heidelberg, Germany

3 To whom correspondence should be addressed at: Institute of Pharmacology, University of Heidelberg, Im Neuenheimer Feld 366, D-69120 Heidelberg, Germany. e-mail: sabina.lewicka{at}urz.uni-heidelberg.de


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
BACKGROUND: The glucocorticoid status of ovarian follicular fluid has been linked to oocyte quality. The aim of this study was to examine whether the concentrations of cortisone and cortisol and their calculated ratios in the follicular fluid and serum samples are predictive of IVF outcome. METHODS: In the prospective study of 387 patients (420 treatment cycles) undergoing IVF treatment the concentrations of cortisone and cortisol were measured with specific assays, and their calculated ratios in the follicular fluid and serum samples obtained after ovarian stimulation and induced ovulation were determined. RESULTS: In 75 patients, treatment resulted in clinical pregnancy and was associated with significantly lower follicular cortisone (24 ± 12 versus 29 ± 16 nmol/l, P < 0.002) and higher cortisol/cortisone ratio (7.24 ± 2.22 versus 6.45 ± 2.17 nmol/l, P < 0.007). In addition, the ratios of serum cortisone and cortisol to follicular cortisone and cortisol were significantly higher in those women who became pregnant. CONCLUSIONS: We propose that the follicular fluid glucocorticoid concentration resulting from the conditions in the circulation and the course of the intrafollicular cortisol–cortisone interconversion appear to play a role in the outcome of IVF.

Keywords: cortisol/cortisone/follicular fluid/IVF


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Glucocorticoid hormones are important regulators of cell function in many tissues. A number of studies have suggested that glucocorticoids affect ovarian physiology. Cortisol exerts direct inhibitory action on ovarian steroidogenesis (Michael et al., 1993aGo) but a positive effect on oocyte maturation (Fateh et al., 1989Go; Jimena et al., 1992Go). During the LH surge an increase in total and free cortisol occurs in the ovarian follicle (Harlow et al., 1997Go; Yong et al., 2000Go) and follicular fluid levels of cortisol at the time of oocyte recovery for IVF have been reported to be higher in follicles yielding mature rather than immature oocytes (Fateh et al., 1989Go).

In an effort to improve pregnancy rates several parameters of ovarian function, including serum and intrafollicular concentrations of cortisol and of binding protein (Fateh et al., 1989Go; Jimena et al., 1992Go), have been examined as potential prognostic indicators for the outcome of IVF. The enzyme 11{beta}-hydroxysteroid dehydrogenase (11{beta}-HSD) catalyses the interconversion of cortisol and the biologically inactive cortisone. Thus, if present in glucocorticoid target tissues, 11{beta}-HSD can alter local concentrations of active glucocorticoid and thereby control access to glucocorticoid receptors (Edwards et al., 1996Go). 11{beta}-HSD is known to exist as at least two distinct isoforms that differ in their actions, co-factor requirements and tissue localization. In the human ovary the expression of 11{beta}-HSD has been shown to be differentially regulated across the menstrual cycle (Michael et al., 1997Go; Tetsuka et al., 1997Go; Yong et al., 2000Go), while in the oocyte only expression of type 1 11{beta}-HSD RNA has been detected (Smith et al., 2000Go). The dehydrogenase activity of this enzyme in follicular aspirates after 3 days of incubation was shown to be inversely related to the successful outcome of IVF (Michael et al., 1993bGo; 1995Go) and was proposed as a potential predictor of conception. Later studies in freshly isolated granulosa-lutein cells did not confirm this relation (Thomas et al., 1998Go). Direct measurement of enzyme activity involves time-, skill- and equipment-demanding culture and subsequent incubation of granulosa cells. Thus we have undertaken studies aiming at the indirect assessment of the effects of 11{beta}-HSD activity by measurement of the cortisol and cortisone concentrations in the follicular fluid. Furthermore, we have compared values obtained in follicular fluid with the respective cortisone and cortisol in serum in order to answer the question of whether serum determinations may replace or complement follicular fluid measurements.

While we were conducting our study two reports appeared that also applied measurements of intrafollicular cortisone and cortisol to the assessment of IVF outcome. Yding Andersen et al. (1999)Go did not support the notion that follicular cortisol or cortisone or their ratio reflect the individual ability of oocyte to fertilize, cleave or implant. Michael et al. (1999)Go reported significantly lower follicular cortisol and cortisone and an elevated intrafollicular ratio of cortisol/cortisone in patients that conceived than in non-conception cycles.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Patient group
Glucocorticoid (cortisol and cortisone) concentrations were examined by radioimmunoassay (RIA) in the follicular fluid and serum from 387 consecutive unselected infertile patients (aged 22–46 years) who were treated between 1995 and 1999 in an IVF programme. The study design and use of patients specimens was approved by the clinical ethics committee of the University of Heidelberg.

Stimulation protocol
Patients were stimulated with pure FSH or hMG (step-down regimen) after pituitary down-regulation (starting on day 20 of the previous cycle; leuproreline 0.5 mg/day s.c. until administration of hCG). The step-down regimen is the standard procedure for women <35 years of age in our IVF unit. In the first stimulation cycle, women <35 years of age received a daily dosage of 225 IU of FSH (Fertinorm HP®, Serono, Unterschleissheim, Germany) or hMG (Humegon®, Organon, Unterschleissheim, Germany; Menogon®, Ferring, Kiel, Germany; Pergonal®, Serono) for the first 3 days, followed by 150 IU/day until ovulation induction. Women >35 years old were injected with a daily dosage of 225 IU until ovulation induction. The gonadotrophin dosage in subsequent cycles was adapted according to the observed ovarian response. After adequate follicular growth (at least three follicles >=18 mm mean diameter measured by transvaginal ultrasound), ovulation was induced with 10 000 IU s.c. hCG (Pregnesin 5000®, Serono) and oocyte retrieval was performed 36 h later. From each patient follicular fluids of the first one to five follicles were collected separately, and after oocyte retrieval they were deep-frozen for hormone analysis. On the day of oocyte retrieval blood was drawn (between 10 and 12 a.m.) for the assessment of hormones in serum.

Measurement of steroids
Follicular fluid and serum samples were stored at –20°C until assayed. Cortisol was determined with specific RIA after breaking protein binding with absolute ethanol using the method of Vecsei et al. (1972)Go. The determination of cortisone in the follicular fluid involved an additional pre-extraction step in order to remove excess of progesterone and 17{alpha}-hydroxyprogesterone present in high amounts in the ovulatory follicle. Briefly, to 100 µl of the follicular fluid were added 3000 dpm tritium-labelled cortisone and 2 ml phosphate buffer before extraction of steroids with 15 ml of 5% dichloromethane in cyclohexane. The organic phase was discarded and the remaining aqueous phase extracted with chloroform and further processed in the same manner as the serum samples.

Similarly, to 20 µl of serum were added 3000 dpm tritiated cortisone, 1.5 ml of phosphate buffer and 5 ml of chloroform. After 1 h mixing, the organic phase was separated, evaporated to dryness and taken up in 500 µl of phosphate buffer. For procedural loss determination a 100 µl aliquot was measured in a scintillation counter. RIA was carried out in duplicate, two 100 µl aliquots were incubated overnight at 4°C with an excess of tritiated cortisone and specific antibody raised against cortisone-3-oxime–bovine serum albumin complex (crossreactivity with cortisol <0.2%, progesterone and 17{alpha}-hydroxyprogesterone <0.02%, estradiol 0%). Intraassay variation was <7% and interassay variation <8%. Unbound steroids were then precipitated with dextrane-coated charcoal and antibody-bound radioactivity was measured in a scintillation counter. Each result was corrected with individually determined procedural loss.

Since our antibodies showed low cross-reactivity to certain steroids present at high concentrations in the follicular fluid (progesterone, 17{alpha}-hydroxyprogesterone, estradiol) we examined whether the relatively simple determination procedure was acceptable. We compared results with respective values after chromatographic separation of nine serum and eight follicular fluid samples (Figure 1). The excellent comparability confirmed sufficient accuracy of the measurements without chromatography.



View larger version (8K):
[in this window]
[in a new window]
 
Figure 1. Comparison of the cortisol values obtained after chromatography (diamonds) or with the simplified method applied in the study (squares), and cortisone after chromatography (circles) or with the simplified method applied in the study (triangles) (see Materials and methods). Samples 1–9 are serum; samples 11–18 are follicular fluid.

 
The determinations in the follicular fluid were carried out in every single follicle. The mean values of several follicles aspirated from the same patient on the day of oocyte retrieval were subsequently calculated and applied for all comparisons. The determination of steroids was carried out prior to the assessment of the outcome of the IVF. Since the IVF Unit does not perform single embryo cultures it was impossible to correlate the values obtained for each follicle with the performance of the oocytes during fertilization and embryo transfer.

Statistical analyses
The data presented here are expressed as mean ± SD. The statistical significance of the difference between the means of the two groups was determined and compared using Student’s t-test.

Univariate analysis of variance was performed with SAS program version 8.1 for Windows (SAS Institute Inc., Cary, NC, USA). Estimation and comparison of the intra- and inter-individual variances was carried out separately in the subgroups of patients (pregnant and not pregnant) by means of one-way analysis of variance (ANOVA) with the individual patients as factor levels. The intra-individual variance was 2.55 [95% confidence interval (CI) 2.08–3.20] in the pregnant group and 1.45 (95% CI 1.31–1.62) in the not pregnant group. In both groups the intra-individual variances were significantly smaller than the inter-individual variances (pregnant: F = 6.9, P < 0.0001; not pregnant: F = 10.32, P < 0.0001). Relying on the univariate analysis of variance showing significantly lower intra-individual than inter-individual variances in all cycles analysed, we considered that use of the calculated mean value of each patient cycle for the analysis of group differences was justified.

Correlations between follicular fluid and respective serum concentrations of glucocorticoids were performed using linear regression analysis. A P-value of < 0.05 was considered statistically significant.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Steroid concentrations were measured in 1221 follicular fluid and 387 serum samples obtained from 387 patients. Two hundred and forty follicular fluid samples originated from 75 women who became pregnant after IVF; 981 originated from 312 women who did not.

Cortisone in the follicular fluid was significantly lower in conception cycles (P < 0.002). Cortisol was not statistically different between the groups. Cortisol/cortisone ratio was significantly higher in follicles from cycles resulting in pregnancy (P < 0.007).

There was no statistical difference in cortisol, cortisone and cortisol/cortisone ratios in serum from women in whom treatment resulted in pregnancy and those who failed to become pregnant.

Concentrations of cortisone and cortisol in follicular fluid was approximately half those concentrations in serum. The respective cortisol/cortisone ratios in follicular fluid and in serum were not different. The ratios of serum cortisone/follicular cortisone, serum cortisol /follicular cortisol and serum cortisol/follicular cortisone were significantly higher in pregnant cycles (P < 0.0006, P < 0.04 and P < 0.0002, respectively) (Table I).


View this table:
[in this window]
[in a new window]
 
Table I. Concentrations of cortisone, cortisol and their calculated ratios in the follicular fluid (mean values from several follicles in one cycle) and serum samples
 
Serum cortisol and cortisone were significantly correlated with the respective follicular values in all patients, and pregnant and not pregnant groups separately. Cortisol/cortisone ratios in serum correlated significantly with respective follicular fluid ratios only in the not pregnant group (Table II).


View this table:
[in this window]
[in a new window]
 
Table II. Correlation between respective serum and follicular fluid concentrations of cortisone, cortisol and their calculated ratios
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The present study confirms that glucocorticoids, cortisol and cortisone, are present in the follicular fluid of the follicle recovered after stimulated cycles. Mean values of the follicular fluid cortisone were significantly lower and cortisol/cortisone ratio higher in patients, who became pregnant after IVF than in those who did not. Cortisone and cortisol in the follicular fluid are approximately half, and correlate significantly with the respective serum values.

Since there is no 21-hydroxylase and 11{beta}-hydroxylase activity in the human ovary (Omura and Morohashi, 1995Go), glucocorticoids in follicular fluid are derived from circulation. Thus the differences between respective serum and follicular fluid values in single patients and serum/intrafollicular glucocorticoids from patient to patient result from the different blood–follicle penetration and activity of the transforming enzymes in the ovary. Intrafollicular cortisone and cortisol were lower in our patients than in those studied by Michael et al. (1999)Go and Yong et al. (2000)Go. Yding Andersen et al. (1999)Go reported lower cortisol and higher cortisone concentrations than we found. These differences may be attributed to a different patient population, assay methods and lower number of patients. Serum concentrations of cortisone and cortisol (Table I) agree with all reports to date for these steroids except that of Yong et al. (2000)Go, whose values were 2-fold higher. The higher concentrations of cortisol and cortisone in serum than in follicular fluid confirm earlier reports (Yding Andersen et al., 1999Go; Yong et al., 2000Go). In contrast to Yong et al. (2000)Go, we found correlation between follicular fluid and serum concentrations of cortisone and cortisol.

The higher ratios of cortisol/cortisone in the follicular fluid of patients who became pregnant than in those who did not (Table I) confirm the data reported by Michael et al. (1999)Go and contradict those reported by Yding Andersen et al. (1999)Go, who found no difference in markedly lower number of studied patients and samples.

The cortisol/cortisone ratio reflects the activity of 11{beta}-HSD catalysing interconversion of both steroids and modulating local glucocorticoid activity. In the early phase of the follicular development excess of cortisol may disturb the FSH-stimulated granulosa cell function, so in this period the oxidation of cortisol via type 2 11{beta}-HSD is critical. As ovulation approaches, intrafollicular cortisol concentration rises reflecting the switch from type 2 11{beta}-HSD to type 1 11{beta}-HSD activity, inducible in human granulosa cells by HCG administration (Tetsuka et al., 1997Go; Yong et al., 2000Go). In the final stage of oocyte maturation cortisol plays a beneficial role and ovarian type 1 11{beta}-HSD assures its adequate concentration, while the presence of type 2 11{beta}-HSD may result in follicular immaturity. Very recently the physiological importance of high levels of follicular free cortisol in the establishment and function of corpus luteum was postulated (Yding Andersen, 2002Go). We have observed significantly lower follicular cortisone and higher follicular cortisol/cortisone ratios in patients who became pregnant, while there was no difference in serum values (Table I). At the same time the ratios of serum cortisone/follicular cortisone, serum cortisol/follicular cortisol and serum cortisol/follicular cortisone were significantly higher in the pregnant than in the not pregnant group (Table I). The lower glucocorticoid concentrations in the follicle in women achieving pregnancy could be due to their higher serum cortisol binding protein, as reported by Yding Andersen (1990)Go. It seems that patients combining higher ovarian type 1 11{beta}-HSD activity transforming cortisone to cortisol with lower extraction of glucocorticoids from the circulation to the follicular fluid have a better chance of pregnancy. However, despite statistically significant differences in follicular cortisone, follicular cortisol/cortisone ratio and serum glucocorticoids/follicular glucocorticoids ratios between pregnant and non-pregnant patients, absolute values overlap quite considerably and practically their predictive value is currently limited.

The shortcoming of this study is the lack of glucocorticoid values of the single follicles giving rise to transferred embryos resulting or not in pregnancy. Since we are not able to follow the fate of the single follicles we were forced to approximate the values by calculation of the mean for each cycle, which, although it was statistically justified, likely reduced the significance of the difference. In conclusion, the different relationships between steroids and their ratios in women who conceived after IVF and those who did not support previous reports postulating a link between activity of the ovarian 11{beta}-HSD and the outcome of IVF.

For the first time this notion is reinforced by the measurements of glucocorticoid concentrations in follicular fluid samples, thus in the closest proximity of an oocyte, and in the corresponding serum samples in a large number of patients.

Another important finding of this study is that measurements of follicular cortisone, not cortisol, and ratios of serum glucocorticoids to follicular glucocorticoids provide a good parameter for the assessment of the IVF outcome.


    Acknowledgements
 
We are grateful to Dr Maria Pritsch from the Institute of Medical Biometry and Informatics, University of Heidelberg, for help in the statistical evaluation and interpretation of results. This study was supported by the Research Fund of the Medical Department of the University of Heidelberg (Project 172/95).


    FOOTNOTES
 
* Dr Paul Vecsei died on July 7, 2002. Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Edwards, C.R.W., Benediktsson, R., Lindsay, R.S. and Seckl, J.R. (1996) 11{beta}-Hydroxysteroid dehydrogenases: Key enzymes in determining tissue-specific glucocorticoid effects. Steroids, 61, 263–269.[CrossRef][ISI][Medline]

Fateh, M., Ben-Rafael, Z., Benadiva, C.A., Mastroianni, L., Jr and Flickinger, G.L. (1989) Cortisol levels in human follicular fluid. Fertil. Steril., 51, 538–541.[ISI][Medline]

Harlow, C.R., Jenkins, J.M. and Winston, R.M.L. (1997) Increased follicular fluid total and free cortisol levels during the luteinizing hormone surge. Fertil. Steril., 68, 48–53.[CrossRef][ISI][Medline]

Jimena, P., Castilla, J.A., Peran, F., Ramirez, J.P., Vergara, F., Jr, Molina, R., Vergara, F. and Herruzo, A. (1992) Adrenal hormones in human follicular fluid. Acta Endocrinol. (Copenh), 127, 403–406.[ISI][Medline]

Michael, A.E., Pester, L.A., Curtis, P., Shaw, R.W., Edwards, C.R. and Cooke, B.A. (1993a) Direct inhibition of ovarian steroidogenesis by cortisol and the modulatory role of 11{beta}-hydroxysteroid dehydrogenase. Clin. Endocrinol., 38, 641–644.[ISI][Medline]

Michael, A.E., Gregory, L., Walker, S.M., Antoniw, J.W., Shaw, R.W., Edwards, C.R. and Cooke, B.A. (1993b) Ovarian 11{beta}-hydroxysteroid dehydrogenase: potential predictor of conception by in-vitro fertilisation and embryo transfer. Lancet, 342, 711–712.[ISI][Medline]

Michael, A.E., Gregory,L., Piercy, E.C., Walker, S.M., Shaw, R.W. and Cooke, B.A. (1995) Ovarian 11{beta}-hydroxysteroid dehydrogenase activity is inversely related to the outcome of in vitro fertilization-embryo transfer treatment cycles. Fertil. Steril., 64, 590–598.[ISI][Medline]

Michael, A.E., Evagelatou, M., Norgate, D.P., Clarke, R.J., Antoniw, J.W., Stedman, B.A., Brennan, A., Welsby, R., Bujalska, I., Stewart, P.M. et al. (1997) Isoforms of 11{beta}-hydroxysteroid dehydrogenase in human granulosa-lutein cells. Mol. Cell. Endocrinol., 132, 43–52.[CrossRef][ISI][Medline]

Michael, A.E., Collins, T.D., Norgate, D.P., Gregory, L., Wood, P.J. and Cooke, B.A. (1999) Relationship between ovarian cortisol:cortisone ratios and the clinical outcome of in vitro fertilization and embryo transfer (IVF-ET). Clin. Endocrinol., 51, 535–540.[CrossRef][ISI][Medline]

Omura, T. and Morohashi, K. (1995). Gene regulation of steroidogenesis. J. Steroid Biochem. Mol. Biol., 53, 19–25.[CrossRef][ISI][Medline]

Smith, M.P., Mathur, R.S., Keay, S.D., Hall, L., Hull, M.G. and Jenkins, J.M. (2000) Periovulatory human oocytes, cumulus cells, and ovarian leukocytes express type 1 but not type 2 11{beta}-hydroxysteroid dehydrogenase RNA. Fertil. Steril., 73, 825–830.[CrossRef][ISI][Medline]

Tetsuka, M., Thomas, F.J., Thomas, M.J., Anderson, R.A., Mason, J.I. and Hillier, S.G. (1997) Differential expression of messenger ribonucleic acids encoding 11{beta}-hydroxysteroid dehydrogenase types 1 and 2 in human granulosa cells. J. Clin. Endocrinol. Metab., 82, 2006–2009.[ISI][Medline]

Thomas, F.J., Thomas, M.J., Tetsuka, M., Mason, J.I. and Hillier, S.G. (1998) Corticosteroid metabolism in human granulosa-lutein cells. Clin. Endocrinol., 48, 509–513.[CrossRef][ISI][Medline]

Vecsei, P., Penke, B., Katzy, R. and Baek, L. (1972) Radioimmunological determination of plasma cortisol. Experientia, 28, 1104–1105.[ISI][Medline]

Yding Andersen, C. (1990) Levels of steroid-binding proteins and steroids in human preovulatory follicle fluid and serum as predictors of success in in vitro fertilization-embryo transfer treatment. J. Clin. Endocrinol. Metab., 71, 1375–1381.[Abstract]

Yding Andersen, C. (2002) Possible new mechanism of cortisol action in female reproductive organs: physiological implications of the free hormone hypothesis. J. Endocrinol., 173, 2311–2317.

Yding Andersen, C., Morineau G., Fukuda, M., Westergaard, L.G., Ingerslev, H.J., Fiet, J. and Byskov, A.G. (1999) Assessment of the follicular cortisol:cortisone ratio. Hum. Reprod., 14, 1562–1568.[Medline]

Yong, P.Y., Thong, K.J., Andrew, R., Walker, B.R. and Hillier, S.G. (2000) Development-related increase in cortisol biosynthesis by human granulosa cells. J. Clin. Endocrinol. Metab., 85, 4728–4733.[Abstract/Free Full Text]

Submitted on September 13, 2002; resubmitted on February 24, 2003; accepted on May 7, 2003.