Unit for Human Reproduction, 1st Department of Obstetrics and Gynaecology, Papageorgiou General Hospital, Medical School, Aristotle University of Thessaloniki, Thessaloniki, Greece
Correspondence: EM Kolibianakis, Unit for Human Reproduction, 1st Department of Obstetrics and Gynaecology, Papageorgiou General Hospital, Medical School, Aristotle University of Thessaloniki, Ring Road, Nea Efkarpia, 56429 Thessaloniki, Greece.
To investigate the kinetics of serum vascular endothelial growth factor (VEGF) following gonadotrophin-releasing hormone (GnRH) antagonist administration in the luteal phase in women with established severe early ovarian hyperstimulation syndrome (OHSS).
Pilot observational cohort study.
Private in vitro fertilisation (IVF) Unit.
Twelve IVF women diagnosed with established severe early OHSS 5 days post oocyte retrieval (POR).
Women undergoing IVF diagnosed with severe early OHSS 5 days POR were given 0.25 mg GnRH antagonist for 4 days, from day 5 until and including day 8 POR, combined with elective blastocyst cryopreservation. Serum VEGF was measured from the day of oocyte retrieval until day 11 POR. Ovarian volume, ascites, serum estradiol and progesterone, haematocrit and white blood cells were monitored during the same period.
Main outcome measures
Kinetics of VEGF following luteal GnRH antagonist administration in women with established severe early OHSS.
The concentration of VEGF was highest (390.9 ± 137.4 pg/ml) 5 days POR, coinciding with the day of diagnosis of severe OHSS. There was a significant decline of VEGF on day 7 (302.8 ± 104.9 pg/ml; P = 0.026), day 9 (303.3 ± 148.3 pg/ml; P = 0.007), and day 11 (252.6 ± 182.7 pg/ml; P = 0.010) compared with day 5 POR. This decline was associated with an improvement of ultrasound and laboratory parameters, indicating regression of severe OHSS. All women were managed at an outpatient level.
GnRH antagonist administration in the luteal phase is associated with a significant decline of VEGF and with regression of established severe early OHSS.
Ovarian hyperstimulation syndrome (OHSS) is a serious complication of ovarian stimulation for in vitro fertilisation (IVF) treatment, and may be induced solely when exogenous human chorionic gonadotrophin (hCG) is administered for final oocyte maturation. Early-onset OHSS occurs within 3–7 days after hCG administration,[1, 2] and is classified in mild, moderate and severe forms.[3-5]
Increase of vascular permeability is a critical event in the pathophysiological cascade of OHSS, allowing shift of fluid to the third space, leading to clinical development of OHSS.[6-9]
Vascular endothelial growth factor (VEGF) secreted by multiple corpora lutea of hyperstimulated ovaries seems to be the primary factor responsible for the increase in vascular permeability associated with development of OHSS.[9-11] Indeed, VEGF synthesis is upregulated by hCG,[11, 12] and higher serum VEGF levels have been found in women diagnosed with OHSS.[6, 13-16]
Numerous OHSS prevention methods have been described. Until recently, however, only conservative treatment has been available if severe OHSS occurs, involving correction of fluid and electrolyte imbalance, prevention of thromboembolism and aspiration of the ascitic fluid, combined with prolonged hospitalisation. In this respect, the administration of gonadotrophin-releasing hormone (GnRH) antagonist administration in the luteal phase was proposed as the tertiary management level of severe OHSS.
Increasing evidence suggests that luteal GnRH antagonist administration in women with established severe early OHSS appears to result in quick regression of the syndrome at an outpatient level.[18-22] It was hypothesised that GnRH antagonist may induce luteolysis, thereby drastically reducing the secretion of angiogenic ovarian factors, including VEGF, which have a causal effect in the pathophysiology of the syndrome. However, until today no data regarding the changes of serum VEGF levels following antagonist administration exist.
The current study aimed to assess the kinetics of serum VEGF following luteal GnRH antagonist administration in women diagnosed with severe early OHSS.
Patient population and management
This is a pilot observational cohort study, performed at Eugonia IVF Unit, in which 12 women treated with IVF were diagnosed with severe early OHSS, 5 days post oocyte retrieval (POR). Women were younger than 40 years of age, with polycystic ovaries and were characterised as high risk for severe OHSS because of the presence of at least 18 follicles >11 mm on the day of triggering final oocyte maturation.
A flexible GnRH antagonist protocol or a long GnRH agonist protocol were used for controlled ovarian stimulation, as previously described. All women received 150 IU/day recombinant follicle-stimulating hormone (rFSH, Gonal-F; Merck Serono, Geneva, Switzerland) as starting dose. The dose was adjusted after day 5 of stimulation according to the follicular development, assessed by ultrasound and estradiol levels.
In the preceding cycle, the oral contraceptive pill (Trigynera; Bayer Hellas, Athens, Greece) was administered to all women daily for 21 days, from the second day of spontaneous menses following confirmation of a baseline hormone profile.
Coasting or cortisone administration were not used in any of the women.
Patient information and triggering of final oocyte maturation
Due to the high risk for OHSS, the potential risks of severe OHSS were fully explained to women on the last day of stimulation, and women were informed about the modern prevention methods described in the recent literature.[18-22, 26, 27]
According to our Unit's information policy for women at high risk for OHSS women, the following options were offered: (a) to withhold hCG and cancel the cycle, which is the only way to avoid OHSS, (b) to replace hCG with GnRH agonist for triggering final oocyte maturation for women pretreated with a GnRH antagonist protocol and cryopreserve all embryos, and (c) to proceed at least to oocyte retrieval using low-dose hCG for triggering final oocyte maturation, followed by luteal GnRH antagonist administration and embryo cryopreservation in case severe OHSS developed. If severe OHSS did not occur, as is the case for the majority of high-risk patients (88.7%), women could proceed to embryo transfer.
All women included in the current study opted for triggering final oocyte maturation with hCG because they wished to proceed at least to oocyte retrieval and retain the probability of a fresh embryo transfer.
The study was approved by our institutional ethics review board. An informed consent was obtained from all women included in this study.
Final oocyte maturation was triggered by intramuscular administration of 5000 IU hCG[28, 29] (Pregnyl; MSD, Oss, the Netherlands) when at least three follicles of diameter ≥17 mm were present.
Oocyte retrieval and laboratory procedures
Transvaginal ultrasound-guided oocyte retrieval was performed 36 hours later by double-lumen needle aspiration. Intracytoplasmic sperm injection was performed only in cases with severe male factor infertility or previous fertilisation failure. Embryos were cultured in sequential media (Medicult/Origio, Måløv, Denmark) for 5 days to the blastocyst stage.
Diagnosis of severe OHSS
Women were monitored for development of severe OHSS by measuring haematocrit and white blood cells (WBC) on the day of oocyte retrieval, on day 3 and on day 5 POR, as well as by ultrasound assessment of ovarian size and ascetic fluid on day 3 and on day 5 POR.
Based on previously published criteria,[18, 19] severe early OHSS was diagnosed in the presence of moderate/marked ascites and at least two of the following criteria: enlarged ovaries (>100 mm maximal diameter), haematocrit > 45%, WBC >15 000/mm3, hydrothorax, dyspnoea, oliguria or abnormal liver function tests.
Ascites was classified according to the quantity of fluid accumulation in the peritoneal cavity with the patient in the anti-Trendelenburg position (Table 1), as already described.[18, 19]
Table 1. Classification of ascites
The classification of ascites distinguishes different levels of ascites, depending on the accumulation of ascetic fluid with the woman in the anti-Trendelenburg position.
No presence of fluid
Small amount of fluid, barely detectable by ultrasound in the pouch of Douglas
Increased amount of fluid located in the small pelvis
Large amount of fluid reaching the level of the umbilicus
Significant accumulation of fluid reaching Morrison's pouch
Significant accumulation of fluid up to the level of the diaphragm with/without hydrothorax
Women were diagnosed with established early severe OHSS on day 5 POR, and cryopreservation of all blastocysts was performed on the same day. Concomitantly, 0.25 mg of the GnRH antagonist Cetrorelix (Cetrotide, Merck-Serono, Geneva, Switzerland) was administered daily for 4 days, from day 5 until and including day 8 POR.[19-22]
During the observation period, no woman underwent cortisone or cabergoline administration, or ascetic fluid aspiration. In addition, daily 4500 anti-Xa IU (0.45 ml) tinzaparin sodium (Innohep; LEO Pharmacutica Products Hellas Ltd, Kifisia, Greece) were administered for thromboprophylaxis, from day 5 POR until resolution of the syndrome.
Follow up of patients
Patient follow up after luteal GnRH antagonist administration included measurements of VEGF, haematocrit and WBC as well as ultrasound assessment of ovary size and ascetic fluid on day 7, day 9 and day 11 POR. In addition, serum estradiol and progesterone were also evaluated on the day of oocyte retrieval and on day 3, day 5, day 7, day 9 and day 11 POR.
Ultrasound and laboratory assays
All ultrasound measurements were performed using a 7.5 MHz, 6 MHz or 5 MHz vaginal probe (Sonoline Adara, Siemens), the latter being especially useful in cases of excessively enlarged ovaries. Alternatively, a 3.5 MHz, 2.6 MHz or 5 MHz abdominal probe (Sonoline Adara, Siemens) was used only when visualisation using a vaginal probe was compromised. Ovarian volume was calculated using the prolate ellipsoid formula V = D1 × D2 × D3 × 0.523, where D1, D2 and D3 are the three maximal longitudinal, anteroposterior and transverse diameters, respectively.
Haematocrit and WBC count were determined by flow cytometry (Coulter AC.T diff™ Analyzer, Coulter Corporation, Miami, FL, USA). Coefficient of variation for WBC and red blood cell counts was 3%. Measurements of FSH, luteinising hormone (LH), estradiol and progesterone were performed using an Immulite analyser (DPC, Los Angeles, CA, USA). Analytical sensitivity was 0.1 mIU/ml for FSH assay, 0.1 mIU/ml for LH assay, 15 pg/ml for estradiol assay and 0.2 ng/ml for progesterone assay. Intra-assay and inter-assay precision (coefficients of variation) were 2.6% and 5.8% for FSH, 5.9% and 8.1% for LH, 6.3% and 6.4% for estradiol and 7.9% and 10% for progesterone, respectively.
Levels of VEGF were measured on the day of oocyte retrieval and on days 3, 5, 7, 9 and 11 POR. Serum VEGF was measured using a solid-phase sandwich enzyme-linked immunosorbent assay (DRG Diagnostics, DRG Instruments GmbH, Marburg, Germany). Sensitivity of the assay was 5 pg/ml. Intra- and inter-assay coefficients of variation were 4.7% and 10.8%, respectively. No cross-reactivity was found against human cytokine standards.
The primary outcome was the kinetics of serum VEGF following luteal GnRH antagonist administration in women with established severe early OHSS. Secondary outcomes included changes in haematocrit, WBC count, ovarian volume, ascites, estradiol and progesterone during the same time period.
The outcome measures were subjected to repeated measures analysis of variance followed by post-hoc pairwise comparisons with Bonferroni correction. The frequency distributions of the ascites levels were analysed using the Wilcoxon test. The level of significance was set at 0.05.
Baseline characteristics, ovarian stimulation and embryological data are shown in Table 2.
Table 2. Baseline characteristics, ovarian stimulation and embryological data
2PN, two pronuclei; ICSI, intracytoplasmic sperm injection; TSH, thyroid-stimulating hormone.
Body mass index (kg/m2)
Duration of infertility (years)
Number of previous IVF attempts
Baseline FSH (IU/l)
Baseline LH (IU/l)
Baseline estradiol (pg/ml)
Baseline progesterone (ng/ml)
Baseline TSH (mIU/ml)
Baseline prolactin (ng/ml)
Antral follicle count
Long protocol (n)
Antagonist protocol (n)
Duration of stimulation (days)
Total FSH (IU)
Number of follicles on day of hCG
Estradiol on day of hCG (pg/ml)
Progesterone on day of hCG (ng/ml)
Number of oocytes retrieved
Mature oocytes (in ICSI patients)
Type of fertilisation (IVF/ICSI/IVF+ICSI)
Number of 2PN
The concentration of VEGF was highest (390.9 ± 137.4 pg/ml) on day 5 POR, coinciding with the day of diagnosis of severe OHSS. There was a significant decline of VEGF on day 7 (302.8 ± 104.9 pg/ml; P = 0.026), day 9 (303.3 ± 148.3 pg/ml; P = 0.007), and day 11 (252.6 ± 182.7 pg/ml; P = 0.010) compared to day 5 POR (Figure 1, Table 3). Similarly, ovarian volume, haematocrit, WBC, estradiol and progesterone reached their highest levels on day 5 POR (Table 3), but declined in a statistically significant manner on day 7. This observed decline persisted for all the above parameters until day 11 POR (Figure 1; Table 3).
Table 3. Kinetics of VEGF, haematocrit, WBC count, ovarian volume, women with ascites, estradiol and progesterone during the monitoring period
Values are expressed as mean ± SD unless otherwise stated. Oocyte retrieval was performed on Day 0. Diagnosis of severe OHSS and initiation of antagonist administration was performed on Day 5.
N/P, not performed.
Statistical significance compared to Day 5: *P < 0.05, **P < 0.01, ***P < 0.001.
All women (12/12, 100%) with severe OHSS were diagnosed with significant ascites on day 5 POR. Following GnRH antagonist initiation, a significant proportion of women had low/undetectable ascitic fluid on day 9 (6/12, 50%; P = 0.014) and on day 11 (9/12, 75%; P = 0.0003) compared with day 5 POR (Table 3).
No woman required hospitalisation.
Pearson correlation analysis (Table 4) demonstrated a significant positive correlation between serum VEGF and haematocrit, right/left ovarian volume, proportion of women with ascites, estradiol and progesterone, but not WBC count.
Table 4. Pearson correlation analysis between serum VEGF and various parameters monitored
White blood cell count
Right ovarian volume
Left ovarian volume
Women with ascites
The present pilot study demonstrates that luteal administration of GnRH antagonist for the management of established severe early OHSS is associated with a significant and rapid decline of high serum VEGF levels starting as early as 2 days after GnRH antagonist administration. There was also a progressive simultaneous reduction of ascites, as well as of ovarian volume, haematocrit, WBC, estradiol and progesterone, indicating regression of severe OHSS, confirming previous studies.[18-22] None of the women required hospitalisation.
Strengths and limitations
The present findings provide evidence that GnRH antagonist may lead to regression of severe early OHSS by inducing luteolysis and suppressing VEGF secretion. These are preliminary data on a small number of women with established severe early OHSS. Future controlled studies in a larger patient population are necessary to verify the present findings.
To our knowledge, this is the first study to describe the kinetics of VEGF following luteal GnRH administration in women with established severe early OHSS. Currently, the only existing study in the literature used a rat model of OHSS, and reported reduced VEGF/VEGF receptor expression, serum estradiol and peritoneal vascular permeability following GnRH antagonist administration in the luteal phase.
It has been shown that VEGF levels are significantly higher in women who develop OHSS compared with those who do not and persist at high levels without declining for at least 10–14 days POR, if no intervention is administered in these women.[13, 14, 16] This observation was reported not only in women who developed severe early OHSS, but also, interestingly, in patient groups with milder forms of the syndrome.[13, 14] Therefore, it is likely that in women with severe OHSS, as in the current study, the levels of VEGF might be even higher and require a longer period to subside if no intervention is administered.
The time period required for regression of severe OHSS in the present study appears to be shorter compared with the course of OHSS in the absence of treatment described by previous studies.[13, 31] This rapid severe OHSS resolution, and the associated decline of VEGF levels, might be attributed to the luteal GnRH antagonist administration performed in the current study rather than to the natural course of the syndrome. However, in the absence of a control group in the present pilot study, resolution of the syndrome as part of its natural course cannot be firmly excluded.
No women required hospitalisation in our study, whereas patients diagnosed with severe early OHSS of similar aggression in previous studies required prolonged hospitalisation ranging from 11 to 23 days, occasionally in intensive care units, accompanied by multiple ascites punctures, human albumin administration, and correction of intravascular fluid volume and electrolyte imbalance.[4, 32]
Regarding the kinetics of VEGF, our study is in agreement with the study by Abramov et al., showing that VEGF reached its highest levels on the day of diagnosis of severe OHSS, and its decline was associated with improvement of clinical and laboratory findings and patient symptoms. However, in our study, using luteal GnRH antagonist administration, the time needed for VEGF to decline below 200 pg/ml after severe OHSS diagnosis was 6 days (day 11 of the monitoring period) at an outpatient level, compared with a mean of 9.6 days of inpatient care reported by Abramov et al.
The ovary has been shown to be the primary source of VEGF secretion responsible for development of OHSS.[10, 11] It seems that there is a direct action of GnRH antagonist on the ovary, which may lead to a decline of serum VEGF levels, as well as estradiol, progesterone and ovarian volume, suggesting a luteolytic effect, as has been recently proposed.[18, 19, 21, 22]
The ovary appears to be targeted by GnRH antagonists, as indicated by the presence and function of ovarian GnRH receptors.[33, 34] Indeed, GnRH antagonists have been shown to inhibit the expression of VEGF in human granulosa luteal cell cultures, supporting the hypothesis of direct action of GnRH antagonist on the ovary. It seems unlikely that the GnRH antagonist exerts a luteolytic action by decreasing LH activity, as LH levels are very low during the luteal phase following ovarian stimulation.
Following a general consensus that VEGF plays a key role in the increase of vascular permeability, other approaches have also been proposed with the aim to target VEGF synthesis, bioavailability or downstream signalling, in an attempt to prevent development of OHSS. The dopamine agonist cabergoline, and more recently quinagolide, have been shown to block VEGF signalling by inhibiting VEGF-receptor-2 phosphorylation, leading to reduced incidence and severity of OHSS.[37-41] In addition, reduced VEGF gene and protein expression was demonstrated following coasting in women who were at high-risk for OHSS, partly explaining why coasting may be effective in the prevention of OHSS.
Luteal GnRH antagonist administration is associated with a significant decline of VEGF and a quick regression of established severe early OHSS, supporting the hypothesis that GnRH antagonist may induce luteolysis via a direct action on the ovary. The present data provide further support for low-dose GnRH antagonist administration in the luteal phase as tertiary management of severe OHSS.
Disclosure of interests
Nothing to declare.
Contribution to authorship
GTL participated in the study design, acquisition and analysis of data and writing of the manuscript. EMK participated in the analysis and interpretation of data, writing and revision of the manuscript. IAS participated in the acquisition, analysis and interpretation of data, writing and revision of the manuscript, and performed embryology work. IZZ and GKP participated in the interpretation of data and performed clinical work. TGL had the original conception and general supervision of the study, participated in study design, acquisition, analysis and interpretation of data, writing and revision of the manuscript, and performed clinical work. BCT participated in the interpretation of data and revision of the manuscript and had overall supervision. All authors read and approved the final manuscript.
Details of ethics approval
The study was approved by the ethics review board of Eugonia ART Unit. Approval number: 11/31-1-09.
No external funding was obtained.
The authors wish to thank Ms Georgia Stavropoulou for patient coordination, Mr Miltiades Kyprianou for statistical analysis, Ms Efi Vourvoulia for VEGF and hormone assays, and Ms Ioanna Voulgari and Marina Panagopoulou for data entry.