Ultrasound and hematological changes during early luteal phase in women at high risk for developing ovarian hyperstimulation syndrome

Ultrasound Obstet Gynecol 2018; 51: 126 133 Published online in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/uog.18949 Ultrasound and hematological changes during early luteal phase in women at high risk for developing ovarian hyperstimulation syndrome G. T. LAINAS 1,T.G.LAINAS 1, C. A. VENETIS 2,I.A.SFONTOURIS 1,I.Z.ZORZOVILIS 1, E. ALEXOPOULOU 3, B. C. TARLATZIS 4 ande.m.kolibianakis 4 1 Eugonia Assisted Reproduction Unit, Athens, Greece; 2 Department of Women s and Children s Health, St George Hospital, School of Women s and Children s Health, University of New South Wales, Kogarah, NSW, Australia; 3 2 nd Department of Radiology, Attikon University Hospital, National and Kapodistrian University of Athens, Athens, Greece; 4 Unit for Human Reproduction, 1 st Department of Obstetrics & Gynaecology, Papageorgiou General Hospital, Medical School, Aristotle University of Thessaloniki, Thessaloniki, Greece KEYWORDS: white blood cells ascites; classification; criteria; hematocrit; high risk; OHSS; OHSS diagnosis; ovarian diameter; pathophysiology; ABSTRACT Objective To assess ultrasound and hematological changes during the early luteal phase following triggering of final oocyte maturation with human chorionic gonadotropin (hcg) in women at high risk for developing ovarian hyperstimulation syndrome (OHSS). Methods This was a retrospective cohort study of 319 women undergoing in-vitro fertilization who were at high risk for OHSS following administration of hcg for the triggering of final oocyte maturation. Patients were treated with a gonadotropin-releasing hormone agonist or antagonist protocol and were monitored for 5 days post-oocyte retrieval (early luteal phase). Severe OHSS was diagnosed in the presence of at least moderate ascites and two or more of the following: maximum ovarian diameter (MOD) > 100 mm, hematocrit (Ht) > 45%, white blood cell count (WBC) > 15 000/mm 3, hydrothorax, dyspnea and oliguria. Outcome measures included change in Ht, ascitesgrade,wbcandmod,aswellastheassociation between these changes during the early luteal phase. Results Ascites grade, Ht and WBC increased significantly (P 0.001) during the early luteal phase, both in patients who developed and in those who did not develop severe early OHSS. MOD increased significantly (P = 0.001) only in patients who developed severe early OHSS. On multivariable analysis, both time following oocyte retrieval and whether severe early OHSS developed were significantly associated with ascites grade, Ht, WBC and MOD; furthermore, there was also a significant interaction between time and development of severe early OHSS for all four variables (P 0.001). Conclusions In women at high risk of OHSS, ascites grade, Ht and WBC significantly increased with time over the 5-day observation period, in line with the pathophysiology of the syndrome. Our data support the use of MOD in the diagnosis of severe early OHSS, and provide novel evidence for the role of change in Ht as a patient-specific hemoconcentration marker during development of OHSS. Copyright 2017 ISUOG. Published by John Wiley & Sons Ltd. INTRODUCTION Ovarian hyperstimulation syndrome (OHSS) is a complication of ovarian stimulation during in-vitro fertilization (IVF). OHSS is triggered solely by human chorionic gonadotropin (hcg) 1,2 and is attributed mainly to excessive ovarian secretion of vascular endothelial growth factor and other angiogenic factors, which increase vascular permeability and cause leakage of fluid into the third space 3. Severe OHSS is the most serious form of the syndrome and may become critical or life-threatening 4 6. In an attempt to eliminate OHSS, replacing hcg with gonadotropin-releasing hormone (GnRH) agonist to trigger final oocyte maturation, combined with a freeze-all strategy, has been proposed for high-risk women pretreated with GnRH antagonist protocols (OHSS-free clinic) 7,8. However, despite the increasing use of a GnRH agonist trigger (5.2 36.1% of cases worldwide), hcg administration remains the most popular triggering option 9,10. There have been several proposals for diagnosing and classifying OHSS since its initial description Correspondence to: Dr G. Lainas, Eugonia Assisted Reproduction Unit, 7 Ventiri Street, 11528 Athens, Greece (e-mail: g.lainas@yahoo.co.uk) Accepted: 30 October 2017 Copyright 2017 ISUOG. Published by John Wiley & Sons Ltd. ORIGINAL PAPER

Changes in OHSS in early luteal phase 127 Potentially eligible for inclusion (n = 416) Examined for eligibility (n = 416) Confirmed eligible (n = 355) Included in study (n = 355) Complete follow-up (n = 350) Analyzed (n = 319) Ineligible (n = 61) Agonist trigger (n = 61) Withholding hcg (n = 0) Incomplete follow-up (n = 5) Missing data (n = 31) METHODS Study design and patient population This was a retrospective cohort study of patients who underwent IVF treatment between January 2009 and December 2012 at Eugonia ART Unit in Athens. We included 319 patients at high risk for OHSS following hcg triggering of final oocyte maturation (Figure 1). High risk for OHSS was defined as the presence of at least 19 follicles 11 mm in diameter on the day of hcg triggering 27,28. Patients could enter the study only once and only those with a complete set of data for all parameters were included in the analysis. Patients receiving GnRH agonist trigger or withholding hcg were excluded from the study to reduce bias. No other OHSS-reducing intervention, such as coasting, cabergoline administration or ascites paracentesis, was used in any of the patients. It should be noted that a proportion (32%) of the women reported here (103 women: 21 with and 82 without severe early OHSS) had also been included in a previous study evaluating the efficiency of luteal GnRH antagonist administration after day 5 of in-vitro culture (i.e. after completion of the monitoring period of the present study) for the management of severe early OHSS 18. The institutional review board of the Eugonia ART Unit provided ethics approval for this study (approval number 02/28-09-13). Figure 1 Flowchart of study participants. in 1967 1,4,6,11 21. These classification systems have widespread acceptance and have formed the basis of various official guidelines (Royal College of Obstetricians and Gynaecologists 22 and American Society for Reproductive Medicine 6 ). However, all of them are characterized by an empirical approach and lack of statistical analysis, include arbitrarily proposed criteria and, in fact, represent consecutive reorganizations and modifications of a single original report, by Rabau et al. 11, of clinical observations in 14 women with OHSS. There are several studies available on the prediction and prevention of OHSS 6,17,21,23 25. However, a major contributing factor that may explain the diversity in OHSS classification systems is the absence from all of them of a detailed description of the ultrasound and hematological changes associated with the evolution of the syndrome until its diagnosis. Limited information regarding changes in hematocrit (Ht), hemoglobin and ascites is available from a small study of high-risk patients 26. The aim of the present study, therefore, was to describe, in women at high risk for OHSS, changes in ascites grade, Ht, white blood cell count (WBC) and maximum ovarian diameter (MOD) during the early luteal phase (until day 5 post-oocyte retrieval), following triggering of final oocyte maturation with hcg. Ovarian stimulation and triggering of final oocyte maturation Patients underwent ovarian stimulation for IVF/ intracytoplasmic sperm injection (ICSI) using either a long GnRH agonist down-regulation protocol or a flexible GnRH antagonist protocol, as described previously 29. Ovarian stimulation was performed with a starting dose of 150 IU/day of recombinant follicle-stimulating hormone (Gonal-F; Merck Serono, Geneva, Switzerland). This dose was adjusted after day 5 of stimulation, depending on the ovarian response, as assessed by estradiol levels and ultrasound. The presence of three or more follicles with a mean diameter 17 mm was the criterion for triggering with hcg; once this was observed, the following options were offered to all patients at high risk for OHSS: (a) withholding of hcg and cancelation of the cycle; (b) triggering of final oocyte maturation using GnRH agonist instead of hcg and cryopreservation of all embryos (only if patients had been treated with a GnRH antagonist protocol); or (c) proceeding to oocyte retrieval using low dose of hcg (5000 IU) for triggering of final oocyte maturation 30. The latter was followed either by embryo transfer (ET), if severe OHSS did not develop by day 5 post-oocyte retrieval, or by luteal GnRH antagonist administration and embryo cryopreservation, if severe OHSS had developed by day 5 post-oocyte retrieval 18. Only patients who opted for triggering of final oocyte maturation with hcg were eligible for inclusion in the current study. They did

128 Lainas et al. Table 1 Classification of ascites 18,19,31 Grade Ascites Fluid characteristics 1 None Absence of fluid 2 Low Small amount of fluid, barely detectable by ultrasound, in pouch of Douglas 3 Moderate Increased amount of fluid, located in small pelvis 4 Marked Large amount of fluid, reaching level of umbilicus 5 Massive Significant accumulation of fluid, reaching Morrison s pouch 6 Tense Significant accumulation of fluid, up to level of diaphragm, with/without hydrothorax so since they wished to proceed at least to oocyte retrieval and to retain the possibility of a fresh embryo transfer. Transvaginal ultrasound-guided oocyte retrieval was performed 36 h following hcg administration by double-lumen needle aspiration. ICSI was performed only in cases with severe male factors or previous fertilization failure. Embryos were cultured for 5 days in sequential media (Medicult/Origio, Denmark) to the blastocyst stage. Diagnosis of severe early OHSS Severe early OHSS was diagnosed in the presence of at least moderate (Grade 3) ascites and two or more of the following: Ht > 45%, WBC > 15 000/mm 3, hydrothorax, dyspnea and oliguria (all as proposed by Navot et al. 4, who reorganized previous classifications by Rabau et al. 11, Schenker and Weinstein 12 and Golan et al. 13 ), and enlarged ovaries (MOD > 100 mm, a modification of the 5 12-cm range proposed by Golan et al. 13 ). Dyspnea was assessed subjectively by the patients. Oliguria was defined as urine output < 500 ml/24 h, and was measured by the patients. Ascites was classified into six grades (1, no ascites; 2, low amount of ascites; 3, moderate amount; 4, marked amount; 5, massive amount; 6, tense) depending on the quantity of fluid accumulation in relation to anatomical regions in the peritoneal cavity (Table 1), with the patient in the anti-trendelenburgh position, as described previously 18,19,31. Patients diagnosed with severe early OHSS were managed on an outpatient basis by administration of GnRH antagonist and cryopreservation of all blastocysts 18,31,32. GnRH antagonist was administered for 3 5 days, starting on the day of diagnosis. Patient follow-up All patients were monitored for the development of severe early OHSS on the day of oocyte retrieval (day 0), 3 days post-oocyte retrieval (day 3) and 5 days post-oocyte retrieval (day 5), according to the center s standard policy for all high-risk patients. The present analysis includes the first 5 days post-oocyte retrieval (day 0 until day 5). Administration of antagonist from day 5 onwards in the 43 women with severe early OHSS may be considered to introduce bias. However, by restricting the study period to the first 5 days, luteal GnRH antagonist administration is unlikely to have affected hematological and ultrasound parameters. Ultrasound and laboratory assays Ultrasound assessment of ascites was performed on the day of oocyte retrieval (day 0), 3 days post-oocyte retrieval (day 3) and 5 days post-oocyte retrieval (day 5). Ovarian size was assessed by transvaginal sonography 3 days and 5 days post-oocyte retrieval. The operator scanned the ovaries in at least three different planes and identified the plane with the largest diameter for each ovary. These two diameters were compared and only the largest diameter of the two ovaries was recorded as MOD. Ultrasound measurements of ascetic fluid and MOD were performed using a 7.5-, 6- or 5-MHz vaginal probe (Sonoline Adara, Siemens, Erlangen, Germany). Measurements of Ht and WBC were performed on the day of oocyte retrieval (day 0), 3 days post-oocyte retrieval (day 3) and 5 days post-oocyte retrieval (day 5). Ht and WBC were determined by flow cytometry using a Coulter A C.T diff Analyzer (Coulter Corporation, Miami, FL, USA). The coefficient of variation, specifying imprecision limits for WBC and red blood cell (RBC) counts was 3%. Ht was computed from the mean cell volume of erythrocytes (MCV): Ht (%) = RBC MCV/10. Outcome measures Outcome measures included the changes in absolute values of Ht, ascites grade, WBC and MOD, as well as the associations between these changes during the early luteal phase following hcg triggering of final oocyte maturation. Statistical analysis Multivariable analysis was performed using generalized estimating equations, with ascites grade, Ht, WBC or MOD as the dependent variable and time (day 0, day 3, day 5) and development or not of severe early OHSS as independent variables. Binary variables are expressed as proportions (95% CI), while continuous variables are expressed as mean (95% CI). A paired t-test was used to compare differences in ascites grade, Ht and WBC, between day 0, day 3 and day 5 post-oocyte retrieval. Pairwise correlations were calculated by estimating Pearson s correlation coefficient, while Bonferroni adjustment was used to calculate significance levels. Statistical significance was set at P 0.05. RESULTS A flow diagram reporting patient numbers during the process of inclusion in the study is shown in Figure 1.

Changes in OHSS in early luteal phase 129 All 319 high-risk patients included in the analysis had a complete set of data for all variables. Severe early OHSS was diagnosed in 43 (13.5%; 95% CI, 10.2 17.7) women on day 5 post-oocyte retrieval. No patient developed severe early OHSS prior to day 5. Details on how the diagnosis of severe early OHSS was made in each patient are shown in Table S1. The characteristics of patients included in the analysis are shown in Table 2 and the effect of parameters found to have a statistically significant association with development of severe early OHSS is shown in Table 3. Ascites Proportions of patients who developed OHSS and those who did not, according to grade of ascites, for each day of monitoring, are shown in Table 4. Mean values of ascites grade in all patients, and in subgroups according Table 2 Patient, ovarian stimulation and embryological characteristics for study population of women undergoing in-vitro fertilization (IVF) who were at high risk for ovarian hyperstimulation syndrome following administration of human chorionic gonadotropin (hcg) for triggering of final oocyte maturation Characteristic Value Age (years) 33 (32 33) BMI (kg/m 2 ) 22.6 (22.1 23.2) Duration of infertility (years) 3 (2 3) Number of previous IVF attempts 0 (0 0) Dose of gonadotropins (IU) 1775 (1675 1837.6) Antral follicle count 22 (21 22) Basal LH (IU/L) 5.8 (5.6 6.0) Basal FSH (IU/L) 7.0 (6.7 7.4) Basal estradiol (pg/ml) 30 (29 31) Basal progesterone (ng/ml) 0.5 (0.4 0.5) Estradiol on hcg day (pg/ml) 2909 (2698.9 3022.1) Progesterone on hcg day (ng/ml) 0.95 (0.9 1.0) Days of stimulation 11 (11 11) Number of follicles on hcg day 29 (28 29) Number of oocytes retrieved 24 (23 25) Number of 2PN oocytes 14 (14 15) Antagonist protocol 148 (46.4) Agonist protocol 171 (53.6) Data are presented as median (95% CI) or n (%). 2PN, two pronuclei; BMI, body mass index; FSH, follicle stimulating hormone; LH, luteinizing hormone. to whether they developed severe early OHSS, are shown in Table 5. There was a statistically significant increase (P < 0.001) in grade of ascites during the early luteal phase (from day 0 to day 3 to day 5) in all patients at high risk for OHSS. This increase was also observed when patients were categorized according to whether they in fact developed severe early OHSS. Changes in ascites grade (casc) between day 3 and day 0 (casc D3 D0 ), between day 5 and day 0 (casc D5 D0 )and between day 5 and day 3 (casc D5 D3 ) are shown in Table S2. casc D5 D0 was significantly greater than casc D3 D0 andthancasc D5 D3, in all patients as well as in both patients who developed severe early OHSS and those who did not. casc D5 D3 was significantly smaller than casc D3 D0 in patients who developed severe early OHSS. However, this was not the case in patients who did not develop severe early OHSS (Table S2). casc D3 D0 and casc D5 D0 were significantly higher in patients who did vs those who did not develop OHSS (P < 0.0001), while such a difference was not present for casc D5 D3 (P = 0.54). Both time following oocyte retrieval and whether severe early OHSS developed were associated with ascites grade on multivariable analysis, and there was a significant interaction between time following oocyte retrieval and development of severe early OHSS (Table S3). Patients who developed severe early OHSS showed a higher rate of increase in ascites grade per day (coefficient, +0.326) compared with patients who did not eventually develop severe early OHSS (Table S3). The distribution of the severity of ascites on days 0, 3 and 5 in all patients, as well as in subgroups of those who developed and those who did not develop severe early OHSS, is shown in Figure S1. On day 0, ascites was not detectable (Grade 1) in the majority of high-risk patients, both those who subsequently did not develop (97.1%; 95% CI, 94.5 98.7%) and those who developed (95.4%; 95% CI, 84.2 99.4%) severe early OHSS. On day 5, ascites Grade 5 and Grade 6 were present in 9.3% (95% CI, 3.7 21.6%) and 16.3% (95% CI, 8.1 30.0%), respectively, of women with non-severe early OHSS. Ascites Grades 5 and 6 were absent in women with non-severe early OHSS during the observation period (0%; 95% CI, 0 1.4%). The direction of change in ascites grade for all high-risk patients, as well as in subgroups according to whether they developed severe early OHSS, is shown in Figure S2. Table 3 Logistic regression for parameters found to have statistically significant association with development of severe early ovarian hyperstimulation syndrome (OHSS) in women undergoing in-vitro fertilization who were at high risk for OHSS following administration of human chorionic gonadotropin (hcg) for triggering of final oocyte maturation Parameter Odds ratio (95% CI) Robust SE z P Dose of gonadotropins 1.00 (0.99 1.00) 0.00 0.29 0.775 Basal FSH 0.81 (0.63 1.05) 0.11 1.61 0.107 Estradiol on hcg day 1.00 (1.00 1.13) 0.00 2.24 0.025* Number of oocytes retrieved 1.08 (1.04 1.13) 0.24 3.57 < 0.0001* GnRH analogs (agonist protocol) 0.49 (0.23 1.04) 0.19 1.86 0.062 Model P < 0.0001. Pseudo R 2 = 0.154. *Statistically significant. FSH, follicle stimulating hormone; GnRH, gonadotropin-releasing hormone; SE, standard error.

130 Lainas et al. Table 4 Proportions of women undergoing in-vitro fertilization who were at high risk for ovarian hyperstimulation syndrome (OHSS) following administration of human chorionic gonadotropin for triggering of final oocyte maturation, and who did or did not develop severe OHSS, according to grade of ascites, for each day of monitoring following oocyte retrieval Grade of ascites 1 2 3 4 5 6 P Day 0 0.264 Non-severe early OHSS 268 (97.10) 4 (1.45) 4 (1.45) 0 0 0 Severe early OHSS 41 (95.35) 2 (4.65) 0 0 0 0 Day 3 < 0.0001 Non-severe early OHSS 93 (33.7) 143 (51.81) 37 (13.41) 3 (1.09) 0 0 Severe early OHSS 0 10 (23.26) 12 (27.91) 21 (48.84) 0 0 Day 5 < 0.0001 Non-severe early OHSS 31 (11.23) 102 (36.96) 51 (18.48) 92 (33.33) 0 0 Severe early OHSS 0 0 6 (13.95) 26 (60.47) 4 (9.30) 7 (16.28) Data are presented as n (%). Table 5 Variables associated with development of ovarian hyperstimulation syndrome (OHSS) during early luteal phase (days 0, 3, 5 following oocyte retrieval) in women undergoing in-vitro fertilization who were at high risk for OHSS following administration of human chorionic gonadotropin for triggering of final oocyte maturation Variable n Day 0 Day 3 Day 5 P* Ascites grade All patients 319 1.04 (1.02 1.07) 2.01 (1.91 2.11) 2.95 (2.82 3.07) < 0.001 No severe early OHSS 276 1.04 (1.01 1.07) 1.82 (1.74 1.90) 2.74 (2.62 2.86) < 0.001 Severe early OHSS 43 1.05 (0.98 1.11) 3.26 (3.01 3.50) 4.28 (4.01 4.55) < 0.001 Ht (%) All patients 319 38.27 (37.91 38.63) 39.86 (39.48 40.24) 40.91 (40.51 41.31) < 0.001 No severe early OHSS 276 38.12 (37.74 38.50) 39.55 (39.14 39.95) 40.20 (39.82 40.57) < 0.001 Severe early OHSS 43 39.23 (38.15 40.31) 41.86 (41.03 42.69) 45.5 (44.61 46.39) < 0.001 WBC All patients 319 9813 (9496 10 129) 12 131 (11 749 12 514) 14 336 (13 871 14 802) 0.001 No severe early OHSS 276 9553 (9247 9860) 11 669 (11 314 12 025) 13 503 (13 104 13 904) 0.001 Severe early OHSS 43 11 476 (10 270 12 682) 15 099 (13 654 16 543) 19 685 (18 086 21 283) 0.001 MOD (mm) All patients 319 81.99 (77.88 86.10) 83.52 (81.83 85.21) 0.454 No severe early OHSS 276 80.79 (76.09 85.48) 79.73 (78.26 81.19) 0.643 Severe early OHSS 43 89.72 (85.13 94.31) 107.88 (105.33 110.44) 0.001 Data are presented as mean (95% CI). *Generalized estimating equations with dependent variable of ascites grade, hematocrit (Ht), white blood cell count (WBC) or maximal ovarian diameter (MOD) and independent variable of time (day 0, day 3, day 5). Hematocrit Mean Ht values in all patients at high risk for OHSS, as well as in the subgroups of patients according to whether they developed severe early OHSS, are shown in Table 5. There was a statistically significant increase (P < 0.001) in Ht during the early luteal phase (from day 0 to day 3 to day 5) in all groups. Ht changes (cht) between day 3 and day 0 (cht D3 D0 ), between day 5 and day 0 (cht D5 D0 ) and between day 5 and day 3 (cht D5 D3 ) are shown in Table S4. cht D5 D0 was significantly greater than cht D3 D0 and than cht D5 D3, overall as well as in both patients who developed severe early OHSS and those who did not. cht D3 D0 was significantly greater than cht D5 D3 in patients who did not develop severe early OHSS. This, however, was not the case in patients who developed severe early OHSS (Table S4). cht D3 D0,cHt D5 D3 and cht D5 D0 were significantly greater in patients who did vs those who did not eventually develop severe early OHSS (P 0.001) (Figure S3). Both time following oocyte retrieval and whether severe early OHSS developed were significantly associated with Ht on multivariable analysis. Furthermore, there was a significant interaction between time and development of severe early OHSS (Table S5). Patients who developed severe early OHSS showed a higher rate of Ht increase per day (coefficient, +0.780) compared with patients who did not develop severe early OHSS (Table S5). White blood cell count Mean WBC in all patients at high risk for OHSS, as well as in the subgroups of patients according to whether they developed severe early OHSS, is shown in Table 5. There was a statistically significant increase (P = 0.001) in WBC during the early luteal phase (from day 0 to day 3 to day 5) in all groups. Changes in WBC (cwbc) between day 3 and day 0 (cwbc D3 D0 ), between day 5 and day 0 (cwbc D5 D0 ) and between day 5 and day 3 (cwbc D5 D3 ) are shown

Changes in OHSS in early luteal phase 131 in Table S6. cwbc D5 D0 was significantly greater than WBC D3 D0 and than WBC D5 D3, overall as well as in patients who did and those who did not develop severe early OHSS. cwbc D3 D0 was not significantly different from cwbc D5 D3 in any group (Table S6). cwbc D3 D0, cwbc D5 D0 and cwbc D5 D3 were significantly higher in patients who developed severe early OHSS vs those who did not (P 0.001) (Figure S4). Both time following oocyte retrieval and whether severe early OHSS developed were associated significantly with WBC on multivariate analysis. Moreover, there was a significant interaction between time following oocyte retrieval and development of severe early OHSS (Table S7). Patients who developed severe early OHSS showed a higher rate of increase in WBC per day (coefficient, +884) compared with patients who did not eventually develop severe early OHSS (Table S7). Maximum ovarian diameter MOD in all patients at high risk for OHSS, as well as in the subgroups of patients according to whether they developed severe early OHSS, is shown in Table 5. There was a statistically significant increase (P = 0.001) in MOD from day 3 to day 5 in patients who developed severe early OHSS, but not in those who did not develop severe early OHSS. No significant difference was detected when all patients were analyzed (Table S8). On multivariable analysis, MOD was significantly associated with whether severe early OHSS developed, controlling for the effect of time following oocyte retrieval; the latter did not affect MOD significantly (Table S9). However, there was a significant interaction between time and development of severe early OHSS. Patients who developed severe early OHSS showed a higher rate of increase per day in MOD (coefficient, +9.61) compared with patients who did not eventually develop severe early OHSS (Table S9). Correlation between OHSS-associated variables Pairwise correlations between OHSS-associated variables are presented in Table S10. The highest correlation was between MOD and ascites grade on day 5 followed by that between MOD and WBC on day 5. Pairwise correlations between changes in OHSS-associated variables are presented in Table S11. The highest correlation was between the change in ascites grade between day 0 and day 5 and the change in Ht level over the same period. DISCUSSION This study describes the absolute values and changes in ascites grade, Ht, WBC and MOD during the early luteal phase in women with 19 follicles on the day of triggering of final oocyte maturation with hcg, and thus who were at high risk for developing severe early OHSS. In all high-risk women, the values of ascites grade, Ht and WBC increased significantly with time, within the 5-day observation period following oocyte retrieval, in line with the presumed pathophysiology of the syndrome 4,33,34. Elevation of Ht reflects intravascular volume depletion, and has a reciprocal relationship with ascetic fluid accumulation 4. This relationship between ascites and Ht change was confirmed in our study, since the highest correlation of change from day 0 to day 5 was observed between the change in ascites grade and that in Ht. The only variable which increased between day 3 and day 5, only in patients who developed severe early OHSS, but not in patients who did not, was MOD. Historically, ovarian enlargement has been considered as a marker of OHSS 11,12. The classification by Golan et al. 13,which was the first to use ultrasound to evaluate ovarian enlargement and ascites, proposed an ovarian diameter range of 5 12 cm as a criterion for diagnosing all forms of OHSS. In contrast, Navot et al. 4 did not consider ovarian enlargement to be a reliable factor for diagnosing severe early OHSS. Our present data support the value of ovarian enlargement in OHSS diagnosis. Ht change has been used as a criterion for diagnosing severe early OHSS, although not in modern classifications. The increase in Ht of 6.27% (16.51% relative increase) observed in the current study between day 0 and day 5, in patients who developed severe early OHSS, is in agreement with previous arbitrarily proposed values of 10% 35 and 6 10% 12 Ht increase over the baseline, but not, however, with a 30% increase, also proposed arbitrarily for diagnosis of severe OHSS by Navot et al. 4. Our current data suggest that cht is directly associated with the quantity of ascetic fluid accumulation. Therefore, cht may be used as a marker of hemoconcentration, and possibly as a new additional criterion in a modern OHSS classification system. In this regard, cht, expressed as relative change, may facilitate a patient-specific approach for the diagnosis of severe early OHSS, since this would also take into consideration the starting Ht level of each patient on the day of oocyte retrieval. This may provide more accurate pathophysiological insight into the development of severe early OHSS, and may be particularly useful for patients with either very low Ht (e.g. 28%) or conversely very high Ht, approaching 45%, on the day of oocyte retrieval. Nevertheless, the absolute value of Ht (> 45%) remains a significant marker of severe early OHSS from an epidemiological perspective, since this is related to an increased risk of complications (deep vein thrombosis/pulmonary embolism). In this study, we used a detailed classification of ascites. Older classification systems used the vague description, ultrasound evidence or clinical evidence of ascites 1,4,11 13. More recently, there have been efforts to classify ascites in an objective and reproducible manner, based on the quantity of fluid accumulation within body cavities observed during ultrasound examination 17 20,26,31,36. However, all of these studies assessed ascites only on the day of severe early OHSS diagnosis. To our knowledge, this is the first study to

132 Lainas et al. describe the timecourse of fluid accumulation during the evolution of OHSS in the first 5 days after oocyte retrieval. Interestingly, ascites Grade 5 (massive) and Grade 6 (tense) were observed only in women who were diagnosed with severe early OHSS. Thus, the presence of ascites Grade 5 or 6 alone, which is essentially the equivalent of clinical evidence of ascites, as mentioned in previous classifications 1,4,11 13, could indicate that severe early OHSS is already established. In the present study, severe early OHSS was diagnosed on day 5, and no earlier, in all affected women. However, follow up on day 3 is still of clinical value and should be performed in all high-risk women, as severe early OHSS may well manifest on day 3 or earlier in more aggressive cases. In all high-risk patients, it is recommended to perform blastocyst culture until day 5 or 6 and to avoid performing transfers on day 3 as OHSS may still be evolving. Embryo culture on day 5 allows time for an extensive and accurate assessment of the evolution of OHSS. It is known that early OHSS may be self-limiting and resolve by day 5 post-oocyte retrieval 37. If severe early OHSS has not developed by day 5, the decision to proceed to ET can be made more safely, thus reducing significantly the chance of late OHSS 18,38. Using this management approach, the incidence of late OHSS was shown to be as low as 1.1% among high-risk patients 18. If severe early OHSS develops by day 5, then ET should be canceled, and luteal GnRH antagonist administration can be performed, in combination with freezing of all blastocysts. The limitation of the present study is its retrospective design. It cannot be excluded that non-apparent sources of bias might be present. Severe early OHSS was diagnosed using widely accepted predetermined criteria of ascites, Ht, WBC and MOD, and was not based on patient hospitalization, as women were managed as outpatients by luteal antagonist administration 18. As a result, the diagnostic value of these parameters on the day of severe early OHSS diagnosis cannot be confirmed. In conclusion, ascites grade, Ht and WBC increased significantly with time following oocyte retrieval, within the 5-day study observation period, in all women at high risk for OHSS, in line with the pathophysiology of the syndrome. Our data support the use of MOD in the diagnosis of severe early OHSS, and provide novel evidence for the role of cht as a patient-specific hemoconcentration marker during OHSS development. These findings may be helpful to the clinician and may be useful for the establishment of new criteria in a universally accepted OHSS classification system. ACKNOWLEDGMENTS We wish to thank Dr G. K. Petsas for assisting with clinical work, Mrs G. Stavropoulou for patient coordination, and Mrs M. Panagopoulou and Mrs I. Voulgari for data entry. Disclosure B.C.T. has received unrestricted research grants, honorarium and travel fees from Merck Serono and Merck, Sharp and Dome and travel fees from IBSA & Ferring, unrelated to this work. C.A.V. has received honoraria and travel grants from Merck, Sharp and Dome and Merck Serono as well as consultation fees from IPSEN Hellas S.A., unrelated to this work. REFERENCES 1. Rizk B, Aboulghar MA. Classification, pathophysiology and management of ovarian hyperstimulation syndrome. In In-Vitro Fertilization and Assisted Reproduction, Brinsden P (ed). The Parthenon Publishing Group: New York, London, 1999. p. 131 55. 2. Aboulghar MA, Mansour RT. Ovarian hyperstimulation syndrome: classifications and critical analysis of preventive measures. Hum Reprod Update 2003;9: 275 289. 3. Rizk B, Aboulghar M, Smitz J, Ron-El R. The role of vascular endothelial growth factor and interleukins in the pathogenesis of severe ovarian hyperstimulation syndrome. Hum Reprod Update 1997; 3: 255 266. 4. Navot D, Bergh PA, Laufer N. Ovarian hyperstimulation syndrome in novel reproductive technologies: prevention and treatment. Fertil Steril 1992;58: 249 261. 5. Aboulghar MA, Mansour RT, Serour GI, Sattar MA, Amin YM, Elattar I. Management of severe ovarian hyperstimulation syndrome by ascitic fluid aspiration and intensive intravenous fluid therapy. Obstet Gynecol 1993; 81: 108 111. 6. Practice Committee of the American Society for Reproductive Medicine. Ovarian hyperstimulation syndrome. Fertil Steril 2008; 90: S188 193. 7. Devroey P, Polyzos NP, Blockeel C. An OHSS-Free Clinic by segmentation of IVF treatment. Hum Reprod 2011; 26: 2593 2597. 8. Youssef MA, Van der Veen F, Al-Inany HG, Mochtar MH, Griesinger G, Nagi Mohesen M, Aboulfoutouh I, van Wely M. Gonadotropin-releasing hormone agonist versus HCG for oocyte triggering in antagonist-assisted reproductive technology. Cochrane Database Syst Rev 2014: Cd008046. 9. De Ziegler D, Shoham Z. Survey results: triggering of ovulation with GnRh-a in ART: worldwide feedback on an emerging new option with great potential. IVF Worldwide Newsletter 2013; October. www.ivf-worldwide.com [Accessed 20 May 2017]. 10. Tobler KJ, Zhao Y, Weissman A, Majumdar A, Leong M, Shoham Z. Worldwide survey of IVF practices: trigger, retrieval and embryo transfer techniques. Arch Gynecol Obstet 2014; 290: 561 568. 11. Rabau E, David A, Serr DM, Mashiach S, Lunenfeld B. Human menopausal gonadotropins for anovulation and sterility. Am J Obstet Gynecol 1967; 98: 92 98. 12. Schenker J, Weinstein D. Ovarian hyperstimulation syndrome: a current survey. Fertil Steril 1978; 30: 255 268. 13. Golan A, Ron-el R, Herman A, Soffer Y, Weinraub Z, Caspi E. Ovarian hyperstimulation syndrome: an update review. Obstet Gynecol Surv 1989; 44: 430 440. 14. Borenstein R, Elhalah U, Lunenfeld B, Schwartz ZS. Severe ovarian hyperstimulation syndrome: a reevaluated therapeutic approach. Fertil Steril 1989; 51: 791 795. 15. Alvarez C, Alonso-Muriel I, Garcia G, Crespo J, Bellver J, Simon C, Pellicer A. Implantation is apparently unaffected by the dopamine agonist Cabergoline when administered to prevent ovarian hyperstimulation syndrome in women undergoing assisted reproduction treatment: a pilot study. Hum Reprod 2007; 22: 3210 3214. 16. Golan A, Weissman A. Symposium: Update on prediction and management of OHSS. A modern classification of OHSS. Reprod Biomed Online 2009; 19: 28 32. 17. Humaidan P, Quartarolo J, Papanikolaou EG. Preventing ovarian hyperstimulation syndrome: guidance for the clinician. Fertil Steril 2010; 94: 389 400. 18. Lainas G, Kolibianakis E, Sfontouris I, Zorzovilis I, Petsas G, Tarlatzi T, Tarlatzis B, Lainas T. Outpatient management of severe early OHSS by administration of GnRH antagonist in the luteal phase: an observational cohort study. Reprod Biol Endocrinol 2012; 10: 69. 19. Lainas GT, Kolibianakis EM, Sfontouris IA, Zorzovilis IZ, Petsas GK, Lainas TG, Tarlatzis BC. Pregnancy and neonatal outcomes following luteal GnRH antagonist administration in patients with severe early OHSS. Hum Reprod 2013; 28: 1929 1942. 20. Pau E, Alonso-Muriel I, Gomez R, Novella E, Ruiz A, Garcia-Velasco JA, Simon C, Pellicer A. Plasma levels of soluble vascular endothelial growth factor receptor-1 may determine the onset of early and late ovarian hyperstimulation syndrome. Hum Reprod 2006; 21: 1453 1460. 21. Nastri CO, Teixeira DM, Moroni RM, Leitao VM, Martins WP. Ovarian hyperstimulation syndrome: pathophysiology, staging, prediction and prevention. Ultrasound Obstet Gynecol 2015; 45: 377 393. 22. Royal College of Obstetricians and Gynaecologists. The Management of Ovarian Hyperstimulation Syndrome. Green-top Guideline, No5, 2016. 23. Leitao VM, Moroni RM, Seko LM, Nastri CO, Martins WP. Cabergoline for the prevention of ovarian hyperstimulation syndrome: systematic review and meta-analysis of randomized controlled trials. Fertil Steril 2014; 101: 664 675. 24. Bechtejew TN, Nadai MN, Nastri CO, Martins WP. Clomiphene citrate and letrozole to reduce follicle-stimulating hormone consumption during ovarian stimulation: systematic review and meta-analysis. Ultrasound Obstet Gynecol 2017;50:315 323. 25. Enskog A, Henriksson M, Unander M, Nilsson L, Brannstrom M. Prospective study of the clinical and laboratory parameters of patients in whom ovarian hyperstimulation syndrome developed during controlled ovarian hyperstimulation for in vitro fertilization. Fertil Steril 1999; 71: 808 814. 26. Alvarez C, Marti-Bonmati L, Novella-Maestre E, Sanz R, Gomez R, Fernandez-Sanchez M, Simon C, Pellicer A. Dopamine agonist cabergoline reduces hemoconcentration and ascites in hyperstimulated women undergoing assisted reproduction. J Clin Endocrinol Metab 2007; 92: 2931 2937.

Changes in OHSS in early luteal phase 133 27. Mannaerts B, Van Kuijk J, Griesinger G. Prediction of OHSS in patients treated with corifollitropin alfa or rfsh in a GnRH antagonist protocol. Hum Reprod 2012; 27: ii26 ii8. 28. Griesinger G, Verweij PJ, Gates D, Devroey P, Gordon K, Stegmann BJ, Tarlatzis BC. Prediction of ovarian hyperstimulation syndrome in patients treated with corifollitropin alfa or rfsh in a GnRH antagonist protocol. PLoS One 2016; 11: e0149615. 29. Lainas TG, Sfontouris IA, Zorzovilis IZ, Petsas GK, Lainas GT, Alexopoulou E, Kolibianakis EM. Flexible GnRH antagonist protocol versus GnRH agonist long protocol in patients with polycystic ovary syndrome treated for IVF: a prospective randomised controlled trial (RCT). Hum Reprod 2010; 25: 683 689. 30. Kolibianakis EM, Papanikolaou EG, Tournaye H, Camus M, Van Steirteghem AC, Devroey P. Triggering final oocyte maturation using different doses of human chorionic gonadotropin: a randomized pilot study in patients with polycystic ovary syndrome treated with gonadotropin-releasing hormone antagonists and recombinant follicle-stimulating hormone. Fertil Steril 2007; 88: 1382 1388. 31. Lainas GT, Kolibianakis EM, Sfontouris IA, Zorzovilis IZ, Petsas GK, Lainas TG, Tarlatzis BC. Serum vascular endothelial growth factor levels following luteal gonadotrophin-releasing hormone antagonist administration in women with severe early ovarian hyperstimulation syndrome. BJOG 2014; 121: 848 855. 32. Lainas TG, Sfontouris IA, Zorzovilis IZ, Petsas GK, Lainas GT, Iliadis GS, Kolibianakis EM. Management of severe OHSS using GnRH antagonist and blastocyst cryopreservation in PCOS patients treated with long protocol. Reprod Biomed Online 2009; 18: 15 20. 33. Neulen J, Yan Z, Raczek S, Weindel K, Keck C, Weich HA, Marme D, Breckwoldt M. Human chorionic gonadotropin-dependent expression of vascular endothelial growth factor/vascular permeability factor in human granulosa cells: importance in ovarian hyperstimulation syndrome. J Clin Endocrinol Metab 1995;80:1967 1971. 34. Pellicer A, Albert C, Mercader A, Bonilla-Musoles F, Remohi J, Simon C. The pathogenesis of ovarian hyperstimulation syndrome: in vivo studies investigating the role of interleukin-1beta, interleukin-6, and vascular endothelial growth factor. Fertil Steril 1999; 71: 482 489. 35. Fabregues F, Balasch J, Manau D, Jimenez W, Arroyo V, Creus M, Rivera F, Vanrell JA. Haematocrit, leukocyte and platelet counts and the severity of the ovarian hyperstimulation syndrome. Hum Reprod 1998; 13: 2406 2410. 36. Giles J, Requena A, Garcia-Velasco JA, Pacheco A, Pellicer J, Pellicer A. GnRH analogue for the prevention of ovarian hyperstimulation syndrome: a pilot study. Fertil Steril 2009; 91: 1366 1369. 37. Lyons CA, Wheeler CA, Frishman GN, Hackett RJ, Seifer DB, Haning RV, Jr. Early and late presentation of the ovarian hyperstimulation syndrome: two distinct entities with different risk factors. Hum Reprod 1994; 9: 792 799. 38. Papanikolaou E, Humaidan P, Polyzos N, Kalantaridou S, Kol S, Benadiva C, Tournaye H, Tarlatzis B. New algorithm for OHSS prevention. Reprod Biol Endocrinol 2011; 9: 147. SUPPORTING INFORMATION ON THE INTERNET The following supporting information may be found in the online version of this article: Figure S1 Grade of ascites on days 0, 3 and 5 following oocyte retrieval in all patients, in patients without severe ovarian hyperstimulation syndrome (OHSS) and in patients with severe early OHSS. Error bars depict 95% CI. Figure S2 Direction of change in severity of ascites between days 0, 3 and 5 following oocyte retrieval in all patients, in patients without severe ovarian hyperstimulation syndrome (OHSS) and in patients with severe early OHSS. Error bars depict 95% CI. Figure S3 Change in hematocrit (cht) between days 3 and 0 (cht D3 D0 ),days5and0(cht D5 D0 ) and days 5 and3(cht D5 D3 ) following oocyte retrieval in women with and those without severe ovarian hyperstimulation syndrome (OHSS). Error bars depict 95% CI. Figure S4 Change in white blood cell count (cwbc) between days 3 and 0 (cwbc D3 D0 ), days 5 and 0 (cwbc D5 D0 )anddays5and3(cwbc D5 D3 ) following oocyte retrieval in women with and those without severe ovarian hyperstimulation syndrome (OHSS). Error bars depict 95% CI. Table S1 Diagnosis of severe early OHSS in 43 high-risk women, undergoing ovarian stimulation for IVF, following triggering of final oocyte maturation with hcg Table S2 Changes in ascites grade (casc) during the early luteal phase (days 0, 3, 5 following oocyte retrieval) in all patients as well as in those who did or did not develop severe early OHSS Table S3 Multivariable analysis with dependent variable ascites grade and independent factors: time (day 0, day 3, day 5 following oocyte retrieval) and development or not of severe early OHSS Table S4 Changes in hematocrit (cht) during the early luteal phase (days 0, 3, 5 following oocyte retrieval) in all patients as well as in those who did or did not develop severe early OHSS Table S5 Multivariable analysis with dependent variable Ht and independent factors: time (day 0, day 3, day 5 following oocyte retrieval) and development or not of severe early OHSS Table S6 Changes in WBC during the early luteal phase (days 0, 3, 5 following oocyte retrieval) in all patients as well as in those who did or did not develop severe early OHSS Table S7 Multivariable analysis with dependent variable WBC and independent factors: time (day 0, day 3, day 5 following oocyte retrieval) and development or not of severe early OHSS Table S8 Changes in maximal ovarian diameter during the early luteal phase (days 3, 5 following oocyte retrieval) in all patients as well as in those who did or did not develop severe early OHSS Table S9 Multivariable analysis with dependent variable MOD and independent factors: time (day 3, day 5 following oocyte retrieval) and development or not of severe early OHSS Table S10 Pairwise correlations between OHSS-associated variables Table S11 Pairwise correlations between changes in OHSS-associated variables