In vitro fertilization outcome in severe ovarian hyperstimulation syndrome: An age‐matched contemporaneous control study

Abstract

Aim:   Ovarian hyperstimulation syndrome (OHSS) is a potentially life‐threatening, iatrogenic complication of assisted reproduction and has been associated with poor in vitro fertilization outcome. The aim of the present study was to evaluate the pregnancy rate and outcome following severe OHSS, at a single center over a three‐year period.

Methods:   The incidence of severe OHSS at the IVF Center, National University of Singapore, in Singapore, was 4% (48 cases over 1200 cycles) during the period of 1997–2000. The present retrospective study compared 48 cases of severe OHSS to 144 age‐matched, contemporaneous controls without OHSS.

Results:   The total gonadotropin required for severe OHSS group was found to be lower than for that of controls (2664.06 ± 768.29 IU vs 3349.58 ± 2003.73 IU), although duration of stimulation was similar. The OHSS group was associated with a fivefold increase (OR 5.293, 95% CI: 2.116–13.238) in pregnancy rate compared to controls (87.5% vs 56.9%; P < 0.05). Late OHSS was more common (38/48 cases) and had a pregnancy rate of 97% per embryo transfer. There was no significant difference in the multiple pregnancy (54% vs 48%; P > 0.05) and miscarriage rates (14% vs 7.3%; P > 0.05) between the groups.

Conclusion:   Severe OHSS at our center were mostly late onset. The pregnancy rate was significantly higher, but multiple pregnancy and miscarriage rates were not significantly increased when compared to the age‐matched contemporaneous controls. (Reprod Med Biol 2005; 4: 207–211)

INTRODUCTION

SEVERE OVARIAN HYPERSTIMULATION syndrome (OHSS) is an iatrogenic condition arising from ovarian hyperstimulation, as a result of assisted reproductive treatment. ¹ Despite the many strategies taken to reduce this complication, the incidence of severe OHSS still remains at 0.5–2.0% of all cycles. The exact pathogenesis is unknown but human chorionic gonadotropin (hCG) has been implicated to trigger this event. The source of hCG might be exogenous after its administration prior to oocyte retrieval. Alternatively, severe OHSS can also be precipitated by hCG from an implanting gestation and is associated with conception. ² The clinical manifestations are as a result of increased vascular permeability associated with loss of fluid into ‘third space’ and intravascular dehydration. There might be thromboembolism, renal failure and adult respiratory distress syndrome, in addition to enlarged ovaries caused by multiple follicles.

The high estradiol level, systemic coagulopathy, presence of pro‐inflammatory cytokines and endothelial growth factors ³ are believed to create an unfavorable environment for implantation of the embryo and are associated with increased pregnancy loss. Hypotension, hypoxemia and electrolyte imbalance associated with severe OHSS are postulated to aggravate further the deleterious in vitro fertilization (IVF) outcome. Results of clinical studies addressing the IVF outcome complicated by OHSS remain variable. Higher pregnancy rate has been reported in women who developed severe OHSS, ⁴ , ⁵ , ⁶ however, these comparisons included women managed at different centers and the womens’ ages were not controlled. On the contrary, a higher miscarriage rate has been reported, ⁷ , ⁸ even though other studies found no difference. ⁹ , ¹⁰

However, most studies have been limited by small sample sizes, lack of appropriate controls and compounded by numerous factors. Some controls were from the general population, which provided a different profile compared with pregnancy in infertile couples after IVF treatment. The age of the woman, an important determinant of IVF outcome, was not taken into account in most studies. Advanced age has been associated with less successful IVF and higher pregnancy loss. Until the present study, there have been no age‐matched comparisons observed in the literature.

The present study aimed to determine the IVF outcome after severe OHSS, comparing with those without this complication using an age‐matched, contemporaneous group as controls. To our best knowledge, this is the first study that used an age‐matched control group.

METHODS

BETWEEN JANUARY 1997 to December 2000, a total of 48 cases of severe OHSS were hospitalized following 1200 IVF cycles (4%). A contemporaneous control group was selected from the registry of our own IVF unit at the IVF Center, National University of Singapore, in an effort to reduce bias. For every single case of IVF complicated by severe OHSS, three contemporaneous cases were selected and matched for the maternal age group. The IVF outcome was blinded to the investigators during the selection process. The IVF records and gynecological notes of all the participants were reviewed and recorded on a standard form.

Diagnosis of severe ovarian hyperstimulation syndrome

The severity of OHSS was diagnosed according to criteria proposed by Golan et al. ¹¹ OHSS was considered severe if the patients presented with clinical evidence of ascites or hydrothorax as well as abdominal discomfort and dyspnea. Biochemical evidence included hemoconcentration (hematocrit >50%), leucocytosis (count >25 000/L), increased viscosity, coagulation disorders and reduced renal/liver perfusion. Those who developed severe OHSS less than nine days after oocyte retrieval were considered as early onset, whereas late OHSS was diagnosed if it occurred later. ¹² Patients with severe OHSS were hospitalized and received supportive treatment.

IVF stimulation protocol

Long protocol with gonadotropin release hormone analog (GnRH) down‐regulation was used for controlled ovarian hyperstimulation. Subcutaneous buserelin 0.5 mg was given for two weeks prior to administration of gonadotrophin injection. Down‐regulation was confirmed when serum estradiol was less than 50 IU/mL or luteinizing hormone (LH) was less than 10 IU. Subcutaneous recombinant follicle stimulating hormone 225 IU was initiated for ovarian hyperstimulation. Transvaginal ultrasound scan and serum estradiol were used to monitor the follicular growth from Day 7 onwards. The dosage of gonadotrophin was adjusted according to the ovarian response. The cycle was cancelled when more than 20 follicles or other evidence of severe OHSS was noted. Intramuscular hCG 10 000 IU was given when one or more follicles attained 18 mm and two others of 16 mm size. Oocyte retrieval was carried out 34–36 h later. Three to four embryos were transferred on the third day after oocyte retrieval. Progestogen suppositories and hCG injection were used for luteal supplementation.

Pregnancy was defined as the presence of a gestational sac on an ultrasound scan 3–4 weeks after embryo transfer. Any pregnancy loss before the period of viability (24 weeks) was classified as a miscarriage.

Ethical consideration

There was no significant ethical issue involved as this was a retrospective study. Nevertheless, approval was obtained from the university's research committee.

Statistical analysis

All the data were analyzed using the spss 9.0 statistical package (SPSS, Chicago, IL, USA). Student's t‐test, multiple mean comparison and χ² test were used with P < 0.05 being considered to be significant.

RESULTS

THE PATIENT PROFILE and IVF stimulation are summarized in Table 1. There was no significant difference between the two groups in terms of age, primary subfertility, duration of subfertility and the incidence of polycystic ovarian disease. The severe OHSS group used less gonadotropin but responded with more oocytes. The duration of stimulation was similar in both groups. Although a higher number of oocytes were retrieved in the severe OHSS group, the fertilization rate was similar for both groups. There was also no difference in the number of fertilizations by intracytoplasmic sperm injection and total embryos transferred.

Table 1.

Patient profile and IVF stimulation

	OHSS group (n = 48)	Control group (n = 144)
Mean age (years)	33.73 ± 4.00	33.87 ± 3.61
Primary subfertility (%)	75	69.4
Duration of subfertility (years)	5.31 ± 2.85	6.06 ± 2.88
PCOD (%)	20.8	16.7
Total gonadotropin (iu)*	2664.06 ± 768.29	3349.58 ± 2003.73
Estradiol level (pg/mL)	3372.69 ± 1606.634	3267.85 ± 2262.34
Mean duration (days)	11.60 ± 1.75	12.10 ± 1.95
Mean oocytes retrieved**	13.17 ± 3.82	10.51 ± 5.80
ICSI (%)	60.4	61.1
Fertilization rate	0.663 ± 0.223	0.7182 ± 0.3015
Embryo transfer
Less or equal two	6 (12.5%)	36 (25.0%)
Three	38 (79.2%)	99 (69.2%)
More than three	4 (8.3%)	8 (5.6%)

^{*, **} P < 0.05.

ICSI, intracytoplasmic sperm injection; OHSS, ovarian hyperstimulation syndrome; PCOD, polycystic ovarian disease.

Pregnancy was achieved in 42/48 patients with severe OHSS, which gave a pregnancy rate per embryo transfer of 87.5% (see Table 2). This was significantly higher compared with controls (82/144; 56.9%; P < 0.05). The likelihood of pregnancy was fivefold higher (OR 5.293, 95% CI: 2.116–13.238, P < 0.05) in IVF cases complicated by OHSS compared with controls. When the severe OHSS group was analyzed for timing of onset, there were 10 (20.8%) cases of early OHSS, of which five resulted in pregnancy (pregnancy rate of 50% per embryo transfer). Thirty‐eight (79.2%) of the cases were late OHSS and 37 of these were pregnant, making the pregnancy rate per embryo transfer 97.3% (See Fig. 1). The miscarriage rate was twofold higher but similar between the severe OHSS and controls (14.0% vs 7.3%; P > 0.05). The multiple pregnancy rate was higher in the severe OHSS group (59.5% vs 41.5%; P > 0.05) but not statistically significant.

Table 2.

IVF outcome

	OHSS group (n = 48)		Total	Controls (n = 144)
	Early	Late
Pregnant***	5	37	42 (87.5%)	82 (56.9%)
Non pregnant	5	1	6	62
Ongoing pregnancy	5	31	36	76
Miscarriage	0	6	6	6 (7.3%)
No. gestation
Singleton		17 (40.5%)		48 (58.5%)
Multiple		25 (59.5%)		34 (41.5%)

^*** P < 0.05.

Figure 1.

Scatterplot showing interval to ovarian hyperstimulation syndrome and in vitro fertilization outcome. Case index is a randomly allotted number with no specific chronology reference. (▿), Not pregnant; (▪), pregnant.

DISCUSSION

SEVERE OHSS IS an iatrogenic, potentially life‐threatening complication after IVF treatment involving the use of ovulation induction agent and GnRH agonist. It is also often associated with poor IVF outcome. Despite efforts to prevent this complication, the incidence of severe OHSS at our center was 4% compared with the general occurrence of 1%. ¹³ Most of the severe OHSS occurred later (10 or more days after oocyte retrieval) and were strongly associated with pregnancy. The pregnancy rate per embryo transfer was 87.5% in the severe OHSS group compared with the age matched contemporaneous controls (56.9%). Other investigators had reported similar findings. ⁴ , ⁵ , ⁹ Ariel et al. attributed the high pregnancy rate partly to younger OHSS patients. ⁴ In an attempt to explain the occurrence of OHSS, all the controls in the present series were matched for age to the OHSS group. The possible reason for the incidence of severe OHSS in the present study was the successful pregnancy rate at our center. However, we caution that severe OHSS should not be advocated to improve IVF outcome.

Mathur et al. had distinguished the two distinct forms of OHSS based on the interval after oocyte retrieval. ¹² Severe OHSS, which occurs 10 or more days after oocyte retrieval, is considered ‘late’ whereas early OHSS is diagnosed when the interval is less than 10 days. Early OHSS can be predicted by the high preovulatory serum estradiol level, indicating excessive ovarian response. This condition can be prevented by various measures, which include coasting, withholding hCG, abandoning the IVF cycle or freezing all the embryos. In contrast, late OHSS is not related to serum estradiol level and might not be predicted easily. Thus, prevention of this condition is difficult. Late OHSS is associated with successful IVF, that is, pregnancy, as compared with early OHSS, where overzealous stimulation or excessive response is a factor. At our center, most of the OHSS occurred late (79.1%) with a high association of pregnancy (97.8%). This showed that pregnancy was the main reason for the high incidence of severe OHSS at our center. On the contrary, early OHSS had a pregnancy rate of only 50% per embryo transfer. This observation seemed to suggest that the hemodynamic derangements and presence of high inflammatory mediators affected the implantation rate rather than the pregnancy.

We realized that some potential early OHSS cases were inevitably excluded because of cycle cancellation. It is our center's policy to abandon cases identified as having a high risk of developing severe OHSS prior to hCG administration. Although some had advocated measures such as coasting or albumin infusion to prevent OHSS, the advantage of an improved pregnancy rate is yet to be proven. ¹⁴ , ¹⁵ Despite strict adherence to our protocol, there were 10 cases of early OHSS in the apparently low risk group. The lower pregnancy rate associated with this group enabled prognostication of IVF outcome following the complication of OHSS.

The incidence of multiple pregnancy was higher in the OHSS group as compared with controls (59.5 vs 49.5%). Initial reports described late OHSS occurring only in multiple gestations. ¹⁶ Our results showed late OHSS occurred in both singleton and multiple pregnancies, with a higher incidence in the latter. This finding concurred with that of other investigators. ⁴ , ¹² The higher endogenous hCG found in multiple pregnancies could explain the tendency to trigger late OHSS. In contrast, the occurrence of OHSS with singleton pregnancy suggests that the absolute level of hCG required to initiate this complication might vary from individual to individual and relates to each individual's unique cytokine profile. ³ Other explanations for increased susceptibility to OHSS included the mutation of the hCG/LH receptor ¹⁷ and a disproportionately lower serum α₂‐Macroglobulin to vascular endothelial growth factor ratio. ¹⁸ The present study was not able to comment on the cytokines and molecular profile as they were not assessed.

The miscarriage rate was not significantly different when compared with the age‐matched, contemporaneous controls (14.3 vs 7.3%, P > 0.05). These findings were similar to that of Mathur et al. ⁹ and Chen et al., ¹² who also compared their results to a contemporaneous group. It appeared that when pregnancy (implantation) had occurred, the inflammatory response following late OHSS did not affect the miscarriage rate. However, there were also other investigators who reported a higher miscarriage rate, ranging from 25 to 40% in the severe OHSS group. ⁵ , ⁷ These studies lacked controls except for that by Abramov et al., ⁵ who used the national IVF registries of several countries as controls. This comparison was disadvantaged by possible confounding factors such as ascertainment bias, patient age, criteria for diagnosis of pregnancy, incidence of polycystic ovarian disease and multiple pregnancy. Another possible explanation for the higher miscarriage rate might be attributed to the higher multiple pregnancy rate following OHSS, although this was not observed in the present study.

As in most studies with OHSS, ours had the similar problem of small sample size. This was inevitable as complication of severe OHSS was uncommon. In the present study, we had taken measures to minimize other confounding variables. The uniformity in the management of IVF and OHSS were assured as all cases were managed by the same IVF team, in a single center. We also included a contemporaneous control group, which were selected from our own IVF registry besides matching the age. The importance of the woman's age as the determinant of IVF outcome, including pregnancy, miscarriage and multiple gestation rates was also addressed in the present study.

In conclusion, severe OHSS at our center were mostly of late onset and associated with a high pregnancy rate. Although multiple gestations and miscarriage showed a higher trend, these were not statistically significant when compared with the age‐matched contemporaneous control group.