Severe OHSS: How many cases are preventable?

During 1995, >50 000 cycles of ovarian hyperstimulation for the purpose of in-vitro fertilization (IVF), intracytoplasmic sperm injection (ICSI) or gamete intra-Fallopian transfer (GIFT) were reported in the World Collaborative Report on IVF (1997) and it is believed that equally as many cycles of ovulation induction are undertaken annually. As the overall incidence of clinically relevant ovarian hyperstimulation syndrome (OHSS) occurs in up to 10% of cycles, some of which (0.5–2%) are life-threatening (Forman et al., 1990), it is important for all clinicians, not only infertility specialists, to be aware of strategies in its prevention. OHSS is iatrogenic, and thus places an overwhelming onus on the clinician to avert it if at all possible. But how many cases are actually preventable without compromising the patient's chances of successful outcome of the intended treatment? A balance between optimum ovarian stimulation and successful treatment outcome in the absence of severe OHSS or multifetal pregnancies is desirable in the practice of assisted reproduction technology (ART). In their debate article, Abramov et al. noted a disproportionate increase in the incidence of severe OHSS that has surpassed the increased use of controlled ovarian stimulation for IVF. They attributed this to the current `high' dose gonadotrophin protocols which are commonly used in ART programmes (Abramov et al., 1999). Although clinicians generally tailor the administration of gonadotrophins for ovarian stimulation according to a woman's age, weight, serum concentration of follicular phase follicle stimulating hormone (FSH), the presence or absence of polycystic ovaries (PCO) and any evidence of ovarian response to previous stimulation, it is not uncommon to encounter patients with excessive multiple ovarian follicular recruitment culminating in a serious risk of severe OHSS. The events in gonadotrophin ovarian stimulation represent a spectrum with minimal response at one end and excessive follicular growth and the risk of severe OHSS at the other. In this context, severe OHSS is inherent in the current, commonly-used, ovarian stimulation protocols that employ high doses of gonadotrophins. Despite this presumption, it has been possible to identify cases where severe OHSS can be prevented. OHSS is mostly a direct consequence of excessive multiple follicular response to ovarian stimulation in the presence of a trigger dose of exogenous human chorionic gonadotrophin (HCG) or pregnancy-generated endogenous HCG (Navot et al., 1988). Excessive multiple follicle recruitment and severe OHSS can be prevented in cases who experienced the same response in previous controlled stimulation treatment cycles when they are treated with bilateral or unilateral ovarian diathermy during the luteal phase of the menstrual cycle, and prior to starting subsequent gonadotrophin stimulation (Rimington et al., 1997; Egbase et al., 1998; Grudzinskas and Egbase, 1998). However, this strategy, yet to be explored further, has the drawback of being surgically invasive with an attendant risk of substantial destruction of ovarian tissue that is required to mute the subsequent response to gonadotrophins. Comparative study of clinical and biological characteristics indicate that among OHSS patients PCO-like conditions, i.e. hyperandrogenism, anovulation, and a luteinizing hormone (LH)/FSH ratio of >2, are more frequent (Delvigne et al., 1993) in patients with PCO compared with controls have significantly higher serum oestradiol concentrations on the day of HCG administration, developed more follicles and produced more oocytes (MacDougall et al., 1993). Thus, it is important to diagnose PCO before ovarian stimulation is initiated as these patients are more likely to develop severe OHSS (Dale et al., 1991; Buyalos and Lee, 1996). There is evidence that severe OHSS is preventable in patients with polycystic ovarian syndrome (PCOS) when ovarian stimulation is achieved with a chronic low-dose gonadotrophin protocol, compared with a conventional regimen (Homburg et al., 1995; Balasch et al., 1996; Edwards et al., 1996; ESHRE Capri Workshop, 1997). However, there seems to be a subset of obese (body mass index >30), clomiphene citrate-resistant, PCOS patients in our clinical practice with inherent idiosyncratic excessive multiple follicular response to ovarian stimulation, despite administering low-threshold doses of gonadotrophin. The pattern of ovarian follicular growth in these patients appears to be an `all or none' phenomenon to low threshold doses of gonadotrophin; when there is a response it is always excessive. These patients usually require prolonged stimulation, culminating with large doses of gonadotrophins to achieve the criteria for the administration of a trigger dose of HCG, i.e. a leading follicle of ⩾18 mm. In this category of patients, the risk of severe OHSS could be considered `endemic' with treatment cycles subjected to repeated cancellations and, if completed, to oocyte retrieval, severe OHSS becomes inevitable (Egbase et al., 1999). These are the cases in which severe OHSS is probably not preventable with current strategies except cancellation of the cycle. None of the strategies currently employed to avert severe OHSS when excessive multiple follicular growth is encountered in the course of controlled ovarian stimulation completely prevents the condition. We have confirmed in prospective randomized, controlled trials that follicular aspiration either prior to or after HCG administration does not prevent severe OHSS (Egbase et al., 1997, 1999). Numerous studies have also demonstrated the apparent ineffectiveness of human albumin administered around the time of oocyte retrieval to always prevent severe OHSS (Ndukwe et al., 1997; Chen et al., 1998). Likewise, cryopreservation of all resulting embryos when there is serious risk of severe OHSS at the best prevents pregnancy-associated late OHSS but not the early onset of the condition that is precipitated with the trigger HCG dose (Lyons et al., 1994; Queenan et al., 1997). There is renewed optimism in the strategy of withholding gonadotrophins whilst continuing pituitary down-regulation (coasting) in patients with excessive multiple follicular response encountered in the course of controlled ovarian stimulation for the prevention of severe OHSS (Sher et al., 1995; Benadiva et al., 1997). Patients are commonly identified to be at risk of severe OHSS when the serum oestradiol concentration is high (>3000 pg/ml) and ovarian follicles >20. Neither the serum oestradiol concentration nor the number of ovarian follicles accurately predict the incidence of OHSS and this is possibly related to the lack of complete understanding of its pathophysiology (Meirow et al., 1996). The literature on coasting is replete with differing criteria that have been used either to start or conduct the strategy, the serum oestradiol or the leading ovarian follicle at onset of coast varying from 2500–6000 pg/ml and 15–18 mm respectively whilst the duration of coasting has ranged from 1–11 days (Sher et al., 1995; Benadiva et al., 1997; Dhont et al., 1998; Lee et al., 1998; Tortoriello et al., 1998; Fluker et al., 1999; Waldenstrom et al., 1999). In addition, none of these studies were randomized, controlled trials. In the only prospective randomized, controlled trial reported to date (Egbase et al., 1999) that compared prolonged coasting with early unilateral follicular aspiration, the incidence of severe OHSS was similar in both strategies (20 versus 26.6% respectively). Detailed analysis of the daily serum oestradiol concentrations during the period of coasting in this study confirmed the poor predictive value of serum oestradiol in the prevention of severe OHSS. The oestradiol concentration at the onset of coasting, percentage decrement per day and rate of decline during coasting was similar in patients with or without severe OHSS (Egbase et al., 1998). Notwithstanding, it is possible that coasting could be more effective in the prevention of severe OHSS if the events of coasting were monitored with a more reliable marker that could accurately predict severe OHSS. In this respect, the vascular endothelial growth factor (VEGF) could probably be the appropriate candidate as there is increasing evidence linking serum, follicular or ascitic concentrations of VEGF to the risk, incidence or clinical course of OHSS (Abramov et al., 1997; Lee et al., 1997; Agrawal et al., 1999). The fall in serum oestradiol concentrations during coasting is thought to be related to apoptosis or atresia of the granulosa cells (Benadiva et al., 1997). It would be interesting to study what happens to serum VEGF concentrations during coasting. Overall, the risk of severe OHSS appears to be inherent in the current commonly employed ovarian stimulation protocols that utilise relatively high doses of gonadotrophins, coupled with high-risk patient groups, e.g. those with PCOS. To this extent, the assertions of Abramov et al. (1999) are probably appropriate. However, it is possible to prevent severe OHSS in cases identified at risk of the condition when ovarian diathermy is performed in the index cycle prior to ovarian stimulation or chronic low-dose gonadotrophin protocols are used in clomiphene citrate-resistant PCOS patients.