Ovarian hyperstimulation syndrome

"Ovarian hyperstimulation syndrome" is an L2 condition from Section 2.1.12 in the second edition of the CICM Syllabus for the Second Part Examination. It appeared in the papers suddenly, as Question 13 from the first paper of 2025, so shortly following the development of the syllabus that many were probably unaware that it was even an option. The result was that only 41.3% of the candidates cleared what was already a low (3.76) bar set by the Angoff committee. To be fair, a borderline candidate would be unlikely to readily yield four risk factors for this condition, and even a highly motivated candidate would not have been able to summon "examination features and key investigation results" to score highly, as this condition is a rarity in Intensive Care as it is practiced in Australia. Considering that it is usually a generalised abdominal complaint by a young woman, presenting with OHSS to the Emergency Department will have you dismissed with paracetamol and GP followup as an anxious princess with dysmenorrhoea, and nobody in the ICU will ever have become aware of your existence unless you have a cardiac arrest.

In summary:

Amin et al (2025), if one is able to acquire a copy, is an excellent overview which contains enough material on its own to answer Question 13 without any additional literature. Palomba & Caserta (2023) also had the kind of detail that made it easy to write this summary. 

Definition of ovarian hyperstimulation syndrome

OHS or OHSS doesn't really have an accepted definition as much as a catchy name, like

"life-threating iatrogenic complication of the early luteal phase and/or early pregnancy, caused by an excessive response to ovarian stimulation"

- Palomba & Caserta, 2023

One occasionally sees these "definitions" include the words "iatrogenic", but in fact an ovary can snap and go berserk without any obstetric provocation, as in the case of Dey et al (2015). The authors found many cases of spontaneous OHSS, occuring both in normal pregnancy and in hydatiform moles. They also described hyperreactio luteinalis, some kind of horrifying Harry Potter curse which results in multicystic enlargement of the ovaries, and which mimics OHSS during spontaneous pregnancy. In short, to be completely pedantic, one would want to call this a postovulatory syndrome due to the simultaneous triggering of multiple follicles.

Of course, most of the time it is iatrogenic.  We owe this interesting complication to the pioneering work of Brune Lunenfield in the 1960s. The modern-day trainee may be unaware that only this time lastc century we were doing pregnancy tests by the Ascheim-Zondek reaction, where instead of irrigating a plastic testing stick, the urine of a pregnant woman would be injected into a sexually immature female mouse. The hCG in the urine would be recognisable to the mouse, and its ovaries would enlarge to 2-3 times their normal size (hence the name "gonadotropin"). The demand for the test is said to have been prodigious in the 1930s, such that the Daily Mirror reported on one mouse farm which apparently contained up to 200,000 animals, sacrificing up to 3,000 of them each day. The moral and financial burden of 200kg of dead mice every week was clearly too much and they eventually moved on to Xenopus laevis,  the African clawed frog, who will simply extrude eggs when stimulated with pregnant urine, and does not need to be killed each time. 

Observing the reaction of the toads, some obstetricians took notice, and must have said to themselves, why not our clients also? Lunenfeld extracted hCG from urine and gave it to amenorrhoeic human volunteers to induce ovulation. Some ovulated safely; some had sextuplets; several died. In the present day the main difference seems to be the awareness of this possibility as a complication. 

Aetiololgy of ovarian hyperstimulation syndrome

How does this happen? Well. The author, who is the first to admit that he is not an authority on reperoductive health, hides behind the like of Martinez et al (2021) and Jain & Singh (2022) to avoid being held responsible for the following summary:

• Ordinarily, a dominant antral follicle will sabotage the development of its other sister follicles by hosing them down with oestrogen, thereby limiting ovulation to a solo performance. But assistive reproductive technology calls for many ova, because this increases the yield of artifical insemination and in vitro fertilisation. The reason for this is manifold, but mostly related to the generally low likelihood of the implantation of a single egg. Fire more shots and you might hit something, they say. Ergo, we must have many ova. 
• Superovulation is the term that is usually used in reference to livestock and test animals, whereas the exact same techniques are rebranded as "controlled ovarian stimulation" for human use. It requires the administration of a large amount of recombinant FSH, 100–450 IU per day. Compared to the usual blood concentration of  4 -20 mIU/mL, that seems like a lot (i.e enough for 20L of blood). The stimulation of many follicles results, and no amount of oestrogen will stop them. 
• As these mature, they are observed via ultrasound. At around 18mm in diameter, a dose of hCG is administered to complete their maturation (hCG binds to the same receptor as LH, but is a much more potent signal, as it has a longer half life). This hCG injection seems to be the whole problem in OHSS, as without it,  you can have as many follicles as you like without the OHSS.
• The sustained stimulation of the follicles with hCG results in an excessive overproduction of VEGF by granulosa-lutein cells.  Neulen et al (1995) explain it better, but in summary,  these are corpus luteum cells which are mostly responsible for hoarding cholesterol (which gives them the yellow colour), and using it to power steroidogenesis, producing tons of progesterone. And the progesterone in turn sustains and empowers the development of the highly vascular uterine lining necessary for implantation and development of the syncytiotrophoblast. This is where the VEGF becomes important (angiogenesis is essential for this process.

So, in summary, OHSS is really VEGF hypersecretion syndrome, because VEGF seems to be responsible for the rest of the problems. VEGF, Vascular Endothelial Growth Factor, is a small-ish protein hormone with multiple isoforms ranging from 121 to 206 amino acids in size, which triggers angiogenesis when it binds to the tyrosine kinase receptor on target cells.  It is described as a"mitogen", i..e it stimulates mitosis. This might sound posive and down-farm homestyle wholesome, but in fact after about 150 minutes, an endothelium exposed to this cytokine will become more permeable to water by a factor of 20, liberate enough nitric oxide to vasodilate everything, and become so sticky to neutrophils that they will marginate and adhere to it. This is fine in the local environment of the ovary and uterus, but when enough VEGF is released, the results are systemic.

Complications and known sequelae of OHSS

Notably, in Question 13 from the first paper of 2025, the examiners asked for the clinical features and investigations that would be consistent with severe OHSS. The best way to describe these would probably be via this table from the RCOG classification:

Mild OHSS

• Abdominal bloating
• Mild abdominal pain
• Ovarian size usually < 8 cm

Moderate OHSS

• Moderate abdominal pain
• Nausea ± vomiting
• Ultrasound evidence of ascites
• Ovarian size usually 8–12 cm

Severe OHSS

• Clinical ascites (± hydrothorax)
• Oliguria (< 300 ml/day or < 30 ml/hour)
• Haematocrit > 0.45
• Hyponatraemia (sodium < 135 mmol/l)
• Hypo-osmolality (osmolality < 282 mOsm/kg)
• Hyperkalaemia (potassium > 5 mmol/l)
• Hypoproteinaemia (serum albumin < 35 g/l)
• Ovarian size usually > 12 cm

Critical OHSS

• Tense ascites/large hydrothorax
• Haematocrit > 0.55
• White cell count > 25 000/ml
• Oliguria/anuria
• Thromboembolism (venous and arterial)
• Acute respiratory distress syndrome

But these are just the severity criteria. Other complications which are not listed here include:

• Abdominal compartment syndrome
• Ovarian rupture and haemoperitoneum
• Ileus
• Cerebral oedema
• Pericardial effusion and tamponade
• Relative immunodeficiency (low circulating immunoglobulins)

Risk factors for ovarian hyperstimulation syndrome

Pakhomov et al (2021) and Sun et al (2021) were used to improve this table from Palomba & Caserta, 2023:

• Ovarian stimulation with gonadotropins
• previous OHSS
• Young age
• Black race
• Lean
• PCOS/PCOM
• Hyperinsulinemia
• Hyperandrogenism
• High peak of oestradiol 
• Multifollicular development (>18-20)
• High number of oocytes retrieved (>24)

Assessment of ovarian hyperstimulation syndrome

From the risk factors and complications listed above, the history examination and investigations become fairly selfexplanatory. An excellent resource for this is a paper by Timmons et al (2019):

History:

• Timeframe related to ART hormone injection (a subacute onset is probably something else)
• If possible, details of the agents used for ART, number of follicles on the last scan, etc.
• Un-ovarian sounding features, eg. bloody diarrhoea, dysuria/haematuria, productive cough, well established pregnancy
• Background risk factors, eg PCOS

Examination:

• Airway (oedema, plus also because intubation may be imminent)
• Features of pleural effusion and pulmonary oedema: creps, dull percussion note, etc. 
• Features of distributive or obstructive shock (the latter due to PE or tamponade), eg, tachycardia, distended neck veins, pulsus paradoxus, and so on.
• Level of consciousness (may be depressed due to cerebral oedema)
• Oliguria and generalised oedema
• Features that suggest ascites (eg. shifting dullness); a tense abdomen
• The absence of features to suggest liver disease
• Evidence of DVTs

Investigations:

• Intra-abdominal pressure measurement
• FBC (haemoconcentration, or worse, anaemia due to haemorrhage)
• EUC (renal failure), 
• LFT (liver failure),
• ABG (lactic acidosis)
• CXR (pleural effusion, but realistically, a bedside US will tell you more)
• Ultrasound of the abdomen and pelvis (to confirm ascites and see the size of the ovaries, plus it could give clues re. ovarian torsion)
• A pregnancy test
• Coags, as well as a group & hold

Management of ovarian hyperstimulation syndrome

"Supportive", a pointless term that can describe everything from hugs to ECMO, is the adjective usually applied to describe the management of this condition. It could be made more 

• Intubation. Treating the patient as somebody at an extremely high risk of aspiration, the intubation could be challenging, but may be required if the respiratory rate becomes very high and the tidal volume very small (because of worsening abdominal distension). Additionally, it has the benefit of permitting muscle relaxants, which could help your intraabdominal pressure. 
• This could be an ARDS that responds to steroids. Aman et al (2000) gave a huge dose (methylprednisolone 30 mg/kg) and "two days after starting the steroid therapy she was released from ICU in good condition"; other case reports also discuss varying doses with little apparent harm. Considering the risk of infection in these patients is increased by their lack of humoral immunity, you would have to be damn sure there is no infectious aetiology before ablating their immune system like that.
• Haemodynamic correction: note that correction rather than "support" was used, as often "support" implies "leave them on a high dose of noradrenaline". In this scenario many factors conspire against organ perfusion (volume shifts, vasodilation, pericardial fluid, IVC compression, abdominal compartment syndrom,e haemorrhage) of which some have immediate and relieving interventional options. Drain the pericardial fluid and ascites, and the noradrenaline may be unnecessary. Roll the patient on a 30° wedge. 
• Paracentesis. Many of the organ perfusion problems are coming from the tense ascites. Respiratory failure and oliguria should definitely be improved by draining it. The problem is that the ovaries are massive, and highly vascular, which makes the procedure risky. Moreover the ascitic fluid will continue to accumulate. Some sources recommend the placement of a soft pigtail catheter to allow the drainage to continue for days if needed. The mechanism of ascites is different from what one might see in usual liver disease associated ascites (for one, there is no portal hypertension) and so the usual treatments for ascites are otherwise ineffective. Or at least we do not know whether they would be effective. Only one study so far (Айзятулова, 2014) reported on the use of terlipressin in a mouse model of OHSS.
• Albumin is recommended by almost everybody for the replacement of vascular volume, but if one looks at each reference that is listed for this, one finds very eccentric  reasoning. Budev et al (2005), for example, claim that it "can be used as a plasma expander in cases of severe hemoconcentration", because high haematocrit; and refers to no specific research to support this. Chen et al (2011) leans on the Practice Committee of American Society for Reproductive Medicine (2008), who cite Forman et al (1990), a case series that does not make any recommendations other than to report that seven of their eight patients "received low salt albumin in an attempt to maintain colloid pressure".
• Thromboprophylaxis: the risk of venous thrombosis is extremely high; this is "supportive management" bordering on therapeutic preemptive strike. The risk of ovarian rupture and haemorrhagic shock is the only thing that keeps people from recommending a full dose of anticoagulation, it seems. Even outpatient management with self-injected prophylactic LMWH  seems to be popular (Wormer et al, 2018).

And, moving down into the realm of untested and experimental stuff which one might resort to under extreme circumstances, the following have been reported:

• Cabergoline: a dopamine agonist which Naredi et al (2014) reported on, seems to antagonise some of the effects of VEGF. It seems that dopamine regulates the downstream second messenger pathway of the VEGF ligand-receptor complex (the D2 receptor is involved). Cabergoline is a small molecule dopamine agonist with some haemodynamically counterproductive effects.  Álvarez et al (2007) used this in actual human women and it seemed to reduce the leakiness of their vessels.
• Anti-VEGF monoclonal antibody: seems logical? Bevacizumab is used to starve tumours of their oxygen supply by interfering with their prolific angiogenesis. It hits all isoforms of VEGF.  Ozdemir et al (2025) had promising results in a rat model, but to extend that to clinical use in humans seems some way off.
• GnRH antagonists, eg. ganirelix, work by rapidly disrupting the supply of LH which stimulates the VEGF-secreting granulosa-lutein cells. Rollene et al (2009) used this in their case series, and reported good effects. Nobody knows what this sort of crude interference with the endogenous endocrine voodoo of pregnancy will do to the developing embryo and endometrium, so it is perhaps better to freeze the collected eggs instead of trying to inseminate or implant fertilised embryos.
• Octreotide, if you are not interested in leaving any hormones, could be a brutally repressive final solution to the hormonal hyperstimulation problem. Karapanou et al (2012) reported an effective use of this drug to reduce spontaneous OHSS due to a pituitary macroadenoma. One should not be using this stuff in people who expect to immediately concieve.