Pelvic venous pressure
Personal or family history of thromboembolism without thrombophilia
Impaired vascular reactivity
collection for IVF. These two late complications suggest a generalized effect on
the coagulation system that may persist for several weeks.
Etiopathology of Thromboembolic Complications
A number of factors have been implicated in the pathogenesis of thrombo-
embolism following gonadotrophin therapy (Table V.1). These include:
hemoconcentration, a hypercoagulable state, the presence of thrombophilia,
pressure on pelvic vessels and impairment of vascular activity.
HEMOCONCENTRATION AND OHSS
Hemoconcentration is an important sign in severe OHSS (Rizk et al., 1990).
Hemoconcentration and increased hematocrit values, in the presence of
normal coagulation parameters, were found in nine of 25 patients by Schenker
and Weinstein (1978). Hemoconcentration was found in 95.2% of 209 patients
affected by severe OHSS (Abramov et al., 1999) and in 71.1% of 128 patients
affected by mild or moderate forms of OHSS (Delvigne et al., 1993).
Mechanical Factors and Thrombosis in OHSS
Venous compression due to enlarged ovaries and ascites, together with
immobility, may contribute to pelvic and lower limb thromboses in patients
with severe OHSS (Rizk 2001, 2002).
The term hypercoagulable state is used to refer to hematological changes that
may result in increased risk of thrombogenesis. The natural assumption in
assisted conception and ovulation induction cycles is that ÔÇ˜raised estrogenÔÇ™
concentration is to blame (Rizk 1993a, b).
Exogenous estrogens promote thrombogenesis in a dose-dependent fashion.
Endogenous estrogens are usually considered the cause of the increased risk of
HEMOCONCENTRATION AND OHSS
thromboembolism in pregnancy. During pregnancy, hematological changes
include increased concentrations of factor VII, VIII, IX, X and XII and ´¬übrin-
ogen. There is also reduced concentration of protein S and antithrombin III.
Kim et al. (1981) studied the response of blood coagulation parameters
to elevated estradiol induced by hMG. They reported a rise in ´¬übrinogen
concentration which correlated with the estradiol rise as the cycle progressed.
There were no signi´¬ücant changes in the prothrombin time or activated partial
thromboplastin time. There was a signi´¬ücant rise in Von Willebrand factor
and reduction of antithrombin III, indicating the presence of a relatively
hypercoagulable state. All these measurements were conducted after hCG
administration, and therefore considered not representative of the times at
which many women appear to experience thromboembolic complications of
treatment (Stewart et al., 1997b).
Aune et al. (1991) reported that ovarian stimulation for IVF induced
a hypercoagulable state. They followed up 12 IVF cycles, measuring whole
blood clotting time (WBCT), whole blood clot lysis time (CLT), antithrombin
III, plasma ´¬übrinogen and factor VII, both before stimulation and after hCG
administration, at the peak of estradiol concentration. Their ´¬ündings showed
a signi´¬ücant increase in ´¬übrinogen and reduction in antithrombin III
concentration, and a signi´¬ücant increase in CLT over this time, implying
a disruption of the balance of coagulation and thrombolysis leading to a relative
increase in coagulability. In contrast, Lox et al. (1995) assessed coagulation
parameters up to 14 days after hCG and found that none of the changes were
out of their laboratoryÔÇ™s normal ranges. Although they noted signi´¬ücant
correlations of some factors with changes in estradiol, and in prothrombin time
(PT) and partial prothrombin time (PTT), they concluded that these changes
could not constitute a promotion of coagulability, challenging the ´¬ündings
of Aune et al. (1991).
Stewart et al. (1997b) proposed an interesting question. If hyperestrogen-
ism alone is responsible for the risk of thromboembolic disease, then why
are many of these cases diagnosed some time after the expected peak estradiol
concentration that is usually the time of hCG administration? Stewart et al.
(1997b) proposed several explanations to that question. The ´¬ürst explanation is
that OHSS simply prolongs the risk period of thromboembolism; the second
explanation is that prolonged hepatic dysfunction secondary to OHSS might
contribute to the hypercoagulable state (Ryley et al., 1990; Mills et al., 1992;
Stewart et al., 1997b). The third explanation relies on the fact that high
endogenous estradiol may increase free protein S and therefore be thrombo-
protective, and hence explains delayed thrombogenesis if the fall in estradiol
is prolonged in OHSS (Stewart et al., 1997b).
Hypercoagulable State in OHSS
Kodama et al. (1996) studied the plasma hemostatic markers in OHSS. They
found that the levels of thrombin├Çantithrombin III and plasma a2-antiplasmin
complexes in the plasma began to rise within a few days after hCG
98 COMPLICATIONS OF OVARIAN HYPERSTIMULATION SYNDROME
administration, with signi´¬ücantly higher levels during the midluteal phase.
In OHSS patients who became pregnant, elevation of these markers continued
for three weeks after the onset of the disease. There were also some
characteristic changes in OHSS cycles in other hemostatic markers, such as
a decrease in the levels of antithrombin III and prekallikrein, and shortened
activated partial thromboplastin time. The practice in that unit was to provide
preoperative subcutaneous heparin (5000 IU, single dose) in women under-
going oocyte retrieval, and in women hospitalized with symptomatic OHSS
(subcutaneous 5000 IU, twice daily). Balasch et al. (1996) found an increased
expression of induced monocyte tissue factor by plasma from patients with
severe OHSS. This increase in the procoagulant activity of blood monocytes,
which is mediated principally by tissue factor expression, may be important in
thrombotic events associated with the syndrome.
Increased levels of factor V, platelets, ´¬übrinogen, pro´¬übrinolysin and
´¬übrinolytic inhibitors were observed by Phillips et al. (1975). Thrombophlebitis
occurred in two out of 25 patients. Kaaja et al. (1989) described deep venous
thrombosis in a case of severe OHSS. They suggested that raised Von
Willebrand factor could result from increased platelet adhesion potentiated by
hemoconcentration, and therefore could be an impending marker for OHSS
development. Todorow et al. (1993) performed a retrospective study to evaluate
the changes in Von Willebrand factor in OHSS.
Kodama et al. (1997) studied the status of the plasma kinin system in
patients with OHSS in order to investigate whether activation of the plasma
kinin system correlates with increased blood coagulability. They concluded
that activation of the plasma kinin system occurs speci´¬ücally and occasionally
in OHSS patients, and is associated with increased blood coagulability, and
that, when an OHSS patient demonstrates a low value of plasma kinin, more
careful management is required to prevent thromboembolic complications.
Thrombophilia and Thromboembolism in OHSS
Thrombophilias are important because they could potentiate the risk of
thrombosis under conditions of relative hypercoagulability that occur in
assisted conception, particularly when complicated with OHSS (Rizk, 1993a;
Ryo et al., 1999). The major thrombophilias include factor V Leiden mutation,
protein C, protein S and antithrombin III de´¬üciencies and antiphospholipid
syndrome (Table V.2).
Dulitzky et al. (2000) prospectively evaluated the prevalence of markers
of thrombophilia in 20 women hospitalized for severe OHSS. The following
markers were assessed: plasma levels of antithrombin III, protein S and protein
C, antiphospholipid antibodies, factor V Leiden mutation, and mutation of
the methyltetrahydrofolate reductase (MTHFR) gene. In patients with OHSS,
85% were carriers of one or more positive markers of thrombophilia. All
the thrombotic events occurred in women who had more than one marker
for thrombophilia (40%). Furthermore, 27% of controls were carriers of one
marker of thrombophilia and none carried more than one marker.
HEMOCONCENTRATION AND OHSS
Table V.2 Thrombophilia and OHSS
Author Year Thrombophilia factor
Kaaja et al. 1989 Low antithrombin III
Beni´¬‚a et al. 1994 Decreased protein S activity
Hollemaert et al. 1996 Leiden factor V mutation
Hortskamp et al. 1996 Leiden factor V mutation
2003 Prothrombin 30 UTR and factor V Leiden mutation
McGowan et al.
In contrast, Delvigne et al. (2002) using a matched control study of 25
patients with OHSS and observed no increase in the prevalence in the markers
of thrombophilia for several months after the OHSS resolved. Delvigne and
Rozenberg (2003) suggested that these discordant results may be due to the
different times of screening for the thrombophilia markers. Whereas Dulitzky
et al. (2000) evaluated their patients in the acute phase of OHSS, Delvigne et al.
studied the thrombophilia markers after the OHSS episode. During OHSS, the
coexisting hyperestrogenemia may contribute to the decrease in antithrombin
III or protein S levels.
Fabregues et al. (2004) conducted a case control study on prevalence of
thrombophilia in women with severe OHSS and cost-effectiveness of screening.
The cost of preventing one thrombotic event in a patient developing severe
OHSS after IVF and having factor V Leiden or prothrombin G20210A
mutations was calculated. None of the OHSS patients or controls had anti-
thrombin, protein C, or free protein S de´¬üciencies. All of them tested negative
for antiphospholipid antibodies.
Todros et al. (1999) reported a case of spontaneous OHSS occurring in
a pregnant woman carrying the Factor V Leiden mutation. Even though
prophylactic treatment for thrombo-embolism was adopted by administering
low-molecular-weight heparin, the pregnancy was complicated by thromboses
of the left subclavian, axillary, humeral and internal jugular veins during
the second trimester of gestation. The pregnancy was managed conservatively,
and a healthy baby was delivered at term. In order to avoid unnecessary
laparotomy the authors emphasized the importance of careful diagnosis
in order to differentiate spontaneous OHSS from ovarian carcinoma, as well as
the necessity to look for the presence of coagulation disorders in women
affected by OHSS.
A 33-year-old female developed OHSS with thrombosis of the right
internal jugular vein, subclavian vein and superior vein cava following IVF
(Lamon et al., 2000). As pregnancy progressed, edema, pain and tingling
sensation developed. CT scan con´¬ürmed thrombus with the right internal
jugular and subclavian vein and a free ´¬‚oating tip in the superior vein
cava. Following treatment with intravenous heparin and subcutaneous
low-molecular-weight heparin until delivery, the symptoms improved.
100 COMPLICATIONS OF OVARIAN HYPERSTIMULATION SYNDROME
Andrejevic et al. (2002) reported a 28-year-old patient with polycystic
ovary syndrome who presented with fever and laboratory markers of
in´¬‚ammation. Intracardiac thrombosis was diagnosed in the presence of
antiphospholipid antibodies. The authors suggested that primary antiphos-
pholipid syndrome was possibly triggered by ovulation induction.
Elford et al. (2002) presented a case of a previously healthy woman who
underwent in-vitro fertilization and experienced a middle cerebral artery
thrombosis that was subsequently lysed with intra-arterial recombinant tissue
plasminogen activator (rt-PA). To the authorsÔÇ™ knowledge this was the ´¬ürst
reported case of successful use of rt-PA to lyse a cerebral arterial thrombus
resulting from severe OHSS. The patient made a nearly complete neurologic
recovery and delivered a healthy infant at term, illustrating that intra-arterial
thrombolysis can be used with relative safety even in very early pregnancy. Ulug
et al. (2003) reported a case of bilateral internal jugular venous thrombosis
following successful assisted conception in the absence of OHSS.
McGowan et al. (2003) reported a case of deep vein thrombosis followed
by internal jugular vein thrombosis as a complication of IVF in a woman
heterozygous for the prothrombin 30 UTR and factor V Leiden mutations.
They suggested that neck pain and swelling in a pregnant woman, especially
one that has undergone IVF, should be taken seriously and investigated with
duplex scanning and/or MRI. Women with a personal or family history
of thrombosis who are undergoing IVF should be made fully aware of the
potential thrombotic risks, and should be considered for a thrombophilia
Ou et al. (2003) reported a case of thromboembolism after ovarian
stimulation and successful management of a woman with superior sagittal sinus
thrombosis after IVF and embryo transfer. The authors recommended dose-
adjusted heparinization as the ´¬ürst-line treatment of choice, while intravascular
thrombolysis or operative thrombectomy is an aggressive but effective treat-
ment. Continuation of pregnancy is considered safe without any increased risk
of fetal congenital anomalies.
Nakauchi-Tanaka (2003) reported on a 31-year-old nulligravida woman
who developed an acquired factor VIII inhibitor associated with severe OHSS.
She developed hematuria, ecchymosis and intramuscular bleeding following
the severe OHSS. Laboratory examinations showed a markedly prolonged
activated partial thromboplastin time and a low level of factor VIII activity.
Treatment with prothrombin complex concentrate and factor VIII inhibitor
bypassing agent was successful in reducing the inhibitor, so that she delivered
a healthy baby via spontaneous vaginal delivery. Acquired hemophilia is a
life-threatening disorder. This was the ´¬ürst case report of acquired hemophilia
Personal or Family History of Thromboembolic Disease
Several cases of thromboembolic complications have been reported in patients
who had either a personal or family history of thromboembolic disease without
RARE VASCULAR COMPLICATIONS ASSOCIATED WITH OHSS
evidence of familial thrombophilia (Dalrymple et al., 1982├Ç3; Boulier et al.,
1989; Thill et al., 1994; Benshushan et al., 1995; Stewart et al., 1997a, b).
Impairment of Vascular Reactivity
Foong et al. (2002) performed measurement and quanti´¬ücation of the
cutaneous arteriolar vasoconstrictor response using laser doppler ´¬‚uximetry.
Women with OHSS have impaired vascular reactivity when compared with
normal women. Foong et al. (2002) suggested that OHSS is associated
with reversible impairment of vascular reactivity that may contribute to the
increased prevalence of thromboses.
RARE VASCULAR COMPLICATIONS
ASSOCIATED WITH OHSS
Central Retinal Artery Occlusion
Turkistani et al. (2001) reported a case of severe OHSS presenting with central
retinal artery occlusion.
Mancini et al. (2001) reported a case of forearm amputation after ovarian
stimulation for IVF-ET. The patient underwent many cycles of IVF-ET. She
had a coagulation disorder as a result of OHSS, with thrombosis of the axillary
vein recurring after thromboarterectomy and leading to the paradoxical result
of the amputation of an arm.
Cerebral infarction represents the most serious complication of thrombo-
embolism associated with OHSS (Neau et al., 1989; Rizk and Aboulghar, 1991;
Yoshii et al., 1999; Dumont et al., 2000). The fatal case reported by Cluroe and
Synek (1995) represents the worst of a spectrum of cases. More than 14 cases of
stroke have been reported following OHSS. Six cases were treated with
intravenous heparin and at least one case treated with interarterial recombinant
tissue plasminogen activator (Elford et al., 2002).
Myocardial infarction associated with severe OHSS is extremely rare (Ludwig
et al., 1999; Akdemir et al., 2002). Ludwig et al. (1999) reported the ´¬ürst case
of myocardial infarction after ovarian hyperstimulation. A 35-year-old patient,
77 kg in weight, and a heavy smoker (30 cigarettes/day) underwent an IVF
Intracytoplasmic Sperm Injection (ICSI) cycle. Decapeptyl depot 3.75 mg
was administered and recombinant FSH 150 IU was started two weeks later.
102 COMPLICATIONS OF OVARIAN HYPERSTIMULATION SYNDROME
The dose was increased to 225 IU on the ´¬üfth day and hCG 10 000 IU was
given on the sixteenth day of stimulation. The serum estradiol level was
12 892 pmol/l on the day of hCG, and 16 oocytes were retrieved and three
embryos were fertilized and transferred two days later. Human CG 5000 IU
was given again on the day of transfer and micronized progesterone 600 mg was
administered intravaginally each day. The patient returned at midnight to
the emergency department with severe backache and dyspnea and a hematocrit
of 48%. At the coronary care unit, 5000 IU of heparin and 500 mg of
acetylsalicylate were given immediately, and a coronary angiogram demon-
strated a distal occlusion of the left anterior descending coronary artery. Due
to intracoronary thrombotic material, recannulization of the distal left anterior
descending (LAD) artery by percutaneous transluminal coronary angioplasty
was unsuccessful. A stent to the LAD artery only slightly improved the
blood ´¬‚ow. The patient had an otherwise uncomplicated clinical course, and
a follow-up coronary angiogram after 10 days and again after six months
revealed no reanastomosis or thrombotic material, and the stented segment
of the proximal LAD coronary artery remained open while the distal LAD
coronary artery remained occluded. Results of all coagulation pro´¬üles were
within normal limits and the patient did not become pregnant. Akdemir et al.
(2002) reported a case of a patient with myocardial infraction associated
Respiratory complications can present with a variety of symptoms and signs in
OHSS (Rizk, 1992, 1993b, 2001).
Abramov et al. (1999a) observed pulmonary complications in 7.2% of
severe OHSS cases (Table V.3). Dyspnea and tachypnea were the most common
symptoms appearing in 92% of these cases. A small proportion of patients
that had pulmonary manifestations presented with complications, such
as local pneumonia (4%), adult respiratory distress syndrome (2%) and
pulmonary embolism (2%) (Abramov et al., 1999a). Pleural effusion was
observed in 21% of 128 cases of OHSS in a Belgian multi-center study
(Delvigne et al., 1993).
Unilateral Pleural Effusion
Jewelewicz and Van de Wiele (1975) reported the ´¬ürst case of pleural effusion
as the sole presentation of OHSS following superovulation. Following
the introduction of IVF, Kingsland et al. (1989) from the Hallam Medical
Center reported the ´¬ürst case of right unilateral pleural effusion in an IVF
patient. Wood et al. (1998) reported a case of symptomatic pleural effusion
following an ICSI cycle in a patient who had previously undergone
right hemicolectomy for CrohnÔÇ™s disease. The ´¬ürst aspirate removed 800 ml
of serous ´¬‚uid. The protein content was 53 mg/l which contrasts with our case
Table V.3 Pulmonary and extrapulmonary
features of 209 patients hospitalized with
severe OHSS in Israel from 1987 to 1996
Reproduced with permission from Abramov
et al. (1999a). Hum Reprod 71:645├Ç51
No. (%) of patients
Dyspnea 193 (92)
Pneumonia 8 (4)
ARDS 5 (2)
Pulmonary embolism 4 (2)
Ascites 207 (99)
Gastrointestinal 112 (54)
Oliguria 62 (30)
Peripheral edema 28 (13)
Peritoneal irritation 13 (6)
Acute renal failure 3 (1)
(Kingsland et al., 1989). Rabinerson et al. (2000) reported severe unilateral
hydrothorax as the only manifestation of the OHSS. A 35-year-old woman
presented with mild dyspnea two weeks after ovarian stimulation with hMG
and hCG and IVF-ET. Chest X-ray revealed a large pleural effusion on the
right side. Three consecutive thoracocenteses were needed to drain a total
of 6800 cc of ´¬‚uid. Following drainage, the respiratory symptoms disappeared.
An uneventful pregnancy progressed.
Cordani et al. (2002) discussed a case of massive unilateral hydrothorax
as the only clinical manifestation of OHSS. A decade later, we have described
a case very similar to the ´¬ürst case following IVF (Gore et al., 2002) which
we reported in the late 1980s and summarized the published literature
regarding OHSS presenting only as pleural effusion (Table V.4).
Pathophysiology of Unilateral Pleural Effusion in OHSS
Several authors have investigated the pathophysiology of unilateral pleural
effusions in severe OHSS, particularly in the absence of ascites (Kingsland et al.,