OHSS has two primary components: (a) enlargement of the ovaries accompanied by the formation of follicular, luteal, and hemorrhagic ovarian cysts and edema of the stroma, and (b) an acute shift of fluid out of the intravascular space (3). The syndrome is more frequent in cycles resulting in pregnancy (4). The exact pathogenesis of this syndrome is not clear. At one time, it was thought that the OHSS resulted from high local concentrations of estrogen in the ovaries causing altered capillary permeability and ascites, which, in turn, led to the pleural effusion. This does not appear to be the sole explanation, however, because the syndrome can still be produced in rabbits when the ovaries are exteriorized (5). This indicates that there must be systemic effects involved in the fluid shifts into the peritoneal and pleural cavities. It is now believed that the syndrome is precipitated by an ovarian product, vasoactive peptide, or cytokine that has been released into the peritoneal cavity by the ovary or that has gained access to the systemic circulation directly from the corpus luteum or serosal vessels. Three likely candidates are vascular endothelial growth factor (VEGF) (6,7), interleukin 8 (IL-8) (7), and interleukin 6 (IL-6) (6). All three have been shown to be markedly elevated in the follicular fluid and ascites. Antibodies to VEGF reduce the ascitesinduced endothelial permeability to 44% of control and antibodies to IL-8 reduce the ascites-induced endothelial permeability to 34% of control (7). Antibodies to IL-6 do not significantly affect ascitesinduced endothelial permeability (7). It has also been shown that hCG upregulates the VEGF expression of granulosa cells in the OHSS, but not in control groups (8). The serum levels of VEGF are elevated in patients with OHSS (8). The mean levels of VEGF in ascitic fluid with the OHSS (7) are higher than the mean level of VEGF in any type of pleural effusion (7). The mean level of VEGF in the pleural fluid with the OHSS is elevated but is only about 60% that in ascitic fluid (9).

There are probably two factors responsible for the accumulation of pleural fluid with OHSS. In patients with bilateral effusions, the probable mechanism is a generalized capillary leak syndrome. In patients with large right-sided pleural effusions, it is probable that the fluid moves directly from the peritoneal space to the
pleural space. Evidence supporting this mechanism is the fact that the effusions are frequently large and rightsided, many patients have ascites, and the observation in one patient that the pleural fluid IL-6 level was more than 100 times higher than the simultaneously obtained serum level (10). Obviously, the pleural fluid in this case was not due to a generalized capillary leak syndrome.

Patients with OHSS initially develop abdominal discomfort and distension, followed by nausea, vomiting, and diarrhea. As the syndrome worsens, the patients develop evidence of ascites and then hydrothorax or breathing difficulties. In the most severe stages, the patients develop increased blood viscosity due to hemoconcentration, coagulation abnormalities, and diminishing renal function (3). Respiratory symptoms develop 7 to 14 days after the hCG injection (4).

The pleural effusions with OHSS are usually right sided. In the series of 33 patients with pleural effusions reported by Abramov et al. (11), 17 effusions (52%) were right sided, 9 (27%) were bilateral, and 7 (21%) were unilateral left sided. At times, a pleural effusion may be the sole manifestation of OHSS (12). The pleural effusion can be a significant problem, as evidenced by one patient who had 8,500 mL pleural fluid aspirated from her pleural space over 14 days (13). The pleural fluid in patients with OHSS is an exudate. In the series of Abramov et al. (14), the mean pleural fluid protein was 4.1 g/dL, whereas the mean serum protein was 4.4 g/dL.

The incidence of the syndrome can be reduced if the serum estrogen levels and the number of ovarian follicles are monitored. If the serum estrogen levels are very high or if there are more than 15 ovarian follicles with a high proportion of small and intermediate size follicles, hCG should be withheld (2).

The pathogenesis of fetal pleural effusion is not known and is possibly multifactorial. There is some evidence that the fetal pleural effusions are actually chylothoraces. Benacerraf and Frigoletto (20,21) analyzed the pleural fluid from two fetuses and reported abundant lymphocytes in both, but one had almost all T lymphocytes, whereas the other had a mixture of T and B lymphocytes. The presence of the lymphocytes was suggestive of a chylous effusion. A ratio of the pleural fluid to serum IgG that exceeds 0.6 is very suggestive of a neonatal pleural effusion (22). In addition, most neonatal pleural effusions, which in many cases are continuations of fetal pleural effusions, are chylothoraces (23). Analysis of the pleural fluid for chylomicrons is not useful in the diagnosis of fetal chylothorax because the fetuses are not eating any lipids.

In a review of the literature for isolated fetal pleural effusion in 1998, 204 cases were found (24). The fetal pleural effusions were bilateral in 74%, unilateral right sided in 11%, and unilateral left sided in 14% (24). Polyhydramnios was noted in 72% and hydrops fetalis in 57% (24). The reason for the polyhydramnios is not clear, but it has been suggested that the increased intrathoracic pressure may interfere with normal fetal swallowing. There is a high frequency of chromosomal abnormalities in fetuses with pleural effusions. In one series, the prevalence of chromosomal abnormalities was 50% in 152 fetuses with other sonographic abnormalities and 12% in 94 patients with isolated pleural effusion (25).