Immune hydrops fetalis is a serious fetal condition in which abnormal fluid collections accumulate in at least two different fetal compartments, and in which circulating antibodies against red cell antigens are detected in the mother. Hydrops fetalis is associated with a pathologic increase in interstitial and total fetal body water, usually appearing in fetal soft tissues and serous cavities. The cause may be either immunologic or nonimmunologic, depending on the presence or absence of maternal antibodies against fetal red cell antigens. While previously it was considered that the majority of cases of hydrops fetalis were secondary to maternal–fetal blood group incompatibilities, it is now estimated that the causes are nonimmunologic in more than 90% of cases (see Chapter 128) (Santolaya et al., 1992).

Immune hydrops is most likely a result of maternal–fetal Rhesus (Rh) blood group incompatibility, in which maternal antibodies to certain fetal blood group antigens cross the placenta, causing hemolysis of fetal blood and profound fetal anemia. This process is also referred to as Rh isoimmunization. An in-depth discussion of the pathogenesis of Rh blood group incompatibility is beyond the scope of this textbook. Briefly, the five antigens that make up the Rh system are D, C, c, E, and e. The presence of the D antigen confers Rh blood group positivity, while its absence indicates Rh negativity. Approximately 15% of Caucasians are Rh negative, while 30% of Basques are Rh negative, and 4% to 8% of blacks are Rh negative (Bowman, 1999). A total of 45% of Rh-positive individuals are homozygous, so all their offspring will be Rh positive. By contrast, 55% of Rh-positive individuals are heterozygous, so there is a 50/50 chance that their offspring will be Rh positive. If a first-time mother is Rh negative and her fetus is Rh positive, there is a 16% risk that fetal Rh antigen will stimulate the maternal immune system to produce anti-D antibody (Bowman, 1999). This first immune response generally consists of anti-D IgM antibody, which cannot cross the placenta to cause fetal hemolysis. In a subsequent pregnancy, if the fetus is again Rh positive, a more rapid immune response occurs, consisting of high titers of anti-D IgG antibody, which rapidly crosses the placenta to produce fetal hemolysis and profound fetal anemia.

The mechanism by which maternal–fetal Rh incompatibility produces immune hydrops fetalis is complex and the exact pathophysiologic process has not yet been elucidated. It is known that maternal anti-D IgG antibody attaches itself to the Rh antigen present on fetal red cells. This results in chemotaxis of phagocytes in the fetal spleen, which leads to destruction and hemolysis of fetal red cells. In response, the fetus produces more erythropoietin, which stimulates the fetal bone marrow to increase red cell production. Eventually, marrow capacity is reached, and extramedullary erythropoiesis occurs, with fetal red cell production in the fetal liver, spleen, kidney, adrenal glands, and intestine. Fetal hepatosplenomegaly is, therefore, common. Red cells produced at these sites are often immature, are nucleated, and appear in the circulation as erythroblasts. Hence, the synonym erythroblastosis fetalis for immune hydrops.

The majority of cases of Rh isoimmunization lead to mild or moderate fetal or neonatal hemolytic disease. However, approximately 20% to 25% of cases result in severe hemolytic disease, with immune hydrops developing in utero (Bowman, 1999). Hydrops develops in approximately half of these fetuses between 18 and 34 weeks of gestation, with hydrops developing in the remaining fetuses between 34 weeks and term. The pathophysiology of fetal hydrops is unclear. Several hypotheses have been suggested. One thought is that simple congestive heart failure secondary to fetal anemia leads to hydrops. Another more complicated hypothesis is that severe fetal anemia leads to extensive extramedullary erythropoiesis, with associated hepatosplenomegaly and distortion of intrahepatic architecture. This distortion results in portal and umbilical venous distortion, portal hypertension, placental edema, and placental hypoperfusion (Bowman, 1999). With deteriorating hepatic synthesis, progressive hypoalbuminemia occurs, adding to the generalized edema, anasarca, and pleural and pericardial effusions. However, a recent study calls into question whether or not hypoalbuminemia is the cause of immune hydrops (Pasman et al., 2006). Pasman et al. evaluated fetal blood samples taken at the first fetal blood transfusion in 244 Rh-D alloimmunized pregnancies. Hemoglobin concentration, albumin concentration, and severity of hydrops were assessed. Most fetuses (71%) with immune hydrops had an albumin concentration within the normal range suggesting that decreased albumin is unlikely to cause hydrops. Nonetheless, there was a negative correlation between the degree of fetal hydrops and the fetal serum albumin level suggesting that hypoalbuminemia occurs as a secondary effect in hydrops. The authors suggest that cardiac failure and lymphatic flow obstruction may be more important causative factors of hydrops.

Effective prevention programs to reduce the incidence of Rh(D) isoimmunization, using passive maternal immunization with Rh immunoglobulin, have significantly decreased the numbers of fetuses with immune hydrops. Because of the efficacy of these programs, non-Rh(D) blood group isoimmunization is becoming proportionately more frequent as a cause of immune hydrops. Immune hydrops secondary to ABO blood group incompatibility is extremely rare, as the resulting hemolytic disease is almost always mild in nature (Gilja and Shah, 1988; Bowman, 1999). Isoimmunization due to non-Rh(D) and non-ABO incompatibility usually occurs as a result of blood transfusions, with such atypical antibodies developing in 1% to 2% of recipients (Bowman, 1999). At least 60 such atypical antibodies have been identified as potentially causing hemolytic disease of the fetus or newborn, with anti-c, anti-E, and anti-Kell antibodies being the most important causes of severe disease, including hydrops (Bowman, 1990). Fetal anemia secondary to Kell isoimmunization differs from that due to Rh(D) isoimmunization, as the mechanism for anemia is most likely erythroid suppression rather than hemolysis (Vaughan et al., 1994).

The risk of immune hydrops in a first Rh-sensitized pregnancy is approximately 8% to 10% (Bowman, 1999). The incidence of immune hydrops has decreased significantly with the widespread use of passive immunization using Rh immunoglobulin for Rh-negative mothers at 28 weeks of gestation, following suspected fetomaternal hemorrhage, and postpartum following the delivery of an Rh-positive infant. The efficacy of this prevention program has been demonstrated by a decline in the incidence of Rh hemolytic disease of the fetus or newborn, from 65 in 10,000 births in the United States in 1960 to 10.6 in 10,000 births in 1990 (Chavez et al., 1991). The relative proportion of cases of immune hydrops secondary to non-Rh(D) atypical antibodies is increasing, as the use of Rh immunoglobulin has become widespread, with Kell isoimmunization now affecting 0.1% of all pregnancies (Caine and Mueller-Heubach, 1986).

The diagnosis of hydrops is made following the detection of abnormal or increased fluid accumulation in at least two distinct fetal body cavities. Examples include pericardial effusion, pleural effusion, ascites, subcutaneous edema, cystic hygroma, polyhydramnios, and placental thickening (Figures 127-1 to 127-3). In general, skin thickness of at least 5 mm is required to diagnose subcutaneous edema, and a placental thickness of at least 6 cm is required to diagnose placentamegaly (Romero et al., 1988). These features do not necessarily indicate hydrops; it should be noted that skin thickening of at least 5 mm may be commonly seen in macrosomic fetuses. If abnormal fluid accumulation is confined to only one site, then the diagnosis of hydrops should not be used, and the case should be described simply in terms of the involved site, such as isolated ascites or isolated pleural effusion.

Fetal ascites is diagnosed sonographically by the visualization of an echolucent rim encompassing the entire fetal abdomen in a transverse view. Loops of bowel and the outline of the fetal liver, spleen, bladder, and diaphragm are generally more easily seen in the presence of ascites. Pericardial effusion is diagnosed by the appearance of an echolucent rim at least 1 to 3 mm thick around both cardiac ventricles.

In general, fetal hydrops is not seen sonographically until the fetal hematocrit has fallen to below one-third of its normal range. However, there is no direct relationship between actual fetal hemoglobin level and sonographic appearance or severity of hydrops. The correlation between various sonographic findings and the degree of fetal anemia has been evaluated. No correlation has been found between the degree of fetal anemia and placental thickness or umbilical vein diameter, and poor correlation has been found between fetal liver and spleen dimensions and the degree of anemia (Nicolaides et al., 1988; Roberts et al., 1989; Oepkes et al., 1993). The use of Doppler sonography to predict fetal anemia has been more successful (Mari et al., 1995). An elevated peak systolic velocity (PSV) measured in the middle cerebral artery (MCA) is associated with an increased likelihood of fetal anemia (Figures 127-4 and 127-5). Mari et al. measured hemoglobin concentration from cordocentesis and PSV of blood flow in the MCA by Doppler velocimetry in 111 fetuses at risk for anemia due to maternal red cell alloimmunization (Mari et al., 2000). Hemoglobin values were compared to 265 normal fetuses. Forty-one fetuses did not have anemia, 35 had mild anemia, 4 had moderate anemia, and 31 had severe anemia (12 of which had hydrops). The sensitivity of an increased PSV of blood flow in the MCA for the prediction of moderate or severe anemia was 100%, including all cases of hydrops. The false-positive rate of PSV in an MCA of 1.5 MoM was 12%. The authors concluded that moderate-to-severe fetal anemia can, therefore, be reliably detected noninvasively. Several studies have suggested that Doppler ultraonography can replace invasive testing (measurement of Dela OD450) in the management of Rh-alloimmunized pregnancies (Zimmerman et al., 2002; Pereira et al., 2003; Bullock et al., 2005; Oepkes et al., 2006). Such Doppler measurements may be useful for surveillance following intrauterine transfusions to aid in the timing of repeat fetal blood sampling or transfusion (Mari et al., 2005). PSV in an MCA is also useful in detecting fetal anemia secondary to Kell alloimmunization (van Dongen et al., 2005).