Intravascular intrauterine transfusion for severe erythroblastosis fetalis using different techniques

Sonographic demonstration of brain injury in fetuses with severe red blood cell alloimmunization undergoing intrauterine transfusions

OBJECTIVE

To assess sonographically brain anatomy in fetuses with severe anemia due to red blood cell alloimmunization undergoing intrauterine intravascular transfusions.

METHODS

Multiplanar neurosonography was performed in seven consecutive hydropic fetuses undergoing intrauterine transfusions (mean gestational age 22 +/- 2.5 weeks; mean hemoglobin concentration at the first transfusion 2.3 +/- 1.0 g/dL).

RESULTS

Abnormal cerebral findings were identified in four out of seven fetuses. An intracerebellar hemorrhage developed in two fetuses after the first transfusion and one fetus that had severe brain edema before the first transfusion was later found to have cystic periventricular leukomalacia. In one fetus unilateral ventriculomegaly was noted after the first transfusion. Two fetuses were terminated. The remaining pregnancies had an uneventful course, the infants were delivered between 34 and 36 gestational weeks and were alive and well at the time of writing. Prenatal diagnosis of brain injury was always confirmed except for the case with ventriculomegaly that underwent spontaneous intrauterine resolution.

CONCLUSIONS

Fetuses with extreme anemia due to red blood cell alloimmunization can be salvaged by intrauterine transfusion. In some of these cases brain injury may occur prenatally, and the risk seems to be particularly high when the hemoglobin concentration at the time of the first transfusion is <or= 2 g/dL. We suggest that in these pregnancies detailed fetal neuroimaging by either multiplanar sonography and/or magnetic resonance imaging is indicated.

Survival analysis of transfused fetuses affected by Rh-alloimmunization

The aim of the present study was to evaluate the survival rate of a group of 86 fetuses affected by Rh-alloimmunization submitted to intrauterine red blood cell transfusion. All the women had antibody titres> or = 1:32 at the time of their enrollment in the study. Crude fetal survival rate was 89.5% (77/86 cases). Data were stratified according to specific cut-off points of (1) pre-transfusion fetal haemoglobin expressed as the rate between the observed and the estimated value for each gestational age at the time of the first transfusion; (2) the difference between the haemoglobin at the beginning of the second-transfusion less that at the end of the first transfusion (delta haemoglobin); and (3) presence of ultrasound detected anomalies. Statistically significant stratification of the survival rate was observed for the level of pre-transfusion fetal haemoglobin (95% and 76.9%, respectively, p= 0.009) using a cut-off value of < 70% and > or = 70% of the expected value. Again, delta haemoglobin showed a different survival rate when a cut-off value of 6 g/dl was used to generate subgroups of fetuses: 94.6% and 80%, respectively (p= 0.0145). Among the ultrasound anomalies, the presence of hydrops showed a correlation with the survival rates. The quoted values were 97.83% (absence) and 80.0% (presence) respectively (p= 0.0058). Cox regression showed a significant association of the studied variables with the outcome (survival). The presence of hydrops was the best predictor (Odds ratio= 8.7, p= 0.0073) followed by Delta haemoglobin (Odds ratio= 2.0, p= 0.0422). The rate of pre-treatment fetal haemoglobin < 70% of the expected value did not add any significant valu and was thus removed from the final model. Weight at delivery expressed in grams showed a direct correlation with the survival rate (Odds ratio= 0.9, p= 0.1529) and was added into the model as an adjustment quantitative variable.