The predictive value of maternal serum testing for detection of fetal anemia in red blood cell alloimmunization

OBJECTIVE: Current management protocols for pregnancies complicated by red blood cell alloimmunization use the maternal antibody titer to predict the need for invasive testing for detection fetal anemia. We investigated the use of three maternal serum tests to assess their usefulness in predicting fetal disease: indirect Coombs' titer, Marsh score, and monocyte monolayer assay. STUDY DESIGN: Foryt-seven serum samples from pregnant women with red blood cell antibodies associated with fetal anemia were analyzed at cordocentesis. Fetal blood was analyzed for hematocrit (correlated for gestational age) and antigen status. Fetal anemia was defined as a hematocrit value of <2 SD from the mean value for gestational age. Fetuses were classified into three groups: Antigen positive with anemia (n = 19), antigen positive without anemia (n = 17), antigen negative (n = 11). Statistical methods included Kruskal-Wallis test, Newman-Keuls test, Spearman's rank correlation, and receiver-operator characteristic curves; p < 0.05 was considered significant. RESULTS: The median monocyte monolayer assay (phagocytosis, adherence, and association) did not differ among the three groups. Both maternal titers and Marsh scores were significantly higher in fetuses with anemia compared with the other two groups of fetuses (256 vs 64 vs 64, p < 0.001, and 86 vs 69 vs 64, p = 0.02, respectively). Both titer and Marsh score exhibited significant correlations with corrected fetal hematocrit (r = -0.70, p < 0.001; r = -0.63, p < 0.001, respectively). Comparison of the overall receiver-operator characteristic curves for titer and Marsh score revealed no statistical difference; however, a Marsh score of 57 was noted to have a superior specificity than a titer of 16 (p = 0.02). CONCLUSION: The maternal Marsh score can be performed in conjunction with standard indirect Coombs' titers to enhance the predictability of fetal anemia.