Replacement Transfusion as a Treatment of Erythroblastosis Fetalis, by Louis K. Diamond, MD,Pediatrics, 1948;2:520–524

The treatment of hemolytic anemia of the newborn often requires support by compatible blood transfusions. Before the discovery of the Rh factor such blood was often only ABO compatible, and for convenience, the patient's father, who we now know is always Rh positive, was often used as the donor. We noted that the infant's hemoglobin level often fell rapidly, even with repeated transfusions. Because Rh positive cells are destroyed much more rapidly than Rh negative cells, it seemed logical to use only Rh negative blood as a means to achieve the highest blood levels for protection of the patient, and we began to do this after 1941.We have recently been able to show that the serum of severely affected infants contains free maternal anti Rh agglutinins and the infant's red cells are often completely coated with this antibody. The removal of much of the baby's blood should theoretically diminish the damage resulting from this antibody. Such reasoning lead us to try replacement transfusion as treatment for erythroblastosis. Although it is not a new procedure, to perform it in a newborn adds difficulty to an inherently difficult technic. Earlier methods employed the longitudinal sinus and a peripheral vein or a vein and an artery, requiring prolonged exposure of the infant on an operating table with resultant cold exposure. To prevent clotting, patients also received heparin with its risk of hemorrhage.The pediatrician cannot help but be attracted to the large and readily accessible vein in the umbilical cord. The development of a bland polyethylene catheter by Dr Ingraham of our neurosurgery department overcame some of the major difficulties of the exsanguination procedure. This catheter can be easily threaded through the umbilical vein into the vena cava through the ductus venosus or by-passing the ductus venosus into the portal vein. The catheter is connected to a needle and then to a three way stopcock. The first valve of the stopcock is connected by a tube to an exhaust pail. The second valve is connected to a bottle of blood and the third to a syringe. By alternate maneuvers of first removing and then putting in blood, 20 cc at a time, the infant can be gradually exsanguinated and most of its blood volume replaced with relatively little trauma. The procedure is carried out with the baby in a warm Hess bed with oxygen administration as needed. We usually use about a pint of Rh negative blood and the procedure takes an hour to an hour and a half. By actual measurement this replaces 90 to 95% of the infant's own red cells which are coated with antibody. At the end of the procedure, the infant is given calcium to overcome the alkalotic tetany that might result from the sodium citrate in the blood.In a quarter of cases, it is necessary to transfuse an infant again in the 2nd-4th week because of anemia, but we have rarely had to offer more than a simple transfusion. Infants can be discharged with the mother, sometimes before the eighth day has passed. Previously, the average period of hospitalization for infants with severe erythroblastosis requiring multiple transfusions was three weeks or more, during which time secondary infections could occur.Not every infant with erythroblastosis requires replacement transfusion and in the past year we have treated only about one in eight babies sent to us. We insist that we receive a specimen of the mother's blood to prove that the mother has Rh agglutinins which can be on the infant's Rh positive red cells. Further evaluation is based on clinical signs: anemia, jaundice, edema, and hepatosplenomegaly. The presence of these clinical signs with antibodies in the mother's circulation implies the need for immediate replacement transfusion.Even in the absence of these clinical signs, certain laboratory evidence may indicate the need for immediate treatment. If the baby's Rh positive cells are coated with antibody and the baby has free antibody in its circulation, our selected data over several years indicate that replacement transfusion should be done because such infants quite regularly develop severe symptoms within the second to fourth day of life. The absence of clinical signs and the absence of the laboratory findings suggest that the infant will do well enough without such drastic treatment.As a background for an evaluation of our present mode of therapy, we have statistics collected over the past 15 years. Before 1941, taking all of the infants sent to us, there was a 35–40% mortality. Between 1941 and 1946, with Rh negative transfusions, our mortality dropped to 30% or less. Our present data covers about 15 months of experience and 85 infants treated personally by replacement transfusions through the umbilical vein. Sixty-five are now living and well with follow ups of 3 to 15 months. A number of the surviving infants came from families where a previous infant had died of erythroblastosis. In such families the chance of a subsequent affected infant surviving is less than 10%. The survival of such children in our series was 70%. Seven patients died, but not of erythroblastosis. The causes were immaturity, tentorial tears or other factors which cannot be directly related to erythroblastosis. Seven patients died of the disease; six of these showed kernicterus.This mode of therapy has the advantage of ease and greater safety in performance. An added advantage is earlier discharge from the hospital at 7 or 8 days rather than 2 to 3 weeks. Although this is not a panacea for erythroblastosis, replacement transfusion seems to be a better method of treatment for many cases.