Extended exchange transfusion in pre-hydropic infants.

Exchange transfusion in the immediate treatment of haemolytic disease of the newborn is now a fully established procedure with more or less agreed indications for its use (Allen and Diamond, 1957). Up till now the procedure has been mainly used to prevent kernikterus and in this it has been eminently successful both in the 'clinical ' and the ' pathological' varieties of this condition. It can be said that in any unit in which facilities for proper exchange transfusion are available kernikterus should no longer be a problem in haemolytic disease due either to Rhesus, ABO or other incompatibility. The deaths still encountered in haemolytic disease of the newborn are of two types: those which occur in utero resulting in stillbirth and those in which the babies are born alive but die within a few hours either before, during, or after the first exchange transfusion. The main cause of death in the first type is hydrops foetalis, which is essentially severe heart failure occurring in utero; in the second type the cause of death is probably also heart failure so that the term 'pre-hydrops' would be a fairly accurate description of this clinical condition. It is mainly with this group of pre-hydropic infants that we are concerned in this communication. Before discussing this group further, however, we must explain that in our view these deaths are in quite a different category from those described by Campbell (1955). In his cases the babies were relatively mildly affected and it was only because of the fear of later jaundice that an exchange was undertaken. In them the babies were perfectly well until they suddenly collapsed and it was felt that there was probably a severe biochemical upset affecting perhaps the serum potassium (Campbell, 1955) or serum citrate (Kelsall and Vos, 1955). We have been fortunate in having no experience of this type of case and it seems that no satisfactory explanation for them is, as yet, forthcoming (Farquhar and Smith, 1958). In contrast, the pre-hydropic babies are born with a low cord haemoglobin (under 10 g. %) and are clinically pale and either not jaundiced or only mildly so. The cord blood bilirubin is usually high and may be very high. In addition, the babies are usually ill with rapid respirations which often become grunting in character and there is often some oedema. The liver and spleen are both enlarged and there is sometimes a profuse purpura. The fact that haemolysis is proceeding rapidly is shown by the high reticulocyte count in the cord blood. These babies are on the verge of, or in, heart failure (Mollison and Cutbush, 1949), as is shown by the high venous pressure recorded when the catheter is passed up the umbilical vein. Even with the use of exchange transfusion the mortality in this group is still high. The occurrence of pre-hydropic infants is well documented in the literature and they are usually shown as dying from heart failure. In the report of Mollison and Walker (1952), out of 79 deaths during the first week of life 17 were due to this cause: eight of these died in spite of treatment by simple or exchange transfusion. Walker and Neligan (1955) in their own series of 250 treated cases of haemolytic disease, reported deaths attributable to the disease and, of these, five were due to heart failure, comprising the largest single cause of death. Walker and Mollison (1957), recently reviewed the deaths from the disease in England and Wales and reported that in two years (1953 and 1955), there were 312 and 315 deaths respectively: of these, 58 and 55 were attributed to heart failure although all of them did not necessarily have ' proper' treatment. In fact, of these 113 infants, in only 75 was exchange transfusion attempted, 57 of them dying during the procedure. They reported a further group of 69 infants in the two years who were said to die of ' hydrops foetalis ', this is probably the severest degree of the condition we are discussing and, in