574 THE LEWIS SYSTEM: NEW HISTOCOMPATIBILITY ANTIGENS IN RENAL TRANSPLANTATION R. ORIOL Institut d’Immuno-Biologie, H&ocirc;pital Broussais 96 Rue Didot, 75674 Paris Cedex 14 J. CARTRON Centre National de Transfusion Sanguine, H&ocirc;pital Saint-Antoine, Paris J. YVART Service de Transfusion Sanguine, H&ocirc;pital Saint-Joseph, Paris J. BEDROSSIAN A. DUBOUST J. BARIETY Transplantation Unit, H&ocirc;pital Broussais, Paris J. C. GLUCKMAN Transplantation Unit, H&ocirc;pitaux Piti&eacute;-Salp&eacute;tri&egrave;re et Foch, Paris M. F. GAGNADOUX Transplantation Unit, H&ocirc;pital des Enfants-Malades, Paris Summary Lewis antigen types (Le, le) were retros- pectively determined in 255 first-kidney- transplant recipients. Actuarial survival of grafts at two years was significantly lower in the le/le recipients than in the Le recipients. This indicates that mismatching of these antigens contributes to rejection of kidney trans- plants. The effects of mismatching for the Lewis and HLA antigen systems seemed to be additive. Introduction NATURAL Lewis antibodies account for half of all the irregular erythrocyte antibodies detected. They can lyse red cells’ and may be lymphocytotoxic.2 The biosyn- thesis of A, B, H, and Lewis antigenic determinants is genetically controlled by four independent genes acting on the same molecule (fig. 1). Kidney transplants have much lower survival-rates in homozygous lelle recipients than in L recipients (P<0.01). 3 Fig. I-A, B, H, and Lewis antigenic determinants. Those with functional or active genes A, B, H, Se, Le can synthesise corresponding antigenic determinants. Homozygotes for the amor- phous or silent genes 0, h, se, le do not have glycosyltransferases needed for synthesis of antigenic determinants. Action of the Se-se sys- tem is restricted to some external secretions. Patients and Methods Patients.-255 first-kidney-transplant recipients were stud- ied retrospectively. Transplantation had -been performed at 4 French centres and all the patients typed were included. Although the donors were not typed, transplants in Le and Se recipients should always be compatible for these systems, while transplants in lelle and se/se recipients are very likely to be in- compatible because of the low frequency of lelle (0.1) and se/se (0.2) in the French population. Transplant-survival rates were calculated by the actuarial method.4 Saliva.-Samples were heated in a boiling water-bath for 10 min, centrifuged, and stored at &mdash;20&deg;C. Saliva samples capable of inhibiting A, B, H, or Lewis agglutinin reactions were referred to as Se and Le, respectively, while saliva samples Fig. 2-Percentage actuarial survival of 255 renal transplants in Le (.) and lelle (8) recipients as a function of time. ‘ Fig. 3-Survival-rates of transplants shown in fig. 2 in terms of HLA match. Transplants with less than 2 incompatibilities are considered matched (0) and transplants with two or more incompatibilities, are regarded as being mismatched (0).