Neutropenia is common in patients with systemic lupus erythematosus (SLE) but mechanisms of cell depletion remain obscure. To investigate the possible autoimmune aetiology of neutropenia in SLE, sera from 31 patients with this disorder were tested for anti-granulocyte activity. Granulocyte-binding immunoglobulins were detected by indirect immunofluorescence, and the ability of patient sera to opsonize granulocytes was determined by measuring the chemiluminescent response of human monocytes to granulocytes sensitized by test sera. Sera from 22 of the 31 patients bound IgG to granulocyte cell membranes and/or to nuclei, but only membrane-binding antibodies opsonized the cells for recognition by monocytes. There was no correlation between neutrophil count and the level of granulocyte-binding IgG as measured by indirect immunofluorescence. In contrast, opsonic activity and neutrophil count were inversely correlated (r = 0.5; P less than 0.05). However, opsonic activity was present in sera from most non-neutropenic patients. In patients with SLE, impaired reticuloendothelial system function may allow sensitized granulocytes to remain in the circulation.

Summary. Five examples of human anti‐Lea were subjected to DEAE‐cellulose chromatography. In each case antibody activity was demonstrable in the IgG fractions by the binding of 125I‐Lea glycoprotein. The Lea specificity of these reactions was established by absorption experiments with red cells and with serum from donors of known Lewis phenotypes. Antibody activity was also detected in IgG concentrates of two anti‐Lea antisera by gel diffusion against Lea glycoprotein. Inhibition curves obtained in a radioimmunoassay system suggest that IgC anti‐Lea has a significantly lower binding constant than IgM anti‐Lea. Failure to detect anti‐Lea by serological techniques with red cells may be related to the low binding constant.

Sera from patients with unexplained neutropenia have been assayed for anti-granulocyte opsonic activity using a chemiluminescence technique which measures the metabolic response of human monocytes to antibody-coated granulocytes. This rapid and simple technique was more sensitive than indirect immunofluorescence in the detection of anti-granulocyte antibodies. Anti-granulocyte opsonic activity was detected in sera from 17 of 31 patients, suggesting that their neutropenia may have had an autoimmune basis. The opsonic activity of five of the 17 sera was increased when granulocytes were sensitized in the presence of fresh serum. Four of these sera bound IgM and C3b to granulocytes in the immunofluorescence test. Human IgG when added to the monocyte suspension medium inhibited monocyte response to IgG antibody-opsonized granulocytes. This inhibition was less when granulocytes were opsonized with sera containing IgM and complement granulocyte-binding activity. This observation may be relevant to the selection of neutropenic patients for therapeutic use of intravenous immunoglobulin.

In seven exercises of blood grouping the overall rates of major error were 0.19% and 0.25% in ABO and D grouping respectively. In ABO grouping this represents an increase in error rate over that observed in 1982-1983 but the increase was due to an unusually high error rate with one particular group A2B cell. An improvement in performance was observed in simple D grouping and was largely due to a lower incidence of false positive grouping of D-negative cells in the antiglobulin test. An improvement in performance observed in D grouping IgG-coated D-negative cells appeared to be due to a better understanding of the problem rather than to any change in serological practice. Error rates in antibody screening were somewhat lower than in 1982-1983 but this may or may not represent an improvement in performance as the test materials were not the same in the two periods. The direct antiglobulin test with IgG-coated cells was reliably performed with polyspecific and with anti-IgG reagents but an excess of false positive results was obtained with anti-C3d. Error rates in antibody identification varied from 0.6% for anti-D to 74% for anti-c + E.

In seven surveys of blood grouping the overall rates of major error were 0.12% and 0.37% for uncomplicated ABO and D grouping respectively. Of 17 errors of ABO grouping, 13 were errors of transposition or interpretation and four were apparently technical. Of 52 errors of D grouping, 20 appeared to be errors of transposition or interpretation and 32 were apparently technical. Of the 32 technical errors of D grouping, 31 were D-negative grouped as Du (29) or D-positive (2) and most of these errors were due to misgrouping in the antiglobulin test. Causes of error in D grouping by antiglobulin test include anti-Bg and other contaminating immune antibodies, residual unabsorbed anti-A and the inherently high rate of false positive results obtained in the antiglobulin test. In view of the lack of benefit of Du testing to blood recipients or to pregnant women and of the possible adverse consequences of misgrouping D-negative patients as Du or D-positive, it is recommended that Du testing be abandoned in these groups of patients. The surveys of antibody screening demonstrated lack of standardisation and error rates similar to those previously reported in the UK for compatibility testing.