ResultsThe patient's and control bone marrow samples grew normally in the absence of antimalarial drugs. No antimalarial had any effect on normal marrow growth. Proguanil, pyrimethamine, and quinine had no effect on growth of the patient's marrow. Amodiaquine, chloroquine (a related 4-aminoquinoline), and sulfadoxine, however, produced a dose dependent reduction in the number of colonies (figure). DiscussionThese results suggest that amodiaquine was responsible for the agranulocytosis since growth of colonies of granulocytes and macrophages was sensitive to this agent and recovery ensued on stopping the amodiaquine alone. Another patient with agranulocytosis had shown in vitro inhibition of growth of granulocytes and macrophages by amodiaquine early in the recovery phase.' A similar in vitro inhibitory effect was seen in a case of agranulocytosis induced by quinine.3 A specific effect on marrow colony growth is suggested,.separate from any systemic abnormality ofdrug handling.Our patient wished to know if she could take any medication to enable her to travel safely to areas where malaria was endemic. Clearly, amodiaquine, chloroquine, and sulfadoxine were to be avoided. Twhough we could not be certain that the other drugs tested would be safe, our results suggested that in vitro marrow culture may have important predictive value in some cases of drug induced ulocytosis.We thank Dr J B Wood, of Hereford County Hospital, for referring the patient and for permission to report the case. We also thank Imperial Chemical Eight women showed positive lymphocyte transformation responses and gave birth to uninfected babies. Six showed negative responses and four ofthe babies were born congenitally infected. Cellular immunity therefore plays a part in preventing intrauterine transmission of cytomegalovirus, and its depression after primary infection in the mother during pregnancy may be used as an early marker offetal infection. Introduction Cytomegalovirus is the most common known cause of congenital infection in man and is as important, or more important, than rubella as a cause of handicap. In the United Kingdom 40-50% of women of childbearing age are seronegative and rather fewer than 1% of these undergo pnrmary infection at some time during

SUMMARY The ELISA technique has been found to be reliable for the detection and titration of cytomegalovirus-specific IgG antibody in serum. It is about six times more sensitive than the CF test although some discrepancies were found between the antibody titres determined by the two methods. (Dienstag et al., 1976), and radioimmunoassay (Forghani et al., 1976;Knez et al., 1976), have also been used and found to be satisfactory but have not been generally adopted.The latest technique is the enzyme-linked immunosorbent assay (ELISA) (Engvall and Perlmann, 1972;, which has been shown to be more sensitive than passive haemagglutination for detecting IgG antibody (Castellano et al., 1977) and comparable in sensitivity to immunofluorescence for detecting IgM antibody (Schmitz et al., 1977). The present report compares the ELISA technique with the CF test in examining for CMV-specific IgG antibody in sera from patients.Material and methods PREPARATION OF ANTIGENSConfluent monolayers of MRC-5 diploid human fibroblasts (Jacobs et al., 1970) were grown in Roux bottles. They were infected with CMV strain AD 169, at a multiplicity of 0-1 plaque-forming units per cell, and then incubated in Eagle's minimal Received for publication 19 July 1978 essential medium with 10% fetal calf serum at 37'C. After 7-10 days, when gross cytopathic effects had developed, the cells were scraped off the bottles into a small volume of phosphate-buffered saline (PBS), pH 7-2, washed three times in PBS, and resuspended in a volume of 0-1 M glycine-NaClNaOH buffer, pH 10, equivalent to a final concentration of 1 ml per culture bottle of cells. The concentrated cells were immediately disrupted by ultrasonication for 5 s and the cellular debris was removed by centrifugation at 800 g for 10 min. The final supernatant was stored, in 0 25 ml quantities, at -70'C. Control antigen was prepared in the same way from uninfected MRC-5 cells. These antigens were used for both the CF and ELISA tests. COMPLEMENT-FIXATION TESTThis was carried out by the microtitre technique (Sever, 1962) with 2 units of antigen and 3 MHD5o of complement. Antibody titres were read as the reciprocal of the highest dilution of serum giving 50% (2+) fixation. ELISA PROCEDUREThis was done by the method of . Flat-bottomed wells in polystyrene microtitre plates (Cooke Microtitre M29AR, Dynatech Ltd) were coated with viral or control antigen by the addition of 0 3 ml of the antigen, diluted to optimal concentration in 01 M carbonate-bicarbonate buffer, pH 9-6, and incubated overnight. The plates were then washed three times in PBS containing 0 05% Tween 20 and 0-02 % NaN3 (PBST) and shaken dry.

SUMMARY The radioimmunoassay (RIA) and enzyme-linked immunosorbent assay (ELISA) techniques were found to be comparable in sensitivity and specificity for detecting cytomegalovirus IgG antibody, and 10 to 100 times more sensitive than complement-fixation (CF), anticomplement immunofluorescence (ACIF) and passive haemagglutination (PHA). In screening tests for antibody, the frequency of false-positive and -negative results was 0 6% for RIA and ELISA, 1-5% for CF, 1 6% for ACIF and 3 6% for PHA. PHA was the least satisfactory test, largely because of technical problems.Cytomegalovirus (CMV) infection is an important cause of congenital brain damage' and is also a major complication of both prolonged immunosuppressive therapy, especially in patients with organ transplants,2 and multi-donor blood transfusions.3 For serological diagnosis of infection, as well as for screening for antibody in patients and in blood donors, the solid-phase indirect radioimmunoassay (RIA)" and enzyme-linked immunosorbent assay (ELISA)7-9 techniques offer distinct improvements in sensitivity over previous methods. Although the principle of both tests, based on the detection of antigen-antibody reactions by means of a labelled anti-antibody, is the same, each possesses its own particular technical advantages and disadvantages, and both require their own expensive equipment for the reading of the results. There is still a lack of data on how they compare in sensitivity and specificity. The present study was undertaken to compare the two methods for the detection of CMV IgG and to evaluate them against the older techniques of complement-fixation (CF), passive haemagglutination (PHA) and anticomplement immunofluorescence (ACIF).Material and methods virus-specific binding index (SBI) of >2 was taken to indicate the presence of virus-specific antibody and S E R A antibody titres were calculated graphically from the These were collected from women attending the ante-reciprocal of the serum dilution having an SBI of 2-0.

Cytomegalovirus (CMV)-specific IgM that fixes complement in the presence of CMV antigen was demonstrated in sera from five patients with primary CMV infection. The CF reaction, demonstrable in the IgM fractions of the sera, was not affected significantly by absorption with aggregated IgG but was abolished by treatment with 2-mercaptoethanol. The IgM antibody was readily detected in CF tests with crude cell-extract antigens, prepared from CMV-infected tissue culture cells, or with purified enveloped-virion antigen but not with CMV-soluble antigens.

No abstract