Our interest in the complement content of sera from healthy and diseased individuals was initiated by chance observations on the cephalincholesterol flocculation test (1). 'In our first performance of the test we inadvertently included several sera which had been inactivated and stored in the refrigerator over-night. All of these gave positive readings at 24 hours. Repetition of these tests, using sera from the same individuals, but fresh samples which were less than four hours old, yielded some negative and some positive reactions.It appeared that heating sera at 560 C. for 30 minutes might have played a part in the nonspecific reactions with the cephalin-cholesterol antigen. Sera from 20 healthy individuals were therefore tested before and after inactivation.None of the 20 sera gave positive readings before inactivation, whereas five gave readings after heating which would be considered significant.These results demonstrated that certain normal sera may be altered by inactivation to the extent that they give false positive cephalin-cholesterol reactions. The mechanism of inactivation is not altogether understood. One important result has to do with the destruction of the hemolytic property of the serum. Denaturation of proteins may occur and there may be loss of stabilizing property. However, determination of the complement content of these 20 sera, according to methods described later in this report, revealed no differences which could be related to the cephalin-cholesterol flocculation test. Later it was found that the addition of varying amounts of complement or albumin to positively reacting sera, or to normal sera which gave positive reactions after inactivation, did not inhibit the positive reaction.

Summary Cultures of Donovania granulomatis (Anderson strain) grown in a Locke-embryonic-yolk medium have been used for preparation of several types of antigens. Bacterial suspensions and boiled filtrate antigens were used in complement fixation tests with sera of 88 individuals. The highest order of sensitivity and specificity was obtained with the latter preparation. Eighty-three per cent of 24 patients with verified granuloma inguinale gave positive reactions with this antigen. In sera from 28 cases of early syphilis, 6 positive tests, mostly weak in character, were obtained with the bacterial suspension antigen, and 5 with the control antigen. One weak (1+) non-specific reaction was obtained with the boiled filtrate antigen in this group. No positive tests were obtained with either the boiled filtrate or bacterial suspension antigens in 5 cases of lymphogranuloma venereum, 11 patients with miscellaneous genital lesions, and 20 healthy controls. Skin-tests with the two antigens have not yielded conclusive data. The bacterial suspension would appear to be a better test reagent than is the boiled filtrate antigen. Rabbits, immunized with heat-killed suspensions of Donovania granulomatis gave positive complement-fixation with both antigens but positive skin-tests with only the bacterial suspension. It is our opinion that serological studies and skin-tests employing both the bacterial suspension and boiled filtrate antigen would add to precise information regarding granuloma inguinale, especially if such tests were conducted at different stages of the disease. The boiled filtrate antigen is recommended because of the ease of preparation, the sensitivity, and the specificity exhibited in various other clinical conditions.

No abstract