Complement Fixation Tests in Pertussis.

Bordet and Gengou (1), in 1906, performed serological tests on pertussis employing the organisms which they discovered and claimed to be the causative agent of whooping cough as the antigen for the complement fixation test. They got a positive reaction in most cases of the disease. Yet, later investigations failed to confirm these results. Among those who confirmed their results were Renaux (2) who examined 73 cases, Chievietz and Meyer (3) with 112 cases and Giese (4) with 123 cases. In recent years good results were reported particularly by German authors who employed a method which will be described later. Keller, Klopstock and Klopstock (5) examined 131 sera of 88 cases with pertussis; 77 sera showed a positive reaction, 17 were doubtful, and 37 sera were negative. Gundel and Schliiter (6, 7) reported a study of 70 pertussis patients; 54 of these were positive, 3 doubtful, and 13 negative. In a later series of cases they found that of 140 sera from children with pertussis and suspicious cases, 105 gave a positive reaction. The reason given to explain the negative reactions were, either that the children did not suffer from whooping cough or were too young or too sick to form antibodies. The method employed by the German authors mentioned above consists in preparing an antigen by using whole pertussis bacilli supended in alcohol to which a small amount of lecithin is added. The addition of lecithin tends to render the result more specific and at the same time to decrease the anticomplementary effect of the antigen. In a recently published paper, Hansing (8), using the same method on the sera of 189 patients, found that 48 per cent were positive, 22 per cent doubtful, and 29 per cent negative. The failure of many workers to obtain good results with complement fixation in pertussis may be explained by the fact that no method was available to differentiate between pertussis and influenza bacilli. Only recently Povitzki (9) found that Bordet-Gengou bacilli grew well on a definitely acid medium (pH 5.0) whereas the Pfeiffer bacillus was inhibited by it. Furthermore, Gundel and Schliiter (10) found that genuine pertussis bacilli when injected intradermally into white rabbits always produced severe necrosis while influenza bacilli only caused infiltration and erythema. Because, in the past, there was uncertainty regarding the differentiation of both organisms the proper antigen was not always employed for the complement fixation test. The agglutination method, on the other hand, did not prove successful owing to the insufficient amount of antibodies during the natural disease. Besides, spontaneous agglutination is often observed and nonspecific results thus obtained. If it could be shown that the complement fixation test becomes positive in pertussis and also follows the injections of bacillus pertussis, this would seem to be another point in favor of the Bordet-Gengou bacillus as the etiological factor in whooping cough.' The problems studied were the specificity of the reaction, its value in the diagnosis of whooping cough, its occurrence during the disease and the duration of the antibody content.