Immunogenicity and reactogenicity of a Q fever chemovaccine in persons professionally exposed to Q fever in Czechoslovakia

Bulletin of the World Health Organization, 60 (3): 389- 394 (1982) Immunogenicity and reactogenicity of a Q fever chemovaccine in persons professionally exposed to Q fever in Czechoslovakia J. KAZAR,' R. BREZINA,2 A. PALANOVA,3 B. TVRDA,4 & S. SCHRAMEK5 An immunization trial carried out on 1310 persons professionally exposed to Qfever confirmed thesuitability ofa chemovaccineforfield use. Its immunogenicity and reactogen- icity differed in threegroups ofsubjects who differed in theirprevious experience ofQfever and who were given different lots of the vaccine. There was some correlation between reactogenicity ofthe vaccine andpre-immunization positivity in the microagglutination and skin tests as indicators ofaprevious exposure. The skin test was more reliableforpredicting post-immunization reactions. Q fever is primarily a zoonosis that may become established in domestic animals and represent a continuing and significant public health problem, particularly in professionally exposed groups of the population such as livestock handlers and the processors of animal products (4). The fact that it is the most widespread rickettsial infection is confirmed by recent reports on its occurrence throughout the world, e.g., in Africa (20), Europe (13), India (19), South America (14), and in the United States of America (8). Q fever poses an important public health problem in Slovakia (17), and long-term studies have been carried out there on the epidemiology of the disease. Studies of the antigenic properties of Coxiella burnetii carried on in our laboratory (2) have resulted in the development of Q fever chemovaccine suitable for use in man. The results of vaccinating volunteers exposed to a laboratory challenge with C. burnetii were satisfactory (5) and now we have tested the immunogenicity and reactogenicity of this chemo- vaccine in a large-scale field trial. This paper presents the results of field vaccination of human subjects carried out in Central Slovakia in 1977 and 1978, and in East Slovakia and South Moravia in 1980. ' Head, Immunology Laboratory, Department of Rickettsiae, Institute of Virology, Slovak Academy of Sciences, Bratislava, Czechoslovakia (WHO Collaborating Centre for Rickettsial Refer- ence and Research). 2 Head, Department of Rickettsiae, Institute of Virology, Bratislava. 3Regional Hygiene and Epidemiology Centre, Banska Bystrica, Czechoslovakia. 4 District Hygiene and Epidemiology Station, Vel'ky Krtis, Czechoslovakia. I Head, Biochemistry Laboratory, Department of Rickettsiae, Institute of Virology, Bratislava. MATERIALS AND METHODS The chemovaccine used was the soluble phase-I antigenic component extracted by trichloroacetic acid (TCA) from highly purified phase-I C. burnetii orga- nisms, as described earlier (3). One vaccine dose contained the soluble material extracted from 1 mg of dried rickettsial particles dissolved in 1 ml of physiological saline. The antigenic capacity, immunogenicity, and pro- tective capacity of the vaccine have been described elsewhere (2). Briefly, one vaccine dose contained at least 200 units of antigen when examined by the complement-fixation (CF) test with an immune rabbit reference serum against Q fever. The vaccine induced both phase-I and phase-1I antibody responses, as detected by the microagglutination (MA) test, in more than 90%0 of intraperitoneally inoculated mice, and protected more than 9007 of 12 guinea-pigs against a virulent phase-I C. burnetii challenge with 105 ID50 for embryonated eggs. Macroscopically the vaccine was a colourless, slightly opalescent fluid; microscopically, no rickett- sial bodies were visible in the vaccine preparation. Vaccine safety control tests included the deter- mination of specific sterility on chick embryo yolk sacs (1), control of bacterial sterility by inoculation on blood agar and thioglycollate broth, and obser- vation for adverse effects in intraperitoneally inocu- lated guinea-pigs. In the latter, the vaccine caused only a transient elevation of body temperature with- out any apparent signs of sickness and any loss of body weight. The vaccine was stable after storage for at least one year at 40 C with thiomersal (1:10 000). 418 -389- J. KAZAR ET AL. Three lots of the vaccine were used for the immun- ization of three groups of people differing in their previous exposure to Q fever: (a) agricultural workers, district veterinarians, and veterinary tech- nicians in Central Slovakia, who were exposed to C. burnetii in nature, and who were likely to have had contact with Q fever in domestic livestock (17); (b) personnel of the State Veterinary Institute in East Slovakia; (c) and workers of a cotton processing plant in South Moravia, who were vaccinated at their request because of an outbreak of Q fever in the plant in 1980 (Kazar et al., unpublished data, 1981). The immunization trials were carried out with due regard to the provisions of the Helsinki Declaration concern- ing experimentation on human subjects. One day before immunization with 1 ml of vaccine the subjects were bled for serological examination, and skin-tested on the forearm with one-hundredth of the vaccine dose in 0.1 ml of physiological saline, to detect any previous exposure to Q fever and any hypersensitivity. The skin test was read after 24 hours and was considered positive if there was erythema with a diameter of 0.5 cm or more and/or induration at the site of intradermal inoculation. Some of the volunteers were immunized with one vaccine dose irrespective of the presence of pre-immunization anti- bodies and/or regardless of a positive skin test. A total of 201 volunteers from Central Slovakia with negative results in pre-immunization tests and with- out any local or systemic reaction following the first dose of the vaccine were given a second vaccine dose two weeks later. In general, pregnant women and individuals with acute febrile illness or with allergy were not vaccinated. Post-immunization antibody response was deter- mined 2 or 5 weeks after immunization with one dose in the group of vaccinees from Central Slovakia, and at 5 weeks after immunization in the groups of vaccinees from East Slovakia and South Moravia. The vaccinees from Central Slovakia given two vaccine doses 2 weeks apart were bled only once 5 weeks after the first dose of vaccine. The pre- and post-immunization sera were examined by the "warm" CF test and by the MA test using an artificial phase-II C. burnetii antigen prepared by potassium periodate treatment of purified phase-I C. burnetii suspensions (18). Sera with antibody titres > 4 in either serological test were considered positive. Because more sera were positive in the MA test before and after immunization than in the CF test (12), in this study, only the results obtained by the MA test are presented here. The statistical analysis of data obtained was per- formed using Student's t-test. In the case of propor- tions lower than 20%, the statistical significance of differences was calculated by the method of cyclo- metric transformation. Reactogenicity of the vaccine was determined 24 hours after immunization by reading the local reac- tion (erythema with a diameter of 1 cm or more with oedema, local pain and induration, and regional lymphadenopathy) and by evaluating the systemic reactions (body temperature (axillary) higher than 37 °C and general complaints). Vaccinees with com- plaints that occurred later or persisted were referred to local physicians; their complaints were recorded when the post-immunization blood samples were collected for serological examination. RESULTS Pre-immunization exposure to Qfever In a pre-immunization trial, previous exposure toQ fever was examined by the skin test and the MA test. Of the 1256 persons tested, 214 (17.0%) were positive by the MA test and 202 (16.1%) by the skin test. The results of the two tests were in agreement as regards 78.5% of the 1256 persons tested: 5.8% were positive in both tests and 72.7% were negative in both. Immunogenicity of the vaccine In the first group of vaccinees from Central Slovakia, the post-immunization antibody response was investigated in those who received one dose of vaccine, at 2 weeks or at 5 weeks. In those who received two doses of the vaccine, antibody response was measured 5 weeks after the first dose (Table 1). Though the proportion of positive sera was higher at 5 weeks (70.5%) than at 2 weeks (62.7%), and was higher after two vaccine doses (74.4%), the differ- ences were not significant (P> 0.05). For this reason, Table 1. Serological conversion as detected by the microagglutination (MA) test in different population groups immunized with one or two doses of Q fever chemovaccine Time between No of No. and Groups immunization No- proportion immunized and MA test se ra *tive (weeks) sera One vaccine dose Central Slovakia 2 209 131 (62.7%) 5 105 74 (70.5%) East Slovakia 5 86 39 (45.3%) South Moravia 5 314 149 (47.4%) Total 714 393 (55.0%) Two vaccine dosesa Central Slovakia 180 134 (74.4%) a 2 weeks apart; bled 5 weeks after the first dose. 390 Q FEVER CHEMOVACCINE in the other two groups of vaccinees serological conversion was investigated only at 5 weeks after immunization with one dose of the vaccine. The proportions of vaccinees showing serocon- version in East Slovakia (45.3%) and South Moravia (47.4%) were lower than in Central Slovakia (70.5%) and this may be a result of the use of three different lots of vaccine in the three different groups. In general, however, the vaccine was fairly immuno- genic, because it resulted in seroconversion in 393 (55%) of 714 persons given one dose of vaccine. Some of the persons in whom the post-immunization reac- tion was evaluated were not available at the time the serum samples were taken, and thus the total number of subjects in Table 1 is less than the total in Table 2. Reactogenicity of the vaccine Local and systemic post-vaccination reactions in persons receiving one or two doses ofchemovaccine. The incidence of local and systemic post-immuniz- ation reactions was followed in three groups of sub- jects with different histories of previous exposure to Q fever and immunized with three different lots of vaccine. Most of the immunized persons were chosen because they were MA-test and skin-test negative before immunization. Some subjects were, however, immunized with one dose of the vaccine irrespective of the presence of pre-immunization antibodies and/ or regardless of skin-test positivity, to determine whether reactogenicity of the chemovaccine was influenced by prior experience of C. burnetii antigens. As shown in Table 2, systemic reactions occurred in 48 of 1109 immunized persons (4.3%), the propor- tions being similar (3.7%, 5.8%, and 5.1%) in the three groups under study. Local reactions were seen in 439 persons out of 1109 (39.6%). The proportions of people with local reactions were almost identical in the groups from Central Slovakia and East Slovakia (28.1% and 27.9%, respectively), but a much higher proportion (63.9%) had local reactions in the group from South Moravia. This was not a result of higher reactogenicity of the particular lot of vacine, because when the same lot of vaccine was used to immunize laboratory personnel in West Slovakia (results not given), local reactions occurred in 6 (30.0%) and sys- temic reactions in 1 of 20 vaccinees. The possibility of a higher sensitization with C. burnetii in this group, because of a recent outbreak of Q fever in the cotton processing plant should be borne in mind, even though the proportions of subjects positive in theMA test (1.7%) and the skin test (10.4%) before immuniz- ation in South Moravia were lower than or compar- able with the proportions in the remaining two groups of vaccinees. Nevertheless, previous exposure to Q fever may be of importance as regards the development of post- immunization reactions. When, for administration of the second vaccine dose to the group from Central Slovakia, only the MA-test and skin-test negative individuals were selected, only one inexplicable sys- temic reaction occurred. The number of local reac- tions in 201 vaccinees was reduced to 13, a much lower proportion (6.5%) in comparison with the 28.1% of 668 non-selected subjects with reactions after one dose of the vaccine (P<0.001). Local and systemic reactions in relation to pre- immunization experience of Q fever. Though the severity of the local reactions varied from slight transient erythema at the vaccination site to local pain and erythema persisting for several days, they did not pose any serious problems, because neither abscesses and fistulae nor swelling of the draining axillary lymph notes occurred. The systemic reactions were Table 2. Local and systemic post-immunization reactions in the three different groups given one dose of Q fever chemovaccine No. and proportion positive No. and proportion with pre-immunization post-immunization reactions Group immunized No. immunized MA test Skin test local systemic (1) Agricultural workers, district veterinarians, and veterinary technicians, Central Slovakia 668 97 (14.5%) 56 (8.4%) 188 (28.1%) 25 (3.7%) (2) Personnel of the State Veterinary Institute, East Slovakia 86 2 (2.3%) 9 (10.5%) 24 (27.9%) 5 (5.8%) (3) Workers of the cotton processing plant, South Moravia 355 6 (1.7%) 37 (10.4%) 227 (63.9%) 18 (5.1%) Total 1109 105 (9.5%) 102 (9.2%) 439 (39.6%) 48 (4.3%) 391 J. KAZAR ET AL. mainly similar to the symptoms of early influenza, i.e. sensations of chill, weakness, malaise, headache, myalgia, arthralgia, insomnia, nausea, and elevation of body temperature up to 390 C. The systemic reac- tions varied as to combination and severity, but only in one case was the attention of a physician required -in a women who forgot to declare her general allergic predisposition. In several cases, the systemic reactions or some of the more severe local reactions were accompanied by the recall of the dermal reaction at the site of the skin test. In general, the local reactions were more severe in persons also exhibiting systemic reactions, and in those with a previous experience of Q fever, as demonstrated by positivity in the pre-immunization MA test and/or skin test; although as many as 68.8% of the local reactions occurred in MA-test and skin- test negative individuals (Table 3). A higher propor- tion of local reactions was observed in skin-test positive (21.4%o) than in the MA-test positive (13.9%7o) subjects (P<0.01). Interesting data were obtained by a detailed ana- lysis of the systemic reactions in skin-test and MA-test positive and negative subjects (Table 4). Of 48 sys- temic reactions, 37 (77.1 %) occurred in vaccinees Table 3. Number and proportion of local post- immunization reactions in relation to the results of the pre-immunization MA and skin tests MA test MA test Total positive negative Skin test positive 18 (4.1%) 76 (17.3%) 94 (21.4%) Skin test negative 43 (9.8%) 302 (68.8%) 345 (78.6%) Total 61 (13.9%) 378(86.1%) 439(100.0%) Table 4. Number and proportion of systemic post- immunization reactions in relation to the results of the pre-immunization MA and skin tests MA test MA test positive negative Total Skin test positive 4 (8.3%) 24 (50.0%) 28 (58.3%) Skin test negative 9(18.8%) 11 (22.9%) 20 (41.7%) Total 13 (27.1%) 35 (72.9%) 48 (100.0%) who were either MA-test or skin-test positive and 11 (22.9%) in individuals who were negative by both tests (P<0.001). The proportion of systemic reac- tions was significantly higher (P<0.01) in skin-test positive (58.31o) than in MA-test positive (27.1%) vaccinees, indicating the value of the skin test for predicting the likelihood of post-vaccination systemic reactions. DISCUSSION Q fever is not considered sufficiently dangerous and widespread to warrant the expense of mass vacci- nation campaigns. However, because C. burnetii may cause a considerable number of infections in profes- sionally exposed persons, there is a clear need for a protective vaccine for human use, and immunization of persons at risk (laboratory personnel, livestock handlers, processors of animal products) may be the best method of prevention. Such a vaccine should fulfil two basic criteria: high immunogenicity and low reactogenicity. There have been several attempts to prepare such a vaccine (15). Formalin-killed, purified C. burnetii suspensions were found to have satisfactory protec- tive properties, but they were too reactogenic, particularly in persons with specific sensitivity prior to the immunization (15). An attenuated variant of C. burnetii (M-44) iso- lated by Soviet authors was very promising, pro- ducing a high proportion of seroconversions with small dosages, and causing only minor, transitory systemic reactions (9). However, the use of attenuated, live rickettsial vaccine is still subject to question, in particular as regards the stability of attenuation (22). In fact, the M-44 strain appeared to survive for a considerable length of time in guinea- pigs, retaining a potential for reactivation of infection (10); in addition it produced hepatitis, splenitis, and myocarditis in experimental guinea-pigs (11). The suitability of the attenuated M-44 variant for use in a vaccine is questionable also from the stand point of the phase variation phenomenon of C. burnetti, because serial passaging in guinea-pigs did not result in the development of a phase-I antigenic component (10). Strains of C. burnetii found in nature are in phase I, and phase-I vaccines have been found to be 100- 300 times more efficacious than phase-II vaccines (16). The importance of preparing vaccines from phase-I organisms was also evident from our studies on the antigenic structure of C. burnetii (2). These studies showed the antigen to have a complex protein-lipo- polysaccharide character, phase-I antigenicity, excel- lent protective properties against challenge with 392 Q FEVER CHEMOVACCINE phase-I organisms, and low reactogenicity; the anti- gen was extracted by TCA in our laboratory as early as in 1961 (3), and later recommended for the prepar- ation of a Q fever chemovaccine for human use (5). The results reported here of field immunization trials of this chemovaccine confirmed the low reacto- genicity and fair immunogenicity observed in our previous study (5) and in a trial in Romania (6) in which the same type of TCA-extracted, soluble Q fever vaccine was used. In neither study, however, were reactions observed that were as severe as those reported for persons who were skin-test positive before immunization in a trial of a Q fever partic- ulate vaccine (15). In general, we observed more severe reactions in individuals who had been exposed previously to Q fever, as demonstrated by pre- immunization in the MA and/or skin tests. As many as 77.1% of systemic reactions occurred in the vaccinees who were either the MA or skin-test positive. All these findings are in agreement with those of another Romanian study with a soluble Q fever vaccine (7). On the other hand, not all the individuals in our study who reacted positively in one of the tests before immunization developed post-immunization reac- tions, and as many as 68.8% of local reactions occurred in those who were negative in those tests. This agrees with the results of some American authors with a phase-II particulate Q fever vaccine, that the results of a skin test and pre-immunization serological positivity are not reliable indicators of reactivity (21). However, we think that the exclusion from immuniz- ation of all persons positive in the skin test and/or the MA test would reduce the number and severity of the adverse effects to Q fever chemovaccine, and we recommend the use of both pre-immunization tests in future Q fever immunization trials. The post-immunization antibody response is influenced by the immunization schedule (the dose of vaccine and the number of vaccine doses, the interval between vaccine doses, and the interval between vaccination and serological examination) and by the serological test used. Cracea et al. found serological conversion by the CF test in 63- 1007o and by the mouse protection test in 93- 100% of the vaccinees depending on the protocol of immunization (6). In our study the immunogenicity of the Q fever vaccine was fairly high, resulting in serological conversion detectable by the MA test in 55.0% of subjects given one dose of the vaccine. Two vaccine doses did not result in a substantial increase in the serological conversion. However, the number of vaccine doses and the occurrence of post-immunization serological conver- sion are not necessarily decisive for the development of protection against Q fever. Volunteers highly exposed to infection under laboratory conditions did not contract this disease, irrespective of whether they received one or two doses of the vaccine and whether they had developed antibodies at the times of examination. It is probable that cell-mediated immunity is also elicited by immunization with Q fever chemovaccine and its assessment is to be the subject of future studies. ACKNOWLEDGEMENTS The kind help of the local physicians, Dr J. Gaborik in Vel'ky Krtig, Central Slovakia, and Dr Z. Krunert in Uherske Hradigt6, South Moravia, in the evaluation of post-immunization reactions is greatly appreciated. RESUME IMMUNOGENICITE ET REACTOGENICITE D'UN VACCIN FRACTIONNE CONTRE LA FIEVRE Q CHEZ DES PERSONNES PROFESSIONNELLEMENT EXPOSEES A CETTE MALADIE EN TCHECOSLOVAQUIE Dans des essais pratiques portant sur des personnes professionnellement expos&es a la fievre Q en Tchecoslo- vaquie, on a 6tudi6 l'immunog6nicit6 et la reactogenicite d'un vaccin contre cette maladie, represente par le consti- tuant antig6nique soluble extrait par l'acide trichloracetique a partir de Coxiella burnetii, phase I. Les sujets vaccines etaient a) des travailleurs agricoles, des v6t6rinaires de district et des techniciens v6t6rinaires, en Slovaquie centrale; b) des membres du personnel de l'Institut vet6rinaire d'Ptat en Slovaquie orientale, et c) des travailleurs de l'industrie cotonniere, en Moravie meridionale. Ces sujets differaient quant a leur contact ant6rieur avec la fievre Q et ont ete vaccines avec diff6rents lots de vaccins. L'immunogenicite, d6termin6e par la conversion s6rolo- gique dans l'epreuve de micro-agglutination, variait entre 45,3% et 70,5% parmi les personnes ayant recu une dose de 393 394 J. KAZAR ET AL. vaccin. Les proportions de s6roconversions observ&es 2 semaines (62,7%) ou 5 semaines (70,5%) apres la vaccina- tion, et a la suite d'une (70,5%) ou de deux doses (74,4%) de vaccin, ne diff6raient pas notablement. Sur les 1109 personnes vaccin6es avec une dose de vaccin, il y a eu des r6actions locales (6rytheme, odeme, induration et douleur au point de vaccination) chez 439 (39,6%) et des r6actions g6n6rales (6l6vation de la temperature jusqu'a 39,0 °C, frisson, c6phal6e, myalgies, arthralgies, insomnie et nausees) chez 48 (4,3%). II n'y a pas eu de r6actions locales graves, telles qu'abces ou fistules, comme celles qui ont ete d6crites apres l'utilisation de vaccins particulaires de C. burnetii. On notait une certaine corr6lation entre la r6actog6nicit6 du vaccin (c'est-a-dire la gravit6 des reactions locales et le nombre des r6actions g6n6rales) et le degr6 de positivit6 dans l'6preuve de micro-agglutination et les tests cutanes, indicateurs d'une exposition ant6rieure a la fievre Q; les tests cutan6s permettaient de pr6voir de maniere plus flable l'apparition de r6actions post-vaccinales. Lorsque les sujets positifs a l'une ou l'autre 6preuve ne recevaient pas de deuxieme dose de vaccin, la proportion de r6actions post- vaccinales etait seulement de 13 sur 201 vaccin6s (6,5%) et une seule personne a presente une r6action gen6rale. Pour les futurs essais de vaccination contre la fievre Q, il est recommand6 de pratiquer au pr6alable des 6preuves cuta- n6es et un examen s6rologique par la m6thode de la micro- agglutination. REFERENCES 1. BERMAN, S. ET AL. Safety test for Q fever vaccine. Journal of bacteriology, 79: 747 - 751 (1960). 2. BREZINA, A. Antigens and immunity in Q fever (in Slovak). Bratislava, VEDA, 1977, 116 pp. 3. BREZINA, R. & URVOLGYI, J. Extraction of Coxiella burneti phase I antigen by means of trichloroacetic acid. Acta virologica, 5: 193 (1961). 4. BREZINA, R. ET AL. Rickettsiae and rickettsial diseases. Bulletin of the World Health Organization, 49: 433-442 (1973). 5. BREZINA, R. ET AL. Q fever chemovaccine for human use. Acta virologica, 18: 269 (1974). 6. CRACEA, E. ET AL. Immunization in man with a soluble Q fever vaccine. Archives roumaines de pathologie experimentale et de microbiologie, 32: 45 - 51 (1973). 7. CRACEA, E. ET AL. Q-fever soluble vaccine effects in Coxiella burneti sensitized humans. Zentralblatt fur Bakteriologie Parasitenkunde, Infektionskrankheiten und Hygiene, I., Abt. Orig. A, 238: 413-418 (1977). 8. D'ANGELO, L. J. ET AL. Q fever in the United States, 1948- 1977. Journal of infectious diseases, 139: 613-615 (1979). 9. GENIG, V.A. ET AL. Experience on the mass immuniz- ation of humans with the M-44 live vaccine against Q fever. I. Subcutaneous method of immunization (in Russian). Voprosy virusologij No. 10, pp. 319-323 (1965). 10. JOHNSON, J. W. ET AL. Biological properties of the M-44 strain of Coxiella burnetti. Journal of infectious dis- eases, 133: 334 - 338 (1976). 11. JOHNSON, J. W. ET AL. Lesions in guinea pigs infected with Coxiella burneti strain M-44. Journal ofinfectious diseases, 135: 995-998 (1977). 12. KAZAR, J. ET AL. Suitability of the microagglutination test for detection of post-infection and post-vaccination Q fever antibodies in human sera. Acta virologica, 25: 235 - 240 (1981). 13. LIEBISCH, A. Ecology and distribution of Q-fever rickettsiae in Europe with special reference to Germany. In: Proceedings ofthe Vth Congress ofAcarology, East Lansing, Michigan, 6-12 August 1978. New York, Academic Press, 1979, pp. 225 - 231 (Recent advances in acarology, Vol. 2). 14. LORBACHER DE RUIZ, H. Q fever in Colombia, S.A. A serological survey of human and bovine populations. Zentralblatt fur Veterindrmedizin (Reihe B), 24: 287 - 292 (1977). 15. MARMION, B. P. Development ofQ fever vaccines, 1937 to 1967. Medical journal of Australia, 54: 1074- 1078 (1967). 16. ORMSBEE, R. A. ET AL. The influence of phase on the protective potency ofQ fever vaccine. Journal ofimmu- nology, 92: 404-412 (1964). 17. PALANOVA, A. ET AL. Epidemiology of Q fever in Slovakia. In: Proceedings ofthe2nd International Sym- posium on Rickettsiae and Rickettsial Diseases, Smol- enice, Czechoslovakia, 21 -25 June 1976. Bratislava, VEDA, 1978, pp. 435-441. 18. SCHRAMEK, S. ET AL. Improved method of preparing phase II Coxiella burneti antigen for the microaggluti- nation test. Acta virologica, 14: 415 (1970). 19. SEN, G. P. ET AL. Sero-epidemiological evidence of Coxiella burnetii infections among selected human population in Calcutta. Indian journal of medical research, 68: 911 - 916 (1978). 20. VANEK, E. & THIMM, B. Q fever in Kenya. Serological investigation in man and domestic animals. East African medicaljournal, 53: 678 - 684 (1976). 21. VIVONA, A. ET AL. Report of a field study with Q fever vaccine. American journal of hygiene, 79: 143- 153 (1964). 22. WISSEMAN, C. L., JR. Prevention and control of rickett- sial diseases, with emphasis on immunoprophylaxis. In: Proceedings of the 2nd International Symposium on Rickettsiae and Rickettsial Diseases, Smolenice, Czechoslovakia, 21 - 25 June 1976. Bratislava, VEDA, 1978, pp. 553-583.