Surveillance of meningococcal infections and other forms of purulent meningitis: a 4-year study in the USSR*

Bulletin of the World Health Organization, 64 (2): 263-269 (1986) © World Health Organization 1986 Surveillance of meningococcal infections and other forms of purulent meningitis: a 4-year study in the USSR* A. A. DEMINA,' V. I. POKROVSKIJ,2 T. I. ILJINA,3 L. I. LARINA,3 & N. P. DEVJATKINA3 The laboratory examination, by microscopic or bacteriological methods and by counterimmunoelectrophoresis, of CSF and blood from 2653 patients with purulent bacterial meningitis, including those with clinically diagnosed meningococcal infection, was carried out between June 1980 and June 1984. The results showed three main etiological agents: meningococci (79.9%o), pneumococci (10.8%1)oJ and Haemophilus influenzae type b (5.25o). Out of488 Neisseria meningitidis strains isolatedfrom the CSF or blood, 58.2% belonged to serogroup A, 17.2% to serogroup B, and 14% to serogroup C; infection due to theB serogroup reached nearly 59% among children in 1984. Serotypes were determined in 131 out of 151 strains of Streptococcus pneumoniae and the most frequent were types 1, 19 and 3; type 34, which was isolated from 4 patients, is not a component of thepneumococcal vaccine. The age groups at high risk were children under 5 years old (for meningococcal infection), adults and babies in the first year (pneumo- coccal meningitis), and children under3 years, especially between 6 months and 2years old (H.influenzae type b infection). A notable feature of the increase in meningococcal infections in many countries during the 1960s and 1970s was their multiple etiology. Thus, in addition to epidemics due to Neisseria meningitidis of theA sero- group, less intensive outbreaks provoked by organ- isms of the C serogroup were recorded, as well as rises in infection due to the B serogroup meningococci, especially in European countries (6, 10, 15). Some authors have proposed that other agents might be responsible for a substantial number of meningitis cases (1, 3, 9). It appears therefore that the health care services in the affected countries should take im- mediate preventive measures against purulent menin- gitis, including meningococcal infections, especially when the epidemiological situation is grave. The present study describes: - the main etiological agents of purulent bacterial meningitis, based on hospital data from the USSR; -the groups at highest risk in relation to the etiology; * From the Central Research Institute of Epidemiology, Ministry of Health of the USSR, Novogireevskaja 3a, Moscow 111-123, USSR. ' Head, Meningococcal Laboratory, Central Research Institute of Epidemiology. Requests for reprints should be sent to this author. 2 Director, Central Research Institute of Epidemiology. 3 Scientist, Meningococcal Laboratory. -the serogroups and serotypes of the agents isolated from infected patients. MATERIALS AND METHODS Samples of cerebrospinal fluid (CSF) and blood were taken for laboratory examination from patients who were clinically diagnosed as having a meningo- coccal infection or some other form of purulent bacterial meningitis, or were suspected to have such a diagnosis. These patients were admitted from the city of Moscow or the Moscow region to the 2nd Infection Clinic in the period between June 1980 and June 1984. Of the total of 2653 patients, 1544 were under and 1109 were over 14 years old. Newborn patients with meningitis were admitted to a special hospital for infants; in our study these cases were rare. The above numbers of cases presenting with meningo- coccal infection and purulent bacterial meningitis do not give the total incidence of these infections in Moscow, since some patients (especially with purulent bacterial meningitis) could have been admitted to other hospitals. The CSF and blood samples were examined by established microscopic or bacteriological methods and by counterimmunoelectrophoresis (CIE). In the 4662 -263- A. A. DEMINA ET AL. cultures, three types of agar media were used: 200/o serum medium, 5% blood medium, and chocolate agar, and the colonies were identified by routine methods. The following antisera were used: meningo- coccal, produced by the Stavropol Institute for Vac- cines and Sera; pneumococcal, produced by the Statens Seruminstitut, Copenhagen, Denmark, and from experimental pneumococcal antisera made in 1983 by the Metchnikov Central Institute for Vaccines and Sera (Moscow); and antiserum to Haemophilus influenzae type b (provided by Dr J. Robbins, Bureau of Biologics, Food and Drug Administration, Bethesda, MD, USA). The sensitivity ofthe isolated cultures to antibiotics was established by the disc method, using commercial discs with pre-set antibiotic concentrations and a standard nutrient medium AGV (supplied by the Dagestan Institute for nutrient media). Depending on the diameter of the zone of bacterial growth inhib- ition round the disc, the bacteria were differentiated as sensitive, highly sensitive, or resistant to the antibiotic. RESULTS Because the study was focused on the hospital admitted patients or suspected patients with meningococcal infection, this etiology prevailed over the others -the diagnosis being confirmed in 1401 out of 1956 cases (71.6%) (840 children and 561 adults) (Table 1). In the remainder, we failed to isolate meningococci or to discover their antigens in the CSF or blood, the diagnosis having been made clinically. Out of 697 patients with a clinical diagnosis of purulent bacterial meningitis, a pneumococcal etiology was confirmed in 190 (27.3%) and H. in- fluenzae type b in 92 (13.2%). In 70 cases (10%), staphylococci, streptococci, Gram-negative rods, Candida and other organisms were confirmed, but we failed to determine the etiology in 345 of these patients with a clinical picture of meningitis (Table 1). Thus, out of 1753 cases with laboratory confir- mation, 79.9% were due to N. meningitidis, 10.8% to S.pneumoniae, 5.25% to H. influenzae type b, and Table 1. Distribution of meningitis patients according to disease etiology and age (under or over 14 years) Number of patients' Diagnosis and etiology Under 14 years Over 14 years Total N. meningitidis 840 (70.2)b 561 173.9) 1401 1197 759 1956 Purulent meningitis: 161 46.4) 191 (54.6 352 150.5)(46.4)350 (5.)697 ( 5 Str. pneumonia 42 148 190 H. influenzae type b 85 7 92 Candida 15 1 16 Streptococci 5 5 10 Staphylococci 8 14 22 Gram-negative rods - 11 11 Escherichia coli 1 2 3 Salmonella "c" 2 - 2 Ps. aeruginosa 2 1 3 Proteus 1 - 1 Klebsiella - 1 1 Haemophilus - 1 1 Totals 1001 752 1753(64.8) (67.8) 265 (66.1) a Where there are two figures, the numerator indicates the number of patients with laboratory confirmation of the diagnosis; the denominator gives the number of patients with a clinical diagnosis whose CSF and blood were examined at the laboratory. b Figures in parentheses are percentages. 264 MENINGOCOCCAL INFECTIONS AND MENINGITIS IN THE USSR Table 2. Age distribution of patients with purulent meningitis according to the three main etiological agents Age group (years) Etiology <1 1-2 3-4 5-6 7-13 14-19 20-49 50 Total N. meningitidis 435 409 157 78 114 261 350 152 1956 Total 1001 (51.2)a 453 (23.2) 502 (25.7) Str. pneumoniae 19 9 6 1 9 9 85 52 190 Total 34 (17.9) 19 (10) 137 (72.1) H. influenzae type b 34 40 4 3 5 1 4 1 92 Total 78 (84.8) 9 (9.8) 5 (5.4) a Figures in parentheses are percentages. 3.997o to other organisms. Our data confirm the results of others on the etiology of purulent bacterial meningitis (1, 9, 12, 14). Table 2 shows the distribution of patients with purulent meningitis by age and the 3 main etiological agents. Over half the patients with meningococcal infection were children under 5 years, 43.1% being under 3 years. The proportions of patients aged be- tween 5 and 20 years and over 20 years were roughly equal. About 18% of the patients with pneumococcal meningitis were children under 5 years, 10% being under 1 year (Table 2). The majority of these patients were over 20 years (72.17o), 27.4% being over 50 years old. In the H. influenzae type b group, 84.8% were children under 5 years old. Since the H. influen- zae polysaccharide vaccine is effective in infants up to the age of 18 months, the following detailed age- specific incidence data may be noted: under 3 months (2 children), between 3 and 6 months (6), between 6 and 12 months (26), between 1 year and 18 months (19), between 18 months and 2 years (15), and be- tween 2 and 3 years (6). The most affected age group was between 6 months and 2 years. Among the other etiological agents, Candida was detected in the CSF of 15 children aged under 2 months and in one adult over 50 years. These patients had been transferred from other hospitals and this organism must have become associated as a secon- dary infection. Other agents were identified both in the children and in the one adult. Considering the marked group-specific protection given by the meningococcal capsulated polysac- charide vaccine (4, 5), we investigated the meningo- coccal serogroups responsible for the infection in our patients by two methods: isolated cultures were tested Table 3. Percentage distribution of patients with meningococcal infection according to serogroup (after AR)' and meningococcal antigens (after CIE)' between June 1980 and June 1984 Percentage belonging to serogroups No. of cases NAG or Patients studied b A B C Y Other PAG' AAG' Under 14 years old 553 (192)c 42.8 21.8 21.1 3.0 2.2 5.9 2.9 Over 14 years old 601 (296) 69.4 14.1 9.5 1.6 0.8 2.6 1.8 Total 1154 (488) 56.6 17.8 15.0 2.3 1.5 4.2 2.3 a AR = agglutination reaction; CIE = counterimmunoelectrophoresis; PAG = polyagglutinable; NAG = not agglutinable; AAG = auto- agglutinable. b Includes all cases obtained by AR (488) and by CIE (666). c Figures in parentheses are the numbers of cases obtained by bacterial agglutination only. 265 A. A. DEMINA ET AL. by the agglutination reaction (AR), and antigens in 1984 only occasionally. At the same time, in 1983 found in the CSF were tested by counterimmuno- the B serogroup meningococci increased, these in- electrophoresis. The latter method gave 1.4 timfes fections among the children reaching nearly 59% more positive samples than the number by AR in 1984. Among adults (>14 years old), the pre- cultures: 666 and 488, respectively (Table 3). Out of dominant causative organism was N. meningitidis of the 488 cultures, 284 (58.2%) were related to the A serogroup A. serogroup, 370 of the 666 CIE-positive samples The sensitivity to antibiotics was studied in 107 (55.6%) showing antigens of this serogroup. In a con- strains ofN. meningitidis: 106 were sensitive or highly siderable number of adult cases this serogroup was sensitive to benzathine benzylpenicillin (10 itg), 103 predominant (68.2% byAR and 70.5% by CIE), with to ampicillin (10 jLg) and erythromycin (15 jug), and fewer cases in children (42.7% and 42.9%, respec- 101 to oxacillin (10 ,g). All strains were highly tively). On the other hand, the B and C serogroup sensitive to chloramphenicol (30 gg) and tetracycline meningococci and their antigens were more often (30 ,tg). detected in patients aged under 14 years. Other sero- In the surveillance of purulent bacterial meningitis groups were also encountered, as shown in Table 3. cases, the determination of pneumococcal serotypes Table 4 shows the changes in the distribution of from these patients is very important and this sero- patients according to meningococcal serogroups typing was commenced in 1981, according to WHO during the four years of the study. Some differences recommendations. After the pneumococci were may be noted. Thus, the proportion of children under isolated from the CSF or blood, thus confirming their 14 years in the meningococcal A serogroup (AR- role in the etiology, they were differentiated and iden- proven) dropped from 51.9% in 1980 to 37% in 1984 tified. Specimens were also sent to the regional WHO (CIE-proven: from 66.6% in 1980 to 30.5% in 1984). collaborating centre (Dr J. Henrichsen, Staatens While serogroup C meningococcal infections were Seruminstitut, Copenhagen) for a parallel study. Out initially in the second place, they were found of 190 such patients with a pneumococcal etiology, Table 4. Distribution of patients diagnosed by AR and CIE according to the meningococcal serogroups or their antigens, by year and age group No. of patients diagnosed by AR No. of patients diagnosed by CIE in each serogroup:' in each serogroup:' Year and age group A B C Totalb A B C Totalb 1980 (June-December): Children 14 0 6 27 18 0 3 27 Adults 28 0 8 38 25 1 3 29 1981: Children 20 1 15 44 36 4 31 75 Adults 32 1 4 44 40 1 7 56 1982: Children 20 1 14 42 34 1 14 55 Adults 58 2 4 72 60 0 2 65 1983: Children 1 1 12 4 33 35 14 29 99 Adults 51 12 7 80 55 14 15 86 1984 (January-June): Children 17 27 1 46 32 61 10 105 Adults 33 23 5 62 35 31 2 69 a AR = agglutination reaction; CIE = counterimmunoelectrophoresis. b The totals include small numbers of patients in other serogroups than A, B and C. 266 267MENINGOCOCCAL INFECTIONS AND MENINGITIS IN THE USSR Table 5. Serotypes of pneumococci isolated from meningitis patients during the period 1981-84 Age of patients (years) No. of strains isolated Under Over Serotypes 14 14 Total CSF Blood Total 1 4 18 22 22 7 29 2 1 5 6 6 1 7 3 - 9 9 10 4 14 4 - 3 3 2 3 5 5 1 5 6 5 1 6 6 2 7 9 9 3 12 7 1 - 1 1 - 1 9 - 1 1 1 1 2 10 - 1 1 1 - 1 12 5 3 8 8 - 8 13 - 1 1 1 - 1 14 1 1 2 3 - 3 15 - 2 2 2 1 3 18 1 5 6 6 - 6 19 3 9 12 10 5 15 20 - 3 3 4 - 4 21 1 - 1 1 - 1 23 1 1 2 2 - 2 25 - 2 2 2 - 2 27 1 - 1 1 - 1 29 - 1 1 1 - 1 34 1 3 4 4 - 4 35 - 2 2 2 - 2 42 1 - 1 1 - 1 Total 24 82 106 105 26 131 the diagnosis was confirmed bacteriologically in 163 (85.8%), and microscopically in 27 (14.2%). The pneumococci were serotyped in 106 patients: in 4 cases the agent was isolated in the rough form and 13 cases were not agglutinable; and there was one case each that was polyagglutinable, or agglutinable only with "omniserum" or with pool C, while in 37 patients we failed to study the serotypes of pneumo- cocci. In Table 5 are presented the results of typing of pneumococci isolated from patients between 1981 and 1984. Of these patients, 24 were children under 14 years old and 82 were adults. A total of 131 strains were studied, 105 isolated from CSF and 26 from blood. In 22 patients the pneumococci were isolated from both blood and CSF, in 4 patients from blood only. The typing showed that the pneumococci were related to 24 known types. The frequency of the iso- lated types was as follows (in decreasing order): 1, 19, 3, 6, 12, 2, 5, 18 which together accounted for 73.3% of the total typed. Type 12 was more often isolated in children. In addition to the results presented in Table 5, we studied 5 strains, 2 of which had been isolated from the CSF of children in the city of Minsk; these were types 5 and 34. The three remaining strains had been obtained from some other Moscow clinics and belonged to types 1 (2 cases) and 3. The serotypes of pneumococci isolated from the CSF and blood simultaneously or obtained for re- peated laboratory examinations were also investi- gated. In 22 out of 34 patients the agents belonged to the same type. In the others the serological features of the pneumococci differed. Of those that had been iso- lated from the blood, 6 strains were not agglutinable, 3 strains were in the rough form, and in 2 cases dif- ferent serotypes were isolated. The results of repeated pneumococcal isolations from the CSF of one 46- year-old patient with a prolonged course of illness are interesting because three serotypes were identified: 3, 14, and 19 (twice). The sensitivity to antibiotics of 15 pneumococcal strains was examined at the time of isolation. All of them were found to be sensitive to benzathine benzylpenicillin, rifampicin, chloram- phenicol, oxacillin, and meticillin, i.e., they were sensitive to all the antibiotics used in treating menin- gitis cases of this etiology. H. influenzae type b was the causative organism of purulent bacterial meningitis in 92 patients: in 56 (60.9%) the organism was isolated in culture and in 36 (39.1%) the etiology was confirmed by microscopy and CIE. The high yield of the latter method may be noted because it made possible the identification of non-viable bacteria and their antigens. In all 92 cases, for instance, the CIE results were positive and in agreement with bacteriological data (56 cases) and with microscopy (23 cases). Besides, in 23 out of 25 negative samples on microscopy, the specific antigen was found by the CIE test and in the 2 remaining cases the culture was grown. The sensitivity to anti- biotics was tested in 19 strains only: 10 were sensitive or highly sensitive to chloramphenicol (30 ,g) and 13 were sensitive to tetracycline (30 ptg), i.e., they were sensitive to the antibiotics used for treating meningitis of this etiology. Five strains were found to be sensitive to benzathine benzylpenicillin. DISCUSSION Acute bacterial meningitis, which is mainly due to meningococcal infection, poses a serious problem for the health services in many countries (7, 15). Deter- mination of the main etiological agents responsible (including their type and group specificity) and the age groups at maximum risk is very important both for epidemiological surveillance and for the develop- ment of vaccines with an optimum antigen com- position. The results of our study of pathological material from 2653 patients with purulent bacterial meningitis showed that meningococci, pneumococci and H. influenzae type b were responsible for 96.2% A. A. DEMINA ET AL. of laboratory confirmed diagnoses. The predomi- nance of meningococci (79.9%) in the etiology was not due to an epidemic situation, which is supported by the age distribution pattern of these patients. According to Makela et al. (8), 50% of the patients with meningococcal illness in a non-epidemic period were children aged under 5 years. In our results of similar patients, 51.2% were under 5 years, 23.2% were between 5 and 20 years, and 25.7% were over 20 years old. Thus, the age group with maximum risk was the 0-5-year-old group. The serogroup of the isolated meningococci ap- peared to vary with the age of the patients. In adults, serogroup A was predominant, although the rate fell from 76.5-80.5% in 1980-81 to 54% in 1984. This means that in a serious epidemic situation the use of meningococcal A polysaccharide vaccine must be focused on adults. In patients aged under 14 years the A serogroup was isolated in less than half of the cultures studied (e.g., in 1984, 37%). Other meningo- coccal serogroups, however, were isolated from patients in this age group (e.g., in 1981-82, 30-38% in C serogroup; in 1984, up to 59% in B serogroup). This means that application of the A serogroup vaccine alone in this age group cannot be sufficiently protective, compared with the A + C vaccine. The increase in the number of infections caused by the B serogroup meningococcus supports the need for development of a specific vaccine against this agent. It is known that a pneumococcal vaccine will be satisfactory if it contains at least 80% of the pneumococcal serotypes that are frequently isolated from patients. This calls for a study to establish the serotypes of pneumococci that are responsible for disease during the surveillance study of patients with purulent bacterial meningitis. In the present study, standard methods were followed and for the first time in the USSR the serotypes of pneumococci isolated from the CSF and blood of patients are reported (see Results section). Of the 24 known serotypes of pneumococci, 16 are represented in the vaccine preparation (13). Serotypes 8, 11, 17 and 22, which are included in the vaccine, were not encountered in our study, while types 7 and 9 were isolated only once and type 34, which is not represented in the vaccine, was found in Moscow (4 cases) and in patients in Minsk. Thus, depending on the prevalent pneumo- coccal serotypes in any area, the composition of the vaccine preparation has to be modified. This view was also expressed by Denis et al. (2). H. influenzae type b was responsible for 92 cases in our study, this organism having unexpectedly been separated from the CSF of 7 adult patients. Taking into consideration the prolonged course of meningitis due to this etiology, the identification of specific antigens in the samples of pathological material is important. The epidemiological surveillance of purulent bacterial meningitis cases should take into account the possible resistance of the bacteria to antibiotics (11). Our results showed a high sensitivity among the pneumococci and meningococci to the antibiotics normally used in the treatment of meningitis of this etiology, while only 19 strains of H. influenzae type b could be tested in this way. These investigations are continuing. ACKNOWLEDGEMENTS We thank Dr G. Causse and Dr T. Kereselidze (of WHO), Dr J. Robbins (of the Bureau of Biologics, Food and Drug Administration, Bethesda, MD, USA), and Dr J. Henrichsen (of the Staatens Seruminstitut, Copenhagen, Denmark) for their encouragement and help. RP,SUMt SURVEILLANCE DES INFECTIONS A MENINGOCOQUES ET D'AUTRES FORMES DE MENINGITES PURULENTES: UNE ETUDE DE 4 ANS EN URSS De juin 1980 k juin 1984, on a proc6d -A 1'examen de laboratoire-au microscope ou par des m6thodes bact6rio- logiques, et par 6lectrosynerese-du LCR et du sang de 2653 malades atteints de m6ningite purulente, le diagnostic cinique d'infection i m6ningocoques ayant deja 6t6 pose pour un certain nombre d'entre eux. Les r6sultats ont revele trois principaux agents etiologiques: meningocoques (79,9%), pneumocoques (10,8%) et Haemophilus in- fluenzae type b (5,25%). Sur 488 souches de Neisseria meningitidis isol6es du LCR ou du sang, 58,2% appar- 268 MENINGOCOCCAL INFECTIONS AND MENINGITIS IN THE USSR 269 tenaient au serogroupe A, 17,2% au s6rogroupe B et 14% au s6rogroupe C; en 1984, la proportion des infections dues au s6rogroupe B atteignait pres de 59% chez les enfants. Le s6rotype a ete d6termin6 pour 131 des 151 souches de Streptococcus pneumoniae; les types les plus frequents 6taient le type 1, le type 19 et le type 3; le type 34, isole chez quatre sujets, n'entre pas dans la composition du vaccin antipneumococcique. Les groupes d'Age les plus exposes 6taient les enfants de moins de 5 ans (pour les infections a m6ningocoques), les adultes et les enfants de moins d'un an (pour les infections a pneumocoques), et les enfants de moins de 3 ans, plus parti- culierement ceux ag6s de 6 mois a 2 ans (pour les infections a H. influenzae type b). REFERENCES 1. Cmwoz, M. ET AL. Etude 6pid6miologique des cas de m6ningites purulentes hospitalis6s a Dakar pendant la d6cennie 1970-1979. Bulletin of the World Health Organization, 59: 575-584 (1981). 2. DENIS, F. A. Er AL. Etude multicentrique des s6rotypes de pneumocoques en Afrique. Bulletin of the World Health Organization, 61: 661-669 (1983). 3. GEISELER, P. J. ET AL. Community-acquired purulent meningitis: a review of 1316 cases during the antibiotic era, 1954-1976. Reviews of infectious diseases, 2: 725-745 (1980). 4. GOLD, R. & ARTENSrEIN, M. Meningococcal infec- tions. 2. Field trial of group C meningococcal poly- saccharide vaccine in 1969-1970. Bulletin of the World Health Organization, 45: 279-282 (1971). 5. GOLD, R. ET AL. Immune response of human infants to polysaccharide vaccines ofGroupsA and C ofNeisseria meningitidis. Journal of infectious diseases, 136 (suppl.): S31-S37 (1977). 6. HOLTEN, E. Serotypes ofNeisseria meningitidis isolated from patients in Norway during the first six months of 1978. Journal of clinical microbiology, 9: 186-188 (1979). 7. LAPEYSSONNIE, L. Epid6miologie de la m6ningite c6re- brospinale a m6ningocoques: le point de vue de Sirius. Medecine tropicale, 43 (hors s6rie, No. 1): 9-14 (1983). 8. MAKELA, P. ET AL. The importance and control of meningitis in Finland. In: New Developments in vaccines. Proceedings of an International Symposium, Cairo, 1979, pp. 21-29. 9. MINER, W. & EDMAN, D. Acute bacterial meningitis in Cairo, Arab Republic of Egypt: 1 January 1971 through 31 December 1975. American journal of tropical medicine and hygiene, 27: 986-994 (1978). 10. MOCCA, L. F. ET AL. Serotypes and polyacrylamide gel electrophoresis types among disease-associated isolates of group B Neisseria meningitidis in Spain, 1976-1979. Journal of infectious diseases, 148: 249-253 (1983). 11. PARKER, M. T. Antibiotic resistance in pathogenic bacteria. WHO Chronicle, 36: 191-196 (1982). 12. REY, M. ET AL. Aspects epid6miologiques des m6nin- gites purulentes en Afrique tropicale (d'apres 1052 cas observ6s a Dakar). Lyon medical, 228: 503-508 (1972). 13. ROBBINS, J. ET AL. Considerations for formulating the second-generation pneumococcal capsular polysac- charide vaccine with emphasis on the cross-reactive types within groups. Journal ofinfectious diseases, 148: 1136-1159 (1983). 14. SANBORN, W. Meningitis diagnostic bacteriology. Journal of the Egyptian Public Health Association, 44: 385 404 (1969). 15. VELIMIROVIC, B. The situation of meningococcal meningitis in Europe. MeLdecine tropicale, 43 (hors s6rie, No. 1): 15-20 (1983).