Spread of Streptococcus pneumoniae in Families. II. Relation of Transfer of S. pneumoniae to Incidence of Colds and Serum Antibody

Transcription

1 THE JOURNAL OF INf'ECTlOUS DISEASES VOL. 132, NO.1 JULY 197:5 197:5 by the University of Chicago. All rights reserved. Spread of Streptococcus pneumoniae in Families. II. Relation of Transfer of S. pneumoniae to Incidence of Colds and Serum Antibody Jack M. Gwaltney, Jr-,; Merle A. Sande, Robert Austrian, and J. Owen Hendley From the Departments of Internal Medicine and Pediatrics, University of Virginia School of Medicine, Charlottesville. Virginia; and the, Department of Research Medicine, University of Pennsylvania, Philadelphia, Pennsylvania Factors that affect the spread of Streptococcus pneumoniae and the antibody responses associated with colonization were studied in families for periods of eight to 52 weeks. Surveillance included daily recording of respiratory symptoms and bimonthly pharyngeal cultures for identification of the pneumococcal currier state. Rhinovirus cultures were included for a portion of the study period. Intrafarnilial carriage of a single type ofs. pneumoniae and simultaneous spread to more than one family member were common. Spread often occurred in association with an upper respiratory tract infection; simultaneous transmission of S. pneumoniae and a rhinovirus was documented. Preexisting, type-specific serum antibody did not prevent acquisition of homotypic S. pneumoniae but did appear to shorten the duration of pharyngeal carriage. Sera of all 11 adults had> 150 ng of antibody nitrogen/ml of homotypic serum antibody (measured by a radioimmunoassay) before colonization. In contrast, only one of 13 preschool children had homotypic antibody concentrations of this magnitude before colonization. A threefold or greater rise in the concentration of homotypic antibody occurred in 13 of 2 children (5%) after acquisition of S. pneumoniae ; the increase in antibody concentration was associated with illness in six of the children. On the other hand, acquisition of S. pneumoniae in adults was not associated with an increase in concentration of homotypic serum antibody. Little is known about the mechanisms for the spread of Streptococcus pneumoniae among normal people. In this investigation and in the accompanying paper [1], the epidemiologic behavior ofs. pneumoniae in families was studied. The possibility that S. pneumoniae is disseminated in association with viral colds was examined by determining the relation of colds to transmission of S. pneumoniae, Special emphasis was given to rhinovirus infections because Received for publication November 12, 197, and in revised form February This work was sponsored in part by NIH-NIAID contract no. NIH-NIAID from the Infectious Disease Branch, National Institute of Allergy and Infectious Diseases. National Institutes of Health; by NIH-NIAID contract no. PH NO I AI 32517; and by contract no. DADA 17-8-C-9019 from the U.S. Army Medical Research and Development Command. Please address requests for reprints to Dr. Jack M. Gwaltney, Jr., Department of Internal Medicine, University of Virginia School of Medicine, Charlottesville, Virginia of their relatively high prevalence. In addition, the relation of serum antibody to the carriage of S. pneumoniae was studied by means of a radioimmunoassay of type-specific capsular antibodies. Materials and Methods Population. The two populations have been described [1]. Fifteen families with children were observed for approximately one year (December 1971 to December 1972) for the spread of S. pneumoniae and for symptomatic respiratory disease. In addition, 52 families composed of 92 adults and 105 children were tested for rhinovirus infection during an eight-week period (August 28 to October 23, 1972) in an attempt to obtain specific information on a possible relation between the spread ofrhinoviruses and the transfer of S. pneumoniae. Surveillance. The occurrence of respiratory disease was determined by the use of a symptom record card, which covered a two-week period 2

2 Pneumococcal Spread in Families /I 3 for each family member. To encourage accurate reporting of symptoms, the mother in each family was contacted frequently by telephone and visited weekly by a nurse-epidemiologist. The occurrence of a cold was defined by criteria detailed previously [2]. The minimal criterion for an illness was the occurrence of two respiratory symptoms on one or more days or of one respiratory symptom on two or more days. Collection of specimens. Pharyngeal swabs for isolation ofs. pneumoniae were obtained by methods described previously [I]. During the 12-month surveillance period, culture specimens were collected every two weeks. During the eight-week study of rhinovirus transmission in the fall, samples for isolation of S. pneumoniae were collected weekly, and samples for rhinovirus isolation were obtained from persons with acute respiratory symptoms. For the isolation of rhinovirus, nasal and pharyngeal secretions on cotton-tipped swabs were combined in 5 ml of collecting broth [3]. Samples of serum from either capillary or venous blood were obtained as described previously []. Sera were collected at three-month intervals from individuals in the 15 families in the one-year study and at the beginning and end of the study in the fall. In addition, specimens of acute and convalescent serum were obtained at appropriate times in relation to an acute respiratory infection. Isolation of S. pneumoniae. Details of the methods used are described in the accompanying paper [1]. Rhinovirus isolations. Rhinoviruses were isolated in human embryonic lung cells (WI-38) and identified by characteristic CPE, by acid lability, and by neutralization with hyperimmune antisera. Neutralization tests with paired sera were performed in WI-38 cells by methods described previously []. Microtitration and/or standard techniques were selected, depending on the volume of the serum specimen available and the ability of the virus strain to grow well in cell culture. Three to 10 TCID so of virus were used. End points are reported as reciprocals of the initial dilutions of serum. Serologic tests. Antibody to capsular polysaccharides of S. pneumoniae was measured by a radioimmunoassay [5] based upon the Farr technique []; biosynthetically radiolabeled (lc) capsular antigens were employed. Specimens from individuals who had donated sera prior to acquisition of and during colonization with a new type ofs. pneumoniae were selected for testing. Results Spread of s. pneumoniae in families. A prominent feature of the carriage of S. pneumoniae was intrafamilial spread, which resulted in the clustering of specific types within families. Twenty-five episodes of spread of a single type of S. pneumoniae from one family member to another were noted during the period of observation. Examples of the epidemiologic behavior of the organism in representative families are shown in figure I. In the family described in the top panel of figure I, S. pneumoniae type 23 was detected in the mother and the younger child when surveillance was begun. In January and February, the organism spread to the other child and to the father. Type 19 was introduced into the family by the older child in April and spread to the mother and to the younger child within a month. The rapid spread of a newly introduced type of S. pneumoniae, like that which occurred with type 19. was common. Seven of the 25 transfers took place within two weeks after introduction of a new type into the home. Temporal clustering of specific types of S. pneumoniae is evident in the top panel of figure I; type 23 was isolated from all members of the family during one period, and type 19 was isolated from three members during another. In another family (figure 1, middle panel), each offour types (19, 11,, and 3) was isolated from two or more family members on different occasions. The oldest child carried five different types ofs. pneumoniae during 10 months of observation. The behavior of type 3 was of particular interest. In June type 3 was introduced into the family by the oldest child. It was carried without transfer to other family members until early August, when it was disseminated to the mother and to the youngest child. Six additional episodes of apparently simultaneous dissemination of a single type to more than one member were seen in other families. In addition, five

3 Gwaltney et al " " " " 0- M TYPES ISOLATED N N S N N N CATEGORY ::::. ~ ~ ;::: /29/71- /12/72 N on... ~ N N on on on 00 Father 0 0 NT NT Preschool Child Preschool N :e.... ~ ~ ~ ~ '" 0 ~ ~ '" N ~.. TYPES ISOLATED 0::. 0::. s ~ 0::. s..... "- "- ~ "- :::. "- <, 0::. ~ CATEGORY N ~ ~... '"... e- e- S! S! S! ::: ::: 2/7-11/27/72 Father NT NT 0 0 NT NT NT 0 0 Mother NT NT NT Grammar School o#j NT NT Grammar School NT NT NT 0 0 NT NT Preschool NT NT NT NT ~ ~... l!... ~ ~ ~ l! ~ A l:: ~ ~ TYPES ISOLATED s '- ~ "- t:. ~ ~ "- ~ =:: ~ s S ~ ~ ~ CATEGORY ~ li!....!:! 1/10-12/10/72 Father NT NT NT Mother NT High SChool child a , NT High School Child NT Gram",ar SchOol Child NT Figure 1. Spread of Streptococcus pneumoniae in families. Top, intrafamiiial clustering of types 23 and 19. Middle, intrafamiiial clustering of types I I,, 19. and 3. with an episode ofdissemination of type 3 from the oldest child to the mother and preschool child. Bottom. intrafamilial dissemination oftype 15 with gaps in recovery of the organism on serial culture. NT = not tested. episodes were documented in which two or more members simultaneously acquired a type of S. pneumoniae that was new to the family. This finding suggests that the organism was acquired from a common source outside of the family. In a third family (figure 1, bottom panel), S. pneumoniae was not isolated for five months (between December and May). In May, however, the grammar school child acquired type 15, which he continued to carry through July without spread. In August type 15 spread to the mother and to an older sibling. Finally, in November, the third child in the family acquired type 15. The findings in this family demonstrated that S. pneumoniae was not always isolated consistently, although intermittent recovery of the same type suggested that carriage was actually chronic. Although inoculation of mice for detection of the carriage of S. pneumoniae was the most sensitive technique employed, it was not completely reliable at all times. In an attempt to minimize this problem, acquisition of a serotype of S. pneumoniae was defined as isolation of that type with two or more antecedent samples negative for that type. This criterion was used for calculation of acquisition rates and for purposes of relating the presence of serum antibody to acquisition of S. pneumoniae. Spread ofs. pneumoniae in relation to colds. The temporal sequence of symptomatic respiratory disease and transfer of S. pneumoniae was examined. Data from a representative family are shown in figure 2. There were five acquisitions of S. pneurnoniae and 23 colds in this family during seven winter and spring months. The excess of respiratory illnesses over acquisitions of S.

4 Pneumococcal Spread in Families ;... '"., 0 on 0- C'O TYPES ISOLATED «:;. ~ 0 C'O c::i 00 '" ~ «:;. «:;. CATEGORY ;? ::::. ~ ~ S ;:: C'I C'I 11/29/71. /12/12 '" '"., o:r on 00 Father 0 0 NT NT @@ Preschool Child l!}@ Preschool Figure 2. Relation of carriage of Streptococcus pneumoniae to incidence of respiratory symptoms (dark bars) in a representative family. NT = not tested. pneumoniae seen in this family was the usual finding in the other families in the study. Comparison of the number of respiratory illnesses with the number of acquisitions of S. pneumoniae in the 15 families studied for one year revealed a ratio of at least three colds to each acquisition of the bacteria. In 25 episodes of transfer of S. pneumoniae, both a presumed donor and a recipient could be identified. In 1 of these episodes (5%), the donor had symptoms of an upper respiratory tract infection (sneezing, rhinorrhea, nasal congestion, or cough) during the two-week period in which transfer occurred. For comparison, the frequency of respiratory tract symptoms in the donors during all other two-week periods of the study was examined to determine whether the donors had respiratory symptoms more commonly during periods of transfer of S. pneumoniae than at other times. Symptoms were present during 159 of the other 23 two-week periods of observation (37%). Persons who transmitted S. pneumoniae, therefore, tended to have upper respiratory symptoms more frequently during the period of transfer than at other times (X 2 = 3.3, P = 0.0). Spread of rhinovirus and S. pneunomiae. During the study in the fall of the spread ofs. pneumoniae and of rhinovirus colds, a rhinovirus was introduced into II of the 52 families. In seven of the II families, the individual who introduced the rhinovirus into the home was not carrying S. pneumoniae at the time. In one family in which coprimary infections caused by rhinovirus occurred, one of the infected individuals was carrying S. pneumoniae and the other was not. Neither rhinovirus nor S. pneumoniae spread to other members of this family. In another family, the child who introduced a rhinovirus and his sibling, who subsequently became infected with it, were both carrying S. pneumoniae type 15 before they acquired the rhinovirus. In another family, a grammar school child had a symptomatic infection caused by rhinovirus and was carrying s. pneumoniae. The organism spread to her older sister, who developed a cold due to a different type of rhinovirus during the same period. In one family, simultaneous spread of a rhinovirus and of S. pneumoniae was documented (figure 3). Rhinovirus type 0, introduced by a grammar school child, spread to three family members, while S. pneumoniae type 3, carried by the same child, spread to one of three family members. Serologic results. Concentrations of antibody to S. pneumoniae before colonization varied with age (table I). All adults had> 150 ng of antibody nitrogen (AbN)/ml to the type of S. pneumoniae with which they were colonized. ANTIBODY TO - RV 00 /J,(UUjCONV WORKING RV 0 ADULT (fo.he'l <2-8 WORKING ADULT (Mothe,) 8-8 HIGH SCHOOL.,..- RV 0 2- CHILD JUNIOR RV 0 HIGH SCHOOL <2-32 CHILD GRAMMAR SCHOOL <-1 CHILD ~ I' 1 1 " 'I 8/28 9/ 9/11 9/18 9/2~ Figure 3. Simultaneous transmission of a rhinovirus and Streptococcus pneumoniae. R V 0 = rhinovirus isolation; dark bars =0" respiratory = isolation of S. pneumoniae. Rhinovirus infection in the grammar school child was established by a rise in titer of antibody.

5 Gwaltney et al, Table 1. Concentration of homotypic antibody to Streptococcus p neumoniae in the sera of adults and children prior to colonization. Population (age in years) Adults (""17) Schoolchildren (-1) Preschool children «).,, ,000 >2,000 5 (3) 2 () 11 () 20 () 27 () 27 () (7) 11(7) 7.5 (3) 0() 0() 0() 0 () 100 () 12 () 5 () (19) 100 (19) 15 (23) 0 (23) Antibody concentration" (capsular type) 275 (3) 295 (3) 00 () 12 (23) 2.50 (3) 20 (3) 10 (18) 175 (23) 950 (7) 82.5 (9) 1,000 (19) (19) 700 (23) 550 (8) '::100 (8) 2,50 (19) 2,50 (8) 3,00 (19) Antibody concentration is expressed as ng of antibody nitrogenlrnl. This finding differed from that in preschool children; only one of 13 had initial concentrations of ""150 ng of AbN/ml (P < 0.001, Student's r-test). Values for schoolchildren fell hetween those for the adults and those for preschool children but were significantly greater than those of the preschool children (P < 0.05, Student's r-test). The rise in level of type-specific antibody after colonization appeared to be influenced by age and/or the concentration of antibody before colonization. Thirteen (5%) of 2 children (~1 years old) had threefold or greater increases in concentrations of homotypic serum antibody after colonization. Antibody responses of this magnitude were not seen in any of nine adults (? < 0.01, Fisher's exact test). The rise in titer of serum antibody after colonization also varied with the initial concentration of antibody. Titers rose in II of 18 family members of all ages who had initial concentrations of ~ 100 ng of AbN/ml but in only two of 15 family members who had initial concentrations above this level (P < 0.01, Fisher's exact test). The same analysis of results from children only showed that II of 1 children with initial concentrations of ~ 100 ng of AbN/ml had threefold or greater rises in titer of antibody, compared with rises in two of seven children with initial concentrations of > 100 (P = 0.09, Fisher's exact test). On the basis of these data, one cannot say with certainty that low concentrations of antibody per se predisposed to homotypic antibody responses after colonization, although the trend in children suggested that this correlation might be valid. The duration of carriage was associated with both frequency of rise in titer of antibody and concentration of antibody before colonization. The mean duration of carriage was.5 ± weeks in persons without measurable antibody increases, whereas the duration in persons with threefold increases was 11.3 ± 10 weeks (P < 0.02, Student's r-test), At the same time, persons with initial concentrations of < 100 ng of AbN/ml carried homotypic S. pneumoniae for 7.3 ± 7.9 weeks, whereas in those with initial concentrations of >500 ng of AbN/ml, carriage lasted an average of only 3.3 ± 1. weeks (P < 0.03, Student's r-test). The correlation of duration of carriage both with antibody concentrations before colonization and with antibody increases was not unexpected in view of the relation between these two variables described above. The presence of respiratory symptoms at the time of acquisition of S. pneumoniae did not appear to influence the antibody response. Respiratory symptoms were recorded in six of 10 children who had threefold or greater rises in titer ofantibody and in six of 11 children who did not have rises of this magnitude after colonization. The frequency of colds in children, however, made it difficult to draw conclusions from this analysis with confidence. Thirteen children had threefold or greater increases in antibody concentration in association with the acquisition of a serotype of S. pneumoniae (table 2). A spectrum of antibody responses was observed in these children. Three of the children (D.G., J.H., and L.H.) had undetectable levels at first but later developed relatively high concentrations of antibody to S. pneumoniae type and had respiratory illnesses. Two other children (E.M. and M.M.), playmates of D. G., J.H., and L.H., also had brisk antibody

6 Pneumococcal Spread in Families l/ 7 Table 2. Thirteen children with serum antibody responses of threefold or greater associated with acquisition of Streptococcus pneumoniae. Individual (age) D.V. (13) S.S. (10) D.G. () J.H. (5) L.H. (3)t E.M. () M.M. (5) C.S. (9) S.F. (8) L.H. (3) D.V. (13) S.S. (10) A.N.. (3) Type of S. pneumoniae acquired Antibody concentration' Initial Final Comments 5 20 oo o , , , * Antibody concentrations expressed as ng of antibody nitrogen/mi. t Sister of J.H.; S. pneumoniae type not isolated from this child. responses to S. pneumoniae type and had respiratory symptoms during or close to the time of initiation of carriage. E.M. and M.M. received benzathine penicillin from a private physician. Two children (D. V. and C.S.) had low initial concentrations of serum antibody and responded to colonization with slight increases. Another child (S.S.) first had a moderate and later a high concentration of antibody to S. pneumoniae type 3 during a period when no symptoms were reported. Discussion The most prominent feature of the ecology ofs. pneumoniae in this study was the clustering of serotypes in families, a phenomenon which resulted from the apparent spread of specific types among family members. This observation is in accordance with findings by earlier workers. Brimblecornbe et ai. [7] and Masters et ai. [8] noted the importance of the family setting in the spread ofs. pneumoniae and examined the effeet of a variety of conditions (including crowding, climate, degree of dampness in the home. ventilation, social contacts, and general level of hygiene) on spread. Of these factors, only crowding was shown to enhance spread. Dowling et ai. [9] studied the epidemiology of No illness during first three months of carriage Illness during period spanned by serum specimens; penicillin therapy No symptom record from period of initial carriage; later illness treated with benzathine penicillin ; benzathine penicillin therapy Symptom record incomplete, no known illness during carriage S. pneumoniae in families, with particular emphasis on types 3 and 19, and observed an increased incidence of acquisition in the fall. Finland [10] observed spread of specific types of S. pneumoniae in the families of patients with pneumococcal pneumonia and with otitis media. Several patterns of spread of S. pneumoniae were observed in the current study. Two distinctive patterns (i.e., the apparently simultaneous acquisition of a particular type by two or more family members and the prolonged carriage by one family member with sudden dissemination to several other household contacts) suggested that a specific event led to dissemination of the bacteria. One of the goals of the current study was to determine whether a cold might be such an event. Straker et al. [II] suggested that S. pneumoniae migrates from the pharynx to the nose in adults with colds, an event evidenced by an increased rate of recovery ofthe bacteria from nasal cultures during respiratory illness. Brimblecombe et al. [7] noted that rates of carriage of S. pneumoniac increased after the onset of coryza. Characterization of the relation between colds and the dissemination of S. pneumoniae in this investigation was difficult because there were approximately three times as many colds as episodes of transfer of S. pneumoniae. How-

7 8 Gwaltney et al. ever, persons who were the presumed donors of S. pneumoniae to other family members tended to have respiratory symptoms at the time of transfer more often than they did at other times. This finding suggests that the enhanced production of respiratory secretions or some other mechanism associated with presumed viral colds may play a role in the transmission of S. pneumoniae from carriers to other susceptible family members. The study of specific rhinovirus colds in the fall was undertaken to define more precisely concomitant viral and bacterial spread. Rhinoviruses and S. pneumoniae were frequently found simultaneously in families in the fall. In addition, strains of both organisms were recovered from individual family members on multiple occasions, and, in one instance, definite simultaneous spread of the two organisms was demon strated. Serum antibody to S. pneumoniae may influence the epidemiological behavior of the bacteria in the family environment. Serum antibody to colonizing types of S. pneumoniae was present in relatively high concentrations in all adults, whereas children tended to have low or undetectable concentrations. Whether the lower rate of colonization with S. pneumoniae and the less persistent carriage in adults are due to a relatively high concentration of antibody remains to be determined. Of interest in relation to the acquisition of natural immunity to infection with S. pneumoniae was the finding that threefold or greater rises in titer of serum antibody occurred in more than half of the children during carriage. Many of these subjects were asymptomatic during this time. In contrast, no adults had rises in antibody titer associated with colonization. In the family studies of Finland [10], homologous antibody responses, as measured by agglutination and by mouse protection tests, were seen in children who carried S. pneumoniae. In that study, antibody responses were also seen occasionally in adult carriers of S. pneumoniae. Homotypic antibody responses are known to occur in cases of pneumonia caused by S. pneumoniae [10]; recently, up to 50% of children with acute otitis media caused by S. pneumoniae have been shown to develop serum antibody in association with that infection [12]. Therefore, type-specific serum antibody may result both from clinical infection with S. pneumoniae and from apparently asymptomatic colonization of the upper respiratory tract. The latter mechanism, because of its frequent occurrence, may play an important part in the development of natural immunity to infection and disease caused by S. pneumoniae. References I. Hendley, J. 0., Sande, M. A., Gwaltney, J. M., Jr. Spread of Streptococcus pneumoniae in families. I. Carriage rates and distribution of types. J. Infect. Dis. 132:55-1, Gwaltney, J. M., Jr., Hendley, J. 0., Simon, G., Jordan, W. S., Jr. Rhinovirus infections in an industrial population. I. The occurrence of illness. N. Engl. 1. Med. 275: , Gwaltney, J. M., Jr., Jordan, W. S., Jr. Rhinoviruses and respiratory illnesses in university students. Am. Rev. Resp. Dis. 93:32-371, 19.. Hendley, J. 0., Gwaltney, J. M., Jr., Jordan, W. S., Jr. Rhinovirus infections in an industrial population. IV. Infections within families of employees during two fall peaks of respiratory illness. Am. J. Epidemio!. 89: 18-19, Schiffman, G., Austrian, R. A radioimmunoassay for the measurement of pneumococcal capsular antigens and of antibodies thereto [abstract]. Fed. Proc, 30:58, Farr, R. S. A quantitative immunochemical measure of the primary interaction between I*BSA and antibody. J. Infect. Dis. 103:239-22, Brimblecombe, F. S. W., Cruickshank, R., Masters, P. L., Reid, D. D., Stewart, G. T. Family studies of respiratory infections. Br. Med. 1. 1: , Masters, P. L., Brumfitt, W., Mendez, R. L., Likar, M. Bacterial flora of the upper respiratory tract in Paddington families, Br. Med. J. 1: , Dowling, J. N., Sheehe, P. R., Feldman, H. A. Pharyngeal pneumococcal acquisitions in "normal" families: a longitudinal study. J. Infect. Dis. 12:9-17, Finland, M. Excursions into epidemiology: selected studies during the past four decades at Boston City Hospital. 1. Infect. Dis. 128:7-12, Straker, E., Hill, A. B., Lovell, R. A study of the nasopharyngeal bacterial flora of different groups of persons observed in London and south-east England during the years 1930 to Reports on Public Health and Medical Subjects, no. 90. His Majesty's Stationery Office, London, England, 1939, p Sloyer, J. L., Jr., Howie, V. M., Ploussard, J. H., Amman, A. J., Austrian, R., Johnston, R. B., Jr. Immune response to acute otitis media in children. I. Serotypes isolated and serum and middle ear fluid antibody in pneumococcal otitis media. Infec. Immun. 9: , 197.