Serum Antibodies to Pneumolysin in Patients with Pneumonia

Transcription

1 JOURNAL OF CLNCAL MCROBOLOGY, Jan. 1988, p /88/ $02.00/0 Copyright 1988, American Society for Microbiology Vol. 26, No. 1 Serum Antibodies to Pneumolysin in Patients with Pneumonia KRZYSZTOF KANCLERSK,12* SVEN BLOMQUST,3'4 MARTA GRANSTROM,' AND ROLAND MOLLBY' National Bacteriological Laboratory, S Stockholm,' Department of nfectious Diseases, Danderyd lospital, S Danderyd,3 and Vasteras Central Hospital, S Vasteras,4 Sweden, and National nstitute of Hygiene, Warsaw, Poland2 Received 15 June 1987/Accepted 18 September 1987 Serum antibodies to purified pneumolysin were determined by enzyme-linked immunosorbent assay (ELSA) in paired samples from 406 adult patients with community-acquired pneumonia and in samples from 184 healthy controls. A high sensitivity (83%) was obtained in patients with blood culture-confirmed pneumococcal pneumonia. n patients with a tentative pneumococcal diagnosis based on culture of samples from the sputum or the nasopharynx, 45% were positive by ELSA. The difference likely reflected the different relevance of cultural findings for the diagnosis of pneumococcal pneumonia. A significant rise in ELSA titer was found in 17% of the patients. When the diagnosis was also based on high titers, 25% were positive. Pneumococcal pneumonia diagnosed by the pneumolysin ELSA was significantly more common in the patients with a more severe disease and who required hospitalization (21 versus 5% for outpatients). Younger patients were more often positive for pneumococci as determined by high titers, while older patients showed titer rises. Mixed infections with other infectious agents were not uncommon. The finding of low titers in acute-phase samples from positive patients and in the youngest and oldest age groups of healthy controls were unexpected, indicating that further studies on the role of pneumolysin in pneumococcal disease are warranted. Pneumococcal infections are known to be common, but epidemiological studies have been hampered by limitations in the diagnostic methods (1, 2, 7). Diagnosis based on blood culture alone has a high specificity but low sensitivity and results in underestimates of incidence. Cultures from other sites may not yield representative samples because of the high rate of healthy carriers (5, 10, 11, 16, 21). Serological assays could therefore be of great help in the diagnosis of pneumococcal infections. Antibody response to type-specific capsular polysaccharides and to the C polysaccharide have been investigated, but the methods have disadvantages for routine use (4, 12, 17). n earlier studies (15, 18), we have shown that an enzymelinked immunosorbent assay (ELSA) with purified pneumococcal hemolysin (pneumolysin) as antigen is of use for the serological diagnosis of pneumococcal pneumonia. n the present study, this newly developed method was applied to the diagnosis of pneumococcal pneumonia in a group of mainly adult patients with community-acquired pneumonia. MATERALS AND METHODS Patients. Paired serum samples from 406 patients with community-acquired pneumonia were examined. The patients were seen as outpatients or admitted to the Department of nfectious Diseases, Danderyd Hospital, Danderyd, Sweden. The Department of nfectious Diseases serves a population of 720,000 people who live in the northern part of Stockholm County, Sweden. The study period was 15 months (December 1978 to February 1980). All patients over 6 years of age with respiratory symptoms were included in the study if they fulfilled the following criteria: (i) an infiltrate judged as pneumonic on chest X-ray; (ii) symptoms and signs of respiratory illness for less than 3 weeks; (iii) no hospitalization for the previous month; (iv) acute- and convalescent-phase sera available for serological testing. The acute-phase samples were drawn at the first visit or on * Corresponding author. admission (mean, 7 days, and range, 1 to 20 days, after the onset of disease), and the mean time for the convalescentphase sample was 3 weeks (range, 8 days to 12 weeks) after the first sample was obtained. The majority of patients were adults. Only 23 of 406 of the patients were children below 16 years of age. The mean age was 48 years (range, 9 to 98 years) for all patients (n = 406), 34 years (range, 9 to 79 years) for the outpatients (n = 99), and 53 years (range, 6 to 98 years) for the hospitalized patients (n = 307). The sex distribution was 49% males and 51% females. Healthy controls. Sera from 184 healthy individuals (age, 0 to 80 years) were examined as controls to find the upper limits of normal immunoglobulin G (gg) titers to pneumolysin in serum. Pneumolysin ELSA. Production and purification of pneumolysin, as well as the ELSA for the detection of gg antibodies to purified pneumolysin in serum, have been described previously in detail (9, 13, 15, 18). Briefly, highly purified pneumolysin, at a concentration of 2.5 ktg/ml, was used to coat cobalt-irradiated polystyrene microplates (M 129 B; Dynatech Laboratories, Plochingen, Federal Republic of Germany). A single point determination was made routinely at a serum dilution of 1:1,000, but higher or lower dilutions were also used when appropriate. Goat anti-human gg antisera conjugated to alkaline phosphatase (Sigma Chemical Co., St. Louis, Mo.) was used with p-nitrophenyl phosphate as the substrate. Volumes of 100,ul were used in each step, and all washings were done three times. The ELSA titer was defined as the A405 multiplied by the serum dilution factor. A significant rise in titers was defined as at least a twofold (100%) increase between the acute- and the convalescent-phase samples. The intra- and interassay variations were 10 to 15% and 15 to 20%, respectively. Serologic diagnosis of other infectious agents. Paired samples were tested by the complement fixation test for influenza virus types A and B, parainfluenza virus types 1 to 3, respiratory syncytial virus, adenovirus, Mycoplasma pneumoniae, and ornithosis and by the indirect immunofluores- 96

2 VOL. 26, 1988 DAGNOSS OF PNEUMOCOCCAL PNEUMONA BY ELSA 97 cent-antibody test for Legionella pneumophila. A significant rise in titers, i.e., at least a fourfold rise in titers by the complement fixation or the immunofluorescent-antibody test was found in 135 patients (see Table 2). Furthermore, 52 patients had high antibody titers to these antigens (see Table 3). Culture. A conventional culture of a sample from the nasopharynx was performed in all patients. From those who produced sputum, a culture to detect bacteria was performed as well. f the patient had not received antimicrobial therapy prior to the first visit as an outpatient or on admittance as an inpatient, one or three blood samples, respectively, were drawn for culture. Of 406 patients, 12 had pneumococcal bacteremia, while 56 had Streptococcus pneumoniae found in the nasopharynx, the sputum, or both (see Table 1). n addition, four patients were positive by immunoelectroosmophoresis for pneumococcal polysaccharide antigens in sputum. Haemophilus influenza was found in blood culture from 2 patients and in the sputum, the nasopharynx, or both from 50 patients. RESULTS Anti-pneumolysin titers in a healthy population. The occurrence of antibodies to pneumolysin was studied in 184 healthy individuals (age, 0 to 80 years) (Fig. 1). The lowest antibody levels were recorded in infants of less than 1 year of age and in adults older than 70 years. The highest titers were found in the group of children who were 8 to 15 years old. n sera from patients in these age groups, there were only a few (5 of 41) with titers lower than 200. The upper limit of normal values, set at the 95th percentile of the healthy controls, was 500 for children under 7 years of age, 700 for patients 8 to 50 years of age, and 500 for adults older than 50 years of age. The chosen 95% specificity for diagnosis by high titers allowed for 5% false-positive results by high titers in a patient material. Anti-pneumolysin titers in patients with pneumonia. Positive serology by determination of a significant rise in gg titers to pneumolysin was found in 70 of 406 (17%) patients with clinical pneumonia (Table 1). ncluding diagnosis by high titers, a total of 101 of 406 (25%) patients were positive. Patients with pneumococcal bacteremia were positive by pneumolysin ELSA in 10 of 12 (83%) cases (Table 1). Patients with S. pneumoniae in the nasopharynx, the spu s ' CD 100. * S W40 Age groups (years).rn s 8 WF B0 FG. 1. Age distribution of gg anti-pneumolysin titers in serum from healthy controls. Dots indicate actual values, and bars indicate mean values for each age group. 0 TABLE 1. Antibody response to pneumolysin in patients with clinical pneumonia Pneumococci n ELSA for pneumolysin with a titer: detected in: Blood 12 9 (75) 1 (8) 10 (83) Nasopharynx and (35) 6 (10) 27 (45) sputum Negative (12) 24 (7) 64 (19) Total (17) 31 (8) 101 (25) tum, or both were positive by serology in 27 of 60 (45%) cases. n addition, positive anti-pneumolysin titers were found in 64 of 334 (19%) patients that were negative on routine culture for pneumococci. Patients positive by both culture and a titer rise in ELSA received antibiotics prior to culture in 6 of 29 (21%) cases, while patients positive by a titer rise in ELSA but negative by culture or immunoelectroosmophoresis received antibiotics in 17 of 39 (44%) cases (P < 0.07, by the Fischer exact test). Antibody response to pneumolysin. (i) n relation to age. The distribution of the antibody response pattern used for diagnosis, i.e., rise of titers or high titers, as it correlates with age is shown in Fig. 2. n patients under 21 years of age, 24% were positive by high titers and 8% were positive by a rise in titers. n those above 30 years of age, a significant rise in titers became the dominant serologic finding. The rate of high titers in these patients in the older age groups was 5% or lower, with the exception of the patients of 41 to 50 years of age, in which 9% positive results by high titers was found. (ii) n relation to severity of disease. Of the 406 patients, 304 required hospitalization and 102 were treated as outpatients. Pneumococcal pneumonia based on a rise in titers (Table 1) was found in 65 of 304 (21%) hospitalized patients, and was significantly higher in this group (P < 0.001, by the chisquare test with the Yates correction) than in the 5 of 102 (5%) outpatients. Diagnosis of pneumococcal disease based on high titers showed an opposite distribution in the two groups, with 6 of 304 (2%) positive results in hospitalized tn c,3 30- w 20. c W N=1B / / f / N=31 l/ / ',- /p N= 66 _- l43 N- 52 Nz 59 N. 49 N= _ c i- l >00 Age group (years) FG. 2. Percentage of patients with rising and high antibody titers to pneumolysin in relation to age. Hatched bars indicate patients with high titers, and open bars indicate patients with rising titers. A significant rise in titers was at least a 100% increase between two samples, and high titers were those above the 95th percentile found in age-matched healthy controls.

3 98 KANCLERSK ET AL. patients and 14 of 102 (14%) positive results in outpatients (P < 0.01). Serologic evidence of mixed infection. A significant rise in titers to other infectious agents of the respiratory tract was obtained in 135 of 406 (33%) patients (Table 2). Rising titers to pneumolysin were seen in 18 patients, and high titers only were found in another 11 patients. A total of 29 (22%) patients were positive based on serology, and 8 of these 29 patients were also positive for S. pneumoniae based on culture. n an additional 52 patients, high titers to other infectious agents were suggestive of current or recent infection (Table 3). Among these patients, 10 had rising and 5 had high antibody titers to pneumolysin. Positive cultures of S. pneumoniae as well were found in 2 of 15 patients. Patients that were positive by blood culture and sputum and/or nasopharynx culture to H. influenza were positive by serology to pneumolysin in 6 of 52 (12%) cases (Table 4). Concomitant positive cultures of S. pneumoniae were obtained in two of six of these patients. Pneumolysin titers in early acute-phase samples. Acutephase samples drawn within 5 days after the onset of respiratory symptoms were available from 37 patients (median age, 63 years; range, 14 to 90 years), with a significant rise in titers. Titers above 200 were found in only nine (24%) of these patients. This rate was significantly lower than that (61 of 118 [52%]) found among all other patients who were negative by pneumococcal culture and ELSA (P < 0.01, by the chi-square test). The same significant difference was obtained in these patients in comparison with that obtained in 26 of 47 (5%) healthy age-matçhed controls. Also, very low titers (below 100) were found in as many as 11 of 37 (30%) patients with an ELSA titer rise, but in only 16 of 118 (14%) of the other patients and jn 4 of 47 (9%) of the control group (P < 0.05 for both comparisonss. DSCUSSON This is the first study in which ELSA has been used to detect antibodies to pneumolysin in a large prospective patient population with'community-acquired pneumonia. n this study, the sensitivity of the assay in cases confirmed by blood culture was 83%, which is similar to our previous results of 82% (15, 18). Patients with a presumptive diagnosis TABLE 2. Rise in titer of other infectious agents ELSA for pneumolysin nfectious agent n with a titer: nfluenza virus type A nfluenza virus type B il Parainfluenza virus types 1 to 3 Respiratory syncytial virus Adenovirus M. pneumoniae Ornithosis L. pneumophila Total (13) il (8) 29a (22) a Eight patients were also positive by culture for pneumococci, including one patient with S. pneumoniae in blood culture. TABLE 3. High titer of other infectious agents ELSA for pneumolysin nfectious agent n with a titer: nfluenza virus type A O O O O nfluenza virus type B Parainfluenza virus types 1 to 3 Respiratory syncytial virus Adenovirus O O O Q M. pneumoniae Ornithosis L. pneumophila Total (19) 5 (10) 15a (29) a Two patients were culture positive for pneumococci. of pneumococcal pneumonia based on culture of pneumococci from the nasopharynx, by culture or antigen detection in sputum, or both were positive by ELSA at a much lower rate (45%). This discrepancy in the rate of positive results by ELSA in cases verified by blood culture and in patients diagnosed by culture of S. pneumoniae from other sites has several explanations. One possibility is that bacteremic patients have a better antibody response than do nonbacteremic patients. This possibility cannot be excluded, but it seems less likely. Bacteremic patients often have a more severe clinical presentation and receive antimicrobial therapy rapidly after the onset of symptoms. Although early treatment with antimicrobial agents could even result in the absence of or a decreased antibody response, bacteremic patients were positive 83% of the time. A more likely explanations could reflect the often quoted difficulty of obtaining representative samples from patients with suspected pneumococcal pneumonia (2, 10, 14).`Blood culture remains the most (if not the only) reliable method; i.e., it has the highest specificity. Transtracheal aspirates are said to yield more representative samples than sputum (4). The relevance of the finding of pneumococci in cultures of nasopharyngeal swabs and sputum, i.e., its specificity for pneumococcal pneumonia, has not been established. The results from our two studies (15, 18) seem to indicate that culture from these sites has a low degree of specificity for the diagnosis of pneumococcal pneumonia. TABLE 4. Number of patients culture positive for H. influenza ELSA for pneumolysin H. influenzae with a titer: detected in: n Rising High Orishingh Blood 2 la 0 i Nasopharynx Sputum Nasopharynx and sputum Total 52 4 (8) 2 (4) 6b (12) a Pneumococci were also cultured from sputun. b Two patients were culture positive for pneumococci. J. CLN. MCROBOL.

4 VOL. 26, 1988 DAGNOSS OF PNEUMOCOCCAL PNEUMONA BY ELSA 99 The incidence rate of 17 to 25% pneumococcal disease by detection of pneumolysin by ELSA lies within the range of 11.5 to 76% in hospitalized adult patients given in earlier reports (1, 10, 19). Swedish studies have reported frequencies from 34 to 54% (3, 6). Several factors could influence the rate of pneumococcal pneumonia in samples from patients. Some studies were conducted in patients who acquired pneumonia in the hospital, while our study was entirely conducted in patients with community-acquired infections. The severity of the cases and the age group under study were also of importance. n the present study, one-fourth of the patients had a mild case of disease that did not require hospitalization, and the patients were younger (mean age, 48 years) than those studied previously. The epidemic situation for other agents that cause pneumonia could also influence the relative rate of pneumococcal disease. Our study was conducted during a year with an unusually high incidence of M. pneumoniae infections and with an unusually low rate of influenza virus type A activity. Mixed infections have been reported not to be rare (3, 7, 8, 14). Evidence of mixed infection was also found in the present study and in our earlier studies of patients with M. pneumoniae infection (15, 18). n the present study, the rate of patients positive for pneumolysin by ELSA for patients with M. pneumoniae infection was similar or slightly lower than that in the earlier studies (15, 18) (33 versus 22%, respectively). n the earlier studies (15, 18), the patients with viral infections were all negative by ELSA. The main difference seems to be the severity of the disease in the patient samples in the two studies. While all patients in the first studies (15, 18) with M. pneumoniae infection required hospitalization, only 53 of 120 (44%) required hospitalization in the present study. All patients with viral infections were not hospitalized in the earlier studies (15, 18), while 40 of 47 (85%) were hospitalized in the present study. The distribution of rising titers versus high titers in different age groups was an unexpected finding. Also, the finding that rising titers were significantly more common among the patients that required hospitalization was intriguing. The data could suggest that younger patients have more rapid antibody responses, and it is possible that they also start with higher antibody levels. The rapid antibody response could result in a milder clinical course of infection. Other observations also seem to raise the question of whether antibodies to pneumolysin could be of importance in the host defense against pneumococcal disease. The antibody pattern in the healthy controls showed very low titers in the youngest children and patients in the oldest age groups, who are most prone to develop pneumococcal disease. Also, the finding of very low titers in acute-phase samples of patients positive in the pneumolysin ELSA could indicate a relation to the immune status. Although these first observations in humans represent indirect indications, they imply that further studies on pneumolysin are warranted. To our knowledge, only one study has been done in experimental animals immunized with purified pneumolysin (20). The results indicated increased survival time after challenge with virulent pneumococci. n conclusion, diagnosis of pneumococcal disease in clinical samples could be greatly improved by the introduction of a serologic assay. The ELSA for pneumolysin, in which only one antibody class is used, could readily be used in routine diagnostic work. This assay showed a high sensitivity in cases of infection verified by blood culture. The lower sensitivity in pneumococcal disease tentatively diagnosed by culture from other sites is likely to be due to the lower specificity of these cultural findings. Some unexpected observations seem to indicate that antibodies to pneumolysin could have a more important role in host defenses than earlier suspected. ACKNOWLEDGMENTS Goran Sterner is gratefully acknowledged for stimulating discussions and support during this study. We also thank Eivor Norstrom and Maj Ringman for skillful technical assistance. This study was supported by grant 16X-2562 from the Swedish Medical Research Council and by grant from the Swedish Society for Medical Sciences. LTERATURE CTED 1. Austrian, R Some observations on the pneumococcus and on the current status of pneumococcal disease and its prevention. Rev. nfect. Dis. 3: Bassett-Connor, E The nonvalue of sputum culture in the diagnosis of pneumococcal pneumoniae. Am. Rev. Respir. Dis. 103: Berntsson, E., J. Blomberg, T. Lagergârd, and B. Trollfors Etiology of community-acquired pneumonia in patients requiring hospitalization. Eur. J. Clin. Microbiol. 4: Coonrod, J. D., and D. P. Drennan Pneumococcal pneumonia. Capsular polysaccharide antigenemia and antibody responses. Ann. ntern. Med. 84: Davidson, M., B. Tempest, and D. L. Palmer Bacteriologic diagnosis of acute pneumonia: comparison of sputum, transtracheal aspirates, and lung aspirates. J. Am. Med. Assoc. 235: Fransén, H Clinical and laboratory studies on the role of viruses, bacteria, Mycoplasma pneumoniae and Bedsonia in acute respiratory illness. Scand. J. nfect. Dis. Suppl Fransén, H., and G. Tunevail Bacteria and serologic reactions against bacteria in patients hospitalized with acute respiratory illnesses. Scand. J. nfect. Dis. 1: Fransén, H., and S. Wolontis nfections with viruses, Mycoplasma pneumoniae and bacteria in acute respiratory illness. Scand. J. nfect. Dis. 1: Granstrom, M.,. G. Julander, S.-A. Hedstrom, and R. Mollby Enzyme-linked immunosorbent assay for antibodies against teichoic acid in patients with staphylococcal infection. J. Clin. Microbiol. 17: Gray, B. M., G. M. Converse, and H. C. Dillon Epidemiologic studies of Streptococcus pneumoniae in infants: acquisition, carriage and infection during the 24 first months of life. J. nfect. Dis. 142: Hendley, J. O., M. A. Sande, P. M. Stewart, and J. M. Gwaltney Spread of Streptococcus pneumoniae in families.. Carriage rates and distribution of types. J. nfect. Dis. 132: Holmberg, H., A. Krook, and A.-M. Sjogren Determination of antibodies to pneumococcal C polysaccharide in patients with community-acquired pneumonia. J. Clin. Microbiol. 22: Julander,. G., M. Granstrom, S.-A. Hedstrom, and R. Mollby The role of antibodies against alpha-toxin and teichoic acid in the diagnosis of staphylococcal infection. nfection 11: Kaijser, B., E. Berntsson, and B. Trollfors Serological diagnosis of pneumococcal pneumonia. J. nfect. 1(Suppl. 2): Kalin, M., K. Kanclerski, M. Granstrom, and R. Moliby Diagnosis of pneumococcal pneumonia by enzyme-linked immunosorbent assay of antibodies to pneumococcal hemolysin (pneumolysin). J. Clin. Microbiol. 25: Kalin, M., and A. A. Lindberg Diagnosis of pneumococcal pneumonia: a comparison between microscopic examination and expectorate, antigen determination and cultural procedures. Scand. J. nfect. Dis. 15: Kalin, M., and A. A. Lindberg Antibody response against the type specific capsular polysaccharide in pneumococcal

5 100 KANCLERSK ET AL. pneumonia measured by enzyme linked immunosorbent assay. Scand. J. nfect. Dis. 17: Kanclerski, K., M. Granstrom, M. Kalin, and R. Mollby Enzyme-linked immunosorbent assay (ELSA) for detection of antibodies to purified pneumococcal hemolysin (pneumolysin) in patients with pneumonia, p n Y. Kimura, S. Kotami, and Y. Shiokawa (ed.), Recent advances in streptococci and streptococcal diseases. Redbooks Ltd., Berkshire, England. 19. MacFarlane, J. T., R. G. Finch, M. J. Ward, and A. D. MacRae. J. CLN. MCROBOL Hospital study of adult community-acquired pneumonia. Lancet ii: Paton, J. C., R. A. Lock, and D. J. Hansman Effect of immunization with pneumolysin on survival time of mice challenged with Streptococcus pneumoniae. nfect. mmun. 40: Rosén, C., P. Christensen, B. Hovelius, and K. Prellner A longitudinal study of the nasopharyngeal carriage of pneumococci as related to pneumococcal vaccination in children attending day-care centres. Acta Otolaryngol. Stockholm 98: