UMEÅ UNIVERSITY MEDICAL DISSERTATIONS New series No 365 ISSN 036-6612

From the Department of Infectious Diseases University of Umeå, Umeå, Sweden

Streptococcus pneumoniae

Epidemiological, clinical, and serological studies

AKADEMISK AVHANDLING som for avläggande av medicine doktorsexamen vid Umeå Universitet offentligen kommer att försvaras i Infektionsklinikens föreläsningssal,

fredagen den 7 maj 1993, kl 09.00

av

Lars Å Burman

STREPTO CO CCU S PNEUM ONIAE - EPID EM IO LO G IC A L, CLIN IC A L, AND SEROLOGICAL STUDIES

Lars Å Burman, Department of Infectious Diseases, University Hospital of Umeå, Sweden

ABSTRACT

A retrospective study of invasive pneumococcal disease in patients from Greater Göteborg in 1964- 1980 identified 125 cases of meningitis, 305 of pneumonia, 61 of septicemia with unknown focus, and 17 with other manifestations, all verified by cultures from normally sterile body fluids. The incidence was several times higher in infants and in the elderly than in any other age-group. A wide variety of underlying conditions were present in 23% of the infants, 34% of the children, and 81% of the adults. In adults alcoholism was known in one third of the cases. The case fatality rate was 24% among patients with underlying conditions and 9% among previously healthy individuals. The case fatality rate was 50% in patients with hospital-acquired infection.

Twohundred-fifteen pneumococcal strains, isolated from blood or CSF from 1971 to 1983 at the laboratories of clinical bacteriology of Göteborg, Malmö, and Umeå were serotyped by coagglutination (COA). Of all isolates, 89% belonged to serotypes represented in the 23-valent vaccine. In a separate study COA was compared with counterimmunoelectrophoresis (CIE). COA was found to have several advantages; rapidity, lower cost, and ability to disclose serotypes with neutral charge, which constituted 19% of all strains.

In a prospective study the etiology was determined in 196 hospitalized patients with pneumonia, most of them community-acquired. Culture of specimens from blood, transtracheal aspirate (TTA), sputum, and nasopharynx, assays of antigen in sputum, urine, and TTA, and assays of pneumococcal antibodies to capsular polysaccharide, C-polysaccharide, and pneumolysin in paired sera were performed. The etiology was established in 64% of the patients. Streptococcus pneumoniae was the most common agent (32%).

In a serological study of patients with pneumococcal infection, diagnosed by culture of CSF, TTA, or blood, IgG antibodies against C-polysaccharide and pneumolysin were determined by ELISA. The diagnostic sensitivity was only 51% and 60%, respectively.

In conclusion, invasive pneumococcal disease is strongly overrepresented at tender and high age and in patients with concomitant conditions, notably alcoholism. S. pneumoniae remains a predominant causative agent of community-acquired pneumonia in adults needing hospitalization. Due to the low sensitivity and/or specificity of individual microbiological techniques, a combined use of several techniques is necessary when trying to assess the relative importance of pneumococci and other agents in pneumonia. Extended use of the currently available pneumococcal vaccine and development of improved pneumococcal vaccines seem highly warranted.

Key w ords: Streptococcus pneumoniae, incidence, predisposing factors, prognosis, serotyping, etiology of pneumonia, diagnostic methods, COA, CIE, pneumococcal serology.

Umeå 1993 ISBN 91-7174-776-1

Streptococcus pneumoniae

Epidemiological, clinical, and serological studies

by

Lars Å Burman University of Umeå

Umeå 1993

UMEÅ UNIVERSITY MEDICAL DISSERTATIONS New series No 365 ISSN 036-6612

From the Department of Infectious Diseases University of Umeå, Umeå, Sweden

Printed in Sweden by Solfjädern Offset AB

Umeå 1993

ISBN 91-7174-776-1

The pneumococcus - still going strong

This thesis is based on the following papers, which vili be referred to in the text by Roman numerals:

I Burman LÅ, Norrby R, Trollfors B.Invasive pneumococcal infections: Incidence, predisposing factors, and prognosis.Rev Infect Dis 1985;7:133-42.

II Trollfors B, Burman LÅ, Dannetun E, Llompart J, Norrby R.Serotyping of Streptococcus pneumoniae strains by coagglutination and counterimmunoelectrophoresis.J Clin Microbiol 1983;18:978-80.

III Burman LÅ, Trollfors B, Norrby R, Falsen E, Haidl S, Henrichsen J.Serotype distribution of Streptococcus pneumoniae strains isolated from blood and cerebrospinal fluid in Sweden.Scand J Infect Dis 1986; 18:45-8.

IV Burman LÅ, Trollfors B, Möllby R.Antibody response to two pneumococcal antigens, C-polysaccharide and pneumolysin, in patients with invasive pneumococcal infections.In manuscript.

V Burman LÅ, Trollfors B, Andersson B, Henrichsen J, Juto P, Kallings I, Lagergård T, Möllby R, Norrby R.Diagnosis of pneumonia by cultures, bacterial and viral antigen detection tests, and serology with special reference to antibodies against pneumococcal antigens.JInfectDis 1991;163:1087-93.

STREPTOCOCCUS PNEUMONIAE - EPID EM IO LO G IC A L, CLIN ICAL, AND SEROLOGICAL STUDIES

Lars Å Burman, Department of Infectious Diseases, University Hospital of Umeå, Sweden

ABSTRACT

A retrospective study of invasive pneumococcal disease in patients from Greater Göteborg in 1964- 1980 identified 125 cases of meningitis, 305 of pneumonia, 61 of septicemia with unknown focus, and 17 with other manifestations, all verified by cultures from normally sterile body fluids. The incidence was several times higher in infants and in the elderly than in any other age-group. A wide variety of underlying conditions were present in 23% of the infants, 34% of the children, and 81% of the adults. In adults alcoholism was known in one third of the cases. The case fatality rate was 24% among patients with underlying conditions and 9% among previously healthy individuals. The case fatality rate was 50% in patients with hospital-acquired infection.

Twohundred-fifteen pneumococcal strains, isolated from blood or CSF from 1971 to 1983 at the laboratories of clinical bacteriology of Göteborg, Malmö, and Umeå were serotyped by coagglutination (COA). Of all isolates, 89% belonged to serotypes represented in the 23-valent vaccine. In a separate study COA was compared with counterimmunoelectrophoresis (CIE). COA was found to have several advantages; rapidity, lower cost, and ability to disclose serotypes with neutral charge, which constituted 19% of all strains.

In a prospective study the etiology was determined in 196 hospitalized patients with pneumonia, most of them community-acquired. Culture of specimens from blood, transtracheal aspirate (TTA), sputum, and nasopharynx, assays of antigen in sputum, urine, and TTA, and assays of pneumococcal antibodies to capsular polysaccharide, C-polysaccharide, and pneumolysin in paired sera were performed. The etiology was established in 64% of the patients. Streptococcus pneumoniae was the most common agent (32%).

In a serological study of patients with pneumococcal infection, diagnosed by culture of CSF, TTA, or blood, IgG antibodies against C-polysaccharide and pneumolysin were determined by ELISA. The diagnostic sensitivity was only 51% and 60%, respectively.

In conclusion, invasive pneumococcal disease is strongly overrepresented at tender and high age and in patients with concomitant conditions, notably alcoholism. S. pneumoniae remains a predominant causative agent of community-acquired pneumonia in adults needing hospitalization. Due to the low sensitivity and/or specificity of individual microbiological techniques, a combined use of several techniques is necessary when trying to assess the relative importance of pneumococci and other agents in pneumonia. Extended use of the currently available pneumococcal vaccine and development of improved pneumococcal vaccines seem highly warranted.

Key words: Streptococcus pneumoniae, incidence, predisposing factors, prognosis, serotyping, etiology of pneumonia, diagnostic methods, COA, CIE, pneumococcal serology.

Umeå 1993 ISBN 91-7174-776-1

CONTENTS

ABBREVIATIONS 8

INTRODUCTIONStreptococcus pneumoniae as a human pathogen 9Clinical manifestations of pneumococcal disease 11

AIMS 14

STUDY DESIGNS, PATIENTS AND METHODSInvasive pneumococcal infection 15Serotyping of pneumococci and serotype distribution 16Procedures used in etiological studies of pneumonia includingC-polysaccharide and pneumolysin serology 17

RESULTSInvasive pneumococcal infection 24Serotyping of pneumococci and serotype distribution 27Procedures used in etiological studies of pneumonia includingC-polysaccharide and pneumolysin serology 28

DISCUSSION 30

SUMMARY AND CONCLUSIONS 43

ACKNOWLEDGEMENTS 45

REFERENCES 46

PAPERS I-V

8

ABBREVIATIONS

CF Complement fixation

CIE Counterimmunoelectrophoresis

COA Coagglutination

CRP C-reactive protein

CSF Cerebrospinal fluid

ELISA Enzyme-linked immunosorbent assay

IFA Immunofluorescent antibody

PEEP Positive end-expiratory pressure

PID Passive immunodiffusion

TTA Transtracheal aspiration

9

INTRODUCTION

Discovery of the pneumococcus as an etiologic agent of pneumonia

In 1819 Laennec described physical signs and pathologic changes in pneumonia (66) and in 1875 gram-positive diplococci were seen by Klebs during microscopy of bronchial secretions from patients who recently had died in severe pneumonia.

In 1881, pneumococci were isolated by Pasteur (”Microbe septicémique de la salive”) and by Sternberg (Micrococcus pasteuri) from the blood of rabbits, which had previously been inoculated with saliva of patients with pneumonia (13). In 1886 Weichselbaum observed gram-positive diplococci in sections of lungs from patients with fatal pneumonia (171). Pneumonia was recognized as a dangerous disease with high mortality and in his famous textbook from 1901, Sir William Osier termed pneumonia ”captain of the men of death” (9). In 1917 Avery showed that pneumococci were responsible for 95% of all cases of lobar pneumonia in adults, a finding which was confirmed by Cecil in 1926 and Sutliff and Finland in 1933 (66).

Taxonomy

Due to its diploid form the organism was initially named Diplococcus pneumoniae. In the current nomenclature the name Streptococcus pneumoniae is used. The organism tends to grow in short chains, as best seen in cultures in poor media or in specimens directly obtained from sputum or different body fluids. Together with other streptococci they belong to the family Lactobacteriaceae.

Identification

Pneumococci are encapsulated, gram-positive, lancet-shaped cocci. In aging cultures they become gram-negative, due to autolytic enzymes. On blood agar plates pneumococci form round, glistening, unpigmented colonies within 24 to 36 hours, surrounded by hemolysis. When aging the centers of the colonies are prone to collapse because of autolysis. Pneumococci, which produce large amounts of capsular material, form mucoid colonies with a large diameter. The metabolism of pneumococci is dependent on anaerobic glycolysis. Lactic acid is produced, resulting in decreasing pH (173). Both pneumococci and alpha-streptococci are alpha-hemolytic and to discriminate between the two species the increased sensitivity of pneumococci to surface-active agents is utilized. In the routinely performed optochin test, paper discs with 5 microgram of ethylhydrocupreine hydrochloride are used to distinguish pneumococci from alpha-streptococci. It should be noted, however, that some exceptional pneumococcal strains are resistant to optochin. Another test for identifying pneumococci, based on the same principle, is the bile solubility test (130). Pneumococci, but not alpha-streptococci, are lysed by bile. This test is more laborious and is not routinely used.

10

Pneumococcal virulence factors and antigens

Outside the plasma membrane of the pneumococcus, there are two major surface components, the cell wall and the capsule. The cell wall is probably of major importance as inducer of the intense inflammatory response seen in invasive pneumococcal disease. The predominant cell wall component is the C-polysaccharide, which is common to all pneumococci. It is a ribitol teichoic acid containing

phosphorylcholine and galactosamine (86), and is covalently bound to the pepddoglycan of the cell wall. It is strongly antigenic and the phosphorylcholine residue is an immunodominant epitope. Antibodies to the pneumococcal phosphorylcholine may to some extent cross-react with antigens from bacteria in normal oro-intestinal flora, notably non-hemolytic streptococci, alpha-streptococci (113,180), Escherichia coli (67), Klebsiella species (68), and H. influenzae type b (38). The phosphorylcholine residue of the C-polysaccharide is also the component recognized by the C- reactive protein (CRP). CRP is an acute phase protein, which is produced by hepatic cells in response to interleukin-6 during inflammatory processes. Actually, the protein was named CRP due to the initial observation, that it binds to C-polysaccharide (160). The binding resulted in precipitation and differed from antigen-antibody reactions insofar as it required the presence of calcium ions. Although the CRP is known to activate complement (98) and to bind and opsonize various bacterial species including pneumococci (76,122,123), its role in the host defense has not yet been fully explained.

The invasive capacity of pneumococci is determined mainly by the bacterial capsule. In the absence of opsonizing antibodies, the pneumococcal capsule cannot be recognized by phagocytic cells and there is no other mechanism available to evade the infection. In host interaction, the pneumococcal species is favoured by a heterogeneity in capsular antigens among various strains. There are 84 known serologically distinguishable capsular serotypes. The antigenic type and the quantity of capsular polysaccharide are factors which determine the ability of the strain to escape recognition by humoral antibodies and complement and thus determines the invasive capacity of a given strain.

Already in the 1920s much effort was made to determine the carbohydrate structures of the capsule polysaccharide (15,173). Methods to immunize animals for production of antisera for serum therapy were introduced (8), which in many cases of severe pneumococcal pneumonia was life saving.

Besides the pneumococcal capsule, toxins are also important for invasion. Pneumolysin and neuraminidase are both sulfhydryl-activated hemolytic cytotoxins. Pneumolysin shows strong homology to streptolysin O of group A streptococci. It can generally be found in the cytoplasm of pneumococcal isolates. Purified pneumolysin has been found to inhibit the respiratory burst of phagocytes and the activation of the alternative complement pathway (128), and pneumolysin- defective pneumococci are less virulent for mice than are the isogenic parent strains (23).

11

Serotyping

The reference method for serotyping of S. pneumoniae is the capsular swelling test. This test was developed by Neufeld in 1902 (10,130). In the test, homologous antibodies render the capsule swollen as disclosed by microscopy. The capsular swelling test is costly since fairly large amounts of the expensive antisera are needed. It is only performed at a few reference laboratories in the world.

Counterimmunoelectrophoresis (CIE) (41) and coagglutination (COA) (105) depend on immunoprécipitation and agglutination, respectively. These methods are easy to perform, require minimal amounts of antisera and can be used for serotyping of capsulated bacteria and for detection of capsular polysaccharide antigens in body fluids.

Clinical manifestations of pneumococcal disease

Pneumococci are isolated from the respiratory tract of individuals with a wide variety of clinical manifestations but may also be isolated from healthy carriers. Asymptomatic carriage of pneumococci may occur in preschool children in a frequency as high as 21 to 35% (33,70,82). The carriage rate in healthy adults in Sweden is as low as 3% (104). The most important clinical manifestations of pneumococcal infection are pneumonia, septicemia, meningitis, otitis, and sinusitis. Arthritis and cellulitis are less common entities.

Pneumococci belong to the most common etiological agents of meningitis and septicemia in all age-groups and all populations. The incidence of invasive pneumococcal infections differs, however, from country to country and over time, largely due to socio-economic conditions (32,50,99). Serious pneumococcal infections are more common in the youngest and oldest age-groups.

In contrast to infection caused by some other capsulated bacteria, e.g. H. influenzae type b and meningococci (39,136), invasive pneumococcal infections are mainly seen in patients with underlying conditions (50,80,102,124,133). A wide variety of predisposing factors, e.g. alcoholism, immunological diseases, chronic pulmonary and cardiovascular diseases, and malignant diseases, have been described.

Modern intensive care of patients with meningitis and sepsis syndromes has improved the prognosis of severe pneumococcal disease. However, the case-fatality rate of invasive pneumococcal infections is still as high as 28% (124) as compared to 1-2% in meningitis caused by H. influenzae type b (39,162) and 5-10% in meningococcal infections (136). Austrian (8) studied the survival of patients with invasive pneumococcal pneumonia receiving different types of treatment Among those given only symptomatic therapy, 15% were alive at the end of the third week. Of those given type- specific antiserum, about 50% survived, compared to 83% of those given penicillin. Of those patients who died in spite of administration of penicillin, 60% died within 5 days. Death occurred early during the course of disease also among patients receiving immunotherapy and among untreated patients. Also in patients with no known risk factors, the clinical course of pneumococcal disease may be fulminant and beyond treatment when the patient arrives to hospital.

12

Diagnosis of pneumococcal pneumonia

The most common clinical manifestation of pneumococcal infection is pneumonia. Several studies have suggested that S. pneumoniae is the most common cause of pneumonia. However, the true incidence of pneumococcal pneumonia is difficult to establish, due both to a low sensitivity of various diagnostic procedures and to the fact that pneumococci may be part of the normal upper respiratory tract flora. Culture can be done from blood, sputum and nasopharyngeal secretion. Antigen detection test can be performed on sputum, serum and urine. Antibody determinations in paired serum samples can also be used for diagnosis.

Blood culture is insensitive, because septicemia is relatively uncommon. Not more than 25-30% of hospitalized patients with pneumococcal pneumonia have positive blood culture (92,132). Collecting sputum from a coughing patient is often difficult and there is risk for contamination from the upper respiratory tract flora. However, if purulent sputum is obtained, i.e. if more than 5 leucocytes per epithelial cell are found by microscoping homogenized sample, sputum culture tends to be informative (24,69,92,158). Reliable sampling methods for bronchial secretion are fiberoptic bronchoscopy (176), transtracheal aspiration (TTA) (138,143,144), and lung puncture (1). However, the first method requires expensive instruments and the other methods are unpleasant and may even be dangerous for the patient. Nasopharyngeal culture is easy to perform but is less valuable, particularly in children, due to the occurrence of pneumococci in the nasopharyngeal flora of healthy carriers.

Detection of pneumococcal antigen in sputum and urine is useful in the diagnosis of pneumococcal pneumonia and can be used even after initiation of antibiotic treatment. Two methods are available, CIE and agglutination tests, e.g. COA. To detect pneumococci irrespective of serotype, ”omniserum” containing antibodies against 83 capsular polysaccharides is used. For identification of the serotype of one given strain, various pooled antisera are used in a first step, then ”group”- or type-specific sera to identify the organism. CIE is less useful than COA when testing serotypes 7, 14, 33, and 37, because their capsular polysaccharides are neutral at pH 8.6 and do not migrate in an electrical field. COA is more rapid and easy to perform and has somewhat higher specificity and sensitivity than CIE (90). The sensitivity is high when used on purulent sputum but low when used on urine, also when the urine is concentrated (92).

The serum antibody response to capsular polysaccharides of pneumococci has been extensively studied both after vaccination and in natural disease. In diagnostic assays individual type-specific antigens or antigen mixtures of several capsular types (usually the 14-or 23-valent vaccines) have been used. For diagnostic purposes these methods have several shortcomings. Use of several individual antigens is costly. Use of antigen mixtures may lead to results which are difficult to interprete and infection with strains which are not included in the antigen mixture will remain undiagnosed. Furthermore, children younger than approximately two years have a poor antibody response to most polysaccharides (127).

Theoretically, determination of antibodies against C-polysaccharide and pneumolysin, two antigens which are common to all pneumococcal types, should be more suitable for diagnostic purposes. When the present studies were planned, there was little information on the use of these methods for diagnosing pneumococcal infection.

13

Reappraised interest in the diagnosis of pneumococcal pneumonia and other pneumococcal infections

In the pre-antibiotic era, very few patients with invasive pneumococcal infection survived. The only specific treatment available was immune serum, which could be given after serotyping of the pneumococcal strain isolated in each individual case. The introduction during the 1940s of sulphonamides and penicillin led to a drastic decrease in the case-fatality rate (8,55). Therefore, the interest in diagnostic procedures including serotyping of pneumococcal isolates faded. Gradually, however, it became clear that the pneumococci still require diagnostic work and that the procedures even have to be further developed. Since penicillin treatment is often instituted at an early stage, pneumococcal pneumonia may not develop as fulminant as in the old days and therefore becomes less easy to distinguish from pneumonia of other etiologies.

There is reason to believe that the incidence of pneumococcal pneumonia is presently underestimated. In a Swedish study of 205 patients with community-acquired pneumonia (92), pneumococci were suspected or proven as the etiological agent only in 93 (43%) of the patients, in spite of the use of culture specimens from several sites, assays of antigen in secretions, and direct microscopy. We found it worthwhile to perform further studies on the incidence of pneumococcal pneumonia with special efforts to evaluate a wide diagnostic arsenal for the purpose.

New data on the incidence, the underlying host factors, and the case-fatality rate of pneumococcal disease are required for several reasons. One is related to the current appearance of penicillin-resistant pneumococci in some parts of the world, a situation which merits special surveillance of the organism. The development of pneumococcal capsular polysaccharide vaccines is another reason for studying pneumococcal epidemiology. The present 23-valent vaccine remains to be evaluated for various groups of individuals and serotype distribution must be continously surveyed. An evaluation of the vaccine for extended use will require new data on incidence, underlying conditions, and case- fatality rate in pneumococcal disease. The expected development of conjugated pneumococcal vaccines, which presumably will be more effective than the pure polysaccharide vaccines, further stimulates research.

14

AIMS

1. To study the incidence of invasive pneumococcal infection in a Swedish region.

2. To study the age distribution and define risk groups for invasive pneumococcal infection.

3. To study the case-fatality rate in patients with invasive pneumococcal infection.

4. To compare two inexpensive methods for serotyping of Streptococcus pneumoniae; coagglutination and counterimmunoelectrophoresis.

5. To study serotype distribution of Streptococcus pneumoniae strains isolated from blood and cerebrospinal fluid in different parts of Sweden.

6. To study the serum antibody response to two pneumococcal antigens, C- polysaccharide and pneumolysin, in patients with invasive pneumococcal infection.

7. To study the role of pneumococci as etiological agents in adults hospitalized because of pneumonia.

8. To compare and evaluate different methods for diagnosing pneumococcal pneumonia.

15

STUDY DESIGNS, PATIENTS AND METHODS

Invasive pneumococcal infections: Incidence, predisposing factors, and prognosis(I)

In a retrospective analysis of case records from 494 patients with 508 episodes of invasive pneumococcal infection, incidence, predisposing factors, and prognosis were studied.

The patients were found in the files of the Department of Clinical Bacteriology in Göteborg, which covers the Greater Göteborg region, and in the diagnosis registers of all departments of internal medicine, pediatrics, and infectious diseases in the cities of Göteborg, Kungälv, and Mölndal, with surrounding communities. The population in the study area, (Göteborg, Härryda, Kungälv, Mölndal, Partilie, and Öckerö), increased during the 17-year period studied 1964-1980, from 551,636 to 573,564. A total of 524 patients, from whom S. pneumoniae had been isolated in blood, CSF, pleural fluid, or synovial fluid, were found. Eighteen patients not living in the defined area were excluded. The records of 12 patients were not found, but the data of sex, age, diagnosis, and home adress were known and these patients were therefore included in the study of incidence but not in the study of manifestations and outcome.

Table 1. Sources of pneumococcal isolates in normally sterile body fluids from 520 episodes of pneumococcal infections in 506 patients.

Bodv fluid No of episodesBlood 380Blood + CSF 65CSF 55Blood+ pleural fluid 5Pleural fluid 11Blood+ synovial fluid 2Synovial fluid 1Blood + peritoneal fluid 1Total 520

Definition of invasive pneumococcal disease.Meningitis: pneumococci were either isolated from CSF or isolated from blood, combined with clinical symptoms and CSF findings consistent with bacterial meningitis.Pneumonia: pneumococci were found in blood and/or pleural fluid, combined with a chest x-ray showing signs of pneumonia.Sinusitis: pneumococci were identified in blood together with clinical and roentgenological findings of sinusitis.Osteomyelitis: pneumococci were isolated from blood combined with roentgenological and isotope scanning findings.

16

Arthritis and cellulitis: pneumococci were isolated from blood in combination with clinical signs. Endocarditis: diagnosed at autopsy.Septicemia with unknown focus: pneumococci were isolated from blood but no focal infection could be found.

Serotyping of Streptococcus pneum oniae stra ins by coagglutination (COA) and counterimmunoelectrophoresis (CIE) (II)

To compare CEE and COA for serotyping of pneumococci with respect to concordance, rapidity and cost, 217 pneumococcal strains were used. Most of the strains were isolated from blood or CSF at the Departments of Clinical Bacteriology in Göteborg, Umeå and Malmö. Sixteen strains of uncommon serotypes were obtained from Statens Seruminstitut, Copenhagen.

The performance of CEE, COA, and passive immunodiffusion (PID) is described on pages 20-21.

Serotype distribution of Streptococcus pneumoniae strains isolated from blood and cerebrospinal fluid in Sweden (III)

To compare the current serotype distribution of pneumococci in Sweden with the serotypes included in the 23-valent vaccine, 215 pneumococcal strains from 214 patients were serotyped by COA. The pneumococcal strains were obtained from blood and/or cerebrospinal fluid, and the isolates were collected at the bacteriological laboratories in Göteborg during 1971 to 1980, and in Malmö and Umeå during 1982 and 1983.

Antisera were obtained from Statens Seruminstitut in Copenhagen, Denmark, at which laboratory typing of serotypes within groups was done. COA was performed as described on page 20.

The male/female ratio was 1.48:1. The age distribution and clinical manifestations of the patients are shown in table 2.

Table 2. Age distribution and clinical manifestations of 214 patients with invasive pneumococcal infection, from whom S. pneumoniae isolates were serotyped.

Clinical manifestation No. patients

Meningitis 40I Pneumonia 139

Septicemia with unknown focus 32Arthritis/Osteomyelitis 3

I___________________________

Age No. patients

<2 years 172-17 years 4

18-50 years 5951-65 years 65

>66 years 69

17

A ntibody response to two pneum ococcal an tigen s, C -p olysaccharide and pneumolysin, in patients with invasive pneumococcal infection (IV)

To study the IgG antibody response to two pneumococcal antigens, C-polysaccharide and pneumolysin, 2-3 sera were obtained from 43 patients with pneumococcal infections verified by cultures from blood, CSF, TTA, and synovial fluid. The male/female ratio was 25/18. Forty-two of the patients were adults or adolescents with a median age of 68 years, range 15-90 years, while one patient was an infant, 6 months old. The source of the pneumococcal infections and clinical manifestations are shown in table 3.

Table 3. Clinical manifestations and sources of isolates of S. pneumoniae in 43 patients with pneumococcal infections.

Clinical mani­festations

No of patients Sources of pneumococcal isolatesBlood CSF TTA Svnovial fluid

Pneumonia 34 26 9Pneumonia and arthritis 2 1 2Meningitis 6 5 6Peritonitis 1 1

Total 43 33 6 9 2

Twenty-seven patients had one or more chronic diseases: chronic lung disease (9 patients), chronic alcoholism (5), cardiovascular disease (4), diabetes mellitus (4), hematologic malignancy (4), autoimmune or immunologic disease (4), traumatic head injury (2), splenectomy (2), and nonhematologic malignancy (1). Of the 16 patients without chronic disease, 3 had laboratory-verified influenza or parainfluenza infection.

IgG antibodies against pneumolysin and C-polysaccharide were determined by ELISA. A 2-fold increase in antibodies was considered significant (pages 21-22).

Diagnosis of pneumonia by cultures, bacterial and viral antigen detection tests, and serology with special reference to antibodies against pneumococcal antigens (V)

Patients. We studied prospectively the etiology of pneumonia in 196 adult patients (>15 years old), admitted to the Department of Infectious Diseases in Umeå, during the period December 1, 1982 to November 30, 1984. The criteria for definition of pneumonia were fever >38^, clinical signs of respiratory infection, and a chest X-ray showing infiltrates, which within 6-8 weeks had diminished or disappeared. Six patients with pulmonary tuberculosis and 22 patients with other infiltrative pulmonary diseases were excluded as were 16 patients who fulfilled the inclusion criteria, but in whom the initial diagnosis was unclear and diagnostic procedures were not performed.

18

The male/female ratio was 102/94 and the median age was 68 years (range 16-94). In 181 (92%) cases the pneumonia was community-acquired and in 15 cases hospital-acquired. Ten patients died (median age 83 years). The most common underlying and predisposing conditions are shown in Table 4.

Table 4. Underlying conditions in 196 patients with pneumonia.

Underlying condition No. patients (%)

cardiovascular disease 39 (20)chronic obstructive lung disease 32 (16)chronic alcoholism 16 (8)malignant disease 15 (8)diabetes mellitus 15 (8)

smoker 83 (42)bird owner 8 (4)

Criteria for etiological diagnosis.A. Pneumococcal pneumonia: One of the following four criteria was required:1. Isolation of S. pneumoniae from blood.2. Isolation of S. pneumoniae from transtracheal aspirate.3. Significant increase in antibodies against at least one pneumococcal antigen in combination with isolation of S. pneumoniae from sputum or nasopharynx or detection of pneumococcal capsular antigen in sputum, transtracheal secretion, or urine.4. Significant increase in antibodies against both C-polysaccharide and pneumolysin.B. Other etiologies: Infections with Mycoplasma pneumoniae, legionella, and chlamydia were diagnosed by serology. Other bacteria were considered pathogens only if isolated from blood or transtracheal secretion. Virus infections were diagnosed by serology and also by culture or detection of viral antigen by immunofluorescence. Pneumocystis infection was detected at autopsy by microscopic examination of lung tissue. Diagnostic techniques used at admittance to hospital are shown in table 5.

Blood cultures were obtained by direct inoculation of 10-ml samples of venous blood divided in one aerobic and one anaerobic biphasic blood-culture bottle (Department of Clinical Bacteriology, Sundsvall, Sweden). Transtracheal aspirate was obtained as described by Schreiner 1972 (143). A needle was used to perforate the interstitium between the thyreoid and the cricothyreoid cartilage. Physiological saline (1.5-2 ml) was injected into the trachea and immediately aspirated when the patient coughed.

Table 5. Diagnostic techniques used at admittance.

Technique No.(%! of patients investigated

CulturesBlood 187 (95)Transtracheal secretion 61 (31)Sputum 138 (70)Nasopharynx 187 (95)

Coagglutination (capsular antigen detection!Sputum 100 (51)Transtracheal secretion 57 (29)Urine 177 (90)

VirologyVirus isolation 179 (91)Immunofluorescence (nasopharyngeal cells) 70 (36)

Nonspecific testsErythrocyte sedimentation rate 190 (97)C-reactive protein 164 (84)Peripheral white blood cell count 193 (98)

Fig.l. Practical performance of TTA. \ ' \

U

20

Nasopharyngeal specimens were obtained by perinasal swab for bacteriological culture and by nasal washings for virus isolation and viral antigen detection on exfoliated nasopharyngeal cells. Sputum specimens were obtained by trained physiotherapists and collected in sterile plastic vials. Serum samples were obtained on three occasions and frozen for later antibody determinations; acute phase, one week thereafter and about one month after the first sample.

Sputum, secretions obtained by TTA, and urine were frozen at -2QPC. The urine was concentrated 10-fold by ethanol precipitation (41).

Culture techniques and techniques used for demonstration of microbial antigen.Blood cultures were incubated at 37° C for 10 days and were examined daily for signs of growth.

Sputum specimens were transported to the laboratory and processed immediately or after storage at +4° C for a maximum of 18 hours (h). All samples were examined microscopically after Gram- staining of the most purulent part. If the leucocyte/squamous epithelial cell ratio was > 5/high power field (hpf), culture was done. If the ratio was < 5 (hpf) the sample was considered consisting mostly of saliva, and culture was not performed. Before culture, an equal volume of sputum and sterile 2% N-acetyl-L-cystein, (Sigma A-7250, Sigma Chemical Co) were mixed. The mixture was agitated on a Vortex mixer for 10 seconds, then allowed to stand at room temperature for 10 minutes, and finally Vortex-mixed for a further 15 seconds. The homogenized mixture was diluted 1/10 and 1/100 in a meat extract broth (Lab Lemco Powder, Oxoid). An inoculum of 0.1 ml of each dilution was spread on 5% blood agar and hematin agar. The plates were incubated overnight at 37° C in a moist atmosphere with 5% CO2 . Isolates of suspected pathogens were subcultured for definite identification if found in numbers of >105 cfu/ml in the culture. Identification was performed according to microbiological routine methods. TTA specimens were managed in the same manner as sputum specimens. Nasophaiyngeal specimens were cultured on 5% horse blood agar for aerobic, and on 5% horse blood and hematin- agar in moist air with 5% CO2 for anaerobic culture.

Virus isolation was done in roller tubes with cell cultures of green monkey kidney cells, A 549 cells (lung cancer cell), and human diploid fibroblasts. Virus antigen detection was performed on aspirated cells from nasopharynx by indirect immunofluorescence by use of polyclonal antiserum against influenza A and B virus, respiratory syncytial virus, and parainfluenza 1 and 3 virus.

Coagglutination was used to demonstrate pneumococcal antigen in sputum, TTA, and urine and for the serotyping of S. pneumoniae isolates. The coagglutination technique was based on the use of antibody-coated Staphylococcus aureus Cowan 1 (Pharmacia Diagnostics AB Uppsala, Sweden) as a reagent for agglutination of antigen-containing samples. S. aureus Cowan 1 exposes protein A, which binds strongly to the Fc portion of IgG without regard to antigen specificity.

All samples and pneumococcal strains were first tested against nine different COA reagents, each consisting of a pool of pneumococcal group- or type-specific antisera, together covering 83 serotypes. Then the sample or strain was tested against the individual sera of the reacting pool. Serotyping of strains within groups was performed at Statens Seruminstitut, Copenhagen, Denmark, by use of the Neufeld capsular swelling test (115,130).

21

Fig. 2. Principle of coagglutination by use of S. aureus.

Counterimmunoelectrophoresis (CIE) was performed in Veronal buffer (pH 8.6). Glass slides were coated with agarose gel, in which parallel columns of 40 wells, each with a volume of 12 pi, were punched. Ten colonies of each over night cultured pneumococcal strain was then diluted in 0.5 ml PBS and incubated for 2-6 h. A suspension of pneumococci was added to wells on the cathodic side. Pneumococcal antisera were diluted 1:2 and added to wells on the anionic side. The slides were placed on a plastic surface and cooled to 0° C, whereafter a current of 15 V/cm was applied for 15 minutes. The plates were then examined without staining (41).

For serotyping of pneumococci which could not be serotyped by CIE (II), passive immunodiffusion (PID) was used. PID was performed in low-ash agar layered on a glass slide. A hexagonal pattern of wells surrounding a central well was used, each of them containing 18 pi. Undiluted antiserum was added to the central well and suspensions of pneumococci were placed in the surrounding wells. Slides were examined after 24 h and 48 h.

Serplpgy,To evaluate the involvement of pneumococci in pneumonia as thoroughly as possible, we

determined antibodies against four different antigens, using ELISA.For determination of antibodies against pneumococcal capsular polysaccharides, polystyrene

sticks (NUNC, Roskilde, Denmark) were coated with type- or group-specific pneumococcal polysaccharides at 37° C for 4 h and stored overnight at 4° C. Human serum was diluted 1/100 with PBS, 0.5% bovine albumin, and 2.5% rabbit serum. Tween 20 (0.01%) was added to reduce the nonspecific adhesion of human immunoglobulin to the sticks. After incubation for 2 h at 37° C, the sticks were allowed to react with horseradish peroxidase-conjugated rabbit anti-human immunoglobulin for 2 h at 37° C and finally with a substrate consisting of 5-amino-2-hydroxybenzoic acid in PBS and hydrogen peroxide for 30 minutes at room temperature. A 1.3-fold increase in optical density between samples was considered significant (134). Antibodies against pneumococcal

22

capsular polysaccharides were only determined if pneumococci had been isolated or if pneumococcal antigen had been detected in a sample from the patient.

Antibodies against C-polysaccharide were determined by a similar procedure, although the C- polysaccharide was linked to the polystyrene stick surface by coating the sticks with the purified Ig fraction of the anti-pneumococcal C-polysaccharide serum. A 1.3-fold increase in optical density between samples was considered significant (134).

In study IV antibodies against C-polysaccharide were determined with a modified ELISA. A 2- fold increase was considered significant (78).

Antibodies against pneumolysin were determined by use of purified pneumolysin in phosphate- buffered saline (pH 7.4). Cobalt-irradiated polystyrene microtiter plates (Dynatech M 129 B, Plochingen, West Germany) were coated with the antigen overnight at room temperature. Sera from patients were diluted 1:1000 and added to the wells. For IgG and IgA antibody determinations, plates were incubated for one hour at room temperature, whereas IgM antibody determination were done after incubation for two hours at 37° C. Ig class-specific alkaline phosphatase-conjugated antisera (Orion Diagnostica, Oy, Helsinki, Finland) were added for incubation overnight at room temperature. After incubation with substrate, the ELISA value was defined as the absorbtion value at 405 nm multiplied by serum dilution (1000). A two-fold difference between two serum samples was considered significant (94).

Table 6. Methods to diagnose microorganisms other than pneumococci by antibody determination.

Organism Assay Significant increase (fold)

Reference

H. influenzae type b ELISA 4 (45)capsular polysaccharide(IgG, IgM)

L. pneumophila IFA 4* (131)

M. pneumoniae CF 4** (81)IgM ELISA (74)

C. psittaci CF 4** (81)Influenza A and B CF 4** (81)Parainfluenza 1 and 3 CF 4** (81)Respiratory syncytial virus CF 4** (81)Adenovirus CF 4** (81)

CF= complement fixation IFA= immunofluorescent antibody* and titer of > 128 in second serum sample ** and titer of >32 in second serum sample

23

Antibodies were also determined against phosphorylcholine (5), which is a component of the C- polysaccharide. Since these determinations did not correlate with other methods used to diagnose pneumococcal pneumonia, these determinations will not be further discussed.

H. influenzae type b serology was performed in sera from patients from whom H. influenzae had been isolated, irrespective of serotype.

24

RESULTS

Invasive pneumococcal infections: Incidence, predisposing factors, and prognosis(I)

Incidence and age-distribution: The incidence of pneumococcal meningitis remained virtually unchanged during the 17-year period studied; 1.2-1.4 cases/100,000/year. The documented incidence of non-meningitic invasive pneumococcal infections increased from 1.9 to 6.1/100,000/year, probably due to increased diagnostic efforts. Pneumococcal infection was seen in all age-groups. Four patients were neonates, aged between 6 hours and 14 days and the oldest patient was 94 years. The age-specific incidence of pneumococcal meningitis and other invasive infections during 1970- 1980 is shown in Table 7. (For details see tables 2-3 in paper I).

T able 7. Age-specific incidence (no. of cases per 100,000 inhabitants per year) of invasive pneumococcal infection in Göteborg, Sweden, 1970-1980.

Age (years)MeningitisIncidence

Other infections Incidence

All infections Incidence

0-1 12.0 13.8 25.82-9 0.7 1.6 2.310-19 0.4 1.6 2.020-29 0.6 1.8 2.430-39 0.9 3.5 4.440-49 1.6 5.0 6.650-59 1.4 6.2 7.660-69 1.2 10.4 11.670-79 2.2 14.2 16.4>80 1.9 12.2 14.1

Clinical manifestations: The most common clinical manifestation of invasive pneumococcal disease among infants, children, and adolescents, 0-17 years, was meningitis (44%), whereas pneumonia was the most common in adults (69%). Septicemia with unknown focus was seen in 27% of the children and 9% of the adults. Otitis, sinusitis, and cellulitis with concurrent septicemia were mostly documented in children (2.2% vs 0.4%), while arthritis and osteomyelitis were only found among adults (1.2%). Endocarditis, verified at autopsy, was found in 4 of the the adult patients with pneumococcal meningitis. Two of the children had facial cellulitis. (For details see table 4 in paper I). Sex distribution: In all age-groups men were more often affected than women. In adults the male:female ratio of invasive pneumococcal disease was 2.1:1. If the alcoholics were excluded the ratio was 1.5:1. Among alcoholics the ratio was 11.8:1. In infants 0-23 months the ratio was 2.1:1 and in children and adolescents (2-17 years) 1.9:1.

25

Underlying conditions: In 12 (23%) of 52 infants. 0-1 year of age, serious compromising conditions like malformations, hydrocephalus, allergic bronchitis, granulocytopenia, IgG2 deficiency, galactosemia, hepatopathia, or morbilli were known. In 5 (10%) of them the pneumococcal infection was hospital-acquired. There were 4 neonates, all with signs of meningitis. One had galactosemia, whereas the other 3 were full term and had no underlying condition.

In 11 (34%) of 32 children and adolescents. 2-17 years, congenital cardiac valvular malformation, allergic bronchitis, acute lymphoblastic leukemia, posttraumatic splenectomia, IgG2 deficiency, mental retardation and chronic pyelonephritis were seen. In 5 (16%) of the 32 children the pneumococcal infection was hospital-acquired.

In 332 of 411 (81%) adult patients one or more compromising conditions were documented while no risk factors were known in only 79 patients (Table 8).

The most common predisposing condition was alcoholism, present in 131 (32%) of the adult patients. These patients were heavy abusers and many of them had alcohol-related diseases such as liver cirrhosis, pancreatitis, epilepsia and duodenal ulcers. Hematologic malignancy, diabetes mellitus, and other systemic diseases were common. Eleven patients were splenectomized, 5 posttraumatically or accidentally during abdominal surgery, and 6 for medical reasons. Twenty-three episodes of pneumococcal infection occurred in patients treated with corticosteroids and 23 in patients treated with cytotoxic drugs. In 35 (9%) of the 411 adult patients the pneumococcal infection was hospital-acquired. Ten patients, one child and 9 adults had 2-4 episodes of invasive pneumococcal infection. (For details, see tables 5-7 in paper I).

Table 8. Underlying conditions in 411 adult patients with invasive pneumococcal infection.

Underlying conditionNo.of patients No.of patients

who died (%)Alcoholism (131 pat), drug abuse (2 pat.) 133 33 (24)Smoking 115 12 (10)Cardiovascular and arteriosclerotic disease 79 23 (29)Pulmonary disease (incl. lung tumor) 54 13 (24)Hematologic malignancy 29 7(19)Diabetes mellitus 22 6(27)Autoimmune and immunosuppr. disease 27 8(30)Malignant disease (excl. lung tumor) 7 2(29)Head injury and chronic otitis 23 6(23)Other conditions 10 1(10)Splenectomy 11 4 (31)

Patients with one or more risk factors 332 82 (24)Patients with no known risk factors 79 7 (9)All adult patients 411 89(21)

26

Case fatality rate. The overall fatality rate in invasive pneumococcal disease was 20% (100 of 508 episodes). The case fatality rate was unchanged during the observation period and similar in all age- groups. Twenty-five (25%) patients of 100 died within 24 hours after the culture had been obtained and 67 (67%) patients died within one week.

The case fatality rate for patients with pneumococcal meningitis, pneumonia, and septicemia with unknown focus was 33%, 15%, and 21%, respectively. None of the 15 patients with otitis, sinusitis, arthritis, osteomyelitis, or cellulitis died.

Among adults the case fatality rate was 24% in patients with underlying conditions and 9% in patients without such conditions (82/263 vs 7/72, pcO.Ol, x^test). In patients without underlying conditions the case fatality rate in the age interval 18-65 years was 2%, whereas the rate among patients older than 65 was 27%, (1/57 vs 6/22, p<0.01, Fisher's exact-test).

In 52 infants, 0-1 year old, 3 (25%) out of 12 with underlying condition died, whereas 5 (13%) of 40 without risk factor died. In 31 children, 2-17 years old, 3 (30%) of 10 with underlying conditions died compared with no death among the 21 children and adolescents without known risk factor.

Of 20 patients, noi treated with antibiotics, 14 had died within hours after admission, and 6 patients had febrile infections initially thought to be of viral origin. When results of culture were known, they had already recovered and left the hospital.

Table 9. Number of patients with, and without predisposing conditions and case fatality rate (%) among patients with invasive pneumococcal pneumonia/empyema, (unpublished data derived from the raw material of paper I).

ageyears

withpredisposing conditions

withoutpredisposing conditions

total

0 - 1 1/4 (25) 0/1 1/5 (20)2 -17 1/5 (20) 0/5 1/10 (10)18-50 9/64(14) 1/48 (2) 10/112(9)51-65 13/65 (20) 1/19 (5) 14/84 (17)

S66 16/78 (21) 4/16 (25) 20/94 (21)

Total 40/216 (19) 6/89 (7) 46/305 (15)

Among 91 alcoholic patients with pneumococcal pneumonia 16 (18%) died, and among 35 adult patients with vascular diseases in heart and CNS, 12 (34%) died. Eleven adult patients with pneumococcal pneumonia had malignant hematologic diseases; 7 with chronic lymphatic leukemia and 4 myeloma, of whom one (9%) died.

27

S e r o ty p i n g of S t rep tococcus p n e u m o n ia e by coa gg lu t ina t ion and coun te r - immunoelectrophoresis (II)

1. Concordance. There was a complete agreement between COA and CIE performed by us and the capsular swelling test performed at Statens Seruminstitut, Copenhagen, in 217 pneumococcal strains. Forty strains could not be identified by CIE performed at pH 8.6. All these strains were successfully serotyped by passive immunodiffusion (PID) and COA, and were shown to belong to serotypes with neutral charge, namely 7F, 7A, 14, 33F, 33A, and 37.2. Cost. Each COA-test was performed within 30-60 seconds, whereas 15 minutes were required for

CIE. Materials used for COA were cheaper than materials used for CIE. For instance, 5 jil of type- or groupspecific antisera was enough for 130 COA-tests but only for one CIE-test.

Serotype d i s t r ibu t ion of Streptococcus pneumoniae i solated from blood and cerebrospinal fluid in Sweden (III)

The most common serotypes were type 3 (12%), type 4 (10%), type 14 (10%), type 7F (9%), and type 23F (7%). Serotypes 7F, 14, and 33F, which cannot be detected by CIE, together constituted 19 % of all strains.

Table 10. Serotype distribution of pneumococci isolated from patients with invasive disease in relation to the composition of the 23-valent vaccine.

Serotypes No of isolates(%) No of fatal cases

Serotypes includedin the 23-valent vaccine 191(89) 22Serotypes not included in but related totypes of the 23-valent vaccine 9(4) 0Serotypes neither related to norincluded in the 23-valent vaccine 15(7) 1

All serotypes 215(100) 23

No difference in serotype distribution was found related to age-group, year (1971-80 compared with 1982-83) or place of isolation (Göteborg, Malmö, or Umeå).

Overall, 191 strains were covered by the 23-valent vaccine (89%). Moreover, 9 of the remaining 24 strains belonged to serotypes related to types of the 23-valent vaccine. Only 15 strains had serotypes not immunologically related to vaccine types.

Thus, 200 of the 215 strains (93%) had a capsular polysaccharide antigenically identical with or related to those of the currently available 23-valent vaccine.

28

A ntibody response to two pneum ococcal an tigen s, C -polysaccharide and pneumolysin, in patients with invasive pneumococcal infection (IV)

IgG antibodies against pneumolysin and C-polysaccharide were determined in sera from a total of 43 patients with invasive pneumococcal infection. Twenty-six (60%) of the patients showed significant increase in antibodies against pneumolysin compared to 22 (51%) against C- polysaccharide in the convalescent serum obtained 1-2 months after the acute serum. In 12 patients (28%) no increase against any antigen was detected.

Diagnosis of pneumonia by cultures, bacterial and viral antigen detection tests, and serology with special reference to antibodies against pneumococcal antigens (V)

Pneumococcal pneumoniaPneumococcal pneumonia was diagnosed in 63 (32%) of 196 patients fulfilling one or more of the following criteria:1. Isolation of S. pneumoniae from blood.2. Isolation of S. pneumoniae from transtracheal aspirate.3. Significant increase in antibodies against at least one pneumococcal antigen in combination with isolation of S. pneumoniae from sputum or nasopharynx or detection of pneumococcal capsular antigen in sputum, transtracheal secretion, or urine.4. Significant increase in antibodies against both C-polysaccharide and pneumolysin.The sensitivity of the diagnostic methods used are shown in table 11.

Table 11. Sensitivity of the diagnostic methods for pneumococcal pneumonia.

Diagnostic method no pos/no tests (%)

CultureBloodTranstracheal aspirate Sputum Nasopharynx

Coagglutination Sputum UrineTranstracheal aspirate

Serology Capsular polysaccharide C-polysaccharide Pneumolysin

13/62 (21)18/25 (72)

25/48 (52)22/62 (35)

27/34 (79)7/60 (12)

14/24 (58)

40/41 (98)55/62 (89)28/59 (47)

29

Antibiotics had been given to 58 (30%) of the 196 patients before inclusion in the study. Of 70 patients in whom no etiologic agent was demonstrated, 28 (40%) had been treated before admittance. In 126 patients, where one or more agents were demonstrated or suspected, 30 (24%) had received antibiotics before admittance (28/70 vs 30/126, p<0.05, x2-test).

Apart from the 63 patients fulfilling the criteria for pneumococcal etiology, another 71 patients had significant increases in C-polysaccharide antibodies. These patients, however, had no other test positive for pneumococcal infection. Among these 71 patients, 31 (43%) were treated with antibiotics before admission. Three further patients had positive sputum culture or nasopharynx culture as only evidence of pneumococcal infection. Altogether 74 patients who did not fulfil our criteria for pneumococcal infection had one test indicating pneumococcal infection.

Other agentsNoncapsulated H. influenzae was the etiological agent in 9 patients (5%), M. pneumoniae in 17

patients (9%), C. psittaci in 6 patients (3%), M. catarrhalis in 3 patients, S. aureus in 3 patients, K. pneumoniae in one patient and a proteus species in one patient

Four patients (2%) had legionella infection and one patient had Pneumocystis carinii infection. Viral infections were found in 42 patients (21%).

Altogether, the etiology was determined in 64%, of which pneumococci (32%) were by far the most common. If the 74 patients with only one indication of pneumococcal infection (nasopharynx or sputum culture, or with positive C-polysaccharide serology) were included, the total number of patients with pneumococcal pneumonia would be 137 (70%). Table 12.

Table 12. Cause of pneumonia in 196 adults.

Agents no (%)

S. pneumoniae alone 46 (23)S. pneumoniae + another agent 17 ( 9)Agents other than S. pneumoniae 63 (32)Unknown etiology 70 (36)

Total 196 (100)

S. pneumoniae indicated although not provenaccording to criteria used 74 (38)

30

DISCUSSION

Epidemiology of invasive pneumococcal infections (I)

Streptococcus pneumoniae is one of the most important human pathogens in all age-groups all over the world. This is the reason for the intense research efforts to develop vaccines and new antimicrobial agents against this organism, efforts which have become even more important with the emergence of penicillin-resistant pneumococci and the improved survival of patients with immunological disorders who are highly susceptible to pneumococcal infections. In order to enable an optimal use of the currently available and future pneumococcal vaccines and the initiation of an optimal antimicrobial chemotherapy it is necessary to study the epidemiology, including incidence, risk factors, and age-distribution of pneumococcal infections.

The true incidence of pneumococcal infections is not possible to determine with currently available diagnostic methods. In particular there are difficulties in diagnosing noninvasive infections, since pneumococci can be part of the normal respiratory tract flora. The incidence of invasive infections is easier to determine since isolation of pneumococci from a normally sterile body fluid, e.g. blood, CSF, pleural fluid, and synovial fluid is proof of pneumococcal etiology. However, many febrile patients, in particular children, are not subjected to blood cultures and pleural puncture is often omitted in patients with small exudates. Another confounding factor is the frequent use of antibiotics, which, when given to a septic patient before culture, can result in negative blood cultures. The most reliable incidence figures probably come from studies of meningitis, since lumbar puncture with culture of CSF is almost always performed in patients with meningitis and suspected meningitis in industrialized countries. The fact that the incidence of pneumococcal meningitis was unchanged during the 17-year period study (I), while the documented incidence of non-meningitic infections increased, probably reflects that blood cultures were more often obtained during the late 1970s than previously, while CSF cultures were regularly performed already during the 1960s.

In our study (I) the incidence of pneumococcal meningitis was 1.2 to 1.4/100,000 per year, which was very similar to the incidence reported from the USA and the Netherlands (57,58,151). In Dakar, Senegal, the incidence of pneumococcal meningitis was 11/100,000 per year (36), indicating that in developing countries the incidence is much higher than in industrial countries.

The overall incidence of non-meningitic invasive pneumococcal infection (I) was 1.9 to 6.1/100,000, which is lower than incidence rates found in prospective studies. Broome (30) reported an incidence of 9.5/100,000 in USA 1980. In a study from Charlestone County in 1987 Breiman (27) found the incidence to be 18.7, which was more than 2-fold higher than the incidence estimated by Filice (50) 10 years earlier in the same area. The increase was most pronounced among young children, patients of old age, and black people.

The difference between various regions in incidence of pneumococcal disease may to some extent be explained by socioeconomic factors, but also by genetic factors and age distribution in the population. Heffron (66) has reviewed evidence suggesting that blacks are more susceptible to pneumococcal infection than whites, a relation which was also seen in a study limited to children (102). Fraser (57) accordingly noted that pneumococcal meningitis was more common among blacks

31

than in whites. In contrast, Teele (154) found that black children had significantly fewer episodes than whites of otitis media, a disease where pneumococci predominate. The latter finding was tentatively suggested to be due to a different position of the bony eustachian tube in the two racial groups. Due to the difference in incidence of pneumococcal disease among geographic regions, results from one part of the world cannot easily be extrapolated to another pan.

The age distribution of invasive pneumococcal infections does not vary much from one study to another. The most susceptible age-group is children between 3 months and 2 years (1,27,62).

The incidence among children 0-2 years (25.8) was more than ten-fold compared with children and adolescents aged 2-17 years (I). The reason for this age-dependence is only partly understood. Underlying conditions per se seem not to be decisive. Among children 0-2 years, 12/52 (23%) had underlying conditions as compared to 10/31 (32%) among children and adolescents aged 2-17 years (I). In other geographical regions, underlying conditions may be more important in neonates and infants for acquisition of invasive pneumococcal disease. In a study from the USA, sickle cell disease was reported to be an important factor (83).

One important reason for the high incidence of invasive pneumococcal disease in children <2 years old is probably the lack of specific opsonizing antibodies at this age. In infants younger than 18-24 months there is a poor response to the polysaccharides of most pneumococcal serotypes except type 3 (4 4 ,4 7 ). The relatively low incidence of pneumococcal infection among newborns as compared to infants 3 months to 2 years old (101,145,155), can be explained by the presence of protective maternal IgG antibodies. However, the neonatal period is not completely free from severe pneumococcal disease. We recorded 4 cases at the age of 6 hours and 4, 7, and 14 days, respectively (I). Thus, severe infections with gram-positive bacteria during the neonatal period are not limited to group B streptococci but may in a few cases be caused by pneumococci. There is in general a relative immaturity of the host defence in neonates with a lower concentration of complement factors and a lower capacity to direct neutrophils to an infected site (175). Moreover, only IgG but not IgM antibodies are passively transferred from the mother. After the age of 2 years, young individuals develop antibodies against the pneumococcal capsular polysaccharides by repeated exposure to pneumococci and possibly to bacteria with cross-reactive antigens. Gold-miners in South Africa have a high risk of developing pneumococcal pneumonia as young adults, when coming to the mines from rural districts. However, after 6 months, the risk to develop pneumonia falls dramatically (151). This study and other studies indicate that most infections occur during the first week of carriage and that immunity follows carriership of pneumococci after some time (52,64,126).

In adults the incidence of invasive pneumococcal disease increases markedly after the age of 60 years, which is somewhat surprising, since this age-group should have high antibody levels. We found the incidence of invasive pneumococcal disease in aged patients to be 11.6 to 16.4 (I), which is in accordance with other studies (27,50). At high age, various diseases and other factors such as smoking habits may favour the invasion of pneumococci. The neutrophils, which are the crucial effector cells in pneumococcal infection, do, however, not seem to be debilitated by high age (43).

Pneumococci colonize the nasopharynx and this is the natural source of infection in pneumococcal disease. A study of military recruits (75) led to the introduction of the following simplified formula to predict the rate of pneumococcal pneumonia: Pneumonia rate = pneumococcus carrier rate x nonbacterial respiratory disease rate x K, where K varies with the serotype of the pneumococcus

32

and the natural immunity of the individuals in the population.In a study from 1932, Gundel (63) found that colonization of pneumococci could be seen as early

as some hours after birth in newborns, and a carrier rate as high as 59% was found at the age of 12 days. In a study from 1980, Gray (60) reported an onset of carriership of pneumococci at a mean age of 6 months and nearly all the children were colonized with pneumococci on at least one occasion during the first two years of life. Hendley (70) found upper respiratory carrier rates as high as 35% in preschool children and 18% in their parents, whereas adults without preschool children in the family showed only 2-9% carriage rate. Thus, pneumococci may occur in the nasopharyngeal tract at least from a few months of age, and if the patient is susceptible, may cause pneumonia or other infections.

The distribution of clinical manifestations of invasive pneumococcal infection shows a wide variation among different studies (Table 13).There is also a difference in distribution between children and adults (Tables 13, 14). We found (I) meningitis to be a more common manifestation among children than adults (44% versus 21%), and especially in children in the age group 0-2 years (48%). As many as 33% of infants with invasive pneumococcal disease had bacteremia with unknown focus; second to meningitis this was the most common manifestation in children of low age. In adults bacteremia without known focus was seen in only 9%.

In study I, pneumonia (68%) was found to be overrepresented in adults as compared to children, which is in accordance with results of other studies (Table 14). A difference in distribution of clinical manifestations between children and adults may partly be explained by a difference in tendency to perform diagnostic tests. Even venipuncture for blood culture can be felt as traumatic and in a less threatening disease, the medical staff is certainly more reluctant to draw blood for culture in children than in adults. It seems probable, therefore, that blood culture in pneumonia and non-focal infection may be omitted in children, rather than CSF culture in meningitis (153). However, a difference between children and adults in the frequency of pneumococcal pneumonia may also depend on extensive and prolonged exposure of the adult lung to smoke and air pollution, followed by decreased local resistance in the airway epithelium.

It is well-known that pneumococci predominantly attack individuals with conditions predisposing to infection. Apart from tender and advanced age a wide variety of underlying conditions have been described which increase the susceptibility to pneumococcal infections. Of all these predisposing conditions alcoholism seems to be the most important in Sweden.

In study I we found one or more underlying conditions in 332(81%) of the 411 adult patients, which is comparable with the findings of other authors (37,50,80,124). I will here focus on the predominant underlying condition, alcoholism, noted in 131 patients with 135(27%) episodes of pneumococcal bacteremia. Alcoholism was known in 91(31%) of all 290 adult patients with blood- culture positive pneumonia. Of these 91 patients, 16(18%) died (Table 15). The death rate was especially high (29%) in patients 51-65 years old.

33

T a b le 13. Invasive pneumococcal infection in infants and children.

Author (ref.) Year Country No. of pat. PneumoniaNo (%)

MeningitisNo (%)

Bacteremia with unknown focus

No (%)

Otitis Sinusitis Other localized foci

No (%)

Burke (35) 1971 USA 111 49(44) 2(2) 16(14) 38(34)Jacobs (83) 1979 USA 305 107(35) 70(23) 100 (33) 28(9)McIntyre (120) 1983 England 29 8(27) 2(7) 17 (59)Burman (I) 1985 Sweden 84 15(18) 37(44) 23 (27) 7(8)Siegman-Igra(149) 1986 Israel 50 19 (38) 10 (20) 18 (36) 2(4)Gray (61) 1986 USA 1310 307(23) 44(3) 172(13) 787(60)Davidson (46) 1989 Alaska 86 47(55) 12(14) 8(9) 19(22)Jette'(87) 1989 Canada 192 43(22) 29(15) 46(24) 74(39)Breiman (27) 1990 USA 47 11(23) 3(6) 13(28) 20(43)

Table 14. Invasive pneumococcal infection in adults.

Author (ref.) Year Country No. of. pat

PneumoniaNo (%)

Meningitis No (%)

Bacteremia with unknown focus

No (%)

Otitis Sinusitis Other localized foci

No (%)

Jansson (85) 1971 Finland 110* 92(84) 16(14) 1(1) KDFilice (50) 1980 USA 48 36(75) 3(6) 6(12) 2(4)Finkelstein (51) 1980 USA 187 85(46) 23(12) 22(12) 5(3)Burman (I) 1985 Sweden 424 290(68) 88 (21) 38(9) 8(2)Davidson (46) 1989 Alaska 28 22(79) 2(7) 2(7) 2(7)Jette'(87) 1989 Canada 276 212(77) 20(7) 14(5) 28(10)Breiman (27) 1990 USA 56 45(80) 0 11(20) 0

* three of the patients were younger than 15 years.

34

Table 15. Invasive pneumococcal pneumonia among alcoholics and non-alcoholics in different age- groups (unpublished data derived from the raw material of paper I).

AgeYears

No of patients with/ without alcoholism

No of fatal cases among alcoholics/non-alcoholics

18-50 52/60 7/351-65 28/56 8/6

>66 11/83 1/19all 91/199 16/28

In a Swedish study (181) of 279 patients (285 episodes) with bacteremic pneumococcal pneumonia, Örtquist found alcoholism among 33% of the patients, with a similar share of fatal cases in different age-groups as we found (I). Similar results were also reported in a Finnish study (85). In our prospective study (V) of hospitalized patients with pneumonia, 13 patients had invasive pneumococcal pneumonia, of whom 4 were alcoholics, all of them males (Table 16). Thus, there is no doubt that abusers of alcohol are prone to develop bacteremic pneumococcal pneumonia.

The liability of alcoholics to contract severe pneumococcal infection certainly has an influence on the sex ratio of the disease. In a study (181) of bacteremic pneumococcal pneumonia the male/female ratio was 9:1 among alcoholics, which is in accordance with the ratio of 11:1 found by us, (unpublished data derived from raw data of paper I). The male/female ratio of all invasive pneumococcal infections in study I was 2.1:1. If the alcoholics, who were all heavy abusers, were excluded, the ratio was 1.5:1 (I), which is in agreement with other studies (8,54).

However, male sex in itself predisposes to pneumococcal infections, since these infections are more common in boys than in girls (1,26,30,35,83).

Table 16. Alcoholic/non-alcoholic patients hospitalized for pneumonia (unpublished data derived from raw material of paper V).

No of patients with/ % alcoholics_______________________________ without alcoholism____________________with non-invasive pneumococcalpneumonia 7/43 14with invasive pneumococcalpneumonia 4/9 31

patients with pneumoniairrespective of etiology 16/180 8

An unfavourable influence of alcoholism on the outcome of pneumococcal disease has not been a consistent finding. In two studies of invasive pneumococcal pneumonia, comparing alcoholics with

35

non-alcoholics, no significant difference was found in rate of bacteremia or death (159,169). However, alcoholics more often had prolonged fever, slow resolution of infiltrates, and empyema (169). Austrian (8) in a study of 529 cases of bacteremic pneumococcal pneumonia, pointed out that the presence of alcoholism did not affect the prognosis unless accompanied by leukopenia or liver failure.

The case fatality rate among patients with invasive pneumococcal disease without known risk factors was 9%, whereas the rate was 24% among patients with risk factors (I).The highest case fatality rate (30%) was seen in patients with autoimmune disease, cancer, and splenectomized patients, as well as patients with cardiovascular and arteriosclerotic diseases. In contrast to study I, a study by Bruyn (34) did not show splenectomy to be associated with mortality more often than other predisposing factors. However, in that study the reason for splenectomy and data on pneumococcal vaccination were not given. Örtquist (181) reported that low serum albumin and absence of leucocytosis were important factors connected with high fatality rate. S. pneumoniae has been described as a nosocomial pathogen in elderly patients and in cancer patients (2,18,20), with a significantly higher mortality (73.9%) than in community-acquired infections (45.4%). This was in accordance with study I, where 44 patients with hospital-acquired pneumococcal infection had 50% mortality compared with an overall mortality of 20%. Thus, underlying conditions do indeed contribute to the very high mortality in invasive pneumococcal disease.

Pneumococci may have a similar potential to cause fulminant life-threatening disease as primarily recognized in meningococcal and other gram-negative septicemia. In some cases, the course may even be hyperacute with rapidly developing high fever, chills, and septic shock. Hence, of the 67 patients who died within the first week of admission, 25 died already within one day after the culture specimen had been obtained (Fig. 3) (I). However, the mechanism behind the induction of septic shock is at least partly different Gram-negative bacteria release endotoxin, which causes a dramatic inflammatory response in the host. Gram-positive bacteria such as pneumococci lack endotoxin. The predominant virulence factor of the pneumococci is the bacterial capsule, which evades recognition by the humoral and cellular defense of the susceptible patient. The pneumococci may therefore rapidly grow into excessive numbers resulting in a septic shock. The fact that many patients with invasive pneumococcal infections are beyond treatment when arriving to hospital, clearly demonstrates that effective prophylaxis should be more important than improved therapy.

Study I and other studies (17,164) show that the impact of pneumococcal infections in Sweden is considerably higher than of H. influenzae typ b and meningococci. The incidence of H. influenzae meningitis is about 2/100,000/year, which is also the combined incidence of meningococcal meningitis and septicemia. This is higher than the incidence of pneumococcal meningitis, but since S. pneumoniae is a very important cause of septicemia and pneumonia, the total incidence of pneumococcal infections is much higher than of the two other organisms. An important difference in the epidemiology of these three organisms is that pneumococcal infections principally occur in patients with risk factors, while infections with the other two capsulated microorganisms are mainly seen in previously healthy individuals.

36

Figure 3. Number of patients dying in the first week after culture (I).

No

25

12

Day 0 1 2 3 4 5 6

Streptococcus pneumoniae, serotypes and serotyping (II,III)

Pneumococci are divided in types based on the antigenic composition of capsular polysaccharides. Today, 84 serotypes are recognized. Unfortunately, the nomenclature for enumeration of the types has developed along two separate lines. The American nomenclature simply denotes serotypes 1 to 84, whereas the Danish nomenclature considers the antigenic relation among some of the types and brings these types together into groups. For instance, serogroup 6 includes serotypes 6A and 6B, whereas serotype 1 or serotype 3 are antigenically more unique and therefore designed as individual serotypes. The Danish nomenclature, which includes 46 different serogroups or -types, has been used here.

In study HI, 215 strains of S. pneumoniae from blood or CSF were examined. They had been isolated during 1971 to 1983 at the clinical bacteriological laboratories of Malmö, Göteborg, and Umeå. Serotypes 3, 4, 14, 7F, 23F, 6A, 6B, and 8 were the most common, together comprising 62% of all strains.

What is the rationale for undertaking epidemiological studies with regard to pneumococcal serotype? The most obvious reason relates to the current use and the future development of pneumococcal vaccine. In each region, where the present 23-valent vaccine is or may be used, the distribution of serotypes is of pertinent interest. In 1983, when study HI was planned, the 14-valent pneumococcal vaccine was replaced by a 23-valent vaccine mainly based on serotype distribution of

37

European (Danish) and North American strains (142).Out of the 215 strains described in paper IE, 89% belonged to serotypes included in or completely

cross-immunogenic with serotypes in the 23-valent vaccine. This agrees well with studies from the USA (65), where the vaccine is probably more used than in Europe. There was a remarkable similarity in distribution of serotypes of our study (El) and studies of blood and CSF strains isolated during 1965-1976 in the USA (11), during 1969-1977 in England (165), during 1955-1970 in Denmark (114), and during 1976-1978 in Stockholm (89). Notably, our study (El) agreed well with that from Stockholm, thus supporting the data as representative for Sweden. The two Swedish studies (El,89) showed differences in details from the other studies cited here, insofar as serotype 1 was less prevalent and serotype 14 more prevalent in Sweden than in the other countries. Furthermore, we found (III) (in Göteborg) serotype 25 in 4% (9/215), which was surprising, because this type was rare (0.8%) in a world-wide survey of serotypes (142). In study V, we determined serogroups or -types in 44/196 patients with pneumococcal pneumonia, and found serotype 3 in as many as 17 (39%) patients, whereas serogroups and -types 4, 6, 8, 9, 10, 11, 14, 18, 19, 22, and 23 were found in varying frequencies from 2-9%.

Thus, a substantial amount of data indicates that during the 1960s and 1970s, there was a general conformance between the distribution of pneumococcal serotypes in the Western countries on one hand and the composition of the vaccine on the other. So, what about the 1980s? In an extensive study (177) of blood and CSF isolates from Belgium, Denmark, France, the Netherlands, Norway, Sweden, the United Kingdom, and Israel during 1982-1987, the 10 most prevalent serotypes comprised all the 8 serotypes found to be overrepresented in our study from 1971-1983 (IE). This indicates that the distribution of serotypes of invasive pneumococcal disease is not only similar in Western countries but is also relatively unchanged over several decades (Fig. 4).

The present 23-valent pneumococcal vaccine was based on epidemiological data mostly from North America and Europe. There is data to indicate, however, that the vaccine may be well suited for use in other parts of the world. Hence, the vaccine comprises more than 95% of strains isolated in blood or CSF in South Africa (103). Limited studies of pneumococcal strains and efficacy studies of the vaccine indicate that also in Papua New Guinea, the vaccine has the optimal composition (140).

The distribution of serotypes among pneumococcal isolates from blood and CSF in Sweden showed no överrepresentation of a certain serotype in any age-group (in). This is in general accordance with the extensive study of Vinther-Nielsen et al (177). In that study, the 23 most common isolates from adults and children >14 years old were almost the same. However, type 3 was more common in adults, whereas type 18C and serotypes 6A and 6B were more common in children. This knowledge is important if separate vaccines for children should be developed. The general inability of children < 2 years old to evoke good antibody responses to polysaccharide vaccines was overcome in work on Haemophilus influenzae, where the H. influenzae type b polysaccharide was conjugated to a protein. This has stimulated work to improve the pneumococcal vaccine with conjugation of pneumococcal polysaccharides to proteins. With regard to a putative conjugate pneumococcal vaccine for children, there is another problem. The present data on serotypes refer mostly to blood and CSF isolates of pneumococci. With the aim to prevent invasive disease, this is logical. An important indication for a conjugate vaccine would, however, be acute otitis media, and the distribution of serotypes of such strains in various countries is less well known. In a Swedish

38

study (89), there were significant differences between the distribution of blood/CSF strains and nasopharyngeal/nasal strains. Fortunately though, most of the nasopharyngeal/nasal strains, i.e. strains supposed to be important as a source of pneumococcal otitis, belonged to serotypes represented in the 23-valent vaccine.

Fig. 4. Percentage of various serogroups and -types of pneumococci isolated from blood and CSF in Sweden during 1971-83 (IH) and in Western Europe and Israel during 1982-87 (177).

Sweden 1971-83 W. Europe 1982-87and Israel

Serotypes-groups

25242322212019181716151413121110987 -

654321

-□■3

10 15 0 10 15

* Other serotypes and serogroups

39

Our study (III) was performed by use of coagglutination for serotyping of pneumococci. This technique was found cheaper and less time-consuming than CIE, and also allowed the detection of neutrally charged capsular antigens of serotypes 7F, 7A, 14, 33F, 33A, and 37, which were not identified by CIE (II). We found that 19% of the pneumococci belonged to these serotypes. The techniques will not here be discussed in detail. Suffice it to say that a simple and cheap technique is necessary when, in the future, vaccine development and problems with resistant strains will necessitate further performance of typing pneumococci in many populations.

Diagnosis of pneumococcal pneumonia (V) and use of pneumococcal serology (IV)

Community-acquired pneumonia is a common disease with a significant morbidity and mortality, caused by a wide variety of microorganisms. A better understanding of the microbial etiology of pneumonia is fundamental for choice of antibiotic treatment. This knowledge is of special importance in emergency situations. The percentage of pneumonia with unknown etiology has varied from 3 to 45% in different reports from the 1980s, which shows that there is still a gap in our knowledge (Table 17).

Table 17. Proportion of microbial agents in eight studies of hospitalized adult patients with pneumonia.

________________ Ref no/No of cases________________________(117) (28) (7) (178) (79) (21) (V) (182)

Microbial agent 127 453 207 236

%

147 127 196 277

S. pneumoniae 76 42 39 36 47 54 32 46H. influenzae 3 6 0.9 10 9.5 4 5 4M. catarrhalis 2 1.5 1M. pneumoniae 18 17 1 5.4 14 9 10C. psittaci 3 6.2 1 1.4 2 3 1L. pneumophila 15 2 6.2 2.7 2 4S. aureus 2.4 1 0.4 1 0.7 0.7 1.5 1Gram negative rods 0.8 1 3.3 1.5 0 1 1Virus 7 4.1 13 11 18 21 16Unknown etiology 3 33 20 45 29 21 36 32

A general finding in the studies was a marked predominance of pneumococci. The difference in the proportion of pneumococcal etiology among the 8 studies may depend on use of different sampling and culture techniques and different definitions of pneumococcal pneumonia but also on different use of antibiotics before admittance. In our study (V) of 196 hospitalized adult patients with pneumonia, 70 patients remained without etiological diagnosis. Of these patients, 40% had received antibiotics

40

before admittance. Of the 126 patients, in whom the etiology was determined, only 24% had been treated before admittance.

Another pitfall in etiological studies of pneumonia relates to radiological examination; visible infiltrates on chest X-ray may be due to a non-infectious inflammatory disease and the patient may be incorrectly classified as a case of pneumonia with unknown etiology. A variation in diagnostic investigations among studies may also be a cause of diverging results (Table 18). Some diagnostic techniques with high sensitivity are expensive or difficult to perform and are therefore not routinely used, for instance culture via fiber bronchoscopy and TTA.

In study V, we found a microbiological etiology of pneumonia in 64% of the 196 hospitalized adult patients, using culture, antigen detection test, and serology.

Table 18. Sensitivity of diagnostic methods in different studies of pneumococcal pneumonia in adults.

Diagnostic method

Ref no (117) (92) (110)

%(V)

CultureBlood 14 25 16 21Sputum 19 80 60 52Nasopharynx - 59 - 35

Antigen detection testSputum 72* 56* 6811 79§Urine 64* 18* OH 128

* CIE flLatex aggi. §COA

Most patients with pneumococcal pneumonia are not bacteremic; thus the sensitivity of blood culture is low (92,8). However, this technique has high priority in all cases with emergent febrile infection due to its excellent specificity. Blood culture was done in 95 % of all 196 patients with pneumonia. In 13/62 (21%) patients with pneumococcal pneumonia, pneumococci were isolated from blood, results which are congruent with those of other studies (Table 18).

Sputum culture is, when done with optimal technique, valuable (126). However, sometimes the patient cannot cough or there is a ”dry” cough, without production of sputum. There are techniques by which the amount of expectorate can be increased. The patient can simply be given additional oral or intravenous fluids, or be allowed to breathe in a PEEP-bottle, or to use mucolytics. Inhalation of a 3% saline solution may also be effective. In spite of all these efforts only saliva is sometimes obtained. The presence of inflammatory cells has to be microscopically confirmed in homogenized samples, so as to exclude saliva samples with epithelial cells (24,92). In study V, sputum culture was performed in 70% of all patients and found to be positive in 25/48 (52%) of patients with pneumococcal pneumonia. The sensitivity of sputum culture for pneumococcal pneumonia has varied from 19 to 80% (Table 18).

41

In etiological studies of pneumonia, nasopharynx culture is generally regarded as less reliable than sputum culture, due to the fact that pneumococci often colonize the nasopharynx in the absence of pneumococcal disease (48,70). However, Kalin (91) found pneumococci in 43% in adults with invasive pneumococcal pneumonia both in nasopharynx and blood, and within each patient the isolates from both sources belonged to the same serotype. In non-pneumococcal bacteremia and mycoplasma pneumonia, on the other hand, nasopharynx culture yielded pneumococci in only 3% and 4% respectively. In a Swedish study (104) of adult carriers a similar low carrier rate was reported. In our study (V), nasopharynx culture was done in 95% of the patients. In 5/12 (42%) patients with invasive pneumococcal pneumonia and 17/50 (34%) patients with non-invasive pneumococcal pneumonia, pneumococci were isolated by this method. In 13 patients, pneumococci were isolated from nasopharynx and concomitantly detected in at least one more source. The pneumococcal strains identified from different sources of each of these patients belonged to an identical serogroup or -type (unpublished data derived from the raw material of paper V). Thus, in adult patients with pneumonia, nasopharynx culture is at least of some value. In children nasopharyngeal carriership of pneumococci is as high as 20-35% (33,70), and nasopharynx culture is therefore of little or no value for diagnostic purposes in pneumonia.

Transtracheal aspiration (TTA) is an invasive technique, which can be apprehended as laborious, but if all precautions are taken it is an easy, rapid, and mostly painless procedure to obtain secretion from the lower respiratory tract (143). In our study (V), it was performed in 61 (31%) of the patients. In 18 of these cases pneumococci were isolated. In 17 of the 18 cases, the pneumococcal etiology would have been determined irrespective of TTA. In 7 patients with proven pneumococcal pneumonia, TTA culture was negative. Two of these patients had been treated with penicillin before TTA was performed, whereas 5 showed growth of pneumococci in sputum or nasopharynx culture. False negative results in TTA cultures have been reported also by Schreiner (143). The reason for the false negative results is unknown. Tests for detection of pneumococcal antigen in TTA secretion did not improve the possibility to diagnose pneumococcal pneumonia over the results obtained by culture. However, the technique was valuable when diagnosing bacteria other than pneumococci, i.e. noncapsulated H. influenzae and Moraxella (Branhamella) catarrhalis, which were found in 9 and 3 cases each. Similar conclusions have been reached by Verghese (172) in studies of bacterial pneumonia in the elderly. Currently, there are no methods except blood cultures, which can reliably diagnose infections with these organisms.

We studied the presence of capsular antigens in sputum, tracheal secretion, and concentrated urine by COA (V). The sensitivity of COA for sputum was 79%, i.e higher than of sputum culture (52%). The specificity also seemed to be high, because all except two of the patients with positive COA had other tests indicating pneumococcal disease. In a study, where CIE (92) was used, the sensitivity was 56%. In two other Swedish studies (77,106), using ELISA, the sensitivities were reported to be as high as 82-96% with high specificity (92-94%).

In our study (V) the sensitivity of coagglutination for urine specimens was only 12%. Of the 7 patients positive by COA in urine, 5 had bacteremic pneumococcal pneumonia and thus probably high amounts of antigen. Kalin (92), using CIE, similarly reported a sensitivity of 18% in a study of the

42

pneumococcal antigen in concentrated urine, whereas Lim (110), using latex agglutination on unconcentrated urine, found 0% (Table 18). Altogether, even when using concentrated urine, the sensitivity of techniques for detection of pneumococcal antigen in urine is low.

The experience of serology for diagnosing pneumococcal infection is hitherto limited. Pneumolysin and C-polysaccharide are two potentially valuable antigens which can be used in the ELISA. In study V, we used monovalent, typespecific capsular polysaccharides, first of all to evaluate the significance of pneumococci found in isolates or by COA. The sensitivity of this ELISA was extremely high; in fact 40/41 patients with other evidence of pneumococcal infection responded with antibodies to the antigen of the infecting strain. However, this technique cannot be used routinely because of the variety of antigen preparations required. We therefore also determined antibodies against C- polysaccharide and pneumolysin, antigens which are common to all pneumococci. In bacteremic pneumonia the sensitivity was 92% and 64%, respectively. In previous studies (84,94) on patients with bacteremic pneumococcal pneumonia, the sensitivity of pneumolysin serology has been 82- 90%. It may be difficult to reach a higher sensitivity by use of these serological methods. Because, in study IV, including 43 patients with invasive pneumococcal infection, 63% of whom had underlying conditions, the sensitivities of C-polysaccharide and pneumolysin serology were as low as 51 and 60%, respectively. When in study V, all patients with pneumococcal pneumonia were considered, the sensitivity of the C-polysaccharide and pneumolysin serology was 89 and 48%, respectively. Due to the difficulties to establish the pneumococcal etiology in non-invasive disease, these figures are less valid than those confined to patients with bacteremia. In previous studies on pneumococcal pneumonia, lower sensitivity has been recorded by use of the two antigens (78,84,94). Altogether, in the form tried sofar, these serological tests seem to be of limited diagnostic value.

An important finding (V) was that 71 (36%) patients had significant increases in C-polysaccharide antibodies but were culture negative. Many of them had been treated with antibiotics, which may be an explanation why neither culture nor COA could demonstrate pneumococci. If these patients and 3 further patients with positive nasopharynx cultures or sputum COA were included among patients with pneumococcal infection, 70% would have been diagnosed as pneumococcal pneumonia and only 19% would have been denoted as cases of unknown origin.

Study V has confirmed the importance of pneumococci as etiological agents of pneumonia in adult patients requiring hospitalization. In spite of the utilization of a wide arsenal of diagnostic methods, the etiology could, however, not be determined in a significant proportion of the patients. The present studies (IV,V) together with others undertaken during the 1980s suggest that techniques based on culture, antigen detection, and serology may not be sufficient to diagnose all cases of pneumococcal disease and infection with other well-recognized agents.

43

SUMMARY AND CONCLUSIONS

The impact of invasive pneumococcal infections, with special reference to incidence, predisposing factors and prognosis, was studied retrospectively. Furthermore, the serotype distribution of pneumococcal isolates in Sweden was elucidated. Finally, a prospective study showed a prevailing importance of pneumococci as etiological agents in hospitalized adults with pneumonia. The study allowed us to evaluate different methods used to diagnose pneumococcal pneumonia.

In the retrospective study of invasive pneumococcal infections pneumococcal meningitis was diagnosed in 125 (25%) cases, pneumonia in 305 (60%), septicemia with unknown focus in 61 (12%), and septicemia with other foci in 17 (3%) cases. Pneumococcal meningitis was the most common clinical manifestation in infants, whereas pneumonia was more common in adults.

The incidence of pneumococcal meningitis was 1.2-1.4/100,000/year, which was similar to that of studies from the USA. The incidence of non-meningitic invasive pneumococcal infections, 1.9-6.1, was somewhat lower than incidence rates found in prospective studies from the USA. The difference may be due to socioeconomic and genetic factors, but different routines for obtaining blood culture from febrile patients might also be of importance. The age-specific incidence of invasive pneumococcal disease was more than ten-fold higher among infants (25.8/100,000/year) than among children 2-17 years and young to middle aged adults. At high age the incidence was 16.4. Underlying conditions were present in 23% of the infants, 34% of children aged 2-17 years, and in 81% of the adult patients with bacteremic pneumococcal infection. In infants and children no dominating underlying condition was identified, while alcoholism was common (32%) among the adults. These observations were all in agreement with those of other studies.

The overall case fatality rate was 20%. Of the 100 patients who died, 25 died within 24 hours and 67 within one week of admission. The case fatality rate was high in patients with predisposing factors (24%) as compared with patients without such factors (9%). The case fatality rate for patients with pneumococcal meningitis was 33%, bacteremic pneumococcal pneumonia 15%, and pneumococcal septicemia with unknown focus 21%.

Two techniques for serotyping were compared, counterimmunoelectrophoresis (CIE) and coagglutination (CO A). CO A was more rapid and cheaper and was also more useful due to its superior ability to discern the uncharged serotypes 7F, 7A, 14, 33F, 33A, and 37, which were found in 19% of invasive pneumococcal infections. Strains of Streptococcus pneumoniae isolated from blood and cerebrospinal fluid in Sweden during the years 1971-80 and 1982-83 were studied. No difference in serotype distribution was found, in relation to age-group, year, or place of isolation (Göteborg, Malmö, and Umeå). Overall, 89% of the pneumococcal strains were covered by the 23- valent vaccine. The serotype distribution was similar to studies from 1955 to 1980 of European and North American pneumococcal strains, with some minor local differences in the occurrence of separate serotypes.

Due to limitations in microbiological techniques the etiology of pneumonia is only partly known. In a prospective study from 1983 to 1984 of 196 hospitalized adult patients with chest X-ray verified pneumonia, culture from blood, TTA, sputum, and nasopharynx, together with COA for antigen detection of sputum, TTA, and urine were performed. In addition pneumococcal serology in paired

44

sera, with C-polysaccharide and pneumolysin as antigens, was performed. Efforts to diagnose other pathogens were also made. Pneumococcal etiology was verified in 32% of the patients. In 32%, other agents were responsible for the etiology, and in 36% of the patients the etiology was unknown. If patients with some, although not unquestionable evidence of pneumococcal etiology were included, the proportion of pneumococcal etiology would be as much as 70%.

Blood culture had a low sensitivity (21%). Sputum culture seemed to be useful, with a sensitivity of 52%, but was hampered by difficulties to obtain specimens. TTA had a sensitivity of 72%. Coagglutination had a high sensitivity (79%), when performed on sputum, but very low when used for concentrated urine (12%).

The pneumococcal serology showed somewhat contradicting results. When applied on adult patients with pneumococcal pneumonia, rises in IgG and/or IgM antibodies against C-polysaccharide were found in 89%, while only 47% had increase in IgG antibodies against pneumolysin. However, when applied on patients with invasive pneumococcal infection and severe TTA verified pneumococcal pneumonia, the corresponding sensitivities were 51% and 60%, respectively.

This project has shown the importance of 5. pneumoniae as an invasive pathogen with an obvious impact on the panorama of infections in all age-groups. Furthermore, the project has shown the special ability of the pneumococci to attack patients with underlying conditions, resulting in a high mortality, despite the availability of effective antimicrobial agents. The results emphasize the need for more effective pneumococcal vaccines and improved diagnostic methods.

45

ACKNOWLEDGEMENTS

I wish to direct my sincere thanks to:

Birger Trollfors for his patience, his encouraging and didactic manner to conduct me through all

difficulties, and for his brilliant ideas, when the problems were too heavy.

Ragnar Norrby for his generous support, enthusiasm, and never failing capacity to find conceptions.

Arne Tämvik for his positive criticism, scientific spirit, and ability to turn difficulties into easiness.

My other co-authors for most skilful help with laboratory work.

Monica Jonsson, Sonia Lundberg, and Mona Lisa Jonsson for their valuable assistance in laboratory

work, and collecting and assorting samples.

Rolf Lundholm and the staff at the Clinical Bacteriological Laboratory, Umeå, for invaluable help

with microbiological laboratory work.

The staff at the Clinical Virological Laboratory, Umeå for their initiated diagnostics of virus.

Berit Norrman and Gunborg Eriksson for excellent secreterial work.

All the staff at the Department of Infectious Diseases, Umeå for assistance to obtain samples from the

patients and other splendid support to complete this work.

This work was also supported by grants from the Faculty of Medicine, University of Umeå.

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

46

REFERENCES

Adamson JS Jr, Bates JH. Percutaneous needle biopsy of the lung. Arch Intern Med 1967;119:164-9.

Alvarez S, Guarderas J, Shell CG, Holtsclaw-Berk S, Berk SL. Nosocomial pneumococcal bacteremia. Arch Intern Med 1986;146:1509-12.

Ammann AJ, Addiego J, Wara DW, Lubin B, Smith WB, Mentzer WC. Polyvalent pneumococcal-polysaccharide immunization of patients with sickle-cell anemia and patients with splenectomy. N Engl J Med 1977;297:897-900.

Ammann AJ, Schiffman G, Addiego JE, Wara WM, Wara DW. Immunization of immunosuppressed patients with pneumococcal polysaccharide vaccine. Rev Infect Dis 1981 ; 3(Suppl) : S160-7.

Andersson B. Adhesion of Streptococcus pneumoniae to human pharyngeal epithelial cells [Thesis]. Göteborg, Sweden: University of Göteborg. 1988:25-6.

Arstila PP, Halonen P. Direct antigen detection. In: Lennette E, Halonen P, Murphy F, eds. Laboratory diagnosis of infectious diseases. Vol ü. Viral, rickettsial, and chlamydial diseases. New York: Springer Verlag 1988:60-75.

Ausina V, Coll P, Sambeat M, Puig I, Condom MJ, Luquin M, Ballester F, Prats G. Prospective study on the etiology of community-acquired pneumonia in children and adults in Spain. Eur J Clin Microbiol Infect Dis 1988;7:343-7.

Austrian R, Gold J. Pneumococcal bacteremia with especial reference to bacteremic pneumococcal pneumonia. Ann Intern Med 1964;60:759-76.

Austrian R. Random gleanings from a life with the pneumococcus (Maxwell Finland lecture). J Infect Dis 1975;131:474-84.

Austrian R. The quelling reaction, a neglected microbiologie technique. Mt Sinai J Med (NY) 1976;43:699-709.

Austrian R, Douglas RM, Schiffman G, Coetzee AM, Koomhof HJ, Hayden-Smith S, Reid RDW. Prevention of pneumococcal pneumonia by vaccination. Trans Assoc Am Physicians 1976;89:184-94.

Austrian R. Some observations on the pneumococcus and on the current status of pneumococcal disease and its prevention. Rev Infect Dis 1981;3(Suppl):Sl-17.

Austrian R. Pneumococcus: The first one hundred years. Rev Infect Dis 1981 ;3:183- 9.

Austrian R. Editorial retrospective: A reassessment of pneumococcal vaccine. N Engl J Med 1984;310:651-3.

Avery O, Heidelberger M. Immunological relationships of cell constituents of pneumococcus. Paper II. J Exp Med 1925;42:367-76.

Barrett DJ, Ammann AJ. Pneumococcal vaccine in sickle cell disease: IgG and IgM antibody response. Rev Infect Dis 1981;3(Suppl):S179-82.

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

47

Berg S, Trollfors B, Alestig K, Jodal U. Incidence, serogroups and case-fatality rate of invasive meningococcal infections in a Swedish Region 1975-1989. Scand J Infect Dis 1992;24:333-8.

Berk SL, Gallemore GM, Smith JK. Nosocomial pneumococcal pneumonia in the elderly. J Am Geriatr Soc 1981;29:319-21.

Berk SL, Holtsclaw SA, Wiener SL, Smith JK. Nontypeable Haemophilus influenzae in the elderly. Arch Intern Med 1982;142:537-9.

Berk SL, Gage KA, Holtsclaw-Berk SA, Smith JK. Type 8 pneumococcal pneumonia: An outbreak on an oncology ward. South Med J 1985;78:159-61.

Bemtsson E, Blomberg J, Lagergård T, Trollfors B. Etiology of community-acquired pneumonia in patients requiring hospitalization. Eur J Clin Microbiol 1985;4:268-72.

Bemtsson E, Lagergård T, Strannegård Ö, Trollfors B. Etiology of community- acquired pneumonia in out-patients. Eur J Clin Microbiol Infect Dis 1986;5:446-7.

Berry AM, Yother J, Briles DE, Hansman D, Paton JC. Reduced virulence of a defined pneumolysin-negative mutant of Streptococcus pneumoniae. Infect Immun 1989;57:2037-42.

Boemer DF, Zwadyk P. The value of the sputum Gram's stain in community- acquired pneumonia. JAMA 1982;247:642-5.

Bortolussi R, Thompson TR, Ferrieri P. Early-onset pneumococcal sepsis in newborn infants. Pediatrics 1977;60:352-5.

Bratton L, Teele DW, Klein JO. Outcome of unsuspected pneumococcemia in children not initially admitted to the hospital. J Pediatr 1977;90:703-6.

Breiman RF, Spika JS, Navarro VJ, Darden PM, Darby CP. Pneumococcal bacteremia in Charleston County, South Carolina: A decade later. Arch Intern Med 1990;150:1401-5.

British Thoracic Society and the Public Health Laboratory Service, (Research Committee). Community-acquired pneumonia in adults in British hospitals in 1982- 1983: A survey of aetiology, mortality, prognostic factors and outcome. Q J Med, new series 62 1987;239:195-220.

Broome CV, Facklam RR, Fraser DW. Pneumococcal disease after pneumococcal vaccination: An alternative method to estimate the efficacy of pneumococcal vaccine. N Engl J Med 1980;303:549-52.

Broome CV, Facklam RR, Allen JR, Fraser DW, Austrian R. From the Center for Disease Control: Epidemiology of pneumococcal serotypes in the United States, 1978-1979. J Infect Dis 1980;141:119-23.

Broome CV. Efficacy of pneumococcal polysaccharide vaccines. Rev Infect Dis 1981;3(Suppl):S82-8.

Broome CV, Facklam RR. Epidemiology of clinically significant isolates of Streptococcus pneumoniae in the United States. Rev Infect Dis 1981;3:277-80.

Brorson J-E, Malmvall B-E. Branhamella catarrhalis and other bacteria in the nasopharynx of children with longstanding cough. Scand J Infect Dis 1981;13:111-3.

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

48

Bniyn GAW, van der Meer JWM, Hermans J, Knoppert W. Pneumococcal bacteremia in adults over a 10-year period at University Hospital, Leiden. Rev Infect Dis 1988;10:446-50.

Burke JP, Klein JO, Gezon HM, Finland M. Pneumococcal bacteremia: Review of 111 cases, 1957-1969, with special reference to cases with undetermined focus. Am J Dis Child 1971;121:353-9.

Cadoz M, Denis F, Mar ID. Etude épidémiologique des cas de méningites purulentes hospitalisés à Dakar pendant la décennie 1970-1979. Bulletin de l'Organisation mondiale de la Santé 1981;59(4):575-584.

Center for Disease Control. Bacterial meningitis and meningococcemia-United States, 1978. MMWR 1979;28:277-9.

Chu C, Schneerson R, Robbins JB, Rastogi SC. Further studies on the immunogenicity of Haemophilus influenzae type b and pneumococcal type 6A polysaccharide-protein conjugates. Infect Immun 1983;40:245-56.

Claesson B, Trollfors B, Jodal U, Rosenhall U. Incidence and prognosis of Haemophilus influenzae meningitis in children in a Swedish region. Pediatr Infect Dis 1984;3:35-9.

Claesson BA, Trollfors B, Brolin I, Granström M, Henrichsen I, Jodal V, Juto P, Kallings I, Kanclerski K, Lagergård T, Steinwall L, Strannegård Ö. Etiology of community-acquired pneumonia in children based on antibody responses to bacterial and viral antigens. Pediatr Infect Dis J 1989;8:856-62.

Coonrod JD, Rytel MW. Detection of type-specific pneumococcal antigens by counterimmunoelectrophoresis. I. Methodology and immunologic properties of pneumococcal antigens. J Lab Clin Med 1973;81:770-7.

Coonrod JD, Drennan DP. Pneumococcal pneumonia: Capsular polysaccharide antigenemia and antibody responses. Ann Intern Med 1976;84:254-60.

Corberand JX, Laharrague PF, Filliola G. Neutrophils are normal in normal aged humans. J Leukoc Biol 1986;40:333-4.

Cowan MJ, Ammann AJ, Wara DW, Howie VM, Schultz L, Doyle N, Kaplan M. Pneumococcal polysaccharide immunization in infants and children. Pediatrics 1978;62:721-7.

Dahlberg T. Quantitation of antibodies to Haemophilus influenzae type b in humans by enzyme-linked immunosorbent assay. J Clin Microbiol 1981;13:1080-7.

Davidson M, Schraer CD, Parkinson AJ, Campbell JF, Facklam RR, Wainwright RB, Lanier AP, Heyward WL. Invasive pneumococcal disease in an Alaska native population, 1980 through 1986. JAMA 1989;261:715-8.

Douglas RM, Paton JC, Duncan SJ, Hansman DJ. Antibody response to pneumococcal vaccination in children younger than five years of age. J Infect Dis 1983;148:131-7.

Dowling JN, Sheehe PR, Feldman HA. Pharyngeal pneumococcal acquisitions in "normal" families: A longitudinal study. J Infect Dis 1971;124:9-17.

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

49

Everett ED, Rahm Jr AE, Adaniya R, Stevens DL, McNitt TR. Haemophilus influenzae pneumonia in adults. JAMA 1977;238:319-21.

Filice GA, Darby CP, Fraser DW. Pneumococcal bacteremia in Charleston County, South Carolina. Am J Epidemiol 1980;112:828-35.

Finkeistein MS, Petkun WM, Freedman ML, Antopol SC. Pneumococcal bacteremia in adults: Age-dependent differences in presentation and in outcome. J Am Geriatr Soc 1983;31:19-27.

Finland M, Brown JW, Barnes MW. Immune reactions of carriers and noncarriers of type-specific pneumococci: A bacteriological and immunological study of hospital contacts. Am J Hygiene 1950;32B:24-37.

Finland M, Barnes MW. Changes in occurrence of capsular serotypes of Streptococcus pneumoniae at Boston City Hospital during selected years between 1935 and 1974. J Clin Microbiol 1977;5:154-66.

Finland M, Barnes MW. Acute bacterial meningitis at Boston City Hospital during 12 selected years, 1935-1972. J Infect Dis 1977;136:400-15.

Finland M. Pneumonia and pneumococcal infections, with special reference to pneumococcal pneumonia. Am Rev Respir Dis 1979;120:481-502.

Foy HM, Cooney MK, McMahan R, Grayston JT. Viral and mycoplasmal pneumonia in a prepaid medical care group during an eight-year period. Am J Epidemiol 1973;97:93-102.

Fraser DW, Darby CP, Koehler RE, Jacobs CF, Feldman RA. Risk factors in bacterial meningitis: Charleston County, South Carolina. J Infect Dis 1973; 127:271-

Fraser DW, Geil CC, Feldman RA. Bacterial meningitis in Bernalillo County, New Mexico: A comparison with three other American populations. Am J Epidemiol 1974;100:29-34.

Giebink GS, Le CT, Cosio FG, Spika JS, Schiffman G. Serum antibody responses of high-risk children and adults to vaccination with capsular polysaccharides of Streptococcus pneumoniae. Rev Infect Dis 1981 ;3(Suppl):S 168-78.

Gray BM, Converse III GM, Dillon HC Jr. Epidemiologic studies of Streptococcus pneumoniae in infants: Acquisition, carriage, and infection during the first 24 months of life. J Infect Dis 1980;142:923-33.

Gray BM, Dillon HC Jr. Clinical and epidemiologic studies of pneumococcal infection in children. Pediatr Infect Dis 1986;5:201-7.

Gruer LD, McKendrick MW, Geddes AM. Pneumococcal bacteraemia - a continuing challenge. Q J Med 1984;210:259-70.

Gundel M, Schwarz FKT. Studien über die Bakterienflora der oberen Atmungswege Neugeborener (imvergleich mit der Mundhöhlenflora der Mutter und des Pflegepersonals) unter besonderer Berücksichtigung ihrer bedeutung für das Pneumonieproblem. Zeitschrift fur Hygiene und Infektionskrankheiten 1932; 113:411-

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

50

Gwaltney JM Jr, Sande MA, Austrian R, Hendley JO. Spread of Streptococcus pneumoniae in families. II. Relation of transfer of S. pneumoniae to incidence of colds and serum antibody. J Infect Dis 1975;132:62-8.

Hager HL, Woolley TW, Berk SL. Review of recent pneumococcal infections with attention to vaccine and nonvaccine serotypes. Rev Infect Dis 1990;12:267-72.

Heffron R. Pneumonia with special reference to pneumococcus lobar pneumonia. The commonwealth Fund, New York 1939. 2nd printing, Harvard University Press, Cambridge, Mass. 1979:1, 92, 312.

Heidelberger M, Jann K, Jann B, Orskov F, Orskov I, Westphal O. Relations between structures of three K polysaccharides of Escherichia coli and cross-reactivity in anti pneumococcal sera. J Bacteriol 1968;95:2415-7.

Heidelberger M, Nimmich W. Additional immunochemical relationships of capsular polysaccharides of Klebsiella and pneumococci. J Immunol 1972;109:1337-44.

Heineman HS, Chawla JK, Lofton WM. Misinformation from sputum cultures without microscopic examination. J Clin Microbiol 1977;6:518-27.

Hendley JO, Sande MA, Stewart PM, Gwaltney JM Jr. Spread of Streptococcus pneumoniae in families. I. Carriage rates and distribution of types. J Infect Dis 1975;132:55-61.

Henrichsen J. The pneumococcal typing system and pneumococcal surveillance. J Infect 1979(Suppl2); 1:31-7.

Henrichsen J, Bemtsson E, Kaijser B. Comparison of counterimmunoelectrophoresis and the capsular reaction test for typing of pneumococci. J Clin Microbiol 1980;11:589-92.

Hilleman MR, Carlson AJ Jr, McLean AA, Velia PP, Weibel RE, Woodhour AF. Streptococcus pneumoniae polysaccharide vaccine, age and dose responses, safety, persistence of antibody, revaccination and simultaneous administration of pneumococcal and influenza vaccines. Rev Infect Dis 1981;3(Suppl):S31-42.

Hirschberg L, Krook A, Pettersson CA, Vikerfors T. Enzyme-linked immunosorbent assay for detection of Mycoplasma pneumoniae specific immunoglobulin M. Eur J Clin Microbiol Infect Dis 1988;7:420-3.

Hodges RG, MacLeod CM. Epidemic pneumococcal pneumonia: V. Final consideration of the factors underlying the epidemic. Am J Hygiene 1946;44:237-43.

Hokama Y, Coleman MK, Riley RF. In vitro effects of C-reactive protein on phagocytosis. J Bacteriol 1962;83:1017-24.

Holmberg H, Holme T, Krook A, Olsson T, Sjöberg L, Sjögren A-M. Detection of C polysaccharide in Streptococcus pneumonia in the sputa of pneumonia patients by an enzyme-linked immunosorbent assay. J Clin Microbiol 1985;22:111-5.

Holmberg H, Krook A, Sjögren AM. Determination of antibodies to pneumococcal C polysaccharide in patients with community-acquired pneumonia. J Clin Microbiol 1985;22:808-14.

Holmberg H. Aetiology of community-acquired pneumonia in hospital treated patients. Scand J Infect Dis 1987;19:491-501.

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

51

Hook ID EW, Horton CA, Schaberg DR. Failure of intensive care unit support to influence mortality from pneumococcal bacteremia. JAMA 1983;249:1055-7.

Hsiung GD. Complement fixation test. In: Hsiung GD, ed. Diagnostic virology. New Haven, CT: Yale University Press 1982:42-9.

Ingvarsson L, Lundgren K, Ursing I. The bacterial flora in the nasopharynx in healthy children. Acta Otolaryngol 1982;386(Suppl):94-6.

Jacobs NM, Lerdkachomsuk S, Metzger WI. Pneumococcal bacteremia in infants and children: A ten-year experience at the Cook County Hospital with special reference to the pneumococcal serotypes isolated. Pediatrics 1979;64:296-300.

Jalonen E, Paton JC, Koskela M, Kerttula Y, Leinonen M. Measurement of antibody responses to pneumolysin-a promising method for the presumptive aetiological diagnosis of pneumococcal pneumonia. J Infect 1989;19:127-34.

Jansson E. A 10-year study of bacteremia. Scand J Infect Dis 1971;3:151-5.

Jennings HJ, Lugowski C, Young NM. Structure of the complex polysaccharide C- substance from Streptococcus pneumoniae type 1. Biochemistry 1980;19:4712-9.

Jetté LP, Lamothe F. Surveillance of invasive Streptococcus pneumoniae infection in Quebec, Canada, from 1984 to 1986: Serotype distribution, antimicrobial susceptibility, and clinical characteristics. J Clin Microbiol 1989;27:1-5.

Johnson MK, Hamon D, Drew GK. Isolation and characterization of pneumolysin- negative mutants of Streptococcus pneumoniae. Infect Immun 1982;37:837-9.

Kalin M, Lindberg AA. Distribution of pneumococcal types in the Stockholm region 1976-1978. Scand J Infect Dis 1980;12:91-5.

Kalin M, Lindberg AA, Olausson EH. Diagnosis of pneumococcal pneumonia by coagglutination and counterimmunoelectrophoresis of sputum samples. Eur J CUn Microbiol Infect Dis 1982;1:91-6.

Kalin M. Bacteremic pneumococcal pneumonia: Value of culture of nasopharyngeal specimens and examination of washed sputum specimens. Eur J Clin Microbiol Infect Dis 1982;1:394-6.

Kalin M, Lindberg AA. Diagnosis of pneumococcal pneumonia: A comparison between microscopic examination of expectorate, antigen detection and cultural procedures. Scand J Infect Dis 1983;15:247-55.

Kalin M, Lindberg AA. Antibody response against the type specific capsular polysaccharide in pneumococcal pneumonia measured by enzyme linked immunosorbent assay. Scand J Infect Dis 1985;17:25-32.

Kalin M, Kanclerski K, Granström M, Möllby R. Diagnosis of pneumococcal pneumonia by enzyme-linked immunosorbent assay of antibodies to pneumococcal hemolysin(pneumolysin). J Clin Microbiol 1987;25:226-9.

Kallings I, Blomqvist S , Sterner G. Seroreactivity towards Legionella pneumophila in patients with pneumonia: possible cross-reactions. In: Thomsberry C, Balows A, Feeby JC, Jaloibowski W, eds. Legionella. Proceedings of the second international symposium. Washington, DC: American Society for Microbiology 1983:33-5.

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

52

Kanclerski K, Möllby R. Production and purification of Streptococcus pneumoniae hemolysin (pneumolysin). J Clin Microbiol 1987;25:222-5.

Kanclerski K, Blomquist S, Granström M, Möllby R. Serum antibodies to pneumolysin in patients with pneumonia. J Clin Microbiol 1988;26:96-100.

Kaplan MH, Volonakis JE. Interaction of C-reactive protein complexes with the complement system. 1.Consumption of human complement associated with the reaction of C-reactive protein with pneumococcal C-polysaccharide and with the choline phosphatides, lecithin and sphingomyelin. Immunol 1974;112:2135-47.

Kaufman P. Pneumonia in old age: Active immunization against pneumonia with pneumococcus polysaccharide; results of a six year study. Arch Intern Med 1947;79:518-31.

Kerttula Y, Leinonen M, Koskela M, Mäkelä PH. The aetiology of pneumonia. Application of bacterial serology and basic laboratory methods. J Infect 1987; 14:21- 30.

Klein JO, Marcy SM. Bacterial infections. In: Remington JS, Klein JO, eds. Infectious diseases of the fetus and newborn infant. Philadelphia: WB Saunders 1976;747-891.

Klein JO. The epidemiology of pneumococcal disease in infants and children. Rev Infect Dis 1981;3:246-53.

Klugman KP, Koomhof HJ. Drug resistance patterns and serogroups or serotypes of pneumococcal isolates from cerebrospinal fluid or blood, 1979-1986. J Infect Dis 1988;158:956-64.

Krantz I, Alestig K, Trollfors B, Zackrisson G. The carrier state in pertussis. Scand J Infect Dis 1986;18:121-3.

Kronvall G. A rapid slide-agglutination method for typing pneumococci by means of specific antibody adsorbed to protein A-containing staphylococci. J Med Microbiol 1973;6:187-90.

Krook, A, Holmberg H. Pneumococcal antigens in sputa: ELISA for the detection of pneumococcal C-polysaccharide in sputa from pneumonia patients. Diagn Microbiol Infect Dis 1987;7:73-5.

Lehtomäki K, Leinonen M, Takala A, Hovi T, Herva E, Koskela M. Etiological diagnosis of pneumonia in military conscripts by combined use of bacterial culture and serological methods. Eur J Clin Microbiol Infect Dis 1988;7:348-54.

Leland DS, French MLV. Virus isolation and identification. In: Lennette E, Halonen P, Murphy F, eds. Laboratory diagnosis of infectious diseases. Voi ü. Viral, rickettsial, and chlamydial diseases. New York: Springer Verlag 1988:39-59.

Lewis RJ, Richmond AS, McGrory JP. Diplococcus pneumoniae cellulitis in drug addicts. JAMA 1975;232:54-5.

Lim I, Shaw DR, Stanley DP, Lumb R, McLennan G. A prospective hospital study of the aetiology of community-acquired pneumonia. Med J Aust 1989;151:87-91.

53

111 Loda FA, Collier AM, Glezen WP, Strangert K, Clyde WA Jr, Denny FW. Occurrence of Diplococcus pneumoniae in the upper respiratory tract of children. J Pediatr 1975;87:1087-93.

112 Lorian V, Waluschka A, Popoola B. Pneumococcal beta-hemolysin produced under the effect of antibiotics. Appi Microbiol 1973;25:290-4.

113 Lund E. Antigenic relationship between pneumococci and non-hemolytic streptococci. Acta Pathol Microbiol Scand 1950;27:110-8.

114 Lund E. Types of pneumococci found in blood, spinal fluid and pleural exudate during a period of 15 years (1954-1969). Acta Pathol Microbiol Scand 1970;78:333- 6 .

115 Lund E, Henrichsen J. Laboratory Diagnosis, serology and epidemiology of Streptococcus pneumoniae. In: Bergan T, Norris JR, eds. Methods in Microbiology. London: Academic Press 1978;12:241-62.

116 Lund E. Distribution of pneumococcus types at different times in different areas. In Bayer Symposium HI. Springer-Verlag, New York 1971:49-56.

117 MacFarlane JT, Ward MJ, Finch RG, MacRae AD. Hospital study of adult community-acquired pneumonia. Lancet 1982;2:255-8.

118 Masters PL, Brumfitt W, Mendez RL, Likar M. Bacterial flora of the upper respiratory tract in Paddington families, 1952-4. Br Med J 1958;1:200-5.

119 McGowan JE Jr, Barnes MW, Finland M. Bacteremia at Boston City Hospital: Occurrence and mortality during 12 selected years (1935-1972) with special reference to hospitalacquired cases. J Infect Dis 1975;132:316-35.

120 McIntyre P, Kennedy R, Harris F. Occult pneumococcal bacteraemia and febrile convulsions. Br Med J 1983;286:203-6.

121 McNabb WR, Shanson DC, Williams TDM, Lant AF. Adult community-acquired pneumonia in central London. J R Soc Med 1984;77:550-5.

122 Mold C, Edwards KM, Gewürz H. Effect of C-reactive protein on the complement- mediated stimulation of human neutrophils by Streptococcus pneumoniae serotypes 3 and 6. Infect Immun 1982;37:987-92.

123 Mold C, Rodgers CP, Kaplan RL, Gewürz H. Binding of human C-reactive protein to bacteria. Infect Immun 1982;38:392-5.

124 Mufson MA, Kruss DM, Wasil RE, Metzger WI. Capsular types and outcome of bacteremic pneumococcal disease in the antibiotic era. Arch Intern Med 1974;134:505-10.

125 Mujais S, Uwaydah M. Pneumococcal cellulitis. Infection 1983;11:173-4.

126 Musher DM. Infections caused by Streptococcus pneumoniae: Clinical spectrum, pathogenesis, immunity, and treatment. Clin Infect Dis 1992;14:801-7.

127 Mäkelä PH, Leinonen M, Pukander J, Karma P. A study of the pneumococcal vaccine in prevention of clinically acute attacks of recurrent otitis media. Rev Infect Dis 1981 ;3(Suppl):S 124-30.

54

128 Nandoskar M, Ferrante A, Bates EJ, Hurst N, Paton JC. Inhibition of human monocyte respiratory burst, degranulation, phospholipidmethylation and bactericidal activity by pneumolysin. Immunology 1986;59:515-20.

129 National Central Bureau of Statistics. Official statistics of Sweden. Population. Part3. 1970-1980. S-10250 Stockholm, Sweden.

130 Neufeld F. Über die Agglutination der Pneumokokken und über die Theorien der Agglutination. Zeitschrift fur Hygiene und Infektionskrankheiten 1902;40:54-72.

131 Nordstrom K, Kallings I, Dahnsjö H, Clemens F. An outbreak of Legionnaires' disease in Sweden: Report of sixty-eight cases. Scand J Infect Dis 1983;15:43-55.

132 Norrby SR, Pope KA. Pneumococcal pneumonia. J Infect 1979;1:109-20.

133 Ort S, Ryan JL, Barden G, D'Esopo N. Pneumococcal pneumonia in hospitalized patients: Clinical and radiological presentations. JAMA 1983;249:214-8.

134 Pedersen FK, Henrichsen J. Detection of antibodies to pneumococcal capsular polysaccharides by enzyme-linked immunosorbent assay. J Clin Microbiol 1982;15:372-8.

135 Pedersen FK, Henrichsen J. Pneumococcal meningitis and bacteraemia in Danish children 1969-1978. Serotypes, incidence, and outcome. Acta Pathol Microbiol Immunol Scand [B] 1983;91:129-34.

136 Peltola H. Meningococcal disease: still with us. Rev Infect Dis 1983;5:71-91.

137 Poxton IR, Tarelli E, Baddiley J. The structure of C-polysaccharide from the walls of Streptococcus pneumoniae. BiochemJ 1978;175:1033-42.

138 Pratter MR, Irwin RS. Transtracheal aspiration. Guidelines for safety. Chest 1979;76:518-20.

139 Riley I, Douglas RM. An epidemiological approach to pneumococcal disease. Rev Infect Dis 1981;3:233-45.

140 Riley ID, Lehmann D, Alpers MP. Pneumococcal vaccine trials in papua New Guinea: Relationships between epidemiology of pneumococcal infection and efficacy of vaccine. Rev Infect Dis 1991;13(Suppl6):S535-41.

141 Robbins JB, Lee C-J, Rastogi SC, Schiffman G, Henrichsen J. Comparative immunogenicity of group 6 pneumococcal type 6A(6) and type 6B(26) capsular polysaccharides. Infect Immun 1979;26:1116-22.

142 Robbins JB, Austrian R, Lee CJ, Rastogi SC, Schiffman G, Henrichsen J, Mäkelä PH, Broome CV, Facklam RR, Tiesjema RH, Parke JC Jr. Considerations for formulating the second-generation pneumococcal capsular polysaccharide vaccine with emphasis on the cross-reactive types within groups. J Infect Dis 1983;148:1136-59.

143 Schreiner A, Digranes A, Myking O. Transtracheal aspiration in the diagnosis of lower respiratory tract infections. Scand J Infect Dis 1972;4:49-52.

144 Schmerber J, Deltenre M. A new fatal complication of transtracheal aspiration. Scand J Respir Dis 1978;59:232-5.

55

145 Sell SH, Wright PF, Vaughn WK, Thompson J, Schiffman G. Clinical studies of pneumococcal vaccine in infants. I. Reactogenicity and immunogenicity of two polyvalent polysaccharide vaccines. Rev Infect Dis 1981;3(Suppl):S97-107.

146 Shann F, Grätten M, Germer S, Linneman V, Hazlett D, Payne R. Aetiology of pneumonia in children in Goroka hospital, Papua New Guinea. Lancet 1984;2:537- 41.

147 Shapiro ED, Clemens JD. Pneumococcal vaccine efficacy against serious pneumococcal infections in patients with vaccine indications [Abstract No 237]. Clin Res 1983;31:237A.

148 Siber GR, Weitzman SA, Aisenberg AC. Antibody response of patients with Hodgkin’s disease to protein and polysaccharide antigens. Rev Infect Dis 1981 ;3(Suppl) : S144-59.

149 Siegman-Igra Y, Schwartz D, Alperin H, Konforti N. Invasive pneumococcal infection in Israel: Review of 90 cases. Scand J Infect Dis 1986;18:511-7.

150 Spanjaard L, Bol P, Ekker W, Zanen HC. The incidence of bacterial meningitis in the Netherlands-a comparison of three registration systems, 1977-1982. J Infect 1985;11:259-68.

151 Stuart-Harris CH. Pneumonia-present day experiences and trends. Trans Med Soc Lond 1968;84:27-50.

152 Szu SC, Oravec LS. Detection of "neutral" type 7F and type 14 pneumococcal capsular polysaccharides by immunoelectrophoresis. J Clin Microbiol 1982; 15:1172-5.

153 Teele DW, Marshall R, Klein JO. Unsuspected bacteremia in young children. Pediatr Clin North Am 1979;26:773-84.

154 Teele DW, Klein JO, Rosner BA. Epidemiology of otitis media in children. Ann Otol Rhinol Laryngol 1980;89(Suppl 68):5-6.

155 Teele DW, Klein JO, Bratton L, Fisch GR, Mathieu OR, Porter PJ, Starobin SG, Tarlin LD, Younes RP. Use of pneumococcal vaccine for prevention of recurrent acute otitis media in infants in Boston. Rev Infect Dis 1981;3(Suppl):Sl 13-8.

156 Tempest B. Pneumococcal meningitis in mother and neonate. Pediatrics 1974;53:759- 60.

157 Thirumoorthi MC, Asmar BI, Dajani AS. Violaceous discoloration in pneumococcal cellulitis. Pediatrics 1978;62:492-3.

158 Thorsteinsson SB, Musher DM, Fagan T. The diagnostic value of sputum culture in acute pneumonia. JAMA 1975;233:894-5.

159 Tilghman RC, Finland M. Clinical significance of bacteremia in pneumococcic pneumonia. Arch Intern Med 1937;59:602-19.

160 Tillet WS, Francis T Jr. Serological reaction in pneumonia with a non-protein somatic fraction of pneumococcus. J Exp Med 1930;52:561-71.

161 Tillet WS, Goebel WF, Avery OT. Chemical and immunological properties of a species-specific carbohydrate of pneumococci. J Exp Med 1930;52:895-900.

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

56

Todd JK, Bruhn FW. Severe Haemophilus influenzae infections: spectrum of disease. Am J Dis Child 1975;129:607-11.

Tomasz A. Surface components of Streptococcus pneumoniae. Rev Infect Dis 1981;3:190-211.

Trollfors B, Claesson BA, Strangert K Tarangers J. Haemophilus influenzae meningitis in Sweden 1981-1983. Arch Dis Child 1987;62:1220-3.

Turk, DC. Frequencies of pneumococcal types causing serious infections in patients admitted to the Radcliffe Infirmary, Oxford, 1969-77. J Hyg Camb 1978;81:227-38.

Wadsworth C. A slide microtechnique for the analysis of immune precipitates in gel. Int Arch Allergy Appi Immunol 1957;10:355-60.

Walder M, Svantesson B, Ursing J, Cronberg S, Johnsson T, Forsgren A. Incidence of Legionella pneumophila in acute lower respiratory tract infections. Scand J Infect Dis 1981;13:159-60.

Van der Auwera P, André A, Bulliard G, Legrand IC, Gordts B, Van Landuyt H, Schuyteneer F. Comparison of latex agglutination and counterimmunoelectrophoresis in the diagnosis of acute Streptococcus pneumoniae infections. Eur J Clin Microbiol Infect Dis 1983;2:534-40.

Van Metre TE Jr. Pneumococcal pneumonia treated with antibiotics: The prognostic significance of certain clinical findings. N Engl J Med 1954;251:1048-54.

Ward J. Antibiotic-resistant Streptococcus pneumoniae: clinical and epidemiologic aspects. Rev Infect Dis 1981;3:254-66.

Weichselbaum A. Ueber die Aetiologie der acuten Lungen- und Rippenfeilenzündungen. Medizinische Jahrbücher (series 3)1886;l:483-554.

Verghese A, Berk SL. Bacterial pneumonia in the elderly. Medicine 1983;62:271-85.

White B. The biology of pneumococcus. The commonwealth Fund, New York 1938. 2nd printing, Harvard University Press, Cambridge, Mass. 1979:44, 238-50.

Willems JS, Sanders CR, Riddiough MA, Bell JC. Cost effectiveness of vaccination against pneumococcal pneumonia. N Engl J Med 1980;303:553-9.

Wilson CB. Immunologic basis for increased susceptibility of the neonate to infection. J Pediatr 1986;108:1-12.

Wimberley N, Faling LJ, Bartlett JG. A fiberoptic bronchoscopy technique to obtain uncontaminated lower airway secretions for bacterial culture. Am Rev Respir Dis 1979;119:337-43.

Vinther Nielsen S, Henrichsen J. Capsular types of Streptococcus pneumoniae isolated from blood and CSF during 1982-1987. Clin Infect Dis 1992;15:794-8.

Woodhead MA, MacFarlane JT, McCracken JS, Rose DH, Finch RG. Prospective study of the aetiology and outcome of pneumonia in the community. Lancet 1987;i:671-4.

57

179 Wright PF, Sell SH, Vaughn WK, Andrews C, McConnell KB, Schiffman G. Clinical studies of pneumococcal vaccines in infants. II. Efficacy and effect on nasopharyngeal carriage. Rev Infect Dis 1981 ;3(Suppl):S 108-12.

180 Yurchack A, Austrian R. Serologic and genetic relationship between pneumococci and other respiratory streptococci. Trans Assoc Am Physicians 1966;79:368-75.

181 Örtqvist À, Grepe A, Julander I, Kalin M. Bacteremic pneumococcal pneumonia in Sweden: Clinical course and outcome and comparison with non-bacteremic pneumococcal and mycoplasmal pneumonias. Scand J Infect Dis 1988;20:163-71.

182 Örtqvist À, Grillner L, Hedlund J, Jalonen E, Kallings I, Leinonen M, Kalin M. Aetiology, outcome and prognostic factors in community-acquired pneumonia requiring hospitalization. Eur Respiri 1990;3:1105-13.