978 951 44 7744 7 Dikonversi

MASSIMILIANO DON
Pediatric Community-Acquired Pneumonia
A serologic study on etiology, with special focus

on newly identified agents
ACADEMIC DISSERTATION
To be presented, with the permission of
the Faculty of Medicine of the University of Tampere,
for public discussion in the Auditorium of Finn-Medi 1,
Biokatu 6, Tampere, on August 21st, 2009, at 12 o’clock.
UNIVERSITY OF TAMPERE
ACADEMIC DISSERTATION
University of Tampere, Medical School
Tampere University Hospital, Department of Paediatrics
Finland
University of Udine, School of Medicine
University Hospital Udine, Pediatric Clinic
Italy
Supervised by Reviewed by
Professor Matti Korppi Docent Marjo Renko
University of Tampere University of Oulu
Finland Finland
Doctor Mario Canciani Docent Eeva Salo
University of Udine University of Helsinki
Italy Finland
Distribution Tel. +358 3 3551 6055

Bookshop TAJU Fax +358 3 3551 7685
P.O. Box 617 taju@uta.fi
33014 University of Tampere www.uta.fi/taju
Finland http://granum.uta.fi
Cover design by
Juha Siro
Acta Universitatis Tamperensis 1423 Acta Electronica Universitatis Tamperensis 851

ISBN 978-951-44-7743-0 (print) ISBN 978-951-44-7744-7 (pdf )
ISSN-L 1455-1616 ISSN 1456-954X
ISSN 1455-1616 http://acta.uta.fi
Tampereen Yliopistopaino Oy – Juvenes Print

Tampere 2009
To my wife Barbara
To my son Michele
To my parents Margherita and Paolo
3
4
CONTENTS
LIST OF ORIGINAL PUBLICATIONS ............................................................................ 7
ABBREVIATIONS .......................................................................................................... 8
ABSTRACT .................................................................................................................... 9
TIIVISTELMÄ ................................................................................................................. 11
RIASSUNTO ................................................................................................................... 13
INTRODUCTION ............................................................................................................ 15
COMMUNITY-ACQUIRED PNEUMONIA IN THE PEDIATRIC AGE:

LITERATURE REVIEW .................................................................................................. 18
DEFINITION ......................................................................................................... 18
EPIDEMIOLOGY .................................................................................................. 18
RISK FACTORS ................................................................................................... 20
PATHOGENESIS ................................................................................................. 22
ETIOLOGY OF PEDIATRIC CAP ......................................................................... 23
GENERAL CONSIDERATIONS ................................................................ 23
ETIOLOGICAL CLASSIFICATION OF PEDIATRIC CAP .......................... 25
VIRAL ETIOLOGY ..................................................................................... 27
HUMAN METAPNEUMOVIRUS...................................................... 27
HUMAN BOCAVIRUS ..................................................................... 31
BACTERIAL ETIOLOGY............................................................................ 35
SIMKANIA NEGEVENSIS ............................................................... 35
ETIOLOGY BY AGE .................................................................................. 38
CLINICAL PICTURE ............................................................................................ 40
DIAGNOSIS ......................................................................................................... 42
CHEST RADIOGRAPHY ........................................................................... 42
GENERAL INVESTIGATIONS................................................................... 43
OXYGEN SATURATION (SaO2) ..................................................... 43
PROCALCITONIN AND OTHER NON SPECIFIC SERUM
INFLAMMATORY MARKERS ......................................................... 44
SPECIFIC MICROBIOLOGICAL INVESTIGATIONS ................................. 46
TYPICAL BACTERIAL PNEUMONIA .............................................. 46
ATYPICAL BACTERIAL PNEUMONIA ........................................... 48
VIRAL PNEUMONIA ....................................................................... 49
AIMS OF THIS STUDY................................................................................................... 51
PATIENTS AND METHODS........................................................................................... 52

STUDY DESIGN .................................................................................................. 52
PATIENTS ............................................................................................................ 52
CLINICAL DEFINITIONS ..................................................................................... 53
BACTERIOLOGY ................................................................................................. 54
SIMKANIA NEGEVENSIS SEROLOGY .................................................... 55
VIROLOGY........................................................................................................... 55
5
HUMAN METAPNEUMOVIRUS SEROLOGY ........................................... 56
HUMAN BOCAVIRUS SEROLOGY........................................................... 56
NON SPECIFIC LABORATORY MARKERS ........................................................ 57
STATISTICAL ANALYSIS .................................................................................... 58
ETHICS ................................................................................................................ 59
RESULTS ....................................................................................................................... 60
DEMOGRAPHIC, LABORATORY AND RADIOLOGICAL DATA ......................... 60
OVERALL ETIOLOGY ......................................................................................... 60
BACTERIAL INFECTIONS ................................................................................... 61
SIMKANIA NEGEVENSIS INFECTIONS ................................................... 63
VIRAL INFECTIONS ............................................................................................ 65
HUMAN METAPNEUMOVIRUS INFECTIONS ......................................... 65
HUMAN BOCAVIRUS INFECTIONS ......................................................... 68
MIXED INFECTIONS ........................................................................................... 72
ETIOLOGY BY HOSPITALIZATION .................................................................... 73
PROCALCITONIN ................................................................................................ 74
DISCUSSION ................................................................................................................. 81
GENERAL CONSIDERATIONS ........................................................................... 82
SIMKANIA NEGEVENSIS .................................................................................... 84
HUMAN METAPNEUMOVIRUS........................................................................... 85
HUMAN BOCAVIRUS .......................................................................................... 87
PROCALCITONIN ................................................................................................ 89
STRENGTHS AND SHORTCOMINGS OF THE STUDY ..................................... 91
FUTURE PERSPECTIVES .................................................................................. 93
CONCLUSIONS ................................................................................................... 93
ACKNOWLEDGMENTS ................................................................................................. 95
REFERENCES ............................................................................................................... 99
ORIGINAL PUBLICATIONS ......................................................................................... 123
6
LIST OF ORIGINAL PUBLICATIONS
The present thesis is based on the following papers, which are referred to in the text by
their Roman number.
I. Don M, Fasoli L, Paldanius M, Vainionpää R, Kleemola M, Räty R, Leinonen M, Korppi

M, Tenore A and Canciani M.
Aetiology of Community-Acquired Pneumonia: Serologic Results of a Paediatric Survey.
Scandinavian Journal of Infectious Diseases 2005;37(11):806-812.
II. Don M, Valent F, Korppi M, Falleti E, De Candia A, Fasoli L, Tenore A and Canciani M.
Efficacy of serum procalcitonin in evaluating severity of community-acquired
pneumonia in childhood.
III. Fasoli L, Paldanius M, Don M, Valent F, Vetrugno L, Korppi M and Canciani M.

Simkania negevensis in community-acquired pneumonia in Italian children.
IV. Don M, Korppi M, Valent F, Vainionpää R and Canciani M.

Human metapneumovirus pneumonia in children: results of an Italian study and mini-
review.
V. Don M, Söderlund-Venermo M, Valent F, Lahtinen A, Hedman L, Canciani M, Hedman

K and Korppi M.
Serologically verified human bocavirus pneumonia in children.
Submitted.
7
ABBREVIATIONS
CAP community-acquired pneumonia

CF complement fixation
CRP C-reactive protein
DNA deoxyribonucleic acid
EIA enzyme immunoassay
ELISA enzyme-linked immunosorbent assay
ESR erythrocyte sedimentation rate
HBoV human bocavirus
Hib type B Haemophilus influenza
hMPV human metapneumovirus
IFA immunofluorescence assay
LRTI lower respiratory tract infection
MIF microimmunofluorescence method
PCR polymerase-chain-reaction
PCT procalcitonin
ROC receiver operator characteristic (curve)
RRIs recurrent respiratory infections
WBC white blood cells
WHO World Health Organization
8
ABSTRACT
BACKGROUND: Microbe-specific diagnosis of community-acquired pneumonia
(CAP) in childhood is difficult in clinical practice. Chest X-rays and non-specific
inflammatory markers have been used to separate presumably bacterial from viral
infection but the results have been inconsistent. Serological methods are routinary in the
diagnosis of certain viral and atypical bacterial respiratory infections. Simkania negevensis
is an atypical bacterium recently found to be associated with respiratory infections in
children and pneumonia in adults. In the last 25 years, serological methods have been
applied also to typical bacteria, like Streptococcus pneumoniae. Current knowledge on two
newly identified viruses, human metapneumovirus (hMPV) and human bocavirus (HBoV),
which are also involved in pediatric lower respiratory tract infections, is mainly based on
studies performed in respiratory samples by polymerase chain reaction.
AIMS: The aim of this prospective study is to determine, by using antibody assays,
the etiology of pediatric CAP in both ambulatory and hospitalized children younger than 16
years, with a special focus on the role of S. negevensis, hMPV and HBoV. The sero-
positivity rates to these three pathogens were also evaluated in Northern-Italian children.
In addition, the usefulness of procalcitonin (PCT) in assessing the severity of CAP, as well
as its ability to discriminate the etiology of CAP, was studied.
METHODS: During a 15-month study period, 124 children were admitted for
presumptive pneumonia, and in 101 of them pneumonia was radiologically confirmed.
Twenty-seven children were treated as inpatients and 74 as outpatients. The pulmonary
infiltrate was considered to be alveolar in 63 and interstitial in 38 cases, according to the
radiologic diagnosis. The bacterial and viral etiology of CAP was studied by an extensive
serologic test panel to 16 microbes, including microimmunofluorescence to S. negevensis,
9
an enzyme immunoassay (EIA) to hMPV and a newly developed EIA to HBoV. Serum
PCT was measured by an immunoluminometric assay in 100 CAP children.
RESULTS: A potential causing agent was detected in 75 patients (74%) . Evidence
of viral, bacterial and mixed viral-bacterial infection was demonstrated in 49%, 46% and
22% of the CAP cases, respectively. Mycoplasma pneumoniae (27%), S. pneumoniae
(18%) and respiratory syncytial virus (RSV) (17%) were the most commonly found agents.
The age distribution of pneumococcal infections was rather even. Serological evidence of
acute infection due to S. negevensis, hMPV and HBoV was achieved in 5 (5%), 5 (5%)
and 12 (12%) cases, respectively. HBoV was the second most common virus, and 5 (42%)
were mixed infections. In all, 20-30% of the children had measurable antibodies to S.
negevensis, with no association with age. The sero-positivity rate to hMPV and HBoV
increased with age, reaching nearly 100% at school age. No differences were found in
PCT concentrations between the four etiological (pneumococcal, atypical bacterial, viral,
unknown) and the three (<2, 2–4 and ≥5 years) age groups. PCT was higher in inpatients
than in outpatients (medians 17.81 vs. 0.72 ng/mL respectively, p<0.01) and higher in
alveolar than in interstitial pneumonia (medians 9.43 vs. 0.53 ng/mL respectively, p<0.01).
CONCLUSIONS: Our results confirm the importance of S. pneumoniae in pediatric
CAP at all ages. The high proportions of mycoplasmal and mixed viral-bacterial infections
highlight the need to consider antibiotics in all children suffering from CAP and to monitor
clinical responses. S. negevensis and hMPV seem to be real but rare causes of CAP in
children, while HBoV may be considered a fairly common cause. Sero-conversion to
hMPV and HBoV in most children takes place in early childhood. Serum PCT values are
related to the severity of CAP in children even though they do not seem to be capable, at
any serum concentration, to differentiate between bacterial and viral etiology.
10
TIIVISTELMÄ
JOHDANTO: Lasten avosyntyisen keuhkokuumeen aiheuttajan mukainen diagnostiikka on
vaikeaa kliinisessä työssä. Keuhkokuvaa ja tulehduskokeita on käytetty erottamaan bakteerien
ja virusten aiheuttamia keuhkokuumeita, vaikka tutkimustulokset ovat olleet ristiriitaisia. Vasta-
ainetutkimukset kuuluvat rutiinitutkimuksiin tiettyjen virusten ja atyyppisten bakteereien
diagnostiikassa. Simkania negevensis kuuluu atyyppisiin bakteereihin, ja sen on havaittu
liittyneen lasten hengitystieinfektioihin ja aikuisten keuhkokuumeisiin. Viimeisten 25 aikana
vasta-aine- ja antigeenitutkimuksia on sovellettu tavallisiin hengitysteiden bakteereihin, kuten
pneumokokkiin (Streptococcus pneumoniae). Nykyinen käsitys kahden vastikään löydetyn
viruksen, human metapneumoviruksen (hMPV) ja human bocaviruksen (hBoV), jotka näyttävät
liittyvän etenkin lasten hengitystieinfektioihin, perustuu pääosin tutkimuksiin, joissa on
sovellettu geenimonistustestiä (PCR, polymerase chain reaction) hengitysteistä otettuihin
näytteisiin.
TUTKIMUKSEN TARKOITUS: Tämän prospektiivisen tutkimuksen tarkoitus oli selvittää,
käyttämällä vasta-ainetutkimuksia, lasten avosyntyisen keuhkokuumeen aiheuttajia sekä
polikliinisesti että sairaalassa hoidetuilla eri-ikäisillä lapsilla, ja erityinen huomio kiinnitettiin
kolmeen uuteen aiheuttajaan (S. negevensis, hPMV ja hBoV). Seropositiivisuus näille kolmelle
mikrobille pyrittiin myös arvioimaan pohjois-italialaisilla lapsilla. Lisäksi arvioitiin seerumin
prokalsitoniinin (PCT) merkitystä sekä keuhkokuumeen vaikeuden arvioinnissa että bakteeri-
ja virusetiologian erottamisessa.
AINEISTO JA MENETELMÄT: Viidentoista kuukauden aikana 124 lasta lähetettiin
poliklinikalle todennäköisen keuhkokuumeen takia, ja heistä keuhkokuume varmistettiin
keuhkokuvalla 101 lapsella. Näistä 27 hoidettiin sairaalassa ja 74 kotona. Varjostuma
keuhkokuvassa oli alveolaarinen 63 ja interstitiaalinen 38 tapauksessa. Virus- ja
bakteerietiologia tutkittiin laajalla testipaketilla 16 mikrobia vastaan, mukaan lukien
11
immunofluoresenssiin perustuva menetelmä S. negevensis -vasta-aineille ja uudet entsyymi-
immunologiset menetelmät hMPV- ja HboV -vasta-aineille. Seerumin PCT mitattiin
immunoluminometrisellä menetelmällä 100 lapsen seerumista.
TULOKSET: Todennäköinen aiheuttaja löytyi 75 (74%) potilaalta. Näyttöä virusinfektiosta,
bakteeri-infektiosta ja virus-bakteeri-sekainfektiosta saatiin 49 %:ssa, 46 %:ssa ja 22 %:ssa
keuhkokuumeista. Mycoplasma pneumoniae (27%), S. pneumoniae (18%) ja respiratory
syncytial virus (RSV) (17%) olivat yleisimmin löytyneet aiheuttajat. Pneumokokin aiheuttamat
keuhkokuumeet jakautuivat tasaisesti iän suhteen. Serologinen näyttö siitä, että S.
negevensis, hMPV tai HBoV oli aiheuttajana, saatiin 5 (5%), 5 (5%) and 12 (12%)
tapauksessa. HBoV oli toiseksi yleisin virus, ja 5 (42%) tapausta oli sekainfektioita. Kaiken
kaikkiaan, 20-30 %:lla lapsista oli mitattavia vasta-aineita S. negevensis -bakteeria vastaan,
eikä mitään yhteyttä ikään havaittu. Sero-positiivisuus hMPV- ja HBoV-viruksia vastaan
lisääntyi suhteessa ikään ja oli lähes 100 % kouluiässä. Seerumin PCT ei eronnut
merkittävästi neljässä etiologisessa ryhmässä (pneumokokin, atyyppisten bakteerien tai
virusten aiheuttamat infektiot, tai infektiot joiden etiologiaa ei löytynyt), eikä myöskään neljässä
ikäluokassa (<2, 2–4 and ≥5 vuotta). PCT oli korkeampi sairaalassa hoidetuilla kuin kotona
hoidetuilla (mediaani 17.81 vs. 0.72 ng/mL, p<0.01) ja korkeampi alveolaarisessa kuin
interstitiaalisessa keuhkokuumeessa (mediaani 9.43 vs. 0.53 ng/mL, p<0.01).
JOHTOPÄÄTÖKSET: Tutkimuksen tulokset vahvistavat pneumokokin merkitystä
avosyntyisessä keuhkokuumeessa kaikenikäisillä lapsilla. Lisäksi mykoplasman sekä virusten
ja baakteerien aiheuttamien sekainfektioiden yleisyys korostaa, että kaikki lapset joilla on
keuhkokuume, tulee hoitaa antibiootreilla, ja että hoidon vaikutusta on syytä seurata. S.
negevensis ja hMPV näyttävät olevan todellisia mutta harvinaisia lasten keuhkokuumeen
aiheuttajia, kun taas HboV saattaa olla melko yleinen aiheuttaja. Vasta-aineiden
muodostuminen hMPV- ja HBoV-viruksia vastaan tapahtuu ennen kouluikää. Seerumin PCT
on yhteydessä keuhkokuumeen vaikeuteen, mutta PCT:n avulla ei voida millään seerumin
pitoisuudella erottaa bakteerin aiheuttamaa keuhkokuumetta viruksen aiheuttamasta.
12
RIASSUNTO
INTRODUZIONE: la diagnosi eziologica delle polmoniti pediatriche acquisite in comunità
(CAP) è sempre difficile nella pratica clinica. In passato i reperti radiografici del torace ed i
valori degli indici di flogosi sono stati usati per cercare di differenziare l’eziologia batterica da
quella virale, ma i risultati sono stati inconcludenti. Le metodiche diagnostiche sierologiche per
alcuni tipi di infezione respiratoria virale e da germi atipici sono routinarie. La Simkania
negevensis è un germe atipico recentemente associato alle infezioni respiratorie pediatriche
ed alle polmoniti in età adulta. Negli ultimi 25 anni, tali metodiche sierologiche sono state
applicate anche ai germi tipici, quali lo Streptococcus pneumoniae. Le attuali conoscenze su
due virus di recente identificazione, quali il metapneumovirus umano (hMPV) ed il bocavirus
umano (HBoV), sono basate principalmente su studi eseguiti su campioni respiratori analizzati
tramite polymerase chain reaction.
SCOPI: scopo di questo studio prospettico è stato quello di determinare, tramite saggi
anticorpali, l’eziologia delle CAP in pazienti italiani, di qualunque età pediatrica esclusa quella
neonatale, trattati a regime di ricovero e a livello ambulatoriale, con particolare attenzione al
ruolo della S. negevensis, di hMPV e HBoV. Sono stati anche valutati i tassi di sieropositività
verso questi tre agenti patogeni e la capacità della procalcitonina (PCT) di valutare la severità
delle CAP oltre la sua abilità a discriminarne l’eziologia.
METODI: in un periodo di 15 mesi, sono stati arruolati 124 bambini con verosimile polmonite.
Dei 101 casi in cui si è avuta una conferma radiologica, 63 presentavano un infiltrato
alveolare, gli altri 38 un infiltrato interstiziale. Ventisette bambini sono stati ricoverati, gli altri 74
sono stati gestiti ambulatorialmente. L’eziologia batterica e virale delle polmoniti è stata
studiata mediante un pannello sierologico esteso a 16 microorganismi, inclusivo di
microimmunofluorescenza per la S. negevensis, saggio immunoenzimatico (EIA) per hMPV e
un nuovo EIA per HBoV. La PCR sierica è stata misurata tramite un saggio
immunoluminometrico in 100 bambini con CAP.
13
RISULTATI: un agente patogeno è stato riscontrato in 75 pazienti (74%), con evidenza di
infezioni virali, batteriche e miste virali-batteriche nel 49%, 46% e 22% dei casi di CAP,
rispettivamente. Mycoplasma pneumoniae (27%), S. pneumoniae (18%) e virus respiratorio
sinciziale (RSV) (17%) sono stati i microorganismi più spesso evidenziati. Le infezioni
pneumococciche sono risultate comuni a qualunque età. E’ stata raggiunta un’evidenza
sierologica di infezione acuta da S. negevensis, hMPV e HBoV in 5 (5%), 5 (5%) e 12 (12%)
casi, rispettivamente. L’HBoV è risultato essere il secondo virus più commune, coinvolto nel
42% dei casi in infezioni miste. In totale, il 20-30% dei bambini aveva titoli anticorpali
misurabili nei confronti della S. negevensis, a prescindere dall’età. I tassi di sieropositività per
hMPV e HBoV sono invece aumentati all’aumentare dell’età, raggiungendo un valore del
100% dopo i 6 anni di vita. Non sono state riscontrate differenze nei valori di concentrazione
sierica della PCT tra i quattro gruppi eziologici (pneumococcico, germi atipici, virale,
sconosciuto) ed i tre gruppi suddivisi per età (<2, 2–4 e ≥5 anni). La PCT è risultata più elevata
nei pazienti ricoverati rispetto a quelli gestiti ambulatorialmente (mediana pari a 17.81 contro
0.72 ng/mL rispettivamente, p<0.01) e più elevata in quelli con infiltrato radiologico di tipo
alveolare rispetto ai pazienti con infiltrato di tipo interstiziale (mediana pari a 9.43 contro 0.53
ng/mL rispettivamente, p<0.01).
CONCLUSIONI: i nostri risultati confermano l’importanza dello S. pneumoniae nelle
CAP a tutte le età pediatriche. L’alta percentuale di infezioni da micoplasma e di infezioni
miste virali-batteriche sottolinea la necessità di impostare una terapia antibiotica in tutti i
bambini con CAP, monitorizzandone attentamente la risposta clinica. S. negevensis e hMPV
sembrano essere cause infrequenti ma reali di CAP in età pediatrica, mentre il HBoV può
esserne considerata una causa abbastanza frequente. La sieroconversione verso hMPV e
HBoV capita precocemente nell’infanzia in molti bambini. I valori sierici di PCT sono correlati
alla severità delle CAP pediatriche anche se non sembrano essere in grado, a nessuna
concentrazione, di differenziare tra eziologia batterica e virale.
14
INTRODUCTION
Acute respiratory infections represent the most common disease during childhood in
both developing and developed countries (Zar et al. 2003) and although the majority of
these infections are limited to the upper respiratory tract (Monto 2002), lower respiratory
tract infections (LRTI) are the main cause of morbidity, mortality and hospitalization in
infancy all over the world (Williams et al. 2002). Among LRTI, community-acquired
pneumonia (CAP) is still one of the most important causes of mortality in children,
especially among those under the age of 5 in developing countries, where incidence rates
are up to 10 times greater than in developed countries. Even among children of developed
countries, CAP is an important public health problem (Stein and Marostica 2007), due to
the need of rationalizing the use of antibiotics, the continuous fight against resistant
bacteria and the unavoidable complications, the endless developement of new vaccines
and new diagnostic tools, in order to constantly improve the management of this infectious
disease.
Viruses are the most common causes of pneumonia in young children, but the
prevalence decreases with increasing age (Michelow et al. 2004). Respiratory syncytial
virus (RSV) is the highest prevalent viral agent affecting the lower airways and may cause
pneumonia and present with pulmonary infiltrates, not necessarily indicating pneumonia
(Korppi 1999). Other viruses like influenza A-B, parainfluenza 1-2-3, rhinovirus, adenovirus
and cytomegalovirus are also often implicated in the etiology of CAP in children (Zar
2004). Streptococcus pneumoniae and Haemophilus influenza type b (Hib) are the typical
bacteria with the most important clinical impact. Nowadays, S. pneumoniae remains the
main agent causing pneumonia not only in children under the age of 5, but at all ages
(McCracken 2000b, Stein and Marostica 2007). On the other hand, the role of Hib
15
decreased in recent years, because the widespread use of anti-Hib vaccines in developed
countries had a great impact on CAP due to the same microorganism (Vuori-Holopainen et
al. 2002). The atypical bacteria, such as Mycoplasma pneumoniae (M. pneumoniae) and
Chlamydia pneumoniae (C. pneumoniae), tend to be more prevalent in older children and
adolescents (McCracken 2000b, Stein and Marostica 2007).
The etiology of respiratory infections often remains unknown (20-80%) (Jennings et
al. 2004, British Thoracic Society 2002, Louie et al. 2005), thus suggesting that unknown
respiratory pathogens have existed for a long time, even though the negative findings and
the variability in agent-specific incidence rates also depend on methodological differences
of the published studies. Determining the etiology of these diseases is a problem in
pediatric age, because obtaining adequate samples for the microbiological detection is not
often simple and many diagnostic tests, such as cellular cultures or viral antigen detection
in respiratory samples obtained by a non-invasive way, are not always sensitive enough to
identify the etiological agent (Jennings et al. 2004, Arnold et al. 2006, McCracken 2000b);
this stresses the importance of epidemiological data.
Recent progress in epidemiological research and in molecular biology, as the
introduction of polymerase chain reaction (PCR) performed on samples from the upper
respiratory tract (URT) in which viruses and atypical bacteria are not commensal (Korppi
et. 1992), have permitted the quick recognition and identification of many respiratory
emerging pathogens. Therefore, some of these, as human metapneumovirus, coronavirus-
NL63 and coronavirus-HKU1 have circulated unidentified for decades and are more
properly defined as “emerging” rather than “new” pathogens. On the other hand, other
viruses, such as SARS-coronavirus, avian influenza virus H5N1 and human bocavirus
seem to have a recent zoonotic origin, and are, therefore, considered as “new” pathogens.
In contrast, typical bacteria are common commensals of the URT of healthy children
16
(Ghaffar et al. 1999), thus other approaches such as specific antibody assays have been
tested as aids for the etiologic diagnosis of pneumonia.
In the present study, a prospectively collected patient material, including more than
one hundred consecutive cases of childhood CAP diagnosed in an Italian university
pediatric ward during a 15-month study period, was available for an etiological analysis.
This analysis firstly focused on well-known agents in the etiology of pediatric pneumonia,
then deepened the focus on three individual “emerging” and “new” viral and bacterial
pathogens, such as Simkania negevensis, human metapneumovirus and human
bocavirus. In addition to the conventional serological methods, newly developed methods
were used for the diagnosis of these three pathogens. Serum procalcitonin (PCT)
concentrations were also prospectively measured and compared between the etiological
groups in order to seek a potential association of this non-specific serum inflammatory
marker with the etiology of CAP.
17
COMMUNITY-ACQUIRED PNEUMONIA IN THE
PEDIATRIC AGE: LITERATURE REVIEW
DEFINITION
Community-acquired pneumonia (CAP) in the pediatric age is a disease clinically
characterized by the presence of signs and symptoms of pneumonia due to an infection
acquired outside the hospital. In developed countries this infectious disease may be
diagnosed by the radiographic detection of a parenchymal consolidation; in developing
countries a more practical term – lower respiratory tract infection (LRTI) – is used because
of the difficulty to obtain chest X-rays (British Thoracic Society 2002). The World Health
Organization (WHO) defines pneumonia solely on the basis of clinical signs and
respiratory rate (Authors not listed 1981).
EPIDEMIOLOGY
Incidence of pneumonia decreases when children grow up. The incidence of CAP in
Finnish children studied during 1981-1982 was 36/1000/year in <5-year-old children and
about 16/1000/year in the 5-14 years age group. A strong male prevalence was found
below 5 years of age (11.2/1000 among males and 5.7/1000 among females), and the
mortality rate was about 0.1/1000/year (Jokinen et al. 1993). The estimates were similar in
the american pediatric population (1483 CAP episodes studied in 1964-1975): 40
cases/1000/year in the age group of 6 months to 5 years, and 22 cases/1000/year in 5-9
years, 11 cases/1000/years in 9-12 years and 7 cases/1000/year in 12-15 years old
children (Murphy et al. 1981). The overall observed hospitalization rate for pediatric
18
pneumonia is about 1-4/1000 per year (Farha and Thomson 2005) and it coincides with
that of 14.7/10000 calculated for the North European pediatric population, being
32.8/10000 for children under five and 42.1/10000 for children under two (Senstad et al.
2009). In developing countries pneumonias are not only more common than in Europe and
North America (Figure 1) (Riley et al. 1983, Berman and McIntosh 1985, Selwyn 1990,
Rudan et al. 2008), but they are also more severe and characterized by a higher mortality
rate in the pediatric population (Figure 2) (Bulla and Hitze 1978, Baqui et al. 1998, Rudan
et al. 2008).
Figure 1. Differences in incidence of pediatric clinical pneumonia in the world at country

levels (according to Rudan et al. 2008).
19
Figure 2. The proportion of pneumonia-attributed deaths varies widely between WHO
regions and significantly increases in relative importance. The African Region has, in
general, the highest burden of global child mortality (according to Rudan et al. 2008).
RISK FACTORS
The only case-control study on risk factors for CAP in developed countries is the
Finnish study (Heiskanen-Kosma et al. 1997) which found that the risk factors for CAP <5-
year-old children were recurrent respiratory infections (RRI) during the previous 12 months
and the history of episodes of wheezing or otitis media treated by tympanocentesis during
the first 2 years of life. For the 5 to 15 age group, such risk factors were RRIs during the
previous year and a positive history for wheezing episodes. In temperate climates,
pneumonia is more common in cold months, presumably reflecting not only an enhanced
person-to-person droplet spread of respiratory pathogens due to crowding but also a
diminished host resistance due to impaired mucociliary clearance from dry indoor and cold
outdoor air (Durbin and Stille 2008). Children from low socioeconomic levels, those who
20
suffer from underlying chronic disease (e.g. sickle cell disease, bronchopulmonary
dysplasia, gastroesophageal reflux, asthma, cystic fibrosis, congenital heart diseases,
immunodeficiency syndromes, neuromuscular diseases, seizure disorders) and those who
are exposed to cigarette smoking, are at higher risk for acquiring pneumonia (Durbin and
Stille 2008).
Other risk factors for CAP, typical for developing countries and explaining the higher
incidence found in these countries, are shown in Table 1 (Rudan et al. 2008).
Table 1. Risk factors related to the host and the environment that impact on the incidence
of childhood clinical pneumonia in the community in developing countries (according to
Rudan et al. 2008).
DEFINITE RISK FACTORS LIKELY RISK FACTORS POSSIBLE RISK FACTORS
Malnutrition (weight-for-age Parental smoking Low mother’s education
under 2 SD)
Low birth weight (≤ 2500 g) Zinc deficiency Day-care attendance
Non-exclusive breastfeeding Mother’s experience as a Rainfall (humidity)
(during the first 4 months of caregiver
life)
Lack of measles Concomitant diseases (e.g. High altitude (cold air)
immunization (within the first diarrhoea, heart disease,
12 months of life) asthma)
Indoor air pollution Vitamin A deficiency
Crowding Birth order
Outdoor air pollution
21
PATHOGENESIS
Pneumonia typically follows an upper respiratory tract infection. Microrganisms that
cause LRTI usually are transmitted by droplet spread directly from close personal contact.
Following initial colonization of the nasopharynx, organisms may be inhaled, leading to a
pulmonary focus of infection; less commonly, bacteremia results from the initial upper
airway colonization, with subsequent seeding of the lung parenchyma (Durbin and Stille
2008). Normal pulmonary host defence system consists of multiple mechanical barriers,
including saliva, nasal hairs, the mucociliary escalator, the epiglottis and the cough reflex.
Humoral immunity, including secretory immunoglobulin A (IgA) and serum IgG, defends
against pneumonia, and other airways constituents such as surfactant, fibronectin and
complement play roles in microbial killing. Phagocytic cells, including polymorphonuclear
cells and alveolar macrophages, play an important defensive role, and cell-mediated
immunity is important in the defence against certain pathogens, such as viral agents and
other intracellular organisms (Durbin and Stille 2008). Although the inflammatory response
based on such cells is necessary to reinforce innate immunity and to rid the lungs of
microbes, it contributes directly to lung injury and abnormal pulmonary function (Figure 3)
(Mizgerd 2008).
22
Figure 3. Inflammation cascade and genesis of lung injury in case of pulmonary
infection. Neutrophils are effector cells of innate immunity, killing microbes through
phagocytosis and neutrophil extracellular traps (NETs). Neutrophils also generate a
variety of immune mediators to direct immune responses, influencing other cells of
innate and adaptive immunity. Finally, neutrophils damage tissues, with products
such as proteases and reactive oxygen species injuring cells and digesting matrix.
TNF denotes tumor necrosis factor (according to Mizgerd 2008).
ETIOLOGY OF PEDIATRIC CAP
GENERAL CONSIDERATIONS
The determination of the etiology of CAP in the pediatric age, as shown in some
prospective studies on specific pathogens of pneumonia (Claesson et al. 1989, Isaacs
1989, Juven et al. 2000, Ruuskanen et al. 1992, Wubbel et al. 1999), is difficult for several
reasons summerized in Table 2 (British Thoracic Society 2002, McIntosh 2002).
23
Table 2. Factors that make the etiological diagnosis of CAP in children difficult.
 paucity of positive findings from blood or pleural cultures
 low specificity of antigenic tests (e.g. in urine samples)
 dependence of antibody response on age
 difficulty to obtain sputum samples from children
 scarce utility of culture of upper respiratory tract samples (normal flora includes
bacteria commonly responsible for pneumonia)
 invasive examinations (lung biopsy, bronchoalveolar lavage) are not feasible
The etiological findings in pediatric CAP studies derive from four important matters,
that are the study design (inclusive or not inclusive of specific epidemics, the ambulatory
and/or hospital setting, the inclusion and exclusion criteria, the possible epidemics of some
studied microbes), age distribution, the severity of disease and the available test panel for
pathogens (Korppi 2003).
The more tests are available, the more potential causes emerge (Juven et al. 2000,
McIntosh 2002). For example, a potential cause of CAP has been found only in 24% of
cases and mixed infections in 0.3% of cases by means of cultures of respiratory secretions
(Murphy et al. 1981), whilst the identification of an etiological agent has been
demonstrated in 85% of cases and mixed infections in 41% of cases, (that generally are
viral-bacterial infections) using antibody tests for bacteria and sensitive solid-phase
immunoassays and a polymerase-chain-reaction (PCR) assay for viruses (Juven et al.
2000). Anyway, in 20-60% of cases any pathogen is recognized (British Thoracic Society
2002). Two recent studies, that used advanced diagnostic tools for the etiological
identification of CAP, were able to find an etiologic agent in 43% (Wubbel et al. 1999) and
in 85% (Juven et al. 2000) of cases.
24
Serological methods belong to the hospital routine in the diagnosis of respiratory
infections caused by certain viruses and atypical bacteria (such as Mycoplasma
pneumoniae and Chlamydia pneumoniae) (British Thoracic Society 2002). During the last
15 years serological methods have also been applied to study the etiology of pediatric
CAP caused by typical bacteria (above all Streptococcus pneumoniae) (Juven et al. 2001).
To study the overall etiology of CAP in children, assays for viruses and for both typical and
atypical bacteria should be applied in the same study, and ambulatory as well as
hospitalized children of different ages should be included. After the pioneer study of
Claesson et al. from Sweden (Claesson et al. 1989), in which pneumococcal serology was
applied for the first time, most papers on the etiology of pediatric CAP, combining viral and
bacterial data, have been originated from Finland (Heiskanen-Kosma et al. 1998, Vuori et
al. 1998, Juven et al. 2000) and Texas (USA) (Wubbel et al. 1999), thus reflecting only two
western populations. The available serology-based data on overall etiology of CAP in
childhood (Heiskanen-Kosma et al. 1998, Vuori et al. 1998, Korppi et al. 1993), with the
exception of the data in hospitalized children of Juven et al. (Juven et al. 2000) and in
ambulatory children of Wubbel et al. (Wubbel et al. 1999), are from the 1980`s. There is a
need for newer data on current distribution of etiological agents in pediatric CAP including
other populations too.
ETIOLOGICAL CLASSIFICATION OF PEDIATRIC CAP
A high number of infective agents may cause CAP in childhood (Tables 3 and 4)
(McIntosh 2002).
25
Table 3. Common causes of CAP in otherwise healthy children (* in developing countries).
Respiratory syncytial virus (RSV), Influenza virus A or B,
VIRUSES parainfluenza viruses 1,2 or 3, adenovirus, rhinovirus, measles
virus*
ATYPICAL BACTERIA M. pneumoniae, C. trachomatis, C. pneumoniae
TYPICAL BACTERIA S. pneumoniae, S. aureus*, H. influenzae type b*,
M. TUBERCULOSIS*
Table 4. Uncommon causes of CAP in otherwise healthy children.
Varicella-zoster virus (VZV), coronavirus, enterovirus, cytomegalovirus
VIRUSES (CMV), Epstein-Barr virus (EBV), mumps virus, herpes simplex virus
(HSV), hantavirus
CHLAMYDIA C. pittaci
COXIELLA C. burnetii
S. pyogenes, anaerobic mouth flora, non-typable H. influezae, B.
pertussis, K. pneumoniae, E. coli, L. monocytogenes, N. meningitidis,

BACTERIA
Legionella, Pseudomonas pseudomallei, F. tularensis, Brucella
abortus, Leptospira
Coccidioides immitis, Histoplasma capsulatum, Blastomyces

FUNGI
dermatitidis
A significant number of CAP (8-40%) are mixed infections, in which many of the
above mentioned pathogens may be involved (British Thoracic Society 2002). Juven et al.
found mixed viral-bacterial infections in 30%, mixed viral-viral infections in 13% and mixed
bacterial-bacterial infections in 7% of cases (Juven et al. 2000). The same seasonal
distribution of respiratory infections by different viruses and bacteria frequently lead for
26
mixed infections. Anyway, the diagnosis of co-infection must be carefully evaluated, also
from a methodological point of view. For example, the positivity of PCR may be indicative
of the presence of viral genome in respiratory secretions for many weeks after the
resolution of infection, rather than an acute infection. In addition, materials are easily to be
contaminated in laboratory settings.
VIRAL ETIOLOGY
Viruses account for 14-35% of all CAP cases in the pediatric age, while the estimate
of the percentage of CAP caused by bacteria remains always difficult. Viral pneumonia are
common in children living not only in developed (Smyth 2002) but also in developing
countries (Nascimento-Carvalho et al. 2008). A number of viruses appear to be associated
with CAP, the predominant one being respiratory syncytial virus (RSV). Others isolated
include influenza and parainfluenza viruses, adenovirus, rhinovirus, measles, varicella
zoster virus, cytomegalovirus, Epstein-Barr virus, herpes simplex virus and enteroviruses
(British Thoracic Society 2002, Smyth 2002).
HUMAN METAPNEUMOVIRUS
Among viruses, hMPV is an example of an infectious agent which probably has
existed for a long time but has been only recently recognized. Initially found in 2001 in
Dutch children with respiratory infections (van den Hoogen et al. 2001), hMPV has been
successively evidenced also in the rest of the world. This virus can be identified by cultural
methods, PCR and serology, and its discovery was probably delayed due to its slow
growth on continuous cellular lines (van den Hoogen et al. 2001). hMPV infection typically
occurs during infancy, with an incidence of seropositivity that increases by age (van den
Hoogen et al. 2001). Retrospective studies, conducted by stored sera, allowed to establish
27
that specific antibodies against this virus were present in 1958 in 100% of >8-year-old
children (Crowe 2004).
In taxonomy, hMPV has been classified as the first human virus of the genus
Metapneumovirus, subfamily Pneumovirinae. hMPV shares with RSV many biological
properties, including the human being as an only natural host, the same seasonality (with
a peak between December and February in temperate regions) (Crowe 2004) and nearly
identical clinical and laboratory features (van den Hoogen BG et al. 2004). These two
viruses are not even distinguishable in electronic microscopy (Figure 4), because both of
them have a cytopathic effect in cell cultures. Two genotypes of hMPV exist (serotype A
and B) and the heterogeneity of its viral genome may be responsible of the incomplete
immunity and of repeated infections.
Figure 4. Electron micrograph of hMPV particles

(according to van den Hoogen et al. 2001).
In all, hMPV is responsible for 5-7% of respiratory infections in hospitalized children.
Also immunocompromised patients and the elderly are at risk of this infection (van den
Hoogen BG et al. 2003). Many studies have confirmed that <5-year-old children are the
28
most susceptible to hMPV infection and those aged less than 2 years of age are at high
risk of severe infections (van den Hoogen BG et al. 2004). The few studies performed on
sero-prevalence to hMPV have indicated that virtually all children are infected by 5-10
years of age (van den Hoogen et al. 2001, Ebihara et al. 2003, Leung et al. 2005).
hMPV infection may clinically present in many ways, from mild upper respiratory
tract infections to moderate clinical pictures characterized by flu-like symptoms with high
fever, myalgia and vomit, or to more severe infections as bronchiolitis or pneumonia. The
majority of hMPV infections are subclinical and the acquired infection does not protect
from re-infection, even if acquired infection likely produces a partial protection to more
severe diseases that affect only children during the first year of life as a primary infection
(Alto 2004, Williams et al. 2004). In asymptomatic subjects the positivity for hMPV is
extremely rare. In Table 5, the data deriving from the prospective studies after analyzing
the role of hMPV in pediatric LRTI including CAP, are summarized.
At present, the relationship between hMPV and bacterial infections is not clear:
bacterial pneumonias are not frequently reported as complications of hMPV infections,
although the number of studied patients is untill now low. Anyway, it seems that hMPV
causes pneumonia even more frequently than RSV which, on the contrary, more
frequently evolves in bronchiolitis (Døllner et al. 2004).
29
Table 5. The role of human metapneumovirus in lower respiratory tract infections in children.
Type of Methods for Clinical No. of No. of etiological % of all % of co-

Authors & year Setting Country
study hMPV diagnosis patients agents studied cases infection
Madhi et al. 2003 P RT-PCR LRTI 191 9V 7.1 17 I South Africa
Viazov et al. 2003 P RT-PCR ARI 63 2V 17.5 42 I Germany
Peiris et al. 2003 P RT-PCR ARI 587 8V 5.5 20 I Hong Kong
Boivin et al. 2003 P RT-PCR; culture ARI 208 4V 5.8 6.4 I Canada
Maggi et al. 2003 P RT-PCR LRTI 90 9V 25 54 I Italy
Williams et al. 2004 R RT-PCR LRTI 248 1V 20 N.A. I/O U.S.A.
Mullins et al. 2004 P RT-PCR ARI 641 7V 3.9 11.8 I U.S.A.
Konig et al. 2004 P RT-PCR; culture LRTI 620 8V 0 N.A. I/O Germany
Lin et al. 2005 P RT-PCR CAP 116 8V3B 5.2 13.3 I Taiwan
Kim et al. 2005 R RT-PCR LRTI 166 8V 15.7 44.8 I Korea
Choi et al. 2006 P RT-PCR LRTI 515 11 V 4.7 7.7 I/O Korea
P: prospective; R: retrospective; LRTI: lower respiratory tract infection; ARI: acute respiratory infections; CAP community-acquired pneumonia. RT-
PCR: reverse transcription-polymerase chain reaction; V: virus; B: bacteria; I: in-patients; O: out-patients; N.A.: not applicable.
HUMAN BOCAVIRUS
HBoV was discovered in 2005, by “molecular virus screening”, a procedure based
on DNase treatment of the samples, random amplification and cloning, followed by large
scale sequencing and bioinformatics analyses (Allander et al. 2005). HBoV has been
recognized in biological samples from nasopharynx, blood and stools, evidencing the
possibility it could induce not only respiratory but also intestinal and systemic disease.
Rarely, HBoV virions have been observed by electron microscopy (Figure 5).
Figure 5. Electron microscopy pictures of HBoV in three different samples from

nasopharyngeal aspirates positive for HBoV by PCR (a and d, b and e, and c and f)
after negative staining with 4% uranyl acetate, observed at 60000 (a and b), 100000
(c), and 150,000 magnification (d, e, and f). The boxes outlined in white within
micrographs a, b, and c are magnified in micrographs d, e, and f, respectively. The
particles evocative of parvovirus capsids are hexagonally shaped and their centers
show heavier staining. The mean (standard deviation) size, calculated by measuring
45 particles across the apices at the 150000 magnification, are 25 (4) nm (according to
Brieu et al. 2007)
At present, the sole validated diagnostic method for HBoV is quantitative PCR,
which helped to understand that a high viral load in the respiratory tract seems to be
associated with the state of acute infection, while a low viral load is most likely linked to a
viral carrier condition (Allander et al. 2007, Kantola et al. 2008). Anyhow, respiratory
infections due to HBoV may be also serologically diagnosed (Kantola et al. 2008). The
reliance of PCR on not yet standardized sampling modalities at the respiratory tract and
the fact that it does not allow to distinguish reliably between acute infection and simple
asymtomatic viral persistence, make the interpretation of the real patogenicity of HBoV
difficult (Allander 2008). Moreover this virus has not been yet replicated in vitro, animal
models are not available, only preliminary experience exists on sero-epidemiology (Endo
et al. 2007, Kahn et al. 2008, Lin et al. 2008), and the studies based on healthy control
subjects are rather few (Kesebir et al. 2006). Thus, the modified Koch’s postulates
(Fredericks and Relman 1996) have not yet been confirmed for HBoV.
HBoV is a virus classified in the Parvoviridae family, Parvovirinae sub-family,
Bocavirus genus. HBoV is a parvovirus linked, even if remotely, to Parvovirus B19, which
is the etiological agent of “fifth disease” and some other diseases in human beings
(Allander 2008). Applying PCR on biological respiratory samples, the presence of HBoV
has been demonstrated in 19% of children suffering from wheezing (Allander et al. 2007)
and in 1.5 to 11.3% of children with other respiratory infections in North America, Europe,
Asia and Australia, suggesting a world wide circulation of the virus, without any precise
seasonality (Allander 2008). The variable data on prevalence and seasonality of HBoV
may depend on differences in the population studies, on sampling strategies (sampling
has often been performed to evaluate other viral epidemics) and on the different sensitivity
of the methods used (as those showed on PCR) (Allander 2008).
When HBoV is sought by PCR in respiratory samples, the percentage of mixed
infections varies form 34.6% to 76%, and HBoV is often involved in mixed viral infections
32
with RSV, thus drawing final conclusions on its patogenicity is rather difficult (Allander
2008). The highest percentage of mixed infections (83%) has been identified in HBoV
positive pneumonia cases studied in Thailand (Fry et al. 2007).
Till now, few studies based on IFA or ELISA methods have deepened the
knowledge on HBoV sero-epidemiology, confirming its capacity to stimulate the immune
system by B lymphocytes (Endo et al. 2007, Kahn et al. 2008, Lin et al. 2008). In general,
seropositivity rates for HBoV have resulted similar to those of hMPV in the pediatric age
(van den Hoogen et al. 2001), with percentages that increase with the age. These data,
along with those on viral prevalence, suggest that HBoV is an endemic virus, with a high
attack rate among children, likely followed by various immunity degrees, thus explaining
why almost 100% of population is infected in the first years of life (Allander 2008).
HBoV is more frequently identified in patients with respiratory symptoms than in
asymptomatic subjects (Allander et al. 2007, Kesebir et al. 2006, Fry et al. 2007), typically
in wheezing cases, but its DNA has been detected in blood samples (Allander et al. 2007)
as well as in stool samples from children with acute gastroenteritis (Vicente et al. 2007).
However, the current impossibility to fulfill Koch’s postulates does not allow us to consider
HBoV as a sure etiological agent of the disease.
In Table 6, the main data deriving from all the prospective studies that analyzed,
using various methods, the role of HBoV in pediatric LRTI are summarized.
33
Table 6. The role of human bocavirus in LRTI in children.
Methods
Type of Clinical No. of No. of etiological % of all % of co-
Authors & year for Setting Country
study diagnosis patients agents studied cases infection
HBoV
Bastien et al. 2007 R PCR URTI, LRTI 1265 * - 5.1 N.A . I/O Canada
Fry et al. 2007 P RT-PCR CAP 1168 ** 13 V 4.5 83 I Thailand
Maggi et al. 2007 R PCR ARD 200 9V 4.5 44 I Italy
Völz et al. 2007 P PCR LRTI 389 9V 2.8 36 I Germany
Allander et al. 2007 P RT-PCR wheezing 259 16 V 19 75 I Finland
Cilla et al. 2008 P PCR CAP 315 14 V 3 B 14.2 68.8 I/O Spain
Christensen et al 2008 P RT-PCR URTI, LRTI 376 14 V 3 B 12 78 I Norway
Longtin et al. 2008 R RT-PCR RTI 225 4V 13.8 71 I Canada
Zhang et al. 2008 P PCR LRTI 333 1V 4.8 18.7 I China
P: prospective; R: retrospective; (L)RTI: (lower) respiratory tract infection; ARD: acute respiratory disease; CAP community-acquired pneumonia.
RT-PCR: reverse transcription-polymerase chain reaction; V: virus; B: bacteria; I: in-patients; O: out-patients; N.A.: not applicable.
* respiratory specimens; ** all ages

BACTERIAL ETIOLOGY
S. pneumoniae is the most commonly detected bacterium in pediatric CAP (16-
37%) (British Thoracic Society 2002), followed by M. pneumoniae (4-39%) and C.
pneumoniae (0-20%) (Heiskanen-Kosma et al. 1999, British Thoracic Society 2002), and
also H. influenzae (5%) (Claesson et al. 1991) and M. catarrhalis (1.5%-3%) (Claesson et
al. 1994, Korppi et al. 1992b). Co-infections of typical with atypical bacteria are common,
thus both S. pneumoniae and M. pneumoniae should be taken into account when starting
antibiotics for children with CAP (Toikka et al. 2000b, Korppi et al. 2004).
SIMKANIA NEGEVENSIS
A recently discovered atypical bacterium is Simkania negevensis, previously called
“Simkania Z” (Kahane et al. 1999). It is an obligatorily intracellular bacterium and was
identified for the first time in Israel in 1993 (Kahane et al. 1993). S. negevensis belongs to
the Simkaniaceae family, in the order of Chlamydiales and it shares many characteristics
with bacteria of Chlamydia species but it also presents differences, regarding the growth
cycle (Figure 6) (Friedman et al. 2003, Kahane et al. 2002). The genomic identity is equal
to Chlamydia species in 80-87% (Kahane et al. 2002) and the sensitivity to antibiotics is
almost totally equal (sensitive to macrolides and tetracyclines and resistant to penicillins)
(Kahane and Friedman 2000).
S. negevensis may be identified by means of cultural methods, PCR and enzyme
immunoassays. At present, the determination of serum specific IgG by ELISA is the most
reliable screening method for the identification of past infection. On the other hand, the
diagnosis of acute infection is obtained by seeking the microorganism in clinical samples
by culture or PCR or by serology (significant seroconversion of IgG titre in paired sera or
detection of IgA or both). However ELISA is not a suitable method with very young children
(Friedman et al. 2003).

Figure 6. Vero-cell monolayers fixed and stained with May-
Grunwald/Giemsa at various times after infection with Simkania
negevensis. (a) Uninfected cells; (b) 2 days; (c) 5 days; (d) 13 days. Bar 50
mcm (according to Kahane et al. 2002).
Simkania is associated with LRTI in the pediatric age. The connection was identified
for the first time, in the pediatric age, in an Israeli bronchiolitis study in 15% of the patients
by IgA antibodies (Kahane et al. 1998). Thereafter, S. negevensis was isolated in 63.6% of
Canadian <6-month-old infants admitted for pulmonary infection: in 9% of cases it was the
only microorganism, while in 85.7% of cases it was involved in mixed infections (above all
with RSV). The high number of co-infections suggests the possibility that Simkania may
not be a real pathogen in case of low respiratory tract infections, but may only act as an
opportunistic agent (Greenberg et al. 2003, Friedman et al. 2003). In Table 7, the data
deriving from the prospective studies on the role of S. negevensis in pediatric LRTI are
summarized.
36
Table 7. The role of Simkania negevensis in acute LRTI in children.
No. of other
Methods for Clinical No. of % co-
Authors & year etiological % of all cases Setting Country
Simkania diagnosis patients infections
agents studied
Lieberman et al. 1997 Serology CAP 308 * 7B6V 2.6 50 I Israel
Kahane et al. 1998 PCR; culture; serology bronchiolitis 239 3V 25 37 O Israel
Greenberg et al. 2003 PCR bronchiolitis 22 2 B 10 V 63.6 85 I Canada
CAP, asthma,
Kumar et al. 2005 PCR; culture; serology 188 1B N.A. N.A. I/O U.S.A.
bronchiolitis,
Friedman et al. 2006 PCR; culture; serology bronchiolitis 34 ** - 15 § N.A. I UK
Kahane et al. 2007 PCR; culture; MEIA CAP 34 4V 91 N.A. I Israel
Heiskanen-Kosma et
MIF CAP 174 6B7V 10 67 I/O Finland
al. 2008
LRTI: lower respiratory tract infection; CAP community-acquired pneumonia; PCR: reverse transcription-polymerase chain reaction; MEIA:
membrane immunosorbent assay; MIF: microimmunofluorescence; V: virus; B: bacteria; I: in-patients; O: out-patients; N.A.: not applicable.
* adult patients; ** children aged 1-4 years; § percentage of serum IgG detection.
A recent Finnish study on case-control aimed at evaluating the association of
Simkania infection with the development of asthma in the pediatric age, demonstrated that
the presence of specific antibodies to S. negevensis were common among 1-6 year-old
boys, but there did not seem to be any association with future asthma onset (Korppi et al.
2006). Studies in adults are based on serology, which reflects the presence of an infection
more accurately than cultures or PCR, because many microorganisms may colonize the
respiratory tract without being necessarily involved in the infectious process (Friedman et
al. 2003). Besides causing respiratory infections, Simkania has also been found in the
cardiovascular system (Quinn and Gaydos 1999).
Rates of seropositivity to S. negevensis varied from 39% to 80% in adulthood.
Primary infection is contracted very early: sero-prevalence is about 30% in <2-year-old
children, in comparison to 2% among children infected by Chlamydia in the same age
group (Friedman et al. 2003).
ETIOLOGY BY AGE
A useful classification in clinical setting is based on age, which correctlypredicts the
likely etiology of CAP in children (Table 8) (British Thoracic Society 2002, Jadavji 1997). S.
pneumoniae remains the single most common bacterial cause across all age groups, and
RSV the commonest microorganism in pediatric CAP (Juven et al. 2000, Ruuskanen et al.
1992, Korppi et al. 1993, Chetty and Thomson 2007). CAP cases in the first three weeks
of age are rare and associated with perinatal infections. Between three weeks and three
months of age, Chlamydia trachomatis and Bordetella pertussis are frequently involved in
CAP (McIntosh 2002); viral infections, above all RSV, are more frequently found in
children from 4 months to 4 years of age (Claesson et al. 1989, Juven et al. 2000, Wubbel
et al, 1999, Korppi et al, 1993, Heiskanen-Kosma et al. 1998, Russel 2001), while M.
pneumoniae and C. pneumoniae are more typical in children over 5 years of age
(Claesson et al. 1989, Wubbel et al. 1999, Heiskanen-Kosma et al. 1998, Gendrel et al
1997, Harris et al. 1998), being detectable also in children aged <5 years (Principi et al.
2001). A graphic representation of the likeliest causes of CAP, according to age, is shown
in Figure 7.
Table 8. Microbiological causes of pediatric CAP, in reference to age (etiological agents
are shown in decreasing order of frequency) (according to McIntosh 2002).
AGE CAUSES
Group B streptococci, gram-negative enteric bacteria, CMV, Listeria

Birth - 20 days
monocytogenes
C. trachomatis, RSV, parainfluenza virus 3, S. pneumoniae, B.

3 weeks - 3 months
pertussis, S. aureus
RSV, parainfluenza viruses, influenza viruses, adenovirus, rhinovirus,

4 months - 4 years
S. pneumoniae, H. influenzae, M. pneumoniae, M. tuberculosis
5-15 years M. pneumoniae, C. pneumoniae, S. pneumoniae, M. tuberculosis
39
Mycoplasma
pneumoniae and
Streptococcus pneumoniae Chlamydia
pneumoniae
Haemophilus influenzae
0 3 6 9 12 15
Anni
Years
< 3 weeks 3 weeks – 3 months

B group Streptococcus Chlamidya trachomatis
Gram negative bacteria RSV / parainfluenza viruses
Cytomegalovirus Streptococcus pneumoniae
Listeria monocytogenes Bordetella pertussis
Staphylococcus aureus
Figure 7. Graphic representation showing the etiology of

pediatric CAP in relation to age (according to McIntosh 2002).
CLINICAL PICTURE
In <5-year-old children, tachypnoea defined as for the WHO criteria (respiratory rate
>50 breaths/min in <12-month-old infants, >40 breaths/min in 1-5-year-old children and
>30 breaths/min in ≥6-year-old children) (Mulholland et al. 1992) has had the highest
sensitivity (74%) and specificity (67%) for radiologically defined pneumonia, even if
sensitivity and specificity are low in the initial phase of the disease (less than 3 days of
duration) (Palafox et al. 2000). In addition, the severity of tachypnea correlates with the
severity of the disease, although pneumonia may present without tachypnea (Cherian et
al. 1988). Fever is another important sign, generally >38.5°C in infants and children living
40
not only in developed but also in developing countries (Campbell et al. 1989, Turner et al.
1987). Clinical symptoms may help more than clinical signs, and, therefore, dyspnea has
been considered to be included in the WHO guidelines for the diagnosis of pneumonia
instead of respiratory rate (Pereira and Escuder 1998).
Observational signs are more helpful than auscultatory ones (Margolis and
Gadomski 1998), and in accordance, rales and bronchial breathing have only a 75%
sensitivity and 57% specificity for CAP (Smyth et al. 1998). Wheezing does not belong to
the clinical picture of CAP (Harari et al. 1991), although it has been found in up to 30% of
CAP due to Mycoplasma (that may be confused with acute asthma) (Broughton 1986,
British Thoracic Society 2002). Cough, tachypnoea, flaring of the alae of the nose, chest
retractions, grunting, dullness of percussion and crackles on lung auscultation have been
recently noted as signs suggesting pneumonia on admission (Senstad et al. 2009).
Thus, the absence of all clinical signs (tachypnoea, auscultation and respiratory
effort signs) seems to exclude pneumonia (Jadavji 1997). Bacterial CAP should be
considered in <3-year-old children with fever >38.5°C along with chest recession and
respiratory rate > 50 breaths/min; in older children the difficulty in breathing is more helpful
than clinical signs. Clinical characteristics of bacterial, viral and mycoplasma LRTI are
shown in Table 9 (British Thoracic Society 2002). Anyway, there is a large overlapping of
clinical findings in viral and bacterial pneumonia. This similarity and the frequent
occurrence of mixed infections make it impossible to differentiate between viral and
bacterial pneumonia and between typical and atypical pneumonia simply by clinical
symptoms and signs (Principi and Esposito 2001, Principi and Esposito 2002, Esposito et
al. 2002, Juven et al. 2003).
41
Table 9. Clinical characteristics of bacterial, viral and mycoplasma LRTI (according to
British Thoracic Society 2002, modified).
Bacterial LRTI Viral LRTI Mycoplasma LRTI
Fever >38.5°C Infants and young children School chi ldren
Respiratory rate > 50 Wheezing Cough, wheezing, signs
breaths/min of pneumonia
Chest recessions Fever <38.5°C Interstitial infiltr ates,
Clinical and radiological Marked chest recessions lobar consolidation and
signs of consolidation hilar adenopathy
Hyperinflation
Respiratory rate normal or raised
Radiological signs of hyperinflation
Lobar collapse when severe
DIAGNOSIS
CHEST RADIOGRAPHY
Radiographic findings are poor indicators of the etiology of CAP (British Thoracic
Society 2002, McIntosh 2002). It has been demonstrated that chest radiography is too
insensitive to discriminate between bacterial and viral CAP (McCarthy et al. 1981, Courtoy
et al. 1989, Clements et al. 2000), even if Virkki et al. found that most children with
alveolar pneumonia, particularly those with lobar infiltrates, show laboratory evidence of a
bacterial infection, while interstitial infiltrates were seen in both viral and bacterial
pneumonia (Virkki et al. 2002). Moreover, pathognomonic radiological patterns lack for
CAP due to M. pneumoniae (Guckel et al. 1989). Radiological findings also have a limited
42
capacity to differentiate atypical infections (Esposito et al. 2001, Principi and Esposito
2001, Principi and Esposito 2002) and typical from atypical infections (Esposito et al.
2002) in children with CAP.
Although among radiologists there is a variation between intra- and inter-observer
agreement on the radiographic features used for diagnosis and a variation in how these
features are used in picture interpretation, the main finding of consolidation for the
diagnosis of pneumonia seems to be highly reliable (Davies et al. 1996), not very sensitive
but reasonably specific for bacterial CAP (74%) (Korppi et al. 1993b). Lateral X-rays are
unhelpful (Kiekara et al. 1996).
Thus, chest radiography should not be performed routinely in children with
uncomplicated acute LRTI (Virkki et al. 2005) and a radiological follow-up should be
considered only in case of lobar collapse, apparent round pneumonia or continuing
sympthoms, but not after the resolution of uncomplicated acute CAP in asymptomatic
patiens (British Thoracic Society 2002, Gibson et al. 1993, Heaton and Arthur 1998).
GENERAL INVESTIGATIONS
OXYGEN SATURATION (SaO2)
SaO2 should be measured in every child hospitalized for pneumonia. An appropriate
measure is obtained when the child is quiet, and the signal must be present and stable for
at least thirty seconds (British Thoracic Society 2002). In an African prospective study, a
significant increase of mortality for CAP was found when hypoxemia was present (Smyth
et al. 1998).
43
PROCALCITONIN AND OTHER NON SPECIFIC SERUM INFLAMMATORY MARKERS
Procalcitonin (PCT), the prohormone of calcitonin, is a 116-aminoacid protein
(Figure 8) (Le Moullec et al. 1984) recognized as a non-specific marker of bacterial
infection (Gendrel and Bohuon 2000), and its serum concentration seems to correlate with
the severity of microbe invasion (Assicot et al. 1993, Karzai et al. 1997). In fact, serum
PCT reaches much higher concentrations in invasive, bacteremic infections, such as
bacterial meningitis (Gendrel et al. 1997b, Hatherill et al. 1999, Gendrel et al. 1996,
Gendrel et al. 1999) or septicemia (Assicot et al. 1993, Gendrel et al. 1999, Chiesa et al.
1998) than in less invasive, localized, non-bacteremic infections, represented by most
cases of urinary tract infection or pneumonia (Gendrel et al. 1999).
Figure 8. Schematic description of the aminoacid sequence of PCT (according to

Le Moullec 1984).
44
Serum PCT has been used as a non-specific laboratory marker for distinguishing
between viral and bacterial infections and to help decide whether or not to prescribe
antibiotics. Earlier studies have suggested that serum PCT concentrations remain low in
viral infections (Assicot et al. 1993, Gendrel et al. 1999) and in autoimmune inflammatory
diseases (Eberhard et al. 1997) without superimposed bacterial infections, even if a study
on adults showed that PCT does not clearly discriminate severe infections from a systemic
inflammatory syndrome (Gendrel and Bohuon 2000, Assicot et al. 1993, Suprin et al.
2000). The authors of a recent systematic review concluded that PCT is superior to other
non-specific inflammatory markers, such as C-reactive protein (CRP), in differentiating
bacterial infections both from viral infections and non-infectious inflammatory diseases,
among patients hospitalized for suspected bacterial infectious diseases (Simon et al.
2004).
Clinical features, radiological findings and acute phase reactants have proved to be
poor indicators of the etiology of pneumonia in children (British Thoracic Society 2002,
McIntosh 2002). Microbe-specific diagnosis, based on blood culture, fluid or samples
obtained directly from the focus of the infection in the lungs, is possible only in a small
minority of CAP cases. In studies published over the last 15 to 20 years, serologic tests
based on antigen, antibody and immune complex detection have been used for microbe-
specific diagnosis of CAP in children (Heiskanen-Kosma et al. 2003, Heiskanen-Kosma et
al. 1998, Korppi and Leinonen 1998). In these studies, attention has also been paid to
serum non-specific inflammatory markers, such as CRP, erythrocyte sedimentation rate
(ESR) and white blood cell (WBC) count, in trying to distinguish between viral and bacterial
CAP in children but the results have been inconsistent (Nohynek et al. 1995, Korppi et al.
1997, Heiskanen-Kosma and Korppi 2000). These laboratory parameters were not
informative for the distinction among atypical infections (Principi et al. 2001, Esposito et al.
2001, Principi and Esposito 2001, Principi and Esposito 2002) and between typical and
45
atypical infections (Esposito et al. 2002) in children with pneumonia. Similarly, conflicting
results exist concerning serum PCT: some authors have neglected its role (Korppi and
Remes 2001, Korppi et al. 2003), some have concluded that PCT has no additional value
over CRP (Toikka et al. 2000), while others have stressed its role as a useful marker for
identifying bacterial CAP in basically healthy children (Moulin et al. 2001, Pratt et al. 2003).
In the pediatric age many acute LRTI may not be very invasive, mucosa being limited and
inducing a low host inflammatory response. On the other hand, few viral agents, such as
adenovirus or influenza virus, are able to cause invasive infections and systemic
responses that are very similar to those determined by bacterial infections. For this
reason, serum acute phase reactants are generally not able to distinguish between
viral and bacterial infections, they should not be measured routinely but only used to be
checked in case of lack of clinical treatment improvement (British Thoracic Society 2002).
SPECIFIC MICROBIOLOGICAL INVESTIGATIONS
TYPICAL BACTERIAL PNEUMONIA
Blood cultures should be obtained from all children showing a reasonable suspicion
of bacterial CAP, even if their positivity is less than 10% (British Thoracic Society 2002,
Russel 2001). Bacterial growth in nasopharyngeal sample is not indicative for the etiology
of LRTI (Jadavji 1997, Korppi et al. 1992).
When present in sufficient amount, pleural fluid should be sampled in order to
perform a microbiological and cultural analysis; more samples should also be kept for
likely further bacterial antigen investigations (British Thoracic Society 2002, Russel 2001).
In a study on 840 pleural fluid samples derived from <12-year-old children with clinical
46
signs suggestive for acute pneumonia, bacterial growth was found in 17.7% of all the
microbiological cultures (Requejo et al. 1997).
Recent studies on adult patients have found that the detection of pneumococcal
antigen in urine by immunochromatographic test may be particularly useful and valid (with
high sensitivity and specificity) for a quick diagnosis of pneumococcal pneumonia
(Gutiérrez et al. 2003, Domìnguez et al. 2001) The test is not useful in children, at least not
before school age. Studies performed in healthy children, who were nasopharyngeal
carriers of S. pneumoniae, demonstrated by immunochromatographic test the presence of
the pneumococcal urinary antigen in more than half of them. Thus, colonization seems to
be the major source of pneumococcal antigen in urine in children (Faden et al. 2002), and
a positive result of this test does not necessarily indicate an active pneumococcal disease
in children (Adegbola et al. 2001). Recent pediatric studies on CAP have underlined that
the specificity of this test is highly limited by nasopharyngeal colonization to be useful for
the diagnosis of pneumonia among children (Domìnguez et al. 2003, Dowell et al. 2001,
Ramsey et al. 1986).
In Finnish research laboratories, many assays have been developed and tested for
the serological diagnosis of pneumococcal infection based on the specific antibody and
immuno-complex detection in paired sera (Korppi et al. 1993c, Korppi and Leinonen 1998,
Heiskanen-Kosma et al. 2003, Korppi et al. 2008). Nevertheless, no test has showed a
sufficiently high sensitivity and specificity to be diagnostic alone, and many serological
tests needed to be used at the same time in order to identify pneumococcal cases (Korppi
et al. 1993c).
Diagnosis of pneumococcal LRTI may be also obtained by performing pneumolysin-
based PCR on blood and pleural fluid samples. This assay is particulary useful to identify
promptly children with pretreated or nonbacteremic pneumococcal LRTI (Michelow et al.
2002), it is the most sensitive one among all assays for the diagnosis of invasive
47
pneumococcal infections in children (Toikka et al. 1999), but it identified more frequently S.
pneumoniae in pleural fluid than in whole blood samples taken from chidren with
pneumococcal pneumonia and empyema (Lahti et al. 2006).
ATYPICAL BACTERIAL PNEUMONIA
PCR and serologic tests, the two primary methods for detecting M. pneumoniae and
C. pneumoniae in the clinical setting, may produce contrasting results, even if the
combination of PCR and serology may be necessary to achive maximum sensibility
(Dorigo-Zetsma et al. 1999, Smyth 2002). Positive PCR tests often occur in the presence
of negative serologic tests (Principi et al. 2001, Esposito et al. 2001) but, on the other
hand, a positive PCR for M. pneumoniae or C. pneumoniae in the upper respiratory tract
does not necessarily imply that it is the etiologic agent of the patient’s lower respiratory
disease. Positive PCR tests for M. pneumoniae or C. pneumoniae have been found also in
healthy children (McCracken 2000, Falck et al. 1997). Moreover, PCR for identification of
M. pneumoniae and C. pneumoniae is not available in all microbiological laboratories and
even though its positivity does not produce a definite diagnosis (McCracken 2000,
Ostapchuk et al. 2004), it may give a rapid diagnosis after testing a single specimen
(Principi and Esposito 2001, Principi and Esposito 2002).
Serologic tests for atypical bacteria often provide only a retrospective diagnosis,
they may take 4-9 weeks for a seroconversion and are more useful in establishing the
causing agent during an outbreak than in treating individual children (Principi and Esposito
2001). The two serological tests mainly used for the diagnosis of M. pneumoniae infection
are complement fixation and EIA. EIA tests are more sensitive and specific and they can
also reveal reinfections, because they measure class-specific IgM or IgG antibodies. MIF
which also measures both IgG and IgM, is the serological assay of choice for the diagnosis
of C. pneumoniae infections, but it is technically complex, its interpretation is subjective
48
and there are no standarized reagents or diagnostic criteria (Tsolia et al. 2004, Esposito et
al. 2001, Principi and Esposito 2001, Principi and Esposito 2002). Cultures for M.
pneumoniae and C. pneumoniae are not routinely recommended (Ostapchuk et al. 2004),
also because the specificity of the culture greatly depends on the subjective ability to
perform and interpret the assay (Principi and Esposito 2001).
VIRAL PNEUMONIA
Nasopharyngeal aspirate, replaceable by nasal lavage (Balfour-Lynn et al. 1995),
used for viral antigen assays (such as immunofluorescence), has high specificity for RSV,
parainfluenza virus, influenza virus and adenovirus, with a sensitivity equal to 80% (Korppi
et al. 1993c). Samples from nasopharyngeal aspirates or nasal washes may also be used
for viral culture and, above all, for the determination of viral genomic material by PCR-
based detection techniques (Tsolia et al. 2004), this fact increases the rate of diagnosis of
viral pneumonia from 13% to 31% (Clements et al. 2000). These viruses are identifiable
also by seroconversion in paired serum samples by means of serological assays, but other
important viruses which cause CAP in children, like rhinoviruses and enteroviruses, are
not. The results of viral tests are particularly useful to quantify the number of infected
children in case of outbreaks and for epidemiological purposes; nasopharyngeal aspirates
should be obtained from all children <18 months old for viral antigen detection with or
without viral culture (British Thoracic Society 2002, Russel 2001).
Table 10 summarizes all serological studies on the role of HBoV and S. negevensis
in pediatric LRTI. Untill now, the role of hMPV in pediatric LRTI has not yet been studied
by serology and only sero-epidemiological studies on hMPV are available (van den
Hoogen et al. 2001, Ebihara et al. 2003, Leung et al. 2005). Respiratory HBoV infection
has been studied by serology in only one study in wheezing children (Allander et al. 2007).
49
Table 10. Serological studies that analyze the role of HBoV and S. negevensis in LRTI in children. No serological studies are available on hMPV
pneumonia or even respiratory infection in children.
% of antibody positive
% of antibody
No. of Serological Specific Single or cases among the
Authors & year Pathogen Type of LRTI positive
patients assay antibodies paired sera PCR or culture
cases
positive cases
Kantola et al. 2008 HBoV wheezing 117 IB IgM, IgG P 51 * 59
Lieberman et al. 1997 Simkania CAP 308 ELISA IgA, IgG P 2.6 N.A.
Kahane et al. 1998 Simkania bronchiolitis 92 IPX IgA S/P 15 27
CAP, asthma,
Kumar et al. 2005 Simkania 188 ELISA IgG S 18 N.A.
bronchiolitis,
Friedman et al. 2006 Simkania bronchiolitis 34 ** ELISA IgA, IgG S 15 § N.A.
Heiskanen-Kosma et al.
Simkania CAP 174 MIF IgM, IgG P 10 N.A.
2008
LRTI: lower respiratory tract infection; CAP community-acquired pneumonia. S: single serum sample; P: paired serum samples; N.A. not
applicable. IB: immunoblot assay; ELISA: Enzyme-Linked Immunosorbent Assay; IPX: immunoperoxidase assay; MIF: microimmunofluorescence.
* percentage of serum IgG detection (28% of serum IgM detection); ** children aged 1-4 years; § percentage of serum IgG detection (3% of serum
IgA detection in the same age group).

AIMS OF THIS STUDY
The general aim of this thesis was to study, by using antibody assays, the etiology
of community-acquired pneumonia in a prospectively collected patient material, including
more than one-hundred consecutive ambulatory and hospitalized Northern-Italian children
of different ages.
The specific aims were:
1. to investigate the role of 16 microbes (6 bacteria and 10 viruses) in the etiology of
pediatric CAP;
2. to evaluate the role of S. negevensis in the etiology of pediatric CAP, to estimate the
sero-prevalence of S. pneumoniae and to assess the possible cross-reactivity between
S. negevensis and C. pneumoniae serology;
3. to analyze the role of hMPV, by using a newly developed antibody assay, in pediatric
CAP, and to estimate the sero-prevalence of hMPV ;
4. to analyze the role of HBoV, by using a newly developed antibody assay, in pediatric
CAP, and to estimate the sero-prevalence of HBoV;
5. to clarify the usefulness of serum PCT in assessing the severity of CAP and in
differential diagnosis of bacterial and viral CAP.

PATIENTS AND METHODS
STUDY DESIGN
This prospective observational study was conducted at the Department of
Pediatrics, School of Medicine DPMSC, University Hospital of Udine, Italy, during a 15
months period, from October the 1st, 2001 to December the 31st, 2002. The study period
covered one winter season with frequent infections, from 21 st December 2001 to 21st
March 2002. On admission, all patients underwent a clinical evaluation (including medical
history and physical examination), a blood sample and chest X-rays, interpreted as
described below. A blood culture was obtained in 35 children who had high fever or
appeared severely ill according to the physician on duty. Serum samples for serologic
assays, obtained on admission and on average 35 days later (range 18-56 days), were
stored at –20°C until transported to the reference laboratories and analysed
simultaneously.
PATIENTS
In all, 125 consecutive and previously healthy children, aged less than 16 years,
who arrived at the Pediatric Emergency Room (ER) of the University Hospital of Udine,
Italy, and were considered to have CAP, were eligible for the study. The University
Hospital of Udine serves a population of about 50.000 children in the province of Udine,
near the north-eastern Italian border; the patients either come to the ER based on the
decision of the parents or they are sent by a general paediatrician who usually manages at
his/her office the mild and uncomplicated cases only.
The inclusion criteria of the present study were signs and/or symptoms compatible
with respiratory infection (fever ≥38°C, tachypnea, cough and/or findings of crackles,
52
bronchial breathing or diminished/silenced breath sounds on auscultation) and radiological
infiltrations consistent with pneumonia.
Wheezing was considered to be an exclusion criterion because, in infants, it is more
likely a sign of bronchiolitis, wheezy bronchitis or asthma exacerbation than of pneumonia
(British Thoracic Society 2002). The other exclusion criteria included newborns, severe
and chronic underlying diseases (such as neoplasia, cardiovascular and respiratory
diseases, kidney and liver diseases, immunosuppression, mental retardation and
malnutrition) and hospital-acquired pneumonia.
Chest X-rays were taken on admission in all children with suspected pneumonia
and were interpreted by a senior radiologist who classified the findings into alveolar,
interstitial and no infiltrations. Only children with an infiltration consistent with pneumonia
were included. Seventy cases (69%) with alveolar and 31 (31%) with interstitial infiltrations
were analysed as combined. In 2005, three experienced radiologists, unaware of the
etiology and the interpretation of the initial radiographs, individually and independently
evaluated all the X-rays and the final diagnosis was considered when at least two of them
agreed. At that evaluation, 62% cases were interpreted as alveolar and 38% as interstitial
pneumonia.
CLINICAL DEFINITIONS
On inspection, tachypnoea was defined by the WHO criteria, i.e. respiratory rate
>60 breaths/minute in children aged <2 months, >50 breaths/minute in children aged 2-12
months and >40 breaths/minute in children aged >12 months (Palafox et al. 2000). On
auscultation, lung sounds were assessed as either normal or silenced and the rales or
crackles, whether fine, medium or coarse, were called together as crackles. Children with
wheezing were excluded and rhonci and ruttles were not registered nor included in the
analyses (Elphick et al. 2000).
53
BACTERIOLOGY
The serological assays for S. pneumoniae, H. influenzae, Moraxella catarrhalis and
C. pneumoniae were carried out at the National Public Health Institute, Oulu, Finland. For
S. pneumoniae infection, IgG antibodies to pneumococcal pneumolysin and
pneumococcal C-polysaccharide were measured by enzyme immunoassay (EIA) and a
two-fold or greater rise in antibody titers between paired sera was considered diagnostic
(Nohynek et al. 1995b). For H. influenzae and M. catarrhalis infections, Ig (polyvalent)
antibodies against whole bacterial cell antigens (a mixture of 10 different strains) of
nontypeable H. influenzae and M. catarrhalis were measured by EIA and a three-fold
or greater antibody rise between paired sera was considered diagnostic (Nohynek et al.
1995b).
In-house microimmunofluerescence (MIF) was used to measure IgG and IgM
antibodies to C. pneumoniae, using purified, formalized elementary bodies of strain K6 as
antigen (Wang 2000). The diagnosis was based on a four-fold or greater rise in IgG
antibodies between the paired sera or on the presence of IgM of ≥1:16 in any serum. C.
pneumoniae-specific IgG and IgM were also measured by EIA antibody kits
(AniLabsystems Ltd. Oy, Helsinki, Finland) based on an indirect solid-phase technique
with horseradish peroxidase as a marker enzyme (Persson and Boman 2000). In the
assay, 1.5-fold changes in enzyme immunounits (EIU) in the zones below 130 EIU in IgG
were considered as diagnostic. A 1.3-fold change was considered diagnostic when the EIU
values were more than 130 EIU in IgG. The IgM EIA results were expressed as a
signal/cut-off (S/CO) ratio, and a S/CO ratio of more than 1.1 was considered as positive,
0.5-1.1 as equivocal, and less than 0.5 as negative.
The serologic assays for M. pneumoniae were carried out at the Department of
Microbiology, National Public Health Institute, Helsinki, Finland. Complement fixation (CF)
test was performed by a standard micromethod using in-house antigen prepared according
54
to Kenny and Grayston (Kenny and Grayston 1965). A four-fold or greater increase in titers
between paired sera and high titers (1:128 or higher) in either serum without a four-fold
rise were considered diagnostic. The CF test was supplemented by measuring IgG and
IgM antibodies to M. pneumoniae by EIA, using the procedure recently published by
Korppi et al. (Korppi et al. 2004).
SIMKANIA NEGEVENSIS SEROLOGY
In 2005, serological studies were supplemented by MIF to Simkania negevensis in
frozen samples at the National Public Health Institute, Oulu, Finland. The culture of S.
negevensis strains and the consequent purification of S. negevensis antigen were done as
described earlier (Kahane et al. 2002), and specific antibodies to elementary bodies of S.
negevensis (ATCC, VR-1471) were measured by an in-house MIF (National Public Health
Institute, Oulu, Finland) (Korppi et al. 2006, Korppi et al. 2004b). The detection limits, used
in the assessment of sero-prevalence, were 1:8 for IgG and 1:10 for IgM antibodies. The
diagnostic criteria of S. negevensis infection were ≥4-fold or greater increases or
decreases in IgG or IgM antibodies between paired sera.
VIROLOGY
Conventional viral serology was performed at the Department of Virology, University
of Turku, Finland. IgG antibody titers against parainfluenza viruses type 1-3, influenza A
and B viruses and adenoviruses were determined by EIA using antigen-coated solid phase
and horseradish peroxidase-conjugated rabbit antihuman IgG (Dako, Glostrup, Denmark),
as described earlier (Vuorinen and Meurman 1989). A two-fold or greater increase in
antibody titers between paired sera was considered diagnostic. IgG and IgM antibodies
against cytomegalovirus were measured as described by Ziegler et al (Ziegler et al. 1989)
55
and IgG antibodies against RSV as described by Laine et al (Laine et al. 2003). For RSV a
four-fold or greater increase in IgG titers was considered diagnostic.
HUMAN METAPNEUMOVIRUS SEROLOGY
IgG antibodies to hMPV were determined by an in-house EIA at the Department of
Virology, University of Turku, Finland. The micro-titer plates were coated with hMPV-
infected cell lysate. Serum dilutions (1:40-1:2560) were added to the plates and incubated
for two hours at +37°C, after which horse-radish-pe roxidase conjugated anti-human IgG
antibodies (at dilutions of 1:12000 and 1:3000, Dako, Germany) were added. Ortho-
phenylene-diamine was used as a color substrate and the optical density was measured at
492 um by a Multiskan Analyzer (Eflab Oy, Helsinki, Finland). The last dilution which gave
an OD value higher than 2.5 times to OD value of negative controls was considered
positive. The detection limit, used in the assessment of sero-prevalence, was 1:40 for IgG
antibodies. The diagnostic criterion of hMPV infection was a four-fold or greater increase
or decrease in IgG antibodies between paired sera.
HUMAN BOCAVIRUS SEROLOGY
HBoV serology was performed at the Department of Virology, University of Helsinki,
Finland. HBoV IgG EIA was set up as described previously for parvorvirus B19 IgG EIA
(Kaikkonen et al. 1999). For IgM EIA, a µ-capture IgM EIA format was used (Kallio-Kokko
et al. 1998). Recombinant virus-like particles of VP2 were used as antigen in both HBoV
IgG EIA and µ-capture IgM EIA (Söderlund-Venermo et al. in press). The cut-off
absorbances for antibody presence (positive: mean of negative controls +4 SD) were
1.888 for IgG and 0.167 for IgM, and for the antibody absence (negative: mean of negative
controls + 3SD) 0.154 for IgG and 0.136 for IgM (Söderlund-Venermo et al. in press). The
diagnostic criteria of an acute HBoV infection were either the presence of IgM or a four-
56
fold or greater increase in the apparent concentration (determined by reference titration
curves) of IgG in paired sera.
NON-SPECIFIC LABORATORY MARKERS
Serum PCT was measured in serum samples obtained on admission. The blood
samples were centrifuged and the serum was separated and stored at –20°C until assayed
for PCT at the Laboratory Department, University of Udine, Italy. Serum PCT levels were
measured by immunoluminometric assay with 2 monoclonal antibodies (PCT LIA TEST,
Brahms Diagnostics, Henningsdorf, Germany), one against the catacalcin region of PCT
and the other against calcitonin, as described previously (Gendrel et al. 1996). By this
method the detection limit was 0.1 ng/mL and the upper limit for normal serum PCT
suggested by the manufacturer was 0.5 ng/mL. We applied this latter value, as well as the
values of 1.0 ng/mL suggested by Moulin et al (Moulin et al. 2001) and 2.0 ng/mL
suggested by Toikka et al (Toikka et al. 2000) and Prat et al (Pratt et al. 2003) as cut-off
limits for elevated PCT concentration.
In addition, serum CRP was measured by a particle-enhanced immunoturbidometric
assay (CRPLX reagent on Hitachi-Roche Modular P-System, Roche Diagnostics,
Mannheim, Germany) and the limit of 60 mg/L was used as the cut-off limit for an elevated
concentration (Korppi et al. 1997, Korppi and Remes 2001, Korppi et al. 2003, Moulin et al.
2001). ESR and WBC count were studied by routine methods and the cut-off limits used
were 30 mm/h for ESR and 15000 u/L for WBC (Korppi et al. 1997, Korppi and Remes
2001, Moulin et al. 2001).
The 100 patients with CAP and serum PCT determinations available were classified
into four etiological groups (Figure 9). The pneumococcal group consisted of all cases with
S. pneumoniae etiology, including sole pneumococcal infections and mixed infections with
viruses and with other bacteria. The atypical bacterial group consisted of cases with M.
57
pneumoniae, C. pneumoniae or S. negevensis etiology (viral coinfections included, but
pneumococcal coinfections not included). The viral group consisted of cases with sole viral
or dual viral etiology (pneumococcal coinfections and coinfections with atypical bacteria
not included). The cases with no serological findings formed the unknown etiology group.
Community acquired
pneumonia (N=125)
Enrolled patients
(N=101)
Eligible patients for

the PCT study
(N=100)
Pneumococcal Atypical group Viral group Unknown

group etiology group
Pneumococcal Mycoplasma, Viral cases, No serological

cases, mixed Chlamydia and mixed evidence of
infections with Simkania cases, infections with pneumococcal,
viruses and mixed infections with pneumococci atypical
other bacteria viruses included, or with atypical bacterial or
included. with pneumococci bacteria not viral infection.
not included. included.
Figure 9. Classification of the patients into four etiological groups for the PCT study.
STATISTICAL ANALYSIS
The statistical significances of the differences in categorical variables between the
three age groups (<2 years; 2-5 years; 5 years), and between the children treated as
58
inpatients and outpatients, were analysed by Chi-square or Fisher’s exact tests. Fisher’s
exact test was used in the statistical analysis of the S. negevensis, hMPV and HBoV data,
too. Two-tailed tests were used in all analyses, and the differences were considered as
statistically significant at the level of p <0.05.
Exploratory data analysis showed that PCT concentrations were non-normally
distributed. Therefore, the results are presented as medians and 25 th-75th percentiles. In
statistical analysis of the data, Wilcoxon and Kruskal-Wallis tests were used for continuous
variables and Chi-square and Fisher`s exact tests for discrete variables. The diagnostic
parameters of sensitivity, specificity, predictive value (PV) and likelihood ratio of a positive
test result (LR+) were calculated by using routine equations. The LR+ value of >3 is
considered to change moderately and of >5 significantly pre-test probability of a presumed
diagnosis. The receiver operator characteristic (ROC) curve was used to evaluate the
ability of serum PCT concentrations to discriminate pneumococcal and atypical CAP from
viral pneumonia. Multiple linear regression was used to evaluate the association between
PCT expressed as a continuous variable and etiological group, hospitalisation and
radiological group, after adjusting for age groups. The correlation between PCT
concentration and age was expressed by Spearman’s correlation coefficient.
ETHICS
The protocol was approved by the Ethics Committee of the School of Medicine
DPMSC, University of Udine, Italy and an informed consent was obtained from at least one
parent for all children.
59
RESULTS
DEMOGRAPHIC, LABORATORY AND RADIOLOGICAL DATA (I)
Of the initial 125 consecutively enrolled patients, 24 (19%) were excluded, 16
because convalescent serum samples were not obtained and 8 because chest X-rays
were not diagnostic for pneumonia. Therefore, 101 children with CAP confirmed by chest
X-rays, at the mean age of 4.7 years (median 3.6 years; range 0.3 to 16 years), attended
the serological study (Table 11). None of the patients had more than one episode of CAP
during the surveillance period.
Table 11. Basic data of the 101 children with community-acquired pneumonia, by age.
Age groups (years) <2 2-5 ≥5 Total
Number of patients 19 44 38 101
Age (years) 1.3±0.4 3.3±0.7 8.0±3.2 4.7±3.4
Male gender 12 (63.1) 22 (50.0) 16 (42.1) 50 (49.5)
Ambulatory patients 12 (63.2) 30 (68.2) 32 (84.2) 74 (73.3)
Hospitalized patients 7 (36.8) 14 (31.8) 6 (15.8) 27 (26.7)
Data are presented as means±SD or number (percentage).
No significant differences were observed.
OVERALL ETIOLOGY (I)
A potential causative agent was detected in 75 (74%) of the 101 patients (Table 12).
Bacterial infection was demonstrated in 47 (46%) and viral infection in 50 (49%) cases.
Sole bacterial infections were found in 17 (17%) cases, being more frequent at ≥5 years of
60
age, whereas sole viral infections were found in 23 (23%) patients, being more frequent at
<5 years of age (Table 12). Mixed infections were found in 35 cases and appeared to be
evenly distributed among age groups; 22 (22%) of them were mixed viral-bacterial
infections. Two-thirds of the CAP cases occurred from November to March, with peaks in
November (16/101) and December (19/101).
Table 12. Etiology of pediatric community-acquired pneumonia by age.
Age groups (years) <2 years 2-5 years ≥5 years Total
Sole viral 7 (36.8) 14 (31.8) 2 (5.2) 23 (22.7)
Sole bacterial 1 (5.3) 7 (15.9) 9 (23.7) 17 (16.8)
All mixed * 9 (47.3) 12 (27.3) 14 (36.8) 35 (34.6)
Any etiology detected 17 (89.5) 33 (75.0) 25 (65.8) 75 (74.3)
Data are presented as number (percentage). The distribution of viral, bacterial and mixed
infections differed significantly between the three age groups (p<0.05, Fisher`s exact
test). * Mixed infections include viral-bacterial, dual bacterial and dual viral infections.
BACTERIAL INFECTIONS (I)
As seen in Table 13, the most common bacterial agent was M. pneumoniae,
detected in 27 (27%) patients. S. pneumoniae was found in 18 (18%) patients, and blood
culture was positive in a 5-year old boy only; in this case, pneumococcal antibody assays
remained negative. In addition, C. pneumoniae was diagnosed in 8 cases, S. negevensis
in 5 cases, H. influenzae in 3 cases and M. catharralis in 1 case, all these last 4 cases
being mixed infections. Empyema was diagnosed in one child, Streptococcus pyogenes
being cultured from pleural fluid. M. pneumoniae was found in 18% of 2-5 years old and in
61
47% of ≥5 years old children. C. pneumoniae was found only in ≥5 years old children,
causing 18% of the cases at that age group. Pneumoccal infections were distributed rather
equally between age groups, causing 16 to 21% of the CAP cases.
Table 13. Etiology of pediatric community-acquired pneumonia by specific microbes.
Serologic Number of cases with diagnostic findings in three age groups

findings All ages <2 years 2-5 years ≥5 years p*
[N=101] [N=19] [N=44] [N=38]
M. pneumoniae
All§ 27 (26.7) 1 (5.2) 8 (18.1) 18 (47.3) <0.001
Single# 12 (11.8) 1 (5.2) 4 (9.1) 7 (18.4)
S. pneumoniae
All 18 (17.8) 4 (21.0) 7 (15.9) 7 (18.4) ns
Single 4 (3.9) 0 2 (4.5) 2 (5.2)
C. pneumoniae
All 8 (7.9) 1 (5.2) 0 7 (18.4) <0.01
Single 1 (0.9) 0 0 1 (2.6)
S. negevensis
All 5 (4.9) 1 (5.2) 3 (6.8) 1 (2.6) ns
Single 2 (1.9) 0 2 (4.5) 0
RSV
All 17 (16.8) 7 (36.8) 7 (15.9) 3 (7.9) <0.05
Single 7 (6.9) 3 (15.7) 4 (9.1) 0
Parainfluenza 1,2
and 3 viruses
All 12 (11.8) 3 (15.7) 8 (18.1) 1 (2.6) 0.05
Single 7 (6.9) 1 (5.2) 6 (13.6) 0
Bocavirus
All 12 (12) 5 (26.3) 7 (15.9) 0 <0.01
Single 7 (6.9) 3 (15.7) 4 (9.1) 0
Influenza viruses
All 9 (8.9) 1 (5.2) 4 (9.1) 4 (10.5) ns
Single 2 (1.9) 0 1 (2.3) 1 (2.6)
Metapneumovirus
All 5 (4.9) 2 (10.5) 2 (4.5) 1 (2.6) ns
Single 2 (1.9) 2 (10.5) 0 0
Data are presented as number (percentage). *Fisher’s exact test; ns = not significant. § all cases,
including cases with two or more findings; # single cases, including cases with only one finding.
62
SIMKANIA NEGEVENSIS INFECTIONS (III)
Serological evidence for acute S. negevensis infection was found in 5/101(4.9%)
children, based on IgM antibodies in 3 and on IgG antibodies in 2 cases (Table 14). The
patients were 16 months to 10 years old, one was treated in hospital, and he only had
alveolar pneumonia. Two cases were single and 3 mixed infections. At least one non-
specific inflammatory marker was raised in all cases. The cases appeared evenly during
the year.
Table 14. CAP patients with serological evidence of acute Simkania negevensis infection.
Findings Case 1 Case 2 Case 3 Case 4 Case 5
Age (months) 36 16 57 120 52
Sex (F/M) F M M F M
Type of infiltration Interstitial Interstitial Interstitial Alveolar Interstitial
CRP (mg/L) 210 29.5 ND 116 ND
WBC (cells/mm3) 36500 10660 14850 27990 11870
PCT (ng/ml) 12.62 3.4 0.34 17.98 0.21
ESR (mm/h) 120 70 43 82 57
In/out-patient In Out Out Out Out
Co-infections Mycoplasma RSV None Mycoplasma None
Seasons October March May June November
Diagnostic 8-fold increase 4-fold decrease 4-fold increase 8-fold decrease 4-fold decrease
criteria IgM ABs IgG ABs IgG ABs IgM ABs IgM ABs
Abbreviations: ND, not done; RSV: respiratory syncytial virus; ABs, antibodies. The
suggested cut-off limits for high PCT concentration are 0.5 ng/mL, 1 ng/mL or 2 ng/mL.
63
The prevalence of S. negevensis IgM or IgG antibodies by the detection limits of the
MIF test was 20% in <2 y, 23% in 2-4 y, 16% in 5-9 y and 30% in >10 years old children.
IgG antibodies were positive in 20%, 13%, 4% and 0% (p=0.2), respectively, suggesting a
decreasing trend.
There were no cases with concomitant S. negevensis and C. pneumoniae infection
by the present criteria (Table 14). Microbe-specific IgM and IgG antibodies by MIF for S.
negevensis, separately in acute and convalescent sera, were compared also by the
detection limits of the MIF and EIA tests for C. pneumoniae (Table 15). When the MIF was
used for both pathogens, there were no class-specific cross-reactions. When C.
pneumoniae antibodies were measured by EIA, 5 children had IgG antibodies to both
pathogens in the acute (56% of S. negevensis and 11% of C. pneumoniae positive cases,
p=0.3) and 3 children in the convalescent serum (30% of S. negevensis and 7% of C.
pneumoniae positive cases, p=0.3).
Table 15. Antibodies to S. negevensis by MIF, compared with antibodies to C. pneumoniae by
MIF and/or EIA.
Cpn IgM pos Cpn IgG pos
MIF (N=2) EIA (N=4) MIF (N=17) EIA (N=44)
Acute samples
p=0.8
Sn MIF IgM pos (N=8) 0 1 p=0.3 2 5
Sn MIF IgG pos (N=9) 0 0 0 p=0.2 5 p=0.3
Convalescent samples
p=0.8
Sn MIF IgM pos (N=10) 0 0 p=0.7 2 3
Sn MIF IgG pos (N=10) 0 0 0 p=0.2 3 p=0.3
Cpn, C. pneumoniae; Sn, S. negevensis; MIF, micro immunofluorescence. Statistical significance:
p-values were calculated for the differences between the presence and absence of class-specific
antibodies to the two pathogens in paired sera, respectively.
64
VIRAL INFECTIONS (I)
As seen in Table 9, RSV (17 cases, 17%), parainfluenza 1, 2 or 3 virus (12 cases,
12%), HBoV (12 cases, 12%) and influenza A or B virus (9 cases, 9%) were the most
common findings. Nearly all RSV cases and all HBoV cases were in ≤5 years old children,
while hMPV infections were seen at all age groups (Table 13).
HUMAN METAPNEUMOVIRUS INFECTIONS (IV)
As shown in Table 16, serological evidence of acute hMPV infection, based on ≥4-
fold increase in IgG titers in paired sera, was found in 5 (4.9%) CAP patients. Their
average age was 2.6 years (11-70 months). Four of them were males, all but one were
treated as outpatients, none developed complications and all recovered completely. All
cases occurred during winter, from November to March. Single infection was present in
one case, while S. pneumoniae, M. pneumoniae, C. pneumoniae, influenza B virus and
HBoV were involved in the other 4 infections. Thus, there were no cases of mixed hMPV
and RSV infection. On admission, all patients had fever and cough, 4/5 had tachypnoea on
inspection, 4/5 had crackles and 3/5 silenced lung sound on auscultation. The infiltrates on
chest radiographs were alveolar in 3 and interstitial in 2 patients. Serum non-specific
inflammatory markers varied widely, being highest in the only hospitalized patient. He had
single hMPV infection, presenting with interstitial infiltration on chest X-rays.
The distribution of the hMPV-specific antibody titres by age on admission are
presented in Table 17. The titres were significantly associated with age. High acute titres
≥1:2560 were seen in 23 (19.3%) children; the titre was ≥1:5120 in 13 (10.9%) and
≥1:10240 in 9 (7.5%) cases. Despite the high titres on admission, a significant ≥4-fold
decrease indicating an acute infection was not seen in any case.
65
Table 16. Characteristics of the 5 patients suffering from CAP with serological evidence of
acute hMPV infection.
Cases No-1 No-2 No-3 No-4 No-5
Gender/age (months) M/70 M/37 M/15 F/25 M/11
Admission IgG titre 1:640 <1:40 <1:40 1:160 1:40
Convalescent IgG titre 1:2560 1:640 1:160 1:10240 1:160
In/out-patient Out Out In Out Out
Co-infections C.pn., M.pn S.pn, Infl.B HBoV S.pn, Infl.B None
Month of detection November January January February March
BT on admission (°C) 38 39.30 38.80 38 39.90
Symptoms on admission C C C C, FR C, FR
Respiratory signs on T, CR, SLS R T, CR, SLS T, CR, SLS T,CR
admission
Type of Rx infiltration alveolar alveolar interstitial interstitial alveolar
WBC (cells/mm3) 6490 11900 16870 6350 7900
PCT (ng/mL) 0.31 5.74 28.67 0.80 1.81
ESR (mm/h) 50 58 75 20 90
CRP (mg/L) ND 20 112 ND 103
F: female, M: male; In: in-patients, Out: out-patients; BT: body temperature; C: cough, FR:
food refusal; R: rhinorrhea, T: tachypnoea, CR: crackles, SLS: silenced lung sound; I:
interstitial, A: alveolar; RSV: respiratory syncitial virus; C.pn: C. pneumoniae; M.pn: M.
pneumoniae; S.pn: S. pneumoniae; Infl.B: influenza B virus; HBoV: human bocavirus; ND:
not done.
66
Table 17. Distribution of IgG titres to hMPV in acute sera by age in 119 children with data
on admission available.
<2 yrs 2-4 yrs 5-9 yrs ≥10 yrs

Age groups p-value *
(N=25) ** (N=43) (N=44) (N=12)
N°of cases ** 24 41 42 12
IgG ≥1:40 14 (58) # 30 (73) 40 (95) 12 (100) <0.0001
IgG ≥1:160 8 (33) § 22 (54) 36 (86) 12 (100) <0.0001
IgG ≥1:640 2 (8) 15 (37) 28 (67) 8 (67) 0.0027
IgG ≥1: 2560 2 (8) 6 (15) 13 (31) 2 (17) 0.8631
Data are presented as number (percent). * Fisher’s exact test. ** Cases with available
serum samples on admission. # 3 children ≤12 months of age; § 1 child ≤12 months of
age.
The prevalence of children positive for IgG to hMPV, by the detection limit of the EIA
test (≥1:40), increased constantly in relation to age (Table 18). When data from acute and
convalescent samples were combined, 68% of the <2 years old, 72% of the 2-4 years old
and 98% of the 5-9 years old children were positive for hMPV.
67
Table 18. Distribution of the subjects with detectable IgG to hMPV in paired sera by age,
among the 124 subjects with data available on admission and/or on convalescence.
Age groups <2 yrs 2-4 yrs 5-9 yrs ≥10 yrs p-value *
No of cases 25 43 44 12
IgG in acute sera 14/24 (58) ** 30/41 (73) 40/42 (95) 12/12 (100) <0.0001
IgG in convalescent sera 14/22 (63) # 24/35 (69) 38/39 (97) 12/12 (100) <0.0001
Total 17/25 (68) § 31/43 (72) 43/44 (98) 12/12 (100) <0.0001
Data are presented as number of sero-positive children / total number of children in each age
group (percent). The detection limit of 1:40 of the EIA test was used as the cut-off limit for
sero-positivity. * Fisher’s exact test. ** 3 children ≤12 months of age; # 3 children ≤12 months
of age; § 4 children ≤12 months of age.
HUMAN BOCAVIRUS INFECTIONS (V)
Serological evidence of an acute HBoV infection was found in 12 (12%) children
(Table 19). A significant increase in IgG antibodies between paired sera was found in 6
children. IgM antibodies were detected in 11 children: 7 were positive in acute and 8 in
convalescent sera. Both IgG and IgM findings were diagnostic in 5/12 children.
As shown in Table 19, the ages of the 12 CAP patients with HBoV infection ranged
from 10 months to 3 years and 9 months. Seven of these patients (58%) were treated as
inpatients, and all 12 children recovered completely. Seven of the 12 HBoV cases
occurred during 3 months, from November 2001 to January 2002.
68
Table 19. Characteristics of the 12 patients suffering from community-acquired pneumonia with serological evidence of acute human bocavirus infection, among the 101 children
with radiologically confirmed diagnosis.
Cases No-1 No-2 No-3 No-4 No-5 No-6 No-7 No-8 No-9 No-10 No-11 No-12
Age (months) / sex 19/F 36/M 45/F 15/M 40/F 37/M 33/F 10/M 27/M 18/M 23/M 34/M
Acute IgM 0.180 1.102 0.038 0.019 0.229 ND 0.031 ND 1.288 0.348 1.358 0.782
Convalescent IgM 0.077 0.522 0.044 0.911 0.383 0.369 0.529 0.295 0.050 0.046 0.323 0.351
Acute IgG 3.671 0.173 0.197 0.014 0.025 ND 0.023 ND 0.193 0.512 2.483 0.060
Convalescent IgG 3.677 1.574 1.986 1.491 2.569 0.024 1.612 1.669 0.033 0.506 3.076 1.835
Setting / month at diagn. In/II In/IX In/XI In/I Out/XII In/XI In/XI Out/XI Out/XII Out/III In/III Out/X
Co-infections S.pn,RSV None None hMPV None M.pn, AV S.pn, M.ca None None None None Para
BT on admission (°C) 39,2 39,80 39,80 38,80 38,20 39,90 40,30 40,20 38,60 36,00 37,30 36,50
Symptoms and signs C,IL C,IL C,IL C,IL C C,IL C,IL IL C C C,IL C,IL
Respiratory findings T T,SLS CR,SLS T,CR,SLS SLS T T,SLS None None CR CR CR
Infiltration type Alv Alv Alv Int Alv Alv Alv Alv Alv Alv Int Int
WBC (cells/mm3) 30310 39480 30900 16870 18790 28420 22270 31620 26810 11200 11930 8110
PCT (ng/mL) 62,73 16,14 18,58 28,67 0,49 23,01 2,63 74,99 17,04 0,50 0,22 0,87
ESR (mm/h) 128,00 85,00 40,00 75,00 65,00 115,00 94,00 115,00 60,00 50,00 90,00 27,00
F: female, M: male; ND: not done; Setting In: in-patients, Out: out-patients; M.pn: M. pneumoniae; hMPV: human metapneumovirus; AV: adenovirus; S.pn: S. pneumoniae;
M.ca: M. catarrhalis; PIV: parainfluenzavirus 1-2-3; BT: body temperature; C: cough; IL: ill looking; T: tachypnoea, CR: crackles, SLS: silenced lung sound; Alv: alveolar; Int:
interstitial. The diagnostic IgM and IgG titres in paired sera used to define an acute HBoV infection are evidenced in bold.
HBoV infection was single in 7 cases (58%), while 4 (33%) had co-infections with
other viruses and 3 (25%) with bacteria (Table 19). There was only one mixed infection
with RSV, although RSV infections were common and occurred during the same months
as most of the HBoV infections. On admission, all but one of the HBoV patients had
cough, 10/12 had fever, 5/12 had tachypnea, 5/12 had silenced lung sounds and 5/ 12
crackles on auscultation (Table 19). The infiltrates on chest radiographs were alveolar in 9
(75%) patients. PCT was >2.0 ng/mL in 8 (67%) patients, and WBCs were >22000
cells/mm3 in all but one of the patients with high PCT. Half of the cases with elevated PCT
and WBC were single HBoV infections with an alveolar infiltration on the chest radiograph.
Interestingly, ESR was >40 mm/h in 11/12 cases.
The proportion of children positive for IgG to HBoV increased constantly in relation
to age (Table 20). When the data from acute and convalescent samples were combined,
72% of the children aged <2 years, 74% of those 2-4 years and 98% of those 5-9 years
were positive for IgG to HBoV. On the contrary, the proportion of children positive for IgM
to HBoV decreased constantly in relation to age (Table 21). In all, 24% of the children
aged <2 years, 14% of those 2-4 years, 5% of those 5-9 years and 0% of those aged ≥10
years showed HBoV IgM. The combined HBoV specific IgG and IgM seropositivity rates
were 72% of the children aged <2 years, 79% of those 2-4 years and 98% of those 5-9
years, respectively.
Table 20. Children with detectable IgG to HBoV in paired sera by age, among those 124 with
available data on admission and/or convalescence.
Age groups <2 yrs 2-4 yrs 5-9 yrs ≥10 yrs
p-value *
No of cases 25 43 44 12
IgG in acute sera 15/23 (65)** 27/40 (67) 41/42 (98) 10/11 (91) <0.0001
IgG in convalescent sera 14/21 (67) # 26/35 (74) 38/39 (97) 11/12 (92) 0.0028
§
32/43 (74) 43/44 (98) 11/12 (92) 0.0231
Total 18/25 (72)
Data are presented as number of seropositive children / total number of children in each age
group (percent). The detection limit of 0.188 in the EIA test was used as the cut-off limit for
sero-positivity for IgG antibodies to HBoV. * Fisher’s exact test. ** 5 children ≤12 months of
age; # 5 children ≤12 months of age; § 6 children ≤12 months of age.
Table 21. Children with detectable IgM to HBoV in paired sera by age, among those 124 with
data available on admission and/or on convalescence.
Age groups <2 yrs 2-4 yrs 5-9 yrs ≥10 yrs
p-value*
No of cases 25 43 44 12
IgM in acute sera 3/23 (13)** 4/40 (10) 2/42 (5) 0/11 (0) 0.48
#
IgM in convalescent sera 5/35 (14) 0/39 (0) 0/12 (0) 0.0121
4/21 (19)
§
Total 6/25 (24) 6/43 (14) 2/44 (5) 0/12 (0) 0.0253
Data are presented as number of seropositive children / total number of children in each age
group (percent). The detection limit of 0.167 in the EIA test was used as the cut-off limit for
sero-positivity of IgM antibodies to HBoV. * Fisher’s exact test. ** 5 children ≤12 months of
age; # 5 children ≤12 months of age; § 6 children ≤12 months of age.
71
MIXED INFECTIONS (I)
Serologic evidence of mixed infections was found in 35 cases (35%); 22 were mixed
viral-bacterial infections, 8 were dual bacterial infections and 5 were dual viral infections. In
twelve patients, three or more microbial agents were detected (Table 12). In all, 44% of the
50 children with viral infection had evidence of bacterial infection, and respectively, 47% of
the 47 children with bacterial infection had evidence of viral infection. A concomitant viral
infection was diagnosed in 61% of the 18 patients with pneumococcal and in 36% of the
28 patients with atypical CAP. Mixed viral-bacterial infections occurred twice as often in <2
years old children (6/19, 31.6%) as in older children (14/82, 17%), even though the
difference was not statistically significant (p=0.19). As shown in Table 22, the most
frequently involved bacteria in mixed viral-bacterial infections were S. pneumoniae (15
cases) and M. pneumoniae (11 cases), being most often combined with RSV (4 cases for
both bacteria) or influenza viruses (3 cases for both bacteria). Correspondingly, more than
half of the children with RSV and influenza infections had evidence of concomitant
bacterial infection. Surprisingsly, M. pneumoniae was involved in all the 8 dual bacterial
infections, with S. pneumoniae and C. pneumoniae in 3 cases each and with S.
negevensis in 2 cases; all but one of these 8 cases occurred in children older than 5 years
of age.
72
Table 22. Mixed viral-bacterial infections in the etiology of 101 cases of community-acquired
pneumonia in childhood.
Any bacterial Streptococcus Mycoplasma Chlamydia Simkania

Viruses
agent pneumoniae pneumoniae pneumoniae negevensis
RSV (N=17) 10 (59) 4 (23) 4 (23) 1 (6) 1 (6)
Parainfluenza 1,2 or 3 (N=12) 4 (33) 3 (25) 1 (8) 1 (8) 0
Human bocavirus (N=12) 3 (25) 2 (17) 1 (8) 0 0
Influenza A or B (N=9) 5 (55) 3 (33) 3 (33) 0 0
Metapneumovirus (N=5) 3 (60) 2 (40) 1 (20) 1 (20) 0
Adenovirus (N=3) 2 (67) 1 (33) 1 (33) 0 0
Data are presented as number (percentage). H. influenzae was involved in two RSV cases
and one parainfluenza case. M. catarrhalis and cytomegalovirus were found in no cases.
ETIOLOGY BY HOSPITALIZATION (I)
The age distributions of hospitalized (27 children, 27%) and ambulatory (74
children, 73%) patients are shown in Table 11; the hospitalization rates did not differ
significantly between the three age groups. Twenty-one (33%) of the 63 children aged <5
years (Table 23) and 6 (16%) of those 38 aged ≥5 years were treated in the hospital. The
majority of S. pneumoniae CAP cases (15/18, 83%) were treated as outpatients. Eighteen
(66%) of the 27 children with M. pneumoniae infection were treated as outpatients in the
≥5 years age group whereas, in the younger age group, 6 of the 9 cases (p <0.05) were
treated in the hospital (Table 23). Seven of the 8 C. pneumoniae cases were treated as
outpatients. The majority of RSV cases, (17/29, 85%) were treated as outpatients, though
nearly all were <5 years of age.
73
Table 23. Etiology of community-acquired pneumonia in the 63 hospitalized and
ambulatory patients less than 5 years of age.
Inpatients, N=21 Outpatients, N=42 p-value *
S. pneumoniae 2 (9.5) 9 (21.4) ns
M. pneumoniae 6 (28.6) 3 (7.1) < 0.05
C. pneumoniae 0 1 (2.4) ns
S. negevensis 1 (5.7) 3 (7.1) ns
RSV 4 (19.0) 10 (23.8) ns
Human metapneumovirus 1 (5.7) 0 ns
Human bocavirus 7 (33.3) 5 (11.9) ns
Mixed infections 4 (19.0) 13 (30.9) ns
Unknown etiology 9 (42.8) 13 (30.9) ns
Data are presented as number (percentage). *Fisher’s exact test; ns = not significant.
PROCALCITONIN (II)
The median (25th-75th percentile) serum PCT concentration was 1.70 ng/mL (0.62-
20.53), 1.88 ng/mL (0.42-17.23) and 0.83 ng/mL (0.39-17.98) in the three age groups (<2,
2-4 and ≥5 years), respectively (p=0.83). There was no correlation between serum PCT
and age of the patients when analysed as continuous variables (rs = -0.14, p=0.16). Within
the youngest age group, the values did not differ significantly between <12 months and 12-
24 month-old children (data not shown). PCT was >0.5 ng/mL in 69 children, >1.0 ng/mL in
54 children and >2.0 ng/mL in 47 children. Thus, serum PCT concentration <0.5 ng/ml,
suggested as normal by the manufacturer, was found in 31% of the CAP patients. In the
74
only patient with positive blood culture to S. pneumoniae, serum PCT level was 9.66
ng/mL.
The severity of pneumonia as defined by the need for hospital treatment and the
presence of an alveolar infiltration on chest radiograph was associated with a significant
elevation of serum PCT concentration in comparison to outpatiens and children with
interstitial CAP (Tables 24 and 25). Likewise, the differences between these groups were
statistically significant, when PCT data were analysed as categorized by 0.5 ng/mL, 1.0
ng/mL and 2.0 ng/mL limits (Tables 24 and 25). In multiple linear regression adjusted for
age, the need for hospital treatment remained significant, but the presence of alveolar
pattern of pneumonia lost its significance (data not shown).
Table 24. Serum PCT concentrations in 100 children with CAP divided on the basis
of the setting of treatment.
Place of treatment
Outpatients Inpatients P value
Number of subjects 74 26
Median 0.72 17.81 0.0012 *
25th-75th percentile 0.39-7.75 2.63-28.67
Serum PCT
<0.5 ng/mL 27 (36) 4 (15)
0.5-1 ng/mL 13 (18) 2 (8)

0.0049 **
1.0-2.0 ng/mL 7 (9) 0 (0)
>2.0 ng/mL 27 (36) 20 (77)
Data are presented as number (percentage).
* Wilcoxon Two-Sample Test; ** Fisher's Exact Test.
75
Table 25. Serum PCT concentrations in 100 children with CAP divided on the basis
of radiological findings.
Radiological pattern
Interstitial Alveolar P value
Median 0.53 9.43 0.0003 *
25th-75th percentile 0.31-1.04 0.54-22.87
Serum PCT
<0.5 ng/mL 18 (47) 13 (21)
0.5-1 ng/mL 8 (21) 7 (11)

<0.0001 **
1.0-2.0 ng/mL 4 (11) 3 (5)
>2.0 ng/mL 8 (21) 39 (63)
* Wilcoxon Two-Sample Test; ** Fisher's Exact Test
As seen in Table 26, median PCT concentrations varied from 0.74 ng/mL (viral
group) to 7.75 ng/ml (unknown group) in the four etiological groups (p=0.62). Serum PCT
was >1.0 ng/mL in 55.5% of the patients in the pneumococcal, in 60.7% of those in the
atypical bacterial and in 41.4% of those in the viral CAP group, respectively (p=0.46).
Median PCT concentration was 6.39 ng/mL (25th-75th percentile 0.47-19.35) in the 8
patients with chlamydial etiology (mixed infections with M. pneumoniae included) and 6.37
ng/mL (25th-75th percentile 0.36-18.39) in the 17 patients with mycoplasmal etiology (mixed
infections with C. pneumoniae excluded) (p=0.93). The results on association between
serum PCT concentration and microbe-specific etiology remained negative when analysed
by multiple linear regression adjusted for age (data not shown).
76
Table 26. Serum PCT concentrations in 100 children with CAP divided on the basis of the
etiology.
Etiological groups of infection
Pneumococcal Atypical Viral Unknown P value
Number of subjects 18 28 29 25
Median 1.21 4.34 0.74 7.75 0.62 *
25th-75th percentile 0.62-33.07 0.34-18.28 0.41-5.79 0.47-12.82
Serum PCT
<0.5 ng/mL # 4 (22.2) 9 (32.1) 11 (37.9) 8 (32.0)
0.5-1 ng/mL # 4 (22.2) 2 (7.1) 6 (20.7) 2 (8.0)
1.0-2.0 ng/mL 2 (11.1) 1 (3.6) 3 (10.3) 1 (4.0) 0.46 **
>2.0 ng/mL 8 (44.4) 16 (57.1) 9 (31.0) 14 (56.0)
* Kruskal-Wallis Test; ** Fisher's Exact Test; # mixed infections were found in 7 of the 8
cases of pneumococcal pneumonia with serum PCT <1 ng/mL.
PCT concentrations ranged from 0.11 to 74.99 ng/mL, except for one child with a
very high PCT (330.80 ng/mL). This patient was a 37 months old girl, admitted for pleural
empyema caused by group-A beta-haemolytic Streptococcus and all her non-specific
inflammatory markers were very high (WBC 31400/µL, neutrophiles 92%, ESR 95 mm/h
and CRP 426.60 mg/L). Serum PCT concentration was >35 ng/mL in 9 additional cases.
Among these 10 patients with high serum PCT, five were treated in hospital, all but one
had an alveolar infiltration on chest X-rays and the etiology of CAP varied (pneumococcal
in 3, mycoplasmal in 2, single viral in 2 and unknown etiology in 3 cases). The median
values of serum CRP, blood ESR and WBC were 294.30 mg/L, 100.50 mm/h and
28000.00/µL, respectively. We made supplementary analyses by first excluding the case
77
with PCT >330 ng/mL and then all the 10 cases with serum PCT >35 ng/mL and the result
remained similar; there was no significant association between serum PCT concentration
and the etiology of CAP (data not shown).
Serum PCT values of 0.5 ng/mL, 1.0 ng/mL and 2.0 ng/mL were also tested as
screening limits between pneumococcal and viral pneumonia (Table 27). In this analysis,
mixed viral-pneumococcal infections were included into the pneumococcal CAP group.
The best LR value was 1.37 at the PCT level of 2.0 ng/mL, corresponding to a 44%
sensitivity and a 69% specificity.
Table 27. Diagnostic parameters for serum PCT concentrations in differentiating
pneumococcal from viral pneumonia.
Cut-off limits S. pneumoniae Sensitivity Specificity PV+ LR+

(N=18) % %
>0.5 ng/mL (N=32) 14 77.78 37.93 43.75 1.22
>1.0 ng/mL (N=22) 10 55.56 58.62 45.45 1.31
>2.0 ng/mL (N=17) 8 44.44 68.97 47.06 1.37
PV+: predictive value of a positive test result; LR+: likelihood ratio of a positive result.
Four single and 14 mixed infections made up the S. pneumoniae group, while the viral
group consisted of 23 single and 6 mixed infections (for more details see the text).
A similar analysis was made between atypical and viral pneumonia (Table 28).
Mixed viral-mycoplasmal, mixed viral-chlamydial and mixed viral-simkania cases were
included in the atypical bacterial group and all pneumococcal cases were excluded from
this analysis. Again, the best LR value was 1.78 at the PCT level of 2.0 ng/mL,
corresponding to a 57% sensitivity and a 69% specificity. Further, a similar analysis was
78
made between pneumococcal (coinfections with atypical bacteria included) and atypical
CAP (coinfections with pneumococci excluded). The best LR value was 1.11 at the PCT
level of 0.5 ng/mL, corresponding to a 76% sensitivity and a 32% specificity (data not
shown). Among the 8 patients with pneumococcal pneumonia and serum PCT <1 ng/mL
(Table 26), no one was treated in hospital and developed complications, three of them had
an alveolar infiltration on chest radiograph and mixed infections were found in all cases but
one (4 dual and 3 triple mixed infections, with M. pneumoniae involved in 4 and viruses in
5 cases). The median values of serum CRP, blood ESR and WBC were 59.95 mg/L, 35.00
mm/h and 10735.00/µL, respectively.
Table 28. Diagnostic parameters for serum PCT concentrations in differentiating atypical
from viral pneumonia.
Mycoplasma or chlamydia Sensitivity Specificity PV+ LR+

Cut-off limits (N=28) % %
>0.5 ng/mL (N=37) 19 67.86 37.93 51.35 1.06
>1.0 ng/mL (N=29) 17 60.71 58.62 58.62 1.42
>2.0 ng/mL (N=25) 16 57.14 68.97 64.00 1.78
PV+: predictive value of a positive test result; LR+: likelihood ratio of a positive result.
Thirteen single and 15 mixed infections made up the atypical group, while the viral group
consisted of 23 single and 6 mixed infections (for more details see the text).
The receiver operator characteristic (ROC) curve for serum PCT to separate
bacterial CAP (pneumococcal and atypical cases combined) from viral CAP had an area
under curve of 0.586 (Figure 10). The area under ROC curves for serum PCT to separate
79
pneumococcal from viral CAP, pneumococcal from atypical CAP and atypical from viral
CAP were also calculated and their values were 0.600, 0.556 and 0.588, respectively.
0.9
0.8
0.7
0.6
sensitivity
0.5
0.4
0.3
0.2
0.1
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
1-specificity
Figure 10. Receiver operator characteristic (ROC) curve for

serum PCT concentrations in discriminating combined
pneumococcal plus atypical CAP from viral pneumonia
(area under ROC curve = 0.586).
80
DISCUSSION
The present thesis on the etiology of CAP in childhood confirms, above all, the
importance of S. pneumoniae at all ages and in both hospital (Claesson et al. 1989, Juven
et al. 2000, Korppi et al. 1993, Nohynek et al. 1991, Ruuskanen et al. 1992) and
ambulatory settings (Heiskanen-Kosma et al. 1998, Wubbel et al. 1999). Moreover, as it
has been previously observed in Italian children (Esposito et al. 2001), mycoplasmal
infections are numerous and, particularly, M. pneumoniae was the most common etiologic
agent of CAP. Our data confirm that a high number of CAP (35%) is caused by mixed
infections (Heiskanen-Kosma et al. 1998, Juven et al. 2000) and that mixed viral-bacterial
infections are common not only in children less than 2 years of age, as suggested earlier
(Juven et al. 2000, Nohynek et al. 1991), but also in older children up to 5 years of age.
Secondly, the present study focused on three specific microbes, S. negevensis,
hMPV and HBoV in the etiology of pediatric CAP. The results suggest that S. negevensis
may be a true although rare cause of CAP in children (in less than 5%), and provide
serological evidence that hMPV and HBoV, previously found by PCR in respiratory
samples from children with respiratory disease (Boivin et al. 2003, Williams et al. 2004,
Kim et al. 2005, Konig et al. 2004, Mullins et al. 2004, Maggie et al. 2003, Viazov et al.
2003, Choi et al. 2006, Madhi et al. 2003, Lin et al. 2005, Peiris et al. 2003, Allander et al.
2007, Kantola et al. 2008, Völz et al. 2007, Lin et al. 2007, Bastien et al. 2007, Foulongne
et al. 2006, Ma et al. 2006, Sloots et al. 2006, Arnold et al. 2006, Choi et al. 2006,
Weissbrich et al. 2006), may be genuine causative agents of children`s pneumonia (hMPV
in less than 5%, but HBoV even in more than 10%). hMPV and HBoV were often involved
in mixed infections, as also earlier observed (Boivin et al. 2003, Kim et al. 2005, Mullins et
al. 2004, Maggie et al. 2003, Viazov et al. 2003, Choi et al. 2006, Madhi et al. 2003, Lin et
81
al. 2005, Peiris et al. 2003, Fry et al. 2007, Maggi et al. 2007, Völz et al. 2007, Allander et
al. 2007, Cilla et al. 2008, Christensen et al 2008, Longtin et al. 2008, Zhang et al. 2008)
and their sero-prevalence rates increased by age, reaching a 100% sero-positivity by age
10.
Thirdly, the present study confirmed earlier findings (Korppi and Remes 2001,
Korppi et al. 2003) indicating that serum PCT is unable to distinguish between bacterial
and viral CAP in children, the conclusion being similar for both ambulatory and
hospitalised children at different ages. Likewise, the ROC curves demonstrate the absence
of any cut-off limit useful to differentiate pneumococcal, mycoplasmal or chlamydial from
viral pneumonia. On the other hand, our results suggest that serum PCT is a useful
indicator of the severity of pediatric CAP when comparing the PCT values between
children who needed hospitalisation and those who did not and in comparing the alveolar
and interstitial pulmonary involvment on chest radiograph (Gendrel et al. 1999). Moreover,
very high values (>35 ng/mL) of serum PCT were found only among patients with severe
pneumonia, as found by Gendrel et al (Gendrel et al. 1999) in children with bacteremic
and/or other invasive bacterial infections.
GENERAL CONSIDERATIONS
The initial study population included 125 consecutive and previously healthy, less
than 16 years old patients with presumptive CAP. Pneumonia was radiographically
confirmed in 101 of them, and they formed the final study group. The mean age (4.7 years)
of the patients was higher than that found in previous studies on etiology of CAP in
hospitalized children from Finland (mean age 3.8 years) (Juven et al. 2001), USA (median
33 months) (Michelow et al. 2004) and Norway (median age 23 months) (Senstad et al.
2009).
82
Pneumococcal CAP is common at all ages and this is in agreement with two
previous studies, one from Finland (Juven et al. 2001) and the other one from Texas, USA
(Wubbel et al. 1999). However, pneumococcal etiology is less frequent (18%) than the one
found by Wubbel et al. (27%) and Juven et al. (37%) (Wubbel et al. 1999, Juven et al.
2001), who applied both pneumococcal antibody and immune complex assays in paired
sera (Wubbel et al. 1999, Juven et al. 2001), compared to the single antibody assay
applied in the present study. Pneumocococal immune complex assays have been
criticized for low specificity, complicated laboratory technique and difficulty to standardize
the method (Musher et al. 2001). Likewise, pneumococcal antigen detection tests, though
promising in adults, are not sufficiently sensitive and specific for pediatric use (Domìnguez
et al. 2003), thus they were not used in the present study. PCR and other methods for
pneumococcal nucleic acids are very sensitive and they have been promising in three
recent CAP studies in children (Toikka et al. 1999, Michelow et al. 2002, Michelow et al.
2004). Finally, the wide use of the seven-valent pneumococcal conjugate vaccine (PCV7)
and its demonstrated effectiveness against invasive pneumococcal disease (Whitney et al.
2006) may have influenced to the results of the present study.
Regarding M. pneumoniae, our high finding rate (27%) is in line with the results
obtained by Esposito et al. in pediatric CAP (33.5%) (Esposito et al. 2001), who
supplemented serologic tests by PCR. The high proportion of M. pneumoniae infections
found in our study may be explained by epidemiologic situation, that is a local outbreak,
but more likely by the fact that we performed two serologic assays. In line with PCR-based
studies, mycoplasmal infection was common also in young children (Principi et al. 2001).
C. pneumoniae was found in 8% of CAP, mainly in mixed infections and in children over 5
years of age, as also previously observed (Esposito et al. 2001, Heiskanen-Kosma et al.
1999). These results stress the role of atypical bacteria in CAP in school-aged children.
83
Alhough we only based the viral diagnoses on serum antibody rises by EIA, and
therefore no methods were available for e.g. rhinoviruses (Juven et al. 2000), viral
infections were common in young children (Table 12) (Heiskanen-Kosma et al. 1998). The
fairly high number of mixed infections (30%) in the present research is in line with recent
studies. As observed by Juven et al. (Juven et al. 2000) and Korppi et al. (Korppi et al.
1993) mixed viral-bacterial infections were common (32%) in children less than 2 years of
age, reflecting the role of RSV in mixed infections. Moreover, we found that mixed
infections were common (16%) up to 5 years of age, reflecting the role of M. pneumoniae
in dual bacterial infections.
SIMKANIA NEGEVENSIS
All Simkania diagnoses were based on significant increases or decreases of
antibodies in paired sera measured by MIF, which is known to be a specific but non-
sensitive method (Yamaguchi et al. 2005, Korppi et al. 2006). The prevalence of S.
negevensis -specific antibodies was 20-30%, with no expected rise by age, and there was
no substantial cross-reactivity between S. negevensis and C. pneumoniae antibodies.
S. negevensis is common in many areas. Sero-positivity in healthy adults has been
55-80% in Israel (Friedman et al. 1999), 46-65% in the UK and Denmark (Johnsen et al.
2005, Friedman et al. 2006), but only 4% in Japan (Yamaguchi et al. 2005). Age-specific
sero-prevalence in Israel was 13-29% in 1-2 y, 33-36% in 3-6 y, 42-57% in 7-10 y and 55-
75% in 11-15 y old non-symptomatic children, respectively (Friedman et al. 2003). The
present figures from Northern Italy were clearly lower, 20-30%, in accordance to recent
Finnish data (Korppi et al. 2006). Moreover, the sero-positivity did not rise with age. In
addition, the differences may reflect the poor sensitivity of the used MIF test (Yamaguchi
et al. 2005, Korppi et al. 2006), or climatic and other environmental factors.
84
Chlamydial antibodies were seen in no case with S. negevensis infection, speaking
against earlier suspicions that Simkania and Chlamydia species, at least when studied by
EIA, may induce cross-reactions (Korppi et al. 2006). Likewise, there were no significant
associations between IgM or IgG antibodies to S. negevensis and C. pneumoniae in MIF
and/or EIA serology. Despite this, some children with IgG or IgM to C. pneumoniae
detected by EIA alone may have actually had S. negevensis infection, in accordance with
a recent study from Japan (Miyashita et al. 2006).
In a recent study from Israel, S. negevensis was found by PCR in nasal samples in
91% of 34 children with pneumonia, in 88% of drinking water samples from their homes
and in 75% of healthy controls; these results suggest that S. negevensis is common in the
environment and may be transmitted via domestic water (Kahane et al. 2007). On the
other hand, the method used was a nested PCR which is prone to contaminations, and
this may explain the high finding rates (Apfalter et al. 2005). Viral infection was diagnosed
in half of the pneumonia cases, and no methods were available for other bacteria. Thus,
the final role of S. negevensis in pneumonia remained obscure, as it also showed in a
recent PCR study from the US (Kumar et al. 2005).
HUMAN METAPNEUMOVIRUS
To date, the present study is the first one on hMPV in pediatric CAP applying
serological methods. We found 11 earlier studies on hMPV respiratory infections in
children, which at least in part focused on lower respiratory tract (Boivin et al. 2003,
Williams et al. 2004, Kim et al. 2005, Konig et al. 2004, Mullins et al. 2004, Maggie et al.
2003, Viazov et al. 2003, Choi et al. 2006, Madhi et al. 2003, Lin et al. 2005, Peiris et al.
2003). Six studies consisted of children suffering from LRTI only (Williams et al. 2004, Kim
et al. 2005, Konig et al. 2004, Maggi et al. 2003, Choi et al. 2006, Madhi et al. 2003), four
included children with LRTI and other respiratory infections (Boivin et al. 2003, Mullins et
85
al. 2004, Viazov et al. 2003, Peiris et al. 2003) and one included children with CAP only
(Lin et al. 2005). RT-PCR was used in all studies and viral culture in addition to RT-PCR in
two of them. The number of patients have varied from 63 to 641, and hMPV has caused
0% to 25% of all cases and 7.7% to 54% of cases with virological diagnosis. In the CAP
study by Lin et al., upper airway samples were studied by RT-PCR from 116 Chinese
children with pneumonia, and the incidence of hMPV infections was 5.2% (Lin et al. 2005).
In that study, hMPV caused 13.3% of all viral CAP cases, all patients were hospitalized
their age ranging from 7 to 11 years. Our 101 CAP patients were clearly younger but the
figures based on hMPV serology were rather similar. In the study of Williams et al.
(Williams et al. 2004), the patients were less than 4 years old, and as many as 20% had
hMPV infection, but only 8% had pneumonia.
In the present study, hMPV infections mainly occurred during the winter season as
demonstrated by other year-round surveillance studies (Williams et al. 2004, Maggi et al.
2003, Choi et al. 2006). hMPV infections often occur after RSV outbreaks and can be
separated from RSV cases by virological assays only (Wolf et al. 2006). Thus, hMPV
cases have been reported as either virus-negative or simply as RSV-negative before the
year 2001 (Boivin et al. 2003, Williams et al. 2004, Lin et al. 2005, van den Hoogen et al.
2003). In the present study, there were no co-infections with RSV, although other mixed
infections were common.
The diagnosis of hMPV infection was based on the demonstration of a four-fold or
greater rise in the IgG antibodies (Hamelin and Boivin 2005), which proves viral
involvement, either as a cause of CAP or as a preceding infection paving way for bacterial
pneumonia. On admission, the titre of hMPV-specific IgG was ≥1:2560 in about 20% of the
patients, making a 4-fold increase in antibodies unlikely. However, a four-fold decrease
was not seen in any of them. S. pneumoniae was found in two patients, either PCT or ESR
was elevated in four patients, and the recovery after starting antibiotics was appropriate in
86
all cases, suggesting the presence of pneumococcal or other bacterial infection as the
principal cause on CAP. On the other hand, Madhi et al. found that hMPV with no evident
bacterial complication seems also to induce severe inflammatory responses (Madhi et al.
2003).
The few studies performed on sero-prevalence to hMPV from the Netherlands (van
den Hoogen et al. 2001), Japan (Ebihara et al. 2003) and the United States (Leung et al.
2005) have indicated that virtually all children are infected by 5-10 years of age. The
present figures from Northern Italy and the absence of cases of hMPV infection in patients
>6 years of age are in accordance with the Dutch (van den Hoogen et al. 2001), Japanese
(Ebihara et al. 2003) and American (Leung et al. 2005) results. In the study from Israel, the
authors performed a longitudinal serological analysis by immunofluorescence assay in
healthy children during the first two years of life (Wolf et al. 2003) and found that the sero-
positivity rate decreased to a minimum, due to disappearance of maternal antibodies, by
age of 13 months and increased to 52% by age of 24 months.
HUMAN BOCAVIRUS
At present, our study is the first one documenting the role of HBoV in children`s
pneumonia, and the second one documenting serologically the role of HBoV in children`s
respiratory infections. Fry et al. found, by using PCR in respiratory specimens, HBoV in
3.9% of pediatric CAP (Fry et al. 2007). Kantola et al. found significant IgG rises in 45%
and IgM in 49% of HBoV positive wheezing children; the serodiagnoses correlated with
high viral loads in NPA, with viremias and with single PCR findings (Kantola et al. 2008).
Thus, serological responses to HBoV seem to be associated not only with the presence of
infection but also with its invasiveness. In line, non-specific inflammatory markers were
often elevated in the present study, including also HBoV infections with no bacterial co-
infections or complications.
87
Since the first detection of HBoV (Allander et al. 2005), the clinical role of this virus
has been actively investigated. By PCR in respiratory samples, HBoV has been
documented in 19% of wheezing children (Allander et al. 2007, Kantola et al. 2008) and in
2% to 11% of children with other respiratory infections in North America, Europe, Asia and
Australia (Völz et al. 2007, Lin et al. 2007, Bastien et al. 2007, Foulongne et al. 2006, Ma
et al. 2006, Sloots et al. 2006, Arnold et al. 2006, Choi et al. 2006, Weissbrich et al. 2006).
In a recent Danish birth cohort, HBoV was found in 8.2% of infants with respiratory
infection, but equally often, in 8.6% of non-symptomatic controls (von Linstow et al. 2008).
In a population-based PCR study from Thailand, HBoV was detected in 9.6% of children
with respiratory infection and only in 0.4% of non-symptomatic controls. The figure was
10% to 12% in children less than 5 years with pneumonia (Fry et al. 2007). In the present
serological CAP study, HBoV infections were even more prevalent, 12% in the entire
group. Since all patients with HBoV pneumonia were less than 4 years of age, the
frequency in this age group was even higher: 17.6%.
HBoV involvement has been common in young children with respiratory and
intestinal infections (Schildgen et al. 2008, Kahn 2008) but more than half of the HBoV
positive cases have been associated with other viral infections. The figures varied from
35% to 76% in the previous PCR-based studies on respiratory tract infections (Allander et
al. 2007, Foulongne et al. 2006, Sloots et al. 2006, Choi et al. 2006, Weissbrich et al.
2006, Manning et al. 2006). In particular, HBoV findings have been associated with RSV
infections. In the only previous study on pneumonia in children, as many as 83% of the
cases positive by PCR for HBoV also showed infections with other microbes (Fry et al.
2007). In our study, only one-third of the HBoV cases were mixed infections, although
antibodies to as many as 9 other viruses were studied. Only one mixed case with RSV
was found, despite the fact that RSV findings were common, especially from November to
January, when also HBoV findings occurred frequently. These observations stress the
88
independent role of HBoV in the genesis of respiratory infections. Surprisingly, there were
only three mixed infections with bacteria, though respiratory viruses are known to pave the
way for pneumococci and other bacteria (Korppi 2002, Smyth 2002).
Thus far, only three studies have been published on the seroepidemiology of HBoV
infections (Endo et al. 2007, Kahn et al. 2008, Lin et al. 2008). Endo et al. developed an
immunofluorescence assay for HBoV IgG using insect cells expressing recombinant VP1
(Endo et al. 2007). Kahn et al. and Lin et al. set up an EIA for HBoV IgG by using VP2
virus-like particles (Kahn et al. 2008, Lin et al. 2008). In these studies, IgG antibodies were
found in over 90% of infants less than 3 months old, reflecting the high seropositivity rate
in their mothers (Endo et al. 2007, Kahn et al. 2008). In older children, the HBoV IgG
seroprevalences have been about 40% at 12 months, 60-80% at 24-36 months and 90%
at 4-5 years of age (Endo et al. 2007, Kahn et al. 2008). In the present study, the age-
specific figures were quite similar. However, Lin et al. found lower HBoV IgG seropositivity
rates: 22% at <12 months, 31% at 24 months and 60% at 36 months of age (Lin et al.
2008).
PROCALCITONIN
Two most recent studies on serum PCT and CAP in children (Toikka et al. 2000,
Moulin et al. 2001) have concluded that PCT is an informative marker for discriminating
between bacterial and viral pneumonia. Moulin et al (Moulin et al. 2001) suggested the
value of >1 ng/mL and Toikka et al (Toikka et al. 2000) the value of >2 ng/mL as cut-off
limits useful for differentiating between bacterial and viral CAP. These results are not in
agreement with the present paper, since the limits were exceeded in about 40% and 26%
of viral CAP cases, respectively. In accordance, the likelihood ratios at all cut-off limits
were far from the values of >3 or >5, which significantly modify pre-test probabilities in
clinical decision making. In the studies of Moulin et al. (Moulin et al. 2001) and Toikka et
89
al. (Toikka et al. 2000) serum PCT values were analysed only in hospitalised children
which may over-estimate the role of high values due to the over-representation of severe
cases. The studies of Korppi et al. (Korppi and Remes 2001, Korppi et al. 2003) that were
population-based and performed in primary health care settings, gave a negative result.
However, PCT determinations were done many years after the study from frozen serum
samples and this may have caused falsely low results. In those studies (Korppi and
Remes 2001, Korppi et al. 2003), serum PCT values tended to be lower in children under
2 years of age, compared with older children, a finding which is in disagreement with the
present data.
Elevated host-response markers, such as PCT and CRP, may be caused by
unrecognised bacterial co-infection albeit viral etiology alone has been established. In the
present study a good correlation was found between PCT and other non-specific host-
response markers, even if no evidence of bacterial aetiology was present. In a recent
double-blind, randomized, placebo-controlled trial conducted in a large number of children
to evaluate pneumococcal vaccination, Madhi et al (Madhi et al. 2004) showed that the
pneumococcal conjugate vaccine is able to reduce pneumonia associated with respiratory
virus infections, presumably by preventing bacterial co-infections. A significant number of
cases of pneumonia in children, if they need hospital care, seem to have pneumococcal
co-infection, though the infections primarily are of viral origin (Korppi 2002). On the other
hand, serologic evidence of bacterial infection may be present in mild respiratory infections
of airways not needing any antibiotic treatment (British Thoracic Society 2002, Nohynek et
al. 1995, Nohynek et al. 1995b) and, correspondingly, viral infections (for instance those
caused by adenoviruses or influenza virus) may present with severe disease clinically
resembling more an invasive bacterial than a viral infection (British Thoracic Society 2002,
Mistchenko et al. 1994, Vuori et al. 1998, Haterill et al. 2000).
90
In earlier studies, serum PCT has been higher in invasive bacterial infections, such
as bacterial meningitis (Gendrel et al. 1996, Gendrel et al. 1997b, Gendrel et al. 1999,
Hatherill et al. 1999) or septicaemia (Assicot et al. 1993, Gendrel et al. 1999, Chiesa et al.
1998), than in less invasive localized bacterial infections, such as urinary tract infections or
pneumonia (Gendrel et al. 1999). In addition, there is evidence that PCT may be better
than other acute phase reactants, such as CRP, WBC, interleukin 6 and interferon-alpha,
for distinguishing viral from bacterial infections (Gendrel et al. 1999, Hatherill et al. 1999).
The present study underlines the ability of PCT to relieve the severity of pneumonia,
assessed by the need of hospital care and the presence of an alveolar type of pneumonia.
The present results also stress the clinical importance of very high serum PCT values (for
instance >35 ng/mL), since the patients had severe pneumonia regardless of the etiology.
In fact, the median serum PCT detected in patients with alveolar pneumonia and in
hospitalised patients (9.43 and 17.81 ng/ml, respectively) are surprisingly similar to the
17,75 ng/ml reported by Gendrel et al (Gendrel et al. 1999) in 46 pediatric cases of
bacterial septicemia or meningitis. Thus, like other non-specific inflammation markers
(Nohynek et al. 1995), PCT may reflect more the invasiveness and severity than the
etiology of an infection. Recently, serum PCT level was found to be an interesting predictor
of complications and even mortality but not predictive for bacterial infection in adult
patients admitted in an intensive care unit for severe CAP (Boussekey et al. 2005).
STRENGTHS AND SHORTCOMINGS OF THE STUDY
Our results on viral and bacterial etiology derive from serologic assays that give an
indirect evidence of the etiology of infections. The lack of direct detection by PCR and
antigen tests of respiratory viruses in samples from upper respiratory tract was the main
shortcoming of the present study. For this reason, rhinovirus as well as some hMPV and
HBoV cases were evidently missed, particularly in <2 years old children, at the age period
91
in which the viral etiology is more common. The reduction in admission of all-cause
pneumonia in preschool children has been achieved by the routine infant immunisation
with PCV7 (Grijalva et al. 2007, De Wals et al. 2008); the omission of the information of
pneumococcal vaccination is a shortcoming of the present study, even if such vaccination
has been just introduced in Italy in 2001, for high risk children only. Children with wheezing
were excluded, because wheezing is more likely a sign of bronchiolitis or wheezy
bronchitis than of bacterial pneumonia (British Thoracic Society 2002). But since wheezing
seems to belong to the clinical picture of hMPV and HBoV infections in children, including
those with viral pneumonia (Williams et al. 2004, Allander 2008), some hMPV and HBoV
cases with concomitant wheezing and pneumonia were evidently missed. For all these
reasons, we therefore assume that the real frequency of hMPV and HBoV infections was
higher than that of 4.9% and 12% reported in the present study, respectively.
Our results represent only an estimate of the sero-prevalence of S. negevensis,
hMPV and HBoV, since the subjects were not randomly collected from the population. A
control group without respiratory disease was lacking for all the serologic assays of these
three respiratory pathogens. The convalescent serum samples were collected on average
35 days after the CAP diagnosis and this time may be too long for testing rises in
pneumococcal antibodies but too short for testing rises in mycoplasmal and chlamydial
antibodies. It has been shown (Nohynek et al. 1995b) that half of the children hospitalized
for acute LRTI develop a pneumococcal antibody response within 5 days from admission,
while antibodies to atypical bacteria, chlamydial antibodies in particular, may need 6
weeks or more to show diagnostic increases (Mazzulli 2003).
On the other hand, two different IgG antibody assays were available for S.
pneumoniae and two different IgM and IgG antibody assays for M. pneumoniae and C.
pneumoniae. Therefore, we consider that the results, for bacterial etiology in general,
reflect the current situation in Italy, supplementing previous data from Finland and Texas,
92
USA (Wubbel et al. 1999, Juven et al. 2001). Since both hospitalized and ambulatory
patients were included and our patients covered all pediatric age groups except newborns,
the results represent the whole spectrum of pediatric CAP. Other strengths of the present
study are that it was prospective and it employed paired sera, pneumonias were
radiologically confirmed in all cases, and an extensive antibody test panel for 16 microbes
(6 bacteria and 10 viruses) was used. Two new antibody tests, one for hMPV and the
other for HBoV, belonged to the panel.
FUTURE PERSPECTIVES
The further development of molecular technology, the continuation of the studies
performed world-wide, with an access to the available data banks of prospectively
collected clinical samples, not only samples collected from children with an infection but
also from healthy controls, will increase and improve our knowledge on important illnesses
like CAP in children. In order to study the overall etiology of pediatric CAP, direct and
indirect assays for viruses and for both typical and atypical bacteria should be applied in
the same study, and a sufficient number of ambulatory and hospitalized children of
different ages should be included. Such an approach not only improves the knowledge on
the respiratory pathogens till now discovered, but also allows e.g. by “molecular virus
screening” or population-based sero-epidemiology, to discover new pathogens. The
discovery of the emerging, not previously identified pathogens, as well as the monitoring of
the clinical epidemiology and changing characteristics of the known pathogens remain
important challenges for medical research in future years.
CONCLUSIONS
In conclusion, beside the importance of S. pneumoniae in pediatric CAP at all ages
and in both hospital and ambulatory settings, we found a high number of mycoplasmal
93
infections from 2 years of age onwards and a high number of mixed viral-bacterial
infections. These results underline the need to start an empirical antibiotic treatment in all
children suffering from uncomplicated CAP. If the treatment of CAP is started with
macrolide, the clinical response should be carefully monitored, since macrolides are not
optimal drugs for pneumococcal infections.
The present results also show that S. negevensis and hMPV might cause about 5%
each of CAP in children. By MIF, there is no substantial cross-reactivity between
antibodies to S. negevensis and C. pneumoniae, but chlamydial findings by EIA alone, not
confirmable by other tests, may reflect responses to S. negevensis. HBoV infections are
associated with significant antibody responses and they are fairly commonly involved
(12%) in pediatric pneumonia. The finding that only one-third of the cases were mixed
infections with other viruses suggests an independent pathogenetic role of HBoV in
pediatric CAP. Sero-conversion to both these two viruses usually takes place in early
childhood.
Finally, the present study confirms the inability of serum PCT determination to
differentiate between bacterial and viral etiology of CAP in ambulatory and hospitalized
children at different ages. On the other hand, serum PCT concentrations were high in the
most severe pneumonia cases and, in addition, PCT correlated well with other non-specific
inflammatory markers. The measurement of serum PCT seems to be useful in the
assessment of the severity of pneumonia, but it has to be interpreted jointly with other
clinical observations (such as general condition, findings in clinical examination and in
chest radiographs) and with other non-specific host-response markers (like WBC and
CRP).
94
ACKNOWLEDGEMENTS
The clinical ground at the basis of the present thesis has been carried out at the
Pediatric Clinic, Udine University Hospital, Italy, from October the 1 st, 2001 to December
the 31st, 2002. I express my sincere thanks to all my Colleagues and Collaborators I
worked with to reach these unexpected goals.
In particular, I am absolutely indebted and very grateful to:
 Professor Matti Korppi, MD, PhD. What about prof. Korppi? Impressive is his culture in
general pediatrics, pediatric respiratory diseases and infectious diseases, striking is his
ability to make quickly available this knowledge, huge are his strength of will, his skill to
raply istantly to any question and to the many “scientific” e-mail I asked and wrote him,
respectively and his capacity to correct all the drafts I submitted him. Well, he is and
has been an example and the “scientific” guide for me in the last five years;
 Doctor Mario Canciani, MD, Head of the Pediatric Allergology and Pulmonology Unit of
the Pediatric Clinic at the Udine University Hospital, Italy and my first teacher in
pediatric pulmonology. He was always optimistic and encouraging me during the
planning, starting and clinical phases of the study and he believed, right to the end, in
the achievement of the project and in the possibility of carrying the patient’s sera in
Finland for the microbiological analysis, thus creating a fruitful working relationship with
all the Finnish Colleagues;
 Doctor Francesca Valent, MD. She is my “reference point” for all the statistical analysis
included in the present thesis and in the published papers this thesis deals with. She
always replied to my questions clearly, quickly, with great pragmatism and enthusiasm,
although she recently became mother for the second time. I would have missed an
important contribution if Francesca had not adhered to the present project. In my
95
opinion, she likely will become a clear statistical professor… good luck, Francesca, for
your future professional career and many thanks again!
 Professor Giuliana Valerio, MD, valuable paediatrician and pediatric endocrinologist,
for her ideas and cleverness that have been essential in the development of my
specialization pediatric thesis and the two published papers on glycaemic and
natraemic values in pediatric CAP, that (“unfortunately”) were not included in the
present thesis; she also performed all the statistical analysis and she always was
helpful for any sort of comment and correction for my previous thesis and such papers;
for me a great honour has been and presently is knowing Giuliana and having had the
possibility to collaborate with her;
 Professor Alfred Tenore, MD, for the possibility to organize and materialize the work at
the Pediatric Clinic of the Udine University Hospital, Italy and Professor Markku Mäki,
MD, PhD, for the possibility to work in collaboration with the Paediatric Research
Centre, Tampere University and University Hospital, Finland. Professor Tenore, like
Mario, totally believed in the project, encouraged me to joint a collaboration with the
Finnish Colleagues and drafted the first papers about pediatric CAP, those regarding
the general etiology and procalcitonin;
 Professor Maija Leinonen, PhD. Only after having red a little part of her wide scientific
production, I understood how important this person is. She is a Scientist many
researchers world-wide refer to and I feel very proud to have had the possibility to
share, even if shortly, her activity with me;
 Mika Paldanius, PhD. How luck Mika is working together with Prof. Leinonen? He is not
luck, but simply very smart and sharp. In fact, he is strongly skilled in atypical bacteria
infections as proved by his many important papers on this particular topic. I knew him
in person in Oulu, Finland, about five years ago, delivering all the paired sera collected
96
from the Italian children with pneumonia. I have to thank him for all the serological
determinations on the typical and atypical bacteria the present thesis deal with;
 Professor Klaus Hedman, MD, PhD, Professor Raija Vainionpää, PhD, Doctor
Marjaana Kleemola, MD, PhD, Maria Söderlund-Venermo, PhD, Riitta Räty, PhD, Anne
Lahtinen, BSc and Mrs. Lea Hedman, all skilful Finnish virologists and microbiologists,
responsible for the viral and mycoplasmal serology, including sophisticated new tests,
of the present thesis;
 Doctor Lolita Fasoli, MD, for the initial assistance in the data collecting phase and for
the substantial contribution in writing the paper on Simkania negevensis in pediatric
CAP that has been included in the present thesis;
 Doctor Edmondo Falleti, MD, PhD, for the valuable contribution in the laboratoristic
determinations, in particular for those about procalcitonin that led to the paper on
procalcitonin in pediatric CAP included in the present thesis;
 Doctor Alessandro De Candia, MD, for the individual and independent interpretation,
blinded to the etiology and the evaluation of the initial images, of all the chest
radiographs;
 Doctor Maria Luisa Ciattei, for her helpfulness and quickness in the language revision
of the manuscript;
 all my Colleagues of the Pediatric Clinic of the Udine University Hospital, Italy for the
recruitment of the patients;
 all the little patients and their parents that voluntarily and willingly accepted to
participate to the study;
 my parents, Paolo and Margherita, I owe them all;
 my wife Barbara, clever woman that has understood how close I am to the present
thesis which come from a long way off and, in general, to my role of paediatrician; only
her knows how many time I spent to arrive at this point and all the time I stole her
97
instead of staying together, working as a paediatrician and writing as a “little
researcher” but, now, I want to thank her for the comprehension and encouragements;
 my son Michele, “little man” that perhaps will read, a day, the present thesis; I hope he
can discover the pleasure to work with satisfaction as me and understand how beautiful
is studying and deepening what he will do everyday.
Udine, 21st March 2009
Massimiliano Don
98
REFERENCES
Adegbola RA, Obaro SK, Biney E, Greenwood BM. Evaluation of Binax NOW
Streptococcus pneumoniae urinary antigen test in children with a high carriage rate of
pneumococcus. Pediatr Infect Dis J 2001;20:718-9
Allander T, Jartti T, Gupta S, Niesters HG, Lehtinen P, Osterback R, Vuorinen T, Waris M,
Bjerkner A, Tiveljung-Lindell A, van den Hoogen BG, Hyypiä T, Ruuskanen O.
Human bocavirus and acute wheezing in children. Clin Infect Dis 2007;44:904-910
Allander T, Tammi MT, Eriksson M, Bjerkner A, Tiveljung-Lindell A, Andersson B. Cloning
of a human parvovirus by molecular screening of respiratory tract samples. Proc Natl
Acad Sci U S A 2005;102:12891-12896
Allander T. Human bocavirus. J Clin Virol 2008;41:29-33
Alto WA. Human metapneumovirus: a newly described respiratory tract pathogen. J Am
Board Fam Pract 2004;17:466-469
Apfalter P, Reischl U, Hammerschlag MR. In-house nucleic acid amplification assays in
research: how much quality control is needed before one can rely upon the results? J
Clin Microbiol 2005;43:5835-41
Arnold JC, Singh KK, Spector SA, Sawyer MH. Human bocavirus: prevalence and clinical
spectrum at a children’s hospital. Clin Infect Dis 2006;43:283-288
Assicot M, Gendrel D, Carsin H, Raymond J, Guilbaud J, Bohuon C. High serum
procalcitonin concentrations in patients with sepsis and infection. Lancet
1993;341:515-518
Authors not listed. Clinical management of acute respiratory infections in children: a WHO
memorandum. Bull World Health Organ 1981;59:707-16
99
Balfour-Lynn IM, Girdhar DR, Aitken C. Diagnosing respiratory syncytial virus by nasal
lavage. Arch Dis Child 1995;72:58-9
Baqui AH, Black RE, Arifeen SE, Hill K, Mitra SN, al Sabir A. Causes of childhood deaths
in Bangladesh: results of a nationwide verbal autopsy study. Bull World Health Organ
1998;76:161-71
Bastien N, Brandt K, Dust K, Ward D, Li Y. Human Bocavirus infection, Canada. Emerg
Infect Dis 2006;12:848-850
Bastien N, Chui N, Robinson JL, Lee BE, Dust K, Hart L, Li Y. Detection of human
bocavirus in Canadian children in a 1-year study. J Clin Microbiol 2007;45:610-3
Berman S, McIntosh K. Selective primary health care: strategies for control of disease in
the developing world. XXI. Acute respiratory infections. Rev Infect Dis 1985;7:674-91
Boivin G, De Serres G, Côté S, Gilca R, Abed Y, Rochette L, Bergeron MG, Déry P.
Human metapneumovirus infections in hospitalized children. Emerg Infect Dis
2003;9:634-640
Boussekey N, Leroy O, Georges H, Devos P, d’Escrivan T, Guery B. Diagnostic and
prognostic values of admission procalcitonin levels in community-acquired
pneumonia in an intensive care unit. Infection 2005;33:257-263
Brieu N, Gay B, Segondy M, Foulongne V. Electron microscopy observation of human
bocavirus (HBoV) in nasopharyngeal samples from HBoV-infected children. J Clin
Microbiol 2007;45:3419-20
British Thoracic Society of Standards of Care Committee BTS Guidelines for the
management of community acquired pneumonia in childhood. Thorax 2002;57:1-24
Broughton RA. Infections due to Mycoplasma pneumoniae in childhood. Pediatr Infect Dis
J 1986;5:71-85
Bulla A, Hitze KL. Acute respiratory infections: a review. Bull World Health Organ
1978;56:481-98
100
Campbell H, Byass P, Lamont AC, Forgie IM, O'Neill KP, Lloyd-Evans N, Greenwood BM.
Assessment of clinical criteria for identification of severe acute lower respiratory tract
infections in children. Lancet 1989;1:297-9
Cherian T, John TJ, Simoes E, Steinhoff MC, John M. Evaluation of simple clinical signs
for the diagnosis of acute lower respiratory tract infection. Lancet 1988;2:125-8
Chetty K, Thomson AH. Management of community-acquired pneumonia in children.
Paediatr Drugs 2007;9:401-411
Chiesa C, Panero A, Rossi NA, Stegagno M, De Giusti M, Osborn JF, Pacifico L.
Reliability of procalcitonin concentrations for the diagnosis of sepsis in critically ill
neonates. Clin Infect Dis 1998;26:664-672
Choi EH, Lee HJ, Kim SJ, Eun BW, Kim NH, Lee JA, Lee JH, Song EK, Kim SH, Park JY,
Sung JY. The association of newly identified respiratory viruses with lower respiratory
tract infections in Korean children, 2000-2005. Clin Infect Dis 2006;43:585-592
Christensen A, Nordbø SA, Krokstad S, Rognlien AG, Døllner H. Human bocavirus
commonly involved in multiple viral airway infections. J Clin Virol 2008;41:34-7
Cilla G, Oñate E, Perez-Yarza EG, Montes M, Vicente D, Perez-Trallero E. Viruses in
community-acquired pneumonia in children aged less than 3 years old: High rate of
viral coinfection. J Med Virol 2008;80:1843-9
Claesson BA, Lagergard T, Trollfors B. Antibody response to outer membrane of
noncapsulated Haemophilus influenzae isolated from the nasopharynx of children
with pneumonia. Pediatr Infect Dis J 1991;10:104-8
Claesson BA, Leinonen M. Moraxella catarrhalis: an uncommon cause of community-
acquired pneumonia in Swedish children. Scan J Infect Dis 1994;26:399-402
Claesson BA, Trollfors B, Brolin I, Granström M, Henrichsen J, Jodal U, Juto P, Kallings I,
Kanclerski K, Lagergård T, et al. Etiology of community-acquired pneumonia in
101
children based on antibody responses to bacterial and viral antigens. Pediatr Infect
Dis J 1989;8:856-62
Clements H, Stephenson T, Gabriel V, Harrison T, Millar M, Smyth A, Tong W, Linton CJ.
Rationalised prescribing for community acquired pneumonia: a closed loop audit.
Arch Dis Child 2000;83:320-324
Courtoy I, Lande AE, Turner RB. Accuracy of radiographic differentiation of bacterial from
nonbacterial pneumonia. Clin Pediatr (Phila) 1989;28:261-4
Crowe JE. Human metapneumovirus as a major cause of human respiratory tract disease.
Pediatr Infect Dis J 2004;23:S215-21
Davies HD, Wang EE, Manson D, Babyn P, Shuckett B. Reliability of the chest radiograph
in the diagnosis of lower respiratory infections in young children. Pediatr Infect Dis J
1996;15:600-4
De Wals P, Robin E, Fortin E, Thibeault R, Ouakki M, Douville-Fradet M. Pneumonia after
implementation of the pneumococcal conjugate vaccine program in the province of
Quebec, Canada. Pediatr Infect Dis J 2008;27:963-8
Døllner H, Risnes K, Radtke A, Nordbø SA. Outbreak of human metapneumovirus
infection in norwegian children. Pediatr Infect Dis J 2004;23:436-40
Domìnguez J, Blanco S, Rodrigo C, Azuara M, Galì N, Mainou A, Esteve J, Castellvì A,
Prat C, Matas L, Ausina V. Usefulness of Urinary Antigen Detection by an
Immunochromatographic Test for Diagnosis of Pneumococcal Pneumonia in
Children. J Clin Microbiol 2003;41:2161-3
Domìnguez J, Galì N, Blanco S, Pedroso P, Prat C, Matas L, Ausina V. Detection of
Streptococcus pneumoniae Antigen by a Rapid Immunochromatographic Assay in
Urine Samples. Chest 2001;119:243-249
Don M, Paldanius M, Fasoli L, Canciani M, Korppi M. Simkania negevensis and
pneumonia in children. Pediatr Infect Dis J 2006;25:470-471
102
Dorigo-Zetsma JW, Zaat SA, Wertheim-van Dillen PM, Spanjaard L, Rijntjes J, van
Waveren G, Jensen JS, Angulo AF, Dankert J. Comparison of PCR, culture, and
serological tests for diagnosis of Mycoplasma pneumoniae respiratory tract infection
in children. J Clin Microbiol 1999;37:14-7
Dowell S, Garman RL, Liu G, Levine OS, Yang YH. Evaluation of Binax NOW, an Assay
for the Detection of Pneumococcal Antigen in Urine Samples, Performed among
Pediatric Patients. CID 2001;32:824-5
Durbin WJ, Stille C. Pneumonia. Pediatr Rev 2008;29:147-158
Eberhard OK, Haubitz M, Brunkhorst FM, Kliem V, Koch KM, Brunkhorst R. Usefulness of
procalcitonin for differentiation between activity of systemic autoimmune disease
(systemic lupus erythematosus/systemic antineutrophil cytoplasmic antibody-
associated vasculitis) and invasive bacterial infection. Arthritis Rheum 1997;40:1250-
1256
Ebihara T, Endo R, Kikuta H, Ishiguro N, Yoshioka M, Ma X, Kobayashi K. Seroprevalence
of human metapneumovirus in Japan. J Med Virol 2003;70:281-283
Elphick HE, Ritson S, Rodgers H, Everard ML. When a "wheeze" is not a wheeze:
acoustic analysis of breath sounds in infants. Eur Respir J 2000;16:593-597
Endo R, Ishiguro N, Kikuta H, Teramoto S, Shirkoohi R, Ma X, Ebihara T, Ishiko H, Ariga
T. Seroepidemiology of human bocavirus in Hokkaido prefecture, Japan. J Clin
Microbiol 2007;45:3218-23
Esposito S, Blasi F, Bellini F, Allegra L, Principi N; Mowgli Study Group. Mycoplasma
pneumoniae and Chlamydia pneumoniae infections in children with pneumonia. Eur
Respir J 2001;17:241-5
Esposito S, Bosis S, Cavagna R, Faelli N, Begliatti E, Marchisio P, Blasi F, Bianchi C,
Principi N. Characteristics of Streptococcus pneumoniae and atypical bacterial
103
infections in children 2-5 years of age with community-acquired pneumonia. Clin
Infect Dis 2002;35:1345-1352
Faden H, Heimerl M, Varma C, Goodman G, Winkelstein P. Urinary excretion of
pneumococcal cell wall polysaccharide in children. Pediatr Infect Dis J 2002;21:791-
793
Falck G, Gnarpe J, Gnarpe H. Prevalence of Chlamydia pneumoniae in healthy children
and in children with respiratory tract infections. Pediatr Infect Dis J 1997;16:549–54
Farha T, Thomson AH. The burden of pneumonia in children in the developed world.
Paediatr Respir Rev 2005;6:76-82
Foulongne V, Olejnik Y, Perez V, Elaerts S, Rodière M, Segondy M. Human bocavirus in
French children. Emerg Infect Dis 2006;12:1251-1253
Fredericks DN, Relman DA. Sequence-based identification of microbial pathogens: a
reconsideration of Koch's postulates. Clin Microbiol Rev 1996;9:18-33
Friedman MG, Dvoskin B, Kahane S. Infections with the Chlamydia-like microrganism
Simkania negevensis, a possible emerging pathogen. Microb Infect 2003;5:1013-
1021
Friedman MG, Galil A, Greenberg S, Kahane S. Seroprevalence of IgG antibodies to the
chlamydia-like microorganism ‘Simkania Z’ by ELISA. Epidemiol Infect 1999;122:117-
23
Friedman MG, Kahane S, Dvoskin B, Hartley JW. Detection of Simkania negevensis by
culture, PCR, and serology in respiratory tract infection in Cornwall, UK. J Clin Pathol
2006;59:331-3
Fry AM, Lu X, Chittaganpitch M, Peret T, Fisher J, Dowell SF, Anderson LJ, Erdman D,
Olsen SJ. Human bocavirus: a novel parvovirus epidemiologically associated with
pneumonia requiring hospitalization in Thailand. J Infect Dis 2007;195:1038-1045
104
Gendrel D, Assicot M, Raymond J, Moulin F, Francoual C, Badoual J, Bohuon C.
Procalcitonin as a marker for the early diagnosis of neonatal bacterial infection. J
Pediatr 1996;128:570-573
Gendrel D, Bohuon C. Procalcitonin as a marker of bacterial infection. Pediatr Infect Dis J
2000;19:679-688
Gendrel D, Raymond J, Assicot M, Moulin F, Iniguez JL, Lebon P, Bohuon C.
Measurement of procalcitonin levels in children with bacterial or viral meningitis. Clin
Infect Dis 1997b;24:1240-1242
Gendrel D, Raymond J, Coste J, Moulin F, Lorrot M, Guérin S, Ravilly S, Lefèvre H, Royer
C, Lacombe C, Palmer P, Bohuon C. Comparison of procalcitonin with C-reactive
protein, interleukin 6 and interferon-alpha for differentiation of bacterial vs. viral
infections. Pediatr Infect Dis J 1999;18:875-881
Gendrel D, Raymond J, Moulin F, Iniguez JL, Ravilly S, Habib F, Lebon P, Kalifa G.
Etiology and response to antibiotic therapy of community-acquired pneumonia in
French children. Eur J Clin Microbiol Infect Dis 1997;16:388-91
Ghaffar F, Friedland IR, McCracken GH Jr. Dynamics of nasopharyngeal colonization by
Streptococcus pneumoniae. Pediatr Infect Dis J 1999;18:638-46
Gibson NA, Hollman AS, Paton JY. Value of radiological follow up of childhood
pneumonia. BMJ 1993;307:1117
Greenberg D, Banerji A, Friedman MG, Chiu CH, Kahane S. High rate of Simkania
negevensis among Canadian inuit infants hospitalized with lower respiratory tract
infections. Scand J Infect Dis 2003;35:506-8
Grijalva CG, Nuorti JP, Arbogast PG, Martin SW, Edwards KM, Griffin MR. Decline in
pneumonia admissions after routine childhood immunisation with pneumococcal
conjugate vaccine in the USA: a time-series analysis. Lancet 2007;369:1179-1186
105
Guckel C, Benz-Bohm G, Widemann B. Mycoplasma pneumonias in childhood. Roentgen
features, differential diagnosis and review of literature. Pediatr Radiol 1989;19:499-
503
Gutiérrez F, Masià M, Rodrìguez JC, Ayelo A, Soldàn B, Cebriàn L, Mirete C, Royo G,
Hidalgo AM. Evaluation of the Immunochromatographic Binax NOW Assay for
Detection of Streptococcus pneumoniae Urinary Antigen in a Prospective Study of
Community-Acquired Pneumonia in Spain. CID 2003;36:286-92
Hamelin ME, Boivin G. Human metapneumovirus: a ubiquitous and long-standing
respiratory pathogen. Pediatr Infect Dis J 2005;24(11 Suppl):S203-7
Harari M, Shann F, Spooner V, Meisner S, Carney M, de Campo J. Clinical signs of
pneumonia in children. Lancet 1991;338:928-30
Harris JA, Kolokathis A, Campbell M, Cassell GH, Hammerschlag MR. Safety and efficacy
of azithromycin in the treatment of community-acquired pneumonia in children.
Pediatr Infect Dis J 1998;17:865-71
Haterill M, Sykes K, McIntyre AG, Murdoch IA. Procalcitonin may help differentiate
disseminated herpes simplex viral from bacterial sepsis in neonates. Eur J Pediatr
2000;159:168-169
Hatherill M, Tibby SM, Sykes K, Turner C, Murdoch IA. Diagnostic markers of infection:
comparison of procalcitonin with C reactive protein and leukocyte count. Arch Dis
Child 1999;81:417-421
Heaton P, Arthur K. The utility of chest radiography in the follow-up of neumonia. NZ Med
1998;111:315-7
Heiskanen-Kosma T, Korppi M, Jokinen C, Kurki S, Heiskanen L, Juvonen H, Kallinen S,
Sten M, Tarkiainen A, Ronnberg P-R, Kleemola M, Makela H, Leinonen M. Etiology of
childhood pneumonia: serologic results of a prospective, population-based study.
106
Heiskanen-Kosma T, Korppi M, Laurila A, Jokinen C, Kleemola M, Saikku P. Chlamydia
pneumoniae is an important cause of community-acquired pneumonia in school-aged
children: serological results of a prospective, population-based study. Scand J Infect
Dis. 1999;31:255-9
Heiskanen-Kosma T, Korppi M, Leinonen M. Serologically indicated pneumococcal
pneumonia in children: a population-based study in primary care settings. APMIS
2003;111:945-950
Heiskanen-Kosma T, Korppi M. Serum C-reactive protein cannot differentiate bacterial and
viral aetiology of community-acquired pneumonia in children in primary healthcare
settings. Scand J Infect Dis 2000;32:399-402
Heiskanen-Kosma T, Paldanius M, Korppi M. Simkania negevensis may be a true cause of
community acquired pneumonia in children. Scand J Infect Dis. 2008;40:127-30
Isaacs D. Problems in determinig the etiology of community-acquired childhood
pneumonia. Pediatr Infect Dis J 1989;8:143-8
Jadavji T. A practical guide for the diagnosis and treatment of pediatric pneumonia. CMAJ
1997;156:S703-S711
Jennings LC, Anderson TP, Werno AM, Beynon KA, Murdoch DR. Viral etiology of acute
respiratory tract infections in children presenting to hospital: role of polymerase chain
reaction and demostration of multiple infections. Pediatr Infect Dis J 2004;23:1003-
1007
Johnsen S, Birkebaek N, Andersen PL, Emil C, Jensen JS, Østergaard L. Indirect
immunofluorescence and real time PCR for detection of Simkania negevensis
infection in Danish adults with persistent cough and in healthy controls. Scand J
Infect Dis 2005;37:251-5
Jokinen C, Heiskanen L, Juvonen H, Kallinen S, Karkola K, Korppi M, Kurki S, Rönnberg
PR, Seppä A, Soimakallio S, et al. Incidence of community-acquired pneumonia in
107
the population of four municipalities in eastern Finland. Am J Epidemiol
1993;137:977-88
Juven T, Mertsola J, Toikka P, Virkki R, Leinonen M, Ruuskanen O. Clinical profile of
serologically diagnosed pneumococcal pneumonia. Pediatr Infect Dis J
2001;20:1028-33
Juven T, Mertsola J, Waris M, Leinonen M, Meurman O, Roivainen M, Eskola J, Saikku P,
Ruuskanen O. Etiology of community-acquired pneumonia in 254 hospitalized
children. Pediatr Infect Dis J 2000;19:293-8
Juvén T, Ruuskanen O, Mertsola J. Symptoms and signs of community-acquired
pneumonia in children. Scand J Prim Health Care 2003;21:52-6
Kahane S, Everett KD, Kimmel N, Friedman MG. Simkania negevensis strain Z: growth,
antigenic and genome charachteristics. Int J Syst Bacteriol 1999;49:815-820
Kahane S, Friedman MG. Antibiotic sensitivity in vitro of Simkania negevensis, in P.
Saikku (Ed.), Proceedings of the Fourth Meeting of the European Society for
Chlamydia research, Editrice Esculapio, Bologna, Italy, 2000, p. 397
Kahane S, Gonen R, Sayada C, Elion J, Friedman MG. Description and partial
characterization of a new Chlamydia-like microorganism. FEMS Microbiol Lett
1993;109:329-333
Kahane S, Greenberg D, Friedman MG, Haikin H, Dagan R. High prevalence of “Simkania
Z”, a novel Chlamydia-like bacterium, in infants with acute bronchiolitis. J Infect Dis
1998;177:1425-1429
Kahane S, Greenberg D, Newman N, Dvoskin B, Friedman MG. Domestic water supplies
as a possible source of infection with Simkania. J Infect 2007;54:75-81
Kahane S, Kimmel N, Friedman MG. The growth cycle of Simkania negevensis. Microbiol
2002;148:735-742
108
Kahn J. Human bocavirus: clinical significance and implications. Curr Opin Pediatr
2008;20:62-66
Kahn JS, Kesebir D, Cotmore SF, D'Abramo A Jr, Cosby C, Weibel C, Tattersall P.
Seroepidemiology of Human Bocavirus Defined Using Recombinant Virus-Like
Particles. J Infect Dis 2008;198:41-50
Kaikkonen L, Lankinen H, Harjunpää I, Hokynar K, Söderlund-Venermo M, Oker-Blom C,
Hedman L, Hedman K. Acute-phase-specific heptapeptide epitope for diagnosis of
parvovirus B19 infection. J Clin Microbiol 1999;37:3952-56
Kallio-Kokko H, Vapalahti O, Lundkvist Å, Vaheri A. Evaluation of Puumala virus IgG and
IgM enzyme immunoassays based on recombinant baculovirus-expressed
nucleocapsid protein for early nephropathia epidemica diagnosis. Clin Diagn Virol
1998;10:83-90
Kantola K, Hedman L, Allander T, Jartti T, Lehtinen P, Ruuskanen O, Hedman K,
Söderlund-Venermo M. Serodiagnosis of human bocavirus infection. Clin Infect Dis
2008;46:540-546
Karzai W, Oberhoffer M, Meier-Hellman A, Reinhart K. Procalcitonin: a new indicator of the
systemic response to severe infections. Infection 1997;6:329-334
Kenny GE, Grayston JT. Eaton pleuropneumonia-like organism (Mycoplasma
pneumoniae) complement-fixing antigen: Extraction with organic solvents. J Immunol
1965;95:19-25
Kesebir D, Vazquez M, Weibel C, Shapiro ED, Ferguson D, Landry ML, Kahn JS. Human
bocavirus infection in young children in the United States: molecular epidemiological
profile and clinical characteristics of a newly emerging respiratory virus. J Infect Dis
2006;194:1276-1282
Kiekara O, Korppi M, Tanska S et al. Radiological diagnosis of pneumonia in children. Ann
Med 1996;28:69-72
109
Kim YK, Lee HJ. Human metapneumovirus-associated lower respiratory tract infections in
Korean infants and young children. Pediatr Infect Dis J 2005;24:1111-1112
Konig B, Konig W, Arnold R, Werchau H, Ihorst G, Forster J. Prospective study of human
metapneumovirus infection in children less than 3 years of age. J Clin Microbiol
2004;42:4632-4635
Korppi M, Heiskanen-Kosma T, Jalonen E, Saikku P, Leinonen M, Halonen P, Mekela PH.
Aetiology of community-acquired pneumonia in children treated in hospital. Eur J
Pediatr 1993;152:24-30
Korppi M, Heiskanen-Kosma T, Kleemola M. Incidence of community-acquired pneumonia
in children caused by Mycoplasma pneumoniae: serological results of a prospective,
population-based study in primary health care. Respirology 2004;9:109-114
Korppi M, Heiskanen-Kosma T, Leinonen M, Halonen P. Antigen and antibody assays in
the aetiological diagnosis of respiratory infection in children. Acta Paediatr
1993c;82:137-41
Korppi M, Heiskanen-Kosma T, Leinonen M. White blood cells, C-reactive protein and
erythrocyte sedimentation rate in pneumococcal pneumonia in children. Eur Respir J
1997;10:1125-9
Korppi M, Katila ML, Jääskeläinen J, Leinonen M. Role of Moraxella (Branhamella)
catarrhalis as a respiratory pathogen in children. Acta Paediatr 1992b;81:993-6
Korppi M, Katila ML, Kalliokoski R, Leinonen M. Pneumococcal finding in a sample from
upper airways does not indicate pneumococcal infection of lower airways. Scand J
Infect Dis 1992;24:445-51
Korppi M, Kiekara O, Heiskanen-Kosma T, Soimakallio S. Comparison of radiological
findings and microbial aetiology of childhood pneumonia. Acta Paediatr
1993b;82:360-3
110
Korppi M, Leinonen M, Ruuskanen O. Pneumococcal serology in children's respiratory
infections. Eur J Clin Microbiol Infect Dis 2008;27:167-75
Korppi M, Leinonen M. Pneumococcal immune complexes in the diagnosis of lower
respiratory infections in children. Pediatr Infect Dis J 1998;17:992-995
Korppi M, Paldanius M, Hyvarinen A, Nevalainen A, Husman T. Chlamydia pneumoniae
and newly diagnosed asthma: a case-control study in 1- to 6-years old children.
Respirology 2004b;9:255-259
Korppi M, Paldanius M, Hyvärinen A, Nevalainen A. Simkania negevensis and newly
diagnosed asthma: a case-control study in 1- to 6-year-old children. Respirology
2006;11:80-83
Korppi M, Remes S, Heiskanen-Kosma T. Serum procalcitonin concentrations in bacterial
pneumonia in children: a negative result in primary healthcare settings. Pediatr
Pulmonol 2003;35:56-61
Korppi M, Remes S. Serum procalcitonin in pneumococcal pneumonia in children. Eur
Respir J 2001;17:623-627
Korppi M. Community-acquired pneumonia in children: issues in optimizing antibacterial
treatment. Paediatr Drugs 2003;5:821-32
Korppi M. Mixed viral-bacterial aetiology of community-acquired pneumonia in children.
APMIS 2002;110:515-522
Korppi M. Mixed viral-bacterial pulmonary infections in children. Pediatr Pulmonol Suppl
1999;18:110-2
Kumar S, Kohlhoff SA, Gelling M, Roblin PM, Kutlin A, Kahane S, Friedman MG,
Hammerschlag MR. Infection with Simkania negevensis in Brooklyn, New York.
111
Lahti E, Mertsola J, Kontiokari T, Eerola E, Ruuskanen O, Jalava J. Pneumolysin
polymerase chain reaction for diagnosis of pneumococcal pneumonia and empyema
in children. Eur J Clin Microbiol Infect Dis 2006;25:783-9
Laine M, Vainionpää R, Uksila J, Oksi J, Nissilä M, Kaarela K, Luukkainen R, Toivanen A.
Prevalence of Sindbis-related (Pogosta) virus infections in patients with arthritis. Clin
Exp Rheumatol 2003;21:213-216
Le Moullec JM, Jullienne A, Chenais J, Lasmoles F, Guliana JM, Milhaud G, Moukhtar MS.
The complete sequence of human preprocalcitonin. FEBS Lett 1984;167:93-7
Leung J, Esper F, Weibel C, Kahn JS. Seroepidemiology of human metapneumovirus
(hMPV) on the basis of a novel enzyme-linked immunosorbent assay utilizing hMPV
fusion protein expressed in recombinant vesicular stomatitis virus. J Clin Microbiol
2005;43:1213-1219
Lieberman D, Kahane S, Lieberman D, Friedman MG. Pneumonia with serological
evidence of acute infection with the Chlamydia-like microorganism “Z”. Am J Respir
Crit Care Med 1997;156(2 Pt 1):578-82
Lin F, Guan W, Cheng F, Yang N, Pintel D, Qiu J. ELISAs using human bocavirus VP2
virus-like particles for detection of antibodies against HBoV. J Virol Methods
2008;149:110-7
Lin F, Zeng A, Yang N, Lin H, Yang E, Wang S, Pintel D, Qiu J. Quantification of human
bocavirus in lower respiratory tract infections in China. Infect Agent Cancer 2007;2:3
Lin PY, Lin TY, Huang YC, Tsao KC, Huang YL. Human metapneumovirus and
community-acquired pneumonia in children. Chang Gung Med J 2005;28:683-688
Longtin J, Bastien M, Gilca R, Leblanc E, de Serres G, Bergeron MG, Boivin G. Human
bocavirus infections in hospitalized children and adults. Emerg Infect Dis
2008;14:217-21
112
Louie JK, Hacker JK, Gonzales R, Mark J, Maselli JH, Yagi S, Drew WL. Characterization
of viral agents causing acute respiratory infection in a San Francisco University
Medical Center Clinic during the influenza season. Clin Infect Dis 2005;41:822-828
Ma X, Endo R, Ishiguro N, Ebihara T, Ishiko H, Ariga T, Kikuta H. Detection of human
bocavirus in Japanese children with lower respiratory tract infections. J Clin Microbiol
2006;44:1132-1134
Madhi SA, Klugman KP and The Vaccine Trialist Group. A role for Streptococcus
pneumoniae in virus-associated pneumonia. Nat Med 2004;10:811-813
Madhi SA, Ludewick H, Abed Y, Klugman KP, Boivin G. Human metapneumovirus –
Associated lower respiratory tract infections among hospitalized human
immunodeficiency virus type 1 (HIV-1)–infected and HIV-1–uninfected African infants.
Clin Infect Dis 2003;37:1705-1710
Madhi SA, Ludewick H, Kuwanda L, Niekerk N, Cutland C, Little T, Klugman KP.
Pneumococcal coinfection with human metapneumovirus. J Infect Dis
2006;193:1236-43
Maggi F, Andreoli E, Pifferi M, Meschi S, Rocchi J, Bendinelli M. Human bocavirus in
Italian patients with respiratory diseases. J Clin Virol 2007;38:321-5
Maggi F, Pifferi M, Vatteroni M, Fornai C, Tempestini E, Anzilotti S, Lanini L, Andreoli E,
Ragazzo V, Pistello M, Specter S, Bendinelli M. Human metapneumovirus associated
with respiratory tract infections in a 3-year study of nasal swabs from infants in Italy. J
Clin Microbiol 2003;41:2987-2991
Manning A, Russell V, Eastick K, Leadbetter GH, Hallam N, Templeton K, Simmonds P.
Epidemiological profile and clinical associations of human bocavirus and other human
parvoviruses. J Infect Dis 2006;194:1283-1290
Margolis P, Gadomski A. Does this infant have pneumonia? JAMA 1998;279:308-13
113
Mazzulli T. Laboratory Diagnosis of Infection Due to Viruses, Chlamydia, and
Mycoplasma. In Long: Principles and Practice of Pediatric Infectious Diseases, 2nd
ed., 2003. pg 1394
McCarthy PL, Spiesel SZ, Stashwick CA, Ablow RC, Masters SJ, Dolan TF Jr.
Radiographic findings and etiologic diagnosis in ambulatory childhood pneumonias.
Clin Pediatr (Phila) 1981;20:686-91
McCracken GH Jr. Diagnosis and management of pneumonia in children. Pediatr Infect
Dis J 2000;19:924-928
McCracken GH Jr. Etiology and treatment of pneumonia. Pediatr Infect Dis J
2000b;19:373-377
McIntosh K. Community-Acquired Pneumonia in Children. N Engl J Med 2002;346:429-
437
Michelow IC, Lozano J, Olsen K, Goto C, Rollins NK, Ghaffar F, Rodriguez-Cerrato V,
Leinonen M, McCracken GH Jr. Diagnosis of Streptococcus pneumoniae lower
respiratory infection in hospitalized children by culture, polymerase chain reaction,
serological testing, and urinary antigen detection. Clin Infect Dis 2002;34:e1-e11
Michelow IC, Olsen K, Lozano J, Rollins NK, Duffy LB, Ziegler T, Kauppila J, Leinonen M,
McCracken GH Jr. Epidemiology and clinical characteristics of community-acquired
pneumonia in hospitalized children. Pediatrics 2004;113:701-7
Mistchenko AS, Diez R, Mariani AL, Robaldo J, Maffey AF, Bayley-Bustamante G,
Grinstein S. Cytokines in adenoviral disease in children: association of interleukin-6,
interleukin-8, and tumor necrosis factor alpha levels with clinical outcome. J Pediatr
1994;124:714-720
Miyashita N, Obase Y, Fukuda M, Shoji H, Mouri K, Yagi S, Yoshida K, Ouchi K, Oka M.
Evaluation of serological tests detecting Chlamydophila pneumoniae-specific
immunoglobulin M antibody. Intern Med 2006;45:1127-31
114
Mizgerd JP. Acute lower respiratory tract infection. N Engl J Med 2008;358:716-727
Monto AS. Epidemiology of viral respiratory infections. Am J Med 2002;112(suppl 6A):4S-
12S
Moulin F, Raymond J, Lorrot M, Marc E, Coste J, Iniguez JL, Kalifa G, Bohuon C, Gendrel
D. Procalcitonin in children admitted to hospital with community acquired pneumonia.
Arch Dis Child 2001;84:332-336
Mulholland EK, Simoes EA, Costales MO, McGrath EJ, Manalac EM, Gove S.
Standardized diagnosis of pneumonia in developing countries. Pediatr Infect Dis J
1992;11:77-81
Mullins JA, Erdman DD, Weinberg GA, Edwards K, Hall CB, Walker FJ, Iwane M,
Anderson LJ. Human metapneumovirus infection among children hospitalized with
acute respiratory illness. Emerg Infect Dis 2004;10:700-705
Murphy TF, Henderson FW, Clyde WA Jr, Collier AM, Denny FW. Pneumonia: an eleven-
year study in a pediatric practise. Am J Epidemiol 1981;113:12-21
Musher DM, Mediwala R, Phan HM, Chen G, Baughn RE. Nonspecificity of assaying for
IgG antibody to pneumolysin in circulating immune complexes as a means to
diagnose pneumococcal pneumonia. Clin Infect Dis 2001;32:534-8
Nascimento-Carvalho CM, Ribeiro CT, Cardoso MR, Barral A, Araújo-Neto CA, Oliveira
JR, Sobral LS, Viriato D, Souza AL, Saukkoriipi A, Paldanius M, Vainionpää R,
Leinonen M, Ruuskanen O. The role of respiratory viral infections among children
hospitalized for community-acquired pneumonia in a developing country. Pediatr
Infect Dis J 2008;27:939-41
Nohynek H, Eskola J, Kleemola M, Jalonen E, Saikku P, Leinonen M. Bacterial antibody
assays in the diagnosis of acute lower respiratory tract infection in children. Pediatr
Infect Dis J 1995b;14:478-84
115
Nohynek H, Eskola J, Laine E, Halonen P, Ruutu P, Saikku P, Kleemola M, Leinonen M.
The causes of hospital-treated acute lower respiratory tract infection in children. Am J
Dis Child 1991;145:618-22
Nohynek H, Valkeila E, Leinonen M, Eskola J. Erythrocyte sedimentation rate, white blood
cell count and serum C-reactive protein in assessing etiologic diagnosis of acute
lower respiratory infections in children. Pediatr Infect Dis J 1995;14:484-490
Ostapchuk M, Roberts DM, Haddy R. Community-acquired pneumonia in infants and
children. Am Fam Physician 2004;70:899-908
Palafox M, Guiscafré H, Reyes H, Munoz O, Martínez H. Diagnostic value of tachypnoea
in pneumonia defined radiologically. Arch Dis Child 2000;82:41-5
Peiris JS, Tang WH, Chan KH, Khong PL, Guan Y, Lau YL, Chiu SS. Children with
respiratory disease associated with metapneumovirus in Hong Kong. Emerg Infect
Dis 2003;9:628-633
Pereira JC, Escuder MM. The importance of clinical symptoms and signs in the diagnosis
of community-acquired pneumonia. J Trop Pediatr 1998;44:18-24
Persson K, Boman J. Comparison of five serologic tests for diagnosis of acute infections
by Chlamydia pneumoniae. Clin Diagn Lab Immunol 2000;7:739-44
Pratt C, Dominguez J, Rodrigo C, Gimenez M, Azuara M, Jimenez O, Galì N, Ausina V.
Procalcitonin, C-reactive protein and leukocyte count in children with lower respiratory
tract infection. Pediatr Infect Dis J 2003;22:963-968
Principi N, Esposito S, Blasi F, Allegra L; Mowgli study group. Role of Mycoplasma
pneumoniae and Chlamydia pneumoniae in children with community-acquired lower
respiratory tract infections. Clin Infect Dis 2001;32:1281-9
Principi N, Esposito S. Emerging role of Mycoplasma pneumoniae and Chlamydia
pneumoniae in paediatric respiratory-tract infections. Lancet Infect Dis 2001;1:334-44
116
Principi N, Esposito S. Mycoplasma pneumoniae and Chlamydia pneumoniae cause lower
respiratory tract disease in paediatric patients. Curr Opin Infect Dis 2002;15:295-300
Quinn TC, Gaydos CA. In vitro infection and pathogenesis of Chlamydia pneumoniae in
endovascular cells. Am Heart J 1999;S507-511
Ramsey BW, Marcuse EK, Foy HM, Cooney MK, Allan I, Brewer D, Smith AL. Use of
bacterial antigen detection in the diagnosis of pediatric lower respiratory infections.
Pediatrics 1986;78:1-9
Requejo HI, Guerra ML, Dos Santos M, Cocozza AM. Immunodiagnoses of community-
acquired pneumonia in childhood. J Trop Pediatr 1997;43:208-12
Rudan I, Boschi-Pinto C, Biloglav Z, Mulholland K, Campbell H. Epidemiology and etiology
of childhood pneumonia. Bull World Health Organ 2008;86:408-16
Russel G. Community acquired pneumonia. Arch Dis Child 2001;85:445-446
Ruuskanen O, Nohynek H, Ziegler T, Capeding R, Rikalainen H, Huovinen P, Leinonen M.
Pneumonia in childhood: etiology and respons to antimicrobial therapy. Eur J Clin
Microbiol Infect Dis 1992;11:217-23
Sarasini A, Percivalle E, Rovida F, Campanini G, Genini E, Torsellini M, Paolucci S,
Baldanti F, Marchi A, Grazia Revello M, Gerna G. Detection and pathogenicity of
human metapneumovirus respiratory infection in pediatric Italian patients during a
winter--spring season. J Clin Virol 2006;35:59-68
Schildgen O, Müller A, Allander T, Mackay IM, Völz S, Kupfer B, Simon A. Human
bocavirus: passenger or pathogen in acute respiratory tract infections? Clin Microbiol
Rev 2008;21:291-304
Selwyn BJ. The epidemiology of acute respiratory tract infection in young children:
comparison of findings from several developing countries. Coordinated Data Group of
BOSTID Researchers. Rev Infect Dis 1990;12 Suppl 8:S870-S888
117
Senstad AC, Suren P, Brauteset L, Eriksson JR, Hoiby EA, Wathne K-O. Community-
acquired pneumonia (CAP) in children in Oslo, Norway. Acta Paediatrica
2009;98:332-336
Simon L, Gauvin F, Amre DK, Saint-Louis P, Lacroix J. Serum procalcitonin and C-reactive
protein levels as markers of bacterial infection: A systematic review and meta-
analysis. Clin Infect Dis 2004;39:206-217
Sloots TP, McErlean P, Speicher DJ, Arden KE, Nissen MD, Mackay IM. Evidence of
human coronavirus HKU1 and human bocavirus in Australian children. J Clin Virol
2006;35:99-102
Smyth A, Carty H, Hart CA. Clinical predictors of hypoxaemia in children with pneumonia.
Ann Trop Paediatr 1998;18:31-40
Smyth A. Pneumonia due to viral and atypical organisms and their sequelae. Br Med Bull
2002;61:247-62
Söderlund-Venermo M, Lahtinen A, Jartti T, Hedman L, Kemppainen K, Lehtinen P,
Allander T, Ruuskanen O, Hedman K. Bocavirus-induced wheezing in children:
clinical assessment and improved diagnosis by serology. Emerg Infect Dis (in press).
Stein RT, Marostica PJ. Community-acquired pneumonia: a review and recent advances.
Pediatr Pulmonol 2007;42:1095-103
Suprin E, Camus C, Gacouin A, Le Tulzo Y, Lavoue S, Feuillu A, Thomas R. Procalcitonin:
a valuable indicator of infection in a medical ICU? Intensive Care Med 2000;26:1170-
Toikka P, Irjala K, Juven T, Virkki R, Mertsola J, Leinonen M, Ruuskanen O. Serum
procalcitonin, C-reactive protein and interleukin-6 for distinguishing bacterial and viral
pneumonia in children. Pediatr Infect Dis J 2000;19:596-602
118
Toikka P, Juvén T, Virkki R, Leinonen M, Mertsola J, Ruuskanen O. Streptococcus
pneumoniae and Mycoplasma pneumoniae coinfection in community acquired
pneumonia. Arch Dis Child 2000b;83:413-4
Toikka P, Nikkari S, Ruuskanen O, Leinonen M, Mertsola J. Pneumolysin PCR-based
diagnosis of invasive pneumococcal infection in children. J Clin Microbiol
1999;37:633-7
Tsolia MN, Psarras S, Bossios A, Audi H, Paldanius M, Gourgiotis D, Kallergi K, Kafetzis
DA, Constantopoulos A, Papadopoulos NG. Etiology of community-acquired
pneumonia in hospitalized school-age children: evidence for high prevalence of viral
infections. Clin Infect Dis 2004;39:681-686
Turner RB, Lande AE, Chase P, Hilton N, Weinberg D. Pneumonia in pediatric outpatients:
cause and clinical manifestations. J Pediatr 1987;111:194-200
van den Hoogen BG, de Jong JC, Groen J, Kuiken T, de Groot R, Fouchier RA, Osterhaus
AD. A newly discovered human pneumovirus isolated from young children with
respiratory tract disease. Nat Med 2001;7:719-724
van den Hoogen BG, Osterhaus DM, Fouchier RA. Clinical impact and diagnosis of human
metapneumovirus infection. Pediatr Infect Dis J 2004;23(1Suppl):S25-32
van den Hoogen BG, van Doornum GJ, Fockens JC, Cornelissen JJ, Beyer WE, de Groot
R, Osterhaus AD, Fouchier RA. Prevalence and clinical symptoms of human
metapneumovirus infection in hospitalized patients. J Infect Dis 2003;188:1571-7
Viazov S, Ratjen F, Scheidhauer R, Fiedler M, Roggendorf M. High prevalence of human
metapneumovirus infection in young children and genetic heterogeneity of the viral
isolates. J Clin Microbiol 2003;41:3043-3045
Vicente D, Cilla G, Montes M, Pérez-Yarza EG, Pérez-Trallero E. Human bocavirus, a
respiratory and enteric virus. Emerg Infect Dis 2007;13:636-637
119
Virkki R, Juven T, Mertsola J, Ruuskanen O. Radiographic follow-up of pneumonia in
children. Pediatr Pulmonol 2005;40:223-7
Virkki R, Juven T, Rikalainene H, Svedstrom E, Mertsola J, Ruuskanen O. Differentiation
of bacterial and viral pneumonia in children. Thorax 2002;57:438-441
Völz S, Schildgen O, Klinkenberg D, Ditt V, Müller A, Tillmann RL, Kupfer B, Bode U,
Lentze MJ, Simon A. Prospective study of Human Bocavirus (HBoV) infection in a
pediatric university hospital in Germany 2005/2006. J Clin Virol 2007;40:229-35
von Linstow ML, Høgh M, Høgh B. Clinical and epidemiologic characteristics of human
bocavirus in Danish infants: results from a prospective birth cohort study. Pediatr
Infect Dis J 2008;27:897-902
Vuori E, Peltola H, Kallio M, Leinonen M, Hedman K and the SE-TU Group. Etiology of
pneumonia and other common childhood infections requiring hospitalization and
parenteral antimicrobial therapy. Clin Infect Dis 1998;27:566-72
Vuori-Holopainen E, Salo E, Saxén H, Hedman K, Hyypiä T, Lahdenperä R, Leinonen M,
Tarkka E, Vaara M, Peltola H. Etiological diagnosis of childhood pneumonia by use of
transthoracic needle aspiration and modern microbiological methods. Clin Infect Dis
2002;34:583-90
Vuorinen T, Meurman O. Enzyme immunoassays for detection of IgG and IgM antibodies
to parainfluenza types 1, 2 and 3. Journal of Virological Methods 1989;23:63-70
Wang S. The microimmunofluorescence test for Chlamydia pneumoniae infection:
technique and interpretation. J Infect Dis 2000;181(suppl 3):S421-5
Weissbrich B, Neske F, Schubert J, Tollmann F, Blath K, Blessing K, Kreth HW. Frequent
detection of bocavirus DNA in German children with respiratory tract infections. BMC
Infect Dis 2006;6:109
Whitney CG, Pilishvili T, Farley MM, Schaffner W, Craig AS, Lynfield R, Nyquist AC,
Gershman KA, Vazquez M, Bennett NM, Reingold A, Thomas A, Glode MP, Zell ER,
120
Jorgensen JH, Beall B, Schuchat A. Effectiveness of seven-valent pneumococcal
conjugate vaccine against invasive pneumococcal disease: a matched case-control
study. Lancet 2006;368:1495-1502
Williams BG, Gouws E, Baschi-Pinto C, Bryce J and Dye C. Estimates of world-wide
distribution of child deaths from acute respiratory infections. Lancet Infect Dis
2002;2:25-32
Williams JV, Harris PA, Tollefson SJ, Halburnt-Rush LL, Pingsterhaus JM, Edwards KM,
Wright PF, Crowe JE Jr. Human metapneumovirus and lower respiratory tract
disease in otherwise healthy infants and children. N Engl J Med 2004;350:443-450
Wolf DG, Greenberg D, Kalkstein D, Shemer-Avni Y, Givon-Lavi N, Saleh N, Goldberg,
MD, Dagan R. Comparison of human metapneumovirus, respiratory syncytial virus
and influenza A virus lower respiratory tract infections in hospitalized young children.
Wolf DG, Zakay-Rones Z, Fadeela A, Greenberg D, Dagan R. High seroprevalence of
human metapneumovirus among young children in Israel. J Infect Dis
2003;188:1865-7
Wubbel L, Muniz L, Ahmed A, Trujillo M, Carubelli C, McCoig C, Abramo T, Leinonen M,
McCracken GH Jr. Etiology and treatment of community-acquired pneumonia in
ambulatory children. Pediatr Inf Dis J 1999;18:98-104
Yamaguchi T, Yamazaki T, Inoue M, Mashida C, Kawagoe K, Ogawa M, Shiga S,
Nakagawa Y, Kishimoto T, Kurane I, Ouchi K, Ohzeki T. Prevalence of antibodies
against Simkania negevensis in a healthy Japanese population determined by the
microimmunofluorescence test. FEMS Immunol Med Microbiol 2005;43:21-7
Zar HJ, Mulholland K. Global burden of Pediatric Respiratory Illness and the implications
for management and prevention. Ped Pulmonol 2003;36:457-461
121
Zar HJ. Pneumonia in HIV-infected and HIV-uninfected children in developing countries:
epidemiology, clinical features, and management. Curr Opin Pulm Med. 2004;10:176-
82
Zhang LL, Tang LY, Xie ZD, Tan XJ, Li CS, Cui AL, Ji YX, Xu ST, Mao NY, Xu WB, Shen
KL. Human bocavirus in children suffering from acute lower respiratory tract infection
in Beijing Children's Hospital. Chin Med J (Engl) 2008;121:1607-10
Ziegler T, Lindholm T, Scalia G. Causes of false-positive reactions in herpesvirus IgM
assays. Serodiagnosis and Immunotherapy in Infectious Diseases 1989;3:101-109
122
ORIGINAL PUBLICATIONS (I-V)
I. Don M, Fasoli L, Paldanius M, Vainionpää R, Kleemola M, Räty R, Leinonen M,

Korppi M, Tenore A and Canciani M. Aetiology of Community-Acquired Pneumonia:
Serologic Results of a Paediatric Survey. Scandinavian Journal of Infectious
Diseases 2005;37(11):806-812.
II. Don M, Valent F, Korppi M, Falleti E, De Candia A, Fasoli L, Tenore A and Canciani
M. Efficacy of serum procalcitonin in evaluating severity of community-acquired
pneumonia in childhood. Scandinavian Journal of Infectious Diseases
2007;39(29):129-137.
III. Fasoli L, Paldanius M, Don M, Valent F, Vetrugno L, Korppi M and Canciani M.

Simkania negevensis in community-acquired pneumonia in Italian children.
IV. Don M, Korppi M, Valent F, Vainionpää R and Canciani M. Human metapneumovirus

pneumonia in children: results of an Italian study and mini-review. Scandinavian
Journal of Infectious Diseases 2008;40(10):821-826.
V. Don M, Söderlund-Venermo M, Valent F, Lahtinen A, Hedman L, Canciani M,

Hedman K and Korppi M. Serologically verified human bocavirus pneumonia in
children. Submitted.
123
I
Don M, Fasoli L, Paldanius M, Vainionpaa R, Kleemola M, Raty R, Leinonen M, Korppi M,

Tenore A and Canciani M. Aetiology of Community-Acquired Pneumonia: Serologic Results
of a Paediatric Survey. Scandinavian Journal of Infectious Diseases 2005;37(11):806-812
Scandinavian Journal of Infectious Diseases, 2005; 37: 806 − 812 /
ORIGINAL ARTICLE
Aetiology of community-acquired pneumonia: Serological results

of a paediatric survey
MASSIMILIANO DON1, LOLITA FASOLI1, MIKA PALDANIUS2, RAIJA VAINIONPÄ Ä 3,

MARJAANA KLEEMOLA4, RIITTA RÄ TY4, MAIJA LEINONEN2, MATTI KORPPI5,
ALFRED TENORE1 & MARIO CANCIANI1
From the 1Department of Paediatrics, School of Medicine DPMSC, University of Udine, Udine, Italy, 2National Public
Health Institute, Oulu, 3Department of Virology, University of Turku, Turku, 4Department of Microbiology, National Public
Health Institute, Helsinki, and 5Department of Paediatrics, Kuopio University and University Hospital, Kuopio, Finland
Abstract
Serological methods are routinely used in the diagnosis of viral and atypical bacterial respiratory infections. Recently, they
have also been applied to typical bacteria, such as Streptococcus pneumoniae. The aim of this study was to determine the
aetiology of paediatric community-acquired pneumonia (CAP) in both ambulatory and hospitalized patients, by using
antibody assays. During a 15-month prospective surveillance, paired sera were studied for antibodies to 14 microbes in 101
children with symptoms of acute infection and infiltrates compatible with pneumonia on chest radiographs. A potential
causative agent was detected in 66 (65%) patients. Evidence of bacterial, viral and mixed viral-bacterial infection was
demonstrated in 44%, 42% and 20% of the CAP cases, respectively. The most commonly found agents included
Mycoplasma pneumoniae (27%), Pneumococcus (18%) and respiratory syncytial virus (17%). Human metapneumovirus
(hMPV) was detected in 5 (5%) children. Pneumococcal infections were evenly distributed among the age groups studied.
Our results confirm the role of S. pneumoniae in paediatric CAP at all ages, those of M. pneumoniae at >/2 y of age and
emphasize the emerging role of hMPV. The high proportion of mixed viral-bacterial infections highlights the need to treat all
children with CAP with antibiotics.
Introduction applied for the first time, most papers on the

aetiology of CAP in children, combining viral and
Serological methods are routinely used for in-hospi-
bacterial data, have originated from Finland [4 − 6] /
tal diagnosis of respiratory infections caused by
or Texas (USA) [7], thus reflecting only 2 popula-
viruses and atypical bacteria (such as Mycoplasma
tions. The available serology-based data on overall
pneumoniae and Chlamydia pneumoniae) [1].
aetiology of CAP in childhood [4,5,8], with the
During the last 15 y, serological methods have also exception of the data in hospitalized children of
been applied to study the aetiology of paediatric Juven et al. [6] and in ambulatory children of
community-acquired pneumonia (CAP) caused Wubbel et al. [7] are from the 1980s. There is a
by typical bacteria (notably Streptococcus pneumo- need for newer data on current distribution of
niae) [2]. To assess the overall microbial aetiology aetiological agents in paediatric CAP including also
of CAP in children, assays for viruses, typical other populations.
bacteria and atypical bacteria should be performed Human metapneumovirus (hMPV) is a new agent
in the same study, and ambulatory as well as isolated in 2001 from children and adults with acute
hospitalized children of different ages should be respiratory tract infection [9]. Human metapneu-
included. movirus has been found to be associated not only
After the pioneer study of Claesson et al. from with bronchiolitis, croup and asthma exacerbation
Sweden [3], in which pneumococcal serology was but also with pneumonia [10− 12]. /
Correspondence: M. Don, Paediatric Department, School of Medicine DPMSC, University of Udine, P. le S. Maria della Misericordia, 33100 Udine, Italy. Tel:
‡/39 0432 559240-4. Fax: ‡/39 0432 559258. E-mail: max.don@libero.it
(Received 20 July 2005; accepted 20 July 2005)

ISSN 0036-5548 print/ISSN 1651-1980 online # 2005 Taylor & Francis
DOI: 10.1080/00365540500262435
Serological results in paediatric community-acquired pneumonia 807
We carried out a 15-month prospective study to and chronic diseases (such as neoplasia, cardiovas-
investigate, by means of antibody assays, the role of cular and respiratory diseases, kidney and liver
14 microbes (5 bacteria and 9 viruses, including diseases, immunodepression, mental retardation
hMPV) in the aetiology of paediatric CAP in both and malnutrition) and hospital-acquired pneumonia.
ambulatory and hospitalized Italian children of Wheezing was considered as an exclusion criterion
different ages. because it more probably is a sign of bronchiolitis,
wheezy bronchitis or asthma exacerbation than of
pneumonia [1].
Materials and methods
Study design Bacteriology
This prospective study was conducted at the Depart- The serological assays for S. pneumoniae, Haemo-
ment of Paediatrics, School of Medicine DPMSC, philus influenzae, Moraxella catarrhalis and C.
University Hospital of Udine, Italy, during a 15- pneumoniae were carried out at the National Public
month period, from 1 October 2001 to 31 December Health Institute, Oulu, Finland. For S. pneumoniae
2002. The protocol was approved by the ethics infection, IgG antibodies to pneumococcal pneumo-
committee of the University of Udine, School of lysin and pneumococcal C-polysaccharide were
Medicine and informed consent was obtained from measured by enzyme immunoassay (EIA, Labsys-
the parents of all children. On admission, all patients tems, Helsinki, Finland) and a 2-fold or greater rise
underwent a clinical evaluation (including medical in antibody titres between paired sera was consid-
history and physical examination), a blood sample ered diagnostic, according to Nohynek et al. [13].
and a chest radiograph. A blood culture was For H. influenzae and M. catarrhalis infections, Ig
obtained in 35 children who had high fever or (polyvalent) antibodies against whole bacterial cell
appeared severely ill as judged by the physician on antigens (a mixture of 10 different strains) of non-
duty. Serum samples for serological assays, obtained typeable H. influenzae and M. catarrhalis were
on admission and on average 35 d later (range 18 measured by EIA and a 3-fold or greater antibody
− 56 d), were stored at —208C until transported to
/ /
rise between paired sera was considered diagnostic
the reference laboratories and analysed [13].
simultaneously. All patients were treated with In-house microimmunofluorescence (MIF) was
antibiotics (penicillins in 85 cases, cephalosporins in used to measure IgG and IgM antibodies to C.
9 cases, macrolides in 6 cases and glycopeptides in 1 pneumoniae, using purified, formalized elementary
case). bodies of strain K6 as antigen [14]. The diagnosis
was based on a 4-fold or greater rise in IgG
antibodies between the paired sera or on the
Patients
presence of IgM of ]1:16 in any serum. C.
/
In all, 125 consecutive and previously healthy pneumoniae-specific IgG and IgM were also mea-
children, who arrived at the Paediatric Emergency sured by EIA antibody kits (AniLabsystems Ltd. Oy,
Room (ER) of our hospital and were considered to Helsinki, Finland) based on an indirect solid-phase
have CAP, were eligible for the study. Our hospital technique with horseradish peroxidase as a marker
serves a population of about 50,000 children in the enzyme [15]. In the assay, 1.5-fold changes in
province of Udine, in northeast Italy; the patients enzyme immunounits (EIU) in the zones below
either come to the ER spontaneously or are sent by a 130 EIU in IgG were considered as diagnostic. A
general paediatrician. The inclusion criteria of the 1.3-fold change was considered diagnostic when the
present study were signs and/or symptoms compa- EIU values were more than 130 EIU in IgG. The
tible with respiratory infection (fever ]388C, ta-
/ IgM EIA results were expressed as a signal/cut-off
chypnoea, cough and/or findings of crackles, (S/CO) ratio, and a S/CO ratio of more than 1.1 was
bronchial breathing or diminished breath sounds considered as positive, 0.5 − 1.1 as equivocal, and less
/
on auscultation) and radiological infiltrations con- than 0.5 as negative.

sistent with pneumonia. Chest radiographs were The serological assays for M. pneumoniae were
taken on admission in all children with suspected carried out at the Department of Microbiology,
pneumonia and were interpreted by a senior radi- National Public Health Institute, Helsinki, Finland.
ologist who classified the findings into alveolar, Complement fixation (CF) test was performed
interstitial and no infiltrations. Only children with by a standard micromethod using in-house antigen
an infiltration consistent with pneumonia were prepared according to Kenny and Grayston [16]. A
included. The exclusion criteria included infants 4-fold or greater increase in titres between paired
less than 1 month of age, the presence of severe sera and high titres (1:128 or higher) in either serum
808 M. Don et al.
without a 4-fold rise was considered diagnostic. cases. None of the patients had more than 1 episode
The CF test was supplemented by measuring IgG of CAP during the surveillance period.
and IgM antibodies to M. pneumoniae by EIA
(Labsystems, Helsinki, Finland), using the proce-
Overall aetiology
dure recently published by Korppi et al. [17].
A potential causative agent was detected in 66 (65%)
of the 101 patients (Table II). Bacterial infection was
Virology demonstrated in 44 (44%) and viral infection in 42
All viral serology was performed at the Department (42%) cases. Infections caused by a single bacterium
of Virology, University of Turku, Finland. IgG were found in 17 (17%) cases, being more frequent
antibody titres against parainfluenza viruses type at ]5 y of age, whereas infections by a single virus
1− 3, influenza A and B viruses and adenoviruses
/ were found in 19 (19%) patients, being more
were determined by EIA using antigen-coated solid frequent at B5 y of age (Table II). Mixed infections
phase and horseradish peroxidase-conjugated rabbit included viral-bacterial infections in 20 (20%) pa-
antihuman IgG (Dako, Glostrup, Denmark), as tients, dual bacterial infections in 7 and dual viral
described earlier [18]. A 2-fold or greater increase infections in 3 patients, with no significant differ-
in antibody titres between paired sera was consid- ences between the age groups. Two-thirds of the
ered diagnostic. IgG and IgM antibodies against CAP cases occurred from November to March, with
cytomegalovirus were measured as described by peaks in November (16/101) and December (19/
Ziegler et al. [19], and IgG antibodies against RSV 101).
and hMPV as described by Laine et al. [20]. For
RSV and hMPV, a 4-fold or greater increase in IgG Bacterial infections
titres was considered diagnostic.
As seen in Table III, the most common bacterial
agent was M. pneumoniae, detected in 27 (27%)
Statistical analysis patients. Streptococcus. pneumoniae was found in
The significances of the differences among the 3 age 18 (18%) patients and blood culture was positive in
groups ( B2 y; 2 − 5 y; ]5 y), and between the
/ / /
only 1 case; in that case, pneumococcal antibody
children treated as inpatients and outpatients, were assays remained negative. Chlamydia pneumoniae
analysed by x2 or Fisher’s exact tests. Two-tailed was diagnosed in 8 cases and, in addition, H.
tests were used in all analyses, and the differences influenzae in 3 cases and M. catharralis in 1 case.
were considered as statistically significant at the level Haemophilus influenzae and M. catharralis were
of p B0.05.
/
only found in mixed infections. Empyema was
diagnosed in 1 child; Streptococcus pyogenes was
cultured from pleural fluid. M. pneumoniae was
Results found in 18% of 2 − 5-y-old and in 47% of ]5-y-old
/
children. Chlamydia pneumoniae was found only in

Demographic, laboratory and radiological data
children ]5 y of age, causing 18% of the cases in
Of the initial 125 consecutively enrolled patients, 24 that age group. Pneumoccal infections were distrib-
(19%) were excluded, 16 because convalescent uted rather equally between the 3 age groups,
serum samples were not obtained and 8 because ranging from 16 to 21%.
chest radiographs were not diagnostic of pneumonia.
Therefore, 101 children with CAP, at a mean age of
Viral infection
4.7 y (range 0.3 − 16 y), remained eligible for this
/
study (Table I). An alveolar infiltration was present As seen in Tables III and IV, RSV (17 cases, 17%),
in 70 (69%) cases and an interstitial in 31 (31%) parainfluenza 1, 2, 3 viruses (12 cases, 12%), and
Table I. Basic data of the 101 children with community-acquired pneumonia.
Age groups (y) B/2 2 −/5 ]5 Total

Age (y) 1.39/0.4 3.39/0.7 8.09/3.2 4.79/3.4
Male gender 12 (63.1) 22 (50.0) 16 (42.1) 50 (49.5)
Ambulatory patients 12 (63.2) 30 (68.2) 32 (84.2) 74 (73.3)
Hospitalized patients 7 (36.8) 14 (31.8) 6 (15.8) 27 (26.7)
Data are presented as means 9/ SD or number (percentage). No significant differences were observed.
Table II. Aetiology of paediatric community-acquired pneumonia by age.
Age group Number of cases Viral Bacterial Mixeda Any aetiology detected
B /2 y 19 6 (31.6) 1 (5.3) 7 (36.8) 14 (73.7)

2 −/5 y 44 11 (25.0) 6 (13.6) 10 (22.7) 27 (61.4)
]/5 y 38 2 (5.3) 10 (26.3) 13 (34.2) 25 (65.8)
Total 101 19 (18.8) 17 (16.8) 30 (29.7) 66 (65.3)
The distribution of viral, bacterial and mixed infections differed significantly between the 3 age groups (p =0.02). a Mixed infections include
viral-bacterial, dual bacterial and dual viral infections. Data are presented as number (percentage).
influenza A and B viruses (9 cases, 9%) were the of bacterial infection and 45% of the 44 children
most common viral aetiologies. Human metapneu- with bacterial infection had evidence of viral infec-
movirus was found in 5 children; 2 infections were tion. A concomitant viral infection was diagnosed in
single and 3 mixed with bacteria. Nearly all RSV 55% of the 18 patients with pneumococcal and in
cases were in children 55 y of age, while hMPV
/ 33% of the 27 patients with mycoplasmal CAP.
infections were seen in all age groups (Table III). Mixed viral-bacterial infections occurred more fre-
The mean age of children infected with hMPV was quently in children B2 y of age (6/19, 31.6%;
2.6 y and occurred from November to March; only 1 p B0.05) than in older children (14/82, 17%). As
child needed hospitalization. shown in Table IV, the most frequently involved
bacteria in mixed viral-bacterial infections were
S. pneumoniae (14 cases) and M. pneumoniae
Mixed infections (12 cases). Likewise, more than half of the children
Serological evidence of mixed infections was found with RSV and influenza infections had evidence of
in 30 cases (30%); 20 were mixed viral-bacterial concomitant bacterial infection. Mycoplasma pneu-
infections, 7 were dual bacterial infections and 3 moniae was involved in 6 of the 7 dual bacterial
were dual viral infections (Table II). In 9 patients 3 infections, with S. pneumoniae and C. pneumoniae
or more microbial agents were detected. In all, 47% in 3 cases each and exclusively in children older than
of the 42 children with viral infection had evidence 5 y of age.
Table III. Aetiology of paediatric community-acquired pneumonia by specific microbes.
Number of cases with diagnostic findings in 3 age groups
Serological findings All ages B/2 y 2 −/5 y ]/5 y p value
M. pneumoniae
Alla 27 (26.7) 1 8 18 B/0.001
Singleb 12 1 4 7
S. pneumoniae
All 18 (17.8) 4 7 7 0.88
Single 4 0 2 2
C. pneumoniae
All 8 (7.9) 1 0 7 0.004
Single 1 0 0 1
RSV
All 17 (16.8) 7 7 3 0.02
Single 7 3 4 0
Parainfluenza 1,2 and 3 viruses
All 12 (11.8) 3 8 1 0.052
Single 7 1 6 0
Inflenza viruses
All 9 (8.9) 1 4 4 0.91
Single 2 0 1 1
Metapneumovirus
All 5 (4.9) 2 2 1 0.41
Single 2 2 0 0
Data are presented as number (percentage). a All cases, including cases with 2 wo or more findings; b Single cases, including cases with only
1 finding.
810 M. Don et al.
Table IV. Mixed viral-bacterial infections in the aetiology of 101 cases of community-acquired pneumonia in childhood.
Any bacterial Streptococcus Mycoplasma Chlamydia Haemophilus

Viruses agent pneumoniae pneumoniae pneumoniae influenzae
RSV (n =17) 9 (53) 4 (23.5) 4 (23.5) 2 (11.8) 2 (11.8)

Parainfluenza 1,2 or 3 (n =/12) 4 (33) 3 (25) 1 (8.3) 1 (8.3) 1 (8.3)
Influenza A or B (n =/9) 5 (55) 4 (44.4) 5 (55.5) 0 0
Metapneumovirus (n =/5) 3 (60) 2 (40) 1 (20) 1 (20) 0
Adenovirus (n =/3) 2 (67) 1 (33) 1 (33) 0 0
Data are presented as number (percentage). No mixed cytomegalovirus-bacterial or Moraxella catarrhalis-viral infections were found.
Aetiology by place of treatment 8% of paediatric CAP, in line with our 5% finding

[10]. Fourth, our data confirm that a high number of
The age distribution of hospitalized (27 children,
CAP (30%) are caused by mixed infections [4,6] and
27%) and ambulatory (74 children, 73%) patients is
that mixed viral-bacterial infections are common not
shown in Table I; the hospitalization rates did not
only in children less than 2 y of age, as suggested
differ significantly between the 3 age groups. 21
earlier [6,21], but also in older children up to 5 y of
(33%) of the 63 children aged B5 y (Table V) and 6
age.
(16%) of those 38 aged ]5 y were treated in the
hospital. The majority of S. pneumoniae CAP cases Pneumococcal CAP was common at all ages, in
(15/18, 83%) were treated as outpatients. 18 (66%) agreement with previous studies [2,7]. However,
of the 27 children with M. pneumoniae infection pneumococcal aetiology was less frequent (18%)
were treated as outpatients in the ]5 y age group than recently found by Wubbel et al. (27%) [7] and
whereas, in the younger age group, 6 of the 9 cases Juven et al. (37%) [2]; these authors applied both
(p B0.05) were treated in the hospital (Table V). antibody and immune complex assays in paired sera
Seven of the 8 C. pneumoniae cases were treated as [2,7]. Pneumococcal immune complex assays have
outpatients. The majority of RSV cases, (17/29, been criticized for low specificity, complicated la-
85%) were treated as outpatients, though nearly all boratory technique and difficulty to standardize the
were B5 y of age.
/
method [24]. Likewise, pneumococcal antigen de-
tection tests, though promising in adults, are not
sufficiently sensitive and specific for paediatric use;
Discussion for example, tests for urinary antigen have correlated
There are 4 main results in the present study on the more with nasopharyngeal carriage than with aetiol-
aetiology of CAP in childhood. First, the study ogy of CAP [25]. Nasopharyngeal colonization does
confirms the importance of S. pneumoniae at all not cause seroconversion between paired sera, which
ages and in both hospital [3,6,8,21,22] and ambu- was the approach of the present study. On the other
latory settings [4,7]. Secondly, as previously ob- hand, pneumococcal antibody tests are, mainly due
served in Italian children [23], M. pneumoniae was to the lack of a gold standard, insufficiently validated
the most common aetiological agent of CAP from for clinical practice. In addition, they are time-
the age of 2 y onwards. Thirdly, we confimed, by consuming, requiring antibody measurements in
serological means, the involvment of hMPV in paired sera. PCR and other methods for pneumo-
paediatric CAP. In a recent PCR-based study, coccal nucleic acids are very sensitive and appear to
Williams et al. observed that hMPV was found in be the most promising approach for the aetiology of
CAP in children [26 − 28]. /
Table V. Aetiology of community-acquired pneumonia in the 63

With regard to M. pneumoniae, the high rate
hospitalized and ambulatory patients less than 5 y of age. found (27%) is in line with the result of Esposito et
al. (33.5%) [23], who supplemented serological tests
Inpatients, Outpatients, by PCR. The high proportion of M. pneumoniae
n = 21
/ n =42 p infections found in the present study may be
S. pneumoniae 2 (9.5) 9 (21.4) 0.31
explained by epidemiological reasons and by the
M. pneumoniae 6 (28.6) 3 (7.1) 0.049 fact that 2 serological assays were used. In line with
C. pneumoniae 0 1 (2.4) 1.00 PCR-based studies, mycoplasmal infection was
RSV 4 (19.0) 10 (23.8) 0.76 common also in young children [29]. Chlamydia
0.35 pneumoniae was found in 8% of CAP, mainly in
Mixed infections 4 (19.0) 13 (30.9)
Unknown aetiology 9 (42.8) 13 (30.9) 0.41
mixed infections and in children aged over 5 y, as
Data are presented as number (percentage). observed also in previous studies [23,30]. The
results stress the role of atypical bacteria in CAP in only in school-aged outpatients but also in pre-
school-aged children. school-aged inpatients. The recently found patho-
Although the viral diagnoses were based only on gen, hMPV, is the causative agent in B 10% of /
serum antibody rises by EIA, viral infections were paediatric pneumonia. If the treatment of CAP is
common in young children (Table II) [4]. Viral cases started with macrolide, the clinical response should
were missed, particularly in B2-y-old children, due be carefully monitored, since macrolides are not
to the lack of direct detection of viruses in respira- optimal drugs for pneumococcal infections.
tory samples. For hMPV, we noticed similar fre-
quencies, hospitalization rates and occurrence in
Acknowledgements
winter-time as were found in a recent study on lower
respiratory tract infection in children, utilizing PCR We thank Dr. Francesca Valent (Department of
in nasal-wash specimens. [10]. In 2 other recent Hygiene, School of Medicine DPMSC, University
studies based on detection of hMPV by PCR in of Udine, Italy) for statistical support.
upper airways of children with lower respiratory tract
infection [11,12], the proportion of CAP with
hMPV aetiology varied from 12.7% [11] to 15.5% References
[12]. [1] British Thoracic Society Standards of Care Committee. BTS
The fairly high number of mixed infections (30%) Guidelines for the Management of Community Acquired
in the present study is in line with recent studies. As Pneumonia in Childhood. Thorax 2002;57(Suppl 1):i1 −23.
[2] Juven T, Mertsola J, Toikka P, Virkki R, Leinonen M,
observed by Juven et al. [6] and Korppi et al. [8], Ruuskanen O. Clinical profile of serologically diagnosed
mixed viral-bacterial infections were common (32%) pneumococcal pneumonia. Pediatr Infect Dis J 2001;20:
in children less than 2 y of age, reflecting the role of 1028 −/33.
RSV in mixed infections. Moreover, we found that [3] Claesson BA, Trollfors B, Brolin I, Granstrom M, Henrich-
mixed infections were common (16%) up to 5 y of sen J, Jodal U, et al. Aetiology of community-acquired
pneumonia in children based on antibody responses to
age, reflecting the role of M. pneumoniae in dual bacterial and viral antigens. Pediatr Infect Dis J 1989;8:
bacterial infections. 856 −/61.
The convalescent serum sample was collected on [4] Heiskanen-Kosma T, Korppi M, Jokinen C, Kurki S,
average 35 d after the CAP diagnosis, and this time Heiskanen L, Juvonen H, et al. Aetiology of childhood
may be too long for testing pneumococcal antibodies pneumonia: serological results of a prospective, population-
based study. Pediatr Infect Dis J 1998;17:986 −/91.
but too short for testing mycoplasmal and chlamy- [5] Vuori E, Peltola H, Kallio M, Leinonen M, Hedman K, and
dial antibodies. It has been shown [13] that half of the SE-TU Group. Aetiology of pneumonia and other
the children hospitalized for acute lower respiratory common childhood infections requiring hospitalization and
tract infection develop pneumococcal antibody re- parenteral antimicrobial therapy. Clin Infect Dis
sponse within 5 d from admission, while antibodies 1998;27:566 −/72.
[6] Juven T, Mertsola J, Waris M, Leinonen M, Meurman O,
to intracellular bacteria, chlamydial antibodies in Roivainen M, et al. Aetiology of community-acquired
particular, may need 6 weeks or more to show pneumonia in 254 hospitalized children. Pediatr Infect Dis
diagnostic increases [31]. On the other hand, 2 J 2000;19:293 −/8.
different IgG antibody assays were available for S. [7] Wubbel L, Muniz L, Ahmed A, Trujillo M, Carubelli C,
pneumoniae and 2 different IgM and IgG antibody McCoig C, et al. Aetiology and treatment of community-
acquired pneumonia in ambulatory children. Pediatr Infect
assays for M. pneumoniae and C. pneumoniae. Dis J 1999;18:98 −/104.
Therefore, we consider that the results, for bacterial [8] Korppi M, Heiskanen-Kosma T, Jalonen E, Saikku P,
aetiology in particular, reflect the current situation in Leinonen M, Halonen P, Makela PH. Aetiology of commu-
Italy, thus corroborating previous data from Finland nity-acquired pneumonia in children treated in hospital. Eur
and Texas, USA. Since both hospitalized and Respir J 1993;152:24 −/30.
[9] van der Hoogen BG, de Jong JC, Groen J, Kuiken T, de
ambulatory patients were included, the results re- Groot R, Fouchier RA, Osterhaus AD. A newly discovered
present the whole spectrum of paediatric CAP. In human pneumovirus isolated from young children with
addition, a wide spectrum of pathogens causing respiratory tract disease. Nat Med 2001;7:719 −/24.
paediatric CAP, including the recently found virus [10] Williams JV, Harris PA, Tollefson SJ, Halburnt-Rush LL,
hMPV, was covered by the used serological methods. Pingsterhaus JM, Edwards KM, et al. Human metapneu-
movirus and lower respiratory tract disease in otherwise
In conclusion, besides the importance of S. healthy infants and children. N Engl J Med 2004;350:443 −/
pneumoniae in paediatric CAP at all ages and in 50.
both hospital and ambulatory settings, we found a [11] Viazov S, Ratjen F, Scheidhauer R, Fiedler M, Roggendorf
high number of mycoplasmal infections from 2 y of M. High prevalence of human metapneumovirus infection in
age onwards. This result permits to suggest that young children and genetic heterogeneity of the viral isolates.
J Clin Microbiol 2003;41:3043 −/5.
macrolides should be considered as an empirical [12] Maggi F, Pifferi M, Vatteroni M, Fornai C, Tempestini E,
antimicrobial treatment for uncomplicated CAP, not Anzilotti S, et al. Human metapneumovirus associated with
812 M. Don et al.
respiratory tract infection in a 3-y Study of nasal swabs from [23] Esposito S, Blasi F, Bellini F, Allegra L, Principi N, and the
infants in Italy. J Clin Microbiol 2003;41:2987 −/91. Mowgli Study Group. Mycoplasma pneumoniae and Chla-
[13] Nohynek H, Eskola J, Kleemola M, Jalonen E, Saikku P, mydia pneumoniae infections in children with pneumonia.
Leinonen M. Bacterial antibody assays in the diagnosis of Eur Respir J 2001;17:241 −/5.
acute lower respiratory tract infection in children. Pediatr [24] Musher DM, Mediwala R, Phan HM, Chen G, Baughn RE.
Infect Dis J 1995;14:478 −/84. Non-specificity of assaying for IgG antibody to pneumolysin
[14] Wang S. The microimmunofluorescence test for Chlamydia in circulating immune complexes as a means to diagnose
pneumoniae infection: technique and interpretation. J Infect pneumococcal pneumonia. CID 2001;32:534 −/8.
Dis 2000;181(Suppl 3):S421 −/5. [25] Dom`ınguez J, Blanco S, Rodrigo C, Azuara M, Gal`ı N,
[15] Persson K, Boman J. Comparison of 5 serological tests for Mainou A, et al. Usefulness of urinary antigen detection by
diagnosis of acute infections by Chlamydia pneumoniae. an immunochromatographic test for diagnosis of pneumo-
Clin Diagn Lab Immunol 2000;7:739 −/44. coccal pneumonia in children. J Clin Microbiol 2003;41:
[16] Kenny GE, Grayston JT. Eaton pleuropneumonia-like or-
2161 −/3.
ganism (Mycoplasma pneumoniae) complement-fixing anti-
[26] Toikka P, Nikkari S, Ruukanen O, Leinonen M, Mertsola J.
gen: extraction with organic solvents. J Immunol 1965;95:
Pneumolysin PCR-based diagnosis of invasive pneumococ-
19 −25.
cal infection in children. J Clin Microbiol 1999;37:633 −7.
[17] Korppi M, Heiskanen-Kosma T, Kleemola M. Incidence of
[27] Michelow IC, Lozano J, Olsen K, Goto C, Rollins NK,
community-acquired pneumonia in children caused by
Ghaffar F, et al. Diagnosis of Streptococcus pneumoniae
Mycoplasma pneumoniae: serological results of a prospec-
tive, population-based study in primary health care. Respir- lower respiratory infection in hospitalized children by
ology 2004;9:109 −/14. culture, polymerase chain reaction, serological testing, and
[18] Vuorinen T, Meurman O. Enzyme immunoassays for detec- urinary antigen detection. CID 2002;34:e1 −/11.
tion of IgG and IgM antibodies to parainfluenza types 1, 2 [28] Michelow IC, Olsen K, Lozano J, Rollins NK, Duffy LD,
and 3. Journal of Virological Methods 1989;23:63 −/70. Ziegler T, et al. Epidemiology and clinical characteristics of
[19] Ziegler T, Lindholm T, Scalia G. Causes of false-positive community-acquired pneumonia in hospitalized children.
reactions in herpes virus IgM assays. Serodiagnosis and Pediatrics 2004;113:701 −7.
Immunotherapy in Infectious Diseases 1989;3:101 −/9. [29] Principi N, Esposito S, Blasi F, Allegra L, and the Mowgli
[20] Laine M, Vainionpää R, Uksila J, Oksi J, Nissilä M, Kaarela Study Group. Role of Mycoplasma pneumoniae and Chla-
K, et al. Prevalence of Sindbis-related (Pogosta) virus mydia pneumoniae in children with community-acquired
infections in patients with arthritis. Clin Exp Rheumatol lower respiratory tract infections. CID 2001;32:1281 −/9.
2003;21:213 −6. [30] Heiskanen-Kosma T, Korppi M, Laurila A, Jokinen C,
[21] Nohynek H, Eskola J, Laine E, Halonen P, Ruutu P, Saikku Kleemola M, Saikku P. Chlamydia pneumoniae is an
P, et al. The causes of hospital-treated acute lower respira- important cause of community-acquired pneumonia in
tory tract infection in children. Am J Dis Child 1991;145: school-aged children: serological results of a prospective,
618 −/22. population-based study. Scand J Infect Dis 1999;31:255 −/9.
[22] Ruuskanen O, Nohynek H, Ziegler T, Capeding R, [31] Mazzulli T. Laboratory diagnosis of infection due to viruses,
Rikalainen H, Huovinen P, Leinonen M. Pneumonia in Chlamydia, and Mycoplasma. In: Long. Principles and
childhood: aetiology and responses to antimicrobial therapy. Practice of Paediatric Infectious Diseases, 2nd edn, 2003.
Eur J Clin Microbiol Infect Dis 1992;11:217 −/23. p. 1394.
II
Don M, Valent F, Korppi M, Falleti E, De Candia A, Fasoli L, Tenore A and Canciani M.

Efficacy of serum procalcitonin in evaluating severity of community-acquired pneumonia in
childhood. Scandinavian Journal of Infectious Diseases 2007;39(29):129-137
Scandinavian Journal of Infectious Diseases, 2007; 39: 129 −137
ORIGINAL ARTICLE
Efficacy of serum procalcitonin in evaluating severity of

community-acquired pneumonia in childhood
MASSIMILIANO DON 1, FRANCESCA VALENT2, MATTI KORPPI3, EDMONDO

FALLETI4, ALESSANDRO DE CANDIA 5, LOLITA FASOLI1, ALFRED TENORE1 &
MARIO CANCIANI1
From the Departments of 1Paediatrics, 2Hygiene and 4Laboratory, 5Institute of Radiology, School of Medicine, DPMSC,
University of Udine, Italy, and 3Department of Paediatrics, Kuopio University and University Hospital, Kuopio, Finland
Abstract
Microbe-specific diagnosis of community-acquired pneumonia (CAP) in childhood is difficult in clinical practice. Chest
radiographs and non-specific inflammatory markers have been used to separate presumably bacterial from viral infection but
the results have been inconsistent. The aim of the present study was to evaluate the usefulness of procalcitonin (PCT) in
assessing the severity as well as the bacterial or viral aetiology of CAP. Serum PCT was measured by an immunolumino-
metric assay in 100 patients with CAP; 26 were treated as inpatients and 74 as outpatients. The pulmonary infiltrate was
considered to be alveolar in 62 and interstitial in 38 cases, according to the radiological diagnosis. The bacterial and viral
aetiology of pneumonia was studied by an extensive serological test panel. No differences were found in PCT concentrations
between the 4 aetiological (pneumococcal, atypical bacterial, viral, unknown) and the 3 age ( B/2, 2 −4 and ]/5 y) groups.
Serum PCT was >/0.5 ng/ml in 69%, >1.0 ng/ml in 54% and >/2.0 ng/ml in 47% of all patients. PCT was higher in
patients that were admitted than as outpatients (medians 17.81 vs 0.72 ng/ml, respectively, p B/0.01) and higher in alveolar
than in interstitial pneumonia (medians 9.43 vs 0.53 ng/ml, respectively, p B0.01). In conclusion, serum PCT values were
found to be related to the severity of CAP in children even though they were not capable, at any level of serum
concentration, to differentiate between bacterial and viral aetiology.
Introduction low in viral infections [2,7] and in autoimmune

inflammatory diseases [9] without superimposed
Procalcitonin (PCT), the prohormone of calcitonin,
bacterial infections, even if a study on adults showed
is a 116-aminoacid protein recognized as a non-
that PCT does not clearly discriminate severe
specific marker of bacterial infection [1] and its
infections from a systemic inflammatory syndrome
serum concentration seems to correlate with the [1,2,10]. The authors of a recent systematic review
severity of microbe invasion [2,3]. In fact, serum concluded that PCT is superior to other non-specific
PCT reaches much higher concentrations in inva- inflammatory markers, such as the widely used C-
sive, bacteraemic infections, such as bacterial me- reactive protein (CRP), in differentiating bacterial
ningitis [4−7] or septicaemia [2,7,8] than in less infections both from viral infections and non-
invasive, localized, non-bacteraemic infections, re- infectious inflammatory diseases, among patients
presented by most cases of urinary tract infection or hospitalized for suspected bacterial infectious dis-
pneumonia [7]. eases [11].
Serum PCT has been used as a non-specific Community acquired pneumonia (CAP) is a
laboratory marker for distinguishing between viral common localized bacterial infection in childhood,
and bacterial infections and to help decide whether caused by viruses, ‘typical’ respiratory bacteria
or not to prescribe antibiotics. Earlier studies have (as Streptococcus pneumoniae) or ‘atypical’ respira-
suggested that serum PCT concentrations remain tory bacteria (as Mycoplasma pneumoniae and
Correspondence: M. Don, Paediatric Department, School of Medicine DPMSC, University of Udine P.le S. Maria della Misericordia, 33100 Udine, Italy. Tel:
‡/39 0432 559240-4. Fax: ‡/39 0432 559258. E-mail: max.don@libero.it
(Received 11 May 2006; accepted 13 June 2006)

ISSN 0036-5548 print/ISSN 1651-1980 online # 2007 Taylor & Francis
DOI: 10.1080/00365540600951283
130 M. Don et al.
Chlamydia pneumoniae). Clinical features, radiolo- being males. 19% were B24 months old, 43% were
/
gical findings and acute phase reactants have proved between 2 and 5 y and 38% were ]5 y old. 74
to be poor indicators of the aetiology of pneumonia children were treated as outpatients and 26 as
in children [12,13]. Microbe-specific diagnosis, inpatients. Three y later, in 2005, 3 experienced
based on culture of blood, fluid or samples obtained radiologists, unaware of the aetiology and the inter-
directly from the focus of infection in the lungs, is pretation of the initial radiographs, individually and
possible only in a small minority of CAP cases. In independently evaluated all the radiographs and the
studies published over the last 15 to 20 y, serological final radiological diagnosis was considered when at
tests based on antigen, antibody and immune com- least 2 of them agreed. At that evaluation, 62 cases
plex detection have been used for microbe-specific were interpreted as alveolar and 38 as interstitial
diagnosis of CAP in children [14 −16]. In these pneumonia.
studies, attention has also been paid to serum non-
specific inflammatory markers, such as CRP, ery-
throcyte sedimentation rate (ESR) and white blood Aetiological classification
cell (WBC) count, in trying to distinguish between The data collection and the aetiological findings
viral and bacterial CAP in children but the results have been published recently [25]. Streptococcus
have been inconsistent [17 −19]. Similarly, conflict- pneumoniae aetiology of CAP was diagnosed in
ing results exist concerning serum PCT: some 18 patients by significant increases of IgG antibodies
authors have neglected its role [20,21], some have to pneumococcal pneumolysin and/or pneumococcal
concluded that PCT has no additional value over capsular C-polysaccharide in paired sera, measured
CRP [22], while others have stressed its role as a by enzyme immunoassay (EIA) [26].
useful marker for identifying bacterial CAP in Mycoplasma pneumoniae aetiology of CAP was
basically healthy children [23,24]. diagnosed in 27 patients, in 5 cases by significant
In the present paper, we further analyse our increases in antibodies by complement fixation (CF)
recently published data in terms of aetiology of [27] and in 20 cases by significant increases in
CAP [25], in order to evaluate the value of serum antibodies by EIA (among them, 6 had diagnostic
PCT in assessing the severity of CAP and its rises in IgM, 6 in IgG and 8 in both IgM and IgG)
capability to distinguish bacterial from viral CAP in [28]. Correspondingly, C. pneumoniae aetiology was
children. diagnosed in 8 patients, in all cases by EIA [29] and
in 3 cases also by microimmunofluorescence [30].
The EIA test was also performed to measure Ig
Materials and methods (polyvalent) antibodies against whole bacterial cell
Study subjects antigens of non-typeable Haemophilus influenzae (3
cases) and Moraxella catarrhalis (1 case) [26].
This prospective study was conducted in the Haemophilus influenzae and M. catarrhalis were
Department of Paediatrics of the University of only found in mixed infections with viruses or other
Udine, School of Medicine, Italy, during a 15-month bacteria.
period, from 1 October 2001 to 31 December 2002. Viral studies consisted of the determination of IgG
The protocol was approved by the Ethics Committee antibodies to parainfluenza viruses type 1 −3, influ-
of the University of Udine, School of Medicine and enza A and B viruses and adenoviruses [31], IgG and
an informed consent was obtained from the parents IgM antibodies to cytomegalovirus, [32] and IgG
of all children. In total, 125 consecutive and antibodies to RSV and hMPV [33] by EIA in paired
previously healthy children were evaluated for the sera. Viral infection was demonstrated in 42 pa-
study. The patients in whom pneumonia was clini- tients; 22 cases were viral infections alone (including
cally suspected arrived at the Paediatric Emergency 3 dual viral infections) and 20 cases were mixed
Room (ER) of our hospital either autonomously or infections with S. pneumoniae, M. pneumoniae,
were sent by a general paediatrician. The detailed C. pneumoniae or H. influenzae.
inclusion and exclusion criteria for enrolment in the The 100 patients with CAP were classified into 4
study, as well as the radiological definition of aetiological groups. The pneumococcal group con-
pneumonia, have been published recently [25]. 25 sisted of all the 18 cases with S. pneumoniae
children were excluded, because 17 did not have aetiology (4 isolated and 14 mixed infections,
appropriate serum samples available and 8 had chest including viral coinfection in 10 cases, mycoplasmal
radiographs wihch were not diagnostic for pneumo- coinfection in 5 cases and M. catarrhalis coinfection
nia [25]. The median age of the 100 patients in 1 case). The atypical bacterial group comprised
with radiologically confirmed pneumonia that com- 25 cases, including 13 single and 12 mixed infections
prised the actual study group was 3.7 y with 49% caused by M. pneumoniae and/or C. pneumoniae
Serum procalcitonin and pneumonia in childhood 131
(viral coinfection in 9 cases and H. influenzae Statistical analysis

coinfection in 1 case; mycoplasmal-chlamydial coin-
Exploratory data analysis showed that PCT concen-
fection in 5 cases; pneumococcal coinfections not
trations were non-normally distributed. There-
included). The viral group consisted of 23 cases,
fore, the results are presented as medians and
including 19 isolated viral and 3 dual viral infections
25th −75th percentiles. In statistical analysis of the
and 1 case with RSV and non-typeable H. influenzae
data, Wilcoxon and Kruskal-Wallis tests were used
findings. RSV was the most common virus, being
diagnosed in 17 cases. The 34 cases with no sero- for continuous variables and x2 and Fisher’s exact
logical findings formed the unknown aetiology group tests for discrete variables. The diagnostic para-
(Figure 1). meters of sensitivity, specificity, predictive value
(PV) and likelihood ratio of a positive test result
(LR) were calculated by using routine equations.
Laboratory studies The LR value of >3 is considered to change
/
moderately and of >5 significantly pre-test prob-

/
Serum PCT was measured in serum samples ob-
tained on admission. The blood samples were ability of a presumed diagnosis. The receiver opera-
centrifuged and the serum was separated and stored tor characteristic (ROC) curve was used to evaluate
at —208C until assayed for PCT. Serum PCT levels the ability of serum PCT concentrations to discri-
were measured by immunoluminometric assay with minate pneumococcal and atypical CAP from viral
2 monoclonal antibodies (PCT LIA TEST, Brahms pneumonia. Multivariate logistic regression was
Diagnostics, Henningsdorf, Germany), 1 against the performed to analyse the association between PCT
catacalcin region of PCT and the other against ( ]1 vs B1 ng/ml) and aetiological diagnostic group
calcitonin, as described previously [6]. By this (pneumococcal, atypical bacterial, viral, unknown),
method the detection limit was 0.1 ng/ml and the the radiological diagnostic group (alveolar, intersti-
upper limit for normal serum PCT suggested by the tial), hospitalization (inpatient, outpatient) and age
manufacturer was 0.5 ng/ml. In this paper, we category (B2, 2 −4 and ]5 y). Multiple linear
/
applied this value, as well as the values of 1.0 ng/ml regression was used to evaluate the association
suggested by Moulin et al. [23] and 2.0 ng/ml between PCT expressed as a continuous and aetio-
suggested by Toikka et al. [22] and Prat et al. [24] logical group, hospitalization and radiological group,
as cut-off limits for elevated PCT concentration. after adjusting for age. The correlation between
In addition, serum CRP was measured by PCT concentration and age was expressed by Spear-
a particle-enhanced immunoturbidometric assay man’s correlation coefficient.
(CRPLX reagent on Hitachi-Roche Modular P-
System, Roche Diagnostics, Mannheim, Germany)
Results
and the limit of 60 mg/l was used as the cut-off limit
for an elevated concentration [18,20,21,23]. ESR The median (25th−75th percentile) serum PCT con-
and WBC count were studied by routine methods centration was 1.70 ng/ml (0.62 −20.53), 1.88 ng/ml
and the cut-off limits used were 30 mm/h for ESR (0.42 −17.23) and 0.83 ng/ml (0.39 −17.98) in the
and 15,000 u/l for WBC [18,20,23]. 3 age groups ( B2, 2 −4 and ]5 y), respectively
Community acquired
pneumonia (n = 125)
Enrolled patients (n = 101)
Eligible patients for the PCT

study (n = 100)
Mycoplasma - Unknown
Pneumococcal Viral
Chlamydia aetiology
(n = 18; 18%) (n = 23; 23%)
(n = 25; 25%) (n = 34; 34%)
Figure 1. Flow-chart of the present study.

132 M. Don et al.
(p =0.83). There was no correlation between serum Table II. Serum PCT concentrations in 100 children with CAP
PCT and age of the patients, when they were divided on the basis of the radiological findings.
considered as a continuous variable (rs =—0.14, / /
Radiological pattern
p =0.16). Within the youngest age group, the values
/
did not differ significantly between B12 months and Interstitial Alveolar p- value
12 −24-month-old children (data not shown). PCT
was >0.5 ng/ml in 69 children, >1.0 ng/ml in /
Median 0.53 9.43 0.0003a
54 children and >2.0 ng/ml in 47 children. Thus, 25 th−75 th percentile 0.31 −1.04 0.54 −22.87
serum PCT concentration B0.5 ng/ml, suggested as
/
Serum PCT
normal by the manufacturer, was found in 31% of the B/0.5 ng/ml 18 (47) 13 (21) )
CAP patients. In the only patient with blood culture 0.5 −1 ng/ml 8 (21) 7 (11) b
B
positive to S. pneumoniae, serum PCT level was 1.0 −2.0 ng/ml 4 (11) 3 (5) / 0.0001
9.66 ng/ml. >/2.0 ng/ml 8 (21) 39 (63)
The severity of pneumonia, reflected by the Data are presented as number (percentage).
need for hospital treatment and the presence of an a
Wilcoxon 2-sample test; bFisher’s exact test.
alveolar infiltration on chest radiograph, was asso-
ciated with a significant elevation of median serum aetiology (mixed infections with C. pneumoniae
PCT concentration compared to outpatients and excluded) (p=0.93). The results on association
children with interstitial CAP (Tables I, II). Like- between serum PCT concentration and microbe-
wise, the difference between these groups was specific aetiology remained negative when analysed
statistically significant, when PCT data were ana- by multiple linear regression adjusted for age (data
lysed as categorized by 0.5 ng/ml, 1.0 ng/ml and not shown).
2.0 ng/ml limits (Tables I, II). In multiple linear PCT concentrations ranged from 0.11 to 74.99 ng/
regression adjusted for age, the need for hospital ml, except for 1 child with a very high PCT
treatment remained significant, but the presence of (330.80 ng/ml). This patient was a 37-months-old
alveolar pattern of pneumonia lost its significance girl, admitted for pleural empyema caused by group-
(data not shown). A beta-haemolytic Streptococcus and all her non-
As seen in Table III, median PCT concentrations specific inflammatory markers were very high (WBC
varied from 0.74 ng/ml (viral group) to 6.37 ng/ml 31400/ml, neutrophils 92%, ESR 95 mm/h and
(atypical bacterial group) in the 4 aetiological groups CRP 426.60 mg/l). Serum PCT concentration was
(p=0.43). Serum PCT was >1.0 ng/ml in 55.5% of/ >35 ng/ml in 9 additional cases. Among these
the patients in the pneumococcal, in 64.0% of those 10 patients with high serum PCT, 5 were treated in
in the atypical bacterial and in 39.1% of those in the hospital, all but 1 had an alveolar infiltration on chest
viral CAP group, respectively (p=0.24). Median radiograph and the aetiology of CAP varied (pneu-
PCT concentration was 6.39 ng/ml (25th−75th per- mococcal in 3, mycoplasmal in 2, single viral in 2 and
centile 0.47 −19.35) in the 8 patients with chlamydial unknown aetiology in 3 cases). The median values of
aetiology (mixed infections with M. pneumoniae serum CRP, blood ESR and WBC were 294.30 mg/l,
included) and 6.37 ng/ml (25th−75th percentile 100.50 mm/h and 28000.00/ml, respectively. We
0.36 −18.39) in the 17 patients with mycoplasmal made supplementary analyses by excluding first the
case with PCT >330 ng/ml and then all the 10 cases
/
with serum PCT >35 ng/ml and the result remained

/
Table I. Serum PCT concentrations in 100 children with CAP
similar; there was no significant association between
divided on the basis of the setting of treatment.
serum PCT concentration and the aetiology of CAP
Place of treatment
(data not shown).
Serum PCT values of 0.5 ng/ml, 1.0 ng/ml
Outpatients Inpatients p- value and 2.0 ng/ml were also tested as screening limits
between pneumococcal and viral pneumonia
Median 0.72 17.81 0.0012a (Table IV). In this analysis, mixed viral-pneumococ-
25 th−75 th percentile 0.39 −7.75 2.63 −28.67 cal infections were included into the pneumococcal
Serum PCT CAP group. The best LR value was 1.69 at the PCT
B0.5 ng/ml 27 (36) 4 (15)
level of 2.0 ng/ml, corresponding to a 44% sensitivity
0.5 −1 ng/ml 13 (18) 2 (8) ) and a 74% specificity. A similar analysis was made
1.0 −2.0 ng/ml 7 (9) 0 (0) between atypical and viral pneumonia (Table V).
>2.0 ng/ml 27 (36) 20 (77) 0.0049b
Mixed viral-mycoplasmal and mixed viral-chlamydial
Data are presented as number (percentage). cases were included in the atypical bacterial group
a
Wilcoxon 2-sample test; bFisher’s exact test. and all pneumococcal cases were excluded from this
Table III. Serum PCT concentrations in 100 children with CAP divided on the basis of the aetiology.
Aetiological groups of infection
Pneumococcal Atypical Viral Unknown p- value
Number of subjects 18 25 23 34
Median 1.21 6.37 0.74 5.78 0.43a
25th−75th percentile 0.62 −33.07 0.36 −18.39 0.39 −2.09 0.46 −17.04
Serum PCT )
B/0.5 ng/mlc 4 (22) 7 (28) 8 (35) 12 (35)
0.5 −1 ng/mlc 4 (22) 2 (8) 6 (26) 3 (9) 0.24 b
1.0 −2.0 ng/ml 2 (11) 1 (4) 3 (13) 1 (3)
>/2.0 ng/ml 8 (44) 15 (60) 6 (26) 18 (53)
Data are presented as number (percentage). aKruskal-Wallis test; bFisher’s exact test; cmixed infections were found in 7 of the 8 cases of
pneumococcal pneumonia with serum PCTB1 ng/ml.
analysis. Again, the best LR value was 2.31 at the Serum PCT concentration correlated significantly
PCT level of 2.0 ng/ml, corresponding to a 60% with CRP, ESR and PCT, the correlation being
sensitivity and a 74% specificity. Further, a similar good with CRP (rs =0.72) and WBC (rs =0.59) /
analysis was made between pneumococcal (coinfec- and only moderate with ESR (rs=0.43). As seen in
tions with atypical bacteria included) and atypical Table VI, a significant association was found be-
CAP (coinfections with pneumococci excluded). tween serum PCT and CRP or WBC count but not
The best LR value was 1.07 at the PCT level of between serum PCT and ESR.
0.5 ng/ml, corresponding to a 78% sensitivity and Finally, from the logistic regression model includ-
a 28% specificity (data not shown). Among the ing the aetiological group (pneumococcal, atypical
8 patients with pneumococcal pneumonia and serum bacterial, viral, unknown), place of treatment (in-
PCT B1 /ml (Table III), no one was treated in
/
patient and outpatient) and radiological group
hospital and developed complications, 3 of them had (alveolar and interstitial), the likelihood of having
an alveolar infiltration on chest radiograph and PCT ]1 ng/ml in patients with pneumococcal, aty-
mixed infections were found in all but 1 case (5 dual pical bacterial or viral aetiology was not significantly
and 2 treble mixed infections, with M. pneumoniae different from patients with unknown aetiology. It
was significantly higher in hospitalized children than
involved in 4 and viruses in 5 cases). The median
in those treated as outpatients (OR 3.6; 95% CI
values of serum CRP, blood ESR and WBC were
1.2 −10.9) and in children with an alveolar than with
59.50 mg/l, 70.00 mm/h and 15065.00/ml, respec-
an interstitial infiltrate in chest radiograph (OR 4.1;
tively.
95% CI 1.6 −10.4). The inclusion of age into the
The receiver operator characteristic (ROC) curve
model did not significantly modify the results (data
for serum PCT to separate bacterial CAP (pneumo-
not shown).
coccal and atypical cases combined) from viral CAP
had an area under curve of 0.629 (Figure 2). The
area under ROC curves for serum PCT to separate Discussion
pneumococcal from viral CAP, pneumococcal from The results of the present study confirm earlier
atypical CAP and atypical from viral CAP were also studies [20,21] indicating that serum PCT is unable
calculated and their values were 0.624, 0.511 and to distinguish between bacterial and viral CAP
0.632, respectively. in children, the conclusion being similar in both
Table IV. Diagnostic parameters for serum PCT concentrations in differentiating pneumococcal from viral pneumonia.
S. pneumoniae (n =/18) Sensitivity % Specificity % PV‡/ LR‡
Cut-off limits
>/0.5 ng/ml (n =/29) 14 77.78 34.78 48.28 1.20
>/1.0 ng/ml (n =/19) 10 55.56 60.87 52.63 1.41
>/2.0 ng/ml (n =/14) 8 44.44 73.91 57.14 1.69
PV‡/: predictive value of a positive test result; LR‡/: likelihood ratio of a positive result.
Four single and 14 mixed infections made up the S. pneumoniae group, while the viral group consisted of 19 single and 4 mixed infections
(for more details see the text).
134 M. Don et al.
Table V. Diagnostic parameters for serum PCT concentrations in differentiating atypical from viral pneumonia.
Mycoplasma and Chlamydia

(n =/25) Sensitivity % Specificity % PV‡/ LR‡/
Cut-off limits
>/0.5 ng/ml (n =/33) 18 72.00 34.78 54.55 1.11
>/1.0 ng/ml (n =/25) 16 64.00 60.87 64.00 1.64
>/2.0 ng/ml (n =/21) 15 60.00 73.91 71.43 2.31
PV‡/: predictive value of a positive test result; LR‡/: likelihood ratio of a positive result.
13 single and 12 mixed infections made up the atypical group, while the viral group consisted of 19 single and 4 mixed infections (for more
details see the text).
ambulatory and hospitalized children at different terial and viral pneumonia. Moulin et al. [23]
ages. Likewise, the ROC curves demonstrate the suggested the value of >1 ng/ml and Toikka et al.
/
absence of any cut-off limit useful to differentiate [22] the value of >2 ng/ml as cut-off limits useful
/
pneumococcal, mycoplasmal or chlamydial from for differentiating between bacterial and viral CAP.
viral pneumonia. On the other hand, our results These results are not in agreement with the present
suggest that serum PCT is a useful indicator of the paper, since the limits were exceeded in about 40%
severity of CAP in children when comparing the and 26% of viral CAP cases, respectively. In
PCT values between those who needed hospitaliza- accordance, the likelihood ratios at all cut-off limits
tion and those who did not, and in comparing were far from the values of >3 or >5, which modify
/ /
the alveolar and interstitial pulmonary involvment pre-test probabilities in clinical decision-making. In
on chest radiograph [7]. Moreover, very high values the studies of Moulin et al. [23] and Toikka et al.
( >35 ng/ml) of serum PCT were found only among
/ [22], serum PCT values were analysed only in
the patients with severe pneumonia, as found by hospitalized children, which may over-estimate the
Gendrel et al. [7] in children with bacteraemic and/ role of high values due to the over-representation of
or other invasive bacterial infections. severe cases. The studies of Korppi et al. [20,21]
Two most recent studies on serum PCT and CAP that were population-based and performed in pri-
in children [22,23] have concluded that PCT is an mary health care settings, gave a negative result.
informative marker for discriminating between bac- However, PCT determinations were performed
0.9
0.8
0.7
0.6
sensitivity
0.5
0.4
0.3
0.2
0.1
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
1-specificity
Figure 2. Receiver operator characteristic (ROC) curve for serum PCT concentrations in discriminating combined pneumococcal p lus
atypical CAP from viral pneumonia (area under ROC curve =/0.629).
Table VI. Relations between serum PCT concentrations and 3 non-specific inflammatory markers in 100 patients with CAP due to pneum-
ococcal, atypical, viral or unknown aetiology.
Serum PCT concentrations (ng/mL)
B/0.5 0.5 −1 > /1 p- valuea
CRP]/60 mg/lb 7 3 39
Σ
B0.0001
CRPB60/ mg/l 14 6 7 Σ
WBC ]15000/ulc 11 5 43
11 B0.0001
WBCB15000/ul 20 10
ESR ]/30 mm/hd 27 11 52 Σ 0.0737
0.0737
ESR B/30 mm/h 4 3 2
CRP: C-reactive protein; WBC: white blood cells; ESR: erythrocyte sedimentation rate.
x test.
a 2
b
76 patients, c100 patients, d99 patients.
many y after the study from frozen serum samples underlines the ability of PCT to relieve the severity
and this may have caused falsely low results. In those of pneumonia, assessed by the need of hospital
studies [20,21], serum PCT values tended to be care and the presence of an alveolar type of
lower in children under 2 of age compared with older pneumonia. The present results also stress the
children, a finding which is in disagreement with the clinical importance of very high serum PCT values
present data. (for instance >35 ng/ml), since the patients had
Elevated host-response markers, such as PCT severe pneumonia regardless of the aetiology. The
and CRP, may be caused by unrecognized bacterial median serum PCT detected in patients with alveo-
coinfection, albeit viral aetiology alone has been lar pneumonia and in hospitalized patients (9.43 and
established. In the present study a good corre- 17.81 ng/ml, respectively) are surprisingly similar
lation was found between PCT and other non- to that of 17.75 ng/ml reported by Gendrel et al.
specific host-response markers, even though no [7] in 46 paediatric cases of bacterial septicaemia or
evidence of bacterial aetiology was present. In a meningitis. Thus, like other non-specific inflamma-
recent double-blind placebo-controlled trial con- tion markers [17], PCT may reflect more the
ducted in a large number of children to evaluate invasiveness and severity than the aetiology of
pneumococcal vaccination, Madhi et al. [34] showed an infection. Recently, serum PCT level was found
that the pneumococcal conjugate vaccine is able to to be an interesting predictor for complication and
reduce pneumonia associated with respiratory virus mortality but not predictive of any specific bacterial
infections, presumably by preventing bacterial coin- agent in a population of adult patients admitted in an
fections. A significant number of cases of pneumonia intensive care unit for severe CAP [39].
in children, if they need hospital care, seem to As stated previously [25], our results on bacterial
have pneumococcal coinfection, though the infec- aetiology derive from serological assays that give an
tions primarily are of viral origin [35]. On the other indirect evidence of the aetiology of infection. These
hand, serological evidence of bacterial infection may tests have been used in specialized laboratories only
be present in mild respiratory infections of airways for research purposes and their clinical value has not
not needing any antibiotic treatment [12,17,26] yet been assigned [26]. Viral diagnoses were based
and, correspondingly, viral infections (for instance on serum antibody rises and, evidently, many cases
those caused by adenoviruses or influenza virus) may were lost due to the lack of direct detection of viruses
present with severe disease clinically resembling in respiratory samples [40]. Strengths of the present
more invasive bacterial than viral infection [12, paper are that a large panel of microbiological
36 −38]. methods was used, all CAP cases were radiologically
In earlier studies, serum PCT has been higher in confirmed and serum samples for PCT levels were
invasive bacterial infections, such as bacterial me- available in all children, including both mild and
ningitis [4 −7] or septicaemia [2,7,8], than in less severe cases, as well as cases treated as inpatients and
invasive localized bacterial infections, such as urin- outpatients.
ary tract infections or pneumonia [7]. In addition, In conclusion, this study confirms the inability
there is evidence that PCT may be better than other of serum PCT determination in differentiating
acute phase reactants, such as CRP, WBC, inter- between bacterial and viral aetiology of CAP in
leukin 6 and interferon-alpha, for distinguishing viral ambulatory and hospitalized children at different
from bacterial infections [5,7]. The present study ages. On the other hand, serum PCT concentra-
136 M. Don et al.
tions were high in the most severe pneumonia cases plasmic antibody-associated vasculitis) and invasive bacterial
and, in addition, PCT correlated well with other non- infection. Arthritis Rheum 1997;40:1250 −6.
[10] Suprin E, Camus C, Gacouin A, Le Tulzo Y, Lavoue S,
specific inflammatory markers. The measurement Feuillu A, et al. Procalcitonin: a valuable indicator of
of serum PCT seems to be useful in the assessment infection in a medical ICU? Intensive Care Med 2000;26:
of the severity of pneumonia, when the 1170 −1.
information has to be interpreted together with other [11] Simon L, Gauvin F, Amre DK, Saint-Louis P, Lacroix J.
clinical observations (such as general condition, Serum procalcitonin and C-reactive protein levels as markers
of bacterial infection: a systematic review and meta-analysis.
findings in clinical examination and in chest
Clin Infect Dis 2004;39:206 −17.
radiographs) and with other non-specific host-re- [12] British Thoracic Society Standards of Care Committee.
sponse markers (such as white blood cell count and British Thoracic Society Guidelines for the Management of
C-reactive protein). Community Acquired Pneumonia in Childhood. Thorax
2002;57 (Suppl 1):1 −24.
[13] McIntosh K. Community-acquired pneumonia in children.
Acknowledgements N Engl J Med 2002;346:429 −37.
[14] Heiskanen-Kosma T, Korppi M, Leinonen M. Serologically
We are grateful to Mika Paldanius and Maija indicated pneumococcal pneumonia in children: a popula-
Leinonen (National Public Health Institute, Oulu, tion-based study in primary care settings. APMIS 2003;111:
Finland), Raija Vainionpaa (Department of Virol- 945 −50.
ogy, University of Turku, Finland), Marjaana Klee- [15] Heiskanen-Kosma T, Korppi M, Jokinen C, Kurki S,
mola and Riitta Raty (Department of Microbiology, Heiskanen L, Juvonen H, et al. Aetiology of childhood
pneumonia: serological results of a prospective, population-
National Public Health Institute, Helsinki, Finland)
based study. Pediatr Infect Dis J 1998;17:986 −91.
for their valuable work on the aetiological research. [16] Korppi M, Leinonen M. Pneumococcal immune complexes
We also thank Francesco Toso and Martina Urbani in the diagnosis of lower respiratory infections in children.
(Institute of Radiology, School of Medicine, Pediatr Infect Dis J 1998;17:992 −5.
DPMSC, University of Udine, Italy) for the revision [17] Nohynek H, Valkeila E, Leinonen M, Eskola J. Erythrocyte
sedimentation rate, white blood cell count and serum C-
of all the radiographs.
reactive protein in assessing aetiological diagnosis of acute
lower respiratory infections in children. Pediatr Infect Dis J
1995;14:484 −90.
References [18] Korppi M, Heiskanen-Kosma T, Leinonen M. White blood
cells, C-reactive protein and erythrocyte sedimentation rate
[1] Gendrel D, Bohuon C. Procalcitonin as a marker of bacterial in pneumococcal pneumonia in children. Eur Respir J 1997;
infection. Pediatr Infect Dis J 2000;19:679 −88. 10:1125 −9.
[2] Assicot M, Gendrel D, Carsin H, Raymond J, Guilbaud J, [19] Heiskanen-Kosma T, Korppi M. Serum C-reactive protein
Bohuon C. High serum procalcitonin concentrations in cannot differentiate bacterial and viral aetiology of commu-
patients with sepsis and infection. Lancet 1993;341:515 −8. nity-acquired pneumonia in children in primary healthcare
[3] Karzai W, Oberhoffer M, Meier-Hellman A, Reinhart K. settings. Scand J Infect Dis 2000;32:399 −402.
Procalcitonin: a new indicator of the systemic response to [20] Korppi M, Remes S. Serum procalcitonin in pneumococcal
severe infections. Infection 1997;6:329 −34.
pneumonia in children. Eur Respir J 2001;17:623 −7.
[4] Gendrel D, Raymond J, Assicot M, Moulin F, Iniguez JL,
[21] Korppi M, Remes S, Heiskanen-Kosma T. Serum procalci-
Lebon P, et al. Measurement of procalcitonin levels in
tonin concentrations in bacterial pneumonia in children: a
children with bacterial or viral meningitis. Clin Infect Dis
negative result in primary healthcare settings. Pediatr Pul-
1997;24:1240 −2.
monol 2003;35:56 −61.
[5] Hatherill M, Tibby SM, Sykes K, Turner C, Murdoch IA.
Diagnostic markers of infection: comparison of procalcitonin [22] Toikka P, Irjala K, Juven T, Virkki R, Mertsola J, Leinonen
with C-reactive protein and leukocyte count. Arch Dis Child M, et al. Serum procalcitonin, C-reactive protein and
1999;81:417 −21. interleukin-6 for distinguishing bacterial and viral pneumo-
[6] Gendrel D, Assicot M, Raymond J, Moulin F, Francoual C, nia in children. Pediatr Infect Dis J 2000;19:596 −602.
Badoual J, et al. Procalcitonin as a marker for the early [23] Moulin F, Raymond J, Lorrot M, Marc E, Coste J, Iniguez J-
diagnosis of neonatal bacterial infection. J Pediatr 1996;128: L, et al. Procalcitonin in children admitted to hospital with
570 −3. community-acquired pneumonia. Arch Dis Child 2001;84:
[7] Gendrel D, Raymond J, Coste J, Moulin F, Lorrot M, 332 −6.
Guerin S, et al. Comparison of procalcitonin with C-reactive [24] Pratt C, Dominguez J, Rodrigo C, Gimenez M, Azuara M,
protein, interleukin 6 and interferon-alpha for differentiation Jimenez O, et al. Procalcitonin, C-reactive protein and
of bacterial vs viral infections. Pediatr Infect Dis J 1999;18: leukocyte count in children with lower respiratory tract
875 −81. infection. Pediatr Infect Dis J 2003;22:963 −7.
[8] Chiesa C, Panero A, Rossi NA, Stegagno M, de Giusti M, [25] Don M, Fasoli L, Paldanius M, Vainionpää R, Kleemola M,
Osborn JF, et al. Reliability of procalcitonin concentrations Räty R, et al. Aetiology of community-acquired pneumonia:
for the diagnosis of sepsis in critically ill neonates. Clin Infect serological results of a paediatric survey. Scand J Infect Dis
Dis 1998;26:664 −72. 2005;37:806 −12.
[9] Eberhard OK, Haubitz M, Brunkhorst FM, Kliem V, Koch [26] Nohynek H, Eskola J, Kleemola M, Jalonen E, Saikku P,
KM, Brunkhorst R. Usefulness of procalcitonin for differ- Leinonen M. Bacterial antibody assays in the diagnosis of
entiation between activity of systemic autoimmune disease acute lower respiratory tract infection in children. Pediatr
(systemic lupus erythematosus/systemic antineutrophil cyto- Infect Dis J 1995;14:478 −84.
[27] Kenny GE, Grayston JT. Eaton pleuropneumonia-like [35] Korppi M. Mixed microbial aetiology of community-ac-
organism (Mycoplasma pneumoniae) complement-fixing quired pneumonia in children. APMIS 2002;110:515 −22.
antigen: extraction with organic solvents. J Immunol 1965; [36] Mistchenko AS, Diez R, Mariani AL, Robaldo J, Maffey AF,
95:19 −25. Bayley-Bustamante G, et al. Cytokines in adenoviral disease
[28] Korppi M, Heiskanen-Kosma T, Kleemola M. Incidence of in children: association of interleukin-6, interleukin-8, and
community-acquired pneumonia in children caused by tumour necrosis factor alpha levels with clinical outcome. J
Mycoplasma pneumoniae: serological results of a prospec- Pediatr 1994;124:714 −20.
tive, population-based study in primary health care. Respir- [37] Vuori E, Peltola H, Kallio MJY, Leinonen M, Heidman K,
ology 2004;9:109 −14. SE-TU Study Group. Aetiology of pneumonia and other
[29] Persson K, Boman J. Comparison of 5 serological tests for common childhood infections requiring hospitalization and
diagnosis of acute infections by Chlamydia pneumoniae.
parenteral antimicrobial therapy. Clin Infect Dis 1998;27:
Clin Diagn Lab Immunol 2000;7:739 −44.
566 −72.
[30] Wang S. The microimmunofluorescence test for Chlamydia
[38] Haterill M, Sykes K, McIntyre AG, Murdoch IA. Procalci-
pneumoniae infection: technique and interpretation. J Infect
tonin may help differentiate disseminated herpes simplex
Dis 2000;181(Suppl 3):S421 −5.
[31] Vuorinen T, Meurman O. Enzyme immunoassays for detec- viral from bacterial sepsis in neonates. Eur J Pediatr 2000;
tion of IgG and IgM antibodies to parainfluenza types 1, 2 159:168 −9.
and 3. Journal of Virological Methods 1989;23:63 −70. [39] Boussekey N, Leroy O, Georges H, Devos P, d’Escrivan T,
[32] Ziegler T, Lindholm T, Scalia G. Causes of false-positive Guery B. Diagnostic and prognostic values of admission
reactions in herpes virus IgM assays. Serodiagnosis and procalcitonin levels in community-acquired pneumonia in an
Immunotherapy in Infectious Diseases 1989;3:101 −9. intensive care unit. Infection 2005;33:257 −63.
[33] Laine M, Vainionpää R, Uksila J, Oksi J, Nissilä M, Kaarela [40] Jennings LC, Anderson TP, Werno AM, Beynon KA,
K, et al. Prevalence of Sindbis-related (Pogosta) virus Murdoch DR. Viral aetiology of acute respiratory tract
infections in patients with arthritis. Clin Exp Rheumatol infections in children presenting to hospital: role of poly-
2003;21:213 −6. merase chain reaction and demonstration of multiple infec-
[34] Madhi SA, Klugman KP and The Vaccine Trialist Group. A tions. Pediatr Infect Dis J 2004;23:1003 −7.
role for Streptococcus pneumoniae in virus-associated pneu-
monia. Nat Med 2004;10:811 −3.
III
Fasoli L, Paldanius M, Don M, Valent F, Vetrugno L, Korppi M and Canciani M. Simkania

negevensis in community-acquired pneumonia in Italian children. Scandinavian Journal of
Infectious Diseases 2008;40(3):269-272
Scandinavian Journal of Infectious Diseases, 2008; 40: 269−272
SHORT COMMUNICATION
Simkania negevensis in community-acquired pneumonia in Italian

children
LOLITA FASOLI1, MIKA PALDANIUS2, MASSIMILIANO DON 1, FRANCESCA VALENT 3,

LUIGI VETRUGNO1, MATTI KORPPI4 & MARIO CANCIANI1
From the 1Paediatric Department, 3Hygiene Department, School of Medicine, DPMSC, University of Udine, Italy, 2National
Public Health Institute, Oulu, Finland, 4Paediatric Research Centre, Tampere University and University Hospital, Tampere,
Finland
Abstract
Simkania negevensis, a recently found Chlamydia-like organism, has been associated with respiratory infections in children
and adults with pneumonia, but S. negevensis findings have been common also without any infection. The aims of the
present paper were to evaluate S. negevensis in the aetiology of paediatric community-acquired pneumonia (CAP), its
seroprevalence in north Italian children, and whether there is cross-reactivity between S. negevensis and Chlamydia
pneumoniae serology. Antibodies to S. negevensis were measured by microimmunofluorescence (MIF) in 101 frozen paired
sera obtained from children with CAP. Serological evidence (>/=4-fold increase or decrease in IgM or IgG) of acute S.
negevensis infection was achieved in 5 (5%) cases. Two were mixed infections with Mycoplasma pneumoniae and 1 with
respiratory syncytial virus. In total, 20−30% of the children had measurable antibodies to S. negevensis, with no association
with age. No cross-reactivity was observed between antibodies to S. negevensis and C. pneumoniae. S. negevensis appears to
be a real, though rare, cause of CAP in children.
Introduction moniae aetiology by either microimmunofluores-

cence (MIF) or enzyme immunoassay (EIA) was
Simkania negevensis is a recently discovered, intra-
found in 8% of the children [13]. However, the
cellular, penicillin-resistant microorganism [1]. It
specificity of the EIA test has been questioned [9]. In
belongs to a new family, Simkaniaceae, in the order
2005, serological studies were supplemented by MIF
of Chlamydiales [2], sharing many characteristics to S. negevensis in frozen samples.
with Chlamydia species. S. negevensis appears to be The aims of the present paper were to evaluate the
common in the environment, at least in certain role of S. negevensis in the aetiology of paediatric
geographical areas, and to be transmitted via dome- CAP, its seroprevalence in north Italian children and
stic water [3]. Seropositivity rates in healthy adults the possible cross-reactivity between S. negevensis
have been 46% to 80% in Israel, Denmark and the and C. pneumoniaie serology.
UK [4−6], but only 4% in Japan [7]. In preschool-
aged children, the figures have been 13−57% in
Israel [8] and 11−23% in Finland [9]. S. negevensis Materials and methods
has been associated with a variety of respiratory The aetiology of CAP in children was prospectively
illnesses, including bronchiolitis in children [10,11] studied at the Department of Pediatrics, University
and pneumonia in adults [12]. Hospital of Udine, Italy, during a 15-month period in
S. negevensis can be detected in clinical samples 2001−2002, as described recently [13]. In brief, 101
by culture, polymerase chain reaction (PCR) and children admitted to the emergency room of the
serology [6,8]. In our recent study on community hospital for CAP attended the serological study. On
acquired pneumonia (CAP), evidence of C. pneu- admission all patients underwent a clinical evaluation,
Correspondence: M. Don, Paediatric Department, School of Medicine DPMSC, University of Udine P.le S. Maria della Misericordia, 33100 Udine, Italy.
(Tel: ‡39 0432 559240-4. Fax: ‡39 0432 559258. E-mail: max.don@libero.it)
(Received 17 May 2007; accepted 12 August 2007)

ISSN 0036-5548 print/ISSN 1651-1980 online # 2008 Informa UK Ltd. (Informa Healthcare, Taylor & Francis As)
DOI: 10.1080/00365540701642146
270 L. Fasoli et al.
chest radiographs were taken, and acute and con- stitute, Oulu, Finland) [9,17]. The detection limits,
valescent blood samples were collected on average used in the assessment of seroprevalence, were 1:8
35 d (range 18−56) apart. Inclusion criteria were a for IgG and 1:10 for IgM antibodies. The diagnostic
clinical picture compatible with respiratory infection criteria of S. negevensis infection were]4-fold or
and radiological infiltrations consistent with pneumo- greater increases or decreases in IgG or IgM anti-
nia. Exclusion criteria were B1 month of age, chronic bodies between paired sera.
underlying illness, wheezing on admission and pneu- Fisher’s exact test was used in the statistical
monia acquired in hospital [13]. analysis of the data.
The original protocol was approved by the ethics
committee of the School of Medicine DPMSC, Results
University of Udine and informed consent was
obtained from the parents of all children. Serological evidence for acute S. negevensis infec-
The paired sera were stored at —208C until tion was found in 5/101(4.9%) children, based on
studied for antibodies to 9 viruses and 5 bacteria IgM antibodies in 3 and on IgG antibodies in 2
including C. pneumoniae [13]. White blood cells cases, respectively (Table I). The patients were from
(WBC) were counted, and erythrocyte sedimenta- 16 months to 10 y of age, only 1 was treated in
tion rate (ESR), serum procalcitonin (PCT) and hospital, and only 1 had lobar pneumonia. Two
serum C-reactive protein (CRP) were measured by cases were single and 3 mixed infections. At least 1
routine methods [13,14]. non-specific inflammatory marker was elevated in all
IgM and IgG antibodies to C. pneumoniae were cases. No seasonality was found.
studied in paired sera by MIF, and also by a The prevalence of S. negevensis IgM or IgG
commercial EIA test [13]. In the MIF, the diagnosis antibodies by the detection limits of the MIF test
of chlamydial infection was based on a ]4-fold rise was 20% in B2-y, 23% in 2−4-y, 16% in 5−9-y and
in IgG antibodies between paired sera or on IgM 30% in ]10-y-old children. IgG antibodies were
titre ]1:16 in any sample. In the EIA, ]1.5-fold positive in 20%, 13%, 4% and 0% (p =0.2),
EIU (enzyme immune unit) increases in IgG anti- respectively, suggesting a decreasing trend.
bodies in the zones >130 EIU or ]1.3-fold EIU There were no cases with concomitant S. nege-
increases in the zones B130 EIU were considered as vensis and C. pneumoniae infection by the present
diagnostic. The IgM EIA results were expressed as criteria (Table I). Microbe-specific IgM and IgG
signal/cut-off (S/CO) ratios, and the S/CO ratio antibodies, separately in acute and convalescent sera,
more than 1.1 was considered as positive [13]. were compared also by the detection limits of the
In 2005, S. negevensis IgM and IgG antibodies MIF and EIA tests (Table II). When the MIF test
were measured by MIF using elementary bodies of was used for both pathogens, there were no class-
S. negevensis (ATCC, VR-1471) [15]. The growth specific cross-reactions. When C. pneumoniae anti-
of S. negevensis strains and purification of S. bodies were measured by EIA, 5 children had IgG
negevensis antigen were performed as described antibodies to both pathogens in the acute (56% of S.
earlier [16], and specific antibodies were measured negevensis and 11% of C. pneumoniae positive
by an in-house MIF (National Public Health In- cases, p =0.3) and 3 children in the convalescent
Table I. Patients with CAP caused by Simkania negevensis.
Findings Case 1 Case 2 Case 3 Case 4 Case 5
Age (months) 36 16 57 120 52

Gender (F/M) F M M F M
Type of infiltration Interstitial Interstitial Interstitial Alveolar Interstitial
CRP (mg/l) 210 29.5 ND 116 ND
WBC (cells/mm3) 36,500 10,660 14,850 27,990 11,870
PCT (ng/ml) 12.62 3.4 0.34 17.98 0.21
ESR (mm/h) 120 70 43 82 57
In-/out-patient In Out Out Out Out
Coinfections Mycoplasma RSV None Mycoplasma None
Seasons October March May June November
Diagnostic criteria 8-fold increase 4-fold decrease 4-fold increase 8-fold decrease 4-fold decrease
IgM ABs IgG ABs IgG ABs IgM ABs IgM ABs
Abbreviations. ND: not done; RSV: respiratory syncytial virus; ABs: antibodies. The suggested cut-off limits for elevated PCT
concentration are 0.5 ng/ml, 1 ng/ml or 2 ng/ml [14].
Simkania negevensis in paediatric pneumonia 271
Table II. Antibodies to Simkania negevensis by MIF, compared with antibodies to Chlamydia pneumoniae by MIF and/or EIA.
Cpn IgM pos Cpn IgG pos
MIF (n =2) EIA (n =4) MIF (n =17) EIA (n =44)
Acute samples
Sn MIF IgM pos (n =8) 0p=0.8 1p=0.3 2 5
Sn MIF IgG pos (n =9) 0 0 0p=0.2 5p=0.3
Convalescent samples
Sn MIF IgM pos (n =10) 0p=0.8 0p=0.7 2 3
Sn MIF IgG pos (n =10) 0 0 0p=0.2 3p=0.3
Abbreviations. Cpn: Chlamydia pneumoniae; Sn: simkania negevensis; MIF: micro immunofluorescence.
Statistical significance: p-values were calculated for the differences between the presence and absence of class-specific antibodies to S.
negevensis and C. pneumoniae on admission or convalescence, respectively.
serum (30% of S. negevensis and 7% of C. pneu- S. negevensis infection, in accordance with a recent
moniae positive cases, p =0.3). study from Japan [18].
In a recent study from Israel, S. negevensis was
found by PCR in nasal samples in 91% of 34
Discussion
children with pneumonia, in 88% of drinking water
Our results suggest that S. negevensis may be a true, samples from their homes and in 75% of healthy
though rare, cause of CAP in children. All Simkania controls; these results suggest that S. negevensis is
diagnoses were based on significant increases or common in the environment and may be transmitted
decreases of antibodies in paired sera measured by via domestic water [3]. On the other hand, the
MIF, which is known to be a specific but non- method used was a nested PCR, which is prone to
sensitive method [7,9]. The prevalence of S. nege- contaminations, and this may explain the high
vensis-specific antibodies was 20−30%, with no incidence rates [19]. Viral infection was diagnosed
expected rise by age, and there was no substantial in half of the pneumonia cases, and no methods were
cross-reactivity between S. negevensis and C. pneu- available for other bacteria. Thus, the final role of S.
moniae antibodies. negevensis in pneumonia remained obscure, as also
S. negevensis is common in many areas. Seropo- in a recent PCR study from the USA [20]. In the
sitivity in healthy adults has been 55−80% in Israel present paper there were no healthy controls, the
[4], 46−65% in the UK and Denmark [5,6], but only diagnoses were based on a 4-fold or greater increase
4% in Japan [7]. Age-specific seroprevalence in or decrease in antibodies between paired sera, which
Israel was 13−29% in 1−2-y, 33−36% in 3−6-y, 42− is a conventional way to interpret serological tests as
57% in 7−10-y and 55−75% in 11−15-y-old non- positive in the aetiological diagnosis of infection, and
symptomatic children, respectively. [8] The present only a specific MIF test was used in S. negevensis
figures from north Italy were clearly lower (20− serology. These findings suggest that S. negevensis
30%), in accordance with recent Finnish data [9]. might cause about 5% of paediatric CAP.
Moreover, the seropositivity did not rise with age.
On the other hand, our results represent only an
estimate of seroprevalence, since the subjects were
not randomly collected from the population. In References
addition, the differences may reflect the poor sensi- [1] Kahane S, Gonen R, Sayada C, Elion J, Friedman MG.
tivity of the used MIF test [7,9] or climatic and other Description and partial characterization of a new Chlamy-
environmental factors. dia-like microorganism. FEMS Microbiol Lett 1993;109:
Chlamydial antibodies were seen in no case with 329−33.
[2] Everett KD, Bush RM, Andersen AA. Amended description
S. negevensis infection, in contrast to earlier suspi-
of the order Chlamydiales, proposal of Parachlamydiaceae
cions that Simkania and Chlamydia species, at least fam. nov. and Simkaniaceae fam. nov., each containing 1
when studied by EIA, may induce cross-reactions monotypic genus, revised taxonomy of the family Chlamy-
[9]. Likewise, there were no significant associations diaceae, including a new genus and 5 new species, and
between IgM or IgG antibodies to S. negevensis and standards for the identification of organisms. Int J Syst
Bacteriol 1999;49:415−40.
C. pneumoniae in MIF and/or EIA serology. Despite [3] Kahane S, Greenberg D, Newman N, Dvoskin B, Friedman
this, some children with IgG or IgM to C. pneumo- MG. Domestic water supplies as a possible source of
niae detected by EIA alone may have actually had infection with Simkania. J Infect 2007;54:75−81.
272 L. Fasoli et al.
[4] Friedman MG, Galil A, Geenberg S, Kahane S. Seropreva- [12] Lieberman D, Kahane S, Lieberman D, Friedman MG.
lence of IgG antibodies to the Chlamydia-like microorganism Pneumonia with serological evidence of acute infection with
‘Simkania Z’ by ELISA. Epidemiol Infect 1999;122:117−23. the Chlamydia-like microorganism ‘Z’. Am J Respir Crit
[5] Johnsen S, Birkebaek N, Andersen PL, Emil C, Jensen J, Care Med 1997;156:578−82.
Ostergaard L. Indirect immunofluorescence and real time [13] Don M, Fasoli L, Paldanius M, Vainionpää R, Kleemola M,
PCR for detection of Simkania negevensis infection in Räty R, et al. Aetiology of community acquired pneumonia:
Danish adults with persistent cough and in healthy controls. serological result of a paediatric survey. Scand J Infect Dis
Scand J Infect Dis 2005;37:251−5. 2005;37:806−12.
[6] Friedman M, Kahane S, Dvoskin B, Hartely J. Detection of [14] Don M, Valent F, Korppi M, Falleti E, De Candia A, Fasoli
Simkania negevensis by culture, PCR, and serology in L, et al. Efficacy of serum procalcitonin in evaluating severity
respiratory tract infection in Cornwall, UK. J Clin Pathol of community acquired pneumonia in childhood. Scand J
2006;59:331−3. Infect Dis 2007;39:129−37.
[7] Yamaguchi T, Yamazaki T, Inoue M, Mashida C, Kawagoe [15] Don M, Paldanius M, Fasoli L, Canciani M, Korppi M.
K, Ogawa M, et al. Prevalence of antibodies against Simkania negevensis and pneumonia in children. Pediatr
Simkania negevensis in a healthy Japanese population Infect Dis J 2006;25:470−1.
determined by the microimmunofluorescence test. FEMS [16] Kahane S, Kimmel N, Friedman MG. The growth cycle of
Immunol Med Microbiol 2005;43:21−7. Simkania negevensis. Microbiol 2002;148:735−42.
[8] Friedman MG, Dvoskin B, Kahane S. Infections with the [17] Korppi M, Paldanius M, Hyvarinen A, Nevalainen A,
Chlamydia-like microorganism Simkania negevensis, a pos- Husman T. Chlamydia pneumoniae and newly diagnosed
sible emerging pathogen. Microbes Infect 2003;5:1013−21. asthma: a case-control study in 1- to 6-y-old children.
[9] Korppi M, Paldanius M, Hyvärinen A, Nevalainen A. Respirology 2004;9:255−9.
Simkania negevensis and newly diagnosed asthma: a case- [18] Miyashita N, Obase Y, Fukuda M, Shoji H, Mouri K, Yagi S,
control study in 1- to 6-y-old children. Respirology 2006;11: et al. Evaluation of serological tests detecting Chlamydio-
80−3. phila pneumoniae-specific immunoglobulin M antibody.
[10] Greenberg D, Banerji A, Friedman MG, Chiu CH, Kahane Intern Med 2006;45:1127−31.
S. High rate of Simkania negevensis among Canadian Inuit [19] Apfalter P, Reischl U, Hammerschlag MR. In-house nucleic
infants hospitalized with lower respiratory tract infections. acid amplification assays in research: how much quality
Scand J Infect Dis 2003;35:506−8. control is needed before one can rely upon the results? J Clin
[11] Kahane S, Greenberg D, Friedman MG, Haikin H, Dagan Microbiol 2005;43:5835−41.
R. High prevalence of ‘Simkania Z’, a novel Chlamydia-like [20] Kumar S, Kohlhoff SA, Gelling M, Roblin PM, Kutlin A,
bacterium, in infants with acute bronchiolitis. J Infect Dis Kahane S, et al. Infection with Simkania negevensis in
1998;177:1425−9. Brooklyn, New York. Pediatr Infect Dis J 2005;24:989−92.
IV
Don M, Korppi M, Valent F, Vainionpaa R and Canciani M. Human metapneumovirus

pneumonia in children: results of an Italian study and mini-review. Scandinavian Journal of
Infectious Diseases 2008;40(10):821-826
Scandinavian Journal of Infectious Diseases, 2008; 40: 821−826
ORIGINAL ARTICLE
Human metapneumovirus pneumonia in children: Results of an

Italian study and mini-review
MASSIMILIANO DON1, MATTI KORPPI2, FRANCESCA VALENT3, RAIJA VAINIONPAA4

& MARIO CANCIANI1
From the 1Paediatric Department, School of Medicine, DPMSC, University of Udine, Italy, 2Paediatric Research Centre,
Tampere University and University Hospital, Tampere, Finland, 3Hygiene Department, School of Medicine, DPMSC,
University of Udine, Italy, and 4Department of Virology, University of Turku, Finland
Abstract
Human metapneumovirus (hMPV) is a newly identified paramyxovirus causing lower respiratory tract infections (LRTI).
Current knowledge on hMPV is mainly based on retrospective studies performed in stored respiratory and serum samples.
We found 15 previous prospective clinical studies on LRTI (11 clinical and 4 epidemiological studies) that have been
reviewed. Our aims were to analyse the role of hMPV in community acquired pneumonia (CAP) and the seroconversion rate
to hMPV in a prospective study in North Italian children. During a 15-month study period, 124 children were admitted due
to presumptive CAP and, in 116 of them, CAP was radiologically confirmed. The aetiology of CAP was assessed by serology
to 15 microorganisms, including enzyme immunoassay to hMPV. hMPV infection was found in 5 children (4.9%), being
single in 2 and mixed in 3 cases. The seroconversion rate to hMPV increased with age, reaching nearly 100% seropositivity
rate at school age. In conclusion, hMPV caused 0% to 17.5% of LRTI cases in children in the mini-review. The figure was
about 5% in the present and in the only earlier paediatric CAP study. Thus, hMPV is a real but rare cause of paediatric CAP,
although seroconversion to hMPV in most children takes place in early childhood.
Introduction particular focused on hMPV in children with com-

munity acquired pneumonia (CAP) [14].
In 2001, van den Hoogen et al. detected a new virus,
We have recently published our data on the
human metapneumovirus (hMPV), in children with
aetiology [15−17] and severity [18,19] of CAP in
respiratory tract disease [1]. Successively, it has been
North Italian children, treated both as inpatients and
documented that hMPV causes both acute upper [2]
outpatients. In the present paper, we further analyse
and lower respiratory tract infections (LRTI) [3]
the role of hMPV in CAP, including data on
that share many clinical characteristics with respira- seroprevalence of this new virus in the same study
tory syncytial virus (RSV) infections [4,5]. To date, population. In addition, the earlier studies on hMPV
hMPV has been detected in children with rhino- in paediatric LRTI are systematically reviewed.
pharyngitis, laryngitis, bronchitis, bronchiolitis,
pneumonia, acute otitis media and asthma exacer-
bation [2,4,6−12]. Material and methods
There is an emerging number of studies on hMPV
in LRTI in children. The viral diagnosis has mainly Study subjects
been carried out by reverse transcriptase polymerase Initially, 125 consecutive and previously healthy
chain reaction (PCR) [2,6−11,13]. Most studies children admitted to the Department of Paediatrics
have been hospital based [2,4,7,9−11,13] and/or of the University of Udine, Italy, with a clinical
only included young children aged B3 y [4,8,10, suspicion of CAP were prospectively evaluated
11]. To our knowledge, there is only 1 study in during a 15-month period, from 1 October to 31
Correspondence: M. Don, Paediatric Department, Allergology and Pulmonology Unit, School of Medicine DPMSC, University of Udine, P. le S. Maria della
Misericordia, 33100 Udine, Italy. Tel: ‡39 0432 559240 4. Fax: ‡39 0432 559258. E-mail: max.don@libero.it
(Received 17 January 2008; accepted 22 May 2008)

ISSN 0036-5548 print/ISSN 1651-1980 online # 2008 Informa UK Ltd. (Informa Healthcare, Taylor & Francis As)
DOI: 10.1080/00365540802227110
822 M. Don et al.
December 2002. The study protocol was approved peroxidase conjugated anti-human IgG antibodies
by the ethics committee of the University of Udine (at dilutions of 1:12,000 and 1:3000, Dako, Ger-
and an informed consent was obtained from the many) were added. Ortho-phenylene-diamine was
parents of all children. Briefly, patients showing used as a colour substrate and the optical density was
signs and/or symptoms compatible with respiratory measured at 492 um by a Multiskan Analyzer (Eflab
infection and radiological infiltrations consistent Oy, Helsinki, Finland). The last dilution which gave
with pneumonia were included in the clinical study; an OD value higher than 2.5 times the OD value of
newborns, children presenting with wheezing and negative controls was considered positive. The
those with chronic underlying illness or hospital detection limit, used in the assessment of seropre-
acquired pneumonia were excluded. The detailed valence, was 1:40 for IgG antibodies. The diagnostic
inclusion and exclusion criteria for enrolment into criterion of hMPV infection was a 4-fold or greater
the study as well as the radiological definition of increase or decrease in IgG antibodies between
pneumonia, have been published recently [15]. paired sera.
Three experienced radiologists evaluated the radio- Streptococcus pneumoniae, non-capsulated Hae-
graphs and interpreted the infiltrations as alveolar in mophilus influenzae and Moraxella catarrhalis in-
63 and interstitial in 38 cases [18]. fections were also studied by EIA in paired sera [15].
Mycoplasma pneumoniae was studied by EIA and
complement fixation, while Chlamydia pneumoniae
Study design
was studied by EIA and microimmunofluorescence
Among the 125 recruited children, at least 1 of the 2 (MIF) [15]. In 2005, serological studies were
paired serum samples was available in 124 cases and supplemented by MIF to Simkania negevensis in
these children formed the subjects of the seroepide- frozen samples [16,17].
miological study. 23 children were not included in
the clinical study, 15 because convalescent samples
were lacking and 8 because chest radiographs were Laboratory studies
not diagnostic for pneumonia. Thus, 101 children
formed the subjects of the clinical study. Serum Non-specific inflammatory markers were analysed in
samples for serological assays were obtained on blood and serum samples obtained on admission.
admission and on average 35 d later (range 18−56 Serum procalcitonin (PCT) was measured by im-
d) [15]. munoluminometric assay (PCT LIA TEST, Brahms
Diagnostics, Henningsdorf, Germany). In the pre-
sent paper we applied the values of 0.5 ng/ml, 1.0 ng/
Clinical definitions ml and 2.0 ng/ml as cut-off limits for elevated serum
On inspection, tachypnoea was defined by the WHO PCT concentration, as previously suggested [18].
criteria, that are respiratory rate >60 breaths/min in Blood erythrocyte sedimentation rates (ESR) and
children aged B2 months, >50 breaths/min in white blood cell counts (WBC) were measured by
children aged 2−12 months and >40 breaths/min routine methods [18].
in children aged >12 months [20]. On auscultation,
lung sounds were assessed as either normal or
Statistical analysis
silenced and the rales or crackles, whether fine,
medium or coarse, were combined together as Fisher’s exact test was used to assess the statistical
crackles. Children with wheezing were excluded significance of differences in IgG titres to hMPV and
and rhonci and ruttles were not registered nor in the distribution of detectable serum IgG to hMPV
included in the analyses [21]. between children of different ages.
Aetiological studies Results

As described earlier, enzyme immunoassay (EIA) The mean age of the 101 patients was 4.7 y (range
was used to study RSV, influenza A and B virus, 0.3−16 y); 49% of them were males, 73% were
parainfluenza 1, 2 and 3 virus, adenovirus and treated as outpatients and a potential causative agent
cytomegalovirus aetiology of infection in paired was detected in 67% of them [15]. Viral infection
sera [15]. IgG antibodies to hMPV were determined was diagnosed in 42 children, including 17 RSV
by an in-house EIA. The micro-titre plates were cases and 5 hMPV cases, and bacterial infection was
coated with hMPV-infected cell lysate. Serum dilu- diagnosed in 47 patients. Serological evidence of
tions (1:40−1:2560) were added to the plates and mixed infection was found in 30 cases, and 20 of
incubated for 2 h at‡378C, after which horseradish them were mixed viral-bacterial infections [15].
Human metapneumovirus in paediatric pneumonia 823
Table I. Characteristics of the 5 patients suffering from CAP with serological evidence of acute hMPV infection.
Cases No. 1 No. 2 No. 3 No. 4 No. 5
Gender/age (months) M/70 M/37 M/15 F/25 M/11

Admission IgG titre 1:640 B1:40 B1:40 1:160 1:40
Convalescent IgG titre 1:2560 1:640 1:160 1:10240 1:160
In/outpatient Out Out In Out Out
Coinfections C.pn., M.pn S.pn, Infl.B None S.pn, Infl.B None
Month of detection November January January February March
BT on admission (8C) 38 39.30 38.80 38 39.90
Symptoms on admission C C C C, FR C, FR
Respiratory signs on admission T, CR, SLS R T, CR, SLS T, CR, SLS T,CR
Type of Rx infiltration alveolar alveolar interstitial interstitial alveolar
WBC (cells/mm3) 6490 11900 16870 6350 7900
PCT (ng/ml) 0.31 5.74 28.67 0.80 1.81
ESR (mm/h) 50 58 75 20 90
Legend. F: female; M: male; In: inpatients; Out: outpatients; BT: body temperature; C: cough; FR: food refusal; R: rhinorrhoea; T:
tachypnoea; CR: crackles; SLS: silenced lung sound; I: interstitial; A: alveolar; RSV: respiratory syncitial virus; C.pn: Chl amydia
pneumoniae; M.pn: Mycoplasma pneumoniae; S.pn: Streptococcus pneumoniae; Infl.B: influenza B virus; ND: not done. The suggest ed
cut-off limits for elevated PCT concentration are 0.5 ng/ml, 1 ng/ml or 2 ng/ml [18].
There were no cases of mixed hMPV and RSV fever after the beginning of antibiotics was less then
infection. 24 h.
As shown in Table I, serological evidence of acute The distribution of the hMPV-specific antibody
hMPV infection, based on ]4-fold increase in IgG titres by age on admission are presented in Table II.
titres in paired sera, was found in 4.9% of CAP The titres were significantly associated with age.
patients. Their ages ranged from 11 to 70 months. High acute titres ]1:2560 were seen in 23 (19.3%)
Four of them were males, all but 1 were treated as children; the titre was ]1:5120 in 13 (10.9%) and
outpatients, none developed complications and all ]1:10240 in 9 (7.5%) cases. Despite the high titres
recovered completely. All cases occurred during on admission, a significant ]4-fold decrease indi-
winter, from November to March. Single infection cating an acute infection was not seen in any case.
was present in 2 cases, while S. pneumoniae, M. The prevalence of children positive for IgG
pneumoniae, C. pneumoniae and influenza B virus to hMPV, by the detection limit of the EIA test
were involved in the other 3 infections. On admis- (]1:40), increased constantly in relation to age
sion, all patients had fever and cough, 4/5 had (Table III). When data from acute and convalescent
tachypnoea on inspection, 4/5 had crackles and 3/5 samples were combined, 68% of the B2-y-old, 72%
silenced lung sound on auscultation. The infiltrates of the 2−4-y-old and 98% of the 5−9-y-old children
on chest radiographs were alveolar in 3 and inter- were positive for hMPV.
stitial in 2 patients. Serum non-specific inflamma-
tory markers varied widely, being highest in the only
Discussion
hospitalized patient. He had single hMPV infection,
presenting with interstitial infiltration on chest radio- The present paper is the first on hMPV in paediatric
graph. Four patients received oral amoxicillin and 1 CAP studied by serological methods. Our results
parenteral ampicillin. In all cases, the duration of confirm the involvement of hMPV in pneumonia in
Table II. Distribution of IgG titres to hMPV in acute sera by age in 119 children with data on admission available.
Age groups B2 y (n =25)b 2−4 y (n = 43) 5−9 y (n =44) ]10 y (n =12) p-valuea
No. of casesb 24 41 42 12
IgG ]1:40 14 (58)c 30 (73) 40 (95) 12 (100) B0.0001
IgG ]1:160 8 (33)d 22 (54) 36 (86) 12 (100) B0.0001
IgG ]1:640 2 (8) 15 (37) 28 (67) 8 (67) 0.0027
IgG ]1: 2560 2 (8) 6 (15) 13 (31) 2 (17) 0.8631
Data are presented as number (percent).

a
Fisher’s exact test.
b
Cases with available serum samples on admission.
c
3 children 512 months of age.
d
1 child 512 months of age.
824 M. Don et al.
Table III. Distribution of the subjects with detectable IgG to hMPV in paired sera by age, among the 124 subjects with data a vailable on
admission and/or on convalescence.
Age groups B2 y 2−4 y 5−9 y ]10 y p-valuea
No. of cases 25 43 44 12
IgG in acute sera 14/24 (58)b 30/41 (73) 40/42 (95) 12/12 (100) B0.0001
IgG in convalescent sera 14/22 (63)c 24/35 (69) 38/39 (97) 12/12 (100) B0.0001
Total 17/25 (68)d 31/43 (72) 43/44 (98) 12/12 (100) B0.0001
Data are presented as numbers of seropositive children/total numbers of children in each age group (percent). The detection limit of 1:40 of
the EIA test was used as the cut-off limit for seropositivity.
a
Fisher’s exact test.
b
c
d
children. The 4.9% incidence is similar to that found respiratory infections [4,9,11,26] and 1 of children
in 5 studies on paediatric LRTI by RT-PCR per- with CAP only [14]. RT-PCR was used in all studies
formed in nasopharyngeal samples [4,9,12−14]. and viral culture in addition to RT-PCR in 2 of them.
However, hMPV was often involved in mixed infec- Compared to these 11 studies, the present is based
tions with bacterial agents [22]. As observed also on the largest microbiological test panel. The
earlier [1,23−25], seroprevalence to this virus in- numbers of patients have varied from 63 to 641,
creased by age, reaching a 100% seropositivity by the and hMPV has caused 0% to 25% of all cases and
age of 10 y. 7.7% to 54% of cases with virological diagnosis
We found 11 earlier studies on hMPV respiratory (Table IV). In the CAP study by Lin et al., upper
infections in children, totally or in part focused on airway samples were studied by RT-PCR from 116
the lower respiratory tract [4,6−14,26]. Six studies Chinese children with pneumonia, and the incidence
consisted of children suffering from LRTI only [6− of hMPV infections was 5.2% [14]. In that study,
8,10,12,13], 4 of children with LRTI and other hMPV caused 13.3% of all viral CAP cases, all
Table IV. The role of human metapneumovirus in lower respiratory tract infections in children.
hMPV aetiology
No. of
Authors Type of Methods Clinical No. of aetiological (% of all (% of viral
[references] Y study for hMPV diagnosis patients agents studied cases) cases) Setting Country
Madhi et al. 2003 P RT-PCR LRTI 191 9V 7.1 17 I South

[13] Africa
Viazov et al. 2003 P RT-PCR ARI 63 2V 17.5 42 I Germany
[11]
Peiris et al. 2003 P RT-PCR ARI 587 8V 5.5 20 I Hong
[26] Kong
Boivin et al. 2003 P RT-PCR; ARI 208 4V 5.8 6.4 I Canada
[4] culture
Maggi et al. 2003 P RT-PCR LRTI 90 9V 25 54 I Italy
[10]
Williams et al. 2004 R RT-PCR LRTI 248 1V 20 − I/O U.S.A.
[6]
Mullins et al. 2004 P RT-PCR ARI 641 7V 3.9 11.8 I U.S.A.
[9]
Konig et al. 2004 P RT-PCR; LRTI 620 8V 0 − I/O Germany
[8] culture
Lin et al. [14] 2005 P RT-PCR CAP 116 8V3B 5.2 13.3 I Taiwan
Kim et al. [7] 2005 R RT-PCR LRTI 166 8V 15.7 44.8 I Korea
Choi et al. 2006 P RT-PCR LRTI 515 11 V 4.7 7.7 I/O Korea
[12]
Present study P Serology CAP 101 9V6B 4.9 11.6 I/O Italy
Legend. P: prospective; R: retrospective; LRTI: lower respiratory tract infection; ARI: acute respiratory infections; CAP: community
acquired pneumonia; RT-PCR: reverse transcription polymerase chain reaction; V: virus; B: bacteria; I: inpatients; O: outpatients.
Human metapneumovirus in paediatric pneumonia 825
Table V. Comparison of age-related seroprevalence rates to human metapneumovirus in children from different countries.
Age
Months Y
Authors [references] Y 0−12 12−24 2−5 5−9 ]10 Methods Number of patients Country
van den Hoogen et al. [1] 2001 25 55 70 100 100 IFA 120 The Netherlands
Wolf et al. [23] 2003 40a 52b − − − IFA 40 Israel
Ebihara et al. [24] 2003 18c 33 77 93 100 IFA 142 Japan
Leung et al. [25] 2005 55d 36 77 91 96 ELISA 216 U.S.A.
Present study 60 68 72 98 100 EIA 124 Italy
Data are presented as percents. IFA: indirect immunofluorescence assays; ELISA: enzyme-linked immunosorbent assay; EIA: enzyme
immunoassay.
a
80% for the 0−6 months group.
b
c
d
patients were hospitalized and the age ranged from 7 infection as the principal cause of CAP. On the other
to 11 y. Our 101 CAP patients were clearly younger hand, Madhi et al. found that hMPV with no evident
but the figures based on hMPV serology were rather bacterial complication seems also to induce severe
similar. In the study of Williams et al. [6], the inflammatory responses [13].
patients were B4 y of age and as many as 20% had The few studies performed on seroprevalence to
hMPV infection, but only 8% pneumonia. hMPV from the Netherlands [1], Japan [24], and
In the present study, hMPV infections mainly United States [25], have indicated that virtually all
occurred during the winter season as demonstrated children are infected by 5−10 y of age (Table V). The
by other y-round surveillance studies [6,10,12]. present figures from North Italy and the absence of
hMPV infections often occur after RSV outbreaks cases of hMPV infection in patients >6 y of age are
and can be separated from RSV cases by virological in accordance with the Dutch, Japanese [1,24] and
assays only [27]. Thus, hMPV cases have been American [25] results. In the study from Israel, the
reported as either virus-negative or simply as RSV- authors performed a longitudinal serological analysis
negative before the y 2001 [4,6,14,28]. In the by immunofluorescence assay in healthy children
present study, there were no coinfections with RSV, during the first 2 y of life [23] and found that the
although other mixed infections were common. On seropositivity rate decreased to a minimum by age 13
the other hand, hMPV infections were diagnosed by months and increased to 52% by age 24 months.
antibody assays alone, although a sensitive RT-PCR Since we did not study age cohorts, our results give
with no serological confirmation of the result has only an estimate of the real seroprevalence.
been used in other studies. Due to the lack of direct In conclusion, hMPV is a real but rare cause of
detection of viruses in respiratory samples, some paediatric CAP and seroconversion to hMPV usually
hMPV and also other viral cases were evidently takes place in early childhood.
missed [29−31].
Declaration of interest: The authors report no
The diagnosis of hMPV infection was based on
conflicts of interest. The authors alone are respon-
the demonstration of a ]4-fold rise in the IgG
sible for the content and writing of the paper.
antibodies [32], which proves viral involvement,
either as a cause of CAP or as a preceding infection
paving the way for bacterial pneumonia. On admis-
References
sion, the titre of hMPV-specific IgG was ]1:2560 in
about 20% of the patients, making a 4-fold increase [1] van den Hoogen BG, de Jong JC, Groen J, Kuiken T, de
Groot R, Fouchier RA, et al. A newly discovered human
in antibodies non-probable. Moreover, a 4-fold pneumovirus isolated from young children with respiratory
decrease was not seen in any of them. We assume tract disease. Nat Med 2001;7:719−24.
that the real frequency of hMPV infections was [2] Boivin G, Abed Y, Pelletier L, Ruel L, Moisan D, Cote S,
higher than that of 4.9% reported. S. pneumoniae et al. Virological features and clinical manifestations asso-
ciated with human metapneumovirus: a new paramyxovirus
was found in 2 patients, either PCT or ESR was
responsible for acute respiratory-tract infections in all age
elevated in 4 patients, and the recovery after starting groups. J Infect Dis 2002;186:1330−4.
antibiotics was appropriate in all cases, suggesting [3] Freymouth V, Vabret A, Legrand L, Eterradossi N, Lafay-
the presence of pneumococcal or other bacterial Delaire F, Brouard J, et al. Presence of the new human
826 M. Don et al.
metapneumovirus in French children with bronchiolitis. [18] Don M, Valent F, Korppi M, Falleti E, de Candia A, Fasoli
Pediatr Infect Dis J 2003;22:92−4. L, et al. Efficacy of serum procalcitonin in evaluating severity
[4] Boivin G, de Serres G, Cote S, Gilca R, Abed Y, Rochette L, of community acquired pneumonia in childhood. Scand J
et al. Human metapneumovirus infections in hospitalized Infect Dis 2007;39:129−37.
children. Emerg Infect Dis 2003;9:634−40. [19] Don M, Valerio G, Korppi M, Canciani M. Hyper- and
[5] Esper F, Boucher D, Weibel C, Martinello RA, Kahn JS. hypo-glycaemia in children with community acquired pneu-
Human metapneumovirus infection in the United States: monia. J Pediat Endocrinol Metab (In press.)
clinical manifestations associated with a newly emerging [20] Palafox M, Guiscafrè H, Reyes H, Munoz O, Martinez H.
respiratory infection in children. Pediatrics 2003;111:1407− Diagnostic value of tachypnoea in pneumonia defined
10. radiologically. Arch Dis Child 2000;82:41−5.
[6] Williams JV, Harris PA, Tollefson SJ, Halburnt-Rush LL, [21] Elphick HE, Ritson S, Rodgers H, Everard ML. When a
Pingsterhaus JM, Edwards KM, et al. Human metapneu- ‘wheeze’ is not a wheeze: acoustic analysis of breath sounds
movirus and lower respiratory tract disease in otherwise in infants. Eur Respir J 2000;16:593−7.
healthy infants and children. N Engl J Med 2004;350:443− [22] Madhi S, Ludewick H, Kuwanda L, Niekerk NV, Cutland
50. C, Little T, et al. Pneumococcal coinfection with human
[7] Kim YK, Lee HJ. Human metapneumovirus-associated
metapneumovirus. J Infect Dis 2006;193:1236−43.
lower respiratory tract infections in Korean infants and
[23] Wolf DG, Zakay-Rones Z, Fadeela A, Greenberg D, Dagan
young children. Pediatr Infect Dis J 2005;24:1111−2. R. High seroprevalence of human metapneumovirus among
[8] Konig B, Konig W, Arnold R, Werchau H, Ihorst G, Forster young children in Israel. J Infect Dis 2003;188:1865−7.
J. Prospective study of human metapneumovirus infection in [24] Ebihara T, Endo R, Kikuta H, Ishiguro N, Yoshioka M, Ma
children less than 3 y of age. J Clin Microbiol 2004;42:
X, et al. Seroprevalence of human metapneumovirus in
4632−5.
Japan. J Med Virol 2003;70:281−3.
[9] Mullins JA, Erdman DD, Weinberg GA, Edwards K, Hall
[25] Leung J, Esper F, Weibel C, Kahn JS. Seroepidemiology of
CB, Walker FJ, et al. Human metapneumovirus infection
human metapneumovirus (hMPV) on the basis of a novel
among children hospitalized with acute respiratory illness.
enzyme-linked immunosorbent assay utilizing hMPV fusion
Emerg Infect Dis 2004;10:700−5.
protein expressed in recombinant vesicular stomatitis virus. J
[10] Maggi F, Pifferi M, Vatteroni M, Fornai C, Tempestini E,
Clin Microbiol 2005;43:1213−9.
Anzilotti S, et al. Human metapneumovirus associated with
respiratory tract infections in a 3-y study of nasal swabs from [26] Peiris JS, Tang WH, Chan KH, Khong PL, Guan Y, Lau YL,
infants in Italy. J Clin Microbiol 2003;41:2987−91. et al. Children with respiratory disease associated with
[11] Viazov S, Ratjen F, Scheidhauer R, Fiedler M, Roggendorf metapneumovirus in Hong Kong. Emerg Infect Dis 2003;
M. High prevalence of human metapneumovirus infection 9:628−33.
in young children and genetic heterogeneity of the viral [27] Wolf DG, Greenberg D, Kalkstein D, Shemer-Avni Y,
isolates. J Clin Microbiol 2003;41:3043−5. Givon-Lavi N, Saleh N, et al. Comparison of human
[12] Choi EH, Lee HJ, Kim SJ, Eun BW, Kim NH, Lee JA, et al. metapneumovirus, respiratory syncytial virus and influenza
The association of newly identified respiratory viruses with A virus lower respiratory tract infections in hospitalized
lower respiratory tract infections in Korean children, 2000− young children. Pediatr Infect Dis J 2006;25:320−4.
2005. Clin Infect Dis 2006;43:585−92. [28] van den Hoogen BG, van Doornum GJ, Fockens JC,
[13] Madhi SA, Ludewick H, Abed Y, Klugman KP, Boivin G. Cornelissen JJ, Beyer WE, de Groot R, et al. Prevalence
Human metapneumovirus-associated lower respiratory tract and clinical symptoms of human metapneumovirus infection
infections among hospitalized human immunodeficiency in hospitalized patients. J Infect Dis 2003;188:1571−7.
virus type 1 (HIV-1)-infected and HIV-1-uninfected African [29] Jennings LC, Anderson TP, Werno AM, Beynon KA,
infants. Clin Infect Dis 2003;37:1705−10. Murdoch DR. Viral aetiology of acute respiratory tract
[14] Lin PY, Lin TY, Huang YC, Tsao KC, Huang YL. Human infections in children presenting to hospital: role of poly-
metapneumovirus and community acquired pneumonia in merase chain reaction and demonstration of multiple infec-
children. Chang Gung Med J 2005;28:683−8. tions. Pediatr Infect Dis J 2004;23:1003−7.
[15] Don M, Fasoli L, Paldanius M, Vainionpää R, Kleemola M, [30] Crowe JE. Human metapneumovirus as a major cause of
Räty R, et al. Aetiology of community acquired pneumonia: human respiratory tract disease. Pediatr Infect Dis J 2004;
serological results of a paediatric survey. Scand J Infect Dis 23:S215−21.
2005;37:806−12. [31] Sarasini A, Percivalle E, Rovida F, Campanini G, Genini E,
[16] Don M, Paldanius M, Fasoli L, Canciani M, Korppi M. Torsellini M, et al. Detection and pathogenicity of human
Simkania negevensis is a rare but real cause of pneumonia in metapneumovirus respiratory infection in paediatric Italian
children. Pediatr Infect Dis J 2006;25:470−1. patients during a winter-spring season. J Clin Virol 2006;35:
[17] Fasoli L, Paldanius M, Don M, Valent F, Vetrugno L, 59−68.
Korppi M, et al. Simkania negevensis in community [32] Hamelin M-E, Boivin G. Human metapneumovirus: a
acquired pneumonia in Italian children. Scand J Infect Dis ubiquitous and long-standing respiratory pathogen. Pediatr
2008;40:269−72. Infect Dis J 2005;24:S203−7.

978 951 44 7744 7 Dikonversi

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

978 951 44 7744 7 Dikonversi

Încărcat de

Drepturi de autor:

Formate disponibile

MASSIMILIANO DON

Pediatric Community-Acquired Pneumonia

A serologic study on etiology, with special focus

Distribution Tel. +358 3 3551 6055

Acta Universitatis Tamperensis 1423 Acta Electronica Universitatis Tamperensis 851

Tampereen Yliopistopaino Oy – Juvenes Print

LIST OF ORIGINAL PUBLICATIONS ............................................................................ 7

COMMUNITY-ACQUIRED PNEUMONIA IN THE PEDIATRIC AGE:

AIMS OF THIS STUDY................................................................................................... 51

PATIENTS AND METHODS........................................................................................... 52

ORIGINAL PUBLICATIONS ......................................................................................... 123

I. Don M, Fasoli L, Paldanius M, Vainionpää R, Kleemola M, Räty R, Leinonen M, Korppi

III. Fasoli L, Paldanius M, Don M, Valent F, Vetrugno L, Korppi M and Canciani M.

IV. Don M, Korppi M, Valent F, Vainionpää R and Canciani M.

V. Don M, Söderlund-Venermo M, Valent F, Lahtinen A, Hedman L, Canciani M, Hedman

CAP community-acquired pneumonia

BACKGROUND: Microbe-specific diagnosis of community-acquired pneumonia

(CAP) in childhood is difficult in clinical practice. Chest X-rays and non-specific

is an atypical bacterium recently found to be associated with respiratory infections in

studies performed in respiratory samples by polymerase chain reaction.

as its ability to discriminate the etiology of CAP, was studied.

presumptive pneumonia, and in 101 of them pneumonia was radiologically confirmed.

Twenty-seven children were treated as inpatients and 74 as outpatients. The pulmonary

infiltrate was considered to be alveolar in 63 and interstitial in 38 cases, according to the

serologic test panel to 16 microbes, including microimmunofluorescence to S. negevensis,

PCT was measured by an immunoluminometric assay in 100 CAP children.

RESULTS: A potential causing agent was detected in 75 patients (74%) . Evidence

22% of the CAP cases, respectively. Mycoplasma pneumoniae (27%), S. pneumoniae

CONCLUSIONS: Our results confirm the importance of S. pneumoniae in pediatric

children, while HBoV may be considered a fairly common cause. Sero-conversion to

any serum concentration, to differentiate between bacterial and viral etiology.

JOHDANTO: Lasten avosyntyisen keuhkokuumeen aiheuttajan mukainen diagnostiikka on

vaikeaa kliinisessä työssä. Keuhkokuvaa ja tulehduskokeita on käytetty erottamaan bakteerien

ja virusten aiheuttamia keuhkokuumeita, vaikka tutkimustulokset ovat olleet ristiriitaisia. Vasta-

ainetutkimukset kuuluvat rutiinitutkimuksiin tiettyjen virusten ja atyyppisten bakteereien

diagnostiikassa. Simkania negevensis kuuluu atyyppisiin bakteereihin, ja sen on havaittu

liittyneen lasten hengitystieinfektioihin ja aikuisten keuhkokuumeisiin. Viimeisten 25 aikana

vasta-aine- ja antigeenitutkimuksia on sovellettu tavallisiin hengitysteiden bakteereihin, kuten

pneumokokkiin (Streptococcus pneumoniae). Nykyinen käsitys kahden vastikään löydetyn

viruksen, human metapneumoviruksen (hMPV) ja human bocaviruksen (hBoV), jotka näyttävät

liittyvän etenkin lasten hengitystieinfektioihin, perustuu pääosin tutkimuksiin, joissa on

sovellettu geenimonistustestiä (PCR, polymerase chain reaction) hengitysteistä otettuihin

TUTKIMUKSEN TARKOITUS: Tämän prospektiivisen tutkimuksen tarkoitus oli selvittää,

käyttämällä vasta-ainetutkimuksia, lasten avosyntyisen keuhkokuumeen aiheuttajia sekä

polikliinisesti että sairaalassa hoidetuilla eri-ikäisillä lapsilla, ja erityinen huomio kiinnitettiin

mikrobille pyrittiin myös arvioimaan pohjois-italialaisilla lapsilla. Lisäksi arvioitiin seerumin

prokalsitoniinin (PCT) merkitystä sekä keuhkokuumeen vaikeuden arvioinnissa että bakteeri-

AINEISTO JA MENETELMÄT: Viidentoista kuukauden aikana 124 lasta lähetettiin

poliklinikalle todennäköisen keuhkokuumeen takia, ja heistä keuhkokuume varmistettiin

keuhkokuvalla 101 lapsella. Näistä 27 hoidettiin sairaalassa ja 74 kotona. Varjostuma

keuhkokuvassa oli alveolaarinen 63 ja interstitiaalinen 38 tapauksessa. Virus- ja

bakteerietiologia tutkittiin laajalla testipaketilla 16 mikrobia vastaan, mukaan lukien

immunologiset menetelmät hMPV- ja HboV -vasta-aineille. Seerumin PCT mitattiin

immunoluminometrisellä menetelmällä 100 lapsen seerumista.

TULOKSET: Todennäköinen aiheuttaja löytyi 75 (74%) potilaalta. Näyttöä virusinfektiosta,

bakteeri-infektiosta ja virus-bakteeri-sekainfektiosta saatiin 49 %:ssa, 46 %:ssa ja 22 %:ssa

keuhkokuumeista. Mycoplasma pneumoniae (27%), S. pneumoniae (18%) ja respiratory

keuhkokuumeet jakautuivat tasaisesti iän suhteen. Serologinen näyttö siitä, että S.

eikä mitään yhteyttä ikään havaittu. Sero-positiivisuus hMPV- ja HBoV-viruksia vastaan

merkittävästi neljässä etiologisessa ryhmässä (pneumokokin, atyyppisten bakteerien tai

interstitiaalisessa keuhkokuumeessa (mediaani 9.43 vs. 0.53 ng/mL, p<0.01).

JOHTOPÄÄTÖKSET: Tutkimuksen tulokset vahvistavat pneumokokin merkitystä

avosyntyisessä keuhkokuumeessa kaikenikäisillä lapsilla. Lisäksi mykoplasman sekä virusten