Review Ar ticle Journal of Paediatric Respirology and Critical Care

Review on paediatric necrotising pneumonia and its pulmonary

co-morbidities
Danial Wai-Tai KO
Department of Paediatrics, Queen Elizabeth Hospital, Hong Kong

Abstract
Necrotising pneumonia (NP) is defined as destruction of normal lung parenchyma with the presence
of areas of decreased contrast-enhancement +/- small air or fluid cavitations. Apart from the complication
of sepsis, it can also be associated with other pulmonary co-morbidities including parapneumonic
effusion/empyema, bronchopleural fistula and lung abscess. Most literature have reported
Streptococcus pneumoniae and community-acquired methicillin resistant Staphylococcus aureus as
the commonest pathogens. Mycoplasma pneumoniae, however, is becoming another important
causative agent, especially in the setting of increased prevalence of macrolide-resistant strain.
Contrasted computer tomography thorax remains the gold standard for the diagnosis of NP, but its
role may be substituted in the near future by thoracic ultrasonography with no concern of radiation
risk. Intravenous antibiotic is the main line of treatment and surgery has only ancillary role in managing
the complications. Though NP has significant morbidity in its acute stage, which usually leads to
prolonged hospitalisation, extended course of antibiotics or even necessity of surgical intervention,
its long-term prognosis is found to be excellent with low mortality, based on full clinical recovery and
minimal residual changes in follow-up imaging. More studies are however required to have better
assessment on the functional outcomes of children affected.

Keywords: Child, necrotising, pneumonia

Introduction thorax. Different studies had minor variations in their

working definitions but the main features are similar as
Nowadays, dramatic increase in incidence of paediatric follows.2-5
necrotising pneumonia (NP), after the introduction of - Areas of consolidation without loss of volume
7 polyvalent pneumococcal conjugate vaccine (PCV-7) - Destruction of normal parenchyma with the presence
followed by the selection of more virulent non-vaccine of areas of decreased contrast-enhancement
serotypes of Streptococcus pneumonia,1 was reported - Plus or minus presence of multiple small air or fluid
as well as the worldwide epidemic of community- cavities
acquired methicillin resistant Staphylococcus aureus - Excluding abscess, the features of which include
(C-MRSA). Though paediatric community-acquired prominent rim of contrast-enhancement in the
pneumonia has good prognosis in general, the same periphery
situation may not be true in severe NP. In this article
the relevant publications on paediatric necrotising Parapneumonic effusion and empyema are actually
pneumonia will be reviewed. within a continuum. Parapneumonic effusion (PPE) is
the presence of pleural effusion in association with the
underlying pneumonia.6 Empyema is usually defined
Definitions of necrotising pneumonia, when there is macroscopic appearance of purulent
parapneumonic effusion and empyema and pleural fluid, but in broader definition it also includes
pulmonary complications the condition in which bacteria is identified in the pleural
fluid.6
In nearly all studies, necrotising pneumonia was defined
mostly by findings in the computer tomography (CT) of Bronchopleural fistula (BPF) is defined as the presence
of persistent communication between the pleura and
Email: [email protected] the underlying necrotic lung tissue.7 Pneumatocele is a

20
 Volume 10 No. 4, December 2014 Review Ar ticle

thin-walled, air-filled cyst inside the lung parenchyma.8 their Paediatric Emergency Department. Six hundred
Finally lung abscess is defined as the well-delineated and thirty-five patients (13%) required hospitalisation
area of intrapulmonary fluid density or cavity with air- and 41 (0.8%) children was later diagnosed to have NP.
fluid level, with thick (>2 mm) enhancement wall.9 Furthermore the percentage of pneumonia complicated
with NP actually doubled from 4.5% to 9% when
comparing the two periods 2006-2009 and 2009-2011.
Epidemiology and incidence

Reports on necrotising pneumonia in children were Microbiology

scanty before 1990s with the first case report published
in 1994. 2 Since then the publications have been Streptococcus pneumoniae
accumulating but they were mainly case reports or As shown in the Table 1a, as well as the local
retrospective studies of small patient number. Tables experience, Streptococcus pneumonia is the most
1a and 1b have listed some important findings extracted important pathogen causing necrotising pneumonia.
from the relevant published articles on necrotising/ Furthermore the non-PCV-7 serotypes including
necrotizing pneumonia limited to paediatric population serotypes 3, followed by 19A, have become the
(birth-18 years). important serotypes causing the disease after the start
of PCV-7 vaccination.
It can be seen that there is no prospective study on
population basis to evaluate the true incidence of NP. In Utah's study, after the launch of the universal
Nevertheless, several retrospectives studies indicated immunisation of PCV-7, serotype 3 has become the
the rise in the incidence, especially after the launch of most important strain of Streptococcus pneumonia
PCV-7. Data from one Taiwan tertiary university hospital causing NP with OR 14.67 (95% CI 3.39-86.25). 1
showed that there was gradual growth in the annual Furthermore the serotypes 3 was also found to have
number of pneumococcal pneumonia from 1995 to the higher risk in developing empyema (OR 5.38, 95% CI
peak in 2000 followed by modest slip to to 2002. 10 1.11-51.14), requiring chest tube insertion (OR 5.0, 95%
However, the percentage of complicated pneumonia, CI 1.03-47.57), and undergoing surgical procedure (OR
defined as the presence of empyema or necrotising 3.59, 95% CI 0.90-13.28, p=0.025). A letter to the Editor
pneumonia, showed rapid rise from 25% in 1995 to 70% from Texas' researchers also reported 4 cases of NP all
in 2002. Another study in Children's Hospital Boston on caused by the more virulent serotype 19A. 12 When
their confirmed NP also saw a drastic increase in case compared with the 7 NP in the control group, all of which
number from three cases per year in the interval 1993- were retrieved from the past hospital record with no
1996 to 14 cases per year in 2003-2004.5 The strongest pathogens identified, the serotype 19A led to more
evidence demonstrating an increase in both the number complicated disease with longer hospital stay (19 days
of culture-positive pneumococcal pneumonia and the versus 13 days), prolonged duration of intravenous
percentage of paediatric NP actually came from the antibiotics (19.2 days versus 17 days), more admission
study in Utah when data in pre- and post-PCV-7 era to intensive care unit (75% versus 29%) as well as higher
were compared.1 The authors divided the study period intubation rate (75% versus 29%).
into two groups, 1997-2000 and 2001-2006, separated
by the demarcation line of the year 2000 when PCV-7 One Taiwan's study also compared the PCV-7 and non-
vaccine was first launched in the United States. Over PCV-7 strains on their role in all paediatric invasive
the total study period of around nine years, 124 cases pneumococcal disease (IPD) over 12 year at the time
of culture-positive paediatric pneumococcal pneumonia when the compliance of PCV-7 was still poor in
were retrieved. A significant escalation in the percentage Taiwan.13 Serotype 14, covered by PCV-7, remained the
of NP was detected from 5 of 39 cases (13%) in the most important strain identified in IPD (23.8%), followed
earlier time to 28 of 85 cases (33%) in the post-PCV-7 by serotypes 6B (19.0%) and 23F (16.2%). The serotype
period, with odd ratio (OR) 3.34 [95% confidence interval 3, not included in PCV-7, ranked the fourth causing IPD
(95% CI) 1.11-12.03]. (13.3%). Second commonest member in the non-PCV-
7 group associated with IPD was serotype 19A (3.8%).
A recent French study in one tertiary Paediatric Hospital When compared with PCV-7 group, the non-PCV-7
provided a better estimate on their local incidence of serotypes had more pulmonary manifestation (63.4%
necrotising pneumonia.11 Between May 2006 and April versus 38.7%), but also their pneumonia were mostly
2011, there were 4859 consultations of pneumonia in complicated with necrotising pneumonia or empyema

21
 Journal of Paediatric Respirology and Critical Care

Table 1a. Summary of published articles on necrotising pneumonia (NP) in children

Study Reference Year Type Number Age Premorbid Pathogens Duration
of fever
Kerem (Israel) 2 1994 Case report 4 1.3yr, 1.5yr, Normal immunity SP 9-20 days
2.9yr, 7.5yr except one IgG2 deficiency
McCarthy (USA) 40 1999 Case report 3 1.3yr, 2yr, 7yr Healthy SP ND
Wong (Taiwan) 3 2000 Restrospective 21 Mean 28.3mth Healthy 3 SP, 2 SA, 2 HIB, 2 MP, ND
review +/- 15.1(SD) 13 culture-negative
Congiz (Turkey) 26 2004 Case report 1 7 yr Healthy GAS ND
Wang ( Taiwan) 23 2004 Case report 5 3yr, 4yr, 6yr, Healthy MP 3 -20 days
6yr,14yr
Hsieh (Taiwan)** 10 2004 Retrospective 40 Mean 52.3mth 5 asthma, 4 heart disease, SP 13.5 days+/-
review (9-144mth) 2 genetic, 2 malignancy, 8.2 days(SD)
1 haemolytic anaemia, 1 metabolic
Hacimustafaoglu 4 2004 Prospective 36 Mean 3.8yr Healthy 1 GAS, 3 gram 8.9 days +/-
(Turkey) study (9mth-14yr) positive diplococci 4.3 days(SD)
in pleural fluid
Tseng (Taiwan) 20 2005 Case report 1 2yr Healthy C-MRSA 15 days
Chiu (Taiwan) 22 2006 Case report 1 7yr Healthy MP Around 30 days
Hsieh (Taiwan)## 31 2006 Retrospective 15 Mean 49mth Healthy SP Median 9 days
review (9-85mth) (4-30 days)@
Sawicki (USA) 5 2008 Retrospective 80 Median 3.6yr 53 healthy, 18 SP( 22%), Median 6 days
review (IQR 2.4-6.2yr) 2 immunocompromised also (IQR 3-9 days)
MSSA/MRSA,
42 culture-negative
(52%)
Obando (Spain) 27 2009 Case report 1 12yr Healthy C-MRSA + H1N1 flu A 23 days
Kalaskar (USA) 12 2009 Case report 4 2.8-4.1yr 1 asthma, 4 SP serotype ND
3 healthy 19A
Pascual 30 2010 Case report 1 15yr Healthy but Mycoplasma ND
(Switzerland) sexually active hominis
Yazer (Canada) 28 2011 Case report 1 5yr Mild asthma SP+ H1N1 ND
Moreira Silva 29 2012 Case report 1 28mth Healthy Acinetobacter ND
(Portugal) Iwoffii
Wang (China) 24 2012 Case report 5 3yr, 4.5yr, 5yr, ND MP Up to 20 days
5yr, 9yr
Lematre (France) 11 2013 Retrospective 41 Median 14mth Not 13 C-MRSA, Median 7 days
review (1mth-16yr) immunocompromised 7 SP, (1-25 days)
1 Fusobacterium
nucleatum
Abbreviation: SP, Streptococcus pneumoniae; SA, Staphylococcus aureus; MSSA, methicillin-sensitive SA; MRSA, methicillin-resistant SA; C-MRSA, community
acquired MRSA; MP, Mycoplasma pneumoniae; GAS, Group A Streptococcus; HIB, Haemophilus influenzae type B; BPF, bronchopleural fistula; PPE,
parapneumonic effusion; AB, antibiotic; CD, chest drain; VATS, video-assisted thoracoscopic surgery; PFT, pulmonary function test; ND not documented;
FU, follow-up; IQR, interquartile range; SD, standard deviation; yr, year; mth, month.
** Study on complicated pneumonias (with empyema or NP) out of 71 pneumococcal pneumonias.
## Study on NP out of 56 pneumococcal pneumonias. @ Days of fever from hospital admission.

22
 Volume 10 No. 4, December 2014 Review Ar ticle

Table 1b. Summary of published articles on necrotising pneumonia (NP) in children

Study Hospitalisation Complications Duration of Other Mortalithy Outcome FU time
day relevant AB treatments
Kerem (Israel) 12-26 days PPE, empyema 6 weeks 2 CD Nil Full recovery except 1 Up to
CXR with pleural thickening 12mth
McCarthy (USA) 10-15 days PPE, abscess, BPF 20-22 days 2 decortication, Nil CT done at 6 weeks in ND
1 abscess debridement+ 1 patient: small pneumatocele
decortication+closure of BPF and residual pleural thickening
Wong (Taiwan) ND Empyema, loculations, 14-35 days 9 VATS, Nil 19 normal CXR, 6mth
pneumothorax, 2 segmentectomy, 2 with small
BPF, lung abscess 2 abscess debridement residual pneumatocele
Congiz (Turkey) 8 days Empyema 8 days CD Death Death on 8th day Nil
Wang (Taiwan) ND PPE (Massive), 3 weeks in CD Nil 3 full recovery, Some up to
multiple abscess 1 patient 2 residual lung damage 8mth
(1 pneumatocele, 1 partial
atelectasis of left lingular)
Hsieh (Taiwan)** 25.2 days+/- 1/40 NP alone, ND 24/40 3/40 ND except ND
12.0 days (SD) 12/40 NP+empyema thoracotomy (7.5%) 7.5% mortality
Hacimustafaoglu 26 days+/- Empyema/PPE (94%), ND 66% requiring lung 6% 5.5% Mortality, 2mth
(Turkey) 9 days (SD) BPF (55%) excision, fistula repair, others fully recovered
or decortication
Tseng (Taiwan) 8 days Empyema 8 days CD Death Death on day 8th Nil
Chiu (Taiwan) 22 days PPE (massive) 10 days VATS, decortication, CD Nil Full recovery ND
Hsieh Median 18 days Empyema (93%) 16-38 days 93% CD, 1 death FU CXR in 10 patients: 1-6mth
(Taiwan)## (5-40 days) 53% VATS 2 total resolution,
7 minimal fibrosis,
1 cicatrizing atelectasis
Sawicki (USA) Median 12 days PPE, BPF Mmedian 27days CD, VATS, open Nil Full clinical recovery; Median 174
(IQR 9-17 days) (range 3-95) thoracotomy/ 12 with FU PFT: days (IQR77-
decortication, 4 mildly deranged 360 days)
partial lung resection
Obando (Spain) 28 days Empyema, 3 weeks VATS, CD Nil Full recovery except CXR: ND
pneumothorax mutliple small pneumatoceles
and residual pleural thickening
Kalaskar (USA) 15-28 days ND 23-35 days VATS Nil Full recovery ND
Pascual 19 days PPE, 2 weeks CD, Nil Full recovery ND
(Switzerland) pericardial effusion pericardiectomy
Yazer (Canada) ND Empyema, BPF ND CD Nil Minimal respiratory symptoms ND
Moreira Silva (Portugal) ND PPE(mild) 25 days Nil Nil Full recovery 2yr
Wang (China) 17-35 days PPE Up to 4 weeks CD Nil ND but one with persistent Up to
consolidation in CXR 180 days
up to 180 days after FU
Lematre Median 16 days Empyema, Median 42 days 3 CD+ later Nil No deaths ND
(France) (7-43 days) pneumothorax, (31-60 days) VATS, 3 VATS
hydropneumothorax, at outset
pneumatocele
Abbreviation: SP, Streptococcus pneumoniae; SA, Staphylococcus aureus; MSSA, methicillin-sensitive SA; MRSA, methicillin-resistant SA; C-MRSA, community
acquired MRSA; MP, Mycoplasma pneumoniae; GAS, Group A Streptococcus; HIB, Haemophilus influenzae type B; BPF, bronchopleural fistula; PPE, parapneumonic
effusion; AB, antibiotic; CD, chest drain; ; CXR, chest X-ray; VATS, video-assisted thoracoscopic surgery; PFT, pulmonary function test; ND, not documented;
FU, follow-up; IQR, interquartile range; SD, standard deviation; yr, year; mth, month.
** Study on complicated pneumonias (with empyema or NP) out of 71 pneumococcal pneumonias.
## Study on NP out of 56 pneumococcal pneumonias. @ Days of fever from hospital admission.

23
 Journal of Paediatric Respirology and Critical Care

(78.9% compared with 55.2%). Importantly, serotype 3 two out of eight culture-positive NP were caused by
was identified to carry the highest risk of developing Staphylococcus aureus, one being methicillin-sensitive
complicated pneumonia (necrotising pneumonia or whereas another methicillin-resistant.3 Tseng et al also
empyema) with OR 8.8 (95% CI: 1.024-75.59). The reported one fatal case of C-MRSA with severe sepsis
serotype 19A also caused more complicated pneumonia and necrotising pneumonia.20 With respect to the local
than uncomplicated one (12.9% versus 0%), though not incidence, the only article published in 1993 stated that
reaching the statistical significance (p=0.282). According Staphylococcus aureus ranked the third of the
to the authors, the emergence of serotype 19A was commonest bacteria in childhood pneumonia, following
rather related to selection of more resistant-strains after Haemophilus influenzae and Streptococcus
extensive antibiotic usage. It should be reminded that pneumoniae.21 No recent articles in the literature have
the results in this study should be interpreted cautiously reported the current local situation of paediatric
as none of the IPD patients were reported to have Staphylococcus aureus pneumonia, nor are there any
received prior PCV-7. case reports of paediatric NP caused by C-MRSA in
Hong Kong. However from the author's experience, few
One local study also analysed the change in serotypes necrotising pneumonia are caused by C-MRSA.
involved in IPD for all ages (from less than 5 to greater
than 65 years) after initiation of PCV-7 immunisation.14 Mycoplasma pneumoniae
Serotype 14 declined from 36% to 15.7% (p<0.01) Mycoplasma pneumoniae (MP), on the other hand, has
whereas 19A rose exponentially from 0 to 12.9% become an important cause of NP around this region.
(p<0.01). For those younger than 5, the proportion of In Wong's study, two cases of necrotising pneumonia
PCV-7 covered serotypes decreased from 89.5% to were caused by MP.3 Subsequently two further case
65.7%, but those serotypes included in PCV-13 reports from Taiwan 22,23 and one from China 24 also
remained similar (from 91.4-93.2%). Furthermore 19A described NP related to this bacteria, though the total
was associated with penicillin/erythromycin resistance. number of cases remained small. The three case reports
Another similar study in China analysed all hospitalised however revealed the clinical features quite different
children <5 years with pneumococcal pneumonia from from the usual self-limiting atypical pneumonia. First, it
February 2006 to February 2008. 15 The most predominantly affected the younger age, down to the
predominant serotype was 19F (55.6%), followed by 19A age of three.23,24 Second, prolonged fever occurred up
(13.9%). The 19A serotype was also found to have the to 20 days even after starting erythromycin and there
highest drug resistance. was long hospital stay ranged from 17-35 days.24 Third,
though no mortality was reported so far, the massive
To summarise all the studies, the immunisation of parapneumonic pleural effusion actually required chest
Prevnar-7 has actually selected the two more virulent drain insertion 23 and video-assisted thoracoscopic
non-vaccine serotypes 3 and 19A, which has supported surgery (VATS) with decortication.22 Furthermore, there
the concept of serotype replacement.12 The serotype 3 was a recent article reporting an increase in the
also results in more necrotising pneumonia or empyema, prevalence of macrolide-resistant MP in Hong Kong.25
as well as the necessity of surgical treatment. On the
other hand, the emergence of serotype 19A has posed Miscellaneous pathogens
a threat of multiple antibiotic resistance. Fortunately, Other pathogens reported in the literature also included
most of the common serotypes reported to cause NP group A Streptococcus,26 co-infection with human swine
nowadays are still protected by the newer 13-valent influenza (H1N1), 27,28 as well as those rare bacteria
pneumococcal conjugated vaccine. including Acinetobacter species 29 and Mycoplasma
hominis.30
Staphylococcus aureus
The second commonest bacteria causing necrotising
pneumonia in the literature is Staphylococcus aureus, Pathogenesis of necrotising pneumonia
especially community acquired methicillin resistant
Staphylococcus aureus (C-MRSA), which carries the It is still debatable about the pathogenesis of necrotising
gene for Panton-Valentine leukocidin (PVL).16-20 From pneumonia in paediatric population. Panton-Valentine
the French single center study on 41 children with NP, leukocidin-positive Staphylococcus aureus, has been
13 cases were caused by C-MRSA followed by showed to induce extensive necrotic ulcerations of the
7 children with Streptococcus pneumoniae. 11 tracheal and bronchial mucosa, as well as massive
From Taiwan's study in 2000 including 21 cases, haemorrhagic necrosis of inter-alveolar septa in one

24
 Volume 10 No. 4, December 2014 Review Ar ticle

post-mortem study.16 However it is still unclear how presence of immature PMN and elevated CRP, to be
Streptococcus pneumonia causes tissue destruction and risk factors. The identification of the above three
it has been postulated to be related to one unknown conditions actually supports the notion that the host
virulent factor secreted.31 Furthermore, the same authors inflammatory towards the pathogens, as exhibited by
performed the autopsy in their only mortality case and left-shifted neutrophilia and elevated CRP, is probably
identified pulmonary gangrene in the right middle lobe the key element for the tissue injury. Furthermore, as
and intravascular thrombi in the pulmonary artery shown in Table 1a, most children with NP were healthy
draining to the infarcted region. They postulated that without impaired immunity. It further supports the
pulmonary gangrene might be infrequently associated postulation that NP is developed out of the exaggerated
with pneumococcal NP. The gangrene subsequently cytokine-immune response from the host toward the
evolved to liquefaction of pulmonary parenchyma and pathogens, leading to parenchymal destruction. 4
then cavitations. Nonetheless, two rather contradictory findings from that
retrospective study was that neither the serotypes of
In another article, lung resection was performed in 12 Streptococcus pneumonia, nor the penicillin resistant
patients with necrotising pneumonia complicated with strains were found to have significant difference between
bronchopleural fistula. 7 Coagulation necrosis was the complicated pneumonia and lobar pneumonia. 10
detected in 11/12 cases. Eight of those twelve patients Actually the predominant serotype isolated in that study
were also detected with co-existent suppurative necrosis. was serotype 14 and none of the cases was caused by
Furthermore vasculitis and thrombosis formation were serotype 3.
identified histologically in 33.3% (4/12) and 66.7%
(8/12) respectively. These findings further support the With respect to PVL-positive Staphylococcus aureus
theory of macroscopic pulmonary gangrene and necrotising pneumonia, the French researchers
microscopic ischemic necrosis. investigated PVL-positive Staphylococcus aureus
necrotising pneumonia in 50 patients with median age
14.5 years (range 1 month - 78 year). The risk factors
Risk factor for the development of necrotising identified in deceased group when compared with the
pneumonia/complicated pneumonia survived group included airway haemorrhage and
leukopenia.18 Multivariate analysis further calculated that
Only one study investigated the potential risk factors the adjusted relative hazard associated with death for
leading to the complicated pneumococcal pneumonia, total WBC count >1000-3000 /ml and 0-1000 were 7.99
which was defined to be culture-positive and associated (95% CI 1.66-38.43) and 7.38 (95% CI 1.60-34.02)
with necrotising pneumonia or empyema10 Out of 71 respectively. Six year later from the same French
confirmed pneumococcal pneumonia in their 8-year institution, another paper was published and it included
retrospective review, 40 developed parenchymal larger population of 148 patients with median age of 22
necrosis or empyema. Univariate analysis identified years (interquartile range 3.0-43.7 years). 19 Severe
5 characteristics associated with complicated leukopenia (</=3000/ml) increased the mortality with
pneumonia as tabulated (Table 2). adjusted hazard ratio 4.5 (95% CI: 2.38-8.51). The
presence of prior influenza illness increased the risk of
However further multivariate analysis only isolated the severe leukopenia with adjusted odd ratio 4.45 (95%
latter three conditions, namely no-underlying diseases, CI 1.67-11.88).

Table 2. Risk factors associated with complicated pneumonia

Univariate analysis Multivariate analysis
Risk factors OR (95% CI) P OR (95% CI) P
Age >36 months 2.81 (1.05-7.56) 0.038
Thrombocytopaenia 8.71 (1.04-73) 0.035
No underlying disease 7.79 (1.96-30.99) 0.01 5.48 (1.06-28.25) 0.04
Presence of immature polymorphs in peripheral blood 5.14 (1.79-14.74) 0.002 3.67 (1.08-12.63) 0.037
C-reactive protein >12 mg/dL 5.83 (1.64-20.70) 0.04 5.24 (1.10-24.93) 0.037
OR: odd ratio; CI: confidence interval.

25
 Journal of Paediatric Respirology and Critical Care

Tool of diagnosis Growing potential of thoracic ultrasonography

Ultrasonography (USG) of thorax has gained a more
Role of CT for the diagnosis of NP important role nowadays in the management of
In most studies, necrotising pneumonias were pneumonia,34,35 especially in the management of co-
diagnosed by CT. Some papers also used chest X-ray existing complicated pleural effusion/empyema. In
(CXR) for diagnosis in the presence of multiple cavities British Thoracic Society (BTS) guidelines on the
on the background of consolidation.4 In a Switzerland's management of pleural infection in children, 6 thorax
study on 9 children, when CXR findings were compared u l t r a s o n o g r a p h y i s i n d i c a t e d t o d i ff e r e n t i a t e
with those of contrasted CT, there was a delayed consolidation/parapneumonic effusion (PPE), simple
detection of cavitary necrosis in three children by 5-9 versus complicated PPE, and to guide the algorithm
days. 32 In evaluation of detection of the associated into different management approach: antibiotic alone,
complications, CXR also missed 2 cases of chest tube drainage plus intrapleural fibrinolytic, or even
parapneumonic effusion in 5 children confirmed by CT. surgical intervention. Its role on the diagnosis of NP
Three cases of BPF were only identified in that series has not been well established currently. Only one case
and all was diagnosed by CT solely. report from Taiwan showed that the presence of
peripheral hypoechoic spaces (PHES) was specific
Though CT thorax should be the investigational tool sonographic features for childhood NP with the
of choice in diagnosing necrotising pneumonia, its specificity and positive predictive values both of
routine use for all suspected cases is not 100%.36 However the low sensitivity of 35% of PHES
recommended in view of its considerable radiation could not exclude NP in its absence.
dose especially in children who still have higher
number of active dividing cells than adult.33 There is Nevertheless, there was an adult study (aged 33-79
no study on the role of CT thorax guiding on the years) proving the usefulness of thoracic ultrasound in
management of isolated necrotising pneumonia. differentiating between empyema and lung abscess for
However in treating empyema, the United Kingdom peripheral pulmonary air-fluid lesions. 37 The study
randomised controlled trial on conservative approach showed that the ultrasound findings of complex-septated
(percutaneous chest drain insertion plus intrapleural effusions and passive atelectasis both were specific
urokinase) versus video-assisted thoracoscopic (specificity 100%) for empyema, but with low sensitivity
surgery (VATS) on paediatric empyema suggested of 40% and 47% respectively. However when the
that additional findings obtained from CT thorax did additional Colour Doppler technique was applied, the
not alter the management plan nor provide any identification of vessel signals in pericavitary
prognostic value.9 In that study, out of 25 patients with consolidation was found excellent in diagnosing
CXR abnormality showing simple opacification in lung peripheral lung abscess with sensitivity 94%, specificity
parenchyma only, CT was more sensitive in detecting 100%, positive predictive value 100% and negative
one additional cavitation, one more pneumatocele and predictive value 94% respectively. Thus with the vast
four extra cavity-necrosis, but no lung abscess were advancement in the field of thoracic ultrasonography, it
detected by CXR and CT. Those additional CT may be in near future widely utilised as the primary tool
findings, according to the authors' comments, could for investigating the suspected NP and detecting other
still be managed conservatively. Thus they concluded pulmonary complications.
that there was no additional role for CT when those
children with empyema were already managed with Important role of CT thorax in equivocal/complicated
combination of chest tube drainage with intrapleural cases and pre-operative assessment
urokinase. It can always be argued that the results Despite the above discussion, CT thorax still has its
may be biased as the subtle findings of CXR might unique role for more difficult cases.38 USG is not easily
only be picked up by the very experienced paediatric performed in uncooperative children. Presence of
radiologists in that tertiary children hospital, the Great pneumothorax will also hinder the detailed lung
Ormond Street Hospital for Children. However as the imaging beneath the air layer. Moreover, the
Table 1b has pointed out, necrotising pneumonia pathology which is situated in areas not easily
mostly has an excellent prognosis with full clinical accessible by USG (e.g. adjacent to scapula) cannot
recovery, even the confirmation of NP by CT-findings be identified. Finally, some surgeons still prefer to
will usually not alter the conservative approach for request CT scan as a road-map before they proceed
management on those patients. to VATS/decortication.

26
 Volume 10 No. 4, December 2014 Review Ar ticle

Pulmonary complications/co-morbidities one percent (10/47) on pleural drainage developed

BPF. Those with BPF, when compared with no-BPF
Most patients diagnosed to have necrotising pneumonia group, had longer duration of pleural drainage (median
usually suffer from other complications, as showed in 14 versus 6 days, p=0.0007), and prolonged
Table 1b. In addition to those secondary to severe sepsis hospitalisation (median 19 versus 13 days, p=0.001).
(septic shock, disseminated intravascular coagulopathy, Nevertheless all BPF could be treated successfully
acute renal failure/fluid-electrolyte disturbance, acute by chest drainage without surgery.
respiratory distress syndrome), there are important
pulmonary complications that needed to be managed Pneumatoceles
simultaneously, including parapneumonic empyema, Pneumatoceles are thin-walled, air-filled cysts inside the
bronchopleural fistula, pneumatoceles and lung lung parenchyma, resulted from alveolar and bronchiolar
abscess. necrosis, which allow unilateral air entry into the
interstitial tissue through the check-valve effect.8 It is
Parapneumonic effusion and empyema well known to be associated with Staphylococcus aureus
PPE and empyema are within a continuum. Necrotising pneumonia. With respect to necrotising pneumonia, a
pneumonia is nearly universally associated with PPE/ study investigated the 394 children admitted to one
empyema. In the earliest paediatric report including 4 Brazilian university hospital with the diagnosis of
patients, two patients had PPE while empyema was pneumonia complicated with empyema and/or
found in the remaining two children both with positive pneumatocele.39 8.3% (33/394 cases) were found to
pleural culture of Streptococcus pneumoniae . 2 have pneumatoceles with diameter ranging from 5 mm
In Taiwan's study in 2000, from the 11 out of 21 NP to 90 mm. The authors did not describe which pathogens
successfully treated conservatively with antibiotics, nine were more commonly associated with the
patients had chest-drain inserted also for PPE and two pneumatoceles, though in that study the most
of the 9 patients had complicated PPE. 3 One-third commonest three bacteria in all culture-positive pleural
(7/21) of the patients needed surgical intervention for fluid were Streptococcus pneumoniae (18%),
the empyema. From Boston's study including 80 cases Staphylococcus aureus (10%) and Gram negative
of paediatric NP, 86% were found to have appreciable bacteria (7%). Twenty-eight children had their
pleural effusion.5 Nineteen cases (24%) had bacteria pneumatoceles resolved spontaneously but four cases
cultured from pleural fluid and 5 cases was positive for including one tension pneumatocele and three ruptured
pneumococcal antigen in pleural fluid. In a French study cysts required drainage.
on paediatric NP, the percentage of empyema in NP
was 63% (26/41 cases).11 Pneumatoceles can be detected as early complication
or late sequel. In the literature review as shown in Table
Bronchopleural fistula 1, pneumatoceles as an acute complication was only
The presence of bronchopleural fistula actually implies reported in one article.11 Out of 41 cases in the French
the extension of tissue necrosis from the lung study, pneumatoceles were present in 4 (9.7%) patients.
parenchyma to the adjacent pleura. This complication Staphylococcus aureus was the predominant pathogen
should be suspected when the pneumothorax exists (13/21 cases) in all positive culture in NP followed by
in the background features suggestive of necrotising Streptococcus pneumoniae (7/21 cases). On the other
pneumonia. The definitive diagnosis depends on hand, pneumatoceles as small and residual findings in
contrast CT thorax.32 From the Turkey study, 55% of the follow-up chest X-rays were relatively common as
patients with NP had BPF.4 Twenty percent resolved mentioned in 4 reports.3,23,27,40
on conservative management (with small fistula)
while 80% required surgical intervention. In another Lung abscess
study, out of 112 children with culture-confirmed In most studies investigating necrotising pneumonia,
pneumococcal pneumonia, 18 of them (16%) were abscess was excluded and was treated as a separate
detected to have BPF.7 All of them also had empyema disease entity. However, in many patients suffered from
and necrotising pneumonia concurrently. The "complicated" pneumonia, both conditions co-
prognosis in that study was not satisfactory as 15 existed simultaneously. 3,23,40 Further from one case
patients with BPF required surgery, including lung report, the authors suggested that multiple small
resection in 12 cases. On the contrary, the data from lucent lesions coalesced to lung abscess over 3-4
the Boston study showed a better outcome.5 Twenty- days.40

27
 Journal of Paediatric Respirology and Critical Care

Management of necrotising pneumonia and Consequently, the role of surgery in the management
pulmonary co-morbidities of necrotising pneumonia is mainly to treat the
associated pulmonary complications. The main
The management of NP including treating the indications for the surgical intervention include
pneumonia itself, together with the associated decortication for local empyema causing atelectasis,4
pulmonary complications. The management can be the surgical excision of diseased lung with persistent
divided into medical versus surgical interventions. bronchopleural fistula,7 or excision of large
pneumatocele which imposes heavy pressure effect on
Antibiotic treatment the adjacent healthy lung tissues, 46 imaging-guided
The main treatment for NP remains to be medical aspiration or insertion of drainage catheter for lung
therapy with intravenous antibiotics. No randomised abscess in non-responding patients treated with
control trial is available to guide the evidence-based intravenous antibiotics.42
recommendation on this disease management. Thus
the duration of antibiotic treatment is still controversial. Adjunctive treatment for empyema thoracis
Both British Thoracic Society (BTS) and Infectious There has been no change in the approach of the
Diseases Society of America (IDSA) did not state clearly management of empyema since the publication of the
how long the antibiotics are required for NP in their latest BTS guidelines in 2006.6 Conservative management
guidelines on management of community-acquired through chest drain insertion plus intrapleural fibrinolytic
pneumonia in children. From BTS statements, it only should be considered first, and the VATS will only be
mentions that prolonged course of intravenous antibiotic proceeded in those failed cases. The APSA also adopts
may be required until the fever settles.41 IDSA does not the same approach and recommends that the chemical
discuss the antibiotic strategy for NP.42 In contrast, the debridement is the first line treatment and surgical
recommendation on the antibiotic duration for empyema debridement will be reserved for those failed cases as
was written explicitly. BTS suggests that intravenous the last resort.43
antibiotics are continued until the child is afebrile or at
least until the chest drain is removed. 6 Then oral
antibiotics are continued on discharge for 1-4 more Prognosis
weeks, but longer if there is residual disease. IDSA gives
similar recommendation of 2-4 weeks duration of The prognosis of the necrotising pneumonia is usually
antibiotics for empyema, or at least 10 more days after excellent with very low mortality and minimal morbidity.
resolution of fever.42 As the NP and empyema usually
co-exist, the usual practice in my institution is to give a In the literature review, the mortality of NP was found to
similar duration of antibiotics to treat both conditions as be very low. Apart from those isolated case reports, only
follows: two relative large studies including 36-40 children
A total of 4 weeks, or 2 weeks after the patient is reported the mortality from 5.5% to 7.5%.4,10 The causes
afebrile and has improved clinically. of death were not discussed in details in both reports.
In another article focusing on invasive pneumococcal
Role of surgery on NP disease, the mortality was 6.7% and all the death had
It has already been a generally accepted approach that concurrent bacteremia.13
surgical intervention for NP per se should be avoided
as NP usually has an excellent outcome in children. With respect to the morbidity, all recovered clinically up
This is also in accordance with the recommendations to the follow-up period of 12 months in the first paediatric
from IDSA42 and American Pediatric Surgical Association case report.2 Out of the four patients, one had follow-up
(APSA).43 The suggestion is based on the evidence that CXR done at 2 months, which still detected
drainage of necrotic lung tissue actually led to the pneumatoceles but the lesions resolved in 6 months after
development of bronchopleural fistula.44 Nevertheless, discharge. CXR at 2 months in another patient only
in some uncommon circumstances when the sepsis detected pleural thickening with no parenchymal
cannot be controlled after appropriate antibiotic, excision abnormalities. One patient had pulmonary function test
of the necrotic tissue, in the form of segmentectomy, (PFT) showing decreased lung volume during the acute
lobectomy or even bilateral lobectomy, may be illness, but subsequent PFT at 2 months' time had
performed as the last resort.45 normalised.

28
 Volume 10 No. 4, December 2014 Review Ar ticle

Based on the Boston's study, all 80 children had full its prognosis is suggested to be excellent with low
clinical recovery, as well as normal CXR.5 Only a few mortality. The most common pathogen, in this locality,
had follow-up CT thorax, which also revealed complete is Streptococcus pneumoniae with potential serotypes
resolution. In Taiwan's study, follow-up CXR films within already covered by PCV 13 currently. Mycoplasma
1-3 months showed complete resolution or only minimal pneumoniae is another important bacteria, especially
fibrotic changes in all nine patients with the imaging in the setting of emergence of macrolide-resistant
performed.31 strains. Ultrasonography of thorax has gained an
important role in the management of empyema and
It is still arguable that clinical and radiological resolution necrotising pneumonia and will probably replace CT as
are not equivalent to absence of functional pulmonary the primary investigation. Intravenous antibiotic
deficit. Furthermore, CXR is not sensitive enough to treatment is still the first line treatment for NP and
detect residual structural lesions. In Switzerland study surgery is reserved for managing its complications. More
on radiologic imaging, CT still detected pneumatoceles studies are required to have better understanding on
in 25% (2/8) of the patient.32 In another study including the functional outcomes on this disease.
five patients suffering from necrotising pneumonitis
caused by Mycoplasma pneumoniae, two patients had
residual lung damage detected by follow-up CT thorax References
6-8 months after the illness, one with pneumatocele
while another with persistent atelectasis.23 All the above 1. Bender JM, Ampofo K, Korgenski K, Daly J, Pavia AT, Mason
studies have already suggested that CXR may not be a EO, et al. Pneumococcal necrotizing pneumonia in Utah:
appropriate tool for detecting permanent lung damage. does serotype matter? Clinical Infectious Diseases 2008;46
However too liberal use of CT thorax as evaluation is (9):1346-52.
2. Kerem E, Bar Ziv Y, Rudenski B, Katz S, Kleid D, Branski D.
also discouraged in view of the profound radiation risk.
Bacteremic necrotizing pneumococcal pneumonia in children.
Am J Respir Crit Care Med 1994;149(1):242-4.
Pulmonary function test is definitely a better 3. Wong KS, Chiu CH, Yeow KM, Huang YC, Liu HP, Lin TY.
measurement in assessing the functional status but it Necrotising pneumonitis in children. Eur J Pediatr 2000;159
can only be performed for those older and cooperative (9):684-8.
children. Out of 80 patients studied in Bostons study, 4. Hacimustafaoglu M, Celebi S, Sarimehmet H, Gurpinar A,
only 12 patients had PFT done in their follow-up.5 Normal Ercan I. Necrotizing pneumonia in children. Acta Paediatrica
PFT was found in 8 patients, mild obstructive pattern in 2004;93(9):1172-7.
3 patients and mild restrictive pattern in 1 patient. DLCO 5. Sawicki GS, Lu FL, Valim C, Cleveland RH, Colin AA.
study was not performed in that study. Necrotising pneumonia is an increasingly detected
complication of pneumonia in children. Eur Respir J 2008;
31(6):1285-91.
To summarise, necrotising pneumonia can impose a
6. Balfour-Lynn IM, Abrahamson E, Cohen G, Hartley J, King
significant morbidity on paediatric population in terms
S, Parikh D; Paediatric Pleural Diseases Subcommittee of
of acute complications. The current evidence, however, the BTS Standards of Care Committee. BTS guidelines for
suggests an excellent outcome with full clinical recovery, the management of pleural infection in children. Thorax 2005;
normal follow-up CXR as well as normal to mildly 60 Suppl 1:i1-21.
deranged pulmonary function tests. However, without 7. Hsieh YC, Wang CW, Lai SH, Lai JY, Wong KS, Huang YC,
the longer-term follow up, the better imaging technique et al. Necrotizing pneumococcal pneumonia with
such as CT with low radiation dose protocol, as well as bronchopleural fistula among children in Taiwan. Pediatr
better modalities of pulmonary function tests for the Infect Dis J 2011;30:740-4.
appropriate age group (including preschoolers), it is still 8. Quigley MJ, Fraser RS. Pulmonary pneumatocele: pathology
too earlier to predict the long-term pulmonary capacity and pathogenesis. Am J Roentgenol 1988;150(6):1275-7.
9. Jaffe A, Calder AD, Owens CM, Stanojevic S, Sonnappa S.
of those children recovered from NP.
Role of routine computed tomography in paediatric pleural
empyema. Thorax 2008;63(10):897-902.
10. Hsieh YC, Hsueh PR, Lu CY, Lee PI, Lee CY, Huang LM.
Conclusions Clinical manifestations and molecular epidemiology of
necrotizing pneumonia and empyema caused by
Necrotising pneumonia can have significant morbidity Streptococcus pneumoniae in children in Taiwan. Clin Infect
during its acute stage, which may lead to prolonged Dis 2004;38(6):830-5.
hospitalisation, extended course of antibiotics or even 11. Lematre C, Angoulvant F, Gabor F, Makhoul J, Bonacorsi
necessity of surgical intervention. However in long run, S, Naudin J, et al. Necrotizing pneumonia in children: report

29
 Journal of Paediatric Respirology and Critical Care

of 41 cases between 2006 and 2011 in a French tertiary care Med J 2011;17(5):407-9. .
center. Pediatr Infect Dis J 2013;32(10):1146-9. 26. Cengiz AB, Kanra G, Caglar M, Kara A, Gucer S, Ince T.
12. Kalaskar AS, Heresi GP, Wanger A, Murphy JR, Wootton Fatal necrotizing pneumonia caused by group A
SH. Severe necrotizing pneumonia in children, Houston, streptococcus. J Paediatr Child Health 2004;40(1-2):69-71.
Texas, USA. Emerg Infect Dis 2009;15:1696-8. 27. Obando I, Valderrabanos ES, Millan JA, Neth OW.
13. Shen CF, Wang SM, Lee KH, Ho TS, Liu CC. Childhood Necrotising pneumonia due to influenza A (H1N1) and
invasive pneumococcal disease caused by non-7-valent community-acquired methicillin-resistant Staphylococcus
pneumococcal vaccine (PCV7) serotypes under partial aureus clone USA300: successful management of the first
immunization in Taiwan. J Formos Med Assoc 2013;112(9): documented paediatric case. Arch Dis Child 2010;95(4):
561-8. 305-6.
14. Ho PL, Chiu SS, Ang I, LauYL. Serotypes and antimicrobial 28. Yazer J, Giacomantonio M, Macdonald N, Lopushinsky S.
susceptibilities of invasive Streptococcus pneumoniae before Severe necrotizing pneumonia in a child with pandemic
and after introduction of 7-valent pneumococcal conjugate (H1N1) influenza. CMA 2011;183(2):215-9.
vaccine, Hong Kong, 1995-2009. Vaccine 2011;29(17):3270-5. 29. Moreira Silva G, Morais L, Marques L, Senra V. Acinetobacter
15. Yao KH, Wang LB, Zhao GM, Zheng YJ, Deng L, Huang JF, community-acquired pneumonia in a healthy child. Rev Port
et al. Pneumococcal serotype distribution and antimicrobial Pneumol 2012;18(2):96-8.
resistance in Chinese children hospitalized for pneumonia. 30. Pascual A, Perez MH, Jaton K, Hafen G, Di Bernardo S,
Vaccine 2011;29(12):2296-301. Cotting J, et al. Mycoplasma hominis necrotizing
16. Gillet Y, Issartel B, Vanhems P, Fournet JC, Lina G, Bes M, pleuropneumonia in a previously healthy adolescent. BMC
et al. Association between Staphylococcus aureus strains Infect Dis 2010;10:335.
carrying gene for Panton-Valentine leukocidin and highly 31. Hsieh YC, Hsiao CH, Tsao PN, Wang JY, Hsueh PR, Chiang
lethal necrotising pneumonia in young immunocompetent BL, et al. Necrotizing pneumococcal pneumonia in children:
patients. Lancet 2002;359(9308):753-9. the role of pulmonary gangrene. Pediatr Pulmonol 2006;41
17. Gonzalez BE, Hulten KG, Dishop MK, Lamberth LB, (7):623-9.
Hammerman WA, Mason EO Jr, et al. Pulmonary 32. Hodina M, Hanquinet S, Cotting J, Schnyder P, Gudinchet
manifestations in children with invasive community-acquired F. Imaging of cavitary necrosis in complicated childhood
Staphylococcus aureus infection. Clin Infect Dis 2005;41(5): pneumonia. Eur Radiol 2002;12(2):391-6.
583-90. 33. Tiddens H, van Straten M, Ciet P. Computed tomography.
18. Gillet Y, Vanhems P, Lina G, Bes M, Vandenesch F, Floret In: Eber E, Midulla F, eds. ERS Handbook on Paediatric
D, et al. Factors predicting mortality in necrotizing community- Respiratory Medicine. First Edition: HERMES; 2013:166-75.
acquired pneumonia caused by Staphylococcus aureus 34. Darge K, Chen A. Ultrasonography of the lungs and pleurae
containing Panton-Valentine leukocidin. Clin Infect Dis 2007; for the diagnosis of pneumonia in children: prime time for
45(3):315-21. routine use. JAMA Pediatr 2013;167(2):187-8.
19. Khanafer N, Sicot N, Vanhems P, Dumitrescu O, Meyssonier 35. Shah VP, Tunik MG, Tsung JW. Prospective evaluation of
V, Tristan A, et al. Severe leukopenia in Staphylococcus point-of-care ultrasonography for the diagnosis of pneumonia
aureus-necrotizing, community-acquired pneumonia: risk in children and young adults. JAMA Pediatr 2013;167:119-
factors and impact on survival. BMC Infect Dis 2013;13:359. 25.
20. Tseng MH, Wei BH, Lin WJ, Lu JJ, Lee SY, Wang SR, et al. 36. Chiu CY, Wong KS, Lai SH, Huang YH, Tsai MH, Lin YC.
Fatal sepsis and necrotizing pneumonia in a child due to Peripheral hypoechoic spaces in consolidated lung: a specific
community-acquired methicillin-resistant Staphylococcus diagnostic sonographic finding for necrotizing pneumonia in
aureus: case report and literature review. Scand J Infect Dis children. Turk J Pediatr 2008;50(1):58-62.
2005;37(6-7):504-7. 37. Chen HJ, Yu YH, Tu CY, Chen CH, Hsia TC, Tsai KD, et al.
21. Sung RY, Cheng AF, Chan RC, Tam JS, Oppenheimer SJ. Ultrasound in peripheral pulmonary air-fluid lesions. Color
Epidemiology and etiology of pneumonia in children in Hong Doppler imaging as an aid in differentiating empyema and
Kong (Abstract). Clin Infect Dis 1993;17(5):894-6. abscess. Chest 2009;135(6):1426-32.
22. Chiu CY, Chiang LM, Chen TP. Mycoplasma pneumoniae 38. Calder A, Owens CM. Imaging of parapneumonic pleural
infection complicated by necrotizing pneumonitis with effusions and empyema in children. Pediatr Radiol 2009;39
massive pleural effusion. Eur J Pediatr 2006;165:275-7. (6):527-37.
23. Wang RS, Wang SY, Hsieh KS, Chiou YH, Huang IF, Cheng 39. Kunyoshi V, Cataneo DC, Cataneo AJ. Complicated
MF, et al. Necrotizing pneumonitis caused by Mycoplasma pneumonias with empyema and/or pneumatocele in children.
pneumoniae in pediatric patients: report of five cases and Pediatr Surg Int 2006;22(2):186-90.
review of literature. Pediatr Infect Dis J 2004;23(6):564-7. 40. McCarthy VP, Patamasucon P, Gaines T, Lucas MA.
24. Wang Y, Xu D, Li S, Chen Z. Mycoplasma pneumoniae- Necrotizing pneumococcal pneumonia in childhood. Pediatr
associated necrotizing pneumonitis in children. Pediatr Int Pulmonol 1999;28(3):217-21.
2012;54(2):293-7. 41. Harris M, Clark J, Coote N, Fletcher P, Harnden A, McKean
25. Lung DC, Chan YH, Kwong L, Que TL. Severe community- M, et al; British Thoracic Society Standards of Care
acquired pneumonia caused by macrolide-resistant Committee. British Thoracic Society guidelines for the
Mycoplasma pneumoniae in a 6-year-old boy. Hong Kong management of community acquired pneumonia in children:

30
 Volume 10 No. 4, December 2014 Review Ar ticle

update 2011. Thorax 2011;66 Suppl 2:ii1-23. empyema in children: a comprehensive review from the APSA
42. Bradley JS, Byington CL, Shah SS, Alverson B, Carter ER, Outcomes and Clinical Trials Committee. J Pediatr Surg 2012;
Harrison C, et al. Pediatric Infectious Diseases Society and 47(11):2101-10.
the Infectious Diseases Society of America. The management 44. Hoffer FA, Bloom DA, Colin AA, Fishman SJ. Lung abscess
of community-acquired pneumonia in infants and children versus necrotizing pneumonia: implications for interventional
older than 3 months of age: clinical practice guidelines by therapy. Pediatr Radiol 1999;29(2):87-91.
the Pediatric Infectious Diseases Society and the Infectious 45. Netto JC, Tavares E, Andrade Ede O, Silva Mdos S, et al.
Diseases Society of America. Clin Infect Dis 2011;53:e25- Surgical treatment of children with necrotizing pneumonia.
76. J Bras Pneumol 2010;36(6):716-23.
43. Islam S, Calkins CM, Goldin AB, Chen C, Downard CD, 46. Al-Saleh S, Grasemann H, Cox P. Necrotizing pneumonia
Huang EY, et al; APSA Outcomes and Clinical Trials complicated by early and late pneumatoceles. Can Respir J
Committee, 2011-2012. The diagnosis and management of 2008;15(3):129-32.

31