Human Pathology (2007) 38, 1239 1247

www.elsevier.com/locate/humpath

Original contribution

Lung involvement in childhood measles:

severe immune dysfunction revealed by
quantitative immunohistochemistryB
Talib Moyses Moussallem MD, PhDa,1, Fernanda Guedes BSca,2,
Elaine Raniero Fernandes BSca,2, Carla Pagliari PhDa,2,
Carmen Lucia Penteado Lancellotti MD, PhDb,3,
Heitor Franco de Andrade Jr MD, PhDa,*,1, Maria Irma Seixas Duarte MD, PhDa,1
a

Laboratory of Pathology of Transmissible Diseases, Depto Pathology, Faculdade de Medicina da Universidade de Sao Paulo,
CEP 01246-903 Brazil
b
Instituto de Infectologia Emlio Ribas, Sao Paulo, CEP 01246-900 Brazil
Received 14 July 2006; revised 8 January 2007; accepted 8 January 2007

Keywords:
Measles;
BALT;
Cytokines;
Lymphocytes;
Immune system;
Pneumonia

Summary Measles, accounting for nearly 1 million deaths each year, presents intense involvement of
lymphoid organs and the lungs. The immune response in situ in the lungs was determined in blocks
recovered from 42 necropsies of children who died from measles determined by immune cell phenotype
(CD4, CD8, CD20, CD45RO, CD68, natural killer [NK], and antigen S-100 B [S100]) and cytokine
production (interferon, tumor necrosis factor, interleukin [IL]-1, IL-2, IL-4, IL-10, and IL-12).
Compared with the lungs of age-paired controls, patients with measles presented severe depletion of
CD4+, CD20+, CD68+, NK+, and S100+ cells in alveolus- and bronchus-associated lymphoid tissue
without depletion of CD8+ cells. Most of these features were similar in both forms of measles lung
involvement, Hecht giant cell, or interstitial pneumonia, but S100+ cells were depleted in bronchusassociated lymphoid tissue from patients with Hecht pneumonia, which also occurs more frequently in
malnourished children. IL-10 and IL-12producing cells were depleted in patients with measles,
whereas IL-1, interferon-, and IL-4producing cells were more frequently seen in the alveolus of
patients with measles compared with controls. Quantitative in situ immune cell phenotype and function
in the lung in measles demonstrated severe immune dysfunction, with loss of key cells, such as
dendritic, CD4+, and NK+ cells, and deficient cytokine production, which allows for a better
comprehension of local reactions in this process.
D 2007 Elsevier Inc. All rights reserved.

This work was supported by Conselho Nacional de Pesquisas and Instituto dos Laboratorios de Investigacao Medica do Hospital das Clnicas da
Faculdade de Medicina da USP.
* Corresponding author.
E-mail address: hfandrad@usp.br (H. F. de Andrade Jr).
1
The following authors take the responsibility for the integrity of the work.
2
The following authors are responsible for morphometry and immunohistochemistry.
3
The following author collaborated with lung histopathology analysis.
0046-8177/$ see front matter D 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.humpath.2007.01.015

1240

1. Introduction
Measles, or rubeola, is the oldest virus-induced disease
known to cause serious damage to immune function [1].
Despite the efficacy of its vaccine, measles continues to
cause approximately 770,000 deaths around the world each
year, especially in third world countries and with isolated
outbreaks in developed nations [2]. Measles is accompanied
by high morbidity and mortality, mainly in malnourished
children [1] because of lethal complications such as
secondary infections, especially pneumonia and upper
respiratory tract involvement, due to either direct measles
virus (MV) effect or its indirect disabling effects on the
immune system [3].
The immune system is affected by MV in both innate and
adaptive immune responses. MV attacks natural killer (NK)
cells leading to the down-regulation of their lytic activity
[4]. Heat shock proteins are also subverted by MV to induce
viral product transcription, which, in turn, induces more heat
shock proteins to be synthesized, causing a positive
feedback loop where the virus multiplies and raises its
numbers inside the cell [5].
Infection of dendritic cells leads to fusion of these cells
with macrophages and with each other, forming giant cells
that are unable to exert antigen-presenting function and
constitute an ideal place for virus multiplication [6].
Infection of dendritic cells also causes apoptosis, which
diminishes their numbers in the lung tissue, affecting the
transition from the innate phase to the adaptive phase of the
immune response [7].
MV damages the adaptive immune responses, but
interestingly, the virus infection induces a life-long protective immunity, although the acute disease causes severe
impairment of immune responses to other microorganisms
[4]. During this acute phase, Th1 type of immune response
occurs with a drastic antiviral effect, clearing the virus, as
seen in the skin by cytotoxic CD8+ cell infiltration [8].
Interferon c (IFNc) is present in the acute phase of measles
in the prerash period, followed by interleukin (IL)-2,
characteristic Th1 cytokines. During and after the exanthema, the immune system turns its response to the Th2 profile.
IL-4 begins to predominate and acts as a Th2 immunity
trigger [7]. Specific IgG subtype production changes from
IgG1 to IgG4, as well as high IgE plasma levels [9].
Measles, or rubeola, presents 2 main lung involvements.
The classic giant cell pneumonia, or Hecht pneumonia, is
more severe and frequently associated with respiratory
failure and death [1]. Unfortunately, patients often die with
histological lung involvement that is indistinct from
interstitial pneumonia, except for detection by additional
diagnostic immunohistochemistry or electron microscopy
for viral detection. The pathogenic mechanisms of lung
involvement are not well understood, but they are attributed
to both direct MV action and indirect effect causing
generalized immune disruption [4]. The lungs are directly
exposed to the external environment, being protected by

T. M. Moussallem et al.
cilia, mucus, alveolar macrophages, and a powerful bronchus-associated lymphoid tissue (BALT) [10]. This lymphoid lung tissue has been reported to mature and change
with age, but few, if any, studies exist on pulmonary
infections [10,11].
In the present study, we investigate the in situ immune
response caused by MV in the lungs from children who died
due to measles or rubeola, grouped according to main
histological lung involvement. We quantify the numbers of
immune cells and their activation evidenced by cytokine
production, by quantitative immunohistochemistry in areas
of lung structures such as the alveolus, septum, and BALT,
and by the response to viral infection.

2. Methods
2.1. Cases and controls
The necropsies from patients diagnosed as having measles
or rubeola were recovered from the files of Instituto Emilio
Ribas in Sao Paulo (Brazil) and Hospital Infantil Nossa
Senhora da Gloria in Vitoria (Brazil). We excluded cases with
concomitant diagnosis of bacterial pneumonia and patients
more than 20 years of age. We recovered the paraffin blocks
and rechecked lung histology, excluding cases that had
evidence of tissue autolysis, inappropriate fixation with
artifacts, or typical bacterial infection in lung histology. This
selection resulted in adequate lung tissue from 42 patients
diagnosed as having isolated measles, with well-defined
clinical and histological lung involvement. Their mean age
was 29 months, ranging from 4 to 228 months. There was no
bias caused by sex (23/42 females) or ethnicity. For
noninfected controls, we recovered paraffin blocks from the
lungs of age-paired children with congenital heart disease
without significant lung involvement who died during
surgery from the files of the Heart Institute (InCor),
University of Sao Paulo Medical School, Brazil, and whose
data were also described elsewhere [11]. All controls had no
lung infection at necropsy and Heath-Edwards first degree or
no pulmonary artery hypertrophy, with the least intense
pulmonary involvement as possible. A total of 18 controls
were selected, 11 females (61%) and 7 males (39%), with a
mean age of 12 months (range, 5-31 months). This careful
approach for selecting blocks with adequate controls was
performed to minimize variation to block conservation,
storage, fixation, and autolysis, but one can only hope that
these precautions would suffice for subsequent analysis,
which is always presented as individual data.

2.2. Immunohistochemistry and

electron microscopy
The paraffin-embedded lungs were resectioned, and
5-lm sections adhered to sylane-treated slides. Histology
was analyzed in hematoxylin-eosin slides by 2 independent
observers for diagnostic histopathology. Sections were also

Lung involvement in childhood measles

Table 1

1241

Demographic and histological findings in patients with measles and controls

Event

Age, mean F SD, mo

Sex (% males)
Alveoli septum enlargement and edema (%)
Presence of pneumocyte I necrosis (%)
Giant cell presence at each 20 field (%)
Presence of bronchial squamous metaplasia (%)

Noninfected controls
(n = 18)

Measles (n = 40)
Total

Interstitial
pneumonitis
(n = 27)

Hecht
pneumonia
(n = 13)

12.7 F 8.04
38.9
0
0
NA
0

31.5 F 40.5
42.5
80
NA
NA
50

33.5 F 48.7
35.7
70
18
0
43

27.5 F 14.7
50
93
100
100
78.5

Statistical
significance

NS
NS
PI and H N C
H N PI and C
H N PI
H N PI N C

Comparison between groups was tested using the Kruskal-Wallis test with Dunn post test for quantitative data or v 2 test for frequencies.
Abbreviations: C, controls; H, Hecht pneumonia; PI, interstitial pneumonia; NA, not applicable.

submitted to immunohistochemistry, with monoclonal antibodies (mabs) revealed via the streptavidin-biotin peroxidase method using an endogenous biotin blocking system
(DAKO, Carpinteria, CA, USA) [12] with modifications
described elsewhere [11]. The following mabs for cellular
phenotypes and cytokines were used in this process: CD4
(M834/DAKO), CD8 (M7103/DAKO), CD20 (M755/
DAKO), CD68 (M786/DAKO), CD45RO (M742/DAKO),
NK (M1014/DAKO), S100 (Z311/DAKO), IL-1b (LP-712/
Genzyme; Cambridge, MA, USA), IL-2r (M731/DAKO),
IL-4 (AB204/R&D Systems; Minneapolis, MN, USA), IL10 (MAB 217/ R&D Systems), IL-12 (MAB 219/R&D
Systems), tumor necrosis factor (TNF; IP300/Genzyme),
and IFNc (IP 500/Genzyme). Viral antigens were detected
using a direct monoclonal antibody against N protein, a
cytoplasmic antigen from MV (MAB8906/Chemicon), in
the same protocol of streptavidin-biotin peroxidase revealing system. Ultrastructural analysis was performed in
araldite-embedded glutaraldehyde fixed lungs recovered
from paraffin blocks counterstained with osmium tetroxide
as described elsewhere [13]. Ultrathin sections were
analyzed in ZEISS electronic microscopy, targeting intranuclear nucleocapsid viral inclusions.

2.3. Lung histology and immunohistochemistry

The hematoxylin & eosin-stained sections were analyzed
by 2 independent observers in blind sections using
semiquantitative scores: 0absence, 1mild, 2moderate, and 3intense. We scored bronchial epithelial hyperplasia, squamous metaplasia, necrosis, erosion, ulcer, and
infiltrate. In the alveoli wall and lumen, we scored necrosis,
regeneration of pneumocytes, giant cells, edema, hemorrhage, and hyaline membrane. We also scored the septum
thickness, edema, capillary congestion, and fibrosis. We
defined the MV-associated pneumonias as Hecht pneumonia, when alveoli damage with intense presence of giant
cells and type I pneumocyte necrosis predominated despite
the concomitant interstitial involvement; and as interstitial
pneumonia, when these giant cells and pneumocyte necrosis

were scarce with main septal involvement and little

disturbances in alveoli epithelium.

2.4. Quantitative in situ immune response analysis

The quantitative analysis of immunohistochemistry stained
cells was performed using an ocular square grid, which
marked an area of 0.0625 mm2 by field, under 40 objective.
We analyzed at least 10 areas for alveoli and 5 fields for BALT
[13], all were chosen randomly after morphological verification of the presence of available structure.

2.5. Statistical analysis

Analysis of differences in frequencies was performed
using the v 2 test. Comparison among means was tested by the
Kruskal-Wallis test and Dunn post test. We compared all
areas in the noninfected control and patients with measles,
with subsequent subgrouping according to lung involvement
[14]. In all tests, power of 0.90 (type II error) was used. We
considered every difference in which the probability of
similarity less than .05 ( P b .05) was significant.

3. Results
Demographic and histological parameters of our
sample are described in Table 1, with comparison to
an age-matched control group from children with left
ventricle congenital disease without significant lung
disease. Clear histological parameters of measles with lung
involvement allow classification of Hecht pneumonia (14/
42) or interstitial pneumonia (28/42) based on giant cell
presence and pneumocyte I necrosis. Most patients classified with Hecht pneumonia were considered malnourished
at necropsy, which differs from those who died with
interstitial pneumonia.
In histopathology, as shown in Fig. 1, both types of
measles lung involvement had an important interstitial
presence, but Hecht pneumonia was characterized by more
aggressive alveoli damage and giant cells than interstitial

1242

T. M. Moussallem et al.

Fig. 1 Histological and diagnostic features of measles lung involvement. A-C, Hecht pneumonia with intense alveolar epithelial
desquamation and septal thickening with giant cells (C) with viral inclusions in alveolar lumen (A, 100; B, 200; and C, 400). D and E,
Typical interstitial pneumonia with septa with inflammatory infiltrate and congestion, with minor alterations in alveolar epithelium (D, 100;
E, 200). F, Typical squamous metaplasia in measles (200). G, Immunohistochemistry for viral antigen showing positive staining of
cytoplasm in infected cells. H, Electron microscopy showing typical intranuclear RNP inclusions (27,000).

pneumonia (Fig. 1A-C). Quantitative analysis of Hecht

pneumonia showed greater aggressiveness ( P b .001, v 2
test) when compared with interstitial pneumonia with intraalveolar multinucleate giant cells with intranuclear viral
inclusions and increased frequency of pneumocyte I necrosis
and bronchial squamous metaplasia (Fig. 1F), as shown in
Table 1. Interstitial pneumonia is shown in Fig. 1D and E,
with less prominent direct viral effects despite large septum

infiltrate. In all cases, we observed viral antigens by

immunohistochemistry, mainly in the cytoplasm of infected
cells, mostly mononuclear (Fig. 1G). Via electron microscopy, most cases showed paramixovirus ribonucleoprotein
structures occupying most of the nucleus of the pneumocyte,
an aspect of viral inclusion at electron microscopy (Fig. 1H).
The quantitative distribution of immune cells in the main
lung areas as compared with the controls can be seen in

Lung involvement in childhood measles

Table 2

Measles: quantitative analysis of lung cell surface phenotype in selected areas of the lung involved in measles and controls

Surface
marker

Lung area

CD4

Alveoli
BALT
Alveoli
BALT
Alveoli
BALT
Alveoli
BALT
Alveoli
BALT
Alveoli
BALT
Alveoli
BALT

CD8
CD20
CD68
NK
S100
CD45RO

1243

Noninfected
controls

Measles
Total

Hecht pneumonia

Interstitial pneumonitis

Significance
( P b .05)

2.50
24.5
3.10
6.43
0.65
17.1
2.33
3.24
0.60
1.96
0.77
4.52
1.61
14.3

0.18 F 0.30 (42)

2.73 F 7.5 (39)
2.57 F 3.7 (42)
6.27 F 8.2 (38)
0.07 F 0.21 (42)
0.74 F 3.7 (38)
1.6 F 2.9 (41)
0.8 F 2.4 (38)
0.5 F 0.9 (42)
0.8 F 1.3 (38)
0.08 F 0.2 (42)
0.92 F 0.91 (39)
1.15 F 2.1 (42)
4.17 F 5.9 (39)

0.22 F 0.39 (14)

6.10 F 12.3 (13)
3.74 F 5.63 (14)
8.34 F 10.8 (13)
0.07 F 0.18 (14)
2.2 F 6.43 (13)
1.31 F 1.76 (13)
1.33 F 4.12 (12)
0.42 F 0.46 (14)
0.71 F 0.56 (12)
0.05 F 0.13 (14)
0.56 F 0.95 (13)
1.67 F 2.31 (14)
6.30 F 8.97 (13)

0.15
1.03
1.98
5.19
0.07
0.01
1.76
0.53
0.54
0.83
0.08
1.09
0.88
3.10

CN
CN
NS
NS
CN
CN
NS
CN
NS
CN
CN
CN
NS
CN

F
F
F
F
F
F
F
F
F
F
F
F
F
F

1.14 (18)
25.1 (18)
2.76 (18)
4.07 (18)
0.56 (17)
28.3 (18)
1.72 (18)
2.54 (18)
1.02 (18)
2.09 (18)
1.26 (18)
3.45 (18)
1.32 (18)
13.1 (18)

F
F
F
F
F
F
F
F
F
F
F
F
F
F

0.25
1.80
2.05
6.38
0.23
0.05
3.30
0.87
1.02
1.54
0.23
0.85
1.95
3.16

(28)
(26)
(28)
(25)
(28)
(26)
(28)
(26)
(28)
(26)
(28)
(26)
(28)
(26)

M (PI = H)
M (PI = H)

M (PI = H)
M (PI = H)
M (PI = H)
M (PI = H)
M (PI = H)
M (PI N H)
M (PI = H)

Data are expressed as mean of number of positive cell F SEM (analyzed sample size) by area. Comparison between groups is tested using Kruskal-Wallis
test, with Dunn post test.
Abbreviation: M, measles.

Table 2, whereas representative fields of main aspects are

shown in Fig. 2. The distribution of values of main immune
cells from individual patients is shown in Fig. 3.
BALT showed the greatest involvement of the 3
compartments analyzed in lungs involved in measles
demonstrating significant depletions of CD4+ (Fig. 2A),
as well as CD20+ (Fig. 2I), CD68+, NK (Fig. 2G), S100+,
TCD45RO+ cells in both types of pneumonias as compared
with controls. The alveoli of the lungs from patients with
measles had severe depletion of TCD4+ (Fig. 2B), S100+,
and CD20+ as compared with controls, but with similar
numbers of CD8+ cells (Fig. 2E) and other immune cells
tested. Similar findings of depletion of immune cells were
also observed in the epithelium, which was also associated
with depletion of CD68+ and NK+ cells.
Quantitative data on cytokine-producing cells can be
seen in Table 3, with representative fields of immunohistochemistry peroxidase reaction (IH) stains presented in
Fig. 2K to N. Lungs from patients with measles showed
low numbers of IL-12+ cells, unrelated to histological type,
as compared with controls. The distribution of individual
values for IL-4 and IL-10labeled cells can be seen in
Fig. 4. In the alveoli, IL-1b+, IFN+, and IL4+ cells were
found more frequently in patients with measles. In BALT,
there were low numbers of IL-10 cells in patients with
measles compared with controls. No difference was found in
any level between controls and measles in the numbers of
IL-2 and TNF-producing cells.
When the 2 types of measles pneumonias, Hecht
pneumonia and interstitial pneumonia, were compared,
more intense depletion of CD4+ (Fig. 3B) and CD68+
was observed in interstitial pneumonia as revealed by
low numbers of these cells in the BALT of these lungs, but
a relative depletion of dendritic S100+ cells was found

in Hecht pneumonia. No differences in cytokine-produced

cells were observed in these subgroups of measles
lung involvement.

4. Discussion
We observed that the main subtypes of cells detected in
situ and involved in measles lung immune dysfunction are T
helper lymphocytes (CD4+), B lymphocytes (CD20+), and
pulmonary dendritic cells (S100+), which were depleted in
all pulmonary tissues studied such as in alveoli and BALT.
Both measles forms of pneumonia, Hecht giant cell and
interstitial pneumonia, showed this pattern of involvement
of immune cells in the lung, confirmed by viral etiology
performed by immunohistochemistry and electron microscopy. The severe clinical signs and symptoms of these
patients reflect the fact that these cells are cornerstones of
the immune system, which are playing important roles in
antigen presentation and cellular and humoral immunity [7].
CD8+ cells are less affected, probably because they are
preformed and not related to the acute immune response.
The effect on dendritic cells in the lungs probably disables
the phase of transition from the innate to an adaptive
immune response, especially harming the antigen-presenting function. We could observe, in situ, the decrease of
dendritic cell amount in the lungs induced by this infection,
probably also affecting their function and quality. The
reported inhibitory effect of MV upon IL-12 production [8],
observed in peripheral blood, correlates with the findings
observed in situ in measles-affected lungs, where depletion
of IL-12+labeled cells could be demonstrated in BALT and
alveoli, probably affecting the cellular immune response,
due to the crucial role of this cytokine in this response [15].

1244

T. M. Moussallem et al.

Fig. 2 In situ detection of immune cells in the lungs from measles and noninfected children. A-C, CD4 IH; D-F, CD8 IH; A and D, BALT
from patients with measles; B and E, alveoli from patients with measles; C and F, BALT from noninfected controls; G and H, NK IH; I and J,
CD20 IH; G and I, BALT from patients with measles; H and J, BALT from noninfected controls; K-N, lung from patients with measles IH
for cytokines; K, TNF; L, IFN IH; M, IL-10; N, IL-4 (A-I, 200; K-N, 400). Dark brownlabeled positive cells are seen in Harris
hematoxylin counterstain.

This is also supported by the depletion of DC cells in our

patients with measles, similar to those found in IL-12+ cells,
which are also produced by dendritic cells [16].

CD4+ T helper lymphocytes were diminished in the

lungs, similar to dendritic cells, and significantly reduced in
alveoli, bronchial epithelium, and BALT, showing the

Lung involvement in childhood measles

1245

Fig. 3 Individual distribution of relevant immune cells as CD4+ lymphocytes (A), CD8+ lymphocytes (B), CD20 B lymphocytes (C), NK
cells (D), S100+ dendritic cells (E), and CD68+ macrophages (F) in alveolus (open symbols) or BALT (closed symbols) from the lungs of
patients with measles (dots), with Hechts pneumonia (diamonds), or interstitial pneumonia (triangles) as compared with controls (squares).
Bars represent mean values. Statistical significance is shown in Table 2.

Table 3 Measles: quantitative analysis of in situ lung cellular cytokine production in selected areas of the lung involved in measles
and controls
Cellular cytokine
staining

Lung area

IL-1b

Alveoli
BALT
Alveoli
BALT
Alveoli
BALT
Alveoli
BALT
Alveoli
BALT
Alveoli
BALT
Alveoli
BALT

IL-2
IL-4
IL-10
IFNc
TNFa
IL-12

Noninfected
controls

Measles

0.05 F 0.11 (18)

0.16 F 0.30 (18)
0 F 0 (18)
0.01 F 0.04 (18)
0.01 F 0.04 (18)
0.23 F 0.61 (18)
0.04 F 0.08 (18)
0.53 F 0.82 (17)
0 F 0 (18)
0.05 F 0.15 (16)
0.14 F 0.28 (18)
0.84 F 0.92 (18)
0.16 F 0.40 (18)
0.11 F 0.21 (18)

0.42
0.08
0
0.02
0.07
0.12
0.06
0.06
0.01
0.04
0.44
1.04
0.05
0.03

Total
F
F
F
F
F
F
F
F
F
F
F
F
F
F

0.75 (41)
0.2 (37)
0.03 (41)
0.1 (38)
0.16 (42)
0.26 (38)
0.2 (41)
0.16 (36)
0.9 (42)
1.5 (38)
0.9 (42)
1.5 (38)
0.1 (42)
0.1 (38)

Hecht pneumonia

Interstitial pneumonitis

Statistical
significance
( P b .05)

0.36
0.13
0
0.21
0.07
0.21
0.05
0.03
0.01
0.03
0.32
1
0.04
0.01

0.45 F 0.71 (27)

0.04 F 0.10 (24)
0.01 F 0.04 (27)
0.01 F 0.04 (24)
0.06 F 0.13 (28)
0.07 F 0.15 (25)
0.05 F 0.21 (27)
0.06 F 0.18 (24)
0 F 0.02 (28)
0.04 F 0.12 (25)
0.5 F 1.02 (28)
1.06 F 1.58 (25)
0.05 F 0.12 (28)
0.03 F 0.11 (25)

Cb
NS
NS
NS
Cb
NS
NS
CN
Cb
NS
NS
NS
CN
CN

F 0.84 (14)
F 0.33 (13)
F 0 (14)
F 0.37 (13)
F 0.21 (14)
F 0.37 (13)
F 0.09 (14)
F 0.07 (12)
F 0.03 (14)
F 0.07 (13)
F 0.50 (14)
F 1.25 (13)
F 0.09 (14)
F 0.05 (13)

M (PI = H)

M (PI = H)

M (PI = H)
M (PI = H)

M (PI = H)
M (PI = H)

Data were expressed as numbers of cells expressing cytokine F SEM (analyzed sample size) by 0.0625 mm2 microscopic field of selected area.
Comparison between groups is tested using Kruskal-Wallis test with Dunn post test.

1246

Fig. 4 Individual distribution of relevant populations of cytokine-producing cells, such as IL-12 and IL-4, in alveoli (open
symbols) or BALT (closed symbols) from the lungs of patients
with measles (dots), with Hechts pneumonia (diamonds), or
interstitial pneumonia (triangles) as compared with controls
(squares). Bars represent mean values. Statistical significance is
shown in Table 3.

harmful effect of MV to most cells from the immune

system. Interestingly, the number of cytotoxic lymphocytes
(CD8+) was similar in patients and controls, suggesting that
the effect of the virus or the acute inflammation may be
insufficient to reduce these cells or may be more intense in
the CD4 cells, as the infection could more efficiently
promote the death of CD4 lymphocytes by higher expression of FAS-ligand in these cells [17]. CD8+ lymphocytes
are usually activated cells that were previously selectively
grown elsewhere and probably less affected, which is a
scenario similar to that reported in blood cells [17].
Those affected T helper lymphocytes, as demonstrated
previously, express a disturbed production of cytokines
during measles [16]. In peripheral blood, the cytokines shift
from a Th1 to a Th2 pattern, being relevant IFNc and IL-2
production in initial steps of the disease, fading thereafter,
and raising the levels of IL-4 after the rash [3]. IL-12
producing cells, as described above, were severely depleted
in lung involvement in measles, both in the alveoli and
BALT, probably leading to an environment that favors Th2
shifting [18]. Further evidence supporting this phenomenon
is provided by the fact that levels of IL-4+ cells show

T. M. Moussallem et al.
greater expression in patients with measles than in controls.
Despite observing the same pattern with IFNc (the main
Th1 cytokine), it is known by experimental studies that Th2
cytokine effects tend to predominate those of Th1 cytokines
[18]. This fact is probably related to measles-induced
depletion of TCD4+ cells described above and not to the
low expression of cytokines by cells in the lungs from
control patients, which exhibit no inflammation.
BALT at rest has an environment favoring Th2 immune
response, which induces the production of IgA to the alveoli
and bronchial lumen. Incidentally, we observed that IL-10
and CD20+ lymphocytes had significant statistical differences in BALT in controls and patients with measles.
Depleting B cells and IL-10 expression probably leads to
poor IgA production favoring secondary infections in the
lungs affected by measles [19].
Macrophages also presented important depletion in
measles at BALT and bronchial epithelium, perhaps leading
to an inefficient antigen presentation [20]. Similar to other
cytokines, we believe that the absence of significant
differences in IL-1b and TNFa-producing cells in situ in
controls and measles could be secondary to MV immune
suppressor effects on the inflammatory response.
We also found a disruption of innate immunity, partially
demonstrated by depletion of NK cells at bronchial
epithelium and BALT. Losing memory immunity in the
lungs during measles, as shown in the decreased quantities
of CD45RO+ cells at BALT and bronchial epithelium,
may explain reactivation of latent infections such as
tuberculosis [17].
The 2 main histopathological manifestations were Hecht
pneumonia and interstitial pneumonia. Septal thickness and
edema, capillary congestion, alveoli macrophages, bronchial
metaplasia, pneumocyte I necrosis, pneumocyte II regeneration, and giant cells with viral inclusions were all more
severe in Hecht pneumonia than in interstitial pneumonia.
On the other hand, in assessment of IH immune cells, there
are minor differences, such as CD4+ T cells in BALT, greater
in Hecht than in interstitial pneumonia. Probably, the
intensity of lung injury is greater in Hecht pneumonia
because of the higher viral load, as detected by the presence
and numbers of giant cells and intensity of immunohistochemistry for viral antigens. CD4+ T cells are recruited from
circulation in giant cell pneumonia to clear the greater viral
load, but it also results in more inflammation and lung
damage with unsuccessful viral clearance. Our data are
isolated, and we cannot speculate on a sequential relationship
between these lung histological findings, evolving from acute
Hecht pneumonia to interstitial pneumonia, in which the
immune response already cleared most viral infected cells.
The involvement of the lung in measles, either as Hecht
or interstitial pneumonia, provokes intense disruption of
lung immunity. Despite the more severe histopathological
and clinical aspects of giant cell pneumonia, both patterns
exhibit intense depletion of immune cells in bronchial
epithelium, alveoli, and especially in BALT.

Lung involvement in childhood measles

The correlation of malnutrition and Hecht pneumonia
confirms the classic descriptions of a more severe scenario
of measles in the face of malnutrition, probably related to
the characteristic vitamin A deficit in this group of patients,
which renders the epithelia more vulnerable to the aggressive behavior of the virus and also favors secondary
infections. This observation also corroborates the fact that
prescribing vitamin A to patients with measles reduces
morbidity and mortality due to this infection [21]. Recently,
an enormous effort has begun in Africa aimed at vaccine
coverage and vitamin A supplementation to reduce child
mortality to one third its current levels within 15 years [22].
Measles is the oldest virus-induced immune-suppressive
disease known to cause human disease [2]. It continues to
cause severe childhood disease, with half a million deaths
mainly in developing countries [22], and pneumonias are
important complications [1]. Our data demonstrate several
aspects of the cells and their cytokines as inflammatory
components at alveoli and bronchi, in situ, in measles
pneumonias. We believe that these data could enhance
knowledge on measles immune pathogenesis in the lungs,
but understanding these mechanisms requires further indepth studies to eliminate vaccine-preventable disease that
continues to take lives around the world.

Acknowledgments
The authors acknowledge Dr Carlos Musso of the
Hospital Infantil Nossa Senhora da Gloria-Vitoria ES,
Brazil, and the Instituto Emlio Ribas-Sao Paulo for
providing paraffin blocks of the lungs from measles-infected
necropsies; and Dr Vera Demarchi Aiello of the Heart
Institute (InCor), University of Sao Paulo Medical School,
Brazil, for providing blocks of control cases to the present
work. The authors thank James Hesson for providing the
English revision of this article.

References
[1] Oxman MN. Measles. In: Richman DD, Whitley RJ, Hayden F,
editors. Clinical virology. New York7 Churchill Livingstone; 1997.
p. 821 - 61.
[2] Duke T, Mgone CS. Measles: not just another viral exanthem. Lancet
2003;361:763 - 73.
[3] Griffin DE, Bellini WJ. Measles virus. In: Fields BN, Knipe DM,
Howley PM, editors. Fields virology. Philadelphia7 Lippincott-Raven
Publishers; 1996. p. 1267 - 312.

1247
[4] Okada H, Sato TA, Katayama A, et al. Comparative analysis of host
responses related to immunosuppression between measles patients and
vaccine recipients with live attenuated measles vaccines. Arch Virol
2001;146:859 - 74.
[5] Oglesbee MJ, Pratt M, Carsillo T. Role for heat shock proteins in the
immune response to measles virus infection. Viral Immunol 2002;15:
399 - 416.
[6] Grosjean I, Caux C, Bella C, et al. Measles virus infects human
dendritic cells and blocks their allostimulatory properties for CD4+ T
cells. J Exp Med 1997;186:801 - 12.
[7] Peebles Jr RS, Graham BS. Viruses, dendritic cells and the lung.
Respir Res 2001;2:245 - 9.
[8] Schneider-Schaulies S, ter Meulen V. Triggering of and interference
with immune activation: interactions of measles virus with monocytes
and dendritic cells. Viral Immunol 2002;15:417 - 28.
[9] Griffin DE, Cooper SJ, Hirsch RL, et al. Changes in plasma IgE levels
during complicated and uncomplicated measles virus infections. J
Allergy Clin Immunol 1985;76:206 - 13.
[10] Gould SJ, Isaacson PG. Bronchus-associated lymphoid tissue (BALT)
in human fetal and infant lung. J Pathol 1993;169:229 - 34.
[11] Moussallem TM, Guedes F, Fernandes ER, et al. Characterization
of cellular phenotypes and cytokine expression in BALT from
children with congenital heart diseases. Pediatr Pathol Mol Med
2003;22:449 - 59.
[12] Hsu SM, Raine L, Fanger H. Use of avidin-biotin-peroxidase complex
(ABC) in immunoperoxidase techniques: a comparison between ABC
and unlabeled antibody (PAP) procedures. J Histochem Cytochem
1981;29:577 - 80.
[13] Yamamoto JH, Boletti DI, Nakashima Y, et al. Severe bilateral
necrotizing retinitis caused by Toxoplasma gondii in a patient with
systemic lupus erythematosus and diabetes mellitus. Br J Ophthalmol
2003;87:651 - 2.
[14] Motulsky HJ. Analyzing data with GraphPad prism. San Diego, CA7
GraphPad Software Inc; 1999 www.graphpad.com.
[15] Karp CL, Wysocka M, Wahl LM, et al. Mechanism of suppression
of cell-mediated immunity by measles virus. Science 1996;273:
228 - 31.
[16] Fugier-Vivier I, Servet-Delprat C, Rivailler P, Rissoan MC, Liu YJ,
Rabourdin-Combe C. Measles virus suppresses cell-mediated immunity by interfering with the survival and functions of dendritic and T
cells. J Exp Med 1997;186:813 - 23.
[17] Ryon JJ, Moss WJ, Monze M, Griffin DE. Functional and
phenotypic changes in circulating lymphocytes from hospitalized
Zambian children with measles. Clin Diagn Lab Immunol 2002;9:
994 - 1003.
[18] Spellberg B, Edwards Jr JE. Type 1/Type 2 immunity in infectious
diseases. Clin Infect Dis 2001;32:76 - 102.
[19] Hexham JM, Carayannopoulos L, Capra JD. Structure and function in
IgA. Chem Immunol 1997;65:73 - 87.
[20] Servet-Delprat C, Vidalain PO, Bausinger H, et al. Measles virus
induces abnormal differentiation of CD40 ligand-activated human
dendritic cells. J Immunol 2000;164:1753 - 60.
[21] Benn CS, Balde A, George E, et al. Effect of vitamin A
supplementation on measles-specific antibody levels in GuineaBissau. Lancet 2002;359:1313 - 4.
[22] Hoekstra EJ, McFarland JW, Shaw C, Salama P. Reducing measles
mortality, reducing child mortality. Lancet 2006;368:1050 - 2.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.