CASE REPORT

Bronchiolitis
obliterans an
illustrative case
following toxic
fume exposure

Basil J Sher
FCRad(D)SA

Ian C Duncan
FFRad(D)SA

SunninghilI Medical Institute
Johannesburg

Abstract
Bronchiolitis obliterans is a condi-

tion where irreversible narrowing of
small airways leads to chronic airflow
obstruction. Usually an associated
underlying condition can be identi-
fied. The diagnosis depends upon this
together with the appropriate lung
function abnormalities and radiologi-
cal features. We describe a case of
bronchiolitis obliterans that devel-
oped after toxic fume inhalation and
which demonstrates the typical clini-
cal, physiological and radiological fea-
tures on which the diagnosis of this
condition is made.

Introduction
Bronchiolitis obliterans is a condi-

tion where chronic irreversible airflow
obstruction develops in the small air-
ways in response to some associated

noxious stimulus. Making the diag-
nosis depends upon a constellation of
diagnostic features including: (l) the
appropriate history of an associated
underlying causative pathology or
noxious stimulus; (ii) clinical evidence
of diminished lung function; and (ii1)
the appropriate correlative radiologi-
cal features.

Case report
A previously healthy 26-year-old

male presented with shortness of
breath and wheezing. An industrial
accident had involved toxic fume
inhalation some 5 months prior to
this. An obstructive picture was
obtained on lung function testing. A
high-resolution CT scan of the chest
was requested to investigate the lung

abnormality further.
Inspiratory and expiratory chest

radiographs were normal (Fig. I).
High-resolution CT images of the

lungs were obtained using 1.5 mm
collimation and 10 mm spacing dur-
ing both end-inspiration and end-
expiration. The images were initially
viewed at a window level of -645
Hounsfield units (HU) and a window
width of 1 660 HU, and the appear-
ance suggested a mosaic attenuation
pattern accentuated with end-expira-
tion compatible with areas of air trap-
ping (Fig. 2). This appearance was fur-
ther exaggerated by decreasing the
window level to -902 HU and the
window width to 1 023 HU (Fig. 3).
Bronchiectasis was present involving
the lower lobe segmental bronchi
(Fig. 4). There were no areas of con-
solidation, and no reticulonodular
opacities were noted.

On the basis of the lung function
testing and the CT appearance, and
given the history of toxic fume expo-
sure, a diagnosis of bronchiolitis oblit-
erans was made.

Discussion
Bronchiolitis obliterans (also

termed constructive bronchiolitis, or
obliterative bronchiolitis) is charac-

B

Figs te and lb. Inspiratory (1a) and expiratory (Ib) frontal chest radiographs appear normal, particularly with
a normal pulmonary vascular appearance and no evidence of air trapping.

43 SA JOURNAL OF RADIOLOGY • December 2002



CASE REPORT

Figs 2a and b. High-resolution scan slices taken during inspiration (2a) and expiration (2b). Areas of mosa-
ic attenuation are apparent on the expiratory phase scan.

Figs Sa and b. The same scan slices as in Fig. 2 taken at different window settings that further accentuate
the mosaic attenuation pattern. The darker areas are the normal ones indicating areas of air trapping and
hypoperfusion.

Fig. 4. Bronch/al dl/ation and wall thickening are
also present.

terised histologically by fibrosis of the
submucosal and peribronchial tissues
of terminal and respiratory bronchi-
oles producing concentric narrowing
of the bronchiolar lumen.
Characteristically there is very little
associated inflammation. The clinical

severity is dependent upon the num-
ber of bronchioles involved as well as
the degree of narrowing of their lumi-
na. It is non-uniform in distribu-
tion. No inflammatory or infiltra-
tive changes are seen in the adjacent
parenchyma.

The development of bronchiolitis
obliterans is closely related to a num-
ber of associated pathologies (Table I).
It is very rarely truly idiopathic.

The diagnosis is extremely difficult
to make at open lung biopsy, and
relies more upon the combination of
clinical features, lung function testing
and the radiographic findings.
Usually there is a history of some rele-
vant associated pathology. Lung func-
tion testing shows irreversible airflow
obstruction with a forced expiratory
volume in 1 second (FEVI) of less

44 SA JOURNAL OF RADIOLOGY • December 2002

Table I. Aetiological associations of
bronchiolitis obliterans

I. Post infective
Viral

Adenovirus
Respiratory syncytial virus
Influenza
Parainfluenza
Cytomegalovirus
HIV

Non viral
Mycoplasma pneumonia
Pneumocystis carinii
Bacterial pneumonia

2. Toxic fume inhalation
Nitrogen dioxide (silo-filler's dis-
ease)
Sulphur dioxide
Ammonia
Chlorine
Phosgene

3. Drug reaction
Penicillamine
Gold
Lomustine

4. Connective tissue disorders
Rheumatoid arthritis
Sjogren's syndrome
Polymyositis

5. Transplantation
Bone marrow transplant (graft-
versus-host rejection)
Lung transplant
Heart and lung transplant

6. Miscellaneous
Inflammatory bowel disease
Pulmonary neuroendocrine cell
hyperplasia/carcinoid microtu-
mours
Bronchopulmonary dysplasia
S. androgynus ingestion
Bronchiectasis, cystic fibrosis
Cryptogenic - very rare

than 60% of the predicted value in the
absence of any other cause of airway
obstruction.

Plain film chest radiography may
show either no abnormalities or evi-
dence of mild air trapping or hyperin-



CASE REPORT

flation and mild peripheral pul-
monary vascular attenuation.

The high resolution CT (HRCT)
findings include the following:

1. The 'mosaic attenuation' or
'mosaic perfusion' pattern. This con-
sists of multifocal areas of decreased
lung attenuation, i.e. 'darker areas,
which may be either poorly or sharply
defined. These areas of decreased
attenuation result from a combina-
tion of air-trapping secondary to the
bronchiolar narrowing as well as
reduced perfusion due to secondary
hypoxic vasoconstriction in these
poorly ventilated areas. As a result
there is a preferential shunting of
blood to the normally ventilated areas
of the lung which accentuates the
mosaic appearance. With more
extensive lung involvement the mosa-
ic pattern is replaced by a more
homogeneous hypoattenuation pat-
tern. The term 'mosaic attenuation' is
somewhat of a misnomer as the
pathological areas of the lung
parenchyma are actually the darker
ones, i.e. the areas of reduced beam
attenuation. The mosiac pattern
becomes even more exaggerated on
expiratory HRCT images as the areas
of air trapping remain dark whereas
the normally perfused areas (which
are relatively hyperperfused) become
relatively denser and smaller in cross-
sectional area.

2. Decreased calibre of the pul-
monary vasculature in affected
areas. There is a reflex hypoxic vaso-
constriction in the areas where air

trapping is found. With larger con-
fluent areas of involvement this results
in the narrowing of macroscopic pul-
monary arterial branches. Eventually
this vascular constriction becomes
permanent.

3. Bronchiectasis and bronchial
wall thickening. In many cases both
bronchial dilation and wall thickening
are seen in the affected areas.
Occasionally the obstructed bronchi-
oles can fill with inspissated secretions
creating small centrilobular branch-
ing opacities or the 'tree-in-bud'
appearance.

Finally it is important to note the
following points concerning bronchi-
olitis obliterans. Firstly, if the areas of
air-trapping are extensive enough the
characteristic mosaic pattern may not
be seen and the lungs may appear
homogeneously dark on end-expira-
tory scans. On the other hand, subtle
changes may require manipulation of
the image window levels and widths
to show the affected areas adequately.

Secondly, the signs mentioned
above are not unique to bronchiolitis
obliterans and may be seen in other
obstructive pulmonary diseases.
Therefore the diagnosis is made based
upon a combination of the clinical
features, HRCT findings, and history
or clinical evidence of any relevant
associated cause or underlying condi-
tion.

Thirdly, the Swyer-lames or
Macleod's syndrome is probably the
same pathological process occurring
in the relatively immature lungs of

infants and young children, usually in
response to a viral lung infection.

And lastly, as the anatomical
changes related to bronchiolitis oblit-
erans are permanent due to bronchio-
lar fibrosis with little or no associated
inflammatory component there is
generally a poor clinical response to
steroid therapy.

Conclusion
The diagnosis of bronchiolitis

obliterans cannot be made on the
radiological features alone but
requires correlation with the lung
function testing and relevant clinical
information as well. HRCT scanning
is useful in showing areas of air-trap-
ping, secondary circulatory shunting
and vascular attenuation, but requires
meticulous attention to the acquisi-
tion and presentation of the images so
as to optimise and accentuate the
mosaic attenuation pattern and other
associated features.

Recommended reading
I. Webb WR, Muller NL, Naidich OP. Airway's

diseases. In: Webb WR, Muller NL, Naidich OP,
eds. High Resolution CT of the Lung. 3rd ed.
Philadelphia: Lippincott, Williams and Wilkins,
2001: 467-546.

2. Worthy SA, Muller NL. Small airway diseases.
Radial Clin North Am 1998; 36: 163-173.

3. Arakawa H, Webb WR. Expiratory high-resolu-
tion CT scan. Radial Clin North Am J 998;
36:189-209.

4. Hansell OM. HRCT of obliterative bronchioli-
tis and other small airways diseases. Sernin
Roelltgenol200 I; 36: 51-65.

5. Collins J, Stern EI. 'Tree-in-bud' pattern of
bronchiolar disease. TIle Radiologist 1999; 6:
173-181.

45 SA JOURNAL OF RADIOLOGY • December 2002