Biomed Imaging Interv J 2006; 2(2):e21
© 2006 Biomedical Imaging and
Imaging features of fungal infection in immuno-suppressed
patients in a local ward outbreak
S Ahmad Sarji1, FRCR,
WA Wan Abdullah2, MRCP,
ML Wastie1, FRCR
1 Department of Biomedical Imaging, Faculty of Medicine, University of Malaya, Kuala Lumpur, Malaysia
2 Department of Paediatrics, University of Malaya, Kuala Lumpur, Malaysia
Purpose of study: To examine the role of imaging in
diagnosing and assessing fungal infections in paediatric patients undergoing
chemotherapy in a facility, which had high fungal air contamination due to
adjacent building construction work.
Materials and method: Nineteen patients aged five
months to 12 years with various malignancies, mainly leukaemia, along with
probable fungal infection were referred for imaging over a period of 12 months.
The imaging findings from their CT and chest radiographs were reviewed by two
radiologists and correlated with the clinical findings. Blood culture and/or
biopsy of relevant lesions were performed for all patients.
Results: Fungus was positively isolated in 11 out of 19
patients, but the remaining patients clinically had fungal infection. The most
common species isolated was Candida sp. (five patients), followed by Aspergillus
sp. The most common site of fungal infection was the lungs (10 out of 19
patients), where consolidation or cavitating nodules were seen on CT or the
plain chest radiograph. One patient developed pulmonary artery aneurysm as a
complication. The other sites affected were the intra-abdominal organs
(liver, kidneys, and spleen) and the paranasal sinuses, shown on CT. Two
patients with clinical evidence of infection and Candida sp. isolated
from their blood, however, showed no abnormal findings on imaging.
Conclusion: Early diagnosis of fungal infections in
oncology patients undergoing chemotherapy is important, but diagnosis may be
difficult through imaging because of the non-specific changes and the presence of
abnormalities from the underlying disease. Even if a specific diagnosis cannot
be reached, imaging is useful to monitor response to treatment and detect
complications. � 2006 Biomedical Imaging and Intervention Journal. All rights
Keywords: Leukaemia, fungal infection, air contamination,
Fungal infections are a major cause of morbidity and
mortality in patients with haematological and other malignancies. Aspergillus
and Candida species are the most common fungal pathogens, although other
unusual fungi are also becoming recognized. Filamentous fungi, such as,
Aspergillus are soil inhabitants, but the airborne fungal contamination may
increase if the soil is disturbed. An outbreak of fungal infection in a ward
for children with leukaemia and other malignancies undergoing chemotherapy is
reported. This infection was aggravated by the soil being disturbed by adjacent
building construction. Early diagnosis and treatment of fungal infection is
important, but imaging may not provide diagnosis because of the non-specific
changes and the presence of abnormalities from the underlying disease.
MATERIALS AND METHOD
During the period 2002-2003, 19 children in a paediatric
ward were diagnosed with a proven or probable fungal infection. This ward
caters mainly to children with leukaemia; 15 children had leukaemia, one had a
medulloblastoma, one had a sacrococcygeal tumour, and two had neuroblastoma.
Air quality tests for fungal air contaminants were performed
in different parts of the ward.
Plain chest radiography and CT were used for imaging as the
majority of patients had clinical signs of respiratory tract and pulmonary
involvement. The findings on chest radiograph and CT were reviewed to see how
imaging could assist in the diagnosis of fungal infection.
A summary of patient data and findings is presented in Table
1. As the increased number of fungal infections was thought to be due to soil
disturbance at the adjacent building work site, air quality tests for fungal
air contaminants were performed. The measurement was in colony forming units/m3
(cfu/m3). Measurement obtained from the treatment room was 659 cfu/m3, one of
the single bedded rooms was 206 cfu/m3, and the paediatric day care unit was
The species of fungi isolated included Aspergillus,
Fusarium, Chrysosporium, Mycelium, Monilia, and Phaecilomyces.
Patients with haematological malignancies suffer from several
deficits in host defence, which make the patient susceptible to fungal
infections, especially Candida and Aspergillus species.
The opportunistic yeasts of the Candida species are endogenous
flora that can gain access to the bloodstream, usually through
the bowel . There may be a superficial
infection, such as, oesophageal candidiasis, and the disease
may be disseminated to visceral sites. Candida also infects
the lungs, causing a miliary nodular pattern .
The diagnosis can be difficult because other opportunistic infections
may be present and there may be superficial incidental colonisation
by Candida. Most patients with Candida infection have widespread
systemic involvement, and Candida pneumonia without systemic
involvement is rare [3,4].
In our study, Candida was isolated in five patients, but not all of them had
abnormalities on imaging. In fact,
two patients showed no abnormal findings on imaging.
The spores of Aspergillus species enter the body commonly through
the sinuses or the respiratory tract. Aspergillus infects the
airways, resulting in bronchopneumonia, but in the early stages
of the disease the chest radiograph may be normal. As the disease
progresses, a nodular appearance or patchy consolidation may
be evident. Aspergillus is angioinvasive and its most frequent
appearance is a single area or multiple areas of rounded consolidation
that may cavitate (Figure 1). The areas of consolidation are
infarcts resulting from vascular invasion by Aspergillus. Cavitation
is thought to be a favourable sign, indicating a significant
defence response . Of the four patients
with cavitating lung nodules, Aspergillus was isolated in two.
[View this figure]
|Figure 1 An 18-month-old girl with
leukaemia and Aspergillus isolated from the lung.
(a) Chest radiograph showed a round opacity behind
the heart. (b) CT revealed a cavitating nodule in
the left lower lobe.
[View this figure]
|Figure 2 A 10-year-old girl with
leukemia and biopsy proven Aspergillus infection of
the lung. She later developed a mycotic aneurysm in
the chest. (a) CT showed a cavity with a “crescent
sign” in keeping with an Aspergilloma. The lesion
was excised. (b) Chest radiograph two months later
showed a round opacity in the right lung. (c) CT revealed
an aneurysm of the lower lobe branch of the right
Because of the tendency of the organism to invade pulmonary
blood vessels, mycotic aneurysms of the pulmonary artery may
occur, which can be mistaken on the chest radiograph for areas
of consolidation (Figure 2). Extension of the pulmonary parenchymal
lesion to involve the mediastinum can occasionally cause mycotic
aneurysms of the aorta [6,7].
The other main route of Aspergillus into the body is via the
sinuses. This can cause a sinusitis that may become aggressive
and invade blood vessels, resulting in osteomyelitis and bone
destruction of the walls of the sinus. Bone destruction of the
walls of the maxillary antrum was noticed in two out of three
patients with sinus disease (Figure 3). Fungi were isolated
from the sinus wash out of these two patients.
Invasive central nervous system aspergillosis is occuring with
increased frequency , resulting in meningitis,
abscess, or encephalitis although none of these features were
seen in the current series. One child had lesions in the brain,
but the nature of the lesions was never elucidated.
[View this figure]
|Figure 3 Fungal sinusitis of a
four-year-old girl with myeloid leukaemia. CT of the
sinuses revealed opacification of the maxillary sinuses
with bone destruction of the medial wall. Mucor species
was isolated from maxillary sinus washout.
Another fungal infection is caused by the mucormycoses,
which include Mucor. Like Aspergillus, Mucor can cause vascular invasion
resulting in infarction. The sinuses are the most common site of infection, and
this was seen in one patient in our series. Involvement of the lungs can result
in consolidation and nodules that may cavitate.
Besides the lungs and the sinuses, fungal micro-abscesses may be seen in the
liver, spleen, and kidneys (Figure 4). These are most commonly
caused by Candida, but may also occur due to Aspergillus. Among
the three patients with multiple low attenuation areas in the
liver, spleen, or kidneys on CT, Aspergillus was isolated in
one patient and unidentified fungal elements were isolated in
another. But, in the third, there were no isolates.
[View this figure]
|Figure 4 Fungal abscesses in intrabdominal
organs. (a) A two-year-old girl with leukaemia and
Aspergillus sepsis. CT showed multiple hypodense lesions
in both kidneys. (b) A six-year-old girl with leukemia
and presumed fungal sepsis. CT revealed multiple hypodense
lesions in the liver and spleen. Fungal elements were
isolated from liver biopsy, but the species could
not be identified. (c) A three-year-old boy with leukemia
with presumed fungal sepsis. CT abdomen showed multiple
hypodense splenic lesions No fungi were isolated from
Other pathological fungi, such as, Zygomycetes, Trichosporon,
and Fusarium are becoming recognised [9,10],
and some of these may be resistant to the anti-fungal agent
Because of the need for early treatment, subsequent isolation
of the fungus may be difficult . Because
of the controversy over the diagnostic criteria for opportunistic
invasive fungal infection, an American-European group has developed
standard definitions for fungal infections and proposed three
levels of probability: proven, probable, and possible .
Empirical anti-fungal treatment is usually commenced in patients
after 4-6 days of persistent febrile neutropenia not showing
a response to maximum antibiotic treatment. The anti-fungal
agent of first choice is Amphotericin B because of its proven
broad spectrum efficacy, ideally in a lipid formulation. Occasionally,
Itraconazole is used as the first choice anti-fungal agent if
the renal function is impaired and the patient did not develop
the fungal infection while receiving itraconazole prophylaxis.
If a fungal infection is proved, treatment with anti-fungal
agents should be continued for several months.
Pulmonary infection counts for 70% of all fatal complications
in patients treated for acute leukaemia .
One difficulty in diagnosing pulmonary fungal infections radiologically
in immune compromised patients is caused by the multiplicity
of pulmonary disorders that may occur. Besides fungal infections,
there may be pneumonia due to viral, bacterial, or protozoal
organisms, adult respiratory distress syndrome (ARDS), infarction,
and leukaemic infiltration. In patients with leukaemia, lung
haemorrhage occurs and this has been thought to be the sole
cause of pulmonary shadowing in as high as 40% of leukaemic
patients . Other lung changes may occur
with neoplastic involvement (Figure 5) and due to the complications
of therapy. Sometimes no cause of the pulmonary shadowing can
be found, and it may be necessary to resort to lung biopsy.
Through biopsy, a diagnosis of interstitial pneumonitis or organising
pneumonia can be made. In our series, seven patients showed
lung consolidation on imaging and fungi was identified in only
three of these patients.
[View this figure]
|Figure 5 A four-year-old boy with
sacrococcygeal endodermal sinus tumour. Blood culture
grew Peiciliomyces lilacinus. (a) CT lungs showed
tiny nodular lesions, presumed involvement of lungs
by fungus. (b) After two weeks of anti-fungal therapy,
CT showed the lung lesions had increased in size and
number. (c) CT lungs after two months of continuous
anti-fungal therapy showed more and larger lesions,
which were likely to be metastases.
A study to assess the accuracy of the plain chest radiograph in diagnosing
pulmonary complications of leukaemia in children was undertaken
by Winer-Muram and her colleagues, who
compared findings on portable chest radiographs with subsequent
post-mortem findings. The accuracy for diagnosing fungal pneumonia
was only 55% and they attributed this poor diagnostic accuracy
to the variable radiographic appearance of fungal pneumonia
and the absence of characteristic radiographic features.
Similarly, in the abdomen, fungal infection can cause low attenuation areas
on CT in the liver, spleen, or kidneys, which may be mimicked
by metastases, lymphoma, pyogenic abscesses, and tuberculosis.
The diagnosis may be difficult because blood cultures may be
negative and the symptoms non-specific .
The safe environmental level of fungal counts has not been
standardised for immune compromised patients. Rhame 
found that outside, unfiltered air averages 1-15 colony forming
units/m3 (cfu/m3), while highly protected hospital areas with
air filtered at high efficiency had counts as low as 0.01 cfu/m3.
Arnow et al  found an infection rate
of about 3.5% with Aspergillus count levels of 1-2 cfu/m3. The
situation described in the present account had fungal counts
many times in excess of these values.
A similar cluster of fungal infections caused by released fungal
spores from dormant soil reservoirs, due to adjacent construction
work, was reported by Leng et al  when
eight patients with leukaemia were diagnosed with invasive sinusitis,
although none were presented with concomitant pulmonary involvement.
Aspergillus was isolated from seven of these patients. Although
the report is restricted to patients with sinus disease, it
was noted that the incidence of invasive fungal bronchopulmonary
disease was also high in the same ward. Air cultures during
the construction period frequently grew Aspergillus and the
authors advocated increased precautions against airborne pathogens
when susceptible patients are treated. Steifel et al 
found that when a seven-story building was demolished and fungal
counts outdoors increased, with preventive measures to seal
doors and windows and manipulation of the air handling system,
there was no significant rise in fungal counts in nearby hospital
In the present situation, fungal spore counts were performed
when the high incidence of fungal infections in the leukaemic children and
those with other malignant diseases was recognised. The high spore counts were
attributed to release of spores due to adjacent construction work for a new
hospital block. The best way to prevent infection was to reduce the air
contamination by installing a filter or by moving the patients to a clean area.
Neither solution was logistically possible. Measures, however, were taken to
identify high-risk patients. Because of their intensive chemotherapy regimen
and subsequent prolonged neutropenia, patients with acute myeloblastic and
lymphoblastic leukaemia, neuroblastoma, and non Hodgkin�s lymphoma were
considered at high-risk for fungal infection. Admission was discouraged in
patients with these conditions and the risks explained. The high- risk patients
were given itraconazole prophylaxis and also granulocyte colony stimulating
factor (GCSF) to reduce the degree and the duration of neutropenia induced by
chemotherapy. These measures were not highly successful as evidenced by fungal
infection occurring while the patients were on anti-fungal prophylaxis.
Fungal infection should be considered after four to six days
of febrile neutropenia. Sometimes the abnormal findings on imaging
investigations do permit a diagnosis of fungal disease, but even if a specific
diagnosis cannot be reached, imaging can monitor the response to treatment and
detect any further complications. To conclude, imaging remains important for patients
undergoing chemotherapy for leukaemia and other malignant diseases.
Segal BH, Bow EJ, Menichetti F. Fungal infections in nontransplant patients with hematologic malignancies. Infect Dis Clin North Am 2002;16(4):935-64, vii.
Pagani JJ, Libshitz HI. Opportunistic fungal pneumonias in cancer patients. AJR Am J Roentgenol 1981;137(5):1033-9.
DeGregorio MW, Lee WM, Linker CA, et al. Fungal infections in patients with acute leukemia. Am J Med 1982;73(4):543-8.
Rose HD, Sheth NK. Pulmonary candidiasis. A clinical and pathological correlation. Arch Intern Med 1978;138(6):964-5.
Gefter WB, Albelda SM, Talbot GH, et al. Invasive pulmonary aspergillosis and acute leukemia. Limitations in the diagnostic utility of the air crescent sign. Radiology 1985;157(3):605-10.
Koral K, Hall TR. Mycotic pseudoaneurysm of the aortic arch: an unusual complication of invasive pulmonary aspergillosis. Clin Imaging 2000;24(5):279-82.
Silva ME, Malogolowkin MH, Hall TR, et al. Mycotic aneurysm of the thoracic aorta due to Aspergillus terreus: case report and review. Clin Infect Dis 2000;31(5):1144-8.
Guermazi A, Gluckman E, Tabti B, et al. Invasive central nervous system aspergillosis in bone marrow transplantation recipients: an overview. Eur Radiol 2003;13(2):377-88.
Jahagirdar BN, Morrison VA. Emerging fungal pathogens in patients with hematologic malignancies and marrow/stem-cell transplant recipients. Semin Respir Infect 2002;17(2):113-20.
Ruhnke M, Maschmeyer G. Management of mycoses in patients with hematologic disease and cancer -- review of the literature. Eur J Med Res 2002;7(5):227-35.
Young RC, Bennett JE, Vogel CL, et al. Aspergillosis. The spectrum of the disease in 98 patients. Medicine (Baltimore) 1970;49(2):147-73.
Ascioglu S, Rex JH, de Pauw B, et al. Defining opportunistic invasive fungal infections in immunocompromised patients with cancer and hematopoietic stem cell transplants: an international consensus. Clin Infect Dis 2002;34(1):7-14.
Winer-Muram HT, Rubin SA, Fletcher BD, et al. Childhood leukemia: diagnostic accuracy of bedside chest radiography for severe pulmonary complications. Radiology 1994;193(1):127-33.
Tenholder MF, Hooper RG. Pulmonary infiltrates in leukemia. Chest 1980;78(3):468-73.
Francis IR, Glazer GM, Amendola MA, et al. Hepatic abscesses in the immunocompromised patient: role of CT in detection, diagnosis, management, and follow-up. Gastrointest Radiol 1986;11(3):257-62.
Rhame FS. Prevention of nosocomial aspergillosis. J Hosp Infect 1991;18 Suppl A:466-72.
Arnow PM, Sadigh M, Costas C, et al. Endemic and epidemic aspergillosis associated with in-hospital replication of Aspergillus organisms. J Infect Dis 1991;164(5):998-1002.
Lueg EA, Ballagh RH, Forte V. Analysis of the recent cluster of invasive fungal sinusitis at the Toronto Hospital for Sick Children. J Otolaryngol 1996;25(6):366-70.
Streifel AJ, Lauer JL, Vesley D, et al. Aspergillus fumigatus and other thermotolerant fungi generated by hospital building demolition. Appl Environ Microbiol 1983;46(2):375-8.
|Received 14 October 2005; received
in revised form 7 March 2006; accepted 15 March 2006
Correspondence: Department of Biomedical Imaging, Faculty of Medicine, University of Malaya, 50603 Kuala Lumpur, Malaysia. E-mail: firstname.lastname@example.org (Sazilah Ahmad Sarji).
Please cite as: Ahmad Sarji S, Wan Abdullah WA, Wastie ML, Imaging features of fungal infection in immuno-suppressed patients in a local ward outbreak, Biomed Imaging Interv J 2006;2(2):e21
This article has been viewed 23797 times.