Cardiopulmonar y Imaging • Original Research

Downloaded from www.ajronline.org by 2409:4041:2e86:ab64:c93:1ff3:d484:2dec on 06/09/21 from IP address 2409:4041:2e86:ab64:c93:1ff3:d484:2dec. Copyright ARRS. For personal use only; all rights reserved

Hammer et al.
Pulmonary Mucormycosis

Cardiopulmonary Imaging
Original Research

Pulmonary Mucormycosis:
Radiologic Features at
Presentation and Over Time
Mark M. Hammer 1 OBJECTIVE. Pulmonary mucormycosis is an aggressive opportunistic fungal infection.
Rachna Madan We set out to evaluate the CT and MRI features of pulmonary mucormycosis.
Hiroto Hatabu MATERIALS AND METHODS. Through a search of the electronic medical record
from 2007 to 2017, we identified 30 patients with definite or probable mucormycosis. Two ra-
Hammer MM, Madan R, Hatabu H diologists reviewed the initial chest CT examinations for the presence of features including
the “reverse halo” sign, large ground-glass halo, and peripheral lesion distribution. Additional
CT and MRI studies were reviewed to evaluate evolution over time.
RESULTS. The majority (67%) of patients had lesions with the reverse halo sign at some
point in the disease course. A ground-glass halo larger than the lesion was seen in 53% of pa-
tients. Notably, lesions had a peripheral predominance in 87% of cases. Through careful re-
view of images, a perivascular ground-glass precursor lesion was identified in 20% of patients
1–2 weeks before a consolidation developed. In five (17%) patients, CT showed a multifocal
pneumonia appearance. Finally, MRI of two patients showed T2-hypointense rims and central
nonenhancement, a finding we refer to as the “black hole” sign.
CONCLUSION. Large nodules or consolidations with an associated reverse halo sign or
large perilesional ground-glass halos are common in mucormycosis. Lesions tend to show a
peripheral predominance, and a perivascular ground-glass focus preceded nodular lesions in
some cases. In some patients with severe disease, imaging features evolved to show a multifo-
cal pneumonia pattern, and this pattern was associated with a high mortality rate.

ucormycosis is an opportunistic successful. Therefore, imaging findings that

M fungal infection caused by fungi

Keywords: angioinvasive fungal infection,
­mucormycosis, reverse halo sign, Zygomycetes
doi.org/10.2214/AJR.17.18792
Received July 23, 2017; accepted after revision
August 31, 2017.
1
All authors: Department of Radiology, Brigham and
Women’s Hospital, Harvard Medical School, 75 Francis
St, Boston, MA 02115. Address correspondence to
M. M. Hammer (mmhammer@bwh.harvard.edu).
AJR 2018; 210:742–747
0361–803X/18/2104–742
© American Roentgen Ray Society
of the order Mucorales in the
class Zygomycetes. Also known
as zygomycosis, it is most commonly caused
by fungal species in the genera Rhizopus,
­Lichtheimia, and Mucor [1]. Mucormycosis
is far less common than other opportunistic
fungal infections, such as Aspergillus and
Candida infections, although the mortality
rate is much higher. It most commonly in-
volves the sinuses, lungs, and skin. In this ar-
ticle, we focus on pulmonary mucormycosis.
Correct diagnosis of mucormycosis is im-
portant because effective treatment requires
the use of more toxic antifungal drugs than
are needed to treat aspergillosis [2]. Clini-
cal diagnosis of mucormycosis is challeng-
ing because the clinical features overlap with
those of other infections and because of the
lack of blood markers such as galactoman-
nan and β-d-glucan [3]. Hence, definitive di-
agnosis often requires invasive sampling,
which may delay treatment and is not always
suggest the diagnosis play an important role
in the management of these patients.
Mucormycosis is an aggressive angioinva-
sive infection that often leads to pulmonary
infarction. Mucormycosis can manifest as
the “reverse halo” (or “reversed halo”) sign
on CT, which describes a consolidation with
central ground-glass [4]. Several studies have
shown the reverse halo sign to be a frequent-
ly seen and relatively specific sign of mucor-
mycosis on CT [5–8]. However, other imag-
ing manifestations are less well studied, and
the temporal evolution of lesions in mucor-
mycosis is not well elucidated.
In this study, we set out to evaluate CT
and MRI examinations performed in patients
with mucormycosis to determine common
imaging findings and temporal resolution.
Materials and Methods
Patient Selection
This study was approved by the institutional re-
view board. We searched the electronic medical re-

742 AJR:210, April 2018

 Pulmonary Mucormycosis
Downloaded from www.ajronline.org by 2409:4041:2e86:ab64:c93:1ff3:d484:2dec on 06/09/21 from IP address 2409:4041:2e86:ab64:c93:1ff3:d484:2dec. Copyright ARRS. For personal use only; all rights reserved

TABLE 1: Clinical Characteristics of 30 Patients With Pulmonary M

­ ucormycosis radiologists then reviewed all CT studies for each
patient to identify the presence of the reverse halo
Characteristic Value
sign, cavitation, vascular cutoff, and multifocal
Age (y) pneumonia appearance, among other features. The
Median 60 multifocal pneumonia pattern was defined by ex-
tensive bilateral multifocal areas of consolidation.
Range 37–84
Sex, no. (%) of patients Results
Male 13 (43) Patient Characteristics
Female 17 (57) We identified 30 patients with pulmonary
mucormycosis (Table 1). Most of these pa-
Underlying disease or history of transplant, no. (%) of patients
tients had benign or malignant hematologic
Hematologic diseasea 25 (83) diseases (83%), and many (37%) had under-
Bone marrow transplant 11 (37) gone bone marrow transplant. The remain-
Solid organ transplant 4 (13) ing patients had undergone solid organ trans-
plant (in particular, heart and lung) or had
Other malignancy 1 (3)
other malignancies. None was simply a dia-
Absolute neutrophil count at diagnosisb, no. (%) of patients betic patient without other risk factors. Just
< 500 μL 11 (46) under half of the patients (46%) with avail-
≥ 500 μL 13 (54) able complete blood count results were neu-
tropenic at the time of diagnosis.
Method of diagnosis, no. (%) of patients
Diagnoses were made using a variety of
Bronchoalveolar lavage or transbronchial biopsy 8 (27) methods in this group. The majority of pa-
Percutaneous biopsy 10 (33) tients underwent minimally invasive biop-
Surgical or other biopsy 4 (13) sies, either through bronchoscopic evaluation
(bronchoalveolar lavage or transbronchial bi-
Autopsy 5 (17)
opsy, 27%) or via CT-guided percutaneous
Clinical diagnosis 3 (10) lung biopsy (33%). Cultures confirmed Zy-
Death from mucormycosis 16 (53) gomycetes species in eight patients. In pa-
Median time from first CT evidence of disease to death (d) 27 tients with culture-negative findings, the pa-
aHematologic diseases include malignancies such as leukemia and lymphoma as well as benign diseases such thology results were read as consistent with
as aplastic anemia or myelodysplastic syndrome. Mucor species or as favoring Mucor species
bData were available for 24 patients. in 17 cases. In two patients, Mucor infection
was listed in the pathologic differential di-
cord from 2007 to 2017 for patients with radiolo- cavitation. We considered nodules as well-defined agnosis. In two cases, patients had coinfec-
gy or pathology reports mentioning mucormycosis rounded lesions; consolidations were considered tion with other agents (one case of respirato-
or zygomycosis. A careful chart review was then to be geographic and ill defined. The reverse halo ry syncytial virus and one case of Aspergillus
performed to identify 30 patients, all of whom had sign is defined by peripheral consolidation with infection). None of these biopsies were com-
either pathologically proven invasive fungal infec- central ground-glass or clearing. The “large halo” plicated by pneumothorax or significant hem-
tion consistent with or suggestive of Zygomyce- sign is defined by a ground-glass halo around a le- orrhage. Three patients were diagnosed clin-
tes infection or clinically diagnosed Zygomycetes sion that is much larger than the lesion. One of the ically through multidisciplinary consensus,
pneumonia by multidisciplinary consensus [9]. We
recorded the absolute neutrophil count at the time
of initial presentation for patients whose medical
records included complete blood count results with
differential in the medical record.

Image Review
Two thoracic radiologists reviewed the CT stud-
ies available for each patient to identify the ear-
liest CT examination with imaging features con-
vincingly related to the fungal infection. These CT
studies were then reviewed in consensus to iden-
tify the following characteristics: the presence of
nodules, presence of consolidations, number of A B
lesions, size of the largest lesion, presence of the Fig. 1—CT images of patients with pulmonary mucormycosis.
A and B, 58-year-old man with acute myeloid leukemia who presented for follow-up imaging after allogeneic stem
reverse halo sign, presence of a large perilesional cell transplant. CT images of lungs obtained at initial presentation (A) and 1 week later (B) show small ground-glass
halo, peripheral location of lesions, and presence of focus (arrow, A) in right upper lobe that developed into consolidation with “reverse halo” sign (arrow, B).
(Fig. 1 continues on next page)

AJR:210, April 2018 743

 Hammer et al.
Downloaded from www.ajronline.org by 2409:4041:2e86:ab64:c93:1ff3:d484:2dec on 06/09/21 from IP address 2409:4041:2e86:ab64:c93:1ff3:d484:2dec. Copyright ARRS. For personal use only; all rights reserved

with biopsy either not performed or yielding

unrevealing results.
There was high mortality in this cohort
with pulmonary disease attributable to mu-
cormycosis: 16 of 30 patients (53%) died of
a cause related to the infection. Death oc-
curred a median of 27 days after the first CT
evidence of disease. Among the 30 cases of
pulmonary mucormycosis, lung involvement
was seen in all patients, one patient had pul-
monic valve endocarditis, and one patient
had endobronchial involvement.
C D
Parenchymal Disease and Disease Evolution
on CT
We evaluated the sequential imaging find-
ings on CT in this group of 30 patients. In six
patients (20%), a ground-glass nodular lesion
was identified at a site that later developed a
nodule or consolidation. These lesions tend-
ed to be perivascular in location and were
seen as early as 1–2 weeks before the de-
velopment of a full-blown infectious picture
E F
(Figs. 1A, 1C, and 1E).
Fig. 1 (continued)—CT images of patients with pulmonary mucormycosis.
C and D, 61-year-old woman with acute myeloid leukemia who presented for follow-up imaging after allogeneic We then evaluated the characteristics of
stem cell transplant. CT images of lungs obtained at initial presentation (C) and 1 week later (D) show small ground- the fungal lesions on the first CT examina-
glass focus (arrow, C) in left upper lobe on initial image that developed into large consolidation with reverse halo tion that showed convincing evidence of
sign (arrow, D). There is additional consolidation in right upper lobe with reverse halo sign (asterisks).
infection. Both nodular and consolidative
E and F, 45-year-old woman with acute lymphocytic leukemia who presented for follow-up imaging after allogeneic
stem cell transplant. CT images of lungs obtained at initial presentation (E) and 1 week later (F) show small ground- patterns were seen in these patients, with
glass focus in right upper lobe (arrow, E) that developed into consolidation with reverse halo sign (arrow, F). Large nodules slightly more common (67% and
halo sign (asterisks, F) is also shown in F. 57%, respectively) (Table 2). Seven patients
(23%) had both patterns. The majority, 63%,
of patients had multiple lesions; 23% of pa-
TABLE 2: Findings on Initial CT Study of 30 Patients With Pulmonary
­Mucormycosis tients had 10 or more lesions. Lesions ranged
in size from 2.1 to 11.9 cm, with a median of
CT Finding Value 4.2 cm. In patients with consolidations, we
Nodule or nodules present, no. (%) of patients 20 (67) noted a relative paucity of air bronchograms
as compared to patients with bacterial pneu-
Consolidation or consolidations present, no. (%) of patients 17 (57)
monias and consolidations of similar size.
No. of lesions per patient, no. (%) of patients The reverse halo sign, defined as periph-
1 11 (37) eral consolidation with central ground-glass,
2–9 12 (40) was present on the initial CT study in 18 of
30 (60%) patients (Figs. 1B, 1D, and 1F). A
≥ 10 7 (23)
large perilesional halo, defined as a ground-
Multifocal pneumonia pattern 2 (7) glass halo much larger than the solid lesion
Size of largest lesion (cm) at its center, was present in 53% of patients
Median 4.2 (Fig. 1F). Finally, lesions in 26 (87%) patients
had a predominantly peripheral distribution
Range 2.1–11.9
within the lungs.
CT finding, no. (%) of patients We recorded the imaging findings seen on
“Reverse halo” sign 18 (60) subsequent CT studies to document the evo-
Large perilesional haloa 16 (53) lution of CT findings, which are listed in Ta-
ble 3. Cavitation was seen in only three of 30
Peripheral location of lesions 26 (87)
(10%) patients on initial imaging and devel-
Cavitation 3 (10) oped in eight additional patients on subsequent
aDefined as a ground-glass halo much larger than the solid lesion at its center.
CT studies, for a total of 11 patients (37%). Of

744 AJR:210, April 2018

 Pulmonary Mucormycosis
Downloaded from www.ajronline.org by 2409:4041:2e86:ab64:c93:1ff3:d484:2dec on 06/09/21 from IP address 2409:4041:2e86:ab64:c93:1ff3:d484:2dec. Copyright ARRS. For personal use only; all rights reserved

TABLE 3: CT Findings in 30 Patients with Pulmonary Mucormycosis on Initial it showed evidence of susceptibility on lon-
and Subsequent CT Studies ger-TE sequences. Additionally, in two pa-
CT Finding Initial CT Study Initial and Subsequent CT Studies
tients, contrast-enhanced images showed a
complete lack of enhancement of the lesion,
“Reverse halo” sign 18 (60) 20 (67) a finding we term the “black hole” sign (Figs.
Cavitation 3 (10) 11 (37) 4C and 4D).
Multifocal pneumonia pattern 2 (7) 5 (17)
Discussion
“Vascular cutoff” sign — 6 (20)
We evaluated clinical risk factors and im-
Note—Data are reported as number (%) of patients. Dash (—) indicates data not evaluated. aging findings in a series of patients with
pulmonary mucormycosis. The predisposing
note, the reverse halo sign was not seen in all Vascular Findings factors in more than 80% of patients in our
of these patients (present in 7/11). In two pa- In the patient with mucormycosis endo- series were hematologic disorders, particu-
tients (7%), the initial presentation with exten- carditis, pulmonary arterial filling defects larly hematologic malignancies. Although
sive bilateral consolidations mimicked that of (macroscopic septic emboli) were seen, and patients with hematologic malignancies have
multifocal bacterial pneumonia. An additional the patient had developed pulmonary artery always represented a fraction of the cases
three patients developed multifocal pneumo- pseudoaneurysms (Fig. 3A). Pulmonary em- of pulmonary mucormycosis, most articles
nia on subsequent CT studies (Fig. 2) for a to- boli were also identified in another patient have not identified such a preponderance as
tal of five patients (17%). Within the group of on pulmonary CT angiography performed 3 was seen in our series [1, 10]. This difference
patients with multifocal pneumonias, mortal- days after the initial unenhanced CT study. A likely reflects the fact that our institution
ity attributable to mucormycosis was seen in “vascular cutoff” sign (i.e., abrupt termina- serves as a large cancer referral center but
four of five patients (80%); of the remaining tion of a pulmonary artery branch) was iden- may also reflect the increasing life span of
25 patients, 12 (48%) died. Reverse halo signs tified in six patients (20%) (Fig. 3B). patients with hematologic malignancies and
were seen in four of five patients with the mul- new therapies in hematologic malignancies.
tifocal pneumonia pattern. MRI Findings Indeed, the recent use of prophylactic anti-
Finally, because mucormycosis is an ag- Three patients underwent MRI examina- fungal medications in certain patient popula-
gressive infection, it commonly invades tions that included the lesions seen on CT tions has led to a lower incidence of invasive
across normal tissue boundaries. We identified (one chest MRI study, two abdominal MRI Aspergillus infections and relatively more
two patients with mediastinal invasion (one of studies). In two patients, a T2-hypointense breakthrough Zygomycetes infections, be-
whom also had chest wall invasion), one pa- rim was seen along the edge of the lesion cause those antifungal drugs have no activ-
tient with diaphragm invasion, and one patient (Figs. 4A and 4B). The rim was relatively ity against Zygomycetes infections [2]. Only
with extension of infection across a fissure. isointense on T1-weighted ­images, although just under half of patients were neutropenic

Fig. 2—52-year-old man with acute myeloid leukemia

who presented for follow-up imaging after allogeneic
stem cell transplant. CT images show pulmonary
mucormycosis.
A and B, CT images of lungs obtained at initial
presentation (A) and 2 weeks later (B) show initial
single consolidation with reverse halo sign (asterisk,
A) with subsequent evolution into multifocal
pneumonia involving both lungs. This appearance
simulates bacterial pneumonia or diffuse alveolar
damage.
A B

Fig. 3—Vascular findings in patients with pulmonary

mucormycosis.
A, 60-year-old woman with myelodysplastic
syndrome on deferoxamine with Mucor endocarditis
on pulmonic valve (not shown). CT pulmonary
angiogram shows filling defect in right lower lobe
pulmonary artery (arrow) that is compatible with
septic pulmonary embolism.
B, 65-year-old man with multifocal pulmonary
mucormycosis who presented for follow-up imaging
after bilateral lung transplant. CT image shows
consolidation in left upper lobe (asterisk) with abrupt
cutoff of pulmonary artery branch (arrow) feeding it.
A B

AJR:210, April 2018 745

 Hammer et al.
Downloaded from www.ajronline.org by 2409:4041:2e86:ab64:c93:1ff3:d484:2dec on 06/09/21 from IP address 2409:4041:2e86:ab64:c93:1ff3:d484:2dec. Copyright ARRS. For personal use only; all rights reserved

at the time of diagnosis; thus, mucormycosis (20%), a vascular cutoff sign was observed went on to develop a multifocal pneumonia
should not be excluded from a differential di- in association with these lesions. These fea- pattern (seen in just under 20% of cases). This
agnosis if the patient is not neutropenic. We tures are consistent with a perivascular pro- pattern can mimic bacterial pneumonia; thus,
did see a high mortality in our series of 53%, cess that leads to infarction of the lung distal in critically ill immunocompromised patients,
similar to previous series [1, 10], reflecting to the involved vessel. Outside the setting of fungal pneumonia—particularly mucormyco-
the fact that this disease remains highly le- neutropenia, similar features can be seen in sis—should be considered in the differential
thal despite advances in medical therapy. septic emboli, suggesting that in some cas- diagnosis of multifocal airspace disease par-
As described in several previous stud- es, pulmonary mucormycosis begins as a he- ticularly if a reverse halo sign is present. This
ies, the reverse halo sign was seen in the matogenously seeded process. Indeed, in one appearance was associated with a high mor-
majority of patients with mucormycosis in case in our series, a patient developed mucor- tality in our series (80%) and may reflect pro-
our study. However, we identified addition- mycosis endocarditis of the pulmonic valve gressive fungal infection as the host immune
al imaging features present in the majority with macroscopic septic pulmonary emboli. system further deteriorates.
of cases. These CT features include a large The evolution of pulmonary mucormyco- Finally, three patients in our series under-
ground-glass halo, which we defined as a sis over time has been reported in small series went MRI. In two patients, imaging showed
ground-glass halo much larger than the ­solid [11, 12]. However, these studies have focused a T2-hypointense rim at the edge of the le-
lesion at its center, which was present in 53% on the subsequent evolution of the nodular or sions. This MRI finding likely represents
of patients. This halo presumably represents consolidative lesions not on the earlier mani- blood products related to the hemorrhagic in-
hemorrhage, and its large size likely indi- festations. We discovered in six patients that, farction, although it could alternatively rep-
cates that mucormycosis results in greater in locations where the nodular or consolida- resent the fungal organisms which concen-
pulmonary hemorrhage than other angioin- tive lesions later developed, a small ground- trate metals such as iron, magnesium, and
vasive infections, such as Aspergillus infec- glass focus could be identified 1–2 weeks ear- manganese [13]. Additionally, MR images
tions, that typically have only a small halo lier. This ground-glass precursor lesion was showed a complete lack of contrast enhance-
around the nodules. often perivascular and may represent a small ment within the lesions, a finding we termed
Another common imaging feature in our amount of hemorrhage resulting from the fun- the “black hole” sign. To our knowledge, this
patients was the peripheral distribution of le- gus attacking the vessel’s wall. Later in the article is the first report of MRI findings in
sions. In 87% of patients, the lesions had a course of infection, we observed cavitation in pulmonary mucormycosis. Further study
peripheral predominance. In some patients a minority of patients, whereas other patients will be needed to evaluate the sensitivity and
specificity of these findings.
Our study is limited by its retrospective
cohort study with inconsistent imaging and
management of patients. We are also some-
what limited by a small sample size, although
that is to be expected with a rare disease. Our
study included only patients with pulmonary
mucormycosis, so we are not able to assess
for the specificity of our findings. Howev-
er, we do note that the reverse halo sign has
been shown to be specific for mucormycosis
in previous studies [6, 7]. Additionally, our
sample mostly included patients with patho-
A B logically proven mucormycosis, so it is pos-
sible that this introduced a bias in the spec-
trum of imaging features that we saw.
In summary, pulmonary mucormycosis is
an uncommon but deadly opportunistic in-
fection. Although classically described in pa-
tients with diabetes, in the modern era, it is
most commonly seen in patients with hema-
tologic malignancies, particularly in patients
who have undergone stem cell transplant.
Common CT appearances of mucormycosis
include the reverse halo and large halo signs.
Occasionally, one may observe a perivascular
C D ground-glass focus early in the course of in-
Fig. 4—41-year-old man with acute myeloid leukemia and pulmonary mucormycosis. fection, indicating the possible hematogenous
A and B, T2-weighted MR images show two consolidative lesions with T2-hypointense rims (arrows). Note that spread of this infection. In severely immuno-
inferior lesion is invading mediastinum and chest wall.
C and D, Subtracted contrast-enhanced T1-weighted images show complete lack of enhancement within compromised patients or patients with ter-
lesions (asterisks), which we termed “black hole” sign. minal disease, mucormycosis may appear as

746 AJR:210, April 2018


Downloaded from www.ajronline.org by 2409:4041:2e86:ab64:c93:1ff3:d484:2dec on 06/09/21 from IP address 2409:4041:2e86:ab64:c93:1ff3:d484:2dec. Copyright ARRS. For personal use only; all rights reserved
multifocal pneumonia, an appearance insep-
arable from bacterial multifocal pneumonia
or massive aspiration. This pattern of mucor-
mycosis is associated with high mortality and
has not been reported to date in the radiologic
literature. We also described the MRI black
hole sign, which is defined as a T2-hypoin-
tense rim and indicates probable hemorrhagic
infarction. Using these imaging features, radi-
ologists may be able to suggest the diagnosis
of mucormycosis, which will lead to potential
tissue sampling and the appropriate therapy.
References
1. Petrikkos G, Skiada A, Lortholary O, Roilides E,
Walsh TJ, Kontoyiannis DP. Epidemiology and
clinical manifestations of mucormycosis. Clin
­Infect Dis 2012; 54(suppl 1):S23–S34
2. Spellberg B, Ibrahim AS. Recent advances in the
treatment of mucormycosis. Curr Infect Dis Rep
2010; 12:423–429
AJR:210, April 2018 747
Pulmonary Mucormycosis
3. Kontoyiannis DP, Lewis RE. How I treat mucor-
mycosis. Blood 2011; 118:1216–1224
4. Chung JH, Godwin JD, Chien JW, Pipavath SJ.
Case 160: pulmonary mucormycosis. Radiology
2010; 256:667–670
5. Wahba H, Truong MT, Lei X, Kontoyiannis DP,
­Marom EM. Reversed halo sign in invasive pulmonary
fungal infections. Clin Infect Dis 2008; 46:1733–1737
6. Jung J, Kim MY, Lee HJ, et al. Comparison of
computed tomographic findings in pulmonary mu-
cormycosis and invasive pulmonary aspergillosis.
Clin Microbiol Infect 2015; 21:684.e11–684.e18
7. Georgiadou SP, Sipsas NV, Marom EM, ­Kontoyiannis
DP. The diagnostic value of halo and reversed halo
signs for invasive mold infections in compromised
hosts. Clin Infect Dis 2011; 52:1144–1155
8. Legouge C, Caillot D, Chrétien ML, et al. The re-
versed halo sign: pathognomonic pattern of pul-
monary mucormycosis in leukemic patients with
neutropenia? Clin Infect Dis 2014; 58:672–678
9. De Pauw B, Walsh TJ, Donnelly JP, et al Revised
definitions of invasive fungal disease from the Euro-
pean Organization for Research and Treatment of Can-
cer/Invasive Fungal Infections Cooperative Group and
the National Institute of Allergy and Infectious Dis-
eases Mycoses Study Group (EORTC/MSG) Consen-
sus Group. Clin Infect Dis 2008; 46:1813–1821
10. Lee FYW, Mossad SB, Adal KA. Pulmonary mu-
cormycosis: the last 30 years. Arch Intern Med
1999; 159:1301–1309
11. Choo JY, Park CM, Lee H-J, Lee CH, Goo JM, Im
JG. Sequential morphological changes in follow-
up CT of pulmonary mucormycosis. Diagn Interv
Radiol 2014; 20:42–46
12. Nam BD, Kim TJ, Lee KS, Kim TS, Han J, Chung
MJ. Pulmonary mucormycosis: serial morphologic
changes on computed tomography correlate with
clinical and pathologic findings. Eur Radiol 2018;
28:788–795
13. Aribandi M, McCoy VA, Bazan C. Imaging fea-
tures of invasive and noninvasive fungal sinusitis:
a review. RadioGraphics 2007; 27:1283–1296