Vous êtes sur la page 1sur 6

European Journal of Radiology 84 (2015) 2339–2344

Contents lists available at ScienceDirect

European Journal of Radiology


journal homepage: www.elsevier.com/locate/ejrad

The relation between CT findings and sputum microbiology studies in


active pulmonary tuberculosis
Jeong Min Ko a , Hyun Jin Park a,∗ , Chi Hong Kim b , Sun Wha Song a
a
Department of Radiology, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea
b
Department of Internal Medicine, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea

a r t i c l e i n f o a b s t r a c t

Article history: Purpose: To evaluate whether CT findings suggesting active pulmonary tuberculosis correlate with spu-
Received 21 April 2015 tum microbiological studies, and to determine whether CT could predict infectivity.
Received in revised form 17 July 2015 Materials and methods: Total 108 patients with active pulmonary tuberculosis were enrolled. We reviewed
Accepted 27 July 2015
CT findings and sputum microbiological studies. Then, we analyzed the statistical difference in CT findings
between the positive and negative groups of each sputum microbiological study (AFB smear, PCR, and
Keywords:
culture). Also, we divided the patients into five groups according to sputum AFB smear grade and analyzed
Tuberculosis
linear trends of CT findings between the five groups.
Lung
Multidetector computed tomography
Results: Both frequencies and extents of centrilobular micronodules (63% vs 38%, p = 0.011 for frequency;
Sputum 1.6 ± 1.6 vs 0.6 ± 1.1, p = 0.001 for extent), tree-in-bud opacities (63% vs 33%, p = 0.002; 1.6 ± 1.6 vs
Infectivity 0.5 ± 0.9, p < 0.001, respectively), consolidation (98% vs 81%, p = 0.003; 2.7 ± 1.5 vs 1.3 ± 1.1, p < 0.001,
respectively), and cavitation (86% vs 33%, p < 0.001; 1.5 ± 1.2 vs 0.4 ± 0.7, p < 0.001, respectively), were
significantly increased in the sputum AFB-positive group than in the negative group. These four CT find-
ings were increase in frequency and extent in the sputum PCR-positive group with or without statistical
significance. They did not show significant differences between the sputum culture-positive and nega-
tive groups. As the AFB smear grade increased, frequencies and extents of centrilobular micronodules,
tree-in-bud, consolidation, and cavitation also increased.
Conclusion: CT features representing active tuberculosis—centrilobular nodules, tree-in-bud, consolida-
tion, and, cavitation—strongly correlate with the positivity and grading of AFB smear.
© 2015 Elsevier Ireland Ltd. All rights reserved.

1. Introduction been a few studies about the relationship between CT findings and
the positivity/grading of sputum smears [1,5]. In these studies, the
In pulmonary tuberculosis (TB), CT provides valuable informa- presence of cavitation and consolidation correlated with the results
tion for the diagnosis and management of the disease. Furthermore, of sputum studies. However, the presence of micronodules did not.
CT can help distinguish active forms of the disease from its inac- This is unexpected because all micronodules (but not miliary) in
tive forms [1,2]. The most common CT findings of active pulmonary active TB have been considered to have centrilobular distribution
TB are centrilobular micronodules, tree-in-bud opacities, consoli- representing bronchogenic spread [2,6].
dation, and cavitation [2]. Especially, centrilobular micronodules, Recently, we reported that micronodules with perilymphatic
tree-in-bud opacities, and cavitation are considered active disease distribution are common in pulmonary TB which may repre-
processes [2–4]. Patients with sputum smear-positive TB are the sent lymphatic spread of TB [7]. We thought that all of the
most potent sources of infection. Moreover, the bacterial count micronodules visible on CTs in pulmonary TB may not be cen-
from the smear correlates with the degree of infectiousness of trilobular distribution reflecting bronchogenic spread, and that the
the patient as well as the severity of the disease [1]. There have presence of micronodules by exception for ones showing peri-
lymphatic and miliary distribution may correlate with results of
sputum studies. Accordingly, our study aimed to evaluate whether
∗ Corresponding author at: Department of Radiology, College of Medicine, St. CT findings of active TB, including the presence of centrilobu-
Vincent’s Hospital, The Catholic University of Korea, 93 Jungbu-daero, Paldal-gu, lar micronodules, correlate with sputum microbiological results
Suwon-si, Gyeonggi-do 442-723, Republic of Korea. Fax: +82 31 247 5713.
E-mail address: radiodoc@catholic.ac.kr (H.J. Park).

http://dx.doi.org/10.1016/j.ejrad.2015.07.032
0720-048X/© 2015 Elsevier Ireland Ltd. All rights reserved.
2340 J.M. Ko et al. / European Journal of Radiology 84 (2015) 2339–2344

and to determine whether CTs can predict the infectivity of Table 1


The clinical characteristics and prevalence of CT findings in patients.
pulmonary TB.
Patients’ characteristics

2. Materials and methods Number of patients 108


Age 41 ± 17 (16–86)
Sex (M:F) 74:34
This study was approved by The Catholic university of Korea Grade of sputum AFB smear 1.3 ± 1.4
St. Vincent’s Hospital Institutional Review Board (VC14RISI0100), AFB smear -positive sputum 63/105 (60%)
which waived informed consent. We selected 194 consecutive PCR-positive sputum 60/81 (74%)
patients with active pulmonary TB, who obtained CT scans at the Culture-positive sputum 74/89 (83%)
time of diagnosis from January 2010 to February 2013. We ana- CT findings
lyzed the patients’ medical charts, microbiology studies’ results, Micronodules 102 (94%)
and pulmonary parenchymal changes on the CT scans. Centrilobular 58 (54%)
Perilymphatic 92 (85%)
Conventional CT with or without an intravenous administration Peribronchovascular 89 (82%)
of contrast medium (100 mL at 2–2.5 mL/s) and high resolution CT Septal 66 (61%)
(HRCT) were obtained with two MDCT scanners. The decision to Subpleural 54 (50%)
perform contrast enhancement was made by the attending clini- Random 2 (2%)
Tree-in-bud 56 (52%)
cians based on their suspicion for malignancy or TB lymphadenitis
Consolidation 99 (92%)
from chest radiographs of the patients. Parameters for a LightSpeed Cavitation 71 (66%)
VCT (General Electric Medical Systems, Milwaukee, WI, USA) were Bronchovascular bundle thickening 100 (93%)
as follows: detector collimation, 64 × 0.625 mm; rotation time, Septal thickening 73 (68%)
500 ms; pitch 1.375; 120 kV tube voltage; automatic tube current
modulation. Parameters for a SOMATOM Definition Flash (Siemens
to extensive pleural TB, n = 1). We also excluded 37 patients who
Healthcare, Forchheim, Germany) were as follows: detector col-
did not undergo sputum studies and 20 patients had clinical and
limation, 128 × 0.6 mm; rotation time, 280 ms; pitch 1.5; 100 or
radiological findings suggestive of a diagnosis of pulmonary TB and
120 kV tube voltage; automatic tube current modulation. In the
showed a successful response to anti-tuberculous drugs. There-
conventional CT, all images were reconstructed into axial images
fore, we ultimately evaluated CT findings (conventional CT, n = 98;
with 5-mm slice thickness at 5-mm intervals, coronal images with
HRCT, n = 10) from 108 patients with pulmonary TB (74 men and 34
3-mm slice thickness and high-spatial-frequency algorithm. For
women; mean age, 41 years; age range, 16–86 years). Twenty-one
HRCT, axial and coronal images with 1-mm slice thickness at 5-mm
patients had a chronic illness, including diabetes (n = 9), chronic
intervals and high-spatial-frequency algorithm were obtained.
obstructive lung disease (n = 9), malignancy (n = 3), renal failure
CT scans obtained before the administration of anti-tuberculous
(n = 1), and collagen vascular disease (n = 1). Diagnoses of active
medication were analyzed for the presence and distribution of
pulmonary TB were confirmed by positive sputum microbiology
micronodules. The micronodules were defined as small rounded
results (acid-fast bacilli (AFB), n = 63; polymerase chain reaction
opacities with a diameter less than 7 mm and were classified by
(PCR), n = 60; culture, n = 74) in 96 patients. Twelve patients were
their distribution as either centrilobular, perilymphatic, or ran-
diagnosed by positive microbiology results from bronchial lavage
dom [8]. We also noted the presence or absence of consolidation,
(AFB, n = 4; PCR, n = 10; culture, n = 9). The clinical characteristics
cavitation, tree-in-bud, interlobular septal thickening, and bron-
and frequencies of various CT findings in patients are summarized
chovascular bundle thickening which are well-known, common CT
in Table 1. The overall frequencies of micronodules, consolidation,
features of active TB [4,9,10]. The extent of each CT finding was
cavitation, and tree-in-bud opacities which are well-known radi-
recorded as the number of involved lobes (the lingula was consid-
ologic feature of active pulmonary TB were 94%, 92%, 66%, and
ered as a separate lobe).
52%, respectively. Bronchovascular bundle thickening and inter-
All CT scans and medical records were retrospectively reviewed
lobular septal thickening were also commonly seen (93% and 68%,
by two chest radiologists with 4 and 11 years of experience. The
respectively). For the distribution of micronodules, perilymphatic
two radiologists independently and blindly analyzed the CT scans
distributing micronodules (85%) were most common followed by
for the presence or absence and extent of TB features. The inter-
centrilobular micronodules (54%). Only two patients had randomly
pretation of differences in the observed findings was based on a
distributing micronodules (Figs. 1, 2, and 3).
consensus between the two radiologists. We divided the patients
Both frequencies and extents of centrilobular micronodules
into two groups according to the positivity of each sputum studies
(63% vs 38%, p = 0.011 for frequency; 1.6 ± 1.6 vs 0.6 ± 1.1, p = 0.001
(AFB smear, PCR assay, and culture) and compared the CT find-
for extent), tree-in-bud opacities (63% vs 33%, p = 0.002; 1.6 ± 1.6
ings using the Pearson chi-square test or Fisher’s exact test for
vs 0.5 ± 0.9, p < 0.001, respectively), consolidation (98% vs 81%,
frequency and Mann–Whitney U test for extent. Also, we assigned
p = 0.003; 2.7 ± 1.5 vs 1.3 ± 1.1, p < 0.001, respectively), and cav-
the patients to five groups as per grading of the sputum AFB smear
itation (86% vs 33%, p < 0.001; 1.5 ± 1.2 vs 0.4 ± 0.7, p < 0.001,
(0, no AFB seen; 1+, 1–9 AFB/100 fields; 2+, 1–9 AFB/10 fields;
respectively), were significantly increased in the sputum AFB-
3+, 1–9 AFB/field; 4+, >9 AFB/field) according to the number of
positive group than in the negative group (Tables 2 and 3).
visible AFB on slide. Then, we compared CT findings using the chi-
Regarding PCR, consolidation (98% vs 71%, p = 0.001; 2.6 ± 1.6
square test for trend for frequency and Spearman’s rank correlation
vs 1.1 ± 1.0, p < 0.001, respectively), and cavitation (78% vs 33%,
for extent. P values less than 0.05 were considered statistically
p < 0.001; 1.4 ± 1.3 vs 0.4 ± 0.7, p < 0.001, respectively), were more
significant.
frequent and increased extent in the sputum PCR-positive group.
Extents of centrilobular micronodules were significantly increased
3. Results in sputum PCR-positive group (1.5 ± 1.6 vs 0.8 ± 1.2, p = 0.038).
However, frequencies of centrilobular micronodules (63% vs 38%,
Among 194 patients, 29 were excluded from this analysis due p = 0.072) and frequencies and extents of tree-in-bud opacities
to a co-existing pulmonary disease (pneumonia, n = 19; pulmonary (60% vs 38%, p = 0.083; 1.5 ± 1.6 vs 0.8 ± 1.2, p = 0.065, respectively)
edema, n = 4; pulmonary hemorrhage, n = 3; severe emphysema, were insignificantly greater in the sputum PCR-positive group. Fre-
n = 1; diffuse interstitial pneumonitis, n = 1; and lung collapse due quencies and extents of centrilobular micronodules, tree-in-bud
J.M. Ko et al. / European Journal of Radiology 84 (2015) 2339–2344 2341

Fig. 1. A 25-year-old woman with pulmonary TB. Chest CT scans (A–C) show cavitations, multiple consolidations, centrilobular micronodules, and tree-in-bud opacities.
Results of sputum microbiologic tests were 4+ for AFB smear, and positive for PCR assay and culture.

Fig. 2. A 16-year-old man with pulmonary and pleural TB (4+ for sputum AFB smear and positive for sputum PCR and culture). Multiple centrilobular micronodules and
tree-in-bud opacities are seen on chest CT scans (A–B). Note also peribronchial consolidations and a cavitation.

opacities, consolidation, and cavitation did not show significant dif- opacities (r = 0.467, p < 0.001), consolidation (r = 0.531, p < 0.001),
ferences between the sputum culture-positive and negative group. and cavitation (r = 0.585, p < 0.001, Table 5).
As the sputum AFB grade increased, the frequencies of
centrilobular micronodules, tree-in-bud opacities, consolidation, 4. Discussion
and cavitation also increased (p < 0.001, p < 0.001, p = 0.006, and
p < 0.001, respectively, Table 4 and Fig. 4). The sputum AFB grade Of all CT findings of pulmonary TB that we evaluated, cavitation,
also had a statistically significant and moderate positive correlation consolidation, centrilobular micronodules, and tree-in-bud opaci-
with centrilobular micronodules (r = 0.414, p < 0.001), tree-in-bud
2342 J.M. Ko et al. / European Journal of Radiology 84 (2015) 2339–2344

Fig. 3. A 29-year-old man with sputum AFB-negative TB. Chest CT scans (A–C) show extensive perilymphatic micronodules with peribronchiolar nodular thickening and
subpleural micronodules. Also note round, oval or nodular radiolucencies representing open bronchioles. Pulmonary TB was confirmed by positive result of PCR and culture
from bronchial lavage.

Table 2
The frequencies of common CT findings according to the positivity of microbiology tests.

Sputum AFB Sputum PCR Sputum culture

Negative (n = 42) Positive (n = 63) p Value Negative (n = 21) Positive (n = 60) p Value Negative (n = 15) Positive (n = 74) p Value

Micronodules 37 (88%) 62 (98%) 0.037 18 (86%) 58 (97%) 0.107 12 (80%) 73 (99%) 0.014
Centrilobular micronodules 16 (38%) 40 (63%) 0.011 8 (38%) 38 (63%) 0.072 7 (47%) 40 (54%) 0.601
Perilymphatic micronodules 35 (83%) 54 (86%) 0.739 16 (76%) 53 (88%) 0.282 11 (73%) 67 (91%) 0.085
Peribronchovascular 34 (81%) 52 (83%) 0.836 16 (76%) 51 (85%) 0.503 11 (73%) 64 (86%) 0.243
Septal 27 (64%) 37 (59%) 0.568 11 (52%) 34 (57%) 0.734 9 (60%) 46 (62%) 0.875
Subpleural 18 (43%) 34 (54%) 0.265 8 (38%) 31 (52%) 0.284 5 (33%) 41 (55%) 0.119
Tree-in-bud 14 (33%) 40 (63%) 0.002 8 (38%) 36 (60%) 0.083 6 (40%) 39 (49%) 0.370
Consolidation 34 (81%) 62 (98%) 0.003 15 (71%) 59 (98%) 0.001 14 (93%) 66 (89%) 1.000
Cavitation 14 (33%) 54 (86%) <0.001 7 (33%) 47 (78%) <0.001 10 (67%) 46 (62%) 0.742
Bronchovascular bundle thickening 37 (88%) 60 (95%) 0.262 19 (90%) 57 (95%) 0.600 13 (87%) 69 (92%) 0.336
Interlobular septal thickening 27 (64%) 44 (70%) 0.551 13 (62%) 41 (68%) 0.591 10 (67%) 51 (69%) 1.000

AFB = acid-fast bacilli, PCR = polymerase chain reaction.


The chi-square or Fisher’s exact tests were used.

Table 3
The extent (mean number of involved pulmonary lobes) of common CT findings according to the positivity of microbiology tests.

Sputum AFB Sputum PCR Sputum culture

Negative (n = 42) Positive (n = 63) p Value Negative (n = 21) Positive (n = 60) p Value Negative (n = 15) Positive (n = 74) p Value

Micronodules 2.4 ± 1.6 3.7 ± 1.8 0.001 1.9 ± 1.6 3.7 ± 1.8 <0.001 1.9 ± 1.8 3.6 ± 1.8 0.002
Centrilobular micronodules 0.6 ± 1.1 1.6 ± 1.6 0.001 0.8 ± 1.2 1.5 ± 1.6 0.038 0.8 ± 1.0 1.3 ± 1.6 0.388
Perilymphatic micronodules 2.1 ± 1.5 3.1 ± 2.0 0.019 1.8 ± 1.6 3.2 ± 2.0 0.006 1.8 ± 1.7 3.1 ± 1.9 0.014
Peribronchovascular 2.0 ± 1.5 2.9 ± 2.0 0.017 1.8 ± 1.6 3.1 ± 2.1 0.015 1.7 ± 1.6 2.9 ± 2.0 0.018
Septal 1.0 ± 0.9 1.1 ± 1.2 0.256 0.8 ± 1.0 1.0 ± 1.2 0.510 0.9 ± 1.0 1.1 ± 1.2 0.691
Subpleural 0.7 ± 1.0 0.9 ± 0.9 0.675 0.5 ± 0.7 0.9 ± 1.2 0.234 0.4 ± 0.6 1.0 ± 1.2 0.060
Tree-in-bud 0.5 ± 0.9 1.6 ± 1.6 <0.001 0.8 ± 1.2 1.5 ± 1.6 0.065 0.7 ± 1.0 1.2 ± 1.5 0.279
Consolidation 1.3 ± 1.1 2.7 ± 1.5 <0.001 1.1 ± 1.0 2.6 ± 1.6 <0.001 1.6 ± 1.1 2.3 ± 1.6 0.090
Cavitation 0.4 ± 0.7 1.5 ± 1.2 <0.001 0.4 ± 0.7 1.4 ± 1.3 <0.001 1.2 ± 1.3 1.0 ± 1.1 0.667
Bronchovascular bundle thickening 1.4 ± 1.0 2.2 ± 1.3 <0.001 1.3 ± 1.1 2.0 ± 1.2 0.003 1.5 ± 1.1 1.9 ± 1.3 0.194
Interlobular septal thickening 1.1 ± 1.0 1.0 ± 0.9 0.270 1.0 ± 0.9 1.3 ± 1.3 0.404 1.1 ± 1.0 1.4 ± 1.2 0.430

AFB = acid-fast bacilli, PCR = polymerase chain reaction.


The Mann–Whitney U tests were used.

Table 4
The frequencies of common CT findings according to the sputum AFB smear grade.

Sputum AFB smear grade p Value

0 (n = 42) 1+ (n = 22) 2+ (n = 18) 3+ (n = 9) 4+ (n = 14)

Micronodules 37 (88%) 22 (100%) 18 (100%) 9 (100%) 13 (93%) 0.230


Centrilobular micronodules 16 (38%) 11 (50%) 10 (56%) 8 (89%) 11 (79%) <0.001
Perilymphatic micronodules 35 (83%) 20 (91%) 16 (89%) 6 (67%) 12 (86%) 0.772
Peribronchovascular 34 (81%) 20 (91%) 15 (83%) 5 (56%) 12 (86%) 0.657
Septal 27 (64%) 15 (68%) 13 (72%) 3 (33%) 6 (43%) 0.092
Subpleural 18 (43%) 13 (59%) 12 (67%) 5 (56%) 4 (29%) 0.802
Tree-in-bud 14 (33%) 10 (45%) 10 (56%) 8 (89%) 12 (86%) <0.001
Consolidation 34 (81%) 21 (95%) 18 (100%) 9 (100%) 14 (100%) 0.006
Cavitation 14 (33%) 15 (68%) 17 (94%) 9 (100%) 13 (93%) <0.001
Bronchovascular bundle thickening 37 (88%) 21 (95%) 18 (100%) 8 (89%) 13 (93%) 0.477
Interlobular septal thickening 27 (64%) 15 (68%) 16 (89%) 5 (56%) 8 (57%) 0.884

AFB = acid-fast bacilli, N/A = not applicable.


The linear-by-linear association tests (chi-square test for trend) were used.
J.M. Ko et al. / European Journal of Radiology 84 (2015) 2339–2344 2343

Fig. 4. Bar graph showing the relative frequency of each CT finding in the five groups according to the sputum AFB smear grade. The asterisks indicate a significant linear
trend observed between groups (p < 0.05).

Table 5
The extent (mean number of involved pulmonary lobes) of common CT findings according to the sputum AFB smear grade.

Sputum AFB smear grade r Value p Value

0 (n = 42) 1+ (n = 22) 2+ (n = 18) 3+ (n = 9) 4+ (n = 14)

Micronodules 2.4 ± 1.6 3.3 ± 1.7 3.9 ± 1.9 3.3 ± 1.7 4.1 ± 2.0 0.346 <0.001
Centrilobular micronodules 0.6 ± 1.1 1.0 ± 1.1 1.4 ± 1.7 2.4 ± 1.6 2.4 ± 1.8 0.414 <0.001
Perilymphatic micronodules 2.1 ± 1.5 2.9 ± 1.7 3.6 ± 2.3 1.7 ± 1.6 3.7 ± 2.1 0.222 0.023
Peribronchovascular 2.0 ± 1.5 2.6 ± 1.7 3.5 ± 2.4 1.6 ± 1.7 3.7 ± 2.1 0.239 0.014
Septal 1.0 ± 0.9 1.2 ± 1.1 1.6 ± 1.5 0.4 ± 0.7 0.9 ± 1.2 −0.041 0.681
Subpleural 0.7 ± 1.0 1.2 ± 1.4 1.1 ± 1.1 0.7 ± 0.7 0.4 ± 0.8 0.007 0.941
Tree-in-bud 0.5 ± 0.9 0.9 ± 1.0 1.4 ± 1.7 2.3 ± 1.5 2.4 ± 1.7 0.467 <0.001
Consolidation 1.3 ± 1.1 2.1 ± 1.2 2.6 ± 1.4 3.3 ± 1.6 3.5 ± 1.7 0.531 <0.001
Cavitation 0.4 ± 0.7 1.1 ± 1.1 1.3 ± 0.7 2.1 ± 1.7 2.0 ± 1.4 0.585 <0.001
Bronchovascular bundle thickening 1.4 ± 1.0 1.8 ± 1.0 2.6 ± 1.4 1.6 ± 0.9 2.6 ± 1.4 0.381 <0.001
Interlobular septal thickening 1.1 ± 1.0 1.5 ± 1.5 1.7 ± 1.1 0.7 ± 0.7 1.2 ± 1.3 0.046 0.641

AFB = acid-fast bacilli, N/A = not applicable.


The Spearman’s rank correlation were used.

ties were most important CT features significantly associated with micronodules did not correlate with the number and positivity of
positive sputum microbiology results. The frequencies of only these AFB on sputum smear. In active pulmonary TB, all micronodules,
four findings correlated with the sputum AFB grade. Moreover, the except those of random distribution, are generally considered as
extent of only them had moderate positive correlation with the centrilobular micronodules reflecting a bronchogenic dissemina-
sputum AFB grade. It seems understandable and not surprising that tion [2,6]. Kosaka et al. thought this was caused by the smaller
only these four CT features are associated with the infectivity of pul- number of micronodules and the longer distance between the
monary TB, because consolidation and cavitation represent a local affected area and the central airway, rather than by consolidation
invasion of TB and centrilobular micronodules and tree-in-buds and cavitation.
reflect bronchogenic dissemination [4,9,11]. According to our results, the micronodules in cases of TB
Consolidation represents caseous pneumonia which consists did not show a uniform centrilobular distribution, and in fact,
of centrally located granulomas containing caseous necrosis and perilymphatic micronodules were more common than the cen-
marginal nonspecific inflammation [9]. A cavity is formed when a trilobular variety. After excluding the perilymphatic and randomly
tuberculous lesion erodes a bronchial tree and the caseous materi- distributed micronodules, the presence of centrilobular micron-
als soften, flowing out into airway. Most of the caseous material odules were well correlated to the results of sputum studies. The
is expectorated via mucociliary action and respiratory motion. presence of perilymphatic micronodules is not a novel CT feature
However, some may be aspirated into a new area of the lung in pulmonary TB. Poey et al. [13] also reported that interstitial
(bronchogenic dissemination) [11]. The onset of caseous necrosis (67%), peribronchovascular (63%) and subpleural nodules (41%)
is frequently associated with an explosive increase in numbers of were found on HRCT scans of patients with active pulmonary TB,
AFB [12]. and many of them disappeared with treatment. However, they did
There have been a few studies investigating the relationship not explain the meaning of these findings. Pathologically, granulo-
between sputum microbiology results and CT findings of pul- mas along the pulmonary interstitium are not unusual [14,15] and
monary TB. Matsuoka et al. [1] reported that as the number of AFB pulmonary lymphatic involvement of early TB has been identified
on sputum smear increases, consolidation and cavitation become in an animal study [16].
significantly more frequent. Kosaka et al. [5] also documented The major limitation of our study must be acknowledged. We
that consolidation and cavitation were significantly associated retrospectively analyzed CT images with 3–5-mm slice thickness,
with smear-positive pulmonary TB. In both studies, however, the which may cause underestimation of presence of tiny cavitations or
2344 J.M. Ko et al. / European Journal of Radiology 84 (2015) 2339–2344

micronodules. In our institute, conventional contrast-enhanced CT, [4] O.N. Hatipoğlu, E. Osma, M. Manisali, et al., High resolution computed
rather than HRCT, is performed routinely when evaluating micron- tomographic findings in pulmonary tuberculosis, Thorax 51 (4) (1996)
397–402.
odules, as the main reasons for the scan are to examine the presence [5] N. Kosaka, T. Sakai, H. Uematsu, et al., Specific high-resolution computed
of lymphadenitis or other associated findings, or to rule out dis- tomography findings associated with sputum smear-positive pulmonary
eases other than pulmonary TB, such as lung cancer. Nonetheless, tuberculosis, J. Comput. Assist. Tomogr. 29 (6) (2005) 801–804.
[6] A.N. Leung, Pulmonary tuberculosis the essentials, Radiology 210 (1999)
the resolution of MDCT has been improved, which facilitates the 307–322.
evaluation of subtle abnormalities, such as micronodules and sep- [7] J.M. Ko, H.J. Park, C.H. Kim, Clinicoradiologic evidence of pulmonary
tal lines, and if needed, multiplanar reconstruction with thin slice lymphatic spread in adult patients with tuberculosis, AJR Am. J. Roentgenol.
204 (1) (2015) 38–43.
thickness is feasible by applying appropriate software, Syngo.via
[8] D.M. Hansell, A.A. Bankier, H. MacMahon, T.C. McLoud, N.L. Muller, J. Remy,
(Siemens Healthcare, Erlangen, Germany). Fleischner society glossary of terms for thoracic imaging, Radiology 246 (3)
In conclusion, with exclusion of perilymphatic micronodules, all (2008) 697.
[9] J.G. Im, H. Itoh, Y.S. Shim, et al., Pulmonary tuberculosis: CT findings—early
CT features representing bronchogenic spread and local invasion in
active disease and sequential change with antituberculous therapy, Radiology
patients with active pulmonary TB were found to be associated with 186 (3) (1993) 653–660.
the results of sputum studies. Therefore, CT can be a useful tool not [10] K.S. Lee, J.W. Hwang, M.P. Chung, H. Kim, O.J. Kwon, Utility of CT in the
only for accurate diagnoses, but also in the prediction of infectivity evaluation of pulmonary tuberculosis in patients without AIDS, Chest 110 (4)
(1996) 977–984.
in patients with pulmonary TB. [11] P.C. Pratt, Pathology of tuberculosis, Semin. Roentgenol. 14 (3) (1979)
196–203.
Conflict of interest [12] G. Canetti, The Tubercle Bacillus in the Pulmonary Lesion of Man:
Histobacteriology and Its Bearing on the Therapy of Pulmonary Tuberculosis,
Springer Publishing Company, 1955.
The authors have no conflicts of interest. [13] C. Poey, F. Verhaegen, J. Giron, J. Lavayssiere, P. Fajadet, B. Duparc, High
resolution chest CT in tuberculosis: evolutive patterns and signs of activity, J.
Comput. Assist. Tomogr. 21 (4) (1997) 601–607.
References [14] T.C. Allen, Pathology of small airways disease, Arch. Pathol. Lab. Med. 134 (5)
(2010) 702–718.
[1] S. Matsuoka, K. Uchiyama, H. Shima, et al., Relationship between CT findings [15] O.Y. Cheung, J.R. Muhm, R.A. Helmers, et al., Surgical pathology of
of pulmonary tuberculosis and the number of acid-fast bacilli on sputum granulomatous interstitial pneumonia, Ann. Diagn. Pathol. 7 (2) (2003)
smears, Clin. Imaging 28 (2) (2004) 119–123. 127–138.
[2] Y.J. Jeong, K.S. Lee, Pulmonary tuberculosis: up-to-date imaging and [16] R.J. Basaraba, E.E. Smith, C.A. Shanley, I.M. Orme, Pulmonary lymphatics are
management, AJR Am. J. Roentgenol. 191 (3) (2008) 834–844. primary sites of Mycobacterium tuberculosis infection in guinea pigs infected
[3] L. Curvo-Semedo, L. Teixeira, F. Caseiro-Alves, Tuberculosis of the chest, Eur. J. by aerosol, Infect. Immun. 74 (9) (2006) 5397–5401.
Radiol. 55 (2) (2005) 158–172.