DOI:10.2214/AJR.07.2284
AJR 2008; 191:252-259
© American Roentgen Ray Society
Quantitative Investigation of Solitary Pulmonary Nodules: Dynamic Contrast-Enhanced MRI and Histopathologic Analysis
Yu Zou1,
Minming Zhang1,
Qidong Wang1,
Desheng Shang1,
Lijun Wang2 and
Guowei Yu3
1 Department of Radiology, First Affiliated Hospital, College of Medicine,
Zhejiang University, 79 Qinchun Rd., Hangzhou, Zhejiang Province 310003,
China.
2 Department of Pathology, First Affiliated Hospital, College of Medicine,
Zhejiang University, Hangzhou, China.
3 Department of Thoracic and Cardiovascular Surgery, First Affiliated Hospital,
College of Medicine, Zhejiang University, Hangzhou, China.
Received March 20, 2007;
accepted after revision January 17, 2008.
Address correspondence to M. Zhang
(zhangminming{at}163.com).
This study was supported by National Natural Science Foundation of China
grant number 30170284.
Abstract
OBJECTIVE. The purposes of this study were to analyze the relation
between enhancement patterns on dynamic enhanced MRI and histologic
microvessel patterns of solitary pulmonary nodules (SPNs) and to address the
topic of false-positive findings in differentiating SPNs with dynamic MRI.
SUBJECTS AND METHODS. Sixty-eight patients with 68 pathologically
proven SPNs (diameter 
30 mm) underwent dynamic 1.5-T MRI. On
time–signal intensity curves generated after bolus injection of contrast
material, steepest slope, peak height, and enhancement ratios of signal
intensity at the first, second, and fourth minutes were calculated. The
relation between dynamic MRI values and microvessel density was analyzed. The
morphologic differences between malignant SPNs and active inflammatory SPNs
also were analyzed. Threshold dynamic MRI values for differential diagnosis
were determined.
RESULTS. The dynamic MRI values of benign SPNs were significantly
lower than those of the other SPNs (p < 0.01). The enhancement
ratio at the fourth minute for active inflammatory SPNs was significantly
higher than that of malignant SPNs (p < 0.01). A high correlation
coefficient (r = 0.87, p < 0.001) was found between
steepest slope and microvessel density. With steepest slope 1.5%/s or less,
benign SPNs were clearly differentiated from other SPNs. With enhancement
ratio at the fourth minute 65% or less, malignant SPNs were differentiated
from active inflammatory SPNs with high sensitivity (93%) and high specificity
(100%).
CONCLUSION. Dynamic MRI values reflect the quantitative and
morphologic characteristics of microvessels in SPNs and are a useful tool for
differentiating SPNs with little overlap.
Keywords: angiogenesis • lung disease • lung neoplasm • lung nodule • MRI
Introduction
The solitary pulmonary nodule (SPN) is one of the most common findings on
chest radiographs, and accurate evaluation of SPNs is a diagnostic challenge
that has long perplexed clinicians. In clinical practice, it is important to
differentiate malignant nodules from benign nodules in the least invasive way
and to make a specific and accurate diagnosis. Investigators have used imaging
techniques to exploit the differences in vascularity, pharmacodynamics, and
metabolism between malignant and benign SPNs. Despite advances in the
assessment of SPNs with hemodynamic information from CT and MRI and with
biochemical 18F-FDG PET characteristics, the specific diagnoses of
a substantial portion of SPNs remain radiologically indeterminate because some
active inflammatory SPNs yield false-positive results
[1–15].
In this study, we analyzed the relations between the enhancement patterns of
SPNs and histologic microvascular patterns and addressed the topic of
false-positive findings on dynamic contrast-enhanced MRI.
Subjects and Methods
Patient Population
From April 2003 to July 2005, 68 patients (42 men, 26 women; mean age, 64.5
years; age range, 26–80 years) were consecutively enrolled in this
study. All patients had definite SPNs 10–30 mm in diameter detected on
both conventional radiographs and CT scans before MRI exami nations were
performed. Nodule diameter was defined as the largest diameter on conventional
radiographs or CT scans obtained with a lung window setting. On conventional
CT scans (5- to 10-mm section thickness), none of the nodules exhibited
calcifications or fat attenuation at a mediastinal window setting. All nodules
were surgically resected within 1 week after MRI. All final diagnoses were
confirmed with histopathologic examination after surgical resection.
MRI Protocol
The MRI examinations were performed with a 1.5-T superconductive system
(Signa, GE Healthcare) with a phased-array torso coil. Axial or coronal 8-mm
T1-weighted spin-echo images (TR/TE, 600–800/9) followed by axial or
coronal 8-mm T2-weighted fast spin-echo images (TR/effective TE,
2,500–4,000/83; echo-train length, 16; number of signals averaged, two)
without fat suppression were acquired for lesion localization. Both sequences
encompassed the whole thorax.
Dynamic MRI was performed with the following parameters: fast spin-echo
sequence; TR/effective TE, 600–800/9; section thickness, 4 mm with 1-mm
gap; echo-train length, 4; number of signals averaged, 1; matrix size, 256
x 128; field of view, 35–40 cm2; blurring cancellation
selection, on; ECG gating. Dynamic images were acquired in the axial plane.
When SPNs were located close to the diaphragm, dynamic images were acquired in
the coronal plane to minimize the influence of respiration. Gadopentetate dime
glumine (Magnevist, Bayer HealthCare) was administered by hand as a bolus
injection at a dosage of 0.1 mmol/kg body weight through a cubital vein at a
rate of 2 mL/s. Acquisition of the dynamic perfusion images was begun 10
seconds after the start of injection of gadopentetate dimeglumine. Sequential
multiphase images covering the entire SPN were continuously ac quired in the
axial or coronal plane for 4 minutes. Before undergoing MRI, all patients
received careful instruction in breathing technique to produce the precise
degree of inspiration for each imaging series. The total imaging time was
20–25 minutes.
Postprocessing of MR Images and Data Analysis
Time–signal intensity curves of SPNs were plotted from signal
intensity values in the lesions obtained by drawing of regions of interest.
The region of interest drawn over the tumor was as large as possible to
minimize noise but avert partial volume effect. According to this criterion,
the diameter range of the region of interest was 10–28 mm.
The patterns of time–signal intensity curves were classified into
three types (A, B, and C) on the basis of the results of phase analysis of the
peak enhancement and washout of gadopentetate dimeglumine. In a type A curve,
enhancement increased in the early phase (peak enhancement seen within 120
seconds) and a decrease or plateau occurred in the following phase. In a type
B curve, there was no peak enhancement; enhancement increased with time
throughout the examination. In a type C curve, no significant increase over
baseline was present: (SImax – SI0)
/SI0 < 10%, where SI is signal intensity, SImax is
the maximum signal intensity after bolus injection of contrast medium, and
SI0 is signal intensity before bolus injection of contrast medium.
The definition of curve types was reevaluated visually by two independent
observers, who were blinded to all results and used only the plotted
time–signal intensity curves (Fig.
1).
Dynamic contrast-enhanced MRI values were derived as follows. First, the
steepest slope (measured in percentage per second) of the time–signal
intensity curve was determined according to the following formula: steepest
slope = [(SIend – SIprior) x 100%]
/[SI0 x (Tend – Tprior)], where
SIend and SIprior are signal intensity values on the
contrast medium uptake curve that differ the most from image to image in the
dynamic series and Tend and Tprior are the time points
that correspond to SIend and SIprior
(Fig. 2). Second, peak height
was determined according to the following formula: peak height =
SImax – SI0. Third, the enhancement ratio of
signal intensity at the first, second, and fourth minutes after contrast admin
istration were determined. The enhancement ratio of signal intensity, a
percentage, was defined according to the equation: [(SIn
– SI0) x 100] /SI0, where n = 1, 2,
or 4.
Histologic and Microvessel Staining Analysis
As much as possible, pathologic specimens of nodules were obtained at
approximately the same location and orientation as the corresponding MR
images. Vascularization and immunohistochemical staining were performed on the
specimens with anti-CD31 antibody according to the avidin–biotin
peroxidase complex technique for evaluation. The vessel-counting method was
described by Weidner et al.
[16]. In brief, the area of
highest neo vascularization was first identified by imag ing of the tumor
sections at low power (x40 and x100). This area was then located
subjectively, and individual microvessels were counted on a 200-power field in
five to 10 areas with an ocular graticule (area, 0.25 mm2) to
enhance precision. The mean counts in these five to 10 areas were recorded.
Microvessel density (MVD) counts were then determined by one pathologist
without knowledge of MRI findings.
Statistical Analysis
According to the pathologic results, all the nodules were classified into
three groups: benign, malignant, and active inflammatory. All values were
expressed as mean ± SD. A chi-square test was performed to determine
differences in distribution of the curve types among the groups. The
significant difference of dynamic MRI values and MVD between groups was
analyzed with a two-sample Student's t test (if necessary, with
correction for unequal variance with a two-sample Student's t or
t' test). A Pearson correlation test was performed to determine
the strength of the relation between enhancement value and MVD. The software
package SPSS for Windows (version 10.0, SPSS) was used for the statistical
analysis. For all tests, p < 0.05 was considered to indicate a
statistically significant difference.
View larger version (11K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3 —Bar graph shows distribution of time–signal intensity
curve types for malignant, benign, and active inflammatory lesions. Light gray
indicates type A curve; dark gray, type B curve; white, type C curve.
|
|
Receiver operating characteristic (ROC) analyses were performed to test the
usefulness of dynamic MRI values for differentiation of benign from malignant
and active inflammatory lesions and for differentiation of malignant from
active inflammatory lesions. Areas under the curves were calculated.
Sensitivity, specificity, accuracy, and positive and negative predictive
values were calculated for each level by varying the thresholds. The MedCalc
software package (version 6.14, MedCalc) was used for ROC statistical
analysis.
Results
On the basis of the final pathologic diagnosis of the 68 SPNs, 40 nodules
were malignant, including 37 primary pulmonary carcinomas (17 adenocarcinomas,
15 squamous cell carcinomas, two small cell carcinomas, two large cell
carcinomas, and one bronchial carcinoid) and three metastatic lung tumors (one
rectal carcinoma and two primary hepatic cell carcinomas). Sixteen nodules
were benign (five hamartomas, nine tuberculomas, and two granulomas). Twelve
nodules were active inflammatory lesions (six, active tuberculosis; two,
cryptococcosis infection; two, aspergillosis; two, organizing pneumonia). The
tuberculomas were differentiated from active tuberculosis on the basis of lack
of evidence of the presence of Mycobacterium tuberculosis at
microbiologic examination. No significant differences were found in mean SPN
diameter among the groups.
Distribution of Time–Signal Intensity Curve Types
The time–signal intensity curve profiles of the benign, malignant,
and active inflammatory SPNs differed significantly
(Fig. 3). In the malignant
group, the predominant curve profile was type A (36 of 40 cases). Type B
curves were seen in three of 40 cases; a type C curve was seen in only one of
40 cases. In the benign group, the predominant curve profile was type C (13 of
16 cases). A type B curve was identified in only one of 16 cases and type A
curves in two of 16 cases. In the active inflammatory group, the predominant
curve profile was type B (10 of 12 cases). Type A curves were present in two
of 12 cases. None of the active inflammatory nodules had a type C curve. A
chi-square test showed a statistically significant difference in the
distribution of the curve types among the groups (p < 0.001).
Evaluation of Dynamic Images and Histopathologic Patterns
Dynamic MRI values (steepest slope, peak height, and enhancement ratios of
signal intensity at the first, second, and fourth minutes) and MVD of the SPNs
in each group are summarized in Table
1. The dynamic values of steepest slope; peak height; enhancement
ratios at the first, second, and fourth minutes, and MVD of the benign SPN
group were significantly smaller than those of the malignant SPN group
(p < 0.001) and the active inflammatory SPN group (p <
0.001). There were no significant differences in dynamic values or MVD between
the malignant and active inflammatory SPN groups (p > 0.05). The
active inflammatory SPNs had complete overlap with the malignant SPNs when the
most dynamic values were analyzed. Moreover, the active inflammatory SPNs had
even higher dynamic values than did the malignant SPNs. However, the
enhancement ratio at the fourth minute of the active inflammatory SPN group
was significantly higher than that of the malignant SPN group.
View larger version (141K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4C —48-year-old woman with adenocarcinoma. Photomicrograph shows
microvessel density with antibodies against CD31 immunostaining is 47.
Abundance of mostly immature tumor microvessels is evident. Vessels are
stained brown. (x200)
|
|
In detailed analysis of the patterns of microvessels, which stained brown,
a morphologic difference was observed between malignant and active
inflammatory nodules. In malignant nodules, an abundance of mostly immature
tumor microvessels were seen, whereas active inflammatory nodules had an
abundance of dilated mature capillary vessels (Figs.
4A,
4B,
4C and
5A,
5B,
5C).
View larger version (115K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5C —60-year-old woman with active tuberculosis. Photomicrograph
shows microvessel density with antibodies against CD31 immunostaining is 38.
Abundance of dilated capillary vessels is evident. Vessels are stained brown.
(x200)
|
|
Table 2 shows the results of
correlation studies between the various dynamic MRI values and the MVD of the
SPNs. Statistically significant correlations were found between steepest
slope; peak height; enhancement ratios at the first, second, and fourth
minutes; and MVD. In general, the highest correlation coefficient (r
= 0.87, p < 0.001) was found between steepest slope and MVD.
Diameter of SPNs was not significantly correlated with MVD (p >
0.05).
View this table:
[in this window]
[in a new window]
|
TABLE 2: Correlation Coefficients for Comparisons of Dynamic MR Imaging
Parameters or Diameter of SPNs with Vascularization of SPNs
|
|
ROI Analysis and Threshold Determination
To differentiate benign from malignant and active inflammatory lesions,
steepest slope was analyzed with ROC analysis. A threshold level of 1.5%/s
( 1.5%/s indicating a benign SPN) for steepest slope was found suitable.
With this threshold, the benign potential of SPNs was predicted with a
sensitivity of 81% (13 of 16 cases), specificity of 98% (51 of 52 cases),
positive predictive value of 93% (13 of 14 cases), negative predictive value
of 94% (51 of 54 cases), and accuracy of 94% (64 of 68 cases). The area under
the ROC curves was 0.88 (95% CI, 0.78–0.95) (Fig.
6A,
6B,
6C).
View larger version (7K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6A —Results of receiver operating characteristic analysis of
steepest slope. Scattergram shows criteria for selecting cutoff point of
steepest slope for greatest overall accuracy in differentiating benign from
malignant or active inflammatory solitary pulmonary nodules with receiver
operating curve analysis. When threshold value of 1.5%/s ( 1.5%/s
indicating benign nodules) was selected, sensitivity and specificity were 81%
(13 of 16 cases) and 98% (51 of 52 cases).
|
|
View larger version (9K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 6B —Results of receiver operating characteristic analysis of
steepest slope. Graph shows relation between sensitivity and specificity and
steepest slope. Diamonds indicate sensitivity; squares, specificity.
|
|
To differentiate malignant from active inflammatory SPNs, enhancement ratio
at the fourth minute was analyzed with ROC analysis. A threshold level of 65%
( 65% indicating a malignant SPN) for enhancement ratio at the fourth
minute was found suitable. With this threshold for differentiating malignant
SPNs from active inflammatory nodules, we obtained a sensitivity of 93% (37 of
40 cases), specificity of 100% (12 of 12 cases), positive predictive value of
100% (37 of 37 cases), negative predictive value of 80% (12 of 15 cases), and
accuracy of 94% (49 of 52 cases) (Table
3). The area under the ROC curve was 0.96 (95% CI,
0.86–0.99) (Fig. 7A,
7B,
7C).
View larger version (9K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 7A —Results of receiver operating characteristic analysis of
enhancement ratio of signal intensity at fourth minute. Scattergram shows
criterion for selecting cutoff point of enhancement ratios at the fourth
minute for greatest overall accuracy in differentiating malignant from active
inflammatory solitary pulmonary nodules by means of receiver operating curve
analysis. When threshold value of 65% ( 65% indicated malignant nodules)
was selected, sensitivity and specificity were 93% (37 of 40 cases) and 100%
(12 of 12 cases).
|
|
View larger version (9K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 7B —Results of receiver operating characteristic analysis of
enhancement ratio of signal intensity at fourth minute. Graph shows relation
between sensitivity and specificity and enhancement ratio of signal intensity
at fourth minute. Triangles indicate sensitivity; diamonds, specificity.
|
|
View larger version (10K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 7C —Results of receiver operating characteristic analysis of
enhancement ratio of signal intensity at fourth minute. Graph shows area under
receiver operating characteristic curve for enhancement ratio of signal
intensity at fourth minute.
|
|
Discussion
Our results show that the patterns of time–signal intensity curves of
SPNs reflect the quantitative and morphologic characteristics of microvessels
in lung nodules. Steepest slope was the main value associated with quantity of
microvessels. At a steepest slope of 1.5%/s or less, benign SPNs were clearly
differentiated from malignant and active inflammatory nodules. Enhancement
ratio at the fourth minute was an indicator of washout of contrast medium and
was associated with the morphologic features of microvessels. When enhancement
ratio at the fourth minute was 65% or less, malignant SPNs were differentiated
from active inflammatory SPNs with limited overlap.
It is well known that solid tumors proliferate to a size of a few
millimeters without neovascularization. Further expansion requires
angiogenesis
[17–19].
According to biologic and histopathologic findings
[1–12,
15,
19–23],
the blood supply and metabolism of malignant neoplasms are qualitatively and
quantitatively different from those of chronic infection and benign neoplasms.
Thus angiogenesis in malignant tumors plays an important role in the
enhancement of malignant tumors. Less contrast medium is delivered to
hypovascular benign SPNs than to hypervascular malignant SPNs. Encouraging
reports
[13–15,
21–23]
have described increased uptake of radionuclide and contrast material by
malignant SPNs. It has been suggested that imaging of tumor perfusion,
interstitial accumulation of contrast material, and cellular metabolic changes
is accurate in differentiation of benign from malignant SPNs
[8,
10,
13–15].
In radiopathologic, pharmacokinetic, and pathologic studies
[4–11,
19–24],
increased blood flow, perfusion, and capillary permeability have been observed
not only in malignant neoplasms but also in tissues with active inflammation,
although the pathophysiologic mechanisms and theories of these two diseases
may be different. The pharmacokinetics of contrast medium can be approximated
to a two-compartment model with intravascular and extravascular compartments
[25]. In the initial period,
delivery of contrast medium to the tissues largely depends on blood flow. As
time progresses, contrast medium passes from the intravascular space into the
extravascular space. Finally, contrast medium washes out from tissues. In our
study, the initial enhancement (i.e., high steepest slope value) observed in
both malignant SPNs and active inflammatory SPNs accounted for increased
vascularization in either malignant SPNs or active inflammatory SPNs.
Our finding that the mean enhancement ratio at the fourth minute of active
inflammatory nodules was significantly higher than that of malignant nodules
is particularly interesting. The enhancement ratio at the fourth minute
indicated the washout of contrast medium from the nodules. When looking at the
immunostained microvessels in greater morphologic detail with light
microscopy, we found that inflammatory nodules were filled with an abundance
of dilated capillary vessels but that malignant nodules contained mostly
immature tumor microvessels (Figs.
4A,
4B,
4C and
5A,
5B,
5C). Immature tumor
microvessels in malignant tumors are associated with increased capillary
permeability, whereas dilated capillary vessels in inflammatory nodules are
associated with increased blood flow and prolonged presence of contrast medium
in tissues. Therefore, washout of contrast medium may be faster for malignant
SPNs than that for active inflammatory SPNs. This theory may explain the
apparent rapid increase in enhancement of malignant nodules followed by a
decrease during the delayed phase (time–signal intensity curve type A)
and the continued increase in enhancement of inflammatory nodules throughout
the examination without a peak (curve type B). The time–signal intensity
curves of SPNs after contrast injection may display the distinct differences
in the vascularity and vasculature of inflammatory and malignant nodules.
These different curve profiles for characterizing the vascularity of SPNs
allow for visual evaluation.
Significant correlations between MVD and enhancement values on dynamic CT
and MRI have been reported
[10,
19,
24–27].
Our results confirmed those findings. Moreover, a very high correlation
coefficient (r = 0.87, p < 0.001) was found between MVD
and steepest slope, which indicated that steepest slope can be used for
accurate characterization of angiogenesis of SPNs and differentiation of
benign SPNs from malignant SPNs and active inflammatory SPNs. With a steepest
slope value of 1.5%/s or less as a threshold for indicating benign SPNs, we
found high sensitivity (81%), specificity (98%), positive predictive value
(93%), negative predictive value (94%), and accuracy (94%).
Differentiation between malignant and benign SPNs based on threshold values
for maximum enhancement and slope of enhancement has been performed in
previous CT and MRI investigations
[1–11,
24]. Those reports, however,
revealed that accurately differentiating malignant SPNs from benign SPNs was
difficult when those SPNs coexisted with active inflammation. The radiation
dose in dynamic CT is approximately four times higher than that of
conventional CT, and cutoff values are difficult to standardize
[28]. Methodologic
difficulties with artifacts and the spatial resolution of dynamic MRI and
standardization of cutoff values for differentiation have been limitations in
the past [2,
9]. Similar overlap has been
observed at dynamic MRI [10,
24]. In our study, high
specificity (98%) was reached in differentiating benign from malignant and
active inflammatory SPNs because we separated the active inflammatory from the
benign SPNs.
Both malignant and active inflammatory SPNs have increased blood flow,
perfusion, and capillary permeability in radiopathologic, pharmacokinetic, and
pathologic studies. Accurately differentiating malignant from active
inflammatory SPNs on the basis of these biologic and biochemical mechanisms
can be difficult. Even with FDG PET, overlap was seen in differential uptake
ratio between malignant SPNs and active histoplasma infection
[15]. In this study, we
observed an interesting behavior of time–signal intensity curves of
malignant versus active inflammatory nodules. Most of the malignant nodules
exhibited a marked washout phenomenon, but the active inflammatory nodules did
not manifest this phenomenon. With enhancement ratio at the fourth minute of
65% or less as a threshold for malignant SPN, sensitivity, specificity,
positive predictive value, negative predictive value, and accuracy were 93%,
100%, 100%, 80%, and 94%. In two previous studies on washout of lung nodules
and dynamic MRI, Schaefer et al.
[10,
24] set a threshold value
greater than 0.1 SI%/s as an indicator of washout. Those authors found a
specificity of 100%, and no benign nodule reached that level of washout. Our
study revealed specificity equivalent to that reported by Schaefer et al.
[24] but higher sensitivity
(93% vs. 52%). Thus our approach enables management of overlap between
malignant and active inflammatory SPNs and is a novel system for
differentiating malignant from active inflammatory SPNs with high sensitivity
and specificity.
All of the SPNs in this study were diagnosed with pathologic analysis.
Therefore, there was no diagnostic dilemma. Adopting a T1-weighted fast
spin-echo sequence as a dynamic MRI protocol has been reported
[29,
30]. With this protocol,
breath-holding was unnecessary during the dynamic MRI examination. This factor
was important in practice because many patients cannot tolerate relatively
long breath-holding during MRI. Thus our protocol allows satisfactory
measurement of the enhancement of SPNs and provides quantitative information
about tissue perfusion of SPNs. This technique is simple and can be performed
with conventional MRI.
There were limitations to our study. First, even higher temporal resolution
is desired for dynamic MRI because the pulmonary circulation is 4.0–5.0
seconds and the pulmonary capillary circulation is 0.7 second in adults
[11,
31]. Second, observer bias
might have occurred owing to the evaluation of MVD by only one pathologist.
Third, the subgroup populations were relatively small.
The development of new vessels within tumors determines their appearance on
dynamic MRI, which can be used for accurate differentiation of benign from
malignant SPNs. More important, active inflammatory SPNs can be successfully
discriminated from malignant SPNs with minimal overlap, which has been
considered difficult to achieve in previous studies.
References
- Kusumoto M, Kono M, Adachi S, et al. Gadopentetate
dimeglumine–enhanced magnetic resonance imaging for lung nodules:
differentiation of lung cancer and tuberculoma. Invest
Radiol 1994; 29:S255
–S256[Medline]
- Hittmair K, Eckersberger F, Klepetko W, Helbich T, Herold CJ.
Evaluation of solitary pulmonary nodules with dynamic contrast-enhanced MR
imaging: a promising technique. Magn Reson Imaging1995; 13:923
–933[CrossRef][Medline]
- Swensen SJ, Brown LR, Colby TV, et al. Lung nodule enhancement at
CT: prospective findings. Radiology 1996;201
: 447–455[Abstract/Free Full Text]
- Zhang M, Kono M. Solitary pulmonary nodules: evaluation of blood
flow patterns with dynamic CT. Radiology1997; 205:471
–478[Abstract/Free Full Text]
- Swensen SJ, Viggiano RW, Midthun DE, et al. Lung nodule enhancement
at CT: multicenter study. Radiology 2000;214
: 73–80[Abstract/Free Full Text]
- Yamashita K, Matsnobe S, Tsuda T, et al. Solitary pulmonary
nodules: preliminary study of evaluation with incremental dynamic CT.
Radiology 1995;194
: 399–405[Abstract/Free Full Text]
- Midthun DE. Solitary pulmonary nodule: time to think small.
Curr Opin Pulm Med 2000;6
: 364–370[CrossRef][Medline]
- Tateishi U, Nishihara H, Watanabe S, et al. Tumor angiogenesis and
dynamic CT in lung adenocarcinoma: radiologic-pathologic correlation.
J Comput Assist Tomogr 2001;25
: 23–27[CrossRef][Medline]
- Guckel C, Schnabel K, Deimling M, et al. Solitary pulmonary
nodules: MR evaluation of enhancement patterns with contrast-enhanced dynamic
snapshot gradient-echo imaging. Radiology1996; 200:681
–686[Abstract/Free Full Text]
- Schaefer JF, Schneider V, Vollmar J, et al. Solitary pulmonary
nodules: association between signal characteristics in dynamic contrast
enhanced MR imaging and tumor angiogenesis. Lung
Cancer 2006; 53:39
–49[CrossRef][Medline]
- Ohno Y, Hatabu H, Takenaka D, et al. Solitary pulmonary nodules:
potential role of dynamic MR imaging in management initial experience.
Radiology 2002;224
: 503–511[Abstract/Free Full Text]
- Gould M, Maclean C, Kuschner W, et al. Accuracy of positron
emission tomography for diagnosis of pulmonary nodules and mass lesions: a
meta-analysis. JAMA 2001;285
: 914–924[Abstract/Free Full Text]
- Herder GJ, Golding RP, Hoekstra OS, et al. The performance of
(18)F-fluorodeoxyglucose positron emission tomography in small solitary
pulmonary nodules. Eur J Nucl Med Mol Imaging2004; 31:1231
–1236[Medline]
- Nomori H, Watanabe K, Ohtsuka T, et al. Evaluation of F-18
fluorodeoxyglucose (FDG) PET scanning for pulmonary nodules less than 3 cm in
diameter, with special reference to the CT images. Lung
Cancer 2004; 45:19
–27[CrossRef][Medline]
- Christensen JA, Nathan MA, Mullan BP, et al. Characterization of
the solitary pulmonary nodule: 18F-FDG PET versus
nodule-enhancement CT. AJR 2006;187
:1361
–1367[Abstract/Free Full Text]
- Weidner N, Semple J, Welch W, Folkman J. Tumor angiogenesis and
metastasis: correlation in invasive breast carcinoma. N Engl J
Med 1991; 324:1
–8[Abstract]
- Folkman J. What is the evidence that tumors are angiogenesis
dependent? J Natl Cancer Inst 1990;82
: 4–6[Free Full Text]
- Ohta Y, Tomita Y, Oda M, Watanabe S, Murakami S, Watanabe Y. Tumor
angiogenesis and recurrence in stage I non-small cell lung cancer.
Ann Thorac Surg 1999;68
:1034
–1038[Abstract/Free Full Text]
- Yamazaki K, Abe S, Takekawa H, et al. Tumor angiogenesis in human
lung adenocarcinoma. Cancer 1994;74
:2245
–2250[CrossRef][Medline]
- Gupta NC, Aloof J, Gunnel E. Probability of malignancy in solitary
pulmonary nodules using fluorine-18-FDG and PET. J Nucl
Med 1996; 37:943
–948[Abstract/Free Full Text]
- Worsley DF, Seller A, Adam MJ, et al. Pulmonary nodules:
differential diagnosis using 18F-fluorodeoxyglucose single-photon emission
computed tomography. AJR 1997;168
: 771–774[Abstract/Free Full Text]
- Stumps KD, Dazzi H, Schaffner A, von Schulthess GK. Infection
imaging using whole-body FDG-PET. Eur J Nucl Med2000; 27:822
–832[CrossRef][Medline]
- Goo JM, Im JG, Do KH, et al. Pulmonary tuberculoma evaluated by
means of FDG PET: findings in 10 cases. Radiology2000; 216:117
–121[Abstract/Free Full Text]
- Schaefer J, Vollmar J, Schick F, et al. Solitary pulmonary nodules:
dynamic contrast enhanced MR imaging-perfusion differences in malignant and
benign lesions. Radiology 2004;232
: 544–553[Abstract/Free Full Text]
- Hawighorst H, Weikel W, Knapstein P, et al. Angiogenic activity of
cervical carcinoma: assessment by functional magnetic resonance imagingbased
parameters and a histomorphological approach in correlation with disease
outcome. Clin Cancer Res 1998;4
: 2305–2312[Abstract/Free Full Text]
- Buadu L, Murakami J, Murayama S, et al. Patterns of peripheral
enhancement in breast masses: correlation of findings on contrast medium
enhanced MR imaging with histologic features and tumor angiogenesis.
J Comput Assist Tomogr 1997;21
: 421–430[CrossRef][Medline]
- Tateishi U, Kusumoto M, Nishihara H, Nagashima K, Morikawa T,
Moriyama N. Contrast-enhanced dynamic computed tomography for the evaluation
of tumor angiogenesis in patients with lung carcinoma.
Cancer 2002; 95:835
–842[CrossRef][Medline]
- Yi CA, Lee KS, Kim EA, et al. Solitary pulmonary nodules: dynamic
enhanced multi-detector row CT study and comparison with vascular endothelial
growth factor and microvessel density. Radiology2004; 233:191
–199[Abstract/Free Full Text]
- Mayr NA, Yuh WT, Arnholt JC, et al. Pixel analysis of MR perfusion
imaging in predicting radiation therapy outcome in cervical cancer.
J Magn Reson Imaging 2000;12
:1027
–1033[CrossRef][Medline]
- Mayr NA, Yuh WT, Zheng J, et al. Prediction of tumor control in
patients with cervical cancer: analysis of combined volume and dynamic
enhancement pattern by MR imaging. AJR1998; 170:177
–182[Abstract/Free Full Text]
- Goldsmith SJ, Kostakoglu L. Role of nuclear medicine in the
evaluation of the solitary pulmonary nodule. Semin Ultrasound CT
MR 2000; 21:129
–138[CrossRef][Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
Abstract
Introduction
