Appearances of mediastinal and pulmonary hilar lymph nodes on thin-ction CT:comparison with 5mm slice thickness CT
hotcha
Atsushi Nambu a,4,Satoshi Kato b ,Akitoshi Saito c ,Tsutomu Araki a
a
Department of Radiology,University of Yamanashi,Tamaho-cho,Nakakoma-gun,Yamanashi prefecture,Japan b
Department of Radiology,Kofu Municipal Hospital,Masutsubo-cho,Kofu City,Yamanashi prefecture,Japan cadvance
Department of Internal Medicine,Kofu Municipal Hospital,Masutsubo-cho,Kofu City,Yamanashi prefecture,Japan
Received 1May 2006;accepted 22April 2007
Abstract
Aim:This study aimed to evaluate the efficacy of thin-ction CT of the mediastinum in the asssment of thoracic lymph nodes in comparison with conventional CT.Materials and Methods:A total of 193CT examinations from 193patients with suspected pulmonary dia were reconstructed into thin-ction CT and conventional CT.The appearances of the lymph nodes were assd and compared between thin-ction CT and conventional CT.Results:Intranodal fat was more often detected on thin-ction CT than on conventional CT (P b .001).There were no statistically significant differences in the frequencies of inhomogeneous enhancement and bulging margin of the hilar lymph node.Conclusion:Thin-ction CT can improve clinical N-staging of lung cancer due to classification of enlarged mediastinal lymph nodes as benign bad on identification of intranodal fat.D 2007Elvier Inc.All rights rerved.
Keywords:Lung cancer;CT;Mediastinum;N-staging
1.Introduction
rfq
The evaluation of hilar and mediastinal lymph nodes (N-staging)is very important in the management of lung cancer as it can affect the choice of treatment;surgery is usually not appropriate for dias in their advanced stages,but it should be performed in tho limited in scope.CT has been ud for t
his purpo for a long time.However,its nsitivity and specificity for N-staging were unsatisfacto-rily low,ranging from 33%to 77%and from 53to 96%,respectively [1–6].This is largely due to the limitation of the size criteria ud for the evaluation of hilar and mediastinal lymph nodes.
Although positron emission tomography (PET)is emerging as a new modality of choice in N-staging of
lung cancer with a higher accuracy than that of CT and in detecting distant metastasis [1,3–6],it is known that PET shows a significant number of fal-positive results in active nonneoplastic lesions and fal-negative results in smaller lesions less than 10mm in diameter [4,7].In addition,PET is still not an easily available modality in many institutions dealing with lung cancer.Furthermore,information about anatomical details,which is indispen-sable for surgery or radiotherapy planning,is not obtained by PET.Therefore,chest CT cannot be omitted even if PET is performed.
Recent development of multidetector row CT has enabled us to obtain thin-ction CT of the mediastinum as well as that of the lung by reconstructing the raw data without extra radiation exposure to the patients.We thought that detailed obrvation of lymph nodes using thin-ction CT
might provide additional information regarding their nature,which may improve accuracy of the N-staging.In this,study,we tried to find out the merits of thin-ction CT of the mediastinum in the evaluation of the hilar and
0899-7071/07/$–e front matter D 2007Elvier Inc.All rights rerved.doi:10.1016/j.clinimag.2007.05.001
4Corresponding author.Department of Radiology,University of Yamanashi,Shimokawato 1110,Tamaho-cho,Chuo-shi,Yamanashi pre-fecture 409-3898,Japan.Tel.:+81552731111;fax:+81552736744.
E-mail address:nambu-a@jp (A.Nambu).Clinical Imaging 31(2007)375
–378abigale
mediastinal lymph nodes in comparison with conventional 5mm slice thickness CT.2.Materials and methods
Becau the information for this retrospective study was obtained from routine clinical data,including reconstruc-tion,the approval of our institutional review board was not required.Patient informed connt was also not required.A total of 193chest CT examinations from 193patients between September 2004and February 2005were enrolled into this study.There were 71unenhanced and 122enhanced CT scans.Forty-four patients have lung carcino-
mas,78have malignancies other than lung carcinoma,and 71have benign dias.CT was carried out with Aquilion (Toshiba Medical Systems),a 16-detector row helical scanner in all cas.The imaging parameters were as follows:1mm row;current,200–300mA;voltage,120–140kvP;pitch,15.Thin-ction CT was reconstructed from the raw data,with 1mm thickness and a small FOV targeting the mediastinum and hila from the level of the upper margin of the aortic arch through the pulmonary hila,as well as 5mm slice thickness CT of the whole chest with a larger FOV including the thorax,which was routine chest CT in our institution.Two chest radiologists (A.N.and S.K.),who were not aware of the diagnos of the patients,independently evaluated the prence of mediastinal and hilar lymph nodes,intranodal fat and calcification,inhomo-geneous enhancement,an
d convex margin of hilar lymph nodes on thin-ction and 5-mm CT.CT images were evaluated on high-resolution monitors for routine CT reading system.All nodular soft tissue densities other than vasculature in the mediastinal fat tissue were regarded as lymph nodes except for the region of thymus,where residual thymic tissue is another possibility.Intranodal fat was defined as a focal fat density,which was assd visually,within the lymph nodes.When the readers were not confident that the intranodal low-density area was compod of fat,we measured its CT value on the monitor using a tool of pixel CT value measurement function equipped into the system.We could also magnify the images of 5mm CT with a larger FOV in the evaluation of the findings if needed.
Table 1
Mean size and number of the lymph nodes,and frequency of each finding in thin-ction and 5mm slice CT
Thin-ction (1mm)CT (n =193)
5mm slice (n =193)P Number of lymph nodes
15.7F 6.111.0F 5.6b .001Diameter (in mm)of the short axis of the largest lymph node 8.9F 4.97.9F 4.8b .001Bulging margin of the hilar lymph node,n (%)vegetarian
25(13)19(10).109Inhomogeneous enhancement a ,n (%)
22(18)16(8).07Intranodal fat,n (%)
101(52)
56(29)
b .001
a
The number of the cas evaluated for inhomogeneous enhancement was 122with enhanced
CT.
Fig.1.A 63-year-old man with malignant lymphoma.Multiple lymph nodes are en at the paraaortic region and around the trachea on both (A)5-mm CT and (B)thin-ction CT.However,the number of the lymph nodes is larger on thin-ction CT than on 5-mm CT.Note that the margin of each lymph node is more distinct on thin-ction CT.
A.Nambu et al./Clinical Imaging 31(2007)375–378
376
Final decision with regard to the findings was made by connsus.Kappa value between the readers was calculated for each finding.The frequency of each finding was statistically compared between thin-ction and 5-mm CT using McNemar’s test.In addition,the number of the lymph nodes was counted and the short-axis diameter of the largest lymph node was measured by a single radiologist (A.N.).Size measurement was performed using a distance measure-ment tool equipped to the high-resolution monitor.The differences were statistically compared using paired t test.A P value b .05was considered to be significant.
3.Results
Kappa values between the two readers were .50for bulging margin of the hilar lymph node,.48for inhomoge-neous enhancement,and .44for intranodal fat,indicating moderate interreader agreement (Table 1).
At least one lymph node was detected in all cas on both thin-ction and 5-mm CT.The number of the detected lymph nodes was significantly larger on thin-ction CT than on 5-mm CT (Fig.1).
The short-axis diameter of the largest lymph node tended to be greater in thin-ction CT than in 5-mm CT.Lymph nodes with a short-axis diameter of 10mm or more on thin-ction CT were en in 55cas.There were no significant differences in frequencies of bulging margin of the hilar
lymph node and inhomogeneous enhancement.Intranodal fat was much more frequently detected on thin-ction CT than on 5-mm CT,with a significant statistical difference.Of the 55cas with lymph nodes that have a short-axis diameter of 10mm or more,a fat-containing lymph node was en on 5-mm CT in 9cas and on thin-ction CT in 12cas (Fig.2).
4.Discussion
Although it has been shown that PET is superior to CT in the evaluation of the hilar and mediastinal l
四级英语听力下载
ymph nodes [1,3–7],CT is still an indispensable modality for preop-erative asssment of lung cancer as it provides detailed anatomical information about the lymph nodes as well as the primary tumor.The lower accuracy of CT in N-staging is mainly due to its simple size criterion disregarding the nature of the lymph nodes;a lymph node with a short-axis diameter of 10mm or more is considered to be metastasis.Several investigators attempted to add new criteria to N-staging on CT [8,9]to date.Ratto et al.[8]documented that central necrosis or discontinued capsule in addition to a shorter axis (1cm or larger)improved specificity in N-staging of lung cancer.Shimoyama et al.[9]found that a convex margin to the surrounding lung parenchyma of hilar lymph nodes indicates metastasis.It has been also generally known that the prence of intranodal fat,which appears as an age-related change,suggests benignancy of
托福必备词汇a
dairy
Fig.2.A 72-year-old man with squamous cell carcinoma of the lung.There is an enlarged lymph node that is 10mm in short-axis diameter at the pretracheal region on both (A)5-mm CT and (B)thin-ction CT (arrow).Thin-ction CT definitely shows an internal low-density area,with pixel CT values ranging from À19to 12,indicating the prence of fat tissue.Five millimeters CT fails to demonstrate the low density.This lymph node does not change in size on a follow-up CT examination (not shown)after 6months,suggesting the benign nature of the lymph node.A round-shaped lymph node that is 7mm in short-axis diameter without internal fat density is also en at the subaortic region,which increas in size on the follow-up CT.
A.Nambu et al./Clinical Imaging 31(2007)375–378377
lymph node as it indicates prervation of normal lymph node architectures.
Our results demonstrate that thin-ction CT could detect intranodal fat more nsitively than5-mm CT.Intranodal fat was en in101of193cas(52%)on thin-ction CT and in56of193cas(29%)on5-mm CT.Conquently, intranodal fat was detected only on thin-ction CT in45of 101cas(45%).We think that this result is due to less partial volume effect on thin-ction CT.Therefore,thin-ction CT
may decrea fal-positive lymph nodes and, conquently,might improve the specificity in N-staging of lung cancer.Moreover,intranodal fat may fally assume a water density as en in areas of necrosis on5-mm CT due to more partial volume effect even with CT value ,the low-density area may not show a negative CT value).This could be problematic if central necrosis is ud as a criterion in N-staging of lung cancer.Thin-ction CT would exclude this misleading appearance.
Although lymph nodes less than10mm in short-axis diameter are judged to be negative by the size criterion disregarding the prence of intranodal fat,it is not uncommon to e intranodal fat even in lymph nodes 10mm or more in short-axis diameter.Such lymph nodes were en in12of193cas(6%)on thin-ction CT and in 9of193cas(5%)on5-mm CT in our study.Although there were only three cas with a fat-containing lymph node that was detected on thin-ction CT but not on5-mm CT in our study,it could happen becau the area of intranodal fat tissue is not related to the size of the lymph node.
The number of lymph nodes was significantly larger on thin-ction CT than on5-mm CT.We think that this result is due to the following two reasons.First,tiny lymph nodes appear on thin-ction CT,which are not detected on5-mm CT.Second,abutting lymph nodes,which fally appear as one ly
obsolescence
mph node on5-mm CT,are parately defined on thin-ction CT becau of incread spatial resolution.However,the number of lymph nodes does not affect N-staging.
The size of lymph nodes was also measured and was found to be larger on thin-ction CT than on5mm CT.The difference of size measurements is probably due to a propensity of the measurer and is within the range of measurement deviation.However,we think that size measurement would be more accurate on thin-ction CT becau the margins of lymph nodes are more distinct on thin-ction CT due to less partial volume effect.
There was no significant difference in the frequency of inhomogeneous enhancement,calcification,and convexity margin of hilar lymph nodes.We may say that thin-ction CT does not have an advantage in detecting the findings. However,it should be noted that thin-ction CT more frequently detected the findings than5-mm CT and that a larger sample size may reveal statistically significant differences.In addition,our5-mm CT was reconstructed from thinner row raw data in our study and inherent5mm slice thickness CT may have wor images.
Bad on the results,we recommend thin-ction reconstruction CT of the mediastinum when enlarged lymph nodes10mm or more in short-axis diameter are en on routine CT.In addition,if cent
ral necrosis is ud as a criterion in N-staging,thin-ction reconstruction CT of the mediastinum should be performed in all cas to exclude fal positives by small amount of fat tissue.
This study has limitations.First,although our focus was N-staging in lung cancer,our subjects included many patients with dias other than lung cancer.However, we think that the detectability of each finding is not affected by the type of the evaluated dia,and therefore,our results could be applied to the patients with lung cancer. Second,our study did not have a pathological gold standard for benignancy or malignancy of the lymph nodes.
In conclusion,when using intranodal fat as an indicator of benignity for mediastinal lymph nodes,thin-ction CT can improve specificity in clinical N-staging. Acknowledgments
We are grateful to Saburo Akiyama,Yoshihito Aikawa, and Yoshihito Tateda,radiology technologists assigned in CT examination(from the University Hospital of Yamanashi),for their cooperation.
Referencesaggressive的意思
[1]Steinert HC,Haur M,Allemann F,et al.Non-small cell lung cancer:
nodal staging with FDG PET versus CT with correlative lymph node mapping and sampling.Radiolog
y1997;202:441–6.
[2]Kamiyoshihara M,Kawashima O,Ishikawa S,Morishita Y.Mediastinal
lymph node evaluation by computed tomographic scan in lung cancer.
J Cardiovasc Surg(Torino)2001;42:119–24.
[3]Gdeedo A,Van Schil P,Corthouts B,Van Mieghem F,Van
Meerbeeck J,Van Marck E.Comparison of imaging TNM [(i)TNM]and pathological TNM[pTNM]in staging of bronchogenic carcinoma.Eur J Cardiothorac Surg1997;12:224–7.
[4]Konishi J,Yamazaki K,Tsukamoto E,et al.Mediastinal lymph node
staging by FDG-PET in patients with non-small cell lung cancer: analysis of fal-positive FDG-PET findings.Respiration2003;70: 500–6.
[5]Halter G,Buck AK,Schirrmeister H,et al.Lymph node staging in
lung cancer using[18F]FDG-PET.Thorac Cardiovasc Surg2004;
52(2);96–101.
[6]von Haag DW,Follette DM,Roberts PF,Shelton D,Segel LD,Taylor
TM.Advantages of positron emission tomography over computed tomography in mediastinal staging of non-small cell lung cancer.J Surg Res2002;103:160–4.
[7]Fischer BM,Mortenn J.The future in diagnosis and staging of lung
cancer:positron emission tomography.Respiration2006;73:267–76.
[8]Ratto GB,Frola C,Cantoni S,Motta G.Improving clinical efficacy of
computed tomographic scan in the preoperative asssment of patients with non-small cell lung cancer.J Thorac Cardiovasc Surg 1990;99:416–25.
[9]Shimoyama K,Murata K,Takahashi M,Morita R.Pulmonary hilar
lymph node metastas from lung cancer:evaluation bad on morphology at thin-ction,incremental,dynamic CT.Radiology 1997;203:187–95.
A.Nambu et al./Clinical Imaging31(2007)375–378 378
