The size of mediastinal lymph nodes and its relation with metastatic involvement: a meta-analysis.
ABSTRACT Positron emission tomography with 18-fluorodeoxyglucose (FDG-PET) seems to be superior to computed tomography (CT) in staging the mediastinum in patients with non-small-cell lung cancer (NSCLC). However, recent results suggest that FDG-PET performance characteristics are conditional for nodal size as shown by CT: FDG-PET is more sensitive but less specific with lymph node enlargement on CT. The association between size and the probability of malignancy needs to be known to predict the post-test probabilities after PET, and finally, stratify patients for mediastinoscopy or thoracotomy depending on the PET and CT results. Therefore, we performed a meta-analysis of available studies reporting on the prevalence of metastatic involvement for different size categories of enlarged lymph nodes in patients with NSCLC and were able to include 14 studies. The prevalence of metastatic involvement and conditional test performance of CT and FDG-PET were calculated for lymph nodes measuring 10-15 mm, 16-20 mm and >20 mm. We found a post-test probability for N2 disease of 5% for lymph nodes measuring 10-15 mm on CT in patients with a negative FDG-PET result, suggesting that these patients should be planned for thoracotomy because the yield of mediastinoscopy will be extremely low. For patients with lymph nodes measuring > or =16 mm on CT and a negative FDG-PET result a post-test probability for N2 disease of 21% was found, suggesting that these patients should be planned for mediastinoscopy prior to possible thoracotomy to prevent too many unnecessary thoracotomies in this subset.
- SourceAvailable from: mpoullis.com[show abstract] [hide abstract]
ABSTRACT: To determine the test performance characteristics of CT scanning, positron emission tomography (PET) scanning, MRI, and endoscopic ultrasound (EUS) for staging the mediastinum, and to evaluate the accuracy of the clinical evaluation (ie, symptoms, physical findings, or routine blood test results) for predicting metastatic disease in patients in whom non-small cell lung cancer or small cell lung cancer is diagnosed. Systematic searches of MEDLINE, HealthStar, and Cochrane Library databases to July 2001, and of print bibliographies. Studies evaluating the staging results of CT scanning, PET scanning, MRI, or EUS, with either tissue histologic confirmation or long-term clinical follow-up, were included. The performance of the clinical evaluation was compared against the results of brain and abdominal CT scans and radionuclide bone scans. Pooled sensitivities and specificities for staging the mediastinum were as follows: for CT scanning: sensitivity, 0.57 (95% confidence interval [CI], 0.49 to 0.66); specificity, 0.82 (95% CI, 0.77 to 0.86); for PET scanning: sensitivity, 0.84 (95% CI, 0.78 to 0.89); specificity, 0.89 (95% CI, 0.83 to 0.93); and for EUS: sensitivity, 0.78 (95% CI, 0.61 to 0.89); specificity, 0.71 (95% CI, 0.56 to 0.82). For the evaluation of brain metastases, the summary estimate of the negative predictive value (NPV) of the clinical neurologic evaluation was 0.94 (95% CI, 0.91 to 0.96). For detecting adrenal and/or liver metastases, the summary NPV of the clinical evaluation was 0.95 (95% CI, 0.93 to 0.96), and for detecting bone metastases, it was 0.90 (95% CI, 0.86 to 0.93). PET scanning is more accurate than CT scanning or EUS for detecting mediastinal metastases. The NPVs of the clinical evaluations for brain, abdominal, and bone metastases are > or = 90%, suggesting that routinely imaging asymptomatic lung cancer patients may not be necessary. However, more definitive prospective studies that better define the patient population and improved reference standards are necessary to more accurately assess the true NPV of the clinical evaluation.Chest 01/2003; 123(1 Suppl):137S-146S. · 5.85 Impact Factor
- Chest 01/2000; 117(3):773-778. · 5.85 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: The significance of computed tomography of the thorax and mediastinoscopy in pretherapeutic mediastinal assessment for the staging of lung cancer remains controversial. The presents study was designed to establish a standard approach to cervical mediastinoscopy for otolaryngologists, who in Denmark traditionally are involved in the staging of non-small cell lung cancer. Sixty-four potentially operable patients with non-small cell lung cancer underwent thoracic computed tomography prior to bronchoscopy and cervical mediastinoscopy. Thirty-six of the 43 mediastinoscopically negative patients additionally underwent thoracotomy, which in 32 cases was considered curative. Mediastinoscopy alone established the lung cancer diagnosis in 20% of the patients. In diagnosing lymph node metastases in the superior mediastinum, a criterion of 10 mm for abnormal enlargement resulted in an overall sensitivity and specificity of mediastinal computed tomography of 72% and 85%, respectively, and the overall false-negative and false-positive rates appeared to be 18% and 25%, respectively. No clinicopathological characteristics could be identified that influenced the occurrence of mediastinal metastases or the accuracy of computed tomography. It is concluded that mediastinoscopy remains essential in the evaluation of patients with presumed or verified non-small cell lung cancer. For otolaryngologists, the strategy of routine cervical mediastinoscopy, performed under general anesthesia in the same procedure as bronchoscopy, is advocated as a standard approach to preoperative mediastinal assessment for the staging of non-small cell lung cancer.Danish medical bulletin 05/1995; 42(2):192-4. · 0.92 Impact Factor
The size of mediastinal lymph nodes and its relation with metastatic
involvement: a meta-analysis
Adrianus J. de Langena, Pieter Raijmakersa, Ingrid Riphagenb,
Marinus A. Paulc, Otto S. Hoekstraa,d,*
aDepartment of Nuclear Medicine and PET Research, VU University Medical Center, De Boelelaan 1117,
1081HV Amsterdam, The Netherlands
bVU University Medical Center, Amsterdam, The Netherlands
cDepartment of Surgery, VU University Medical Center, Amsterdam, The Netherlands
dDepartment of Clinical Epidemiology and Biostatistics, VU University Medical Center, Amsterdam, The Netherlands
Received 9 August 2005; received in revised form 26 September 2005; accepted 3 October 2005; Available online 6 December 2005
Positron emission tomography with 18-fluorodeoxyglucose (FDG-PET) seems to be superior to computed tomography (CT) in staging the
mediastinum in patients with non-small-cell lung cancer (NSCLC). However, recent results suggest that FDG-PET performance characteristics are
size and the probability of malignancy needs to be known to predict the post-test probabilities after PET, and finally, stratify patients for
mediastinoscopy or thoracotomy depending on the PETand CTresults. Therefore, we performed a meta-analysis of available studies reporting on
the prevalence of metastatic involvement for different size categories of enlarged lymph nodes in patients with NSCLC and were able to include
14 studies. The prevalence of metastatic involvement and conditional test performance of CT and FDG-PETwere calculated for lymph nodes
measuring10—15 mm, 16—20 mm and>20 mm.Wefounda post-testprobability forN2 diseaseof 5% forlymph nodes measuring10—15 mm on CT
in patients with a negative FDG-PETresult, suggesting that these patients should be planned for thoracotomy because the yield of mediastino-
scopy will be extremely low. For patients with lymph nodes measuring ?16 mm on CTand a negative FDG-PETresult a post-test probability for N2
disease of 21% was found, suggesting that these patients should be planned for mediastinoscopy prior to possible thoracotomy to prevent too
many unnecessary thoracotomies in this subset.
# 2005 Elsevier B.V. All rights reserved.
Keywords: Carcinoma, non-small-cell lung; Fluorodeoxyglucose F18; Tomography, X-ray computed; Lymphatic metastasis; Sensitivity and specificity
Since its introduction, computed tomography (CT) has
been used to stage patients with non-small-cell lung cancer
(NSCLC) and many studies have assessed its accuracy in
staging the mediastinum, but with variable results. It has
been shown that for CT, a threshold of 1 cm short-axis
diameter has the best trade-off between sensitivity and
specificity . However, malignant lymph nodes may well be
smaller than 1 cm and this explains the limited sensitivity of
CT, which is estimated to be 57% . Since specificity is about
82%, confirmation of the malignant nature of enlarged nodes
is necessary . Taken together, mediastinoscopy is required
in all patients without distant metastases. Positron emission
tomography with 18-fluorodeoxyglucose (FDG-PET) is super-
ior to CTin staging the mediastinum [3,4]. However, FDG-PET
is not perfect either. False-negative test results occur when
tumour size is well below the system’s highest resolution
(typically 5—7 mm) and false-positive results may occur due
to inflammatory disease and anthracosilicosis .
At present, there is debate about the diagnostic algorithm
which combines performance characteristics of CT, FDG-PET
and mediastinoscopy. In a recent meta-analysis, it was shown
that FDG-PET is more sensitive but less specific with lymph
node enlargement on CT . A Bayesian model suggested
that the post-test probability of malignant involvement is
very low (6%) if the mediastinum is normal at both CT and
FDG-PET. However, with enlarged nodes on CTand a negative
PETscan the situation is less clear: e.g. the reported median
prevalence of malignant involvement in enlarged nodes of
63% corresponds with a 17% post-test probability of
malignancy in case of a negative PET scan . The exact
size may have important implications forthe implementation
of these test results in daily practice. Therefore, we
performed a meta-analysis of available studies to assess
European Journal of Cardio-thoracic Surgery 29 (2006) 26—29
* Corresponding author. Tel.: +31 20 4444214; fax: +31 20 4443090.
E-mail address: email@example.com (O.S. Hoekstra).
1010-7940/$ — see front matter # 2005 Elsevier B.V. All rights reserved.
the prevalence of metastatic involvement for different size
categories of enlarged lymph nodes.
to identify all studies that assessed mediastinal lymph node
size (with CT) and the prevalence of metastatic involvement
in patients with NSCLC (Table 1).
This search coveredthe timeframe from1985 until August
2004, and it was followed by additional extensive cross-
referencing. Studies were included if they presented
sufficient data required to construct 2 ? 2 contingency
tables, and with a language restriction to English, German
and Dutch. Studies were excluded if they comprised more
than 10% small-cell lung carcinoma patients as well as if they
were published before 1985 because of the inferior quality of
the CT-scanners used.
2.1. Data analysis
The following information was extracted from the
included articles: year of study (to be able to account for
potential technological advances with CT), geographic origin
(to account for endemic diseases leading to reduced
specificity of CT), number of patients, histological subtype
of lung cancer (adeno-, squamous cell, small cell carcinoma),
number of hilar—mediastinal lymph nodes resected per
patient, method of size measurement at CT (short-axis or
other), CT-scanner generation, histopathological methods
and prevalence of metastasis.
We set out to calculate the prevalence of metastatic
involvement for the following size categories: 10—15 mm,
16—20 mm and >20 mm. However, since several researchers
used different classifications, we chose to consider lymph
node size groups with a difference of ?1 mm as a single group
(e.g. a size category of 10—15 mm in one study and of 10—
14 mm in another study were considered to be one category).
2.2. Statistical analysis
Chi-square test was applied to test for heterogeneity of
the study results for each lymph node size group. In case of
statistical heterogeneity a random effect model was used for
pooling, whereas in case of homogeneity a fixed effect model
was used. The proportion of histologically proven metastatic
lymph nodes and 95% confidence intervals (CI) were
calculated for each size group. Z-test was used to compare
the proportion of histological proven metastatic lymph nodes
between the different size categories. To predict the
conditional test performance of CT and FDG-PET, we
constructed 2 ? 2 contingency tables for each lymph node
size category, in which the FDG-PET test result and pathology
result were outlined. The post-test probability for N2 disease
after CTaccounted for the percentage of positive test results
(true-positive and false-positive). The sensitivity (91%) and
specificity (78%) of FDG-PET for enlarged lymph nodes 
positive, true-negative and false-negative test results.
Additionally, we calculated the positive and negative
predictive values (PPV and NPV, respectively) and the post-
test probability for mediastinal metastasis.
The initial search yielded 1537 hits, of which 73 articles
were potentially eligible based on title and abstract. Cross-
referencing yielded another 23 potentially eligible studies.
After full-text analysis, 14 were included [7—20], all of which
had been identified with the initial Medline search. In nine
studies lymph node size was measured in the short axis with
CT, in one in the long axis with CT and four studies did not
mention whether the short or long axis was measured. Using
these different methods of size measurement, the pre-
valence of malignant involvement ranged from 9% to 42% for
lymph nodes measuring 10—15 mm [15,16], 19—75% for
lymph nodes measuring 16—20 mm [9,16] and 27—100% for
lymph nodes measuring >20 mm [11,15].
3.1. Description of CT-based studies for measuring lymph
For the 14 CTstudies, the median number of patients per
study was 64 (range, 48—387). The mean number of resected
lymph nodes per patient was 2.2, while it was not recorded in
two studies [9,19]. Two studies also examined hilar lymph
nodes [10,11]. The mean percentage of SCLC patients was
3.6% (range, 0—10%).
3.2. Lymph node size measurement in the short axis with
The percentage of metastatic involvement for all lymph
node size groups was calculated for each included study
(Fig. 1). Four studies [12,15,18,19] provided data on the
A.J. de Langen et al./European Journal of Cardio-thoracic Surgery 29 (2006) 26—29
Medline search strategy
Tomography, X-ray computed[mesh]
ct[tw] OR tomo[tw] OR ‘‘cat scan’’ OR tomograph*[tw]
#1 OR #2
‘‘mediastinal lesion’’ OR ‘‘mediastinal lesions’’
Lymph[tw] OR lymphoid[tw] OR lymphatic[tw] OR mediastin*[tw]
node[tw] OR nodes[tw] OR nodal[tw]
#4 OR #5 OR (#6 AND #7)
Size[tw] OR diameter[tw] OR measur*[tw]
enlarged[tw] OR enlargment*[tw] OR sized[tw] OR volume[tw]
stage[tw] OR stages[tw] OR staging[tw]
#9 OR #10 OR #11
Carcinoma, non-small-cell lung[mesh] OR Lung neoplasms[mesh]
Lung[tw] OR lungs[tw] OR pulmonary[tw] OR bronchial[tw] OR
mediastinal[tw] OR mediastinum[tw] OR thorax[tw] OR thoracic[tw]
#15 OR #16
Cancer[tw] OR cancers[tw] OR carcinoma[tw] OR tumor[tw] OR
malignan*[tw] OR adenocarcinoma[tw] OR neoplas*[tw] OR
#18 OR #19
#17 AND #20
#13 OR #14 OR #21
#3 AND #8 AND #12 AND #22
prevalence of malignancy in lymph node size group 10—
were homogeneous (x2: 6.22; P = 0.101), we pooled it to
obtain a prevalence of metastatic lymph nodes of 29% (95%
CI: 0.23—0.36) (Fig. 2). Three eligible studies [12,15,19]
reported on lymph node size group 16—20 mm to yield 38
lymph nodes. These data were also homogeneous (x2: 0.07;
P = 0.966) and the pooled prevalence of malignancy was 68%
(95% CI: 0.52—0.81). The mean prevalence of metastatic
involvement was higher in the 16—20 mm group than in the
10—15 mm group (Z-test: P < 0.001). For lymph node size
group >20 mm, seven studies were found eligible [7,10—
14,19], for a total of 131 lymph nodes. Here the data were
heterogeneous (x2: 35.7; P < 0.001) due to one outlier 
(Fig. 1). With exclusion of the outlier, the data were
homogeneous (x2: 6.95; P = 0.224). With inclusion of all
seven studies, the mean prevalence of metastatic involve-
ment was higher in the >20 mm group, 66% (95% CI: 0.42—
0.83), than in the 10—15 mm group (Z-test: P = 0.002).
However, it was not higher than that in the 16—20 mm group
(Z-test: P = 0.879). Exclusion of the outlier did not make a
3.3. Modelling conditional test performance of FDG-PET
We calculated the predicted positive and negative values
of FDG-PET for identifying mediastinal lymph node metas-
tasis in patients with enlarged lymph nodes of different size
categories based on the calculated post-test probabilities
after CT (Table 2).
For patientswith lymph nodesmeasuring 10—15 mm onCT
and a negative FDG-PET result, the predicted post-test
probability of malignancy was 5%. A positive FDG-PET result
was predicted to yield a post-test probability of 62%. If CT
showed lymph nodes of ?16 mm, the post-test probability of
malignancy was 21% when the FDG-PET result was negative
and 90% if it was positive. These results were similar for the
subsets of 16—20 mm and >20 mm short-axis diameter.
The results of this meta-analysis suggest that the
probability of lymph node metastasis in mediastinal lymph
consistently two-fold higher in larger ones. We found 14
studies [7—20] that assessed the prevalence of metastatic
involvement of enlarged lymph nodes of different size
categories in patients with NSCLC. Although reports have
been published that show no increase in prevalence of
metastasis with increasing lymph node size or an increase at
larger lymph node size (e.g. >20 mm), most of them
measured either the long axis or did not specify the
measured axis [8,9]. These studies were excluded on
qualitative grounds. With pooling of the included studies,
only one was responsible for heterogeneity . In this study,
the percentage of T2 tumours was relatively high (66%) and
most were staged T2 based on associated atelectasis or an
obstructive bronchitis, both of which can lead to benign
hyperplasia of lymph nodes .Therefore, weclassified this
study as an outlier and calculations were done both with and
without this study.
Our data can be used to predict the conditional test
performance of CTand FDG-PET for different size categories
of enlarged lymph nodes on CT. Gould et al.  evaluated
enlarged lymph nodes as one group and found a post-test
probability for N2 disease of 17% when CT showed enlarged
lymph nodes and the FDG-PET result was negative. The
reported pre-test probability for enlarged lymph nodes was
63%. Our results suggest the added value of a subdivision of
enlarged lymph nodes. We found a post-test probability for
N2 disease of 5% for lymph nodes measuring 10—15 mm in the
short axis on CT in patients with a negative FDG-PET result.
Since, with extensive resection of lymph node stations,
mediastinoscopy reaches a sensitivity of 85% , 25 of such
patients should undergo mediastinoscopy to prevent one
unnecessary thoracotomy. Looking at the morbidity of
mediastinoscopy, we believe that patients in this group
A.J. de Langen et al./European Journal of Cardio-thoracic Surgery 29 (2006) 26—29
Fig. 1. Individual study results of CT-based studies (reference numbers in
brackets [7,10—15,18,19]) for metastatic involvement of all lymph node size
categories (error bars represent 95% CIs).
Fig. 2. Pooled values of metastatic involvement of CT-based lymph node size
categories (error bars represent 95% CIs).
Predicted positive and negative values of FDG-PET in patients with enlarged
lymph nodes of different size categories, assuming FDG-PETsensitivity of 91%
and specificity of 78% for enlarged nodes 
Lymph node size categoryNPV (%) PPV (%)
CT 10—15 mm
CT 16—20 mm
CT >20 mm, with outlier
CT >20 mm, without outlier
NPV: negative predictive value; PPV: positive predictive value.
should be directly planned for surgery, without previous
mediastinoscopy. However, for patients with lymph nodes
measuring?16 mmintheshortaxis onCTandwitha negative
FDG-PET result, the post-test probability for N2 disease was
21%. This suggests that preoperative invasive mediastinal
demonstrates extrathoracic spread). A prospective evalua-
show whether indeed five patients are needed to undergo
invasive staging in order to obtain one positive biopsy. Since
we assumed that the accuracy of FDG-PET is independent of
size above the centimetre level, we performed a sensitivity
analysis recalculating the results for sensitivity and speci-
ficity of FDG-PET of 85% and 70%, respectively. This shows
that post-test probability for N2 disease will only moderately
increase to 7% in patients having 10—15 mm lymph nodes on
CT and a negative FDG-PET scan (vs 31% in patients having
lymph nodes measuring ?16 mm on CT).
We conclude that the prevalence of metastasis strongly
increases above the 15 mm short-axis threshold at CT
scanning, and that this may have implications for the
positioning of mediastinoscopy in patients with negative
lymph nodes on FDG-PET. The data suggest that patients with
nodes measuring <15 mm on CT should be planned for
thoracotomy if FDG-PET does not reveal mediastinal
involvement, since the expected yield of mediastinoscopy
is extremely low. Patients with lymph nodes measuring
?16 mm on CTand a negative FDG-PETresult should undergo
mediastinoscopy before possible thoracotomy.
 Glazer GM, Gross BH, Quint LE, Francis IR, Bookstein FL, Orringer MB.
Normal mediastinal lymph nodes: number and size according to American
Thoracic Society mapping. Am J Roentgenol 1985;144:261—5.
 Toloza EM, Harpole L, McCrory DC. Noninvasive staging of non-small cell
lung cancer:a review of thecurrent evidence.Chest2003;123:137S—46S.
 Halter G, Buck AK, Schirrmeister H, Aksoy E, Liewald F, Glatting G,
Neumaier B, Muhling B, Nussle-Kugele K, Hetzel M, Sunder-Plassmann
L, Reske SN. Lymph node staging in lung cancer using [18F]FDG-PET.
Thorac Cardiovasc Surg 2004;52:96—101.
 Gould MK, Kuschner WG, Rydzak CE, Maclean CC, Demas AN, Shigemitsu
and computed tomography for mediastinal staging in patients with non-
small-cell lung cancer: a meta-analysis. Ann Intern Med 2003;139:879—
 Konishi J, Yamazaki K, Tsukamoto E, Tamaki N, Onodera Y, Otake T,
Morikawa T, Kinoshita I, Dosaka-Akita H, Nishimura M. Mediastinal lymph
node staging by FDG-PET in patients with non-small cell lung cancer:
analysis of false-positive FDG-PET findings. Respiration 2003;70:500—6.
 Aaby C, Kristensen S, Nielsen SM. Mediastinal staging of non-small-cell
lung cancer: computed tomography and cervical mediastinoscopy. ORL J
Otorhinolaryngol Relat Spec 1995;57:279—85.
 Brion JP,DepauwL, Kuhn G,deFracquenP,Friberg J, Rocmans P, Struyven
J. Role of computed tomography and mediastinoscopy in preoperative
staging of lung carcinoma. J Comput Assist Tomogr 1985;9:480—4.
 McKenna Jr RJ, Libshitz HI, Mountain CE, McMurtrey MJ. Roentgeno-
graphic evaluation of mediastinal nodes for preoperative assessment in
lung cancer. Chest 1985;88:206—10.
 Gupta NC, Graeber GM, Bishop HA. Comparative efficacy of positron
emission tomography with fluorodeoxyglucose in evaluation of small
(<1 cm), intermediate (1 to 3 cm), and large (>3 cm) lymph node
lesions. Chest 2000;117:773—8.
 Izbicki JR, Thetter O, Kreusser T, Passlick B, Trupka A, Karg O, Wilker DK,
Schweiber L. The value of computerized tomography and of the surgical
assessment for lymph node staging in bronchial carcinoma. A prospective
study. Chirurg 1991;62:885—90.
 Krestin GP, Krestin M, Gross-Fengels W, Thul H. Conventional X-ray
techniques and computed tomography in the diagnosis of mediastinal
lymph node involvement in non-small cell bronchogenic cancer. Which
method is reliable? Rofo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr
 Kristensen S, Aaby C, Nielsen SM. Mediastinal staging of lung cancer. Is
mediastinoscopy still essential? Dan Med Bull 1995;42:192—4.
 McLoud TC, Bourgouin PM, Greenberg RW, Kosiuk JP, Templeton PA,
Shepard JA, Moore EH, Wain JC, Mathisen DJ, Grillo HC. Bronchogenic
carcinoma: analysis of staging in the mediastinum with CT by correlative
lymph node mapping and sampling. Radiology 1992;182:319—23.
 Mori K, Yokoi K, Saito Y, Tominaga K, Miyazawa N. Diagnosis of mediast-
 Nakata H, Ishimaru H, Nakayama C, Yoshimatsu H. Computed tomography
for preoperative evaluations of lung cancer. J Comput Tomogr 1986;10:
 Poon PY, Bronskill MJ, Henkelman RM, Rideout DF, Shulman HS, Weisbrod
GL, Steinhardt MI, Dunlap HJ, Ginsberg RJ, Feld R. Mediastinal lymph
node metastases from bronchogenic carcinoma: detection with MR ima-
ging and CT. Radiology 1987;162:651—6.
 Takamochi K, Nagai K, Suzuki K, Yoshida J, Ohde Y, Nishiwaki Y. Clinical
predictors of N2 disease in non-small cell lung cancer. Chest 2000;117:
 Wathen CG,Kerr KM, ReidW,WightmanAJ, Best JJ,Millar AM,Walker WS,
Cameron EW, Douglas NJ. A comparison of cobalt (57Co) bleomycin
scanning and contrast-enhanced CT scanning for assessment of the
mediastinum in lung cancer. Chest 1990;97:1148—51.
 Whittlesey D. Prospective computed tomographic scanning in the
staging of bronchogenic cancer. J Thorac Cardiovasc Surg 1988;95:
 Ferguson MK, MacMahon H, Little AG, Golomb HM, Hoffman PC, Skinner
DB. Regional accuracy of computed tomography of the mediastinum in
staging of lung cancer. J Thorac Cardiovasc Surg 1986;91:498—504.
 Hammoud ZT, Anderson RC, Meyers BF, Guthrie TJ, Roper CL, Cooper JD,
Patterson GA. The current role of mediastinoscopy in the evaluation of
thoracic disease. J Thorac Cardiovasc Surg 1999;118:894—9.
A.J. de Langen et al./European Journal of Cardio-thoracic Surgery 29 (2006) 26—29