Efficacy of Helical Dynamic CT Versus Integrated PET/CT for Detection of Mediastinal Nodal Metastasis in Non-Small Cell Lung Cancer
Abstract
OBJECTIVE. The purpose of our study was to compare the diagnostic efficacies of helical dynamic CT and integrated PET/CT for the prediction of mediastinal nodal metastasis in stage T1 non-small cell lung cancer (NSCLC).
MATERIALS AND METHODS. One hundred forty-three patients with stage T1 NSCLC underwent both helical dynamic CT and integrated PET/CT followed by surgical nodal staging. In helical dynamic CT, patients were regarded to have stage N2 disease when a nodule showed a peak enhancement ≥ 110 H or a net enhancement ≥ 60 H. In integrated PET/CT, nodes were regarded as positive for malignancy when they showed 3 3.5 in maximum standardized uptake value with a discrete margin and more 18F-FDG uptake than mediastinal structures. Sensitivities, specificities, and accuracies for mediastinal nodal metastasis detection were compared for helical dynamic CT and integrated PET/CT using the McNemar test.
RESULTS. Of the 143 patients, 34 (24%) had positive mediastinal nodes. The sensitivity, specificity, and accuracy for mediastinal nodal metastasis prediction on helical dynamic CT were 65% (22 of 34 patients), 89% (97 of 109), and 83% (119 of 143), respectively, whereas those on integrated PET/CT were 56% (19 of 34), 100% (109 of 109), and 90% (128 of 143). The p values were 0.664, < 0.001, and 0.015.
CONCLUSION. In stage T1 NSCLC, contrast-enhanced helical dynamic CT better predicts, but not significantly so, mediastinal nodal metastasis than PET/CT, whereas PET/CT shows perfect specificity and higher accuracy than helical dynamic CT.
Introduction
At the initial diagnosis and staging of non-small cell lung cancer (NSCLC), treatment options are determined on the basis of TNM staging. Although there is some controversy concerning the determination of the T stage of NSCLC, the majority of techniques offer relatively high and similar accuracies [1]. However, the same cannot be said of the diagnosis of lymph node metastasis.
As for the mediastinoscopic evaluation of lymph nodes, initial studies raise questions concerning routine mediastinoscopy in patients with stage T1 NSCLC because of the low prevalence of mediastinal nodal metastases and the substantial morbidity and mortality associated with the invasiveness of the procedure. However, several recent studies have reported relatively high frequencies of mediastinal lymph node metastases in stage T1 NSCLC, ranging from 16% to 21%, by complete nodal sampling using either mediastinoscopy or thoracotomy [2-4]. Therefore, the accurate detection of mediastinal nodal metastasis in stage T1 NSCLC helps with decision making concerning staging procedures (mediastinoscopy) and treatment options (surgery vs neoadjuvant chemoradiation therapy).
In the CT diagnosis of mediastinal nodal metastasis, lymph nodes have been evaluated classically using the size criterion of 10 mm in short-axis diameter, which is based on statistical data regarding the normal size range of mediastinal lymph nodes [5, 6]. However, the results of mediastinal nodal metastasis prediction in NSCLC by CT diagnosis are unsatisfactory and show a sensitivity of 41-54% and a specificity of 43-97% [4, 7, 8]. Moreover, in stage T1 lung cancer, CT has a sensitivity of only 27-41% [4, 9].
The likelihood of metastasis increases as the number of intratumoral microvessels increases in lung cancer. Because mean peak attenuation values of helical dynamic CT reflect microvessel densities in lung cancer and are significantly higher in vascular endothelial growth factor (VEGF)-positive lung cancer, the likelihood of metastasis is probably increased in lung cancers that show strong enhancement on helical dynamic CT [9, 10]. Shim et al. [9] suggested that the probability of mediastinal nodal metastasis is high in strongly enhancing malignant lung nodules with 3 110 H of peak enhancement or 3 60 H of net enhancement on helical dynamic CT in stage T1 NSCLC. With these cutoff vales for nodule enhancement of 3 110 H of peak enhancement or 60 H of net enhancement for mediastinal nodal metastases, helical dynamic CT enabled mediastinal nodal metastasis prediction with sensitivity of 62-77%, specificity of 70-76%, and accuracy of 71-74%.
PET with 18F-FDG as a radiotracer has recently been reported to be more effective at detecting tumor involvement in mediastinal and hilar lymph nodes than CT [11, 12]. FDG PET has an advantage in functional imaging because it reflects tissue hypermetabolism but is limited in anatomic detail because of its poor resolution. On the other hand, integrated PET/CT, by combining morphologic CT and functional PET data, provides satisfactory spatial resolution and anatomic and metabolic information. Moreover, integrated PET/CT scanners have shown promising initial oncologic imaging results [13, 14]. However, in stage T1 NSCLC, integrated PET/CT still has a low sensitivity of 47%, although it has 100% positive predictive value and a high negative predictive value [15].
Although helical dynamic CT and integrated PET/CT have been the most reliable techniques in the prediction of mediastinal nodal metastasis in stage T1 NSCLC until now [9, 15], the diagnostic efficacies of these two techniques have not been directly compared. Therefore, the purpose of our study is to compare the diagnostic accuracies of helical dynamic CT and integrated PET/CT for the prediction of mediastinal nodal metastasis in patients with stage T1 NSCLC.
Materials and Methods
Patient Enrollment
From July 2003 to June 2005, a total of 180 patients with histopathologically proven stage T1 lung cancer underwent a dynamic MDCT enhancement study. Of these 180 patients, we retrospectively enrolled 164 nonrandom patients who underwent both helical dynamic CT and integrated PET/CT during this period. Our institutional review board approved the study protocol for both helical dynamic CT and integrated PET/CT studies, and written informed consent for the studies was obtained from all patients.
At our institution, integrated PET/CT is usually recommended for cancer staging before surgery in patients with proven lung cancer or for tissue characterization in patients with clinically indeterminate pulmonary nodules or masses. Twelve patients were excluded from analysis because they had a positive PET scan or brain MRI indicating extrathoracic metastases. Nine patients with pathologically proven positive nodes on cervical mediastinoscopy were also excluded because they had received concurrent neoadjuvant chemoradiation therapy before curative resection in another hospital. The remaining 143 patients (82 men and 61 women; age range, 31-72 years; mean age, 60 years) underwent surgical nodal staging. Eleven of the 143 patients underwent mediastinoscopy only, and the remaining 132 patients underwent lobectomy or pneumonectomy plus lymph node dissection. None had evidence of distant metastasis at the time of surgery.
Helical Dynamic CT Image Acquisition
Helical dynamic CT was performed using a 4-MDCT scanner (LightSpeed QX/i, GE Healthcare) or a 16-MDCT scanner (LightSpeed Ultra or Ultra16, GE Healthcare). Before the IV injection of contrast medium, a series of 13 images was obtained through a nodule for 30 mm along the z-axis with 2.5-mm collimation at 120 kVp, 90 mA, 0.8-second gantry rotation time, and 3.75 mm/s table speed over 8 seconds. Thereafter, an additional six series of images were obtained at 30, 60, 90, and 120 seconds and 5 and 15 minutes after IV injection of contrast medium (iomeprol [Iomeron 300, Bracco]) at 3 mL/s, total of 120 mL, using a power injector (Envision CT, Medrad) and using the same parameters as for the initial unenhanced series (a total series of seven images at 0, 30, 60, 90, and 120 seconds and at 5 and 15 minutes).
Image data were reconstructed with a thickness of 2.5 mm using a standard algorithm. Immediately after the dynamic study at 120 seconds, helical CT (125 mA, 120 kVp, 5-mm collimation, and 15 mm/s table speed) scans were obtained from the lung apices to the level of the middle pole of both kidneys for tumor staging. All thin-section and dynamic CT image data were directly interfaced with our PACS system (PathSpeed, GE Healthcare), which displayed all image data on monitors (four monitors, 1,536 × 2,048 image matrices, 8-foot bit viewable gray scale, and 60-foot-lambert luminescence). Both mediastinal (window width, 400 H; window level, 20 H) and lung (window width, 1,500 H; window level, -700 H) window images were viewed on these monitors.
Integrated PET/CT Acquisition
All 143 patients underwent an integrated FDG PET/CT study within 7 days of helical dynamic CT (mean, 3.2 days; median, 2 days), and all patients fasted for at least 6 hours before the integrated PET/CT examination, although oral hydration with glucose-free water was allowed. After we tested each patient's peripheral blood to ensure a normal blood glucose level, each patient received an IV injection of 370 MBq (10 mCi) of FDG and then rested for approximately 45 minutes before undergoing imaging. Image acquisition was performed using an integrated PET/CT device (Discovery LS, GE Healthcare) consisting of an Advance NXi PET scanner and an 8-MDCT LightSpeed Plus CT scanner. The axes of both systems were mechanically aligned so that shifting the examination table by 68 cm moved the patient from the CT gantry into the PET gantry. The resulting PET and CT images were then coregistered.
CT was performed from the head to the pelvic floor using a standardized protocol involving 140 kV, 80 mA, tube-rotation time of 0.5 second per rotation, pitch of 6, and section thickness of 5 mm, which matched the PET section thickness. Patients were in normal shallow respiration during the CT scan acquisition. No contrast material was administered. Immediately after CT, PET was performed in the identical axial field of view. The acquisition time for PET was 5 minutes per table position. CT data were resized from a 512 × 512 matrix to a 128 × 128 matrix to match the PET data so that the images could be fused and CT transmission maps generated. PET image data sets were reconstructed iteratively using the ordered subsets expectation maximization algorithm and by segmented attenuation correction (two iterations, 28 subsets) using the CT data. Coregistered images were displayed using eNTEGRA software (GE Healthcare).
Helical Dynamic CT Image Evaluation
We measured nodule attenuation values in the same area on the selected image for each cluster on each occasion (from unenhanced images to the 15-minute images). A circular region of interest (ROI) was placed over a nodule, and an ROI covering about one half of the diameter of the nodule at its equator was examined. Calcified, cavitary, or necrotic areas were avoided; thus ROIs were designed to be as large as possible without overlapping such areas. Two radiologists, with 4 years and 8 years of chest CT experience, independently measured attenuation values. All measurements were made at the time of CT examination. The radiologists were unaware of the patient information such as age and clinical history. Two measurements were obtained for each nodule at each imaging phase by each observer.
Mean attenuation values were recorded by each observer and then, by averaging attenuation values by the two observers, the extent of nodule enhancement was analyzed by calculating peak enhancement and net enhancement (wash-in of contrast medium). Peak enhancement attenuation was defined as the maximum attenuation value of a nodule over the entire time course of the dynamic study. Net enhancement attenuation was calculated by subtracting the unenhanced attenuation value from the peak enhancement attenuation value. If a patient had a pulmonary nodule with an ≥ 110 H peak enhancement or an ≥ 60 H net enhancement, the mediastinal nodes were considered metastatic [9].
We also measured the short-axis diameter of each mediastinal nodal group according to the lymph node map definition for lung cancer staging proposed by Mountain and Dresler [16]: group 1, highest mediastinal (1R, right; 1L, left); group 2, upper paratracheal (2R, right; 2L, left); group 3, prevascular or retrotracheal; group 4, lower paratracheal (4R, right; 4L, left); group 5, subaortic (aortopulmonary window); group 6, paraaortic (ascending aorta or phrenic); group 7, subcarinal; group 8, paraesophageal; and group 9, pulmonary ligament (9R, right; 9L, left).
Nodal groups with a lymph node of ≥ 10 mm in the short-axis diameter on CT scans were considered metastatic on a per lymph node basis [6]. Nodes containing nodular or laminated calcification were regarded as benign irrespective of size.
Integrated PET/CT Image Analysis
Integrated PET/CT data sets were evaluated prospectively by one chest radiologist with 16 years of chest CT interpretation experience and 2 years of integrated PET/CT interpretation experience and one nuclear medicine physician with 12 years of PET interpretation experience who were blinded to clinical and pathologic results. Decisions on findings were reached by consensus. Nodal stations were evaluated by allocating them to nine groups according to the lymph node map definition for lung cancer staging proposed by Mountain and Dresler [16] as previously described. Mediastinal nodes with increased glucose uptake satisfying both qualitative (greater than that of the surrounding tissue) and quantitative (a maximum standardized uptake value [SUV] adjusted for patient body weight of ≥ 3.5 with a distinct margin) criteria were considered positive.
By receiver operating characteristic (ROC) curve analysis using different maximum SUV threshold cutoffs, we decided that an SUV of 3.5 was optimum for differentiating benign and malignant nodules on our machines [14, 17]. Mediastinal nodes were divided into four categories according to the integrated PET/CT results: positive uptake without calcification or high attenuation, positive uptake with calcification or high attenuation, negative uptake with calcification or high attenuation, and negative uptake without calcification or high attenuation. Calcification was regarded as present when it was nodular or laminated in pattern and > 200 H. A high-attenuation lymph node was defined as a node that appeared visually to have a higher attenuation than mediastinal vascular structures or had an attenuation of > 70 H by ROI-based measurement. Even if glucose uptake was high (higher than background activity or > 3.5 in SUV), calcified lymph nodes and lymph nodes with a higher attenuation (> 70 H) than surrounding great vessels on CT images on integrated PET/CT were regarded as benign [14].
Surgical-Pathologic Correlation
Surgical staging included mediastinoscopic nodal sampling alone in 11 patients and thoracotomy and lymph node dissection in 132. The surgical staging was performed by one of two experienced thoracic surgeons (one with 17 years of experience and the other with 12). In 11 patients, only mediastinoscopic nodal staging results were available because curative resection was deferred in favor of neoadjuvant concurrent chemoradiation therapy due to the presence of positive nodes. During mediastinoscopy, American Thoracic Society (ATS) lymph node stations [16] of 2R, 4R, 2L, 4L, and 7 were routinely sampled, and during thoracotomy all encountered lymph nodes were removed from the ATS lymph nodal stations of 10R, 9, 8, 7, 4R, 3, and 2R in tumors of the right lung and from areas 10L, 9, 8, 7, 6, 5, and 4L of the left lung. When necessary, station 1 (highest mediastinal) or 2L (when tumors were located in the left lung) nodes were also evaluated during mediastinoscopy or thoracotomy.
A lung pathologist with 10 years of experience described lymph nodes (location and number) according to the surgeons' labeling of dissected lymph nodes [16]. The pathologist then evaluated the nodes for the presence or absence of tumor as numbered in the surgical field and recorded the presence or absence of tumor in the nodes. Specimens were stained with H and E and examined by light microscopy.
Data and Statistical Analysis
Statistical analyses were performed using commercially available software (SAS version 8.2, SAS Institute). Agreement between the two observers in terms of nodule measured attenuation values was analyzed by calculating intraclass correlation coefficients. The Student's t test and the Mann-Whitney test were used to analyze statistical differences in the extents of enhancement of pulmonary nodules on helical dynamic CT and SUVs on integrated PET/CT between those with or without mediastinal nodal metastasis. A p value of < 0.05 was regarded as significant.
Diagnostic characteristics—that is, sensitivity, specificity, accuracy, and positive and negative predictive values—were calculated by considering the peak and net enhancements of nodules (lymph node metastasis with peak enhancement of ≥ 110 H or net enhancement of ≥ 60 H) on helical dynamic CT and by considering maximum SUVs from integrated PET/CT on a per-patient basis. Likewise, diagnostic characteristics on a per-nodal basis were also calculated by measuring short-axis diameter on helical dynamic CT and considering SUVs obtained on integrated PET/CT for each mediastinal lymph node group.
The McNemar test was used to compare helical dynamic CT and integrated PET/CT with respect to sensitivity, specificity, and accuracy. A p value of < 0.05 was regarded as significant.
In patients with pathologically proven positive mediastinal nodal metastasis, helical dynamic CT and integrated PET/CT results concerning the detection of mediastinal nodal metastases were compared. In false-negative cases on integrated PET/CT for the diagnosis of mediastinal nodal metastasis, the size and CT attenuation values of nodes were reviewed to identify the relationship between FDG uptake of nodes on PET and size or morphology on CT correlated with the pathologic results of the nodes.
Results
Results of Imaging Studies and Pathology
We evaluated 453 mediastinal nodal groups in 143 lung cancer patients. The pathologic diagnoses of these 143 patients were as follows: adenocarcinoma (n = 107), squamous cell carcinoma (n = 20), unspecified NSCLC (n =5), large cell neuroendocrine carcinoma (n =5), pleomorphic carcinoma (n = 3), and atypical carcinoid tumor (n = 3). Of the 143 patients, 34 (24%) had more than one metastatic lymph node in the mediastinum. And, of the 453 mediastinal nodal groups, 50 nodal groups were metastatic.
Tumor characteristics are summarized in Table 1, which shows the extent of enhancement on helical dynamic CT and SUVs on integrated PET/CT. Net enhancements on helical dynamic CT (p = 0.049, Student's t test) and SUVs on integrated PET/CT (p = 0.043, Mann-Whitney test) were significantly higher in patients with lymph node metastasis in the mediastinum. Good interobserver agreement was obtained between the two observers in terms of measured nodule attenuation values (intraclass correlation coefficient = 0.863-0.902, p < 0.0001) on helical dynamic CT.
| Nodule Characteristics | With Stage N2 Node Metastasis (n = 34) | Without Stage N2 Node Metastasis (n = 109) | p |
|---|---|---|---|
| Helical dynamic CT | |||
| Peak enhancement (H) | |||
| Mean ± SD | 102 ± 19.4 | 96 ± 17.4 | 0.162a |
| Median | 110 | 96 | |
| Range | 45-135 | 49-151 | |
| Net enhancement (H) | |||
| Mean ± SD | 55 ± 19.3 | 48 ± 16.0 | 0.049a |
| Median | 60 | 48 | |
| Range | 11-95 | 12-101 | |
| Integrated PET/CT | |||
| Maximum standardized uptake value | |||
| Mean ± SD | 9.4 ± 2.8 | 8.5 ± 5.6 | 0.043b |
| Median | 8.7 | 7.1 | |
| Range | 5.6-16.9 | 1.3-24.0 |
a
Student's t test.
b
Mann-Whitney test.
Efficacy of Helical Dynamic CT and Integrated PET/CT in Diagnosing Mediastinal Nodal Metastasis
When applying the diagnostic criteria for the presence of mediastinal nodal metastasis fulfilling pulmonary nodule enhancement of either ≥ 110 H of peak enhancement or ≥ 60 H of net enhancement, the sensitivity, specificity, and accuracy for a prediction of mediastinal nodal metastasis were 65%, 89%, and 83%, respectively, on a per-patient basis on helical dynamic CT. The results for integrated PET/CT were as follows: sensitivity, 56%; specificity, 100%; and accuracy, 90% (Table 2). The sensitivity of helical dynamic CT was higher than that of integrated PET/CT, although the difference was not statistically significant (p = 0.664). The specificity and accuracy of integrated PET/CT were significantly higher than those of helical dynamic CT (p <0.001, p = 0.015, respectively).
| Parameter | Helical Dynamic CT | Integrated PET/CT | p |
|---|---|---|---|
| Sensitivity | 22/34 (65) | 19/34 (56) | 0.664 |
| Specificity | 97/109 (89) | 109/109 (100) | < 0.001 |
| Accuracy | 119/143 (83) | 128/143 (90) | 0.015 |
| Positive predictive value | 22/34 (65) | 19/19 (100) | |
| Negative predictive value | 97/109 (89) | 109/124 (88) |
Note—Data in parentheses are percentages.
In the evaluation of nodal stations by applying the criterion for mediastinal nodal metastasis of ≥ 10 mm in the short-axis diameter on helical dynamic CT, sensitivity, specificity, and accuracy for the diagnosis of mediastinal nodal metastasis were 42%, 99%, and 92%, respectively, on a per-nodal basis. Those for integrated PET/CT were as follows: sensitivity, 44%; specificity, 99%; and accuracy, 93% (Table 3). The sensitivities, specificities, and accuracies of helical dynamic CT and integrated PET/CT were not significantly different from each other (p =0.727 ≈ 1.000).
| Parameter | Helical Dynamic CT | Integrated PET/CT | p |
|---|---|---|---|
| Sensitivity | 21/50 (42) | 22/50 (44) | 1.000 |
| Specificity | 397/403 (99) | 399/403 (99) | 0.727 |
| Accuracy | 418/453 (92) | 421/453 (93) | 1.000 |
| Positive predictive value | 21/27 (78) | 22/26 (85) | |
| Negative predictive value | 397/426 (93) | 399/427 (93) |
Note—Data in parentheses are percentages.
Comparison of Nodal Metastasis Prediction Results Between Helical Dynamic CT and Integrated PET/CT in Positive Cases
Of the 34 NSCLC patients with proven mediastinal nodal metastasis, nodal metastasis was detected concordantly in 10 patients by both helical dynamic CT and integrated PET/CT (Fig. 1A, 1B, 1C, 1D). However, it was not detected by helical dynamic CT or integrated PET/CT in three patients. In the remaining 21 patients, helical dynamic CT and integrated PET/CT results were discordant; helical dynamic CT was positive in 12 patients (Fig. 2A, 2B, 2C, 2D, 2E, 2F) and integrated PET/CT in nine patients (Fig. 3A, 3B, 3C, 3D). Therefore, mediastinal nodal metastasis was suggested by either technique—helical dynamic CT or integrated PET/CT—in 31 of the 34 (91%) patients with proven disease.
False-negative results were obtained in 15 patients by integrated PET/CT. In 12 of the 15 patients with false-negative results, PET showed no significant uptake in the area of the mediastinum, although the CT scan showed visible lymph nodes in the mediastinum of 5.5 mm in average diameter (size range, 2-8 mm; SD, 1.8 mm) in 17 false-negative mediastinal nodal stations. In the remaining three patients with false-negative results, PET/CT was interpreted as showing inflammatory conditions of reactive lymph nodes due to high attenuation on CT, although nodes showed increased FDG uptake.
Discussion
Our study shows a 24% (34 of 143 patients) prevalence of mediastinal nodal metastasis in stage T1 lung cancer, which is slightly higher than that (7-21%) reported previously [2-4]. This higher than expected prevalence of mediastinal nodal metastasis in stage T1 NSCLC emphasizes the need for the development of more reliable, useful indicators for predicting mediastinal lymph node metastasis. However, only about 41% of patients with metastatic mediastinal lymph nodes were correctly diagnosed using the CT size (short-axis diameter > 10 mm) criterion for the prediction of the presence of mediastinal nodal metastasis [4].
Efforts have been made to identify CT findings that indicate a propensity to metastasize based on analyses of morphologic characteristics such as size or the marginal characteristics of primary tumors [14, 18-21]. However, results have been unsatisfactory and controversial, especially in solid NSCLC nodules. Shim et al. [9] suggested that the extent of nodule enhancement is related to a propensity toward mediastinal or hilar nodal metastasis, not tumor size, marginal characteristics, or the presence of necrosis or bronchovascular thickening, which showed no correlation with mediastinal or hilar nodal metastasis. They reported that stage T1 lung cancers showing high peak enhancement or net enhancement on dynamic CT have a high likelihood of mediastinal nodal metastasis, with a sensitivity of 62%, specificity of 76%, and accuracy of 74% after applying a cutoff value of ≥ 110 H of peak enhancement and a sensitivity of 77%, specificity of 70%, and accuracy of 71% after applying a cutoff value of ≥ 60 H of net enhancement.
As was found by meta-analysis, PET is more sensitive than CT for detecting small metastatic nodes, with sensitivity in the range of 79-84% and specificity of 89-91% for mediastinal nodal metastasis [22-24]. With the introduction of the coregistration of morphologic CT and functional PET (integrated PET/CT) data, the overall efficacy of PET/CT in predicting the presence of nodal metastasis (regardless of T staging) has improved to 81-84% accuracy [14, 25]. Shim et al. [14] reported that for the depiction of malignant nodes, including mediastinal and hilar nodal groups regardless of T staging, the sensitivity, specificity, and accuracy of CT were 70%, 69%, and 69%, respectively, whereas those of PET/CT were 85%, 84%, and 84% (p = 0.25, p < 0.001, and p < 0.001). However, according to a study by Kim et al. [15], the sensitivity, specificity, and accuracy of integrated PET/CT for mediastinal nodal staging were 42%, 100%, and 94%, respectively, in stage T1 NSCLC [15]. PET/CT is still limited in the detection of metastatic nodes when the nodes are microscopic in diameter [25, 26]. Our false-negative cases on integrated PET/CT for the prediction of mediastinal nodal metastasis support this limitation of integrated PET/CT and also disclose false-negative results for small lymph nodes. In 12 of 15 false-negative cases in which there was no significant FDG uptake on PET, there were visible lymph nodes in the mediastinum, but they were small, with an average short-axis diameter of 5.5 mm.
In our study, the sensitivity of helical dynamic CT was found to be higher than that of integrated PET/CT for predicting mediastinal nodal metastasis. However, the accuracy of helical dynamic CT is lower than that of integrated PET/CT because of the 100% specificity of integrated PET/CT. In terms of sensitivity for detecting mediastinal nodal metastasis, only three (9%) of 34 patients with proven mediastinal nodal metastasis were not shown to have mediastinal nodal metastasis on either helical dynamic CT or integrated PET/CT. Each of the remaining false-negative cases, nine on helical dynamic CT and 12 on integrated PET/CT, was properly suggested to have mediastinal nodal metastasis by the other technique.
Although integrated PET/CT is currently the best technique available for an imaging diagnosis of nodal staging in NSCLC, its sensitivity is not high for the detection of mediastinal nodal metastasis in stage T1 NSCLC. On the other hand, our study showed up to 24% (34 of 143 patients) prevalence for mediastinal nodal metastasis in stage T1 lung cancer. Consequently, the need for routine mediastinoscopic lymph node biopsy has increased despite its high cost and associated risks. We suggest that mediastinoscopic biopsy of mediastinal lymph nodes should be recommended in patients with highly enhancing NSCLC with ≥ 110 H of peak enhancement or ≥ 60 H of net enhancement, even when the results of integrated PET/CT are negative for mediastinal lymph node metastasis.
Our study may have a selection bias because it was not designed prospectively. However, in our hospital we perform helical dynamic CT for every solitary pulmonary nodule detected on chest radiography that appears solid on thin-section CT. Integrated PET/CT was performed on all stage T1 NSCLC patients and on all patients with an indeterminate-nature nodule on helical dynamic CT or nodule biopsy. We did not obtain respiration-gated PET/CT images; thus, motion-induced registration and FDG uptake quantification errors might have contributed to an incomplete mediastinal nodal characterization evaluation.
In conclusion, our evaluation of the extent of nodule enhancement on helical dynamic CT in stage T1 NSCLC shows better (although not statistically significant) sensitivity for the prediction of mediastinal nodal metastasis on a per-patient basis than PET/CT, whereas PET/CT was found to have perfect specificity and positive predictive values. Therefore, mediastinoscopy may be omitted and direct neoadjuvant therapy given to patients with positive nodal metastasis results on PET/CT. Mediastinoscopy may be recommended in patients with malignant lung nodules showing high enhancement on helical dynamic CT even though PET/CT does not suggest the presence of mediastinal nodal metastasis.
Footnotes
Presented at the 2005 annual meeting of the Radiological Society of North America, Chicago, IL.
Address correspondence to K. S. Lee ([email protected]).
References
1.
Greene FL, Page DL, Fleming ID, et al., eds. AJCC cancer staging manual, 6th ed. New York, NY: Springer,2002 : 165-177
2.
Choi YS, Shim YM, Kim J, Kim K. Mediastinoscopy in patients with clinical stage I non-small cell lung cancer. Ann Thorac Surg 2003; 75:364-366
3.
Tahara RW, Lackner RP, Graver LM. Is there a role for routine mediastinoscopy in patients with peripheral T1 lung cancers? Am J Surg 2000; 180:488-491
4.
Seely JM, Mayo JR, Miller RR, Müller NL. T1 lung cancer: prevalence of mediastinal nodal metastases and diagnostic accuracy of CT. Radiology 1993; 186:129-132
5.
Genereux GP, Howie JL. Normal mediastinal lymph node size and number: CT and anatomic study. AJR 1984; 142:1095-1100
6.
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. AJR 1985; 144:261-265
7.
McLoud TC, Bourgouin PM, Greenberg RW, et al. Bronchogenic carcinoma: analysis of staging in the mediastinum with CT by correlative lymph node mapping and sampling. Radiology 1992; 182:319-323
8.
Libshitz HI, McKenna RJ Jr, Haynie TP, McMurtrey MJ, Mountain CT. Mediastinal evaluation in lung cancer. Radiology 1984; 151:295-299
9.
Shim SS, Lee KS, Chung MJ, Kim H, Kwon OJ, Kim S. Do hemodynamic studies of T1 lung cancer enable the prediction of hilar or mediastinal nodal metastasis? AJR 2006; 186:981-988
10.
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. Radiology 2004; 233:191-199
11.
Gupta NC, Tamim WJ, Graeber GG, Bishop HA, Hobbs GR. Mediastinal lymph node sampling following positron emission tomography with fluorodeoxyglucose imaging in lung cancer staging. Chest 2001; 120:521-527
12.
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 falsepositive FDG-PET findings. Respiration 2003; 70:500-506
13.
Pieterman RM, van Putten JW, Meuzelaar JJ, et al. Preoperative staging of non-small-cell lung cancer with positron-emission tomography. N Engl J Med 2000; 343:254-261
14.
Shim SS, Lee KS, Kim BT, et al. Non-small cell lung cancer: prospective comparison of integrated FDG PET/CT and CT alone for preoperative staging. Radiology 2005; 236:1011-1019
15.
Kim BT, Lee KS, Shim SS, et al. Stage T1 non-small cell lung cancer: preoperative mediastinal nodal staging with integrated FDG PET/CT—a prospective study. Radiology 2006; 241:501-509
16.
Mountain CF, Dresler CM. Regional lymph node classification for lung cancer staging. Chest 1997; 111:1718-1723
17.
Yoon SB, Kim B-T, Choi JY, et al. Role of PET in evaluating indeterminate solitary pulmonary nodule with CT [in Korean]. Korean J Nucl Med 1997; 31:83-89
18.
Heyneman LE, Herndon JE, Goodman PC, Patz EF Jr. Stage distribution in patients with a small (< or = 3 cm) primary nonsmall cell lung carcinoma: implication for lung carcinoma screening. Cancer 2001; 92:3051-3055
19.
Patz EF Jr, Rossi S, Harpole DH. Correlation of tumor size and survival in patients with stage IA non-small cell lung cancer. Chest 2000; 117:1568-1571
20.
Gail MH, Eagan RT, Feld R, et al. Prognostic factors in patients with resected stage I non-small cell lung cancer: a report from the Lung Cancer Study Group. Cancer 1984; 54:1802-1813
21.
Aoki T, Tomoda Y, Watanabe H, et al. Peripheral lung adenocarcinoma: correlation of thin-section CT findings with histologic prognostic factors and survival. Radiology 2001; 220:803-809
22.
Dwamena BA, Sonnad SS, Angobaldo JO, Wahl RL. Metastases from non-small cell lung cancer: mediastinal staging in the 1990s—meta-analytic comparison of PET and CT. Radiology 1999; 213:530-536
23.
Gould MK, Kuschner WG, Rydzak CE, et al. Test performance of positron emission tomography and computed tomography for mediastinal staging in patients with non-small-cell lung cancer: a metaanalysis. Ann Int Med 2003; 139:879-892
24.
Toloza EM, Harpole L, McCrory DC. Noninvasive staging of non-small cell lung cancer: a review of the current evidence. Chest 2003; 123 [suppl 1]:137S-146S
25.
Lardinois D, Weder W, Hany TF, et al. Staging of non-small-cell lung cancer with integrated positron-emission tomography and computed tomography. N Engl J Med 2003; 348:2500-2507
26.
Takamochi K, Yoshida J, Murakami K, et al. Pitfalls in lymph node staging with positron emission tomography in non-small cell lung cancer patients. Lung Cancer 2005; 47:235-242
Information & Authors
Information
Published In
Copyright
© American Roentgen Ray Society.
History
Submitted: December 1, 2005
Accepted: March 25, 2006
First published: November 23, 2012
Keywords
Authors
Metrics & Citations
Metrics
Citations
Export Citations
To download the citation to this article, select your reference manager software.