| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Yedikule Chest Diseases and Thoracic Surgery Education and Research Hospital, Department of Thoracic Surgery, 
stanbul, Turkey
Received 3 June 2007; received in revised form 21 November 2007; accepted 23 November 2007.
* Corresponding author. Address: Yedikule Gogus Hastanesi, Zeytinburnu, stanbul, Turkey. Tel.: +90 505 258 17 72; fax: +90 212 547 22 33. (Email: hmelek77{at}hotmail.com).
 |
Abstract |
|---|
 Top Abstract 1. Introduction2. Materials and methods3. Results4. DiscussionReferences |
|---|
Key Words: Lung cancer • Staging • Mediastinoscopy • PET
|
1. Introduction |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. DiscussionReferences |
|---|
Positron emission tomography (PET), which detects the biologic activities of tumor cells, has been increasingly used for several indications in NSCLC patients. Former studies have suggested the superiority of PET over CT in mediastinal lymphatic staging of NSCLC [4–6]. Due to reported excellent results in mediastinal staging, PET, as a non-invasive method, was expected to replace the role of the invasive mediastinoscopy in detection of MLN metastasis and decrease the use of mediastinoscopy [7,8].
We aimed to define the efficacy of PET in detection of MLN metastasis by comparing the mediastinal findings of PET with the histopathologic results obtained by either mediastinoscopy or thoracotomy in patients with clinically operable NSCLC.
|
2. Materials and methods |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. DiscussionReferences |
|---|
Routine biochemical laboratory tests, respiratory function tests, arterial blood gases tests and flexible bronchoscopy were performed in all patients enrolled in the study. Histopathologic diagnosis of all patients was NSCLC based on the biopsies obtained by bronchoscopy, transbronchial fine needle aspiration or transtracheal fine needle aspiration. Whole body scintigraphy, cranial CT or MRI was performed in some patients to exclude the presence of a distant metastasis.
Patients were suggested to have preoperative PET imaging in order to support primary diagnosis, search for presence of distant metastasis or mediastinal lymphatic metastasis and informed about the costs of the examination. Patients consented for the examination were included in the study and PET imaging was performed in four separate centers. CT-fusion PET scanner (Siemens PET ECAT ART + multi-slice CT fusion) was used for 45 patients, 6-slice multi-detector CT integrated high-resolution PET-CT scanner (Siemens Biograph LSO HI-RES PET/CT) was used for 125 patients. At least 4 h of fasting and adequate hydration was arranged before the examination. Fasting blood glucose levels of the patients were determined; and 296–703 MBq FDG18 was administered intravenously if values were below 150 mg/dl (Flouro-deoxy glucose). Patients were taken to a comfortable environment for 45–90 min and whole body PET imaging was carried out after this resting period. Images acquired to detect mediastinal lymphatic metastasis were visually evaluated. In the visual evaluation, mediastinal foci apart from normal biodistribution that show increased FDG uptake relative to background and adjacent tissues were considered as ‘suspected malignancy’, and maximum standard uptake value (SUVmax) more than 2.5 was considered as ‘mediastinal metastasis’. SUVmax values of primary tumor and the mediastinal lymph nodes, if involved, were recorded in the PET reports. Upper and lower paratracheal lymph nodes at the right and left sides were combined as right and left paratracheal region, subaortic and para-aortic lymph nodes were combined as aortopulmonary region, subcarinal, paraesophageal and pulmonary ligament nodes were combined as inferior mediastinal region in order to overcome the disadvantage of low spatial resolution of PET imaging. After the evaluation of the PET images, patients with the following criteria were excluded from the study:
Cervical mediastinoscopy was performed within 30 days after PET imaging in all patients to avoid the risk of time interval affecting the mediastinal metastasis results. Patients were placed supine and after induction of general anesthesia, skin was incised for 3–5 cm just above the jugular notch to reach the pretracheal fascia. Fascia was incised to introduce mediastinoscope. Stations number 2R, 4R, 2L, 4L and 7 were routinely explored. Extended mediastinoscopy was performed through the same incision if lymph nodes with pathologic size were detected at aortopulmonary region (station number 5–6) by thorax CT and involvement of these lymph nodes was confirmed by PET in the patients with tumor localized at the left upper lobe. Identified lymph nodes were sampled for histopathologic examination. Patients were referred to the oncology department for neoadjuvant or other non-surgical treatments when metastasis was detected in the MLN samples examined by experienced pathologists. If no metastasis was observed by mediastinoscopy, systematic sampling of lymph nodes by thoracotomy and pulmonary resection if possible, was performed.
Out of 173 patients, three patients were excluded from the study: one patient because the time interval between PET and mediastinoscopy exceeded 1 month (PET revealed findings of suspected metastasis in the colon and colonoscopy was performed), one patient that systematic mediastinal lymph node sampling could not be performed on and one patient who refused thoracotomy after mediastinoscopy. Remaining 170 patients formed the study group.
Necessary variables of thorax CT, PET, mediastinoscopy and thoracotomy to evaluate mediastinal lymphatic metastasis were recorded in a database and sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), false positivity (FP), false negativity (FN) and accuracy rates of each for detection of MLN metastasis were calculated with 95% confidence intervals. 95% confidence intervals for proportions are calculated according to the efficient-score method (corrected for continuity) described by Newcombe [9].
Data were analyzed using ‘SPSS for Windows 11.5 (Statistical Package for Social Sciences, SPSS Inc., Chicago, Illinois, USA)’ statistics software. Mann–Whitney U-test, chi-square test and Fisher's exact tests were used for comparison of the parameters. p Values less than 0.05 were accepted as significant.
|
3. Results |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. DiscussionReferences |
|---|
Localization of the tumor was right lung in 85 patients (50%), left lung in 85 patients (50%) and upper lobes in 115 patients (58 right + 57 left). Tumor was located in the central part in 26% of the patients.
Thorax CT revealed MLNs with pathologic size in 57 patients, and no enlarged MLN was observed in 113 patients. Metastasis was detected by mediastinoscopy or thoracotomy in 15 of 113 patients with no findings of MLN metastasis, (false negative n = 15/113) and metastasis could not be shown by mediastinoscopy or thoracotomy in 21 of 57 patients that had suspected MLN metastasis by thorax CT (false positive n = 21/57). These findings indicate that thorax CT has a sensitivity of 70% (95% CI: 58–83), and a specificity of 82% (95% CI: 76–89), PPV of 63% (95% CI: 51–76), NPV of 86% (95% CI: 80–93), accuracy rate of 78% for detection of MLN metastasis.
PET reported MLN involvement in 101 patients and no MLN involvement in 69 patients. Metastasis was detected by mediastinoscopy or thoracotomy in 13 of 101 patients reported as no MLN involvement by PET (false negative n = 13/101). MLN metastasis was demonstrated by mediastinoscopy in eight patients (number 4L in two patients, 4R in two patients, both 4R and 7 in two patients and 5 in one and 7 in one patient) or thoracotomy in five patients (number 8 in three patients, 5 in one patient and 4R in one patient). MLN metastasis was detected by thoracotomy or mediastinoscopy in 38 of 69 patients that PET suggested MLN metastasis, and no metastasis was detected in the remaining 31 patients (false positive n = 31/69). These findings indicate that PET has a sensitivity of 74% (95% CI: 63–86) and a specificity of 73% (95% CI: 66–82), PPV of 55% (95% CI: 43–67), NPV of 87% (95% CI: 80–94), accuracy rate of 74% for detection of MLN metastasis.
Mediastinoscopy sampled a total of 687 stations (mean 4.04, range 2–7). In 127 of the patients MLN metastasis was not detected, whereas mediastinal lymphatic metastasis was detected in 43 patients. Thoracotomy was performed in the patients negative for MLN metastasis. During thoracotomy lobectomy was performed in 94 patients (74%), sleeve lobectomy was performed in nine patients (7%) pneumonectomy was performed in 22 patients (17.3%) together with lymph node dissection; explorative thoracotomy and only lymph node sampling was performed in two patients (1.5%). Thoracotomy revealed metastasis in lymph node number 8 in four patients, in 4R in two patients, in number 7 in one patient and in number 6 in one patient, among 127 patients who were supposed to be negative in mediastinoscopy (mediastinoscopy false negative n = 8/127). These findings indicate that mediastinoscopy has a sensitivity of 84% (95% CI: 74–94), specificity of 100%, PPV of 100%, NPV 94% (95% CI: 89–98), accuracy rate of 95% for detection of MLN metastasis. When only stations of 2R, 2L, 4L, 4R and number 7 that can be reached by standard cervical mediastinoscopy are included in the calculation, standard cervical mediastinoscopy has a sensitivity of 93% (95% CI: 86–100), specificity of 100%, PPV of 100%, NPV of 98% (95% CI: 95–100), FN 6% (95% CI: 2–11), FP 0% and accuracy rate of 98%.
N2 or N3 was detected in 51 of 170 patients (30%) clinically assessed as operable after mediastinoscopy and thoracotomy.
MLN metastasis was detected in 22 of 58 adenocarcinoma patients (38%), and 17 of 79 squamous carcinoma patients (21%). Incidence of MLN metastasis was significantly higher in adenocarcinoma patients (p < 0.05).
When PET results were compared to postoperative pathologic examination results, staging of PET for lymph node metastasis was accurate in 93 patients (55%), lower in 33 patients (20%), higher in 44 patients (25%). When nodal stages of PET were compared to histopathologic nodal stages, MLN was detected in six of 80 patients (7.5%) staged as N0 by PET, in seven of 21 patients (33%) staged as N1 by PET, in 38 of 69 patients (55%) staged as N2 or N3 by PET (Table 1 ). MLN metastasis was significantly higher in the patients reported as N1 by PET compared to patients reported as N0 by PET (p < 0.05).
|
Out of 31 false positive MLN reported by PET, 22 were reactive hyperplasia, three were chronic granulomatous inflammation, and six were necrotizing granulomatous inflammation. One of these patients with MLNs had a history of tuberculosis and one patient had a history of diabetes. Tumor localization in patients with false positive PET reports was peripheral in 19 patients (61%) and central in 12 patients (39%). Histopathologic type of tumor was squamous carcinoma in 14 patients (45%) and adenocarcinoma in 10 patients (32%). MLNs with pathologic diameter were detected in 14 by thorax CT among these patients.
Tumor localization was peripheral in 82 and central in 19 of total 101 patients reported as negative for mediastinal metastasis by PET. Histopathologic type of tumor in this group was squamous carcinoma in 52 patients, adenocarcinoma in 34 patients and other types in the remaining 15 patients. MLN metastasis was detected in 13 (13%) of these patients. Localization of tumor in this false negative group of patients was peripheral in nine patients (9/82; 11%) and central in four patients (4/19; 21%). False negative rates for central tumors were higher than peripheral tumors although this difference was not statistically significant (p > 0.05). Histopathologic type of tumors in patients with false negative results after mediastinal examination was adenocarcinoma in nine patients (69.2%), squamous carcinoma in four patients (30.8%) demonstrating a statistically significant difference (p < 0.05).
When stepped analysis is performed to thorax CT and PET findings, no involvement was detected by PET in 12 patients out of 57 patients that had pathologic lymph nodes detected by thorax CT, and metastasis was detected in five of these 12 patients by mediastinoscopy or thoracotomy. Mediastinoscopy or thoracotomy showed no metastasis in 14 patients out of 45 patients that thorax CT detected pathologic MLNs and reported as metastasis by PET (Fig. 1 ). These results indicate that PET has a sensitivity of 86% (95% CI: 75–97), specificity of 33% (95% CI: 13–53), PPV of 68% (95% CI: 55–82), NPV of 59% (95% CI: 30–86), accuracy rate of 66% for the patients reported suspected MLN metastasis by CT examination.
|
|
4. Discussion |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. DiscussionReferences |
|---|
Thorax CT is the most readily and frequently performed non-invasive method but metastasis cannot be found in 40% of the lymph nodes denoted as pathologic size in thorax CT, and 15–20% of the MLNs that have not reached to pathologic size in CT may be metastatic [3]. Thus efficacy of thorax CT is regarded as inadequate for detection of MLN metastasis [2,3,10].
Mediastinoscopy has the best sensitivity and specificity rates but is an invasive method with the disadvantages of morbidity and mortality, though very low [11]. Mediastinoscopy is reported to have sensitivity above 80% and NPV above 90% in almost all of the studies in the literature [12]. Although reported morbidity and mortality rates are as low as 1% and 0.008%, respectively [11], the disadvantage of invasiveness persists.
Imaging-guided fine needle biopsies have gained popularity in the hope of replacing mediastinoscopy in recent years. But these methods can be used only in some selected patients and also contain some degree of morbidity albeit low.
Consequently, none of these modalities seem to be ideal for mediastinal staging of NSCLC, and various combinations of these methods are employed for a reliable mediastinal staging. A sensitive, specific and non-invasive method to demonstrate MLN metastasis sounds very attractive. But the selected method must be applicable to all patients, have a high sensitivity for the patients with no suspected metastasis, high specificity for the patients with suspected metastasis and a high accuracy rate for all patient groups.
When these objectives were considered, PET was introduced with very encouraging results. Studies comparing the efficacy of thorax CT and PET to demonstrate MLN metastasis implemented between 1990 and 2000 proved PET to be superior to thorax CT [3,5,13,14]. Accuracy of PET to demonstrate MLN metastasis was reported as 93–100% in the study of Scott et al. [14] and Sasaki et al. [15] in 1996. PET was believed to replace mediastinoscopy as a non-invasive method with its high accuracy rate. Studies performed in the following years did not confirm these early good results. For example, in a study by Gonzalez-Stawinski et al. [16] published in 2003 including 202 patients, PET was reported to have a sensitivity of 64%, specificity of 77%, PPV of 44%, and NPV of 88% to detect MLN metastasis and in the study of Reed et al. [17] published in 2003 on 302 patients sensitivity was 61%, specificity was 84%, PPV was 56%, and NPV was 87%. Toloza et al. [3], in his meta-analysis including 1045 patients in 2003 reported PET to have a sensitivity of 84%, specificity of 89%, PPV of 79%, and NPV of 93% to demonstrate MLN metastasis. In our study PET has shown a sensitivity of 74%, specificity of 73%, PPV of 55%, NPV of 87% accuracy rate of 74%, which proved the overall efficacy of PET for mediastinal staging to be low compared to mediastinoscopy.
This difference in support of mediastinoscopy compared to PET, is present despite the disadvantage of mediastinoscopy not able to reach all mediastinal stations. During standard cervical mediastinoscopy, due to technical reasons, mediastinoscope cannot reach the lymph nodes in aortic and inferior mediastinal regions. Still, accuracy rate of mediastinoscopy to reveal MLN metastasis is higher than PET. When only lymph nodes in the range of mediastinoscopy are considered (station numbers 2,4 and 7), standard cervical mediastinoscopy provides a sensitivity of 93%, specificity of 100%, PPV of 100%, NPV of 97.7% and accuracy rate of 98% and the difference between the accuracy rate of PET is further increased.
The major disadvantage of PET to demonstrate MLN metastasis is its inability to discriminate non-tumoral causes of FDG uptakes from tumor metastasis. PET can also yield positive results for benign conditions such as hyperplasia that usually develops mostly secondary to tuberculosis but also to sarcoidosis, coccidioidomycosis, histoplasmosis and pneumonia [4,18]. This figure is reported to be at least 16–55% in the literature [3,4,16]. We determined the false positivity rate of PET as 26% in our study. Reactive hyperplasia was observed in most of the lymph nodes with false positive results (71%). Thus, as generally accepted, we believe that positive PET results must be histopathologically confirmed [4,10]. Positive PET results also can be used as a guide to sampling the mediastinal lymph nodes with suspicious metastasis by invasive methods [19].
Although false positive rates are high, false negative rates of PET are reported to be lower in the literature [4,7]. Thus, PET study reporting negative results for mediastinal metastasis are regarded as a substitute for invasive staging methods [7,8,20]. However major deficiencies of the studies that report very low false negative rates can be listed as implementation of the study in terms of mediastinal lymph node stations, inclusion of patients other than NSCLC, i.e. inclusion of benign diseases, and lower than expected N2 rates of the selected patient groups. Studies based on MLN-stations may lead to some bias. For example, if only one false negative is obtained by PET from four mediastinal regions performed to detect MLN metastasis, this will be recorded as three true negatives and one false negative in a station-based study. But if the same situation is reconsidered based on the patient, PET has to be recorded as only one false negative. When results of our study are analyzed for stations, NPV is 96% and PET accuracy rate is 85%, but if patient-based analysis is performed NPV decreases to 87%, and accuracy rate decreases to 74%. Presence of N2 in the patients clinically considered to be operable is reported to be between 30% and 37% [3]. In our study this ratio was determined as 30%. Inclusion of less than expected patients with N2 and a patient population of predominantly early stages would certainly decrease false negative rates. Thus patient inclusion criteria of the study should be carefully reviewed.
Although overall efficacy of PET is lower compared to mediastinoscopy to demonstrate MLN metastasis, we searched for patient groups where PET could potentially replace mediastinoscopy through some subgroup analysis. In these analyses:
We believe that early results of PET reported for mediastinal staging in the initial studies would be decreased to reasonable levels just like the excellent results of thorax CT for mediastinal staging that were reported in the first years regressed to rational levels by time. Although PET sounds attractive, it does not provide enough accuracy in mediastinal staging of NSCLC compared to mediastinoscopy. Mediastinoscopy still continues to be the gold standard for the mediastinal staging of NSCLC. To obtain better results in mediastinal staging of NSCLC, all available invasive and non-invasive methods should be used together.
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstract1. Introduction2. Materials and methods3. Results4. DiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Bille, E. Pelosi, A. Skanjeti, V. Arena, L. Errico, P. Borasio, M. Mancini, and F. Ardissone Preoperative intrathoracic lymph node staging in patients with non-small-cell lung cancer: accuracy of integrated positron emission tomography and computed tomography Eur. J. Cardiothorac. Surg., September 1, 2009; 36(3): 440 - 445. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. M. Carnochan and W. S. Walker Positron emission tomography may underestimate the extent of thoracic disease in lung cancer patients Eur. J. Cardiothorac. Surg., May 1, 2009; 35(5): 781 - 785. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Hwangbo, S. K. Kim, H.-S. Lee, H. S. Lee, M. S. Kim, J. M. Lee, H.-Y. Kim, G.-K. Lee, B.-H. Nam, and J. I. Zo Application of Endobronchial Ultrasound-Guided Transbronchial Needle Aspiration Following Integrated PET/CT in Mediastinal Staging of Potentially Operable Non-small Cell Lung Cancer Chest, May 1, 2009; 135(5): 1280 - 1287. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
