|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Departments of Thoracic Surgery (Drs. Ladas and Goldstraw) and Histopathology (Dr. Nicholson), Royal Brompton Hospital, London, UK; and Departments of General and Thoracic Surgery (Dr. Tiffet), Histopathology (Dr. Khaddage), and Nuclear Medicine (Drs. Prévot and Dubois), Teaching Hospital of Saint Etienne, France.
Correspondence to: Peter Goldstraw, FRCS, Department of Thoracic Surgery, Royal Brompton Hospital, Sydney Street, London, SW3 6NP, UK; e-mail: p.goldstraw{at}rbh.nthames.nhs.uk
| Abstract |
|---|
|
|
|---|
Design: Prospective study.
Setting: Royal Brompton Hospital, London, UK; and Hôpital Nord, Saint Etienne, France.
Methods: After thoracotomy, a total of 2 mL patent blue dye mixed with 1,600 µCi 99mTc-albumin or 99mTc-colloid was injected into each quadrant of lung tissue immediately surrounding the tumor. Routine lymphadenectomy was carried out. The first lymph nodes to stain blue or radioactive, if any, were considered SLNs.
Results: Twenty-four patients were evaluated. We successfully identified 17 SLNs in 13 patients (detection rate, 54.2%). Mean time from injection to identification of SLNs was 18 min (range, 5 to 30 min). In nine cases, the SLN was blue and radioactive, in six cases only blue, and in two cases only radioactive. The pathologic status of the SLN reflected the pathologic status of other nodes of the routine lymphadenectomy except one case of false-negative SLN (14%). Four SLNs were in N2 stations (23.5%).
Conclusions: The sentinel node mapping in NSCLC with blue dye and radioisotopic techniques is feasible, but the detection rate has to be improved. This technique is an accurate method of identifying the first node draining a tumor, although it is not yet sufficiently sensitive to have a role in reducing the extent of nodal dissection.
Key Words: intraoperative detection lung cancer sentinel lymph node
| Introduction |
|---|
|
|
|---|
| Materials and Methods |
|---|
|
|
|---|
Routine preoperative staging was performed, including clinical examination, chest radiography, CT scans of the chest and the upper abdomen, and cervical mediastinoscopy in case of mediastinal enlarged node (> 1 cm in the short axis). Two patients required mediastinoscopy for enlarged node in homolateral stations 2 and 4.
At thoracotomy, if necessary, a frozen section of the tumor was undertaken to ensure diagnosis, and then a total of 2 mL Patent Blue Dye-V (Laboratoire Guerbet; Roissy, France), mixed with 1,600 µCi 99mTc-albumin (99mTc Nano-colloid; Mallinckrodt; Bicester, UK) or of 99mTc-sulfur colloid (Nanocis; CIS Bio International; Nancy, France; reduced heating time protocol and filter) was injected in four divided doses into each quadrant of lung tissue immediately surrounding the tumor in a collapsed lung. The first two patients had only patent blue dye injected. The systematic nodal dissection before lung resection as described by Graham et al7 was carried out immediately after injection. In a previous study14 in melanoma, preoperative dynamic lymphoscintigraphy correctly identified 90% of the SLNs in < 15 min after a peritumoral injection. Dissection was carried out just after the injection because prolonged anesthesia remained a concern to us. SLN mapping by visual and handheld gamma probe guidance (Navigator intraoperative gamma probe; Tyco; Gosport, UK; or Neoprobe 2000; Ethicon Endo-surgery; Issy les Moulineaux, France) was performed every 5 min or 10 min or at each step of the lymphadenectomy. The first lymph nodes to stain blue or radioactive (three times the background) were considered to be the sentinel nodes. The exact station according the lymph node classification of Naruke et al10 was recorded. Each blue node, if any, was checked with the gamma probe, and the count per second was recorded after removal. We then verified the absence of residual radioactivity; a probe count < 1% of the most radioactive lymph node was regarded as background radiation. The SLNs were sent separately for pathologic examination. In the absence of blue coloration, this was recorded and the probe alone was used to detect the SLN, and vice versa. The planned intervention then continued according to standard practice. On the specimen, all lymph nodes (stations 11, 12, 13, and 14) were dissected from the surrounding parenchyma and connective tissue and checked for blue staining with the gamma probe.
The nodes were processed in their entirety, and sections were stained with hematoxylin-eosin. If required, sections of SLN were stained immunohistochemically using epithelial markers: MNF116 (1/80 dilution; Dako; Cambridge, UK) and Ber-EP4 (1/30 dilution; Dako) to confirm or refute involvement by tumor. Tissues obtained during radioguided SLN biopsy contained low residual radioactivity, and pathologic examination was delayed by 48 h.18
The number of patients in whom one or more SLNs were found was used to calculate the detection rate of the SLN biopsy. The unit of analysis was the patients, not the number of SLNs removed. A false-negative SLN was defined as a negative SLN with other nodes (non-SLN) positive for NSCLC. Sensitivity was calculated by the number of patients with a positive SLN divided by the number of patients with positive SLN plus the number of patients with a negative SLN but positive nodes in the node dissection specimen (false-negative SLN). Specificity was calculated by the number of patients with a negative SLN and negative nodes at resection divided by the number of patients with a negative SLN and negative nodes at resection plus the number of patients with a positive SLN but negative nodes at resection (false-positive SLN).
| Results |
|---|
|
|
|---|
The postoperative pathologic staging ranged from stage IA to stage IV. Eleven patients (45.8%) had pathologic stage I disease: T1N0M0 in 4 patients (16.6%) and T2N0M0 in 7 patients (29.1%). Seven patients (29.1%) had stage II disease: T1N1M0 or T2N1M0 in six patients (25%) and T3N0M0 in one patient (4.1%). Five patients (20.8%) had stage IIIA disease (T1N2M0 or T2N2M0). One patient (4.1%) had stage IV disease (T2N0M1). Median tumor size in the larger diameter was 38 mm (range, 11 to 60 mm).
The first two patients had only blue dye injected, while patient 15 had only radiotracer injected. No complications were related to the SLN biopsy; notably, no adverse effects of the tracers (patent blue dye or 99mTc) occurred.
Median time from injection to identification of the 17 SLNs successfully identified in 13 patients was 18.5 min (range, 5 to 30 min). In nine cases, we found the SLNs during the operation, but in four cases it was only identified on assessment of the resected specimen within the operating theater. The locations of the SLNs picked up on the resected specimen were station 12 (n = 3) and station 10 (n = 1). The intrapulmonary location, the background of the tumor, or the airway diffusion of the 99mTc made the intraoperative detection impossible in these cases.
At the least, one SLN was found in 13 patients; the detection rate of the method was 54.2% (Fig 1 ). We did not find an SLN in 11 cases (45.8%) [Table 1 ]. For two patients, no lymph node uptake of either dye or 99Tc was seen, but a translobar passage from segment 6 to segment 2 of the blue dye was noticed, and for one patient directly to a metastatic nodule (< 1 cm in diameter). For two patients, a blue lymphatic channel was evident in the direction of stations 7 and 2, but the anthracotic nodes hampered a colorimetric identification. In two negative cases, dissection of intense adhesions may have disrupted the lymphatic channels. In four patients, we identified no migration of both tracers, but we did notice blue sputum on bronchial aspiration and a high background of the bronchus. This suggested an airway injection of the tracers.
|
|
|
Among the 17 SLNs, 4 were noted to be mediastinal lymph nodes (23.5%). Two were in stations 5 and 6 for two LUL tumors, and two were in station 7 for the same LLL.
Finally, a translobar passage of the blue was noticed in three patients (patients 6, 23, and 24). In one case, the blue lymphatic channel led to a metastatic nodule.
| Discussion |
|---|
|
|
|---|
When we began this study, we sought to address the following concerns: (1) Is the SLN technique applicable to a deep malignancy such as the lung cancer? (2) Could the adjunction of a radiotracer increase the sensitivity? (3) Is the SLN concept applicable in NSCLC? One of the goals of this study was to see if SLN techniques were practicable in the clinical setting. We therefore made the decision to start the lymphadenectomy as soon as the tracers had been injected in order to avoid prolonged anesthesia. Our data show that the average migration time of 18 min (range, 5 to 30 min) was compatible with the course of the dissection adding only few minutes to the operation. Liptay et al19 reported an average migration time of 63 min (range, 23 to 170 min), while Sugi et al20 reported an average migration time of 135 min (range, 78 to 300 min). Melfi et al21 waited a mean of 1 h before performing any procedures after the intraoperative injection of the isotope.20 This technical difference probably contributes to the relatively low detection rate (54.2%) found in our study, compared to the 82% found by Liptay et al,19 the 64.3% found by Sugi et al,20 or the 96% found by Melfi et al21 using only radioisotope SLN mapping.
However, others factors could explain our low detection rate. We did not find any SLNs in 11 patients, and in 4 patients (36%) we noticed an airway diffusion of the two tracers. The site of injection was peritumoral on the edge of the tumor or in the bed of the tumor if a wedge resection had been required for frozen section to confirm diagnosis. This could have facilitated airway injection. Also, the radioactive background in the bronchus limited the probe detection. Others modalities of injection such as preoperative injection under CT scan control, intraoperative subpleural injection, or preoperative bronchoscopic injection have to be explored in future studies. Sugi and colleagues20 gained access to the tumor with a preoperative injection under imaging guidance in 14 patients. They reported several pneumothoraces, although they did not require chest drainage. The potential risk of complications such as pneumothorax, bleeding, and pleural tumor seeding does, however, remain a concern.
At the onset of this study, Little et al22 had already reported a detection rate of 47% using isosulfan blue dye mapping in lung cancer patients. The direct injection of both dye and 99mTc improves the detection rate. We therefore recommend, as did Schmidt et al,23 to use both tracers. In our experience, three patients in whom both tracers were injected had only blue SLNs, while only one patient had a non-blue but a radioactive SLN. Moreover, visual inspection found a blue translobar pathway on three occasions, in one case leading to a small metastatic nodule. This finding has not been described before.
We found direct mediastinal lymphatic drainage in four cases (23%), which is consistent with others series.1920222324 The intraoperative sentinel node mapping is an accurate method to identify "skip metastasis," although in our study routine systematic nodal dissection had removed the stations where the N2 SLNs were located.
Finally, our study provides supportive data for the SLN concept being applicable to lung cancer with a sensitivity in this study of 75% and a false-negative SLN rate of 14%. The sentinel node status is a reliable predictor of the mediastinal nodes histologic status. Sugi et al20 detected one false-negative SLN in nine patients using the dye technique. The false-negative SLN rate for Little et al22 and Schmidt et al23 was 0%, while it was 5% for Liptay et al19 and 3.8% for Melfi et al.21
If the future goal of this method is to avoid a futile mediastinal lymphadenectomy after a negative SLN on frozen section, one major limitation of the technique is its inability to easily detect SLNs located in stations 12 or 13. This required a dissection of the nodes on the specimen because the background of the tumor hampered the handheld gamma probe detection, and unfortunately the blue dye crosses nodes rapidly with the risk of picking up a satellite node rather than sentinel nodes.
In conclusion, intraoperative detection of the SLN with radioisotopic and blue dye techniques in peripheral NSCLC is feasible and safe. Further larger prospective studies are desirable to answer remaining questions about the site of injection, the timing of injection, and the detection methods in order to refine the technique. If SLN technology is to be useful to the surgeon, it should, ideally, allow the tracer to be injected prior to surgery. This would remove any conflict between the time required for migration and the need to avoid prolonged anesthesia. This would also allow the realization of a preoperative lymphoscintigraphy; in the 25% of patients in whom the SLN is identified in a mediastinal station, selective biopsy by mediastinoscopy prior to thoracotomy might allow the use of induction therapy with a possible improvement in prognosis.
| Footnotes |
|---|
Received for publication September 22, 2003. Accepted for publication September 2, 2004.
| References |
|---|
|
|
|---|
probe guidance. Br J Surg 2004;91,886-892[Medline]
This article has been cited by other articles:
![]() |
C. Nwogu, G. Fischer, D. Tan, M. Glinianski, D. Lamonica, and T. Demmy Radioguided Detection of Lymph Node Metastasis in Non-Small Cell Lung Cancer Ann. Thorac. Surg., November 1, 2006; 82(5): 1815 - 1820. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. L. Chen, D. M. Iddings, R. P. Scheri, and A. J. Bilchik Lymphatic Mapping and Sentinel Node Analysis: Current Concepts and Applications CA Cancer J Clin, September 1, 2006; 56(5): 292 - 309. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |