- Email: [email protected]
Systematic node dissection by VATS is not inferior to that through an open thoracotomy: A comparative clinicopathologic retrospective study
Systematic node dissection by VATS is not inferior to that through an open thoracotomy: A comparative clinicopathologic retrospective study Atsushi Watanabe, MD, Tetsuya Koyanagi, MD, Hisayoshi Ohsawa, MD, Tohru Mawatari, MD, Shinji Nakashima, MD, Noriyuki Takahashi, MD, Hiroki Sato, MD, and Tomio Abe, MD, Sapporo, Japan
Background. Major pulmonary resection with systematic node dissection (SND) for early lung cancer by video-assisted thoracic surgery (VATS) is performed in many institutes, but the feasibility of SND for early lung cancer by VATS remains controversial. The aim of this study was to elucidate the feasibility and safety of SND by VATS. Methods. Three hundred fifty patients with clinical stage I lung cancer who underwent pulmonary major resection with SND between 1998 and 2003 were enrolled in this study. Of these patients, 191 (VATS group) underwent pulmonary resection with SND by VATS; 159 patients (open thoracotomy [OT] group) did so through anterolateral thoracotomy. The clinical and pathologic data, including the number of dissected nodes in each nodal station, of the 2 groups were compared to evaluate the feasibility of SND by VATS. Results. Pathologic data showed that, in the VATS group, more patients had adenocarcinoma (P = .0078) and fewer patients had advanced factors than the OT group. The greatest tumor diameter was 24.5 mm and 29.6 mm in the VATS group and OT group, respectively (P < .0001). The total number of mediastinal nodes dissected in right upper lobectomy plus right middle lobectomy (RUL+RML), right lower lobectomy (RLL), left upper lobectomy (LUL), and lower left lobectomy (LLL) also did not differ between the 2 groups. The total number of mediastinal nodes dissected in RUL+RML, RLL, LUL, and LLL was 19.7 in the VATS group versus 22.0 in the OT group (P = .122), 23.4 versus 21.0 (P = .241), 14.8 versus 17.5 (P = .123), and 18.8 versus 15.8 (P = .202), respectively. The number of dissected nodes in each nodal station in RUL+RML, RLL, LUL, and LLL was similar between the 2 groups. Operative mortality, morbidity, or recurrence did not differ between the 2 groups. Conclusions. With regard to the number of dissected nodes, SND by VATS was not inferior to that of OT. SND by VATS is technically feasible and safe, and seems acceptable for clinical stage I lung cancer. (Surgery 2005;138:510-7.) From the Department of Thoracic and Cardiovascular Surgery, Sapporo Medical University School of Medicine
ALTHOUGH the therapeutic effect of node dissection (ND) for primary lung cancer remains controversial, the diagnostic effect is clear. Lobectomy with mediastinal ND for primary lung cancer was first reported in 1993 by Martini and associates.1 A video-assisted thoracoscopic surgery Accepted for publication April 18, 2005. Reprint requests: Atsushi Watanabe, MD, Department of Thoracic and Cardiovascular Surgery, Sapporo Medical University School of Medicine, South 1, West 16, Chuo-ku, Sapporo 0608543, Japan. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2005 Mosby, Inc. All rights reserved. doi:10.1016/j.surg.2005.04.005
510 SURGERY
(VATS) lobectomy with mediastinal ND for primary lung cancer was first performed in 1995 by McKenna and associates.2 They showed that the VATS lobectomy for bronchogenic carcinoma appears to be a safe operation, with the same survival rate as that expected for a lobectomy done by thoracotomy.3 In this study, we sought to elucidate the feasibility and safety of SND by VATS for stage I primary non–small cell lung cancer (NSCLC). MATERIAL AND METHODS Three hundred fifty patients with clinical stage I NSCLC who underwent major pulmonary resection with SND between 1998 and 2003 were
Surgery Volume 138, Number 3
enrolled in this study. These patients were divided into 2 groups (VATS group and open thoracotomy [OT] group) on the basis of the approach to the pleural space. One hundred ninety-one patients, the VATS group, underwent pulmonary resection with SND by VATS, whereas one hundred fifty-nine patients, the OT group, did so through an anterolateral thoracotomy. Clinical and pathologic data, including the number of dissected nodes in each nodal station, of the 2 groups were compared. Preoperative management. All patients who underwent major pulmonary resection gave written informed consent for performing an ND before the operation; those who refused were excluded from the study. The consenting patients underwent preoperative staging and pulmonary function assessment. The preoperative workup was standardized for the staging and consisted of routine chest roentgenography, computed tomographic (CT) scanning of the thorax and the abdomen, CT scanning or magnetic resonance imaging of the brain, bone scintigraphy, and bronchoscopy. Neither mediastinoscopy nor positron emission tomography were performed for clinical stage I primary lung cancer. If the patients had cardiac symptoms such as chest oppression, pain, or palpitations, a cardiac evaluation was performed. Surgical indications for VATS major pulmonary resection with SND. We started VATS lobectomy with SND for primary lung cancer in January 1997 at our institute. The inclusion criteria for VATS lobectomy were as follows: clinical stage IA lung cancer; neither incomplete fissure area nor extensive pleural adhesion found on the preoperative chest CT or intraoperatively; and the ability to physiologically tolerate 1--lung ventilation. Our inclusion criteria for VATS lobectomy have changed with our knowledge of and ability to perform the procedure. Before December 2000, we passively recommended the VATS approach with SND as a first-line procedure because of our inexperience; 2 patients refused to undergo VATS lobectomy and received a lobectomy through an anterolateral thoracotomy. After January 2001, we recommended the VATS approach as a first-line procedure because of experience in performing the procedure. After January 2002, inclusion criteria were extended as follows: clinical stage I patients whose greatest tumor diameter was 40 mm or less and the ability to physiologically tolerate 1--lung ventilation. Extensive pleural adhesion and incomplete fissure area without emphysematous change in the lung found on the preoperative chest CT or intraoperatively were no longer con-
Watanabe et al 511
sidered inclusion criteria. Therefore, the rate of VATS major pulmonary resections to all major pulmonary resections for clinical stage I primary lung cancer has gradually increased every year: namely, a rate of 5.3% (2/38) in 1997 slowly increased to 81.6% (71/87) by 2003. Furthermore, after January 1999, patients were registered for a VATS segmentectomy if they had the following criteria: adenocarcinoma with greatest tumor diameter of 20 mm or less and 50% or more of the rate of ground glass area to the greatest tumor area on the thin-slice CT scanning, and peripheral squamous cell carcinoma with greatest tumor diameter of 20 mm or less. We performed SND for all clinical stage I primary lung cancer patients without intraoperative assessment of the frozen section of any nodes. Criteria for intraoperative conversion of the VATS procedure to OT. Before December 2001, the criteria for intraoperative conversion of the VATS procedure to OT were as follows: uncontrolled bleeding, dense pleural adhesion, or the need to extend the procedure such as bronchoplasty or pulmonary artery plasty. After January 2002, dense pleural adhesion without emphysematous change in the lung was excluded from the criteria. Unfortunately, 15 patients underwent conversion of VATS lobectomy to lobectomy by OT. The conversion rate was 15/206 (7.3%), the reason being intraoperative uncontrolled bleeding in 6, dense pleural adhesion in 5, extended resection in 3, and an intraoperative cardiac event (ventricular fibrillation) in 1. One hundred ninety-one patients excluding these 15 patients were enrolled in this study as the VATS group. Surgical technique of VATS lobectomy. General anesthesia with selective lung ventilation was performed with the use of a double lumen endotracheal tube. Patients were placed in a decubitus position on an operating table. In the VATS group, 2 thoracoports were placed in the sixth or seventh intercostal space (ICS) on the anterior axillary line and in the seventh or eighth ICS on the posterior axillary line; an anterolateral minithoracotomy (37 cm) was made in the fourth ICS for an upper lobectomy or in the fifth ICS for a middle or lower lobectomy. The site of the minithoracotomy was covered with a Lap Protector Mini (Hakko Medical Co, Tokyo, Japan). Pulmonary vessel management was performed by using forceps or scissors for conventional surgery. These vessels were divided after double ligation with a silk suture or after clipping at the proximal and distal portions without the use of end-stapler devices. In recent cases, the basal segmental artery (with or without
512 Watanabe et al
superior segmental pulmonary artery of the lower lobe) or inferior pulmonary vein were divided by the use of end-stapler devices (Endo-GIA; United States Surgical Corp, Norwalk, Conn; and ETS-45; Ethicon, Cincinnati, Ohio). All the bronchial management was performed with the use of endstapler devices. A lobe or segment with NSCLC was extracted in all cases in a specimen bag made of vinyl that we designed to avoid chest wall contamination. On the right side, an upper mediastinal ND usually was performed without the transection of an azygos vein. The nodes were held with long forceps for conventional thoracic surgery. During the procedure, the superior vena cava or vagus nerves were compressed with a cherry dissector (Ethicon); vagus nerves and azygos veins were sometimes taped and retracted posterolaterally with a silk suture if necessary. When dissecting the most upper mediastinum, we avoided using an electrocautery, so as not to damage the right recurrent laryngeal nerve. Small vessels and small branches of the vagus nerve were clipped with a premium surgiclip (United States Surgical Corp) and then transected, or transected with an electrocautery. Secondary ND at the lower posterior mediastinum (subcarinal and paraesophageal) was performed after the right main bronchus was taped with a silk suture, which was brought out of the thorax through a minithoracotomy and retracted anterolaterally to gain a good operative view. During the dissection, the bronchial artery was clipped with endoscopic clips or transected by using electrocautery; the esophagus was compressed with a cherry dissector or the lung was retracted anteriorly with the use of an ENDLUNG (United States Surgical Corp) if necessary. Nodes in the pulmonary ligament and paraesophageal site were dissected only when a lower lobectomy was performed. On the left side, at first, para-aortic and subaortic nodes were dissected. After that, pretracheal and tracheobronchial nodes were dissected. During the dissection, the left main pulmonary artery was compressed to the caudal side with a cherry dissector, or the aortic arch was pushed to the cranial side. The Botallo’s duct was simply transected after confirming the absence of intraductal blood flow if a good operative field was not gained despite the compression as mentioned above. The dissection of the lower posterior mediastinal nodes and nodes in pulmonary ligaments were performed in the same method as that on the right side. After the procedure, 1 chest tube was placed in the pleural cavity and the wounds were closed.
Surgery September 2005
Node mapping. We used node mapping, which was reported by Naruke et al.4 They assigned a number to each node station, namely, #1 to superior mediastinum, #2 to paratracheal, #3 to pretracheal, #4 to tracheobronchial, #5 to subaortic, #6 to para-aortic, #7 to subcarinal, #8 to paraesophageal, #9 to pulmonary ligament, #10 to hilar, and #11 to interlobar. We defined #1 and #2 together as superior mediastinal nodes (#1+2) for convenience. Postoperative adjuvant therapy. Basically, we performed adjuvant chemotherapy only on the patients who had recurrence even if they were patients with pathologic n1 or n2 disease. We performed re-resection or radiotherapy on the patients with locoregional recurrence when possible. Parameters. We reviewed demographic data, clinical data (stage, tumor location, procedure), and pathologic data (stage, greatest tumor size, number of dissected nodes). In relation to the number of dissected nodes, we counted the number of large nodes that were cut during the ND after excluding small pieces of nodes that were cut the same way. We defined the dissected node number as the sum of the number of nodes that were not cut and the number of large nodes that were cut. Furthermore, we compared the number of dissected nodes in each nodal station, postoperative morbidity and mortality, and recurrence between the 2 groups. Statistical analysis. All continuous variables are expressed as means ± SD. Differences between the 2 groups were assessed by means of the Student t test after the assurance of homogeneity by the Levene test. Categorical data were compared by using the chi-square test or the Fisher exact test. Survival rates were assessed by using the KaplanMeier method; the curves were compared by using a log-rank test. All reported probability values are 2-tailed, and P values of less than 0.05 were considered statistically significant. Statistical analyses were performed with the use of SPSS 10.0 software (SPSS, Inc, Chicago, Ill). RESULTS Clinical data. Clinical data are summarized in Table I. Although the groups did not differ in age (P = .578), tumor location (P = .149), surgical procedure (P = .3843), or concomitant disease, the male ratio (P = .001) and clinical T factor (P < .0001) were lower in the VATS group than the OT group. The 2 groups did not differ with regard to the occurrences of preoperative concomitant pulmonary diseases.
Watanabe et al 513
Surgery Volume 138, Number 3
Table I. Demographic and clinical data Procedure Lobectomy (yes/no) Lobectomy Segmentectomy Bilobectomy Pneumonectomy Concomitant disease Tuberculosis COPD Pneumoconiosis
168/23 168 17 6 0
145/14 145 6 7 1
8 26 12
6 28 17
Table II. Pathologic data .161 .3843
.8437 .3727 .1728
VATS, Video-assisted thoracic surgery; OT, open thoracotomy; RUL, right upper lobectomy; RML, right middle lobectomy; RLL, right lower lobectomy; LUL, left upper lobectomy; LLL, left lower lobectomy; COPD, chronic obstructive pulmonary disease.
Pathologic data. Pathologic data showed that more patients had adenocarcinoma (P = .0078) and fewer patients had pathologically advanced factors in the VATS group than the OT group. The greatest tumor diameter was 24.5 mm and 29.6 mm in the VATS group and OT group, respectively (P < .0001; Table II). The number of dissected nodes. The 2 groups did not differ with regard to the total number of dissected nodes in RUL+RML, RLL, LUL, and LLL (Tables III and IV). The 2 groups also did not differ with regard to the total number of mediastinal nodes dissected in RUL+RML, RLL, LUL, and LLL. The number of dissected nodes in each nodal station in RUL+RML, RLL, LUL, and LLL was similar between the 2 groups. However, the number of dissected nodes in #9 in RUL+RML was greater in the OT group than the VATS group (0.3 vs 0.7, P = .028), and the number of dissected nodes in #10 in LUL was greater in the VATS group than the OT group (3.6 vs 2.5, P = .025). Furthermore, we compared the 2 groups with regard to the total number of dissected nodes and total number of mediastinal nodes dissected in the subgroups: namely the male group, adenocarcinoma group, GDT < 30 mm group, and 0 of pathologic node status group. There were no differences between the 2 groups (Table V). Cut nodes. In the recent 100 cases (50 cases in each group) of this series, the proportions of cut nodes to dissected nodes of #1+2 or #7 between the 2 groups were evaluated and compared by the histopathologists in each group. The definition of cut nodes was nodes with torn capsule in the hematoxylin-eosin staining (original magnification: 320). The proportions of crushed nodes were not evaluated because of the difficultly in estimating the existence of crush in nodes. The pro-
GTD (mm) Histology Ad/non-ad Ad Sq Others p-T 1/>1 1 2 3 4 p-N 0/1,2 0 1 2 p-stage I/II,III I (A/B) II (A/B) III (A/B)
VATS (n = 191)
OT (n = 159)
P value
24.5 ± 11.1
29.6 ± 14.0
<.0001
143/48 143 35 13
98/61 98 43 18
.0078 .0278
142/49 142 41 3 5
93/66 93 52 6 8
.0017 .016
159/32 159 12 20
122/37 122 22 15
.1271 .059
154/37 154 (125/29) 14 (5/9) 23 (17/6)
113/46 113 (78/35) 22 (12/10) 24 (13/11)
.0363 .074
VATS, Video-assisted thoracic surgery; OT, open thoracotomy; GTD, greatest tumor diameter; Ad, adenocarcinoma; Sq, squamous cell carcinoma; p, pathologic.
portions of #1+2 and #7 in the VATS and OT groups were 37/276 versus 30/261 (P = .5897) and 47/348 and 42/354 (P = .6916), respectively. There were no significant differences between the groups. Operative mortality or morbidity. Neither operative mortality nor morbidity differed between the 2 groups. Two patients in the VATS group died of pneumonia and interstitial pneumonia; 3 patients in the OT group died of pneumonia (n = 2) and interstitial pneumonia (n = 1). The groups did not differ in the incidence of postoperative chylothorax or recurrent laryngeal nerve injury, which is caused by SND (Table VI). Recurrence. The incidence of recurrence and requirement of postoperative chemotherapy, radiotherapy, or re-resection are summarized in Table VII. There were no differences between the 2 groups (P = .6437). Furthermore, recurrence-free survival curves (Figure) showed that the 5-year rates were 76.3% and 71.4% in the VATS group and OT group, respectively, and that the curves did not differ between the 2 groups (P = .884). DISCUSSION Although the therapeutic effect of extensive mediastinal lymphadenectomy has been the debated
514 Watanabe et al
Surgery September 2005
Table III. Number of dissected nodes in right side Upper + Middle VATS (n = 78) TNDN TNMND 1+2 3 4 7 8 9 10 11s 11i
30.7 19.7 5.5 4.7 2.4 6.4 2.1 0.3 2.0 2.2 1.1
± ± ± ± ± ± ± ± ± ± ±
11.1 8.4 4.5 2.9 2.7 4.3 1.3 1.0 3.0 2.5 2.1
OT (n = 52) 31.1 22.0 5.2 4.9 2.8 6.9 2.4 0.7 1.2 2.3 0.8
± ± ± ± ± ± ± ± ± ± ±
13.4 8.6 4.1 4.5 2.9 4.1 1.9 1.0 2.0 3.3 2.0
Lower P value .831 .122 .767 .730 .544 .559 .578 .026 .099 .775 .421
VATS (n = 49) 33.8 23.4 5.7 5.1 2.8 7.9 2.4 0.8 1.4 0.6 3.0
± ± ± ± ± ± ± ± ± ± ±
13.1 10.3 4.0 3.3 2.5 5.3 1.8 1.0 2.1 1.3 2.4
OT (n = 41) 30.9 21.0 4.4 5.8 2.2 6.2 2.0 1.2 1.2 1.0 2.0
± ± ± ± ± ± ± ± ± ± ±
12.1 8.3 3.8 5.0 2.5 4.2 1.2 1.0 1.7 1.5 2.0
P value .279 .241 .128 .416 .298 .107 .424 .290 .685 .156 .063
VATS, Video-assisted thoracic surgery; OT, open thoracotomy; TNDN, total number of dissected nodes; TNMND, total number of mediastinal nodes dissected.
Table IV. Number of dissected nodes in left side Upper VATS (n = 48) TNDN TNMND 3 4 5 6 7 8 9 10 11
28.0 14.8 2.1 2.0 2.2 2.2 4.3 2.0 2.5 3.6 3.0
± ± ± ± ± ± ± ± ± ± ±
10.2 8.1 2.5 1.9 1.9 2.2 4.2 1.6 1.0 2.4 3.2
Lower
OT (n = 40) 28.1 17.5 1.9 2.3 2.7 2.6 4.9 2.3 2.5 2.5 2.4
± ± ± ± ± ± ± ± ± ± ±
11.3 8.2 2.5 3.0 2.3 2.8 4.1 2.3 1.5 2.1 2.0
VATS (n = 16) .964 .123 .694 .537 .275 .442 .477 .637 .975 .025 .260
29.7 18.8 2.0 1.7 2.6 2.5 5.8 3.0 2.8 2.3 3.3
± ± ± ± ± ± ± ± ± ± ±
7.1 7.2 2.4 1.6 2.3 3.9 4.9 2.1 2.7 2.3 3.2
OT (n = 26) 25.5 15.8 2.9 1.7 2.0 1.8 4.0 1.3 2.1 2.3 3.0
± ± ± ± ± ± ± ± ± ± ±
11.8 7.2 2.9 1.9 1.8 1.6 3.2 1.4 1.9 2.0 2.9
.212 .202 .293 .940 .363 .425 .157 .063 .389 .978 .777
VATS, Video-assisted thoracic surgery; OT, open thoracotomy; TNDN, total number of dissected nodes; TNMND, total number of mediastinal nodes dissected.
issue,5,6 the diagnostic effect is clear. Some reports7,8 indicate that complete mediastinal ND is important for accurate staging as well as overall survival. A large, prospective randomized multicenter trial of mediastinal node sampling versus complete lymphadenectomy during pulmonary resection in patients with N0 or N1 (less than hilar) NSCLC is currently under way (American College of Surgeons Oncology Group, ACOSOG protocol Z0030). In Japan, we generally perform SND through OT as a standard operation for clinical stage I primary NSCLC. Some institutes refuse to perform SND by VATS; namely, ND is either not performed or only node sampling is performed. Procedure of SND. Certainly, performing ND by VATS for primary lung cancer seems difficult. During dissection of upper mediastinal nodes and subcarinal nodes, exposing the operative field
requires great effort. Kaseda and colleagues9 reported the following ND technique: During dissection at the subcarina, the esophagus and bronchus are compressed by a mini retractor (END MINI-RETRACT, United States Surgical Corp). They reported recently at a conference in Japan they made a new device with a long wound opener for gaining a wide intraoperative field during mediastinal ND. During the dissection of the upper mediastinum, an azygos vein was ligated with a suture and transected. The vagus nerve was taped with silk thread, which was brought outside the thorax with an ENDO CLOSE (United States Surgical Corp), and the right main bronchus was then retracted laterally with the use of an END MINI-RETRACT to obtain a good operative view. We did not transect the azygos veins if we could obtain a good thoracoscopic view and operative field using techniques such as lateral retraction of
Watanabe et al 515
Surgery Volume 138, Number 3
Table V. Number of dissected nodes in subgroups Right side
Male N TNDN TNMND Ad N TNDN TNMND GTD < 30 mm N Mean (mm) TNDN TNMND p-n0 N TNDN TNMND
Left side
VATS
OT
P value
VATS
OT
P value
63 33.8 ± 13.1 22.5 ± 9.1
66 32.8 ± 13.1 22.8 ± 8.6
.672 .881
39 27.6 ± 8.9 15.5 ± 8.3
47 25.6 ± 11.9 16.1 ± 7.8
.378 .721
98 30.3 ± 10.0 19.7 ± 8.3
56 27.6 ± 12.0 19.9 ± 8.6
.133 .892
45 28.8 ± 9.6 16.0 ± 8.0
43 26.5 ± 11.2 16.5 ± 7.9
.307 .754
100 19.9 ± 6.8 31.6 ± 11.2 20.8 ± 8.6
61 20.9 ± 6.2 30.0 ± 12.9 20.9 ± 8.4
.340 .431 .945
48 19.7 ± 5.9 28.5 ± 9.6 15.0 ± 8.2
37 21.1 ± 5.9 27.1 ± 12.2 16.7 ± 9.0
.231 .571 .369
104 32.0 ± 12.4 21.1 ± 9.8
71 31.6 ± 13.1 22.0 ± 9.0
.865 .526
55 28.0 ± 10.0 15.2 ± 7.9
51 27.5 ± 10.3 17.0 ± 8.1
.797 .259
VATS, Video-assisted thoracic surgery; OT, open thoracotomy; TNDN, total number of dissected nodes; TNMND, total number of mediastinal nodes dissected; Ad, adenocarcinoma; GTD, greatest tumor diameter; p, pathologic.
Table VI. Postoperative mortality and morbidity Mortality Morbidity Chylothorax RLNI
VATS
OT
P value
2 (1.1%)
3 (1.9%)
.6624
2 (1.1%) 5 (2.6%)
4 (2.5%) 2 (1.3%)
.417 .462
VATS, Video-assisted thoracic surgery; OT, open thoracotomy; RNI, recurrent nerve injury.
the operative side main bronchus, the vagus nerve, and the azygos veins; compression of the esophagus and superior vena cava; compression of the left pulmonary artery; and transection of Botallo’s ligament to spread the aortopulmonary window if necessary. Estimation of SND by VATS. Although most general thoracic surgeons might think that SND by VATS is inferior to that by OT, surprisingly, there are no reports written in English on the inferiority of ND by VATS to that done through OT. In fact, Sagawa and associates10 reported on the completeness of the ND for primary lung cancer by VATS. After they performed pulmonary resection with hilar and mediastinal ND by VATS, a standard thoracotomy was carried out to complete the SND. On the right side, the average number of resected nodes by VATS and the remnant nodes were 40.3 and 1.2, respectively. The average weights of the dissected tissues by VATS and the remnant tissues were 10.0 g and 0.2 g, respectively. On the left side, there were 37.1 and 1.2 nodes and 8.3 g and 0.2 g
Table VII. Recurrence VATS Recurrence Remote Regional POCTX PORTX Re-resection
25 21 4 23 2 1
(13.1%) (11.0%) (2.1%) (12.0%) (1.1%) (0.5%)
OT 24 19 5 23 3 2
(15.1%) (11.9%) (3.1%) (14.5%) (1.9%) (1.3%)
P value .6437 .8664 .7369 .5283 .6624 .5928
VATS, Video-assisted thoracic surgery; OT, open thoracotomy; POCTX, postoperative chemotherapy; PORTX, postoperative radiotherapy.
of dissected tissues, respectively. From these results, they concluded that ND with VATS was technically feasible and that the remnant (‘‘missed’’ by VATS) nodes and tissues were 2% to 3%, which seems acceptable for clinical stage I lung cancer. Their study has 2 limitations: differences of operators between ND by VATS and that by OT, and prejudice that ND by OT is superior to that by VATS (they set as a premise that if ND by OT was first performed and then ND by VATS was performed, there would be no remnant nodes.). We elucidated the completeness of SND by VATS by comparing dissected node numbers between VATS and OT performed mostly by the same operator (Watanabe). Although the best way to elucidate the completeness is to anatomize the regions, we cannot do this. The literature provides no definitions of completeness. We suppose that many thoracic surgeons agree that ND is complete if they can find
516 Watanabe et al
Figure. Recurrence-free survival curves.
neither a node nor fat tissue around each station. We have aimed at being able to observe the anterolateral wall of the trachea and right main bronchus on the posterior side, the subclavian artery on the upper side, the posterior wall of the superior vena cava on the anterior side, the right main pulmonary artery on the lower side, and the lateral wall of the ascending aorta on the back side after a right superior mediastinal ND; the inferior wall of the aortic arch on the upper side, the anterolateral wall of the tracheobronchial area on the posterior side, the left lateral wall of the ascending aorta on the anterior side, the anterior wall of the descending aorta on the posterior side, and the left main pulmonary artery on the inferior side after a left para-aortic ND; the carina on the upper side, the median wall of the main bronchi on the bilateral sides, the upper edge of the lower pulmonary vein on the lower side, and the retropericardium on the back wall after a subcarinal ND. In our series, the number of #9 dissected nodes in an RUL differed between the VATS and the OT because this procedure is not routinely performed; the number of #10 nodes in an LUL also differed between the 2 groups because differentiation between #10 and #4 nodes is clinically difficult. Actually, the total number of dissected nodes of #4+10 in an LUL did not differ between the 2 groups. In the other nodal stations, the number of dissected nodes did not differ between VATS and OT. The cutting and crushing may cause dissemination of cancer cells to the pleural or mediastinal space if the node is involved; therefore, we have excluded patients with primary lung cancer, whom we suspected had node involvement preopera-
Surgery September 2005
tively, as candidates for VATS lobectomy. There is some doubt whether an assessment by only the number of dissected nodes is as feasible as an assessment of the oncologic treatment. We attempted qualitative assessment of dissected nodes; however, we could not show the incidence of crushed nodes in this study because evaluation of the crushed nodes was very difficult. Although we thought the incidence of cut nodes is greater during ND by VATS than OT because of the need to grasp the node more strongly and increased node traction, unexpectedly the 2 groups did not differ with regard to the incidence of cut nodes during ND of #1+2 or #7. In addition, our data show that, in clinical stage I primary lung cancer, even if the incidence of crushed nodes differs between the 2 groups, it does not affect the recurrence of lung cancer. Study limitation. This study is a clinical review of patients who underwent major pulmonary resection with SND for clinical stage I NSCLC. Although clinical follow-up was complete, patients were not randomly assigned to either group. Therefore, selection bias may affect our results. However, different factors between the 2 groups were male ratio, clinical T factor, histology (rate of adenocarcinoma), and pathologic factors. Specifically, in the VATS group, the male ratio was lower, and clinical and pathologic factors were better than in the OT group. These factors are likely to make the number of nodes and dissected nodes greater in the OT group. We think the fact that, under these background conditions, the number of dissected nodes between the 2 groups did not differ gives evidence for the feasibility of SND for NSCLC by VATS. Of course, the best model is a randomized clinical trial. Furthermore, although the groups did not differ with regard to the occurrence of mortality and morbidity, a type II error of analysis might have occurred because of the very low occurrence of morbidity and mortality. CONCLUSION There was no differentiation with regard to the number of dissected nodes between ND by VATS and that through an OT. There were no differences in operative mortality and morbidity. We conclude that VATS lobectomy with SND is feasible and safe for patients with clinical stage I NSCLC. REFERENCES 1. Martini N, Flehinger BJ, Nagasaki F, Hart B. Prognostic significance of N1 disease in carcinoma of the lung. J Thorac Cardiovasc Surg 1983;86:646-53.
Watanabe et al 517
Surgery Volume 138, Number 3
2. McKenna R Jr. Vats lobectomy with mediastinal lymph node sampling or dissection. Chest Surg Clin N Am 1995;5: 223-32. 3. McKenna RJ Jr, Wolf RK, Brenner M, Fischel RJ, Wurnig P. Is lobectomy by video-assisted thoracic surgery an adequate cancer operation? Ann Thorac Surg 1998;66:1903-8. 4. Naruke T, Suemasu K, Ishikawa S. Lymph node mapping and curability at various levels of metastasis in resected lung cancer. J Thorac Cardiovasc Surg 1978;76:832-9. 5. Sugi K, Nawata K, Fujita N, et al. Systematic lymph node dissection for clinically diagnosed peripheral non-small-cell lung cancer less than 2 cm in diameter. World J Surg 1998; 22:290-5. 6. Keller SM, Adak S, Wagner H, Johnson DH. Mediastinal lymph node dissection improves survival in patients with
7. 8.
9.
10.
stages II and IIIa non-small cell lung cancer. Ann Thorac Surg 2000;70:358-66. Naruke T. Significance of lymph node metastases in lung cancer. Semin Thorac Cardiovasc Surg 1993;5:210-8. Ginsberg RJ, Rubinstein LV. Randomized trial of lobectomy versus limited resection for T1 N0 non-small cell lung cancer. Lung Cancer Study Group. Ann Thorac Surg 1995; 60:615-23. Kaseda S, Hangai N, Yamamoto S, Kitano M. Lobectomy with extended lymp node dissection by video-assisted thoracic surgery for lung cancer. Surg Endosc 1997;11:703-6. Sagawa M, Sato M, Sakurada A, et al. A prospective trial of systematic nodal dissection for lung cancer by video-assisted thoracic surgery: can it be perfect? Ann Thorac Surg 2002; 73:900-4.
Background. Major pulmonary resection with systematic node dissection (SND) for early lung cancer by video-assisted thoracic surgery (VATS) is performed in many institutes, but the feasibility of SND for early lung cancer by VATS remains controversial. The aim of this study was to elucidate the feasibility and safety of SND by VATS. Methods. Three hundred fifty patients with clinical stage I lung cancer who underwent pulmonary major resection with SND between 1998 and 2003 were enrolled in this study. Of these patients, 191 (VATS group) underwent pulmonary resection with SND by VATS; 159 patients (open thoracotomy [OT] group) did so through anterolateral thoracotomy. The clinical and pathologic data, including the number of dissected nodes in each nodal station, of the 2 groups were compared to evaluate the feasibility of SND by VATS. Results. Pathologic data showed that, in the VATS group, more patients had adenocarcinoma (P = .0078) and fewer patients had advanced factors than the OT group. The greatest tumor diameter was 24.5 mm and 29.6 mm in the VATS group and OT group, respectively (P < .0001). The total number of mediastinal nodes dissected in right upper lobectomy plus right middle lobectomy (RUL+RML), right lower lobectomy (RLL), left upper lobectomy (LUL), and lower left lobectomy (LLL) also did not differ between the 2 groups. The total number of mediastinal nodes dissected in RUL+RML, RLL, LUL, and LLL was 19.7 in the VATS group versus 22.0 in the OT group (P = .122), 23.4 versus 21.0 (P = .241), 14.8 versus 17.5 (P = .123), and 18.8 versus 15.8 (P = .202), respectively. The number of dissected nodes in each nodal station in RUL+RML, RLL, LUL, and LLL was similar between the 2 groups. Operative mortality, morbidity, or recurrence did not differ between the 2 groups. Conclusions. With regard to the number of dissected nodes, SND by VATS was not inferior to that of OT. SND by VATS is technically feasible and safe, and seems acceptable for clinical stage I lung cancer. (Surgery 2005;138:510-7.) From the Department of Thoracic and Cardiovascular Surgery, Sapporo Medical University School of Medicine
ALTHOUGH the therapeutic effect of node dissection (ND) for primary lung cancer remains controversial, the diagnostic effect is clear. Lobectomy with mediastinal ND for primary lung cancer was first reported in 1993 by Martini and associates.1 A video-assisted thoracoscopic surgery Accepted for publication April 18, 2005. Reprint requests: Atsushi Watanabe, MD, Department of Thoracic and Cardiovascular Surgery, Sapporo Medical University School of Medicine, South 1, West 16, Chuo-ku, Sapporo 0608543, Japan. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2005 Mosby, Inc. All rights reserved. doi:10.1016/j.surg.2005.04.005
510 SURGERY
(VATS) lobectomy with mediastinal ND for primary lung cancer was first performed in 1995 by McKenna and associates.2 They showed that the VATS lobectomy for bronchogenic carcinoma appears to be a safe operation, with the same survival rate as that expected for a lobectomy done by thoracotomy.3 In this study, we sought to elucidate the feasibility and safety of SND by VATS for stage I primary non–small cell lung cancer (NSCLC). MATERIAL AND METHODS Three hundred fifty patients with clinical stage I NSCLC who underwent major pulmonary resection with SND between 1998 and 2003 were
Surgery Volume 138, Number 3
enrolled in this study. These patients were divided into 2 groups (VATS group and open thoracotomy [OT] group) on the basis of the approach to the pleural space. One hundred ninety-one patients, the VATS group, underwent pulmonary resection with SND by VATS, whereas one hundred fifty-nine patients, the OT group, did so through an anterolateral thoracotomy. Clinical and pathologic data, including the number of dissected nodes in each nodal station, of the 2 groups were compared. Preoperative management. All patients who underwent major pulmonary resection gave written informed consent for performing an ND before the operation; those who refused were excluded from the study. The consenting patients underwent preoperative staging and pulmonary function assessment. The preoperative workup was standardized for the staging and consisted of routine chest roentgenography, computed tomographic (CT) scanning of the thorax and the abdomen, CT scanning or magnetic resonance imaging of the brain, bone scintigraphy, and bronchoscopy. Neither mediastinoscopy nor positron emission tomography were performed for clinical stage I primary lung cancer. If the patients had cardiac symptoms such as chest oppression, pain, or palpitations, a cardiac evaluation was performed. Surgical indications for VATS major pulmonary resection with SND. We started VATS lobectomy with SND for primary lung cancer in January 1997 at our institute. The inclusion criteria for VATS lobectomy were as follows: clinical stage IA lung cancer; neither incomplete fissure area nor extensive pleural adhesion found on the preoperative chest CT or intraoperatively; and the ability to physiologically tolerate 1--lung ventilation. Our inclusion criteria for VATS lobectomy have changed with our knowledge of and ability to perform the procedure. Before December 2000, we passively recommended the VATS approach with SND as a first-line procedure because of our inexperience; 2 patients refused to undergo VATS lobectomy and received a lobectomy through an anterolateral thoracotomy. After January 2001, we recommended the VATS approach as a first-line procedure because of experience in performing the procedure. After January 2002, inclusion criteria were extended as follows: clinical stage I patients whose greatest tumor diameter was 40 mm or less and the ability to physiologically tolerate 1--lung ventilation. Extensive pleural adhesion and incomplete fissure area without emphysematous change in the lung found on the preoperative chest CT or intraoperatively were no longer con-
Watanabe et al 511
sidered inclusion criteria. Therefore, the rate of VATS major pulmonary resections to all major pulmonary resections for clinical stage I primary lung cancer has gradually increased every year: namely, a rate of 5.3% (2/38) in 1997 slowly increased to 81.6% (71/87) by 2003. Furthermore, after January 1999, patients were registered for a VATS segmentectomy if they had the following criteria: adenocarcinoma with greatest tumor diameter of 20 mm or less and 50% or more of the rate of ground glass area to the greatest tumor area on the thin-slice CT scanning, and peripheral squamous cell carcinoma with greatest tumor diameter of 20 mm or less. We performed SND for all clinical stage I primary lung cancer patients without intraoperative assessment of the frozen section of any nodes. Criteria for intraoperative conversion of the VATS procedure to OT. Before December 2001, the criteria for intraoperative conversion of the VATS procedure to OT were as follows: uncontrolled bleeding, dense pleural adhesion, or the need to extend the procedure such as bronchoplasty or pulmonary artery plasty. After January 2002, dense pleural adhesion without emphysematous change in the lung was excluded from the criteria. Unfortunately, 15 patients underwent conversion of VATS lobectomy to lobectomy by OT. The conversion rate was 15/206 (7.3%), the reason being intraoperative uncontrolled bleeding in 6, dense pleural adhesion in 5, extended resection in 3, and an intraoperative cardiac event (ventricular fibrillation) in 1. One hundred ninety-one patients excluding these 15 patients were enrolled in this study as the VATS group. Surgical technique of VATS lobectomy. General anesthesia with selective lung ventilation was performed with the use of a double lumen endotracheal tube. Patients were placed in a decubitus position on an operating table. In the VATS group, 2 thoracoports were placed in the sixth or seventh intercostal space (ICS) on the anterior axillary line and in the seventh or eighth ICS on the posterior axillary line; an anterolateral minithoracotomy (37 cm) was made in the fourth ICS for an upper lobectomy or in the fifth ICS for a middle or lower lobectomy. The site of the minithoracotomy was covered with a Lap Protector Mini (Hakko Medical Co, Tokyo, Japan). Pulmonary vessel management was performed by using forceps or scissors for conventional surgery. These vessels were divided after double ligation with a silk suture or after clipping at the proximal and distal portions without the use of end-stapler devices. In recent cases, the basal segmental artery (with or without
512 Watanabe et al
superior segmental pulmonary artery of the lower lobe) or inferior pulmonary vein were divided by the use of end-stapler devices (Endo-GIA; United States Surgical Corp, Norwalk, Conn; and ETS-45; Ethicon, Cincinnati, Ohio). All the bronchial management was performed with the use of endstapler devices. A lobe or segment with NSCLC was extracted in all cases in a specimen bag made of vinyl that we designed to avoid chest wall contamination. On the right side, an upper mediastinal ND usually was performed without the transection of an azygos vein. The nodes were held with long forceps for conventional thoracic surgery. During the procedure, the superior vena cava or vagus nerves were compressed with a cherry dissector (Ethicon); vagus nerves and azygos veins were sometimes taped and retracted posterolaterally with a silk suture if necessary. When dissecting the most upper mediastinum, we avoided using an electrocautery, so as not to damage the right recurrent laryngeal nerve. Small vessels and small branches of the vagus nerve were clipped with a premium surgiclip (United States Surgical Corp) and then transected, or transected with an electrocautery. Secondary ND at the lower posterior mediastinum (subcarinal and paraesophageal) was performed after the right main bronchus was taped with a silk suture, which was brought out of the thorax through a minithoracotomy and retracted anterolaterally to gain a good operative view. During the dissection, the bronchial artery was clipped with endoscopic clips or transected by using electrocautery; the esophagus was compressed with a cherry dissector or the lung was retracted anteriorly with the use of an ENDLUNG (United States Surgical Corp) if necessary. Nodes in the pulmonary ligament and paraesophageal site were dissected only when a lower lobectomy was performed. On the left side, at first, para-aortic and subaortic nodes were dissected. After that, pretracheal and tracheobronchial nodes were dissected. During the dissection, the left main pulmonary artery was compressed to the caudal side with a cherry dissector, or the aortic arch was pushed to the cranial side. The Botallo’s duct was simply transected after confirming the absence of intraductal blood flow if a good operative field was not gained despite the compression as mentioned above. The dissection of the lower posterior mediastinal nodes and nodes in pulmonary ligaments were performed in the same method as that on the right side. After the procedure, 1 chest tube was placed in the pleural cavity and the wounds were closed.
Surgery September 2005
Node mapping. We used node mapping, which was reported by Naruke et al.4 They assigned a number to each node station, namely, #1 to superior mediastinum, #2 to paratracheal, #3 to pretracheal, #4 to tracheobronchial, #5 to subaortic, #6 to para-aortic, #7 to subcarinal, #8 to paraesophageal, #9 to pulmonary ligament, #10 to hilar, and #11 to interlobar. We defined #1 and #2 together as superior mediastinal nodes (#1+2) for convenience. Postoperative adjuvant therapy. Basically, we performed adjuvant chemotherapy only on the patients who had recurrence even if they were patients with pathologic n1 or n2 disease. We performed re-resection or radiotherapy on the patients with locoregional recurrence when possible. Parameters. We reviewed demographic data, clinical data (stage, tumor location, procedure), and pathologic data (stage, greatest tumor size, number of dissected nodes). In relation to the number of dissected nodes, we counted the number of large nodes that were cut during the ND after excluding small pieces of nodes that were cut the same way. We defined the dissected node number as the sum of the number of nodes that were not cut and the number of large nodes that were cut. Furthermore, we compared the number of dissected nodes in each nodal station, postoperative morbidity and mortality, and recurrence between the 2 groups. Statistical analysis. All continuous variables are expressed as means ± SD. Differences between the 2 groups were assessed by means of the Student t test after the assurance of homogeneity by the Levene test. Categorical data were compared by using the chi-square test or the Fisher exact test. Survival rates were assessed by using the KaplanMeier method; the curves were compared by using a log-rank test. All reported probability values are 2-tailed, and P values of less than 0.05 were considered statistically significant. Statistical analyses were performed with the use of SPSS 10.0 software (SPSS, Inc, Chicago, Ill). RESULTS Clinical data. Clinical data are summarized in Table I. Although the groups did not differ in age (P = .578), tumor location (P = .149), surgical procedure (P = .3843), or concomitant disease, the male ratio (P = .001) and clinical T factor (P < .0001) were lower in the VATS group than the OT group. The 2 groups did not differ with regard to the occurrences of preoperative concomitant pulmonary diseases.
Watanabe et al 513
Surgery Volume 138, Number 3
Table I. Demographic and clinical data Procedure Lobectomy (yes/no) Lobectomy Segmentectomy Bilobectomy Pneumonectomy Concomitant disease Tuberculosis COPD Pneumoconiosis
168/23 168 17 6 0
145/14 145 6 7 1
8 26 12
6 28 17
Table II. Pathologic data .161 .3843
.8437 .3727 .1728
VATS, Video-assisted thoracic surgery; OT, open thoracotomy; RUL, right upper lobectomy; RML, right middle lobectomy; RLL, right lower lobectomy; LUL, left upper lobectomy; LLL, left lower lobectomy; COPD, chronic obstructive pulmonary disease.
Pathologic data. Pathologic data showed that more patients had adenocarcinoma (P = .0078) and fewer patients had pathologically advanced factors in the VATS group than the OT group. The greatest tumor diameter was 24.5 mm and 29.6 mm in the VATS group and OT group, respectively (P < .0001; Table II). The number of dissected nodes. The 2 groups did not differ with regard to the total number of dissected nodes in RUL+RML, RLL, LUL, and LLL (Tables III and IV). The 2 groups also did not differ with regard to the total number of mediastinal nodes dissected in RUL+RML, RLL, LUL, and LLL. The number of dissected nodes in each nodal station in RUL+RML, RLL, LUL, and LLL was similar between the 2 groups. However, the number of dissected nodes in #9 in RUL+RML was greater in the OT group than the VATS group (0.3 vs 0.7, P = .028), and the number of dissected nodes in #10 in LUL was greater in the VATS group than the OT group (3.6 vs 2.5, P = .025). Furthermore, we compared the 2 groups with regard to the total number of dissected nodes and total number of mediastinal nodes dissected in the subgroups: namely the male group, adenocarcinoma group, GDT < 30 mm group, and 0 of pathologic node status group. There were no differences between the 2 groups (Table V). Cut nodes. In the recent 100 cases (50 cases in each group) of this series, the proportions of cut nodes to dissected nodes of #1+2 or #7 between the 2 groups were evaluated and compared by the histopathologists in each group. The definition of cut nodes was nodes with torn capsule in the hematoxylin-eosin staining (original magnification: 320). The proportions of crushed nodes were not evaluated because of the difficultly in estimating the existence of crush in nodes. The pro-
GTD (mm) Histology Ad/non-ad Ad Sq Others p-T 1/>1 1 2 3 4 p-N 0/1,2 0 1 2 p-stage I/II,III I (A/B) II (A/B) III (A/B)
VATS (n = 191)
OT (n = 159)
P value
24.5 ± 11.1
29.6 ± 14.0
<.0001
143/48 143 35 13
98/61 98 43 18
.0078 .0278
142/49 142 41 3 5
93/66 93 52 6 8
.0017 .016
159/32 159 12 20
122/37 122 22 15
.1271 .059
154/37 154 (125/29) 14 (5/9) 23 (17/6)
113/46 113 (78/35) 22 (12/10) 24 (13/11)
.0363 .074
VATS, Video-assisted thoracic surgery; OT, open thoracotomy; GTD, greatest tumor diameter; Ad, adenocarcinoma; Sq, squamous cell carcinoma; p, pathologic.
portions of #1+2 and #7 in the VATS and OT groups were 37/276 versus 30/261 (P = .5897) and 47/348 and 42/354 (P = .6916), respectively. There were no significant differences between the groups. Operative mortality or morbidity. Neither operative mortality nor morbidity differed between the 2 groups. Two patients in the VATS group died of pneumonia and interstitial pneumonia; 3 patients in the OT group died of pneumonia (n = 2) and interstitial pneumonia (n = 1). The groups did not differ in the incidence of postoperative chylothorax or recurrent laryngeal nerve injury, which is caused by SND (Table VI). Recurrence. The incidence of recurrence and requirement of postoperative chemotherapy, radiotherapy, or re-resection are summarized in Table VII. There were no differences between the 2 groups (P = .6437). Furthermore, recurrence-free survival curves (Figure) showed that the 5-year rates were 76.3% and 71.4% in the VATS group and OT group, respectively, and that the curves did not differ between the 2 groups (P = .884). DISCUSSION Although the therapeutic effect of extensive mediastinal lymphadenectomy has been the debated
514 Watanabe et al
Surgery September 2005
Table III. Number of dissected nodes in right side Upper + Middle VATS (n = 78) TNDN TNMND 1+2 3 4 7 8 9 10 11s 11i
30.7 19.7 5.5 4.7 2.4 6.4 2.1 0.3 2.0 2.2 1.1
± ± ± ± ± ± ± ± ± ± ±
11.1 8.4 4.5 2.9 2.7 4.3 1.3 1.0 3.0 2.5 2.1
OT (n = 52) 31.1 22.0 5.2 4.9 2.8 6.9 2.4 0.7 1.2 2.3 0.8
± ± ± ± ± ± ± ± ± ± ±
13.4 8.6 4.1 4.5 2.9 4.1 1.9 1.0 2.0 3.3 2.0
Lower P value .831 .122 .767 .730 .544 .559 .578 .026 .099 .775 .421
VATS (n = 49) 33.8 23.4 5.7 5.1 2.8 7.9 2.4 0.8 1.4 0.6 3.0
± ± ± ± ± ± ± ± ± ± ±
13.1 10.3 4.0 3.3 2.5 5.3 1.8 1.0 2.1 1.3 2.4
OT (n = 41) 30.9 21.0 4.4 5.8 2.2 6.2 2.0 1.2 1.2 1.0 2.0
± ± ± ± ± ± ± ± ± ± ±
12.1 8.3 3.8 5.0 2.5 4.2 1.2 1.0 1.7 1.5 2.0
P value .279 .241 .128 .416 .298 .107 .424 .290 .685 .156 .063
VATS, Video-assisted thoracic surgery; OT, open thoracotomy; TNDN, total number of dissected nodes; TNMND, total number of mediastinal nodes dissected.
Table IV. Number of dissected nodes in left side Upper VATS (n = 48) TNDN TNMND 3 4 5 6 7 8 9 10 11
28.0 14.8 2.1 2.0 2.2 2.2 4.3 2.0 2.5 3.6 3.0
± ± ± ± ± ± ± ± ± ± ±
10.2 8.1 2.5 1.9 1.9 2.2 4.2 1.6 1.0 2.4 3.2
Lower
OT (n = 40) 28.1 17.5 1.9 2.3 2.7 2.6 4.9 2.3 2.5 2.5 2.4
± ± ± ± ± ± ± ± ± ± ±
11.3 8.2 2.5 3.0 2.3 2.8 4.1 2.3 1.5 2.1 2.0
VATS (n = 16) .964 .123 .694 .537 .275 .442 .477 .637 .975 .025 .260
29.7 18.8 2.0 1.7 2.6 2.5 5.8 3.0 2.8 2.3 3.3
± ± ± ± ± ± ± ± ± ± ±
7.1 7.2 2.4 1.6 2.3 3.9 4.9 2.1 2.7 2.3 3.2
OT (n = 26) 25.5 15.8 2.9 1.7 2.0 1.8 4.0 1.3 2.1 2.3 3.0
± ± ± ± ± ± ± ± ± ± ±
11.8 7.2 2.9 1.9 1.8 1.6 3.2 1.4 1.9 2.0 2.9
.212 .202 .293 .940 .363 .425 .157 .063 .389 .978 .777
VATS, Video-assisted thoracic surgery; OT, open thoracotomy; TNDN, total number of dissected nodes; TNMND, total number of mediastinal nodes dissected.
issue,5,6 the diagnostic effect is clear. Some reports7,8 indicate that complete mediastinal ND is important for accurate staging as well as overall survival. A large, prospective randomized multicenter trial of mediastinal node sampling versus complete lymphadenectomy during pulmonary resection in patients with N0 or N1 (less than hilar) NSCLC is currently under way (American College of Surgeons Oncology Group, ACOSOG protocol Z0030). In Japan, we generally perform SND through OT as a standard operation for clinical stage I primary NSCLC. Some institutes refuse to perform SND by VATS; namely, ND is either not performed or only node sampling is performed. Procedure of SND. Certainly, performing ND by VATS for primary lung cancer seems difficult. During dissection of upper mediastinal nodes and subcarinal nodes, exposing the operative field
requires great effort. Kaseda and colleagues9 reported the following ND technique: During dissection at the subcarina, the esophagus and bronchus are compressed by a mini retractor (END MINI-RETRACT, United States Surgical Corp). They reported recently at a conference in Japan they made a new device with a long wound opener for gaining a wide intraoperative field during mediastinal ND. During the dissection of the upper mediastinum, an azygos vein was ligated with a suture and transected. The vagus nerve was taped with silk thread, which was brought outside the thorax with an ENDO CLOSE (United States Surgical Corp), and the right main bronchus was then retracted laterally with the use of an END MINI-RETRACT to obtain a good operative view. We did not transect the azygos veins if we could obtain a good thoracoscopic view and operative field using techniques such as lateral retraction of
Watanabe et al 515
Surgery Volume 138, Number 3
Table V. Number of dissected nodes in subgroups Right side
Male N TNDN TNMND Ad N TNDN TNMND GTD < 30 mm N Mean (mm) TNDN TNMND p-n0 N TNDN TNMND
Left side
VATS
OT
P value
VATS
OT
P value
63 33.8 ± 13.1 22.5 ± 9.1
66 32.8 ± 13.1 22.8 ± 8.6
.672 .881
39 27.6 ± 8.9 15.5 ± 8.3
47 25.6 ± 11.9 16.1 ± 7.8
.378 .721
98 30.3 ± 10.0 19.7 ± 8.3
56 27.6 ± 12.0 19.9 ± 8.6
.133 .892
45 28.8 ± 9.6 16.0 ± 8.0
43 26.5 ± 11.2 16.5 ± 7.9
.307 .754
100 19.9 ± 6.8 31.6 ± 11.2 20.8 ± 8.6
61 20.9 ± 6.2 30.0 ± 12.9 20.9 ± 8.4
.340 .431 .945
48 19.7 ± 5.9 28.5 ± 9.6 15.0 ± 8.2
37 21.1 ± 5.9 27.1 ± 12.2 16.7 ± 9.0
.231 .571 .369
104 32.0 ± 12.4 21.1 ± 9.8
71 31.6 ± 13.1 22.0 ± 9.0
.865 .526
55 28.0 ± 10.0 15.2 ± 7.9
51 27.5 ± 10.3 17.0 ± 8.1
.797 .259
VATS, Video-assisted thoracic surgery; OT, open thoracotomy; TNDN, total number of dissected nodes; TNMND, total number of mediastinal nodes dissected; Ad, adenocarcinoma; GTD, greatest tumor diameter; p, pathologic.
Table VI. Postoperative mortality and morbidity Mortality Morbidity Chylothorax RLNI
VATS
OT
P value
2 (1.1%)
3 (1.9%)
.6624
2 (1.1%) 5 (2.6%)
4 (2.5%) 2 (1.3%)
.417 .462
VATS, Video-assisted thoracic surgery; OT, open thoracotomy; RNI, recurrent nerve injury.
the operative side main bronchus, the vagus nerve, and the azygos veins; compression of the esophagus and superior vena cava; compression of the left pulmonary artery; and transection of Botallo’s ligament to spread the aortopulmonary window if necessary. Estimation of SND by VATS. Although most general thoracic surgeons might think that SND by VATS is inferior to that by OT, surprisingly, there are no reports written in English on the inferiority of ND by VATS to that done through OT. In fact, Sagawa and associates10 reported on the completeness of the ND for primary lung cancer by VATS. After they performed pulmonary resection with hilar and mediastinal ND by VATS, a standard thoracotomy was carried out to complete the SND. On the right side, the average number of resected nodes by VATS and the remnant nodes were 40.3 and 1.2, respectively. The average weights of the dissected tissues by VATS and the remnant tissues were 10.0 g and 0.2 g, respectively. On the left side, there were 37.1 and 1.2 nodes and 8.3 g and 0.2 g
Table VII. Recurrence VATS Recurrence Remote Regional POCTX PORTX Re-resection
25 21 4 23 2 1
(13.1%) (11.0%) (2.1%) (12.0%) (1.1%) (0.5%)
OT 24 19 5 23 3 2
(15.1%) (11.9%) (3.1%) (14.5%) (1.9%) (1.3%)
P value .6437 .8664 .7369 .5283 .6624 .5928
VATS, Video-assisted thoracic surgery; OT, open thoracotomy; POCTX, postoperative chemotherapy; PORTX, postoperative radiotherapy.
of dissected tissues, respectively. From these results, they concluded that ND with VATS was technically feasible and that the remnant (‘‘missed’’ by VATS) nodes and tissues were 2% to 3%, which seems acceptable for clinical stage I lung cancer. Their study has 2 limitations: differences of operators between ND by VATS and that by OT, and prejudice that ND by OT is superior to that by VATS (they set as a premise that if ND by OT was first performed and then ND by VATS was performed, there would be no remnant nodes.). We elucidated the completeness of SND by VATS by comparing dissected node numbers between VATS and OT performed mostly by the same operator (Watanabe). Although the best way to elucidate the completeness is to anatomize the regions, we cannot do this. The literature provides no definitions of completeness. We suppose that many thoracic surgeons agree that ND is complete if they can find
516 Watanabe et al
Figure. Recurrence-free survival curves.
neither a node nor fat tissue around each station. We have aimed at being able to observe the anterolateral wall of the trachea and right main bronchus on the posterior side, the subclavian artery on the upper side, the posterior wall of the superior vena cava on the anterior side, the right main pulmonary artery on the lower side, and the lateral wall of the ascending aorta on the back side after a right superior mediastinal ND; the inferior wall of the aortic arch on the upper side, the anterolateral wall of the tracheobronchial area on the posterior side, the left lateral wall of the ascending aorta on the anterior side, the anterior wall of the descending aorta on the posterior side, and the left main pulmonary artery on the inferior side after a left para-aortic ND; the carina on the upper side, the median wall of the main bronchi on the bilateral sides, the upper edge of the lower pulmonary vein on the lower side, and the retropericardium on the back wall after a subcarinal ND. In our series, the number of #9 dissected nodes in an RUL differed between the VATS and the OT because this procedure is not routinely performed; the number of #10 nodes in an LUL also differed between the 2 groups because differentiation between #10 and #4 nodes is clinically difficult. Actually, the total number of dissected nodes of #4+10 in an LUL did not differ between the 2 groups. In the other nodal stations, the number of dissected nodes did not differ between VATS and OT. The cutting and crushing may cause dissemination of cancer cells to the pleural or mediastinal space if the node is involved; therefore, we have excluded patients with primary lung cancer, whom we suspected had node involvement preopera-
Surgery September 2005
tively, as candidates for VATS lobectomy. There is some doubt whether an assessment by only the number of dissected nodes is as feasible as an assessment of the oncologic treatment. We attempted qualitative assessment of dissected nodes; however, we could not show the incidence of crushed nodes in this study because evaluation of the crushed nodes was very difficult. Although we thought the incidence of cut nodes is greater during ND by VATS than OT because of the need to grasp the node more strongly and increased node traction, unexpectedly the 2 groups did not differ with regard to the incidence of cut nodes during ND of #1+2 or #7. In addition, our data show that, in clinical stage I primary lung cancer, even if the incidence of crushed nodes differs between the 2 groups, it does not affect the recurrence of lung cancer. Study limitation. This study is a clinical review of patients who underwent major pulmonary resection with SND for clinical stage I NSCLC. Although clinical follow-up was complete, patients were not randomly assigned to either group. Therefore, selection bias may affect our results. However, different factors between the 2 groups were male ratio, clinical T factor, histology (rate of adenocarcinoma), and pathologic factors. Specifically, in the VATS group, the male ratio was lower, and clinical and pathologic factors were better than in the OT group. These factors are likely to make the number of nodes and dissected nodes greater in the OT group. We think the fact that, under these background conditions, the number of dissected nodes between the 2 groups did not differ gives evidence for the feasibility of SND for NSCLC by VATS. Of course, the best model is a randomized clinical trial. Furthermore, although the groups did not differ with regard to the occurrence of mortality and morbidity, a type II error of analysis might have occurred because of the very low occurrence of morbidity and mortality. CONCLUSION There was no differentiation with regard to the number of dissected nodes between ND by VATS and that through an OT. There were no differences in operative mortality and morbidity. We conclude that VATS lobectomy with SND is feasible and safe for patients with clinical stage I NSCLC. REFERENCES 1. Martini N, Flehinger BJ, Nagasaki F, Hart B. Prognostic significance of N1 disease in carcinoma of the lung. J Thorac Cardiovasc Surg 1983;86:646-53.
Watanabe et al 517
Surgery Volume 138, Number 3
2. McKenna R Jr. Vats lobectomy with mediastinal lymph node sampling or dissection. Chest Surg Clin N Am 1995;5: 223-32. 3. McKenna RJ Jr, Wolf RK, Brenner M, Fischel RJ, Wurnig P. Is lobectomy by video-assisted thoracic surgery an adequate cancer operation? Ann Thorac Surg 1998;66:1903-8. 4. Naruke T, Suemasu K, Ishikawa S. Lymph node mapping and curability at various levels of metastasis in resected lung cancer. J Thorac Cardiovasc Surg 1978;76:832-9. 5. Sugi K, Nawata K, Fujita N, et al. Systematic lymph node dissection for clinically diagnosed peripheral non-small-cell lung cancer less than 2 cm in diameter. World J Surg 1998; 22:290-5. 6. Keller SM, Adak S, Wagner H, Johnson DH. Mediastinal lymph node dissection improves survival in patients with
7. 8.
9.
10.
stages II and IIIa non-small cell lung cancer. Ann Thorac Surg 2000;70:358-66. Naruke T. Significance of lymph node metastases in lung cancer. Semin Thorac Cardiovasc Surg 1993;5:210-8. Ginsberg RJ, Rubinstein LV. Randomized trial of lobectomy versus limited resection for T1 N0 non-small cell lung cancer. Lung Cancer Study Group. Ann Thorac Surg 1995; 60:615-23. Kaseda S, Hangai N, Yamamoto S, Kitano M. Lobectomy with extended lymp node dissection by video-assisted thoracic surgery for lung cancer. Surg Endosc 1997;11:703-6. Sagawa M, Sato M, Sakurada A, et al. A prospective trial of systematic nodal dissection for lung cancer by video-assisted thoracic surgery: can it be perfect? Ann Thorac Surg 2002; 73:900-4.