The operative details and postoperative complications were demonst rated in Table 3. The pre-operative tissue diagnostic rates were similar. The VATS group demonstrated a significantly longer operation time (p=0.004), less intraoperative blood loss (p=0.029), shorter wound length (p<0.001) and shorter length of hospital stay (p<0.001). No significant difference was found in the location of lung cancer. There was no significant difference between the two groups, but VATS group showed significant prolonged air leak (p=0.048). There was only one surgical mortality on postoperative 23rd day because of pneumonia deteriorated into acute respiratory distress syndrome in the MSVMT group; no surgical death in VATS group.

Lobectomy remained the standard surgical resection for early lung cancer. However, with the increasing prevalence of computed tomography application, early lung cancer with small size nodule became more easily detectable. There was a resurgence of interest in anatomic segmentectomy for very early lung cancer, especially in patients with compromised cardio-pulmonary function, who might not tolerate lobectomy due to inadequate postoperative reserved pulmonary function (13). Growing data suggested that segmentectomy was an alternative to lobectomy in patients with clinical T1N0M0 status, especially when tumor diameter was less than 2 cm. This anatomic segmental resection could be performed safely without compromising oncologic results (13-15). In some institutions, segmentectomy with radical lymph node dissection was performed not only in high-risk patients but also in low-risk patients with clinical T1N0M0 and tumors ≤2 cm in diameter (16-17). It could offer the benefit of significantly better preservation of pulmonary function compared with lobectomy (18-19). In our institution, segmentectomy was designed as an alternative standard resection for peripheral clinical T1N0M0 lung cancer with diameter ≦2 cm regardless of the risk level. According to the published data, we considered segmentectomy could preserve more pulmonary function without compromising cancer survival. In our data, a total of 32 patients underwent segmentectomy. There were 16 patients in the VATS group and the other 16 in the MSVMT group. In this study, we focus the analysis of the postoperative complication difference between VATS and MSVMT, not segmentectomy and lobectomy. We merged the data of VATS lobectomy with VATS segmentectomy before comparing the VATS group with MSVMT group on account of both lobectomy and segmentectomy being considered as radical curative anatomic resection for early lung cancer. We compared the data difference on postoperative complications between the two groups and found no significant difference (18.2% vs. 23.6%, p=0.255). There was no surgical mortality in the VATS group and only one conversion. We concluded that VATS lobectomy/ segmentectomy was a safe and technical feasible surgical approach in our institute based on the present data.

Although the surgical risks of VATS lobectomy/ segmentectomy are considered to be acceptable, this new operative approach has been adopted slowly over the past decade. There seems to lack a generally accepted standard procedure for VATS lobectomy/segmentectomy; however, surgical techniques, differently modified, are proposed from all over the world. The obstacles associated with VATS included enigmatic technique skill, steep learning curve, operative safety and oncologic concerns. There are relatively few VATS reports for lung cancer in Taiwan to date. We started VATS lobectomy/segmentectomy with radical lymph node dissection for lung cancer in 2005. The initial criteria for thoracoscopic surgery was limited to small clinical early lung caner. Gradually, we extended the indications of thoracoscopic surgery because of cumulative experiences through time. In our institution to date, patients considered appropriate for thoracoscopic approach include those tumors smaller than 5 cm in diameter without central airway involvement or chest wall invasion. Radical lymph node dissection should be routinely done for definitively pathologic staging of mediastinal lymph node status. Even those patients with single station N2 status by PET-CT scan staging are considered candidates for thoracoscopic surgery.

Many controversies regarding VATS approach face further debates for consensus, which include the length of utility thoracotomy, the application of rib spreader, the usage of endoscopic instruments versus conventional instruments and visualization through the incision or only by the monitor. Even now, the thoracoscopic techniques vary among nations, which may attribute, to some degree, different results in the outcomes. We performed VATS approach, which composed of video-monitor dependent visualization, non-ribs spreading, and shorter-than-six cm working wound length. A total of two to three chest wall incisions were used in our institution.

Better quality of lymphadenectomy may lead to more accurate tumor staging and therefore influence statistical result. Patients with 15 or fewer mediastinal lymph nodes dissected had worse survival outcome than those with more than 15 (20). Generally, we performed a radical mediastinal lymph node dissection for all patients as much as we can. Our data demonstrates the number of dissected nodes is smaller in the VATS group (p=0.005), but the mean numbers of lymph node was larger than 15, which could indicate the accurate tumor staging. We presumed that patients in the VATS group had earlier stage lung cancer, contributing to reduced numbers of lymph node. Of course, it is impossible to discuss the technical impact of lymph node dissection simply based on the numbers. As a matter of fact, the technical quality of node dissection need to be further analyzed according to long-term loco-regional disease-free survival rate.

Adequate postoperative pain control has been known to decrease postoperative pulmonary complications. Diminish pain from chest wall incisions will improve the ability to breathe deep, effectively cough and prevent correlative atelectasis. VATS requires only two small skin incisions for thoracoscopic insertion and utility thoracotomy window without rib spreading, which lessen a lot of postoperative pain (21). The optimal postoperative pain control methods for thoracoscopic surgery have been controversial. Epidural anesthesia may be the most popular and well known means of choice, however several associated complications have been reported in literatures, such as nausea, vomiting, hypotension, pruritus, constipation and technical related complications (22). Epidural analgesia is no longer used in VATS group in our institution the potential risk could be avoided. We prescribed oral non-steroid inflammatory drugs and oral opioids for postoperative pain control. Some patients needed several additional shots of intravenous opioids on postoperative first day. We didn’t compare the pain scale between the two groups. In fact, epidural anesthesia was only used in MSVMT group