Video-Assisted Thoracoscopic Surgery and Minimal Access Spinal Surgery Compared in Anterior Thoracic or Thoracolumbar Junctional Spinal Reconstruction: A Case-Control Study and Review of the Literature

There are no published reports that compare the outcomes of video-assisted thoracoscopic surgery (VATS) and minimal access spinal surgery (MASS) in anterior spinal reconstruction. We conducted a retrospective case-control study in a single center and systematically reviewed the literature to compare the efficacy and safety of VATS and MASS in anterior thoracic (T) and thoracolumbar junctional (TLJ) spinal reconstruction. From 1995 to 2012, there were 111 VATS patients and 76 MASS patients treated at our hospital. VATS patients had significantly (p < 0.001) longer operating times and significantly (p < 0.022) higher thoracotomy conversion rates. We reviewed 6 VATS articles and 10 MASS articles, in which there were 625 VATS patients and 399 MASS patients. We recorded clinical complications and a thoracotomy conversion rate from our cases and the selected articles. The incidence of approach-related complications was significantly (p = 0.021) higher in VATS patients. The conversion rate was 2% in VATS patients and 0% in MASS patients (p = 0.001). In conclusion, MASS is associated with reduction in operating time, approach-related complications, and the thoracotomy conversion rate.


Introduction
Video-assisted thoracoscopic surgery (VATS) and minimal access spinal surgery (MASS) have been considered primarily as minimally invasive surgery (MIS) for anterior thoracic (T) and thoracolumbar junction (TLJ) spine surgery [1]. VATS was first described by Mack et al. in 1993 [2]; it allows for biopsy, anterior release, abscess drainage, and discectomy [3,4]. VATS has been used to treat anterior thoracic diseases at our hospital since 1995. Over the next 10 years, we used VATS in many spinal procedures: decompression, corpectomy, reconstruction, and stabilization. The microsurgical miniopen anterolateral approach was first introduced in 1997 by Mayer [5] for minimally invasive anterior lumbar interbody fusion. Kossmann et al. [6] reported in 2001 that the anterior column of the thoracic spine could easily be assessed and reconstructed using a minithoracotomy and a table-mounted retractor. At that time, we developed a new VATS approach [7][8][9][10], which we called the "extended manipulating channel method." It allowed us to use a combination of conventional spinal instruments and VATS to enter the chest cavity and to manipulate those instruments as we would for standard open surgical procedures. Furthermore, at our hospital, a refined MASS has been evolving since 2000 from our extended manipulating channel method without VATS [11,12]. MASS has been used to treat vertebral metastasis, osteomyelitis, and 2 BioMed Research International fractures. It is generally believed that because MASS allows direct three-dimensional vision of the surgical field, which seems to make the procedure familiar to spine surgeons used to standard open surgical procedures, it has become more popular than VATS. Thus, we compared the outcomes of MASS in anterior T and TLJ spinal reconstruction and fusion with those of VATS.

Patients.
We identified, in our hospital's Spine Operation Registry, all patients who underwent VATS (Figures 1 and  2) and MASS ( Figure 3). We previously published reports which described both VATS [7,8,10,13] and MASS [12,14] techniques for anterior T or TLJ spinal reconstruction between 1995 and 2012 and retrospectively reviewed their records. The inclusion criteria were anterior intervertebral fusion after a discectomy with a partial or a total corpectomy for treating spinal fractures, vertebral malignancy, infectious spondylitis, thoracic disc herniation, and degenerative spinal diseases. Patients with pediatric scoliosis, a discectomy without fusion, or a biopsy were excluded from the study. All included patients had undergone minimally invasive anterior spine reconstruction performed by one senior surgeon (T. J. Huang). We reviewed the patients' medical records and recorded data on operating time, estimated blood loss, need for intensive care, conversion to standard open thoracotomy, and complications in patients with T and TLJ spinal disorders. Approval for this study was obtained from the Ethics Committee and Institutional Review Board of our hospital (IRB number 101-1238B).

Review of Published
Literature. The English language literature published between 1995 and 2012 was systematically reviewed. The Cochrane Review Database, EMBASE, Medline, PubMed, and Google Scholar were searched. The reference lists of the selected articles were checked. Search terminology included miniopen, MASS, VATS, anterior T spinal surgery, TLJ (T11-L2) spinal surgery, and anterior spinal fusion. We excluded studies associated with pediatric spine surgery, disc excision without fusion, and anterior lumbar surgery (L3-L5). Technical notes, case reports, anatomical descriptions, or a combined surgery of thoracoscopic surgery and thoracotomy was not included. The articles were screened and selected by two independent reviewers (Y. Y. Li and C. C. Cheng) based on the inclusion and exclusion criteria. Disagreements were resolved by discussion or by a consultation with a third reviewer (Ching-Yu Lee). The data of the selected articles were extracted and analyzed in detail by two independent reviewers (M. H. Wu and Chien-Yin Lee). Because data on the surgical complications were going to be analyzed, the interrater agreement about these data was analyzed using the kappa statistic. Disagreements were resolved by discussion or by a consultation with a senior spine surgeon (T. J. Huang).

Data Analysis.
The perioperative parameters of our included sample were operating time, estimated blood loss, complications, conversion to standard thoracotomy, and the need for postoperative admission to the intensive care unit. They were recorded and compared between our VATS and MASS patients. The perioperative data of the selected articles were average operating time, average estimated blood loss, complication rates, and conversions to thoracotomy.
Data of clinical complications and conversions to standard thoracotomy, which were recorded from our cases and the selected articles, were compared between VATS and MASS patients. A minor complication was defined as a minor risk event with no treatment, with medical treatment, or with intraoperative repair but without long-term sequelae. A major complication was defined as a life-threatening or irreversible event requiring invasive treatment or revision surgery. Death was mortality because of associated perioperative complications.
An approach-related complication was defined as intercostal neuralgia, pleural effusion, or air leakage causing subcutaneous emphysema or pneumothorax [15,16].

Statistical Methods
All statistical analyses were done using SPSS 12.0 for Windows. An independent Student -test was used for numerical data. An 2 analysis or a Fisher exact test was used for categorical data. Significance was set at < 0.05. The observed interrater agreement for the data extracted from the selected publications was analyzed using the kappa statistic.

Results
We reviewed the medical records of 187 patients who had undergone minimally invasive surgery (MIS) for anterior T or TLJ spinal fusion at our hospital between 1995 and 2012. VATS was used in 111 patients, and MASS was used in the other 76 patients (Table 1). Operating time was longer in the VATS group than in the MASS group ( < 0.001). There was a significantly higher incidence of conversion to standard open thoracotomy in the VATS group than in the MASS group ( = 0.022). There were no significant differences in average blood loss or the need for postoperative admission to the intensive care unit (ICU).
Perioperative complications were collected from 187 patients of our institute and 1024 patients of the 16 selected    publications ( Table 3). The assessment score agreement between the reviewers was good (kappa statistic: 0.62, < 0.001). There were 126 (17%) perioperative complications in VATS patients and 71 (15%) in MASS patients ( = 0.317): there was not significantly different distribution of no, minor, and major complication ( = 0.567). Revision surgery was the most common major complication in both groups: 11 VATS patients and 8 MASS patients. There were 6 mortalities in this study, 3 in each cohort of VATS and MASS: 1 with pneumonia, 1 with acute thromboembolism, and 1 with intraoperative arrhythmia and acute cardiac infarction in VATS patients; 1 with pneumonia and 2 with acute thromboembolism in MASS patients. The incidence of approach-related complications was significantly higher in VATS patients than MASS patients ( = 0.011). There was no significant difference in the prevalence of pulmonary infection or iatrogenic cardiovascular injury between both surgical procedures. The overall conversion rate from MIS to standard thoracotomy in VATS patients was 2% ( = 15) and 0% in MASS patients ( = 0.001) ( Table 4). The most common cause for unplanned conversion to standard open thoracotomy was severe intrathoracic adhesion (40%), followed by iatrogenic cardiovascular injury (20%) and excessive uncontrollable bleeding from cancellous bone or soft tissue (20%).

Discussion
VATS and MASS are well-known MIS methods for anterior spinal surgeries [1]. It is generally believed that using VATS for spinal surgery entails a learning curve more difficult to negotiate than does using MASS [19]; however, few studies focus on analyzing the advantages and disadvantages of using VATS and MASS to treat anterior spinal disorders. In this study, VATS required longer operating time and a higher incidence of conversion to standard open thoracotomy than did MASS at our hospital. Similarly, our review of the VATS and MASS literature for anterior T and TLJ spinal reconstruction showed that VATS was more likely to need operating time and to increase blood loss. In addition, Molina et al. [32], in a systematic review of MIS in the management of metastatic spine disease, reported that VATS was associated with longer operating time, a longer length of stay in the hospital, and more blood loss than was MASS. We found that since MASS seems more familiar to most surgeons it yields faster and safer decompression, stabilization, and reconstruction than does VATS.
We found that the overall MIS complication rate for anterior T and TLJ spinal reconstruction in the 1211 patients analyzed in the selected articles and in our hospital was 16.2%: 126 perioperative complications in VATS patients (17%) and 71 complications in MASS patients (15%). VATS and MASS patients had similar minor and major complication rates; however, VATS is more associated with approach-related complications. Consistent with the results of previous case series [3,16,19,33], approach-related complications are most common in patients undergoing VATS. This might be true because trocar placement sometimes injures an intercostal nerve or pleural membrane, which leads to intercostal neuralgia, pleural effusion, pneumothorax, or subcutaneous emphysema [15]. Hence, the first thoracoscopic portal, which is not made using endoscopic visualization, is created using a minithoracotomy to make a 1.5 cm skin incision that precludes blind trocar insertion [16,34,35].
Conversion to standard open thoracotomy occurred more frequently in VATS patients than in MASS patients in our hospital and in the selected literature. Consistent with our findings, other studies [35,36] have reported that VATS is restricted because of severe pleural adhesion, poor tolerance of one-lung ventilation, and difficulty with  metastatic chronically infected diseases of the thoracic spine, but thoracoscopic adhesiolysis is a technically demanding procedure that must be done by an expert thoracic surgeon. Besides, intraoperative bleeding was more directly and easily controlled using cauterization, hemoclips, or suture ligation in MASS. Therefore, MASS is a reasonable MIS method for treating anterior T and TLJ spinal reconstruction and fusion, especially for metastatic and infectious spinal diseases. This study has some limitations. First, this is a retrospective study. To minimize the statistical bias, we included patients who had undergone MIS spinal surgery done by the same surgeon (T. J. Huang) at our hospital. Second, the evidence level of the systemic review in this study is low. This is because there is still a paucity of reports that a meta-analysis needs, those that show comparative data of VATS and MASS for treating anterior spinal diseases. Additional comparative VATS and MASS studies that focus on treating anterior spinal diseases are required.
In conclusion, VATS and MASS are effective MIS methods for treating anterior T and TLJ spinal reconstruction, and they have equivalent complication rates. MASS requires less operating time and has fewer approach-related complications. VATS is more likely to have a higher conversion rate from MIS to standard open thoracotomy when severe pleural adhesion and difficulty in endoscopically controlling bleeding are encountered.