Ethics, consent and permissions
The Shizuoka Cancer Center Hospital Institutional Review Board approved the retrospective collection and analysis of data from medical records of patients included in this study (approval ID: 25-J128-25-1-3). The need for informed consent from each patient was waived.
Patients who underwent thoracotomy for lung resection at our institution between March 2013 and March 2014 were examined retrospectively. Cases were discussed by anesthesiologists and surgeons, and those patients determined to be contraindicated for EDA, or whose anatomy would make epidural catheter insertion challenging, were selected for thoracic PVB.
General anesthesia was induced with 1.5–2 mg/kg of propofol, 2 μg/kg of fentanyl, and 0.6 mg/kg of rocuronium and maintained with volatile anesthesia or total intravenous anesthesia. All patients were intubated with a double-lumen endobronchial tube and ventilated mechanically.
Patients were placed in the lateral decubitus position. Before the thoracotomy was performed, thoracoscopic intrathoracic inspection was conducted in each patient as follows: First, a 5-mm port for the camera was placed in the 7th intercostal space at the midaxillary line. After confirming that there was no evidence of malignant pleural effusion or pleural dissemination, a thoracotomy was created in the 4th or 5th intercostal space using a posterolateral or anterior axillary approach. Upon completion of the thoracotomy, a chest tube was placed in the 7th intercostal space through the camera port. Thoracotomy and wound closure were accomplished in the same manner in the PVB and EDA groups.
Patients were begun on a regimen of oral COX-2 inhibitor and pregabalin. The chest tube was removed if there was no leakage and the pleural effusion amounted to less than 200 mL daily. Other protocols of postoperative management were also the same in both groups.
After confirmation that there was no dissemination, but before initiation of the thoracotomy, the paravertebral catheter was inserted in a manner that has been previously described [11–14], as follows: The upper edge of the spinous process of the T5 vertebral body was palpated through the skin. With the assistance of forceps inserted through the same camera port, the paravertebral space was visualized under thoracoscopy. An 18-gauge Tuohy epidural needle was inserted at a point 3 cm lateral to the lateral edge of the vertebra. The Tuohy needle was carefully advanced, without puncturing the parietal pleura, until it reached the paravertebral space over the superior border of the transverse process, where 20 mL of normal saline was injected to expand the paravertebral space. Next, while holding the Tuohy needle steady, the paravertebral catheter was placed through the needle and the needle was removed, making sure that the tip of the catheter remained in place. Advancement of the needle and entry of catheter into the paravertebral space were continuously monitored under thoracoscopic vision. To ensure proper positioning of the catheter tip, 20 mL of 0.375% ropivacaine was injected through the catheter. Correct placement was indicated by expansion of the extrapleural space without leakage of local anesthetic into the pleural space (Video 1). The catheter was secured with 2-0 nylon sutures, and continuous infusion of 0.45% ropivacaine was started as soon as possible. The rate of infusion was initially 6–8 mL/h but was titrated to patient comfort. In addition, intravenous fentanyl 20–40 μg/h was continued was continued for the remainder of the day of surgery.
Additional file 1: Video 1. PVB procedure in the case of right lower lobectomy. The catheter was placed in the sixth intercostal space, followed by the fifth intercostal space of thoracotomy. (MOV 147456 kb)
Selection of the control group
The matched-pair control group (EDA group) was selected on a 1:2 ratio from patients who underwent thoracotomy with EDA in our hospital from April 2011 to February 2013. Thoracic epidural catheters were inserted before the induction of general anesthesia at the level of T6-7 and secured in place. EDA was also used for intraoperative analgesia. Patients received a mixture of 0.2% ropivacaine with fentanyl as local anesthetic. The initial dose was 5 mL/h, but this was titrated to patient comfort.
Matching criteria were sex, age, and type of surgery. The criterion of age was divided into five groups: ≤49, 50–59, 60–69, 70–79, and 80–89 years. The types of surgery were divided into three groups: lobectomy with mediastinal lymph node dissection, lobectomy without mediastinal lymph node dissection, and all others. The control group was selected by a person not otherwise associated with the study with no other information about the patients. When there were more than three matching controls for a PVB patient, we selected patients using a random-number table.
All relevant patient data were recorded before and after surgery, and patients were followed until hospital discharge. The following data were assessed: 1) pain score 48 h after surgery, 2) requirement for intravenous rescue analgesia, 3) the required duration of regional anesthesia, and 4) the amount of fentanyl or ropivacaine administered during the perioperative period.
Categorical variables were compared using Fisher’s exact test and continuous variables using the Mann-Whitney test. IBM SPSS for Windows, Version 22.0 (IBM Corp., Armonk, NY, USA) was used for all statistical evaluations. P values less than 0.05 were considered statistically significant.