A Prospective, Randomized, Double-Blinded Comparison Between Multimodal Thoracic Paravertebral Bupivacaine and Levobupivacaine Analgesia in Patients Undergoing Lung Surgery Vesna Novak-Jankovic, PhD, MD,* Zoka Milan, MD,‡ Iztok Potocnik, MSc, MD,* Tomaz Stupnik, MD,† Stela Maric, MD,* Tatjana Stopar-Pintaric, MD, PhD,* and Boriana Kremzar, MD, PhD* Objectives: To compare the effects of paravertebral anal- gesia with levobupivacaine or bupivacaine on intra- and postoperative pain for thoracic surgery. Design: A prospective, randomized, and double-blinded study. Setting: A university hospital. Participants: Forty patients undergoing thoracic surgery. Interventions: Patients received paravertebral catheteriza- tion and a bolus (14-20 mL) of 0.5% bupivacaine (n  20) or 0.5% levobupivacaine (n  20) with morphine, 60 g/kg, before the induction of general anesthesia that consisted of a propofol infusion. A paravertebral continuous infusion (0.05 mL/kg/h) of 0.25% bupivacaine or 0.25% levobupiva- caine, 100 mL, with added morphine, 10 mg, and clonidine, 0.15 mg, was started at the end of surgery for 72 hours postoperatively. Postoperative rescue diclofenac analgesia was available if required. Measurements and Main Results: The primary outcome was intraoperative fentanyl consumption. Static and dy- namic pain scores measured by a visual analog scale were assessed regularly. Intraoperative fentanyl consumption was significantly lower in the levobupivacaine group com- pared with the bupivacaine group (p  0.001). On all 3 postoperative days, static pain scores were significantly lower in the levobupivacaine group compared with the bu- pivacaine group (p < 0.05). Dynamic pain scores were sig- nificantly lower in the levobupivacaine group compared with the bupivacaine group during the 2 postoperative days (p < 0.05). A smaller proportion of patients in the levobupi- vacaine group used rescue analgesia (p < 0.005). Conclusions: Paravertebral analgesia with levobupiva- caine resulted in less intraoperative fentanyl consumption, lower static (3 days) and dynamic (2 days) pain scores, and less rescue analgesia than analgesia with bupivacaine. © 2012 Elsevier Inc. All rights reserved. KEY WORDS: pulmonary surgical procedures, anesthesia and analgesia, paravertebral block, bupivacaine, levobupivacaine T HORACOTOMY is a painful surgical procedure. Ade- quate pain relief can accelerate mobility, improve ventila- tion and secretion clearance, and decrease atelectasis, chest infection, morbidity, and mortality. 1 The available analgesic options are systemic opioids and epidural and thoracic para- vertebral analgesia (TPA). TPA has been shown to provide less pain, nausea, and vomiting during the immediate postoperative period and greater patient satisfaction compared with intravenous opioid patient-controlled analgesia. 2,3 It results in similar analgesia but greater hemodynamic stability and fewer complications than epidural analgesia. 4 Bupivacaine is the local anesthetic most commonly used for TPA. Numerous studies have sought the optimal dose, volume, and concentration of bupivacaine for continuous TPA, without clear answers. 3 Thus far, levobupivacaine, the L-isomer of bupivacaine that is less toxic than bupivacaine, rarely has been used for TPA in thoracic surgery. 4,5 The high price of levobupi- vacaine with comparable clinical effects 6 is the likely reason for the rare use of levobupivacaine for TPA in thoracic surgery. The aim of this prospective, randomized, double-blinded study was to compare the analgesic efficacy of preoperative and postoperative multimodal thoracic paravertebral bupivacaine analgesia (TPA-B) with thoracic paravertebral levobupivacaine analgesia (TPA-L) by measuring intraoperative fentanyl re- quirements, postoperative pain, postoperative pulmonary func- tions, and hemodynamic stability in patients undergoing lung surgery. Because bupivacaine and levobupivacaine have been re- ported to have a similar potency, 6 the null hypothesis was that TPA-B and TPA-L have an equivalent analgesic effect. METHODS After approval by the republic ethical committee and written in- formed consent, 40 patients with American Society of Anesthesiology physical statuses II and III who were scheduled for thoracic surgery were enrolled in the study. Exclusion criteria were 1 of the following patient characteristics: severe chronic obstructive pulmonary disease with forced expiratory volume in 1 second 40% of the predicted value, central and peripheral neuropathies, empyema, systemic sepsis, allergy to the amide type of local anesthetics, morphine, or clonidine, patients on long-term opioid therapy, patients receiving -blockers, and coagulopathy. All patients were premedicated before surgery with oral midazolam (Dormicum, F. Hoffmann-La Roche Ltd, Basel, Switzerland), 0.1-0.2 mg/kg, 1 hour before surgery. Once in the operating room, an intrave- nous cannula, a radial artery catheter, and a central venous catheter (cubital vein) were inserted while the patients were awake. Patients were allocated randomly to the TPA-B group or the TPA-L group (n = 20 per group) using a computer-generated sequence and sealed envelopes. The technique of unilateral TPA performed at the site of surgery was as follows: an 18-gauge Tuohy needle (epidural kit; Portex–Smiths Medical ASD, Inc, Keene, NH) was inserted perpendicularly into the skin at the level of the T6-7 interspace, 2.5 cm laterally to the tip of the spinous process, and then farther until the transverse process was reached. Then, the needle was withdrawn slightly and redirected ceph- alad at a 45° angle to the skin to 1.5 cm deeper than the depth of bone contact. The catheter was inserted through the needle 1-2 cm beyond its tip. The paravertebral block was initiated in the TPA-B group with a From the Clinical Departments of *Anesthesiology and Intensive Therapy and †Thoracic Surgery, University Medical Center, Ljubljana, Slovenia; and ‡Leeds Teaching Hospitals, St James’s University Hos- pital, Leeds, UK. Address reprint requests to Vesna Novak-Jankovic, MD, PhD, Clinical Department of Anesthesiology and Intensive Therapy, University Medical Center, Zaloska 7, 1000 Ljubljana, Slovenia. E-mail: vnovakjankovic@gmail.com © 2012 Elsevier Inc. All rights reserved. 1053-0770/2605-0018$36.00/0 doi:10.1053/j.jvca.2012.01.050 863 Journal of Cardiothoracic and Vascular Anesthesia, Vol 26, No 5 (October), 2012: pp 863-867