The purpose of the study is to investigate if paravertebral block for breast surgery may improve postoperative analgesia and shorten length of hospital stay. Within a prospective clinical trial sixty women (ASA I - III) undergoing unilateral mastectomy surgery with axillary dissection are enrolled. Exclusion criteria are general infection and local infection at the site of planned puncture, anatomic deformities of the thoracic spine, coagulation disorders, allergy against local anesthetics or contrast agents, pregnancy or breast-feeding and BMI ≥ 30 kg/m2.

Patients are randomly assigned to two groups, 30 patients each, by using block of four method. Group I patients receive paravertebral block at spine level Th 4 via catheter; Group II patients will have general anesthesia without nitrous oxide.

Paravertebral block (PVB) will be entirely performed by one person (1st author) with the patient in prone position. In randomly selected patients the catheter localization will be detected by fluoroscopy after injection of 0.5 ml contrast agent. The catheter positioned at about 8 cm from skin level, a mixture of 10 ml bupivacaine 0.5% plus 20 ml lidocaine 2% with adrenaline 1:200,000 will be injected sequentially 10 ml each while withdrawing the catheter 2 cm after the first and again 2 cm after the second injection. Then the catheter will be removed. Low dose intraoperative sedation is provided with IV midazolam 2-3 mg, ketamine 0.5 mg/kg bolus and continuous propofol using target-controlled infusion (TCI) system aiming at effect site concentration (Ce) of 1 - 1.5 mcg/mL to allow spontaneous breathing. In case surgical approach is not tolerated, the regional block is graded insufficient, and it is switched to general anesthesia.

At postoperative care unit (PACU) the pain score using visual rating scale (VRS) from 0 to 10 (worst imaginable pain) is assessed at rest and on movement of the shoulder every 15 minutes until 60 minutes. Analgesia is provided by IV morphine 0.04 mg/kg boluses with a 15-min dosing interval to maintain VRS - score < 3. Time to first analgesic demand and 24-hr analgesics consumption is recorded as well as postoperative nausea and vomiting (PONV). Complications are recorded including those related to the paravertebral block, such as Horner's syndrome, epidural spreading, hypotension that is not related to another. Patients will be checked 6, 12, and 24hr postoperatively for adverse effects, pain scores, and satisfaction scores.

Multiple studies have shown that breast surgery is associated with significant postoperative pain, nausea, vomiting, and development of chronic pain. General anaesthesia is associated with a 50% incidence of nausea and vomiting among breast cancer surgery patients. The proposed benefits of regional anesthesia including paravertebral block compared to general anaesthesia are decreasing stress response, improving postoperative pain control and postoperative pulmonary function, decreasing cancer recurrence, and reducing thromboembolic events. However, the debate about these benefits is still controversial.

Paravertebral block for breast surgery, either alone or additional to general anaesthesia is favored by many investigators, due to better postoperative analgesia and shorter hospital stay. In an elder investigation thermographic imaging was used to apply paravertebral block. After a single injection of 15 ml 0.5% bupivacaine at spine level Th9-Th10 there was a mean distribution of somatic block of five and of sympathetic block of eight dermatomes without any hemodynamic side effects. No bilateral spread was found. The authors contradicted the suggestions paravertebral block being ineffective and sometimes hazardous. However, their patients were chronic pain patients with a number of six only.

Single-injection paravertebral block may not be sufficient to cover the relevant dermatomes C6 - Th6 for radical mastectomy with axillary dissection, when applied as the sole anesthetic technique. Some study groups performed paravertebral block with several injections. Four injections resulted in sufficient anaesthesia in 97% of these patients, whereas one injection was effective in only 11%. Though there is limited published evidence yet, a single injection for paravertebral block may be insufficient for extensive breast surgery, such as mastectomy including axillary evisceration. However, considering the potential risks of deep paravertebral puncture a limitation of punctures would be preferable. Inserting a paravertebral catheter far enough and in proper position offers the opportunity to apply the anesthetic agent at different sites, by moving the catheter backwards after each injection.

The purpose of our study is to find out if injections via paravertebral catheter at three different levels are advantageous and efficient compared to general anesthesia in patients having mastectomy with axillary dissection.

In addition, instead of 4 cm catheter insertion inside the paravertebral space (PVS) as the recommended, we used ultrasound-guided PVB (transverse view, out of plane technique), to facilitate further insertion of the catheter (8 cm) in the PVS. A contrast agent will be used in randomly selected patients to prove the exact site of the inserted catheter and to exclude/detect obstructions such as kinking.

Methods: This is a prospective clinical trial. Patients have to agree documented by written informed consent. Sixty women, ASA I - III undergoing unilateral mastectomy surgery with axillary dissection, including radical lymph node removal or sentinel node biopsy are enrolled.

Patients are randomly assigned to two groups, 30 patients each, by using block of four method. Group I patients receive paravertebral block at spine level Th 4 via catheter with injections at three different levels; Group II patients will have general anesthesia.

Anesthetic preparation: After arrival at the OR standard monitoring is applied including ECG, noninvasive blood pressure and pulse oxymetry. Midazolam 0.05 mg per kilogram bodyweight is given intravenously to all patients to achieve mild sedation.

Paravertebral block (Group I). Paravertebral block will be entirely performed by one person (first author) with the patient in prone position. The superior aspect of the Th4 spinous process is marked. After skin disinfection an 18-gauge Tuohy needle including attached saline filled syringe is inserted perpendicular to the skin at the Th3-Th4 interspace, using the area 2.5 cm lateral to T4 spinous process as Landmark. Under ultrasound guidance, the needle is advanced by out of plane technique, angled anteriorly until contact with the transverse process, and then redirected either caudal or cephalad and parallel to the spine advanced further until loss of resistance. After placement of the needle an end-hole 20 gauge catheter is inserted and advanced 8 cm beyond needlepoint into the paravertebral space. The catheter insertion is scored as 'easy', 'difficult' or 'impossible'. In randomly selected patients the catheter localization will be detected by fluoroscopy after injection of 0.5 ml contrast agent. The catheter positioned, a mixture of 10 ml bupivacaine 0.5% plus 20 ml lidocaine 2% with adrenaline 1:200,000 will be slowly injected sequentially 10 ml each while withdrawing the catheter 2 cm after the first and again 2 cm after the second injection. Then the catheter will be removed. Patients are returned to the supine position and remain for at least 30 minutes in the preoperative area. Sensory blockade is assessed by using pin-prick method.

When analgesia is adequate patients are moved into the operating room. Intraoperative sedation is provided with low dose IV midazolam, ketamine bolus and continuous propofol using target-controlled infusion system to allow spontaneous breathing. In case skin incision or further surgical approach is not tolerated, the regional block is graded insufficient, and it is switched to general anesthesia.

General Anaesthesia (Group II): Anesthesia is induced intravenously with propofol and fentanyl. Atracurium is used to facilitate tracheal intubation. Anesthesia is maintained with sevoflurane and 50% oxygen in air. Additional Fentanyl will be administered on discretion of the anesthesiologist.

Data collection: Patients' biometric data, details of the clinical course, such as amount of anesthetics used, operation time, surgical difficulties, hemodynamic instability (if any) and similar are recorded. At postoperative care unit (PACU) the pain score using visual rating scale from 0 (no pain) to 10 (worst imaginable pain) is assessed at rest and on movement of the shoulder every 15 minutes until 60 minutes. Analgesia is provided by IV morphine boluses with a 15-min dosing interval to maintain VRS - score < 3. Time to first analgesic demand and 24-hr analgesics consumption is recorded. Postoperative nausea and vomiting is assessed using a 3-point scale; Ondansetron is administered to treat it. Any complications are recorded including those related to paravertebral block, such as Horner's syndrome, epidural spreading, hypotension that is not related to another reason, pneumothorax and adverse effect of local anesthetics. Patients will be checked 6, 12, and 24hr postoperatively for adverse effects, pain scores, and satisfaction scores. During hospital stay, the patients receive one tablet of Ultracet (tramadol plus acetaminophen) and Celebrex 200 mg two times a day for pain relief and oral lorazepam before bedtime. Residents or nurses on postoperative care station and surgical ward both are blinded to the treatment groups.

Sample size estimation: According to the data from a previous study paravertebral block compared to general anesthesia can reduce postoperative analgesia requirement (opioids + non-steroidal analgesics) up to 80%. Taking a 5% significance level as basis 30 patients per group is adequate.

Statistical analyses: Parametric data are expressed as mean and analyzed using Student's T test. The non-parametric data are expressed as median (range) and analyzed using Mann-Whitney U Test. Repeated analysis of variance is used to compare VAS between groups. The association between the techniques and incidence of complications are evaluated by chi-square or Fisher's exact test as appropriate. Statistical analyses will be performed using SPSS v20.0 for Windows (SPSS Inc., Chicago, USA). All tests are two-sided and a p-value of < 0.05 is accepted as statistically significant.

Inclusion Criteria:

• Samples are collected after patients' written informed consent in accordance with the Declaration of Helsinki and institutional research board approval. Sixty women (ASA I - III) undergoing unilateral mastectomy surgery with axillary dissection (radical lymph node removal or sentinel node biopsy) are enrolled. Exclusion criteria are general infection and local infection at the site of planned puncture, anatomic deformities of the thoracic spine, coagulation disorders, allergy against local anesthetics or contrast agents, such as Iopamiro 300, pregnancy or breast-feeding and BMI ≥ 30 kg/m2.

Exclusion Criteria:

• Patients who did not meet the inclusion criteria.