Secondary Logo

Journal Logo

Regional Anesthesia: Research Report

Guided Paravertebral Blocks With Versus Without Clonidine for Women Undergoing Breast Surgery

A Prospective Double-Blinded Randomized Study

Naja, Zoher M., MD*; Ziade, Fouad M., PhD; El-Rajab, Mariam A., MD; Naccash, Nicole, MD§; Ayoubi, Jean-Marc, MD, PhD

Author Information
doi: 10.1213/ANE.0b013e31828f28d6

In the past several years, paravertebral blocks (PVBs) have been introduced as an alternative to general anesthesia for breast cancer surgeries.1,2 PVB was shown to provide minimal airway intervention, less cardiopulmonary depression, decreased postoperative nausea and vomiting (PONV), as well as a reduction in pain, analgesic consumption, and length of hospital stay.3–12

This increased application of PVB, however, entails further improvement. Previous studies indicated that almost half (44%) of the patients undergoing breast surgery under PVB without clonidine consumed analgesics 24 hours after the operation.13 On the contrary, a study investigating the effectiveness of PVB versus general anesthesia in breast surgery reported that the use of PVB with clonidine produced advantageous results.14 This is supported by a number of studies relating clonidine to enhanced sensory blockade and prolonging the duration of surgical analgesia.15–17 Based on this, we hypothesized that the addition of clonidine may reduce analgesic consumption and prolong the postoperative pain-free period.

This study assessed the significance of adding clonidine to the anesthetic mixture for women undergoing mastectomy under PVB in terms of analgesic consumption. Secondary objectives were to assess pain scores, perioperative hemodynamic stability, hospital stay, and return to regular daily activity.


The study was approved by our IRB, and a written informed consent was obtained from all patients. The inclusion criteria consisted of patients with body mass index ≥30 kg/m2 and scheduled for breast cancer surgery (classified as overweight and obese according to the World Health Organization standardized criteria).18 Sixty patients were randomized and included in this study. The sealed envelope technique was used to assign the patient to either the clonidine group (PVB with clonidine) or control group (PVB without clonidine). Uncooperative patients or patients with known allergies to local anesthetics and opioids, anatomical abnormalities in the spine or paravertebral region, and coagulopathies were excluded.

The study was conducted from October 2007 to May 2010. Data collected included patient characteristics such as age, weight, height, ASA status (I, II, or III), type of surgery, intraoperative hemodynamic assessments, intraoperative fentanyl consumption, duration of surgery, total dosage of clonidine used, recovery room stay, incidence of PONV, hospital stay, time needed to resume daily activities, and overall patients’ and surgeons’ satisfaction.

Patients in the operating room were continually monitored for arterial blood pressure and heart rate. Hemodynamic instability was defined as changes of >25% from the baseline. The measurements reported in this study were at baseline preincision and postincision. Patients were moved to the postanesthesia care unit after surgery and were continuously monitored. Patients were moved to a regular ward once they regained full consciousness, hemodynamic stability, and had a pain score <3.

Incidence of PONV was recorded whether the patient had experienced nausea, or retching and/or vomiting for >10 minutes. Pain at rest and during attempted movement of the arm and consumption of supplemental analgesics were recorded at predetermined time intervals: immediately after surgery (in the recovery room) H0, H6 (at 6 hours postoperatively), H12, H24, H36, day 2, day 3, week 1, and week 2.

Pain scores were assessed during hospital stay using a visual analog scale (0 cm: no pain; 10 cm: worst possible pain). After discharge, the patients were instructed on how to record PONV, analgesic consumption, and pain scores. A numeric rating scale (graded 0–10) was used to evaluate their condition regarding motion-related pain and pain at rest. A rescue analgesic tramadol hydrochloride 100 mg (Tramal® retard, Grunenthal, Aachen, Germany) (maximum 2 tablets per day equivalent to 200 mg) was administered for any pain score >4.

The criteria for hospital discharge were pain score <4, hemodynamic stability, ability to drink water, bowel movement, and absence of nausea and vomiting.

Both patients’ and surgeons’ satisfaction were assessed using a scale rating from 1 to 3 (1 “unsatisfied,” 2 “partially satisfied,” 3 “satisfied”). Patients’ satisfaction was assessed verbally during the operation, during hospital stay, and up to 2 weeks after the operation. Immediately after the surgery, surgeons were asked to assess their satisfaction with the operating conditions. The surgeon’s satisfaction was based on the overall preoperative and postoperative clinical status of the patient. Anesthesiologists were not involved in the collection of surgeon or patient data.

All personnel involved in the postoperative management and data collection were blinded to patient group assignment. Neither the surgeons nor the anesthesiologists were aware of the content of the injected solution. The surgeons discharged the patients from the hospital according to the established clinical routine.

Before discharge, patients were told that they would be contacted by phone for postdischarge data collection. A prestructured questionnaire was used. All patients were instructed to note their analgesic needs and the pain they might encounter at home. They were familiarized with the numeric rating scale to report their perceived pain scores and were asked to keep a written record of their analgesic consumption. Nurses who were not involved in the study were responsible for the collection of these data at predetermined time intervals.

Paravertebral Nerve Block Technique

The PVB was performed at thoracic segments T1-T5 according to the type and nature of the surgical procedure. For simple mastectomy, the dermatomes T2-T5 were blocked at each individual level (4 separate injections). For partial mastectomy (removing large part of the breast), the corresponding levels were T2-T4 (3 separate injections). For modified radical mastectomy, the levels T1-T5 were injected (5 separate injections), resulting in an anaesthetized area including the axilla.14 The surgical incision was extensively discussed with the surgeon before the operation and consequently blocking the needed corresponding dermatomes.

In some patients, locating the landmarks at the T1-T5 levels may be difficult. In such cases, we attempted to determine the landmark at different levels (T11-T12) as a guide, and then deduced the landmark at T1-T5 level by assuming (based on our experience) that the distance between each 2 consecutive vertebrae is 2.0 to 2.5 cm. Hence, after locating the T11-T12 level, we moved gradually upwards to the T5 level. After aseptic preparation of the skin, the subcutaneous tissue was infiltrated with lidocaine 1 mL (1%) at each separate injection site. A 21-G insulated needle (10 or 15 cm Stimuplex®; B. Braun, Melsungen AG, Germany), already attached to a nerve stimulator (initially stimulating current 2.5–5 mA, 1 Hz, 9 V), was introduced perpendicularly to the skin at all planes. Contraction of the paraspinal muscles was observed initially. Then, after piercing of the costotransverse ligament, an appropriate muscular response from the intercostal muscles of the corresponding level was sought and the needletip manipulated into a position to allow the same muscular response while the stimulating current was reduced to 0.4 to 0.6 mA. At this point, 0.06 mL/kg of a local anesthetic mixture was injected at each level. Each 20 mL of the injected mixture contained: 6.5 mL lidocaine 2%, 6 mL lidocaine 2% with epinephrine 5 µg·mL−1, 7 mL bupivacaine 0.5%, and either 0.5 mL clonidine 150 µg·mL−1 or 0.5 mL plan saline depending on the assigned group.

In some patients, visual detection of the intercostal muscle response was not possible so we had to rely on manual palpation or verbal confirmation of the patient’s sensation (all patients in this study were kept awake during this procedure). In case of unsuccessful detection of muscle response, the landmark was relocated a few millimeters upwards or downwards from the initial position.

Five minutes after completion of the PVB, adequate distribution of cutaneous anesthesia was verified with a pinprick test. After PVB, on patients request, additional light sedation was administered intraoperatively by incremental IV doses of midazolam of 1 mg to a maximum dose of 5 mg, resulting in a sedation level ranging from “alert but calm,” to “sedated but responsive to verbal command.” Successful PVB block is defined as the complete blockade of the incision area without the need to repeat the PVB, or perform local infiltration during the surgery, or convert to general anesthesia.

Statistical Analysis

Results were reported as mean and SD, or frequency and percentage as appropriate. The 95% confidence interval (CI; Bonferroni corrected for multiple comparisons) for the reduction in the different variables under study was calculated either for the difference between 2 means formula (scale variables) or the difference between 2 proportions formula as appropriate. Data were statistically analyzed by χ2 test, t test (2-tailed; preceded with Levene test for equality of variances) for scale variables or Mann-Whitney test (2-tailed) for ordinal variables. Generalized pivotal methods were used to calculate the 95% CI (corrected for multiple comparisons) for the ratio of means. A mixed effect model (lag 1 covariance, autoregressive) was conducted on both pain at rest and pain with movement to test for the difference between the 2 groups with repeated measurements over time and to assess the interaction between group and time. Specification of the tests used for the different end points is given in the table legends.

The effect of multiple comparisons on P values was corrected using Bonferroni correction and were reported (corrected) in the results. Hence, P <0.05 was considered significant.

A sample size of 27 patients in each study group with a power of 90% and α = 0.05 was found to be necessary to detect a difference between 6.7% and 44%; the previously reported percentages of patients who required analgesics within 24 hours under PVB with clonidine versus without clonidine, respectively.13,14


Sixty patients were enrolled in this study. All patients, in the control and clonidine groups, finished their surgical procedure under PVB, and none required conversion to general anesthesia. Most of the patients (57 patients, 95%) had successful PVB blockade, and only 3 patients needed local infiltration during the surgery due to incomplete block of the axillary area. Due to loss of follow-up and patients’ refusal, 54 patients were included in the final analysis: 28 patients in the clonidine group and 26 patients in the control group (Fig. 1).

Figure 1
Figure 1:
Patients included in the study.

There were no significant differences in patient characteristics (Table 1). The method of confirmation of surface landmarks and motor response are reported in Table 2. There were no significant differences in hemodynamic measurements, intraoperative fentanyl consumption, duration of surgery, recovery room stay, and incidence of PONV and hospital stay between the 2 groups (Table 3). Heart rate and arterial blood pressure measurements for all patients in both groups during the operation remained within 15% from the baseline (preincision).

Table 1
Table 1:
Patients Characteristics
Table 2
Table 2:
Confirmation of Landmark and Muscle Contraction Using Guided Paravertebral Blocks
Table 3
Table 3:
Intraoperative and Postoperative Variables

In terms of analgesic consumption, the clonidine group consumed significantly less opioids than the control group at 48 hours postoperatively (Fig. 2), with the lower magnitude of the 95% CI limits for the difference equal to −6.6%. Differences in postoperative pain scores at rest were observed from H24 with the lower 95% CI limits for the difference equal to −0.4, −1.3, −1.1, respectively (Table 4). However, only magnitudes of >1 reveal a clinically significant change in pain scores, indicating clinically important changes only at 48 and 72 hours postoperatively.19,20 The use of a generalized pivotal method to calculate the 95% CI for the ratio of the 2 means resulted in the following intervals: (0.64–2.96), (0.72–2.59), (0.72–2.22), (1.09–3.61), (2.04–9.04), (2.54–16.55), for pain scores at rest during the follow-up period (H0, H6, H12, H24, H48, and H72) respectively. The lower 95% CI limit for the ratio of the 2 means was superior to one at H24, H48, and H72 revealing that the pain scores at rest were significantly lower in the clonidine group compared with the control group.

Figure 2
Figure 2:
Analgesic consumption.
Table 4
Table 4:
Postoperative Pain Scores (0–10) at Rest and with Movement

With shoulder movement, the average pain scores were lower in the clonidine group from H12 to H72 postoperatively with the lower 95% CI limits for the differences equal to −0.3, −0.5, −1.4, and −1.1, respectively (Table 4), indicating clinically important changes only at 48 and 72 hours postoperatively. The use of a generalized pivotal method to calculate the 95% CI for the ratio of the 2 means resulted in the following intervals: (0.76–2.42), (0.88–2.13), (1.10–3.15), (1.32–6.38), (1.33–8.42), (1.00–9.89), for pain scores with movement during the follow-up period (H6, H12, H24, H48, H72, and week 1) respectively. The lower 95% CI limit for the ratio of the 2 means was superior to one at H24, H48, and H72 revealing that the pain scores with shoulder movement were significantly lower in the clonidine group compared with the control group.

A mixed effect model (lag 1 covariance, autoregressive) was conducted on both pain at rest and pain with movement. Type III tests of fixed effects indicated that both the group and the time factors contributed to the model with P < 0.002 both at rest and with movement. The estimates of fixed effects suggest that pain scores at rest in the clonidine group are lower by 0.89 than the control group (P < 0.0001), and the estimates at time intervals H6, H12, H24, H48, and H72 were 1.09, 1.54, 1.61, 1.43, and 0.94 (all P values <0.0001). The estimates of fixed effects suggest that pain scores on movement in the clonidine group are lower by 1.02 than the control group (P = 0.002), and the estimates at time intervals 6, 12, 24, 48, and 72 hours were 2.24, 2.09, 1.72, 1.16, and 0.46 (all P values <0.0001). However, when the term of interaction between time and group was included, in both models at rest and with movement P < 0.001 was obtained. Pain scores seemed to decrease faster over time in the clonidine group compared with the control group.

The average time needed to resume normal daily activity was shorter in the clonidine group with 95% CI for the difference (−4.6 to −0.2)(Table 3). Levene test indicated unequal variances (P < 0.001) between the 2 groups, and the log-transformation for pain scores indicated better fit (with P = 0.393 using Lilliefors test for normality). Hence, we used a generalized pivotal method to calculate the 95% CI (1.14–1.62) for the ratio of the 2 means, which indicated that the mean time for the control group was longer than the mean time for the clonidine group. There were no differences between the patients’ and surgeons’ satisfaction in the 2 groups (Table 3).


The results of this randomized double-blind study demonstrated that the addition of clonidine as an adjuvant to the local anesthetic mixture increased the effectiveness of the block-analgesic consumption and pain scores were lower from day 1 up to day 3 postoperatively.

Several studies compared the different anesthetic mixtures used in PVB, mainly bupivacaine, ropivacaine, and levobupivacaine; however, 1 study showed that the addition of clonidine to the anesthetic mixture enhanced pain management postoperatively and decreased morphine consumption.15 Clonidine is an (α-2)-adrenoreceptor agonist which produces analgesia by a non-opioid mechanism. It has been shown to prolong the duration of analgesia when administered intrathecally, in the epidural space or in the paravertebral space.16,17 However, other studies reported contradicting results indicating that clonidine combined with a long-acting local anesthetic does not prolong postoperative analgesia.21

In our study, the clonidine group experienced an extended postoperative free-pain period. The proportion of patients consuming analgesics 24 hours postoperatively was 13.3% in the clonidine group which is significantly less than that observed in the previously published case series (44%).13

The prolonged postoperative analgesia experienced by our clonidine group is consistent with previously conducted studies demonstrating that the use of clonidine can reduce hypersensitivity and cytokines expression in established nerve injury and may also have a distinct effect on chronic pain that usually develops after surgery.22,23 In addition, PVB may also reduce the prevalence and severity of pain up to 1 year after breast cancer surgery.10 We speculate that this combination of clonidine and PVB might have played a role in delaying pain development and reducing the effect of chronic pain. However, further follow-up studies are still needed.

Similar studies reported hypotension as a possible adverse effect of clonidine; however, this was not experienced by any patient in this study.15,21 While other studies used a fixed quantity of clonidine per patient, in this study the injected quantity was based on the weight of the patient which was on average 85.6 µg of clonidine per patient.15,24 Thus, these differences in the clonidine administration and quantity may be a possible reason for the conflicting results concerning the incidence of hypotension.

A multilevel PVB approach was used in this study as the sole anesthetic technique. Both single- and multilevel PVB have been reported to provide good postoperative analgesia; however, multilevel injection of PVB has been shown to produce a more reliable sensory block and longitudinal spread for the same volume of anesthetic mixture being injected in 1 single site.25 Most of the patients in this study (57 patients, 95%) completed the operation with PVB, and only 3 patients needed local infiltration before completing the surgery due to incomplete block of the axillary area. The result is supported by other studies reporting that the incidence of failed block is approximately 5% to 10% of the cases.26

A further point of interest is the choice of the PVB technique. Different PVB techniques such as the classical approach, nerve stimulator, and ultrasound have been used in breast cancer surgery with varying results. Conducting PVB on obese patients using the classical approach of bone contact is associated with technical difficulties and an increase of block failure.24,27 Alternatively, a guided PVB technique, such as the nerve stimulator or the ultrasound, increases accuracy, rendering higher success rates and reduced rates of complications.14,28,29

Although all patients included in this study had body mass index ≥30 kg/m2, there is no compelling reason to believe that the effects of clonidine as a local anesthetic adjuvant for PVB would be any different in obese women compared with nonobese women or men.

In summary, our study demonstrates that PVB is an effective anesthetic technique and that the addition of clonidine enhances both the anesthetic and analgesic efficacy of PVB in patients undergoing mastectomy with no reported adverse effects or complications.


Name: Zoher M. Naja, MD.

Contribution: This author helped design the study.

Attestation: Zoher M. Naja approved the final manuscript.

Name: Fouad M. Ziade, PhD.

Contribution: This author helped analyze the data.

Attestation: Fouad M. Ziade approved the final manuscript.

Name: Mariam A. El-Rajab, MD.

Contribution: This author helped conduct the study.

Attestation: Mariam A. El-Rajab approved the final manuscript.

Name: Nicole Naccash, MD.

Contribution: This author helped revision of paper.

Attestation: Nicole Naccash approved the final manuscript.

Name: Jean-Marc Ayoubi, MD, PhD.

Contribution: This author helped in manuscript preparation.

Attestation: Jean-Marc Ayoubi approved the final manuscript.

This manuscript was handled by: Terese T. Horlocker, MD.


1. Vila H Jr, Liu J, Kavasmaneck D. Paravertebral block: new benefits from an old procedure. Curr Opin Anaesthesiol. 2007;20:316–8
2. Najarian MM, Johnson JM, Landercasper J, Havlik P, Lambert PJ, McCarthy D. Paravertebral block: an alternative to general anesthesia in breast cancer surgery. Am Surg. 2003;69:213–8; discussion 218
3. Nielsen KC, Guller U, Steele SM, Klein SM, Greengrass RA, Pietrobon R. Influence of obesity on surgical regional anesthesia in the ambulatory setting: an analysis of 9,038 blocks. Anesthesiology. 2005;102:181–7
4. Coveney E, Weltz CR, Greengrass R, Iglehart JD, Leight GS, Steele SM, Lyerly HK. Use of paravertebral block anesthesia in the surgical management of breast cancer: experience in 156 cases. Ann Surg. 1998;227:496–501
5. Pusch F, Freitag H, Weinstabl C, Obwegeser R, Huber E, Wildling E. Single-injection paravertebral block compared to general anaesthesia in breast surgery. Acta Anaesthesiol Scand. 1999;43:770–4
6. Klein SM, Bergh A, Steele SM, Georgiade GS, Greengrass RA. Thoracic paravertebral block for breast surgery. Anesth Analg. 2000;90:1402–5
7. Boughey JC, Goravanchi F, Parris RN, Kee SS, Kowalski AM, Frenzel JC, Bedrosian I, Meric-Bernstam F, Hunt KK, Ames FC, Kuerer HM, Lucci A. Prospective randomized trial of paravertebral block for patients undergoing breast cancer surgery. Am J Surg. 2009;198:720–5
8. Arunakul P, Ruksa A. General anesthesia with thoracic paravertebral block for modified radical mastectomy. J Med Assoc Thai. 2010;93 Suppl 7:S149–53
9. Naccache N, Jabbour H, Nasser-Ayoub E, Abou Zeid H, Naja Z. Regional analgesia and breast cancer surgery. J Med Liban. 2009;57:110–4
10. Kairaluoma PM, Bachmann MS, Rosenberg PH, Pere PJ. Preincisional paravertebral block reduces the prevalence of chronic pain after breast surgery. Anesth Analg. 2006;103:703–8
11. Greengrass R, O’Brien F, Lyerly K, Hardman D, Gleason D, D’Ercole F, Steele S. Paravertebral block for breast cancer surgery. Can J Anaesth. 1996;43:858–61
12. Weltz CR, Greengrass RA, Lyerly HK. Ambulatory surgical management of breast carcinoma using paravertebral block. Ann Surg. 1995;222:19–26
13. Naja ZM, Naccache N, Ziade F, El-Rajab M, Itani T, Baraka A. Multilevel nerve stimulator-guided paravertebral block as a sole anesthetic technique for breast cancer surgery in morbidly obese patients. J Anesth. 2011;25:760–4
14. Naja MZ, Ziade MF, Lönnqvist PA. Nerve-stimulator guided paravertebral blockade vs. general anaesthesia for breast surgery: a prospective randomized trial. Eur J Anaesthesiol. 2003;20:897–903
15. Burlacu CL, Frizelle HP, Moriarty DC, Buggy DJ. Fentanyl and clonidine as adjunctive analgesics with levobupivacaine in paravertebral analgesia for breast surgery. Anaesthesia. 2006;61:932–7
16. Constant I, Gall O, Gouyet L, Chauvin M, Murat I. Addition of clonidine or fentanyl to local anaesthetics prolongs the duration of surgical analgesia after single shot caudal block in children. Br J Anaesth. 1998;80:294–8
17. Iskandar H, Guillaume E, Dixmérias F, Binje B, Rakotondriamihary S, Thiebaut R, Maurette P. The enhancement of sensory blockade by clonidine selectively added to mepivacaine after midhumeral block. Anesth Analg. 2001;93:771–5
18. World Health Organization. Obesity: Preventing and Managing the Global Epidemic. Report of a WHO Consultation on Obesity. 1997 Geneva, Switzerland World Health Organization
19. Todd KH, Funk KG, Funk JP, Bonacci R. Clinical significance of reported changes in pain severity. Ann Emerg Med. 1996;27:485–9
20. Kelly AM. Does the clinically significant difference in visual analog scale pain scores vary with gender, age, or cause of pain? Acad Emerg Med. 1998;5:1086–90
21. Culebras X, Van Gessel E, Hoffmeyer P, Gamulin Z. Clonidine combined with a long acting local anesthetic does not prolong postoperative analgesia after brachial plexus block but does induce hemodynamic changes. Anesth Analg. 2001;92:199–204
22. Romero-Sandoval A, Eisenach JC. Perineural clonidine reduces mechanical hypersensitivity and cytokine production in established nerve injury. Anesthesiology. 2006;104:351–5
23. Lavand’homme PM, Eisenach JC. Perioperative administration of the alpha2-adrenoceptor agonist clonidine at the site of nerve injury reduces the development of mechanical hypersensitivity and modulates local cytokine expression. Pain. 2003;105:247–54
24. Adams JP, Murphy PG. Obesity in anaesthesia and intensive care. Br J Anaesth. 2000;85:91–108
25. Naja ZM, El-Rajab M, Al-Tannir MA, Ziade FM, Tayara K, Younes F, Lönnqvist PA. Thoracic paravertebral block: influence of the number of injections. Reg Anesth Pain Med. 2006;31:196–201
26. Lönnqvist PA, MacKenzie J, Soni AK, Conacher ID. Paravertebral blockade. Failure rate and complications. Anaesthesia. 1995;50:813–5
27. Cotter JT, Nielsen KC, Guller U, Steele SM, Klein SM, Greengrass RA, Pietrobon R. Increased body mass index and ASA physical status IV are risk factors for block failure in ambulatory surgery—an analysis of 9,342 blocks. Can J Anaesth. 2004;51:810–6
28. Hara K, Sakura S, Nomura T, Saito Y. Ultrasound guided thoracic paravertebral block in breast surgery. Anaesthesia. 2009;64:223–5
29. Naja MZ, Ziade MF, El Rajab M, El Tayara K, Lönnqvist PA. Varying anatomical injection points within the thoracic paravertebral space: effect on spread of solution and nerve blockade. Anaesthesia. 2004;59:459–63
© 2013 International Anesthesia Research Society