Journal of Clinical Anesthesia (2007) 19, 506–511

Original contribution

Interpleural versus epidural analgesia with ropivacaine forpostthoracotomy pain and respiratory functionB

Vedat Yildirim MD (Assistant Professor)a,Hakki Tankut Akay MD (Staff Cardiovascular Surgeon)b,⁎,Hakan Bingol MD (Associate Professor)b, Cengiz Bolcal MD (Assistant Professor)b,Kursad Oz MD (Staff Cardiovascular Surgeon)b, Erkan Kaya MD (Resident)b,Ufuk Demirkilic MD (Professor)b, Harun Tatar MD (Professor and Chairman)b

aDepartment of Anesthesiology and Reanimation, Gülhane Military Medical Academy, Ankara 06552, TurkeybDepartment of Cardiovascular Surgery, Gülhane Military Medical Academy, Ankara 06552, Turkey

Received 16 June 2006; revised 28 March 2007; accepted 11 April 2007

0d

Keywords:Interpleural analgesia;Thoracic epiduralanalgesia;

Thoracotomy;Postoperative pain;Atelectasis

AbstractStudy Objective: To evaluate the impact of interpleural analgesia (IP) on postthoracotomy pain andrespiratory function as an alternative to thoracic epidural analgesia (TEA).Design: Prospective, randomized study.Setting: Tertiary-care military hospital.Patients: Sixty young patients scheduled for elective thoracic surgery (correction of aorta coarctationand patent ductus arteriosus).Interventions: Patients were randomized into two groups to receive either IP or TEA forpostthoracotomy pain management.Measurements: Patients in the IP group (n = 30) had a catheter inserted between the parietal andvisceral pleura by a surgeon, and 0.2% ropivacaine was given through this catheter. In the TEA group,ropivacaine was administered through a thoracic epidural catheter. The impact of both methods on paincontrol, respiratory function, and pulmonary complications was analyzed and compared.Main Results: The frequency of atelectasis and pleural effusion was also significantly high in theIP group (P b 0.01). Respiratory function and postoperative pain scores were better in the TEA group(P b 0.01). Arterial blood gas analysis on the fifth postoperative day was significantly better in theTEA group.Conclusion: Thoracic epidural analgesia has more beneficial effects on respiratory function andpostoperative pain after thoracotomy than does IP.© 2007 Elsevier Inc. All rights reserved.

☆ The study was performed at Gülhane Military Medical Academy (Ministry of Defence). There was no external financial support.⁎ Corresponding author. Tel: +90 312 3045218; fax: +90 3123045200.E-mail address: tankutakay@gmail.com (H.T. Akay).

952-8180/$ – see front matter © 2007 Elsevier Inc. All rights reserved.oi:10.1016/j.jclinane.2007.04.005

507Analgesia for thoracotomy pain

1. Introduction

Thoracotomy is associated with severe and intensive pain,which is considered to occur as a consequence of tissuedamage to the ribs, muscles, and peripheral nerves [1]. Thispain prevents deep breathing and effective coughing, whichoften result in atelectasis and other pulmonary complica-tions. Therefore, postoperative thoracotomy pain is thoughtto be an important factor causing postoperative impairmentof respiratory function and gas exchange. Although systemicopioids traditionally have formed the basis for the treatmentof postthoracotomy pain, continuous epidural administrationof analgesics provides better pain control. Various analgesictechniques have been developed to treat postoperativethoracotomy pain, including the administration of local,regional, or systemic analgesics [1-7]. Thoracic epiduralanalgesia has been accepted as the best method for themanagement of postoperative pain [3,4]; however, it is notsuitable for all patients and it carries potential risks andlimitations such as dural perforation, bleeding, infection,hypotension, and bradycardia [8-10]. Intrapleural adminis-tration of anesthetics may provide an alternative to thoracicepidural anesthesia [11,12].

In this study, we compared the effects of intrapleuralropivacaine with thoracic epidural anesthesia on earlypostoperative pain and respiratory function.

2. Materials and methods

2.1. Patient selection

After receiving study approval from the Gülhane MilitaryMedical Academy's ethics committee and patients' informedconsent, 60 young patients scheduled for thoracic surgery(correction of aorta coarctation [n = 26] and closure of patentductus arteriosus [n = 34]) were allocated in this prospective,randomized study. The ages of the patients are expected to beolder for these anomalies. Because our institution is the mainmilitary medical center in our country, the young patientpopulation in this study consisted of soldiers who hadcongenital defects (not diagnosed in childhood) and thediagnosis of all patients was made in the army hospitals. Allpatients in both groups underwent classic posterolateralthoracotomy, performed by the same surgeon, involving thedivision of the latissimus dorsi and serratus anterior musclesafter a posterolateral incision at the fifth intercostal space.Patients who underwent thoracic surgery (surgery formalignant primary or metastatic pulmonary tumors, chestwall reconstruction, minithoracotomy, and other nonstandardincisions) were not included in the study. Patients with ahistory of allergy to local anesthetics, obesity (weight greaterthan 100 kg), significant central nervous system, hepatic, orrenal disease, those with inadequate respiratory functionaltests (vital capacity [VC], forced expiratory volume in 1 s

[FEV1], and FEV1/VC ratio [%]), or history of previous lungsurgery were also excluded from the study.

2.2. Technique

All patients received a standard premedication consistingof 40 mg of famotidine and 10 mg diazepam orally. Onpatient arrival at the operating room, pulse oximeter andelectrocardiogram (ECG) monitors were attached. Beforeinduction of anesthesia, invasive arterial monitoring usingright radial artery access and central venous catheterizationvia the right internal jugular vein were maintained usinglocal anesthesia and light intravenous (IV) sedation. Afterstarting IV fluid and oxygenation, anesthesia was inducedwith 2.5 mg/kg of propofol and 1.5 μg/kg of fentanyl. Thetrachea was intubated with a double-lumen endobronchialtube (Sherı bronch endobronchial tube, right 39-41Fcatheter, Kendall Healthcare Products Co, Mansfield, MA)after administration of vecuronium 0.1 mg/kg for neuro-muscular relaxation, and the lungs were ventilated mechani-cally. Anesthesia was maintained with isoflurane and nitrousoxide in oxygen. Patients were randomized to receive inter-pleural analgesia (IP group, n = 30) or thoracic epiduralanalgesia (TEA group, n = 30), both with ropivacaine.Randomization was by patient unit or identification number(ending numeral odd or even) and took place in theoperating room.

For group IP, the surgeon inserted an 18-gauge intra-pleural catheter in all patients at the end of the operation,before chest closure (Epidural minipach system 1, size 18 G,Portex Ltd, Hythe, Kent, UK) and under direct vision. Thetip of the catheter (with 20 terminal side ports) was placedtoward the fourth or fifth intercostal space. All catheters weresutured to the posterior parietal pleura. Interpleural anesthe-sia was initiated intraoperatively with infusion of 10 mLropivacaine 0.2% during a 30-min period. A continuousinfusion (0.1 mL/kg per hour) with 0.2% ropivacainesupplemented with fentanyl 2 μg/mL was maintained for aminimum of 3 days.

In group II (TEA group), patients received an epiduralcatheter inserted from the T5-6 interspace with a paramedianapproach in the lateral decubitus position a day before theinduction of general anesthesia. Ten milliliters of 0.2%ropivacaine with 0.5 μg/kg of fentanyl was administeredbefore the induction of general anesthesia and ropivacaine0.2% with fentanyl 2 μg/mL was infused at a rate of 5 to 15mL/h (Abbott Acute Pain Manager-APM, Pain ManagerProvider, North Chicago, IL). The epidural infusion wasplanned to continue for 3 days.

2.3. Operative technique

The same surgeon performed all the operative procedures.Left posterolateral thoracotomy was the surgical approachusing the fourth intercostal space. Patent ductus arteriosus

Table 1 Demographic data of the patients

IP group TEA group P

No. of patients 30 30Age, mean ± SD (y) 23.2 ± 2.1 22.8 ± 2.2 NSWeight (kg) 74.8.1 ± 7.2 76.4 ± 8.1 NSHeight (cm) 173.5 ± 5.8 174.5 ± 6.1 NSIntraoperative fentanyldose (μg)

93 ± 34 95 ± 36 NS

Operation time (min) 93.6 ± 23.5 89.3 ± 25.6 NS

NS indicates nonspecific.

Table 3 Comparison of ABG measurements preoperativelyand in the 5th postoperative day

IP group TEA group P

PaO2 (mm Hg)Preoperative 89.1 ± 4.2 87.9 ± 3.7 0.782Postoperative 74.1 ± 7.4 84.7 ± 6.3 0.031P b0.001 0.754PaCO2 (mm Hg)Preoperative 36.1 ± 3.4 35.3 ± 4.2 0.651Postoperative 39.3 ± 4.4 34.4 ± 5.7 0.039P 0.037 0.698O2 saturation (%)Preoperative 99.1 ± 1.3 99.2 ± 1.7 0.812Postoperative 94.1 ± 1.8 98.5 ± 0.6 0.04P 0.046 0.875

508 V. Yildirim et al.

anomalies were corrected by the double ligation andtransfixion technique. In patients with coarctation of aorta,we preferred to resect the coarctated segment and interpose aprosthetic graft.

2.4. Postoperative period

Extubation was performed when the patient was awake,cooperative, normothermic (core body temperature N36°C),and with acceptable drainage and hemodynamic and arterialblood gas (ABG) analysis (pH N7.35 and PaO2 N75 mm Hgon 40% inspired oxygen, PaCO2 b45 mm Hg). Patients wereobserved in the postoperative care unit for at least 24 h.After the period in the postoperative care unit, patients wereobserved in the ward. The following parameters werecompared between groups: VC, forced expiratory volume in1 s (FEV1), FEV1/VC (%), and ABG (PO2, PCO2, and O2

saturation). Blind spirometric measurements and chest radio-graphs were evaluated routinely 1 day before the operation,then 5 and 30 days after the operation. Radiographs wereexamined by the same radiologist. Atelectases were recordedwhen showing a clear radiological shadow of a width of morethan 15 mm. Linear atelectases were also recorded. Painintensity at rest (Visual Analog Scale [VAS]-R) and duringfunction (VAS-E) was independently evaluated by the samestudy-blinded anesthesiologist in the first 3 postoperativedays using a 10-cm VAS, with end points labeled “no pain”and “worst possible pain imaginable.” Pain with effort wasevaluated by asking the patient to breathe deeply or to cough.VAS scores lower than 4 during function were considered torepresent satisfactory analgesia. Whenever VAS-E scoreswere greater than 5 and patients complained about pain, theyreceived supplemental IV fentanyl doses of 0.5 μg/kg.

Table 2 Comparison of preoperative and postoperative pulmonary fu

Before operation 5 Days postop

IP TEA P IP

FEV1 (%) 82.3 ± 6.5 81.1 ± 4.2 0.674 61.2 ± 2.7VC (%) 95.1 ± 6.2 94.1 ± 2.2 0.863 68.8 ± 4.2FEV1/VC 95.2 ± 8.5 97.2 ± 7.6 0.788 66.1 ± 9.4

Patients were routinely administered nonsteroidal anti-inflammatory drugs orally two times daily.

2.5. Statistical analysis

Statistical analysis was performed with SPSS/PC softwareversion 10.0 (SPSS Inc, Chicago, IL). Clinical data areexpressed as mean values ± SD. Differences between thegroups (demographic data, anesthesia and operation time,total analgesic use, and continuous variables such as hospitalstay time, PaCO2, SpO2, and spirometric values) werecompared by independent-samples t test. For nonparametricdata (pain scores), the Mann-Whitney rank sum test was usedto compare the two groups. A P value less than 0.05 wasaccepted as statistically significant.

3. Results

Patient characteristics are shown in Table 1. There was nostatistical significant difference between the groups indemographic data or preoperative respiratory function(Tables 1 and 2). Preoperative and intraoperative hemody-namic parameters were normal in all groups. Patients'tracheas were extubated after a mean of 12 ± 4.3 h.Mechanical ventilation times were similar in both groups.Pulmonary function tests are compared in Table 2. FEV1

showed a significant decrease 5 days after operation in the IP

nction

eratively 30 Days postoperatively

TEA P IP TEA P

74.1 ± 5.2 0.001 67.4 ± 4.1 79.1 ± 3.4 0.03887.2 ± 6.4 b0.001 82.3 ± 3.4 94.2 ± 3.1 0.03185.7 ± 8.8 b0.001 83.5 ± 7.2 94.6 ± 2.5 b0.001

Fig. 1 Visual Analog Scale scores of patients at rest.

509Analgesia for thoracotomy pain

group (IP, 61.2% ± 2.7%; TEA, 74.1% ± 5.2%; P b 0.001).Vital capacity was significantly more restricted in the IPgroup at 30 days (IP, 82.3% ± 3.4%; TEA, 94.2% ± 3.1%;P b 0.001). Arterial blood gas analysis of both groups issummarized in Table 3. On the fifth postoperative day, PaO2(mm Hg) and O2 saturation (%) were significantly higher(84.7 ± 6.3 vs 74.1 ± 7.4 mm Hg for PaO2), and PaCO2 (mmHg) was significantly lower in the TEA group (34.4 ± 5.7 vs39.3 ± 4.4 mm Hg).

Postoperative hospital stay time was markedly higher inthe IP group when compared with the TEA group (7.7 ± 0.8vs 6.0 ± 1.1, P b 0.01). VAS pain scores at rest and duringeffort were lower in the TEA group in the third postoperativeday (5.8 ± 1.27 vs 4.9 ± 1, P = 0.037, for VAS at rest and 6.5± 1.3 vs 5.1 ± 1.2, P = 0.018, for VAS at effort). VAS scoresof both groups are shown in Figs. 1 and 2. Total additionalfentanyl consumption was significantly found to beincreased in the IP group versus the TEA group (640 ±11.5 μg vs 368 ± 30 μg, P b 0.001). There was noreoperation or death in either study group. Similarly, no

Fig. 2 Visual Analog Scale score

catheter- or drug-related complications were recorded duringTEA or IP anesthesia.

On the 5th and 30th postoperative days, atelectasis wascompared between the groups. The frequency of atelectasiswas significantly higher in the IP group on the 5thpostoperative day (P b 0.01), but there was no differencein atelectasis in either group on the 30th postoperative day(Table 4). Chest radiography on the third postoperative dayshowed no evidence of phrenic nerve injury. Finally, therewas no lung infection noted in either group.

4. Discussion

Postthoracotomy pain is considered to be severe andintense as a consequence of tissue damage to the ribs,muscles, and peripheral nerves. It is a complex phenomenoninvolving multiple neurotransmitters and excitatory andinhibitory pathways that are difficult to target and quantify.Pain is exacerbated by motion and coughing. These actions

s of patients breathing deeply.

Table 4 Postoperative atelectasis and pleural effusion

Atelectasis

Present Absent

Postoperative 5th day, n (%)IP group 14 (46.6) 16 (53.4)TEA group 6 (20) 24 (80)P 0.027Postoperative 30th day, n (%)IP group 2 (6.6) 28 (93.4)TEA group 1 (3.3) 29 (96.3)P 0.413

510 V. Yildirim et al.

result in weak, superficial breathing and nonproductivecoughing. The earliest change in respiratory mechanicsduring the postoperative period is the decrease in FEV1 andforced VC. Decreased functional residual capacity andalveolar collapse during anesthesia may be impaired furtherby restrictive ventilation caused by postoperative pain andabnormal respiration pattern. Therefore, pain managementplays a vital role in decreasing morbidity and alteration oflung function after thoracotomy [1,3].

Various analgesic techniques have been developed to treatpostoperative thoracotomy pain [2-13]. However, the acutepain condition associated with thoracotomy continues to be achallenge to clinicians [14]. Systemic administration ofopioids is the simplest and most common method to provideanalgesia for postoperative pain; unfortunately, systemicopioid administration may not be adequate for treating theintense postoperative pain associated with thoracotomy.

Thoracic epidural analgesia has been accepted for themanagement of postthoracotomy pain. It not only providesexcellent pain control, but also prevents excessive sedationwith systemic opioids. Salomaki et al [15] comparedpatients receiving IV versus thoracic epidural fentanyl;patients in the epidural group had less fentanyl consumptionand better preservation of respiratory function than the IVgroup. In a meta-analysis of 65 studies, Ballantyne et al [16]concluded that postoperative epidural pain control maysignificantly decrease pulmonary morbidity. However, it isnot appropriate for certain patients, especially those withcoagulation disorders, spinal deformities, or neurologicdisorders, or who have received anticoagulant therapy [13].

Interpleural analgesia—the administration of local anes-thetic agents through a catheter positioned inside the pleuralcavity to diffuse across the parietal pleura and anesthetizethe intercostal nerves—may be another alternative [5].Several studies have shown limited improvement inanalgesia with IP [12,17,18]. Explanations for the limitedanalgesic efficacy of IP include loss of local anestheticthrough the chest tube, dilution of local anesthetic withblood and exudative fluid present in the pleural cavity,binding of local anesthetic with proteins, and altereddiffusion across the parietal pleural after surgical manipula-tion and inflammation [18,19].

Another option to minimize postthoracotomy pain isextrapleural regional anesthesia. Extrapleural regionalanesthesia depends primarily on diffusion of the analgesicagent into the paravertebral region [20]. Local anestheticagents affect not only ventral nerve roots but also afferentfibers of the posterior primary ramus. Posterior ligamentsof the posterior primary ramus innervate posterior spinalmuscles and skin. These ligaments are usually traumatizedduring posterolateral thoracotomy [21]. The procedureinvolves intermittent administration of local anesthetic tothis area through a catheter placed in the extrapleuralregion. Depth and width of the anesthetized area dependon diffusion of the local anesthetic. This finding is one ofthe method's objective criteria for providing pain reliefafter thoracotomy.

In our study, we aimed to compare two analgesicmodalities for postthoracotomy pain (epidural vs IP). Wefound that TEA is superior to IP. Respiratory function tests,VAS scores, and frequency of atelectasis were significantlylower in patients who received TEA.

Determining the quality and intensity of pain is veryimportant. The VAS is a reliable method of evaluating painlevel [7]. We used VAS scoring in our patients both at restand in deep breathing. We believe that it is also important toevaluate pain in two different conditions (at rest or witheffort). In both conditions, TEA was found to be superiorto IP.

There are studies in the literature reporting commonlyused opioid–local anesthetic mixtures including fentanyl-bupivacaine, morphine-bupivacaine, and fentanyl-ropiva-caine [6,22,23]. In this study, we preferred to useropivacaine, which has a similar onset and duration ofaction to that of bupivacaine, but with a more enhancedsafety profile because of decreased cardiotoxicity and a lessprofound motor block than bupivacaine. In our study, therewere no systemic side effects related to ropivacaine.

There may be some limitations to our study. The epiduralgroup received local anesthetic before incision, whereas theother group received it after incision. We could have waiteduntil after incision and administer bolus injections to bothgroups at the same time. However, the interpleural groupreceived the local anesthetic approximately 1 h after theincision (just after the end of the dissections), whereas theepidural group received the anesthetic agent 1 h sooner (therewas no wait between induction of general anesthesia andthoracotomy). We are not sure that this timing had anyimportant effect. If it did, then it may also be considered anadvantage of epidural analgesia (not waiting for the incisionand thoracotomy but starting the analgesia just before theinduction of general anesthesia).

Thoracic epidural analgesia with ropivacaine is abeneficial treatment modality for the management ofpostthoracotomy pain. In our study, epidural administrationof ropivacaine provided superior pain relief, with moreimproved respiratory function and a decreased rate ofsupplemental analgesic use, than did IP.

511Analgesia for thoracotomy pain

References

[1] Warner DO. Preventing postoperative pulmonary complications.Anesthesiology 2000;92:1467-72.

[2] Kaiser AM, Zollinger A, Lorenzi D, Largiader F, Weder W.Prospective, randomized comparison of extrapleural versus epiduralanalgesia for post-thoracotomy pain. Ann Thorac Surg 1998;66:367-72.

[3] Craig DB. Postoperative recovery of pulmonary function. AnesthAnalg 1981;60:46-52.

[4] Mourisse J, Hasenbos MA, Gielen MJ, Moll JE, Cromheecke GJ.Epidural bupivacaine, sufentanil or the combination for post-thoracotomy pain. Acta Anaesthesiol Scand 1992;36:70-4.

[5] McKenzie AG, Mathe S. Interpleural local anaesthesia: anatomicalbasis for mechanism of action. Br J Anaesth 1996;76:297-9.

[6] Macias A, Monedero P, Adame M, Torre W, Fidalgo I, Hidalgo F. Arandomized blinded comparison of thoracic epidural ropivacaine,ropivacaine/fentanyl or bupivacaine/fentanyl for postthoracotomyanalgesia. Anesth Analg 2002;95:1344-50.

[7] Takamori S, Yoshida S, Hayashi A, Matsuo T, Mitsuoka M, ShirouzuK. Intraoperative intercostal nerve blockade for postthoracotomy pain.Ann Thorac Surg 2002;74:338-41.

[8] James EC, Kolberg HL, Iwen GW, Gellatly TA. Epidural analgesia forpost-thoracotomy patients. J Thorac Cardiovasc Surg 1981;82:898-903.

[9] Logas WG, el-Baz N, el-Ganzouri A, et al. Continuous thoracicepidural analgesia for postoperative pain relief following thoracotomy:a randomized prospective study. Anesthesiology 1987;67:787-91.

[10] Liu SS, Moore JM, Luo AM, Trautman WJ, Carpenter RL.Comparison of three solutions of ropivacaine/fentanyl for post-operative patient-controlled epidural analgesia. Anesthesiology 1999;90:727-33.

[11] Mann LJ, Young GR, Williams JK, Dent OF, McCaughan BC.Intrapleural bupivacaine in the control of postthoracotomy pain. AnnThorac Surg 1992;53:449-54.

[12] Kambam JR, Hammon J, Parris WC, Lupinetti FM. Intrapleuralanalgesia for postthoracotomy pain and blood levels of bupivacainefollowing intrapleural injection. Can J Anaesth 1989;36:106-9.

[13] Savage C, McQuitty C, Wang D, Zwischenberger JB. Postthoracotomypain management. Chest Surg Clin North Am 2002;12:251-63.

[14] Soto RG, Fu ES. Acute pain management for patients undergoingthoracotomy. Ann Thorac Surg 2003;75:1349-57.

[15] Salomaki TE, Laitinen JO, Nuutinen LS. A randomized double-blindcomparison of epidural versus intravenous fentanyl infusion foranalgesia after thoracotomy. Anesthesiology 1991;75:790-5.

[16] Ballantyne JC, Carr DB, deFerranti S, et al. The comparative effects ofpostoperative analgesic therapies on pulmonary outcome: cumulativemeta-analyses of randomized, clinical trials. Anesth Analg 1998;86:598-612.

[17] Scheinin B, Lindgren L, Rosenberg PH. Treatment of post-thoracotomy pain with intermittent instillations of intrapleuralbupivacaine. Acta Anaesthesiol Scand 1989;33:156-9.

[18] Silomon M, Claus T, Huwer H, Biedler A, Larsen R, Molter G.Interpleural analgesia does not influence postthoracotomy pain.Anesth Analg 2000;91:44-50.

[19] Kreitzer JM, Reuben SS. Central nervous system toxicity in a patientreceiving continuous intrapleural bupivacaine. J Clin Anesth 1996;8:666-8.

[20] Eng J, Sabanathan S. Site of action of continuous extrapleuralintercostal nerve block. Ann Thorac Surg 1991;51:387-9.

[21] D'Amours RH, Riegler FX, Little AG. Pathogenesis and managementof persistent postthoracotomy pain. Chest Surg Clin North Am 1998;8:703-22.

[22] Mahon SV, Berry PD, Jackson M, Russell GN, Pennefather SH.Thoracic epidural infusions for post-thoracotomy pain: a comparisonof fentanyl-bupivacaine mixtures vs. fentanyl alone. Anaesthesia 1999;54:641-6.

[23] Fischer RL, Lubenow TR, Liceaga A, McCarthy RJ, Ivankovich AD.Comparison of continuous epidural infusion of fentanyl-bupivacaineand morphine-bupivacaine in management of postoperative pain.Anesth Analg 1988;67:559-63.