Continuous Preperitoneal Infusion of Ropivacaine Provides Effective Analgesia and Accelerates Recovery after Colorectal Surgery: A Randomized, Double-blind, Placebo-controlled Study

This prospective, randomized, double-blind, placebo-controlled study was approved by the Committee for the Protection of Human Subjects in Biomedical Research. All of the patients signed a written informed consent form. The study was conducted from July 2005 to May 2006. Patients were recruited at St. Antoine Hospital, Paris France (n = 34); Geneva University Hospital, Geneva, Switzerland (n = 13); and Strasbourg University Hospital, Strasbourg, France (n = 2).

Patients included in the study had an American Society of Anesthesiologists physical status I or II, were aged between 18 and 80 yr, and were scheduled to undergo elective open resection of malignant colorectal tumors through a periumbilical midline incision followed by a primary anastomosis. Exclusion criteria were obesity (body mass index > 30 kg/m²), inflammatory bowel diseases, preoperative cognitive dysfunction, chronic pain, preoperative opioid consumption, psychiatric disorders, inability to use the patient-controlled analgesic device, and a priori  indication for dysfunctioning stoma or abdominal suction drains.

Patients were premedicated with oral hydroxyzine (1 mg/kg) given 1 h before the induction of anesthesia. After arrival in the operating room, patients were monitored as usual, and anesthetic induction was performed with intravenous thiopental (3–4 mg/kg), sufentanil (0.2–0.3 μg/kg), and atracurium (0.5 mg/kg). After tracheal intubation, mechanical ventilation was initiated with a mixture of 50% O²and 50% N²O and adjusted to keep end-tidal carbon dioxide tension between 30 and 35 mmHg. Anesthesia was maintained with desflurane or sevoflurane, continuous infusion of atracurium (0.4–0.5 mg · kg⁻¹· h⁻¹) and sufentanil (0.1–0.2 μg · kg⁻¹· h⁻¹). At the end of the procedure, halogenated agents were switched off, and 100% O²was given with 8 l/min fresh gas flow. Residual neuromuscular blockade was reversed, if needed, with a mixture of atropin and neostigmine. A warming forced-air blanket (Bair-Hugger; Arizant Health Care Inc., Eden Prairie, MN) covering the upper part of the body was used routinely to prevent intraoperative hypothermia.

After arrival in the operating room, patients were randomly allocated to receive a continuous wound infusion of either 0.2% ropivacaine (ropivacaine group) or 0.9% saline (control group). The attending anesthesiologist sent the inclusion number to the pharmacist. The inclusion number referred to a sealed envelope, which was opened by the pharmacist and which contained the patient’s allocation group (determined by a computer-generated random list). Randomization was established by blocks of four patients.

The pharmacist prepared a 10-ml syringe for bolus infusion and, at the same time, filled the elastomeric pump (On-Q Pain Buster®, ref. PS12507I; I-Flow Corp., Lake Forest, CA), under aseptic conditions, with 480 ml solution. Both the 10-ml syringe and the elastomeric pump were provided to the treating physician. Only the pharmacist was aware of the code defining the type of solution to be administered. Physicians in charge of the patient, during both intraoperative and postoperative periods, were fully blinded to the patient’s group assignment.

At the end of the surgery, after closure of the parietal peritoneal membrane with running sutures, the surgeon inserted a 20-gauge multiholed Soaker catheter (On-Q Pain Buster®, ref. PS12507; I-Flow Corp.) approximately 3 cm from the lower end of the midline incision through an introducer needle. The catheter was positioned between the previously closed parietal peritoneum and the underside of the transversalis fascia, along the full length of the wound (fig. 1). Thereafter, the surgeon closed the fascia layer and skin and secured the infusion catheter to the skin. When the wound was closed, a 10-ml bolus of test solution was administered through the catheter. The prefilled elastomeric pump, set to deliver a 10-ml/h constant rate during 48 h (infusion pressure = 10 psi corresponding to 517 mmHg), was connected immediately thereafter. The catheter was covered with a transparent dressing.

Except for the medication delivered through the preperitoneal wound catheter, postoperative management was strictly identical for all patients. Tracheal extubation was performed when the patient was conscious, with a respiratory rate between 12 and 30 breaths/min, with a central core temperature greater than 36°C, and without residual muscle weakness. After tracheal extubation, pain was assessed, and those patients with pain greater than 2 on a 4-point verbal rating scale received intravenous boluses of 2 mg morphine as titration, with 5-min intervals, until pain decreased to a maximum verbal rating scale of 1 (0 = no pain, 1 = mild pain, 2 = moderate pain, and 3 = severe pain). A patient-controlled analgesia (PCA) device (Graseby 9300; Watford Herts, United Kingdom) was then connected to an intravenous infusion and set to deliver a 1-mg dose of morphine with a 7-min lockout time. Before surgery, each patient had received information on the PCA device and was able to use it efficiently. PCA was maintained until daily morphine consumption was less than 10 mg.

Nonopioid intravenous analgesics (ketoprofen 50 mg × 3 daily, or acetaminophen 1 g × 4 daily in case of contraindication to non steroidal antiinflammatory drugs) were given as rescue medication if pain was not controlled adequately, as defined by a verbal numerical scale above 4 out of 10 at rest, or at the patient’s request for better pain relief despite the morphine PCA.

Nasogastric tube and urinary bladder catheter were left in place for at least 24 h after surgery. Oral fluids were started as soon as the patients passed flatus. Solid meals were given the day after.

Evaluation started at the end of the wound closure (hour 0 [H0]). At that time, the bolus of test solution was administered.

The primary endpoint was parenteral morphine consumption, which was measured daily on the PCA device.

Secondary outcomes measures were as follows:

• The number of patients requiring morphine titration in the postanesthesia care unit (PACU) and the dose of morphine administered were recorded.

• Pain was measured at rest and at mobilization (defined as pain experienced during coughing) using the verbal numerical scale from 0 (no pain) to 10 (worst pain imaginable) at H2, H6, H12, and thereafter once daily until discharge.

• The modified Aldrete score was used to estimate the time course of initial arousal from anesthesia in the recovery room.14After reaching an Aldrete score greater than 8, patients were considered fit for discharge from the PACU to the ward.

• Time to return of gastrointestinal function was defined as the time from the end of surgery (H0) until the first bowel movement. Time until the first occurrence of flatus was noted.

• Mental function was assessed by the Digit Symbol Substitution Test performed daily until patients attained the preoperative score (determined the day before surgery).15 

• Quality of the night’s sleep was evaluated each morning with a 10-cm visual analog scale from 0 (very poor quality of sleep) to 10 (excellent quality of sleep).

• Duration of hospital stay could be considered as a synthetic index of the postoperative recovery. Patients were considered ready for discharge when they fulfilled all of the criteria from an objective scale (see  appendix). The criteria were checked for each patient twice daily by a surgeon who was blinded to the patient’s group assignment. The duration of hospital stay was therefore assessed to the nearest half-day time interval. Authorization for discharge was rapidly followed by actual hospital discharge.

• All side effects were recorded. The incidence of postoperative nausea or vomiting, requiring specific treatment with intravenous ondansetron (4 mg), was noted. The level of sedation was monitored at H2, H6, H12, H24, and then twice daily with a 4-point rating scale (where 0 = fully alert, 1 = sleepy but easily aroused with verbal stimulation, 2 = sleepy but barely arousable, and 3 = unconscious patient not answering to contact). Special attention was paid to detect any problem with the infusing material.

• To study the diffusion of the local anesthetic in the preperitoneal position, one patient underwent a computer tomography contrast study 24 h after the surgery. A 10-ml mixture of nonionic contrast material and saline solution (1:1) was injected, and transverse and coronal sections were obtained.

• Plasma concentrations of ropivacaine (total and unbound) were measured by chromatography at H24, H48, and H60 in eight patients who were allocated to receive ropivacaine. Those patients were selected during the constitution of the randomization list. The information that blood samples would be needed was notified in the randomization envelopes. To respect the blindness of the study, blood samples were also taken from eight patients allocated to the placebo group. Unbound fraction of ropivacaine was measured in only six patients because the volume of the blood samples after the dosage of the total fraction was not sufficient in two patients. After pH adjustment by equilibration during 2 h in an agitated water bath with 95% N²and 5% CO²at 37°C, protein binding was determined using ultrafiltration at 35°C using YMT membranes (Millipore, Saint-Quentin-en-Yvelines, France). In all patients, an electrocardiogram was performed on the first and second postoperative days to document possible ropivacaine cardiac toxicity.

• Patients were contacted by phone between 8 and 12 weeks after the surgery and asked about any residual wound pain and analgesia requirements.

The calculation of the sample size was based on the primary endpoint, i.e. , morphine consumption during the first postoperative day. Taking into account retrospective data from our institution, showing a morphine PCA consumption of 50 ± 15 mg in a similar population, a sample size of 21 patients in each group was required to detect as significant a between-group difference of 30%, with an α risk of 0.05 and a β risk of 0.1.

Pain intensity between the two groups was compared with repeated-measures (two-way) analysis of variance, the independent within-subject variable being the time of evaluation and the intersubject variability being the verbal numerical scale values. A two-level between-groups factor was used (ropivacaine, control). In case of statistical differences between the two groups, post hoc  pairwise comparisons were performed with the Fisher protected least significant difference test (Statview®; Abacus Concepts, Inc., Berkeley, CA).

Because the morphine consumption was not normally distributed after the 48th postoperative hour (high value of the asymmetric test), the between-group comparisons were performed with nonparametric Mann–Whitney test. Other continuous quantitative variables were analyzed with a two-tailed Student t  test. Categorical data were analyzed using Mann–Whitney test or chi-square contingency table. Logistic regression was performed to test for a possible interaction between the centers of evaluation and the main studied parameter.

Variables are presented as mean ± SD. The threshold for statistical significance was set at P < 0.05. Non–statistically significant differences are abbreviated as NS.

Forty-nine patients were enrolled in the study. Three patients were excluded from analysis because of an intraoperative decision to use a dysfunctioning stoma. In one patient (allocated to the saline group), the catheter was withdrawn at H12 because of severe hyperthermia. This episode resolved spontaneously, without any sign of local wound infection. Microbiologic culture of the tip of the catheter was sterile. A further 3 patients were excluded because of parietal tumor extension (1 patient), lack of peritoneum (1 patient who had undergone previous major intraabdominal surgery), and intraoperative urologic complication (1 patient). Twenty-one patients successfully completed the study in each groups. Logistic regression did not show any significant interaction between centers of evaluation and the main parameter, i.e. , morphine consumption (P = 0.31). There were no missing values, except for the pain assessment at 12 h postoperatively because it was overnight and some patients were sleeping (12 patients in both groups).

As demonstrated by computer tomography, once infused in the preperitoneal position, the radiopaque contrast media remained in the deep layer of the abdominal wall, in close vicinity to the peritoneal injury, without any signs of intraabdominal penetration (fig. 2). The local anesthetic spread toward the upper part of the abdominal wall due to the pressure applied when injecting the radiopaque solution. Indeed, pressure to infuse the computer tomography scan contrast was approximately 10 times higher than the infusion pressure of the elastomeric pump.

Demographic and intraoperative data are presented in table 1. Tracheal extubation was performed 20 ± 9 and 23 ± 10 min after H0, respectively, in the ropivacaine and control groups (NS). In the control group, 20 patients, as compared with 15 patients in the ropivacaine group, needed intravenous morphine titration in the PACU (NS). The total doses of intravenous morphine given as titration in the PACU were 4 ± 3 and 7 ± 5 mg in the ropivacaine and control groups, respectively (P = 0.004). Time to reach an Aldrete score greater than 8 was not different between groups (63 ± 29 and 70 ± 48 min in the ropivacaine and control groups, respectively).

After discharge from the PACU, morphine consumption was significantly reduced in the ropivacaine group as compared with the control group during the first 3 postoperative days (fig. 3). Total morphine consumption over the first 3 postoperative days was 48 ± 23 mg in the ropivacaine group and 84 ± 37 mg in the control group (P = 0.0004).

Pain intensity was significantly reduced in the ropivacaine group compared with control group, both at rest (significant group–time interaction effect on analysis of variance; P < 0.01) and during coughing (significant group–time interaction effect on analysis of variance; P < 0.01). Pairwise comparisons showed that the difference was significant throughout the first 12 h for pain at rest and throughout the first 48 h for pain during coughing (figs. 4A and B). During the first postoperative day, 6 patients in the ropivacaine group and 11 patients in the control group needed rescue analgesic medications (NS). The same applied to 4 and 7 patients, respectively, during the second postoperative day (NS).

No major adverse event occurred. Two patients in the ropivacaine group and six patients in the control group experienced severe postoperative nausea or vomiting requiring treatment (NS). The course of postoperative recovery is presented in table 2. Quality of sleep was rated as better in the ropivacaine group than in the control group during the two first postoperative nights. Time to recover preoperative mental status, assessed by Digit Symbol Substitution Test, was similar between the two groups. Recovery of intestinal transit, assessed by the time to first bowel movement, was faster in the ropivacaine than in the control group (P = 0.02), as well as the time to be eligible for discharge from the hospital (P = 0.02).

Total plasma concentrations of ropivacaine were 2.3 ± 0.9, 1.6 ± 0.9, and 0.4 ± 0.3 μg/ml, respectively, at H24, H48, and H60. Plasma unbound fractions of ropivacaine are given in figure 5. Unbound fractions were below the quantification threshold (0.01 μg/ml) at H60 for all patients except one (0.03 μg/ml).

At the follow-up evaluation, between 8 and 12 weeks after the surgery, one patient in each group reported residual wound pain requiring analgesic medications.

The increasing use of wound infiltration of local anesthetics as part of multimodal analgesia after major surgery is based on the recognition of the important role played by parietal nociceptive afferents in the overall pain and in the pathophysiologic repercussions induced by surgery. In this study, we provide for the first time evidence that continuous infusion of ropivacaine into the preperitoneal space for 48 h has a beneficial effect on pain relief, reduces the parenteral morphine consumption, and accelerates the recovery after open colorectal resection, as compared with parenteral analgesia alone.

Open colorectal surgery induces severe and prolonged postoperative pain, especially during mobilization.16Systemic or even epidural opiates are not effective enough to fully control pain induced by mobilization.17Only epidural local anesthetics have shown a marked benefit in controlling pain at mobilization.18When compared with systemic patient-controlled morphine analgesia, epidural analgesia using local anesthetic is significantly better for pain control during mobilization for the first 2 postoperative days after open colonic resection.17However, several medical conditions preclude the use of epidural analgesia, and approximately 20–30% of eligible patients do not benefit from it because of technical problems or failure in efficiency.19,20These limitations have stimulated the search for alternative ways of pain management in the setting of abdominal laparotomy.

In the current study, we show that continuous preperitoneal infiltration of ropivacaine exhibits a significant benefit over systemic analgesia alone, both on pain at rest and at mobilization. Pain at mobilization was significantly better alleviated throughout the first 48 postoperative hours, corresponding to the duration of the local anesthetic infusion. It is noteworthy that pain intensity did not increase after finishing the local anesthetic infusion. Moreover, daily morphine consumption was still significantly less in the ropivacaine group than in the control group during the 24 h after catheter removal. This is in accordance with some recent data suggesting that the blockade of parietal afferents may reduce spinal dorsal horn neuron sensitization, thereby providing postoperative analgesic effect that may outlast the duration of the wound infusion.21In the current study, there was only one patient in each group with residual long-term wound pain. However, this information was collected by phone interview using simple questions and without clinical examination. This result might have been different if a more detailed questionnaire about the nature and the intensity of the pain had been proposed and if a clinical examination had been performed. Furthermore, it cannot be excluded that other nonpharmacologic parameters, related to the infusion, may have played a role in reducing dorsal horn neuron sensitization and the incidence of long-term postoperative pain in both groups.

The current results emphasize that the peritoneum and the deep muscular layer play a crucial role in the pain induced by abdominal incisions. This assumption is further supported by the failure of epidural analgesia when metameric level is not high enough to block peritoneal nociceptive influx, even after lower abdominal surgery.22Furthermore, recent data from animal studies have shown that parietal pain may sensitize neurons in the spinal cord to visceral colonic pain.23Therefore, it cannot be excluded that blockade of parietal pain influx may even contribute to a reduction of the visceral component of pain.

Injection of radiopaque contrast media through a preperitoneal wound catheter shows that once injected, the local anesthetic remains in close vicinity of the abdominal wound incision, between the injured parietal peritoneum and the muscular layer, thereby effectively blocking peritoneal afferents. Information drawn from computer tomography must be taken with caution because there was only one case and the diffusion of the radiopaque media may not reflect the exact diffusion of the local anesthetic.

When limited to the subcutaneous layers, local anesthetic wound infusion has been disappointing after laparotomy. Cheong et al.  11infused 0.5% bupivacaine at a flow rate of 2 ml/h for 60 h into the subcutaneous layer of a left iliac fossa incision. Only pain at rest, not during movement, was better controlled in patients allocated to receive local infusion during the first postoperative day, and total morphine consumption was reduced only moderately. Fredman et al.  10did not show any benefit for pain relief or for morphine consumption when infusing 0.25% bupivacaine into the subcutaneous space by PCA device set to deliver 9 ml with a 60-min lockout interval, during 24 h after abdominal laparotomy. More recently, Baig et al.  12presented results obtained by continuous subcutaneous infusion of 0.5% bupivacaine at 4 ml/h for 72 h and showed a significant reduction in daily morphine consumption but no difference in overall postoperative pain. These results underline the limited influence of the superficial layer of the abdominal wall on overall postoperative pain after laparotomy.

Ropivacaine and bupivacaine have been the most used local anesthetics for continuous wound infiltration.9In this indication, and at similar doses, analgesic efficacy between these two agents seems comparable.24In the current study, we chose ropivacaine instead of bupivacaine because of its lower systemic toxicity and its shorter elimination half-life, reducing the risk of plasma accumulation during prolonged infusion.25The dose/volume infused was chosen according to the study by Burm et al.  26showing that a constant rate (10 ml/h) epidural infusion of 0.2% ropivacaine for 72 h after major orthopedic surgery was well tolerated and associated with plasma concentrations below the level of toxicity. In the current study, total and unbound fractions of ropivacaine were comparable to those obtained in the study by Burm et al.  26No sign or symptom indicative of systemic toxicity was noted. Both unbound and total fraction of ropivacaine decreased between the 24th and 48th postoperative hours, showing the absence of drug accumulation. The highest unbound ropivacaine concentration after 24 h of infusion was 0.12 μg/ml, which is slightly above the threshold concentration for mild central nervous system toxicity derived after rapid intravenous infusion of ropivacaine in healthy subjects.27This suggests a sufficient margin of safety with the use of the studied infusion regimen, but cautions against using higher ropivacaine doses in this setting.

None of the studies on subcutaneous local anesthetic wound infusion reported any positive influence on postoperative recovery. In contrast, we showed that preperitoneal wound infusion of ropivacaine improved sleep quality during the first two postoperative nights, reduced the duration of paralytic ileus, and shortened the duration of hospital stay. Time to ileus resolution after abdominal surgery is one of the most important factors contributing to the duration of hospital stay. Mechanisms whereby preperitoneal continuous administration of a local anesthetic reduce the duration of ileus may include improvement of analgesia, with concomitant reduction in sympathetic activation, and morphine sparing. However, because some experimental studies suggest that the afferent limb of the reflex leading to a postoperative ileus originates primarily from the peritoneum, it cannot be excluded that a direct effect of local anesthetics on the peritoneal membrane may be directly implicated.28Furthermore, as hypothesized with epidural analgesia, systemic effects of local anesthetics, which are known to have antiinflammatory property, may also speed up the return of bowel function,29although this remains under debate.30In addition to the effect on ileus resolution, it has been recently shown in an animal study that wound infiltration with local anesthetics may partly restore food intake behavior, which is disturbed after an abdominal wall incision.31This point may have significant implications for postoperative recovery but must be evaluated in the human setting.

Sleep quality is important for patients’ comfort and postoperative fatigue. The better sleep quality in patients who received a preperitoneal ropivacaine infusion may be associated with more vigor and contribute to a faster rate of recovery in this subgroup. It may be due to better pain relief but also to a reduction in morphine consumption, because opiates are known to disrupt sleep quality.32Finally, we chose to evaluate the duration of hospital stay because it represents a synthetic index of recovery. To ensure reliable assessment, an objective scale was used. Continuous preperitoneal infusion of ropivacaine reduced the duration of hospital stay on average by more than 24 h, and it may therefore be considered to be among the analgesic techniques that have proven a benefit for postoperative rehabilitation after abdominal surgery. Potential economical benefits should be evaluated in further studies.

Preperitoneal continuous infusion seems to be well tolerated and devoid of unwanted side effects. In agreement with other reports on continuous local anesthetic wound perfusion, no local complications were observed. In the metaanalysis by Liu et al. ,9the overall wound infection rates were similar between catheter with local anesthetic (0.7%) and catheter with placebo or no-catheter control group (1.2%). The incidence of reported catheter or pump failure was 1.1%.9In the current study, no technical problems occurred with the infusion devices. However, larger sample sizes must be evaluated before a definite conclusion can be drawn about the safety of this technique.

The current study reports encouraging results with the use of a continuous infusion of 0.2% ropivacaine at 10 ml/h during 48 h, but further questions will have to be answered in the future, such as the choice of local anesthetic, the optimal dose/volume per time, and the influence of the mode of administration (e.g. , patient-controlled administration) on efficiency. Nonetheless, several limitations of this analgesic technique should be mentioned, such as patients with dysfunctioning stoma, not accessible to local wound perfusion, and those with previous abdominal surgery including peritoneal resection.

Several secondary evaluation parameters, such as the incidence of postoperative nausea or vomiting, or the number of patients who have required rescue analgesic medications, showed a trend in favor of the ropivacaine-group treatment without reaching the threshold of statistical significance. This is probably in relation with the small sample size which had been calculated based on the primary outcome.

In conclusion, preperitoneal continuous infiltration of 0.2% ropivacaine at 10 ml/h during 48 h seems to be an effective method to relieve pain after open colorectal surgery. It reduced morphine consumption and accelerated the postoperative recovery. It is easy to implement and seems devoid of major side effects, making specific supervision unnecessary. It could therefore be considered as an interesting alternative to epidural analgesia in this setting.

The authors thank Malika Benis, M.D. (Staff, Department of Digestive Surgery, St. Antoine Hospital, Paris, France), and Arthur Atchabahian, M.D. (Associate Professor, Department of Anesthesiology, Columbia University, New York, New York), for assistance in writing the article, and Régine Le Guen (Technician, Faculté de Médecine, Le Kremlin Bicêtre, Unité Propre de Recherche de l’Enseignement Supérieur, Equipe d’Accueil 3540, Villejuif, France) for her expertise in ropivacaine dosages.

• Apyrexia defined as central core temperature between 36.7° and 37.8°C

• Leukocyte count less than 12 × 10⁹/l

• Absence of anemia with clinical repercussion (no dyspnea at rest, no orthostatic hypotension)

• Resumption of normal bowel function (bowel movement without diarrhea)

• Lack of nausea and/or vomiting

• Lack of significant pain (verbal rating scale < 2 at movement)

• Ability to wake up and ambulate without help