J Anesth DOI 10.1007/s00540-016-2286-y

ORIGINAL ARTICLE

Intrathecal morphine versus intravenous opioid administration to impact postoperative analgesia in hepato‑pancreatic surgery: a randomized controlled trial Sara Dichtwald1 · Menahem Ben‑Haim2 · Laila Papismedov1 · Shoshana Hazan1 · Anat Cattan1 · Idit Matot1 

Received: 24 August 2016 / Accepted: 13 November 2016 © Japanese Society of Anesthesiologists 2016

Abstract  Purpose Inadequate analgesia following abdominal surgery may affect outcome. Data in patients undergoing liver surgery suggested that postoperative coagulopathy might delay epidural catheter removal. Thus, alternative analgesic techniques should be evaluated. Methods  We compared the analgesic efficacy of intraoperative intrathecal morphine [single injection 4 µg/kg before skin incision (ITM group, n = 23)] to intravenous opioids [iv remifentanil infusion during surgery followed by i.v. bolus of morphine, 0.15 mg/kg before the end of surgery (IVO group, n  = 26)]. Forty-nine adult patients undergoing elective open resection of liver or pancreas lesions in the Tel Aviv Medical Center were randomized into the two analgesic protocols. Postoperatively both groups received i.v. morphine via a patient-controlled analgesia pump. Follow-up was till the 3rd postoperative day (POD). Results  There was no significant difference in demographics and intraoperative data between groups. The primary outcome, pain scores on movement, was significantly worse in the IVO group when compared with the ITM group at various time points till POD3. In the secondary outcomes — need for rescue drugs — the IVO group required significantly more rescue morphine boluses. Complication related to the analgesia and recovery parameters were similar between groups.

* Idit Matot [email protected] 1

Division of Anesthesiology, Intensive Care and Pain, Tel Aviv Medical Center and the Sackler Faculty of Medicine, Tel Aviv University, 6 Weitzman Street, Tel Aviv 64239, Israel

2

Division of Surgery, Shaare Zedek Medical Center, Hebrew University Faculty of Medicine, Jerusalem, Israel





Conclusions The findings suggest that a single dose of ITM before hepatic/pancreatic surgery may offer better postoperative pain control than i.v. opioid administration during surgery. This beneficial effect is maintained throughout the first three PODs and is not associated with a higher complication rate; neither did it influence recovery parameters. ITM provides an appropriate alternative to i.v. morphine during major abdominal surgery. Keywords  Analgesia · Opioids · Liver resection · Spinal morphine

Introduction Inadequate analgesia following upper abdominal surgery may increase postoperative morbidity, delay recovery and increase hospital length of stay. Epidural analgesia is a well-described option in major abdominal surgery [1–4]. However, postoperative coagulopathy following major hepatic resections, even in patients with normal preoperative coagulation profiles, may raise concerns regarding the safe removal of epidural catheters postoperatively, limiting the use of epidural analgesia in these procedures [5–7]. As more hepatic and pancreatic surgeries are being performed nowadays, it is of utmost importance to evaluate and compare other analgesic techniques for this patient population. Two analgesic techniques, i.e., intrathecal morphine (ITM) or continuous remifentanil infusion followed by morphine towards the end of surgery are widely used for major abdominal surgeries. In major hepatectomy, a single injection of ITM prior to surgical incision, followed by postoperative use of i.v. patient-controlled analgesia (PCA) with morphine was demonstrated to be a satisfying analgesic technique, equivalent to continuous epidural analgesia

13



[8, 9]. Another study of patients scheduled for hepato-pancreato-biliary surgery demonstrated lower rates of postoperative hypotension, decreased fluid demand, shorter length of hospital stay and fewer respiratory complications in the group receiving a single injection of ITM when compared with continuous thoracic epidural analgesia [10]. Continuous infusion of the short-acting opioid remifentanil in these procedures was found to be associated with a decreased time interval for extubation, excellent intraoperative hemodynamic stability and rapid recovery [11]. To the best of our knowledge, the benefit of one technique (ITM) over the other (solely i.v. opioids) has not yet been evaluated. The aim of the current study was thus to compare, in a randomized clinical trial, pain relief, recovery parameters and incidence of side effects with the two different pain therapies as described above, in patients undergoing elective open hepatic or pancreatic resections.

Methods This prospective, interventional, randomized clinical study was approved by the Tel-Aviv Medical Centre Institutional Review Board (IRB no. 0073-10-TLV) and registered in ClinicalTrials.gov (ClinicalTrials.gov Identifier: NCT01201499) before inclusion of any patients. Participants Adult (>18 years old) ASA I-III patients, scheduled for elective major hepatic or pancreatic surgeries with midline or bi-subcostal incision, in the Tel Aviv Medical Center were included in the study. Exclusion criteria were patient’s refusal, contraindications for intrathecal analgesia (e.g., preoperative coagulopathy, localized infection), drug allergy (for morphine or remifentanil), pregnancy, morbid obesity with obstructive sleep apnea, and chronic treatment with opioids. Dropout criteria included patients who required postoperative mechanical ventilation for more than 2 h, and patients scheduled for laparoscopic surgery. Intra‑ and postoperative management During surgery, the anesthesia and surgical teams operated according to routine. In addition to the analgesic regimens, as will be described below, the only other anesthetic drugs used were propofol for induction, isoflurane (0.7–1.1 MAC) for maintenance and rocuronium for paralysis. Tracheal extubation was performed at the end of the surgery when weaning criteria were satisfied. Postoperatively, all patients were transferred to a high monitoring area for a minimum of 24 h; this is the standard

13

J Anesth

of care for major hepatic or pancreatic surgeries in our medical facility. During the first three PODs, all patients received the same postoperative physiotherapy daily. All patients were encouraged to sit in an armchair with the help of the nurses, unless medically contraindicated (i.e., if the patient was hemodynamically unstable). Patients were allowed to breathe room air if they were able to maintain oxygen saturation ≥ 92% without additional oxygen (via face mask or nasal cannula). They were allowed to drink water during the first POD, unless they had significant nausea and vomiting. Enteral nutrition was usually begun at the second POD day, via nasogastric tube or orally, according to the instructions of the attending physician. The nasogastric tube was removed when gastric suction provided less than 500 ml/day, and according to instructions of the attending physician. The urinary catheter was removed when urinary output exceeded 0.5 ml/kg/h for more than 6 h, the patient was hemodynamically stable, and urea and creatinine levels were within the patient’s baseline levels. From the third POD until discharge, the patients continued treatment and follow-up according to department protocols. Study groups Patients who consented to participate in the study were randomized by the research assistant away from the study procedure (48 h before surgery) into two treatment groups according to a computer-generated randomization list. Participants were allocated into one of these two treatment groups. Intrathecal morphine group (ITM) Patients received a single dose of intrathecal morphine, 4 µg/kg, at the level of L3–4/L4–5 [10, 12–14], in the operating room. Intraoperatively, hemodynamic changes suggestive of pain were addressed by continuous i.v. remifentanil infusion (5 mg in 50 ml normal saline, starting with 0.1 mcg/kg/min). Remifentanil was infused only when an increase of 20% or more in blood pressure or heart rate from the baseline values was observed, and adjustments were performed in the dosing accordingly. Intravenous opioid group (IVO) Following induction of anesthesia, patients received a continuous infusion of i.v. remifentanil with a starting dose of 0.1–0.2 µg/kg/min. The dosage was adjusted when an increase of 20% or more in blood pressure or heart rate from the baseline values was observed. Twenty to thirty

J Anesth

minutes before the conclusion of surgery, a single bolus of i.v. morphine, 0.15 mg/kg, was given. After randomization and before surgery, patients were informed about their group of allocation. Patients were allowed to leave the study at any time for any reason and without any consequences. Postoperative management Both groups received the same analgesic protocol postoperatively which included i.v. morphine via a PCA device (IVAC®) according to the following protocol: a bolus of 1.5 mg morphine for each patient demand, lockout time of 7 min and a maximal dose of 25 mg morphine within 4 h. There was no continuous basal infusion of morphine in this protocol [15, 16]. The patients were allowed to ask the nurse for additional boluses of morphine up to 0.05 mg/kg, twice per shift (8 h) if they had pain and had exceeded what was allowed at that time point to be delivered from the PCA (so they would not get more from the pump once they pushed the button). According to the protocol in our medical center, nurses would not administer opioids to a patient that requested pain relief in the case of respiratory depression (respiratory rate less than 10 per minute and/or documented hypercapnia, PaCO2 higher than 60 mmHg). If pain persisted despite administration of morphine, the patients could ask for an additional rescue non-opioid analgesic, which was 1 g of i.v. acetaminophen or dipyrone. Follow‑up We followed all patients for 72 h postoperatively. Demographic data (age, gender, weight and ASA score), medical history (pre-existing diseases, chronic medications), etiology of surgery, and surgical procedure were collected from the electronic medical records. In addition, intraoperative data were collected: duration of anesthesia, type and amount of fluids given during surgery, urine output, estimated blood loss, amount of blood products required, need for vasopressor or inotropic support, need for diuretic drugs, major hemodynamic events, and time to extubation. Patients were evaluated twice in the 3 PODs by a research nurse or physician blinded to the treatment group: one visit during the morning (8:00– 10:00 a.m.) and a second during the afternoon (2:00–4:00 p.m.). Data collected included: NRS (numerical rating scale) score [17] [NRS ranging from “0” (no pain) to “10” (worst imaginable pain)] at rest and while coughing, the total amount of morphine delivered via the PCA device in the past 24 h and the amount of rescue drugs (opioids followed by non-opioids) administered by the ward nurse when asked by the patient in the past 24 h. Additionally,

adverse drug reaction or potential complication related to the analgesic technique were assessed; nausea and vomiting, pruritus, respiratory depression (defined as respiratory rate less than 10 per minute, hypercapnea and respiratory acidosis with pH lower than 7.25 and PaCO2 higher than 60 mmHg), post-dural puncture headache, spinal hematoma or abscess. Parameters indicative of postoperative recovery were also retrieved: time to sit (from conclusion of surgery), time to eat, drink, remove urinary catheter, length of intensive care unit stay and overall length of hospitalization. Study endpoints The primary endpoint of our study was the quality of analgesic technique, characterized by pain level as measured by the 0–10 numerical rating scale (NRS, in which 0 represents “no pain” and 10 “worst imaginable pain”) twice daily (morning round, afternoon round), during rest and while coughing. Secondary endpoints included the total amount of morphine delivered via the PCA device, need for rescue opioid or non-opioid analgesic drugs, time to recovery parameters and the incidence of complications and adverse effects related to the analgesic technique as defined above. Postoperative documentation was performed by an investigator blinded to the patients’ study group allocation. Statistical analysis Statistical analysis was performed with SPSS version 22.0. We analyzed the NRS during rest and while coughing separately for each study group. Comparisons were made between days within and between the two groups. Each variable was tested for normality. The t-test for independent samples was used when continuous variables were normally distributed and the sample size was large enough. When the variables were not normally distributed or the sample size was small we used the non parametric Wilcoxon’s test. We used an axis table for the ordinal variables. If the number of observations in each cell was at least 5, the comparison between proportions was calculated using the chi-square test and the Fisher’s exact test for cases in which one of the cells contained fewer than 5 observations. We compared the NRS between the groups by using the general linear regression model, taking repeated measures into account. The dependent variables in each model were either the NRS at rest or the NRS while coughing, and the independent variables were Time 1 (morning visit, afternoon visit) and Time 2 (POD1, POD2, and POD3) and the interaction between them (i.e., for the analysis between the AM and PM visits, between the three POD visits, for each POD between the AM and PM visits, and for the AM and

13

Table 1  Patient demographics, comorbidities, medications and surgical procedure

J Anesth Parameter

ITM (N = 23)

IVO (N = 26)

Age (years) Gender (female/male) Weight (kg) ASA score Comorbidities  Hypertension  Dyslipidemia  IHD  CHF  TIA/CVA  Smoking  COPD  DM  Endocrine disorder Medical therapy  Anticoagulants  β blockers  ACE inhibitor  ARB  Calcium channel blocker  Oral antidiabetics  Insulin  Steroids  Chemotherapy  Radiotherapy Surgical procedure:  Whipple  Distal/partial pancreatectomy  Partial hepatectomy

54 ± 17 11 (48%)/12 (52%) 72 ± 14 2 ± 0.6

60 ± 15 12 (46%)/14 (54%) 71 ± 16 2 ± 0.7

7 (30%) 10 (43%) 2 (8%) 0 0 5 (22%) 1 (4.3%) 6 (26%) 4 (17%)

12 (46%) 13 (50%) 3 (11%) 0 2 (7%) 6 (23%) 2 (7.6%) 7 (27%) 3 (11%)

5 (22%) 2 (9%) 6 (26%) 2 (9%) 1 (4%) 4 (17%) 2 (9%) 1 (4%) 2 (9%) 1 (4%)

7 (27%) 5 (19%) 7 (27%) 0 5 (19%) 4 (15%) 2 (7%) 2 (7%) 5 (19%) 3 (11%)

10 (43%) 2 (9%) 11 (48%)

10 (38%) 1 (4%) 14 (54%)

0

1 (4%)

 Partial hepatectomy and distal pancreatectomy

Data are presented as mean ± SD or as number (%) of patients. P value was non-significant for all these data ITM intrathecal morphine, IVO intravenous opioid, ASA American Society of Anesthesia, IHD ischemic heart disease, CHF congestive heart failure, CVA cerebrovascular accident, TIA transient ischemic attack, COPD chronic obstructive pulmonary disease, DM diabetes mellitus, ACE angiotensin converting enzyme, ARB angiotensin receptor blocker

PM visits for each POD). For the analysis between all 6 postoperative visits, the independent variables were Time 3 (POD1-AM, POD1-PM, POD2-AM, POD2-PM, POD3AM, and POD3-PM). From previous data from our institution, we determined a minimal sample size of 18 patients per group to detect a difference of 2 NRS units between the 2 treatment groups with an estimated standard deviation of 2 by setting the β value at 80% and the α value at 0.05 or less. To compensate for potential exclusions or withdrawals, we decided to include an additional 35 patients. P values of 0.05 were considered statistically significant.

13

Results Seventy-one patients were approached by the study team. Sixty-six patients fulfilled all inclusion criteria, signed informed consents and were randomized: 33 patients into the ITM group, and 33 patients into the IVO group. Seventeen patients were excluded due to laparoscopic surgery and prolonged postoperative mechanical ventilation (>2 h) because of massive blood loss (CONSORT diagram). Demographic data were similar between the groups (Table  1). There were no significant differences between

J Anesth Table 2  Anaesthesia Parameter

ITM (N = 23)

IVO (N = 26)

Length of anesthesia (min) Fluids and non-anesthetic drugs  Crystalloids (ml)  Colloids (ml)  Total fluids (ml)  Urine output (ml)  Red packed cells (units)  Fresh frozen plasma (units)  Platelets (units)  Vasopressors  Inotropes  Diuretics

371 ± 127

365 ± 110

3795 ± 2033 326 ± 387 4226 ± 2348 648 ± 421 0.3 ± 0.7 0.2 ± 0.6 0 9 (39%) 4 (17%) 0

4070 ± 1843 404 ± 394 4666 ± 2114 597 ± 587 1.0 ± 2.1 0.6 ± 2.4 0.2 ± 1.2 14 (56%) 7 (28%) 1 (4%)

22 (96%)

23 (88%)

 Extubation in OR

Data are presented as mean ± SD or as number (%) of patients. P value was non-significant for all these data ITM intrathecal morphine, IVO intravenous opioid, OR operating room

the groups in past medical history, chronic drug therapy, or surgical procedure (Table 1). Ninety percent of patients underwent either the Whipple operation or partial hepatectomy. Duration of anesthesia was about 6 h, and at least half of all the patients required some pharmacological hemodynamic support during surgery (Table 2). Ninety percent of patients were extubated in the operating room. Intraoperative data (duration of anesthesia, fluid and blood product requirements, intraoperative blood loss, urinary output, need for inotropic or vasopressor drug support, and number of patients extubated at the conclusion of surgery) were comparable in both groups (Table 2).

Primary endpoint: postoperative quality of analgesic technique (NRS data) Pain scores (NRS) in the entire population (Table 3). There was no significant difference in pain scores between the morning and the afternoon visits on the same day and between days when the results of both groups were gathered. There were significant differences between pain scores at rest and while coughing at all time points. Pain scores (NRS) in each study group (Fig.  1). Pain scores in the ITM group were significantly lower than those measured in the IVO group at most time points of followup: both at rest and while coughing in the morning and the afternoon visits on POD1, in the afternoon visit on POD2 and while coughing, both in the morning and in the afternoon visits of POD3. Secondary endpoints: analgesic consumption, recovery parameters and side effects Significantly more IVO patients requested additional rescue morphine from the nursing staff compared with the ITM patients: 12 (46%) versus 4 (17%), respectively, P = 0.03. The amount of additional morphine given by the nurse was also significantly higher: 14.7 ± 5.3 mg vs 4.8 ± 3.7, in the IVO compared with the ITM group, respectively. No significant difference was found in the demand and delivery of morphine via the PCA device for POD1, POD2, and POD3 (Fig. 2). The number of patients requesting non-opioid analgesic drugs was also similar: 17 (74%) in the ITM group [mean (SD) cumulative 3 days dose, acetaminophen 3.7 mg (2.1) and dipyrone 2.8 mg (1.4)] and 18 (69%) in the IVO group [acetaminophen 4.1 mg (2.2) and dipyrone 3.2 mg (1.6)] (P = 0.71).

Table 3  Pain level (NRS) at rest and cough at different time points — all patients A. AM vs PM postoperative visit pain level (all patients, all days)

 Rest  Cough

AM visit

PM visit

3.0 ± 3.0 5.8 ± 3.0*

3.1 ± 3.1 5.7 ± 3.1*

B. POD1 vs POD2 vs POD3 pain level (all patients, AM and PM visits) POD1

POD2

POD3

 Rest

3.2 ± 3.3

3.0 ± 2.9

3.0 ± 2.9

 Cough

5.6 ± 3.7*

5.8 ± 2.7*

5.8 ± 2.7*

Data are presented as mean ± SD P value was non-significant for NRS AM visit vs PM visit NRS numerical rating scale, POD postoperative day * P value was <  0.05 for comparisons between NRS at rest vs while coughing

13



J Anesth

Fig. 2  Morphine demand (a) and delivery (b) via the PCA device in the two study groups

Fig. 1  Pain scores according to the numerical rating scale (NRS) at rest (a) and during cough (b) in the two study groups. Pain scores in the ITM group were significantly lower than those measured in the IVO group at most time points as indicated in the figure. *P value was <  0.05 for comparisons between pain scores in ITM group vs. IVO group

The majority of patients sat in an armchair on POD1 and walked with help within 72 h. Drinking started on POD1 in about half of the patients, and some kind of feeding was started within 72 h. Recovery parameters were similar between the two study groups, as were the number of admissions to the intensive care unit (Table 4). There was also no significant group difference in the length of hospital stay (Table 4).The most common side effects were nausea [6 (26%) patients and 8 (31%) patients in the ITM and IVO groups, respectively], pruritus [5 (22%) patients and 5 (19%) patients in the ITM and IVO groups, respectively] and drowsiness [5 (22%) patients and 5 (19%) patients in the ITM and IVO groups, respectively]. None of the ITM patients had post-dural puncture headache. One ITM patient and 2 IVO patients had one episode of hypoventilation, with an average time of 6 h from the end of surgery to the episode.

13

Discussion Pain following open operations on the pancreas and the liver is almost inevitable. The results of the present study that was conducted on this patient population suggest that intraoperative intrathecal morphine may provide superior analgesia compared with an i.v. opioid regimen, when both techniques are followed by an i.v. PCA with morphine. Moreover, even though the analgesic properties of intrathecal morphine lasted up to only 24 h, its advantage was maintained throughout all 3 PODs. This may be attributed to the effect of preemptive analgesia that had been achieved by the administration of intrathecal morphine prior to the first surgical incision [18]. Epidural analgesia is a widespread and effective technique for patients undergoing major abdominal surgeries. It may, however, be problematic in patients following extensive hepatic resections due to postoperative coagulopathy [5, 6]. A common problem is related to the timing of the epidural catheter removal [7], since significant thrombocytopenia and/or an abnormal coagulation profile may last as long as up to 1 week following surgery. The administration of coagulation factors, such as in the form of fresh frozen plasma, has been recommended in order to facilitate safe

J Anesth Table 4  Recovery parameters ITM (N = 23) IVO (N = 26) Time to sit (h)  0–24  24–48  48–72  After 72 Time to walk (h)  0–24  24–48  48–72  After 72 Time to drink (h)  0–24  24–48  48–72  After 72 Time to removal of urinary catheter (h)  0–24  24–48  48–72  After 72 Time to eat (h)  0–24  24–48  48–72  After 72 LOS-ICU (h)  0  0–24  24–48  48–72  After 72 Overall LOS (days)

20 (87%) 2 (9%) 1 (4%) 0

23 (88%) 0 3 (12%) 0

2 (9%) 10 (43%) 10 (43%)

5 (19%) 10 (38%) 9 (34%)

1 (5%)

2 (9%)

12 (52%) 8 (34%) 3 (14%) 0

14 (53%) 8 (31%) 3 (12%) 1 (4%)

5 (22%) 7 (30%) 8 (35%) 3 (13%)

4 (15%) 11 (42%) 7 (27%) 4 (16%)

1 (5%) 10 (43%) 8 (35%) 4 (17%)

2 (8%) 6 (23%) 13 (50%) 5 (19%)

12 (52%) 7 (30%) 2 (10%) 1 (4%) 1 (4%)

18 (70%) 5 (18%) 2 (8%) 1 (4%) 0

11.2 ± 9.5

10.6 ± 8.0

Data are presented as number (%) of patients or as mean ± SD. P value was non-significant for all these data ITM intrathecal morphine, IVO intravenous opioid, LOS length of stay, ICU intensive care unit

removal of the epidural catheter [7]. Therefore, the singledose administration of intrathecal morphine before surgery may well be an appropriate substitute for epidural analgesia in major hepatic resections. Intrathecal morphine for the provision of perioperative pain relief has been widely investigated in obstetric procedures. In a randomized double blind trial on 60 women undergoing cesarean section, Swart et al. [14] found lower postoperative pain scores that were documented beyond the first POD in the group that received intrathecal morphine in addition to spinal local anesthetics when compared with

intrathecal local anesthetics alone. Both groups had i.v. PCA devices for postoperative morphine supplement. The intrathecal morphine group had a higher rate of pruritus, but only during the first 4 postoperative hours. No differences in other adverse effects or complication rates were observed between the groups [14]. The efficacy for postoperative pain relief and safety of intrathecal morphine was reported by Gwirtz et al. [13] who reviewed the files of almost 6000 patients undergoing major abdominal surgery and who gave high satisfaction scores (a mean of 8.5 out of 10 possible points) with this analgesic technique. The most frequent side effect was pruritus, occurring in about a third of the study patients, and 3% of all patients had non-lifethreatening respiratory depression. These findings are in agreement with the results of the present study. Post-dural puncture headache was reported in 0.54% of their cases and in none of the patients in our study. No other significant complications were reported, thus confirming the safety of this analgesic technique in major abdominal surgeries. DePietri et al. [8] compared intrathecal morphine followed by i.v. PCA with continuous epidural analgesia for partial hepatectomy and found no difference in postoperative pain scores between the groups during the first 2 PODs. There were no reports of spinal or epidural hematoma, spinal cord compression or post-dural puncture headache among their 2 groups, nor were there any differences in the rate of adverse effects or complications related to the analgesic technique. The authors’ conclusion was that intrathecal morphine is an appropriate analgesic technique in major hepatic resections, and one that is comparable to epidural analgesia in terms of perioperative pain control. These finding were further supported by the work of Sakowska et al. [10] which included 160 patients who underwent hepatic or pancreatic resections. Their retrospectively retrieved data showed that the use of intrathecal morphine was associated with a reduced incidence of postoperative hypotension, reduced i.v. fluids requirements, shorter hospital stay, and lower incidence of respiratory complications compared with thoracic epidural anesthesia. There was no significant difference in the postoperative pain scores. Only one study compared the use of intrathecal morphine to other non-epidural analgesic techniques in the setting of abdominal surgery. In their study on 60 patients undergoing oncologic abdominal surgeries, Devys et al. [12] reported that intrathecal morphine followed by i.v. PCA morphine provided better analgesia on POD1 compared with a group receiving i.v. morphine via PCA during the postoperative period. Significantly higher percentages of postoperative nausea and vomiting were observed among the intrathecal morphine patients during POD1. In their study, however, unlike the present one, both lower abdominal and upper abdominal surgeries were included. Moreover, there was no documentation of the analgesic

13



technique provided during the operation [12]. The results of the present study extend these previously published data by showing that intraoperative intrathecal morphine may be superior to i.v. opioids in liver and pancreatic surgeries by providing improved analgesia beyond the first 24 postoperative hours and without significant side effects. Hansen et al. [15] reported that continuous infusion of remifentanil intraoperatively during major abdominal surgeries (as in the current study in the IVO group) might cause opioid tolerance in the postoperative period. Those authors found that the group of patients that received continuous infusion of remifentanil intraoperatively had higher rates of hyperalgesia and a trend towards higher demand of opioids postoperatively. Continuous infusion of remifentanil may offer hemodynamic stability in major liver resections, as observed by Milne et al. [11]. They reported that the use of continuous infusion of remifentanil and propofol in hepato-pancreato-biliary surgery was associated with a decreased interval to extubation, excellent intraoperative hemodynamic stability and rapid recovery. However, their report included only nine patients and focused on the intraoperative period. The present study, which extended throughout the first three PODs, shows similar requirements for vasopressor support intraoperatively with intrathecal morphine compared with continuous administration of remifentanil, and no differences between the approaches in the rate of extubation or postoperative recovery parameters. There are several limitations to the present study. The main limitation is the lack of blinding. Postoperative evaluation, though, was performed by an investigator blinded to patient group assignment. Also, the number of study patients was relatively small. Thirdly, we assessed only one dose of morphine in the IVO study group. This is the pain protocol we use routinely in our practice for major abdominal surgeries. Fourth, intraoperative analgesia by remifentanil was titrated in both groups according to predefined changes in hemodynamics which served as surrogate markers. While this is not an accurate method, a better validated tool to adequately monitor and treat pain is not currently in hand. Finally, we have no exact documentation regarding the total amount of i.v. remifentanil received by each patient in the study. We therefore could not compare the total amounts of remifentanil that had been given to the 2 groups, a factor that might influence the patients’ postoperative pain scores, given remifentanil’s reported hyperalgesic properties [19]. In conclusion, intrathecal morphine in the setting of hepatic and pancreatic resections is associated with better postoperative pain control compared with i.v. opioids during most of the early postoperative period. Intrathecal morphine was not associated with a higher rate of adverse effects or complications in the intra- or postoperative period. As suggested in previous studies [5–7], intrathecal

13

J Anesth

morphine may be a suitable alternative to epidural analgesia, especially for patients scheduled for extensive hepatic resections, in which postoperative coagulopathy may complicate the removal of the epidural catheter. It should be noted, however, that the use of intrathecal morphine mandates postoperative observation and monitoring in a highcare facility since postoperative respiratory depression may occur within the first POD.

References 1. Ballantyne JC, Kupelnick B, McPeek B, Lau J. Does the evidence support the use of spinal and epidural anesthesia for surgery? J Clin Anesth. 2005;17(5):382–91. 2. Cousins MJ, Mather LE. Intrathecal and epidural administration of opioids. Anesthesiology. 1984;61(3):276–310. 3. Kehlet H, Holte K. Effect of postoperative analgesia on surgical outcome. Br J Anaesth. 2001;87(1):62–72. 4. Yeager MP, Glass DD, Neff RK, Brinck-Johnsen T. Epidural anesthesia and analgesia in high-risk surgical patients. Anesthesiology. 1987;66(6):729–36. 5. Matot I, Scheinin O, Eid A, Jurim O. Epidural anesthesia and analgesia in liver resection. Anesth Analg. 2002;95(5):1179–81. 6. Pelton JJ, Hoffman JP, Eisenberg BL. Comparison of liver function tests after hepatic lobectomy and hepatic wedge resection. Am Surg. 1998;64(5):408–14. 7. Vandermeulen EP, Van Aken H, Vermylen J. Anticoagulants and spinal-epidural anesthesia. Anesth Analg. 1994;79(6):1165–77. 8. De Pietri L, Siniscalchi A, Reggiani A, Masetti M, Begliomini B, Gazzi M, Gerunda GE, Pasetto A. The use of intrathecal morphine for postoperative pain relief after liver resection: a comparison with epidural analgesia. Anesth Analg. 2006;102(4):1157–63. 9. Redai I, Emond J, Brentjens T. Anesthetic considerations during liver surgery. Surg Clin N Am. 2004;84(2):401–11. 10. Sakowska M, Docherty E, Linscott D, Connor S. A change in practice from epidural to intrathecal morphine analgesia for hepatopancreato-biliary surgery. World J Surg. 2009;33(9):1802–8. 11. Milne SE, Horgan PG, Kenny GN. Target-controlled infusions of propofol and remifentanil with closed-loop anaesthesia for hepatic resection. Anaesthesia. 2002;57(1):93. 12. Devys JM, Mora A, Plaud B, Jayr C, Laplanche A, Raynard B, Lasser P, Debaene B. Intrathecal + PCA morphine improves analgesia during the first 24 hr after major abdominal surgery compared to PCA alone. Can J Anaesth. 2003;50(4):355–61. 13. Gwirtz KH, Young JV, Byers RS, Alley C, Levin K, Walker SG, Stoelting RK. The safety and efficacy of intrathecal opioid analgesia for acute postoperative pain: seven years’ experience with 5969 surgical patients at Indiana University Hospital. Anesth Analg. 1999;88(3):599–604. 14. Swart M, Sewell J, Thomas D. Intrathecal morphine for caesarean section: an assessment of pain relief, satisfaction and sideeffects. Anaesthesia. 1997;52(4):373–7. 15. Hansen EG, Duedahl TH, Romsing J, Hilsted KL, Dahl JB. Intra-operative remifentanil might influence pain levels in the immediate post-operative period after major abdominal surgery. Acta Anaesthesiol Scand. 2005;49(10):1464–70. 16. Schüttler J, Albrecht S, Breivik H, Osnes S, Prys-Roberts C, Holder K, Chauvin M, Viby-Mogensen J, Mogensen T, Gustafson I, Lof L, Noronha D, Kirkham AJ. A comparison of remifentanil and alfentanil in patients undergoing major abdominal surgery. Anaesthesia. 1997;52(4):307–17.

J Anesth 17. Cook KF, Dunn W, Griffith JW, Morrison MT, Tanquary J, Sabata D, Victorson D, Carey LM, Macdermid JC, Dudgeon BJ, Gershon RC. Pain assessment using the NIH toolbox. Neurology. 2013;80(11 Suppl 3):S49–53. 18. Kissin I. Preemptive analgesia. Anesthesiology. 2000;93(4):1138–43.

19. Angst MS, Koppert W, Pahl I, Clark DJ, Schmelz M. Short-term infusion of the mu-opioid agonist remifentanil in humans causes hyperalgesia during withdrawal. Pain. 2003;106(1–2):49–57.

13