Share this article on:

A Comparison of Intrathecal Morphine/Fentanyl and Patient-Controlled Analgesia with Patient-Controlled Analgesia Alone for Analgesia After Liver Resection

Roy, Jean-Denis MD, FRCPC*; Massicotte, Luc MD*; Sassine, Marie-Pascale C PhD†; Seal, Robert F. MD, FRCPC‡; Roy, André MD, FRCPC§

Section Editor(s): Liu, Spencer S.

doi: 10.1213/01.ane.0000238040.41872.7e
Analgesia: Research Report

Continuous epidural anesthesia and analgesia may be considered in liver resection, but is often avoided because of the potential development of coagulopathies and the risk of epidural hematoma. In this prospective, randomized, double-blind study we compared postoperative morphine consumption via patient-controlled analgesia after liver surgery between two groups of patients: patients receiving a preoperative dose of intrathecal morphine (0.5 mg) and fentanyl (15 μg) (treatment group) and patients receiving a sham intrathecal injection (placebo group). Forty patients scheduled for major liver resection (≥two segments) were enrolled. The primary outcome measure was patient-controlled analgesia morphine consumption. Secondary outcomes were evaluation of pain at rest and with movement, scored on a visual analog scale with assessment of sedation, nausea, pruritus, and respiratory frequency. Outcome measures were recorded at 6, 12, 18, 24, and 48 h postspinal anesthesia or simulation. Patients in the placebo group consumed approximately three times more morphine during each time interval than patients in the treatment group (at 48 h: 124 ± 30 vs 47 ± 21 mg, P < 0.0001). Pain evaluation on the visual analog scale was lower for the first 18 h in the treatment group. There was no difference in the incidence of side effects in both groups. Intrathecal morphine (0.5 mg) and fentanyl (15 μg) given before liver surgery significantly decreased postoperative morphine consumption compared to placebo without any increase in side effects.

IMPLICATIONS: This prospective, randomized, double-blind study demonstrated that intrathecal morphine (0.5 mg) and fentanyl (15 μg) given before liver surgery significantly decreased postoperative morphine consumption, without any increase in side effects, compared to placebo.

From the Departments of *Anesthesiology and †Epidemiology and Biostatistics, Centre hospitalier de l’Université de Montréal (CHUM), Hôpital St-Luc, Montreal, Quebec, Canada; ‡Department of Anesthesiology and Pain Medicine, University of Alberta, Edmonton, Alberta, Canada; and §Department of Surgery, Hepatobiliary Service, CHUM-Hôpital St-Luc, Montreal, Quebec, Canada.

Accepted for publication June 12, 2006.

Address correspondence and reprint requests to Dr. Luc Massicotte, CHUM-Hôpital St-Luc, 1058 St-Denis, Montreal, QC, Canada H2X 3J4. Address e-mail to

Postoperative pain is a major concern for patients undergoing surgery. Continuous epidural analgesia can be useful after liver resection (1–5). However, insertion of an indwelling epidural catheter and the timing of its removal are problematic, because the liver resection may result in postoperative coagulopathy (1–3,6–9). Because of the risk of epidural hematoma, continuous epidural analgesia is often avoided. Consequently, patient-controlled analgesia (PCA) is often used to provide postoperative analgesia in patients undergoing liver resection. A single intrathecal injection of morphine 0.5–0.7 mg has been suggested as an alternative (2) to PCA or epidural analgesia. No studies have compared single intrathecal injection of morphine and fentanyl to a standard PCA technique. The aim of the present investigation was to compare the analgesic efficacy of intrathecal morphine (0.5 mg) and fentanyl (15 μg) to sham intrathecal injection, assessing postoperative PCA morphine administration and the incidence of the side effects. We hypothesized that a single dose of intrathecal morphine and fentanyl would significantly decrease PCA morphine consumption during the first 48 h after surgery.

Back to Top | Article Outline


After approval by the Ethics Committee of the Centre Hospitalier de l’Université de Montréal and signed informed consent given by each patient, we performed a prospective, randomized, double-blind study. Forty patients between 18 and 75 yr of age scheduled for major hepatic resection (two segments or more) were enrolled. The exclusion criteria were contraindication to spinal anesthesia or intrathecal morphine or fentanyl, insufficient comprehension to use PCA, and analgesic drug dependence.

No premedication was given. In the operating room (OR), all patients were positioned on the left lateral side. Midazolam 0.05 mg/kg IV was administered for sedation. The same anesthesiologist (L.M.) performed lumbar puncture at L2–3 or L3–4 with a Sprotte no. 25 needle. Patients were randomized to receive morphine 0.5 mg and 15 μg of fentanyl (treatment group) or a sham intrathecal anesthesia (placebo group). For the placebo group, the skin was punctured with the spinal needle, but it was not advanced beyond the subcutaneous tissue. After the procedure, the anesthesiologist (L.M.) left the OR, and was replaced by the anesthesiologist in charge of the case. Monitoring and anesthesia were standardized. A peripheral IV catheter and a right internal jugular catheter were placed in all patients. Electrocardiogram, oxygen saturation, capnometry, invasive and noninvasive arterial blood pressure, and central venous pressure were monitored continuously. All patients received sufentanil 0.4 μg/kg as an induction bolus followed by an infusion of 0.25 μg/kg/h. Induction was completed with propofol 1–2 mg/kg and rocuronium 0.6 mg/kg. After tracheal intubation, anesthesia was maintained with a mixture of air (0.5 L/min) and oxygen (0.5 L/min) plus 1 MAC of desflurane. Additional boluses of sufentanil 0.15 μg/kg were given at 5-min intervals if mean arterial blood pressure exceeded 20% of the baseline. A right subcostal incision was made in all cases.

Each anesthesiologist decreased baseline central venous pressure by about 40% by phlebotomy without intravascular volume replacement. The quantity of blood withdrawn varied according to patient mass (400–800 mL) (10). Collected blood was reinfused at the end of surgery.

On arrival to the postanesthesia care unit (PACU), each patient received a PCA pump programmed to deliver an initial morphine bolus of 0.05 mg/kg at 7-min intervals if pain was more than 60 on the visual analog scale (VAS) (0–100) (11). On discharge from the PACU the pump was reprogrammed for a morphine bolus of 1.5 mg and a 7-min lockout.

The primary outcome measure was PCA morphine consumption. Secondary outcomes were evaluation of pain at rest and with movement on the VAS, with assessment of sedation, nausea, pruritus, and respiratory frequency on a standardized scale (Appendix) (12–14). These variables were also analyzed on a VAS of 0–100. Outcome measures were recorded by the same research nurse at 6, 12, 18, 24, and 48 h postspinal anesthesia or simulation.

Rescue medications were naloxone 0.08 mg IV every 5 min for respiratory frequency <8 min−1 and sedation score ≥3 or Paco2 ≥70 mm Hg; metoclopramide 10 mg IV every 8 h for nausea, and if ineffective, dolasetron 12.5 mg IV each day, to a maximum of 100 mg if ineffective; and diphenhydramine 25 mg IV every 8 h for pruritus.

None of the patients, the anesthesiologists in charge of the cases, the surgeons, the research nurse, the nurses in the OR, in the PACU, in the intensive care unit, or on the floor were aware of patients’ assignments.

In a pilot study, 25 patients with hepatic resection for cancer lesions had morphine consumption of 121 ± 45 mg in the first 48 h through PCA. To achieve a 50% reduction of morphine consumption with an α error of 0.05 and power of 80%, we needed nine patients in each group. The power would increase to 90% with 12 patients in each group. Ultimately, 20 patients were selected for each group to minimize the impact of potential independent variables on morphine consumption. Patients were randomly assigned to the morphine or placebo group.

Data are expressed as mean ± sd of the mean for continuous values or median with interquartile range for discontinuous values. Distributions were examined to ensure proper statistical treatment. Parametric and nonparametric statistical tests were applied as appropriate. P values <0.05 were considered significant. SPSS 10 statistical programs were run for analysis.

Morphine consumption and respiratory frequency were analyzed as continuous variables. The VAS had to be analyzed as a nonparametric variable because of its nonnormal distribution and the significant frequency of zero as a result. Sedation, nausea and pruritus were analyzed as discontinuous variables. We also tried to analyze these variables on a 0–100 scale but the abnormal distribution required calculation as categories. The results remained the same as when using standardized scales (Appendix).

Back to Top | Article Outline


The study was stopped midway after 20 patients were enrolled because of a large difference between the amount of morphine consumed and the side effects in the placebo group versus the treatment group (124 ± 30 vs 47 ± 21 mg, P < 0.0001). In the placebo group, four patients had to discontinue PCA because of excessive side effects after 18, 21, 24, and 32 h (drowsiness, pruritus, aggressive state, and confusion). Morphine consumption at these times was 95.5, 98, 168, and 85 mg, respectively. The data on these patients were gathered until they left the study. All patients received hydromorphone subcutaneously after they left the study. Table 1 lists demographic data and preoperative characteristics of both groups. Table 2 lists the anesthetic and surgical characteristics of both groups. Surgical duration was longer and sufentanil consumption was larger in the treatment group. Figure 1 illustrates the cumulative postoperative morphine received and demonstrates that hourly morphine use was significantly less in the treatment group.

As measured by the VAS, patients in the treatment group had less pain at rest and with movement for up to 18 h (Figs. 2 and 3). This pain diminution appeared to persist but did not achieve statistical significance. There were no significant differences in sedation, nausea, or pruritus between the groups for each study period (Table 3).

There was also no difference in respiratory rate (Table 4). Paco2 was not monitored in every patient, since <20% of the patients were admitted to the intensive care unit. No patient required naloxone. Table 5 shows morphine consumption for each patient in both groups according to time. No postdural puncture headache or neurological complications were reported.

Back to Top | Article Outline


We evaluated the efficacy of intrathecal opioids, rather than epidural opioids, for reducing the risk of epidural hematoma—a risk which could increase after liver resection. Liver resection results in measurable decreases in the concentration of various clotting factors due to transient insufficiency of the remaining liver to synthesize new factors, factor consumption, and intraoperative fibrinolysis (6). The incidence of epidural hematoma using an epidural technique in patients receiving low-molecular-weight heparin is estimated to be 1/1000 to 1/10,000 in an orthopedic population (15). After hepatic resection, biological perturbation of coagulation is probably more significant than that due to heparin; therefore, the risk of epidural hematoma could be increased. A literature review (1,4–8) identified only 279 patients who had an epidural for analgesia after liver resection. This is an inadequate number to support the safety of epidural injection, which is why we elected to evaluate intrathecal analgesia.

We added fentanyl to intrathecal morphine because fentanyl has a faster onset than morphine. Fournier et al. (16) found that the first demand for analgesic after 40 μg of intrathecal fentanyl for total hip replacement was 214 ± 120 min. Celeski et al. (17) found that there was no advantage to giving more than 25 μg. The recommended doses for intrathecal fentanyl are now in the range of 12.5 μg (18) to 20 μg (19). After analyzing theses studies, we decided to use 15 μg.

We found that the placebo group needed three times more morphine in the postoperative period than the treatment group. Four patients had to discontinue PCA because of excessive side effects. Two of these side effects (confusion and aggressive state) were not in our secondary outcomes (sedation, nausea, pruritus, respiratory depression). Considering that four patients in the placebo group discontinued their PCA before the end of the study period, the increase in postoperative morphine use in the placebo group, although very significant, likely underestimated the true difference. Intrathecal morphine has a reported duration of action of about 24 h (20). However, the treatment group continued to consume less morphine between the 24 and 48 h, suggesting a preemptive analgesic effect.

The amount of sufentanil given to the placebo group was appropriate for this surgery (89 ± 31 μg, 1.1 μg/kg) and the postoperative morphine consumed (124 ± 30 mg) was the same as in patients from our pilot study (121 ± 45 mg). Neither the duration of action of sufentanil (21) nor the quantity given to both groups could explain why the placebo group required three times more morphine during the first 48 h after surgery. Pain at rest and with movement remained lower in the treatment group for 18 h. This difference was clinically obvious, particularly in the PACU. After 18 h, the VAS score continued to be lower in the treatment group at rest and with movement, although it did not achieve statistical significance. Perhaps this would have been different had we studied more patients; also, as more time passed after surgery, the more pain decreased, and the VAS result between groups became similar.

At 6 h, more sedation (3 vs 2) was evident in the treatment group but this difference was not significant. There was also no significant difference for the other side effects between groups. Possible reasons for these results are that (a) there was an inadequate number of study subjects; (b) the number was not chosen to detect a significant difference in side effects but rather in morphine consumption; or (c) the power of standardized scales was insufficient discriminative power (for example, the itchiness score: 0–2; Appendix).

Both groups were comparable in all aspects, except that surgery in the treatment group was longer. This difference did not translate into a difference in pain scores.

In conclusion, patients who received 0.5 mg of intrathecal morphine with 15 μg of fentanyl before hepatic resection required significantly less morphine during the first 48 h after the surgery compared to patients in the placebo group. Intrathecal morphine also produced lower pain scores in the first 18 h without an increase in side effects.

Back to Top | Article Outline


The authors thank Mrs. Denise Bois for her clerical work, Mr. Robert Boileau for data processing, Mrs. Josée Gagnon, RN for data collection, and Dr. Danielle Beaulieu and Mr. Ovid Da Silva for their comments during the preparation of this manuscript.

Back to Top | Article Outline


Sedation Score (Ramsay):

1 – Wide awake

2 – Drowsy

3 – Dozing

4 – Mostly sleeping

5 – Needs stimulating to rouse

Back to Top | Article Outline

Nausea Score:

0 – None

1 – Mild

2 – Discomforting

3 – Distressing

4 – Horrible

5 – Worst possible

Back to Top | Article Outline

Itchiness Score:

0 – None

1 – Mild

2 – Discomforting

Cited Here...

Back to Top | Article Outline


1. Matot I, Scheinin O, Eid A, Jurim O. Epidural anesthesia and analgesia in liver resection. Anesth Analg 2002;95:1179–81.
2. Redai I, Emond J, Brentjens T. Anesthetic considerations during liver surgery. Surg Clin N Am 2004;84:401–11.
3. Lentschener C, Ozier Y. Anesthesia for elective liver resection: some points should be revisited. Eur J Anaesthesiol 2002;19:780–8.
4. Cywinski JB, Parker BM, Xu M, Irefin SA. A comparison of postoperative pain control in patients after right lobe donor hepatectomy and major hepatic resection for tumor. Anesth Analg 2004;99:1747–52.
5. Ayanoglu HO, Ulukaya S, Yuzer Y, Tokat Y. Anesthetic management and complications in living donor hepatectomy. Transplant Proc 2003;35:2970–3.
6. Borromeo CJ, Stix MS, Lally A, Pomfret EA. Epidural catheter and increased prothrombin time after right lobe hepatectomy for living donor transplantation. Anesth Analg 2000;91:1139–41.
7. Schuman R, Zabala L, Angelis M, et al. Altered hematologic profiles following donor right hepatectomy and implications for perioperative analgesic management. Liver Transpl 2004;10:363–8.
8. Siniscalchi A, Begliomini B, DePietri L, et al. Increased prothrombin time and platelet count in living donor right hepatectomy: implications for epidural anesthesia. Liver Transpl 2004;10:1144–9.
9. Tsui SL, Yong BH, Ng KF, et al. Delayed epidural catheter removal: the impact of postoperative coagulopathy. Anaesth Intensive Care 2004;32: 545–6, 630–6.
10. Massicotte L, Lenis S, Thibeault L, et al. Effect of low central venous pressure and phlebotomy on blood product transfusion requirements during liver transplantation. Liver Transpl 2006;12:117–23.
11. Choiniere M, Amsel R. A visual analogue thermometer for measuring pain intensity. J Pain Symptom Manage 1996;11:299–311.
12. Parlow JL, Costache I, Avery N, Turner K. Single-dose haloperidol for prophylaxis of postoperative nausea and vomiting after intrathecal morphine. Anesth Analg 2004;98:1072–6.
13. Ramsay MA, Savage TM, Simpson BR, Goodwin R. Controlled sedation with alphaxolone-alphadolone. Br J Med 1974;2:656–9.
14. Melzak R. Measurement of nausea. J Pain Symptom Manage 1989;4:157–60.
15. Horloker TT, Heit JA. Low molecular weight heparin: biochemistry, pharmacology, perioperative prophylaxis regimens, and guidelines for regional anesthetic management. Anesth Analg 1997;85:874–85.
16. Fournier R, Van Gessel E, Weber A, et al. A comparison of intrathecal analgesia with fentanyl or sufentanil after total hip replacement. Anesth Analg 2000;90:918–22.
17. Celeski DC, Heindel L, Haas J, et al. Effect of intrathecal fentanyl dose on the duration of labor analgesia. AANA J 1999;67:239–44.
18. Bogra J, Arora N, Srivastava P. Synergistic effect of intrathecal fentanyl and bupivacaine in spinal anesthesia for cesarean section. BMC Anesthesiol 2005;17:5.
19. Jain K, Grover VK, Mahajan R, et al. Effect of varying doses of fentanyl with low dose spinal bupivacaine for caesarean delivery in patients with pregnancy-incuced hypertension. Int J Obstet 2004;132:215–20.
20. Swart M, Sewell J, Thomas D. Intrathecal morphine for caesarean section: an assessment of pain relief, satisfaction and side-effects. Anesthesiology 1997;52:373–7.
21. Bailey PL, Egan TD, Stanley TH. Intravenous opioid anesthetics. In: Miller RD, ed. Anesthesia, 6th ed. New York: Churchill Livingstone, 2005:379–437.
© 2006 International Anesthesia Research Society