Postoperative pain after resection of hepatocellular carcinoma (HCC) is almost always severe because of the intense noxious stimuli generated by the subcostal incision. Epidural analgesia (EA) and IV opioid analgesia (IA) are the 2 most common forms of postoperative analgesia for patients with HCC. However, it is unclear whether anesthetic technique or the approach to postoperative analgesia affects cancer recurrence and patient survival.
Animal studies have showed that regional anesthesia can reduce the metastatic burden in animals inoculated with breast adenocarcinoma cells.1,2 However, some observational studies in humans showed different results. For example, 2 studies reported beneficial effects with regional anesthesia in breast cancer and prostate cancer,3,4 whereas others showed that regional anesthesia had no significant effect on survival or recurrence in colon cancer, prostate cancer, and cervical cancer.5–7 The contrasting findings from these studies suggested the potential benefits of regional anesthesia on cancer recurrence may depend on the specific tumor type and its mechanism and risk for metastasis.5
Tumor biology varies considerably from organ to organ, and the effects of differing approaches to postoperative analgesia are likely to vary from site to site. HCC is a typical angiogenesis-dependent cancer, whose generation, progression, and invasion are closely related to angiogenesis.8 Due to its particular tumor biology, we performed this retrospective trial to determine whether differing approaches to postoperative analgesia after resection of HCC affect cancer recurrence and/or survival.
We performed a retrospective cohort study involving patients with HCC (n = 1846) receiving hepatic resection at the Cancer Center of Sun Yat-Sen University from January 1997 to December 2007. Follow-up ended in December 2009. The study was reviewed and approved by the IRB. Only records from patients with HCC where hepatic resection was indicated were included in the analysis. Patients were excluded from the analysis if they were lost to follow-up, if postoperative EA or IA was not used, if the number of tumors ≥2, if thrombosis in the portal vein or hepatic vein was noted, or if tumor metastases were present. Of the original 1846 patients, 819 patients met the inclusion criteria and were divided into 2 groups: patients receiving postoperative EA (n = 451) and patients receiving postoperative IA (n = 368) (Fig. 1).
We collected follow-up data from the patients’ medical records, hospital surgery or pathology database (or both), and a letter of introduction followed by telephone contact with the patients or their families. Independent investigators prospectively collected data on each patient. We obtained the demographic, oncologic, and operative characteristics of all patients including age, gender, ASA physical status, preoperative α-fetoprotein, Child-Pugh class, liver cirrhosis, tumor diameter, pathological stage, transfusion, duration of anesthesia, and the dose of intraoperative fentanyl.
The major outcomes include recurrence-free survival and long-term survival. Recurrence was defined as radiologic evidence of local recurrence or distant metastatic disease. Recurrence-free survival was calculated from operation to recurrence. Recurrence-free survival for those patients without a recorded recurrence was defined as the time between the date of operation and the date of last follow-up or the date of death. Long-term survival was calculated from operation to death of any cause.
All 819 patients had the same balanced general anesthesia, including induction with propofol (1.5–2.0 mg/kg), fentanyl (2 μg/kg), and vecuronium (2 mg/kg), or cisatracurium (0.2 mg/kg). Anesthesia was maintained with isoflurane or sevoflurane supplemented with fentanyl.
For postoperative EA, all epidural catheters were inserted in the low thoracic region (T9-T10 or T10-T11) before the induction of anesthesia, and a standard 100 mL solution containing 0.15% ropivacaine combined with 0.07 mg/kg per day morphine was administrated at the same rate of 2 mL/h for 48 hours after surgery. The first 6 mL bolus dose containing 0.33% ropivacaine combined with 2 mg morphine was given through the epidural catheter at the time of fascial closure. The epidural was not used intraoperatively.
For postoperative IA, all patients were administrated standard 100 mL solution consisting of 0.012 mg/kg per day fentanyl IV at the same rate of 2 mL/h for 48 hours after surgery. Patients also received tramadol 150 mg IV at the time of fascial closure.
Tramadol 100 mg IV was given at the patient’s request to supplement analgesia for all patients.
Continuous and categorical variables were reported as mean ± SD or percentages and compared with a 2-sample t tests or a χ2 test (2 tailed), respectively. Univariable association between recurrence-free survival or long-term survival and the type of postoperative analgesia was assessed with Kaplan-Meier survival estimates, and the groups were compared with the log-rank test9 and with univariable Cox proportional hazards regression. Multivariate Cox proportional hazards regression adjusting for propensity score were performed to assess associations of the type of postoperative analgesia and recurrence-free survival or long-term survival.
Because this was an observational study, a propensity score-adjusted analysis was performed to control for selection bias as a result of nonrandom assignment to the 2 groups. A propensity score was derived, reflecting the probability that a patient would receive postoperative EA. This was accomplished by performing a multivariate logistic regression analysis using postoperative EA as the dependent variable and entering all baseline variables as in Table 1.
In this study, the propensity score was used in regression (covariance) adjustment, that is, using a large set of baseline or intraoperative variables as above to estimate the propensity score, and then, the propensity score was subsequently regressed as an independent covariate in the multivariate Cox proportional hazards regression analysis, which was performed by using all relevant variables to identify risk factors for recurrence-free survival and long-term survival. To achieve model parsimony and stability, the backward stepwise selection procedure was applied with the drop-out criterion P > 0.1, but the propensity score and postoperative EA was forced in the model. Results are reported as percentages and hazards ratios (HR) and with 95% confidence intervals (CI). The success of the propensity score was estimated by assessing the balance of baseline characteristics after propensity score matching. The balance of each variable between the 2 groups was evaluated by the standardized difference (i.e., the absolute difference in sample means divided by an estimate of the pooled standard deviation (SD) of the variable expressed as a percentage).10 Balancing was considered successful when all standardized differences were <10%.11
All reported P values were 2-sided, and P values <0.05 were considered to be statistically significant. Because this study was exploratory, no correction for multiple testing was applied.12 Statistical analysis was performed with SPSS 17.0 software for Windows (SPSS Inc., Chicago, IL).
Of 1846 patients in the database, 819 patients met the inclusion criteria and were divided into 2 groups: patients with postoperative EA with morphine (n = 451) and patients with postoperative IA with fentanyl (n = 368) (Fig. 1). Demographic, oncologic, and operative characteristics of patients who received postoperative EA and postoperative IA are presented in Table 1. There was no statistically significant difference between the 2 postoperative analgesia groups regarding baseline variables except Child-Pugh class. Liver function of the patients receiving postoperative EA was worse than that of patients receiving postoperative IA.
The median (quartiles) follow-up time for all patients was 4.2 years (2–9). Cancer recurrence was detected in 37.7% (n = 170) of the patients who received postoperative EA and 30.7% (n = 113) of patients who received postoperative IA during follow-up (P = 0.036). The all-cause mortality was 40.6% (n = 183) among patients who received postoperative EA and 30.4% (n = 112) among patients who received IA during the follow-up (P = 0.003).
In addition, we successfully matched 89 patients in the EA group with 89 patients in the IA group using propensity scores. Their main baseline characteristics are shown in Table 2. The propensity score successfully balanced the baseline characteristics of the 2 groups, and all of the standardized differences were <10%.
The Kaplan-Meier survival estimates of recurrence-free survival for the postoperative EA and the postoperative IA groups are provided in Figure 2 and Table 3. The type of postoperative analgesia (EA versus IA) was not associated with a significant difference in recurrence-free survival (P = 0.170, log-rank test) (Fig. 2) with an unadjusted estimated HR of 1.170 (95% CI, 0.921–1.488).
Multivariate Cox proportional hazards regression analysis was adjusted for propensity score, covariates, and some covariate interactions to assess independent risk factors for recurrence-free survival. After adjusting for propensity score, covariates, and some covariate interactions, transfusion (HR 0.499, 95% CI, 0.273–0.911, P = 0.024) and cirrhosis (HR 1.605, 95% CI, 1.167–2.209, P = 0.004) were found to affect recurrence-free survival of patients after resection of HCC (Table 4). The effect of postoperative analgesia (HR 2.224, 95% CI, 0.207–23.893, P = 0.509) was not significant in the multivariate Cox proportional hazards regression analysis adjusting for propensity score (Table 4). In a Cox model accounting for the correlation within matched sets, the effect of the type of postoperative analgesia (HR 0.328, 95% CI, 0.022–7.236, P = 0.512) was also not significant (Table 5).
The Kaplan-Meier survival estimates of long-term survival for postoperative EA and postoperative IV groups are provided in Figure 3 and Table 6. IA was associated with increased long-term survival (P = 0.043, log-rank test) (Fig. 3), with an unadjusted estimated HR of 1.260 (95% CI, 0.995–1.595).
Multivariate Cox proportional hazards regression analysis was adjusted for propensity score, covariates and some covariate interactions to assess independent risk factors for long-term survival. After adjusting for propensity score, covariates and some covariate interactions, postoperative EA (HR 1.279, 95% CI, 1.007–1.625, P = 0.043) was found to be associated with decreased long-term survival. Other independent risk factors of decreased long-term survival were preoperative α-fetoprotein(+) (HR 1.838, 95% CI, 1.421–2.379, P < 0.001), ASA physical status (HR 3.185, 95% CI 2.318–4.375, P < 0.001), and tumor diameter (HR 1.434, 95% CI, 1.263–1.629, P < 0.001) (Table 4). In a Cox model accounting for the correlation within matched sets, independent risk factors of long-term survival were postoperative EA (HR 1.628, 95% CI, 0.845–3.725, P = 0.031), ASA physical status (HR 5.828, 95% CI, 1.417–17.848, P < 0.001) and tumor stage (HR 3.174, 95% CI, 1.232–5.137, P = 0.015) (Table 5).
A wealth of basic science data supports the hypothesis that the surgical stress response increases the likelihood of cancer dissemination and metastasis during and after cancer surgery. Anesthetic management of the cancer patient, therefore, could potentially influence long-term outcome.13 The effect of anesthetic management on cancer patients is complex; its mechanisms are not completely understood, and to a certain extent, contradictory.
Exadaktylos et al.3 demonstrated that regional anesthesia in combination with general anesthesia was associated with a longer cancer-free interval and a lower incidence of recurrence in a retrospective analysis of patients undergoing surgical treatment for breast cancer. Similar results were obtained in another retrospective study of patients with prostate cancer.2 However, Gottschalk et al.5 found that the use of EA for perioperative pain control during colorectal cancer surgery was not associated with a decreased cancer recurrence. Two other retrospective studies examining prostate cancer6 and cervical cancer7 also demonstrated no association between epidural anesthesia and long-term outcome of cancer patients.
In contrast to previous retrospective studies, the major findings from our observational cohort study suggest that postoperative EA with morphine has a higher incidence of recurrence (37.7% vs 30.7%, P = 0.036) and mortality (40.6% vs 30.4%, P = 0.003). More importantly, in multivariate Cox proportional hazards regression analysis after adjusting for the propensity score, we found that the postoperative EA with morphine was an independent risk factor for decreased long-term survival in patients after resection of HCC (HR 1.279, 95% CI, 1.007–1.625, P = 0.043). However, the use of postoperative EA with morphine did not show a significant effect on recurrence-free survival in this study (HR 2.224, 95% CI, 0.207–23.893, P = 0.509). These results were confirmed by similar analyses of the matched cases. These results also showed that postoperative EA with morphine had a negative effect on survival and cancer recurrence. Long-term survival is the most important end point and is least subject to investigator bias. It was also the primary end point recommended by the expert panel for any phase 3 study in HCC.14
The fact that our observations stand in partial contrast to previous studies can be attributed to at least 3 different mechanisms. First, is the timing of epidural anesthesia. Regional anesthesia, including spinal and epidural anesthesia, reduces the stress response and postoperative immunosuppression associated with surgery15,16 and decreases anesthetic requirements that could result in less immunosuppression.17 These associations may be theoretically beneficial to patients undergoing cancer surgery. In our series, we did not use the epidural intraoperatively to decrease the risk of awareness during anesthesia. Thus, similar suppression of the surgical stress response18 may not have occurred in patients in this study.
Second, we used different opioids in our study. In the EA group, morphine was the primary analgesic, whereas fentanyl and tramadol were used in the IA group. Although it has been established that opioids, morphine in particular, inhibit cellular and humoral immune function in humans,19–21 the synthetic opiates did not seem to exhibit the same immunosuppressive effects. In 1 study, fentanyl increased natural killer (NK) cell activity in healthy volunteers too.22 Tramadol also stimulates NK cell activity, both in rats and humans.20,23 In a rat model, tramadol was shown to block the enhancement of lung metastasis induced by surgery and to prevent surgery-induced suppression of NK cell activity.23 Although plasma levels of morphine after epidural administration are quite low, the central nervous system (CNS) concentration is 100 to 200 times the corresponding plasma levels.24 Morphine cannot only suppress immunity through effects on the CNS19 but can also cause CNS-mediated release of corticosteroids, which may lead to immune suppression.25 Morphine, in clinically relevant concentrations, also stimulates cancer cell survival, cell cycle progression, and endothelial proliferation and angiogenesis,26–28 which may contribute to cancer recurrence or decreased long-term survival.
Finally, the specific tumor type, which may be related to different tumor biology and the risk for and mechanisms of metastasis, may be associated with the effect of different types of analgesia on cancer recurrence or long-term survival. HCC is a hypervascular tumor mainly supplied by hepatic arteries and characterized by neovascularization, which plays an important role in the growth and progression of HCC.29 Morphine induces tumor neovascularization and increases tumor progression,28 which may be a possible reason for the negative effect of postoperative EA on the cancer recurrence rate and long-time survival of patients after hepatic resection.
HCC is often clinically silent for a long period and is often discovered at a very advanced stage. At this stage, the tumor progresses aggressively, and its prognosis may be more dependent on other factors. In our study, we excluded patients if the number of tumors ≥2, if thrombosis in portal vein or hepatic vein was noted, or if tumor metastasis were present. All these factors may affect HCC recurrence or long-time survival. We discovered several other risk factors for increased cancer recurrence and decreased survival including preoperative α-fetoprotein(+), ASA physical status, and tumor diameter. Sala et al.30 developed a staging system according to the results of cohort studies and found treatment indications, including tumor stage, liver functional status, physical status, and cancer-related symptoms. This is inconsistent with our results. In our study, liver function was worse (the ratio of Child-Pugh class B was higher) in patients who received EA. However, the multivariate Cox proportional hazards regression analysis adjusted for propensity score and covariates, Child-Pugh class (HR 1.372, 95% CI, 0.257–5.768, P = 0.832), was not an independent risk factor of long-term survival. This result is supported by Vauthey et al.31’s study, which also suggested the Child-Pugh class did not affect survival.
The present retrospective study has several limitations. It was not randomized, and a selection bias can thus not be definitively excluded, even with the propensity score analysis. Although surgical and anesthetic procedures are standardized in our institution and patients in each group had comparable surgery and anesthesia, we cannot exclude imbalanced baseline characteristics and some covariate interactions. Furthermore, our study does not provide information on the cause of death of patients. Cause of death might have been different in the early postoperative period and in later years. Postoperative EA may be associated with cancer recurrence or survival more significantly in the early period after operation than in the later years. Additional studies are needed to confirm these observations. Data on other types of cancer, other patient populations, adjuvant chemotherapy or radiation, and time and location of recurrence would also be important. Accordingly, we offer our findings as an impetus for future research in this area in random trials.
In this retrospective study, we report the cancer recurrence and long-term survival in patients assigned to receive postoperative EA with morphine or postoperative IA with fentanyl. Cancer recurrence rates and mortality were higher in patients who received EA with morphine than those who received IA with fentanyl. In addition, a negative effect of postoperative EA with morphine on long-term survival was observed, but no significant difference was found in recurrence-free survival. Prospective, randomized, and controlled clinical trials are warranted to reliably assess this important clinical question.
Name: Longhui Cao, PhD.
Contribution: This author helped design and conduct the study, analyze the data, and write the manuscript.
Attestation: Longhui Cao has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.
Name: Yi Chang, MD.
Contribution: This author helped design the study and analyze the data.
Attestation: Yi Chang has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.
Name: Wenqian Lin, MD.
Contribution: This author helped design the study and write the manuscript.
Attestation: Wenqian Lin has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.
Name: Jianhua Zhou, MD, PhD.
Contribution: This author helped analyze the data.
Attestation: Jianhua Zhou has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.
Name: Hongying Tan, MD, PhD.
Contribution: This author helped conduct the study.
Attestation: Hongying Tan has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.
Name: Yunfei Yuan, MD, PhD.
Contribution: This author helped write the manuscript.
Attestation: Yunfei Yuan has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.
Name: Weian Zeng, MD, PhD.
Contribution: This author helped design the study.
Attestation: Weian Zeng has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.
This manuscript was handled by: Edward C. Nemergut, MD.
1. Bar-Yosef S, Melamed R, Page GG, Shakhar G, Shakhar K, Ben-Eliyahu S. Attenuation of the tumor-promoting effect of surgery by spinal blockade in rats. Anesthesiology. 2001;94:1066–73
2. Page GG, Blakely WP, Ben-Eliyahu S. Evidence that postoperative pain is a mediator of the tumor-promoting effects of surgery in rats. Pain. 2001;90:191–9
3. Exadaktylos AK, Buggy DJ, Moriarty DC, Mascha E, Sessler DI. Can anesthetic technique for primary breast cancer surgery affect recurrence or metastasis? Anesthesiology. 2006;105:660–4
4. Biki B, Mascha E, Moriarty DC, Fitzpatrick JM, Sessler DI, Buggy DJ. Anesthetic technique for radical prostatectomy surgery affects cancer recurrence: a retrospective analysis. Anesthesiology. 2008;109:180–7
5. Gottschalk A, Ford JG, Regelin CC, You J, Mascha EJ, Sessler DI, Durieux ME, Nemergut EC. Association between epidural analgesia and cancer recurrence after colorectal cancer surgery. Anesthesiology. 2010;113:27–34
6. Tsui BC, Rashiq S, Schopflocher D, Murtha A, Broemling S, Pillay J, Finucane BT. Epidural anesthesia and cancer recurrence rates after radical prostatectomy. Can J Anesth. 2010;57:107–12
7. Ismail H, Ho KM, Narayan K, Kondalsamy-Chennakesavan S. Effect of neuraxial anaesthesia on tumour progression in cervical cancer patients treated with brachytherapy: a retrospective cohort study. Br J Anesth. 2010;105:145–9
8. Poon RT, Ng IO, Lau C, Yu WC, Yang ZF, Fan ST, Wong J. Tumor microvessel density as a predictor of recurrence after resection of hepatocellular carcinoma: a prospective study. J Clin Oncol. 2002;20:1775–85
9. Peto R, Peto J. Asymptotically efficient rank invariant test procedures. J R Stat Soc A. 1972;135:185–207
10. Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat Med. 2009;28:3083–107
11. Wahbi K, Meune C, Porcher R, Bécane HM, Lazarus A, Laforêt P, Stojkovic T, Béhin A, Radvanyi-Hoffmann H, Eymard B, Duboc D. Electrophysiological study with prophylactic pacing and survival in adults with myotonic dystrophy and conduction system disease. JAMA. 2012;307:1292–301
12. Wuethrich PY, Hsu Schmitz SF, Kessler TM, Thalmann GN, Studer UE, Stueber F, Burkhard FC. Potential influence of the anesthetic technique used during open radical prostatectomy on prostate cancer-related outcome: a retrospective study. Anesthesiology. 2010;113:570–6
13. Gottschalk A, Sharma S, Ford J, Durieux ME, Tiouririne M. Review article: the role of the perioperative period in recurrence after cancer surgery. Anesth Analg. 2010;110:1636–43
14. Llovet JM, Di Bisceglie AM, Bruix J, Kramer BS, Lencioni R, Zhu AX, Sherman M, Schwartz M, Lotze M, Talwalkar J, Gores GJ.Design and Endpoints of Clinical Trials in Hepatocellular Carcinoma.. Panel of Experts in HCC-Design Clinical Trials. J Natl Cancer Inst. 2008;100:698–711
15. Liu SS, Block BM, Wu CL. Effects of perioperative central neuraxial analgesia on outcome after coronary artery bypass surgery: a meta-analysis. Anesthesiology. 2004;101:153–61
16. Tønnesen E, Wahlgreen C. Influence of extradural and general anaesthesia on natural killer cell activity and lymphocyte subpopulations in patients undergoing hysterectomy. Br J Anesth. 1988;60:500–7
17. Hong JY, Lim KT. Effect of preemptive epidural analgesia on cytokine response and postoperative pain in laparoscopic radical hysterectomy for cervical cancer. Reg Anesth Pain Med. 2008;33:44–51
18. Tiouririne M. Epidural analgesia and cancer recurrence: timing matters. Anesthesiology. 2011;114:717–8; author reply 718
19. Yeager MP, Colacchio TA, Yu CT, Hildebrandt L, Howell AL, Weiss J, Guyre PM. Morphine inhibits spontaneous and cytokine-enhanced natural killer cell cytotoxicity in volunteers. Anesthesiology. 1995;83:500–8
20. Sacerdote P, Bianchi M, Gaspani L, Manfredi B, Maucione A, Terno G, Ammatuna M, Panerai AE. The effects of tramadol and morphine on immune responses and pain after surgery in cancer patients. Anesth Analg. 2000;90:1411–4
21. Beilin B, Shavit Y, Hart J, Mordashov B, Cohn S, Notti I, Bessler H. Effects of anesthesia based on large versus small doses of fentanyl on natural killer cell cytotoxicity in the perioperative period. Anesth Analg. 1996;82:492–7
22. Yeager MP, Procopio MA, DeLeo JA, Arruda JL, Hildebrandt L, Howell AL. Intravenous fentanyl increases natural killer cell cytotoxicity and circulating CD16(+) lymphocytes in humans. Anesth Analg. 2002;94:94–9
23. Gaspani L, Bianchi M, Limiroli E, Panerai AE, Sacerdote P. The analgesic drug tramadol prevents the effect of surgery on natural killer cell activity and metastatic colonization in rats. J Neuroimmunol. 2002;129:18–24
24. Nordberg G, Hedner T, Mellstrand T, Borg L. Pharmacokinetics of epidural morphine in man. Eur J Clin Pharmacol. 1984;26:233–7
25. Bryant HU, Bernton EW, Kenner JR, Holaday JW. Role of adrenal cortical activation in the immunosuppressive effects of chronic morphine treatment. Endocrinology. 1991;128:3253–8
26. Farooqui M, Li Y, Rogers T, Poonawala T, Griffin RJ, Song CW, Gupta K. COX-2 inhibitor celecoxib prevents chronic morphine-induced promotion of angiogenesis, tumour growth, metastasis and mortality, without compromising analgesia. Br J Cancer. 2007;97:1523–31
27. Gupta K, Kshirsagar S, Chang L, Schwartz R, Law PY, Yee D, Hebbel RP. Morphine stimulates angiogenesis by activating proangiogenic and survival-promoting signaling and promotes breast tumor growth. Cancer Res. 2002;62:4491–8
28. Singleton PA, Lingen MW, Fekete MJ, Garcia JG, Moss J. Methylnaltrexone inhibits opiate and VEGF-induced angiogenesis: role of receptor transactivation. Microvasc Res. 2006;72:3–11
29. Semela D, Dufour JF. Angiogenesis and hepatocellular carcinoma. J Hepatol. 2004;41:864–80
30. Sala M, Forner A, Varela M, Bruix J. Prognostic prediction in patients with hepatocellular carcinoma. Semin Liver Dis. 2005;25:171–80
31. Vauthey JN, Klimstra D, Franceschi D, Tao Y, Fortner J, Blumgart L, Brennan M. Factors affecting long-term outcome after hepatic resection for hepatocellular carcinoma. Am J Surg. 1995;169:28–34