Afferent neural blockade by epidural anesthesia can decrease the postoperative neuroendocrine stress response (1). A neuroendocrine response blunted by epidural anesthesia also could affect postoperative immune function because the immune and nervous systems bidirectionally communicate and influence each other (2). It has been argued that nociception and proinflammatory cytokines play a mutual up-regulatory role (3). Therefore, pain management may influence the immune response in the postoperative period. It has been reported that the alterations in lymphocyte subsets and the increase in white cell counts induced by surgery and general anesthesia can be prevented by epidural analgesia (4–6). Furthermore, Beilin et al. (7,8) reported that patients treated with patient-controlled epidural analgesia (PCEA) exhibited attenuated proinflammatory cytokine response in the postoperative period. These reports resulted from procedures below the umbilicus. In upper abdominal or major surgery, however, the effect of epidural anesthesia and analgesia on attenuation of the stress response and preservation of immune function is controversial (9,10). Major surgery induces a series of inflammatory responses, and the stress response is often exaggerated after major surgery to the detriment of the patient (11,12). Postoperative immune suppression can predispose patients with cancer to the development of postoperative infection (13) or facilitate postoperative tumor growth and metastasis (14,15).
Abdominothoracic esophagectomy with multiple lymph node dissection (radical esophagectomy) is a major surgical procedure with frequent morbidity and mortality (16). Theoretically, blunting the perioperative stress response and preserving immune function are critical issues for patients undergoing this procedure. Several studies compared postoperative epidural analgesia with postoperative IV analgesia and reported reductions in time spent in the intensive care unit (ICU) and fewer respiratory complications for patients receiving epidural analgesia (17,18). However, the effects of epidural anesthesia and analgesia on the stress response and immune function during and after radical esophagectomy have not been studied.
The present study was conducted to determine whether epidural anesthesia and analgesia affect the stress response and immune function in patients undergoing radical esophagectomy. In previous studies of procedures above the umbilicus, it appears that the incomplete effect of epidural anesthesia on the stress response was largely attributable to incomplete neural blockade at the surgical site (19). It is reported that a block of the sensory fibers included in the phrenic nerves is required for a complete block of the endocrine stress response to upper abdominal surgery (20). Thus, two epidural catheters for extensive nerve blockade (C3-L) were used in this study.
Institutional and ethics committee approval was obtained for this study, and all participants gave written informed consent. The study group comprised 30 patients (ASA physical status I–II) with diagnosed esophageal cancer and scheduled for radical esophagectomy. Because the cough reflex is suppressed for several days after radical esophagectomy, all patients were mechanically ventilated at least until the morning of postoperative day (POD) 3. The surgical procedure included thoracoabdominal esophagectomy, multiple lymph node dissection, and a cervical incision for the anastomosis. Patients were excluded from the study if they had recently taken corticosteroids or nonsteroidal antiinflammatory medications. The participants were randomly assigned to one of two methods of anesthesia and postoperative pain control. Fifteen patients received general anesthesia with continuous epidural infusion (CEI) through two epidural catheters during surgery; CEI was continued for postoperative pain control (group E). Fifteen patients received general anesthesia during surgery with postoperative analgesia accomplished by continuous IV morphine (group G).
On the day of surgery, patients in both groups were premedicated IM with 50 mg hydroxyzine and 0.5 mg atropine. An IV cannula was placed for fluid administration, and a radial arterial cannula was placed for continuous arterial blood pressure monitoring and blood sampling. In group E, 2 epidural catheters were inserted at the T3-4 and T10-11 interspaces. The analgesic area was checked by pinprick after 20 min of epidural injection of 8 mL 1.5% lidocaine via each epidural catheter and was confirmed to extend at least from the C3 to L2 dermatomes in all patients. Then, CEI of 1% lidocaine with 4 μg/mL fentanyl was administered at 6 mL/h during surgery. In both groups, general anesthesia was induced IV with 4 mg/kg thiopental and 50–100 μg fentanyl. Intubation of the tracheal was facilitated with 0.1 mg/kg vecuronium. Ventilation was controlled mechanically to maintain the partial pressure of expiratory carbon dioxide between 30 and 35 mm Hg, as measured by capnography. General anesthesia was maintained with 50% nitrous oxide in oxygen and isoflurane. Intermittent boluses of 50 μg fentanyl were given as necessary. Vecuronium was used to maintain muscular blockade. After the induction of general anesthesia, a double-lumen catheter was inserted via the right internal jugular vein for continuous central venous pressure monitoring and fluid administration. Heart rate (by electrocardiography), CO2 production (by capnography), arterial blood pressure, central venous pressure, blood oxygen saturation (by pulse oximetry), and bladder temperature were monitored. Lactated Ringer’s solution was infused at 20 mL · kg−1 · h−1 during the induction of anesthesia and to maintain preoperative central venous pressure. Hypotension (arterial blood pressure <90 mm Hg) was corrected by increasing the rate of IV fluid infusion and by administration of ephedrine. Blood was replaced as necessary; the hemoglobin level was maintained at more than 8.0 g/dL with transfusion of packed red blood cells, and the albumin level was maintained at more than 3.0 g/dL with administration of 5% albumin or fresh-frozen plasma. Bladder temperature was maintained between 36°C and 37°C with the use of a warming device (Bair Hugger; Augustine Medical Inc., Minneapolis, MN). Arterial blood gases, serum electrolytes (Na+, K+, Cl−, Ca2+), and blood glucose levels were analyzed (Stat Profile M; Noba Medical, Waltham, MA) at least every 1 h, and these values were maintained within normal ranges by appropriate treatments.
After surgery, patients in both groups were transferred to the ICU and mechanically ventilated. While on the ventilator, patients were lightly sedated (Ramsey scale 2; cooperative and tranquil) (21) by continuous infusion of propofol.
CEI of 0.2% ropivacaine with 4 μg/mL fentanyl was administered at 4 mL/h via each epidural catheter after surgery for postoperative pain control. If a patient complained of pain, an epidural injection of 5 mL 0.2% ropivacaine was administrated as a supplemental analgesic. In group G, patients received an initial loading dose of morphine (4–5 mg) for pain relief on arrival at the ICU and a continuous infusion of 1 mg/h morphine was started. If a patient complained of pain, a 2.5 mg IV bolus of morphine was administered as a supplemental analgesic. Pain intensity was assessed at the end of surgery and every 12 h after surgery with a box scale (0–10, where 0 is no pain and 10 is the worst pain ever). The independent observer who was not involved in this study assessed pain analgesic area by pinprick and pain intensity.
Arterial blood samples (10 mL) were collected before and at the end of surgery and on POD1 and POD3. Blood samples were treated with EDTA, and 2 mL of blood was used for the analysis of lymphocyte subsets and blood cell counts. The rest of the sample was centrifuged, and the plasma was frozen at −80°C until further analysis. The plasma concentrations of epinephrine, norepinephrine, cortisol, and adrenocorticotropic hormone (ACTH) were measured, the number of leukocyte was counted, the distribution of lymphocyte subsets was analyzed, and the plasma levels of interleukin (IL)-1β, IL-6, IL-10, tumor necrosis factor (TNF)-α, and C-reactive protein (CRP) were measured before and at the end of surgery and on POD1 and POD3.
Lymphocyte subsets were analyzed by flow cytometry (EPICS ELITE; Coulter, Miami, FL) with fluorescence-labeled antibodies specific to the cell markers (Coulter). A 0.1-mL blood sample was incubated for 30 min with monoclonal antibodies at 4°C in the dark. The samples were processed with a Q-prep Immunology Station (Coulter), which lyses the erythrocytes in a semiautomatic fashion, stabilizes the leukocytes, and fixes the cells. The percentage of lymphocytes relative to the total leukocyte count was determined through differential gating after triple-color staining. The following antibodies to lymphocyte antigens were used and cell types determined: cluster of differentiation (CD)3-CD19+ (B cells), CD3+CD19- (T cells), CD3+CD4+ (inducer and helper T cells), CD3+CD8+ (suppressor and cytotoxic T cells), and CD3-CD16+CD56+ (natural killer [NK] cells). This method of lymphocyte subset analysis is accurate, with a high degree of specificity and precision; the coefficient of variation we obtained was <2.5%. The total leukocyte number and the percentage of total lymphocytes were measured with a cell counter (MAXM-Retic; Coulter).
The plasma concentrations of epinephrine and norepinephrine were measured by high-performance liquid chromatography with electrochemical detection according to the method described by Weicker et al. (22). The sensitivity limit of this method was 1 pg/mL for each catecholamine. Commercially available radioimmunoassay kits were used to measure the plasma concentrations of ACTH (Allegro HS-ACTH; Nichols Institute, San Juan Capistrano, CA) and cortisol (Amerlex Cortisol RIA Kit; Amersham, Arlington Heights, IL). The sensitivity limits were 1 pg/mL for ACTH and 0.1 μg/dL for cortisol. The plasma concentrations of IL-1β, IL-6, IL-10, and TNF-α were measured by enzyme-linked immunosorbent assay kits, and the detection levels were as follows: 0.125 pg/mL for IL-1β, 0.15 pg/mL for IL-6, 0.5 pg/mL for IL-10, and 0.35 pg/mL for TNF-α.
Sample size was determined based on our preliminary study of patients undergoing esophageal surgery. Power calculation had indicated that 15 patients would be required per group to detect a difference of 25% in proportion of lymphocyte, white cell count, and plasma levels of epinephrine, cortisol, and IL-1β with a power of 80%. We felt the smallest clinically significant difference in pain score and in CRP would be 2 and 2.0 mg/dL, respectively. Two groups of 15 patients were also sufficient to detect both these differences with a power of 80%. Because the values of ACTH, IL-6, IL-10, and TNF-αvaried so widely, more than 100 patients would be required to detect a difference of 25% in these values with a power of 80%. We therefore showed these differences with an α error of 0.05 and an insufficient power. Data are expressed as mean ± sd. Student’s t-test (unpaired) was used to analyze between-group differences in patients characteristics (age, weight, height), duration of surgery, amounts of fluid and blood infused, and amount of blood lost during surgery. Pain scores and plasma levels of epinephrine, norepinephrine, cortisol, ACTH, IL-1β, IL-6, IL-10, TNF-α, and CRP, and the leukocyte count, and the percentage of lymphocyte subsets were analyzed by the Kruskal-Wallis test followed by Dunn’s method within and between groups. Differences were considered statistically significant at P < 0.05.
Patient characteristics, duration of surgery, amounts of fluid and blood infused, and amount of blood lost during surgery were similar between the groups (Table 1). Pain scores in group G were significantly higher than in group E at the end of surgery (P < 0.01), but there was no significant different between groups thereafter (Fig. 1). The epidural-blocked area at the end of surgery was not assessed exactly because of surgical dressing and slight sedation, but no pain at the cervical, thoracic and abdominal regions was affirmed in all patients of group E. Doses of fentanyl administered in group E and the dose of morphine administered in group G for the first 24 h after surgery were 797 ± 47 μg and 35.8 ± 3.4 mg, respectively. The mean dose of propofol administered for the first 24 h after surgery was similar between groups (1.0 ± 0.1 mg · kg−1 · h−1 in group E and 1.0 ± 0.0 mg · kg−1 · h−1 in group G). All variables were similar between groups before surgery.
The plasma concentration of epinephrine increased significantly in group G (P < 0.05) to a level significantly more than that in group E at the end of surgery (P < 0.05; Fig. 2A). The plasma concentration of norepinephrine increased significantly in group G (P < 0.01) at the end of surgery and on POD1 and POD3 to a level significantly more than that in group E (P < 0.01; Fig. 2B). The CD4+/CD8+ ratio decreased significantly at the end of surgery in group G (P < 0.05) to a value significantly less than that in group E (P < 0.05; Fig. 5D). However, there were no significant differences in any other values between the groups (Figs. 2–5). In both groups, the following postoperative values differed significantly in comparison to the preoperative values. The plasma levels of cortisol and ACTH were increased at the end of surgery (each group P < 0.01; Figs. 2C and D), and IL-1β, IL-6, IL-10, and the CRP levels were increased at the end of surgery and on POD1 and POD3 (each group: P < 0.01; Figs. 3A, B, C and Fig. 4C). The leukocyte count was increased on POD1 (each group P < 0.05) and POD3 (each group P < 0.01) (Fig. 4A), whereas the proportion of lymphocytes decreased from the end of surgery to POD3 (each group P < 0.01; Fig. 4B). Among the lymphocyte subsets, the proportion of B cells was increased on POD1 (each group P < 0.01; Fig. 5A), and that of NK cells decreased on POD1 and POD3 (each group P < 0.01; Fig. 5C).
Our findings indicate that extensive epidural block by two epidural catheters was not able to suppress leukocytosis, lymphopenia, increases in cytokines and CRP, or changes in proportions of lymphocyte subsets after radical esophagectomy. Thus, it appears that tissue damage and inflammation overcome the effects of extensive epidural block on stress responses and immune functions in patients undergoing radical esophagectomy.
There is indirect evidence that increased plasma norepinephrine changes reflect increased sympathetic activity during stress (23). The increase in the plasma level of catecholamines was suppressed and postoperative pain was not observed in group E at the end of surgery. Thus, CEI via two epidural catheters was able to suppress activation of the sympathetic and somatic nervous systems during surgery. However, CEI did not suppress the increase in plasma levels of cortisol and ACTH. These results indicate that the stress response was not completely suppressed during surgery by epidural anesthesia, despite blockade of the nervous system.
It is reportedly difficult to prevent the surgical stress response in procedures above the umbilicus (9,10). There are a few likely causes of the incomplete effect of epidural anesthesia on the stress response to procedures above the umbilicus (19). First, the doses of local anesthetic used may not be sufficient to produce complete neural blockade at the surgical site (20). Second, cytokines directly activate the stress response and are released in larger quantities after surgery above the umbilicus (24). Segawa et al. (20) reported that a block of the sensory fibers included in the phrenic nerves (C3–5 spinal nerves) is required for a complete block of the endocrine stress response to upper abdominal surgery. We used two epidural catheters to provide extensive neural blockade (C3-L2) because most patients in our preliminary study could not develop analgesia of C3-L2 by the one catheter method described by Segawa et al. (20). Although administration of 1.5% lidocaine provided extensive neural blockade before surgery, we could not confirm the blocked area during continuous infusion of 1% lidocaine with fentanyl. Although the analgesic area was assessed at the end of surgery, it was difficult to check the blocked area exactly by pinprick because of the surgical dressing and incomplete emergence from anesthesia. The doses may not have been adequate to provide complete neural blockade at the surgical site during surgery. However, suppression of catecholamines and absence of pain were achieved. Taken together, our findings indicate that the direct effects of cytokines were the likely cause of the surgical stress response after radical esophagectomy.
The postoperative neuroendocrine response to stress is now believed to be mediated by cytokines such as IL-1β, IL-6, and TNF-α, as well as by some of the more classic stress hormones including cortisol and the catecholamines (24–26). IL-6 is the primary stimulus for acute responses, and plasma levels of IL-6 are reportedly related to the severity of surgical trauma (27). Silverman et al. (28) demonstrated recently that IL-6 can directly stimulate corticosteroid secretion from the adrenal cortex, and there is evidence suggesting that it can also stimulate ACTH secretion from the pituitary gland. We found that plasma levels of IL-6 increased significantly after surgery in both of our groups. Therefore, the increase in stress hormones during surgery appears to be mediated by cytokines that are not suppressed by epidural blockade. IL-10, which has strong antiinflammatory and immunoinhibitory activities, was also not affected by epidural analgesia, and the peak level of IL-10 was similar to that of IL-6.
CRP was increased significantly in both groups after surgery; the peak level was observed on POD3. The consistently high level of CRP after surgery indicates that surgical inflammation continues for several days after radical esophagectomy. Volk et al. (5) also reported that PCEA, in comparison with patient-controlled analgesia (PCA), had no influence on altered levels of circulating cytokines (IL-6, IL-8, IL-10) or indicators of the stress response (CRP and cortisol) after major spine surgery.
In our study, leukocytosis and lymphopenia were not prevented by epidural anesthesia and analgesia. Similar changes in lymphocyte subsets were observed in both groups, with the exception of the CD4/CD8 ratio at the end of surgery. The absence of a significant decrease in the CD4/CD8 ratio during surgery might have been attributable to epidural anesthesia. However, the ratio returned to the presurgical value even in the group without epidural analgesia, and the values remained at presurgical levels on POD1 and POD3 in both groups. The proportion of NK cells decreased in both groups during and after surgery. Although Volk et al. (5) reported that postoperative epidural anesthesia preserves lymphocytes after major spine surgery, our findings suggest that CEI only affected immune function slightly after radical esophagectomy.
There are several limitations to the current study. First, a small participant size might have affected the results. We could not provide a sufficient power with regard to IL-6, IL-10, and TNF-α. However, it was difficult to recruit more than 100 patients scheduled for radical esophagectomy for short-term study in our hospital. We, thus, showed these differences with an α error of 0.05 and an insufficient power. Second, it was impossible to blind this study because one group had two epidural catheters. However, the independent observer who was not involved in this study collected pain scores by a box scale. This protocol likely did not cause a bias between groups. Third, sedation during intubation/ventilation might have affected results. However, patients were slightly sedated (Ramsey scale 2) and were able to give pain scores easily with a box scale. The dose of propofol administered was similar in both groups and the dose was small (1.0 mg · kg−1 · h−1), so it was unlikely to cause changes in pain scores and other valuables.
The present findings indicate that combined regional/general anesthesia with epidural anesthesia for blockade of afferent neural impulses does not attenuate stress-induced cytokine production during and after radical esophagectomy. The use of two separate epidural catheters with independent risks for complications may remove most clinical relevance. However, this does not mean that epidural anesthesia and analgesia by a single catheter are useless for postoperative care. Tsui et al. (17) observed fewer cardiovascular complications, less frequent morbidity and mortality, and a shorter hospital stay for patients undergoing esophageal surgery with epidural analgesia compared with patients with PCA. Smedstad et al. (18) compared postoperative epidural analgesia with postoperative IV analgesia and reported a reduction in the time spent in the ICU and in the total time spent in the hospital for patients treated epidurally. A multimodal approach for patients undergoing abdominothoracic esophagectomy including thoracic epidural analgesia, early tracheal extubation, and forced mobilization could reduce the ICU stay (16).
In conclusion, our data demonstrated that extensive epidural block by two catheters was not able to preserve immune function and suppress acute inflammatory responses after radical esophagectomy. We recommend epidural analgesia by a single catheter to improve recovery after radical esophagectomy as a multimodal approach.
1. Kehlet H. Modification of responses to surgery by neural blockade: clinical implications. In: Cousins MJ, Bridenbaugh PO, eds. Neural blockade in clinical anesthesia and management of pain. Philadelphia, JB Lippincott, 1988:145–90.
2. Basedovsky HO, Del Rey A. Immune-neuro-endocrine interactions: facts and hypotheses. Endocr Rev 1996;17:64–102.
3. Watkins LR, Maier SF, Goehler LE. Immune activation: the role of pro-inflammatory cytokines in inflammation, illness responses, and pathological pain states. Pain 1995;63:289–302.
4. Tonnesen E, Wahlgreen C. Influence of extradural and general anaesthesia on natural killer cell activity and lymphocyte subpopulations in patients undergoing hysterectomy. Br J Anaesth 1988;60:500–7.
5. Volk T, Schenk M, Voigt K, et al. Postoperative epidural anesthesia preserves lymphocyte, but not monocyte, immune function after major spine surgery. Anesth Analg 2004;98:1086–92.
6. Moor CM, Desborough JP, Powell H, et al. Effects of extradural anaesthesia in interleukin-6 and acute phase response to surgery. Br J Anaesth 1994;72:272–9.
7. Beilin B, Shavit Y, Trabekin E, et al. The effects of postoperative pain management on immune response to surgery. Anesth Analg 2003;97:833–7.
8. Beilin B, Bessler H, Mayburd E, et al. Effects of preemptive analgesia on pain and cytokine production in the postoperative period. Anesthesiology 2003;98:151–5.
9. Tsuji H, Asoh T, Takeuchi Y, Shirasaka C. Attenuation of adrenocortical response to upper abdominal surgery with epidural blockade. Br J Surg 1983;70:122–4.
10. Hjortso NC, Christensen NJ, Andersen T, Kehlet H. Effects of the extradural administration of local anaesthetic agents and morphine on the urinary excretion of cortisol, catecholamines and nitrogen following abdominal surgery. Br J Anaesth 1985;57:400–6.
11. Kehlet H. The surgical stress response: should it be prevented? Can J Surg 1991;34:565–7.
12. Bone RC. Toward an epidemiology and natural history of SIRS (systemic inflammatory response syndrome). JAMA 1992;268:3452–5.
13. Meakins JL. Surgeons, surgery, and immunomodulation. Arch Surg 1991;126:494–8.
14. Pollock RE, Lotzova E, Stanford SD. Mechanism of surgical stress impairment of human perioperative natural killer cell cytotoxicity. Arch Surg 1991;126:338–42.
15. Eggermont AM, Steller EP, Sugarbaker PH. Laparotomy enhances intraperitoneal tumor growth and abrogates the anti-tumor effects of interleukin-2 and lymphokine-activated killer cells. Surgery 1987;102:71–8.
16. Brodner G, Pogatzki E, Van Aken H, et al. A multimodal approach to control postoperative pathophysiology and rehabilitation in patients undergoing abdominothoracic esophagectomy. Anesth Analg 1998;86:228–34.
17. Tsui SL, Law S, Fok M, et al. Postoperative analgesia reduces mortality and morbidity after esophagectomy. Am J Surg 1997;173:472–8.
18. Smedstad KG, Beattie WS, Blair WS, Buckley DN. Postoperative pain relief and hospital stay after total esophagectomy. Clin J Pain 1992;8:149–53.
19. Liu S, Carpenter RL, Neal JM. Epidural anesthesia and analge-sia. Their role in postoperative outcome. Anesthesiology 1995;82:1474–506.
20. Segawa H, Mori K, Kasai K, et al. The role of the phrenic nerves in stress response in upper abdominal surgery. Anesth Analg 1996;82:1215–24.
21. Ramsey MAE, Savage TM, Simpson BRJ, Goodwin R. Controlled sedation with alfaxolone-alphadolone. Brit Med J 1974;2:656–9.
22. Weicker H, Feraudi M, Hägele H, Pluto R. Electrochemical detection of catecholamines in urine and plasma after separation with HPLC. Clin Chim Acta 1984;141:17–25.
23. Bravo EL, Tarazi RC. Plasma catecholamines in clinical investigation: a useful index or a meaningless number? J Lab Clin Med 1982;100:155–60.
24. Naito Y, Tamai S, Shingu K, et al. Responses of plasma adrenocorticotropic hormone, cortisol, and cytokines during and after upper abdominal surgery. Anesthesiology 1992;77:426–31.
25. Sapolsky R, Rivier C, Yamamoto G, et al. Interleukin-1 stimulates the secretion of hypothalamic corticotropin-releasing factor. Science 1987;238:522–4.
26. Baumann H, Prowse KR, Marinkovic S, et al. Stimulation of hepatic acute phase response by cytokines and glucocorticoids. Ann NY Acad Sci 1989;557:280–95.
27. Cruickshank AM, Fraser WD, Burns HJ, et al. Response of serum interleukin-6 in patients undergoing elective surgery of varying severity. Clin Sci 1990;79:161–5.
© 2005 International Anesthesia Research Society
28. Silverman MN, Miller AH, Biron CA, Pearce BD. Characterization of an interleukin-6 and adrenocorticotropin-dependent, immune-to-adrenal pathway during viral infection. Endocrinology 2004;145:3580–9.