Effects of Preemptive Analgesia on Pain and Cytokine Production in the Postoperative Period

Healthy female patients (American Society of Anesthesiologists physical status I or II) aged 40–70 yr old were included in this study after obtaining approval from the Hospital Human Studies Committee and informed consent from the patients. Patients were hospitalized to undergo transabdominal hysterectomy (for details, see table 1). On the preoperative visit by the anesthesiologist, patients were randomly assigned to one of two perioperative pain management techniques (treatment lasted until 48 h postoperatively): One group (N = 20) received PCEA in the postoperative period; the other (N = 21) received preemptive epidural followed by PCEA (PA + PCEA). On the same visit, patients were familiarized with the visual analog pain scale (VAS) and were instructed on the use of the PCEA pump.

Patients were premedicated with lorazepam (1–2 mg orally) 90 min before induction of anesthesia, followed by 2–3 mg intravenous midazolam on arrival in the operating theater. An epidural catheter was placed in all patients via  the L2–L4 interspaces and was advanced 3 to 4 cm cephalad. The position of the epidural catheter was tested with 3 ml lidocaine, 2%. Patients of the PA + PCEA group received an epidural mixture of 12 ml bupivacaine (0.5%) plus fentanyl (50–100 μg) 20–25 min before incision.

General anesthesia was induced using 2–3 μg/kg fentanyl, 4–6 mg/kg thiopental, and 0.1 mg/kg vecuronium, intravenously. Anesthesia was maintained with nitrous oxide, isoflurane, and additional fentanyl. Mean arterial blood pressure was maintained within 20% of baseline values with isoflurane and fentanyl. Patients received upper body forced-air warming, and intravenous fluids were warmed to 37°C. Patients requiring blood transfusion during the perioperative period were not included in this study.

On arrival to the postanesthesia care unit, patients of both groups were connected to the PCEA pump (Pain Management Provider; Abbott, Chicago, IL) and received 3 ml bupivacaine (0.1%) plus 2 μg/ml fentanyl per demand (lockout time, 10 min), with continuous background infusion of 6 ml/h.

A 10-cm VAS (with end points labeled “no pain” and “worst possible pain”) was used to assess pain intensity in rest and after coughing at 4, 8, 12, 24, 48, and 72 h after completion of surgery.

Venous blood samples (15 ml) were collected on the morning of the surgery and at 24, 48, and 72 h following surgery. Peripheral blood mononuclear cells (PBMCs) were isolated from heparinized venous blood using a histopaque (Sigma-Aldrich, Rehovot, Israel) gradient centrifugation. PBMCs were washed twice in RPMI-1640 medium containing 1% penicillin, streptomycin, and nystatin and supplemented with 10% fetal calf serum (designated complete medium). Cells were suspended in fetal calf serum containing 10% dimethyl sulfoxide (Sigma) and were frozen at −70°C until used. On the day of assay, cells were thawed quickly and washed three times in complete medium, and their viability was tested by trypan blue dye exclusion. The viability was over 95%.

Peripheral blood mononuclear cells (2 × 10⁶) suspended in 1 ml RPMI-1640 supplemented with 5% fetal calf serum were incubated for 24 h in the presence of 10 ng/ml lipopolysaccharide (E. coli ; Sigma-Aldrich) for IL-1β, IL-1ra, IL-6, IL-10, and tumor necrosis factor (TNF)-α. For IL-2 production, 2 × 10⁶PBMCs were suspended in 1 ml complete medium and were incubated for 48 h with 1% phytohemagglutinin (PHA-M; Difco Laboratories, Detroit, MI). Following the incubation period, culture media were collected, cells were removed by centrifugation, and the supernatants were kept at −70°C until assayed for cytokine content.

Cytokine concentration in the supernatant was assessed using ELISA kits specific for human IL-1β, IL-1ra (Biosource International, Camarillo, CA), IL-6, (Pharmingen, San Diego, CA), IL-2 (R&D Systems, Minneapolis, MN), and IL-10 (Genzyme Corporation, Cambridge, MA), as detailed in the guidelines provided by the manufacturers. The detection levels of the cytokines in the assays were 30 pg/ml for IL-1β and IL-2, 15 pg/ml for IL-6, and 60 pg/ml for IL-1ra and TNF-α.

The number of observations varies slightly among assays as a result of an occasional missing sample or an occasional problem in a particular assay. Data were analyzed for each measure, using analysis of variance with repeated measures (time periods before and after surgery). 5A post hoc  Bonferroni procedure was conducted as appropriate, correcting for multiple comparisons. 5Probability values of P < 0.05 were considered significant. The results are expressed as mean ± SEM.

The groups were similar in body weight, age, duration of surgery, and total fentanyl consumption in 48 h postoperatively (table 1).

Pain scores reveal similar trends between the measures taken at rest or during coughing. There were significant differences among the study groups (F^1,39= 20.71 or 23.67;P < 0.0001 for VAS at rest or VAS during coughing, respectively;figs. 1A and B). Patients of the PA + PCEA group experienced less severe postoperative pain throughout the postoperative observation period.

Analysis of variance with repeated measures for IL-1β levels revealed significant main effects of groups (PCEA vs.  PA + PCEA; F^1,39= 13.63, P < 0.001) and trials (preoperative, 24, 48, and 72 h postoperative; F^3,117= 6.08, P < 0.001). There was also a significant interaction (group × trial; F^3,117= 3.11, P < 0.03). The analysis of variance test for IL-6 revealed significant main effects of groups (F^1,39= 4.25, P < 0.05) and trials (F^3,117= 9.72, P < 0.001), and significant interaction (group × trial; F^3,117= 2.89, P < 0.04). These results indicate that proinflammatory cytokine production by stimulated PBMCs was elevated in the postoperative period and that such elevation was significantly less pronounced in patients of the PA + PCEA group compared with patients of the PCEA group for both IL-1β and IL-6 (figs. 2A and B, respectively). Levels of TNF-α exhibited a similar trend, which was not statistically significant (fig. 2C). It is known that the maximal release of TNF-α by lipopolysaccharide-stimulated PBMCs occurs earlier than 24 h. This may account for the lack of significant group differences at the times observed in the present study.

Ex vivo  production of antiinflammatory cytokines was also assessed by stimulated PBMCs. IL-10 and IL-1ra levels were elevated in the postoperative period (significant trial effects in both groups), but this elevation was significantly less pronounced in patients of the PA + PCEA group (group effects: F^1,38= 6.16, P < 0.02 and F^1,36= 5.22, P < 0.03, respectively;figs. 3A and B). The IL-1ra/IL-1β ratio exhibited a slight, but not significant, linear increase from the preoperative to the 72 h postoperative measures (0.74–1.23). There were no significant main effects or a significant interaction. That is, the two groups did not differ in the IL-1ra/IL-1β ratio at any time point (data not shown).

Ex vivo  IL-2 production was assessed in PHA-stimulated PBMCs. There was a significant trial effect (F^3,117= 6.18, P < 0.001), indicating a significant suppression of IL-2 levels in 24 and 48 h postoperatively, examined over the two study groups. There was a significant difference between the study groups in IL-2 levels (F^1,39= 5.41, P < 0.025;fig. 4). IL-2 levels remained almost unchanged in the PA + PCEA group, while they were significantly suppressed in the PCEA group in the first 48 h postoperatively, gradually recovering by 72 h.

Patients of the PA + PCEA group exhibited less severe postoperative pain, compared to the PCEA group. Ex vivo  stimulated PBMCs of these patients produced less elevated levels of both proinflammatory and antiinflammatory cytokines in the postoperative period. In addition, production of IL-2 by stimulated PBMCs was significantly less suppressed in the PA + PCEA group compared to the PCEA group. The findings of increased production of proinflammatory cytokines and suppressed production of IL-2 observed in the postoperative period in our present and previous studies 6are in agreement with reports of alterations in cytokine serum levels following surgery (e.g. , Baxevanis et al.  7).

Surgery-associated tissue injury sets off a cascade of related events, including nociception and inflammatory reaction. Tissue and peripheral nerve injury leads to local inflammatory reaction, accompanied by elevated levels of proinflammatory cytokines, including IL-1β and IL-6. 8These cytokines can induce peripheral and central nerve system sensitization, leading to pain augmentation (hyperalgesia). 8Peripherally, IL-1β induces long-lasting synthesis and release of substance P from peripheral nerve terminals of primary afferent neurons, which may contribute to neurogenic inflammation. 9,10 

Within minutes of injury, glial cells in the central nervous system respond with increased production of immune factors, including proinflammatory cytokines. 11IL-1β can induce central sensitization via  IL-1 receptors on neurons 12or via  activated glia cells, which produce pain mediators, including substance P, glutamate, and nitric oxide synthase, 13all of which can alter pain processing within the central nervous system. Elevated IL-1β in the central nervous system also leads to the production of COX2 by neurons in the brain and spinal cord and further synthesis of PGE2, which is known to increase pain sensitivity. 14Similarly, IL-6 levels are also elevated following nerve injury, both peripherally and centrally, contributing to hyperalgesia by direct spinal nociceptive mechanisms or by glial activation. 15,16IL-1β and IL-6 are involved in the mechanisms of allodynia and possibly in the development of postoperative neuropathic and chronic pain. 11,17 

It has been argued that preemptive analgesia, commencing before surgery and continuing in the postoperative period, prevents the establishment of peripheral and central sensitization. 3In the present study, patients receiving preemptive analgesia exhibited less severe pain and attenuation of proinflammatory cytokine production throughout the postoperative period. It is difficult to determine whether blocking the pain contributes to the reduced production of proinflammatory cytokines or, vice versa , reduced production of proinflammatory cytokines results in less severe pain experience.

It is noteworthy that patients receiving preemptive analgesia also exhibited less suppression of IL-2 production in the postoperative period. Since IL-2 is secreted by the TH1 subset of lymphocytes, the present finding may suggest attenuated suppression of the TH1 phenotypic response in the PA + PCEA pain management group. It has been shown that IL-2 structurally resembles opiate peptides and has analgesic effects in both the peripheral and central nervous system. 18Thus, it is possible that less suppressed IL-2 levels seen in the PA + PCEA group contributes to the less severe pain reported by these patients.

In conclusion, preemptive analgesia results in reduced pain intensity and attenuated production of proinflammatory cytokines by stimulated PBMCs in the postoperative period. These findings are interrelated, demonstrating correlation between subjective measure of pain and objective measures of immune reactivity. Proinflammatory cytokines are key mediators of illness symptoms, including fever, reduction in synthesis and release of acute phase proteins, deceased food and water intake, and hyperalgesia. This study suggests that reduced secretion of proinflammatory cytokines following preemptive analgesia may contribute to reduced postoperative pain and possibly to attenuated illness response. 19