Antiinflammatory Effect of Peripheral Nerve Blocks after Knee Surgery: Clinical and Biologic Evaluation

Our study was approved for the two institutions by an ethics committee (Comité de Protection des Personnes, Boulogne Billancourt, France), and all patients gave written informed consent. We recruited consecutive patients scheduled to undergo total knee arthroplasty, regardless of age. Patient inclusion criteria were an indication for total knee arthroplasty due to knee arthritis and surgery performed during general anesthesia. Exclusion criteria were previous surgery or trauma of the knee or preoperative use of corticosteroids. Nonsteroidal antiinflammatory drugs were discontinued 48 h before surgery for all patients and substituted with acetaminophen. The main outcome measure was IL-6 plasma concentration 24 h after surgery. Based on previous reports,13,1520 patients were needed in each group to detect a 50% reduction in IL-6 plasma concentration at 24 h after surgery with a power of 80% and an α risk of 5%. Secondary clinical outcome measures for inflammation were knee circumference and temperature.

All patients were given 100 mg hydroxyzine before surgery. Surgery was performed during balanced general anesthesia combining propofol, sufentanil, a muscle relaxant, and sevoflurane. The same surgeon (P.P.) performed all operations. Postoperative pain was controlled for all patients by intravenous morphine patient-controlled analgesia (PCA) in combination with intravenous acetaminophen (1 g every 6 h; Perfalgan®; BMS Laboratory, Rueil Malmaison, France). PCA delivered a 1-mg morphine bolus with a 5-min lockout time. In cases of poor pain control by the patient, the PCA bolus was increased to 2 mg. PCA analgesia was discontinued 72 h after surgery. A randomization list was generated, and patients were assigned consecutively to two groups: The PCA group was treated by patient-controlled morphine and the block group was treated with combined sciatic (single-shot) and femoral (continuous for 48 h after surgery) peripheral nerve blocks. Femoral block was performed preoperatively according to the technique of Capdevila et al. ,4with an initial 20-ml bolus of 0.75% ropivacaine (Naropein®; Astra Zeneca, Rueil Malmaison, France) followed by a continuous infusion of 0.2% ropivacaine, 0.15 ml · kg⁻¹· h⁻¹for 48 h. The sciatic block was performed preoperatively according to the technique of Winnie with 20 ml ropivacaine, 0.75%.32The absence of sensory response to cold tested before surgery in the area of the femoral and sciatic nerves confirmed that the nerve blocks were effective. Patients failing this test were excluded.

Clinical evaluations were performed before surgery (D0) and at 1 (D1), 4 (D4), and 7 days (D7) and 1 (M1) and 3 months (M3) after surgery. Evaluations were performed on the mornings of D1, D4, and D7 and when the patient came back for follow-up surgical consultation at M1 and M3. During clinical evaluation, the patient was in the supine position in a quiet room, at a constant temperature (22°C), with the lower limb in neutral position. The circumference and skin temperature of the operated knee and contralateral knee were monitored at each follow-up visit. Skin temperature was measured in the center of the patella with a Thermopoint device (Protechnique, Quebec, Canada). The temperature was recorded as the mean of two successive measures. The knee circumference was measured by a thread, to the nearest half centimeter, along a horizontal line crossing the middle of the patella. All circumference and skin temperature measurements were performed by the same investigator (D.F.). Knee circumference and skin temperature were measured after surgery, after removal of all dressings. Pain was evaluated at rest and when moving (flexion/extension of the knee during physical therapy) by a 100-mm visual analog scale graduated from 0 (no pain) to 100 mm (worst imaginable pain). Pain scores and duration of preoperative pain were recorded the day before surgery. Pain scores were then monitored at D1, D4, D7, M1, and M3 and during all physical therapy sessions. Cumulative doses of PCA morphine were recorded at 24, 48, and 72 h. Physical therapy was started 24 h after surgery with passive and active mobilization of the operated knee. The active angle of flexion was recorded during hospitalization and at the follow-up visits at M1 and M3 by a physical therapist not involved in the study. Functional recovery was evaluated by recording the delay for first mobilization out of the bed and first use of the toilet with and without assistance.

Cytokine concentrations were measured in plasma before surgery (H0), four hours after completion of surgery (H4), and at day one (D1), day four (D4) and day seven (D7) after surgery. Cytokines concentrations in the capsule (C1 and C2 respectively) and synovial membrane (S1 and S2 respectively) of the operated knee were also measured at the beginning and at the end of surgery. We measured the proinflammatory cytokines, TNF-α and sTNF-R1, IL-1β and IL-6, and the antiinflammatory cytokine IL-10, in plasma; in tissue, we measured TNF-α, IL-1 β, IL-6, and IL-10. Cytokines were assayed using enzyme-linked immunosorbent assay kits (DuoSet®; R&D Systems, Lille, France). The lower limits for quantification were 15.6 pg/ml for TNF-α, 12.5 pg/ml for sTNF-R1, 3.9 pg/ml for IL-1β, 9.4 pg/ml for IL-6, and 31.2 pg/ml for IL-10.

The analysis used Statview 5.0 software (SAS, Paris, France). The primary outcome was systemic IL-6 concentration at D1. The secondary outcomes were knee circumference and temperature, tissue and plasma cytokine concentrations, pain scores, and functional scores. Clinical and biologic data were compared between groups and over time using two-way analysis of variance (ANOVA; one way for repeated measures). We used unpaired and paired t  tests with a Bonferroni adjustment for multiple comparisons. Data are expressed as mean ± SEM. P < 0.05 was considered statistically significant.

Forty patients were included in the study. Two patients in the PCA group were excluded immediately at the beginning of the study because of absence of a preoperative blood sample for cytokine concentration measurement. A total of 20 patients in the block group and 18 patients in the PCA group were analyzed. Patient characteristics were similar in both groups, with prolonged period of pain before surgery and intense pain when moving (table 1).

Morphine use was significantly lower in the block group than in the PCA group in the postanesthesia care unit (4 ± 1.8 vs.  21 ± 2 mg; P = 0.0001) and at D1 (12 ± 2.3 vs.  25 ± 2.7; P = 0.001). Morphine PCA use per day was also lower in the block group than in the PCA group on the second (18 ± 4 vs.  24 ± 4 mg; P = 0.43) and third postoperative days (10 ± 2.7 vs.  18 ± 6.5; P = 0.23), but it did not reach significance. Pain at rest was lower in the block group for D1 to D7 (ANOVA, group effect, P = 0.005; fig. 1A). Post hoc  analysis showed significant differences between the two groups at D4 (t  test, P = 0.01) and D7 (t  test, P = 0.002). Pain when moving was lower in the block group than in the PCA group for D1 to D7 (ANOVA, group effect, P = 0.0005; fig. 1B). Post hoc  analysis showed significant differences between the two groups at D1 (t  test, P = 0.003), D4 (t  test, P = 0.002), and D7 (t  test, P = 0.001). We did not observe a significant difference in pain score at M1 or M3 between the patient groups (figs. 1A and B).

Patients in the block group had a more rapid recovery of operative knee flexion, between D1 and M1, than PCA patients (ANOVA, group effect, P = 0.0001; fig. 2A). Post hoc  analysis showed significant differences between the two groups at D1 (t  test, P = 0.0001), D4 (t  test, P = 0.0004), D7 (t  test, P = 0.009), and M1 (t  test, P = 0.01). There was no difference in functional recovery between block and PCA groups at M3. Patient in the block group recovered more rapidly than PCA patients, as assessed by autonomy criteria for nonassisted mobilization (t  test, P = 0.04) and toilet use (t  test, P = 0.03; fig. 2B).

The postoperative circumference of the operated knee increased after surgery. This increase was significantly smaller for block group patients than for PCA group patients between D1 and D7 (ANOVA, interaction between group and time, P = 0.01; fig. 3). Post hoc  analysis showed significant differences at D1 (t  test, P = 0.02) and D7 (t  test, P = 0.02). There was no significant difference in postoperative knee circumference increase between the two groups at M1 or M3 (fig. 3).

Operative knee temperature increased after surgery. The increase was significantly smaller in the block group than in the PCA group between D1 and D7 (ANOVA, interaction between group and time, P = 0.03; fig. 4). Post hoc  analysis showed significant differences at D4 (t  test, P = 0.01) and D7 (t  test, P = 0.02). There were no significant differences between the two groups at M1 or M3 (fig. 4).

The skin temperature and circumference of the contralateral knee were stable throughout the study period (figs. 3 and 4).

Plasma concentrations of IL-1, TNF, and IL-10 were below the detection limit for almost all samples and therefore were not analyzed. There was a maximal increase of plasma IL-6 and sTNF-R1 concentration at D1 in both treatment groups, with this increase persisting until D7 for sTNF-R1 (fig. 5), but this increase was not significantly different between the block and PCA groups.

The overall tissue cytokine concentration increased during surgery. This increase was significant for IL-1 (72 ± 136%; paired t  test, P < 0.05) and IL-10 (128 ± 224%; paired t  test, P < 0.01) in the synovial membrane sample, but there was no significant difference between the two treatment groups (table 2).

This study demonstrates a clinical antiinflammatory effect of peripheral nerve blocks after surgery. This effect was supported by a reduction in both operative knee circumference and temperature but was not associated with a reduction of tissue or plasma proinflammatory cytokine concentration.

Continuous postoperative nerve blocks, regardless of catheter location, provide improved postoperative analgesia and fewer opioid-related side effects than opioid analgesia.1Our study suggests that this analgesic effect may be related to both reduced nociceptive inputs and a local antiinflammatory effect. Interestingly, our study is consistent with previous data showing the beneficial effects of postoperative nerve blocks on knee function.4Patients with peripheral nerve block had an improved postoperative knee flexion and more rapid recovery of functional autonomy. As previously observed,4this effect on function persisted after discontinuation of peripheral nerve block: Patients with nerve block had improved operative knee flexion 1 month after surgery. This persistent improvement of function cannot be explained by the pharmacologic effect of ropivacaine. It may be related to the prolonged combined analgesic and antiinflammatory effect of peripheral nerve blocks. In our study, the effect on postoperative pain, knee edema, and temperature was significant after discontinuation of postoperative nerve block (i.e. , after 2 days). Previous animal21or human22,33,34experiments suggest that nerve block may exert significant effects beyond its pharmacologic action on both nociception and inflammation. Our clinical findings extend these previous results by describing the prolonged effect of peripheral nerve block on nociception, inflammation, and postoperative functional recovery.

Peripheral nerve blocks reduced edema and temperature increase in our patients after surgery. This is the first study testing the effect of peripheral nerve block on clinical inflammation after surgery. One limitation of our study is caused by the fact that we could not perform a blinded study because of the sensory effects of the block and the visibility of the catheter during postoperative evaluation. As in a previous clinical study on knee surgery,35we used the combination of circumference and temperature measurements to evaluate clinical inflammation. Previous experimental studies in humans have used similar clinical criteria (i.e. , flare, erythema, temperature) to evaluate inflammation.22,23,26,28–30,36The precise mechanisms underlying the observed antiedematous effect of peripheral nerve block is unknown. First, in the absence of a change in markers of inflammation, one cannot exclude that the reduction of edema may have been due, at least in part, to other factors such as improved function and greater mobility of the operated leg. However, this antiedematous effect was combined with a reduction of skin temperature, suggesting an antiinflammatory effect. Second, the design of our study (i.e. , no systemic administration of ropivacaine) cannot rule out a systemic effect of ropivacaine on edema because we used high doses of ropivacaine and did not measure the plasma concentration of ropivacaine in patients. Systemic lidocaine reduces erythema and flare in acute inflammation models in humans,26–30whereas systemic bupivacaine does not reduce edema in carrageenan-injected rats20; however, the effect of systemic ropivacaine on edema has not been tested in humans or animals. Therefore, although a systemic effect of ropivacaine on edema cannot be ruled out, it seems unlikely. Third, peripheral nerve block may reduce inflammation. Contradictory findings have been obtained in human studies, with a positive effect of lidocaine nerve block in a model of superficial burn injury22but no effect observed for capsaicin-induced inflammation.23Contradictory findings have also been obtained in animal models, with decreased local edema in carrageenan-injected rats after sciatic nerve infusion of bupivacaine18or ipsilateral and contralateral bupivacaine sciatic nerve block20and no effect observed after a tonicaine nerve block.21Overall, these findings from human and animal studies suggest that local anesthetic can reduce peripheral inflammation through a sodium channel–dependent segmental effect; we consider this the most likely mechanism of the antiedematous effect of peripheral nerve blocks observed in our study.

The dissociation between the clinical and biologic markers of inflammation observed in our study also deserves discussion. Peripheral nerve blocks had no detectable effect on the increase of plasma cytokine levels. Our findings are in line with a recent study using a combined continuous lumbar plexus and sciatic nerve blocks with ropivacaine after knee surgery. The authors did not detect a change in IL-6 plasma level associated with a blunted increase in C-reactive protein.31This absence of a detectable effect on plasma IL-6 levels by peripheral nerve blocks may be explained by limited sensitivity of enzyme-linked immunosorbent assay to measure plasma cytokine concentration, as compared with an ex vivo  cytokine assay. However, this technique has been used in previous studies to detect changes in cytokine profiles of surgical patients37or patients with fibromyalgia and painful neuropathy.38,39Alternatively, it is possible that plasma cytokine levels do not reflect the clinical antiinflammatory effect observed on the operated knee. Indeed, this clinical antiinflammatory effect of peripheral nerve block is probably related to inhibition of neurogenic inflammation with reduced peripheral release of substances such as substance P and calcitonin gene–related peptide. The effect on neurogenic inflammation may be independent of plasma cytokine levels. It is also possible that tissue damage and inflammation related to total knee arthroplasty may overcome the effects of nerve block on inflammatory response. Other clinical studies using more extensive nerve blocks, such as epidural analgesia, have demonstrated attenuated proinflammatory cytokine production after visceral surgery.13,37,40Therefore, the effects of nerve blocks on biologic indicators of inflammation seem to depend on several factors, including the type of marker (i.e. , C-reactive protein or IL-6), the technique of regional anesthesia (i.e. , peripheral nerve block or epidural analgesia), the assay used to measure of cytokine concentration (i.e. , in vivo  or ex vivo  assays), and probably the type of surgery.

We did not observe any effect of peripheral nerve block on tissue cytokine concentration, consistent with the absence of significant effects on plasma cytokine levels in patients receiving peripheral nerve block. Cytokine levels in operative area have not been previously studied. Our study demonstrated an overall increase in tissue cytokine concentration during surgery; we detected a significant increase in IL-10 and IL-1 levels in the synovial membrane. We did not find a correlation between preoperative pain or postoperative pain intensity and tissue cytokine concentration.

In conclusion, our study suggests that peripheral nerve block used for postoperative analgesia also exerts a prolonged antiinflammatory effect; the combination of these effects probably participates in the observed improved functional recovery.