Local infiltration analgesia in knee and hip arthroplasty efficacy and safety

1.1 Knee and hip arthroplasty

Total knee arthroplasty (TKA) and total hip arthroplasty (THA) are highly successful surgical procedures which alleviate pain, correct leg deformity and help patients with debilitating arthritis to restore function and resume normal activities [1].

However, postoperative pain is common after total joint arthroplasty and effective pain relief may be hard to achieve [2,3]. Inadequate pain control increases the length of hospital stay, escalates the cost of care, and is associated with venous thrombo embolism, coronary ischaemia, myocardial infarction, pneumonia, insomnia, cognitive dysfunction, poor wound healing, and slowed overall recovery [4]. The most effective pain treatment after knee and hip arthroplasty has traditionally been opioid analgesia and epidural analgesia. Each of this method has its limitation.

1.2 Femoral nerve block (FNB)

Superior analgesia after TKA with fewer side-effects than intravenous opioids or epidural analgesia [5,6] can be obtained by femoral nerve block, a technically demanding method of pain management that requires additional anaesthetic time [7]. However, a disadvantage is the decreased muscle tone of the quadriceps muscles, which counteracts effective rehabilitation and increases the risk of patient falls [8,9].

An alternative method for control of acute postoperative pain following knee and hip replacement surgery is local infiltration analgesia (LIA)

1.3 Local infiltration analgesia (LIA)

Local infiltration analgesia (LIA) was first described by Kerr and Kohan in Sydney, Australia [10]. In contrast to epidural analgesia and peripheral nerve blocks, multimodal infiltration is cheap and requires only limited technical skills.

A mixture of ropivacaine a long-acting amide local anaesthetic agent, ketorolac a non-steroidal anti-inflammatory drug (NSAID), and adrenaline is infiltrated around all the structures subjected to surgical trauma. LIA reduces postoperative pain at its origin without loss of muscle strength, decreases opioid consumption, improves lower-limb function, decreases complications, operating room time and shortens hospital stay [11,12,13].

Some studies have compared LIA after TKA with other methods of postoperative pain treatment, such as systemic analgesia and placebo [14,15]. Limited data was available to compare LIA with femoral block [16]. Information regarding the safety margins and the plasma concentration of ropivacaine after LIA was incomplete [17,18,19] and no information on plasma concentration of ketorolac after local infiltration in LIA was available.

1.4 Aims

The aims were to assess

1. whether pain relief after LIA in knee arthroplasty is as effective as femoral block,

2. whether the plasma concentration of ropivacaine after LIA in knee arthroplasty is higher than after femoral block,

3. whether the plasma concentration of ketorolac after LIA in knee arthroplasty reaches levels linked to toxicity,

4. whether the maximal dose adjusted concentration of ropivacaine after LIA is higher in THA as compared to TKA,

5. whether the plasma concentration of unbound ropivacaine after LIA in hip arthroplasty reaches levels linked to toxicity,

6. whether an increase in AAG after hip arthroplasty decreases the unbound concentration of ropivacaine,

7. whether the plasma concentration of ketorolac after LIA in hip arthroplasty reaches levels linked to toxicity,

8. whether local administration of ketorolac in LIA may be safer as compared to the intramuscular route.

3.1 Pain relief

Both LIA and femoral block methods resulted in a high quality pain relief with a low NRS pain score both at rest (Fig. 1) and upon movement (Fig. 2). The patient controlled demand for morphine via a PCA-pump was slightly lower in the LIA group. Corrected for patient weight no difference could be detected. [20].

Fig. 1

Average pain score (NRS) at rest during 24 h after surgery. No data recorded for sleeping patients. (n = 20 in each group).

Fig. 2

Average pain score (NRS) upon movement during 24 h after surgery. No data recorded for sleeping patients. (n = 20 in each group).

3.2 Ropivacaine plasma concentration after total knee arthroplasty

The maximal total plasma concentration of ropivacaine was below the established toxic threshold after LIA and femoral block for most patients although a few in the LIA group reached potentially toxic concentrations of 1.4–1.7 mg/L (Figs. 3 and 4). The time to maximal detected plasma concentration was around 4–6 h after release of tourniquet in the LIA group (Fig. 3) [21].

Fig. 3

Total plasma concentration of ropivacaine during 24 h. (A) LIA group (n = 20). In this group zero “0” refers to release of the tourniquet. The LIA group received 300 mg ropivacaine. (B) Femoral block group (n = 19). In this group time zero “0” refers to completion of the first injection for femoral block (60 mg), the subsequent doses (30 mg every 4 h) are indicated in the figure.

Fig. 4

Maximal detected concentration of ropivacaine during 24 h. Data expressed as maximal concentration for each patient. The line indicates the median.

All patients in the THA cohort were subjected to the routine LIA protocol. In these patients both the total and unbound plasma concentration of ropivacaine was determined (Fig. 5). The concentration was below the established toxic threshold. As ropivacaine binds to α–1 acid glycoprotein (AAG) we monitored AAG. During the first 24 h after surgery a 40% increase in AAG was detected (Fig. 6). However, the fraction of unbound ropivacaine remained the same. There was a trend towards increased C_max of ropivacaine with increasing age and decreasing creatinine clearance but the statistical power was too low to draw any conclusion [22].

Fig. 5

Individual data of unbound (A) and total (B) plasma concentrations of ropivacaine versus time during 30 h after LIA.

Fig. 6

Individual plasma concentrations of α–1 acid glycoprotein versus time in 14 patients. Time zero “0” indicates a baseline sample priorto surgery.

3.3 Ketorolac plasma concentration

Administration of 30 mg ketorolac according to the LIA protocol both in TKA and THA resulted in a similar C_max as previously reported after 10 mg intramuscular ketorolac in most patients (Figs. 7 and 8). Neither age, nor body weight or BMI, nor creatinine clearance, correlated to maximal ketorolac plasma concentration or total exposure to ketorolac (AUC) (data not shown here) [23].

Fig. 7

Total plasma concentration of ketorolac in the LIA group in TKA, which received 30 mg ketorolac. In this group zero “0” refers to release of the tourniquet.

Fig. 8

Individual total plasma concentrations of ketorolac expressed as mg/L (left Y-axis) and as ng/ml (right Y-axis) versus time during 30 h after injection of 30 mg in LIA (n = 13) in THA.

4.1 Pain intensity and morphine consumption after total knee and hip arthroplasty with LIA

Both LIA and femoral block resulted in low average pain intensity during the first post-operative day after TKA, which is in line with a recent randomized study [24].

However, only 1/20 (5%) of patients in the LIA group reported pain intensity greater than 7/10 on NRS upon movement as compared to 7/19 (37%) after femoral block. No difference between LIA and femoral block was observed with regard to pain at rest. These findings are in line with the observation that differences in efficacy between treatment modalities may appear only when pain is assessed during function, not at rest [25]. Low pain intensity upon movement may be crucial for rapid rehabilitation, particularly after joint replacement.

Femoral nerve block does not cover the posterior part of the knee, which is innervated by the sciatic nerve. Some authors recommend a supplementary sciatic nerve block to achieve better pain relief after TKA [26,27]. However, sciatic nerve block may weaken muscles needed for mobilization after surgery [28].

In addition, femoral nerve block usually produces a partial motor block of the quadriceps femoris muscle, which could delay early postoperative mobilization. This effect is avoided with LIA [11,24,29].

With regard to the analgesic efficacy of LIA in total hip arthroplasty (THA) we did not compare this to any other method of post-operative pain relief. However, we asked all our patients for a general. Eighty-six percent of the patients with THA assessed their satisfaction with the postoperative pain management as reported good or excellent after LIA.

Our data do not permit to draw a conclusion on the exact contribution of ketorolac to analgesia induced by LIA. However, we have observed less efficient postoperative analgesia in patients who received local infiltration analgesia without ketorolac in the mixture. In addition, ketorolac is more effective when given intra-articularly and in LIA compared to other routes of administration [30,31].

4.2 Ropivacaine plasma concentration after LIA in TKA and THA

The maximal ropivacaine concentrations observed in knee arthroplasty were below the established toxic threshold for most of our patients although two individuals reached concentrations of 1.4–1.7 µg/mL.

Similar concentrations have been linked to signs and symptoms of toxicity after i.v. administration in healthy volunteers [32,33]. In addition, the maximal plasma concentration of ropivacaine using the LIA protocol seems to be higher than after femoral block during the first 24 h. however, studies on peripheral and central block have reported even higher plasma concentrations (2–4.2 µg/mL) without adverse reactions [34,35].

Very few investigators have studied safety aspects of LIA or assessed plasma concentrations of ropivacaine after LIA in TKA and THA. Bianconi et al. [19] reported total plasma concentrations of ropivacaine in patients subjected to elective hip/knee arthroplasty. The range of maximum plasma concentration (C_max) was 0.30–1.28 µg/mL, which is comparable to our data. However, the dose used in that study was 200 mg ropivacaine plus a continuous infusion of 10 mg/h for 55 h. The maximal concentration was detected 24 h after clearance, nor age, nor body weight correlated with the AUC of unbound ropivacaine. Renal function may be of greater importance for the AUC during continuous infusion than after a single infiltration as used in our protocol.

4.3 Safety aspects of ropivacaine

Knowledge of the potential risks of cardiac or central-nervous-system side-effects at different concentrations of ropivacaine is based on data obtained from early studies with healthy volunteers who received intravenous infusion of ropivacaine [33].

Side-effects sufficient to stop the intravenous infusion were reported at arterial concentrations of 0.34–0.85 µg/mL. This range has been considered to represent a relevant safety limit or neurological toxicity range for venous plasma concentration of unbound ropivacaine [36]. However, the clinical relevance of this range may be questioned. Considerably higher unbound concentrations without adverse reactions have been reported during epidural or local infusion [37,38].

We did not detect any signs or symptoms of ropivacaine toxicity after LIA in the patients included in the present studies. A potential risk of local anaesthetic toxicity during arthroscopic knee surgery is illustrated by case reports of healthy patients subjected to synovial surgery with local administration of bupivacaine (75 mg and 150 mg) [39,40].

Regarding the effect of tourniquet use, a previous study suggests that a longer duration of tourniquet ischaemia may lead to a faster absorption of local anaesthetics and higher peak plasma level due to enhanced post-ischaemic reperfusion. In contrast, the longer duration of tourniquet inflation after local anaesthetics injection increases tissue binding, and decreases peak serum levels [10,41].

Regarding THA, the tourniquet does not apply but the surgical wound is large and the possibility of absorption of local anaesthetics is greater. A lower dose of ropivacaine (200 mg instead of 300 mg) may decrease the risk of high plasma concentration. The maximal total concentration after LIA with 300 mg in TKA was 0.81 µg/mL, and after LIA with 200 mg in THA was 0.78 µg/mL. Thus, the absorption of ropivacaine is greater after LIA in THA than in TKA.

4.4 Ropivacaine binding to α–1 acid glycoprotein (AAG)

Ropivacaine binds mainly to AAG [42,43]. This protein increases in stress conditions like surgical trauma. We detected AAG levels similar to those in young healthy adults [44]. After 24 h the AAG level in our study had increased by less than 40%, which did not result in any significant change in the unbound concentration of ropivacaine. AAG increases after 24 h, as shown in several studies of prolonged infusion of ropivacaine [37,45,46]. AAG levels may double around 4 days postoperatively and seem to reach a maximal concentration at the sixth to twelfth postoperative day [47,48].

4.5 Ketorolac

In knee arthroplasty the range of the maximal detected plasma concentration of ketorolac was 0.15–0.96 mg/L. In hip arthroplasty the maximum plasma concentration of ketorolac after LIA was 0.82 mg/L (0.31–2.16). This is comparable to the maximal plasma concentrations after 10 mg ketorolac given intramuscularly in healthy volunteers 0.77 mg/L [49].

We could not find any correlation between peak concentration or C_max and the patient age within our cohort. These results are line with the reported similar range after an intramuscular injection of 30 mg ketorolac in young adults (mean age 30 years) and healthy elderly (mean age 72 years) [49].

We could not demonstrate any effect of creatinine clearance on the peak concentration of ketorolac within the range present in our cohort 58–150 ml/min. Renal function is more important for total exposure or AUC than the peak concentration after a single dose. However, we could not find any correlation between creatinine clearance and AUC either. A tendency towards higher AUC after intramuscular injection of 30 mg ketorolac to elderly as compared to younger adults has been reported [49]. We could presume that individuals with creatinine clearance lower than 50 ml/min may have higher AUC. But due to safety concern of ketorolac in patients with reduced renal function a clinical trial on this issue may be ethically questionable.

Our data on ketorolac after LIA do not seem to help identify patients with a higher risk of potential adverse events. Instead it seems reasonable to avoid ketorolac in patients with congestive heart failure treated with ACE inhibitors or ARB and in patients with low creatinine clearance. The optimal cutoff level of creatinine clearance remains to be established.

4.6 Overall efficacy and safety of LIA

Systemic toxicity from local anaesthetics is relatively rare. However, local anaesthetic toxicity can be catastrophic to the individual when it does occur.

Although many anesthesiologists may occasionally see mild manifestations, most never encounter serious intoxication.

In most of our patients the plasma concentration of ropivacaine after LIA in knee and hip arthroplasty did not reach levels linked to toxicity. However, patients undergoing knee and hip arthroplasty are usually old with various medical diseases. Slow incremental and frequent aspiration during LIA mixture infiltration is advisable. Lipid emulsion should be available for use in case of toxicity, since this therapy is effective for treating local anaesthetic toxicity [50]. However, the mechanism of the reversal of toxic effects of local anaesthetics by lipid emulsion is still unclear. One theory is that it creates a lipid plasma phase that essentially extracts the high lipid-soluble local anaesthetic molecules from the aqueous plasma phase [51,52,53].

An important question still not answered is how much local anaesthetic is required in the LIA mixture to produce the optimal therapeutic effect.

The manufacturer Astra Zeneca recommends a maximal dose of 225 mg. Higher doses of ropivacaine are used in various institutions both in knee and hip arthroplasty. It may be difficult to recommend a safe maximal dose of local anaesthetics because individuals vary in their sensitivity to local anaesthetics toxicity, as has already been observed in a ropivacaine toxicity study in healthy volunteers [33]. The correlation between blood levels and signs of toxicity is considered multifactorial as physiological, anatomical and pharmacokinetic factors all contribute [54].

Although we could not find any previously reported case of ropivacaine toxicity after LIA in knee and hip arthroplasty, this does not mean that the possibility of its occurrence is negligible, especially in severely ill patients with renal and hepatic impairment.

It seems that single doses may carry a lower risk of toxicity than continued infusion; but scientific proof for this assumption is at least weak.

Older patients may have a higher peak of unbound ropivacaine than younger ones. We found a trend towards a correlation between age and unbound maximal ropivacaine, at least 29 individuals are needed to get sufficient power based on our data. It may be advisable to use lower doses of ropivacaine inpatients at higher risk.

However, our data are insufficient to provide an exact dose recommendation for these patients. Intuitively, the “one size fits all” approach using the same dose of ropivacaine for a patient weighing a hundred kilos and one weighing fifty may seem inappropriate. However, our data do not indicate that dosage-per-kilo carries a lower risk of peak concentrations of ropivacaine.

Ketorolac plasma concentration after LIA infiltration in both knee and hip arthroplasty is not negligible, and the risk of renal side-effects should be kept in mind. Age 80 years or older is an independent risk factor for NSAID nephrotoxicity, since 50% of 80-year-old patients have already lost half their glomerular filtration rate [5]. Patients with congestive heart failure, hepatic cirrhosis, hypovolemia or underlying renal disease are more susceptible to ketorolac-induced nephrotoxicity [55,56]. Heart failure is increasingly diagnosed in the elderly and 30–50% of these patients with heart failure suffer from some degree of renal insufficiency, making their kidneys even more vulnerable to renal adverse events.

Increasing life expectancy, with a growing geriatric population, produces a new cohort of elderly surgical candidates extremely vulnerable to potential nephrotoxic effects of combinations of drugs, in particular in clinical conditions where renal perfusion is reduced.

In our department we had two cases of renal failure after hip and knee arthroplasty after local infiltration analgesia with ketorolac. Both patients were elderly (80 and 89 years) with concomitant treatment with ACE inhibitors or ARB. One required renal dialysis treatment [57]. Based on these observations we now avoid ketorolac in LIA in patients with renal impairment treated with ACE inhibitors or ARB.