A large European survey mentions a 19 % incidence of chronic pain in the population, 3 % of the individuals reporting surgery as the initial cause of their chronic pain (Breivik et al. 2006). For 22 % of the patients who consult in pain clinics, surgery is the origin of their chronic pain (Crombie et al. 1998). Joint arthroplasties are highly successful when judged by prosthesis-related outcomes (clinical evaluation and radiographic appearance), and several outcome studies have reported improved function and health-related quality of life. However, one in five patients (i.e. 19 %) undergoing primary TKA is not satisfied with the outcome (Bourne et al. 2010), and chronic postsurgical pain appears to be the primary predictor of dissatisfaction (Scott et al. 2010).

Chronic postsurgical pain (CPSP) has been defined by the International Association for the Study of Pain (IASP) as pain that develops after a surgical intervention and that lasts at least 2 months, with other causes for the pain having been excluded (e.g. infection, recurrence of malignancy) as well as pain from a condition preceding the surgery (Merskey 1994). The time frame of 2 months has been strongly debated because the exact duration of postoperative inflammatory processes still remains undetermined. Actually, one may consider CPSP as pain lasting more than 3–6 months (often beyond 6 months) after surgery. CPSP is a prevalent problem involving all types of surgical procedures including minor ones. The incidence varies between 10 and 50 % according to the type of procedure, but more worrisome is the fact that 2–10 % of these patients present with moderate-to-severe pain, which strongly affects their daily quality of life (Kehlet et al. 2006). Johansen et al. reported a prevalence between 6.2 and 18.3 % of moderate-to-severe pain in the area of surgery 3–36 months after the procedure (Johansen et al. 2012). While most patients usually recover and experience pain relief within 3 months after TKA (Vilardo and Shah 2011), about 20 % (10–34 %) of the patients are left with an unfavourable long-term pain outcome according to a recent systematic review (Table 13.1) (Beswick et al. 2012).

The large variability in reported incidences of persistent pain after TKA can be explained by the various definitions used for CPSP. Pain is often reported as an element of functional knee scores (i.e. WOMAC, KOOS) instead of using specific chronic pain questionnaires (i.e. McGill Pain Questionnaire, Brief Pain Inventory). It is also important to consider that pain identified by functional scores, reported during specific activities, may not totally reflect CPSP. Intermittent and intense pain at rest or during the night may also exist and usually has a major negative impact on the daily quality of life (Beswick et al. 2012).

The nature of CPSP remains unclear because inflammatory mechanisms play an important role even in the development of neuropathic pain. Iatrogenic neuropathic pain caused by incision and nerve injury is thought to be the most common cause of CPSP (Kehlet et al. 2006). The average prevalence of neuropathic pain after surgery is around 24 % with a large range from 6 % to more than 50 % in thoracic procedures (e.g. breast surgery and thoracotomy) (Table 13.1). Chronic pain of neuropathic origin is usually associated with higher pain intensity (average visual analogue scale (VAS) score around 7–8/10) than non-neuropathic chronic pain (Torrance et al. 2006). Johansen et al. observed at 4 months to 3 years after knee surgery that 66 % of patients self-reported knee pain was associated with sensory disturbances, 22 % with hypoesthesia and 12 % with hyperesthesia in an area at or close to the incision. The presence of hyperalgesia close to the scar is associated with an increased risk of severe CPSP (odds ratio 6.3). Due to the lack of use of adequate questionnaires, the nature of CPSP after TKA has been rarely questioned. It seems that the incidence of CPSP of a neuropathic origin is rare after TKA, 6 % at 1 year and later (Haroutiunian et al. 2013; Liu et al. 2012; Wylde et al. 2011b). The causes of nerve injury after TKA do not only involve direct surgical trauma of the infrapatellar branch of the saphenous nerve (84 %) or more exceptional the peroneal nerve (Henningsen et al. 2013) but may also be caused by the tourniquet during the procedure or peripheral nerve blocks used for perioperative analgesia (Kinghorn et al. 2012). Not all nerve lesions will cause pain, and CPSP associated with nerve injury will only develop in predisposed individuals (Lavand’homme 2011). Reflex sympathetic dystrophy (RSD), also called complex regional pain syndrome type 1 (CRPS-1), is a pain disorder associated with autonomic dysfunction and characterised by a disproportionate level of pain. CRPS-1 seems to occur more frequently after trauma or surgery of the lower limb. Harden et al. highlighted that “CRPS-like” phenomena, which diagnostic criteria include abnormal sensory modalities like hyperalgesia and allodynia, may affect up to 13–20 % of the patients during the first year after TKA. CRPS-like phenomena, in the early postoperative period, might also be the result of delayed postsurgical healing and persistent inflammatory reaction that would not normally merit the diagnosis of CRPS (Harden et al. 2003). Experimental studies support that hypothesis as focal nerve inflammation induces neuronal signs consistent with symptoms of early CRPS (Bove 2009). Burns et al. reviewed the medical records of 1,280 patients who underwent TKA for osteoarthritis and found a very low incidence of CRPS, as low as 0.7 %. More importantly, when managed early, patients complicated with CRPS after TKA have a similar prognosis to patients with uncomplicated TKA (Burns et al. 2006).

Today, it is still unclear whether a distinct transition period exists between acute and chronic pain after surgery or trauma. Subacute pain, which can last for several weeks after the surgery, is now recognised as a neglected area of clinical investigation. Poorly relieved subacute pain after surgery not only has a negative psychological impact but also might contribute to maintain a state of central sensitisation, which in turn might facilitate the persistence of pain (Lavand’homme 2011). According to Andersen et al., 52 % of patients report moderate pain, and 16 % report severe pain at rest 30 days after TKA, while pain when moving affects as much as 78 % of the patients (Table 13.1). Using the recent concept of pain trajectories (Chapman et al. 2011), Morze et al. have examined the weekly resolution of knee pain during the first 3 months after TKA (Morze et al. 2013). Clearly, all patients are not equal in facing pain and recovery, and 25 % of the patients still have worst average pain score of 4.9/10 by week 12 after surgery. The overall time taken to reduce worst pain was 6 weeks in 52 % of the patients but could be as long as 12 weeks for 32 % of the patients.

The unexplained painful TKA with no obvious cause remains a challenge for the surgeon (Hofmann et al. 2011). Surgical exploration is rarely advised, and only 45 % of the patients have problems related to their implant (Mont et al. 1996). Revision TKA for unexplained knee pain might harm even more. Between 4 and 22 months after surgery, 38 % of patients suffer from daily life disturbing pain after revision arthroplasty, and 40 % of these patients use analgesics (Puolakka et al. 2010). At 2 and 5 years after a TKA revision, pain is still reported three times more frequently than after a primary arthroplasty (Singh et al. 2008).

However, it is encouraging to see that the incidence of CPSP after TKA seems to be falling down since the 1990s, when 22 % of patients experienced pain at 7 years (Murray and Frost 1998) and as much as 51 % at 1 year (Dickstein et al. 1998). Such evolution underlines the progresses that have been realised in perioperative care, both in surgical and anaesthetic management, of these patients. Further improvements are still needed, mainly based on individualised and tailored management thanks to targeting patients at risk and promoting effective perioperative treatments.

Predisposition to chronic pain is multifactorial and includes the severity of postoperative pain, which is the most striking risk factor, pre-existing pain and psychological factors such as catastrophising and hypervigilance to pain (Kehlet et al. 2006). Predictive factors of CPSP do not really differ from those involved in the risk of severe acute postoperative pain.

Early severe postoperative pain is an important predictor for CPSP after TKA (Puolakka et al. 2010). If the degree of pain during the first postoperative week ranges from moderate to severe, the risk to develop persistent pain is 3–10 times higher compared with patients complaining of mild pain during the same period (Puolakka et al. 2010). Not only poorly relieved postoperative pain contributes to central nervous system sensitisation, but it also has a negative impact on the psychological aspects of recovery in predisposed individuals.

Preoperative pain, either at the operative site or elsewhere, is a known risk factor (Liu et al. 2012; Puolakka et al. 2010; Wylde et al. 2011b) for both severe postoperative pain and CPSP. Current research is ongoing to assess and to target by tailored treatments the endogenous pain modulatory processes.

Negative psychosocial conditions like preoperative anxiety and catastrophisation are regularly found in the history of patients suffering from CPSP (Kehlet et al. 2006). The impact of preoperative psychological factors on the persistence of pain after surgery or trauma has received a growing attention (Lautenbacher et al. 2010). A recent systematic review has demonstrated that, effectively, low preoperative mental health and pain catastrophising influence the outcome after TKA in terms of function and pain scores (Vissers et al. 2012). Surprisingly, those factors have shown only limited impact on the outcome after total hip arthroplasty. Recent results support the fact that high, if not unrealistic, expectations of TKA are common and should be moderated to maintain patient satisfaction (Hepinstall et al. 2011). From a clinical point of view, patients with a low preoperative pain score (assessed by WOMAC score) are 2.4 times more likely to be dissatisfied with their operation (Bourne et al. 2010). Morze et al. found that the rate of decline of worst postsurgical pain was slower in patients who reported lower preoperative pain scores (6 % per week versus 13 %) (Morze et al. 2013).

Other common risk factors to develop CPSP are younger age and female gender (Kehlet et al. 2006). Those factors have been incriminated in several large retrospective studies from 1 to 5 years after TKA (Liu et al. 2012; Singh et al. 2008). Female sex as predictive factor for CPSP after TKA might be supported by the fact that women usually wait much longer than men before having surgery, despite greater reported disability (Petterson et al. 2007). Waiting longer before having surgery signifies living longer with the aggravating preoperative pain, which is another known predictive factor of CPSP after TKA (Puolakka et al. 2010; Singh et al. 2008).

Finally, the genetic background of the patient also affects the sensitivity to pain as well as the metabolism of analgesic drugs. Some protective phenotypes have been described (France et al. 2009) but at the present, no genotype screening to identify populations at risk is available.

Whereas numerous studies have assessed the incidence and the risk factors for persistent pain after TKA, very few reports have investigated the real consequences of chronic postsurgical pain in these patients, particularly the impact of long-term analgesic use (Steyaert and Lavand’homme 2013). According to published clinical reports, 56 % of patients were still under analgesics at 30 days after TKA (Andersen et al. 2009), 40 % of patients after 4 months (Puolakka et al. 2010) and around 25 % of patients about 2 years later (Carroll et al. 2012).

Nonsteroidal anti-inflammatory drugs (NSAIDs) are very effective in alleviating pain after orthopaedic procedures. Alam et al. (2012) found that elderly patients who started taking NSAIDs within 7 days of surgery were almost four times more likely to become long-term NSAIDs users. Aside from the risks of gastrointestinal ulcers and bleeding or renal insufficiency, chronic use of NSAIDs carries a non-negligible cardiovascular risk for stroke and myocardial infarction. Recently, the American Geriatric Society has recommended the use of opioid analgesics for elderly patients (Kuehn 2009) instead of NSAIDs. Opioids are not innocent either and there have been some alarming reports in the last years. In a propensity-matched cohort study, Solomon et al. (2010) compared the safety of opioids and NSAIDs in more than 12,000 elderly patients with arthritis where patients on opioids had an increased relative risk for many safety events (fractures, cardiovascular events, hospitalisations) compared with NSAIDs. It is mandatory to keep in mind that the initiation of a short-term opioid therapy may lead to longer-term use after discharge in some patients. Elderly patients receiving an opioid prescription after surgery seem to be 44 % more likely to become long-term opioid users compared to patients who did not receive opioids (Alam et al. 2012). Andersen et al. highlighted the fact that 56 % of patients are still on opioids, including 36 % taking strong opioids, 30 days after TKA (Andersen et al. 2009). Carroll et al. found that 6 % of the patients started taking new opioids more than 150 days after surgery and about a quarter of patients, having undergone total hip or knee arthroplasty, were still taking opioids at the end of the follow-up period 2 years later (Carroll et al. 2012). Interestingly, postoperative pain duration and severity only account for 48 % variance in that long-lasting opioid intake, while preoperative factors like legitimate prescribed opioid use, self-perceived risk of addiction and depressive symptoms are better predictors of prolonged opioid use (Carroll et al. 2012). It is mandatory that caregivers involved in the management of patients undergoing TKA take their responsibilities for the prescription of analgesics in these patients: better identification of the patients who are at risk of needing prolonged postoperative opioids, adaptation of perioperative treatments to reduce the need for opioids to a minimum (preventive analgesia, i.e. multimodal analgesia with regional analgesia and anti-hyperalgesics) and closer follow-up of patients after their discharge (Steyaert and Lavand’homme 2013).

Minimally invasive (MI) techniques gained popularity in the last decade. Minimally invasive TKA is a technique that focuses on using a smaller skin incision, avoiding patella eversion and quadriceps-sparing approaches (Arnout et al. 2009; Khanna et al. 2009; Thienpont 2013a, b). Less tissue damage should reduce the release of inflammatory mediators and its cascade. The length of the skin incision in TKA is proportional with the surface area of sensory change in the front of the knee, which is related to the inability to kneel following surgery (Hassaballa et al. 2012). MI-TKA leads to reduced length of stay (LOS), reduced pain and more rapid rehabilitation (Niki et al. 2009). Although the learning curve of such new techniques seems to be substantial, the benefits in terms of need for opioids and need for rehabilitation and LOS are significant (King et al. 2007). The reduction of tissue damage, the decrease in tourniquet time as well as the reduction of the duration of the surgical procedure may also play a role in postoperative recovery (Niki et al. 2009). Despite that tourniquet use does not have a direct impact on knee pain and analgesics use, tourniquet however induces thigh pain (Wakankar et al. 1999). Quadriceps muscle strength is the parameter, which correlates the best with postoperative functional performance (Khanna et al. 2009). MI-TKA drastically reduces the direct quadriceps damage. In opposite, tourniquet-induced ischaemia represents a substantial contributor to intraoperative muscle damage (Appell et al. 1993). A recent systematic review of the literature showed that as a group, patients undergoing MI-TKA tend to have decreased postoperative pain, rapid recovery of quadriceps function, improved range of motion and shorter LOS (Khanna et al. 2009).

In order to facilitate early discharge from the hospital and more rapid return to the normal daily activities, the concept of fast-track surgery has been developed. Fast-track surgery integrates new modalities in surgery, anaesthesia and nutrition, enforcing early mobilisation and rehabilitation. Along with the use of minimally invasive techniques, the choice of anaesthesia and analgesia techniques has to be adapted, a choice which relies on multimodal balanced techniques allowing a subsequent opioid-sparing effect (White et al. 2007).

The aim of multimodal analgesia is to minimise the need for opioid analgesics (Kehlet and Dahl 1993). Although opioids remain very effective analgesics, their well-known side effects like sedation, nausea and vomiting and respiratory depression may prevent rapid rehabilitation. Furthermore, opioids may also induce hyperalgesia (also called “opioid-induced hyperalgesia”) and acute tolerance that enhances postoperative pain.

Nonsteroidal anti-inflammatory drugs (NSAIDs) and acetaminophen (paracetamol) are commonly associated drug combinations used for multimodal analgesia. Recent meta-analyses have shown that the significant 30–50 % opioid dose-sparing effect of non-selective NSAIDs is associated with a 30 % reduction in opioid-related side effects like postoperative nausea and vomiting and sedation (Elia et al. 2005; Marret et al. 2005). Selective inhibitors of cyclooxygenase 2 (COX-2) also display a 35 % opioid-sparing effect (Straube et al. 2005). However, non-selective and selective NSAIDs are contraindicated in one patient out of five because of the risks of gastrointestinal bleeding or renal insufficiency. For those patients, paracetamol then represents an alternative. Paracetamol has complex analgesic effects, which rely on central endocannabinoid and serotoninergic mechanisms and perhaps on a COX-3 inhibition (Mallet et al. 2008). By itself it demonstrates only a weak opioid-sparing effect of around 20 % not associated with a reduction of opioids’ side effects (Remy et al. 2005).

Corticosteroids represent the “ultimate anti-inflammatory drugs” (Turan and Sessler 2011) and are used in rheumatic disease for their local anti-inflammatory properties. In the perioperative setting, they are often administered for their anti-emetic effect as a preventive strategy for postoperative nausea and vomiting. A recent meta-analysis has demonstrated that a small dose of perioperative dexamethasone (0.1–0.2 mg/kg) reduced postoperative pain at rest and during mobilisation up to 24 h postoperatively, as well as opioid consumption (Waldron et al. 2013). Corticosteroids have previously proven their efficiency in reducing postoperative pain after spine surgery, hip replacement and TKA. A high dose of 40 mg dexamethasone before anaesthesia significantly prevented dynamic pain until 24 h after elective hip replacement (Kardash et al. 2008). In patients undergoing elective TKA, preoperative high dose of methylprednisolone 125 mg significantly reduced pain during walking up to 32 h after surgery (Lunn et al. 2011). The postoperative opioid requirements were also reduced as well as nausea and vomiting. Postoperatively, inflammatory parameters like CRP concentrations were lower. These patients reported less postoperative fatigue although the sleep quality was worse on the first night. Other authors have assessed the modulatory effect of a low dose of hydrocortisone (two doses of 100 mg given at 8 h apart) in patients undergoing bilateral TKA (Jules-Elysee et al. 2012, 2011). The decreased inflammatory response as measured by 40 % reduction in IL-6 production was associated with higher postoperative haemodynamic stability, lowered pain scores and improved knee motion. It is interesting to note that administration of corticosteroids 1 day after surgery may help to control pain even better than regular NSAIDs. Romundstad et al. have assessed the time course and magnitude of the analgesic effect of 125 mg methylprednisolone in patients with moderate-to-severe pain 1 day after orthopaedic surgery (Romundstad et al. 2004). Methylprednisolone analgesic effect was as effective as the effect of 30 mg ketorolac but lasted longer and provided higher opioid-sparing effect up to 72 h after administration. Very few studies have evaluated the benefit of combining NSAIDs with corticosteroids to improve postoperative pain. It seems that such combination may demonstrate superior analgesic effect as reported after knee arthroscopic surgery in patients receiving a low dose of 8 mg dexamethasone and a COX-2 selective inhibitor (Dahl et al. 2012). Finally, experimental data found a preventive effect of intraoperative systemic corticosteroids on the development of persistent neuropathic pain (Liu et al. 2012). Further studies are needed in order to define the optimal doses of corticosteroids to improve postoperative pain as well as the potential long-term benefits. Moreover, although no serious adverse effects have been reported, the safety aspects and implications on the glycaemic profile, risk of infections and wound healing remain largely unknown.

Systemic administration of alpha-2 adrenergic agonists clonidine and dexmedetomidine decreases postoperative opioid consumption, pain intensity and nausea (Blaudszun et al. 2012). The major site of action of α2-adrenergic agonists is the spinal cord, and both their analgesic and anti-hyperalgesic effects result from spinal administration either epidurally or intrathecally. Alpha-2 adrenergic agonists also act locally on peripheral nerve endings and potentiate local anaesthetics effect in peripheral nerve blocks (Marhofer et al. 2013). Intra-articular injection of clonidine demonstrated local anti-inflammatory effect during knee arthroscopic surgery (Gentili et al. 2001). Alpha-2 adrenergic agonists also possess interesting anxiolytic and sedative effects (Crassous et al. 2007). Although they do not induce respiratory depression, they cause hypotension, which may interfere with rehabilitation process (Blaudszun et al. 2012).

Ketamine is an old anaesthetic drug used in clinical practice since the 1960s. The main working mechanism of action of ketamine is not clear but relies on a noncompetitive antagonism of excitatory neurotransmission (NMDA receptors) and potentially on an interaction with opioid, monoaminergic, cholinergic, purinergic and adenosine receptor systems. These multiple interactions account for the various clinical effects observed such as anaesthesia, analgesia, anti-hyperalgesia, induction of psychiatric symptoms (schizophrenia-like) or positive effects on depressive moods. Ketamine also has anti-inflammatory properties as it avoids an exacerbated proinflammatory reaction (De Kock et al. 2013). It influences the immune system and is able to regulate very early local inflammatory processes as it reduces the release of proinflammatory cytokines without affecting the production of anti-inflammatory ones. These anti-inflammatory properties of the drug may be involved in its anti-hyperalgesic effect. Several meta-analyses have highlighted the benefits of intraoperative use of low doses of ketamine (median dose 0.4 mg/kg; range 0.1–1.6 mg/kg). At that dose, a reduction of postoperative pain scores, both at rest and during mobilisation, as well as a 30–55 % reduction in postoperative opioid requirements has been demonstrated. The greatest efficacy of perioperative ketamine administration is observed in painful procedures including thoracic, upper abdomen and major orthopaedic surgeries (Laskowski et al. 2011). After TKA performed under general anaesthesia combined with a femoral nerve block, a single bolus dose of ketamine 0.5 mg/kg significantly reduced opioid use and facilitated rehabilitation (Adam et al. 2005). In TKA under general anaesthesia but without femoral nerve block, ketamine (bolus dose of 0.2 mg/kg followed by an infusion of 60–120 μg/kg/h) administered over 48 h had a significant opioid-sparing effect and decreased pain intensity at both rest and mobilisation (Aveline et al. 2009). Patients receiving ketamine infusion achieved earlier rehabilitation progresses and hospital discharge (Aveline et al. 2009).

Gabapentinoids (gabapentin and pregabalin) are primarily used in the treatment of epilepsia and chronic neuropathic pain syndromes. Gabapentin and its analogue pregabalin have been designed as analogues of γ-aminobutyric acid, but their mechanism of action mostly relies on binding to α2-δ subunit of voltage-gated calcium channels in the central nervous system, preventing the release of several excitatory neurotransmitters like glutamate. Pregabalin, the second generation of calcium channel α2-δ ligands, offers the advantages linked to a more reliable pharmacokinetic profile with a rapid dose-independent absorption. Thereby, pregabalin will be more suitable for a short-term use like the perioperative period, a single preoperative dose of 300 mg oral pregabalin displaying a sufficient central nervous system bioavailability to reach anti-hyperalgesic therapeutic levels within 6 h post-administration (Buvanendran et al. 2003). For a few years, both drugs have been used as part of multimodal analgesia in the perioperative setting where they help to reduce postoperative pain and opioid consumption (around 30 %) as well as opioid-related side effects like nausea and vomiting. More recently, they have also demonstrated promising anti-hyperalgesic effects, which might reduce the risk of persistent pain after surgery (Zhang et al. 2011). As a matter of fact, a preoperative dose of 300 mg pregabalin followed by a twice-daily dose of 150 mg during 14 days reduced the incidence of chronic neuropathic pain after TKA (0 % at 3 and 6 months versus 8.7 and 5.2 %, respectively) (Buvanendran et al. 2010). Patients who received pregabalin treatment required less oral opioids when hospitalised and had greater active flexion on the first 30 days (Buvanendran et al. 2010). Side effects of perioperative gabapentinoids involve greater sedation, dizziness and blurred vision (diplopia). The promising results obtained so far with pregabalin reinforce the need for further studies aimed to determine optimal doses and duration of administration.

The supposed superiority of locoregional anaesthesia and analgesia over general anaesthesia and systemic analgesia for the perioperative management of various surgical procedures including orthopaedic surgery is currently debated, especially in the context of fast-track or enhanced recovery protocols (Carli et al. 2011; Harsten et al. 2013). For TKA, some authors strongly believe that spinal anaesthesia, continuous peripheral nerve blocks and/or wound infiltration represents the recommended standard to achieve the goals of fast-track surgery (Carli et al. 2010a; Memtsoudis et al. 2013). That might not be true anymore for two reasons. First, the drugs and techniques used for general anaesthesia and systemic analgesia have changed these last years, along with the use of less invasive surgical techniques. Modern general anaesthesia seems to favour a more comfortable recovery in terms of nausea, vomiting and dizziness, to reduce pain and morphine consumption pain and to result in shorter LOS (Harsten et al. 2013). Second, regional analgesic techniques provide statistically superior analgesia compared with systemic opioids. However, the benefits of locoregional techniques have been previously, almost exclusively, examined in terms of postoperative analgesia, while other significant clinical outcomes, e.g. performance-based outcomes, should be assessed besides pain intensity in order to determine the success of a chosen technique (Bernucci and Carli 2012). The benefits of locoregional techniques may not be so relevant when considering the later outcomes, which are now considered of major interest in the patient’s rehabilitation process.

Besides an analgesic effect related to nerve conduction block preventing the noxious stimuli from the wound to reach and to further sensitise the central nervous system, locoregional analgesia also modulates the inflammatory response caused by tissue destruction. This beneficial anti-inflammatory effect partly relies on the use of local anaesthetics, which, in addition to blocking nerve conduction, are known to have a variety of anti-inflammatory actions. Inhibition of phagocytosis in macrophages or leucocytes, decrease in adhesion of polymorphonuclear granulocytes and reduction in platelet aggregation are well-known effects of local anaesthetics (Cullen and Haschke 1974; Hu and Muscoplat 1980). A newly discovered anti-inflammatory mechanism is the inhibition of the proton channels of microglia, which are known to play a crucial role in regulating inflammatory responses in the central nervous system (Matsuura et al. 2012).

Lumbar epidural analgesia has been popular over the last decades as there is evidence for lower postoperative thromboembolic complications and other protective effects (Rawal 2012). Nevertheless, there is today little evidence for a decrease in perioperative mortality and morbidity in a low- to medium-risk population in relation to the use of perioperative epidural analgesia. Moreover, the widespread implementation of anticoagulant regimens may not only overcome the benefits of epidural analgesia on thromboembolic complications but also make around 30 % of the patients ineligible for the technique. The failure rate of the technique may reach 28 %. Systemic review of the literature has shown that epidural analgesia provides superior analgesia to systemic opioids at rest and with activities whatever the type of surgery (Liu and Wu 2007). This clinical superiority of epidural analgesia may occur only with mobilisation through the first postoperative day. A previous systematic review in TKA comparing lumbar epidural blockade with systemic opioid analgesia reported better dynamic pain scores in the epidural group although limited to the first 6 h (Choi et al. 2003). As the magnitude of pain relief must be weighed against the frequency of adverse events, patients who received epidural analgesia had more hypotension, urinary retention and pruritus, whereas systemic opioids caused more sedation, but no difference was found for the postoperative respiratory depression or nausea and vomiting (Choi et al. 2003).

Peripheral nerve blocks (PNB) of the major nerves supplying the lower limb represent an attractive alternative to epidural analgesia. With the development of ultrasounds (US), peripheral nerve blocks have regained interest among the anaesthesiologists. Although nerve injuries lasting longer than 1 year are rare, their frequency with both US guidance techniques and nerve stimulator (NS) guidance techniques seems to be similar (Orebaugh et al. 2012). However, US-guided peripheral blocks are associated with a significant increase in the success rate when compared with NS techniques only or other methods (Gelfand et al. 2011) as is the US-guided peripheral nerve catheter placement which proves also a lower risk for accidental vascular puncture when compared with NS-alone guidance (Schnabel et al. 2013). The development of enhanced echogenic needles in the last years has contributed to the improvement of peripheral nerve blocks’ safety and efficacy (Hebard and Hocking 2011).

The most popular analgesic technique after TKA remains the femoral nerve block (FNB) (also referred to as “3-in-1 block”), either single-shot or continuous infusion (United Kingdom, France). For major knee surgery, a femoral nerve block provides postoperative analgesia which is comparable with that obtained with an epidural technique but with an improved side effect profile, i.e. less hypotension, pruritus and urinary retention (Fowler et al. 2008). A preoperative single-shot FNB reduces postoperative opioid use and significantly decreases pain scores with activity, but not at rest, up to 48 h after surgery by comparison with systemic opioids (Paul et al. 2010). The addition of a continuous perineural infusion of local anaesthetic in the postoperative period does not seem to enhance the analgesic benefits observed after a single-shot injection (Paul et al. 2010).

The sensory innervation of the knee is complex and involves the femoral nerve along with contributions from the sciatic and obturator nerves at the posterior and medial aspects. Consequently, a peripheral block of the sciatic nerve may be added to the classical FNB in the aim to improve postoperative analgesia, such combination remaining less invasive and associated with fewer serious complications than the use of a lumbar plexus block. The addition of sciatic nerve block could improve the quality of analgesia during the first 24 h by reducing posterior knee and calf pain (Cappelleri et al. 2011; Cook et al. 2003; Wegener et al. 2011). However, the addition of a sciatic block does not result in better pain scores or lesser opioid consumption than the use of a single-shot FNB alone (Abdallah and Brull 2011; Fowler 2009; Fowler et al. 2008; Paul et al. 2010). The FNB is not always complete, as it does not constantly produce analgesia of the obturator nerve. Nevertheless, although the obturator nerve is far more consistently involved with a lumbar plexus block technique than with an inguinal FNB technique, the obturator block, like the sciatic one, does not seem to translate into improved patient recovery or pain reduction (Bergeron et al. 2009; Macalou et al. 2004). Compared with systemic analgesics, the use of FNB allows faster rehabilitation and reduces time to discharge (Macfarlane et al. 2009). Despite the reduction of local knee inflammation (Martin et al. 2008) and in some extent of the systemic inflammatory reaction (Bagry et al. 2008) associated with the use of locoregional anaesthesia, the impact on patient’s outcomes is disappointing. Better postoperative analgesia does not lead to better knee function as observed by an improved knee range of motion that usually does not extend long after the block’s completion. Better postoperative analgesia does not translate into improved long-term pain reduction after TKA neither. Patients receiving a 48 h continuous FNB achieved better knee flexion in the first 6 days after TKA but not further on (no significant functional difference observed at 3 months) (Kadic et al. 2009). Within the concept of preventive analgesia, some authors have extended the duration of analgesia to 4 days by discharging patients at home with femoral catheter and infusion pumps. The results of the study did not show that extending an overnight continuous FNB to 4 days improves or worsens subsequent health quality of life, assessed by WOMAC scores, between 7 days and 12 months after TKA (Ilfeld et al. 2008; Ilfeld et al. 2011). Other authors have examined the impact of a more complete knee block by adding a 48 h sciatic block to the continuous FNB. They did not find improved postoperative outcomes in terms of long-term pain or functional disability (Wegener et al. 2011).