Severe postoperative pain, as well as causing suffering and discomfort for the patient, delays rehabilitation (Kehlet et al. 2006; Carli et al. 2010), increases hospital length of stay (LOS) and is a risk factor for the development of persistent postsurgical pain (PPSP) (Kehlet et al. 2006), which is associated with higher pain intensity (Torrance et al. 2006). For patients embarking on rehabilitation programmes, it is important to distinguish between pain at rest (PAR) and movement-evoked pain (MEP) (Grosu et al. 2013) as MEP is potentially more than twice as intense as PAR in the first three postoperative days (Srikandarajah and Gilron 2011). Moreover, poorly relieved MEP may enhance central sensitisation and further increases the risk of PPSP (Grosu et al. 2013).

Although the majority of patients experience a reduction in pain within 3 months of undergoing TKA (Vilardo and Shah 2011), approximately 20 % endure pain for a long time (Beswick et al. 2012). Perhaps unsurprisingly, almost the same proportion (19 %) of TKA patients are unsatisfied with the outcome of the operation (Bourne et al. 2010), with persistent postsurgical pain (PPSP) being the apparent primary predictor of dissatisfaction (Bourne et al. 2010). There have been few studies on the consequences of PPSP (Grosu et al. 2013), and in particular on the impact of the long-term use of analgesics (Steyaert and Lavand’homme 2013). Nevertheless, 56 % of patients continue to receive analgesics at 30 days postoperatively (Andersen et al. 2009), 40 % are still taking them at 4 months (Puolakka et al. 2010) and approximately 25 % still require analgesics after 2 years (Carroll et al. 2012).

Efforts to improve effective pain relief after TKA are therefore required (Carli et al. 2010) to improve both patient comfort and early postoperative function and provide the optimal conditions for rehabilitation (McCartney and McLeod 2011; Sisak 2013). Early mobilisation, which is associated with adequate pain control, has been identified as paramount for achieving a good clinical outcome (Dillon et al. 2012), with accelerated rehabilitation achieving early functional recovery in the motivated patient (Sisak 2013). A recent study indicated that starting rehabilitation within the first postoperative 24 h is associated with a reduced length of stay and a more rapid return to independent walking (Labraca et al. 2011).

Over recent decades, advances in implant and instruments design, along with less invasive techniques, have reduced the severity of the surgical impact for the patient (Shen et al. 2009). Improvements in anaesthetic techniques and operative treatment has also led to quicker recovery, improved patient satisfaction and earlier independent walking (Berger et al. 2004; Lombardi et al. 2006).

Innovations in analgesic approaches and techniques have not been so dramatic in recent years. Peri-operative analgesia has traditionally relied on techniques such as general anaesthesia with per- and postoperative intravenous opioids, epidurals with or without catheters, and femoral and/or sciatic regional nerve blocks with or without continuous postoperative infusions (Sisak 2013). All these techniques are still being used, despite known and well-defined failure and complication rates (Sisak 2013).

For example, parenteral opioids may not provide adequate pain relief but are especially affected by adverse events like nausea and vomiting, confusion, constipation, urinary retention, sedation, respiratory depression and pruritus, all of which may prevent rapid rehabilitation and influence the perception of the quality of postoperative care (McCartney and McLeod 2011; Dillon et al. 2012; Grosu et al. 2013). Moreover, opioids may induce hyperalgesia and acute tolerance, and exacerbate postoperative pain (Grosu et al. 2013). Although epidural anaesthesia is used widely and can achieve good pain control (McCartney and McLeod 2011), it could be associated with adverse events such as excessive motor block and the failure rate can be as high as 20 % (McLeod et al. 2001). Moreover, when it is removed the patient experiences pain later on in the postoperative period (Grosu et al. 2013).

In contrast, peripheral nerve blocks, such as continuous femoral nerve block, provides pain control comparable to epidural analgesia, but with fewer adverse effects (Singelyn et al. 1998). Continuous femoral nerve block has therefore become regarded by many as the gold standard for pain relief in TKA (Hadzic et al. 2010). However, it is hampered by being technically demanding, more time consuming and requires acquired skill vs simpler techniques (McCartney and McLeod 2011; Sisak 2013) and it may cause prolonged motor block (Kandasami et al. 2009).

One solution to the issue of pain relief in TKA is that of multimodal analgesia, with the aim of minimising the need for opioid analgesia (Kehlet and Dahl 1993). Furthermore, recognition of the role of the incision in the initiation and maintenance of pain sensitisation has meant that intra-wound analgesia has received increasing attention (Grosu et al. 2013).

For example, adding posterior capsule infiltration to femoral nerve block has been shown to reduce pain scores and improves extension of the knee in the first postoperative 12 h (Krenzel et al. 2009). A systematic review also demonstrated that general anaesthesia plus femoral nerve block or spinal anaesthesia in TKA provides intra-operative anaesthesia and improves postoperative pain relief (Fischer et al. 2008). Importantly, severe complications following peripheral nerve block and spinal anaesthesia are rare (Aromaa et al. 1997; Auroy et al. 1997).

However, a note of caution should be observed, as patients receiving multimodal analgesia may remain in significant pain following TKA (Davies et al. 2004), potentially increasing the risk of chronic pain (Perkins and Kehlet 2000). Furthermore, intra-articular analgesia does not, on its own, improve patients’ pain, satisfaction or range of motion (Joo et al. 2011).

The ongoing quest to find the ideal analgesic approach for TKA has led to the relatively novel technique of local infiltration analgesia (LIA). This involves the intra-operative infiltration of anaesthetic agents locally into the surgical wound to provide an additional form of analgesia (Keijsers et al. 2013; Ibrahim et al. 2013). The term was coined in the late 1990s, when it was developed as part of a multimodal pain management and early mobilisation protocol after joint replacement (Kerr and Kohan 2008; Ibrahim et al. 2013).

LIA is a promising technique which is easy to use, safe and inexpensive, providing simple, effective pain relief with few adverse effects, due to the local administration (Keijsers et al. 2013; Affas et al. 2011; Kerr and Kohan 2008; McCartney and McLeod 2011; Vendittoli et al. 2006). Moreover, it avoids the need for peripheral nerve blocks and/or catheters (Sisak 2013) and may reduce opioid consumption (Kerr and Kohan 2008). Consequently, LIA has become relatively common for several types of surgical procedures (McCartney and McLeod 2011; Keijsers et al. 2013).

Typically, the LIA injection consists of a mixture of an anaesthetic drug and a non-steroidal anti-inflammatory drug (NSAID), to which adrenaline or a corticosteroid may be added (Raeder 2011). This may be followed by small boluses of the same mixture postoperatively using an intra-articular catheter (Ibrahim et al. 2013).

Initially a small randomised controlled trial comparing LIA with placebo established the benefits of local infiltration during TKA (Bianconi et al. 2003). Kerr and Kohan used LIA to manage postoperative pain in 325 patients presenting for THA, elective hip resurfacing or TKA (Kerr and Kohan 2008). They found that two-thirds of patients did not require morphine for postoperative pain control, and few patients experienced limiting adverse effects. The majority of patients were able to walk with assistance 5–6 h after surgery, with independent mobility achieved 13–22 h postoperatively, and 71 % of patients were discharged after a single night in hospital.

Since then there have been a number of studies of LIA in TKA (McCartney and McLeod 2011), several indicating that the technique is associated with lower morphine consumption, reduced pain and shorter LOS (Essving et al. 2011; Andersen et al. 2007). LIA has also achieved favourable outcomes vs epidural analgesia (Andersen et al. 2010; Spreng et al. 2010) and has been advocated as an alternative to femoral nerve block due to the reduced risk of affecting quadriceps function (Chaumeron et al. 2013).

Two meta-analyses have been published recently in an attempt to summarise the evidence so far on the use of LIA in TKA (Gibbs et al. 2012; Keijsers et al. 2013). Gibbs et al. identified 29 randomised controlled trials of LIA following TKA (Gibbs et al., 2012a, b). Thirteen studies compared LIA with control, of which seven concluded that LIA was beneficial in providing analgesia, while six concluded it was of no benefit. LIA was not compared with peripheral nerve block in any of the identified studies; however, it was shown to be beneficial against a background of epidural analgesia in one sufficiently controlled study. Other studies included in the analysis either lacked systematic infiltration or used sub-therapeutic doses. Looking at outcomes for continuous infusion vs multiple analgesic administration, Gibbs et al. identified 11 randomised controlled trials, of which seven used multiple administration and four continuous administration. For multiple administration, five studies reported that local infiltration was effective, as did three continuous infusion studies. However, two studies indicated that local infiltration was less effective than continuous femoral nerve block (Carli et al. 2010; Toftdahl et al. 2007). Interestingly, Andersen et al. and Essving et al. demonstrated that an injection given 24 h postoperatively decreased both PAR and MEP (Andersen et al. 2008b; Essving et al. 2010). Gibbs et al. identified nine studies that examined LOS, in six of which LIA failed to reduce LOS significantly, although two of those reported that the time to be ready for discharge was reduced without affecting LOS.

The meta-analysis by Keijsers et al. involved 7 studies, with a total of 406 TKAs in 374 patients (Keijsers et al. 2013). Postoperative pain visual analogue scales (VAS) scores were in favour of LIA. However, the reductions were modest, at just 12.3 % on the first postoperative day across all studies, with a heterogeneity in scores of 71 %. There were no significant differences in pain scores on postoperative day 2, and no difference in pain scores on activity. Patients in the LIA group used 14.8 % less opioids than those in the placebo group on the first postoperative day. However, there was no difference in opioid usage by 48 h. Keijsers et al. concluded that LIA may be able to reduce pain and opioid use on the first postoperative day, but the effect is small and not clinically relevant (Keijsers et al. 2013).

The lack of a definitive body of evidence for LIA in TKA may be due to inconsistencies in the techniques used in the different studies. The basic procedure, as outlined above and originally defined by Kerr and Kohan, is for the surgeon to infiltrate the joint and surrounding tissues with a high volume of anaesthetic, typically including ropivacaine, adrenaline and adjuvant drugs such as NSAIDs, clonidine, corticosteroids or opioids (Kerr and Kohan 2008; Maheshwari et al. 2009a, b). Specifically, ‘local infiltration’ may be characterised as the systematic injection into all the tissues which are exposed, instrumented or incised during surgery, including the capsule, ligaments and other soft tissues, along with the subcutaneous layers (Andersen et al. 2008b).

A number of different infiltration techniques have been described, involving various agents and methods of administration (Sisak 2013). Agents used in addition to local anaesthetics include steroids (Fu et al. 2010), magnesium sulphate (Chen et al. 2009), morphine (Ritter et al. 1999) and NSAIDS, (Vendittoli et al. 2006a, b), while the methods of administration include intra-articular injection (Badner et al. 1996), infiltration of all traumatised tissues (Andersen et al. 2008a), intra-articular infusion (Ong et al. 2010) or a single shot intra-articular injection (Keijsers et al. 2013).

Due to the lack of consensus over the optimal technique for LIA; with a wide variety in mixture, concentration and volume of analgesics administered but also for the anatomical location where the mixture is infiltrated (Keijsers et al. 2013). In one study, the addition of ketorolac was advocated after it was demonstrated that combining it with ropivacaine reduced morphine consumption, reduced pain intensity and shortened the patients’ readiness for discharge (Andersen et al. 2013). Another study of 100 unilateral TKA patients examined the impact of adding 40 mg triamcinolone to the LIA mixture, with recipients recording lower pain scores and better ranges of motion up to 6 months postoperatively vs controls (Bramlett et al. 2012).

Another issue is that the duration of action of an intra-operative bolus of LIA appears to be short (Busch et al. 2006) and it is unclear whether an indwelling catheter should be placed to extend the duration of action through intermittent boluses of local anaesthetic (McCartney and McLeod 2011; Kehlet and Andersen 2011). There are, however, concerns over the risk of infection with an indwelling catheter (McCartney and McLeod 2011; Grosu et al. 2013).

For their meta-analysis, Gibbs et al. examined the different techniques used for LIA in TKA, focusing on five studies that used the same initial infiltration of ropivacaine and adrenaline of all instrumented tissues (Gibbs et al. 2012a, b). Patients received slight variations in the volume and concentration of analgesia, the use of bandaging, catheter site placement, and additional subcutaneous analgesia administration. None of the variations could demonstrate superiority of efficacy.

The use of peri-articular multimodal drug injection (PMDI) in TKA was proposed 10 years ago with the aim of obtaining a better patient outcome after surgery. Several studies on the use of PMDI showed stable and promising results on pain control, primarily in the first 24 h after surgery, better patient satisfaction, lower morphine consumption during the first 48 post-operative hours, reduction of average length of stay, and improvement of the knee function while minimizing complications (Essving et al. 2010a, b; Husted et al. 2010; Park et al. 2014). However many concerns about this approach need to be elucidated, including the relevance of the drug’s choice, the site and timing for wound delivery of local anesthetic agent after TKA.

One of the best topics is about the choice of the drugs used in PMDI. A large consensus in the literature establishes that the most frequently used is a combined mix of local anaesthetic (generally bupivacaine or ropivacaine), magnesium sulphate, morphine, non-steroidal anti-inflammatory drugs (ketorolac), sometimes steroids (depending on the author’s preference) and epinephrine. These agents cover all the pathways of the pain as opioid receptors, nociceptors and local inflammatory mediators. Peripheral opioid receptors, covered by morphine, are the first expressed after surgical trauma, sending immediately information to central receptors. NSAID agents such as ketorolac inhibit the production of prostaglandin inflammatory mediators, leading to an anti-inflammatory, antipyretic and analgesic effect. A local anaesthetic such as bupivacaine inhibits the nociceptor and its action is prolonged, with less risk of systemic toxicity, by epinephrine’s local vasoconstriction (Lamplot et al. 2014). The use of magnesium sulphate is to “stabilise” the plasma membrane of nociceptors leading to a more difficult transmission of nociceptive stimuli.

Regarding the physiology of pain transmission, the PMDI plays an important role in reducing or avoiding the “early transmission” that leads to the recruitment phenomenon in which painful stimuli induce hyperpolarisation of the neural pathway which makes it difficult to control the subsequent pain.

This pain reduction through its transmission mechanism at peripheral and central levels allows better patient participation in the rehab program and consequently faster discharge from hospital, improving the overall outcome (Maheshwari et al. 2009a, b).

Despite the action of PMDI seldom extending over the first 24–48 h postoperatively, it is helpful, in association with multimodal analgesia program, in assessing the patient’s immediate postoperative program, avoiding the experience of “early” pain with lower opioid consumption and faster home readiness, particularly in a “fast track” program (Husted et al. 2010; Lamplot et al. 2014).

Another important issue is the injection site. Randomised studies have reported no differences of analgesic efficacy of intra-capsular vs intra-articular injection, and intra-articular vs extra-articular wound space (Husted et al. 2010; Ruder et al. 2014).

Moreover, local infiltration of the posterior capsule of the knee with a cocktail based on marcaine, morphine, adrenaline antibiotic and corticosteroids before cementing the implants followed by a second infiltration prior to closure did not reduce the morphine consumption and did not improve the recovery of walking capacity, physical activity and knee function vs continuous femoral block (Meftah et al. 2012). Conversely, peri-articular injection of ropivacaine, morphine, non-steroidal anti-inflammatory drug (ketorolac), epinephrine, and cefuroxime after cementation of the prostheses into the sheath of the medial and lateral collateral ligaments and posterior capsule in simultaneous bilateral TKA improved early post-operative pain control, but not patient satisfaction (Koh et al. 2010, 2012). Similar outcomes have been observed after peri-articular injection of bupivacaine, ketorolac after implantation or before wound closure. Overall these data suggest that it is not the injection site which is important (Table 17.1) but the dose and the type of multiple agents used.