Painful knee arthroplasty: current practice

Abstract

Primary total knee arthroplasty is the treatment for end-stage arthritis of the knee; in the last years, it is becoming more common and reliable, due to technical and implant improvement. With larger implant rates, the overall complications will increase and pain is the most common sign of implant failure. Pain can be related to a lot of different clinical findings, and the surgeon has to be aware of the various etiologies that can lead to failure. Pain does not always mean revision, and the patient has to be fully evaluated to have a correct diagnosis; if surgery is performed for the wrong reason, this will surely lead to a failure. In this paper, the authors revised the more common causes of failure that can have a painful onset proposing an approach for diagnosis and treatment.

Introduction

Nowadays, primary total knee arthroplasty has become a very reliable and durable procedure; for these reasons, total knee arthroplasty (TKA) surgeries are growing in number and patients undergoing surgery are now younger than in the past. In the next years, the expectation is to observe a further increase of this kind of surgery [1]. An increased number of implants will probably lead to an increased number of complications and failures.

Pain is commonly considered as a sign of failure of a TKA, but it does not always have to be treated with a revision surgery, especially if the diagnosis is not clear [2].

Results in revisions are not as good as in primary implants and surgery have to be performed only on the basis of a sure diagnosis to improve results [3•].

The first step in approaching a painful TKA is to differentiate intra-articular from extra-articular causes. Extra-articular causes can originate by the hip or general conditions. Intra-articular causes can be correlated to the implant (failure, poor design), to surgical technique (malalignment, malrotation, incorrect sizing, etc.), or to surrounding tissues (inflammatory diseases, allergy, infection, etc.).

The first paper classifying the classes of failure was published in 1988 [4], but the schematic classification of knee failures was given in 2003 by Vince [3•].

In this paper, we will treat the major causes of failure leading to pain:

• Instability

• Stiffness

• Extensor mechanism failure

• Infection

• Loosening

• Infection

• Anterior knee and patella-related pain

• Aseptic loosening

• Extra-articular causes (mystery knee, neuroma, complex regional pain syndrome (CPRS), allergy, bursitis)

Instability

It has to be considered a very important cause of TKA failure; 10 to 20 % of all knee revision surgeries are due to this problem [3•, 5–10], representing nearly 26 % of overall complications in the first 5 years after surgery. Only infection has a more frequent presentation [11].

Patients usually report discomfort that can degenerate to pain and functional impairment, but instability is infrequently reported in itself [12]. It can be a consequence of surgical errors or poor implant design [9]. Mild instability can be misdiagnosed and confused with infection; however, this should always be ruled out before surgery (see section dedicated).

General medical conditions can also affect stability [5]:

• Obesity

• Muscular pathologies

• Hip or foot deformities

• Neurologic pathologies

Before testing knee objective stability, it is useful to search for extra-articular problems such as neuromuscular disorders, hip or ankle deformities, and areas of tenderness (i.e., pes anserine and Gerdy’s tubercle) [13•].

A complete radiographic analysis of the knee should be performed including anteroposterior (AP) and latero-lateral (LL) radiographs, full-length weight-bearing radiographs, and patellar tangential views. Varus-valgus stress images are very useful in order to assess collateral ligaments integrity and deformities’ reducibility [7, 14] (Fig. 1).

Fig. 1.

Instability in a primary TKA with a postero-stabilized (PS) implant: a AP view X-ray, b full-length weight-bearing, and c patellar view X-rays showing articular space opening on the medial side

In case of suspected malrotation, a CT scan or a MRI should be performed; MRI is more technically demanding but is reported to be more reliable [15–17]. Tibio-femoral instability can be classified according to three different patterns: flexion instability, genu recurvatum, and extension instability.

Flexion instability

More common in cruciate retaining (CR) implants [5, 12], it can occur in posterior stabilized (PS) also [18]. In CR implants, instability can be led by surgical errors (undersized femoral component, excessive tibial slope, and misdiagnosed previous PCL rupture) or later PCL failure. In PS implants, the cam prevents dislocation [19], but with an unbalanced flexion gap, the cam can unhook the femoral stop [5, 13•]. In addition, the cam mechanism does not provide varus-valgus constrain; this has been observed especially in patients with a damaged lateral collateral ligament (LCL) [20]. A strong correlation has been proved with a decreased condylar offset by 4 mm, femoral distal under-resection with distalization of joint line by 6 mm, and an increased tibial slope up to 5° [13•]. Thus, a stepwise approach including three steps can be proposed:

• Correction of the excessive slope (to a mean of 5°)

• Upsizing the femoral component (posterior condyle offset to a mean of 4 mm)

• Increasing distal femoral resection (not always necessary)

Genu recurvatum

Genu recurvatum is identified in only 0.5 to 1 % of patients [21, 22]. It is usually a consequence of underlying conditions like quadricep weakness, paralysis, bone deformity, previous high tibial osteotomies (HTO), instability [22], and plantar foot flexion [22]. Soft tissue balancing is an important part of the treatment with particular attention on the choice of the constraint, exception made for bone deformity that can be corrected with bone cuts. The most common neuromuscular disorder causing genu recurvatum is poliomyelitis [23]. Patients with fixed valgus deformities may have an iliotibial band (ITB) contracture leading to genu recurvatum instability [22].

Extension instability

Extension instability can be classified as symmetrical and asymmetrical according to the shape of the extension gap: rectangular or trapezoidal [20].

Symmetrical instability can originate from excessive distal femoral or proximal tibial resection. Both events result in inadequate space filling by the implant components [24]. In case of excessive tibial resection, to use a thicker polyethylene insert is enough to correct instability [5].

In the case of excessive distal femoral resection, the use of a thicker polyethylene is not suitable for resulting flexion gap tightening [5]. The solution is the joint line distalization by using femoral distal augments. Joint line elevation should be avoided for flexion stiffness, patellar overstuffing, and mid-flexion instability [9, 24, 25].

Asymmetric instability, instead, is a consequence of incomplete correction of a previous angular deformity or surgical ligament damaging. This has to be corrected addressing the soft tissues: several techniques have been proposed. Authors’ preferred technique is the Insall “pie-crusting” of iliotibial band and posterolateral corner for valgus knees [26, 27] and the MCL elevation or multiple needle puncturing for varus deformities.

Stiffness

Stiffness is a rare complication after TKA that has been reported to be from 1.3 % [28] to 60 % [29]. A univocal definition of stiffness however is lacking; every author defines it differently. It should be considered that to sit on a chair without hands requires 93 of flexion and for tying a shoe requires 103 of flexion on average [30]. Christensen et al. [31] define stiffness as a maximum ROM less than 70°. In a high percentage of cases, a stiff knee is also reported as painful by patients.

The etiology of stiffness is not clear but seems to be multifactorial. Risk factors are limited preoperative range of motion, biological predisposition, intraoperative complications, patients’ psychological problems, or a poor postoperative rehabilitation [32–37].

Multiple treatment modalities have been proposed, including manipulation under anesthesia (MUA), arthroscopic arthrolysis, open arthrolysis, and implant revision. The indications are not clear, and the results are still debated. MUA has been proved to have a good result when performed within 12 weeks after the index surgery [38]. Arthroscopic arthrolysis has the same results of MUA [29].

Open arthrolysis has inferior results and greater complications including recurrence and major complications [29].

Extensor mechanism failure

Extensor mechanism failure is reported to occur after primary TKA from 1 to 12 % [39], more commonly in chronic settings [39–41]. Patients usually report knee impotence following sudden pain without or after a minor trauma. Surgical treatment can alter the anatomy, modify patellar tracking, and produce fibrotic scars [42•] that can potentially produce anterior knee pain. Furthermore, some systemic pathologies and medications [43] can affect tendons’ structure and integrity. Predisposing factors for rupture are local corticosteroid injections [43], diabetes [43–45], microtrauma [44], inflammatory disorders (e.g., rheumatoid arthritis) [41], metabolic disorders (e.g., diabetes) [41], iatrogenic lesions (e.g., ACL reconstruction with patellar tendon graft) [46], and fluoroquinolones [47, 48]. For these reasons, the surgeon has to carefully investigate on previous episodes of pain and on comorbidities that can affect tendons and patella.

The site of damage can be various (quadriceps tendon, patella, and patellar tendon), and the treatment has to be chosen and addressed according to the correct diagnosis.

Infection

Infection is one of most frequent causes of pain in TKA [49]. In the USA, a percentage ranging from 25 to 43 % of revisions TKA is for septic failure [50, 51].

At the recent International Consensus Meeting (ICM) on periprosthetic joint infection (PJI), a definition of infection was given and the diagnosis criteria were redefined [52]; a PJI is present when one major criterion or at least three minor criteria are present [52] (Table 1). Exception is made for some low-virulence microorganism as Propionibacterium acnes [49].

Table 1.

Periprosthetic joint infection criteria

Major criteria	• A sinus tract communicating with the joint • Two positive periprosthetic cultures with phenotypically identical organisms
Minor criteria (at least 3)	• Elevated serum C-reactive protein (CRP) AND erythrocyte sedimentation rate (ESR) • Elevated synovial fluid white blood cell (WBC) count or change on leukocyte esterase test strip • Elevated synovial fluid polymorphonuclear neutrophil percentage (PMN%) • Positive histological analysis of periprosthetic tissue • A single positive culture

Comorbidities increasing PJI risk are previous surgery, recent hospitalization, poorly controlled diabetes mellitus (DM), malnutrition, active liver or chronic renal disease, smoking, alcohol consumption, drug abuse, male gender, post-traumatic or inflammatory arthritis, and immunodeficiency [53•]. Risk factors for PJI are recent bacteremia, multiple surgeries, and history of PJI.

Before clinical exam, anamnesis should be collected, and special attention should be given to:

• Pain time onset (early, delayed, or late)

• Previous possible bacteremia (e.g., dental procedures, etc.)

• Perioperative events (prolonged operative time, excessive wound drainage or healing problems, blood transfusion, or additional antibiotics during surgery)

Silent infections have been described [54]; these are characterized by a positive culture at the time of the revision in unlikely infected patients.

Physical examination should be focused on detecting any major or minor criteria of infection reported before. Additional suggestive findings are wound dehiscence, joint warmth, redness, or swelling. Fever may be present but not to necessarily [55]. A complete examination of the joint is necessary.

Standard X-rays should include anterior‒posterior, lateral, merchant, and full-length weight-bearing view in order to evaluate common signs of infection, such as loosening, osteolysis, sub-periostal elevation, or the presence of a trans-cortical sinus tract [53•].

The utility of CT scans and MRI is low but can be useful to exclude other causes of pain [53•].

Nuclear imaging has been confirmed to be effective by some authors [56]; for the low cost-effectiveness ratio, it is excluded from the standard workout [53•].

Blood test has a good sensitivity and negative predictive value [57–59], particularly erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) [60].

In acute setting, ESR is not useful and CRP is positive if greater than 100 mg/L. In chronic settings, ESR should be higher than 30 mm/h and CRP more than 10 mg/L [53•]. When positive, a joint aspiration is indicated [61]. Other possible blood tests are serum interleukin 6 (IL-6) and peripheral white blood count cell (WBC), but they are not so effective in clinical practice [62, 63]. Knee aspiration must be performed to evaluate total WBC and for differential (PMN%). Synovial fluid cultures are indicated and have to be prolonged for at least 14 days.

Additional tools for diagnosis are as follows:

• Sonication

• Leukocyte esterase on the synovial fluid [64]

• Polymerase chain reaction (PCR)

• α-Defensin in synovial fluid

Different surgical options are available, depending on the type of infection.

Irrigation and debridement (I&D)

I&D is indicated in early infection within 3 months from index surgery or after late hematogenous infections no later than 3 weeks from onset [53•]. The reported success rate range is wide with good results ranging from 0 to 89 % [65, 66]. However, after one failed I&D, the revision of the implant should be considered [53•].

One-stage revision TKA

It consists in implant removal, complete and accurate debridement, and revision of the TKA in one single step. It is indicated when effective antibiotics are available in patients without sepsis and in whom resection arthroplasty and reduction of bio burden may be useful [53•].

Two-stage revision TKA

It is indicated in patients with sepsis, with unknown pathogen, in which an effective antibiotic therapy is not available [53•]. No definitive evidence is reported in literature regarding the optimal interval between the stages that ranges between 2 and 8 weeks [67].

Spacers can be articulated or static: a recent systematic review concluded that articulating spacers resulted in better ROM, lower re-infection rate, less bone losses, and easier re-implantation in PJI [68]. Static spacers are indicated in cases of massive bone loss and soft tissue or ligamentous deficiency.

Anterior knee and patella-related pain

Anterior knee pain is a very common problem with difficult solution. Diagnosis can be various and complex.

The onset of the pain is indicative of its origin. A sudden onset after a non-symptomatic period is suggestive for failure of the component or the extensor mechanism; on the contrary, an onset of symptoms occurring immediately or soon after surgery is more likely related to the surgery itself.

Conventional X-rays including a lateral and Merchant’s view are useful [69, 70]. Component alignment and rotation has to be evaluated using CT scan [71].

Patello-femoral instability and maltracking

Patello-femoral instability is a common cause of postoperative pain in TKA, with a reported revision rate of about 1 %.

Maltracking can be related to an insufficient soft tissue balancing, patello-femoral joint overstuffing, asymmetrical resection of the patella, or poor component positioning [70, 72]. Implant design may also play an important role in patellar maltracking [73, 74].

Patellar instability can be caused by various surgical errors like residual valgus alignment, patella alta, malrotation, medial translation of the femoral component, asymmetric patellar resection, excessive lateral placement of the patellar button, and improper soft tissue balancing [70, 75].

Component malrotation

Different studies [76–78] confirmed that an internally rotated femoral or tibial component can be related to lateral tilt angle and excessive lateralization of the patellar component, resulting in a painful TKA. An internal rotation of 3° to 7° is sufficient to cause chronic pain in TKA [69, 79]. An excessive internally rotated tibial component may lead to patello-femoral problems; excessive external rotation is better tolerated [80] but can be painful for excessive soft tissue impingement [81].

Patellar clunk syndrome

The “patellar clunk” syndrome (PCS) is a due to a suprapatellar fibrous nodule that impinges in the femoral component [82]. The incidence is widely variable ranging in between 0 and 12 % [83]. Risk factors can be various: shortened patella, smaller patellar components, decreased patellar thickness, joint line elevation, and flexed femoral component [83]. The clunk typically occurs at approximately 30° to 45° of flexion, when the fibrous nodule impinges the intercondylar notch and determines pain.

Aseptic loosening

Aseptic loosening accounts from 31 to 55 % of TKA failure [84]; the pathogenesis is most of all implant-related [84, 85].

Obesity was found to increase twice the risk for implant failure [86].

Limb malalignment and material debris appear to be important risk factors [87]. An optimal surgical alignment of the limb is fundamental for a better distribution of the forces to the periprosthetic bone [88]. Debris provokes an indirect biological reaction on osteoclasts and osteoblast function causing osteolysis [89–91].

Intra-articular inflammation produces a pseudo-synovial membrane secreting fluid that can increase intra-articular synovial pressure and can weaken the implant‒bone surface [92].

It is associated with pain during gait and weight-bearing and radiolucent lines of at least 2 mm in the X-rays [93]; in case of strong suspicion without radiographic evidence, a CT scan is indicated.

Extra-articular causes

“Mystery knee”

Sometimes, a painful knee after total knee arthroplasty cannot be apparently explained even if the pain is seriously affecting patients’ quality of life [94].

Vince [3•] and Hofmann et al. [95] described diagnostic algorithms for failure analysis. The evaluation of a painful TKA includes an extended clinical history, study of the type of pain, test infiltration, and a total body clinical examination including spine, hip, and ankle. Unusual cause of pain, like obturatory nerve syndrome or “end of stem pain,” must be considered. When a clear reason for pain is not evident, an early or chronic infection must be suspected. Laboratory tests and joint aspiration are the first analyses to be performed in these cases.

CPRS

The complex regional pain syndrome type 1 (CRPS1) is an uncommon cause of pain, representing about 1–2 % of painful TKA [96]. The principal feature of the CRPS1 is a joint neuropathic pain difficult to localize and unproportioned to the rehabilitation phase, associated to sensory alteration and skin burning ache exacerbate by motion or cold [97]. No specific instrumental examination is available for the diagnosis of CRPS, but it can be easier if there is persistent dusky or cyanotic discoloration or atrophy of the skin of the leg, knee, and foot. The discoloration can increase with temperature changes, especially with cold [98].

Allergy

The failure rate of total knee arthroplasty caused by an allergy-related problem is very low, accounting for 1–2 % of all revisions [99]. In the general population, the skin allergy is quite frequent, observed in about 26 % for at least one metal, but the correlation between skin allergy and implant failure is unknown [100].

Unallergic material prosthesis represents a good solution, but the high cost can be an issue.

Neuroma

Iatrogenic nerve injuries after TKA are reported to be 1 to 70 % of cases [101, 102]. The more frequently affected nerve is the infrapatellar branch of the saphenous nerve, but rare cases of neuroma of the medial and inferior cutaneous nerve, of the proximal tibiofibular nerve, and of the common peroneal nerve are described. The suspicion of neuroma should be taken into consideration if there is a cutaneous hypersensibility with a positive Tinel sign, while a more accurate diagnosis can be achieved with a selective anesthetic injection [103].

Peri-articular bursitis (ITB, pes anserinus)

Extra-articular tendinous problems or bursitis can be the cause of pain after TKA implant [104•], more commonly on the pes anserinus bursa [105] or iliotibial band [106].

This pain is usually caused by an excessive stretching of the iliotibial band after TKA implant in the valgus knee and of the pes anserinus in varus knee. Other predisposing factors are tibial component overhang, bone prominence, and instability [20, 107] (Table 2).

Table 2.

Table showing indications for revision of different pathologies and their location

Pathology	Location	Need for revision
Infection	Intra-articular	Yes
Patello-femoral clunk		Not always
Osteolysis		Yes
Instability		Yes
Malalignment		Yes
Arthrofibrosis		Not always
Recurrent hemarthrosis		Not always
Popliteal impingement		Yes
Loose bodies		Not always
Synovitis		Not always
Overhanging component		Not always
Neuroma	Peri-articular	No
Stress fractures		Not always
Bursitis (pes anserinus)		No
Ileo-tibial band		No
Heterotopic ossification		No
“End-of-stem pain”		Not always
Complex regional pain syndrome (CPRS)	Extra-articular	No revision
Hip/spine pathology
Vascular etiology
Unrealistic expectations
Psychological profile
Depression

Conclusion

In conclusion, the approach to a painful TKA is a challenging problem for both surgeon and patient. The quality of life impairment for the patient and emotional implications can be great, and this can drive to an early and sometimes rash indication to revision. The surgeon, then, has to carefully analyze all the aspects and possible causes of pain.

The diagnosis usually takes time, and the patient has to be aware that a revision surgery is not always the right solution. The result of a revision procedure is in a high percentage of cases less satisfying than in case of a primary TKA, especially if performed with wrong or unclear indications. Additional rehabilitation or a longer recover time after surgery may be enough to solve the problem.

For this reason, in some cases, the “wait and see” approach can be effective and can drive to a spontaneous solution.