CC BY-NC-ND 4.0 · Joints 2018; 06(03): 204-210
DOI: 10.1055/s-0038-1675845
Review Article
Georg Thieme Verlag KG Stuttgart · New York

The Challenge of Managing the “Third-Space” in Total Knee Arthroplasty: Review of Current Concepts

Pierluigi Antinolfi
1  Service of Orthopedics and Trauma Surgery, Santa Maria della Misericordia Hospital, Perugia, Italy
Francesco Manfreda
2  Department of Orthopedics and Traumatology, University of Perugia, Perugia, Italy
Giacomo Placella
3  Service of Orthopedics and Traumatology, Azienda Ospedaliera Regionale San Carlo, Potenza, Italy
Julien Teodori
2  Department of Orthopedics and Traumatology, University of Perugia, Perugia, Italy
Giuliano Cerulli
4  I.R.T.A.L. Istituto di Ricerca Traslazionale Apparato Locomotore, Nicola's Foundation, Arezzo, Italy
Auro Caraffa
1  Service of Orthopedics and Trauma Surgery, Santa Maria della Misericordia Hospital, Perugia, Italy
2  Department of Orthopedics and Traumatology, University of Perugia, Perugia, Italy
› Author Affiliations
Further Information

Address for correspondence

Pierluigi Antinolfi, PhD, MD
Service of Orthopedics and Trauma Surgery, Santa Maria della Misericordia Hospital
Piazza Menghini, 1, 06129 Perugia

Publication History

31 January 2018

07 October 2018

Publication Date:
10 December 2018 (online)



Total knee arthroplasty (TKA) is the best treatment for advanced knee osteoarthritis and it has proven to be durable and effective. Anterior knee pain (AKP) is still one of the most frequent complications after TKA, but sometimes no recognized macroscopic causes can be found. The correct treatment of patella is considered the key for a proper management of AKP. The inclusion of patellar resurfacing during TKA has been described as a potential method for the reduction of AKP. After surgeons started to resurface the patella, new complications emerged, such as component failure, instability, fracture, tendon rupture, and soft tissue impingement. Patelloplasty has been proposed as a good alternative to resurfacing but whether or not to resurface the patella is still a controversial topic in the literature. Therefore, patellofemoral joint is a complex critical aspect in TKA and choosing between the several options of treatment of patella could not be sufficient. In this review, evidence-based studies do not succeed in resolving this difficult argument. The accurate management of the so-called “third space” should include an accurate assessment of cartilage layers, balance of soft tissue, preoperative anterior tracking, and positioning of the femoral and tibial components. In fact, the selection of suitable implants and adherence to proper surgical technique are the fundamental principles for the success of TKA.



The patellofemoral (PF) joint (PFJ) is involved in over 45% of cases of knee osteoarthritis (OA), independently or concurrent with OA in the other compartments.[1] Despite its incidence, few papers in the literature have focused on it. Patellar disorders are strictly linked to anterior knee pain (AKP) and sometimes it can be traced back to malalignment of the extensor mechanism.[2] PF malalignments seem to have several predisposing intrinsic factors, such as neural inflammation, altered proprioception, patellar tendinitis, Osgood–Schlatter disease, coronal knee malalignment, and joint hypermobility.[3] [4] [5] [6] [7] There are also some others causes of AKP that are not related to malalignment, such as inflammatory disease, tumors, infections, stress fractures, fat pad inflammations, and psychological causes.[8] [9]

Total knee arthroplasty (TKA) is the most effective treatment for advanced knee OA. The intervention does not always include a resurfacing of the patella, even if PFJ is frequently involved. PFJ is a critical aspect of kinetic features of the implant design in TKAs and is supposed that implant kinematics contributes to PF complications. A successful TKA generally is able to increase knee flexion angle, thus enhancing the PF contact and pressing forces.[10]

Long-term changes for nonresurfaced patella have been described. Laughlin et al demonstrated how the lateral tilt increases over time.[11] Shih et al showed that in patients with long-term follow-up the patellar tracking and the PFJ remained normal in 60% of the TKAs, but progressive degenerative changes and patellar maltracking are common radiographic findings. Retention of the native patella is associated with progressive degeneration of the lateral patellar facet in 85% of cases.[12]

Patients with preoperative patellar maltracking could be considered at risk of developing these abnormalities and clinical symptoms. Indeed, leaving the native patella is related to a high incidence of AKP, up to 47%.[13] [14] For this reason, some authors recommend routinely resurfacing the patella.[15] [16]

Nevertheless, some studies showed that patella complications in TKA are more frequent in the resurfaced group than in nonresurfaced one.[17] [18] [19] Some others suggested a selective decision based on factors such as preoperative AKP, evaluation of patellar thickness, severity of OA in the third compartment, and the experience of the surgeons.[20] [21] Therefore, during surgery, surgeons that do not routinely resurface the patella should consider several issues, including the severity of OA, knee alignment, PF tracking, implant design, and position of the components.

If the surgeon decides to retain the native patella, several options could be chosen to reduce the rate of AKP and improve the outcomes. Options that have been described include patelloplasty, denervation, and circumferential resection of the osteophytes. Patelloplasty has been defined as the surgical options aiming at the improvement of the congruency between the patella and the trochlea, conducting a real reshaping of the patella.[22]

Despite the excellent records about the long-term outcome of TKA, there remains controversy about the correct management of the PFJ.


Patellar Resurfacing

In 1955, McKeever tried to reduce pain in PFJ OA by patella resurfacing as an alternative to patellectomy or debridement.[2] Even though TKA has shown good results almost since its creation, in the 1970s many physicians identified AKP as a frequent complication of TKA. Since then, the implants included the patellar component. This choice resulted in reduction of incidence of AKP, but despite this, complications related to resurfacing started to emerge.

Several designs and materials have been used for patellar resurfacing. At the beginning, all-polyethylene patellar components were used, but they presented with early deformation and accounted for almost 50% of all TKA revisions.[23] Some years later, a metal backing was incorporated into the design to give load transfer to the patellar bone and to allow cementless fixations using a porous metal surface. Unfortunately, new complications arose related to the shearing forces between the polyethylene surface and the metal backing under eccentric loads.[24] One of the most frequent complications, the polyethylene wear, resulted in exposure of the metal backing leading to metal-on-metal contacts. Loosening and migration became frequent, leading to exposition of the femoral metal to the patellar bone. Several studies conducted onto the metal-backed patella reported 5 to 33% failure rate.[25] [26] Later, engineers tried to modify the polyethylene dome shapes to normalize the contact forces, thus improving polyethylene wear patterns.[27] The introduction of a low-contact-stress press-fit metal-backed rotating patella increased the congruity of the PFJ. These innovative designs led to the reduction of contact pressure, which resulted in fewer complications.[28]

Actually, modern prostheses incorporate an all-polyethylene patellar component with pegs (from 1 to 3) and cemented fixation, reducing the high risk of failure with the cementless metal-backed implants.

The implants are available to the surgeons in the range of 8- to 10-mm thickness. A caliber should be used to measure the central patellar thickness, with the aim of maintaining the anatomical thickness. A residual bone thickness of approximately 15 mm is required.[29] The patella resection is performed with the aim of being flat. The implant is placed and generally cemented, because, as we have previously said, complications rate may decrease with cement. Medial placement of the implant improves patellar tracking when compared with central placement.[30] Only if the patella tracks laterally or lifts off medially, a lateral patellar retinacular release should be performed. Some authors describe an additional lateral facetectomy, proper of patelloplasty, and it can be performed when a malaligned patella is addressed. Open partial lateral patellar facetectomy has been shown to improve function but with poor correlation with radiological findings.[31]

Several complications can occur following patellar resurfacing. Instability is probably the first one, requiring revision surgery in about 0.5 to 0.8% cases. Factors that could contribute to patellar instability include malposition of the components, soft tissue imbalance, excessive femoral component size, and inadequate patellar resection.[32] [33] Treatments for instability are component revision, proximal and distal realignment of the soft tissue, and osteotomy of the tibial tubercle. The most dangerous complication is patellar fracture (0.05–8.5% of TKA cases).[34] [35] [36] Decreased bony thickness combined with osteopenia, avascularity, trauma, fatigue, and stress can generate this kind of adverse event, whose risk increases when lateral retinaculum release is simultaneously performed.[37]

Fractures may occur during surgery too: an extremely thin patella is at higher risk of fracture due to the high mechanical pressures. Patellar fractures are generally treated with a conservative approach, especially when an implant is not loose and the extensor mechanism is not disrupted, whereas surgical options include open reduction with internal fixation, excision of fracture fragments, and patellectomy.[38] [39] Another complication is wearing of the patellar component (incidence ranging between 5 and 11%).[23] Factors that can be associated to this complication are maltracking and abnormal contact force. Maltracking is able to induce component loosening in 0.6 to 1.3% of cases, as well as traumas.[40] Soft-tissue impingement, the so-called patellar clunk syndrome, results from the formation of a fibrous nodule over the patella proximal pole and is strictly associated with posterior stabilized design.[38] Generally, it can be treated through an arthroscopic or open resection. The patellar ligament rupture (0.22–0.55%)[41] can be due to excessive dissection, knee manipulation, and trauma. Either staple fixation or grafts can be used in treating acute ruptures, but with poor outcomes.[42]

The advent of modern prosthetic designs with better PF congruence, smoother patellar tracking, and superior patellar fixation, has led toward fewer patellar complications. Recent studies have reported lower PF complication.[43]

Some authors have proposed algorithms to assess the ideal candidate to patella resurfacing. Bourne and Burnett[44] formulated five criteria for choosing the resurfacing option: severe PF OA, poor tracking, inflammatory arthritis, obesity, and age over 60. Several studies have denied all these beliefs: no differences have been found between obese and nonobese[45]; no correlation has been proved with severity of OA[46]; and studies on alignment have also considered the correlation between malalignment and necessity for a resurfacing, without a certain result.[47]

Some gross data derived from randomized clinical trials (RCTs) seem to support those surgeons who are used to resurface the patella.[48] Anyway, some systematic reviews and meta-analysis summarized that the decision is still difficult. Pakos et al[49] reviewed 10 RCTs that favored patellar resurfacing when compared with the relative risk of revision and AKP. Parvizi et al[50] conducted a meta-analysis on 14 studies, which proved that there was no difference between resurfaced and un-resurfaced patellas in reoperation rate, but showed that patellar resurfacing results in the reduction of AKP and achieve better patient satisfaction. Nizard et al[51] demonstrated a lower risk of reoperation for resurfaced patella with respect to un-resurfaced (0.43 vs. 0.71). The Swedish Knee Arthroplasty Register, in 2004, reported a 10-year follow-up in over 30,000 patients, where secondary resurfacing was found to be the main cause for reoperation.[52] In 2007, the Australian National Joint Registry reported a relative risk of reoperation in 4.7% for un-resurfaced patellas, as compared with 3.8% for resurfaced. The same registry shows no differences regarding the type of fixation of the component.[53] Calvisi et al[54] found that resurfacing might reduce the risk of reoperation and AKP. Patil et al[55] evaluated functional scores such as the Knee Society Score (KSS), 36-Item Short Form Health Survey, and satisfaction, and found no differences during comparisons.[55] Li et al conducted a large review founding no differences between the two groups.[56] Chen et al[57] published the most recent, important, and valuable meta-analysis about this issue 4 years ago. They reviewed 14 RCTs, with homogenous parameters, with a cumulative size of over 1,700 patients. Relative risk of reoperation favored the patellar resurfacing (0.50 vs. 0.76), but no significant differences were found for AKP. Anyway, the authors revealed a higher KSS in the resurfaced group.[57]



Patelloplasty consists of different procedures, such as patellar decompression, lateral patellectomy, and patella reshaping, which are aimed at creating a good congruence between the patella and femur. It has the advantage of conforming to the articular facets of the patella toward a large variability of tracking surfaces. In fact, the shape of the trochlea could vary and the shape of the femoral prosthetic component can vary as well according to the implant design.[58]

Reshaping of the patella has the scope of achieving a good matching between the prosthetic trochlea and the articular surface of the patella. This procedure is conducted with or without a resection of cartilage layer.[22] Unfortunately, this is not a well-standardized procedure. There are no specific and worldwide-recognized steps of procedure. There is an excessive variability of gestures. Some author use to reduce the thickness of the patella to less than 20 mm, while some other between 22 and 26 mm. Some authors totally remove the cartilage layer, while some others leave the native one. Therefore, as there is a great variability in the procedure, a great variability could exist in outcomes as well. All these gestures are original, without a real validation in the literature.

Rate of complications is controversial due its variability in procedures. Although this is a technique characterized by rare intraoperative complications,[59] there could be some postoperative ones, such as maltracking for a new patellar shape and a controversial adverse reduction of the strength of the extensor mechanism; but this topic has not yet been investigated in the literature.

The literature is still poor of relevant studies about this issue. Sun et al[59] published a retrospective study with 152 patients, divided in two groups: patelloplasty versus denervation. They found that patelloplasty was better in both relieving AKP and improving patient's satisfaction and knee function.[59] Župan et al[60] reported better clinical outcomes in patients with patelloplasty with respect to the simple denervation and removal of osteophytes. They found better outcomes in Knee injury and Osteoarthritis Outcome Score (KOOS) and in Oxford Knee Score, but revealed no significant differences in KSS and pain scores. They also correlated the results to postoperative patella thickness, but they found no significant correlation. The first systematic review focusing on patelloplasty in primary TKA has been conducted by Cerciello et al[22] about clinical results, the comparison between patelloplasty and simple denervation and osteophytes removal revealed lower AKP in patelloplasty (2.9% vs. 4.6%), as well better clinical outcomes at KOOS, KSS, and Feller patellar score.[22] [61] [62] [63]

Though all these evidences show that patelloplasty is a safe procedure without frequent intraoperative and postoperative complication, the quality and quantity of studies are still poor.

Control groups in the studies are often lacking. In some other studies, preoperative data are missing, so comparisons are extremely difficult.


Other Alternatives

Intraosseous hypertension is another important hypothetical cause of AKP in the third space after TKA.[64] [65] In degenerative pathologies of the knee, an increased intraosseous pressure is described.[66] This patellar hypertension, though not specifically studied yet, is reported to be associated with idiopathic AKP.[67]

In several experimental studies, an impaired venous drainage are linked with hypertension, especially in patients with patellar chondromalacia and OA.[68] Causes for an altered venous drainage include direct compression of vascularization from fat pad, osteophytes of the superior patella, a sustained knee flexion position, chondromalacia, and OA.[67] [69] [70]

Patellar drilling has been proposed for improving impaired drainage and it could indirectly relieve AKP. The first results described after decompression by drilling via the infrapatellar fat pad reported a direct reduction of intraosseous pressure and pain relief.[71]

Ertürk et al[72] published a trial on 49 TKAs treated with patelloplasty and decompression of the patella. They conducted the decompression drilling on the edge of the patella with the aim of not damaging the articular cartilage and drilling at vertical plane to minimize the risk of fracture. The authors reported excellent clinical results, especially for range of motion (ROM) and patellar scores. Anyway, the study did not use a control group.[72]

Lee et al[67] compared patelloplasty to patelloplasty decompression. A 3.5-mm drill has been used via fat pad under tissue protection in the group with decompression. The authors reported AKP in 21.3% of the control group and in 18.5% of the study group. Higher KSS was observed in patients with decompression (78.2 vs. 71.8).



The incidence of AKP is high after TKA, reported between 4 and 49%.[73] Daily activities are impaired due to this complication: they include stair climbing, cycling, and getting up from a chair. Pathogenesis of AKP is still unclear but several potential factors have been proposed to be linked.

Over time, the most studied options for the management of this complication have been the treatments of patella during a TKA.

The gross data reviewed in this article show that the patella resurfacing might reduce pain but is associated to severe complications. Moreover, some systematic reviews and meta-analysis did not succeed in finding significant differences in terms of clinical outcomes.

Unfortunately, patelloplasty is still poorly studied, even if it seems the best alternative solution to resurfacing of the patella. Systematic reviews and meta-analysis are affected by poor quality of RCTs both in the methodology and in uniformity of samples of patients and outcomes.

RCTs are generally conducted through clinical and radiographic evaluations, but few articles are available about biomechanical performances. Smith et al[74] evaluated the gait pattern by a meta-analysis, showing no significant differences between resurfaced and un-resurfaced patellas. Myles et al[75] used an electrogoniometer to evaluate improvements in ROM, but they found no differences between the two groups of patients.

Scientific literature lacks RCTs that focus on comparison between patella resurfacing and patelloplasty. Campbell et al[76] found no significant differences in the International Knee Society and Western Ontario and McMaster Universities Osteoarthritis Index scores at 10-year follow-up. Similar uncertain results have been reported by Burnett el al.[47] However, despite the clinical scores, patient's preference was toward resurfaced patella. Smith et al[77] evaluated 164 patients through KSS, reoperation rate, AKP, and patient's satisfaction. No significant differences have been revealed, but a lower AKP was reported for TKAs with resurfaced patella.[77]

Several other elements have gained importance for the assessment of the PFJ. Abnormal postoperative PF tracking has been described as linked to retropatellar pressure.[78] However, some studies did not find a clear association between patellar tilt or subluxation and AKP.[79]

Generally, in the past, physicians believed that patellar height was linked to AKP. Patella baja, although it has a poor incidence in TKA, is surely a predisposing factor for flexion deficit, but there are no studies that correlated this condition to AKP. On the contrary, patella alta is generally associated with higher contact forces between the trochlea and patella. However, studies have not confirmed that it is a certain predisposing factor for developing AKP.[80] van Houten et al[80] analyzed the effects of patella positioning on AKP in TKA. The authors found that the amount of patellar displacement or patellar tilt is not related to AKP.

Another critical aspect that is generally discussed is the type of bearing surface of TKA. Fixed bearing has been always believed to be of higher risk compared with mobile bearing designs. Type of bearing surface might influence patellar kinematics, because some studies showed lower PF contact stress in the mobile bearing compared with fixed bearing.[81] However, few studies demonstrated a potential benefit of mobile bearing in AKP, some of those suggested that the performances of mobile bearing decline over time.[82] [83] [84] [85]

The tibiofemoral contact point is another potential predisposing factor for AKP. Greater anterior positioning of the tibiofemoral contact point is the complication of the lever arm of the extensor mechanism, which leads to PF disadvantage.[86]

The last important element that is able to create a variability in the onset of AKP is the vascular supply to the patella. In fact, the amount of resection of fat pad can play an important role.[87] [88]



The overall incidence of postoperative anterior knee in TKA is still too high. Evidence-based publications cannot be easily interpreted. There is statistical heterogeneity, perhaps because of methodological diversity. The general trends seem to present a reduction of risk of reoperation in TKA with a resurfaced patella, but without significant benefits to knee function and satisfaction with respect to patelloplasty. So the decision of whether or not to resurface the patella should be assessed according to the evaluation of the PF joint, the design of the prosthesis, the experience of the surgeons, and patient's features.[89] [90] Anyway, treatment of the patella does not seem to be the only factor that can influence AKP in TKA. Positioning of the femoral and tibial components, both in rotation and in AP direction, can influence the contact forces in the third space. The design of the prosthesis and choice of implant may also play an important role.

Treatment of smooth tissues is still poorly considered as a predisposing factor for AKP, such as the treatment of patellar fat pad, treatment of retinacula, and balancing of peripheral compartments could be extremely important both for vascular supply of the extensor mechanism and for biomechanical tracking of the third space.


Conflict of Interest

None declared.

Address for correspondence

Pierluigi Antinolfi, PhD, MD
Service of Orthopedics and Trauma Surgery, Santa Maria della Misericordia Hospital
Piazza Menghini, 1, 06129 Perugia