Proper central tracking of the patella within the trochlear groove requires a complicated balance of forces and is influenced by soft-tissue integrity, patellar resection, component positioning, and implant design. Revision rate for TKA related to patellofemoral complications is around 8%²,⁸; however, recent studies indicate that patellofemoral complications may be decreasing with modern implants and improved surgical technique.⁹,¹⁰ Patient factors that can predispose to patellofemoral instability include medial retinacular laxity, traumatic disruption of the medial patellofemoral ligament, lateral retinacular contracture, weakness of the vastus medialis oblique muscle, and trochlear hypoplasia.¹¹ The medial soft tissue structures are needed to counterbalance the more lateral vector pull of the quadriceps and decrease the functional Q-angle. Patients with a valgus knee deformity or lateralized tibial tubercle are predisposed to patellar instability due to an increased functional Q-angle.¹² If any of these factors are identified, they must be addressed at the time of surgery to ensure proper patella tracking.

Proper surgical technique is paramount to achieving patellofemoral stability. The patellar osteotomy should be a symmetric resection such that the remaining medial and lateral patella facet thicknesses are the same. A proper resection will remove more of the medial facet since the native medial facet is typically thicker than the lateral facet.¹³ The resection should remove enough bone to accommodate the thickness of the patellar button without overstuffing. Patellar tracking is optimized when the patellar component is medialized as far as possible and the femoral and tibial components are rotated appropriately. If the femoral component is internally rotated relative to the transepicondylar line or if the tibial component is internally rotated relative to the tibial tubercle, the patella will track laterally and be at higher risk for dislocation (Fig. 56-1). One should avoid medial translation of both the femoral and tibial components as these errors result in lateralization of the tibial tubercle and a lateral force vector force on the patella.

Advancements in implant designs have greatly improved patellofemoral stability. Previous prostheses with high patellofemoral complication rates typically had a shallow trochlear groove, short and narrow anterior flange, and a small radius of curvature.¹⁴ Newer implant features that have improved patellar tracking include a more laterally oriented trochlear groove, heightened lateral flange, and deepened intercondylar notch. Use of rotating tibial platform (or mobile bearing) implant has shown to improve patellar tracking with less lateral patellar tilt, lower patellofemoral contact stresses, and decreased need for a lateral retinacular release when compared to fixed tibial bearings.¹⁵,¹⁶ Several different types of patellar component designs exist, including a dome-shaped button and conforming biconcave patellar component (also known as an anatomic patella). Cadaveric studies have shown increased inferosuperior shear forces with the domed patella component compared to biconcave components¹⁷; however, this has not shown to affect clinical outcomes thus far.¹⁸,¹⁹ Anatomic patellar designs have sagittal plane kinematics that more closely resemble a nonresurfaced patella when surveilled under fluoroscopy during a deep knee bend.²⁰ However, because of the bicondylar ridge, the patellar component must be
precisely rotated for appropriate tracking in the trochlear groove, whereas the dome component with its symmetric 360° conformity is more forgiving to malrotation.

Instability of the patellofemoral joint should be identified and addressed at the time of surgery. The knee is tested through full range of motion during implant trialing and then again before capsular closure to ensure the patella tracks centrally without lateral tilt or subluxation. The tourniquet should be released at this stage to eliminate the binding effect of the extensor mechanism which can alter patellar tracking.²¹,²² One can use the “no thumb” technique to assess tracking, in which the medial border of the patella should make contact with the medial femoral condyle through knee range of motion without the surgeon having to manually reduce it. This technique has shown to overestimate the number of cases needing lateral release,²³,²⁴ with some preferring the towel clip technique in which a clip is used to hold the quadriceps to the superior patella pole to mimic arthrotomy closure. The senior author prefers the “no thumb” technique to avoid overtightening the extensor mechanism during arthrotomy closure, which can have adverse effects on postoperative knee flexion and patella contact pressures. If patellar tilt or instability is noted, the etiology should be identified and addressed. The most likely causes for instability include imbalance of extensor mechanism soft tissues, component malposition, or anatomic abnormalities. For extensor mechanism imbalance, a lateral retinacular release can be performed to improve tracking.²⁵ The lateral release should be performed at least 1 cm lateral to patellar periphery to protect the superolateral geniculate artery that perfuses the patella.²⁶ Advancement of the vastus medialis muscle or medial retinacular tissues during capsular closure via imbrication can also be performed.²⁷ For patients with severe valgus deformities or a lateralized tibial tubercle, a medial tibial tubercle transfer can be performed. If proper tracking cannot be achieved with any of these realignment procedures, sometimes revision and repositioning of the components may be necessary to improve rotation and tracking.

Loosening of the patellar component accounts for a low percentage of TKA failures, which is reported between 3% and 4.8%.⁹,²⁸,²⁹ Early patellar component designs were associated with higher patellar loosening rates, particularly cementless metal-backed patellar components, with revision rates as high as 13.5%.²⁹ Other factors associated with increased risk for patellar loosening include body mass index over 30 kg/m², preoperative valgus knee alignment of 10° or more, preoperative knee flexion of 100° or more, osteopenia and tibial component thickness 12 mm or greater.²⁸ Certain surgical techniques that have been associated with loosening include component malposition, joint line elevation, performing a lateral retinacular release, uneven patellar resection, limited patellar bone stock, patellar subluxation, loosening of other components, and patellar avascular necrosis.²⁹,³⁰

Diagnosing patellar component loosening can be challenging, as patients can present with subtle symptoms.³⁰ Initial evaluation of symptomatic patellofemoral pain should always include radiographic imaging and infectious work-up. Most standard knee radiograph series used to evaluate the patellofemoral joint include a flexed lateral and Merchant view in which the knee is flexed and the patella component is compressed against the trochlea (Fig. 56-2). If loosening has occurred at the cement-implant interface, this position of compression can reduce a loose patellar component and potentially mask loosening³¹ (Fig. 56-3). If patellar component loosening is suspected but not able to be diagnosed radiographically, a bone scan or arthroscopic evaluation may be required to confirm the diagnosis.

Patients with patellar loosening do not always require treatment, as shown by Berend et al who reviewed 180 loose all-polyethylene patellar implants and found that only 0.3% required revision.³⁰ For those needing operative treatment, options include component revision (assuming there is enough bone stock to support another implant), bone graft augmentation, patellar resection arthroplasty (patelloplasty), gull-wing osteotomy (Fig. 56-4), use of a
porous tantalum metal component, or complete patellectomy.³² Revision with an all-polyethylene cemented patellar component can be considered if the remaining patellar thickness is a minimum of 10 to 12 mm, no fracture is present, and patellar vascularity is preserved.³³ Designs with three peripheral lugs are often useful when revising central lug designs, in which bone loss is typically central. A circular, three-lug design is also often selected when revising a failed three-lug device. In this situation, rotation of the component from its original position often provides adequate bone for fixation of the new patellar component. Outcomes following isolated patellar component revision have shown poor outcomes overall, with high complication rates (up to 45%) and need for reoperation.³⁴,³⁵ Complications reported from isolated revision include patellar fracture, patellofemoral instability, infection, extensor lag, and polyethylene wear.³⁵,³⁶ Considering this, all other etiologies for failure of the patellofemoral component need to be assessed and addressed, including lateral retinacular contracture and femoral/tibial component malposition. Component malposition is best assessed with a computed tomography scan so that rotation of the components relative to anatomic landmarks can be determined. In summary, isolated patellofemoral
component revision should be done with caution and high index of suspicion for other contributing factors. If the patellar component is well-fixed, has no evidence of damage or wear, and tracks appropriately, the patellar component can be safely retained even if revision of other components is being performed.³⁷

Patellar component failure due to polyethylene wear is a complication most commonly reported with metal-backed patellar components, with a reported incidence of 4.7%.²⁹ Metal-backed patellar components can also have polyethylene dissociation, seen as early as 14 months after TKA.³⁸ By nature of their design, metal-backed patellar components have an overall thinner polyethylene-bearing surface than the all-polyethylene implants which predispose to increased wear. The polyethylene is thinnest along the peripheral edge with the metal backing not entirely extending to the periphery of the polyethylene. This creates a sharp cutting edge and bending moment that can result in polyethylene dissociation or excessive wear (Fig. 56-5). Patients at highest risk for these complications include active males with increased body weight and knee flexion greater than 115°.³⁸,³⁹ Patellar tilting, subluxation, increased composite patellofemoral thickness, or a flexed femoral component can further predispose to patellar component failure. If the polyethylene wear is excessive enough, metal-on-metal wear between the patellar component metal plate and femoral component can occur resulting in significant damage.³⁹ One should be prepared to revise all components any time revision of a metal-backed patellar component is needed if there is evidence of destructive wear. Osteolysis secondary to polyethylene wear should also be assessed. In the absence of other findings, isolated revision of a metal-backed patellar component to a cemented all-polyethylene component has been associated with good clinical outcomes and low complication rates.⁴⁰,⁴¹ Wear of an all-polyethylene patellar components is usually well-tolerated. This is often associated with some polyethylene deformation which can improve contact mechanics.

Patellofemoral crepitus and patellar clunk syndrome are frustrating complications following TKA that can result in significant patient dissatisfaction. The reported incidence varies widely from 0% to 18% and is primarily seen with posterior-stabilized (PS) femoral implants that have a large intercondylar box.⁴²,⁴³ Both conditions result from proliferation of peripatellar fibrosynovial tissue at the superior pole of the patella (Fig. 56-6). Patellar clunk occurs when a discrete fibrosynovial nodule develops and gets entrapped within the intercondylar box of the femoral component during knee flexion producing a painful, audible “clunk” as the knee is extended (Fig. 56-7). Patellofemoral crepitus can have subtle clinical exam findings with either asymptomatic or symptomatic crepitation during knee motion. Dennis et al performed a retrospective multicenter case-controlled study of 60 patients requiring surgery for symptomatic patellar crepitus and identified several risk factors including previous knee surgery, use of smaller patellar components, decreased composite patellar thickness, shortened patellar tendon length, smaller femoral components, increased posterior
femoral condylar offset, flexed positioning of the femoral component, and thicker polyethylene inserts. The condition typically develops within 1 year of surgery with range of 3 to 21 months.⁴⁴,⁴⁵,⁴⁶

Nonoperative treatment with avoidance of high patellofemoral loading activities (such as walking stairs or squatting) and use of antiinflammatory medications will suffice in most cases. Many patients with patellar crepitus require no treatment and are often unaware that they have developed this condition.⁴⁷ Some patients with symptomatic patellar crepitus will have resolution of their symptoms within 1 year of symptom onset without surgical intervention.⁴⁵ For patients requiring surgical treatment, either arthroscopic or open synovial débridement can be performed with satisfactory outcomes reported.⁴⁴ Arthroscopic débridement has shown to successfully treat both patellar clunk and crepitus with several studies showing improved pain and functional scores and low recurrence rates.⁴³,⁴⁸,⁴⁹

While patients do well with both conservative and surgical treatments of patellofemoral crepitus, prevention should be a primary focus. Smaller femoral components should be avoided whenever possible and care should be taken to avoid flexing the femoral component during implantation.⁴⁶ Native patella baja can be identified preoperatively, but it can also occur iatrogenically by elevating the joint line with an excessive distal femoral resection. Patients with native patella baja should be counseled on an increased risk for patellofemoral crepitus and care should be taken to avoid elevating the joint line with a minimized distal femoral resection. Patellar resection should be enough to accommodate the thickness of the patellar component without overstuffing and the component should be positioned as superior as possible. Contact between the trochlea and patella shifts superior on the patella as the knee goes into deeper flexion, so superior placement of the patellar component will avoid any contact between the trochlea and nonresurfaced bone.⁴² Removing any exposed bone along the periphery of the patellar component (particularly the lateral facet) has also shown to lower incidence of crepitus.⁵⁰ Any excessive synovial tissue on the posterior aspect of the quadriceps tendon should also be excised as this is the tissue that proliferates to cause these conditions (Fig. 56-8).⁴²

Implant design features can affect the incidence of patellar crepitus. Fukunaga et al⁵¹ described the intercondylar box ratio, defined as the ratio of the intercondylar box height to the anterior-posterior height of the femoral component (Fig. 56-9). Many second- and third-generation PS femoral components have been designed with a decreased intercondylar box ratio resulting in a decreased incidence of patellar clunk and crepitus.⁵¹,⁵²,⁵³,⁵⁴,⁵⁵,⁵⁶ Notably, those prostheses with an intercondylar box ratio less than 0.7 were found to have no patellar clunk.⁵¹ Femoral components with a high intercondylar box ratio allow contact of the distal quadriceps tendon earlier in flexion than components with a lower ratio. The authors concluded that this design feature may be responsible for the high incidence of patellofemoral crepitus and clunk with certain prostheses. Martin et al demonstrated a lower incidence of patellofemoral crepitus with use of a modern PS implant that had a thinner and narrower anterior flange as well as a smaller femoral intercondylar box ratio.⁴⁷ With further improvements in implant design and surgical technique, incidence of symptomatic clunk or crepitus will hopefully continue to decrease.

Injury to the extensor mechanism is a rare but devastating complication following TKA with historically poor outcomes and high complication rates.⁵⁷,⁵⁸ It can occur at any point along the extensor mechanism including the quadriceps tendon, patella, patellar tendon, or tibial tubercle. There are unique risk factors associated with injuries at the various locations and a variety of surgical techniques described for treatment.