Patellofemoral Arthritis and Its Management With Isolated Patellofemoral Replacement: A Personal Experience

Abstract

Trochlear dysplasia is a common precursor of patellofemoral arthritis and may account, in part, for persistent anterior knee pain. After pathologic changes have been demonstrated by radiography, magnetic resonance imaging, or arthroscopy, isolated patellofemoral replacement usually can help, but a modern prosthesis and accurate instrumentation must be used. In addition, patient selection must be precise to achieve successful treatment outcomes. This means ensuring that the arthritic process is confined to the patellofemoral joint, and the patient has a stable mental state. Middle-aged patients with arthritis and instability secondary to trochlear dysplasia are probably the most appropriate group for this procedure.

Isolated symptomatic patellofemoral arthritis, which was considered rare in the past, has been reported in 8% of women and 2% of men older than 55 years.¹ In addition, radiologic changes in the absence of severe symptoms are even more frequent, despite the fact that many physicians do not take skyline radiographs and may therefore miss the radiologic changes.² This article presents studies that seek to establish the etiology of patellofemoral arthritis and discusses a prosthetic development designed to improve the outcome of its treatment, especially in younger patients with extensor mechanism instability.

Etiology of Patellofemoral Arthritis

Anterior knee pain is the most common condition seen in a general knee clinic and can produce considerable resource implications depending on the approach of the managing surgeon.³ The condition had been regarded as a benign, self-limiting condition to be managed primarily with physiotherapy, although some surgeons suggested that minor instability can be corrected.

Few regard adolescent anterior knee pain as a precursor of patellofemoral arthritis. However, because many patients with established patellofemoral arthritis have a history of anterior knee pain, a study was conducted to consider a possible correlation. The study group (group 1) included 118 patients about to undergo patellofemoral replacement who completed a questionnaire about adolescent anterior knee pain and patellar instability. The control group (group 2) comprised 116 patients about to undergo medial unicompartmental replacement. In group 1, 22% had experienced adolescent anterior knee pain compared with 6% of patients in group 2 (P<.001). Although 14% of patients in group 1 had a history of patellar instability, 17% reported adolescent knee pain, but no instability, compared with 6% in group 2.⁴ These results strongly suggest that not all adolescent anterior knee pain is benign and self-limiting. To achieve good treatment outcomes, physicians must be able to determine which patients likely will develop patellofemoral arthritis.

Trochlear dysplasia has long been confirmed and reported in France,⁵ but the condition is not widely recognized in the United Kingdom and the United States. Until 2003 in Bristol, UK, the condition was recorded only retrospectively as a possible diagnosis in 10 of the 385 patellofemoral replacements that had been performed. However, after awareness of the condition increased, it was noted to be the underlying cause in 11 of the next 48 patellofemoral replacements performed during 2004 and 2005. In addition, a retrospective review of 40 patients aged 55 years or younger showed evidence of trochlear dysplasia in 50%, based on operative notes, the presence of a crossing sign, and a sulcus angle >144°. This study also noted that prior extensor mechanism realignment surgery had been performed in only 38 of the 385 patellofemoral replacement cases, thus suggesting that trochlear dysplasia does not present only as recurrent dislocation.⁶

The clinical presentation of trochlear dysplasia has been studied in a group of 71 patients with a confirmed diagnosis of the condition.⁷ Approximately two-thirds reported a history of recurrent dislocation and presented at a younger age than the remaining patients who had never sustained dislocation but who had a variety of vague and disabling symptoms that usually had remained undiagnosed for years. In the United Kingdom, growing interest in patellofemoral pathology led to additional studies and the establishment of trochleoplasty as a treatment option for patients with severe dysplasia and instability. Although the short-term results are encouraging, the long-term outcome regarding prevention of patellofemoral arthritis is unknown. Twenty-five knees assessed for possible trochleoplasty underwent magnetic resonance imaging (MRI) that demonstrated thicker articular cartilage in the center of the trochlear groove than peripherally, thus creating an even less stable situation than is suggested solely by the bony architecture seen on radiography.⁸

Treatment With Isolated Patellofemoral Replacement

Many patients with patellofemoral arthritis have minimal symptoms and require no treatment, even with severe radiologic findings. Others who have symptoms respond to conservative measures or joint-preserving surgery. However, there is a significant number of patients with persistent severe symptoms whose condition can be helped by patellofemoral replacement.

Initial experience was with the Lubinus patellofemoral replacement, which has a narrow asymmetric trochlear component that is generally difficult to fit accurately and does not allow for introduction of any external rotation. Although initial pain relief was usually good, satisfactory results were not maintained. More than 50% of the devices failed within 8 years from a combination of maltracking, polyethylene wear, and development of tibiofemoral arthritis.⁹

A new design, the Avon patellofemoral replacement (Stryker Orthopaedics, Kalamazoo, Mich), was introduced in 1996. This symmetric trochlear component was broad and allowed for some external rotation. The results improved dramatically, with excellent pain relief being maintained and maltracking virtually eliminated, provided adequate soft tissue balancing was achieved at surgery. Among the first 109 patients (average age, 68 years) only 4 underwent revision during the first 5 years.¹⁰ The major cause of failure was progression of tibiofemoral arthritis, although occasional cases of maltracking, lateral catching, and patellar impingement were seen, both as a cause of revision and in some of the less satisfactory unrevised cases.

Because of the success of the Avon prosthesis, its use was extended to younger patients with either severe cartilage wear or recalcitrant instability. A group of 110 patients younger than 55 years was generally helped with reduced pain,¹¹ particularly when instability had been the primary problem that could not be resolved through conservative surgery including patellectomy.¹² These patients generally had trochlear dysplasia and benefited considerably by the creation of a chrome cobalt trochlear groove.

The Patient

To achieve excellent results from patello-femoral replacement, surgeons must consider the patient, the procedure, and the prosthesis.

Selecting appropriate patients for isolated patellofemoral replacement is essential, but the indications are not clearly defined. However patients must have disabling patellofemoral symptoms and significant demonstrable pathology to be considered for isolated patellofemoral replacement.

Ideally, patients should be elderly, but many patients present at a younger age (30-40 years) with severe disability and areas of exposed eburnated bone, usually on the lateral side of the trochlear following episodes of instability. Often such patients can be treated, although it must be recognized that their knees will not function normally again. Young patients with severe anterior knee symptoms but no major demonstrable pathology are unlikely to benefit and should not be treated by isolated patellofemoral replacement.

Once severe radiologic degenerative changes are visible in the patellofemoral joint, the procedure is indicated, but the tibiofemoral joint must be pristine because of the development of tibiofemoral arthritis. A radiographic review of 103 patients showed that at a mean of 7 years, 12% of patients had been revised for progression of tibiofemoral osteoarthritis and that among this group, 7% had evidence of definite but asymptomatic radiologic progression with the medial side predominating. However, tibiofemoral arthritis was not seen in the patients in whom trochlear dysplasia was the primary pathology,¹³ as reported by De Cloedt et al and Suranimala et al.^{14, 15}

After the success of patellofemoral replacement in elderly patients, 10 younger patients were treated, but some failed to thrive. A group of 110 patients younger than 55 years undergoing patellofemoral replacement were assessed with an SF12 and compared with a matched group undergoing medial unicompartmental knee arthroplasty. Although their physical disability scores were similar significantly, more patients in the patellofemoral replacement group had abnormal mental scores.¹¹ Overall, this group had poor outcomes, indicating that surgery should be conducted only after appropriate patient selection, although some had an excellent result.

The Surgery

Excellent results will be achieved only when both the soft tissue and the bony aspects are optimally performed. Realignment of the extensor mechanism is essential to ensure perfect patellar tracking. The procedure should be regarded as a soft tissue operation in which a prosthesis is inserted. A lateral parapatellar release is frequently required, and medial reefing, vastus medialis obliqus advancement, or tibial tubercle transfer may also be needed. Although a medial minimally invasive surgical approach can be used, the disease is usually lateral, and a release is often required. Therefore, it is practical to make a lateral incision to also reduce the altered skin sensation over the front of the knee that may result in poor kneeling capability after patellofemoral replacement.¹⁶

Because many of the previous types of patellofemoral replacements failed because of malalignment of the components,¹⁷ accurate instrumentation is necessary to ensure good reproducible results, especially because this surgery is performed relatively infrequently. The components should be oriented and positioned relative to the patient’s physical landmarks. Most current patellofemoral replacements require freehand trochlear preparation, increasing the risk of malalignment when the surgery is performed by an inexperienced surgeon or in a patient with a significant amount of patellofemoral deformity. A new type of instrumentation has been developed that involves first making a flat anterior femoral cut and then using a reamer to accurately prepare the trochlear region. The orientation of this cut is set by using the intramedullary femoral canal for flexion/extension alignment and the tibial shaft for internal/external rotation alignment. Using the tibial axis has been shown to be very accurate for positioning a component in the correct amount of external rotation.¹⁸

First, a small intramedullary rod is inserted into the femoral canal. The knee is flexed to 90° prior to attaching the cutting block. An extramedullary tibial alignment system is then used to attach to the tibia and orient the cutting guide in the appropriate amount of external rotation (Figure 1). The anterior femur is referenced via a stylus to set the correct resection height. After creating a correctly aligned anterior cut, the trochlear region is prepared using a reaming device and the guide tailored to the shape and orientation of the components (Figure 2). This method accurately prepares the trochlear area to receive the implant.


Figure 1: Tibiofemoral alignment guide is attached to the tibia at the ankle and references the cutting guide, giving accurate rotational control. Figure 2: The reamer reproducibly prepares the trochlear notch, removing minimal bone.

The Prosthesis

Because the patient undergoing patellofemoral replacement is younger than the average patient undergoing arthroplasty, prosthetic design is crucial. Early implant designs had a high degree of constraint (Autocentric, Richards Type I and II, Lubinus, etc) in an attempt to use the implant design rather than soft tissue balancing to provide proper patellar tracking. Unfortunately, these designs were somewhat vulnerable to minor variations in placement and showed some predisposition for maltracking failures.^9,17,19 Broader, less constraining designs have been shown to reduce the malalignment-related failures associated with some of these devices.^10,19

A new implant has been developed that addresses design failures of previous devices. This implant (Journey PFJ; Smith & Nephew, Memphis, Tenn) has a steep trochlear geometry based on the clinically successful design of the Genesis II total knee (Smith & Nephew). The asymmetric trochlear component is designed with a lateralized proximal patellofemoral groove to capture the patella in extension and the groove moves medially on the distal aspect to bring the patella to the midline of the knee in flexion, providing good patellar tracking from full extension to full flexion. The coronal radius of curvature is deeper than that in the previous unconstrained devices, reducing the vulnerability of the implant to subtle component malalignment and possibly allowing the trochlear component to be used with a relatively undamaged native patella (although this has not yet been tested and is not currently recommended).


Figure 3: Journey PFJ implant composed of Oxinium. The trochlear component of the Journey PFJ made in Oxinium. Note that the implant is broad and asymmetric to help capture the patella and aid tracking.

The implant (Figure 3) is composed of Oxinium (Smith & Nephew), an oxidized zirconium metal/ceramic composite, which has been shown to reduce the wear rate of total knee components when combined with polyethylene plastic bearing components under pristine and abraded conditions.²⁰ The underside of the component is abraded to provide cement adhesion, and fixation relies on a convergent 4-peg system similar to that used successfully with the Avon patellofemoral replacement device for the past 10 years. Mid- or long-term results are not yet available for this device, but the early follow-up is encouraging, and the instrumentation and prosthetic design suggest that improved outcomes could be achieved.

Conclusion

Clearly, more work is needed to advance the understanding of the etiology of patellofemoral arthritis and its relationship to patient symptoms. However, the reports presented undermine some traditional beliefs, and perhaps suggest avenues for future study.

Because patellofemoral implants have been used with limited success in the past, some surgeons have seen the proven results of total knee arthroplasty as the only option for treating isolated patellofemoral arthritis. However, advances have been made in the understanding of the disease, and the implants have now reached the third-generation stage (constrained, unconstrained, and now asymmetric unconstrained). A patellofemoral joint implant is a fraction of the size of a total knee replacement, and therefore allows maximal bone preservation, which is vital, especially in younger patients.

References

McAlindon TE, Snow S, Cooper C, Dieppe PA. Radiographic patterns of osteoarthritis of the knee joint in the community. Ann Rheum Dis. 1992; 51:844-849.
Vince AS, Singhania AK, Glasgow MM. What Xrays do we need? a survey of orthopaedic surgeons in the United Kingdom. Knee. 2000; 7:101-104.
Kavadas V, Newman JH. Resource implications of a knee referral. Knee. 2002; 9:281-283.
Utting MR, Davies G, Newman JH. Is anterior knee pain a predisposing factor to patellofemoral osteoarthritis? Knee. 2005; 12:362-365.
Dejour H, Walch G, Neyret P, Adeleine P. Dysplasia of the femoral trochlea [in French]. Rev Chir Orthop. 1990; 76:45-54.
Newman JH, Hendrix M. Trochlear dypslasia: an under recognised cause of patello-femoral arthritis. J Bone Joint Surg Br. 2006; 88(suppl 2):251.
Kulkarni MM, Eldridge JD, Newman JH. Clinical presentation of trochlear dysplasia. Paper presented at British Association Surgery of the Knee (BASK); March 24, 2006; Windsor.
Van Huyssteen AL, Hendrix MR, Barnett AJ, Wakely CJ, Eldridge JE. Cartilage-bone mismatch in the dysplastic trochlear. J Bone Joint Surg Br. 2006; 88:688-691.
Tauro B, Ackroyd CE, Newman JH, Shah NA. The Lubinus patellofemoral arthroplasty: a five to ten year prospective study. J Bone Joint Surg Br. 2001; 83:696-701.
Ackroyd CE, Joslin C, Newman JH. The Avon patello-femoral arthroplasty: The five year results and survivorship. J Bone Joint Surg Br. 2007; 89:310-315.
Clare TD, Newman JH, Ackroyd CE, Evans R. Early results in 110 cases of patello-femoral replacement in patients under 55 years of age. Paper presented at: ESSKA; May 24–27, 2006; Innsbrook.
Ackroyd CE, Smith EJ, Newman JH. Trochlear resurfacing for extensor mechanism instability following patellectomy. Knee. 2004; 11:109-111.
Nicol SG, Loveridge JM, Weale AE, Ackroyd CE, Newman JH. Arthritis progression after patello-femoral joint replacement. Knee. 2006; 13:290-295.
De Cloedt P, Legaye J, Lokietek W. Femoro-patella prosthesis: A retrospective study of 45 consecutive cases with a follow-up of 3-12 years [in French]. Acta Orthrop Belg. 1999; 65:170-175.
Suranimala C, Newman JH. Clinical presentation of trochlear dysplasis. Poster presentation at: British Orthopaedic Association (BOA): September 27-29, 2006; Glasgow.
Hassaballa MA, Porteous A, Newman JH, Rogers CA. Can knees kneel? kneeling ability after total, unicompartmental and patellofemoral knee arthroplasty. Knee. 2003; 10:155-160.
Blazina ME, Fox J, Del Pizzo W, Broukhim B, Ivey F. Patellofemoral replacement. Clin Orthop Relat Res. 2005; 436:3-6.
Shakespeare D, Dikko BA. A simple precise technique for making the anterior cut in patellofemoral resurfacing. Knee. 2005; 12:454-455.
Lonner JH. Patellofemoral arthroplasty: pros, cons and design considerations. Clin Orthop Relat Res. 2004; 428:158-165.
Ries MD, Salehi A, Widding G, Hunter G. Polyethylene wear performance of oxidized zirconium and cobalt-chromium knee components under abrasive conditions. J Bone Joint Surg Am. 2002; 84:129-135.

Author

Dr Newman is from the Avon Orthopedic Centre, Southmead Hospital, Bristol, United Kingdom.

Dr Newman thanks the contributions made by members of the Bristol Knee Group, the Smith & Nephew design team, and Mrs S. Miller, the Bristol Knee Group Coordinator.

Dr Newman is a consultant for and has received compensation from Smith & Nephew.