Although we cannot offer a full explanation for the (apparent) enigma that the preoperative state of the patellofemoral joint has so little long-term predictive power, there are considerations that make it less inexplicable than at first sight.
First, similar lesions to those seen radiographically and intraoperatively in anteromedial OA are common in the joints of most middle-aged and elderly people and, presumably, must be compatible with adequate function. Owre (1936) found flaking and fissuring of some part of the patellar cartilage at necropsy in all but one of 16 subjects aged 60–80 years. Wiles et al. (1956) recorded that nearly all adult patellofemoral joints showed some pathological changes. The medial border of the medial facet was the most frequent site, and severe degeneration was associated with marginal osteophytes. Outerbridge (1961) reported the state of the patellar cartilage during 101 open meniscectomies. He found ‘surface fissuring and fragmentation’ with increasing frequency at each decade in up to 12 of 15 subjects aged 50–69 years. Emery and Meachim (1973) gave a detailed description of the topography of surface degeneration at necropsy. They found fibrillation in almost every knee they examined. In young subjects, degeneration was limited to the articular margins and the medial longitudinal facet of the patella, but in middle-aged subjects fibrillation was seen elsewhere on the patella surface. At these sites it became progressively more common and more severe with increasing age, frequently exposing subchondral bone. The cartilage lesions and marginal osteophytes referred to above were all chance findings at necropsy or at arthrotomy performed for reasons not associated with the patellofemoral joint. Therefore the lesions can be assumed to be generally compatible with adequate patellofemoral function. They are likely to be at least as common in the joints of candidates for unicompartmental replacement as they are in the rest of the middle-aged and elderly population, and to have as little significance.
Figure 5.1 Diagrammatic representation of contact areas on the patella in varying degrees of flexion. (Reproduced with permission and copyright © of the British Editorial Society of Bone and Joint Surgery [Goodfellow JW, Hungerford DS, Zindel M. Patellofemoral joint mechanics and pathology. 1. Functional anatomy of the patellofemoral joint. J Bone Joint Surg [Br] 1976; 58-B: 287–90].)
Secondly, lesions on the medial margin of the patella may have no secondary effect on the rest of the knee joint. There are two unusual features of the medial longitudinal, or ‘odd’, facet. It is the only part of the patella’s surface that articulates with the medial femoral condyle in full flexion (Fig. 5.1(b)) (Goodfellow et al., 1976). In anteromedial OA, the inferior surface of that condyle is devoid of cartilage, and so it is almost inevitable that the odd facet will be secondarily damaged. However, the odd facet only articulates with the femoral condyle, and never with the medial trochlear facet (Fig. 5.1(a)); therefore a lesion on the odd facet has no potential to cause secondary damage to the rest of the patellofemoral joint. After OUKA, the odd facet articulates in full flexion only on the metal prosthetic condyle (Figs. 3.17(c) & 3.18). Therefore its circumstances are similar to those of the retained patella after TKA.
Thirdly, preoperative genu varum tends to overload the medial patellofemoral facets, the most commonly damaged surfaces (Miller et al., 1998). Also, osteophytes at the anterior margin of the erosion on the medial femoral condyle can impinge on the medial facet of the patella in flexion. After OUKA, the intra-articular varus deformity is corrected, tending to unload the medial facet, and the osteophytes are excised during the operation.
It is striking that, after OUKA, problems with the PFJ are very rare (Beard et al., 2007a) whereas after fixed bearing UKA they are relatively common, particularly in the second decade. For example, Argenson et al. (2013) in a series of 160 Miller-Galante UKA with 20 year survival of 74% found that PFJ progression was such an issue in the second decade that they concluded PFJ cartilage loss should be a contraindication to UKA.
In fixed bearing UKA, long term problems with the PFJ may be the result of deteriorating kinematics or impingement. With fixed bearing UKA, the femur tends to wear a ‘divot’ in the polyethylene which affects the kinematics (Ashraf et al., 2004) (see Chapter 2, Fig. 2.19). Argenson et al. (2002) studied the gait of 17 subjects after Miller-Galante medial arthroplasty at 5 – 10 years and demonstrated that about one third of the knees functioned similar to PCL-retaining TKA, as if the ACL was absent. This would dramatically increase the PFJ contact forces (Miller et al., 1998).
In fixed bearing UKA, the femoral component must be mounted on the femur so that its anterior margin is flush with the retained cartilage of the trochlear facet. If this is not achieved, the patella has to negotiate a ridge as it moves distally on the femur. Hernigou and Deschamps (2002) demonstrated that, if such a femoral component is implanted too far anteriorly, the patella can sustain severe damage from impingement in flexion. The technical error may only be revealed by skyline radiographs taken in 90° flexion (Fig. 5.2) and may have been overlooked in the past, with late failure from patellofemoral pain being wrongly attributed to spontaneous progression of degeneration in the patellofemoral joint. Impingement may explain the high incidence of patellofemoral deterioration reported by Berger et al. (2005). Of 49 knees treated by UKA (Miller–Galante) and followed for 10 – 13 years, seven (14%) had radiographic evidence of progressive medial facet patellofemoral joint-space loss. Two of these knees had already been revised (at 7 and 10 years postoperatively) for anterior knee pain attributed to the patellofemoral joint. Of the remaining five, four (one lateral arthroplasty) demonstrated ‘severe patellofemoral joint-space loss secondary to impingement with the femoral component’. These deteriorations occurred despite use of the strict preoperative selection criteria of Kozinn and Scott (1989) which include ‘… only mild radiographic signs of deterioration of the patellofemoral joint’ (Outerbridge grades 1 and 2) and absence of preoperative patellofemoral symptoms.
Figure 5.2 Radiograph, taken 15 years after medial arthroplasty with a polyradial fixed-bearing implant, demonstrating severe erosive changes due to impingement of the femoral component on the patella. (Reprinted with permission from The Journal of Bone and Joint Surgery, Inc [Hernigou P, Deschamps G. Patellar Impingement Following Unicompartmental Arthroplasty. J Bone Joint Surg [Am] 2002; 84-A: 1132–7].)
Unlike polyradial components, the spherical femoral component of the OUKA does not reach the medial trochlear facet of the femur (see Chapter 3, Fig. 3.18 and Appendix, Fig. A10). Therefore, accurate alignment of its surface with that of the articular cartilage is not required as its anterior edge is always buried a few millimetres below the patellofemoral joint line. Furthermore, with the mobile UKA, the kinematics of the knee continues to be normal in the long term so the mechanics of the PFJ will remain normal (Price et al., 2004).
Hernigou and Deschamps (2002) also observed progressive OA of the patella in the second postoperative decade, quite separately from impingement. Joint-space narrowing was most commonly seen in the lateral patellofemoral compartment (Fig. 5.3). It was associated with significant symptoms, and was predicted by lack of joint congruency on the preoperative skyline radiographs. However, the symptoms from this type of degeneration were less severe than those arising from impingement and therefore no revision operation had been performed.
Conclusion
The studies referred to above support the conclusion that the presence of peripheral osteophytes, chondromalacia, fibrillation, and even full-thickness cartilage loss in the patellofemoral compartment need not significantly prejudice the function of the knees of older people. The presence of such lesions in the joints of candidates for UKA is to be expected, and the rarity of revision of OUKA from patellofemoral symptoms suggests that they can be ignored.
We therefore continue to recommend that, in a patient with AMOA, when deciding whether to do an OUKA, the state of the PFJ should be ignored. In particular anterior knee pain, full thickness cartilage loss, or PFJ arthritis seen on skyline radiographs should be ignored. However, we now usually exclude knees in which the lateral side of the patellofemoral joint exhibits bone loss with eburnation and longitudinal grooving, although very few such cases (<1%) have been encountered and we do not have evidence to support this practice.