chapter 9:Patellar Subluxation

Radiologic Features

Standard tangential radiographs may readily demonstrate subluxation, and either the Laurin 20‑degree axial view or the Merchant view, as described in Chapter 4, may show subluxation, particularly if it is more advanced. Some patients with normal tangential (axial) radiographs, however, have significant subluxation that may be detected using CT. Also, there are patients who appear to have subluxation on an axial view, because of image overlap, but CT will not reveal actual subluxation! Flexion of the knee can draw the unstable patella deep into the femoral trochlea so that subtle subluxation problems may be missed. CT may show signs of a patella that is slow to centralize in the trochlea, with notable subluxation on the 10‑ and 20‑degree knee flexion tomographic cuts. This may be important information in the patient who has failed conservative treatment, because such patients may have considerable functional instability with athletic activities despite normal standard radiographs at 45 degrees of knee flexion. CT can be very helpful also in determining if there is tilt associated with subluxation.


Radiographically, there are patients with lateral facet predominance (Wiberg Type III, alpine hunter's cap deformity) and tilt in association with subluxation (Fig. 9.8). Developmentally, a patella that is laterally aligned will form in response to molding pressures such that the result may be lateral facet predominance. It is more likely that this form of dysplasia is secondary, rather than a cause of patellar degeneration. The lateral facet predominance noted in some patients with patellar malalignment is probably analogous to femoral head dysplasia in congenital dislocation of the hip.


The lateral trochlear facet is frequently flat in patients with subluxation. However, such is not always the case. CT with tomographic cuts centered on the midtransverse patella at progressively increasing knee flexion angles will provide a good indication of trochlear morphology in the functional range of motion. Many patients with patellar instability, however, will show no gross evidence of trochlear dysplasia.


Deviation of the patella from its normal femoral groove may be seen on standard tangential (axial) radiographs. We have only rarely observed lateral subluxation on the 60-degree flexion film when the patella was centered on the 20‑ and 40‑degree flexion views. Instability is usually manifest in the first 30 degrees of flexion, which indicates the value of the 20‑degree flexion view and CT in early knee flexion. Those who fail to obtain these studies will undoubtedly miss some significant subluxations. It is unfortunate that an axial radiogram cannot be taken in less flexion, but this is the technical limit. CT is necessary to appreciate fully the details of patellar tracking and to avoid image overlap. Evidence of subluxation can be identified using CT between 0 and 30 degrees as the patella is centering in the trochlear groove. These images may show one of several possibilities: (1) a slight lateral subluxation on the 20‑ to 30‑degree flexion view, with perfect recentering of the patella on further flexion (Figs. 9.9 and 9.10); (2) Type II lateral subluxation (with associated tilt) (Fig. 9.11); (3) a patella that sits astride the lateral trochlear facet (Fig. 9.12), but recenters on further flexion; and (4) permanent lateral subluxation.

In addition, and perhaps most important, one can determine most accurately with CT whether there is tilt associated with subluxation. Using the posterior condyles to determine tilt relative to the lateral facet is possible only with tomography. Although CT is not necessary for all patients with subluxation, it may be very helpful in surgical planning. Magnetic resonance imaging (MRI) has not yet proved any more helpful than CT in this regard, although serial slices of the patella may help in understanding the condition of the articular cartilage. It is our opinion, however, that MRI is unnecessary in the majority of patellofemoral pain patients, and knee flexion beyond 20 degrees is currently a technical problem with MRI. CT, on the other hand, is frequently helpful in differentiating tilt and subluxation, and offers alignment data comparable to that available with MRI but with ability to obtain images in greater degrees of knee flexion.


This type of calcification is noted particularly on the axial views and is quite characteristic of recurrent subluxation (Fig. 9.13). Sometimes these calcifications are well defined, appearing as a discrete medial osteophyte, and need to be carefully distinguished from an artifact. Trillat et al (3) compared them quite correctly to Pellegrini‑Stieda disease of the knee in that they are the soft‑tissue reaction to repetitive ligamentous trauma. They represent a post‑traumatic soft‑tissue calcification and are frequently seen, perhaps even more frequently than with frank dislocations. These calcifications are differentiated from marginal fractures in that the medial border of the medial facet is intact. Moreover, they are covered with synovial membrane and are, therefore, extra‑articular.


Medial subluxation of the patella (15, 16) is caused most commonly by excessive medialization of the extensor mechanism in realignment surgery, or lateral retinacular release in which there was little or no patellar lateralization preoperatively. Rarely, this may occur even after lateral retinacular release in a patient with preoperative patellar lateralization or excessive tilt. In such cases, reattachment of the vastus lateralis obliquus may re‑establish normal patellar tracking. In patients with hypermobile joints, however, particularly if the trochlear groove is shallow, achieving "balance" of the patella in the trochlea may be extremely difficult.

Medial subluxation as an isolated entity, without previous surgery, is extremely rare or non‑existent, in the author's opinion. Radiographic studies (17) have shown that tilting of the patella can simulate medial subluxation by causing rotation of the central ridge of the patella in a medial direction. In an asymptomatic volunteer group, the lower limit for congruence angle was ‑26 degrees by computerized tomography. Consequently, one should exercise extreme caution in making a diagnosis of medial patellar subluxation. Dupont (18) has described a variant in which the patella is lateral with the knee extended and medial with the knee in flexion. In my opinion, medial subluxation is a clinical condition that will be best diagnosed on physical examination.

A clinical test for medial patellar subluxation involves holding the patella in a slightly medial direction with the knee in extension, and then flexing the knee to see if the patient's symptom is reproduced with a sudden entry of the patella into the trochlea from medial to lateral. In patients with symptomatic medial subluxation of the patella, this simple maneuver will usually cause considerable discomfort and reproduction of the painful instability sensation.



Inside Chapter 9: