The patellofemoral joint (PFJ) is a highly complex structure, intimately associated with the dynamic muscle activity, including the global and central control of the limb, which is linked to the pelvis and the spine. The function of the patella is to improve the effective extension capacity of the quadriceps muscle by increasing the moment arm of the patellar tendon. To fulfil this feature, its osseous and soft tissue static and dynamic constraints have to be in perfect balance, thereby conferring a great stability on the PFJ. However, PFJ articulation has a low degree of congruency and anatomical anomalies are not unusual.
The clinical presentation of patellofemoral (PF) instability contains a spectrum of manifestations that can be subtle and very difficult to correctly diagnose and treat. The cornerstone of treatment for this kind of disorder is to obtain an accurate medical history to differentiate patients who present with symptoms of pain, subjective instability (caused by quadriceps inhibition reflex as the result of pain and fear but without loss of contact of the articular surfaces) or pseudolocking (produced by a reflex mechanism of quadriceps and hamstrings contracture because of pain to avoid the contact of the opposed damaged cartilage surfaces) from those who have suffered true episodes of patellar dislocation. Except for cases caused by high-energy trauma, true patellar dislocations are almost always the result of predisposing anatomical anomalies, which eventually lead to recurrent lateral patellar dislocation.
In 1987 the Journées Lyonnaises de Chirurgie du Genou, edited by Henri Dejour and Gilles Walch, set the basic principles and diagnostic rationale on patellar instability, which remain in use today. Their work was the result of a joined effort from the entire Lyon School, who analysed more than 1800 cases of patellofemoral pathological conditions, ranging from painful patellar syndrome to patellar dislocation, patellofemoral arthritis and failures of patellar surgery. By comparing 190 control patients with no orthopaedic antecedents and 147 patients with documented episodes of patellar dislocation, they described four major anatomical factors leading to instability: trochlear dysplasia (present in 96% of the dislocation group), abnormal patellar height (in 30% of patients with patellar instability), a high tibial tubercle to trochlear groove (TT-TG) distance (in 56% of the dislocation group) and high patellar tilt (in 83% of patellar dislocation group) ( Fig. 27.1 ).
In addition to these major predisposing anomalies, secondary instability factors have been identified:
Knee valgus or recurvatum
Increased femoral rotation
Increased external tibial torsion
Excessive subtalar joint pronation
Hip abductor, external rotator and extensor weakness associated with increased hip adduction and internal rotation
Hypoplasia of vastus medialis oblique (VMO) or imbalance of neuromuscular activation and coordination between VMO and vastus lateralis
These secondary instability factors tend to increase the valgus orientation of the extensor mechanism (the so-called Q-angle ), which in turn increases the lateral direct vector forces applied on the patella. However, many of these factors are already considered in the TT-TG ( Fig. 27.2 ) distance evaluation that encompasses the valgus and torsional alignment of the extensor mechanism. Furthermore, they are considered of secondary surgical importance because of insufficient literature reporting on the pathological threshold that leads to surgical treatment.
Another important factor that must be considered is the medial patellofemoral ligament (MPFL) disruption and the consequent insufficiency of the medial restraints. Despite the possible anatomical abnormalities and the consequent overall intrinsic instability that may be present between the trochlea and the patella, dislocations cannot occur if the MPFL is competent.
In regard to the clinical presentation, according to H. Dejour and Walch, patients with patellofemoral pathological conditions associated with instability may be classified in three major groups:
Objective patellar instability (OPI) or objective patellar dislocation (OPD): This group is composed of patients who have had at least one episode of true patellar dislocation, defined as the complete loss of contact between patellar and femoral trochlear articular surfaces, associated with a hemarthrosis in the inaugural instability episode. It is of paramount importance, especially in case of children and adolescents, to actually have a documentation of the dislocation episode or a confirmation received by the patient as the physician shows an actual patellar dislocation; this occurs because a lot of episodes of patellar subluxations may be perceived or described as true dislocation, causing further confusion. The loss of contact may be spontaneously reduced or irreversibly fixed in a dislocated position that needs immediate medical care. This is also extremely important to record; in fact a patella that reduces easily and spontaneously with no pain usually implies the presence of anatomical anomalies that affect the normal restraint of the patella, whereas a patellar dislocation that is difficult to be reduced usually happens over normal PFJ anatomy. These patients will always present at least one of the previously mentioned anatomical abnormalities. Pure traumatic dislocations with no significant underlying abnormalities are excluded from this group.
Potential patellar instability (PPI) or potential patellar dislocation (PPD): These patients typically complain of knee pain or instability, reporting a rather generic feeling of the knee ‘giving way’ but without episodes of true dislocation or hemarthrosis. Symptoms often occur during daily living and low-energy activities such as walking, climbing up and down stairs or rising from a seated position. Patients in this group present at least one instability factor but they do not report documented patellar dislocations in their medical history. This is often the contralateral knee of a patient with OPI.
Painful patellar syndrome (PPS): These patients complain about knee pain or instability but have no objective anatomical abnormalities and no history of subluxation. These patients actually do not belong to the instability spectrum.
Patellofemoral instability may be further classified according to the frequency of patellar dislocations:
Recurrent patellar dislocation: The patella dislocates frequently during knee flexion (more than three episodes are required).
Habitual patellar dislocation: The patella dislocates during early knee flexion (<30 degrees) every time the knee flexes, or the patella relocates after early knee flexion.
Permanent patellar dislocation: The patella is always dislocated throughout normal knee range of motion and never articulates with the femoral trochlea.
The main principle of the ‘à la carte’ treatment is to list and rate the risk factors and their association and then propose an adapted surgical treatment to each anatomical abnormality, which will be combined according to the number of factors to correct. Each risk factor with its imaging presentation and the correspondent surgical treatment that could be addressed is discussed in the following sections.
The femoral trochlea shape becomes flat or convex instead of concave and leads to the loss of the osseous guide to patellar tracking. It is the single factor most associated with patellar instability, present in up to 96% of patients with OPI and only in 3% of the control group in Dejour et al.’s series. The normal trochlear constraint to lateral patellar displacement is lost and as a consequence dislocations can occur. Trochlear dysplasia (TD) seems to be an inherited disorder.
The initial evaluation of the TD has to be done by examining a true lateral x-ray view in monopodal weightbearing with an angle of knee flexion between 15 and 20 degrees. The Blumensaat line is followed anteriorly by the trochlear groove line. In normal knees this line remains posterior to the condylar lines, meaning that the trochlea is deep and congruent. Henri Dejour and Gilles Walch described the ‘crossing sign’ as pathognomonic for TD. It represents the exact location where the deepest point of the trochlear groove reaches the same height as the femoral condyles, meaning in that location the trochlea becomes flat. They also suggested that the first classification of TD with three different types depended on the level of the crossing sign; the lower the crossing sign, the higher the grade of TD. In particular, in type I, the medial and lateral condylar lines cross the sulcus line in the same location in the proximal part of the trochlea; in type II the sulcus line crosses the medial condylar line lower than the lateral one; in type III (the most severe) the condylar lines are symmetrical but the crossing sign is located in the distal part of the trochlea, meaning that the majority of the trochlea is flat (or convex).
Other signs not included in the classification were described: the ‘trochlear bump’ and the depth of the trochlea. In a study performed by H. Dejour et al. it was found that, in knees with TD, the trochlear groove line position is abnormal in relation to the projection of the anterior femoral cortex. The trochlear bump is the distance between the most anterior point of the sulcus and the line tangent to the last 10 cm of the anterior femoral cortex; the measure is positive if the sulcus is anterior and negative if it is posterior to this line. In normal knees the mean distance was –0.8 mm, whereas in knees with dysplastic trochleae the mean value was +3.2 mm. This prominence results in increasing the contact force between the patella and the trochlea with an ‘anti-Maquet’ effect. Because values greater than 3 mm were found in 66% of knees with patellar instability and in just 6.5% of normal knees, 3 mm was taken as the threshold. The depth of the trochlea was measured by the distance from the anterior contours of the condyles to the floor of the sulcus through a line drawn. Because values less than 4 mm were found in 85% of knees with patellar instability and in just 3% of normal knees, 4 mm was taken as the threshold.
This first classification had limitations in its clinical use; in particular, interobserver reproducibility of trochlear analysis was low, especially for type II dysplasia, as shown by the work of François Gougeon, Frank Rémy and Henry Migaud from Lille, France.
This led to a new study performed in 1996 by D. Dejour, Reynaud and Le Coultre, who analysed 177 cases of patellar instability using radiography and pre- and postoperative computed tomography (CT) scans. Two new radiographic signs were added to the crossing sign: the ‘supratrochlear spur’ ( Fig. 27.3 ), which represents a global prominence of the trochlea and plays the role of a ‘ski jump’ while the patella is engaging the trochlea, and the ‘double contour’ sign ( Fig. 27.4 ), which is the radiographic line ending below the crossing sign representing the subchondral condensation of the hypoplastic medial facet on the lateral view. They studied TD in two planes and defined a new and more precise classification with four grades of TD.
This classification is based on the radiographic lateral view combined to slice imaging (like CT scan or magnetic resonance imaging (MRI) axial views) which may assist in the differentiation between types. Based on the three dysplastic signs, four types are described:
Type A (54%): the crossing sign is the only of the three signs present. On axial views the trochlea is still symmetrical, but shallower than normal ones.
Type B (17%): The crossing sign and the supratrochlear spur are present. On axial views the trochlea is flat.
Type C (9%): The crossing sign and the double contour sign are present, but there is no spur. On axial views the lateral facet is convex and the medial facet is hypoplastic.
Type D (11%): All three signs are present: crossing sign, supratrochlear spur and double contour. On axial views there is a ‘cliff pattern’ (asymmetry of the trochlea with a vertical link between lateral and medial facet) ( Fig. 27.5 ).
The same team from Lille tested the inter- and intraobserver agreement of this new classification and concluded that ‘this new classification system is more reproducible than the former 3-type system proposed. The crossing sign and the supratrochlear spur are the most reproducible signs’. This classification was subsequently accepted by the International Patellofemoral Study Group (IPSG) and later validated by other authors. ,
Radiographic axial views can also be evaluated to measure the sulcus angle (defined by Brattstroem ) in the Merchant view; the average sulcus angle measures 138 degrees (standard deviation (SD) ± 6) and is equal in men and women. Values greater than 145 degrees are considered abnormal.
On MRI axial views 3 cm above the joint line, trochlear depth may be measured: The anteroposterior distance between the deepest point of the trochlear groove and the line paralleling the posterior outlines of the femoral condyles is subtracted from the mean of the maximal anteroposterior distance of the medial and lateral condyles from the same posterior line. In the study by Pfirrmann et al., if a threshold is set at 3 mm, a sensitivity of 100% and specificity of 96% could be expected. In the same study, a facet ratio (medial facet/lateral facet) inferior to 40% provided a similar specificity and sensitivity to dysplasia diagnosis.
Carrillon et al. investigated the lateral trochlear inclination (LTI) angle, formed by a line tangential to the subchondral bone of the posterior aspect of the two femoral condyles and a line tangent to the subchondral bone of the lateral trochlear facet. A significant difference between healthy and patellar instability patients was recorded. The mean value in patellar instability patients was 6.17 degrees; in the control group it was 16.9 degrees. When 11 degrees was chosen as the threshold value for LTI, results were excellent in discriminating between the two groups, with a sensitivity of 93%, specificity of 87% and accuracy of 90%.
Biedert et al. proposed the lateral condyle index (LCI), which identifies the length of the lateral trochlear cartilage. This index is measured on sagittal MRI views, with the knee extended, the foot 15 degrees externally rotated and the quadriceps muscle consciously relaxed. First, in the slice showing the entire length of the anterior cruciate ligament (ACL), the femoral shaft axis is determined (drawing two circles and connecting the centre of them). Second, the most lateral image that still shows cartilage of the lateral condyle is used. A tangent line (d) on the distal femoral cartilage is drawn at 90 degrees to the femoral axis (Ca). The length (a and p) to the most anterior (A) and the most posterior (P) aspect of the cartilaginous part is measured in relation to the tangent line. Its ratio ((a:p) × 100) represents the LCI in percentages. Twenty-eight knees in 23 patients with patellar subluxation and 46 controls were analysed. Their mean index values were 86% ± 9% and 93% ± 7%, respectively. The authors concluded that values of LCI 90% or less had a sensitivity of 79% and a specificity of 37% for a lateral cartilaginous condyle that is too short. With this finding, the authors suggested a new form of trochlear dysplasia, a trochlea that is too short, another potential contributor to patellar instability.
Trochleoplasty is a technically challenging surgical procedure that helps reshape the distal femoral articular surface, in particular the trochlea. The goal of trochleoplasty is quite ambitious because this surgery changes what the natural evolution has done during bone growth. This technique involves working directly on the PFJ with a theoretical risk of cartilage damage, modification of the congruency between the two articular surfaces and alteration of joint kinematics. The reason for this drastic surgical choice is to achieve congruency between the patellar and trochlear surfaces in patients with recurrent patellar dislocation for whom it would otherwise not be possible to achieve stability with any other treatment. Continuous stability of the PFJ throughout the whole knee range of motion is of absolute importance for the normal function of the extensor mechanism, human erect stance and bipedalism.
Through the years many trochleoplasty techniques have been described. The ‘sulcus-deepening trochleoplasty’ was first described by Bilton Pollard in 1891, then by Douglas Drew in 1908 and Masse in 1978; modified and formalised by H. Dejour et al. in 1990; and finally described in 2010 by D. Dejour and Saggin, who introduced the lateralisation of the groove as a proximal realignment and the combination of soft tissue procedures for the treatment of recurrent patellar dislocation in patients with underlying high-grade TD (types B and D). The rationale of this surgical procedure is to restore the normal anatomy and to reshape the trochlea by undermining the cancellous bone and deepening the groove. The amount of bone removal is determined to have a new trochlear groove flushed with the anterior femoral cortex and to make the prominence disappear. The trochlea is osteotomised with a scalpel carefully over the position of the new trochlear groove. In this technique the new trochlear groove is positioned according to the preoperative TT-TG value in a more lateral position to reduce an excessive TT-TG value within normal limits. This had been previously shown in 2005 by Schöttle et al., who demonstrated a 10-mm TT-TG reduction in their series after trochleoplasty. This TT-TG reduction may alleviate the need for an additional distal procedure (tibial tuberosity medialisation). The important points of this technique are that a thick osseocartilaginous flap is created like an osteotomy, enough bone can be removed to make the prominence disappear, the sulcus angle is decreased and additionally a proximal realignment procedure is performed by lateralising the trochlear groove. The ideal indication for trochleoplasty is a patient with true and documented patellar dislocations with underlying high-grade (types B and D) TD in whom the patella cannot override the severe trochlear bony prominence during early flexion. Patellofemoral arthritis, skeletally immature patients and isolated patellofemoral pain are contraindications. Trochleoplasty procedures are always combined with other soft tissue procedures (MPFL reconstruction is the intervention recommended by the author) and possibly with bony procedures (tibial tubercle transposition in most cases), when this is needed by the preoperative imaging measurements.
In 1994 Bereiter and Gautier presented another type of trochleoplasty, known as the Bereiter or thin-flap technique . In this procedure a 3- to 5-mm osteochondral flap is first elevated from the whole of the trochlea. Then the underlying cancellous bone is burred and deepened until the prominence disappears, the groove is flush with the anterior femoral cortex and a V-shaped groove is fashioned on the cancellous bone. The osteochondral flap is depressed on the newly shaped bone and is fixed with sutures throughout its perimeter. In 2010 Blønd and Schöttle performed the Bereiter trochleoplasty by arthroscopy: While working through suprapatellar portals, they elevated a thin osteochondral trochlear flake, undermined the cancellous bone beneath it with the use of an arthroscopic shaver and burs and fixed the flap with the use of anchors and absorbable sutures.
The American surgeon Fred H. Albee in 1915 considered that TD was the result of a reduced lateral facet height and that the abnormal flat surface of the trochlea was mostly because of a depressed lateral facet rather than an elevated trochlea in the midline. Therefore he proposed his lateral facet–elevating trochleoplasty in which the lateral facet was osteotomised and the flap created was elevated anteriorly enough to create a mechanical block for pathological lateral patellar translation. This technique is rarely used because of the increased patellofemoral contact forces, which can lead to PF osteoarthritis and pain, but it remains a theoretical indication in cases of TD type C in which a trochlear prominence (‘spur’) is not present. R. Biedert still uses this technique in cases of short lateral trochlea.
In normal knees, patellar engagement with the trochlea occurs at around 20 degrees of knee flexion. Patella alta refers to an abnormally high-riding patella that engages its osseous restraint to dislocation (the trochlear groove) later in flexion, increasing the patellar ‘free’ arch of movement and predisposing to its lateral translation and eventually dislocation. This feature was noticed in 30% of the patients in the dislocation group and in 0% of control group patients in the study by Dejour et al.
Several methods and anatomical references have been used examining radiographic lateral views to quantify the patella height. These are discussed next.
The Blumensaat index was the first published, in 1938, and required measurement with the knee at 30 degrees of flexion, stating that the inferior part of the patella should stay on the projection of the Blumensaat line.
The Insall-Salvati index is the ratio between the length of the patellar tendon and the longest sagittal diameter of the patella. They also investigated the height of the insertion of the patellar tendon from the tibial plateau because a big variation of this parameter would have undermined the value of this index, but they found it constant. It has been, however, criticised for its inability to determine the patella height in patients with a history of Osgood-Schlatter disease or patients who have undergone a tibial tubercle transposition. This ratio is normally 1; a value less than 0.8 indicates patella infera; if greater than 1.2, it indicates patella alta.
The Insall-Salvati modified index was described by Grelsamer and Meadows; they said that the original index did not take into account the possible variations of the distal extraarticular part of the patella. To overcome this issue, they took the ratio between the distance between the tibial tuberosity and the inferior edge of the patellar articular surface and the articular surface of the patella. A ratio greater than 2 indicates patella alta.
The Blackburne-Peel index is the ratio between the length of the perpendicular line drawn from the tangent to the tibial plateau to the inferior pole of the articular surface of the patella and the length of the articular surface of the patella. The authors tried to overcome the precedent ratios’ drawbacks, but this index was also criticised because of the technical issues in perfectly superimposing the medial and lateral tibial plateaus.
Finally, the Caton-Deschamps index , ( Fig. 27.6 ) seems the easiest way to evaluate the patellar height and is the author’s preferred measuring technique, especially for surgical planning. It is the ratio between the distance from the lower edge of the patellar articular surface to the anterosuperior angle of the tibia outline and the length of the articular surface of the patella. Patella alta is diagnosed when the ratio is greater than 1.2, whereas a ratio less than 0.6 indicates patella infera. This method eliminated factors related to patellar inferior pole abnormalities or troubles in identifying the exact location of the distal patellar tendon attachment.