The Anatomy and Kenematics

of the

 Medial

Patellofemoral Ligament

James L. Baldwin, MD

It has been shown that the medial patellofemoral ligament (MPFL) is responsible for 60% of the medial restraining force of the patella [Desio SM. AJSM 1998;26:59-65] and that it is ruptured in virtually every acute patellar dislocation [Nomura E. The Knee 2002;9:139-143]. The treatment of patellar instability has been hampered by a lack of concensus about the true anatomy of the the MPFL. To clarify the anatomy of the MPFL, 69 human knee specimens were dissected between 2002 and 2004. None of the knees had significant pathology. After the first 19 dissections the study was restarted because of new findings. The last 50 knees comprise the basis for this study.

The dissection is approached through an midline anterior longitudinal incision that is "L"ed proximally and distally. It was found that the skin and subcutaneous tissue are best separated from layer I by blunt dissection with an finger or sponge (Figure 1). In the laboratory this is accomplished nicely by attaching Leahy clamps (Gall Bladder Clamps) to the skin and subcutaneous tissue and applying firm traction to define the interval between the areolar tissue and the investing fascia. Attempting to expose layer I by sharp dissection or from the wrong location will usually violate layer I because it is adhered closely to the subcutaneous fat. When this happens, layer II, the home of the MPFL, can be injured.

It is best to tent and enter layer I posterior to the MCL and distal to the medial epicondyle. A blunt intrument is then used to separate and explore the interval between layer I and II (Figure 2). There are many and varied fibrous conections between layer I and II anterior to the medial collateral ligament (MCL).

Layer I is then tented and split longitudinally behind the MCL and reflected where possible. Note in figure 3 that there is a neurovascular complex separating layer I from layer II. These are the capsular branches of the descending genicular artery which is the last branch of the superficial femoral artery just before it passes through the adductor hiatus (Figure 4).

The first branch of the DGA is the saphenous artery which joins the saphenous nerve and travels behind the sartorius muscle. The DGA then passes down the intermuscular septum, giving off several muscular brances to the VMO and then its capsular branches that come to lie on top of the MPFL and is the key to the dissection. After the capsular branch, the DGA continues on to anastamose with the superior medial geniculate artery which then sends osseous brances to the femur. One of the main brances passes under the origin of the transverse portion of the MPFL, so the MPFL is literally "sandwiched" between two vessels. As long as the vessels are undisturbed the ligament will not be injured.

Before proceding further the distal border of the vastus medial obliquus (VMO) is lifted to expose the sharp upper edge of the MPFL (Figure 5). The ligament then can be traced back its attachment on the femur with a blunt instrument. This allows the ligament to be tensioned, protected and defined.

Following the vessels (figure 6), it is found that the upper vessels lay on top of the transverse portion of the MPFL and the lower branches lay over the oblique portion of the MPFL that arises from the upper MCL. Note that a branch of the superior medial genicualte artery passes under the transverse portion of the MPFL separating it from the medial epicondyle, sandwiching the MPFL between the two vessels.

The configuration seen in Figure 7 was seen in all patients except one. It is thought that the one case in which the decussation was not seen, it was damaged in elevating layer I off of layer II. The origin of the MPFL is dual. The upper transverse portion arises from the bony groove between the medial epicondyle and the adductor tubercle (Figure 8).

It is 12 mm high, strong and ribbonlike. The oblique origin arises from the upper 3 cm of the MCL, is also ribbonlike and equal in strength to the transverse portion of the ligament. The two origins combine anterior to the medial epicondyle and form a broad ligament that is 4 cm wide.

As the MCL approaches the patella it combines inseparably with the tendon of the VMO and together they insert strongly into the most ventral aspect of the medial border of the patella all along its articular surface (Figure 9.).

The length of the MPFL from the groove to the medial border of the patella is 60 mm (Figure 10). Typical anatomy is again seen in this photograph. Layer I has been reflected where possible and layer II is seen and includes the MCL, the MPFL and the medial patellotibial ligament. There is no structure that can be called the medial retinaculum. The medial retinaculum is a general term that should be used to refer to the ligamentous structures on the medial side that stabilize the patella [Starok, M. AJR 1997;168:1493-1499].

When the patella dislocates, the MPFL ruptures at is femoral orgin, its weakest link (Figure 11). The oblique portion of the ligament with its soft origin from the MCL dampens the lateral displacement of the patella and guides the transverse portion of the MPFL back to its origin. The MPFL seldom rupures from its strong attachment to the medial patella. When ossicles are seen on the medial side of the patella after a dislocation has occurred, they are usually avulsion fractures of the inferior medial pole of the patella where the medial patellotibial ligament (in layer III) attaches as opposed to avulsions of the conjoined tendon of the VMO and the MPFL off of the medial border of the patella.

In summary (Figure 12), the MPFL was present in all specimens. It has a dual origin. The transverse portion arising from the grove between the adductor tubercle and the medial epicondyle and an oblique portion arising from the upper edge of the MCL. Together they form a broad ligament which combines inseparably with the tendon of the VMO and together they insert strongly into the medial border of the patella all along it articular surface. The MPFL is entirely in Layer II and is defined by vessels that lay on its surface and under its origin. This percise anatomical information will aid the orthopedic surgeon in diagnosis and treatment of patellar instability.

POSTERIOR HORN DETACHMENT OF THE MEDIAL MENSICUS

Posterior Detachment of the Medial Meniscus (PMD) is a meniscal tear not formerly described in the literature. This condition was first recognized at The Portland Knee Clinic in 1984 as a disruption of the extreme posterior attachment (ligamentous) of the medial meniscus (see Fig. 1). It tends to occur in patients after they enter the six decade (above 50 years) but has been observed in younger patients. It occurs occasionally in the lateral meniscus, but the preponderance of cases involve the medial meniscus. The posterior horn of the medial meniscus is the area of maximal stress during most activities and gradual weakening occurs at the posterior horn attachment in 7% of patients with medial meniscus injuries in stable knees. This is an attritional lesion. The knee is usually asymptomatic until the last fibers rupture, then, a seemingly minor event, such as stepping on to an escalator, arising from a chair or applying the car brakes can cause the final injury. When the posterior moorings finally go, the meniscus slides out of the joint with compressive load and the meniscus becomes almost totally non-functional (See Fig. 2). The medial compartment is instantly overloaded by forces usually absorbed by the hoop structure of the medial meniscus and severe medial pain develops. Frequently the patients come into the office on crutches. They are tender medially but frequently don't have an effusion or loss of motion. Plain standing x-rays are usually normal (Fig 3.) because this is an early lesion. The MRI is also usually normal because this area is very hard to image. Sometimes on the axial views the lesion will be seen if the cut is just right. It can also be seen on the posterior coronal views occasionally by looking carefully at the medial meniscus right next to the PCL. If a bone scan is obtained, frequently one will see extreme uptake on the tibial side due to overloading. Prior to MRI it was impossible to distinguish this condition from the early stages of spontaneous osteonecrosis because of the uptake on the bone scan. However, SONC usually involves the femur, whereas the uptake with PMD usually occurs on the tibial side (Figure 4.). The only way to be absolutely sure about the diagnosis is by arthroscopy.
Because the posterior attachment of the medial meniscus curves around into the saddle of for the PCL and because it is so far back it is usually missed by the arthroscopist. This is the reason there has been no description of this lesion in the literature. Sometimes the extreme posterior horn has granulation tissue which has to be cleared. If the knee is tight it is more difficult to see posteriorly. The key is a good leg holder (Such as the Johnson) and careful probing. With careful probing the meniscus will be found to elevate too high off the tibia and it can be moved medially. When this is done, the disruption will be seen (Figure 5.).
Several procedures have been developed to repair this lesion. One is with "T- fix" sutures, first passed through the posterior horn from the medial portal, then pulled through a bony tunnel much like the posterior anchor for a meniscal transplant (Fig 6.). Larger shoulder "T-fix" sutures provide solid fixation (Fig 7.). A blood clot has been used to aid healing. This procedure is not too difficult but has been only about 50% successful. This is probably because the tissue quality is poor and there is extreme force in this area.
The second procedure is to attach the posterior horn to the PCL with "FastFix" sutures.
The third treatment is to roughen up the area, immobilize the knee and put the patient on crutches for 6 weeks.
Since this lesion is so common, the diagnosis is easy to make and in older patients, after full H&P, normal standing films, normal MRI and hot bone scan the patient can be put on crutches or a walker for six weeks and they frequently get better without surgery.
This lesion is the most common cause of a failed arthroscopy for medial pain or "tear of the medial meniscus".
About 50% of this patient group decompensate quickly and go on to a unicompartmental or total knee replacement.
A large number of total knee arthroplasties (TKA) are done at the Portland Knee Clinic and since becoming aware of this condition we have observed a high incidence of disruptions of the posterior horn in patients with medial osteoarthritis who are undergoing TKA. It takes a moment to assess the status of the meniscus before taking it out but the TKA surgeon will be surprised of how commonly it occurs. The body of the meniscus will look normal. There is usually an attempt by the natural repair process to heal this lesion and various levels of ineffective fibrous healing will be present. On close inspection it is noted that the posterior horn of the medial meniscus can easily be lifted off the tibia and displaced medially.
If an effective repair can be developed for PMD, possibly the incidence of medial osteoarthritis can be reduced.

2003 

Medial Subluxation of the Patella after Lateral Release

Lateral Retinacular Release has been in vogue for over fifty years as the treatment for patellofemoral problems associated with abnormal tightness of the structures that hold the patella ( knee cap) in place on the lateral side. The goal of this procedure is deminish the taughtness of these retaining structures whose function is to hold the patella in place and provide static control of tracking. Tightness of the lateral retinacular structures can lead to increased pressure in the lateral aspect of the patellofemoral joint and can cause a wide spectrum of problems from anterior knee pain to patellar dislocation. The former is a more benign form of the problem and the only symptom may be anterior pain (usually anterolaterally) with bent knee activities and the only finding a tight lateral retinacular ligament. More severe forms of the problem are associated with pain with a broad spectrum of activities and findings which may include patellar grinding (crepitation), decreased medial patellar excursion, lateral tracking of the patella and patellar tilting. In patients with associated genu valgus (knock knee), increased Q angle (angle between the vectors of the thigh muscle and the patellar tendon) and various degrees of extensor mechanism dysplasia (maldevelopment) the patella may actually dislocate. Lateral release alone is indicated in many of these circumstances except for those that involve dislocation. When dislocation is involved, lateral release plus more major realignment reconstruction (e.g. Trillat procedure) is indicated. If patients develop anterior knee pain and are diagnosed with tightness of the lateral retinaculum and do not improve with conservative treatment (exercises, medication, bracing, shoe devices) over a 12 to 18 month interval, then arthroscopic lateral release is indicated. At the time other minor procedures (such as smoothing the patella) can be done if needed.
When arthroscopic lateral release was first developed in the late 1970s and early 1980s some pioneers of the procedure recommended that it be done blindly with a pair of surgical scissors starting inferolaterally after enlarging the inferolateral arthroscopic portal. All of the lateral structures were severed until the patella could be stood on its side (everted), 90 degrees to its usual position. Some times this included cutting deeply into the vastus lateralis muscle. This technique turned out to be excessive in many cases and reports of medial subluxation (medial slippage) began to appear in the literature (e.g. Nonweiler & DeLee. AmJSportsMed 1994, 22[5]:680-686). With excessive lateral release, patients began to experience a spectrum of symptoms related to overpull of the medial structures: from medial patellofemoral pain to frank medial dislocation (See figure 1, right knee, medial dislocation). Over time procedures to reconstruct the lateral retinaculum to restore balance were developed ( e.g. Teitge. Arthroscopy 2004, 20[9]:998-1002). Procedures involving spaning a ligamentous structure from the lateral border of the patella to the lateral femoral condyle were found to be non-anatomic. Since the majority of the normal lateral retinacular ligament passes from the Iliotibial Band (ITB) to the lateral border of the bony patella, it seemed logical to recreate this anatomy rather than a bone to bone procedure. A procedure was developed at The Portland Knee Clinic to restore this normal anatomy by attaching a band of ligamentous tissue from the patella to the ITB, a surgery that allows the lateral forces to be modulated by the ITB, restoring stability but avoiding overtightening and recurrence of pain. The most suitable tissue for this reconstruction was found to be doubled fresh frozen iliotibial band. It is sewn to the retinaculum and periosteum of the anterior patella (2.5 cm wide) and to the ITB (4 cm wide) in a trapezoidal construct as shown in the attached photo (figure 2). This has been quite sucessful in a number of cases and has been used to rebalance the extensor mechanism in other situations in which the lateral retinaculum has been found to be deficient. In the case shown here the synovium was plicated, the VLO repaired and a Maquet performed in addition to the LPFL reconstruction.
It is now understood that arthroscopic lateral release is carried out under direct vision with electrocautery from the inferior border of the vastus lateralis tendon proximally, through the lateral meniscopatellar ligament (Ficat & Hungerford. Disorders of the Patello-femoral Joint. p. 17. Williams and Wilkins 1977) distally. This seldom creates problems unless the lateral release is done in the presence of normal patellofemoral soft tissue balance.