This article is more than 5 years old. Information may no longer be current.
MPFL reconstruction for patellar instability in younger patients: Save it for the recurrent dislocators
Click Here to Manage Email Alerts
Introduction
Patellar instability is a common problem, especially in young athletes and dancers. Nonoperative treatment often results in recurrent dislocations and diminished knee function. However, surgical treatments have associated morbidity and are not guaranteed to relieve symptoms. Many surgical techniques, both proximal and distal, have been used with varying success and postoperative morbidity. It has become more apparent recently that in many cases the primary lesion causing symptomatic patellar instability is injury to the medial patella-femoral ligament (MPFL). Surgery to reconstruct this ligament has gained popularity and is the focus of this Round Table discussion.
Peter R. Kurzweil, MD
Moderator
|
||||||||||
Peter R. Kurzweil, MD
Robert T. Burks, MD
Donald C. Fithian, MD
Andrew J. Cosgarea, MD
Robert A. Teitge, MDPeter R. Kurzweil, MD: What physical examination signs and symptoms do you see in patients that may lead you to recommend an MPFL reconstruction?
Andrew J. Cosgarea, MD: The indication for MPFL reconstruction is recurrent instability. The incidence of recurrent instability after a first time dislocation is less than 50% and the literature generally does not support initial stabilization. There are, however, some patients who may benefit from nonreconstructive procedures. I usually recommend surgery in the acut setting only when a large loose body, or other significant intraarticular pathology, such as a meniscal tear, is identified. Occasionally athletes will sustain a concomitant high-grade medial collateral ligament (MCL) tear that may also involve the vastus medialis obliquus (VMO). In this case a strong argument can be made for an anatomic surgical repair of the torn structures (including the MPFL) from the femoral avulsion site. Another indication for surgery would be in patients who have persistent lateral subluxation that doesn’t correct with rehabilitation. This is determined clinically and from serial axial, “sunrise” or Merchant, views in the weeks after injury.
Robert A. Teitge, MD: I believe MPFL reconstruction is an operation to treat recurrent dislocation of the patella — not for use in a primary dislocation. Thus I reserve the procedure for patients with symptomatic recurrent dislocation of the patella. Recurrent dislocation is likely present if on examination: the patient has extreme apprehension when a lateral force is placed on the medial side of the patella with the knee in 20° to 40° of flexion; the patella dislocates when this lateral stress is applied; the patella dislocates when the patient is weight-bearing on the injured limb with the knee flexed 20° to 40° and an internal twist is exerted, or when there is a complete dislocation on stress radiographs. With these findings and a history consistent with a recurrent patellar dislocation, MPFL reconstruction is likely beneficial.
I have some additional comments regarding the indications. When a patient is seen in the office with a first time patellofemoral (PF) dislocation I attempt to answer four questions:
- Was this really a dislocation of the patella or did it simply sublux?
The diagnosis of joint instability requires an examination in which a
displacement force is applied and an abnormal increase in displacement is
noted. For example, it is well known that with a torn ACL the exam reveals
positive Lachman, anterior drawer and pivot shift tests. Verification is
supplied with a KT-1000 difference or with stress radiographs. MRI may indicate
injury, but does not prove instability. To verify a lateral patellar
dislocation a lateral force is applied to the medial border of the patella and
the amount of displacement is noted. The lateral displacement is compared with
the normal or asymptomatic knee subjected to the same stress examination. A
positive Fairbank’s sign of genuine apprehension is usually present and
pain with this maneuver is suggestive of injury, but not necessarily
instability.
Measurement of the increased excursion to verify the instability is best accomplished with stress radiographs. This is an easy test, which is performed by positioning the patient for an axial view of the patella (as in a Merchant view) with knee flexed to 30° to 40°. Apply stress to the medial side of the patella, which displaces the patella laterally, and shoot the axial roentgenogram. Repeat with the normal side and compare the amount of lateral displacement. More than 4 mm of increased excursion on the injured side was shown in one study to correlate with symptomatic, recurrent patellar dislocation. The estimation of medial and lateral quadrant glide has been shown in one study to be an unreliable measure. A radiograph of the dislocated patella gives one confidence in: if the dislocation occurred, what structures were injured and how to treat the injured patient and their injured structures. - Why did this dislocation occur? Patellar dislocation seldom occurs in the normal knee, even with excess trauma. There is nearly always an anatomic predisposition. It is important to realize what the anatomic predisposition was, as this impacts both the treatment and anticipated result of treatment. Stability of a joint is provided by a combination of bony congruity and intact ligaments. In 1962, Kaplan described the two major medial PF ligaments as the epicondylopatellar ligament and the meniscopatellar ligament. A biomechanical study by Conlon confirmed this equating Kaplan’s epicondylopatellar ligament to the MPFL. Brattström pointed out in 1964 that nearly all 131 patients in his series of recurrent dislocation of the patella had a deficiency in the depth of the trochlea, which may be seen on a true lateral radiograph. We call the film a true lateral radiograph when the posterior edges of both the lateral and medial condyles are exactly superimposed. The floor of the center of trochlea and the anterior extent of the medial and lateral trochlear walls can usually be seen. Dejour has popularized the recognition of the trochlear dysplasia on this view. The lateral view may reveal a high riding patella — patella alta. The Merchant axial view may be useful for recognizing trochlear dysplasia; however, in many patients the shallow trochlea in the proximal region becomes near normal distally by 45° of flexion and thus this view often does not reveal the dysplasia. The axial view may reveal lateral displacement of the patella without stress or the presence of an osteochondral fracture fragment. The stress radiograph is important as the usual situation is that the patella normally is reduced in the trochlea and it only dislocates when a displacement force is applied, so a patella sitting central in the trochlea is not evidence of a stable patella.
- What structures have been injured? It is well known that with normal trochlear geometry, limb alignment and patellar height it takes more force to dislocate the patella, so the ligament, bone and cartilage injury potential is higher. After patellar dislocation, the injured structure is clearly the MPFL, probably the medial patellomeniscal ligament (MPML), and often the patella (distal and medial) and the trochlear, chondral or subchondral surface.
- Can any of these be restored to a normal state by surgical intervention? Not likely.
Robert T. Burks, MD: Obviously, the hallmark findings for an MPFL reconstruction indication would be laxity of the ligament. This is determined by lateral translation of the patella relative to the femur. An MPFL in a normal knee has an endpoint much as we might think of in a Lachman examination. This can be compared to the injured knee for “endpoint.” Patients also can have apprehension, obviously with increased translation as well. It is also useful to compare medial and lateral translation as these should be reasonably symmetric in a normal situation.
Kurzweil:If you see a patient in the office with a first time patellar dislocation, do you routinely order an MRI scan or are plain radiographs usually sufficient?
Burks: I do not routinely order an MRI for a first time patellar dislocation. If on plain radiographs there is a hint of an osteochondral injury that I feel an MRI would be better at elucidating, then I would certainly go in that direction.
Cosgarea: My most common scenario for treating a first time dislocator is to see him or her in the training room or office within days of the injury. Typically they have been reduced on the field or in an emergency room, placed in a brace, and are ambulating with the assistance of crutches.
I always include plain radiographs in my initial assessment. Sometimes X-rays from an emergency room are adequate, but often the quality is poor or the series is incomplete. The lateral view is most useful for demonstrating patella alta and trochlear dysplasia. The axial sunrise or Merchant view may demonstrate osteochondral lesions or persistent subluxation. The tunnel view may identify bony loose bodies in the notch.
Not all patients will need an MRI and the decision to order one should be based on whether it will change your diagnosis or approach to treatment. MRIs can be valuable in determining whether acute surgical intervention is warranted. They may identify concomitant intra-articular pathology, ACL tear, meniscus tear; injury to the extensor mechanism, VMO tear; and large chondral loose bodies. Sometimes the patient will have sustained a dislocation with a spontaneous reduction, leaving the diagnosis in doubt. An MRI will usually identify the classic bone bruise pattern with edema seen at the medial facet of the patella and the lateral femoral condyle.
Teitge: I do not routinely obtain an MRI as proper films are generally sufficient and the MRI seldom adds information which alters treatment choices. An MRI is of academic interest as it may show bone bruises and where the MPFL is injured. Nomura has shown that patellar articular cartilage is injured in over 90% of cases with a patellar dislocation. MRI is not necessary to determine if there is cartilage injury. There is seldom anything on the MRI that would change my approach to treatment.
I obtain three routine views and stress radiographs unless pain prevents the stress radiographs. Proper films must include a true lateral, a true AP, a true axial with the knee flexed in the least flexion possible degree for obtaining an axial view and stress radiographs.
Kurzweil: Would you more likely consider an MPFL reconstruction or repair in a first time dislocator?
Burks: At this point I have not done MPFL reconstruction or repair on a first time dislocator. For these patients I have handled that soft tissue aspect conservatively and only focused on osteochondral injury if needed.
Teitge: There is abundant literature to suggest that a significant percentage of first time dislocators do well without surgical treatment. We do not know which patients will not do well. The number needed to treat (NNT) before the benefit of primary surgical treatment is beneficial has not been measured. There have been many studies showing that patients who have sustained a patellar dislocation can continue to function at a high level, including professional athletes. Some are symptomatic and some are not. Successful primary repair may reflect the patient who did not need surgery. There is always a certain amount of morbidity associated with any operative procedure and that risk must be balanced against the risk of nonoperative treatment. In 1996 Sallay published a prospective series of primary repair and did not report recurrent dislocation, but 58% of patients continued to have swelling after exertion, 66% had pain with athletics and 33% likely had recurrent episodes of subluxation. One must be prepared to deal with this large group of complaints if primary repair is selected. Mountney in 2005 showed that suture repair of a transected MPFL is markedly weaker than the native MPFL. Reconstruction, on the other hand, when combined with the inflammatory response seen after the trauma of a dislocation can results in some excess stiffness which may leave the patient and the surgeon frustrated.
Cosgarea: In the unusual case where surgical intervention in the acute setting is deemed necessary, I generally recommend MPFL repair rather than reconstruction. It is known that the MPFL tears whenever the patella dislocates, and that the tear can occur at the patella insertion, femoral insertion, or in midsubstance. It is usually possible to determine the location of the MPFL tear by identifying the site of greatest tenderness to palpation. This location can be confirmed radiographically if an MRI is obtained. I favor performing suture repair with limited imbrication for midsubstance tears, and suture anchor fixation for patellar or femoral avulsions. Other surgeons have reported success with a more general approach where the MPFL and VMO are advanced at the patella, or with arthroscopic capsular plication techniques.
During the initial clinical assessment I try to identify issues that could potentially complicate the standard nonoperative approach to treating a first time dislocation. Obviously, you will need to rule out other ligamentous injuries like ACL tears. It is fairly common to find concomitant low-grade MCL tears. But a high-grade MCL tear, especially when associated with a VMO avulsion, is a much more serious injury and may force your hand surgically. An important clinical point is to identify the point of maximum tenderness along the course of the MPFL. The point of maximum tenderness correlates with the location of the MPFL tear and is valuable information if you subsequently choose to perform an acute MPFL repair.
|
Image: Teitge RA |
Kurzweil: What is your graft preference for MPFL reconstruction? Do you mostly use allografts or autografts and does your graft choice vary with age and gender of your patients?
Donald C. Fithian, MD: The orientation of the MPFL is ideally suited to preventing lateral patellar displacement, so thick grafts are not needed. Semi-tendinosus autograft is my graft of choice. Allografts may be used, but they are bulkier than autograft because most of these patients are not as large as typical donors. The medial side of the knee does not accommodate excessively thick grafts, particularly in thin, petite women. Autografts reflect the patient’s own body habitus, which is convenient for sizing. On occasions where the patient insists on allograft, I choose a graft of suitable length, and remove (strip) some thickness so it is not too bulky.
Teitge: Most commonly I use the adductor magnus tendon for MPFL reconstruction. [See Figure] I have used adductor magnus grafts, quadriceps tendon — patellar bone grafts, hamstring grafts, and allografts of patellar tendon, hamstring, Achilles, anterior tibialis and posterior tibialis. When I originally began MPFL reconstruction in the 1980s I routinely used quadriceps tendon with a patellar bone block, which was countersunk into the epicondylar region of the medial condyle. The graft was exceedingly reliable, but the choice was not optimal if there was a defect in the patella, which I usually bone grafted. However, this would create a stress riser in the patella which could lead to a transverse fracture of the patella on occasion. The patellar bone block was countersunk into the medial femoral epicondylar region and fixed with a lag screw which usually had to be removed.
Occasionally I use a hamstring graft, but dislike sacrificing normal hamstrings. I have evolved about 15 years ago to using the adductor magnus tendon as it is near the femoral attachment of the MPFL — the muscle seems to have a broad insertion so I have not yet recognized symptoms due to loss of the tendon, it is long enough to reach through the patella, it can be left attached to the adductor tubercle so a strong attachment is guaranteed and, although it looks small, it is significantly stronger than the normal MPFL. Neither age nor gender makes any difference in graft selection.
I do not feel that graft selection is the major cause of failure, but rather trochlear dysplasia, which we do not know how to change with reliable safety, is the major challenge. With a significant trochlear dysplasia (a flat or convex sulcus) asking the MPFL to do all the work is asking a lot and so, in that instance, I am inclined to look to a stronger, larger, graft such as an anterior or posterior tibialis allograft, hamstring or quadriceps tendon. More important is the technique: Proper graft location, proper graft tension and proper fixation of graft to patella and femur.
Burks: Typically, we like to use semitendinosus grafts. Since these are young patients, I think most of the time autograft usage is reasonable. There are occasions where it seems reasonable to use an allograft or a patient may have a preference in that direction.
Cosgarea: I use ipsilateral gracilis tendon autograft as my primary graft source for the majority of my patients. Semitendinosus is a suitable alternative if the length or diameter of the gracilis is insufficient. Hamstring grafts are readily available from the surgical field and do not add additional expense to the procedure or disease transmission risk. I offer hamstring allografts as a good alternative to patients with clinical findings suggestive of systemic hyperlaxity.
Kurzweil: What are your landmarks in surgery for the patellar and femoral insertions and how do you identify them in surgery? Do you routinely recommend the use a mini C-arm for fluoroscopic guidance?
Fithian: The insertions of the MPFL on the medial femur and proximal patella have been well studied and published. I do my best to place the graft anatomically. Because the femoral attachment is very close to the point about which the patella rotates during flexion, graft isometry is rather sensitive to variations in the femoral attachment point. Many studies have used normal cadaver knees to determine the “best” femoral attachment spot. But I don’t think this spot can be prescribed for a specific knee, particularly in patella alta or trochlear dysplasia which are common in these patients. I believe that isometry should be checked during the procedure. I use a mini C-arm as well as isometry to check the position of the guide pin prior to reaming the femoral socket. A short Bieth pin such as that found in a Bio-Tenodesis set [Arthrex] is useful for this purpose because of the limited space within the gantry of the mini C-arm.
Attachment to the patella has little or no effect on isometry, but the MPFL is attached only to the proximal two-thirds of the patella. I try to limit the graft attachment to the proximal half of the patella, so the distal pole is not over-constrained and does not “plow.”
Burks: We are publishing our landmarks for surgery of patellofemoral insertions in the American Journal of Sports Medicine in the next few months. We routinely use the mini C-arm for fluoroscopic guidance. On the femoral side we determine attachment points reasonably similar to the article published by Schottle from 2007. On the patellar side we choose a location at the proximal-third, mid-third junction of the patella.
Cosgarea: Recent radiographic and anatomic dissection studies have attempted to define normal anatomy and ideal femoral tunnel placement. During surgery, I localize the adductor tubercle and medial epicondyle, then use and a mini C-arm to verify positioning. I place a K-wire at the presumed femoral tunnel position, and then wrap the graft around the K-wire and cycle the knee to confirm appropriate graft tension behavior. If necessary, I change the position of the K-wire before I drill the femoral tunnel. It is important to remember that the MPFL graft should function as a passive check rein to prevent excessive lateral translation. There should normally be very little tension in the graft, as excessive tension will result in elevated joint forces. It is the surgeon’s obligation to optimize tunnel placement, graft tension and patellar tracking intraoperatively prior to final graft fixation. This is more challenging than it may seem.
Kurzweil:What is your preferred form of femoral and patellar fixation?
Cosgarea: I have used a variety of fixation techniques at the patellar end of the graft including transpatellar tunnel with a metal button, the docking technique, and most recently, interference screw, usually 4.0 mm to 6.0 mm. I fix the patellar end of the graft first.
Graft fixation at the femoral end is crucial as this is where I determine optimum graft tension. I use a blind femoral tunnel and attach a “pull through” suture through a loop at the end of the graft. I advance the “pull through” suture out the lateral side of the knee using a transfemoral eyelet K-wire. After I bring the end of the graft into the femoral tunnel I range the knee and feel the tension in the graft through the “pull through” suture. I fix the graft at the knee flexion angle that corresponds to the greatest tension in the graft. I have found that this knee flexion angle is not constant and varies from patient to patient. Before I advance the screw and fix the graft I confirm that the resultant lateral patellar translation in full extension — the glide test — is equal to or greater than the contralateral normal knee. It is extremely important that we don’t over tighten the graft as this has been shown in multiple studies to create excessive forces across the patellofemoral articular surfaces. As is often the case in knee surgery, a little too loose is better than too tight.
Fithian: I double the graft and place the looped end into a socket on the femur, fixing it with a 20 mm to 25 mm absorbable interference screw. The two free ends are passed through the retinaculum between layers two and three, then into the medial patella, exiting on the ventral surface and doubling back on itself. Then each arm is sewed back onto itself with nonabsorbable suture. Whatever methods are used for securing the graft, adjustment of graft tension should be possible before final fixation so that the graft is under no tension at any time throughout flexion/extension, when the patella is centered within the groove. In other words, the graft should see tension only if the patella displaces laterally.
Kurzweil: Is there a role for concomitant procedures with MPFL reconstruction?
Cosgarea: I usually perform isolated MPFL reconstruction for recurrent instability if the primary problem is medial soft tissue insufficiency. If the primary pathoanatomy is bony malalignment, I favor tibial tuberosity medialization (Elmslie-Trillat osteotomy) or anteromedialization (Fulkerson osteotomy). This is determined by preoperative clinical assessment of the Q-Angle and J-sign, and especially radiographic measurement of the tibial tubercle to trochlear groove (TT-TG) distance. If after completing the osteotomy the patella is still able to dislocate, I perform an MPFL reconstruction. In this unusual circumstance, I always perform the osteotomy first, as I believe that it changes MPFL graft tensioning.
If the primary pathoanatomy is patella alta, then a distalization osteotomy alone may be enough to correct the instability. Because this is a more complex procedure with a higher risk of complications, I would generally not consider this technique unless the Insall-Salvati ratio, the ratio of the patella tendon length to the length of the patella, is 1.40 or greater. Lateral retinacular release is indicated if the lateral retinaculum is determined to be excessively tight by preoperative radiographic or intraoperative criteria.
Fithian: Lateral release has a limited role in patellar stabilization, and no role at all as an isolated procedure. In our series, lateral release is needed about 10% of the time, when the patella cannot be centered passively in the groove. Such cases of chronic lateral patellar shift can generate continuous tension in the MPFL graft if the lateral tissues are not lengthened or released to accommodate passive centering of the patella.
Since my sabbatical visit to France in 2005, I have used Dejour’s specific criteria to medialize or distalize the tibial tubercle. These criteria are based on standardized measures of patella alta and offset of the tubercle on axial tomographic images. It is not clear to me that these procedures have a direct effect on patellar mobility, but they may play a role in subsequent arthrosis by realigning the patella with the groove. This needs further study.
Teitge: I do not believe there is any reason for routine concomitant procedures and would stay away from these:
- Lateral retinacular release: This is almost never indicated, particularly as it has been shown that the lateral retinaculum provides a significant contribution (up to 20%) of the resistance to lateral patellar displacement and therefore lateral patellar instability is increased with this procedure. This is because as the patella moves laterally up the face of the lateral trochlea, the lateral PF ligament becomes tighter thus limiting further patellar excursion. When the lateral edge of the patella lines up with the lateral epicondyle the lateral PF ligament actually restrains the patella from any further lateral displacement. This reconstruction is often beneficial in the patient with recurrent lateral dislocation who has undergone a prior lateral retinacular release.
- Tibial tubercle transfer: The lateral position of the tibial tubercle is normal and needed. If it is moved medially, it increases loading of the medial knee compartment, pulls the tibia into external rotation and it may increase medial facet loading of the patella. It has been shown in a biomechanical study by Ostermeier that moving the tibial tubercle medially has no effect on reducing strain in the MPFL.
- Distalization for patella alta: Occasionally there is a significant patella alta and the patella does not engage the trochlea until quite late in flexion. In this instance I will consider a distalization of the tubercle in line with the fibers of the patellar tendon maintaining the normal Q-angle. There are no biomechanical studies which prove when or how much support is provided by the trochlea when the patella is engaged. Erasmus has pointed out patella alta will cause an MPFL reconstruction to be pulled loose as the patella moves proximally with quadriceps contraction when the length of the reconstructed ligament is shorter than the distance from the medial epicondyle to the superior medial border of the high patella. In this instance a distalization of the patella is clearly indicated. A distal transfer may also be used at times if a patellar dislocation has knocked off a large osteochondral fragment from the inferior portion of the patella. This missing fragment increases instability which may be compensated for by placing more normal patellar cartilage more distally in the trochlea, but this effect has also not been studied biomechanically.
- Quadriceps lengthening: There are instances in recurrent patellar dislocation when the lateral dislocation occurs with flexion and reduction occurs with extension. This is usually due to a short quadriceps muscle and tendon such that there is inadequate length to allow knee flexion unless the patella is dislocated laterally. In this situation, a lengthening of the quadriceps tendon is needed although it is associated with prolonged quadriceps weakness.
Kurzweil: What are some of the technique modifications you have made as you have gained experience with this procedure?
Burks: In general, I used to use a double-banded semitendinosus graft for an MPFL. Although there is some appeal to spread a graft out over a larger medial patellofemoral ligament attachment point, in particular on the patella, I simply feel this is overbuilt. A single strand of semitendinosus graft is much more significant than the MPFL itself, and I feel that with double grafts any small error in the procedure gets magnified because it is such a stiff, strong graft.
In early procedures I drilled across the patella for fixation of the graft, but having seen a couple of patellar fractures have stopped doing this. I currently use a biotenodesis-type approach with a socket approximately 15 mm to 18 mm deep at the medial aspect of the patella and a screw holding the graft. The upside of this is having good early fixation which makes early rehabilitation possible. However, there are some proponents who advocate just suture fixation to the patella which saves any bone work on the patella, but perhaps has the disadvantage of stretching, pulling, or not being as secure with early motion. The other issue of complications is tightness. One of the things that can happen with this procedure is that people will get pain on the medial side. Part of this could be the manner in which the graft is drilled or fixed, but I think many times it can come from an over-tensioned graft. It is difficult for surgeons who are used to pulling on grafts tightly in ACL and PCL procedures, and assuming that there will be some “stretch.” The tendency is to make it just a little tight. It is important that the patella is simply reduced and excess redundancy removed from the graft before the graft is fixed. The graft should not be tensioned or tightened a little bit with the assumption that it will stretch.
Fithian: Before I started doing MPFL reconstructions in May 2001, I noticed several Japanese iterations of the procedure that suggested it sometimes over-constrained the patella. So I have always avoided over-tightening and have always been careful to understand graft isometry — length-change behavior during knee motion — that the graft is never tense if the patella is in contact with the medial trochlear facet.
I have never drilled across the patella, so I have not experienced any transverse patella fractures. The one fracture I have seen was an avulsion of the ventral-medial corner of the patella — the bridge over which the free graft ends pass before being sewn back on themselves — in a patient who experienced a re-dislocation. This was easily repaired, and it demonstrated the strength of the graft construct.
The first thing I noticed was how prone these knees are to stiffness. This is familiar to any surgeon who has worked around the medial femoral epicondyle, and MPFL surgery is no exception. Patients need to move immediately postoperatively. Our postoperative rehabilitation is not altered if the tubercle is shifted medially, but if distalization is added, then care must be exercised to regain motion without allowing the tubercle fragment to shift proximally.
Cosgarea: The greatest technical challenge with MPFL reconstruction is to provide stabilization without creating excessive joint forces across the patellofemoral articular surfaces. It is easy to over-tighten the graft, or place the graft tunnels in the wrong place. This can result in loss of motion as an early postoperative complication. Treatment of this complication with overly aggressive rehabilitation could subsequently result in cartilage damage or graft failure. My colleague, John Elias, PhD, used computational modeling to simulate surgical technical errors and demonstrated that placing the femoral tunnel in the wrong place or fixing the graft in place with too much tension leads to abnormally high joint forces, especially in the medial patellofemoral joint. The concern is that these excessive forces could then result in patellofemoral pain, cartilage overload and subsequent arthrosis.
Another modification that I have made to my technique is to simplify patellar fixation. I had previously used a full-length transpatellar tunnel with metal button fixation on the lateral side of the patella. Although I have never had a patient fracture their patella, this is a well known postoperative complication and a valid concern. Fixation techniques that use single, short, blind tunnels — the “docking technique” or interference screws — have adequate pullout strength and minimize the risk of patella fracture.
Kurzweil: Tell us about your rehabilitation programs.
Burks: With the approach and fixation that we use, there is absolutely no problem with immediate weight-bearing. We place no range of motion (ROM) limitations postoperatively; in fact, we push ROM as much as possible. We use a knee immobilizer early on until the patient demonstrates adequate quadriceps control or alternatively they can use crutches just to be sure that they are safe with ambulation. As they work on motion and quad strengthening, I think return to unrestricted activity is answered by when each patient is ready. When they have full ROM, no swelling, good quadriceps control, and the ability to work out on an exercise bike or elliptical-type machine without limitation, they can begin to return to what activity they desire.
Cosgarea: The advantage of MPFL reconstruction over MPFL repair or capsular plication techniques is that the graft is strong enough that there is no need to limit knee ROM or weight-bearing. We have learned with ACL reconstruction surgery that it is appropriate and necessary to begin early knee motion and weight-bearing in order to optimize functional outcome and decrease the risk of postoperative complications. I place patients in a hinged ROM brace in the operating room and allow them to progress with motion and weight-bearing as tolerated over the first week. Most patients gain enough strength and limb control that they are out of the brace by 4 to 6 weeks. Lower extremity function is usually good enough to allow a running progression by 10 to 12 weeks and return to sports by 4 to 5 months.
Teitge: The MPFL is normally tighter in extension and subjected to more stress in extension, so flexion is the position which protects the graft. I use CPM postoperatively, usually from 20° knee flexion to the maximum which is comfortable, typically 60° to 80°, full weight-bearing as soon as pain allows, and no brace or immobilization unless it encourages early weight-bearing. Atrophy as a result of disuse secondary to surgical pain and inflammation usually takes many months to be reversed. I suggest vigorous functional exercise at 3 months and return to full activity when strength and agility are good, but seldom before 3 months.
- Robert T. Burks, MD, can be reached at University Orthopaedics Center, 590 Wakara Way, Salt Lake City, UT 84108; 801-587-5455; e-mail: robert.burks@hsc.utah.edu.
- Andrew J. Cosgarea, MD, can be reached at Johns Hopkins Sports Medicine, 10753 Falls Road, Suite 215, Lutherville, MD 21093; 410-583-2864 e-mail: acosgar@jhmi.edu.
- Donald C. Fithian, MD, can be reached at Kaiser Permanente Medical Center, 4647 Zion Ave., San Diego CA 92120; e-mail: donald.c.fithian@kp.org.
- Peter R. Kurzweil, MD, can be reached at 2760 Atlantic Ave., Long Beach, CA 90806-2755; 562-424-6666; e-mail: PKurzweil@aol.com.
- Robert A. Teitge, MD, can be reached at Wayne State University, Department of Orthopaedic Surgery, Heritage Hospital Education Administration, 10000 Telegrapgh Road, Taylor, MI ; e-mail: rteitge@med.wayne.edu.