Modifications of the Fulkerson Osteotomy: A Pilot Study Assessment of a Novel Technique of Dynamic Intraoperative Determination of the Adequacy of Tubercle Transfer

Abstract

Medial transfer of the tibial tubercle is commonly implemented to correct patellar alignment in patients with patellar instability. However, the extent of transfer needed is difficult to determine. This article reports a pilot-study experience with a novel technique employing intraoperative femoral nerve stimulation to better determine the distance of tubercle transfer required for proper patellar tracking. this pilot study is a case series involving seven knees, all with a clinical history of dislocation, evidence of maltracking, and excessive medial patellofemoral ligament (MPFL) laxity to the point of producing a positive apprehension sign. All seven knees received femoral nerve stimulation for patellar tracking assessment as part of a modified Fulkerson osteotomy. All knees received clinical follow-up for a minimum of 24 months. six of the seven cases in this series remained stable during two years of follow-up. through these findings we conclude that the use of femoral nerve stimulation for patellar tracking assessment may be associated with a sufficiently high rate of success to warrant more extensive investigation.

INTRODUCTION

Patellar instability is a common and complex problem that is at times both difficult to characterize and difficult to treat. There are many underlying predisposing factors that may lead to patellar instability. These factors include abnormalities in muscle origin and insertion and the dynamic forces created by them as well as significant anatomic malalignment of the femur, tibia, or patella. Additionally, deficiencies of passive restraints (e.g., medial patellofemoral ligament), patella alta, trochlear dysplasia, soft tissue dysplasia, abnormal foot/ankle alignment, genu valgum, and excessive lateral tibial torsion are thought to contribute to patellar instability. Due to its complexity, over 100 surgical methods have been described to treat this condition.⁷ It is clear that no single surgical option is appropriate for every patient with patellar instability and treatment decisions must be made based on the underlying pathology.

It is the senior author's experience that a lateral retinacular release along with direct medialization of the tibial tubercle and a “sliding dovetail” procedure (described by Southwick et al.) are very successful in stopping dislocation episodes.^5,10 However, it became apparent that this procedure resulted in a mechanical displacement of the tubercle attachment posteriorly in the sagittal plane. Although the sliding dovetail procedure had been highly successful in relieving instability, it did little to relieve the fairly common symptom of patellofemoral pain. Many surgeons sought ways to improve the procedure by providing more symptomatic pain relief.⁴

Anteromedialization of the tubercle provides stabilization of the patella without mechanical displacement as seen in dovetail direct medialization. This realignment procedure has also been previously demonstrated to produce a high percentage of positive outcomes in correcting instability.² For these reasons, the anteromedialization procedure has gained popularity in recent decades. The Fulkerson osteotomy is a realignment procedure combining an anteromedial transfer of the tibial tuberosity with or without release of the lateral retinaculum.⁴

The Fulkerson procedure theoretically reduces patellofemoral pain by anteriorizing the tibial tubercle, thereby reducing joint contact forces, while at the same time medializing the extensor mechanism.⁴ However, in the senior author's experience, the success rate for achieving stability with the Fulkerson procedure was lower than the rates reported in the literature, and lower than the success achieved with the dovetail procedure.^2,8,9 It was also noted in the dovetail follow-up study that roughly one in five patients, while not experiencing dislocations, did exhibit a positive apprehension sign. These observations led to modifications in the existing operative techniques.

First, in order to eliminate the positive postoperative apprehension sign, an addition to the traditional Fulkerson was needed. Recent emphasis on the existence and role of the MPFL as a stabilizing structure for the patella made its repair an attractive addition to the Fulkerson procedure. Therefore, the senior author employed reefing and reattachment of the medial capsule and MPFL as a routine method of improving results. Second, a more exacting method was sought to see how far the tibial tubercle should be transferred to achieve proper alignment. With the availability of regional anesthesia at our institution, we developed a method of intraoperative quadriceps activation by femoral nerve stimulation.⁶

The purpose of this pilot study is to assess the results of modifying the Fulkerson osteotomy with the use of intraoperative femoral nerve stimulation. We hypothesized that adding dynamic assessment to the more traditional passive manipulations would lead to a stable extensor mechanism with optimal alignment of dynamic forces.

METHODOLOGY

There were seven knees in six patients (one bilateral) in this series that had a history of recurrent lateral patellar dislocation. Patient age range was 15-37 with an average age of 22. All patients were in good general health. Patients with Dejour type III trochlear dysplasia,3 the most severe type, or those with obvious “miserable malalignment” (excessive femoral anteversion compensated by external tibial torsion) were not included. All of the patients in this study were indicated for treatment based on a clinical history of recurrent dislocations and correlating examination findings. They all demonstrated excessive lateral patellofemoral laxity and a positive apprehension sign. All were also found to have signs of lateral maltracking including combinations of lateralization of the patella on merchant view, abnormal Q-angle, and/or a ?-sign with active extension. Patients who received reconstruction of the medial patellofemoral ligament (MPFL) or lateral patellofemoral ligament (LPFL) were eliminated from this study. All patients in the study were followed for a minimum of two years postoperatively at The University of Iowa Sports Medicine Clinic. Follow-up range was from 25 months to 148 months with an average follow up time of 47 months. At clinical follow-up visits, these patients were assessed for recurrent dislocation, apprehension sign, and J-sign for maltracking. Patients who had one or more of these three criteria were considered outcome failures. Patients who had none of these symptoms were considered to have a successful outcome.

These cases were gathered between March of 1994 and November of 2005 at the University of Iowa Hospitals and Clinics. The novel surgical procedure performed on these knees is described below.

Technique: Fulkerson with limited lateral release, medial capsular repair/plication, and femoral nerve stimulation

Patients in this case series underwent a modification of the traditional Fulkerson osteotomy with the addition of plication of the medial capsule for repair. The technique initially described elsewhere1 is detailed here. Preoperatively, the patient had a “stimulating” femoral catheter placed on the femoral nerve. After induction of general anesthesia (or spinal anesthesia) but before surgery, the settings of the stimulating catheter were calibrated so the quadriceps muscle contraction elicited was of physiologic speed and force. At that time, a first assessment of patellar tracking was made by contraction of the quadriceps muscle with the patient under anesthesia. That assessment was compared to the preoperative assessment in the conscious patient.

Following this calibration, a single long incision lateral to the patella was made and the patellar tendon was identified and dissected so its distal attachment had full exposure medially and laterally. Moving then to the tibia, the bone was cut beneath the tibial tuberosity at an oblique angle posteriorly, from medial to lateral. The proximal end of the tubercle was then cut in a “dovetail” shape while the distal end remained attached. A tamp was then used to slide the tibial tubercle medially and anteriorly along the oblique cut to a desired distance that was measured with a ruler at the proximal end of the cut. Once it was in the desired position, the tubercle was temporarily fixed in place. After this initial bone transfer was complete, lateral patellofemoral structures were evaluated. Lateral retinacular release was then performed when required, based on the tightness of the lateral structures, to achieve soft tissue balancing. The buttress effect of the proximal dovetail along with the temporary fixation of the tubercle, with temporary pin fixation and thumb pressure on the proximal tibia, allowed the surgeon to assess tracking in a variety of ways.

Assessments of patellar tracking in these cases were done in two ways. First, using passive intraoperative manipulation of the knee, and second, by observing active tracking with femoral nerve stimulation.⁶ For the passive intraoperative manipulation, the first maneuver to observe patellar tracking was moving the knee from full extension to full flexion without any rotation and watching the patella for maltracking medially or laterally. Next, to check for under-correction of tubercle placement, the patella was manually displaced laterally in the fully extended knee and then released at the moment passive knee flexion was initiated, with the coupled forces of external rotation and valgus applied to the knee. Lateral dislocation with this motion indicated under-correction. Finally, to assess over-correction, the patella was displaced medially with the knee extended, then the knee was flexed and varus and internal rotation forces were applied. Medial dislocation in this maneuver indicated over-correction and/or excessive lateral release.

After passive assessment of patellar tracking, intraoperative femoral nerve stimulation was performed.⁶ Through the calibrated femoral nerve catheter the quadriceps muscles were stimulated to actively move the knee from flexion to extension against gravity. Special care was taken to note the presence or absence of a J-sign in active extension of the knee. Based on observation of the active patellar tracking, adjustments were then made in the distance of tubercle transfer. A final assessment of tracking was done before the tubercle was secured with two bicortical screws at the position that demonstrated the best tracking.

Following tibial tubercle transfer, all patients in this pilot study received plication of the medial capsule to reduce redundancy of the medial structures created by the tubercle transfer.

The integrity of the medial capsules in all seven cases was judged to be sufficient to avoid restretching postoperatively. The medial patellofemoral ligament was assessed manually with the medial structures under tension. In all seven cases, the MPFL was believed to be both present and of adequate integrity to allow for repair as opposed to reconstruction. The repair was performed in each case after assessing patellar tracking. The capsule and MPFL were attached to the patella with bone suture anchors.

Patellar tracking was always assessed prior to plication of medial structures and after release of the lateral retinaculum. This was done so that tracking could be assessed without any medial or lateral restraints. This allowed for adjustments to be made in both the extent of the lateral retinaculum release, and the tightness of medial reefing, in order to fine tune patellar tracking.

RESULTS

There were seven total cases in this pilot study. Of these seven cases, one patient had a postoperative positive J-sign and recurrent dislocation at the two-year follow-up visit. The six positive outcomes in this group had no apprehension or ?-sign, and no recurrent dislocation.

DISCUSSION

In the years prior to the development of this technique, patients at our institution received only the modified Fulkerson osteotomy without dynamic assessment of patellar tracking. During that time period it was noted that nearly all of the few patients who continued to show patellar instability following the Fulkerson without femoral nerve stimulation had a positive ?-sign on follow-up. Our theory is that by using femoral nerve stimulation to dynamically assess patellar tracking and appropriate placement of the tibial tubercle, intraoperative maltracking should be eliminated as well as any postoperative J-sign.

This first series of patients has shown promising results. We have been impressed with the ability of the femoral nerve stimulation technique to reproduce the same pattern of quadriceps firing demonstrated by the conscious patient. Although we have not yet validated this observation of reproducibility between conscious and anesthetized patients, that validation study is currently under way.

In looking at our novel technique in relation to other possible ways for achieving patellar stability, the most important difference to note is this technique's ability to balance dynamic structures. Other authors, such as Teitge,11 have shown that MPFL reconstruction alone can eliminate instability in patients with or without an adequate trochlear groove. However, those studies do not directly address the additional problem of dynamic and bony malalignment. It is our contention that principles learned during the polio era still hold true; static repairs will not last forever if the dynamics involved are not also correct. It is speculated that with the passage of time, MPFL repairs with persistent pathologic dynamics will eventually loosen again. Additionally, simple correction of bony malalignment, such as the traditional Fulkerson, without consideration of pathologic dynamics may end in the same poor result in the long term. Certainly this short-term pilot study cannot illuminate that issue.

Implementation of this technique has been successful as shown in the results of this case series. There was a high success rate (six of seven) in cases where femoral nerve stimulation was implemented. During these cases the transfer distance of the tibial tubercle was changed anywhere from zero to six millimeters because of the observations made during the femoral nerve stimulation. When the tubercle transfer distance was changed based on the femoral nerve stimulation it was always transferred further medially. This often resulted in a medial transfer distance that was larger than originally thought necessary. There were no medial dislocations postoperatively despite this, suggesting another possible advantage to using this technique. As muscle strength improves postoperatively there may be a tendency for a stronger VMO to pull the patella medially. Assessing forceful contraction of the VMO intraoperatively may predict this change at the time of surgery and allow for adjustments to be made. To determine the significance of using dynamic assessment as part of the Fulkerson osteotomy, a large prospective study that evaluates outcomes with and without use of femoral nerve stimulation for dynamic balancing is needed.