Observations of the Proximal Tibia in Total Knee Arthroplasty

Abstract

Relatively little has been written concerning the proximal tibia in total knee arthroplasty. Few authors have looked at landmarks and guidelines for tibial tray preparation and tibial tray orientation. The current study showed that a line drawn 1 mm medial to the medial border of the tibial tubercle and going through the midsulcus of the tibial spines (the midsulcus line) provided a reproducible landmark for the tibia, and when a perpendicular cut was made relative to this line, 46 of 50 knees were cut in appropriate alignment. In addition, it has been said that the tibial tray should be rotated externally to approximately the medial ⅓ of the tubercle to maximize function. The current study showed that when the tibia is allowed to float in a functional position relative to the femoral implant, the tibial external rotation was only 2 mm lateral from the medial age of the tibial tubercle; this is far less than the medial ⅓ of the tubercle and close to the starting point of the midsulcus line.

The overall function of a total knee replacement is dependent on many variables; mechanical alignment, implant alignment, joint line restoration, and soft tissue balancing are a few of the more important variables. Much attention has been given to establishing reproducible and reliable landmarks for setting appropriate rotation of the femoral implant. ^1,6,7,12 Useful landmarks that have proven to be reliable include the posterior condylar axis, the intercondylar axis, and the midtrochlear line. ^{3,4,8,9,11,15} All of these landmarks have been shown to be helpful in establishing correct femoral rotation. ^2,10 This is important to ensure correct patellofemoral mechanics and equal flexion and extension gaps.

The relationship of the tibial tray to the femoral component is another important determinant of the overall function of a knee arthroplasty. If a tibial tray is rotated internally or externally, there will be impingement on the polyethylene, which can lead to loss of motion and perhaps third body wear. The relationship of the tibial and femoral components also is a prime determinant of patellar tracking and function.

In addition, if a proximal tibial cut is placed in too much flexion and associated internal or external rotation relative to the long axis of the tibia, there can be instability in flexion. Thus, it is important that the tibial cut reflect rotation and flexion.

Compared with the femur, relatively little attention has been given to establishing similar guidelines for alignment and orientation of the tibial tray. Observations of the tibial landmarks in total knee arthroplasty are reported, and a reproducible landmark that is useful for tibial preparation is identified.

MATERIALS AND METHODS

Fifty consecutive total knee replacements were done by one surgeon. All 50 patients had osteoarthritis. There were 34 women and 16 men in the study group. The average age of the patients was 72 years (range, 54–85 years), and the average weight was 154 lb for the women and 190 lb for the men. The average preoperative range of motion (ROM) was 3° to 122°, and the average preoperative alignment was 2° varus (range, 5° valgus–8° varus) (Table 1). All knees were approached using a midline incision, and 48 of 50 knees were entered with the midvastus arthrotomy.

TABLE 1: Patient Profiles

All knee replacements were cemented tricompartmental replacements (P.F.C. Sigma, Depuy Orthopedics, Warsaw, IN), and all were posterior cruciate ligament-retaining replacements. Intramedullary alignment was used for femoral preparation in all patients. The tibias in men were cut in 7° valgus and the tibias in women in 5° valgus relative to the long axis of the femur. Femoral component rotation was aligned using posterior femoral condylar referencing, epicondylar axis, or the midtrochlear line, and alignment usually was checked with all three. All femoral components were rotated externally approximately 3° to achieve a rectangular flexion space to improve flexion stability and patellar tracking.

Once the femoral trial implant was seated satisfactorily, it was removed and attention was given to the tibia. External alignment jigs were used in all knees. Adequate visualization of the entire proximal tibia was obtained in all cases. The proximal tibial cut was made perpendicular to the long axis of the tibia. The tibial cutting guide was set over the midportion of the talus and in line with the second ray to approximate the tibial alignment (Fig 1). The posterior slope of the proximal tibial cut was made using the anterior tibial crest as an external alignment guide and was set to be approximately 7°.

Fig 1.: View of the relationship of the midsulcus line to the tibial cutting jig is shown.

Orientation of the tibial implant was set relative to the fixed femoral trial. With the femoral trial in place and the proximal tibia cut and all ligaments balanced, the tibial trial and plastic insert were placed in the knee. The knee was placed through a full arc of motion several times, allowing the tibial tray to float into a position were there was no impingement of the tibial plastic insert on the femoral trial. This orientation was marked, and the keel was cut into the tibia. The patella was prepared manually by resecting adequate amounts of bone and cartilage to accommodate an all-polyethylene implant.

Once all three components were cemented, the knee was placed again through an arc of motion, and patellar tracking and stability in extension and throughout the flexion arc were observed. Standard postoperative closure and rehabilitation were done. Patients were examined at 6 and 12 weeks after surgery, and radiographs (anteroposterior [AP], lateral, and sunrise views) were obtained at both visits.

RESULTS

All patients were available for followup. There were no complications. Specifically, there were no wound healing problems, infections, fractures, neurovascular compromise, or need for manipulation. There were no lateral releases done. The average ROM was 2° to 115° at 12 weeks. Radiographs were used to measure alignment of the implants. The average tibiofemoral alignment was 6° (range, 4°–7°), and the average tibial tray alignment was 1° varus (range, 2° varus–2° valgus), as seen on 18-inch radiographs obtained at 12 weeks with the patient standing. The average posterior slope of the tibial cut was 6° (range, 4°–9°). There were no subluxations or dislocations of the patella seen on the sunrise view.

The average width of the tibial tubercle, measured at its maximal thickness using a caliper, was 21 mm for the women (range, 19–23 mm) and 24 mm for the men (range, 21–25 mm).

The alignment of the tibial keel (marked by a line on the tibial tray placed by the manufacturer) set according to the floating tray concept was 1 to 2 mm lateral to the medial border of the tibial tubercle (range, 0–4 mm lateral) (Fig 2).

Fig 2.: Alignment of the tibial keel relative to the medial border of the tibial tubercle is shown.

The location for the setting of the midportion of the tibial alignment guide was 1 mm medial to the tibial tubercle along the medial edge of the tibial tubercle (Fig 1). When this location was set and a direct line from this starting point through the deepest point of the sulcus of the tibial spines was drawn, (Fig 3) cutting perpendicular to this line gave the correct orientation of the tibial resection in the coronal plane in 46 of 50 knees.

Fig 3.: View of the tibial surface from above, showing the midsulcus line.

There were four patients in whom the midsulcus line was not an accurate alignment guide. In one of the patients, the medial tibial spine was worn away completely from impingement on the femur, so no sulcus could be identified. In the other three patients, the correct orientation of the tibial cut was lateral to the sulcus and on the lateral tibial spine.

Continuation of the line through the sulcus of the tibial spines and back into the bony recess of the posterior cruciate ligament insertion on the tibia revealed that the line intersected the deepest portion of the crescent of bone around the posterior cruciate ligament in 39 of 50 knees. In 10 of the remaining 11 knees, the line was oriented between 10° and 15° laterally (relative to the deepest point), and in one of the knees it was oriented 10° internally.

DISCUSSION

The current study examined landmarks on the proximal tibia that can be useful for achieving correct rotational alignment for total knee arthroplasty. Relatively little has been written concerning tibial anatomy. ¹⁴ The current study showed that in most patients, there is a consistent relationship between the starting point for the tibial cut, the sulcus between the tibial spines, and a point 1 mm medial to the tibial tubercle. Cutting the tibia perpendicular to this line gave accurate tibial alignment (relative to the midtalar or second ray position) in 46 of 50 knees. Acknowledging this line at the time of tibial preparation gives the surgeon assurance of correct tibial alignment in the coronal plane.

Failure to achieve this alignment is most dangerous when any posterior slope is applied to the tibial cut. Many modern knee arthroplasty systems have a proximal tibial cutting surface with medial and lateral extensions that are meant to make the tibial cut more reproducible by providing a broader surface for cutting. Unfortunately, occasionally these medial and lateral extensions can be caught on the tibial tubercle and cause the tibial tray to become internally rotated relative to the tibial tubercle. If the tibial rotation is malrotated and a posteriorly sloped cut is applied (as commonly is suggested with many knee replacement systems), there will be a relative excess amount of tibial bone resected posterolaterally, and the knee will be at risk for midflexion instability. This is particularly important if a posterior stabilized or rotating platform design is to be used because midflexion instability can lead to dislocation of a tibial polyethylene stabilized or stemmed insert.

The midsulcus landmark also is helpful when planning cuts for a unicompartmental replacement or if any sort of minimally invasive procedure is contemplated. The midsulcus of the tibial spines is visualized easily and palpated arthroscopically, and the starting point (1 mm medial to the tibial tubercle) is easily palpable through the skin. These two points give a reliable set of coordinates if the surgeon is not going to expose the entire proximal tibia. A perpendicular cut to this line at the appropriate depth gives correct tibial alignment coronally.

Extending the midsulcus line back to the tibial recess for the posterior cruciate ligament was predictable in only 39 of 50 knees and was a less helpful landmark. There often are bony abnormalities around the posterior cruciate ligament, which makes this recess on the tibia a less consistent shape.

The external rotation of tibial tray averaged only 2 mm (range, 0–4 mm) relative to the medial border of the tibial tubercle. This point is close to the starting point of the midsulcus line. This is less than the medial ⅓ of the tubercle (average, 21 mm for women and 24 mm for men) that is mentioned in at least one technical brochure of the knee implant systems (P.F.C. Sigma Knee System Revision Surgical Technique, page 12). Although internal rotation of the tibia relative to the femur has been associated with patella complications, such as subluxation and dislocation, the true function of the tibia polyethylene is how it relates to the fixed femur. Allowing the tibia to float into an alignment relative to the femur means there should be less polyethylene impingement on the femoral component and presumably better patellofemoral function. ¹³

In these patients, using this method with this particular knee replacement system, did not seem to cause any patellofemoral complications. Functionally the patients regained an acceptable ROM, and there were no subluxations, dislocations, or maltracking seen on radiographs or by direct examination at 12 weeks. There were no lateral releases in this group, which could be attributable to the surgical approach used (the midvastus has been reported to have a decreased need for lateral releases ⁵) or the design of the implant. However, this also could be attributable to allowing the tibia to assume a functional position that minimizes femoral impingement and maximizes femoral and tibial contact.

In the current study, two observations were made of the proximal tibial anatomy in the patient with a total knee arthroplasty. The first observation is that there is a reproducible line from a point 1 mm medial to the tibial tubercle and extending back through the deepest point of the sulcus between the tibial spines, which when cut perpendicularly, provides accurate tibial preparation in the coronal plane. The second observation is that the orientation of the tibial tray, when allowed to float into a position where there is no impingement on the femoral implant trial, is a position that on average is only 2 mm lateral to the medial border of the tibial tubercle. This point is similar to the starting point of the midsulcus line.

Both of these findings were found to be consistent with various sizes and shapes of knee implants and were not related to any preoperative deformity. Keeping these observations in mind can help the surgeon achieve a more predictably aligned knee replacement.