CORR Insights®: Is There a Force Target That Predicts Early Patient-reported Outcomes After Kinematically Aligned TKA?

Where Are We Now?

The current study by Shelton and colleagues [9] examines the intersection of two pressing issues—subject-specific component alignment and measurements of tibiofemoral compartment forces—and may illustrate that those two issues were never intended to intersect in the first place.

Much of the content of this paper can be attributed to the mixed results associated with computer-assisted tools like robotic-assisted surgery and surgical navigation systems. Less than two decades ago, such surgical systems were on the cutting-edge of potentially improving the accuracy and precision of implant alignment. While the systems led to more-accurate component alignment with fewer outliers [1, 10], the best evidence we have suggests that such improvement in alignment does not result in improved patient-reported outcomes or survivorship [3, 6]. Combined with the cost of those systems and the increase in surgical time associated with their use [6], enthusiasm for computer-assisted tools has cooled in recent years.

Those concerns led clinical researchers to contemplate other ways to improve alignment and refine surgical techniques. One such approach, the use of kinematic alignment (KA), aligns the femoral and tibial components to restore the native joint lines and native alignments of the limb and knee without ligament release [2, 9]. Since the KA approach is based on individual anatomy, its ideal component alignment varies from patient to patient and contrasts the alignment goals of navigation systems where the center of femoral head, the center of the knee, and the center of the ankle are within ± 3° of a straight line.

Wireless communication technology and sensors like the Verasense^TM (Orthosensor Inc, Dania Beach, FL, USA), are other innovations that provide objective measures of tibiofemoral compartment contact forces that serve as a measure of ligamentous knee laxity. As Shelton and colleagues note, such objective measurements of tibiofemoral compartment forces can theoretically be applied to all patients [9].

Thus, the authors set out to determine whether there is a single quantifiable measure of laxity that can be applied to subject-specific component alignment approaches. The authors concluded that they could not establish a threshold loading level for KA that was associated with better outcomes.

Where Do We Need To Go?

After reading the study, several questions remain. First, what if these instrumented sensors are making the wrong kind of measurements? The tibiofemoral contact forces are typically recorded during a passive ROM or with the knee at discrete flexion angles [9]. However, neither of those conditions place the collateral ligaments under a substantial load. In the 1970s, Markolf and colleagues [4] argued that “clearly, continuous recordings of force-displacement and moment-rotation relationships will give more information on knee stability than one measurement of the deformation at a single load level” [4]; the senior author on the present study has performed similar in-vivo characterizations of knee laxity [5]. I contend (and future studies are needed to confirm or refute this contention) that such measurements of angular or linear displacement under a load would be a more-appropriate indicator of the laxity of the knee than tibiofemoral contact and be a better predictor of how the knee would behave under a varus thrust, or during challenging activities like stepping down from a curb or rising from a chair.

A new generation of surgical tools is needed for measuring knee kinematics and laxity. These tools should have the ability to record component alignment and the motion of the knee under a known load that is applied by the surgeon in both the osteoarthritic knee and following component implantation. Early versions of such tools already exist for use in cadavers [7, 8, 11].

Second, what if the measurements are being made at the wrong time during the surgery? All of these instrumented tibial trays interface with trial implants, meaning that the surgeon has already performed most, if not all, of the bone cuts. While slight alternations in alignment are possible at that point in the surgery, the surgeon has already committed to a certain approach and cannot go back to the native knee. Most surgeons may prefer to assess knee laxity early to further inform an overall plan of component alignment and ligament releases. Moreover, the authors have recently published that, in cadavers, KA does not substantially alter the laxity from the native knee [7].

Could it be that there is no single laxity threshold that is appropriate for all patients, but that all patients have individual targets for component alignment and laxity that are based on the preoperative condition of the knee? Such information can only be determined with measurements of both the osteoarthritic and TKA knee.

How Do We Get There?

The development of this next generation of surgical tooling should take several steps. In order to ensure wide-spread adoption, these tools should be designed by an interdisciplinary team of engineers and surgeons. The ability to accurately record knee laxity as force-displacement data should be verified against mechanical surrogates of a leg that are instrumented with load cells and in cadavers before they are used on patients during a TKA [11]. Lastly, it is critical that these data are able to be quickly recorded, perhaps with not more than 5 minutes to 10 minutes of additional operative time.

Once developed, these new tools can help answer several key questions related to knee laxity. An obvious first experiment could determine what, if any, relationship exists between moment-rotation laxity data and tibiofemoral compartment forces from sensors like the Verasense^TM. Suppose we model the collateral ligaments in different knees as mechanical springs with different stiffnesses and different initial lengths. It is then easy to imagine knees that have similar tibiofemoral contact forces at 0° varus/valgus alignment but different moment-rotation data over a range of varus/valgus angles. Simultaneously recording both types of data and comparing them against each other would be a simple way to explore this relationship. Additionally, both these moment-rotation and compartmental-force data could be related to a patient’s postoperative self-reported and functional outcomes to determine which type of data, and what “threshold” characteristics of those data, has greater predictive capabilities.

Finally, these new tools can help explore the interaction between knee laxity and component alignment. Another study could investigate whether laxity in the osteoarthritic knee, laxity following component implantation, or the change in laxity from the osteoarthritic to TKA knee is the best predictor of postoperative outcomes. Given the results of the present article [9], and since KA was found to not substantially alter knee laxity [7], I contend that a small change of laxity between the osteoarthritic and TKA knees may be a better predictor of superior post-operative outcomes than any single postoperative measure of knee alignment or laxity. Should my contention prove correct, additional experimental and computational studies could then determine how changes in component alignment interact with laxity in the osteoarthritic knee to produce a desired amount of TKA laxity. Understanding the interaction between laxity and alignment could help determine a “safe zone” of component alignment that would inform surgeons how subtle variations in their actions interact with the preoperative condition of a patient’s knee to affect the postoperative functional course for a given patient.