CORR Insights®: Does Robotic-assisted TKA Result in Better Outcome Scores or Long-Term Survivorship Than Conventional TKA? A Randomized, Controlled Trial

Where Are We Now?

In my experience, a successful total knee replacement is determined by implant positioning, leg alignment, and soft-tissue balance, which includes medial-lateral and AP stability. My definition of a well-done TKA has not changed since the early 1980s, when our specialty—and patients’ lives—were improved by the development of precision mechanical alignment guides, and by the tireless work of David S. Hungerford MD who taught surgeons how to use them. The principles of successful rotational alignment of the implants, and soft-tissue treatment and balance were taught by Chitranjan S. Ranawat MD, and John N. Insall MD for posterior cruciate ligament sacrificing knees, and Richard D. Scott MD and Tom S. Thornhill MD for posterior-cruciate ligament retaining knees. These principles of total knee replacement have not appreciably changed through four decades, nor has implant design resulted in anything other than evolutionary change.

The authors of the current study do not change the principles of the operation, but describe more-precise instrumentation, specifically for the bone cuts in the coronal plane [4]. Since the success of total knee replacement is dependent on rotational mating of the femoral and tibial implants, and the soft-tissue balance of the knee, both of which remain dependent on surgeon decisions no matter the instrumentation, it is unreasonable to expect a difference in clinical scores or revisions between a surgeon who performed 340 total knee replacements per year (as did the surgeon in this study) and the use of high-tech instruments. Indeed, no difference was found. But that does not mean that robotic instrumentation offers no value to low volume or inexperienced surgeons.

Where Do We Need To Go?

We are still trying to find answers for the 25% or so of patients who are not happy with the results of their total knee replacements. To achieve what some have termed the “forgotten knee” (a knee replacement so good that the patient forgets (s)he has undergone the procedure), I believe the operation needs to be personalized for each patient. Unfortunately, as yet, surgeons haven’t figured out just how to do that effectively. We have learned that malalignment of the knee, either in the coronal plane > 3° or rotationally, can cause pain [5]. A second clinical cause of pain is malrotation of the femoral component to the tibia, which commonly causes anteroposterior instability, again increasing strain on capsule and ligaments. Medial-lateral instability can come from incorrect bone cuts (so the medial and lateral compartments or the flexion-extension gaps are not balanced) and/or insufficient knee alignment (which can result is ligamentous imbalance) or polyethylene thickness. A stiff knee may occur when the surgeon makes choices that result in a tight knee with respect to soft-tissue balance, including insufficient releases or use of a a polyethylene bearing that is too thick. If we combine all of these known causes, only coronal and rotational bone cuts can potentially improve instrumentation, while the soft-tissue balance is under the surgeon’s control.

There is also controversy regarding the limits of coronal alignment. For example, 26% of the knees in the current study are considered outliers, as they have a mechanical axis of +/- 3° using conventional instruments. This drops to 6% if we consider an outlier to be +/- 5°. But we don’t know the limits of outlier alignment before pain or dissatisfaction with alignment becomes statistically relevant. Just because these authors decreased the percentage they considered outliers to 5°, that does not mean that knees at that malalignment should be acceptable. A recent study [1] found no difference in the frequency of revision between knees neutrally aligned and outliers +\- 3° 20 years post-operation. By contrast, a study using radiostereometric analysis to measure migration found increased loosening in knees that were +\- 3° outliers [6]. It’s clear that we have a lot to learn on this topic.

One of my favorite teaching axioms is: Instruments do not have brains. A successful operation requires that the surgeon understand the procedure well enough to accomplish it without instruments, and then becomes a better surgeon by the assistance of appropriate instruments. In the current study, the mean leg alignment was 2° to 3° of valgus [4], not the usually quoted 7° +/- 3° of anatomic axis (to be within 3° of the mechanical axis), because they determined that this was the alignment needed for their Korean population. Femoral tools generally promote a 6° or 7° distal femoral cut. Bellemans and colleagues [2] found that 32% of men and 17% of women have normally varus-aligned (varus of neutral) knees. But it is understood that to constitutionally reconstruct these knees requires the kinematic alignment technique with different femoral (placed in internal rotation) and tibial (placed in 2° to 3° varus) implant positions [3]. Therefore, if the authors performed the usual instrument mechanical alignment of the implants, the incidence of loosening with reconstruction of constitutional alignment would not be different from varus total knee replacement which is an outlier of ≥ 3° from the mechanical axis [4]. All of this shows that personalizing total knee replacement requires knowing different techniques, some of which we may not yet understand within the limits of the implants we use.

How Do We Get There?

It is unrealistic to expect that 100% of patients will achieve a “forgotten knee”, but I think we should attain somewhere close to 90%. Although reaching that percentage is realistic for high-volume surgeons, not many surgeons are going to do 300 total knee replacement a year. Perhaps the first step in reducing patient dissatisfaction is achieving improved, reproducible surgical techniques in the hands of lower-volume surgeons.

Let’s consider two examples: First, is the surgeon married to instruments such that even the position of the hole in the intercondylar notch for the femoral rod has to be in only one position? For this surgeon, the rod inside the femur is controlling the cut, not his or her brain, and of course, if this cut is “off”, the rest of the operation is affected. A second example is the surgeon who does 50 total knee replacement per year, and takes 4 hours to do an operation. The lack of confidence in making decisions prolongs the surgery, and risks resulting in bad decisions during surgery. These two surgeons can certainly benefit from computerized instrumentation like a robot. The pre-operative plan done from the CT scan avoids the mistakes made by being married to the instruments, and being frozen by indecision. For these examples, it seems to me that computerized instrumentation would be a benefit, and I have experienced this in my teaching of young, inexperienced surgeons with both computer navigation and robots. Because all intraoperative decisions are easier when the bone cuts are precise, I believe that the entire operation for the surgeon, and outcome for the patient, would be improved.

But surgeons are stubborn, and do not like criticism of their surgical ability, so to arrive at an understanding of the benefit of computerized robotic instrumentation will necessitate studies of proficiency in performance of total knee replacement with lower volume and inexperienced surgeons. Studies done by high-volume, experienced total knee replacement surgeons are unlikely to reflect relevant differences in alignment or patient satisfaction, since they already achieve these endpoints in a high proportion of patients without use of these tools.

Registry studies are a good source for learning the results of different techniques. However, in the United States, there are still limitations on this data because not enough hospitals participate. Change in orthopaedics is easiest at the resident training level, and so acceptance of computers and robots in the operating room will require a cultural shift among leaders of training programs and their hospitals. Orthopaedics is behind other specialties in integrating computers into their operating rooms. Change in surgical habits usually begin in residency programs, and most have not adopted robots because of poor literature support, the expense of the equipment, and satisfaction with their results. Computers and robots will increasingly populate our operating rooms, and I suspect that will be a result of young surgeons in training who want the experience (Because they have grown up with confidence in computers). I also suspect that as these young surgeons become proficient with these tools they will make our operations more precise, and show benefit for the expense. That is a challenge for their generation that will be fun to watch.

Residents can only use this technology if their training program exposes it to them. We can only get to that point if leaders and joint replacement organizations recognize the cultural shift taking place among surgeons today.