Dunbar v. Lewallen: Kinematic Alignment Optimizes Patient Outcome

This week’s Orthopaedic Crossfire® debate was part of the 34th Annual Current Concepts in Joint Replacement® (CCJR®), Winter meeting, which took place in Orlando. This week’s topic is “Kinematic Alignment Optimizes Patient Outcome.” Affirming is Michael J. Dunbar, M.D., F.R.C.S.(C), Dalhousie University, Halifax, Nova Scotia, Canada. Opposing David G. Lewallen, M.D., Mayo Clinic, Rochester, Minnesota. Moderating is Daniel J. Berry, M.D., Mayo Clinic, Rochester, Minnesota.

Dr. Dunbar: I do have trepidation debating Dave, he’s such a thoughtful and learned knee surgeon. I’ll do my best.

To start, neutral mechanical axis is a paradigm of remarkable evolution of total knee arthroplasty. Millions of patients around the world and all of us in this room owe a great debt of gratitude to those who brought us this standard of care.

But I would ask…is it really the end of the road for evolution of TKAs [total knee arthroplasty]? Are we at a dead end when we consider that we’ve made no substantial improvement over the last 15 years of survivorship in total knee arthroplasty?

Arguably, our results in 2003 are as good or better than they were in the most recent reported rounds.

The human condition is individual variation. When we were first exposed to statistics our teacher would line up our classmates and show how height can fall into a bell curve.

Which brings me to the Bellemans, et al. award-winning paper (CORR 2012) that looked at the alignment (using 3-foot standing films) of 500 Belgium patients, aged 20-29 years. They found—go figure—a frequency distribution approximating normal distribution of a bell curve.

Importantly, this normal, healthy population had an average varus of 1.3 degrees and if you took the males out of those, they were at 1.8 or 2 degrees of varus. Healthy individuals. More importantly looking at plus/minus 3 degrees, a full one-third of that population was outside of that window. So, clearly, not one target fits all.

Adolph Lombardi, world-class knee surgeon and president of the Knee Society, said that “Aiming for neutral provides the safest margin for error, but the foremost objective of total knee arthroplasty is a durable joint, and this is important. Not necessarily one that replicates normal or the patient’s native condition.”

If you want to make something last, make it straight. Go back, look at the most promulgated paper (Jeffery et al, JBJS-Br 1991), it’s based on old designs and old implants. They found that a large portion of patients who were outside the 3 degrees of varus window failed—but this was a rheumatoid patient population. Mostly females, using implants that we clearly don’t use today on a population we’re not operating on today.

A very important paper from Barrack’s group (J Arthroplasty 2016) found more or less the same findings that if you’re plus or minus 3 degrees outside this window you have an increased risk of failure.

But my most important argument against this is short leg films, one implant, single center. When long leg films were looked at in a single series over a 15-year follow-up, (Parrette et al. JBJS-Am 2010) a post-operative mechanical axis of plus or minus 3 degrees did not improve 15-year survivorship.

A very powerful counter argument about this neutral target we’re aiming for and fitting to a widely variable patient population.

This is data from our center where we did a randomized trial looking at the shape/match system with navigation. We have one group in the navigation going for neutral mechanical axis and the shape/match was wherever we put them. What was interesting to us was that even though we were trying for neutral, we ended up in a bit of varus in both populations. And the two distributions looked more or less the same for hip/knee/ankle angle.

When we looked at the tibial component angle, that’s where the action is, we saw the difference between kinematic alignment and neutral mechanical alignment. It’s not necessarily the residual alignment. It’s where you put the components within the soft tissue envelope.

You don’t have to make something straight to make it last. We look at coronal alignment as if that’s the way our patients are going to stand straight. But our patients have the audacity to climb stairs, get in and out of bathtubs and cars. This is not a 2-dimensional or 3-dimensional; it’s a 4-dimensional operation.

I can’t help but think that this might have something to do with the fact that we have this glass ceiling of around 18% dissatisfaction with total knee arthroplasty—which is not improving with time.

The number one reason for patients are dissatisfied is unmet expectations.

Think about it. Neutral mechanical alignment. This is how we’re taught. I would submit this is wrong target for patients.

Consider a patient, now standing, mid-stance, or walking, their newly implanted low friction, slippery surface is loading all their proprioceptive fibers in valgus, when their whole life they’ve been loading in varus. The patient says, “This sucks. I don’t know what’s going on, but it feels wrong.” Unmet expectations.

Only 5% of the population is naturally in neutral tibia. Yet, we’re told to cut them all in neutral because that’s the safest thing we can do. We don’t want to be outside that ±3 degrees. Then we transpose that into flexion, set up a whole series of machinations where we have to chase our tail by externally rotating the femoral component. We make non-anatomic cuts, then we put an implant in place.

We take time to level these cuts at 0 and 90 degrees and we balance the knee both in flexion and extension and we think we’re doing a good job, but we completely forget about what’s happening mid-flexion. It’s not a coincidence, in my opinion, that instability is a leading cause of early revision in total knee arthroplasty.

There is evidence coming out. In an RCT on 60 patients; 30 and 30 in each group (Matsumoto et al., Bone Joint J 2017). No increased complications in the kinematic alignment (KA) group. Also, better flexion in the KA group and a better Knee Society functional score indicating better function. Perhaps the component is in the right position.

A meta-analysis on the same topic (Courtney et al., J Arthroplasty 2017), looking at nine studies; 877 patients; 38 months follow-up. No increased complications with the KA group. Using the Knee Society functional score, KA was favored over mechanical alignment.

In conclusion…neutral mechanical alignment is the historic paradigm. But it ignores individual alignment, morphology, and biomechanics variation. Under the current surgical model, zero degrees is often considered the safest position. But that’s changing as the new tools are becoming available.

We’re not at a dead end. We can do better.

Dr. Lewallen: Let me begin and maybe confuse my opponent a little bit by agreeing with him. I think the future of our profession, over the decades ahead, will be individualized hip and knee reconstruction. We will evolve towards a prescription based on their individual anatomic variability.

But we’re not there, not even close. And it could be dangerous to get in front of the parade.

I don’t have any conflicts directly related to this except being old and resistant to change.

The historical gold standard was chosen because surgeons could do it in a reproducible fashion. The outcome results supported the procedure and you’ve seen the data…variations around a target of this sort do not impact outcome. Importantly, having some varus on your tibia didn’t seem to make it any better either.

In my practice, there’s a lot of other things that can go wrong with primary knees and that are wrong with the ones I revise. How do we avoid those things? If we look at causes of failure, actually malalignment isn’t that big of a player, though admittedly early on perhaps it had a bigger role.

But what are the common errors?

• Under resection of the distal femur.
• Deviation from the alignment target. Not everybody hits the target and all of us have a bad day in the OR when we miss the target.
• Femoral malrotation and epicondylar axis problems.
• Tibial malrotation.
• Flexion extension gap imbalance.

It’s unclear how changing the target and kinematic knee alignment is going to solve any of those issues.

This operation demands attention to detail and small changes from what you’re doing in the surgery can combine to make big differences and the errors are iterative.

Failure to achieve rotation and balance at one place may affect the component on the other side. We’re told to resect exactly the same amount of bone from the back of the condyles on someone who, remember, had their knee wear out at age 50. That’s a normal knee? I don’t think so. So, we’re going to take bone off of the hypoplastic lateral condyle and mal-rotate that femoral component?

A few years ago, the most common cause of problems with total knees from the podium at the Academy was patellofemoral problems. That’s largely been solved by attention to detail and implants, but also surgical technique to avoid malrotation. I don’t think we want to go back.

Malrotation of the tibia, which is facilitated by problems with the femoral side, is a very, very frequent cause of problems. How does kinematic alignment solve that problem?

At the beginning of our careers many of us observed that a lot of folks are walking around a little varus in the tibia, and a little more valgus on the femur and the joint is a little bit oblique. What would go wrong with sort of having that as a target?

I can tell you what went wrong. Surgeons weren’t able to hit the target of the 3 degrees very reliably. They couldn’t see it very well. We had a lot of variability. So, 3 degrees of varus is okay, but 5 or 6 degrees is not. And if you shoot for 3, you’re going to end up over there some of the time.

One of the things I will compliment the kinematic folks on…here’s the first time in my career that I’m giving this talk and there are four RCTs [randomized clinical trials] on the topic that have come out in the last couple of years. The problem? They’re all different. Some say it’s good. Some say it’s okay, but you have more outliers and poor outcome.

Kinematic alignment is an interesting concept. We will watch this with interest. But it increases the technical difficulty and complexity for the average surgeon. Meantime, follow the steps and put it in straight.

Moderator Berry: Michael, you know the concept is appealing. That said, when we look at the best prospective randomized trials on this, and Simon Young is probably the best, it has failed to show any demonstrable benefit of that concept. Why is that?

Dr. Dunbar: Unfortunately, I think it’s the metrics as Meneghini was alluding to. We’re using subjective outcomes. Survivorship data takes a while. I think we need measurement stats from EMG [electromyography]; fluoroscopic getting in and out of chairs; gait analysis, etc., which show proportional distributions. These randomized controlled trials by definition are small numbers of patients, so the averages can be misleading.

Moderator Berry: If we just had a little better way to test it, more numbers, you’d see a difference?

Dr. Dunbar: I think so, but I also would like to say that the other important caveat is that we didn’t see any increased complications or problems. I agree with David that the gold standard is neutral mechanical alignment, but I do think it’s safe to explore this topic.

Moderator Berry: Fair enough. Dave, there seems to be some relative degree of safety in exploring the idea and none of us like to do huge releases. As this topic has come up, would you admit to cheating just a little tiny bit and letting that tibial component slide into a degree or two of varus just so you don’t have to do that extra huge release that you might have had to do to get that last couple of degrees of neutral alignment on the tibia. Or would you just never do that?

Dr. Lewallen: No, I would agree. And I think they are on to something because within the range of pathology that we see in total knees, there’s a subset of patients that have this valgus on the femur, varus on the tibia, joint obliquity, straight leg…people look at the X-ray and they think it’s a varus knee, but it’s not really. Also, in my opinion, there are some rotational problems in those joints too as they’ve developed.

Moderator Berry: It reminds me Rob Trousdale’s humorous quip. Somebody asked him, “How do you decide whether you’re going to do kinematic alignment on the knee?” He said, “On the post-operative X-ray.”

Thank you both very much for an informative debate. I think this is one to stay tuned on. Emerging information. Controversial, but an area that has the potential to make a big difference to our patients.

Please visit www.CCJR.com to register for the 2019 CCJR Spring Meeting, – May 8 – 11 in Cleveland.