To The Editor:
I read with great interest the Instructional Course Lecture "Mobile-Bearing
Knee Replacement. Concepts and Results" (82-A: 1020-1041,
July 2000), by Callaghan et al. I found the section concerning the in
vivo kinematic studies particularly pertinent and complete.
However, the section entitled "Why Should We Question
the Enthusiasm for Mobile-Bearing Knees?" astonished me
with its negative conclusion about the mobile-bearing knee and its statement
that surgeons who use this type of prosthesis "must be willing
to accept a 1 to 2 percent rate of mechanical failure associated
with use of a mobile tibial insert."
The arguments employed are biased and do not support such conclusions.
One could draw a very different conclusion; i.e., the results of
the mobile-bearing knee are at least equivalent to the best
results of the fixed-bearing prosthesis. I feel that the authors’ arguments
are questionable both in their form and their substance. First,
I will address the functional arguments and then I will briefly
address matters of substance.
Mobile-bearing knees are criticized on the basis of their paradoxical
movement (anterior translation of the femur on the tibia) which
is responsible for the unfavorable kinematics of the extensor mechanism
and, consequently, a higher risk of polyethylene wear. This paradoxical
movement is not characteristic of mobile-bearing knees but rather
of posterior cruciate ligament-retaining prostheses (without retention
of the anterior cruciate ligament), whether they have fixed or mobile
bearings. Among the prostheses with mobile meniscal bearings, the
posterior cruciate ligament-retaining prosthesis does have this
characteristic of unfavorable kinematics, but the rotating-platform
design does not have this drawback.
The argument about the flexion angle is also very pernicious. The
authors refer to an article by Dennis et al.1 that
gives a flexion angle of 105 for the meniscal-bearing knee with
posterior cruciate ligament-retention, and then they compare this
to a 110° to 120° flexion angle reported elsewhere for the fixed-bearing
knee. This is an unusual comparison of a one-sided, unfavorable
report on mobile-bearing knees with the most favorable results of
fixed-bearing knees. No valid conclusion can be drawn from such
a comparison.
Having done a meta-analysis of all pertinent articles published over
a two-year period in The Journal of Bone and Joint Surgery,
Clinical Orthopaedics and Related Research, the Journal
of Arthroplasty, and Acta Orthopaedica Scandinavica,
I and my colleagues calculated that out of 1654 patients, the average
range of motion is 100° for posterior cruciate ligament-retaining
prostheses (whether equipped with fixed or mobile bearings), 100
for posterior stabilized fixed-bearing designs, and 110 for rotating-platform knees.
These are not excellent results, but they certainly do not place
the mobile-bearing knees at any particular disadvantage.
Finally, it has been suggested that the mobile-bearing knee may
better tolerate minor errors of rotation of the tibial component.
This argument has never been used by any designer of mobile-bearing
knees. It has been shown that the mobile insert does not correct malrotation
of the tibial component, and malrotation results in identical patellofemoral
problems for both mobile-bearing knees and fixed-bearing knees2.
Regarding the statistical arguments, the authors compare the
series they have selected and conclude: "Clearly, the mobile-bearing
design is not superior with regard to the prevention of mechanical
failure and revision." It is not acceptable to conclude that
a result of 98.7% with a fourteen-year follow-up is superior
to a result of 97.5% with a twelve-year follow-up. When
comparing the survivorship graphs of two treatments, one is confronted with
the risk of concluding that there is no statistical difference when,
indeed, there is one3.
Above all, one must compare comparable groups; that is to say, posterior
cruciate ligament-retaining fixed bearings must be compared with
posterior cruciate ligament-retaining meniscal bearings, and posterior-stabilized fixed
bearings must be compared with rotating platforms or posterior-stabilized
mobile bearings.
When using such comparisons in the previously mentioned meta-analysis,
one finds that the results of the mobile-bearing knees are at least
equivalent to the results of the fixed-bearing knees, and it is
impossible to prove the clinical superiority of one or the other.
We can only be surprised that the authors seem to demand more from
the mobile-bearing design: "one must have data that overwhelmingly
supports its superiority to its temporal peers."
In reference to a point of substance, the authors are questioning
the concept of congruency as a major factor in diminishing long-term
wear. To test this concept, one must measure the wear, not on explanted
prostheses but in terms of the annual rate of reduction of the thickness
of the polyethylene. When one uses this parameter, and only this
one, as a percentage or as a survivorship curve end point, in homogeneous
groups of prostheses with equivalent qualities of polyethylene,
it appears that congruency is a fundamental advantage with a highly
significant influence on the wear of the prosthesis.
J.J. Callaghan, A.S. Greenwald, R.B. Bourne, C.H. Rorabeck, and
L.D. Dorr reply:
We appreciate Dr. Bercovy’s comments. The purpose of
the last section of the paper was to point out the potential concerns with
mobile-bearing knees. In the Instructional Course Lecture, this
section represented the case against surgeons necessarily jumping
on the mobile-bearing knee bandwagon. Rhetorically it was intended
to represent the argument for continuing to perform fixed-bearing
knee replacements. Hence, the authors used the best data to present
the point. All of the authors would agree that there are no statistical data
to demonstrate the superior function of either fixed or mobile-bearing
knees. With both fixed-bearing and mobile-bearing knees, the authors recognize
that there are and will be further data to demonstrate both good
and poor results. As more mobile-bearing knees are introduced to
the market, it will be important for surgeons to perform prospective,
randomized trials to test for any measurable differences between
the results of fixed and mobile-bearing knee replacements. It may
also be important to perform randomized trials comparing newer mobile-bearing
designs with older, proven designs, such as the Low-Contact-Stress rotating-platform
knee. The authors hope that Dr. Bercovy recognizes that the intent
of the final section of our Instructional Course Lecture was purely
to play "the devil’s advocate" to temper
the enthusiasm for the future of the mobile-bearing knee replacement.
All of Dr. Bercovy’s comments are justified, and
once again, demonstrate the need for prospective, randomized trials.