Another study using RSA shows that what happens is the non-cemented total knees may migrate in the first 3 months, but then they stabilize (Henricson et al.; JBJS Br 2008). Whereas a cemented total knee continues to migrate.
Yet, the common perception is that cementless total knees are inferior to cemented total knees. And it’s based on old literature. A study out of the Mayo Clinic (Duffy et al.; CORR 1998) looking at PFC knees showed only 72% 10-year survivorship with cementless total knees.
A study looking at Miller Galante I knees (Berger et al.; CORR 2001) at 11 years showed problems with metal-backed patellae, tibial component loosening and osteolysis around screws.
I think these poor results reported with cementless total knees are design related. They had cobalt-chrome porous interfaces. Poor initial tibial component stability using metal-backed patellae and use of poor polyethylene.
My personal experience with cementless knees is over 1,300 of them done in the last 30-something years. This still represents only 20% of all the total knees I’ve performed because I use cementless knees in patients who are young, active and heavy, where I am looking for durability over 20 years. My average age in this population is 57.
We recently reviewed my database and developed Kaplan-Meier survivorship curves over the various different designs of cementless knees. The implant that I used had a bimetal femoral component with titanium porous coating for reliable bone ingrowth and a cobalt-chrome articular surface for better wear. The tibial side was an asymmetrical baseplate with continuous porous coating. Early on they had no central stem. And then later we added a stem, all had 4 peripheral pegs.
The patellae used were metal-backed in some and all-polyethylene in others. And the poly design initially was non-highly crosslinked and later on highly crosslinked.
And the titanium surface that we used was consistent throughout the entire series, about 60% porosity, which is very similar to normal bone and also very similar to the trabecular metal-type interfaces that we use today.
Here’s what happened when we eliminated certain design aspects over the years. Eliminating stemless baseplates improved the survivorship. Eliminating metal-backed patellar components significantly improved survivorship. Eliminating tibial screws that offered travel pathways for particulate and the potential for osteolysis eliminated some bad results. And now with highly cross-linked polyethylene my results over 10 years are better than 99.5% survivorship.
If I look over the last 1,000 cementless total knees that I used, I’ve had 1 revision since 2002.
And this is supported in the literature. Studies like Buechel’s report, 18-year, 99% survivorship (CORR 2002). Ritter’s study (JOA 2010) if I eliminate the metal-backed patellae they have 97% survivorship. Hofmann’s study (CORR 2002) in young patients at 9 years had no revisions for loosening. And Whiteside’s study (CORR 2007) looking at both heavy patients and young patients, had no loosening at a 5- to 10-year period.
In summary, cemented interfaces eventually fail biologically because the cement-bone interlock loosens over time. We have concerns for these young and heavy patients. RSA studies support the premise that cementless fixation maybe better for the long term. Biological fixation, I think today, is more reliable because we have made improvements in implant design and polyethylene.
Cementless fixation, I feel, is the preferable method for total knee replacements in the young, active and heavy patient.
As a physical therapist specializing in the rehab of patients who undergo total joint replacement, that was a great debate to read. I’m curious if we should or need to change our approach to rehab in terms of loading fully, gradually, or more delayed spending upon the use of cement or in the cementless cases. Now I’ll be able to ask my question intelligently when I see the surgeons up on the floors after their cases. Thanks for publishing.