Revision problems after resurfacing

[John C]

http://informahealthcare.com/doi/full/10.3109/17453671003667176

Here is a very interesting study done from a very large number of cases that questions one of the most basic assumptions regarding the advantage of easy revisions from resurfacing. Though the news is not great, I am afraid that this study must be taken into account in any future discussions on the advantages of resurfacing over THA.

[Bionic]

I have to wonder whether this study is subject to selection bias because it compares revision rates of THRs after failed resurfacings to revision rates of primary THRs.  In some cases, perhaps most, resurfacings fail because of poor bone quality or other problems that are not seen in the typical patient.  Isn't it likely then that the same patients who have had failed resurfacings will also have trouble tolerating their total hips?

The problem with the study is that it compares the general population of THR patients with the problem population of resurfacing patients.  I admit I've only skimmed the study, but on its face this doesn't appear to be a fair comparision.

[resurface]

I just read the report.  Here is my take.  First, IMHO, this does casuse me to consider further my surgery in May.  It has not changed what I will do however.  9% total revision rate.  Not great by THR standards but there is a great deal to consider:

(1) 13 different system were used.  Consider the learning curve of each and the surgeon variability.  Are the failures system, surgeon, patient selection related???
(2) BHR revisioin rate was 0.8%.
(3) 43% of failures occured at the femoral neck. Recent JBJS article that discusses HR in women suggests a higher rate.  What is the impact of this populaton if factored out?  What is the impact of better patient selection that occurs today?
(4) Diagnosis - femoral neck and acetabular loosening, again, goes back to patient selection and technique [possibly]
(5) 247 / 62% (out of 397) Femoral failures is, again, likely patient selection. 

The last chart shows survisorship of THR to that or HR... not pretty, but again, consider the above.  As many have said, patient selection and surgeon experience are very important.  Unfortunately, this retrospective study does not parse the data enough IMHO.  That said, it is part of the discussion and going into this procedure with your eyes open.

[Bionic]

From the study,

The Australian Joint Replacement Registry does not specifically identify pseudotumor as a diagnosis, but uses the diagnosis of metal sensitivity to cover all cases of metallosis. We suspect that metal sensitivity is under-diagnosed and that a proportion of revisions for early loosening/lysis may be the result of metal sensitivity.

The highest re-revision rates were associated with replaced cups.  One has to wonder whether cup angle may again play a role here.  When a surgeon replaces a loose cup, does he or she replace it at the correct angle?  Perhaps surgeon skill with MOM implants is as important in revisions as it is in primary resurfacings.

Tip: Go to an experienced hip resurfacing surgeon or to an experienced MOM THR surgeon if you ever need to have your resurfacing revised.

[obxpelican]

Yes, and keep in mind, you are still saving your original femoral neck, if something causes a need for a revision you still have a 94% chance of the first revision working fine and I am sure with an experienced surgeon you will fair better than 94%---- might be a good question to an experienced surgeon.  What % of your revisions are succesful?

If you read the report you'll find in the beginning a 3.3% rate  of all combined issues other than infection.  That includes fractures, metal sensitivity, bad installed implants, that is EVERYTHING other than infection.  If you consider the report included surgeons that were not experienced and devices that have had problems I am sure the numbers would be less, significantly, had the implants been BHRs with experienced surgeons.

Look at it this way, you have a 96.7 chance of a good result if you use an average surgeon, your % will be better with experience, if you need a revision you stand a 94% chance of a good result with an average surgeon, again, with an experienced surgeon I am sure your results will be better.

Personally, I do not look at the study as any kind of issue one should worry about.  If anything, this report further tells us all that we should seek out experienced surgeons.


Chuck

I just read the report.  Here is my take.  First, IMHO, this does casuse me to consider further my surgery in May.  It has not changed what I will do however.  9% total revision rate.  Not great by THR standards but there is a great deal to consider:

(1) 13 different system were used.  Consider the learning curve of each and the surgeon variability.  Are the failures system, surgeon, patient selection related???
(2) BHR revisioin rate was 0.8%.
(3) 43% of failures occured at the femoral neck. Recent JBJS article that discusses HR in women suggests a higher rate.  What is the impact of this populaton if factored out?  What is the impact of better patient selection that occurs today?
(4) Diagnosis - femoral neck and acetabular loosening, again, goes back to patient selection and technique [possibly]
(5) 247 / 62% (out of 397) Femoral failures is, again, likely patient selection. 

The last chart shows survisorship of THR to that or HR... not pretty, but again, consider the above.  As many have said, patient selection and surgeon experience are very important.  Unfortunately, this retrospective study does not parse the data enough IMHO.  That said, it is part of the discussion and going into this procedure with your eyes open.

[Bionic]

Yes, and keep in mind, you are still saving your original femoral neck, if something causes a need for a revision you still have a 94% chance of the first revision working fine and I am sure with an experienced surgeon you will fair better than 94%---- might be a good question to an experienced surgeon.  What % of your revisions are succesful?

I think the report is confusing, and I'm having some trouble understanding it.  I can see where you get your 94% number from the study:

Of the 397 revisions of primary resurfacing hip arthroplasty undertaken for reasons other than infection, 24 underwent a further revision. The reasons for re-revision included loosening/lysis (n = 9), infection (n = 6), and dislocation (n = 4). The most common type of re-revision was femoral-only revision, followed by acetabular and femoral, and acetabular-only (Table 4).

24 revisions out of 397 surgeries is a 6% revision rate (94% success rate).  But I'm having trouble squaring that with the statements from the results summary:

Acetabular-only revision had a high risk of re-revision compared to femoral-only and both acetabular and femoral revision (5-year cumulative per cent revision of 20%, 7%, and 5% respectively).

Also, from the detailed report:

The subsequent cumulative per cent revision of the 397 revisions of primary resurfacing hip arthroplasty was 9% (6–13) at 5 years[emphasis added]. This varied depending on the type of revision, with acetabular-only revision having the highest risk of re-revision.

Could it be that the 94% data is based on actual revisions to date as recorded in the hip registry, whereas the 91% failure rate is based on statistical predictions of the expected 5-year cumulative revision rate?

You would expect the statistically predicted revision rate to be greater than the actual revision rate because many first revisions from resurfacings are recent, and actual failure rate data is not yet available.

[Bionic]

36 of the 397 revisions (9%) were of the acetabular component only.  For those cases, the revisions had to be resurfacings themselves.  Of all the revisions, those second resurfacings were the most likely to fail a second time and upwardly skewed the overall failure rate.

So, one explanation for the poorer-than-expected success rate of revisions could be that patients received second resurfacings that did not resolve the underlying problems that lead to the first revision (or maybe the same surgeon made the same error the second time around).

It's also worth pointing out that those same 36 of 397 patients got to keep their resurfaced hips (a good thing).  Those patients essentially got a "second bite at the apple," i.e., a second chance to make the resurfacings work.  Of those, it appears that 80% are expected to still be going strong after 5 years.  And even if the second resurfacing fails, the first THR is not installed until the third surgery.

And then there's the obvious explanation:

Resurfacing patients are younger and more active than THR patients.  Maybe it shouldn't be any surprise then that active, young patients will wear out their THRs faster than an older, less active patients.

In addition, the younger resurf patients who get THRs as revisions have histories with resurfaced hips and expectations for what they can and can't do.  Unless they pare down those expectations, they may run into a whole lot of trouble when they get THRs.

Also worth noting: The 9% cumulative failure rate does not exclude infection as the reason for the second revision.  Infection accounted for a large percentage of the total (6 of the 24 re-revisions, or 25% ! ).  

Therefore, a big reason revisions of resurfacings fail is that the revisions get infected.

[Bionic]

Sorry for the multiple posts.  I don't mean to be annoying, but I'm having trouble getting this study out of my head.

As I said above, I think it's unfair that this study compared patients who already had failed hip surgeries to the general hip patient population.  Past problems with hip surgeries are certainly a predictor of future problems, and that alone could go a long way toward explaining the revision failure rates.

Another problem is that the study fails to compare the rate of re-revision when the original surgery was resurfacing to the rate of re-revision when the original surgery was THR.  Patients generally need one surgery or the other, so it doesn't make sense to look at resurfacing in a vacuum.  You have to look at it in comparison to THR.

We know that THR revisions have much higher complication rates than initial THRs.  For example, http://www.ejbjs.org/cgi/content/abstract/85/1/27

This study says that the 10-year failure rate for THR revisions is 26% !!!
http://www.jbjs.org.uk/cgi/reprint/86-B/4/504.pdf

If I'm interpreting things correctly, this compares to a merely 9% failure rate at 5 years when the first surgery was a resurfacing.  Now, 5 years isn't the same as 10 years, but I think it's still pretty clear that the THR re-revision rate is significantly worse than the resurfacing re-revision rate.

Doesn't this demonstrate that, even when it comes to re-revision rate alone, resurfacing is still the better choice for the initial surgery?

[muzza]

Hi Bionic,

Thanks for your analysis of this study.  You've done a good demolition job on a flawed piece of research.

Neil

 

[Bionic]

Thanks, Muzza.  Here's what I've learned by going through this exercise:

All hip revisions have a shorter life expectancy than original hip procedures, and this is true both when the original hip surgery is a resurfacing and when it's a THR.  

But in general patients are better off when the first surgery is a resurfacing.  Aside from the fact that the resurfacings usually perform better than THRs over their operating life, they also are easier to revise if they ever fail and the life expectancy of the resulting revision is longer than it would be if the first surgery had been a THR.

On top of that, the original resurfacing preserves bone as compared to the THR and therefore benefits younger patients who may need more than 1 revision over their lifetimes.

I still feel pretty comfortable with my decision to have a resurfacing.

[Pat Walter]

Hi Guys

Great discussion about some very technical stuff and revision results of hip resurfacing.

I still would prefer to have a hip resurfacing.  My surgeon, Dr. De Smet, along with many other surgeons feel their patient hip resurfacing devices will last a lifetime.  I guess only time will tell that.

My one thought about all the statistics presented in the study was that many of the earlier revisions were from early hip resurfacings.  There is a possiblity that the surgeons just were not as experinced and did not place the components as well as the really experienced hip resurfacing surgeons do today.  Perhaps, if the study started today, using the results of resurfacings from the most experinced surgeons - the stats might be different.  We all know that the early placements of the components were not always perfect and that could make the outcome different then than now.

Again, I still prefer a resurfacing to keep my femur bone and have a more natural gait.  I will take my chances if my revision might be more diffiuclt than if I had a THR.  It is not enough possible negative information to change my mind.

We don't normally have a surgery so that the revision of the surgery will be good.  We have the surgery to fix a problem and get on with our lives.

There have been many studies and information I received from surgeons that suggest that a less expeirnced surgeon doing a THR can not do an optimal surgery and the resulting revision could be more diffiuclt.  A THR or resurfacing surgery has a better outcome in the hands of the most experienced surgeons.  I tell people on the THR site that they still need to choose the best surgeons.  Dr. De Smet has explained a poorly done THR results in a poor results for a revision of that THR.  Having a revision of a THR is diffiuclt and again, the most expiernced surgeons can perform a better surgery that makes the next revision easier. Less experineced surgeons can makethe revision more difficult.  It always comes down to using the most expeirnced surgeons available.

Pat

[Bionic]

Re-revision rate when first surgery was a resurfacing: 9% after 5 years
Re-revision rate when first surgery was a THR:           26% after 10 years

Resurfacing wins hands down.

[Tim Bratten]

I just saw this today. Sorry if it's old, but I took some time to look it over in detail and have a few coments. The case that interested me was the complete revision to THR. I'd have to say, I'm not at all impressed with the way the authors present their  data (by the way, I'm a mathematician). These guys look at 114 cases of total revisions (not due to infection) and found a 5% failure rate after 5 years. This is compared to more than 141,000 primary THRs that have an almost 3% failure at 3 years. Definitely comparing apples to oranges, and hardly worthy of the title "Poor outcome of revised resurfacing hip arthroplasty". If anything, I'd say this small set of data looks pretty promising for people with HR who revise to THR. One thing that is apparently true and confirmed by the study (Dr. De Smet had talked to me about this risk before he did my revision), is that cup only revision is riskier than total revision. I think it's important for people to realize this. However, if a cup only revision fails, there's still the possibility of revision to THR. In the end, the choice should also depend on the patient's personal decision about whether or not they want to risk more surgeries. When De Smet presented this risk to me before my revision, I told him that I didn't want anymore surgeries in the near future and I said he should decide, from his experience what would be best for me. So far, I'm way happy with the result. At any rate, he found infection, so that was the trump card. 

All this makes me wonder. with more than 800 revisions under the belt, how does De Smet's data compare to this smaller sample?

[Dannywayoflife]

I've quickly read the study I'll read it properly later. For me there's too many variables in there for it to mean anything. And I think the moral is go to the best surgeon you can find just as you should for a hr in the first place! If you do that then your % will be far better

[Pat Walter]

Note about Dr. De Smet - in the previous post it was mentioned Dr. De Smet had 800 revisions under his belt.  Please remember that those are not all caused by his poor hip resurfacings requiring revisions - he does many revisions of other surgeons errors!

Just wanted to make a note for perspective patients.

Pat

[Tim Bratten]

Note about Dr. De Smet - in the previous post it was mentioned Dr. De Smet had 800 revisions under his belt.  Please remember that those are not all caused by his poor hip resurfacings requiring revisions - he does many revisions of other surgeons errors!

Just wanted to make a note for perspective patients.

Pat

From the guy (me) who made "the previous post":
Pat's comment got me to expand on my signature note as well, just in case there was some confusion. I would like to emphasize this: I'm convinced Dr. Smet is one the best hip surgeons in the world and I would recommend him to a potential patient in a millisecond.