Abstract
Background: Second-generation total ankle arthroplasty has been
reported to have good intermediate-term results. The purpose of the present
study was to report on the cause and frequency of reoperation and failure
after total ankle arthroplasty and to determine demographic and clinical
predictors of reoperation and failure.
Methods: Three hundred and six consecutive primary total ankle
arthroplasties were performed with use of the DePuy Agility Total Ankle System
between 1995 and 2001. At a mean of thirty-three months after the
arthroplasty, we retrospectively reviewed the records with regard to patient
age, gender, the indications for the index procedure, adjuvant procedures, the
timing and frequency of reoperation, and the indications for and the type of
reoperations performed. Kaplan-Meier analysis was performed to determine the
rate of prosthetic survival, and Cox regression analysis was performed to
determine predictors of reoperation and failure.
Results: Eighty-five patients (28%) underwent 127 reoperations
(involving 168 procedures) after primary total ankle arthroplasty. The most
common procedures at the time of reoperation were débridement of
heterotopic bone (fifty-eight), correction of axial malalignment (forty), and
component replacement (thirty-one). Eight patients underwent below-the-knee
amputation. Age was found to be the only significant predictor of reoperation
and failure after total ankle arthroplasty. The five-year survival rate with
reoperation as the end point was 54%. The five-year survival rate with failure
as the end point was 80% for all patients and 89% for patients who were more
than fifty-four years of age. The prosthesis could not be salvaged in nine
ankles (2.9%); the inability to salvage the prosthesis was most often due to
loosening or infection.
Conclusions: We noted a relatively high rate of reoperation after
total ankle arthroplasty with this second-generation device. Younger age was
found to have a negative effect on the rates of reoperation and failure. Most
prostheses could be salvaged; however, the functional outcome of this
procedure is uncertain.
Level of Evidence: Therapeutic study, Level IV (case
series [no, or historical, control group]). See Instructions to Authors for a
complete description of levels of evidence.
Second-generation prostheses have made total ankle arthroplasty a viable
alternative to ankle arthrodesis in the treatment of severe arthrosis. The
most commonly implanted device in the United States, and the only device
approved by the Food and Drug Administration, is the Agility Total Ankle
System (DePuy, Warsaw, Indiana), a semiconstrained two-component design.
Pyevich et al.1
studied the intermediate-term results associated with this system and reported
that 93% of patients had little or no pain at an average of 4.8 years after
implantation. However, they also noted high rates of component migration and
delayed union or nonunion of the syndesmosis. Approximately 6% of the patients
underwent a reoperation involving removal of components and arthrodesis (1%)
or component revision (5%).
First-generation total ankle arthroplasties frequently
failed2-14.
For example, Kitaoka and
Patzer8 noted a high
rate of complications and failure in association with the Mayo total ankle
arthroplasty, with a 41% rate of reoperation. Most reoperations involved
removal of components and arthrodesis for the treatment of persistent pain.
Other
investigators15-17
also have reported high rates of salvage arthrodesis.
We noted a high rate of reoperation in a large series of total ankle
arthroplasties performed with use of a second-generation implant design. The
purposes of the present study were to determine the cause and frequency of
reoperation and failure (defined as removal or replacement of components,
ankle arthrodesis, or below-the-knee amputation) after a second-generation
primary total ankle arthroplasty and to determine demographic and clinical
predictors of reoperation and failure.
Three hundred and six consecutive primary total ankle arthroplasties in 303
patients were performed at our institution between April 18, 1995, and March
28, 2001. All arthroplasties were performed or directly supervised by the
senior author (S.T.H. Jr.). Forty-eight percent of the patients were female,
and 52% were male. The average age of the patients was 53.5 ± 14.2
years (range, nineteen to eighty-five years). The indication that had led to
the total ankle arthroplasty was posttraumatic osteoarthrosis for 198 (65%),
primary osteoarthrosis (including arthrosis due to chronic ligamentous laxity)
for seventy-seven (25%), prior ankle arthrodesis for seventeen (5.6%),
systemic joint disease for thirteen (4.2%), and both posttraumatic
osteoarthrosis and systemic joint disease for one. Two patients with primary
osteoarthrosis also had an underlying musculoskeletal affliction
(hemochromatosis and Apert's syndrome in one patient each). Fifty-eight
percent of the patients underwent adjuvant surgical procedures to treat axial
malalignment or instability at the time of the total ankle arthroplasty and
were therefore considered to have had a "complex" total ankle
arthroplasty according to our criteria. These data are summarized in
Table I. In all cases, the
indication for primary total ankle arthroplasty was either (1) severe pain in
the ankle joint that had been refractory to nonoperative measures and that was
associated with radiographic evidence of degenerative joint changes or (2)
patient dissatisfaction with a previous ankle arthrodesis.
The prosthesis that was used in all cases was the Agility Total Ankle
System (DePuy), which is a semiconstrained implant with a titanium tibial
component and a cobalt-chromium talar component, both of which are placed
without cement. Implant fixation is secured by means of bone ingrowth, which
requires an arthrodesis of the tibiofibular syndesmosis (performed at the time
of the total ankle arthroplasty). The components were available in small,
medium, and large sizes until February 1998. After that time, the prosthesis
was available in six different sizes. The operative technique used for the
total ankle arthroplasty was performed according to the principles dictated by
the developers of the
device18.
In some patients, one or several adjuvant procedures (such as tibial
osteotomy) were performed simultaneously with, or occasionally prior to, the
total ankle arthroplasty in order to correct major malalignment, instability,
muscle imbalance, or joint contracture. These adjuvant procedures included
osteotomies, joint fusions, tendon transfers, and tendon-lengthening
procedures involving the distal part of the tibia, the hindfoot, or the
midfoot. In patients with varus malalignment of the hindfoot, the components
were maximally medialized and a peroneus longus-to-peroneus brevis tendon
transfer (or in extreme cases, a posterior tibial tendon-to-peroneus brevis
transfer) often was performed to improve eversion strength. In some cases, a
valgus-producing calcaneal osteotomy also was performed, sometimes with a
dorsiflexion osteotomy of the first metatarsal. In patients with valgus
malalignment of the hindfoot, the components were lateralized and correction
of the alignment and stabilization of the medial column were achieved with
medial and/or lateral column arthrodeses as well as augmentation of the
posterior tibial tendon, usually with the flexor digitorum longus tendon.
Occasionally, a subtalar or triple arthrodesis was performed to stabilize
valgus or varus malalignment of the hindfoot that was not correctable with
soft-tissue procedures or osteotomy. For patients with preexisting lateral
ligamentous insufficiency with anterior extrusion of the talus, a lateral
ligament reconstruction was
performed19 with or
without a peroneus longus-to-peroneus brevis transfer, and cast immobilization
or external fixation was used to maintain the position of the talus. Nearly
all patients underwent an Achilles tendon lengthening or gastrocnemius
recession
procedure20 to
correct a tight heel cord or a gastrocnemius equinus contracture before or at
the time of the total ankle arthroplasty. In some cases, a previous ankle
fusion was resected in order to implant the prosthesis. This could be done
only for patients who had had retention of certain anatomic structures, in
particular, the lateral malleolus, the medial malleolus, and the deltoid
ligament.
The indication for reoperation after the primary total ankle arthroplasty
was substantial ongoing ankle pain that was refractory to pain-management
measures and that was associated with radiographic and/or clinical evidence of
various complications or failure. Components were replaced when there was pain
and no radiographic sign of osseointegration within the expected time-course
or when the components were malpositioned or loose as indicated by subsidence,
migration, or tilting. Removal of impinging osseous overgrowth within the
joint was indicated when there was painful impingement or ankylosis and
radiographic evidence of osseous overgrowth in the medial gutter, lateral
gutter, and/or posterior tibiotalar articulation. Repair of the site of the
tibiofibular syndesmotic arthrodesis was performed when there was radiographic
evidence of nonunion of that area, with substantial pain. Extra-articular
realignment procedures were indicated when there was a new or recurrent gross
axial deformity that prevented the foot from achieving a plantigrade position.
These procedures included periarticular osteotomies (e.g., tibial, calcaneal,
and midtarsal osteotomies), tendon transfers and lengthenings, and arthrodeses
(e.g., subtalar and midfoot arthrodeses). The indication for irrigation and
débridement of the joint with retention of the components was acute
deep infection within the joint (that is, a deep infection occurring within
three months after the procedure). For late infection (that is, an infection
occurring more than three months after the procedure), the implants were
removed and an antibiotic spacer was inserted. In the one patient in whom a
late infection occurred, an ankle arthrodesis was performed with use of
structural bone graft from the posterior iliac crest. Open reduction and
internal fixation was performed when the talus fractured intraoperatively
during component exchange or during the postoperative course. The decision to
undergo an amputation was made by the patient, after consultation with our
limb viability and amputation team, when the foot remained dysfunctional
following numerous failed reconstructive attempts.
The data on each patient were obtained from a review of the preoperative
clinical notes, primary surgical report, followup reports, and reoperation
reports. The evaluation was carried out independent of the surgeon who had
performed the operative procedure. The chart review took place at a mean of 33
± 18 months (range, four to seventy-five months) after the index total
ankle arthroplasty.
Regression analyses using Cox proportional-hazards modeling were performed
to determine whether patient age at the time of primary total ankle
arthroplasty, gender, the indication for total ankle arthroplasty, and a
factor that we termed the "complexity" of the total ankle
arthroplasty had an impact on the hazard of reoperation or failure (that is,
the instantaneous probability of reoperation or failure at any time). The
primary total ankle arthroplasty was considered to be "complex" if
adjuvant procedures other than heel-cord lengthening or gastrocnemius
recession were performed to correct axial malalignment or instability, as
previously described. Kaplan-Meier survivorship analyses were performed with
reoperation and failure (defined as removal or replacement of components,
ankle arthrodesis, or below-the-knee amputation) as end points. All patients
who did not have the outcome of interest (reoperation or failure) before the
end of follow-up were censored.
During the study period, eighty-five ankles in eighty-five patients (28%)
required a total of 127 reoperations (see Appendix). Forty-five percent of the
patients who required a reoperation were female, and 55% were male. The
average age of these patients at the time of the primary procedure was 49.7
± 12.3 years (range, twenty-six to seventy-six years). The indication
for total ankle arthroplasty was posttraumatic osteoarthrosis for fifty-four
patients (64%), primary osteoarthrosis for twenty-two (26%), previous ankle
fusion for six (7%), systemic joint disease for two, and posttraumatic
osteoarthrosis and systemic joint disease for one. Fifty-two percent of the
primary total ankle arthroplasties in the patients who required a reoperation
were considered to have been "complex."
Fifty-seven patients had one reoperation at a mean of 17.8 months (range,
0.5 to sixty-one months) after the primary procedure. Eighteen patients had
two reoperations; the mean time between the primary procedure and first
reoperation was 13.7 months (range, one to forty-three months), and the mean
time between first and second reoperations was 10.6 months (range, 0.1 to
thirty-two months). Nine patients had three reoperations; the mean time
between the primary procedure and the first reoperation was 10.8 months
(range, two to thirty-four months), the mean time between first and second
reoperations was 7.2 months (range, 0.1 to thirty-two months), and the mean
time between second and third reoperations was five months (range, 0.2 to
twenty-six months). One patient required seven reoperations at four, twelve,
eight, one, 0.1, one, ten-month intervals, respectively.
The 127 reoperations involved 168 surgical procedures, including
fifty-eight gutter débridements, forty extra-articular procedures,
thirty-one component replacements, fourteen syndesmotic nonunion repairs,
eleven irrigation and débridement procedures for the treatment of
infection, eight below-the-knee amputations, three talar fracture fixations,
one ankle arthrodesis, one split-thickness skin-grafting procedure for the
treatment of wound breakdown, and one prosthetic removal and spacer
implantation (Fig. 1).
Kaplan-Meier analysis as a function of time since the primary total ankle
arthroplasty revealed that the cumulative five-year (sixty-one-month) survival
rate (and 95% confidence interval) with reoperation as the end point was 54%
± 11.5% (Fig. 2).
Thirty-three ankles (10.8%) were considered to have had a failed total
ankle arthroplasty (defined as removal or replacement of components, ankle
arthrodesis, or below-the-knee amputation). Kaplan-Meier analysis revealed
that the cumulative five-year (sixty-one month) survival rate (and 95%
confidence interval) with failure as the end point was 80% ± 8.7%
(Fig. 3). Ultimately, all but
nine joints (2.9%) were salvaged with reoperation. The leading cause of
inability to salvage a prosthesis was loosening of the talar component, which
could not be corrected in six of the twenty-two joints in which it was
evident. The next most common reason for failure was infection, which led to
failure in three of the five ankles in which it was evident.
Seven of the eight patients who underwent below-the-knee amputation had a
history of severe trauma and multiple surgical procedures prior to the total
ankle arthroplasty. All but one of these patients had been considering an
amputation as a surgical option before undergoing total ankle arthroplasty.
The total ankle arthroplasties in these patients failed because of persistent
pain associated with osseous overgrowth and failure of the talar component
(four patients) or because of infection (three patients). An amputation also
was performed in a patient who had had a severe cavovarus clawfoot deformity
before surgery that was difficult to correct. This patient had subsidence of
the talar component and inexplicable persistent severe edema without proven
infection. No patient with normal preoperative alignment underwent an
amputation.
Only one patient had an ankle arthrodesis. In this patient, an ankle fusion
was resected at the time of the total ankle arthroplasty. Osseous impingement
and loosening of the talar component developed, and the patient underwent a
second operation. Postoperatively, a deep infection developed and the patient
was treated with removal of the components and arthrodesis with use of
structural bone graft from the iliac crest. At the time of the last follow-up,
the ankle was successfully fused.
Cox regression analysis revealed that age at the time of the primary total
ankle arthroplasty was the only covariate that had an impact on the hazard of
reoperation and failure after total ankle arthroplasty. Each one-year increase
in age corresponded with a 1.9% relative decrease in the hazard of reoperation
(p < 0.05) and a 3.5% decrease in the hazard of failure (p < 0.05). The
one patient who had a diagnosis of both posttraumatic osteoarthrosis and
systemic joint disease was excluded from the analysis.
Cox regression modeling also was done to determine the impact of a
categorical predictor representing age on the hazard of failure, with time
elapsed since the primary total ankle arthroplasty as the dependent variable
of interest. The validity of including an age-group variable was verified with
use of the technique described by Kalbfleisch and
Prentice21. The
results of this analysis showed that patients with a median age of fifty-four
years or less had a 1.45-times greater risk of reoperation and a 2.65-times
greater risk of failure than did patients who were older than fifty-four
years, at any time, with all other covariates being held constant (p <
0.05). The estimated five-year survival rate with reoperation as the end point
was 0.51 (95% confidence interval, 0.37 to 0.70), at sixty-one-months, for
patients who were fifty-four years old or younger and 0.65 (95% confidence
interval, 0.52 to 0.82), at sixty months, for those who were more than
fifty-four years old. The estimated five-year survival rate with failure as
the end point was 0.74 (95% confidence interval, 0.60 to 0.91), at sixty-one
months, for patients who were fifty-four years old or younger. In comparison,
the estimated forty-seven-month survival rate with failure as the end point
was 0.89 (95% confidence interval, 0.80 to 0.99) for those who were more than
fifty-four years old. This function stops at forty-seven months because no
failures were observed in this age-group beyond that time, which led to
difficulty in predicting five-year survival. This finding is depicted in the
cumulative survivorship curves in the Appendix and is summarized in
Table II.
No evidence of a learning-curve effect was noted when the results for the
first patients who underwent total ankle arthroplasty were compared with the
results for later patients. The first fifty patients fared no worse than the
next 150 patients in terms of the proportion who had a reoperation and/or
failure (Fig. 4). Beyond the
first 200 patients, a decreased proportion of reoperation and failure is
evident because of a shorter duration of follow-up. These observations were
statistically confirmed with use of Cox proportional-hazards modeling, which
demonstrated that the hazard of reoperation or failure was not increased in
the earlier patients compared with the later patients, with all other
covariates held constant.
Avariety of reoperations were performed after primary total ankle
arthroplasty, in slightly more than one-fourth of the patients. The types of
procedures that were performed were dictated by the indications. The most
common procedure was joint débridement for osseous impingement; the
next most common procedures were extra-articular procedures for axial
malalignment and component replacements (usually involving the talar
component). These three types of procedures accounted for 77% of the total
procedures performed and were roughly fourfold more common than all
reoperations for syndesmotic nonunions, infection, talar fractures, and
wound-healing problems combined. In contrast with the high rate of
wound-healing complications requiring operative intervention that is
associated with first-generation devices, only one wound-healing problem that
required surgical treatment was incurred in our patient population.
Component failure involving migration or subsidence nearly always involved
the talar component. Osseous over-growth often was associated with a loose
talar component and was discovered at the time of gutter débridement
or, occasionally, at a later date. The component often was found to have
subtle signs of loosening rather than gross loosening. In these cases, the
component was not well bonded to the talus, allowing a thin elevator to be
inserted in the interface and the component to be detached without difficulty.
This sort of loosening was not recognized in many of the early
débridements but was noted later in some patients who underwent
multiple débridements. It may have been missed initially, or it may not
have developed until a later date. If it was missed and the osseous overgrowth
was a consequence of a loose talar component, then perhaps an early exchange
of the component would have avoided further débridements for
overgrowth. Few problems were encountered with the tibial component.
The revision rate in the present study was higher than that in the study by
Pyevich et al.1, but
there were important differences between the patient populations. The revision
rate in that study was 5%, with a 1% rate of component removal and
arthrodesis, at an average of 4.8 years. In the present study, 10% of the
patients underwent component revision at an average of twenty-one months.
However, the patients in the present study were younger than those in the
study by Pyevich et al. (average age, 53.5 compared with sixty-three years).
Apparently, none of the patients in that study had substantial malalignment
that required correction before or at the time of total ankle arthroplasty,
although this factor may not have affected the rate of revision, as discussed
later in this section. The percentage of patients who underwent total ankle
arthroplasty because of posttraumatic arthrosis was higher in the present
study than it was in the other study (66% compared with 45%). In addition, the
study by Pyevich et al. focused on the results of procedures performed by the
inventor of the Agility ankle prosthesis, who had intimate familiarity with
the implant. Any combination of these factors may have led to a higher rate of
revision in our patient population.
Interestingly, the performance of adjuvant procedures for substantial
malalignment or instability, or the resection of a fusion, did not increase
the risk of reoperation or failure. This finding indicates that total ankle
arthroplasty can be performed just as successfully in patients with complex
reconstructions to correct preexisting malalignment and instability or in
patients who have undergone a previous ankle arthrodesis and have an intact
deltoid ligament.
It is possible that additional patients in this study may have gone
elsewhere for reoperation without our knowledge. We are not aware of any such
cases, but the possibility exists nevertheless. For this reason, the
reoperation and failure rates that we have reported should be regarded as the
lower limits of such rates, with the actual rates perhaps being somewhat
higher.
The adverse effect of age on the risk of reoperation and failure after
total ankle arthroplasty is consistent with the well-known effect of age on
these rates after total knee and total hip
arthroplasty22,25.
However, the present study also indicates that the rates of reoperation and
failure after primary total ankle arthroplasty are substantially higher than
those associated with most modern total hip and knee designs. Berry et
al.22 noted that
the five-year survival rate after Charnley total hip arthroplasty was 96% with
reoperation as the end point and 98% with component removal or revision as the
end point. High survival rates (>95% at ten years) also have been reported
in association with other types of total hip
prostheses23,24.
Rand et al.25 noted
an overall five-year survival rate of 96% for later-generation total knee
arthroplasty designs, although some designs had much lower survival rates in
younger patients. The complex kinematics of the ankle joint and the small
weight-bearing surface of the prosthesis are likely to be major contributing
factors to the lower survival rate after total ankle arthroplasty and
therefore present important design challenges.
In conclusion, we noted a relatively high rate of reoperation due to
complications that arose after total ankle arthroplasty. Age was the only
patient-related factor that was found to have an adverse effect on both the
reoperation rate and the failure rate. In most patients with complications,
the prosthesis could be salvaged. The functional outcome of these salvaged
prostheses remains to be seen.
A table showing specific data on all patients undergoing reoperation as
well as illustrations showing Kaplan-Meier survival curves are available with
the electronic versions of this article, on our web site at
(go to the article citation and click on "Supplementary Material")
and on our quarterly CD-ROM (call our subscription department, at
781-449-9780, to order the CD-ROM).
Pyevich MT, Saltzman CL, Callaghan
JJ, Alvine FG. Total ankle arthroplasty: a unique design. Two to
twelve-year follow-up. J Bone Joint Surg Am.1998;80:
1410-20.801410
1998
[PubMed]
Bolton-Maggs BG, Sudlow RA, Freeman
MA. Total ankle arthroplasty. A long-term review of the London Hospital
experience. J Bone Joint Surg Br.1985;67:
785-90.67785
1985
[PubMed]
Buechel FF, Pappas MJ.
Survivorship and clinical evaluation of cementless, meniscal-bearing total
ankle replacements. Semin Arthroplasty.1992;3:
43-50.343
1992
[PubMed]
Demottaz JD, Mazur JM, Thomas WH,
Sledge CB, Simon SR. Clinical study of total ankle replacement with gait
analysis. A preliminary report. J Bone Joint Surg Am.1979;61:
976-88.61976
1979
[PubMed]
Dini AA, Bassett FH 3rd.
Evaluation of the early result of Smith total ankle replacement.
Clin Orthop.1980;146:
228-30.146228
1980
[PubMed]
Helm R, Stevens J. Long-term
results of total ankle replacement. J Arthroplasty.
1986;1:
271-7.1271
1986
[PubMed][CrossRef]
Herberts P, Goldie IF, Korner L,
Larsson U, Lindborg G, Zachrisson BE. Endoprosthetic arthroplasty of the
ankle joint. A clinical and radiological followup. Acta Orthop
Scand. 1982;53:
687-96.53687
1982
[CrossRef]
Kitaoka HB, Patzer GL. Clinical
results of the Mayo total ankle arthroplasty. J Bone Joint Surg
Am.1996;78:
1658-64.781658
1996
Lachiewicz PF, Inglis AE, Ranawat
CS. Total ankle replacement in rheumatoid arthritis. J Bone
Joint Surg Am.1984;66:
340-3.66340
1984
Newton SE 3rd. Total ankle
arthroplasty. Clinical study of fifty cases. J Bone Joint Surg
Am.1982;64:
104-11.64104
1982
Stauffer RN, Segal NM. Total
ankle arthroplasty: four years' experience. Clin
Orthop. 1981;160:
217-21.160217
1981
Takakura Y, Tanaka Y, Sugimoto K,
Tamai S, Masuhara K. Ankle arthroplasty. A comparative study of cemented
metal and uncemented ceramic prostheses. Clin Orthop.
1990;252:
209-16.252209
1990
[PubMed]
Unger AS, Inglis AE, Mow CS, Figgie
HE 3rd. Total ankle arthroplasty in rheumatoid arthritis: a long-term
follow-up study. Foot Ankle.1988;8:
173-9.8173
1988
[PubMed]
Wynn AH, Wilde AH. Long-term
follow-up of the Conaxial (Beck-Steffee) total ankle arthroplasty.
Foot Ankle.1992;13:
303-6.13303
1992
[PubMed]
Groth HE, Fitch HF. Salvage
procedures for complications of total ankle arthroplasty. Clin
Orthop. 1987;224:
244-50.224244
1987
Kitaoka HB. Fusion techniques for
failed total ankle arthroplasty. Semin Arthroplasty.1992;3:
51-7.351
1992
[PubMed]
Stauffer RN. Salvage of painful
total ankle arthroplasty. Clin Orthop.
1982;170:
184-8.170184
1982
[PubMed]
Saltzman CL, Alvine FG. The
Agility total ankle replacement. Instr Course Lect.2002;51:
129-33.51129
2002
[PubMed]
Hansen ST Jr.Functional reconstruction of the foot and ankle.
Philadelphia: Lippincott Williams and Wilkins; 2000.
Posttraumatic and degenerative problems in the joints; p
146-9.146
2000
Hansen ST Jr.Functional reconstruction of the foot and ankle.
Philadelphia: Lippincott Williams and Wilkins; 2000. Tendon
transfers and muscle-balancing techniques; p
415-7.415
2000
Kalbfleisch JD, Prentice RL.The statistical analysis of failure time data. 2nd ed.
Hoboken, NJ: J Wiley; 2002. p
121.121
2002
Berry DJ, Harmsen WS, Cabanela ME,
Morrey BF. Twenty-five-year survivorship of two thousand consecutive
primary Charnley total hip replacements: factors affecting survivorship of
acetabular and femoral components. J Bone Joint Surg
Am.2002;84:
171-7.84171
2002
Archibeck MJ, Berger RA, Jacobs JJ,
Quigley LR, Gitelis S, Rosenberg AG, Galante JO. Second-generation
cementless total hip arthroplasty. Eight to eleven-year results. J
Bone Joint Surg Am.2001;83:
1666-73.831666
2001
Urban JA, Garvin KL, Boese CK, Bryson
L, Pedersen DR, Callaghan JJ, Miller RK. Ceramic-on-polyethylene bearing
surfaces in total hip arthroplasty. Seventeen to twenty-one-year results.
J Bone Joint Surg Am.
2001;83:
1688-94.831688
2001
[PubMed]
Rand JA, Trousdale RT, Ilstrup DM,
Harmsen WS. Factors affecting the durability of primary total knee
prostheses. J Bone Joint Surg Am. 2003;
85: 259-65.85259
2003
[PubMed]