Keywords
wrist arthroplasty - pyrocarbon implant - wrist osteoarthritis - interposition - failure
Performing a wrist interposition arthroplasty after a bony resection is an old concept.
Ollier had described its principles as early as the late 18th century: the use of
an interposition intended to avoid bone fusion after an arthroplastic resection.[1] Only autologous regional fibrous tissue (capsule, anterior annular ligament) has
been recently used in arthroplastic resections of rheumatoid wrists.[2]
[3]
[4] Today, new concepts of arthroplastic interposition arise with improved knowledge
of anatomy and wrist biomechanics, as well as the use of new biomaterials such as
pyrocarbon.[5]
Pyrocarbon, with its mechanic and biotolerance characteristics (low friction coefficient,
wear resistance, density and elastic module close to those of cortical bone, and lack
of osteointegration), is today an ideal material for an interposition implant.
The recent use of pyrocarbon implants in the wrist and thumb base has proven its medium-term
reliability and validated the free interposition concept as a means for arthroplasty.[6]
[7]
[8]
[9]
[10]
Very recently, a new pyrocarbon interposition implant, the Amandys (Tornier SAS–Bioprofile,
Grenoble, France), has been proposed in an extensive articular destructions of the
wrist as an alternative to more radical or invasive surgical options such as total
wrist fusion (TWF) or total wrist arthroplasty (TWA).[11]
The concept of the Amandys arthroplasty is that of a free interposition radiocarpal
and midcarpal spacer implant made out of pyrocarbon. It replaces the lunate, the two
proximal thirds of the scaphoid, and the head of the capitate. It is a mobile spacer
with no bone fixation, thus avoiding peak constraint forces on the radius or the carpus.
It restores the congruency with the radius and with the distal carpal row. The implant
is quadric elliptic in shape. Its proximal ellipsoid curves are perpendicular to one
another and are close to the proximal anatomical curves of the scaphoid and lunate,
thus achieving optimal congruency with the radial fossa surface.
The two distal curves are less pronounced than the proximal ones. The implant has
three main axes: an anteroposterior one, corresponding to its width; a transverse
one, corresponding to its length; and a proximal distal one, corresponding to its
thickness. The implant exists in eight sizes varying in length (24 to 26 mm) and in
thickness (S, M, L, and XL).
The bony surfaces in contact with the implant can glide, roll, and even slightly rotate,
especially the carpal surface of the implant, which is less convex than the radial
one.
Thanks to minimal bony resection and to the preservation of the triquetrum and the
distal third of the scaphoid, an arthroplasty performed with this implant spares the
main extrinsic wrist ligaments such as the dorsal radiocarpal ligament, the dorsal
intercarpal ligament, and volarly the triquetrohamate ligament and the radioscaphocapitate
ligament.[12] These ligaments play an important part in the kinematics and the stability of the
wrist in several movements and especially in the dart-throwing motion.[13]
The instrumentation includes an ovoid burr, a curved periosteal elevator, a corkscrew
extractor, and eight trial implants.
Patients and Methods
Since November 2008, a prospective study was performed on a consecutive series of
16 patients complaining of pain and functional impairment of their wrist due to extensive
joint damage following a prior surgery.
The series included 14 men and 2 women with an average age of 56 years (38 to 77 years).
The dominant side was involved in eight cases. Ten patients were working at the time
of the surgery. Before surgery, all patients had standard posteroanterior (PA) and
lateral radiographs and a wrist computed tomography scan or arthroscan.
The mean delay from the first surgery was 11 years and 10 months (13 months to 32
years).
The mean follow-up after the pyrocarbon arthroplasty was 24 months (6 to 41 months).
Indications for surgery for prior failed surgery included the following:
In seven cases, the major indications were partial arthrodesis with nonunion (four
cases), major osteoarthritic degradation of the radiocarpal (two cases), or midcarpal
joints (one case). The seven initial procedures were four scaphoidectomies with a
four-bone fusion ([Fig. 1]), a scaphoidectomy with a capitolunate arthrodesis, a radiolunate arthrodesis, and
a radioscapholunate arthrodesis.
Figure 1 Four-bone fusion failure in the right wrist of a 66-year-old man. (A, B) preoperative
X-rays. (C–F) postoperative standard and dynamic X-rays at 1-year follow-up.
In five cases, the major indications were carpal silicone synovitis with extensive
wear of the radiocarpal and midcarpal joints following partial silicone implants (two
for the lunate, two for the scaphoid, and one for the trapezium) ([Fig. 2]).
Figure 2 Scaphoid silicon implant failure after 15 years in the left wrist of a 71-year-old
lady. (A, B) preoperative X-rays. (C, D) postoperative X-rays at 3-year follow-up.
Carpal bone cysts have been filled with cancellous bone graft from the radius. Range
of motion was 60 degrees in flexion extension and 40 degrees in ulnoradial inclination.
Grip strength was 88% to the contralateral side, PRWE score was 3 of 100, QuickDASH
score was 9.09 of100, and the patient was very satisfied. PRWE, Patient-Rated Wrist
Evaluation; DASH, Disability Arm Shoulder and Hand.
In two cases, the major indications were advanced Kienböck disease (Lichtman stage
IV) that had already undergone a radial shortening.
In one case, a proximal row carpectomy, and in another case, an isolated scaphoidectomy
were previously performed.
Contraindications for this arthroplasty were distal radial malunions, radiocarpal
or midcarpal instability, major bone loss, incompetent musculotendinous wrist extensors
or flexors, and ongoing infection.
Operative Technique
The procedure was done under local or regional anesthesia with a tourniquet on the
proximal part of the arm. The arthroplasty can be performed through a dorsal or a
radial approach.[11] In cases of a failed previous procedure on the wrist, only the dorsal approach was
used.
The previous skin incision was used and extended distally or proximally as necessary.
The dorsal extensor retinaculum was incised on the radial or ulnar side of the fourth
extensor compartment to preserve ligament tissue if a capsular reinforcement was needed
at the end of the procedure.
Either a midline or ligament sparing capsulotomy was used.[14]
All previous material (pins, staples, screws, plates, and implant) were removed before
any procedure on the joint.
The carpal bone resection was performed after lysis of any adhesions and a synovectomy.
First, the two proximal third of the scaphoid were removed with an oscillating saw
just above the ligament insertions of the dorsal radiocarpal ligament and the radioscaphocapitate
ligament, which must be preserved. The cut was parallel to the frontal and transversal
slopes of the radial fossa.
The lunate was then freed from all ligamentous attachments. The corkscrew was used
as a joystick during this step to preserve the anterior capsule. The head of the capitate
was then removed. The cut should be level with that of the scaphoid.
In cases of protuberant radial styloid osteophytes, a partial styloidectomy was performed
with the oscillating saw.
The radial fossa was smoothed with the ovoid burr to remove the crest between the
scaphoid fossa and the lunate fossa to obtain a homogenous concave ovoid surface along
both axes. On the carpal side, the new midcarpal joint was smoothed and reamed to
obtain a perfectly smooth and slightly concave sliding surface.
Any large dorsal and volar radial osteophytes were excised taking care to preserve
the ligament attachments. The synovectomy was completed, and any tears in the volar
capsule were repaired. A redundant volar was plicated with nonabsorbable sutures.
Any large bone cavities in the radius or the carpus were curetted and bone grafted.
The size of the implant was assessed preoperatively with transparent tracings. The
final size was determined intraoperatively, although the trial implant was in place
ensuring that there was full passive flexion extension and full radial/ulnar deviation,
the implant did not tend to rotate on its proximodistal axis, which would otherwise
indicate excess length or a lateral impingement through a cam effect, and the implant
did not tend to subluxate volarly or dorsally, which would indicate lack of radial
congruency, too little thickness, insufficient radial concavity, or redundant anterior
capsule. The distal carpal row did not dorsally subluxate due to a torn or deficient
capsule, and the distance between the ulna and the triquetrum were unchanged in comparison
with the preoperative radiographs since this indicated a lack of excess implant thickness,
and there was no impingement between the radial styloid and the scaphoid.
Additional procedures in our series included the following: removal of surgical material
in 13 patients (silicone implants, circular four-bone fusion plate, shape memory staples,
and screws), bone grafting of carpal cysts in 8 cases, and treating an ulnocarpal
impingement in 2 cases (one Wafer procedure and the other partial triquetrum resection).
If the dorsal capsule was insufficient secondary to the multiple previous operations,
it may be reinforced with a half of the extensor retinaculum flap slipped under the
extensor tendons. This procedure was performed in three of our cases.
Postoperatively, the wrist was immobilized in neutral for 15 days, followed by gentle
home motion exercises. A protective splint was worn during the night and during the
day for comfort until the fourth week postoperatively, at which time, formal therapy
was started if necessary. Unrestricted activity was allowed after 6 weeks.
Clinical Data
All measurements were performed on both operated and contralateral wrist.
Two self-administered questionnaires were used to assess functional scores: the Patient-Rated
Wrist Evaluation (PRWE) and the QuickDASH (Disability Arm Shoulder and Hand).[15]
[16]
Pain was rated on a pain scale between 0 and 10 by extracting the pain score from
the Visual Analog Scale pain scale of the PRWE questionnaire (of 50 points).
Satisfaction was rated very satisfied, satisfied, fairly satisfied, or unsatisfied.
For patients with a professional activity, the time lapse before returning to work
was noted.
Pre- and postoperative radiological measurements were done on PA and lateral wrist
radiographs including the head of the third metacarpal to assess:
-
Carpal sagittal subluxation measured on the lateral X-ray by the relation D/L
1, where D is the distance between the axis of the radius and the axis of the third metacarpal
and L
1 is the length of the third metacarpal.
-
Carpal radial deviation measured on the PA view by Shapiro's angle between the line
from the distal ulnar aspect of the radius to the tip of the radial styloid and the
line of the long axis of the second metacarpal.[17]
-
Ulnar carpal translation measured with the Youm index.[18]
[19]
-
The height of the carpus measured with the Mac Murphy and Youm index.[19]
The postoperative radiographs were also analyzed for bone reaction (lysis or periosteal
reaction) or periarticular ossification and for the implant position (subluxation
and subsidence).
Statistical Analysis
All data were analyzed (average, standard deviation [SD], median, maximum, minimum)
and then compared. Excel (Microsoft) and Statistica (StatSoft, Tulsa, OK) were used.
The paired Student t test, the Wilcoxon test, and the Pearson correlation coefficient were used to compare
data.
Results
The average range of motion was 32 degrees in extension and 36 degrees in flexion
on the operated wrist at the last follow-up ([Table 1]). The gain in extension was 1 degree but not statistically significant. The average
range of motion in flexion was comparable with the preoperative data.
Table 1
Clinical Results
|
Preoperative Data (SD)
|
Postoperative Data (SD)
|
Difference
|
p
|
|
Flexion (°)
|
35 (12)
|
36 (13)
|
+1
|
>0.05
|
|
Extension (°)
|
32 (21)
|
32 (9)
|
0
|
>0.05
|
|
Flexion extension ROM (°)
|
67 (26)
54% CL
|
68 (18)
55% CL
|
+1
|
>0.05
|
|
Ulnar inclination (°)
|
24 (18)
|
21 (8)
|
−3
|
>0.05
|
|
Radial inclination (°)
|
10 (7)
|
15 (6)
|
+5
|
0.02
|
|
Ulnoradial inclination ROM (°)
|
34 (21)
61% CL
|
36 (10)
64% CL
|
0
|
>0.05
|
|
Pronation (°)
|
75 (7)
|
78 (5)
|
+3
|
>0.05
|
|
Supination (°)
|
69 (11)
|
74 (5)
|
+5
|
>0.05
|
|
Grip (kg)
|
17 (9)
46% CL
|
19 (10)
53% CL
|
+2
|
>0.05
|
SD, standard deviation; ROM, range of motion; CL, contralateral.
The postoperative radial and ulnar deviation was not significantly different to the
preoperative one. Radial deviation increased an average of 5 degrees and ulnar deviation
decreased by 3 degrees.
The average pronation and supination improved by 7 degrees but this was not significant.
Postoperative grip strength was an average of 19 kg or 53% of the grip strength on
the opposite side. The average increase in grip strength was 2 kg, which was not statistically
significant.
Pain improved from an average of three points from 7 of 10, preoperatively, to 4 of
10, postoperatively. The average postoperative pain score was 19 of 50 (minimum = 3,
maximum = 31, SD = 9) ([Table 2]). The improvement between pre-and postoperative scores was 15 points on average.
This difference was very significant (p = 0.00035). The average postoperative total PRWE score was 33% (4 to 64, SD = 17),
namely an average decrease of 24 points. There was a significant improvement in the
global score after the procedure (p = 0.0034).
Table 2
Functional Results
|
Preoperative Data (SD)
|
Postoperative Data (SD)
|
Difference
|
p
|
|
PRWE pain score/50
|
34 (7)
|
19 (9)
|
−15
|
0.00035
|
|
PRWE function score/50
|
24 (9)
|
14 (10)
|
−10
|
0.0034
|
|
PRWE total score/100
|
57 (14)
|
33 (17)
|
−24
|
0.00034
|
|
QuickDASH score/100
|
59 (12)
|
39 (18)
|
−20
|
0.0022
|
SD, standard deviation; PRWE, Patient-Rated Wrist Evaluation; DASH, Disability Arm
Shoulder and Hand.
The average postoperative QuickDASH score was 39 (9.09 to 68.18, SD = 18). The average
decrease was of 39 points (p = 0.0022). The average score for each question improved significantly.
Patients were either very satisfied (nine patients), satisfied (six patients), or fairly satisfied (one patient) with the surgical outcome.
Among the 10 patients who were still professionally active, at the last follow-up
5 had resumed their work at the same position, 3 had changed positions, and 2 were
still on sick leave.
The mean postoperative carpal sagittal subluxation was 0.07 (0 to 0.19, SD = 0.07).
The difference (0.01) with the mean preoperative value was not significant ([Table 3]).
Table 3
Radiographic Results
|
Preoperative Data (SD)
|
Postoperative Data (SD)
|
Difference
|
p
|
|
Carpal sagittal subluxation
|
0.08 (0.09)
|
0.07 (0.07)
|
0.01
|
>0.05
|
|
Radial deviation
|
115° (7°)
|
109° (10°)
|
−6°
|
0.039
|
|
Carpal ulnar translation
|
0.29 (0.04)
|
0.33 (0.109)
|
+0.04
|
>0.05
|
|
Carpus height
|
0.42 (0.05)
|
0.40 (0.06)
|
−0.02
|
>0.05
|
SD, standard deviation.
The average radial deviation angle (Shapiro's angle) was 109 degrees after surgery
(94 to 125 degrees, SD = 10 degrees). The difference (−6 degrees) with the preoperative
measurements was significant.
The Youm index was on average 0.33 (0.17 to 0.55, SD = 0.109), which is within the
normal values for this index. The difference with the average preoperative value was
not statistically significant.
The average carpal height was 0.40 (0.28 to 0.5, SD = 0.06). The difference (−0.02)
with the average preoperative value was not statistically significant.
No migration or subsidence was observed at the final follow up, even in cases of bone
cysts filled with cancellous bone grafts ([Fig. 2]). No dislocation was observed.
In one patient, there was moderate palmar subluxation (below a third of the total
implant width) in the neutral position radiographically, but this was not seen with
motion studies ([Fig. 3]).
Figure 3 Anterior subluxation of the implant seen on the lateral view in neutral position.
Relocation of the implant on dynamic radiographs. (A) neutral position; (B) extension;
and (C) flexion.
There was one case of major palmar subluxation (half of the implant width). The implant
position did not change over time and did not seem to alter the final clinical and
functional results in these two cases.
Subchondral calcification around the implant was found in nine cases at the latest
X-rays.
No periarticular ossification was observed.
There were no cases of infection. Two patients required repeat surgery.
In one case of a 38-year-old man with advanced Kienböck disease, the implant rotated
90 degrees on its proximodistal axis. This patient had undergone a radius osteotomy
13 years before. A repositioning of the implant with a repeat reaming of the radial
fossa was required twice in this patient. At the last follow-up (36 months), the flexion-extension
range of motion was 70 degrees, grip strength was 80% of the opposite side, the PRWE
score was 14.5 of 100, the QuickDASH score was 18.8 of 100, and the patient was very
satisfied.
In another case, a 52-year-old man had a failed four-corner fusion 8 years before.
At 6 weeks, the implant was subluxated anteriorly and had rotated 90 degrees on its
proximodistal axis. The implant was changed to a thinner one, the articular surfaces
were reamed again, and a volar capsule was plicated with a micro anchor. At the 6-month
follow-up, the patient was still recovering with decreasing pain and had not yet returned
to work.
Discussion
The Amandys arthroplasty was used in this series as a salvage solution for mid-term
or long-term failures of previous wrist surgeries of nonrheumatoid wrists. We did
not use this arthroplasty in failed rheumatoid wrist surgery, because in most of these
indications, the soft tissues, the carpal alignment, or the bone stock are not suitable.
The aim of this new arthroplasty is a pain free and mobile wrist.[11] The lack of wrist motion is correlated with functional impairment to perform daily
tasks and increases the perceived disability.[20]
[21] Preserving wrist mobility is in fact the most important functional priority for
patients along with pain relief.[22]
[23] This new motion-preserving procedure was therefore proposed as an alternative to
TWF, which could be in the past, the standard procedure for our patients, who were
relatively young (aged 56 years on average) and most of them still working. Since
we used the Amandys implant, we did not perform any more TWF in these indications.
Our preliminary results showed a functional improvement for all patients. The decrease
in pain was the main factor for this. Range of motion was nearly on average the same
as preoperatively and no patient lost mobility after the surgery. Average global strength
increased but the improvement was not significant in comparison with preoperative
data.
The implant was well tolerated clinically and radiographically. After the arthroplasty,
the wrists remained correctly aligned without any anterior or posterior subluxation
on the lateral view or ulnar deviation on the PA view. Radial deviation decreased
significantly. This may explain the postoperative gain of 5 degrees in radial inclination
and the postoperative loss of 3 degrees in ulnar inclination.
The implant did not subside and the carpal was maintained. The revision rate was low
(12.5%) and no implant removal or conversion to a total arthrodesis was required.
Complications and the volar subluxation of the implant seen in two cases were mostly
linked to technical errors in relation with the surgical learning curve of this new
technique. These problems were related to the size of the implant, the evaluation
of the anterior capsular distension, or the bone preparations.
The treatment reported in the literature for failed partial carpal arthrodesis, carpal
silicone implant, first-row resection, or advanced Kienböck disease surgery is usually
TWF.[24]
[25]
[26]
[27]
[28]
[29] TWF results are not always satisfactory.[30]
[31] This procedure has complications and the results on pain and function, especially
in patients with multiple surgeries are sometimes disappointing.[31] Wrist motion preservation is the main advantage of the Amandys procedure compare
with TWF. However, functional results should be compared in the future for the same
indications.
The other options such as TWA are exceptionally proposed in these indications given
the high risks of complications and revision in patients with high functional demands.[32] The indications of TWA are limited to patients with low functional demand, usually
rheumatoid arthritis patients, because the recommended maximum load allowed should
be low. These restrictions do not apply to the implant in our series. Furthermore,
the Amandys procedure is less invasive compared with a TWA. Bone and ligament resections
are minimal. In cases of failure, TWF or TWA can be easily performed as salvage procedures.
Because it is a free interposition, the Amandys procedure requires an efficient anterior
and dorsal capsule to ensure both wrist and implant stability. Correct wrist alignment
is mandatory as well. TWA does not imply such demanding restrictions and TWF even
less.
According Mulford and coworkers' meta-analysis, the rates of conversion to TWF after
a four-bone fusion or a first-row carpectomy are respectively 2.9 and 3.9%.[33]
Failures of four-bone fusions are often the consequence of nonunion due to bad technique
or failure of the osteosynthesis material. A revision for a new arthrodesis is often
possible.[34]
[35]
[36] When the lunate cannot be salvaged or that the radiolunate joint is degenerative,
an interposition arthroplasty with a pyrocarbon implant has proven to be a good alternative
to total arthrodesis ([Fig. 1]). This arthroplasty is also possible in cases of failed first-row resections,[11] which are more at risk for arthritic degeneration than four-corner fusions.[33]
Failures after partial silicone implants are often due to silicone debris leading
to foreign body synovitis and intracarpal bone cysts. Kaarela et al observed a 41%
rate of this type of complication with lunate implants at 5.6 years follow-up.[37] Vinnars et al had a 62% rate at 10 years follow-up for scaphoid implants[29] and Minami et al had a 33% rate at 15.3 years follow-up with trapezial implants.[38] Wrist arthrodesis is a salvage solution in cases of progressive symptomatic degeneration.
The pyrocarbon interposition arthroplasty performed in five of our patients maintained
useful wrist mobility and had a satisfactory functional outcome ([Fig. 2]).
For both patients with advanced Kienböck disease who had had a prior radius shortening,
the results were similar. Several technical points need to be highlighted to carry
out this pyrocarbon interposition arthroplasty:
In revision cases of four-bone fusion, we recommend to be particularly careful about
the anterior capsule distension in front of the lunate that is often in DISI position.
Plicating or reinserting the capsule on the anterior edge of the radius prevents anterior
subluxation or 90 degrees proximodistal rotation of the implant. The absence of the
scaphoid did not increase radial implant deviation. The scaphoid space is filled with
a fibrous scarring tissue that should be preserved.
In cases of failed first-row carpectomy, we recommend to check for a stylotrapezial
impingement and to treat it subsequently with a radial styloidectomy.
In cases of failed silicone implant, complete synovectomy should be performed and
the bone cysts facing the implant should be empty and filled with impacted cancellous
bone ([Fig. 2]).
The limitations of our study are its short follow-up and the limited number of patients
in the series. Clinical and radiological studies on larger series with a longer follow-up
will help determine the advantages of this arthroplasty and its specific indications
in relation to arthrodesis indications. The small size of this arthroplasty and its
minimally invasive procedure does not burn the bridge for a TWF in case of failure.
The Amandys interposition arthroplasty is a new therapeutic option in cases of difficult
revisions procedures on the wrist. Preliminary results demonstrated that it is a reliable
alternative to TWF. It is a salvage procedure that maintains wrist mobility and improves
greatly pain and function in patients.
Conflict of Interest
Dr. Bellemère is a consultant agreement with Tornier—Bioprofile (Grenoble, France).
Occasional involvements include expert reports and advisory services (Tornier—Bioprofile).
Dr. Bellemere has attended conferences as contributor (Tornier—Bioprofile). Dr. Gaisne
has attended conferences as audience member (Tornier—Bioprofile). Dr. Loubersac, Dr.
Maes-Clavier, Dr. Kerjean, and Dr. Collon have no conflict of interest to disclose.
