Long-Term Survival Rate and Outcomes for Amandys Implant: A 10-Year Minimum Follow-up of 84 Cases

Abstract

Amandys is a radiocarpal pyrocarbon interposition implant indicated for degenerative wrists, showing promising medium-term results. The aim of this study was to evaluate the outcomes of the implant with a minimum follow-up of 10 years. This retrospective monocentric study included 84 implants in 76 patients who underwent the procedure. Preoperative and follow-up assessments included clinical evaluation, functional scores, and X-rays measurements. A total of 41 implants in 39 patients were evaluated at a mean follow-up of 12.64 years. At the final follow-up, average scores for pain (visual analogue scale) was 2.3, Patient-Rated Wrist Evaluation 26.3, QuickDASH 30.8. Grip strength was 71% of contralateral side. Flexion and extension were, respectively, 32 and 37 degrees. The satisfaction rate was 94.9%. There were 10 revision surgeries (13.6%); the survival rate was 86.4%. Radiological findings remained stable over time. The Amandys implant proves to be a durable alternative to wrist fusion and arthroplasty. The survival rate does not deteriorate at long term.

Level of Evidence  IV.

The therapeutic management of advanced wrist collapse and major wrist destruction is challenging and remains a controversial issue today. There are many therapeutics available ranging from wrist denervation to total wrist fusion (TWF) or total wrist arthroplasty (TWA). The goal being to balance the grip strength, range of motion, and pain.

TWF is still considered as a gold standard for advanced wrist destruction despite impaired personal hygiene, functional deficits, and incomplete pain relief. ¹

The last decade has seen a significant decline in the number of TWA procedures, ² whereas number of TWF remained constant. ³

When looking at the long-term review of fourth generation of TWA, we can see a slow decay of implant survival rates for the Remotion, the Maestro, the Universal 2, and the Motec prosthesis ^{4 5 6 7 8 9 10 11 12 13} with a survival rate ranging from 35 to 95% on the mid-term and long-term review.

The objective of this study is to position the Amandys implant among fourth generation of total prostheses.

Amandys is a pyrocarbon implant used as a radiocarpal interposition arthroplasty. This last generation of wrist motion-preserving procedures has been proposed as therapeutic alternative to TWF and TWA for extended wrist destruction. ^{14 15}

Preliminary, short-term and medium-term outcomes of Amandys were encouraging. At 5 years, the survival rate using any revision as endpoint was estimated at 89%; ¹⁶ however, the long-term outcomes have not yet been reported.

The aim of this study was to report the survival rate, the clinical and radiological outcomes with Amandys implants with a minimum follow-up of 10 years.

Our main hypothesis was to observe a stable curve in the long-term survival rate after 5 years with the Amandys implant, along with stable functional results.

Materials and Methods

We retrospectively reviewed continuous, prospective, single-center data from all patients who underwent interposition arthroplasty with an Amandys implant between November 2008 and May 2013, a procedure conducted by four senior hand surgeons. ¹⁷

The inclusion criterion was all patients who were operated on with an Amandys implant with a minimum follow-up of 10 years. There was no exclusion criterion.

Implant Description

Amandys (Stryker, Montbonnot, France) is a pyrocarbon implant for radiocarpal interposition arthroplasty. It replaces two-thirds of the proximal scaphoid, the lunatum, and a portion of the capitate head. It preserves the triquetrum and the distal third of the scaphoid, and thus the main extrinsic wrist ligaments. The implant is quadrielliptic in shape. The two convex joint surfaces are asymmetrical, with the radial surface being more pronounced than those of the midcarpal surface. It is a mobile spacer with no bone or ligament fixation. Its stability is ensured by its congruence with bone surfaces of the radius and the midcarpus, laterally by bone contours of triquetrum and radial styloid, and by the anterior and posterior capsuloligamentous structures of the wrist.

Amandys is available in eight sizes depending on length (24 or 26 mm) and thickness (S, M, L, and XL).

Operation Technique

The arthroplasty was performed under locoregional anesthesia using an arm tourniquet. Two approaches were used: dorsal or radial.

The dorsal approach was mainly used especially if removal of material, capsular reinforcement, or bone reconstruction was planned. The incision was italic “S-shaped.” The annular ligament was incised between the third and the fourth extensor compartment. Posterior interosseous nerve was resected on 2 cm. The capsulotomy was longitudinal if the capsule was scared or with a capsular L-shaped flap in virgin wrist.

The radial approach was used by one surgeon only in virgin wrists without major osteoarticular destruction and if no capsular reinforcement was envisaged. The wrist was placed in a lateral position with a support under the ulnar side to allow ulnar inclination. An “S-shaped” incision of approximately 5 cm was made on the radial wrist opposite the radial styloid and the anatomical snuffbox. The sensitive branch of the radial nerve was protected. The capsulotomy was longitudinal between the first and second compartments.

A 5-mm distal styloidectomy was achieved with an oscillating saw, parallel to the frontal slope and parallel to the glenoid facet of the radius. A partial resection of the scaphoid was performed with the oscillating saw at the junction of the proximal two-thirds and the distal third. The slice's orientation was parallel to sagittal and coronal slopes of the radius. Resection of the lunatum was performed after releasing it from all the capsuloligamentous attachments taking care not to damage the anterior and posterior capsules. The final step, a partial resection of the capitate, was performed, continuous with the scaphoidectomy.

The glenoid surface of the radius was prepared using the ovoid burr until a homogeneous ovoid surface was obtained by erasing the ridge separating the lunate and scaphoid fossae. On the carpal slope, the same burr was used to obtain a slightly concave homogeneous surface. Voluminous osteophytes could be resected at the same time, taking care to preserve capsuloligamentous insertions. A synovectomy could be performed, if necessary, at the end, and the capsular surface was inspected to repair any breach or plicated, in the case of distension, by suture reinforcement points.

Implants

The size of the implant required was determined using preoperative templates. The implant comes in eight sizes, varying in length (24 or 26 mm) and thickness (S, M, L, and XL). The final size was chosen preoperative under fluoroscopic examination. Different criteria had to be defined to determine the correct size.

• Passive flexion–extension is obtained using gravity alone, and ulnar and radial inclinations are complete.

• The implant does not rotate on its vertical axis during passive mobilization of the wrist (a sign that the implant is too long or impinges on bony spur) and there is no cam effect with the lateral structures.

• The implant does not tend to dislocate anteriorly or posteriorly (a sign of incongruence with the radial surface, either insufficient concavity of the radial glenoid or by lack of implant thickness).

• There is no dorsal subluxation of the second carpal row (a sign of capsular failure by rupture or distension).

• The interval between the ulna and the triquetrum on frontal view is not altered compared with preoperative views (a sign of the implant that is too thick or too thin).

• There is no styloscaphoid conflict.

Closure and Postoperative Care

Capsular closure was achieved using resorbable 3–0 or 4–0 suture. In the case of fragility of the capsule, it was reinforced dorsally by a flap from the dorsal retinaculum passed deeply to the extensor or a flap from the first extensor compartment if the approach was radial.

The wrist was immobilized in neutral position by a splint during the first 15 days, then gentle active mobilization of the wrist is started and a protective splint is worn at night and during the day at intervals until the fourth week. Physical therapy could then be started, adapted to each case. No particular restrictions were imposed beyond the sixth week.

Population

Population characteristics were collected, including etiology, age, sex, and surgical history involving the wrist.

Evaluation

All patients were evaluated at inclusion by the operator, then reevaluated at months 1, 3, 6, 12, and 18 and at years 2, 3, 5, and 10. The intermediate and last clinical assessments were conducted by an independent reviewer with a minimum follow-up of 10 years.

Survival Rate of the Implant

During each evaluation, we looked for any surgical revision by reviewing the medical records and asking the patient.

Satisfaction

Satisfaction was noted by the patient as: very satisfied, satisfied, fairly satisfied, or unsatisfied.

Clinical Evaluation

Measurements for both wrists were recorded. Active mobility in flexion, extension, radial, and ulnar inclination was measured using a goniometer.

Grip strength was measured using a Jamar dynamometer (Kit Baseline, Arex, France), and three successive measurements were recorded for each hand. The maximum value was retained.

Functional Evaluation

Two self-administered questionnaires were used for noting functional evaluation scores from 0 to 100: the short version of the Disability of Arm Shoulder and Hand (QuickDASH) and the Patient-Rated Wrist Evaluation (PRWE). Pain was assessed by taking the PRWE pain score divided by 5 for getting a score ranging from 0 (no pain) to 10 (maximum pain).

Radiological Measurements

Anteroposterior and lateral views of the wrist showing the metacarpal head were taken before surgery and during the follow-up period and were inspected for:

• Carpal sagittal subluxation measured on a lateral radiograph by the relation D/L1, where D is distance between the axis of the radius and the axis of the third metacarpal and L1 is the length of the third metacarpal. ¹⁸

• Radial deviation of the carpus measured by Shapiro's angle. ¹⁹

• Height of the carpus measured by the index of McMurtry et al. ²⁰ This index also allowed us to assess subsidence of the implant in the midcarpal. We hypothesized that a decrease during postoperative follow-up would be linked to the subsidence in the midcarpal.

• Ulnar carpal translation, measured by the index of Youm et al. ²¹

• Height of the radial epiphysis, measured in millimeters by the distance of two parallel lines orthogonal to the axis of the radius, one at the proximal part of the ulnar notch of the radius and the other passing through the distal pole of the implant. Subsidence in the radius is defined as change in height over time.

Statistical Analysis

Quantitative data were expressed as mean standard deviation, with the minimum and maximum in brackets. Qualitative data were expressed as frequency. The paired Student's t - test, the Wilcoxon test, and Pearson's correlation coefficient were used for data analysis. A p -value < 0.05 was considered statistically significant.

The survival analysis was done according to the Kaplan–Meier method, with the main event being surgical revision. Patients lost to follow-up were censored to the right. The latest follow-up examination was used for them.

Results

Population

Our series included 84 implantations in 76 patients, 60.5% of them were males, with an average age of 59 years (25–84). Among the 76 patients, 39 (51.3%) were reviewed with a minimum of 10 years of follow-up, 37 (48.6%) were either dead or could not be contacted after 10 years and were considered lost to follow-up. The clinical analyses were performed on 39 patients (41 procedures) with a mean follow-up of 12.64 ± 1.47 years (10.18–14.45).

Comparison of data before surgery, at 2 years, and at 5 years were also performed on these same patients when the data were available.

All patients lost to follow-up were unreachable, because they did not respond to messages by phone or email despite several unsuccessful follow-up visit reminders.

Table 1 shows the characteristics of the population and Table 2 summarizes the etiologies of the different cases dominated by scapholunate advanced collapse lesions.

Table 1. Population.

Sex	Male	46	60.5%
	Female	30	39.5%
Mean age (y)		59 (25–84)
Side	Right	49	58.3%
	Left	35	41.7%
Dominant side		64	84.2%

Table 2. Etiologies.

Etiologies
SLAC III	SLAC III	2	23.8%
SLAC IV	SLAC IV	18
SNAC IV		5	6.0%
SCAC III		1	1.2%
Rheumatoid arthritis		9	10.7%
Kienböck disease, Lichtman stage III–IV		10	11.9%
Articular malunion of the distal radius		12	14.3%
Failure of silicone implant with synovitis		6	7.1%
Septic arthritis		3	3.9%
Sequelae of perilunate injury		3	3.6%
Failure of partial wrist fusion	Four bones	7	13.1%
	Lunocapitate	1
	Radiolunate	1
	Scaphocapitate	1
	STT	1
Failure of proximal row carpectomy		3	3.6%
Failure of scaphoidectomy		1	1.2%
Total		84	100%

Abbreviations: SCAC, scaphoid chondrocalcinosis advanced collapse; SLAC, scapholunate advanced collapse; SNAC, scaphoid nonunion advanced collapse; STT, scaphotrapeziotrapezoid.

Survival Rate

The cumulated 10-year survival rate, using any revision as the endpoint, was estimated at 86.4%. None of the cases reviewed required removal of the implant or conversion to TWA or TWF.

Fig. 1 shows the Kaplan–Meier survival plot.

Fig. 1.

Kaplan–Meier survival plot.

Satisfaction

According to the survey, 94.9% were very satisfied or satisfied with the procedure.

Clinical Results

Active mobility was preserved between the preoperative measurements and those taken at 5-year follow-up and at the last follow-up visit. The difference was −4 degrees for flexion, 7 degrees for extension, −2 degrees for inclination, with a final range of motion (ROM) of 69 degrees for flexion–extension and 30 degrees for inclination. The outcomes are shown in detail in Tables 3 and 4 .

Table 3. Clinical and functional outcomes (preoperative vs. last follow-up).

	Preoperative, mean (minimum to maximum)	LFU (mean = 12.64 y)	Difference	p
Mobility
Flexion (degrees)	36.2 (10–85)	31.9 (0–80)	−4	NS
Extension (degrees)	29.6 (−10 to 80)	37.1 (0–90)	7	NS
Ulnar inclination	20.0 (5–80)	15.0 (0–35)	−5	NS
Radial inclination	9.28 (−10 to 20)	12.4 (0–30)	3	<0.05
Grip strength (degrees)	14.2 (4–32)	20.8 (1–50)	7	<0.05
% Strength	55.2 (18–107)	71.7 (5–108)	17	<0.05
Function and pain
QuickDASH/100	64.8 (40–92)	19.0 (2–88)	−38	<0.05
PRWE/100	65.6 (28–94)	29.5 (0–78)	−36	<0.05
Pain/10	6.40 (2.6–9)	2.00 (0–9)	−4	<0.05

Abbreviations: LFU, loss to follow-up; PRWE, Patient-Rated Wrist Evaluation; NS, nonsignificant.

Table 4. Clinical and functional outcomes at 5-year follow-up and at the last follow-up.

	5-year follow-up, mean (minimum to maximum)	LFU (mean = 12.64 y)	Difference	p
Mobility
Flexion (degrees)	35.0 (0–80)	31.9 (0–80)	−4	NS
Extension (degrees)	40.0 (20–75)	37.1 (0–90)	−1	NS
Ulnar inclination (degrees)	20.0 (5–32)	15.0 (0–35)	−2	NS
Radial inclination (degrees)	10.0 (0–25)	12.4 (0–30)	2	NS
Grip strength (kg)	18.0 (8–49)	20.8 (1–50)	1	NS
% Strength	81.8 (31–150)	71.7 (5–108)	−4	NS
Function and pain
QuickDASH/100	33.5 (0–70)	19.0 (2–88)	−4	NS
PRWE/100	27.6 (0–80	29.5 (0–78)	−1	NS
Pain/10	2.00 (0–7)	2.00 (0–9)	0	NS

Abbreviations: LFU, loss to follow-up; PRWE, Patient-Rated Wrist Evaluation; NS, nonsignificant.

Pain was significantly improved, with a gain of 4 points between preoperative and the last follow-up, and remained stable between 5-year follow-up and the last follow-up. Strength was also significantly improved, with a gain of 7 kg and 17% compared with the contralateral side between preoperative and the last follow- up, and a loss of 1 kg ( p  < 0.05) between 5-year follow-up and the last follow-up.

Functional Results

Functional scores were significantly improved between preoperative and the last follow-up. The QuickDASH improved by 38 points and PRWE by 36 points. Scores remained stable between the 5-year follow-up and the last follow-up, with a loss of 1 point for QuickDASH ( p  > 0.05) and 4 points for PRWE ( p  > 0.05).

Radiological Assessment

The radiological assessment did not show significant instability of the carpus over time. There was no difference for carpal sagittal subluxation, carpus height, ulnar translation, or radial deviation ( Table 5 ).

Table 5. Radiological outcomes (preoperative vs. last follow-up).

	Preoperative	LFU (mean = 12.64 y)	Difference	p
Palmar subluxation	0.1 (0–10)	0.1 (−1 to 0.4)	0	NS
Radial deviation	114 (8–125)	111 (90–130)	−4	NS
Carpal height	0.4 (0.3–0.5)	0.4 (0.2–0.4)	0	NS
Ulnar translation	0.3 (0.2–22)	0.3 (0.1–0.5)	0	NS
Height of radial epiphysis	21.1 (16–26)	20.8 (16–26)	0.3	NS

Abbreviations: LFU, loss to follow-up; NS, nonsignificant.

The radiological results remained stable between the 5-year follow-up and the last follow-up with nonsignificant differences ( Table 6 ; Fig. 2 ).

Table 6. Radiological outcomes (5-year follow-up vs. last follow-up).

	5-year follow-up	LFU (mean = 12.64 y)	Difference	p
Palmar subluxation	0.1 (−1 to 0.2)	0.1 (−1 to 0.4)	0	NS
Radial deviation	109 (80–130)	111 (90–130)	2	NS
Carpal height	0.4 (0–0.5)	0.4 (0.2–0.4)	0	NS
Ulnar translation	0.3 (0.2–0.4)	0.3 (0.1–0.5)	0	NS
Height of radial epiphysis	20.9 (16–26)	20.8 (16–26)	−0.1	NS

Abbreviation: NS, nonsignificant.

Fig. 2.

A 73-year-old woman at 13.5 years of follow-up for her left wrist intervention for rheumatoid arthritis. ( A ) and ( B ) One-month postoperative X-rays; ( C ) and ( D ) X-rays at 13.5 years of follow-up.

Complications

There were 10 revision surgeries (11.9%), with only one occurring after 15 months: 9 for implant instability (5 rotation and 4 dislocation) treated by repositioning the implant with or without changing its size, and one for ulnocarpal conflict ( Fig. 3 ).

Fig. 3.

A 38-year-old man who underwent a procedure for stage IV Kienböck disease. ( A ) and ( B ) One-month post-operative X-rays; ( C ) and ( D ) X-rays showing an early rotation of the implant. ( E ) and ( F ) X-rays after revision by palmar capsule reinforcement and changing the size of the implant.

The outcomes of patients who underwent revision surgery did not differ from those of the rest of the study population in terms of ROM improvement, functional scores, pain relief, or strength recovery ( Table 7 ).

Table 7. Comparison of improvement in outcomes between patients with revision surgery versus patients without revision surgery.

	No revision ( n  = 31)	Revision ( n  = 10)	Average difference	p
Mobility
Flexion (degrees)	30 (0–80)	30 (20–70)	5	NS
Extension (degrees)	40 (10–90)	35 (0–50)	−5	NS
Ulnar inclination (degrees)	19 (0–35)	12 (0–20)		NS
Radial inclination (degrees)	12 (0–30)	12 (0–30)		NS
Grip strength (kg)	20 (1–50)	20 (1–30)	1	NS
% Strength	79 (5–108)	67 (57–83)	4	NS
Function and pain
QuickDASH/100	31 (2–88)	25 (0–57)	0	NS
PRWE/100	18 (0–78)	36 (0–51)	2	NS
Pain/10	2 (0–9)	1 (0–6)	0	NS

Abbreviations: PRWE, Patient-Rated Wrist Evaluation; NS, nonsignificant.

Discussion

Amandys implant arthroplasty has shown, in preliminary, short-term and mid-term studies, that it was a valid alternative to TWA and TWF for advanced wrist destruction. ^{14 15 16 22} The present study confirmed this finding showing wrist motion preservation and significant pain relief, function improvement, and grip strength recovery over time. This confirmed the main hypothesis of our study.

We also observed that between 5 years and 12 years of follow-up, pain relief, mobility, and functional scores stay stable. Implant was well tolerated and radiological biometric measurements of the wrist remained stable. Furthermore, the survival rate did not deteriorate at long term.

The previous generations of total wrist prostheses like the Universal 1, Biaxial, and Meuli have demonstrated low long-term survival rates, ranging from 40 to 83%. ²³

A systematic review of clinical outcomes for fourth-generation TWA was conducted by Zijlker et al. ⁸ The study analyzed implant survival rates for various prostheses, defining these rates based on implant removal incidents. For the Universal2 implant, the mean 5-year implant survival rate was higher than 92% and declined to a mean 78% at 15-year follow-up. For the Motec, the mean 10-year implant survival rate is 86%.

Using those criteria, the survival rate of Amandys would be virtually 100%, as no Amandys implantation has been converted into another prosthesis or into a TWF.

Functional range of wrist motion was estimated to 5 degrees of flexion, 30 degrees of extension, 10 degrees of radial deviation, and 15 degrees of ulnar deviation. ²⁴ In our study the mean flexion is 31.9 degrees, 37.1 degrees of extension, 15.0 degrees of ulnar inclination, 12.4 degrees of radial inclination, enough for the everyday life use of a wrist.

Pain and functional scores of our series were comparable to those reported for TWA with the same minimal follow-up with a gain of 38 points on the QuickDASH score, 36 points on the PRWE score, and 4 points for pain. The grip strength gets better, approximately 7 kg in our series compared with the preoperative findings, versus 4 kg in some TWA. ¹³

About the radiological findings, there were no bone deterioration at the contact of the implant, and we did not report any significant implant subsidence or carpal migration. The subsidence of the implant into the radius occurred slightly during the first 2 years, then stabilized.

The primary conceptual distinction between the Amandys implant and the fourth-generation implant lies in the absence of osteointegration of the pyrocarbon wrist implant. With an elasticity modulus closely resembling that of cortical bone and a lack of osteointegration, it enables unrestricted movement without imposing pick of constraints on the bone surface, thereby eliminating the risk of loosening. On the contrary, the risk of loosening exists in the long run in fourth-generation implants in the long run. ²⁵

The survival rate for Amandys was stable over time, with a Kaplan–Meier curve that stabilized at around 15 months. On the contrary, TWA and TWF complications worsened over time. Wagner et al ²⁶ in a retrospective review of 215 patients who underwent TWF with a plate and had a 6-year mean follow-up, found 19% of revisions and a survival rate of 76% at 5.8 years. Reigstad et al ²⁷ also reported significant rates of TWF revisions after 11 years of follow-up (63%).

Our study found no carpal misalignment because the operative indications were on the aligned wrist, especially in the anteroposterior axis. Special attention was given in case of significant preoperative dorsal intercalated segment instability, where anterior capsular dehiscence may be present. The technique preserves the triquetrum as well as the dorsal radiocarpal and dorsal intercarpal ligaments, which are important for the wrist stability. This may explain that we did not find any occurrence of radiographic wrist instability postoperatively and with the follow-up.

Most of the revisions of the Amandys implant occurred within the first 2 years of use, which is consistent with other findings. Revision of Amandys implant did not affect in our series the long-term outcomes. This revision consisted almost exclusively of removing bone spurs for correction of a cam effect, reinforcing palmar or dorsal capsule, or changing the size of the implant. These secondary procedures did not affect the final results and no significant difference in final outcomes was found between the patients requiring surgical revision and the other patients.

The present study does suffer from a fairly high rate of loss to follow-up, which can induce nonresponse bias. We did not find any significant difference between the characteristics of the patients lost to follow-up and those of the overall population in our study. Nevertheless, lost to follow-up in our review looks acceptable according to Tang et al. ²⁸

Conclusion

Interposition arthroplasty using an Amandys implant offers a valid alternative to TWA or TWF for the treatment of advanced wrist destruction when the wrist is centered. Wrist motion is preserved, and there is a significant improvement in pain, functional scores, and strength, which were confirmed over time. Patients were highly satisfied with the intervention.

Complications occurred early but do not worsen over time and are likely related to the learning curve. These results validate the concept of pyrocarbon wrist interposition.

This study suggests that unlike TWA, pyrocarbon interposition is not subject to the long-term risks of loosening and osteolysis.