Pyrocardan Trapeziometacarpal Joint Arthroplasty—Medium-Term Outcomes

Abstract

Objective  Pyrocardan trapeziometacarpal interposition implant is a free intra-articular spacer composed of pyrocarbon. This biconcave resurfacing implant, both ligament and bone-stock sparing, is indicated for use in early-to-moderate stage trapeziometacarpal osteoarthritis. It was hypothesized that the postoperative outcome measures of the Pyrocardan implant would be comparable to those seen with ligament reconstruction and tendon interposition (LRTI) surgeries and those reported by the designer of the implant, Phillipe Bellemère, but that strength would be greater than for LRTI.

Methods  In this prospective case series, 40 Pyrocardan implants were performed in 37 patients. Average age was 58 years (range: 46–71). Patients were assessed preoperatively, 3 months, 6 months, 1 year, 2 years, and beyond (long term) wherever possible.

Results  There have been no major complications or revision surgeries for the series. Average follow-up was 29 months (range: 12 months–7 years). Twenty-eight joints were assessed at over 2 years post index surgery. Outcome measure scores improved from preoperative assessment to the most recent follow-up equal or greater than 2 years. Average grip strength at 2 years was 30 kg, as compared with 19.6 kg in an age-matched cohort who underwent trapeziectomy and 25 kg in Bellemère's original series of Pyrocardan implants.

Conclusions  Pyrocardan interposition arthroplasty appears to be a safe, effective treatment for trapeziometacarpal arthritis. Patient-reported clinical outcomes were at least equivalent to LRTI and are comparable to Bellemère's original series. Grip and pinch strength appear to be better than LRTI.

Level of Evidence  This is a Level III, prospective observational cohort study.

Trapeziometacarpal joint (TMJ) osteoarthritis (OA) is an exceedingly common condition leading to pain, weakness, and loss of function. The TMJ is second only to the facet joints of the spine in frequency of degenerative change. ¹ Surgical treatments for TMJ OA include arthroscopic debridement with or without capsular shrinkage, excision of the trapezium, ^{2 3 4} excision of the trapezium, and ligament reconstruction with or without tendon interposition, ^{2 5 6 7 8} fusion of the carpometacarpal joint, ^{4 9} osteotomy of the metacarpal, ^{10 11 12 13} and prosthetic arthroplasty. Joint replacements have been performed utilizing several materials and designs. Implants can be broadly grouped under the following headings ¹⁴ :

Total Joint Replacement

These prostheses consist of a separate trapezial and metacarpal component and utilize a metal on polyethylene/metal bearing as the articulating surface. Both cemented and uncemented fixation has been employed. ^{15 16}

Hemiarthroplasty

These prostheses replace the proximal segment of the metacarpal. They can be further divided in to anatomic or nonanatomic. In an anatomic implant, the concave saddle shaped bearing surface aims to replicate the contours of the native metacarpal. Both anatomic and nonanatomic hemiarthroplasties utilize a stem into the metacarpal to secure the implant.

Interposition Arthroplasty

These stem-less components involve the insertion of a low friction material between the arthritic joint surfaces to prevent pain and preserve motion. To make room for the material to be interposed, a degree of bone resection is required. This can be the entire trapezium or just a part there of. Partial resection implants can be further divided into unconstrained implants that rely on the surrounding soft tissues to maintain their position or constrained implants where the geometry of the prosthesis confers a degree of stability. Total trapezium replacement is a further class of interposition arthroplasty requiring complete excision of the trapezium and replacement with an artificial component. ^{14 17 18 19}

Excision arthroplasty remains the mainstay for the surgical treatment of TMJ OA. While trapeziectomy provides effective and prolonged relief from pain, this comes at the cost of loss of thumb length and strength. ^{9 20 21} Osteotomy is only indicated in very early stage disease and may be complicated by nonunion, hardware problems and inaccurate cuts including intra-articular breach of the osteotomy. ^{22 23}

The Pyrocardan (Bioprofile/Tornier), ²⁴ designed by P. Bellemère, is a pyrolytic carbon unconstrained interposition arthroplasty. ^{14 19} The perpendicular biconcave surfaces convert the saddle joint of the base of the thumb into a universal (or cardan) joint. It is available in seven sizes from 12 to 18 mm but in all sizes the central thickness is just 1 mm to minimize bone resection and stress shielding. Pyrolytic carbon has excellent biocompatibility and a modulus of elasticity that closely resembles that of cortical bone. ¹⁴ Animal models have demonstrated excellent wear properties with pyrocarbon on articular cartilage articulation and in vitro studies have demonstrated superiority of pyrocarbon over metal when articulating against bone. ^{25 26 27} Bellemère et al reported on their experiences with the Pyrocardan in Eaton and Littler grade 1 and 2 trapeziometacarpal OA. ²⁴ In a consecutive series of 27 patients with a mean follow-up of 16.6 months (range: 12–27), they demonstrated excellent improvements in functional outcomes and pain. No implants were revised nor were any significant complications reported. ²⁴

We hypothesized that the Pyrocardan interposition arthroplasty would provide excellent relief of pain from trapeziometacarpal OA while preserving both motion and grip strength. In fact, we felt that justification of arthroplasty in the treatment of trapeziometacarpal OA would require demonstration of superior function such as strength when compared with LRTI.

Patients and Methods

Surgical Technique

We present a single-institution two surgeon (xx, xx) case series with prospectively collected data of Pyrocardan trapeziometacarpal implantation for Eaton and Littler grade 1, 2, and 3 arthritis with mild-to-moderate joint subluxation. All cases were performed using the same surgical technique. A curved dorsoradial incision is made at the junction of dorsal and volar skin in the region of the abductor pollicis longus (APL) tendon crossing the first carpometacarpal joint ( Fig. 1A, B ). APL is split in longitudinally and carefully dissected off the joint capsule. A Z-shaped capsulotomy is performed to expose the TMJ ( Fig. 1C ). The capsulotomy creates two capsular flaps, a proximally based dorsal ulnar flap and a distally based volar radial flap ( Fig. 1D ). The metacarpal base and trapezium are shaped with a sagittal saw and/or a high-speed burr to accept the implant with care being taken to excise the medial horn of the trapezium and the associated medial osteophyte that emerges from it ( Fig. 2A ). Bony contours are shaped with a microsagittal saw or burr. The goal is to remove the volar and dorsal horns from the metacarpal base creating a convexity from radial to ulnar and to remove the medial horn and lateral prominence from the trapezium, thus creating a convexity running from dorsal to volar that is perpendicular to the convexity of the metacarpal base ( Fig. 2B ).

Fig. 1.

( A, B ) Skin incision. ( C ) Z-shaped capsulotomy to reveal first carpometacarpal joint. ( D ) Diagram showing the two capsular flaps created by the capsulotomy, the proximally based dorsoulnar capsular flap (DU), and the distally based volar–radial capsular flap (VR). MC, metacarpal; TM, trapezium.

Fig. 2.

( A ) A microsagittal saw or high-speed burr is used to shape the trapezium and first metacarpal to accept the prosthesis. (Reprinted with permission from Bellemère P). ( B ) Diagram demonstrating the perpendicular convex surfaces created by the bone cuts. Lat, lateral; MC, metacarpal; PA, posteroanterior; TM, trapezium.

The trial implant is inserted. The trials are labeled with the trapezial and metacarpal sides. The correct size implant should fully cover the trapezial surface ( Fig. 3A ). The trial is replaced with the same size definitive implant. The definitive pyrocarbon implant is unmarked; therefore, a useful tip is to place the definitive implant in the empty slot from which the trial came to ensure correct orientation ( Fig. 3B ). The capsular flaps are repaired to the metacarpal and trapezium using 3, 1 mm, all suture anchors (Juggerknot Zimmer Biomet, Warsaw, IN) ( Fig. 4 ). Metal anchors should be avoided due to potential damage to the pyrocarbon surface from contact with the metal. It is important to ensure that the alignment is corrected by the bone resection to ensure that the trapezium and metacarpal are co-linear ( Fig. 5 ). Two capsular flaps allow adjustment of soft tissue tension where there has been a degree of preoperative subluxation. This technique also permits robust side-to-side closure of the longitudinal limbs of the capsulotomy. The APL tendon is closed side to side.

Fig. 3.

( A ) Trial implant. The correct size should cover the entire trapezium and be stable when performing a full passive range of movement under image intensifier screening. ( B ) The trial is replaced with the corresponding definitive implant. Note the black unmarked definitive implant sitting in the 15 mm slot with the purple15 mm trial sitting on the tray. This makes aligning the unmarked prosthesis far simpler.

Fig. 4.

( A–C ) The capsule is repaired with 1 mm all suture Juggerknot anchors. ( D ) Diagram showing anchor placement. DU, dorsoulnar capsular flap; MC, metacarpal; TM, trapezium; VR, volar–radial capsular flap.

Fig. 5.

Collinear alignment of the carpometacarpal joint is restored.

The skin is closed with interrupted nylon. A plaster slab is applied in theater then converted to a forearm-based thermoplastic splint within 1 to 2 days by hand therapy. Active range of movement is commenced immediately but there is some restriction applied in terms of thumb opposition and adduction/flexion depending on the preoperative degree of subluxation. In cases with 20% to one-third subluxation, tip-to-tip motion is only permitted to the middle finger for the first 4 weeks, otherwise full opposition to the little finger tip is encouraged. At 4 weeks, most patients, with the exception of those with marked preoperative subluxation, are transitioned to a Push Brace (Nea International Push Braces, Maastricht, the Netherlands). At 6 weeks, full-time splinting is ceased and activities allowed, although the Push Brace is still recommended for heavier loaded activities until 12 weeks.

Participant Sample

This is a consecutive series of patients undergoing Pyrocardan interposition arthroplasty for trapeziometacarpal OA at a tertiary hand clinic. Patients were offered the procedure as an alternative to trapeziectomy with ligament reconstruction and tendon interposition when they expressed a desire for greater recovery of strength and activity. Inclusion criteria for the study were symptomatic trapeziometacarpal arthritis that had failed nonoperative measures including splinting, hand therapy, and ultrasound-guided joint injection. Patients with Eaton stages 1, 2, and 3 radiographic disease were included and with subluxation up to one-third of the width of the metacarpal base was included.

We implanted 40 prostheses in 37 patients ( Table 1 ). The mean age was 58 years. There was a preponderance of females and an even split of dominant and nondominant hands. Mean follow-up was 29 months with 28 patients having greater than 24 months follow-up. Employment was recorded. The cohort included four farmers, two builders, a miner, two cleaners, and five healthcare professionals. The remaining patients were retired or worked in sedentary professions.

Table 1. Patient demographics.

Number of patients	37
Number of implants	40
Age at surgery, years—mean (range)	58 years (range: 46–71)
Female— n (%)	27 (67.5%)
Dominant side treated— n (%)	21 (52.5%)
Follow-up, months—mean (range)	29 months (range: 12–84)

Outcome Measurements

Patients were assessed preoperatively, at 3, 6, 12 months postoperatively, and followed up annually thereafter. Subjective outcome measures were pain with normal activities and satisfaction using a visual analogue scale from 0 to 100. Function was assessed using validated patient reported outcome questionnaires, the Patient-Related Wrist and Hand Evaluation (PRWHE best 0, worst 100), ²⁸ and the short version of the Disabilities of the Arm, Shoulder, and Hand Questionnaire (QuickDASH best 0, worst 100). ²⁹ The self-reported Global Rating of Change ³⁰ score measures the patient's improvement or deterioration in a Likert scale from +7 to -7 following the intervention in terms of symptoms and function. The Global Rating of Change was used to help demonstrate whether an improvement in symptoms and function was achieved. Objective outcome measures included active range of motion and hand strength. Active TMJ range of motion was assessed with a standard goniometer. Grip strength was measured using a hand hydraulic dynamometer (Jamar, J. A. Preston Corporation, Clifton, NJ) and recorded as the average of three trials. Pinch strength was measured (one trial) with a hydraulic pinch gauge (Saehan Corp, Masan, Korea). Adverse events, recorded at every interval, included infection, chronic pain, excessive scarring or hypersensitivity, implant failure, or other. X-rays were assessed by the consulting surgeon at each assessment interval. Results were compared with an age-matched cohort that underwent trapeziectomy/LRTI, as well as to Bellemère’s original series. ²⁴

Statistical Analysis

Continuous outcome variables are shown as mean (standard deviation), and Student's t -test was used to test for changes in continuous outcome variables between preoperative examination (40 joints) and last follow-up at greater or equal than 24 months (28 joints). ³¹ Global Rating of Change scores are shown as median (range) and no statistical analysis was conducted. For bilateral patients, the two joints were treated as independent observations. A p -value of < 0.05 was considered statistically significant.

Results

A summary of patient-reported outcome measures along with the follow-up intervals is shown in Table 2 . Mean pain with normal activities as assessed using a visual analogue scale from 0 to 100 (where 0 is no pain and 100 is worst pain ever) declined from 58 (21) prior to surgery to 27 (23) at 3 months. Pain continued to improve reaching a mean of 17 (18) at final follow-up greater or equal to 24 months ( p  < 0.001). Mean patient satisfaction with their joint, measured on a visual analogue scale from 0 to 100, was 17 (19) prior to surgery, increasing to 70 (30) at 3 months and 83 (23) at final follow-up ( p  < 0.001). PRWHE and QuickDASH showed a significant and sustained improvement in patient reported function, as they respectively improved from 64 (18) and 48 (16) at preoperative assessment to 20 (17) and 23 (17) at final follow-up (both p  < 0.001). The Global Rating of Change score showed a median improvement in symptoms of +4 and function of +3 at the 3-month postoperative review, followed by ongoing improvements reaching a mean increase of +6 points at final follow-up greater than or equal to 2 years for both symptoms and function.

Table 2. Subjective outcome measures.

	Pre-op ( n  = 40)	3 months ( n  = 38)	6 months ( n  = 36)	12 months ( n  = 35)	≥ 24 months ( n  = 28)	p -Value
Pain (VAS 0–100)	58 (21)	27 (23)	23 (16)	21 (26)	17 (18)	<0.001
Satisfaction (VAS 0–100)	17 (19)	70 (30)	78 (26)	78 (26)	83 (23)	<0.001
PRWHE (0–100)	64 (18)	36 (20)	29 (21)	24 (23)	20 (17)	<0.001
QuickDASH (0–100)	48 (16)	34 (19)	25 (16)	23 (21)	23 (17)	<0.001
Global Rating of Change (−7–+7)
Symptoms	n/a	4 (−4–7)	5 (−4–7)	6 (−5–7)	6 (2–7)	n/a
Function	n/a	3 (−4–7)	5 (−4–7)	5 (−5–7)	6 (2–7)	n/a

Abbreviations: PRWHE, Patient-Related Wrist and Hand Evaluation; QuickDASH, Disabilities of the Arm, Shoulder, and Hand Questionnaire; VAS, Visual Analogue Scale.

Note: Values indicate mean (standard deviation), except for Global Rating of Change, where they indicate median (range).

Assessment of grip strength, lateral pinch, and tip-to-tip pinch demonstrated an initial decrease at the 3-month follow-up as patients recovered from the surgery and brief period of immobilization and functional restriction ( Table 3 ). Tip-to-tip pinch strength remained stable at 3 months at a mean of 3 kg. By 12 months, however, the mean result for all three metrics shows a substantial increase. At 2 years postoperatively, mean grip strength was 30 kg, mean lateral pinch was 7 kg, and mean tip-to-tip pinch was 5 kg. Range of motion was preserved both in terms of opposition measured by Kapandji score and TMJ flexion and extension. Mean Kapandji score at all time points from 3 months postoperatively was 9 indicating almost full opposition. TMJ active extension remained unchanged with a mean range of 31 degrees at 3 months postoperatively. It slightly increased at all follow-up intervals with a mean of 35 degrees at last follow-up. TMJ flexion improved very slightly from a preoperative mean of –3 degrees to 1 degree at 3 months. This improvement was maintained throughout the follow-up period.

Table 3. Objective outcome measures.

	Pre-op	3 months	6 months	12 months	≥ 24 months	p -Value
Grip strength operative side (kg)	19 (11)	16 (10)	22 (11)	26 (10)	30 (9)	<0.001
Lateral pinch strength (kg)	5 (2)	4 (2)	4 (2)	6 (2)	7 (2)	<0.001
Tip-to-tip pinch strength (kg)	3 (2)	3 (1)	3 (2)	4 (2)	5 (1)	0.008
Kapandji score (0–10)	8 (2)	9 (2)	9 (2)	9 (2)	9 (1)	0.07
TMJ extension (deg)	36 (12)	31 (12)	31 (13)	34 (11)	35 (16)	0.14
TMJ flexion (deg)	−3 (14)	1 (10)	2 (10)	0 (12)	4 (13)	0.20

Abbreviation: TMJ, trapeziometacarpal joint.

Note: The values indicate mean (standard deviation).

Regarding complications, there were no reported deep or superficial infections. There were no reported episodes of either subluxation or dislocation of the implant. Nine patients reported some ongoing pain with activity for up to 12 months following the procedure; however, none of these have required further intervention or revision surgery. One patient subsequently underwent ipsilateral metacarpophalangeal joint fusion for progressive hyperextension.

Discussion

Trapeziectomy with or without ligament reconstruction and tendon interposition for TMJ OA is a safe and reliable procedure that provides effective and long-lasting relief from pain; however, this comes at the cost of shortening of the thumb and loss of strength. ^{9 20 21} Some authors have recommended TMJ arthrodesis to preserve thumb length and grip strength. ³² Other studies have failed to demonstrate the benefits in grip strength from arthrodesis. ³³ A high rate of complications has been recorded, in particular nonunion. ^{32 33 34}

Arthrodesis potentially offers preservation of grip strength but at the cost of movement and with a high rate of complication. ³⁴ Prosthetic arthroplasty has been proposed as an alternative with the aim of resolving the pain of TMJ OA, maintaining first ray length while preserving movement. By preserving the joint fulcrum, it is proposed that tip pinch and grip strength may be improved compared with resection arthroplasty. ¹⁴ However, to date total joint replacement or hemiarthroplasty, while producing promising clinical results, has had reasonably high rates of complications. Silicon interposition implants were complicated by silicon synovitis, joint instability, and long-term implant failure. ^{35 36 37 38} Metal on polyethylene/metal total joint arthroplasty, while often producing good functional outcomes, has demonstrated high levels of aseptic loosening or osteolysis in several series looking at a range of different prostheses. ^{15 16 20 39 40 41 14}

Pyrocarbon, due to its excellent wear properties and biocompatibility, is a very appealing material to use in small joint arthroplasty. It is not, however, without its problems. Implant instability, loosening, and persistent pain have been reported following stemmed pyrocarbon hemiarthroplasty. In a series from the Mayo Clinic, there was an 80% survival rate at 1.8 years with 15 cases out of the 49 performed requiring revision for pain or instability. ⁴²

Nonstemmed interposition implants such as the PI2 (Tornier) or the Pyrodisk (Integra) have demonstrated good relief of pain and preservation of joint strength ^{43 44} but have been troubled by instability with rates of dislocation reported between 12.5 and 33%. ^{45 46 47 48}

Our results demonstrate that the Pyrocardan interposition arthroplasty for trapeziometacarpal arthritis appears to provide excellent, rapid reduction in pain and a substantial improvement in patient satisfaction. A significant proportion of the gains was seen in the first 3 months reinforcing Phillipe Bellemère's assertion that the recovery is more rapid than that seen following trapeziectomy. ²⁴ However, in keeping with our experience using pyrocarbon in other joints, outcomes continued to improve beyond 2 years. ²⁴

We compared outcomes in our series with a cohort of patients from our carpometacarpal joint OA database matched for age, gender, and preoperative scores, undergoing trapeziectomy and Scheker flexor carpi radialis suspensionplasty. ⁴⁹ Satisfaction, pain at rest, patient-reported outcome measures, and range of motion were equivalent between the two groups. Strength, however, was greater in the Pyrocardan group. When assessed 2 years following surgery, the mean grip strength in the Pyrocardan group was 30 kg compared with 19.6 kg in the trapeziectomy/suspension group. Bellemère's group of Pyrocardan patients had a mean grip strength of 25.0 kg. ²⁴ We demonstrate preserved range of motion in both flexion–extension and opposition up to 2 years. This is in contrast to arthrodesis where a loss of motion is to be expected. ³⁴ Reduction in pain, preservation of motion, and improved grip strength combine to permit the functional improvements reported by patients in the PWRHE and QuickDASH outcome scores.

We had no cases of implant subluxation clinically or radiographically. This is in keeping with Bellemère's experience but in contrast to other published series. Lauwers et al reported on two patients with dislocation of Pyrocardan implants in a series of 27 cases. ⁵⁰ They performed the surgery through a volar approach using a tendon graft to reinforce the capsule repair. They hypothesized that this variation in technique might account for the disparity between their and Bellemère's results. ⁵⁰ Our technique is similar though not identical to Bellemère's. The dorsal capsular arthrotomy allows differential tensioning and robust capsular repair to account for preoperative subluxation. However, we would not recommend undertaking cases with significant subluxation early in the surgeon's experience. Attendance at a cadaver laboratory may be beneficial to shorten the learning curve.

Studies have reported patients with persistent or worsened pain following the insertion of a Pyrocardan implant. Odella et al had 12% of patients report worsening pain and 2 patients underwent revision surgery, ⁵¹ while Lauwers et al revised four patients to a trapeziectomy for persistent pain. ⁵⁰ In our series, while nine patients reported mild pain on activity as a persisting symptom up to 12 months postoperatively, none have required further intervention or revision surgery to date. In keeping with our experience of pyrocarbon as a bearing surface for hemiarthroplasty in other joints, pain relief continues to improve beyond 12 months. ⁵²

Conclusion

Pyrocardan interposition arthroplasty for trapeziometacarpal arthritis appears to be effective in reducing pain and improving patient function in the short term. Grip and pinch strength appear to be superior to trapeziectomy in a matched cohort, while range of motion is preserved 2 years after surgery. We have shown a low rate of complications. Our outcomes are comparable to the design surgeon's initial published results. Longer-term follow-up and true comparative studies are required to demonstrate sustained benefits of Pyrocardan interposition arthroplasty over other treatments for thumb base arthritis; however, this study does suggest that this particular arthroplasty achieves the stated goal of increased strength in comparison to LRTI, while maintaining comparable pain relief, patient satisfaction, and minimal complications.