Long-term Outcomes of Silastic Arthroplasty of the Thumb Metacarpophalangeal Joint

Abstract

Background: Silastic metacarpophalangeal arthroplasty (SMPA) has proven to be a durable option for end-stage arthritis in the non-thumb digits, while fusion has been the mainstay procedure for the thumb metacarpophalangeal joint (MP). Few studies exist to comment on the viability of thumb MP arthroplasty. This study reports both survival and objective outcomes following SMPA of the thumb. Methods: In an institutional review board-approved retrospective study, we identified 18 patients who underwent thumb SMPA at a tertiary academic center by 3 board-certified hand surgeons. Primary outcome measures were implant survival and post-operative complications. Secondary outcomes measures were quick Disabilities of the Arm, Shoulder, and Hand (quickDASH) scores, brief Michigan Hand Questionnaire (bMHQ), and postoperative pain as rated by the numerical rating scale. Results: Mean quickDASH and bMHQ scores at final follow-up were 35.6 and 70.6, respectively. The most common short-term complication was clinical deformity, followed by instability. The sole long-term complication was an implant dislocation in a previously asymptomatic patient. All patients reported reduction in pain. Three patients were indicated for revision surgery, 2 for persistent instability, and 1 for implant dislocation. Primary survivorship was 83% at mean follow-up of 5.8 years. Conclusions: Thumb SMPA is a viable option for end-stage arthritis. Pain relief in our series was unanimous. Among those that reported persistent symptoms or required revision, a majority had one or more key preoperative risk factors for failure as currently reported in literature. Larger, prospective series are needed to prove superior longevity and functional outcomes of thumb SMPA versus fusion.

Introduction

Silastic metacarpophalangeal joint arthroplasty (SMPA) has become one of the most common treatments for end-stage arthritis of the metacarpophalangeal joints (MPs).^1,2 The original silicone “dynamic spacer” as developed by Swanson³ in 1966 has undergone only subtle changes since its inception.¹ Early follow-up studies have shown high rates of implant fracture,⁴ although this subsequently decreased with introduction of the high-performance elastomer.¹ Many patients also showed evidence of bony resorption at long-term follow-up.^1,4 However, a subsequent radiographic study has shown that neither bony resorption nor implant fracture correlate with patient satisfaction—with over 80% saying they would have the procedure again.⁵ Lastly, due to inadequate soft tissue support in rheumatoid arthritis (RA) patients, recurrence of both ulnar drift and extension deficits is noted at long-term follow-up.^1,2,6 Nevertheless, SMPA provides nearly unanimous pain relief and improvement in outcome scores.^2,7,8

While this procedure is commonly studied in the non-thumb digits, few studies have reported on SMPA of the thumb.^4,9 Thumb MP fusion has been the more commonly performed procedure in the last half century.^1,10-12 However, fusion does come with considerable morbidity. The incidences of failed fusion and symptomatic hardware have been reported to be as high as 17%¹³ and 23%,¹⁴ respectively. Moreover, fusion is known to exacerbate adjacent joint degeneration and may be contraindicated in patients with compromised interphalangeal (IP) and carpometacarpal (CMC) joint mobility.^10,11,15

Since Figgie et al in 1990 reported favorable outcomes for Silastic arthroplasty of the thumb MP, there has been little to no new literature on this topic. The purpose of this study is to examine the long-term patient-reported and objective outcomes following silastic arthroplasty of the thumb MP.

Materials and Methods

After obtaining institutional review board approval, a billing database query was performed to identify all patients who underwent thumb metacarpophalangeal joint arthroplasty from 2000 to 2018 using Current Procedural Terminology (CPT) codes 26530 and 26531 with modifiers FA and F5.

Retrospective chart review was performed. Patient characteristics such as age, gender, and indication for surgery were recorded. Charts were reviewed for complications and any subsequent surgery relating to the operative digit. Complications were defined by infection, neurologic deficits, instability, deformity, osteolysis, and persistent pain. In order to differentiate immediate from delayed post-operative sequelae, these complications were characterized as either short-term (less than or equal to 1 year) or long-term (greater than 1 year). Patients were determined lost to follow-up if there was no contact outside the immediate post-operative period.

Patients were contacted via telephone for quick Disabilities of the Shoulder, Arm, and Hand (quickDASH) scores, brief Michigan Hand Questionnaire (bMHQ) scores, post-operative pain scores via the numerical rating scale (NRS), and inquiry into any further surgeries on the operative digit by an outside surgeon.

The authors adhered to the Strengthening the Reporting of Observation Studies in Epidemiology (STROBE) guidelines. Descriptive statistical analysis was performed with 95% confidence intervals (CIs) where appropriate. Survivorship was determined using an endpoint of either planned or executed revision surgery.

Results

A total of 20 SMPAs were performed in 18 patients from 2000 to 2018. These procedures were performed by 3 certificate of added qualification (CAQ)-certified hand surgeons at 2 hospitals within the same parent institution. Eight out of 20 thumbs had concomitant SMPA of the non-thumb digits at the index procedure, and 6 thumbs underwent simultaneous IP fusion. Of the 20 procedures, 17 (85%) were performed in women and 3 (15%) in men. Mean age at time of operation was 61.3 years (range: 33-79). The indication for surgery was RA in 12 (60%), osteoarthritis (OA) in 4 (20%), non-RA inflammatory arthritis in 3 (15%), and post-traumatic instability in 1 (5%).

Of the 18 patients (20 procedures), 2 patients (2 thumbs) were excluded due to lack of sufficient post-operative follow-up. Two patients (2 thumbs) were deceased at the time of telephone follow-up, but chart data up through most recent follow-up were included.

Of the remaining 14 patients (16 thumbs) available for telephone contact, 11 (12 thumbs) were reached. None had undergone further surgery on their hands since their last examination. All 11 of these patients provided post-operative pain scores via the NRS. Average post-operative pain was rated at 0.8 points (CI, -0.12 to 1.7) out of 10. Nine of the 11 patients were willing to participate in the bMHQ and quickDASH assessments. Mean bMHQ and quickDASH scores were 71.1 (CI, 55.6-86.5) and 35.6 (CI, 18.2-53.0), respectively.

Short-term complications were seen in 3 of 18 thumbs. The most common complication was clinical deformity (2 thumbs). Clinical deformity manifested as recurrence of the preoperative thumb posture—pseudo-Boutonniere deformity in 1 patient, and MP flexion deformity in another. Lastly, one patient experienced instability with persistent ulnar laxity and inability to maintain functional arc of motion.

Of the 3 thumbs with early recurrent deformity or instability, none resolved over time. One patient who was previously asymptomatic suffered an implant dislocation 9 years post-operatively, thus leaving a long-term complication rate of 5.6% (1 of 18 thumbs). Of the 2 thumbs with early recurrent deformity, 1 developed instability as defined by ulnar subluxation at the MP, and the other developed a stiff flexion deformity. The latter continued to endorse the stiff flexion deformity at telephone follow-up, although this did not significantly interfere with daily activities. The patient with early post-operative instability continued to be symptomatic after the 1-year mark.

Among the 4 patients with complications, 3 were indicated for revision surgery (Table 1). The patient with the recurrent pseudo-Boutonniere deformity progressing to late instability was scheduled for revision surgery, but this was never performed (Figure 1). At telephone follow-up 12 years post-operatively, the patient reported no pain but declined participation in bMHQ and quickDASH surveys. The second indicated revision was in a patient that underwent the index procedure for a failed MP fusion for post-traumatic arthritis. The patient continued to experience instability at the MP following SMPA and elected to proceed with revision fusion and CMC arthroplasty (Figure 2). This patient was asymptomatic at the last clinic encounter 9 years from the revision fusion. The third indicated revision was performed in a previously asymptomatic arthroplasty that sustained a dislocation at 9 years post-operatively. This patient underwent revision SMPA and reported no pain at telephone follow-up 9 months from the revision procedure (Figure 3); bMHQ and quickDASH scores were 47.9 and 36.4, respectively.

Table 1.

Postoperative Complications and Their Sequelae.

Patients with complications	Short-term (<1 year)	Long-term (> 1year)	Resolution vs revision
Patient A	Stiffness	None	Resolution with continued therapy.
Patient B	Stiffness	None	Resolution with continued therapy.
Patient C	Deformity (Boutonniere)	Instability	Revision: scheduled revision SMPA but aborted.
Patient D	Deformity (MP flexion)	Stiff flexion deformity	Persistent flexed posture; not requiring revision. Patient functionally asymptomatic.
Patient E	Instability	Persistent instability	Revision: revised to fusion.
Patient F	None	Implant dislocation at 9y	Revision: revised to SMPA.

Note. SMPA = silastic metacarpophalangeal arthroplasty.

Figure 1.

Patient A. (a) Comparison of preoperative and (b) 6-year post-operative radiographs. Note recurrent pseudo-Boutonniere deformity of left thumb in post-operative imaging. Wrist arthrodesis performed prior to index procedure.

Figure 2.

Patient B. 1-year post-operative (a) AP and (b) lateral radiographs from revision fusion with concomitant CMC arthroplasty. CMC = carpometacarpal.

Figure 3.

Patient C. 1-year post-operative radiograph from revision thumb SMPA. SMPA = silastic metacarpophalangeal arthroplasty.

Including planned and executed revisions, survivorship was 83% at mean follow-up of 5.8 years (4.4-7.1)

Discussion

In end-stage degeneration of the thumb MP, surgical management is aimed at pain relief, preservation of function, and preventing additional insult to neighboring joints.¹⁰ In the past 50 years, thumb MP fusion has been the preferred procedure for all-cause joint degeneration.^1,10-12 Techniques have involved various methods of Kirschner-wire (K-wire) fixation, tension-banding, compression screw fixation, and plate fixation.^{10,11,13,14,16-21} The success of this procedure is attributed to its ability to provide pain relief and stability and the fact that loss of motion at this joint is generally tolerated well.^10,16

Stanley et al¹³ reported one of the first series of K-wire instrumented fusions using the Omer method in 1982. Among 42 thumbs, 83% were “thought to be successful” as determined by both pain relief and return to function at approximately 2-year follow-up. They reported a 16.7% complication rate, with the most common—71.4%—being non-union. Bicknell et al¹⁷ reported on 27 patients who underwent K-wire instrumented fusion using a single longitudinal K-wire with cancellous autograft and noted a 100% rate of radiographic union. In all, 89% of the patients were able to return to their previous occupations. However, many reported decreased pinch strength, thumb opposition, and overall dexterity.

As a favorable alternative to K-wire fixation, tension band constructs have been shown to yield superior outcomes with the caveat of longer operative time and greater technical challenge. In 1992, Ijsselstein et al¹⁸ directly compared outcomes of K-wire fixation versus tension-banding and reported fusion rates of 85% versus 95%, respectively.

Thumb MP arthrodesis via cannulated intramedullary screw fixation has also been shown to yield superior time to fusion than tension banding and was first reported by Schmidt et al¹⁹ in 2004. In their series of 26 patients, average time to union was 10 weeks. Of the 26 patients, 25 reported successful return to work and pain relief. One patient, however, developed non-union and was revised to a tension band construct.

In 2006, Proubasta et al²⁰ introduced arthrodesis via Herbert screw fixation. In this method, the joint is osteotomized to place the joint in 10 degrees of flexion and pronation and the medullary canal of the proximal phalanx is tapped. An appropriately sized screw is then placed anterograde through the dorsal cortex of the distal first metacarpal for compression. In their series of 30 patients, they reported a 100% fusion rate without any complications.

Use of fixed angle intramedullary compression devices has also been described with promising short-term results. In 2014, Vanderzanden et al²¹ reported on 17 patients using a 25° fixed angle device. They reported clinical and radiographic evidence of union by 7.9 weeks in all patients with no complications. However, mean follow-up was limited at 4.9 months.

Of more recent, plate fixation with or without compression has also had proven success. In 2012, Rao²² reported on 30 patients using a dorsal spanning plate with transarticular compression screw fixation. Average time to union was 11.8 weeks and at 1-year follow-up there were no reported instances of infection or symptomatic hardware.

Despite improvements in technique, fusion comes with infrequent but considerable and often lasting collateral morbidity. Several studies suggest that fusion at the MP may exacerbate disease at the adjacent CMC and IP joints and may be contraindicated in patients with compromised CMC and IP mobility.^10,11,15 This was first demonstrated in Inglis’ series of RA patients in 1972, in which over 60% of patients who underwent fusion developed adjacent CMC pain and IP hyperextension deformities.¹⁵ A more recent series in chronic, non-rheumatoid thumbs found that over 80% cited one or more “difficulties” postoperatively, with 26% “regretting” the procedure entirely.²³

Although the mechanics thumb and non-thumb MPs certainly differ, SMPA has yielded high rates of success for end-stage RA in the non-thumb digits. Pain relief and patient satisfaction are nearly unanimous among long-term follow-up studies.^2,7,8 In 1 multicenter clinical trial, SMPA showed superior improvement in MHQ scores compared to medical management.⁷ This was most emphasized in the functional, aesthetic, and satisfaction MHQ domains.⁷ Favorable outcomes have also been demonstrated in non-rheumatoid MPs. Neral et al⁸ performed a retrospective review on 38 digits with combined post-traumatic and osteoarthritic etiologies and noted >85% pain relief, statistically significant improvement in DASH scores, and an 89% survival rate at mean 4.6-year follow-up.

In the thumb MP itself, arthroplasty has favorable published outcomes despite fusion being the mainstay surgical option. In our review of the literature, only 4 previous studies have reported results of thumb MPA. Swanson and Herndon were the first to report their case series of SMPA in 34 patients (44 thumbs) in 1977.⁴ Only 1 out of the 44 procedures was performed for non-rheumatic etiology. The primary indication was for pain and all reported “good to excellent” relief at follow-up. An 11% implant fracture rate was reported, although 4 out of 5 of these patients remained asymptomatic and no implant fractures occurred in the more high-performance elastomer implant. Nevertheless, 3 of these patients went on to revision—2 fusions and 1 revision arthroplasty—for survival rate of 89% at mean 3-year follow-up.

Figgie et al⁹ reported on their series of 38 patients (43 thumbs) in 1990. Indications for surgery were RA in all patients, with >70% having flexible (Nalebluff Type 1, Stage 1) deformity of the MP at time of operation. Out of 43, 40 also involved ipsilateral procedures, most commonly MP arthroplasty of the adjacent digits. Pain improvement was unanimous, with nearly 80% reporting no pain. No implant fractures were reported. In all, 1 patient required revision to fusion for reported survival of 98% at mean 6.5- year follow-up. In addition, 6 of these patients had a contralateral thumb MP fusion and preferred the arthroplasty side due to preserved motion. In all, 3 of these patients complained of CMC pain on the fusion side, with 1 also reporting “difficulty” with the IP joint.

McGovern et al in 2001 published a follow-up study to the initial report by Beckenbaugh and Steffee in 1981.^24,25 The “Steffee” implant was a metal-on-poly design that allowed hinged motion at the MP, while also providing multi-axial stability. Their rationale for arthroplasty was that fusion often resulted in loss of both thumb length and functional capacity. Forty-nine out of the 54 procedures were performed for RA, 3 for connective tissue etiologies, and 2 for OA. More than 75% of the patients had one or more complaints of pain, instability, or clinical deformity preoperatively, and nearly all of these findings resolved at short-term follow-up. At long-term follow-up, the complication rate was 7% (4 of 54). Two patients required revision to fusion for infection, resulting in 89% survivorship at 10 years.

Our series demonstrated 83% survivorship at a mean follow-up of 5.8 years. All patients reported excellent pain relief at follow-up. Only one late complication was noted (joint dislocation). Of the 3 patients that were indicated for revision, 2 had preoperative risk factors for failure as currently reported in the literature—1 with post-traumatic instability and failed fusion,¹⁴ the other with significant preoperative instability and IP hyperextension deformity.^10,11 The patient that developed long-term stiffness not requiring revision did have a fixed flexion deformity of the thumb MP preoperatively.

While our mid-term results show that thumb SMPA is a durable option for end-stage arthritis, further prospective study is certainly needed to compare SMPA and fusion in the MP joint of the thumb. The 2 major limitations of our study are the retrospective acquisition of data and the limited number of patients. Also, limited preoperative outcome scores were available to assess post-operative change.

While the numbers in this study are smaller in comparison to both prior thumb MP fusion and thumb MPA series, the data do support SMPA as a durable option in end-stage disease. Our current indications to consider thumb SMPA are patients with degenerative changes of the thumb MP joint, with adjacent thumb joints likely to require fusion in whom the morbidity of the loss of total thumb axis motion is likely to be functionally significant.

Our results do support that SMPA may be contraindicated in patients with deficient soft tissue stabilizers.^1,10-12 In these cases, fusion is indicated.