Radiographic Thumb Metacarpal Subsidence Following Ligament Reconstruction With Tendon Interposition and Suture-Only Suspension Arthroplasty in the Treatment of Basal Joint Arthritis

Abstract

Background:

The thumb carpometacarpal (CMC) joint is a common source of osteoarthritis. Following trapeziectomy, ligament reconstruction with tendon interposition (LRTI) is considered a “gold standard” treatment, but suture-only suspension arthroplasty (SSA) has recently emerged as a simpler alternative. Currently, there is no objective radiographic study comparing subsidence following these 2 techniques.

Methods:

This study is a retrospective review of 23 patients (10 LRTI, 13 SSA) that had at least 6 months of radiographic follow-up following thumb CMC arthroplasty. Posteroanterior radiographs at a preoperative timepoint, and at the 2-week and greater than 6-month postoperative timepoints were evaluated for actual trapezial height, as well as trapezial height normalized to capitate, thumb metacarpal, and proximal phalangeal heights. Normalized trapezial heights were calculated, and preoperative values were compared with greater than 6-month postoperative values. In addition, actual and normalized trapezial heights following LRTI and SSA were compared at each timepoint.

Results:

Mean trapezial height decreased from approximately 12 to 5 mm (reduction of ~60%, P < .05) in both groups with no differences when comparing LRTI and SSA at each timepoint. All normalized trapezial heights revealed differences from preoperative to greater than 6-month postoperative timepoints, but no differences between LRTI and SSA.

Conclusions:

Ligament reconstruction with tendon interposition and SSA exhibit equivalent actual and normalized trapezial heights over a greater than 6-month postoperative time course.

Introduction

Thumb carpometacarpal (CMC) joint osteoarthritis is very common, with 25% of men and 40% of women older than 75 years exhibiting radiographic evidence of the diagnosis.^1,2 Owing to the resulting pain and disability associated with end-stage basal joint osteoarthritis, it is the most commonly reconstructed joint in the upper extremity. ³ Following trapeziectomy, hand surgeons have many reconstructive options, with the mainstay of treatment being ligament reconstruction with tendon interposition (LRTI).^3,4 However, suture-only suspension arthroplasty (SSA) has emerged as a simpler alternative.^5,6 Both LRTI and SSA are effective at suspending the thumb metacarpal after trapeziectomy.^3-6

Standard wrist radiographs remain a widely available tool for preoperative workup of individuals with CMC arthritis. Postoperatively, they provide objective measures of thumb metacarpal subsidence, which may correlate with functional outcomes in the postoperative period. ⁷ While a recent study indicates that TightRope suspensionplasty is superior to LRTI for maintenance of trapezial height at 3 months postoperatively, analyses have not been conducted to compare long-term radiographic outcomes following LRTI and SSA. ⁸ This retrospective radiographic cohort study reviews thumb metacarpal subsidence on radiographs following patients who have undergone LRTI or SSA. We hypothesize that radiographic thumb metacarpal subsidence will not differ following either LRTI or SSA.

Methods

Following approval by an institutional review board (University of Pittsburgh IRB#17090632), all study participants provided their informed consent for retrospective review of their medical records. All recruited individuals had advanced stage thumb basal joint arthritis and underwent thumb CMC arthroplasty by a senior fellowship-trained hand and upper extremity surgeon. Patient underwent either an LRTI or SSA, which reflected their surgeon’s preferred technique. Patients with no postoperative radiographs were excluded.

Two surgeons performed LRTI as they have for the last decade, harvesting the flexor carpi radialis (FCR) tendon in its entirety. Approximately half of the FCR tendon was threaded through bone tunnels in the base of the first metacarpal, and the remaining FCR tendon was rolled into an anchovy and used to fill the trapezial space. Another 2 surgeons adopted SSA approximately 5 years ago and have since used it as their standard-of-care operative intervention for advanced thumb CMC arthritis. They perform this technique as described by DelSignore and Accardi ⁵ and Weiss et al, ⁶ which involves suture tenodesis of the abductor pollicis longus and FCR tendons.

For all patients, trapezial height and subsidence measurements were done using standard posteroanterior wrist radiographs obtained preoperatively, 2 weeks postoperatively, and at a clinic visit greater than 6 months postoperatively. All radiographs were taken under nonloaded static conditions (ie, no pinch). The height of the trapezium or trapezial space was measured as described by Yao (Figure 1). ⁹ To account for magnification, normalized trapezial height was calculated with respect to the height of the capitate, thumb metacarpal, and thumb proximal phalanx, as measured on each radiograph. ¹⁰ The normalized subsidence ratio was also determined by computing the ratio of postoperative to preoperative normalized trapezial height.

Figure 1.

Example radiograph depicting how trapezial height and normalized trapezial height measurements were done with respect to the capitate, thumb metacarpal, and thumb proximal phalanx heights both preoperatively and postoperatively.

Statistical analysis was done using GraphPad Prism (GraphPad software, San Diego, California). The Wilcoxon matched-pairs signed rank test was used to evaluate mean differences in radiographic measurements for the same procedure across postoperative timepoints. The unpaired Mann-Whitney test was used to evaluate differences between the 2 procedures at each postoperative timepoint. All statistical analyses were nonparametric and 2-tailed with the significance level defined at P < .05.

Results

This study included 23 patients, with 10 having undergone LRTI and 13 having undergone SSA (Table 1). There were no differences in the trapezial height preoperatively between groups (12.9 ± 0.4 mm for LRTI vs 12.3 ± 0.3 mm, P = .2). The final radiographs were taken at a mean postoperative time of 32.7 ± 8.1 weeks (~8.1 months) for the LRTI group and 37.7 ± 5.2 weeks (~9.4 months) for the SSA group, with no differences (P = .14). Radiographically measured trapezial height decreased from preoperative to final postoperative radiographs for both LRTI (P ≤ .002) and SSA (P ≤ .0002) (Figure 2). However, there were no differences in the final postoperative trapezial height between groups (5.0 ± 0.7 for LRTI vs 5.2 ± 0.5 for SSA, P = .6). Normalized trapezial height, regardless of the reference used (capitate, thumb metacarpal, or proximal phalanx), decreased significantly from preoperative to the final postoperative timepoint following both LRTI (P < .0001) (Figure 3) and SSA (P < .001) (Figure 4). There was no difference in the normalized trapezial heights between LRTI and SSA from preoperative to the final postoperative timepoint (Figure 5).

Table 1.

Demographic Data and Mean Follow-up Time.

	LRTI	SSA
Number	10	13
Mean age (±SD)	63.3 (±12.6)	59.8 (±10.1)
Female sex (%)	9/10 (90.0%)	10/13 (76.9%)
Active or former smoker	4/10 (40.0%)	10/13 (76.9%)
Diabetic	1/10 (3.8%)	3/13 (21.1%)
Right hand dominance	10/10 (100.0%)	11/13 (84.6%)
Treatment done on dominant hand	5/10 (50.0%)	5/13 (38.5%)
Mean follow-up, wk (±SD)	32.3 (±25.5)	37.7 (±18.7)

Note. LRTI = ligament reconstruction with tendon interposition; SSA = suture-only suspension arthroplasty.

Figure 2.

Differences in trapezial height, as measured on preoperative and final postoperative radiographs, between patients having undergone LRTI and SSA.

Note. LRTI = ligament reconstruction with tendon interposition; SSA = suture-only suspension arthroplasty.

Figure 3.

Comparison in normalized trapezial height, as measured using the capitate, metacarpal, or proximal phalanx as a reference standard, between preoperative and final postoperative radiographs in patients having undergone LRTI.

Note. LRTI = ligament reconstruction with tendon interposition.

Figure 4.

Comparison in normalized trapezial height, as measured using the capitate, metacarpal, or proximal phalanx as a reference standard, between preoperative and final postoperative radiographs in patients having undergone SSA.

Note. SSA = suture-only suspension arthroplasty.

Figure 5.

Differences in normalized subsidence ratio, as measured using the capitate, metacarpal, or proximal phalanx as a reference standard, between patients having undergone LRTI and SSA.

Note. LRTI = ligament reconstruction with tendon interposition; SSA = suture-only suspension arthroplasty.

There were significant decreases in the normalized trapezial heights from the preoperative to the 2-week postoperative timepoint in the LRTI group (normalized to capitate height, 0.53 ± 0.01 to 0.21 ± 0.02 [P = .002]; normalized to thumb metacarpal height, 0.28 ± 0.01 to 0.11 ± 0.01 [P = .002]; normalized to thumb proximal phalanx height, 0.41 ± 0.01 to 0.16 ± 0.02 [P = .002]). Similarly, there were significant decreases in the normalized trapezial heights from the preoperative to the 2-week postoperative timepoint in the SSA group (normalized to capitate height, 0.55 ± 0.02 to 0.35 ± 0.02 [P = .0002]; normalized to thumb metacarpal height, 0.26 ± 0.01 to 0.17 ± 0.01 [P = .0002]; normalized to thumb proximal phalanx height, 0.41 ± 0.01 to 0.17 ± 0.02 [P = .007]). When comparing the LRTI and SSA groups from the 2-week to the final (>6-month) postoperative timepoint, there were differences in the subsidence over that time period. For the LRTI group, the normalized trapezial heights from the 2-week to the final postoperative timepoint were no different (normalized to capitate height, 0.21 ± 0.02 to 0.2 ± 0.02 [P = .68]; normalized to thumb metacarpal height, 0.11 ± 0.01 to 0.11 ± 0.02 [P = .93]; normalized to thumb proximal phalanx height, 0.16 ± 0.02 to 0.07 ± 0.02 [P = .95]). For the SSA group, the normalized trapezial heights from the 2-week to the final (>6-month) postoperative timepoint were different (Figure 6) (normalized to capitate height, 0.35 ± 0.02 to 0.23 ± 0.03 [P = .001]; normalized to thumb metacarpal height, 0.17 ± 0.01 to 0.11 ± 0.01 [P = .002]; normalized to thumb proximal phalanx height, 0.25 ± 0.01 to 0.17 ± 0.02 [P = .007]). Normalized measurements between LRTI and SSA groups were different depending on the reference. There were differences between groups at the 2-week postoperative timepoint when normalized to capitate height (0.21 ± 0.02 for LRTI vs 0.35 ± 0.02 for SSA, P = .0002) or thumb metacarpal height (0.11 ± 0.01 for LRTI vs 0.17 ± 0.01 for SSA, P = .001). However, there were no differences between groups at the 2-week postoperative timepoint when normalized to the thumb proximal phalanx height (0.16 ± 0.02 for LRTI vs 0.25 ± 0.01 for SSA, P = .66).

Figure 6.

Normalized trapezial heights based on capitate, thumb metacarpal, and thumb proximal phalanx at preoperative, 2-week, and greater than 6-month postoperative timepoints.

Note. The significant decrease for all normalized trapezial heights following both LRTI and SSA from the preoperative to the 2-week postoperative timepoints. However, there was less subsidence following SSA within the first 2 weeks postoperatively, but which equalized over the subsequent ~6 months. LRTI = ligament reconstruction with tendon interposition; SSA = suture-only suspension arthroplasty.

Discussion

We conclude that the average trapezial height decreases comparably between the two techniques. Interestingly, it appears that the thumb metacarpal subsidence may progress more rapidly over the first 2 weeks postoperatively following LRTI when compared with SSA. However, the degree of subsidence appears to equalize between LRTI and SSA over greater than 6 months postoperatively and may therefore be inconsequential. Furthermore, and in general, trapezial heights normalized to capitate, thumb metacarpal, or thumb proximal phalanx height appear to reflect equivalent outcome metrics for subsidence following basal joint arthroplasty. Despite being a well-established treatment of basal joint arthritis, LRTI requires a second incision in the forearm and may be associated with a more prolonged recovery. Consequently, there have been efforts to study alternative techniques that may be simpler, including SSA.

The technical objective of any basal joint arthroplasty is to suspend the thumb metacarpal following trapeziectomy. Durable thumb metacarpal suspension and prevention of subsidence particularly to the degree of impingement onto the distal scaphoid may be important to recurrence of pain. ⁷ In this study, LRTI and SSA exhibit comparable static thumb metacarpal subsidence at the greater than 6-month timepoint. Interestingly, early (2-week) thumb metacarpal subsidence was worse for LRTI but that appears not to have any long-term clinical or radiographic consequences. It is unclear why there is earlier subsidence in the LRTI group as compared with the SSA group. There were no differences in postoperative immobilization protocol between groups, as all patients were splinted for 2 to 3 weeks and did not have the opportunity to load the reconstruction before the first postoperative radiograph. It may simply be that the SSA reflects a tighter initial suspension, but which seems to loosen with postoperative loading.

Limitations of this study include a relatively small sample size and the nonrandomized study design. The approximately 6-month postoperative time frame is a reasonable follow-up but there may be additional subsidence over the long term. It is possible that the trapezial height would continue to shorten with postoperative time frame greater than 6 months, but additional study would need to be done to confirm or refute this hypothesis. Furthermore, static radiographs may not reflect conditions under which the thumb is loaded, as with daily activity. Additional clinical radiographic or even biomechanical study may also include evaluating the effect of varying degrees of loading on the trapezial height following LRTI and SSA.

In conclusion, LRTI and SSA exhibit equivalent thumb metacarpal subsidence over a greater than 6-month postoperative time frame. It may be that the thumb metacarpal does not subside as quickly following SSA, but this may be of no clinical consequence when comparing the two techniques.