Arthroscopic Abrasion Arthroplasty Is Not Superior to Ligament Reconstruction and Tendon Interposition for Thumb Carpometacarpal Arthritis

Abstract

Background: A few arthroscopic options have been proposed for the treatment of early stages of the arthritis of the thumb carpometacarpal (CMC) joint. The purpose of this study was to compare the results of arthroscopic abrasion arthroplasty with ligament reconstruction and tendon interposition (LRTI). Methods: In this retrospective cohort study, 11 patients who underwent thumb CMC joint arthroscopic abrasion arthroplasty were compared with 15 patients who were randomly selected from a group of 80 LRTI arthroplasty patients during the same study period, with a minimum 1 year follow-up. Preoperative and postoperative evaluations included radiographs and measurements of grip strength, visual analog scale (VAS) pain scores, and Disabilities of the Arm, Shoulder and Hand (DASH) scores. Results: The mean preoperative VAS score in both groups was 6.8. Mean preoperative DASH scores were 61.1 in the arthroscopy group and 67.4 in the LRTI group. Postoperative VAS score at final follow-up was 4.8 in the arthroscopy group and 1.2 in the LRTI group (P < .05). Postoperative DASH scores 9 months after surgery were 23 for the LRTI group and 55.2 for the arthroscopy group (P < .05). Eight patients in the arthroscopy group had a second surgery due to persistent pain. Conclusions: Patients undergoing arthroscopic abrasion arthroplasty had high revision rates, higher postoperative pain, and lower patient-rated outcomes than patients undergoing LRTI procedure. The poor results in the arthroscopy group may be secondary to the inherent instability of the CMC joint and lack of use of any biological or artificial interposition material.

Introduction

The basal joint, or thumb carpometacarpal (CMC) joint, is responsible for many of the unique features of the human thumb. The joint is comprised of two shallow saddles with little bony congruity, which allows for substantial range of motion at the expense of structural stability. This semiconstrained design enables motion in 3 planes, which, when combined, permit thumb opposition, retropulsion, palmar abduction, radial abduction, palmar adduction, and radial adduction.⁹ The multidirectional forces concentrated on the thumb CMC joint predispose it to the development of osteoarthritis (OA). It is not surprising, therefore, that up to 1 in 3 women and 1 in 8 men can develop early arthritic change in this joint, due to hypermobility.⁴

The basic goals of surgical treatment are elimination of pain with preservation of thumb function. A variety of techniques have been described to achieve this goal over the years, including arthrodesis, metacarpal osteotomy, silicone arthroplasty, total joint arthroplasty, and ligament reconstruction.²³ Open resection of the trapezium with or without a ligament reconstruction or an interposition is commonly performed for advanced disease.²³ However, complete removal of the trapezium may potentially lead to subsidence, and thus shortening of the first metacarpal axis.

The first clinical descriptions of the use of arthroscopy for the treatment of thumb CMC arthritis surfaced in mid-1990s with reports by Menon,¹⁹ Berger,⁷ and Osterman et al.²¹ With an arthroscope, the joint can be assessed without violation of the capsule. In addition, the pathology can be addressed without complete removal of the trapezium. Arthroscopic treatment of basal thumb arthritis may include joint debridement or hemitrapeziectomy, with or without an interposition. The results of these arthroscopic approaches to the treatment of arthritis of the thumb CMC joint have been favorable in several clinical series.^2,10,13,15

We performed trapeziometacarpal joint arthroscopic synovial and osteophyte debridement with biological resurfacing by means of abrasion and microfracture in a limited number of patients at our institution. The goal of this study was to compare the clinical results of arthroscopic abrasion arthroplasty with the clinical results of trapezium excision, ligament reconstruction and tendon interposition (LRTI) arthroplasty for the treatment of thumb CMC OA primarily in terms of postoperative pain relief. We also aimed to compare the 2 procedures using specific objectives such as grip strength, patient-rated outcome measures, number of complications, and revision surgeries. We hypothesized that the clinical results of the arthroscopic intervention would be comparable to or superior to LRTI.

Materials and Methods

This retrospective cohort study was approved by an institutional review board and all patients granted informed consent. All patients who underwent thumb CMC joint arthroscopy with biological resurfacing by the senior author (T.O.) in the period 2010-2013 were identified. The clinical results were compared with the outcome of a control group of patients treated with LRTI arthroplasty. Patients with a minimum postoperative follow-up of 1 year were included in both groups. Radiographic staging of trapeziometacarpal arthritis was performed utilizing the Eaton and Glickel classification system.¹¹ All patients underwent a minimum 6-month trial of conservative management which included nonsteroidal anti-inflammatory medication, splinting, corticosteroid injections, and activity modification prior to consideration of surgical treatment in either group. Patients with stage II and stage III arthritis who failed a trial of conservative treatment were offered arthroscopic biological resurfacing. The surgical indications were patients with a younger age and less severe arthritis. Advanced arthritis (Eaton and Glickel stage IV or Badia stage III) and advanced age were exclusion criteria for arthroscopic treatment.

To compare results of arthroscopic management with a more standard and established surgical treatment for thumb CMC OA, all patients who underwent LRTI by the senior author (T.O.) during the same study period were identified. This search yielded 80 patients; 62 patients had more than 1 year follow-up and appropriate documentation. We aimed to match the size of the arthroscopy group and the control group. We numbered the control patients in their chronological order and chose every fourth patient to include in the cohort group. Accordingly, 15 were randomly selected to be included in the study.

In the arthroscopic group, all 11 patients satisfied the inclusion criteria of at least 1 year follow-up. There were 4 male patients and 7 female patients. The average age was 57 years. Preoperatively, 8 patients presented with stage II thumb CMC arthritis and 3 patients had stage III disease. In the LRTI group, there were 2 male patients and 13 female patients. The average age at the time of surgery was 63 years. Four patients presented with stage II disease, 10 patients were stage III, and 1 patient had stage IV disease. The 2 groups were evaluated for age, gender, and hand dominance (Table 1).

Table 1.

Demographic Information.

Demographics	Abrasion arthroplasty	LRTI
Patients	11	15
Gender (M:F)	4:7	2:13
Age, y (mean)	57	63
Hand dominance (L:R)	4:7	4:11
Radiographic stage
Eaton II	8	4
Eaton III	3	10
Eaton IV		1

Note. LRTI = ligament reconstruction and tendon interposition.

Preoperative demographic information was gathered, which included patient’s age and gender. Preoperative and postoperative evaluations included radiographs and measurements of grip strength, visual analog scale (VAS) pain scores, and Disabilities of the Arm, Shoulder and Hand (DASH) scores. Evaluations from the preoperative visit and the final follow-up were compared using Mann-Whitney U test.

Arthroscopic Abrasion Arthroplasty Surgical Technique

Surgeries were performed under regional anesthesia using a standard sterile preparation and a pneumatic tourniquet. The hand was suspended vertically in an arthroscopic traction tower using a single finger trap secured to the thumb with about 5 to 8 lbs traction applied. The thumb CMC joint was entered at the 1U portal and insufflated with 1 mL of sterile saline. Standard 1U and 1R portals were then created using blunt dissection in the technique described by Berger.⁷ A 1.9-mm arthroscope (Karl Storz, Tuttlingen, Germany) was introduced from the ulnar portal and the trapeziometacarpal joint was explored. A synovectomy was performed followed by removal of any loose bodies or bone spurs. The cartilage surface of the trapezium and metacarpal base was then assessed (Figure 1). Using Badia’s classification system, if the CMC arthritis was found to be stage II or less, we proceeded with abrasion arthroplasty.⁶ With the arthroscope in the ulnar portal and a shaver inserted in the radial portal, a synovectomy was done first. All loose bodies, bone spurs, and flapping cartilage on the trapezium were removed until subchondral bone was exposed (Figure 2). At the end of the procedure, the CMC joint was reduced, and a 0.054 Kirschner wire was placed. A thumb spica splint was applied. We kept the pin and immobilized the thumb for 6 weeks. Patient’s with Badia stage III arthritis were additionally treated with a biological spacer but were not included in this study.⁵

Figure 1.

Arthroscopic assessment of the articular surface of the trapezium reveals Badia stage II arthritis in this patient.

Figure 2.

To complete the abrasion arthroplasty, synovectomy, loose body removal, bone spur excision, and osteochondroplasty was done.

LRTI Arthroplasty Surgical Technique

Like the arthroscopic procedures, surgeries were performed under regional anesthesia using a standard sterile preparation and a pneumatic tourniquet. A chevron-shaped incision was made over the dorsum of the thumb centered over the CMC joint. The superficial branches of the radial nerve and the radial artery were identified and protected. The capsule was incised between the extensor pollicis brevis and abductor pollicis longus tendons. The trapezium was excised in a piecemeal fashion. The distal flexor carpi radialis (FCR) tendon was identified near its insertion. A 2-cm incision was then made in the volar radial aspect of the proximal forearm to retrieve a strip of FCR. A folded 0 stainless steel wire was passed through the FCR sheath from the proximal incision to the CMC joint. Another wire was then used to harvest a strip of FCR comprising about half of the tendon, which was retrieved into the proximal incision.²² A 3.5-mm drill was then used to make 2 intersecting tunnels at the thumb metacarpal base. The first tunnel was drilled from the dorsal radial aspect of the metaphysis, parallel to the base. The second tunnel was drilled from the base, perpendicular to the first, intersecting it to form a “T.” The FCR tendon was passed into the transverse tunnel from the ulnar side of the thumb metacarpal and out the tunnel made in the base. It was then looped around the intact portion of the FCR tendon and secured with a 3-0 Ethibond suture (Ethicon, Somerville, New Jersey). The tendon was then looped around several more times to form a tendon ball and again secured with Ethibond suture. The capsule was repaired and the skin closed in layers. A thumb spica brace was applied for 6 weeks and the thumb progressively mobilized for 6 weeks afterward.

Results

The average follow-up of the arthroscopic arthroplasty group was 15 (range, 12-22) months and the average follow-up of the LRTI group was 14 (range, 12-19) months. The mean preoperative VAS score in both groups was 6.8. The mean preoperative DASH scores were 61.1 in the arthroscopy group and 67.4 in the LRTI group. Postoperative VAS score at final follow-up was better in both groups: 4.8 in the arthroscopic arthroplasty group (P < .05) and 1.2 in the LRTI group (P < .001). The mean postoperative DASH score 9 months after surgery was 23 for the LRTI group (P < .001) and 55.2 for the arthroscopic arthroplasty group (P = .25) (Table 2). Grip strength in the LRTI group was significantly better postoperatively (P < .0001), while the grip strength in the arthroscopy group was similar postoperatively (P = .17). When 2 groups were compared with each other, the LRTI group fared better than the arthroscopy group in terms of pain relief (P < .0001) and DASH scores (P < .0001); however, there was no significant difference between groups in terms of postoperative grip strength.

Table 2.

Results.

	Abrasion arthroplasty	LRTI
Follow-up, months (mean [range])	15 (12-22)	14 (12-19)
Pre-op VAS (mean)	6.8	6.8
Post-op VAS (mean)	4.8	1.2
Pre-op DASH (mean)	61.1	67.4
Post-op DASH (mean)	55.2	23
Grip strength (lbs)
Preoperative	39	29
Postoperative	49	47
Secondary surgery	8	0
Complications	1	0

Note. LRTI = ligament reconstruction and tendon interposition; VAS = visual analog scale; DASH = Disabilities of the Arm, Shoulder and Hand.

Preoperative radiographs were used in staging of patients in both groups. The postoperative radiographs in the arthroscopy group showed excision of the impinging bone spurs but the loss of joint space continued. The postoperative radiographs in the LRTI group showed an excised trapezium and slight collapse but maintained space between the first metacarpal and the scaphoid.

Eight patients in the arthroscopic arthroplasty group had a second surgery due to lack of relief in symptoms compared with none in the LRTI group (P < .0001). Six patients underwent an LRTI as previously described; 2 patients underwent arthrodesis because of their age and level of activity. All 8 patients were relieved of their symptoms after the second surgery. These 8 patients were followed for another year after their second surgery. The mean pain score was 1.5 at last follow-up for these 8 patients and the DASH score was 24.3. No major complications were observed. One pin site infection was observed in the arthroscopic arthroplasty group, which resolved with oral antibiotics.

Discussion

The clinical results of thumb CMC arthroscopic abrasion arthroplasty showed inferior results compared with the LRTI arthroplasty in this study. We attempted arthroscopic abrasion arthroplasty in our patients with the optimism that it could have provided good pain relief while keeping the support of the trapezium in place, and that it could have delayed transition to LRTI arthroplasty in young patients. Osterman et al reported satisfactory pain relief with arthroscopic debridement in degenerative arthritis of the thumb CMC and proposed it could delay salvage arthroplasty, especially in the young patient.²¹ In our study, the pain scores improved in both groups after the surgery, but LRTI arthroplasty group reported significantly better pain control, DASH scores, and grip strength with less reoperation rate (P < .0001).

Thumb CMC arthritis is a very common problem faced by the hand surgeon. However, a consensus has not been reached regarding the optimal surgical treatment approach. Gervis first advocated for the simple trapeziectomy in 1949.¹⁴ Subsequent series, however, reported subsidence of the thumb metacarpal and loss of grip strength.²⁰ A desire to prevent shortening of the thumb axis and prevent late contact between the thumb metacarpal and scaphoid led to development of numerous methods to reconstruct the ligamentous restraints of the CMC joint and interpose biological material in the trapezial space. Eaton and Littler developed one of the earliest methods to stabilize the trapeziometacarpal joint.¹² Burton and Pelligrini⁸ and Weilby²⁴ popularized 2 different methods of ligament reconstruction and tendon interposition in 1986 and 1988, respectively. Variants of these methods are widely used to this day.

Alongside these open treatments, arthroscopy of the thumb CMC joint began to emerge as a treatment option in the 1990s. Arthroscopy has both a diagnostic and therapeutic role in the treatment of CMC arthritis.^3,21 The theoretical advantages of arthroscopic treatment included a shorter recovery period, maintenance of the length of the thumb axis, and preservation of the integrity of the joint capsule, which may be important in patients with instability.⁵

Arthroscopic abrasion arthroplasty with or without interposition of biological materials is recommended for patients with Eaton stage II or III thumb CMC arthritis. Abrasion arthroplasty has long had a role in the treatment of early arthritis of the knee.¹⁶ Both microfracture and abrasion arthroplasty involve the stimulation of bleeding from subchondral bone to create a stable fibrin clot and subsequent transformation into fibrocartilage.¹⁸ The critical points of our procedure include thorough debridement of the synovium, removal of loose bodies and osteophytes, and debridement of remaining trapezial cartilage down to bleeding subchondral bone. Similar to other joints, abrasion is expected to result in the formation of fibrocartilage over the debrided trapezium, thereby eliminating the source of pain while retaining the capsular stabilizers and maintaining thumb length.¹⁷ However, our retrospective comparison of arthroscopic abrasion arthroplasty with LRTI performed by the same surgeon has shown a significantly higher VAS and DASH scores at final follow-up. Eight patients underwent secondary surgery after abrasion arthroplasty with subsequent relief of symptoms. It is possible that the abrasion and microfracture of the trapezium did not produce an adequate layer of fibrocartilage to eliminate pain in the thumb basal joint. When compared with the other joints where abrasion arthroplasty has successfully been applied, the thumb basal joint is extremely mobile and unstable. The eroded surfaces on the trapezium and the first metacarpal continued impinging on each other, causing pain. The authors believe the inherent instability is the reason for failure of the procedure. The more stable ankle and knee joints may be more conducive to the creation of sufficient fibrocartilage.

If the cause of persistent pain after arthroscopic abrasion arthroplasty is impingement and inadequate formation of fibrocartilage due to inherent instability, a possible solution may indeed be the interposition of biologic or artificial material. Menon in 1996 presented his results of arthroscopic partial trapeziectomy and interposition with either tendon, Gore-Tex, or fascia lata and achieved complete pain relief in 75% of patients.¹⁹ The results of arthroscopic hemitrapeziectomy, with and without tendon interposition, debridement of the joint with interposition of GRAFTJACKET, and hemitrapeziectomy with capsular treatment have all been good in recent clinical series.^2,10,13,15 Interposition was used with success in conjunction with arthroscopic partial removal of the trapezium in several earlier reports.^1,2,10,19 Abzug and Osterman recommended a suture button technique as an alternative to pinning when the joint was found unstable.¹ Although we did not interpose any biological spacer in the joint in the arthroscopy group, reduction was maintained by a temporary pin inserted percutaneously.

This study has several limitations. The sample size in the arthroscopic arthroplasty group was small. It was performed as a retrospective study, and certain objective clinical outcome data, such as key pinch, were not gathered during clinic visits; however, the focus of the study was treatment of pain. Though the patients in the LRTI group were chosen randomly, a possibility of selection bias exists, as those chosen for arthroscopy may have had less severe disease. Nevertheless, their outcomes were worse compared with the LRTI group.

Although the role of arthroscopy in the treatment of CMC arthritis continues to evolve, abrasion arthroplasty of the trapezium alone did not achieve the goals of a pain free and stable joint in our comparative study. Based on our results, we cannot recommend arthroscopic abrasion arthroplasty as a treatment for any stage of CMC arthritis.