Introduction

Trapeziometacarpal osteoarthritis is a common entity, often bilateral and predominantly affecting postmenopausal women [2, 11]. Conservative treatment is mostly successful, and surgical treatment is only indicated in resistant cases. A number of surgical procedures have been described for basal joint osteoarthritis, whereby simple excision was first reported in 1949 [20]. Other surgical procedures for osteoarthritis of the trapeziometacarpal joint include fusion [1, 4, 7, 12, 14, 15, 25, 27, 33], trapeziectomy combined with interposition arthroplasty. For interposition two main procedures are described: soft tissue [30] or silicone [26, 32, 34]. Also, reasonable to good results of total joint replacement have been reported [9, 10]. To the best of our knowledge there have been no reports on simple excision of both joint surfaces of the trapezio-metacarpal joint. The rationale behind this simple resection was based on the experience of one of the authors (AM) that a non-union of the arthrodesis of the trapeziometacarpal joint did not result in complaints. So in the simple resection a non painful non-union is aimed for.

In this study the results of simple resection of the surfaces of the trapeziometacarpal joint are reported and compared to trapeziectomy with tendon interposition and trapeziometacarpal arthrodesis.

Operative technique

For the resection arthroplasty, the joint is approached through a straight radial incision over metacarpal I and the trapezium to the wrist. After exposure of the joint, only the joint surfaces of the trapezium and the metacarpal bone are removed, care must be taken to resect all osteophytes. A minimal distance of 1 cm under traction between the two bones is aimed for. Care is taken to resect all osteophytes but not to resect joint capsule and ligaments. Spongostan (Ferrosan A/S, Soeborg, Denmark) is used as a spacer to fill the gap, and the wound is closed in one layer. After-treatment consists of two weeks of immobilisation in a forearm splint (below the elbow) followed by functional treatment (Fig. 1a,c).

Fig. 1
figure 1

a, b Patient A, preoperative X-rays of a 61 year old female. Eaton stage 2 joint destruction. c, d Patient A one year postoperative, no complaints. e, f Patient A, 10 years postoperative. DASH, VAS for pain and satisfaction all 0 points (maximum score)

Full size image

For the arthrodesis the joint space is approached in a similar way and the joint cartilage and the adjacent sclerotic subchondral bone is removed. A small laminar spreader was interposed to achieve lengthening and correction of the adduction. A stable fixation is achieved by a special miniplate technique. The first screwhole of a 3–4 hole AO mini plate is bent 90° and sunk into the metacarpal bone, acting as a washer to achieve maximal compression with a 2.7 mm cortical screw crossing the former joint space. With a second screw at the level of the trapezium, further compression and buttressing can be achieved by asymmetric drilling [29]. In some cases a corticocancellous graft was interposed. A removable forearm splint allows partial function after treatment (Fig. 2a,c).

Fig. 2
figure 2

a, b Patient B, preoperative X-rays of a 61 year old female. Eaton stage 2 joint destruction. c, d Patient B, direct postoperative X-rays showing the trapezio-metacarpal arthrodesis with the plate washer technique. e, f Patient B, X-rays after five years of follow-up showing a completely united arthrodesis. DASH 54, VAS for pain 30 and for satisfaction 25

Full size image

To perform the resection-interposition arthroplasty, a similar incision is used. A trapeziectomy is performed. Through a second small incision approximately 10 cm proximal to the wrist, the flexor carpi radialis tendon (FCR) is identified and split with a tendon stripper from its muscle belly proximally, leaving the distal insertion intact. The split tendon is rolled and stitched to form an anchovy-like structure of the appropriate size to fill the gap created to the first metacarpal base. This procedure is followed by a scaphoid type of plaster fixation for 6 weeks [33] (Fig. 3a,b).

Fig. 3
figure 3

a, b Patient C, preoperative X-rays of a 62 year old female. Eaton stage 2 joint destruction. c, d Patient C, X-rays after five years follow-up of trapeziectomy with tendon interposition showing a slight calcification on the lateral X ray. DASH 10 points, VAS for pain and for satisfaction both 0 points

Full size image

Patients and methods

In 63 consecutive patients with osteoarthritis of the trapezio-metacarpal joint 74 thumbs were surgically treated at the Hilversum Hospital and the Academic Medical Centre in Amsterdam between 1977 and 2000. The indication for surgical intervention was failure of conservative treatment with idiopathic degenerative osteoarthritis, except for one case of posttraumatic osteoarthritis. 54 patients (63 thumbs) were seen for follow-up, 7 patients (9 thumbs; 12%) had died and 2 patients (2 thumbs; 3%) were lost to follow-up (see Table 1). According to the preference of the surgeon, either resection arthroplasty, trapeziectomy with tendon interposition or trapezio-metacarpal arthrodesis was performed (Table 1). The resection arthroplasty was only performed by one surgeon, the other two techniques were performed by several surgeons but each surgeon performed only one of the techniques. All surgeons were experienced and fully qualified orthopaedic surgeons. At follow-up the patients were evaluated by one independent researcher. The results of the patients seen for follow-up were analysed: 18 thumbs (7 male and 11 female patients) had a resection arthroplasty, 17 thumbs (3 male and 14 female) had a trapeziectomy with tendon interposition and 28 thumbs (4 male and 24 female) had a trapezio-metacarpal arthrodesis. The average age was 58 years in the resection group (range 38–76 years), 65 years in the trapeziectomy group (range 47–80) and 61 years in the arthrodesis group (range 46–78 years). The average follow-up was 13 years (range 3–23 years) for the resection group, 8 years (range 5–12 years) for the trapeziectomy group and 9 years (range 4–27 years) for the arthrodesis group (p=0.025 for resection versus trapeziectomy and trapeziectomy versus arthrodesis, p=0.046 for resection versus arthrodesis). Thirteen dominant hands were operated on in the resection group, with 6 and 12 in the trapeziectomy and arthrodesis groups, respectively. The average preoperative Eaton stage was 2.6 (2–3) in the resection group, 2.4 (0–4) and 1.6 (0–4) in the trapeziectomy and arthrodesis groups, respectively.

Table 1 Patients
Full size table

At follow-up all patients were asked to complete a visual linear analogue scale (VAS) for satisfaction with the result of the operation and persisting pain from the thumb. These scales were graded from 0 to 100 millimeters with “0” indicating high satisfaction and no pain. At follow-up pain relief after operation and pain frequency were scored (never pain, pain sometimes during heavy work, frequent pain and continuous pain). The Dutch validated Disability of Arm Shoulder and Hand questionnaire was used to evaluate function of the affected limb [24, 36].

Range of motion was measured with the neutral-zero method. Grip strength was measured with the use of a Jamarmeter (Therapeutic instrument™, Clifton, N.J. 07012, U.S.A.). Also tip pinch strength was recorded (see Table 2). Pressure pain on the trapezio metacarpal joint was noted at follow-up. Instability of the joint in the resection arthroplasty and the trapeziectomy group was recorded.

Table 2 Scoring tip pinch strength
Full size table

The pre-operative X-rays were scored according to the Eaton classification [17, 18] (Table 3). At follow-up the fusion status of the arthrodesis was recorded.

Table 3 Eaton classification [17, 18]
Full size table

The patients records were used for the evaluation of complications. This was also checked at follow-up.

Statistical analysis was performed with an unpaired 2-tailed Student t-test (confidence level was set at 0.05) and the Pearson’s correlation coefficient was calculated (relevant > 0.7) using Microsoft Excel XP.

Results

Results resection arthroplasty (n=18)

The visual analogue scores for pain and satisfaction were 14 and 8, respectively. Compared to preoperative, 16 thumbs (89%) were significantly improved and 2 thumbs (11%) were slightly improved. At follow-up 8 thumbs (44%) never had pain, 7 thumbs (39%) had some pain during heavy work, 2 thumbs (11%) had frequent pain and 1 (6%) had continuous pain. The average score on the DASH was 26 points (range 0–70).

The radial abduction of the thumb was 45°. The average grip strength with the Jamar-meter was 22 kg (range 6–39 kg). The average tip pinch strength was 3. In 4 thumbs (22%) there was pressure pain at follow-up while pressing on the trapeziometacarpal. No instability was found.

Sensory deficit due to damage to terminal branches of the radial nerve was experienced in 2 thumbs (11%). There was no sensory alteration on the tip of the thumb in any of the groups (compromising the tip pinch). One patient was treated with re-resection for recurrent symptoms 8 years after initial surgery with a good result.

Results trapeziectomy (n=17)

The visual analogue scores for pain and satisfaction were 10 and 2, respectively. Compared to preoperative, 16 thumbs (94%) were significantly improved and 1 thumb (6%) was slightly improved. At follow-up 10 thumbs (59%) never had pain, 5 thumbs (29%) had some pain during heavy work, 2 thumbs (12%) had frequent pain and no patient had continuous pain. The average score on the DASH was 27 points (range 0–73).

The radial abduction of the thumb was 64°. The average grip strength with the Jamar-meter was 20 kg (range 8–37 kg). The average tip pinch strength was 3. In 4 thumbs (24%) there was pressure pain at follow-up while pressing on the trapeziometacarpal. No instability was found.

Sensory alteration due to damage to terminal branches of the radial nerve was experienced in 4 thumbs (24%). There was 1 patient with Sudeck’s dystrophy. None of the patients had reoperations due to complications.

Results arthrodesis (n=28)

The visual analogue score for pain and satisfaction was 23 and 10. Compared to preoperative, 20 thumbs (71%) were significantly improved, 6 thumbs (21%) were slightly improved and 2 thumbs (7%) unchanged. At follow-up 8 thumbs (29%) never had pain, 11 thumbs (39%) had some pain during heavy work, 2 thumbs (7%) had frequent pain and 7 (25%) had continuous pain. The average score on the DASH was 25 points (range 0–83). The radial abduction of the thumb was 50°. The average grip strength with the Jamar-meter was 17 kg (range 3–41 kg). The average tip pinch strength was 3. Six thumbs (21%) had pain at follow-up while pressing on the trapeziometacarpal joint.

Sensory alteration due to damage to terminal branches of the radial nerve was experienced in 5 thumbs (18%). Seven out of 28 patients had an arthrodesis with a corticocancellous bone graft at the primary operation; non-unions were not seen in those patients. In the patients who did not have a corticocancellous bone graft in the primary operation, six non-unions were seen. Three patients had a stiff non-union without complaints, the other 3 patients had a re-arthrodesis which proved to be unsuccessful. All 3 patients had a further re-arthrodesis at that time combined with a corticocancellous bone graft which was successful in 2 out of 3 patients. The last patient still has a very poor result and a non-union after 4 operations. Fifteen out of 28 patients had hardware removal due to local complaints. Three patients developed dystrophy. The total number of reoperations due to complaints or due to a complication was 22 (79% of the patients).

Statistics

There was no statistical difference when comparing the resection and trapeziectomy groups and the resection and the arthrodesis groups concerning the VAS for pain and satisfaction, the pain frequency and the pain compared to preoperative levels. Comparing the trapeziectomy and the arthrodesis group a statistical difference was observed concerning the VAS for pain (p=0.02) and satisfaction (p=0.04), the pain frequency (p=0.01) and the pain compared to preoperative (p=0.03). There was no statistical difference in the DASH scores among the groups (Table 4).

Table 4 Results of the operated thumbs
Full size table

The radial abduction in the resection arthroplasty versus the arthrodesis was not significantly different, while trapeziectomy compared to resection arthroplasty was significantly better (p=0.0001) and also compared to arthrodesis (p=0.0046). There was no difference of grip strength and pressure pain. The tip pinch was best in the trapeziectomy group compared to resection (p=0.03) and arthrodesis (p=0.02), while between the resection and the arthrodesis there was no difference (Table 4).

No significant correlations were found in any of the group comparing age, duration of follow-up, preoperative Eaton classification, sex, dominant side and bilateral operation compared to the VAS for pain and satisfaction and the DASH score (see Table 5).

Table 5 Pearson’s correlation coefficient
Full size table

Discussion

Osteoarthritis of the basal joint of the thumb commonly affects middle aged women [2]. There is a disabling pain and swelling of the base of the thumb associated with deformity, instability, crepitus and loss of motion. If conservative treatment fails surgical intervention usually has good results and leads to satisfied patients independent of the technique used, either trapeziectomy with tendon interposition or arthrodesis [1, 4, 7, 12, 14, 15, 20, 21, 25, 27, 30, 33].

Simple trapeziectomy introduced by Gervis [20] is also an easy and quick procedure which is still favoured [19, 21, 33]. The results are comparable with the results we found for all three procedures. To overcome problems of instability and shortening, interposition of a rolled flexor carpi radialis tendon in the space of the removed trapezium was introduced [19]. In more recent literature ligament reconstruction-tendon-interposition arthroplasties are popular [35]. Although, in the randomised studies of Belcher and Nichol [3] and Davis et al. [13], no statistical difference for trapeziectomy alone or combined with ligament reconstruction and tendon interposition was found. Manske [28] raised the question whether any of the stabilising procedures proposed by Eaton and Littler [16], and Burton and Pelligrini [5] are in fact necessary after trapeziectomy, and whether splinting alone might be sufficient. In our study we did not immobilise the thumb with a K-wire; all patients with a resection arthroplasty were immobilised for two weeks in a splint and for 6 weeks if they had a trapeziectomy with tendon interposition. The results of the trapezectomy might have been better if no tendon interposition had been performed.

Arthrodesis of the trapeziometacarpal joint proved to be a reliable procedure with good long term results. In the literature different methods of fixation have been described [1, 4, 6–8, 12, 15, 25, 27, 33]. The problems of the arthrodesis are failure of the hardware and non-union. Also, complaints of pain due to the thin skin overlying the hardware are frequent. In our series these problems were also noticed resulting in a high rate of re-intervention. Most of the patients with a non-union, however, did not have complaints of pain and therefore had no re-intervention. No non-unions were observed in the cases were a corticocancellous bone graft was used. Better results could possibly have been observed if all patients had a corticocancellous bone graft.

Until now there have been no reports of simple resection of the joint surfaces of the metacarpal bone and the trapezium. This study shows that this procedure leads to good results and satisfied patients in the long term. An explanation can be found in the early functional treatment after a short period of immobilisation. This is in accordance with the results presented by Horlock and Belcher who also advocate early functional treatment [23]. The procedure itself is very simple since no hardware is used; this reduces risk of a re-interventions for hardware failure or complaints.

Although there were some differences in the baseline characteristics, concerning Eaton stage, dominant side and length of follow-up, we consider the groups as comparable since we did not find positive correlations for those characteristics.

Because of the small number of patients we could not establish significant differences between the patient groups comparing satisfaction, pain experience and functional outcome. Comparing the complication and re-intervention rates we noticed considerable differences with worse results in the arthrodesis patients. In the published comparative studies no differences were found between trapeziectomy with or without the addition of ligament reconstruction [3, 13]. Also, comparison of arthrodesis with ligament reconstruction revealed no differences except for a higher complication rate [22]; apart from the publication of Mureau et al. [31] who found that patients undergoing tendon interposition arthroplasty had better thumb opposition, interphalangeal joint mobility, and radial and palmar TMC joint range of motion.

Conclusion

Although we have a small number of patients we show that the resection arthroplasty is a simple procedure and has equally good long term results compared to trapeziectomy combined with tendon interposition and arthrodesis. The trapezectomy with tendon interposition requires one extra incision for harvesting the tendon in comparison with resection arthroplasty. Therefore we consider resection arthroplasty to be the most simple procedure. Resection arthroplasty has less complications and less re-interventions compared to arthrodesis.

The resection arthroplasty is the most simple procedure in our clinic, and is therefore the treatment of choice for osteoarthritis of the trapeziometacarpal joint. We consider this a reasonable alternative to the “gold standard” of the trapezectomy in the treatment of osteoarthritis of the trapeziometacarpal joint. Further randomised trials of the resection arthroplasty and trapezectomy should answer the question as to the outcome compared to the gold standard and if the results remain as good with different surgeons.