Outcomes of Thumb Carpometacarpal Joint Osteoarthritis Treated with Arthroscopic Fusion

Abstract

Background  The thumb carpometacarpal joint (CMCJ) osteoarthritis is one of the most common pathologies in the hand with controversial treatment options.

Description of Technique  Describe the use of arthroscopic technique for thumb CMCJ arthrodesis and the clinical outcome.

Patients and Methods  Cases with Eaton III thumb CMCJ osteoarthritis treated with arthroscopic arthrodesis were reviewed. Patient evaluations include: grip strength, pinch strength, range of motion, Kapandji score, Disabilities of the Arm, Shoulder, and Hand (DASH) questionnaire, and the visual analog scores for pain. All cases were assessed before the surgery and at 3, 6, 12, and 24 months after surgery. Radiographs were reviewed.

Results  There were total 16 patients with 18 arthrodesis performed. The average age was 62.2 years with M:F ratio of 2:7. The average follow-up time was 57.2 months. There was improvement of pain score as early as at postop 3 months ( p  < 0.001) and continued to improve at 6, 12, and 24 months. There was improvement of grip strength and pinch strength at 12 and 24 months (p < 0.001). The DASH score showed improvement as early as at 3 months ( p  = 0.012). There was a reduction of Kapandji score and interphalangeal joint motion at 3 months postop, but these returned to normal at 6 months. There was no major complication. There was one case of nonunion (5.6%).

Conclusion  Arthroscopic arthrodesis is a feasible treatment option and provides excellent pain relief, restore thumb strength and stability, retain functional thumb mobility, and hence improvement in hand function.

The thumb carpometacarpal joint (CMCJ) osteoarthritis is one of the most common pathologies in the hand. This disease is found in up to 25% of postmenopausal women. ¹ According to Pellegrini, this is the most frequent site in the upper extremity that needs surgery. ² The treatment options have been controversial, and there exists a variety of different treatment options. The most common treatment options include ligament reconstruction and tendon interposition, trapeziectomy, arthrodesis, and arthroplasty. ³

Muller first described arthrodesis of the thumb CMCJ in 1949. ⁴ Traditionally, thumb CMCJ arthrodesis was performed as open surgery and is not without pitfalls and complications. With the advancement in hand arthroscopy, the use of arthroscopic-assisted treatment for thumb CMCJ arthritis has become more popular. However, arthroscopic thumb CMCJ arthrodesis had never been reported before. The aim of our study is to prospectively evaluate the effectiveness of arthroscopic arthrodesis for treating Eaton stage III thumb CMCJ osteoarthritis.

Materials and Methods

This study prospectively recruited cases from January 2015 to November 2020 and the study has been approved by the institute's research ethics committee. All cases with Eaton III thumb CMCJ osteoarthritis treated with arthroscopic arthrodesis were recruited. All cases were defined as Eaton III CMCJ osteoarthritis by radiographs using the standard posteroanterior view, lateral view, and the Robert's view. Patients with preoperative adductive deformity were included. The minimal follow-up time was 24 months. Cases with known inflammatory arthritis or previous open surgery on the same thumb were excluded.

The patients were evaluated objectively with grip strength, pinch strength, range of motion (maximal thumb palmar abduction, maximal thumb radial abduction, metacarpophalangeal joint [MCPJ] motion, and interphalangeal joint [IPJ] motion), and Kapandji score. They were also evaluated subjectively with the Disabilities of the Arm, Shoulder, and Hand (DASH) questionnaire and pain scores (Visual Analogue Scale [VAS]). The physiotherapist and occupational therapist assessed all cases before the surgery and at 3, 6, 12, and 24 months after surgery. Radiographs were taken during each follow-up and at a final follow-up.

The change in pain score, DASH score, grip strength, pinch strength, range of motion, and Kapandji score were calculated with the Wilcoxon rank sum test using SPSS v.26, with p  < 0.05 as statistically significant.

Surgical Techniques

The surgery is performed under general anesthesia or brachial plexus block. A tourniquet is placed over the upper arm for standby. Tourniquet is not necessary for this operation. The dorsoradial (1R) and dorsoulnar (1U) portals are identified before the application of the traction tower. The first CMCJ can be identified by moving the mobile first metacarpal bone over the relatively immobile trapezium, with the 1R portal radial to abductor pollicis longus tendon and 1U portal ulnar to the extensor pollicis brevis tendon. A 19-G needle is inserted into the 1U portal and the location of the joint is confirmed with fluoroscopic guidance. After confirmation of the joint location, traction is applied with a traction tower. Chinese finger traps are used for traction. A transparent semiocclusive dressing (e.g., Tegaderm) adheres around the thumb before application of the Chinese finger traps to prevent slippage. Adhesive tapes are then applied between the hand dorsum and the traction tower to stabilize the position of the hand in a pronated position, so that the dorsum of the thumb (or the nail) is facing the surgeon. With the 19-G needle in the 1U portal, the joint is injected with 1 mL of normal saline to inflate the joint. A transverse 2-mm skin incision just passed through the dermis was made over 1U portal, then the subcutaneous tissue is bluntly dissected, followed by puncture of the joint capsule using a hemostat. This will avoid injury to the cutaneous branches of the radial nerve. A blunted-end trocar and cannula for the 2.7-mm arthroscope is inserted. A 2.7-mm arthroscope is preferred to achieve a better arthroscopic view. Continuous saline irrigation with a 3-L normal saline bag is used. This hydrostatic pressure is adequate to provide hemostasis even when arthroscopy is performed without a tourniquet ( Fig. 1 ).

Fig. 1.

Clinical photo of performing 1st CMCJ arthroscopic fusion.

The 1R portal can be created and a 2.0-mm shaver is inserted for debridement of synovitis. Additionally, a radiofrequency probe can be used for further debridement of the joint until the joint surfaces can be assessed clearly. Any loose bodies should be removed. If the amount of cartilage remaining is less than 30% of the total joint surfaces, fusion should proceed ( Figs. 2 and 3 ). The remaining cartilage is removed and the subchondral bone is burred until the cancellous bone is seen ( Fig. 4 ). The saline irrigation should be stopped to observe for adequate bleeding from the cancellous bone. After the preparation of the joint surfaces, the hand is taken off from the traction tower. Under fluoroscopic guidance, the thumb CMCJ is reduced manually to correct the dorsal and radial subluxation. In cases of preoperative adductive deformity, the joint should be able to reduce to the desired functional position after adequate removal of osteophytes. If the joint cannot be reduced to the desired tip-pinch functional position, more bone removal is necessary. The joint is then compressed manually and transfix with two 1.6-mm Kirschner wires (k-wires). Additional headless compression screws can be added to improve the joint compression. The position of the thumb should be placed with the tip of the thumb resting comfortably over the lateral aspect of the middle phalanx of the index finger, which is the position for key-pinch. After insertion of the k-wire and screw, the wound is closed and a thumb spica plaster cast is applied.

Fig. 2.

Arthroscopic photo showing complete eburnation of cartilage at proximal trapezium.

Fig. 3.

Arthroscopic photo showing nearly complete eburnation of cartilage at 1st metacarpal base.

Fig. 4.

Burring of subchondroal bone using arthroscopic burr.

The k-wires and thumb spica cast are kept for 6 weeks. After removal of k-wires and cast, active and gentle passive mobilization of the thumb is allowed. Strengthening exercise is allowed starting at 12 weeks.

Results

There were total 18 thumbs in 16 patients recruited in this study, 2 patients had bilateral thumb CMCJ arthrodesis done. The average age was 62.2 years with a male-to-female ratio of 2:7. The average follow-up time was 57.2 months (24–90 months). The pretreatment pain score (VAS) was average at 6.42. There was an improvement of pain score at postop 3 months (VAS 2.78, p  < 0.001), 6 months (VAS 1.06, p  < 0.001), 12 months (VAS 0.19, p  < 0.001), and 24 months (VAS 0.11, p  < 0.001) ( Fig. 5 ). The DASH score improved at 3 months from 47.4 to 36.4 with p  = 0.012. The DASH score continued to improve at 6 months ( p  < 0.001), 12 months ( p  < 0.001), and 24 months ( p  < 0.001) ( Fig. 6 ). The change in grip strength and pinch strength were similar ( Figs. 7 and 8 ). In both parameters, there was an initial decrease at 3 months. But there was a statistically significant improvement in key pinch at 6, 12, and 24 months, while there was a statistically significant improvement of grip strength at 12 and 24 months postop. It was also noted that there was a reduction of Kapandji score and IPJ motion at 3 months postop. However, this reduction returned to normal at 6 months and remained static at 12 and 24 months. For the other thumb range of motion, including the maximal thumb palmar abduction, maximal thumb radial abduction, and MCPJ motion, there were no changes throughout the whole period of time.

Fig. 5.

Graph showing improvement of pain score from pre-op and post-op (VAS scare 0-10).

Fig. 6.

Graph showing improvement in DASH Score.

Fig. 7.

Graph showing change in Grip strength (units: kgf).

Fig. 8.

Graph showing change in pinch strength (units: kgf).

There were no major complications such as infection or implant-related complication. There was one case of nonunion, resulted in a nonunion rate of 5.6%. Table 1 shows a summary of the 18 cases. Case examples are shown in Figs. 9A–C and 10A–F . Table 2 shows a summary of the clinical results.

Table 1. Summary of 18 cases.

Case	Age	Gender	Occupation	FU time (mo)	Implant use	Fusion	STT arthritis (radiological)	STT arthritis (symptomatic)
1	58	F	Housewife	90	2 k-wires	Yes	No	No
2	55	F	Factory worker	84	2 k-wires	Yes	No	No
3	63	F	Packaging worker	24	2 k-wires	No	No	No
4	66	F	Housewife	84	2 k-wires	Yes	Yes	No
5	76	F	Housewife	72	2 k-wires	Yes	Yes	No
6	78	F	Housewife	72	2 k-wires	Yes	No	No
7	57	M	Manual worker	75	1 k-wire + 1 screw	Yes	No	No
8	52	M	Manual worker	63	1 k-wire + 1 screw	Yes	No	No
9	56	F	Sales	68	1 k-wire + 1 screw	Yes	No	No
10	69	F	Housewife	42	2 k-wires	Yes	No	No
11	54	F	Clerk	68	1 k-wire + 1 screw	Yes	No	No
12	60	F	Housewife	57	1 k-wire + 1 screw	Yes	No	No
13	61	F	Housewife	60	2 k-wires	Yes	No	No
14	52	M	Manual worker	51	1 k-wire + 1 screw	Yes	No	No
15	73	F	Housewife	39	2 k-wires	Yes	No	No
16	53	F	Health care assistant	24	2 k-wires	Yes	No	No
17	74	M	Retired	36	1 k-wire + 1 screw	Yes	No	No
18	62	F	Post office	24	1 k-wire + 1 screw	Yes	No	No
Average	62.2	14 F:4 M		57.2		5.6%	11.1%	0%

Abbreviations: F, female; FU, follow-up; k-wire, Kirschner wire; M, male; STT, scapho-trapezium-trapezoid.

Note: Case 1 and case 13 belong to same patient.

Case 8 and case 14 belong to same patient.

Fig. 9.

( A ) A 52-year-old manual worker presented with bilatateral 1st CMCJ osteoarthritis (Case 8 and Case 14) ( B ) Left 1st CMCJ arthroscopic fusion was performed and showed good fusion ( C ) Right 1st CMCJ arthroscopic fusion was done 1 year later, x-ray showed good fusion.

Fig. 10.

( A ) and ( B ) A 55-year-old factory worker (case 2). Pre-operative x-ray showed left 1st CMCJ osteoarthritis ( C ) 1st CMCJ arthroscopic fusion was performed and fixed with 2 k-wires ( D ) X-ray at 1 year post-op ( E ) X-ray at 4 years post-op ( F ) X-ray at 7 years post-op, no STTJ arthritis.

Table 2. Summary of the clinical results comparing preoperative, 3 months postop, 6 months postop, 12 months postop, and 24 months postop.

	Preop	3 mo postop	6 mo postop	12 mo postop	24 mo postop
Pain score (VAS)	6.42	2.78	1.06	0.19	0.11
		p < 0.001*	p < 0.001*	p < 0.001*	p < 0.001*
DASH score	47.4	36.4	23.7	16.1	9.2
		p = 0.012*	p < 0.001*	p < 0.001*	p < 0.001*
Grip (kgf)	16.0	10.6	17.4	22.0	22.5
		p = 0.002*	p  = 0.349	p < 0.001*	p < 0.001*
Key pinch (kgf)	3.2	2.5	4.1	5.1	5.3
		p = 0.015*	p  = 0.035*	p < 0.001*	p < 0.001*
Kapandji score	8.9	8.1	8.9	9.4	9.4
		p = 0.012*	p  = 0.850	p  = 0.066	p  = 0.078
Maximal palmar abduction (degree)	45.6	40.8	42.5	45.8	41.6
		p  = 0.109	p  = 0.229	p  = 0.149	p  = 0.482
Maximal radial abduction (degree)	45.8	41.1	43.1	39.7	42.8
		p  = 0.264	p  = 0.461	p  = 0.062	p  = 0.515
MCPJ motion (degree)	48.3	50.8	51.4	50.3	51.9
		p  = 0.281	p  = 0.240	p  = 0.248	p  = 0.148
IPJ motion (degree)	65.2	58.1	63.1	65.0	68.0
		p = 0.033*	p  = 0.291	p  = 0.687	p  = 0.444

Abbreviations: DASH, Disabilities of the Arm, Shoulder, and Hand; IPJ, interphalangeal joint; kgf, kilogram-force; MCPJ, metacarpophalangeal joint; VAS, Visual Analogue Scale.

Note: * = statistically significant

There were two cases noted to have scapho-trapezium-trapezoid joint (STTJ) arthritis on X-ray. These two cases were noted to have STTJ arthritis at 84 months ( Fig. 11 ) and 72 months postop. However, both cases got no clinical symptoms at the STTJ and no further treatment was needed.

Fig. 11.

A 66-year-old housewife with left 1st CMCJ arthroscopic fusion done (Case 4). X-ray taken at 7 years post-op showed radiological STTJ osteoarthritis, but no symptoms clinically.

Discussion

Thumb CMCJ osteoarthritis is a common problem. Previous systemic reviews and meta-analysis have compared the surgical outcomes of different surgical treatments for first CMCJ osteoarthritis. These methods include volar ligament reconstruction, metacarpal osteotomy, trapeziectomy with or without tendon interposition, trapeziectomy with or without ligament reconstruction, arthrodesis, and joint replacement. The studies showed that none of these surgical treatments was superior to another. ^{5 6}

Thumb CMCJ (basal joint or first CMCJ) arthroscopy was first described by Menon ⁷ and Berger ⁸ in 1996 and 1997. Currently, the arthroscopic treatment of thumb CMCJ includes debridement with or without capsular shrinkage and partial trapeziectomy with or without interposition. ⁹ Arthroscopic thumb CMCJ arthrodesis has not been reported in the literature before.

Thumb CMCJ arthrodesis is traditionally done under an open technique with the joint capsule widely opened and fusion is commonly done with plate and screws. Open thumb CMC arthrodesis provides good relief of pain, restores stability, and maintains strength. However, there is a compromise of motion and fine dexterity. From our study, arthroscopic thumb CMCJ arthrodesis had shown to provide excellent pain relief, restore thumb stability and strength, but at the same time preserve good thumb motion and dexterity, as shown by the thumb range of motion and Kapandji score.

Previous studies had also indicated that thumb CMCJ arthrodesis will not affect the thumb motor function because the CMCJ motion can be well compensated for by the motion at adjacent joints: the STTJ and the thumb MCPJ. ¹⁰ It was claimed that 75% of the compensation comes from the first MCPJ, while the other 25% come from the STTJ. ¹¹

In our study, the patients had disease-free STTJ and thumb MCPJ preoperatively. As a result, the normal STTJ and thumb MCPJ can compensate well for thumb motion for most activities of daily living. Moreover, with arthroscopic thumb CMCJ arthrodesis, the surrounding soft tissue plane is preserved and free of scar adhesion. Also, we used a minimal number of implants (2 k-wires or k-wires plus one headless compression screw), which avoided the problems of surrounding soft tissue scar formation and tendon adhesions which might contribute to joint stiffness. Therefore, with minimal soft tissue scarring and the preoperatively normal STTJ and thumb MCPJ, the thumb mobility can be compensated well and remain mobile after the first CMCJ is fused.

The major pitfall of thumb CMCJ arthrodesis reported in the literature is failure of union, and it had been reported to be as high as 50%. In our study, there was only one case of nonunion (5.6%). Another merit of this procedure is that with arthroscopic treatment, the joint capsule is not disturbed. As a result, the surrounding soft tissue and vasculature is preserved and provides a better environment for bone healing and hence joint fusion.

With the use of arthroscopic arthrodesis, a minimum amount of bone is removed and therefore, it avoids the problem of shortening. Moreover, bone graft was not necessary in any of our cases. Another crucial aspect of arthroscopic fusion involves positioning the thumb in the functional key-pinch position, with around 30 degrees in palmar abduction, 20 degrees in radial abduction, and 10 degrees in extension. This can be easily accomplished by placing a rolled piece of gauze at the first web space, allowing the tip of the thumb to rest comfortably on the radial side of the middle phalanx of the index finger. Unlike arthrodesis in other joints like the ankle, which requires precise alignment, achieving fusion in the thumb does not necessitate a rigid alignment. Moreover, complete conformity of the fusion surfaces is not necessary when arthrodesis is performed arthroscopically. This is in contrast to open arthrodesis.

Adjacent joint osteoarthritis is a commonly anticipated problem after thumb CMCJ arthrodesis. This is because the surrounding joints need to compensate for the loss of movement at the CMCJ, thereby increasing the rates of osteoarthritis in these joints. ¹² However, the STTJ arthritis are mostly presented on radiographs without any subjective symptoms, as a result, no further treatment was necessary. ¹³ These radiological and clinical findings are compatible with our study. We had two cases of STTJ arthritis shown on X-ray but none of these cases got subjective symptoms and no additional treatment was needed.

Previous studies have shown that the results of first CMCJ arthrodesis were suboptimal with high complication rates. The complications mentioned include (1) nonunion and (2) reoperation for implant removal. Table 3 shows the complication rates and functional outcome for first CMCJ fusion. Comparing with previous studies, the nonunion rate in our series is relatively low and none of our cases required any reoperation for removal of implant.

Table 3. Summary of previous studies on thumb CMCJ arthrodesis and the complication rates, comparing with our own study.

Study	Year	Number of cases	FU time	Open/Arthroscopic	Nonunion rate	Implant	Reoperation for removal of implant	STT arthritis	STT arthritis need reoperation
Muller 4	1949	8	1–6 y	Open	12.5%	N/A	N/A	N/A	N/A
Leach and Bolton 14	1968	20	3–18 mo	Open	5%	N/A	N/A	N/A	N/A
Mattsson 15	1969	22	1–5.5 y	Open	50%	Wire loop	N/A	4.5%	N/A
Eiken and Carstam 16	1970	24	5 y	Open	12.5%	Staple	N/A	8.3%	N/A
Carroll and Hill 10	1973	39	3–25 y	Open	7.7%	k-wire	N/A	N/A	N/A
Stark et al 17	1977	30	12 y	Open	6.7%	k-wire	N/A	0	N/A
Pardini et al 18	1982	6	N/A	Open	0	Tension band wire	N/A	N/A	N/A
Cavallazzi and Spreafico 19	1986	43	10 y	Open	21%	k-wires	N/A	14%	0
Alberts and Engkvist 20	1989	34	5 y	Open	17.6%	k-wire/cerclage wire	44%	23.5%	N/A
Clough et al 21	1990	14	1 y	Open	50%	Herbert screw	N/A	5.5%	N/A
Karlsson 22	1991	43	35 mo	Open	9.3%	k-wires	N/A	N/A	N/A
Bamberger et al 23	1992	39	4 y	Open	7.7%	Various	N/A	5.13%	N/A
Caputo and Bennett 24	1993	20	19 mo	Open	10%	Staple	0	5%	0
Chamay and Piaget-Morerod 25	1994	32	6 y 7 mo	Open	12.5%	Various	N/A	25%	0
Lisanti et al 26	1997	52	4 y	Open	7.6%	Staple	N/A	0	0
Hartigan et al 27	2001	58	86 mo	Open	16%	Various	3.4%	24%	N/A
Fulton and Stern 28	2001	59	7 y	Open	6.8%	k-wire	0	11.9%	1.7%
Forseth and Stern 29	2003	26	40 mo	Open	8%	Plate and screw	15.4%	N/A	N/A
Rizzo et al 30	2009	126	11.2	Open	13.5%	Various	N/A	31%	1.6%
De Smet et al 31	2010	26	2 y	Open	8%	Plate and screws	15.4%	N/A	N/A
Smeraglia et al 32	2015	83	9.3 y	Open	30%	k-wires	0	25%	0
Abe and Tokunaga 33	2015	18	2 y	Open	11.1%	Locking plate	16.7%	N/A	N/A
Hayashi et al 34	2015	13	2–5 y	Open	0	T-plate	0	15.4%	0
Spekreijse et al 35	2016	11	5.3 y	Open	27%	2.3 mm T-plate	9.1%	9.1%	9.1%
Jiménez-Díaz et al 36	2017	85	60 mo	Open	4.7%	Quadrangular plate	0	0	0
Li et al 37	2019	22	2.5 y	Open	0	Cannulated screws	0	9.1%	0
van Laarhoven et al 38	2019	25	10.8	Open	4%	2.0 mm T-plate	32%	4%	N/A
Our study	2023	18	57.2 mo	Arthroscopic	5.6%	k-wire or screw	0%	11%	0%

Abbreviations: CMCJ, carpometacarpal joint; FU, follow-up; k-wire, Kirschner wire; N/A, not available; STT, scapho-trapezium-trapezoid.

Conclusion

This prospective study is the very first report of thumb CMCJ arthroscopic arthrodesis for stage III thumb CMCJ osteoarthritis. This study confirms that arthroscopic arthrodesis is a feasible treatment option for stage III thumb CMCJ osteoarthritis. It provides excellent pain relief, restores of thumb strength and stability, retains functional thumb mobility, and hence improvement in hand function.

Advantages of arthroscopic arthrodesis include: (1) allow clear assessment of joint before deciding for fusion; (2) small scar and minimal disturbances of surrounding joint capsule and soft tissue; (3) avoid tendon adhesions and tendonitis; and (4) better preservation of blood supply resulting in better union rate.

Future study is needed for longer term results.