Treatment of First Carpometacarpal Osteoarthritis with Arthroscopy: A Systematic Review

Abstract

Background First carpometacarpal (CMC) osteoarthritis or trapeziometacarpal osteoarthritis is a common debilitating hand condition. No one surgical technique has demonstrated superiority in managing this disease.

Purpose This study performed a systematic review of arthroscopic techniques for treating first CMC arthritis to assess the effectiveness of different arthroscopic techniques.

Methods Grip strength, pinch strength, visual analog scale, the Disability of Arm, Shoulder, and Hand (DASH) score, range of motion (ROM), and complications were recorded. Two subgroup analyses were performed, comparing outcomes of (1) trapeziectomy of any type versus debridement alone and (2) trapeziectomy alone versus interposition versus suspension techniques.

Results Preoperative and postoperative scores significantly improved for DASH scores and pain at rest and with activity with variable improvements in ROM. Complications occurred in 13% of cases in publications that reported complications. When comparing studies that utilized techniques with any type of trapeziectomy to debridement alone, only the trapeziectomy subgroup showed significant improvements in pain. When comparing trapeziectomy alone to interposition and suspension techniques, mean DASH scores and pain levels significantly improved in interposition and suspension subgroups.

Conclusions The existing literature describes a predominantly female population with Eaton-Littler stage II and III disease. In the subgroup analysis, arthroscopic techniques involving a trapeziectomy seem to be more effective at lowering pain scores compared to techniques involving debridement alone. Likewise, interposition and suspension techniques may show improved outcomes compared to techniques involving trapeziectomy alone.

Level of evidence This is a Level III study.

Thumb carpometacarpal (CMC) osteoarthritis is one of the most common sites of arthritis in the body, estimated to affect up to 66% of women and 25% of men by 80 years of age. ^{1 2} It is characterized by thumb-based pain, grip and pinch strength weakness, and loss of function. ^{2 3 4} Surgical interventions for thumb CMC arthritis are among the top three most commonly performed procedures by hand surgeons in the United States. ⁵ After failing nonoperative treatment, there are many treatment options, including metacarpal osteotomy, ⁶ arthrodesis, ^{7 8 9} ligament reconstruction tendon interposition (LRTI), ^{10 11 12 13 14 15 16} implant arthroplasty, ^{17 18 19 20 21 22 23 24} suspensionplasty with a tendon, ²⁵ and Tightrope reconstruction. ^{26 27 28 29 30} Many of the most commonly utilized treatments are open procedures involving a trapeziectomy with or without capsular/ligamentous repair and with or without interposition. ³¹ However, the open procedures require capsular disruption, more extensive soft tissue damage, and tendon harvest in the cases of ligament reconstructions or interpositions. ³² Thus, many surgeons prefer to immobilize them longer, and these procedures are often considered some of the most painful procedures performed by hand surgeons in the early postoperative setting. ^{33 34 35 36 37} The more recently developed arthroscopic techniques offer several advantages over their open counterparts. They potentially overcome some of the aforementioned issues by avoiding a capsulotomy and more extensive soft tissue disruption while instead preserving the native anatomy to provide stability and potentially an expedited recovery. ^{32 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52}

Arthroscopy of the small joints of the wrist and hand was first described in the late 1970s. ⁵³ However, its popularity started to rise markedly in 1996 after Menon published the first article on the use of arthroscopy for the first CMC joint. ⁵² The authors demonstrated an arthroscopic hemitrapeziectomy to have similar results to open treatment in patients with predominantly late stage disease. Since then, multiple arthroscopic techniques have been described and studied in small studies consisting of mostly nonrandomized cohort studies. ^{32 38 39 41 42 43 44 45 46 47 48 49 50 51} To date, there has been a lack of large, prospective controlled trials comparing these methods. Additionally, efforts to compare these smaller studies have failed to encompass all published papers, included poorly designed studies, or reported results from a broad range of techniques without focusing on arthroscopy alone. ^{40 54 55}

This paper aims to determine if there is an arthroscopic method that is superior to other described methods and to determine the effectiveness of arthroscopic methods in treating thumb CMC arthritis. We present a systematic review of all published evidence to date on the arthroscopic treatment of different stages of thumb CMC arthritis. We hypothesized that hemitrapeziectomy leads to improved outcomes compared with debridement alone. Secondarily, we hypothesized that interposition would improve outcomes of hemi- or complete trapeziectomy alone.

Methods

Search Strategy and Study Selection

An electronic search was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines on February 12, 2020, in PubMed and Embase ( Fig. 1 ). The initial search yielded a total of 123 studies, with 82 studies remaining after removing duplicates. References of key articles were searched to identify further potential literature. ⁴⁰ Two independent reviewers screened the studies based on title and abstract. Remaining studies underwent full-text examination by the same team of reviewers. Studies that were included reported patient outcomes of arthroscopic surgical treatment of thumb CMC osteoarthritis. Case reports, review articles without respective patient outcomes, articles not in the English language, and articles that did not report a minimum follow-up of 6 months were excluded. Any disputes for inclusion were resolved by a third independent reviewer.

Fig. 1.

Study selection. CMC, carpometacarpal; OA, osteoarthritis.

Three studies were found during manual citation cross-reference and met all inclusion criteria. ^{28 45 50} Search terms were reexamined and no errors were detected. Two studies ^{28 43} that did not report minimum follow-up but deemed to have adequate mean follow-up were included. Two studies published by the same author presented overlapping patient cohorts, ^{42 56} so the paper containing the most recent and larger cohort ⁴² was included alone. That study also included patients with pantrapezial arthritis which was deemed acceptable as the paper found no difference in outcomes between pantrapezial and isolated trapeziometacarpal arthritis. ⁴² Two studies ^{57 58} were removed after full-text analysis. One study ⁵⁸ reported a series of cases performing arthroscopic-assisted arthrodesis. The other study ⁵⁷ reported cases performing a closing wedge osteotomy. These two surgical techniques were deemed too dissimilar to those reported in the rest of the studies.

Data Extraction

Two independent reviewers extracted data regarding study design (surgical technique, follow-up, and level of evidence), participant demographics (sex, age, Eaton–Littler stage, and handedness), and pre- and postoperative outcome measures (grip strength, oppositional pinch strength, appositional pinch strength, pain scores, Disability of Arm, Shoulder, and Hand [DASH] scale, and range of motion [ROM]). Pain scores combined both visual analog scales (VASs) and numeric pain rating scales into a standard 0 to 10 scale. Complications were also noted. If a study did not report the number of individual operations, it was assumed that each patient in that study corresponded with one operation.

Quality Assessment

A quality assessment of the 22 studies included in the final analysis was performed using the Methodological Index for Non-Randomized Studies Criteria. ⁵⁹ Overall, 20 papers were noncomparative, meaning they did not have a control group, while the remaining two studies contained controls. Of the 16 possible points for noncomparative, the 20 papers scored an average of 10.05 points (range: 3–13). The two comparative studies scored 19 and 20 out of 24 possible points. The quality assessment did not play a role in exclusion criteria ( Appendix A ).

Appendix A. Minors quality assessment.

Study	Clearly stated aim	Inclusion of consecutive patients	Prospective collection of data	End points appropriate to aim of study	Unbiased assessment of the study endpoint	Follow-up period appropriate to the aim of the study	Loss to follow-up less than 5%	Prospective calculation of the study		Adequate control group	Contemporary group	Baseline equivalence	Adjequate statistic analysis	16 patients total	Comparative 24 patients total
Adams et al 60	2	2	2	2	0	2	2	0						12
Chu et al 61	2	2	2	2	0	2	2	0						12
Chuang et al 41	1	2	2	2	0	2	2	0						11
Cobb et al 42	2	2	2	2	1	2	2	0		1	2	2	2		20
Culp and Rekant 62	0	2	0	2	0	2	2	0						8
Desmoineaux et al 43	0	0	0	2	0	1	0	0						3
Diaconu et al 44	1	0	0	2	0	2	2	0						7
Edwards and Ramsey 38	2	2	2	2	0	2	1	0						11
Ehrl et al 63	2	0	2	2	1	2	2	0						11
Furia 32	2	2	2	2	0	2	2	2		1	2	2	0		19
Logli et al 39	2	2	2	2	1	2	1	0						12
Lucet et al 48	2	2	2	2	0	2	2	0						12
Menon 52	2	0	0	2	0	2	0	0						6
Özçelik et al 49	2	2	2	2	0	2	1	0						11
Park et al 46	2	2	2	2	0	2	0	0						10
Pegoli et al 47	2	0	2	2	0	2	2	0						10
Rog et al 64	2	2	2	2	0	2	2	1						13
Tsujii et al 51	2	0	2	2	0	2	2	0						10
Yao and Cheah 29	2	2	2	2	0	2	1	0						11
Hofmeister et al 45	2	2	2	2	0	2	1	0						11
Landes, et al 28	2	1	2	0	0	2	2	0						9
Ogawa et al 50	2	1	2	2	0	2	2	0						11

Statistical Analysis

Two separate subgroup analyses were performed. The first analysis compared the results of studies that utilized surgical techniques containing any type of trapeziectomy (complete or partial) to results of studies that performed debridement alone. The second analysis, which contained three subgroups, compared studies with trapeziectomy alone to those that concerned trapeziectomy with interposition or trapeziectomy combined with suspensionplasty. One study compared arthroscopic resection with insertion of an arthroplasty implant to interposition. ⁴² The data from that study were separated and patients were entered into the respective subgroups for analysis.

Statistical analysis was conducted on demographic information, as well as preoperative and postoperative outcome measurements previously described. Means for variables (age at surgery, follow-up, grip strength, oppositional pinch strength, appositional pinch strength, VAS at rest, VAS during activity, and DASH) were pooled and frequency weighted by the number of operations in each study. The differences in preoperative and postoperative frequency-weighted outcome means were compared using two sample two-tailed t -tests with unequal variances with an α significance level of p  < 0.05.

Results

Search Selection

Twenty-two studies (with 23 sets of data) met the inclusion and exclusion criteria and thus were included in the analysis ^{28 29 32 38 39 41 42 43 44 45 46 47 48 49 50 51 52 60 61 62 63 64} to compare postoperative outcomes between the different techniques ( Table 1 ). In the trapeziectomy versus debridement alone subgroup analysis, the trapeziectomy subgroup contained 16 studies, ^{28 29 38 39 41 42 43 45 46 47 48 49 51 52 60 62} and the debridement subgroup contained six studies. ^{32 44 50 61 63 64} In the trapeziectomy alone versus interposition and suspension analysis, the trapeziectomy alone group contained four studies, ^{38 42 45 62} the interposition had eight studies, ^{39 41 42 46 47 48 52 60} and suspension had five studies. ^{28 29 43 49 51}

Table 1. All included studies.

Study	Year	Surgical technique	Patients	Operations	Eaton–Littler stages
Adams 60	2007	Debridement + interposition	17	17	II, III
Chu et al 61	2009	Electrothermal shrinkage of joint capsule and volar ligaments	17	17	I, II
Chuang et al 41	2015	Debridement + partial trapeziectomy + thermal capsulorrhaphy + interposition	23	23	II, III
Cobb et al 42	2015	Debridement + synovectomy + Partial trapeziectomy ± interposition	125	–	–
Culp and Rekant 62	2001	debridement + capsular shrinkage	22	24	–
Desmoineaux et al 43	2012	Partial trapeziectomy + suspensoplasty	49	49	I–IV
Diaconu et al 44	2011	debridement + synovectomy + implant	25	25	–
Edwards and Ramsey 38	2010	Partial trapeziectomy + thermal capsular modification	28	28	III
Ehrl et al 63	2016	Denervation + synovectomy + debridement	42	42	II, III
Furia 32	2010	synovectomy + debridement	23	23	I, II
Logli et al 39	2018	partial trapeziectomy + interposition	30	–	–
Lucet et al 48	2018	Partial trapeziectomy + interposition	20	20	II–IV
Menon 52	1996	Partial trapeziectomy + interposition	31	33	–
Özçelik et al 49	2019	Partial trapeziectomy + suspensoplasty	21	–	II, III
Park et al 46	2012	Partial trapeziectomy + interposition	9	–	–
Pegoli et al 47	2007	Partial trapeziectomy + interposition	16	16	I, II
Rog et al 64	2019	Capsular modification (“biological resurfacing”)	11	–	II, III
Tsujii et al 51	2020	Partial trapeziectomy + suspensionplasty	27	29	II, III
Yao and Cheah 29	2017	Partial trapeziectomy + interposition	5	5	III
Hofmeister et al 45	2009	Synovectomy + partial trapeziectomy	18	–	I–III
Landes et al 28	2016	Partial trapeziectomy + suspensoplasty	136	153	II–IV
Ogawa et al 50	2018	Synovectomy	13	15	II–IV

Study Characteristics

Indications for surgery were based off of patient symptoms, clinical and radiologic findings, and failure of conservative management (splinting, activity modification, and anti-inflammatories). Twenty studies were level-IV nonrandomized cohort studies without controls, while two studies contained comparison groups. ^{32 42} Cobb et al compared patients who underwent arthroscopic resection arthroplasty with interposition to patients who underwent arthroscopic resection arthroplasty alone. ⁴² No significant difference in outcomes was detected between the two groups. Furia compared patients who underwent an arthroscopic synovectomy and debridement to patients who received conservative, nonsurgical management. Outcome measurements in the surgical group were significantly improved compared with the nonoperative management group. ³²

Patient Demographics

Patient demographic information included across all studies is listed in Table 2 (individual study demographic data shown in Appendix B ). There was a total of 708 patients with 733 operations performed. Of them, 76% patients were female, and the mean cohort size was 31 patients. The mean follow-up was 38.7 months (range: 12–91 months). Most studies included patients with Eaton ⁶⁵ stage-2 or -3 disease. Patient demographics in the three subgroups analyzed are included in Table 2 .

Table 2. Combined reported patient demographics.

Measure	Number of patients	Male	Female	Eaton-Littler Stage				Average age (y)	Handedness
				1	2	3	4		Right	Left	Dominant	Nondominant
Reported total	708	154	483	26	145	214	75	59.2	112	80	184	154
% of reported total	–	24.2	75.8	5.7	31.5	46.5	16.3	–	58.3	41.7	54.4	45.6

Appendix B. Individual patient data.

Study	No. of patients	No. of operations	Male	Female	Eaton's stage 1	Eaton's stage 2	Eaton's stage 3	Eaton's stage 4	Average age (y)	Right	Left	Dominant	Nondominant
Adams et al 60	17	17	5	12					61.7	8	9
Chu et al 61	17	17	11	6	11	6	0	0	35.3
Chuang et al 41	23	23	3	20	0	10	13	0	59	21	2	21	2
Cobb et al 42	125		28	97					60			65	60
Culp and Rekant 62	22	24
Desmoineaux et al 43	49	49			2	11	22	14	64
Diaconu et al 44	25	25	5	20					60.5	24	1	11	14
Edwards and Ramsey 38	28	28	8	20	0	0	28	0
Ehrl et al 63	42	42	8	34	0	15	17	0	62.7			20	22
Furia 32	23	23	3	20	5	18	0	0	53.7
Logli et al 39	30		3	27					58.8			13	17
Lucet et al 48	20	20	2	18	0	8	10	2	55	7	13	9	11
Menon 52	31	33	4	27					59.03	11	22
Özçelik et al 49	21		5	16	0	9	12	0	52.3			14	17
Park et al 46	9		5	4					59.4
Pegoli et al 47	16	16	3	13	6	10	0	0	58			12	1
Rog et al 64	11		4	7	0	8	3	0	57
Tsujii et al 51	27	29	5	22	0	11	18	0	66.5
Yao and Cheah 29	5	5	4	1	0	0	5	0	56.2	2	3	3	2
Hofmeister et al 45	18		4	14	2	14	6	0	48
Landes et al 28	136	153	44	92	0	21	70	58	62
Ogawa et al 50	13	15	0	13	0	4	10	1	64.5	9	6	9	4

Clinical Outcomes of All Studies

The outcomes analyzed across all included studies were grip strength, oppositional pinch strength, appositional pinch strength, VAS at rest, VAS during activity, and DASH. Preoperative ^{32 38 39 41 42 48 49 51 61 64} and postoperative ^{32 38 39 41 48 49 51 60 61 64} resting VAS were reported in 10 studies each. Four studies reported preoperative and postoperative active VAS. ^{41 48 50 61} Nine studies reported preoperative DASH scores, ^{32 38 47 48 49 51 63 64 65} and 12 reported postoperative DASH scores. ^{29 32 38 39 43 47 48 49 50 51 63 64} Preoperative grip strength was measured in 12 studies. ^{38 42 43 44 45 47 48 49 50 51 60 64} Postoperative grip strength was reported in 11 studies. ^{38 39 43 44 45 47 48 49 51 60 64} Oppositional pinch strength was measured in 10 studies, both preoperatively ^{32 42 43 45 47 48 49 51 52 60} and postoperatively. ^{32 39 43 45 47 48 49 51 52 60} Appositional pinch strength was recorded in four studies preoperatively ^{38 44 45 51} and five studies postoperatively. ^{38 39 44 45 51}

In the combined analysis of all included studies, overall preoperative and postoperative means were significantly different for resting pain scores, active pain scores, and DASH scores ( Table 3 ). There were significant improvements from preoperative to postoperative weighted averages for the pain at rest, pain with activity, and DASH scores of 5.36, 5.89, and 19.61 points, respectively. However, the sample size for active pain scores was 75 operations or 10.2% of the entire cohort. The differences between preoperative- and postoperative-weighted means were not significant for grip strength nor for oppositional or appositional pinch strength.

Table 3. Combined analysis outcome measures.

Outcome measure	Value	Sample size (% of all operations)	p -Value
Grip strength (kg)
Preoperative	18.36	374 (51.0)	0.098
Postoperative	22.58	264 (36.0)
Oppositional pinch strength (kg)
Preoperative	4.25	351 (47.9)	0.206
Postoperative	5.34	256 (34.9)
Appositional pinch strength
Preoperative	2.76	100 (13.6)	0.407
Postoperative	4.70	130 (17.7)
Pan score at rest
Preoperative	6.64	327 (44.6)	<0.002
Postoperative	1.28	219 (29.8)
Active pain score
Preoperative	7.19	75 (10.2)	<0.002
Postoperative	1.30	75 (10.2)
DASH
Preoperative	42.73	195 (26.6)	<0.02
Postoperative	23.12	289 (39.4)

Abbreviation: DASH, disability of arm, shoulder, and hand score.

ROM was assessed in 14 studies, ^{29 38 41 43 44 45 46 48 49 50 51 52 60 61} though the specific motion measured varied greatly between studies. The most common motion assessed was opposition to the fifth metacarpal which was recorded in seven studies. ^{29 43 44 46 48 49 60} Four studies assessed palmar and/or radial abduction. ^{29 45 50 51} Three studies recorded flexion and extension. ^{41 45 61} Two studies did not specify a particular ROM assessed. ^{38 52} Four studies reported significant improvements in postoperative ROM. ^{41 44 48 61} Of these, two studies assessed flexion and extension. Chuang et al reported significantly improved flexion and extension, and Chu et al reported significantly improved flexion. ^{41 61} In the latter, extension was decreased, though this was not significant. The other two studies assessed opposition. Diaconu et al reported significantly improved abduction and opposition. Lucet et al reported significantly improved opposition alone. Patient satisfaction was assessed in 10 studies via a variety of metrics. ^{32 38 41 42 43 47 48 57 60 62} In these studies, 83.5 to 100% of patients reported that they were at least satisfied with their arthroscopic procedure. ^{38 41 42 43 48 60 62}

Three studies did not report complications. ^{47 50 62} In total, 85 complications were reported out of a possible 638 operations, so out of all studies that tracked complications, 13.3% of operations observed complications. There were 50 cases of persistent nonspecific pain or irritation, composing 58.1% of all reported complications, or an incidence of 8% overall. Superficial branch of the radial nerve irritation or neuroma accounted for 20 cases, composing 23.3% of all complications, or an incidence of 3% overall. All reported complications are categorized in Table 4 . Specific complications varied in severity and ranged from transient surgical site pain to sepsis and surgical failure requiring reoperation.

Table 4. Total combined complications.

Complication	Neuropathy	Nonspecific persistent pain/irritation (including CRPS types I and II)	Infection	Tendinopathy (rupture, tendonitis)	Bleed	Symptomatic hardware
Number	20	50	6	3	1	5
Percent of overall complications	23.3	58.1	7.0	3.5	1.2	5.8

Abbreviation: CRPS, complex regional pain syndrome.

Outcomes of Subgroup Analysis

In the subgroup analysis comparing trapeziectomy to debridement alone, the trapeziectomy subgroup contained 600 operations while the debridement alone group contained 133 ( Table 5 ). Average follow-up was 40.8 months in the trapeziectomy group and 30.3 months in the debridement group. There was a significant improvement in preoperative to postoperative pain scores at rest and with activity in the trapeziectomy group, without such observations in the debridement group ( Table 6 ). The functional outcome measures improved but did not reach statistical significance for the grip and pinch strengths or DASH scores.

Table 5. Subgroup analysis 1 demographics.

Measures	No. of patients	Male	Female	Eaton–Littler stage				Average age (y)
				Stage 1	Stage 2	Stage 3	Stage 4
Trapeziectomy
Reported total	577	123	383	10	94	184	74	59.7
% of reported total	–	24.3	75.7	2.8	26.0	50.8	20.4	–
Debridement alone
Reported total	131	31	100	16	51	30	1	57.0
% of reported total	–	23.7	76.3	16.3	52.0	30.6	1.0	–

Table 6. Subgroup analysis: trapeziectomy versus debridement alone.

	Trapeziectomy (hemi or complete)			No trapeziectomy (debridement alone)
Outcome measure	Value	Sample size (% of all operations)	p -Value	Value	Sample size (% of all operations)	p -Value
Grip Strength (kg)
Preoperative	18.5	53.8	0.225	17.2	38.3	0.434
Postoperative	22.4	38.0		24.0	27.1
Oppositional pinch strength (kg)
Preoperative	4.3	54.7	0.257	-	-	-
Postoperative	5.3	38.8		-	-
Appositional pinch strength (kg)					-
Preoperative	2.5	12.5	0.611	-	-	-
Postoperative	4.3	61.1		-	-
Pan score at rest					-
Preoperative	7.3	16.8	<0.005	6.6	38.3	0.819
Postoperative	1.1	16.8		5.3	38.3
Active pain score
Preoperative	7.1	7.2	<0.007	7.3	24.1	0.077
Postoperative	1.0	7.2		1.7	24.1
DASH
Preoperative	37.1	19.8	0.059	51.3	57.1	0.159
Postoperative	20.0	33.0		26.0	68.4

Abbreviation: DASH, disability of arm, shoulder, and hand score.

The second subgroup analysis compared the preoperative to postoperative outcome measure improvements in the trapeziectomy alone ( n  = 143) to interposition ( n  = 200) and suspensionplasty ( n  = 257) techniques ( Table 7 ). Average follow-up for trapeziectomy alone, interposition, and suspension subgroups was 71.9, 46.7, and 25.8 months, respectively. Although the postoperative outcome measures improved in the trapeziectomy alone group, none reached statistical significance ( Table 8 ). There was a significant improvement in preoperative to postoperative pain scores with activity and DASH scores in the interposition group, while there was a similarly significant improvement in pain scores at rest in the suspensionplasty group.

Table 7. Subgroup analysis 2 demographics.

Measures	No. of patients	Male	Female	Eaton–Littler stage				Average age (y)
				Stage 1	Stage 2	Stage 3	Stage 4
Trapeziectomy alone
Reported total	166	33	111	2	14	34	0	56.7
% of reported total	–	22.9	77.1	4.0	28.0	68.0	0	–
interposition
Reported total	198	37	161	6	28	23	2	59.0
% of reported total	–	18.7	81.3	10.2	47.5	39.0	3.4	–
Suspension
Reported total	238	58	131	2	52	127	72	62.0
% of reported total	–	30.7	69.3	0.8	20.6	50.2	28.5	–

Table 8. Subgroup analysis: trapeziectomy alone versus interposition versus suspension.

	Trapeziectomy alone			Interposition			Suspension
Outcome measure	Value	Sample size (% of all operations)	p -Value	Value	Sample size (% of all operations)	p -Value	Value	Sample size (% of all operations)	p -Value
Grip strength (kg)
Preoperative	18.5	83.2	0.325	19.9	52.5	0.720	17.2	38.5	0.271
Postoperative	20.3	32.2		23.4	41.5		22.5	38.5
Oppositional pinch strength (kg)
Preoperative	4.8	63.6	0.854	3.8	69.0	0.469	4.3	38.5	0.479
Postoperative	5.9	63.6		4.8	58.0		5.8	38.5
Appositional pinch strength (kg)
Preoperative	1.6	32.2	0.405	–	–	–	–	–	–
Postoperative	2.5	32.2		–	–		–	–
Pan score at rest
Preoperative	6.6	70.6	0.210	6.3	62.5	0.085	7.6	19.5	<0.03
Postoperative	1.0	70.6		0.8	45		1.2	19.5
Active pain score
Preoperative	–	–	–	6.3	21.5	<0.02	–	–	–
Postoperative	–	–		1.0	21.5		–	–
DASH	–	–
Preoperative	–	–	–	54.6	18	<0.004	39.5	21.4	0.220
Postoperative	–	–		16.2	33		14.5	40.5

Abbreviation: DASH, disability of arm, shoulder, and hand score.

Discussion

Thumb CMC arthritis is one of the most common arthritic conditions of the hand, ⁶⁶ occurring in more than 80% of people by 70 years of age. ⁶⁷ These numbers will likely increase as our population continues to live longer and remain active at older ages. ⁶⁸ Furthermore, the numbers seeking treatment will continue to rise, as currently more than 20% of people with CMC arthritis are evaluated and treated by hand specialists, with many eventually requiring surgery. ^{68 69} Surgical interventions involving one of the many described techniques of open CMC arthroplasty has been demonstrated to involve higher pain levels and require more postoperative opioids than most other commonly performed procedures in hand surgery. ^{33 34 35 36 37} The reason for this is likely multifactorial, but likely, in part, involves the capsular and soft tissue disruption associated with the open techniques. Thumb CMC arthroscopic hemi- or complete trapeziectomy theoretically overcomes some of these issues, preserving the native static stabilizers of the CMC joint, including the capsule, dorsoradial, and volar beak ligaments. ^{70 71 72 73 74} Although there have been many case series on CMC arthroscopic outcomes, ^{32 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52} there remains a paucity of large or comparative studies. Therefore, this study assessed the overall outcomes of thumb CMC arthroscopic interventions utilized to treat CMC arthritis, as well as compared the need to perform a soft tissue interposition or suspension after hemitrapeziectomy.

Overall, in 22 studies involving 733 arthroscopic interventions for predominantly Eaton stage-2 and -3 CMC arthritis, patients experienced improvements in their preoperative to postoperative pain and functional outcome scores. Complications occurred in 13% of cases that reported complications, with nonspecific pain or irritation being the most common complication (8%). In the trapeziectomy versus debridement subgroup analysis, there were improvements from preoperative to postoperative for all pain and functional outcomes, but the only values that reached significance occurred in the pain at rest and with activity in the trapeziectomy subgroup. In the subgroup analysis comparing trapeziectomy alone to interposition and suspension techniques, the DASH scores and pain with activity significantly improved in the interposition subgroup, and pain scores at rest significantly improved in the suspension subgroup.

In the surgical treatment of patients with Eaton ⁶⁵ stage-III CMC arthritis, surveys of hand surgeons performed by Brunton and Wilgis in 2010 ⁷⁵ and Wolf and Delaronde in 2012 ^{31 75} found open trapeziectomy with LRTI to be the most commonly preferred procedure. ^{31 75} However, the use of arthroscopy has been on the rise since it was first described by Menon in 1996, ⁵² with up to 10% of hand surgeons stating that it is their preferred method for treating young patients with mild disease. ³¹ A more recent study by Deutch et al in 2018 suggested that 50% of physicians have changed their preference toward arthroscopy in the past 5 years, in large part due to their opinions that evidence has shown these techniques to have the best outcomes with the lowest complication rates. ⁷⁶ However, these opinions seem to be based more on training and experience rather than peer-reviewed evidence, as currently there remains a lack of comparative studies examining the open to arthroscopic approach. ⁷⁷ With this evolving interest and utilization of CMC arthroscopy, particularly in young or active patients, it has become important to understand the indications and technical considerations for this novel approach to CMC arthritis.

The arthroscopic technique has multiple inherent potential advantages, including little disruption of the soft tissues ⁷⁸ and potentially less pain and stability with the need for a ligament reconstruction or suspensionplasty. Furthermore, in young patients, it provides more options for future salvage procedures if necessary. Additionally, given that it does not rely on soft tissue healing like the open procedures, patients can start moving earlier in the postoperative period. Although there have not been any comparative studies, the early outcomes have been promising with this minimally invasive technique. ^{32 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52} The original description by Menon involved an arthroscopic partial trapeziectomy with interposition in 33 patients, reporting 88% pain relief, maintenance of thumb motion, and an improvement of pinch strength from 6 to 11.1 lbs. ⁵² A year later, Berger demonstrated that arthroscopic debridement alone in seven patients achieved similarly promising results. ⁷⁹ Other more recent studies have reinforced these initial promising results with ^{39 41 42 43 44 45 46 47} or without ^{32 38 50 51 52} interpositions performed. In our systematic review of 733 patients treated for CMC arthritis arthroscopically, we found similar improvements in pain, thumb, and hand functional scores from preoperative to postoperative time points.

When considering the type of arthroscopic intervention, there is no unified algorithm to help direct the surgeon or the patient toward the optimal intervention. Arthroscopic debridement alone has the advantage of no bony procedure and focusing on the soft tissue inflammation and loose bodies generated by the arthritis, possibly enabling a faster and less painful initial recovery. However, this does not help on the actual arthritic joint, so seems to be logically limited to much milder disease states. Partial trapiezectomy helps to remove half of the arthritic joint, allowing either hematoma or other tissue to now function as the new joint surface. As seen in this systematic review, promising outcomes in stage-2 and -3 arthritis can be achieved with this technique. ^{32 38 50 51 52} Alternatively, partial trapeziectomy, combined with interposition of either autologous or allogeneic graft, replaces the arthritic joint with a soft tissue articulation, enabling the thumb to move without bone-on-bone contact. As we found, many case series have shown promising pain and functional outcomes of the thumb and hand with this technique. ^{39 41 42 46 47 48 52 60} Finally, another alternative to interposing tissue into the resected articulation is to suspend the first metacarpal to the second metacarpal using either a graft or synthetic material. This helps to preserve the metacarpal and ultimate height, and prevents subsidence. Similar to the interposition procedures, the case series on trapeziectomy and suspensionplasty have also shown promising improvements in pain and function. ^{28 29 43 49 51} Unfortunately, due to the low quality of prospective comparison studies, it is not possible to determine which arthroscopic technique is superior to the others in producing better pain and functional outcomes.

Another consideration is the use of arthroscopic techniques in place of the traditional open CMC arthroplasty techniques. Although our review of 22 papers involving 733 arthroscopies demonstrated that the arthroscopic intervention is successful in reducing pain and functional outcomes, it is not known how this compares to open techniques. Rog et al compared 11 patients who had undergone arthroscopic debridement with biological resurfacing to 15 patients who underwent trapeziectomy with LRTI. Compared with the arthroscopic group, the LRTI group had statistically significant better outcomes in VAS scores, DASH scores, and grip strength, while 8 of the 11 patients in the arthroscopic group required further surgery. ⁶⁴ The arthroscopic group in this study included only debridement, without trapeziectomy alone or in combination with interposition or suspensionplasty techniques. As our systematic review demonstrates, there are a lot of considerations when deciding the optimal arthroscopic intervention to compare with any open procedure. Additionally, as has been the issue with other systematic reviews on this topic, the sample sizes of all of these studies have been very small, with the average being approximately 32 patients. Thus, larger, prospective, and even randomized controlled trials are really needed to determine the superiority of one technique over another.

Limitations

The main limitation of this study is the lack of high-quality evidence currently reported in the literature. The majority of the included studies are level-4 case series, and there exists a lack of prospective trials directly comparing surgical techniques. Additionally, many of the case series are small in size, each reporting outcomes of a relatively low number of cases. The lack of top-quality evidence detracts from the external validity of the findings presented. Another limitation to the study is the heterogeneity of the dataset. This is due to a lack of consistent outcome reporting, as well as incomplete follow-up in some case series. Future research will have difficulty accurately comparing these arthroscopic techniques without more standardized outcome measures.

Conclusion

Arthritis of the thumb CMC joint is a common condition that can confer significant disability. There are presently several viable surgical options for treating this disease. Our present systematic review of arthroscopic techniques has demonstrated that partial or total trapeziectomy with soft tissue insertion significantly improved VAS scores providing adequate pain relief and thus is a viable option for managing this disease. Further studies with larger patient cohorts and randomized control trials comparing treatment modalities are needed to determine the superiority of one technique over another.

Funding Statement

Funding None.