- Email: [email protected]
Five- to 18-Year Follow-Up for Treatment of Trapeziometacarpal Osteoarthritis: A Prospective Comparison of Excision, Tendon Interposition, and Ligament Reconstruction and Tendon Interposition
SCIENTIFIC ARTICLE
Five- to 18-Year Follow-Up for Treatment of Trapeziometacarpal Osteoarthritis: A Prospective Comparison of Excision, Tendon Interposition, and Ligament Reconstruction and Tendon Interposition Soham Gangopadhyay, MSc, Helen McKenna, Dip COT, Frank D. Burke, MBBS, Tim R. C. Davis, ChM
Purpose To investigate whether palmaris longus interposition or flexor carpi radialis ligament reconstruction and tendon interposition improve the outcome of trapezial excision for the treatment of basal joint arthritis after a minimum follow-up of 5 years. Methods We randomized 174 thumbs with trapeziometacarpal osteoarthritis into 3 groups to undergo simple trapeziectomy, trapeziectomy with palmaris longus interposition, or trapeziectomy with ligament reconstruction and tendon interposition using 50% of the flexor carpi radialis tendon. A K-wire was passed across the trapezial void and retained for 4 weeks, and a thumb spica was used for 6 weeks in all 3 groups. We reviewed 153 thumbs after a minimum of 5 years (median, 6 y; range, 5–18 y) after surgery with subjective and objective assessments of thumb pain, function, and strength. Results There was no difference in the pain relief achieved in the 3 treatment groups, with good results in 120 (78%) patients. Grip strength and key and tip pinch strengths did not differ among the 3 groups and range of movement of the thumb was similar. Few complications persisted after 5 years, and these were distributed evenly among the 3 groups. Compared with the results at 1 year in the same group of patients, the good pain relief achieved was maintained in the longer term, irrespective of the type of surgery. While improvements in grip strength achieved at 1 year after surgery were preserved, the key and tip pinch strengths deteriorated with time, but the type of surgery did not influence this. Conclusions The outcomes of these 3 variations of trapeziectomy were similar after a minimum follow-up of 5 years. There appears to be no benefit to tendon interposition or ligament reconstruction in the longer term. (J Hand Surg 2012;37A:411–417. Copyright © 2012 by the American Society for Surgery of the Hand. All rights reserved.) Type of study/level of evidence Therapeutic I. Key words Ligament reconstruction, osteoarthritis, thumb, trapeziectomy, trapezium.
From the Department of Trauma and Orthopaedics, Nottingham University Hospitals, Queen’s Medical Centre, Nottingham; and the Pulvertaft Hand Centre, Royal Derby Hospital, Derby, UK. Received for publication August 3, 2011; accepted in revised form November 29, 2011. The authors are grateful to Apostolos Fakis, Medical Statistician, Department of Research and Development,RoyalDerbyHospital,UnitedKingdom,forinvaluablecontributiontothestatisticalanalyses of the data presented.
No benefits in any form have been received or will be received related directly or indirectly to the subject of this article. Corresponding author: Soham Gangopadhyay, MSc, Department of Trauma and Orthopaedics, Nottingham University Hospitals, Queen’s Medical Centre, Derby Road, Nottingham NG7 2UH, UK; e-mail: [email protected]. 0363-5023/12/37A03-0001$36.00/0 doi:10.1016/j.jhsa.2011.11.027
© ASSH 䉬 Published by Elsevier, Inc. All rights reserved. 䉬 411
412
O
TRAPEZIECTOMY RCT 5- TO 18-YEAR RESULTS
STEOARTHRITIS AT THE TRAPEZIOMETACARPAL
joint is common in postmenopausal women1 and its surgical treatment has evolved over the years. In 1949, Gervis2 reported simple excision of the trapezium, a procedure that is reported to produce satisfactory results.3–5 However, there have been concerns that this procedure may result in shortening of the thumb, owing to proximal migration of the metacarpal, and lead to weakness. Subluxation and osteoarthritis at the pseudarthrosis between the scaphoid and the thumb metacarpal are other concerns.6 To overcome these potential problems, various modifications have been proposed. Placement of a rolled-up length of tendon in the trapezial void7–12 has been used to act as a biological spacer, whereas creation of a ligament between the thumb and index metacarpal bases13,14 has been proposed to prevent instability at the pseudarthrosis. Excision of the trapezium with ligament reconstruction with or without tendon interposition is a commonly performed procedure for trapeziometacarpal osteoarthritis and has good results.14 –19 However, there may be no advantage to either tendon interposition or ligament reconstruction over simple trapeziectomy.19 –24 Most studies that show no benefit to either adjunct have relatively short average follow-up times ranging between 1 and 4 years, and it may be that with longer follow-up the theoretical advantages of tendon interposition or ligament reconstruction may become clinically relevant. The primary purpose of this prospective randomized study was to investigate the outcome and compare the results of trapeziectomy alone (T), trapeziectomy with palmaris longus (PL) tendon interposition (T⫹PL), and trapeziectomy with ligament reconstruction and tendon interposition (LRTI) using half of the flexor carpi radialis (FCR) tendon (T⫹LRTI) in 153 thumbs after a minimum follow-up of 5 years (median, 6 y; range, 5–18 y). Because the same cohort of patients had also been assessed 1 year after surgery,24 we were able to compare the results at 1 year with those after a minimum follow-up of 5 years. MATERIALS AND METHODS The local ethics committees approved this study. Women with painful trapeziometacarpal osteoarthritis who had failed to respond to nonoperative treatment were recruited between 1992 and 2001. We excluded men from this study because they experience this condition much less frequently, and we wanted to have as homogeneous a group as possible. All had Eaton and Littler13 grade 2 through 4 osteoarthritis. A total of 153 women had 174 procedures; 21 had surgery on both thumbs at different times during the recruitment period.
Of the 174 operated thumbs, 153 (88% of the original cohort) were assessed at a median of 6 years (range, 5–18 y). Three patients had moved and could not be traced, 5 patients had died, and 13 patients did not want to be reviewed. Although it was our intention to assess all patients at 5 years, this could not be achieved, and we reviewed 50 thumbs at 5 years, 81 at 6 to 9 years, and 22 at 10 to 18 years after surgery. We randomized patients into 3 groups to undergo T, T⫹PL, or T⫹LRTI. We achieved randomization at the induction of anesthesia by opening the next sequentially numbered sealed, opaque envelope that contained instructions as to which operation should be performed. It was stratified such that each set of 9 consecutive surgeries would include (in random order) 3 of each of the 3 surgical options. Table 1 shows that the 3 groups were well matched before surgery and outlines the operative details and additional procedures performed. We excised the trapezium piecemeal using a dorsal approach. When used, the PL was rolled up and sutured to itself before placement in the trapezial void. We performed the FCR ligament reconstruction and tendon interposition as described by Burton and Pellegrini14 using 50% of the tendon, with 2 exceptions: the trapezium was completely excised in all cases irrespective of the extent of pantrapezial involvement,17 and in the cases done after 1994, the base of the first metacarpal was not resected. Instead, we made a drill hole for passage of the tendon from the dorsal aspect of the thumb metacarpal base to exit the palmar ulnar edge of its articular surface adjacent to the index metacarpal. The tendon was passed through this drill hole from its insertion on the index metacarpal base to exit the thumb metacarpal base dorsally, where it was sutured to the periosteum and joint capsule. The remaining length was then rolled up to act as a spacer. During each of the 3 procedures, we inserted a percutaneous K-wire through the base of the thumb metacarpal and passed it longitudinally across the trapezial void into the distal scaphoid to hold the metacarpal base at the level of the trapezoid–index carpometacarpal joint. We performed additional procedures as needed (Table 1). In most instances, when metacarpophalangeal (MCP) hyperextension was greater than 30°, we stabilized it with a K-wire for 4 weeks; when greater than 40°, we performed palmar capsulodesis or MCP arthrodesis.25 The K-wire across the trapezial void was cut outside the skin and removed at 4 weeks. We used a plaster of Paris cast that maintained the thumb in abduction with the wrist in neutral for a total of 6 weeks. At that stage,
JHS 䉬 Vol A, March
TRAPEZIECTOMY RCT 5- TO 18-YEAR RESULTS
TABLE 1. Patient Details, Additional Operations Performed, and Tourniquet Times in the 3 Treatment Groups for the 153 Procedures With a Minimum of 5 Years of Follow-Up
413
TABLE 2. Pain Levels (Median [Interquartile Range]) Before Surgery and After a Minimum Follow-Up of 5 Years in the 3 Groups of Patients Preoperative
Final
T
T⫹PL
T⫹LRTI
T (n ⫽ 53)
5 (4–6)
0 (0–1)
Thumbs
53
46
54
T⫹PL (n ⫽ 46)
5 (4–6)
1 (0–2)
Dominant
30
22
22
T⫹LRTI (n ⫽ 54)
5 (4–6)
0 (0–1)
.407
.383
Median age (range) (y)
57 (44–74) 57 (40–75) 57 (44–75)
P value
Status Housewife/retired
33
30
33
Office worker
5
1
6
Fine skill
0
0
3
Light manual
15
Mean trapeziometacarpal 3.5 (2–4) osteoarthritis grade (range) Scaphotrapezial joint osteoarthritis (n)
15 3.6 (2–4)
12 3.7 (2–4)
6
5
5
Consultant
27
33
36
Supervised trainee
26
13
18
Surgeon
Treatment center Derby
11
5
11
Nottingham
42
41
43
10
17
12
3
3
5
Additional procedures Carpal tunnel decompression MCP K-wire MCP capsulodesis
0
3
6
MCP arthrodesis
1
1
2
de Quervain release
2
1
1
Trigger thumb release
2
0
2
0
1
0
Trigger finger release Median tourniquet time (interquartile range) (min)a
We assessed the level of pain at every stage using the following scoring system: 0 ⫽ no pain; 1 ⫽ pain only with and after use, no restrictions; 2 ⫽ pain with and after use, some restrictions; 3 ⫽ rest pain without activity restriction; 4 ⫽ rest pain, some restrictions; 5 ⫽ rest pain, severe restrictions; and 6 ⫽ rest pain with night wakening.
43 (36–50) 50 (46–61) 63 (55–75)
a
Tourniquet time was the only variable among the data presented in this table that was significantly different among the 3 groups (P ⬍ .001).
the patients were shown a set of exercises to mobilize and strengthen the thumb. We assessed patients after a minimum of 5 years with the same subjective and objective measures that had been used preoperatively and at 1 year. Two independent observers who were not involved in the surgical procedure carried out all assessments at final followup. Although the observers could not be truly kept unaware of the study protocol owing to the presence of
forearm scars from tendon harvest, no attempt was made to identify these scars before assessing the thumb. We invited patients to attend the outpatient clinics; for those unable to attend, we organized home visits. We made subjective assessments of pain and restriction of activity. We performed objective measurements of grip strength and key and tip pinch strengths with calibrated grip and pinch meters (Jamar, Jackson, MO) using standardized techniques.26 We assessed thumb opposition27 and measured MCP hyperextension. Postoperative complications persisting after 5 years were recorded. We used the Kolmogorov-Smirnov test and histograms to assess the distribution of the variables. None of the variables were normally distributed; as such, we used nonparametric tests for the analyses. We compared the continuous variables among the 3 surgical groups using the Kruskal-Wallis test and between 2 groups using the Mann-Whitney U test. We compared categorical variables at baseline using the chi-square test. Results are presented as median values with the interquartile ranges and were considered to be significant at a P value of ⬍ .050. RESULTS Table 2 shows the preoperative and final pain levels in the 3 groups. We assessed pain on a scale of 0 to 6 as defined in Table 2. There was no difference among the 3 groups with regard to pain relief, and most patients had no pain or mild discomfort with use that did not restrict activity. At the final assessment, 14 patients (7T⫹LRTI, 4T⫹PL, and 3T) had rest pain with or without activity restriction, although all were better than before surgery. Of these, 4 had undergone revision operations. Two patients (T⫹LRTI) had been com-
JHS 䉬 Vol A, March
TRAPEZIECTOMY RCT 5- TO 18-YEAR RESULTS
414
TABLE 3. Key Pinch (Median [Interquartile Range]) (kg) of Affected/Operated Thumb and Contralateral Normal Thumb in the 3 Groups Before Surgery, at 1 Year and at Final Review After a Minimum of 5 Years Thumb Assessed Preoperative
Affected Contralateral
1y Final
a
a
Operated
T
T⫹PL
T⫹LRTI
P Value
3.6 (2.3–4.5) (n ⫽ 53)
3.5 (2.2–5.5) (n ⫽ 46)
3.2 (2.0–4.1) (n ⫽ 54)
.415
5.5 (3.6–6.4) (n ⫽ 47)
4.5 (2.7–6.8) (n ⫽ 44)
5.3 (3.5–6.8) (n ⫽ 49)
.967
4.5 (3.6–5.9) (n ⫽ 53)
4.3 (2.7–5.9) (n ⫽ 46)
4.5 (3.2–5.5) (n ⫽ 54)
.388
Contralaterala
6.3 (4.5–7.2) (n ⫽ 36)
5.0 (3.6–6.8) (n ⫽ 35)
5.5 (4.1–6.4) (n ⫽ 35)
.236
Operated
4.1 (2.7–5.0) (n ⫽ 53)
3.4 (1.8–5.5) (n ⫽ 46)
3.6 (2.7–5.0) (n ⫽ 54)
.471
Contralaterala
4.1 (2.7–5.2) (n ⫽ 37)
2.7 (1.8–5.0) (n ⫽ 35)
3.6 (2.7–5.0) (n ⫽ 37)
.200
We obtained contralateral strength data from patients who had not had surgery and did not have painful osteoarthritis of the contralateral thumb.
TABLE 4. Tip Pinch (Median [Interquartile Range]) (kg) of Affected/Operated Thumb and Contralateral Normal Thumb in the 3 Groups Before Surgery, at 1 Year, and at Final Review After a Minimum of 5 Years Thumb Assessed Preoperative
Affected Contralateral
1y Final
a
a
Operated
T
T⫹PL
T⫹LRTI
P Value
2.3 (1.4–3.2) (n ⫽ 53)
2.3 (1.4–3.6) (n ⫽ 46)
2.3 (1.4–2.7) (n ⫽ 54)
.541
3.2 (1.6–4.1) (n ⫽ 47)
3.2 (0.2–4.1) (n ⫽ 44)
3.2 (1.7–4.1) (n ⫽ 49)
.990
3.2 (2.3–4.1) (n ⫽ 53)
3.1 (1.8–4.1) (n ⫽ 46)
3.6 (2.7–4.1) (n ⫽ 54)
.544
Contralaterala
4.1 (3.2–4.5) (n ⫽ 36)
3.6 (2.3–4.5) (n ⫽ 35)
3.6 (3.2–4.5) (n ⫽ 35)
.164
Operated
2.7 (1.8–3.2) (n ⫽ 53)
2.5 (1.4–3.6) (n ⫽ 46)
2.7 (1.8–3.3) (n ⫽ 54)
.681
Contralaterala
2.7 (1.8–3.6) (n ⫽ 37)
1.8 (0.9–3.2) (n ⫽ 35)
2.7 (1.9–3.2) (n ⫽ 37)
.289
We obtained contralateral strength data from patients who had not had surgery and did not have painful osteoarthritis of the contralateral thumb.
pletely pain free for 12 years before deteriorating. Both were revised to metacarpal hemiarthroplasties, which improved their pain. One (T) had failed early with proximal subluxation of the metacarpal base and was revised with additional LRTI at 1 year with no benefit. The fourth patient (T) had a superficial radial nerve neuroma that was excised early; she subsequently underwent revision to a metacarpal hemiarthroplasty at 4 years, but neither procedure improved the pain. Two patients with continuing pain were being treated for fibromyalgia, and 2 others had severe osteoarthritis affecting the hands. No specific cause for the continuing pain could be identified in the remaining 6 patients. Ability to perform activities such as writing, turning a key, opening a screw-top jar, handling coins, and knitting did not differ between groups at any stage. Tables 3 to 5 show the key pinch, tip pinch, and grip strengths of the operated and the contralateral hands before surgery, at 1 year, and after a minimum of 5 years of follow-up. There was no significant difference
in these strengths among the 3 treatment groups before surgery, at 1 year, or at final follow-up. All 3 types of surgery improved the key and tip pinch strengths. Before surgery, the contralateral normal thumb was considerably stronger than the affected thumb, but by 5 years, the key and tip pinch strengths were similar in both thumbs. Similarly, the grip strengths at final review were much better than the preoperative values and in fact were slightly better than the contralateral hand at this stage. There was no difference in opposition between groups, and at final follow-up, 137 of the 153 thumbs could reach the proximal interphalangeal joint skin crease of the little finger or beyond (P ⫽ .870). Metacarpophalangeal hyperextension had a median value of 10° (interquartile range, 0° to 20°) before surgery (P ⫽ .341) and 5° (interquartile range, 0° to 15°) at final review (P ⫽ .411), with no difference among the 3 groups. Table 6 shows the complications at 1 year and after 5 years. The incidence of complications persisting in the longer term was similar in all 3 groups.
JHS 䉬 Vol A, March
TRAPEZIECTOMY RCT 5- TO 18-YEAR RESULTS
415
TABLE 5. Grip Strength (Median [Interquartile Range]) (kg) of Affected/Operated Hand and Contralateral Normal Hand in the 3 Groups Before Surgery, at 1 Year, and at Final Review After a Minimum of 5 Years Thumb Assessed Preoperative
Affected Contralateral
1y
Operated
Final
a
a
T
T⫹PL
T⫹LRTI
P Value
14 (11–18) (n ⫽ 53)
13 (7–20) (n ⫽ 46)
13 (9–17) (n ⫽ 54)
.624
20 (14–24) (n ⫽ 47)
18 (13–26) (n ⫽ 44)
20 (15–25) (n ⫽ 49)
.770
19 (14–26) (n ⫽ 53)
20 (11–24) (n ⫽ 46)
22 (16–26) (n ⫽ 54)
.423
Contralaterala
23 (16–28) (n ⫽ 36)
20 (15–26) (n ⫽ 35)
22 (18–28) (n ⫽ 35)
.329
Operated
20 (14–25) (n ⫽ 53)
18 (8–26) (n ⫽ 46)
20 (12–24) (n ⫽ 54)
.908
Contralaterala
16 (12–20) (n ⫽ 37)
14 (8–20) (n ⫽ 35)
18 (10–22) (n ⫽ 37)
.471
We obtained contralateral strength data from patients who had not had surgery and did not have painful osteoarthritis of the contralateral thumb.
TABLE 6. Complications of Surgery at 1 Year and After a Minimum Follow-Up of 5 Years in the 3 Groups of Patients T (n ⫽ 53)
T⫹PL (n ⫽ 46)
T⫹LRTI (n ⫽ 54)
1y
5
4
1
5y
2
2
1
1y
2
3
3
5y
2
2
3
1y
0
5
3
5y
0
1
2
1y
0
1
3
5y
0
0
2
1y
1
0
1
5y
0
0
0
Complications Superficial radial nerve dysfunction
Palmar cutaneous branch of median nerve dysfunction
FCR/pollicis longus pulling sensation
Tender scar
Complex regional pain syndrome
DISCUSSION This study showed that after a minimum follow-up of 5 years, the outcome after excision of the trapezium is not improved by either tendon interposition alone or LRTI. The individual treatment options compared in this study are all known to produce good results in case series,3–5,8,9,11,12,16 –18 including studies with relatively longer follow-up.3–5,12,17 However, the superiority of
any 1 procedure over another has never been proven. Existing prospective randomized studies that have compared trapeziectomy alone with other treatment options have failed to establish a difference.20 –24 A recent update of the Cochrane systematic review28,29 has concluded that there is no established superiority of 1 procedure over another and has recommended trapeziectomy alone as the safest option with good outcomes. The patients in this study had been reviewed at 1 year after surgery24; as such, we were able to compare the outcomes in the same patients to see whether the results changed with time. The excellent pain relief achieved at 1 year was maintained in the longer term, with no difference among the 3 groups. At 1 year, 81% of the thumbs were either pain free or had pain only with use without activity restriction. At final follow-up, this level of pain relief was maintained in 78%. In 10 patients, the pain relief achieved at 1 year had deteriorated. Two of these had been revised with relative improvement in pain, 3 had osteoarthritis of the hand, and 1 had cervical spondylosis. For the other 4 patients, we could find no explanation; specifically, they did not have painful instability of the pseudoarthrosis. Although 7 thumbs had rest pain with severe activity restriction or night pain at 1 year, no thumb had this level of pain after 5 years. Although the thumb key and tip pinch strengths were not significantly different between groups after 5 years, both were weaker than at 1 year. The contralateral thumbs had also become weaker, and it is probable that the deterioration in thumb strength is age related and not a function of the surgical procedure. The hand grip strength did not deteriorate with time. These findings differ from those of another long-term report on 25 patients with T⫹LRTI,17 where the grip and thumb tip pinch strengths steadily improved up to 6 years before
JHS 䉬 Vol A, March
416
TRAPEZIECTOMY RCT 5- TO 18-YEAR RESULTS
leveling off, whereas the key pinch strength reached a maximum at 6 years and then showed a slight decline. The Cochrane reviews28,29 reported that T⫹LRTI has the highest complication rate. Nevertheless, our study found no difference in complication rates among the 3 surgical procedures. Moreover, there were fewer ongoing complications after 5 years than at the 1-year review in all 3 groups. Radial nerve dysfunction and pulling sensation in the forearm had resolved in 50%, and chronic regional pain syndrome had resolved in both patients by 5 years, compared with the 1-year results. In contrast, numbness in the distribution of the palmar cutaneous branch of the median nerve persisted after 5 years in most. Four patients in this series underwent revision operations. The metacarpal hemiarthoplasties were performed using a pyrocarbon implant inserted into the metacarpal base, articulating with a socket created in the distal scaphoid using a burr. This study can be criticized for various reasons. We were unable to review all patients at 5 years as initially intended, and the results represent the findings at a median time of 6 years. Fifty thumbs (20 T, 14 T⫹PL, and 16 T⫹LRTI) were seen at 5 years, 81 (29 T, 26 T⫹PL, and 26 T⫹LRTI) at 6 to 9 years, and 22 (4 T, 6 T⫹PL, and 12 T⫹LRTI) at 10 years or more after surgery. However, we reviewed most patients between 5 and 9 years, and the 3 treatment groups were proportionately represented at the different time frames within the entire range (5–18 y). Methods for assessing pain and functional disabilities are crude. However, this study began in 1992, before the advent of the Disabilities of the Arm, Shoulder, and Hand questionnaire30 and the Patient Evaluation Measure,31 and the subjective assessment techniques used in this study are comparable with those used at the time. Because we wanted to maintain uniformity and comparability of the data, we continued the same outcome measures in this study. Furthermore, we placed a K-wire across the trapezial void for 4 weeks, and a thumb spica cast was maintained for 6 weeks in all patients. As such, the simple trapeziectomy performed in this study was dissimilar to that described by Gervis2 but similar to the subsequently reported hematoma and distraction technique that produces good results.32,33 As described by Burton and Pellegrini,14 T⫹LRTI uses a K-wire and a thumb spica cast as in the present study; when this study was designed, we decided to follow the same postoperative rehabilitation for all 3 operations. It is possible that the K-wire used in conjunction with simple excision of the trapezium may have resulted in a more stable pseudar-
throsis between the scaphoid and the thumb metacarpal base, and thus it is not possible to claim that the results after simple trapeziectomy without a K-wire and with less than 4 to 6 weeks of immobilization will produce results such as seen in this study. However, 2 other prospective randomized studies22,34 comparing trapeziectomy without a K-wire or thumb spica cast to trapeziectomy with LRTI using a K-wire for 4 weeks have failed to show any advantage of the latter procedure at 1 year. Although most patients had the additional procedures as described earlier for addressing MCP hyperextension, toward the end of the study the senior author became less certain of the need to treat MCP hyperextension of 45° or less; as such, these procedures were not routinely performed.25 We did not assess the thumbs radiographically as part of this study, because several investigators have shown that trapezial height maintenance has no correlation to the clinical outcome in terms of both pain relief and thumb pinch strength.18,22,32,35,36 Despite these criticisms, the results of this study show no benefit to LRTI after excision of the trapezium after a minimum follow-up of 5 years. Although the use of additional soft tissue procedures was not detrimental to the results of trapeziectomy, this study suggests that they are not necessary and that, even in the longer term, excision of the trapezium remains the most important part of the surgery. REFERENCES 1. Armstrong AL, Hunter JB, Davis TRC. The prevalence of degenerative arthritis of the base of the thumb in post-menopausal women. J Hand Surg 1994;19B:340 –341. 2. Gervis WH. Excision of the trapezium for osteoarthritis of the trapeziometacarpal joint. J Bone Joint Surg 1949;31B:537–539. 3. Dhar S, Gray ICM, Jones WA, Beddow FH. Simple excision of the trapezium for osteoarthritis of the carpometacarpal joint of the thumb. J Hand Surg 1994;19B:485– 488. 4. Varley GW, Calvey J, Hunter JB, Barton NJ, Davis TRC. Excision of the trapezium for osteoarthritis at the base of the thumb. J Bone Joint Surg 1994;76B:964 –968. 5. Gibbons CER, Gosal HS, Choudri AH, Magnussen PA. Trapeziectomy for basal thumb joint osteoarthritis: 3 to 19-year follow-up. Int Orthop 1999;23:216 –218. 6. Conolly WB, Rath S. Revision procedures for complications of surgery for osteoarthritis of the carpometacarpal joint of the thumb. J Hand Surg 1993;18B:533–539. 7. Froimson AI. Tendon arthroplasty of the trapeziometacarpal joint. Clin Orthop Relat Res 1970;70:191–199. 8. Menon J, Schoene HR, Hohl JC. Trapeziometacarpal arthritis— results of tendon interpositional arthroplasty. J Hand Surg 1981;6: 442– 446. 9. Amadio PC, Millender LH, Smith RJ. Silicone spacer or tendon spacer for trapezium resection arthroplasty— comparison of results. J Hand Surg 1982;7:237–244. 10. Dell PC, Muniz RB. Interposition arthroplasty of the trapeziometacarpal joint for osteoarthritis. Clin Orthop Relat Res 1987;220:27–34.
JHS 䉬 Vol A, March
TRAPEZIECTOMY RCT 5- TO 18-YEAR RESULTS
11. Atroshi I, Axelsson G. Extensor carpi radialis tendon arthroplasty in the treatment of primary trapeziometacarpal arthrosis. J Hand Surg 1997;22A:419 – 427. 12. Kaarela O, Raatikainen T. Abductor pollicis longus tendon interposition arthroplasty for carpometacarpal osteoarthritis of the thumb. J Hand Surg 1999;24A:469 – 475. 13. Eaton RG, Littler JW. Ligament reconstruction for the painful thumb carpometacarpal joint. J Bone Joint Surg 1973;55A:1655–1666. 14. Burton RI, Pellegrini VD. Surgical management of basal joint arthritis of the thumb. Part II. Ligament reconstruction with tendon interposition arthroplasty. J Hand Surg 1986;11A:324 –332. 15. Eaton RG, Lane LB, Littler JW, Keyser JJ. Ligament reconstruction for the painful thumb carpometacarpal joint: a long-term assessment. J Hand Surg 1984;9A:692– 699. 16. Nylen S, Johnson A, Rosenquist A-M. Trapeziectomy and ligament reconstruction for osteoarthrosis of the base of the thumb. A prospective study of 100 operations. J Hand Surg 1993;18B:616 – 619. 17. Tomaino MM, Pellegrini VD, Burton RI. Arthroplasty of the basal joint of the thumb. Long-term follow-up after ligament reconstruction with tendon interposition. J Bone Joint Surg 1995;77A:346 – 355. 18. Lins RE, Gelberman RH, McKeown L, Katz JN, Kadiyala RK. Basal joint arthritis: Trapeziectomy with ligament reconstruction and tendon interposition arthroplasty. J Hand Surg 1996;21A:202–209. 19. Gerwin M, Griffith A, Weiland A, Hotchkiss R, McCormack R. Ligament reconstruction basal joint arthroplasty without tendon interposition. Clin Orthop Relat Res 1997;342:42– 45. 20. Belcher H, Nicholl J. A comparison of trapeziectomy with and without ligament reconstruction and tendon interposition. J Hand Surg 2000;25B:350 –356. 21. De Smet L, Sioen W, Spaepen D, van Ransbeeck H. Treatment of basal joint arthritis of the thumb: trapeziectomy with or without tendon interposition/ligament reconstruction. Hand Surg 2004;9:5–9. 22. Field J, Buchanan D. To suspend or not to suspend: a randomized single blind trial of simple trapeziectomy versus trapeziectomy and flexor carpi radialis suspension. J Hand Surg 2007;32E:462– 466. 23. Kriegs-Au G, Petje G, Fojtl E, Ganger R, Zachs I. Ligament reconstruction with or without tendon interposition to treat primary thumb carpometacarpal osteoarthritis: a prospective randomized study. J Bone Joint Surg 2004;86A:209 –218.
417
24. Davis TRC, Brady O, Dias JJ. Excision of the trapezium for osteoarthritis of the trapeziometacarpal joint: a study of the benefit of ligament reconstruction or tendon interposition. J Hand Surg 2004; 29A:1069 –1077. 25. Poulter RJ, Davis TRC. Management of hyperextension of the metacarpophalangeal joint in association with trapeziometacarpal joint osteoarthritis. J Hand Surg 2011;36E:280 –284. 26. Hook WE, Stanley JK. Assessment of thumb to index pulp to pulp pinch grip strengths. J Hand Surg 1986;11B:91–92. 27. Kapandji A. Citation critique de l’opposition et de la contraposition du ponce. Ann Chir Main 1986;5:67–73. 28. Wajon A, Carr E, Edmunds I, Ada L. Surgery for thumb (trapeziometacarpal joint) osteoarthritis. Cochrane Database Syst Rev 2009; 4:CD004631. 29. Wajon A, Ada L, Edmunds I. Surgery for thumb (trapeziometacarpal joint) osteoarthritis. Cochrane Database Syst Rev 2005;4:CD004631. 30. Hudak PL, Amadio PC, Bombardier C. Development of an upper extremity outcome measure: the DASH (disabilities of the arm, shoulder and hand) [corrected]. The Upper Extremity Collaborative Group (UECG). Am J Ind Med 1996;29:602– 608. 31. Macey AC, Burke FD, Abbott K, Barton NJ, Bradbury E, Bradley A, et al. Outcomes of hand surgery. British Society for Surgery of the Hand. J Hand Surg 1995;20B:841– 855. 32. Kuhns CA, Emerson ET, Meals RA. Hematoma and distraction arthroplasty for thumb basal joint osteoarthritis: a prospective, single-surgeon study including outcome measures. J Hand Surg 2003; 28A:381–389. 33. Gray KV, Meals RA. Hematoma and distraction arthroplasty for thumb basal joint osteoarthritis: minimum 6.5-year follow-up evaluation. J Hand Surg 2007;32A:23–29. 34. Davis TR, Pace A. Trapeziectomy for trapeziometacarpal joint osteoarthritis: is ligament reconstruction and temporary stabilization of the pseudarthrosis with a Kirschner wire important? J Hand Surg 2009;34E:312–321. 35. Kadiyala RK, Gelberman RH, Kwon B. Radiographic assessment of the trapezial space before and after ligament reconstruction and tendon interposition arthroplasty. J Hand Surg 1996;21B:177–181. 36. Downing ND, Davis TR. Trapezial space height after trapeziectomy: mechanism of formation and benefits. J Hand Surg 2001;26A:862– 868.
JHS 䉬 Vol A, March
Five- to 18-Year Follow-Up for Treatment of Trapeziometacarpal Osteoarthritis: A Prospective Comparison of Excision, Tendon Interposition, and Ligament Reconstruction and Tendon Interposition Soham Gangopadhyay, MSc, Helen McKenna, Dip COT, Frank D. Burke, MBBS, Tim R. C. Davis, ChM
Purpose To investigate whether palmaris longus interposition or flexor carpi radialis ligament reconstruction and tendon interposition improve the outcome of trapezial excision for the treatment of basal joint arthritis after a minimum follow-up of 5 years. Methods We randomized 174 thumbs with trapeziometacarpal osteoarthritis into 3 groups to undergo simple trapeziectomy, trapeziectomy with palmaris longus interposition, or trapeziectomy with ligament reconstruction and tendon interposition using 50% of the flexor carpi radialis tendon. A K-wire was passed across the trapezial void and retained for 4 weeks, and a thumb spica was used for 6 weeks in all 3 groups. We reviewed 153 thumbs after a minimum of 5 years (median, 6 y; range, 5–18 y) after surgery with subjective and objective assessments of thumb pain, function, and strength. Results There was no difference in the pain relief achieved in the 3 treatment groups, with good results in 120 (78%) patients. Grip strength and key and tip pinch strengths did not differ among the 3 groups and range of movement of the thumb was similar. Few complications persisted after 5 years, and these were distributed evenly among the 3 groups. Compared with the results at 1 year in the same group of patients, the good pain relief achieved was maintained in the longer term, irrespective of the type of surgery. While improvements in grip strength achieved at 1 year after surgery were preserved, the key and tip pinch strengths deteriorated with time, but the type of surgery did not influence this. Conclusions The outcomes of these 3 variations of trapeziectomy were similar after a minimum follow-up of 5 years. There appears to be no benefit to tendon interposition or ligament reconstruction in the longer term. (J Hand Surg 2012;37A:411–417. Copyright © 2012 by the American Society for Surgery of the Hand. All rights reserved.) Type of study/level of evidence Therapeutic I. Key words Ligament reconstruction, osteoarthritis, thumb, trapeziectomy, trapezium.
From the Department of Trauma and Orthopaedics, Nottingham University Hospitals, Queen’s Medical Centre, Nottingham; and the Pulvertaft Hand Centre, Royal Derby Hospital, Derby, UK. Received for publication August 3, 2011; accepted in revised form November 29, 2011. The authors are grateful to Apostolos Fakis, Medical Statistician, Department of Research and Development,RoyalDerbyHospital,UnitedKingdom,forinvaluablecontributiontothestatisticalanalyses of the data presented.
No benefits in any form have been received or will be received related directly or indirectly to the subject of this article. Corresponding author: Soham Gangopadhyay, MSc, Department of Trauma and Orthopaedics, Nottingham University Hospitals, Queen’s Medical Centre, Derby Road, Nottingham NG7 2UH, UK; e-mail: [email protected]. 0363-5023/12/37A03-0001$36.00/0 doi:10.1016/j.jhsa.2011.11.027
© ASSH 䉬 Published by Elsevier, Inc. All rights reserved. 䉬 411
412
O
TRAPEZIECTOMY RCT 5- TO 18-YEAR RESULTS
STEOARTHRITIS AT THE TRAPEZIOMETACARPAL
joint is common in postmenopausal women1 and its surgical treatment has evolved over the years. In 1949, Gervis2 reported simple excision of the trapezium, a procedure that is reported to produce satisfactory results.3–5 However, there have been concerns that this procedure may result in shortening of the thumb, owing to proximal migration of the metacarpal, and lead to weakness. Subluxation and osteoarthritis at the pseudarthrosis between the scaphoid and the thumb metacarpal are other concerns.6 To overcome these potential problems, various modifications have been proposed. Placement of a rolled-up length of tendon in the trapezial void7–12 has been used to act as a biological spacer, whereas creation of a ligament between the thumb and index metacarpal bases13,14 has been proposed to prevent instability at the pseudarthrosis. Excision of the trapezium with ligament reconstruction with or without tendon interposition is a commonly performed procedure for trapeziometacarpal osteoarthritis and has good results.14 –19 However, there may be no advantage to either tendon interposition or ligament reconstruction over simple trapeziectomy.19 –24 Most studies that show no benefit to either adjunct have relatively short average follow-up times ranging between 1 and 4 years, and it may be that with longer follow-up the theoretical advantages of tendon interposition or ligament reconstruction may become clinically relevant. The primary purpose of this prospective randomized study was to investigate the outcome and compare the results of trapeziectomy alone (T), trapeziectomy with palmaris longus (PL) tendon interposition (T⫹PL), and trapeziectomy with ligament reconstruction and tendon interposition (LRTI) using half of the flexor carpi radialis (FCR) tendon (T⫹LRTI) in 153 thumbs after a minimum follow-up of 5 years (median, 6 y; range, 5–18 y). Because the same cohort of patients had also been assessed 1 year after surgery,24 we were able to compare the results at 1 year with those after a minimum follow-up of 5 years. MATERIALS AND METHODS The local ethics committees approved this study. Women with painful trapeziometacarpal osteoarthritis who had failed to respond to nonoperative treatment were recruited between 1992 and 2001. We excluded men from this study because they experience this condition much less frequently, and we wanted to have as homogeneous a group as possible. All had Eaton and Littler13 grade 2 through 4 osteoarthritis. A total of 153 women had 174 procedures; 21 had surgery on both thumbs at different times during the recruitment period.
Of the 174 operated thumbs, 153 (88% of the original cohort) were assessed at a median of 6 years (range, 5–18 y). Three patients had moved and could not be traced, 5 patients had died, and 13 patients did not want to be reviewed. Although it was our intention to assess all patients at 5 years, this could not be achieved, and we reviewed 50 thumbs at 5 years, 81 at 6 to 9 years, and 22 at 10 to 18 years after surgery. We randomized patients into 3 groups to undergo T, T⫹PL, or T⫹LRTI. We achieved randomization at the induction of anesthesia by opening the next sequentially numbered sealed, opaque envelope that contained instructions as to which operation should be performed. It was stratified such that each set of 9 consecutive surgeries would include (in random order) 3 of each of the 3 surgical options. Table 1 shows that the 3 groups were well matched before surgery and outlines the operative details and additional procedures performed. We excised the trapezium piecemeal using a dorsal approach. When used, the PL was rolled up and sutured to itself before placement in the trapezial void. We performed the FCR ligament reconstruction and tendon interposition as described by Burton and Pellegrini14 using 50% of the tendon, with 2 exceptions: the trapezium was completely excised in all cases irrespective of the extent of pantrapezial involvement,17 and in the cases done after 1994, the base of the first metacarpal was not resected. Instead, we made a drill hole for passage of the tendon from the dorsal aspect of the thumb metacarpal base to exit the palmar ulnar edge of its articular surface adjacent to the index metacarpal. The tendon was passed through this drill hole from its insertion on the index metacarpal base to exit the thumb metacarpal base dorsally, where it was sutured to the periosteum and joint capsule. The remaining length was then rolled up to act as a spacer. During each of the 3 procedures, we inserted a percutaneous K-wire through the base of the thumb metacarpal and passed it longitudinally across the trapezial void into the distal scaphoid to hold the metacarpal base at the level of the trapezoid–index carpometacarpal joint. We performed additional procedures as needed (Table 1). In most instances, when metacarpophalangeal (MCP) hyperextension was greater than 30°, we stabilized it with a K-wire for 4 weeks; when greater than 40°, we performed palmar capsulodesis or MCP arthrodesis.25 The K-wire across the trapezial void was cut outside the skin and removed at 4 weeks. We used a plaster of Paris cast that maintained the thumb in abduction with the wrist in neutral for a total of 6 weeks. At that stage,
JHS 䉬 Vol A, March
TRAPEZIECTOMY RCT 5- TO 18-YEAR RESULTS
TABLE 1. Patient Details, Additional Operations Performed, and Tourniquet Times in the 3 Treatment Groups for the 153 Procedures With a Minimum of 5 Years of Follow-Up
413
TABLE 2. Pain Levels (Median [Interquartile Range]) Before Surgery and After a Minimum Follow-Up of 5 Years in the 3 Groups of Patients Preoperative
Final
T
T⫹PL
T⫹LRTI
T (n ⫽ 53)
5 (4–6)
0 (0–1)
Thumbs
53
46
54
T⫹PL (n ⫽ 46)
5 (4–6)
1 (0–2)
Dominant
30
22
22
T⫹LRTI (n ⫽ 54)
5 (4–6)
0 (0–1)
.407
.383
Median age (range) (y)
57 (44–74) 57 (40–75) 57 (44–75)
P value
Status Housewife/retired
33
30
33
Office worker
5
1
6
Fine skill
0
0
3
Light manual
15
Mean trapeziometacarpal 3.5 (2–4) osteoarthritis grade (range) Scaphotrapezial joint osteoarthritis (n)
15 3.6 (2–4)
12 3.7 (2–4)
6
5
5
Consultant
27
33
36
Supervised trainee
26
13
18
Surgeon
Treatment center Derby
11
5
11
Nottingham
42
41
43
10
17
12
3
3
5
Additional procedures Carpal tunnel decompression MCP K-wire MCP capsulodesis
0
3
6
MCP arthrodesis
1
1
2
de Quervain release
2
1
1
Trigger thumb release
2
0
2
0
1
0
Trigger finger release Median tourniquet time (interquartile range) (min)a
We assessed the level of pain at every stage using the following scoring system: 0 ⫽ no pain; 1 ⫽ pain only with and after use, no restrictions; 2 ⫽ pain with and after use, some restrictions; 3 ⫽ rest pain without activity restriction; 4 ⫽ rest pain, some restrictions; 5 ⫽ rest pain, severe restrictions; and 6 ⫽ rest pain with night wakening.
43 (36–50) 50 (46–61) 63 (55–75)
a
Tourniquet time was the only variable among the data presented in this table that was significantly different among the 3 groups (P ⬍ .001).
the patients were shown a set of exercises to mobilize and strengthen the thumb. We assessed patients after a minimum of 5 years with the same subjective and objective measures that had been used preoperatively and at 1 year. Two independent observers who were not involved in the surgical procedure carried out all assessments at final followup. Although the observers could not be truly kept unaware of the study protocol owing to the presence of
forearm scars from tendon harvest, no attempt was made to identify these scars before assessing the thumb. We invited patients to attend the outpatient clinics; for those unable to attend, we organized home visits. We made subjective assessments of pain and restriction of activity. We performed objective measurements of grip strength and key and tip pinch strengths with calibrated grip and pinch meters (Jamar, Jackson, MO) using standardized techniques.26 We assessed thumb opposition27 and measured MCP hyperextension. Postoperative complications persisting after 5 years were recorded. We used the Kolmogorov-Smirnov test and histograms to assess the distribution of the variables. None of the variables were normally distributed; as such, we used nonparametric tests for the analyses. We compared the continuous variables among the 3 surgical groups using the Kruskal-Wallis test and between 2 groups using the Mann-Whitney U test. We compared categorical variables at baseline using the chi-square test. Results are presented as median values with the interquartile ranges and were considered to be significant at a P value of ⬍ .050. RESULTS Table 2 shows the preoperative and final pain levels in the 3 groups. We assessed pain on a scale of 0 to 6 as defined in Table 2. There was no difference among the 3 groups with regard to pain relief, and most patients had no pain or mild discomfort with use that did not restrict activity. At the final assessment, 14 patients (7T⫹LRTI, 4T⫹PL, and 3T) had rest pain with or without activity restriction, although all were better than before surgery. Of these, 4 had undergone revision operations. Two patients (T⫹LRTI) had been com-
JHS 䉬 Vol A, March
TRAPEZIECTOMY RCT 5- TO 18-YEAR RESULTS
414
TABLE 3. Key Pinch (Median [Interquartile Range]) (kg) of Affected/Operated Thumb and Contralateral Normal Thumb in the 3 Groups Before Surgery, at 1 Year and at Final Review After a Minimum of 5 Years Thumb Assessed Preoperative
Affected Contralateral
1y Final
a
a
Operated
T
T⫹PL
T⫹LRTI
P Value
3.6 (2.3–4.5) (n ⫽ 53)
3.5 (2.2–5.5) (n ⫽ 46)
3.2 (2.0–4.1) (n ⫽ 54)
.415
5.5 (3.6–6.4) (n ⫽ 47)
4.5 (2.7–6.8) (n ⫽ 44)
5.3 (3.5–6.8) (n ⫽ 49)
.967
4.5 (3.6–5.9) (n ⫽ 53)
4.3 (2.7–5.9) (n ⫽ 46)
4.5 (3.2–5.5) (n ⫽ 54)
.388
Contralaterala
6.3 (4.5–7.2) (n ⫽ 36)
5.0 (3.6–6.8) (n ⫽ 35)
5.5 (4.1–6.4) (n ⫽ 35)
.236
Operated
4.1 (2.7–5.0) (n ⫽ 53)
3.4 (1.8–5.5) (n ⫽ 46)
3.6 (2.7–5.0) (n ⫽ 54)
.471
Contralaterala
4.1 (2.7–5.2) (n ⫽ 37)
2.7 (1.8–5.0) (n ⫽ 35)
3.6 (2.7–5.0) (n ⫽ 37)
.200
We obtained contralateral strength data from patients who had not had surgery and did not have painful osteoarthritis of the contralateral thumb.
TABLE 4. Tip Pinch (Median [Interquartile Range]) (kg) of Affected/Operated Thumb and Contralateral Normal Thumb in the 3 Groups Before Surgery, at 1 Year, and at Final Review After a Minimum of 5 Years Thumb Assessed Preoperative
Affected Contralateral
1y Final
a
a
Operated
T
T⫹PL
T⫹LRTI
P Value
2.3 (1.4–3.2) (n ⫽ 53)
2.3 (1.4–3.6) (n ⫽ 46)
2.3 (1.4–2.7) (n ⫽ 54)
.541
3.2 (1.6–4.1) (n ⫽ 47)
3.2 (0.2–4.1) (n ⫽ 44)
3.2 (1.7–4.1) (n ⫽ 49)
.990
3.2 (2.3–4.1) (n ⫽ 53)
3.1 (1.8–4.1) (n ⫽ 46)
3.6 (2.7–4.1) (n ⫽ 54)
.544
Contralaterala
4.1 (3.2–4.5) (n ⫽ 36)
3.6 (2.3–4.5) (n ⫽ 35)
3.6 (3.2–4.5) (n ⫽ 35)
.164
Operated
2.7 (1.8–3.2) (n ⫽ 53)
2.5 (1.4–3.6) (n ⫽ 46)
2.7 (1.8–3.3) (n ⫽ 54)
.681
Contralaterala
2.7 (1.8–3.6) (n ⫽ 37)
1.8 (0.9–3.2) (n ⫽ 35)
2.7 (1.9–3.2) (n ⫽ 37)
.289
We obtained contralateral strength data from patients who had not had surgery and did not have painful osteoarthritis of the contralateral thumb.
pletely pain free for 12 years before deteriorating. Both were revised to metacarpal hemiarthroplasties, which improved their pain. One (T) had failed early with proximal subluxation of the metacarpal base and was revised with additional LRTI at 1 year with no benefit. The fourth patient (T) had a superficial radial nerve neuroma that was excised early; she subsequently underwent revision to a metacarpal hemiarthroplasty at 4 years, but neither procedure improved the pain. Two patients with continuing pain were being treated for fibromyalgia, and 2 others had severe osteoarthritis affecting the hands. No specific cause for the continuing pain could be identified in the remaining 6 patients. Ability to perform activities such as writing, turning a key, opening a screw-top jar, handling coins, and knitting did not differ between groups at any stage. Tables 3 to 5 show the key pinch, tip pinch, and grip strengths of the operated and the contralateral hands before surgery, at 1 year, and after a minimum of 5 years of follow-up. There was no significant difference
in these strengths among the 3 treatment groups before surgery, at 1 year, or at final follow-up. All 3 types of surgery improved the key and tip pinch strengths. Before surgery, the contralateral normal thumb was considerably stronger than the affected thumb, but by 5 years, the key and tip pinch strengths were similar in both thumbs. Similarly, the grip strengths at final review were much better than the preoperative values and in fact were slightly better than the contralateral hand at this stage. There was no difference in opposition between groups, and at final follow-up, 137 of the 153 thumbs could reach the proximal interphalangeal joint skin crease of the little finger or beyond (P ⫽ .870). Metacarpophalangeal hyperextension had a median value of 10° (interquartile range, 0° to 20°) before surgery (P ⫽ .341) and 5° (interquartile range, 0° to 15°) at final review (P ⫽ .411), with no difference among the 3 groups. Table 6 shows the complications at 1 year and after 5 years. The incidence of complications persisting in the longer term was similar in all 3 groups.
JHS 䉬 Vol A, March
TRAPEZIECTOMY RCT 5- TO 18-YEAR RESULTS
415
TABLE 5. Grip Strength (Median [Interquartile Range]) (kg) of Affected/Operated Hand and Contralateral Normal Hand in the 3 Groups Before Surgery, at 1 Year, and at Final Review After a Minimum of 5 Years Thumb Assessed Preoperative
Affected Contralateral
1y
Operated
Final
a
a
T
T⫹PL
T⫹LRTI
P Value
14 (11–18) (n ⫽ 53)
13 (7–20) (n ⫽ 46)
13 (9–17) (n ⫽ 54)
.624
20 (14–24) (n ⫽ 47)
18 (13–26) (n ⫽ 44)
20 (15–25) (n ⫽ 49)
.770
19 (14–26) (n ⫽ 53)
20 (11–24) (n ⫽ 46)
22 (16–26) (n ⫽ 54)
.423
Contralaterala
23 (16–28) (n ⫽ 36)
20 (15–26) (n ⫽ 35)
22 (18–28) (n ⫽ 35)
.329
Operated
20 (14–25) (n ⫽ 53)
18 (8–26) (n ⫽ 46)
20 (12–24) (n ⫽ 54)
.908
Contralaterala
16 (12–20) (n ⫽ 37)
14 (8–20) (n ⫽ 35)
18 (10–22) (n ⫽ 37)
.471
We obtained contralateral strength data from patients who had not had surgery and did not have painful osteoarthritis of the contralateral thumb.
TABLE 6. Complications of Surgery at 1 Year and After a Minimum Follow-Up of 5 Years in the 3 Groups of Patients T (n ⫽ 53)
T⫹PL (n ⫽ 46)
T⫹LRTI (n ⫽ 54)
1y
5
4
1
5y
2
2
1
1y
2
3
3
5y
2
2
3
1y
0
5
3
5y
0
1
2
1y
0
1
3
5y
0
0
2
1y
1
0
1
5y
0
0
0
Complications Superficial radial nerve dysfunction
Palmar cutaneous branch of median nerve dysfunction
FCR/pollicis longus pulling sensation
Tender scar
Complex regional pain syndrome
DISCUSSION This study showed that after a minimum follow-up of 5 years, the outcome after excision of the trapezium is not improved by either tendon interposition alone or LRTI. The individual treatment options compared in this study are all known to produce good results in case series,3–5,8,9,11,12,16 –18 including studies with relatively longer follow-up.3–5,12,17 However, the superiority of
any 1 procedure over another has never been proven. Existing prospective randomized studies that have compared trapeziectomy alone with other treatment options have failed to establish a difference.20 –24 A recent update of the Cochrane systematic review28,29 has concluded that there is no established superiority of 1 procedure over another and has recommended trapeziectomy alone as the safest option with good outcomes. The patients in this study had been reviewed at 1 year after surgery24; as such, we were able to compare the outcomes in the same patients to see whether the results changed with time. The excellent pain relief achieved at 1 year was maintained in the longer term, with no difference among the 3 groups. At 1 year, 81% of the thumbs were either pain free or had pain only with use without activity restriction. At final follow-up, this level of pain relief was maintained in 78%. In 10 patients, the pain relief achieved at 1 year had deteriorated. Two of these had been revised with relative improvement in pain, 3 had osteoarthritis of the hand, and 1 had cervical spondylosis. For the other 4 patients, we could find no explanation; specifically, they did not have painful instability of the pseudoarthrosis. Although 7 thumbs had rest pain with severe activity restriction or night pain at 1 year, no thumb had this level of pain after 5 years. Although the thumb key and tip pinch strengths were not significantly different between groups after 5 years, both were weaker than at 1 year. The contralateral thumbs had also become weaker, and it is probable that the deterioration in thumb strength is age related and not a function of the surgical procedure. The hand grip strength did not deteriorate with time. These findings differ from those of another long-term report on 25 patients with T⫹LRTI,17 where the grip and thumb tip pinch strengths steadily improved up to 6 years before
JHS 䉬 Vol A, March
416
TRAPEZIECTOMY RCT 5- TO 18-YEAR RESULTS
leveling off, whereas the key pinch strength reached a maximum at 6 years and then showed a slight decline. The Cochrane reviews28,29 reported that T⫹LRTI has the highest complication rate. Nevertheless, our study found no difference in complication rates among the 3 surgical procedures. Moreover, there were fewer ongoing complications after 5 years than at the 1-year review in all 3 groups. Radial nerve dysfunction and pulling sensation in the forearm had resolved in 50%, and chronic regional pain syndrome had resolved in both patients by 5 years, compared with the 1-year results. In contrast, numbness in the distribution of the palmar cutaneous branch of the median nerve persisted after 5 years in most. Four patients in this series underwent revision operations. The metacarpal hemiarthoplasties were performed using a pyrocarbon implant inserted into the metacarpal base, articulating with a socket created in the distal scaphoid using a burr. This study can be criticized for various reasons. We were unable to review all patients at 5 years as initially intended, and the results represent the findings at a median time of 6 years. Fifty thumbs (20 T, 14 T⫹PL, and 16 T⫹LRTI) were seen at 5 years, 81 (29 T, 26 T⫹PL, and 26 T⫹LRTI) at 6 to 9 years, and 22 (4 T, 6 T⫹PL, and 12 T⫹LRTI) at 10 years or more after surgery. However, we reviewed most patients between 5 and 9 years, and the 3 treatment groups were proportionately represented at the different time frames within the entire range (5–18 y). Methods for assessing pain and functional disabilities are crude. However, this study began in 1992, before the advent of the Disabilities of the Arm, Shoulder, and Hand questionnaire30 and the Patient Evaluation Measure,31 and the subjective assessment techniques used in this study are comparable with those used at the time. Because we wanted to maintain uniformity and comparability of the data, we continued the same outcome measures in this study. Furthermore, we placed a K-wire across the trapezial void for 4 weeks, and a thumb spica cast was maintained for 6 weeks in all patients. As such, the simple trapeziectomy performed in this study was dissimilar to that described by Gervis2 but similar to the subsequently reported hematoma and distraction technique that produces good results.32,33 As described by Burton and Pellegrini,14 T⫹LRTI uses a K-wire and a thumb spica cast as in the present study; when this study was designed, we decided to follow the same postoperative rehabilitation for all 3 operations. It is possible that the K-wire used in conjunction with simple excision of the trapezium may have resulted in a more stable pseudar-
throsis between the scaphoid and the thumb metacarpal base, and thus it is not possible to claim that the results after simple trapeziectomy without a K-wire and with less than 4 to 6 weeks of immobilization will produce results such as seen in this study. However, 2 other prospective randomized studies22,34 comparing trapeziectomy without a K-wire or thumb spica cast to trapeziectomy with LRTI using a K-wire for 4 weeks have failed to show any advantage of the latter procedure at 1 year. Although most patients had the additional procedures as described earlier for addressing MCP hyperextension, toward the end of the study the senior author became less certain of the need to treat MCP hyperextension of 45° or less; as such, these procedures were not routinely performed.25 We did not assess the thumbs radiographically as part of this study, because several investigators have shown that trapezial height maintenance has no correlation to the clinical outcome in terms of both pain relief and thumb pinch strength.18,22,32,35,36 Despite these criticisms, the results of this study show no benefit to LRTI after excision of the trapezium after a minimum follow-up of 5 years. Although the use of additional soft tissue procedures was not detrimental to the results of trapeziectomy, this study suggests that they are not necessary and that, even in the longer term, excision of the trapezium remains the most important part of the surgery. REFERENCES 1. Armstrong AL, Hunter JB, Davis TRC. The prevalence of degenerative arthritis of the base of the thumb in post-menopausal women. J Hand Surg 1994;19B:340 –341. 2. Gervis WH. Excision of the trapezium for osteoarthritis of the trapeziometacarpal joint. J Bone Joint Surg 1949;31B:537–539. 3. Dhar S, Gray ICM, Jones WA, Beddow FH. Simple excision of the trapezium for osteoarthritis of the carpometacarpal joint of the thumb. J Hand Surg 1994;19B:485– 488. 4. Varley GW, Calvey J, Hunter JB, Barton NJ, Davis TRC. Excision of the trapezium for osteoarthritis at the base of the thumb. J Bone Joint Surg 1994;76B:964 –968. 5. Gibbons CER, Gosal HS, Choudri AH, Magnussen PA. Trapeziectomy for basal thumb joint osteoarthritis: 3 to 19-year follow-up. Int Orthop 1999;23:216 –218. 6. Conolly WB, Rath S. Revision procedures for complications of surgery for osteoarthritis of the carpometacarpal joint of the thumb. J Hand Surg 1993;18B:533–539. 7. Froimson AI. Tendon arthroplasty of the trapeziometacarpal joint. Clin Orthop Relat Res 1970;70:191–199. 8. Menon J, Schoene HR, Hohl JC. Trapeziometacarpal arthritis— results of tendon interpositional arthroplasty. J Hand Surg 1981;6: 442– 446. 9. Amadio PC, Millender LH, Smith RJ. Silicone spacer or tendon spacer for trapezium resection arthroplasty— comparison of results. J Hand Surg 1982;7:237–244. 10. Dell PC, Muniz RB. Interposition arthroplasty of the trapeziometacarpal joint for osteoarthritis. Clin Orthop Relat Res 1987;220:27–34.
JHS 䉬 Vol A, March
TRAPEZIECTOMY RCT 5- TO 18-YEAR RESULTS
11. Atroshi I, Axelsson G. Extensor carpi radialis tendon arthroplasty in the treatment of primary trapeziometacarpal arthrosis. J Hand Surg 1997;22A:419 – 427. 12. Kaarela O, Raatikainen T. Abductor pollicis longus tendon interposition arthroplasty for carpometacarpal osteoarthritis of the thumb. J Hand Surg 1999;24A:469 – 475. 13. Eaton RG, Littler JW. Ligament reconstruction for the painful thumb carpometacarpal joint. J Bone Joint Surg 1973;55A:1655–1666. 14. Burton RI, Pellegrini VD. Surgical management of basal joint arthritis of the thumb. Part II. Ligament reconstruction with tendon interposition arthroplasty. J Hand Surg 1986;11A:324 –332. 15. Eaton RG, Lane LB, Littler JW, Keyser JJ. Ligament reconstruction for the painful thumb carpometacarpal joint: a long-term assessment. J Hand Surg 1984;9A:692– 699. 16. Nylen S, Johnson A, Rosenquist A-M. Trapeziectomy and ligament reconstruction for osteoarthrosis of the base of the thumb. A prospective study of 100 operations. J Hand Surg 1993;18B:616 – 619. 17. Tomaino MM, Pellegrini VD, Burton RI. Arthroplasty of the basal joint of the thumb. Long-term follow-up after ligament reconstruction with tendon interposition. J Bone Joint Surg 1995;77A:346 – 355. 18. Lins RE, Gelberman RH, McKeown L, Katz JN, Kadiyala RK. Basal joint arthritis: Trapeziectomy with ligament reconstruction and tendon interposition arthroplasty. J Hand Surg 1996;21A:202–209. 19. Gerwin M, Griffith A, Weiland A, Hotchkiss R, McCormack R. Ligament reconstruction basal joint arthroplasty without tendon interposition. Clin Orthop Relat Res 1997;342:42– 45. 20. Belcher H, Nicholl J. A comparison of trapeziectomy with and without ligament reconstruction and tendon interposition. J Hand Surg 2000;25B:350 –356. 21. De Smet L, Sioen W, Spaepen D, van Ransbeeck H. Treatment of basal joint arthritis of the thumb: trapeziectomy with or without tendon interposition/ligament reconstruction. Hand Surg 2004;9:5–9. 22. Field J, Buchanan D. To suspend or not to suspend: a randomized single blind trial of simple trapeziectomy versus trapeziectomy and flexor carpi radialis suspension. J Hand Surg 2007;32E:462– 466. 23. Kriegs-Au G, Petje G, Fojtl E, Ganger R, Zachs I. Ligament reconstruction with or without tendon interposition to treat primary thumb carpometacarpal osteoarthritis: a prospective randomized study. J Bone Joint Surg 2004;86A:209 –218.
417
24. Davis TRC, Brady O, Dias JJ. Excision of the trapezium for osteoarthritis of the trapeziometacarpal joint: a study of the benefit of ligament reconstruction or tendon interposition. J Hand Surg 2004; 29A:1069 –1077. 25. Poulter RJ, Davis TRC. Management of hyperextension of the metacarpophalangeal joint in association with trapeziometacarpal joint osteoarthritis. J Hand Surg 2011;36E:280 –284. 26. Hook WE, Stanley JK. Assessment of thumb to index pulp to pulp pinch grip strengths. J Hand Surg 1986;11B:91–92. 27. Kapandji A. Citation critique de l’opposition et de la contraposition du ponce. Ann Chir Main 1986;5:67–73. 28. Wajon A, Carr E, Edmunds I, Ada L. Surgery for thumb (trapeziometacarpal joint) osteoarthritis. Cochrane Database Syst Rev 2009; 4:CD004631. 29. Wajon A, Ada L, Edmunds I. Surgery for thumb (trapeziometacarpal joint) osteoarthritis. Cochrane Database Syst Rev 2005;4:CD004631. 30. Hudak PL, Amadio PC, Bombardier C. Development of an upper extremity outcome measure: the DASH (disabilities of the arm, shoulder and hand) [corrected]. The Upper Extremity Collaborative Group (UECG). Am J Ind Med 1996;29:602– 608. 31. Macey AC, Burke FD, Abbott K, Barton NJ, Bradbury E, Bradley A, et al. Outcomes of hand surgery. British Society for Surgery of the Hand. J Hand Surg 1995;20B:841– 855. 32. Kuhns CA, Emerson ET, Meals RA. Hematoma and distraction arthroplasty for thumb basal joint osteoarthritis: a prospective, single-surgeon study including outcome measures. J Hand Surg 2003; 28A:381–389. 33. Gray KV, Meals RA. Hematoma and distraction arthroplasty for thumb basal joint osteoarthritis: minimum 6.5-year follow-up evaluation. J Hand Surg 2007;32A:23–29. 34. Davis TR, Pace A. Trapeziectomy for trapeziometacarpal joint osteoarthritis: is ligament reconstruction and temporary stabilization of the pseudarthrosis with a Kirschner wire important? J Hand Surg 2009;34E:312–321. 35. Kadiyala RK, Gelberman RH, Kwon B. Radiographic assessment of the trapezial space before and after ligament reconstruction and tendon interposition arthroplasty. J Hand Surg 1996;21B:177–181. 36. Downing ND, Davis TR. Trapezial space height after trapeziectomy: mechanism of formation and benefits. J Hand Surg 2001;26A:862– 868.
JHS 䉬 Vol A, March