Keywords
four-bone arthrodesis - SLAC - SNAC - wrist - locking plate
Introduction
Scapholunate (SL) ligament integrity is crucial in maintaining the stability of the
wrist. In fact, a lesion of this ligament could also lead to a degenerative change
called scapholunate advanced collapse (SLAC) with a severe articular impairment.
The SLAC condition was first described in 1984 by Watson and Ballet,[1] who described an initial SL incompetence with an initial degenerative condition
of the radioscaphoid joint (stage I), followed by a more important degeneration of
it (stage II) with a progressive involvement to the midcarpal joint (stage III). After
these three stages, a stage IV was introduced to indicate a condition with a pan-radiocarpal
osteoarthritis (OA)[2] ([Fig. 1]).
Fig. 1 Stage III scapholunate advanced collapse wrist. Anteroposterior (A) and lateral views (B) show damage to the radioscaphoid and lunocapitate articular surfaces.
A similar progressive wrist degeneration is due to a scaphoid nonunion in undiagnosed
fracture. The same progressive degenerative stages are described and the same surgical
treatment is indicated in case of scaphoid nonunion advanced collapse (SNAC).
The treatment of all these conditions is surgical since no conservative treatment
can restore both the competency of the SL ligament and the degenerated cartilage surfaces.
In advanced stages (stage III), the treatment is represented mostly by two salvage
procedure: proximal row carpectomy (PRC) or scaphoid excision with fusion of the lunate,
triquetrum, hamate and capitate, so-called four-bone arthrodesis (FBA).
The aim of this study is to evaluate the clinical and radiological outcomes in patients
who underwent a FBA with dorsal locking plate because of a stage III SLAC/SNAC.
Methods
The inclusion criteria of this retrospective study were: stage III SLAC/SNAC wrist,
surgical treatment performed between 2005 and 2010, and use of a locking dorsal plate
(LDP) in performing the FBA. The diagnosis of SLAC/SNAC wrist was performed on the
evaluation of standard X-rays.
Twenty patients (18 males and 2 females) were enrolled in this study. The mean age
of the patients was 53.1 years. Fourteen patients were affected by a stage III SNAC
and six by a stage III SLAC. No previous surgery to the affected wrist had been performed
in 18 patients, one patient had open reduction and internal fixation with staple in
scaphoid fracture, and one had proximal scaphoid pole excision and prosthesis substitution.
In 14 cases, surgery was performed on the dominant upper arm, while in 6 cases the
nondominant side was involved.
Surgery was performed through a dorsal approach to the carpus: a longitudinal midcarpal
skin incision was performed, then the third extensor retinaculum was opened with retraction
of the extensor pollicis longus. Triangular capsulotomy was achieved following the
fiber of the radiotriquetral and dorsal intercarpal ligaments, then a scaphoidectomy
was performed and the whole cartilage surfaces between capitate, hamate, triquetrum,
and lunate were removed. All the four bones were temporarily fixed with K-wires, paying
attention to place in neutral position the lunate, thus reducing the risk of dorsal
intercalated segment instability and maintaining the carpal height, and then the locking
plate was placed. When all the screws were positioned, the K-wires were removed. After
removal of the cartilage surface, cancellous bone from the scaphoid was used to fill
all the gaps to promote fusion. Attention was paid to mill the carpal bones, in particular
the lunate that is more sclerotic than other carpal bones, to allow the plate to be
placed under the dorsal margin of the bones, thus avoiding dorsal impingement with
radial edge during wrist extension.
Three different type of dorsal locking plate were used: Arthrodesis Plate 2.0 (Medartis,
Basel, Switzerland) in 5 patients, Hub-Cap TM (Acumed, Hillsboro, Oregon, United States)
in 10 patients, and Flower plate (KLS Martin, Freiburg, Germany) in 5 patients.
After surgery, all patients wore an immobilization cast for 6 weeks. Rehabilitation
was allowed only when healing was confirmed on X-rays and after positive clinical
check.
All patients were clinically and radiographically re-evaluated at the final follow-up.
The clinical outcomes were assessed by the evaluation of pain (visual analog scale,
VAS score), range of motion (ROM) (flexion and extension of the wrist) and grip strength,
and by collecting the Mayo wrist score and the Disability of the Arm, Shoulder and
Hand (DASH) score, while the imaging evaluation was based on standard X-rays of the
wrist (anteroposterior and lateral views).
Results
The mean follow-up was 6 years (range: 1–11 years). The mean DASH score was 16.6 ± 11
points; the Mayo wrist score was good for 2 patients, fair for 16 patients, and bad
for 2 patients. In the latter cases, surgery decreased the ROM of the wrist due to
the dorsal prominence of the plate. VAS was two in case of heavy activity. The mean
grip strength was 75% compared with the healthy side.
No differences were noticed using different kinds of locking dorsal plate. No ulnar
translation of the carpal bone was detected in this series.
The mean ROM in flexion was 42° ± 18.5° (range: 20–70°) and in extension was 37° ± 12.7°
(range: 20–60°) ([Table 1]).
Table 1
Descriptive statistics and follow-up data
|
M/F
|
Age
|
Involved wrist (R dominant)
|
Extension
|
Flexion
|
|
Patient 1
|
M
|
58
|
R/L
|
60
|
40
|
|
Patient 2
|
M
|
57
|
R/R
|
40
|
70
|
|
Patient 3
|
F
|
73
|
R/R
|
30
|
20
|
|
Patient 4
|
M
|
59
|
R/R
|
55
|
60
|
|
Patient 5
|
M
|
64
|
R/R
|
20
|
30
|
|
Patient 6
|
M
|
67
|
R/L
|
40
|
70
|
|
Patient 7
|
M
|
62
|
R/L
|
20
|
20
|
|
Patient 8
|
M
|
57
|
R/R
|
30
|
60
|
|
Patient 9
|
M
|
24
|
R/L
|
40
|
50
|
|
Patient 10
|
M
|
49
|
R/L
|
20
|
20
|
|
Patient 11
|
M
|
55
|
R/L
|
50
|
20
|
|
Patient 12
|
M
|
88
|
R/R
|
30
|
50
|
|
Patient 13
|
M
|
49
|
R/R
|
40
|
20
|
|
Patient 14
|
M
|
65
|
R/R
|
30
|
30
|
|
Patient 15
|
M
|
60
|
R/R
|
60
|
60
|
|
Patient 16
|
M
|
38
|
R/R
|
40
|
30
|
|
Patient 17
|
M
|
46
|
R/R
|
40
|
70
|
|
Patient 18
|
M
|
30
|
R/R
|
30
|
50
|
|
Patient 19
|
F
|
60
|
R/R
|
30
|
40
|
|
Patient 20
|
M
|
53
|
R/R
|
55
|
30
|
|
18 M /2 F
|
Mean: 53.1
|
14 dominant
|
Mean: 37°
|
Mean: 42°
|
In one case, a revision surgery was needed to remove the device because of the loosening
of the screws. In this case, a solid fusion between lunate and capitate was achieved
and no further surgery was necessary ([Fig. 2]).
Fig. 2 Stage III scaphoid nonunion advance collapse wrist. (A) Anteroposterior and lateral view. (B) Surgical treatment with dorsal locking plate. (C) Screws loosening 9 years after surgery. (D) Plate and screws removal: lunocapitate joint fusion is evident. (E) Clinical outcome 2 months after hardware removal.
Discussion
The SL ligament, as the scaphoid, has a key role in maintaining a normal kinematics
of the wrist and in making possible a full ROM without pain. When this ligament is
torn, it should be reconstructed to avoid degenerative changes to midcarpal articular
surfaces. Similar degenerative changes can occur in case of unrecognized scaphoid
nonunion.
Several techniques were presented to reconstruct this ligament[3]
[4]
[5] and to promote healing of scaphoid nonunion, and all of them could reach good clinical
outcomes if both radiocarpal and midcarpal articular surfaces are preserved (stage
I and II). If degenerative changes in articular surfaces take places in a wrist suffering
from SL dissociation, SLAC, or from SNAC, wrist could develop radioscaphoid and midcarpal
OA. These pathological conditions are classified in four stages (from I to IV), and
in stage III (radioscaphoid and capitolunate OA, preserved radiolunate articular surface)
the surgical salvage treatments are represented by PRC and FBA. Several studies compared
the two techniques,[6]
[7]
[8]
[9]
[10]
[11] highlighting their advantages and disadvantages.
PRC restores a greater ROM and there is no need of implant or bone fusion, but it
can preserve less grip strength compared with FBA.[7] However, FBA, which maintains a normal carpal height, can assure more preservation
of grip strength, but guarantees less ROM and has higher rate of complications, such
as implant failure or carpal nonunion.
In FBA, several technique were presented in the literature using different implant
such as using K-wire, Herbert's screws, staples, or specific plates. Multiple pinning
was initially used, then staples were introduced in 1990s and different designs were
studied in 2000s. Dorsal circular plates with locking design were recently presented.
The goal was to provide stable fixation and early mobilization, thus improving final
ROM.
The use of specific plate, as shown in studies by Kendall et al[12] and Merrell et al,[13] showed good results in the short- and mid-term follow-up, with a low rate of complications,
particularly nonunion of the carpal bones.
Several studies compared different hardware designs to understand if there is a real
advantage in using LDP. Literature referred about a low nonunion rate in case of K-wire
technique, ranging from 3% to 16%;[1]
[2]
[3]
[4]
[5]
[6]
[7] in contrast, migration, entry-point infection, and new surgery to remove K-wire
were more or less systematically reported, with discomfort for the patients.[14] Even with staples or compression screws the nonunion rate was similar, and time
of immobilization after surgery was 6 to 10 weeks. In case of staples, removal was
required because of dorsal impingement.[15]
Nonlocking dorsal plates (NLDPs) were introduced to obtain greater stability and to
allow early mobilization and to gain greater ROM, but results were less satisfactory
than expected. De Smet et al[6] reported approximately 62.5% of nonunion and no improvement in ROM. Merrell et al[13] observed that this high rate of nonunion is partly due to the use of plate to compensate
for imprecision in performing surgical technique. It is necessary a high precision
in reaming into cancellous bone, grafting from ipsilateral radius or excised scaphoid,
elimination of riming debris, correct position of the plate to avoid dorsal impingement
and fixation of each carpal bone by two screws. They also recommended 6 weeks of immobilization.
Kraisarin et al[16] in a cadaveric study compared NLDPs, K-wires, and LDPs. They demonstrated the biomechanical
superiority of LDP. Rhee and Shyn[17] reported approximately 4% of nonunion rate after 6 weeks of immobilization. On the
contrary, Luegmair and Houvet[18] reported 8% and 9% nonunion with 1 to 2 weeks of strict immobilization followed
by a removable cast or 2 more weeks and rehabilitation after 4 weeks.
In opposite of what expected, using DLP does not give advantages in term of ROM, strength,
or pain after surgery if compared with staples. The real improvement is in time to
return to work. Pauchard et al[14] showed that return to work was 3 to 4 months before with DLP than with staples or
NLDP;[14] even the rate to return to work was higher in LDP.
Ritt et al and Cayci and Carlsen showed that the block effect of the four-corner procedure
and the biomechanical effectiveness of this surgery were achieved even if the lunotriquetral
joint did not fuse.[19]
[20] Nevertheless, the goal of four-corner procedure is to fuse the four joint lines
and it is necessary for a meticulous cartilage resection to obtain a complete arthrodesis.
The main reason for revision surgery is dorsal impingement between the fixation device
and the posterior edge of radius. However, Le Corre et al[21] observed no correlation between hardware failure and surgical revision in the plate
group, as 44% of failed implant were asymptomatic.
The duration of postoperative immobilization is still controversial in the literature.
Some authors suggest that stable fixation allows early rehabilitation;[22] however, others recommended 6 to 8 weeks of cast after surgery to obtain a complete
healing.[23] Tielemans et al[24] established that a shorter postoperative immobilization results in better recovery
of function.
In our series, the use of these plates showed good clinical results, with a DASH score
that could be compared with the score of the healthy population,[25] with a very low rate of surgical revision and complications.
As stated in several clinical studies and confirmed also in biomechanical studies,
the FBA needs a good reduction in the lunate before the arthrodesis[14]
[15]
[26] to improve ROM. Indeed, in all of our cases, the first surgical step was to get
a good reduction in the lunate before performing the fusion. The use of bone chips
from the excised scaphoid was preferred to avoid harvest site morbidity and to shorten
the operative time. Furthermore, we played great attention in reaming cartilage, sinking
the plate, and maintaining the height of carpal bones.
In conclusion, the use of dorsal plate in performing FBA seems to be a safe and effective
procedure with a low rate of complications and good clinical and imaging outcome at
mid-term follow-up in treating stage III SLAC/SNAC wrist.
