Keywords
Kienböck's treatment - revascularization - flap - osteotomy - prosthesis - fusion
Introduction
This review focuses on the surgical management of Kienböck's disease and critically
examines the impact of specific disease and patient features on treatment selection.
Intentionally omitted is the discussion of background, history, and etiology that
can easily be found in other reviews. A common error induced by classifying Kienböck's
into discrete radiographic stages is failing to appreciate that the disease process
is a continuous state of flux within the lunate. Early and late Kienböck's are relatively
easy to conceptualize; the complexity occurs during the transitional phase of the
disease. In early Kienböck's perfusion is compromised but structure is maintained;
treatment is directed at restoring perfusion.[1]
[2] In late Kienböck's the lunate is fragmented beyond potential for reconstruction
into multiple necrotic segments involving both the proximal and distal articular surfaces.
Treatment involves removal of the necrotic lunate fragments and selecting from a short
list of alternative forms of articulation. Most of the errors in evaluation and management
occur when the lunate is in transition. Following the initial interruption of vascular
perfusion, a complex interplay then ensues between the structural integrity of the
bone and its viability.
Transitional Kienböck's must be evaluated not only for its presenting features on
the date of the patient encounter, including the three-dimensional renderings of magnetic
resonance imaging (MRI) and computed tomography (CT) scan, but understood as a process
continuously changing through the fourth dimension of time. The interaction between
compromised structural integrity and tissue viability/perfusion pathways continues
with each increment in time.[3]
[4] The surgeon's planned intervention will not immediately halt ongoing changes that
have inertia and continue to worsen before the surgical manipulations have an opportunity
to take effect. This review will examine how to recognize specific key features of
the continuously changing structure to viability relationship that guide the application
of various treatment strategies.
There are numerous problems in reviewing the past literature on Kienböck's treatment,
not the least of which is that the overwhelming majority of publications are level
IV evidence uncontrolled case series. Publications typically promote a favored treatment
strategy applied to a range of patients assessed at final outcome with subjective
scores, range of motion (ROM), grip strength (GS), and plain radiographic appearance.
The studies usually lack more discriminating information about the details of the
disease state both prior to intervention and at final assessment, rendering most comparative
analyses to conclude no discernible difference between treatments.[5]
[6] Accordingly, this review does not propose to recommend that any one particular treatment
strategy is preferred over others. Instead, the purpose is to examine how specific
identifiable features may influence the analysis and explanation by the surgeon as
well as the ultimate treatment selected by the patient.
Early Kienböck's
The features of early Kienböck's are diffuse perfusion abnormalities on MRI with normal
plain radiographs. The most common error is to interpret the MRI signal changes of
ulnocarpal impaction as representing Kienböck's. The natural history of early Kienböck's
has not been well defined, but the best evidence supports a greater potential for
spontaneous resolution in pediatric patients (especially those 12 years and younger)
compared with adults.[7]
[8]
[9] Even though a cohort of 9 untreated adult patients followed at 27 years reported
a mean Disabilities of the Arm, Shoulder, and Hand (DASH) score of 11.3, the authors
still concluded that observation of natural history was not a valid strategy.[10] Adult patients can go from first symptoms to full fragmentation in less than 6 months,
and adjacent hyaline cartilage degeneration can occur within 12 months.[11] Consequently, the decision analysis for the early Kienböck's patient lies between
observation of natural history versus surgical intervention with the intention of
promoting revascularization: pure unloading; direct revascularization without bone
replacement; indirect revascularization without unloading; indirect revascularization
with unloading. Spontaneous or indirect revascularization requires restoration of
perfusion through existing vascular channels running through the architectural framework
of the bone and will not occur across coronal plane structural fissures. Plain films
are unable to reveal such subtle cleavage planes, but development of these coronal
fissures within the bone is the key feature that signals the transition from very
early Kienböck's toward a worsening prognosis ([Fig. 1]). The presence of fissuring places a stronger emphasis on surgical intervention
because natural revascularization is less likely once the original vascular channels
have been interrupted. The necessary examination is a CT scan, as 33% of Lichtman/Stahl
(L/S) stage II patients may have a coronal fissure undetectable on plain radiographs.[12]
[13] If there is no cavitary bone loss seen on the CT scan (a feature of transitional
Kienböck's), then treatment is primarily aimed at revascularizing the lunate. Cavitary
bone loss requires vascularized replacement of missing bone stock, whereas a two-dimensional
coronal fissure merely creates a watershed separation of vascular pathways with no
requirement to replace bone tissue.
Fig. 1 Coronal plane fractures not visible on plane radiographs are demonstrated on sagittal
computed tomographic images.
Pure Unloading
The least invasive surgical option falls between the choices of observation and revascularization.
No vascular manipulations are performed. The scaphoid is extended and pinned across
the scaphotrapeziotrapezoid (STT) or scaphocapitate (SC) joint; external fixation
in distraction can also be used. By reducing the compressive loads on the lunate for
some planned period of time, the hope is that spontaneous revascularization will occur
through the existing vascular channels. This strategy would best fit a very young
patient whose parents desire to take some proactive step but who does not yet have
coronal fissures on the CT scan that interrupt the original channels. Unloading alone
is unlikely to promote the development of new vascular pathways (a response that has
yet to be anatomically confirmed in live patients). Normalization of MRI signal intensity
begins by 3 months after pinning and should be complete by 6 months (at which time
pins were removed).[14] Six patients with a mean age of 14 all normalized the MRI signal intensity with
3 to 6 months of pinning.[15] Six other patients with a mean age of 15 followed at 7 years all normalized the
MRI signal intensity.[16]
Direct Revascularization without Bone Replacement
In the absence of three-dimensional bone loss, there is no need to replace bone volume.
But with fissures interrupting the original vascular channels, new vascular pathways
will be required to achieve revascularization. Implantation of a vascular bundle into
the lunate has been shown to induce new vascular channels that can form connections
with the original intraosseous vascular pathways.[17] With the location of the fissure(s) demonstrated preoperatively, the bundle can
be placed across the fissure to reach the far segment of bone territory and deliver
new blood supply to a sector that has been structurally isolated from perfusion.[12] This strategy would best fit a patient with fissuring on CT scan but without imaging
features that would suggest greater risk of focal overload. Imaging features concerning
for overload include a high degree of ulnar negative variance, high radial inclination,
a wedge-shaped lunate narrow radially or dorsally, and more than 40% of the lunate
ulnar to the margin of the radius. Of interest, despite repeated historical assertions
that negative ulnar variance promotes the development of Kienböck's, critical examination
of the data, including multiple meta-analyses, concludes that the evidence fails to
support causation.[18]
[19]
[20]
Indirect Revascularization without Unloading
Once fissures have developed to interrupt the original vascular channels, revascularization
will be more likely through the induction of new vascular ingrowth to the lunate.
Operative intervention that violates bone structure adjacent to the lunate may induce
vascular ingrowth and has been proposed as the basis for the observed success of core
decompression of the radius.[21]
[22]
[23] Core decompression has also been reported for the capitate and even the lunate itself.[24]
[25] The indirect revascularization response may also be the basis for the observed success
of full radius osteotomy regardless of shortening, opening wedge, or closing wedge
varieties that all seem to produce about the same result as a core decompression.[26] This strategy best fits an early Kienböck's patient without imaging features indicating
potential for focal overload. Whether or not indirect stimulation of revascularization
can proceed equally well with or without fissures remains unclear as we do not know
whether or not revascularization utilizes existing pathways or forms new pathways.
Fifteen patients, L/S stage IIIA, followed at 13 years after radius core decompression
demonstrated wrist ROM 77% of contralateral and GS 80% of contralateral.[21] Twenty-two patients followed at 10 years after radius and ulna core decompression
demonstrated wrist ROM 77% contralateral and GS 75% contralateral.[22]
Indirect Revascularization with Unloading
This strategy seeks to influence both sides of the interplay between structural integrity
and viability/perfusion. By unloading the lunate from compressive forces, the goal
is to buy time for the revascularization process to occur before any further deterioration
or structural collapse. Beyond simple pinning of the scaphoid, unloading strategies
involve osteotomy of bones adjacent to the lunate that have been studied in the laboratory
to decrease relative compression across either a portion of the lunate (typically
the radial side) or the whole lunate. Radial shortening osteotomy (RSO) reduces radiolunate
load 45% but increases ulnolunate 50% and ulnotriquetral 78%.[27] Capitate shortening osteotomy reduces radiolunate load 66% but increases ulnotriquetral
114%, triquetrohamate 149%, STT 69%, and radioscaphoid 26%.[28] Full capitate osteotomy causes carpal collapse; partial capitate osteotomy increases
radioscaphoid load 39% while decreasing radiolunate 53%.[27]
[28]
[29]
Because these osteotomy procedures may also induce an indirect revascularization response,
it has been all but impossible to determine whether the observed clinical benefits
are attributable to the mechanical alteration of forces, the stimulation of vascular
ingrowth, or a combination. This will likely remain an unanswered question short of
a well-controlled study comparing osteotomy with core decompression. This strategy
would best fit a patient with imaging features indicating potential focal overload.
Since no one unloading strategy has been demonstrated superior to others, it makes
sense to tailor the method of unloading to the observed imaging features placing the
patient at risk. For a patient with high ulnar negative variance, a RSO balances the
compressive forces between the radial and ulnar portions of the lunate. Thirteen patients
followed for 21 years after RSO demonstrated that half progressed in stage (initial
stages: 6 stage II, 4 stage IIIA, 3 stage IIIB).[30] Eleven patients followed for 14.3 years after RSO were reported to have no progression
(2 stage IIIA, 8 stage IIIB, 1 stage IV).[31] Thirty-six patients starting at L/S stage IIIA were followed for 12 years after
RSO, 8 progressed to stage IV, and 7 progressed to stage IIIB.[32] In 16 patients followed for 25 years after RSO, 56% demonstrated no change in stage.[33] The most likely reason for these and other disparate results is a difference in
the initial status of the lunate as well as the inability of plain radiographs to
accurately represent lunate status at both the initial and final assessments.[34]
[35]
[36]
For a patient with high radial inclination or greater than 40% of the lunate off the
ulnar margin of the radius, a closing wedge osteotomy shifts more of the ulnar portion
of the lunate onto the radius and reduces radial sided overload.[37] Six patients demonstrated a lunate covering ratio shift from 61 to 90% without evidence
of osteoarthritis in the distal radial–ulnar joint (DRUJ) at 32 months.[38] Thirteen patients followed for 14 years demonstrated an inability of the radial
wedge osteotomy to prevent carpal collapse, with eight progressing in stage.[39] Comparisons of clinical outcomes between RSO, radial wedge osteotomy, or just cutting
the radius without changes in length or angle have found no differences.[26]
[40]
[41] A concern with any radius osteotomy that alters position is development of DRUJ
osteoarthritis.[42]
For a patient with multiple abnormalities or an ulnar positive variance, capitate
shortening osteotomy unloads the entire lunate approximately 50%.[43]
[44]
[45]
[46] Seven patients, L/S stage II and IIIA, followed at 38 months after partial capitate
osteotomy all demonstrated MRI revascularization.[47] Ten out of 19 patients followed at 16 months after partial capitate osteotomy demonstrated
complete revascularization.[48] Twenty-two patients followed at 30 months after capitate osteotomy demonstrated
that L/S stage II patients improved to a wrist ROM of 73% contralateral and GS 75%
of contralateral, but L/S stage III patients did not improve and further deteriorated.[49]
Transitional Kienböck's
The key feature of worsening Kienböck's is three-dimensional bone loss and corruption
of structural integrity to a greater extent than just thin fissures that interrupt
vascular channels ([Fig. 2]). This is the most difficult phase to assess because the inertia continues at a
variable and unpredictable rate for both the structural collapse and the progression
from merely ischemic bone onwards to a state of terminal necrosis. Spontaneous or
indirect revascularization is not going to occur for regions architecturally separated
from a source of perfusion by planes of cleavage in bone structure. Terminally necrotic
regions (that initially demonstrate trabecular architecture on imaging) will not remain
structurally intact over time. Absent a formal study, the clinical experience of high-volume
specialists is that terminally necrotic bone cannot be revascularized even by an adjacent
new vascularized bone flap.
Fig. 2 In transitional Kienböck's central cavitation precedes subsequent proximal subchondral
collapse.
The distinction between ischemic bone and terminally necrotic bone can only be confirmed
by direct inspection at the time of surgery, not by preoperative imaging (MRI or CT).
Ischemic bone still retains its porous cancellous architecture and retains some degree
of marrow. Terminally necrotic bone is absolutely dry and chalky, lacking any marrow
whatsoever, and appears visually sclerotic but is weak and friable to contact. The
specific bone stock, upon which treatment success or failure hinges, is the subchondral
bone. If the surgeon introduces new vascularized bone as a flap into the core of a
collapsing lunate, time is required for the flap to incorporate and heal to the ischemic
subchondral bone ([Fig. 3]). Despite trabecular patterns by preoperative imaging, the inertia of progressive
necrosis can proceed rapidly enough that the vascularized flap will have arrived too
late to salvage subchondral bone. For this reason, when planning for vascularized
reconstruction of the lunate, alternate treatments must be arranged in advance.
Fig. 3 Successful 4,5-extensor compartment pedicled, vascularized bone flap restores normal
signal intensity on magnetic resonance imaging.
Preoperative Evaluation and Planning
Substantial confusion exists among surgeons that MRI plays a role in treatment selection.
MRI is useful in two contexts: establishing an initial diagnosis of Kienböck's and
outcomes assessment of revascularization procedures. Patients must, of course, select
preferred treatments, but it is the surgeon's job to understand what treatments are
appropriate, along with specific features that eliminate an otherwise desired treatment
from consideration. This information is procured from two sources: CT scan and intraoperative
tissue assessment. There is no role for MRI. A surgeon relying only on the L/S classification
might easily become confused that arthroscopy is somehow a useful step in planning
treatment; it is not. Arthroscopy can demonstrate hyaline cartilage wear on the head
of the capitate and lunate fossa of the radius, but this only occurs after fragmentation,
collapse, and incongruence of the opposing surface of the lunate, a fact easily demonstrated
by CT scan (a noninvasive test that does not require anesthesia and a trip to the
operating room). Arthroscopy can demonstrate lack of subchondral support for hyaline
cartilage by direct probing of the surface, directing the surgeon away from a vascularized
flap that depends on integrity of that articulation. However, the converse is not
true: stability to arthroscopic probing of the hyaline cartilage surface does not
predict success of a vascularized flap joined to that section of subchondral bone.
The reason is that the probed bone may be terminally necrotic and unable to heal to
the flap. The same exact information can be obtained by CT scan. The key features
to observe on CT scan are dark lines running in the subchondral bone parallel and
close to the articular surface that indicate structural cleavage planes ([Fig. 4]). The cleavage planes indicate insufficient bone supporting the hyaline cartilage
to participate in a healing process with the adjacent flap. Like arthroscopy, the
absence of these lines on preoperative CT does not guarantee the success of a vascularized
flap, and the subchondral bone must always be assessed intraoperatively. Because intraoperative
fact checks remain, the preoperative planning must always include patient choice of
alternate procedure. This necessity of quality preoperative planning obviates any
role for advance arthroscopy.
Fig. 4 Proximal subchondral cleavage planes indicate loss of bone support to the hyaline
cartilage surface and preclude use of a core bone flap only for treatment.
Direct Revascularization with Core Bone Replacement
With loss of bone volume, treatment focuses on the dual objectives of revascularization
and replacing bone stock with pedicled or free vascularized bone flaps ([Fig. 5]).[50]
[51]
[52]
[53]
[54]
[55] For a successful purely osseous flap, both the proximal and distal articular surfaces
must be supported by subchondral bone that is merely ischemic but not terminally necrotic.[3]
[13]
[56]
[57] Twenty-six patients followed for 31 months after distal radius 4,5-extensor compartment
vascularized bone flaps achieved 71% of contralateral wrist ROM, 89% of contralateral
GS, and 77% demonstrated no further collapse.[58] Thirteen patients (11/13 L/S stage IIIA) followed at 32 months after distal radius
4,5-extensor compartment vascularized bone flaps without additional unloading demonstrated
no collapse or progression with GS improving from 60 to 88% of contralateral, wrist
flexion from 39 to 53 degrees, and wrist extension from 41 to 56 degrees.[59] However, if the inertia of the vascular decline causes critical subchondral bone
to progress to terminal ischemia before the bone flap has an opportunity to provide
revascularization, the procedure will fail. Also, if the inertia of the structural
collapse causes fragmentation in the subchondral region before revascularization occurs,
the procedure will fail.[60] Twenty-six patients followed for 62 months after vascularized bone graft with 4
months of pinning to unload the lunate still lost additional height if starting from
L/S stage IIIA; already collapsed L/S stage IIIB patients had no additional height
to lose.[61] These fundamental truths are what require the surgeon to see the case as occurring
through the fourth dimension of time and not just as a snapshot of its original presenting
features.
Fig. 5 Pedicled 4,5-extensor compartment bone flap reconstructs central cavitary loss.
Direct Revascularization with Replacement of Bone and Cartilage
The proximal half of the lunate is supplied retrograde by vessels entering distally
([Fig. 6]).[62] In addition, the trabecular structure predisposes the proximal lunate to collapse
prior to the distal articular surface.[63] The key features to observe on CT scan are cleavage planes parallel to the proximal
articular surface indicating insufficient subchondral support to incorporate a core
bone flap only, and the notable absence of such cleavage planes near the distal articular
surface ([Fig. 7]). The distal articular surface must either be intact or reconstructible to a sufficient
concave area capable of supporting the capitate. A common finding is fragmentation
of the dorsal horn of the lunate, but if the size of this fragment is small enough,
the remainder of the volar articular concavity is sufficient to receive load transfer
from the capitate ([Fig. 8]). If the distal articulation proves sufficient, then the remainder of the lunate
can be replaced by an osteocartilaginous free flap from the medial (or lateral) femoral
trochlea that reconstructs the proximal articular surface ([Fig. 9]). Sixteen patients followed at 19 months after medial trochlear flaps achieved 50 degrees
wrist extension, 38 degrees wrist flexion, and GS 85% of contralateral.[64] Two of the 16 progressed and worsened. Another cohort of 18 patients had only 8
available for follow-up at 2 years, yielding 37 degrees wrist extension, 38 degrees
wrist flexion, and GS 68% of contralateral.[65] Transferring vascularized autogenous tissue to reconstruct the collapsing lunate
is very conceptually appealing, but whether or not the long-term outcome holds any
advantage over alternatives remains to be proven.
Fig. 6 Intraosseous vascular channels enter the lunate from volar and dorsal, anastomose
in the distal portion of the bone, and then branch to supply the proximal lunate.
Fig. 7 Transitional Kienböck's proceeds to proximal subchondral collapse with an intact
distal articular surface.
Fig. 8 Even if the dorsal horn fragment is removed, as long as sufficient distal articular
surface remains to support the capitate, medial femoral trochlear reconstruction of
the proximal articular surface can still be performed.
Fig. 9 Healed medial femoral trochlear osteocartilaginous flap reconstruction of the entire
proximal lunate (not same patient as Fig. 8).
Late Kienböck's
Late Kienböck's is characterized by the features of extensive fragmentation and terminally
necrotic bone such that neither the proximal nor distal articular surfaces are capable
of reconstruction or healing to any newly imported vascularized bone flaps ([Fig. 10]). Patients in transitional Kienböck's may also reject flap reconstruction strategies
for personal reasons or be unsuitable candidates based on medical comorbidities. Those
patients may then elect to proceed to the treatment options associated with removal
of the entire lunate: total wrist fusion, partial wrist fusion, proximal row carpectomy
(PRC), PRC with capitate prosthesis, and prosthetic replacement of the lunate with
carpal stabilization. These treatment options also serve as alternate strategies to
planned vascularized bone flaps following intraoperative assessment for terminally
necrotic subchondral bone.
Fig. 10 Once both articular surfaces have fully collapsed and fragmented, the lunate can
no longer be reconstructed.
Total Wrist Fusion
Total wrist fusion (TWF) is typically rejected by most patients who are loathe to
sacrifice all motion. Despite that one major drawback, TWF otherwise offers the advantages
of a durable solution that will never have to be converted, the best pain relief,
and the ability to apply loads without reservation. The key patient feature that best
corresponds to TWF is a young patient that desires to continue high force impact loading
for many decades.
Partial Wrist Fusion
The key patient features that best correspond to partial wrist fusion are a patient
who desires to load the wrist without reservation, has high pain tolerance, and is
willing to experience moderate loss of motion and a moderate likelihood of subsequently
converting the partial fusion after progression of radioscaphoid arthritis. Either
STT or SC fusion permits load to be transmitted through the radioscaphoid articulation.[66]
[67] However, with the inability for the scaphoid to flex relative to the distal row,
the proximal scaphoid is forced to sublux incongruently, leading to accelerated wear
with the radius.[68]
[69] Trapeziometacarpal arthritis, carpal collapse, poor motion, and ulnar translocation
after lunate removal have also been reported.[68]
[70]
[71] Fifty-nine patients followed 4 years after STT fusion demonstrated a flexion–extension
arc 60% of contralateral, radial–ulnar deviation arc 52% of contralateral, and a mean
DASH score of 28.[72] Eleven patients with STT fusion were compared with 19 natural history patients over
13 years with worse results for both pain and motion in the fusion group.[73] Direct comparison of STT fusion to PRC found inferior results for the fusion with
an arc of motion 39% contralateral versus 57% and GS 62% contralateral versus 68%.[74]
Proximal Row Carpectomy
Perhaps the most commonly performed treatment for late Kienböck's is PRC. The key
anatomic feature to support this option is preservation of hyaline cartilage on the
head of the capitate and the lunate fossa of the radius ([Fig. 11]). The key patient feature is a low-to-moderate demand middle aged adult. With the
three bones of the proximal row removed, the shorter radius of curvature capitate
articulates incongruently with the larger radius of curvature lunate fossa. Even if
both surfaces begin with pristine hyaline cartilage, the high load concentration leads
to accelerated wear and eventual loss of the remaining cartilage. The younger the
patient, the higher demand user, and any existing deficiencies on either surface create
a higher likelihood of an eventual conversion or worse outcome.[75]
[76]
[77] Thirteen patients followed for 15 years after PRC all demonstrated degenerative
arthritis but without correlation to subjective good to excellent clinical results
in 12/13, a mean arc of motion 73% of the uninvolved side, with 7/13 in manual labor
positions.[78] Twenty-four patients followed 116 months after PRC demonstrated a motion arc of
76 degrees, GS 78% contralateral, and a DASH score of 31.[79] Systematic review of six studies capturing 147 patients yielded a GS 68% of contralateral,
arc of motion 73 degrees, DASH of 21.5, patient-rated wrist evaluation (PRWE) of 28.7,
osteoarthritis in 79%, and 14% converting to TWF.[80] Followed over a period of 12 years, 24% of 62 patients converted to a TWF, but did
so at a mean of 22 months following the original PRC.[81]
Fig. 11 After proximal row carpectomy the shorter radius of curvature head of capitate articulates
with the larger radius of curvature lunate fossa.
PRC with Capitate Prosthesis
Patients may prefer PRC over other strategies but face cartilage surface deficiencies
too great to predict an acceptable outcome from a conventional PRC. In such cases,
a pyrocarbon prosthesis can be placed at the radius to capitate articulation with
a report of no implant failures or dislocations in six patients followed for 28 months.[82] Twenty-five patients were followed for over 2 years demonstrating 27 degrees of
wrist flexion, 33 degrees of wrist extension, GS 54% of contralateral, DASH of 20,
and PRWE of 28.[83] An alternative design offers an osseointegrated titanium threaded post fit with
a modular cobalt-chrome articular surface replacing the head of the capitate in multiple
sizes that can be optimized for congruence with the lunate fossa ([Fig. 12]). Published data are lacking so it remains unknown whether the higher modulus of
cobalt-chrome will ultimately lead to erosion into the radius and the level of discomfort
patients may experience from this articulation. Currently, without evidence of performance
over time, this model may be better suited for older patients with a shorter exposure
to future loading.
Fig. 12 The head of the capitate is resurfaced with a modular cobalt–chrome prosthesis fixed
with a titanium threaded post.
Prosthetic Replacement of the Lunate with Carpal Stabilization
The key features are a patient who personally rejects the other more traditional options
for late Kienböck's. A prosthesis made of pyrocarbon and shaped similar to the native
lunate has the positive features of an extremely low friction surface that can articulate
successfully, even with eburnated bone, as well as a modulus of elasticity similar
to cortical bone for load transfer. These two attributes are titrated against its
primary detriment of the absolute inability to heal to any tissue. The prime attraction
for patients is not removing or fusing any of the other intact carpal bones. But this
same attraction is exactly what poses the greatest problem, the remaining carpals
are left intrinsically unstable following the loss of scapholunate and lunotriquetral
ligament integrity. An earlier model made of titanium (performed without carpal stabilization)
demonstrated a 20% rate of complete dislocation and the authors discontinued the procedure.[84] Since nothing will adhere to pyrocarbon, establishing sufficient intercarpal stability
requires forming new soft tissue connections between the scaphoid and triquetrum that
bypass the prosthesis ([Fig. 13]). Additional reconstruction, connecting to the radius, may also be needed to resist
ulnar translocation. Thirteen patients, implanted and stabilized by double bundle
scaphoid-trapezium flexor carpi radialis tenodesis, and followed for 30 months demonstrated
improvements by all measures to 43 degrees wrist flexion, 53 degrees wrist extension,
GS 85% contralateral, and a DASH score of 7.7.[85] Maintenance of stability over time showed no loss of surgical improvement achieved
in the scapholunate angle (46.4 degrees), radioscaphoid angle (45 degrees), and modified
carpal height ratio (1.59). One patient was unable to be rendered stable and intraoperatively
converted to the alternate choice of PRC. One patient developed avascular necrosis
of the proximal scaphoid and chose conversion to PRC. The authors reported that the
first-generation stabilization technique used in these13 patients has since been modified.[85]
Fig. 13 Carpal stability has been restored through tenodesis reconstruction, bypassing the
pyrocarbon lunate prosthesis.
Evaluating Outcomes
Without more discriminating assessment of lunate status either before or after treatment,
it is not surprising that systematic reviews have failed to identify any superior
treatment for Kienböck's.[5] Uncertainty even remains as to whether surgical treatment confers advantages over
natural history.[73] Multiple individual studies and a systemic review of 17 studies with greater than
10 years follow-up have demonstrated that RSO does not reverse or prevent collapse.[86]
[87]
[88]
[89] Twenty-five patients with RSO demonstrated osteoarthritis in 54% by 5 years and
73% by 14.5 years.[90] Forty-four nonsurgically tracked patients reported a late DASH score of 20 compared
with 18 surgically treated patients with a score of 23.4.[6] Although more recent studies are starting to incorporate patient-rated outcome measures
(PROMs), the lunate itself has been assessed using the imprecise tool of plain radiography
that fails to adequately demonstrate lunate architecture (CT scan) or vascular status
(MRI).[3]
[35] As the lunate collapses, venous congestion is alleviated, forces are offloaded to
adjacent articulations, and the residual fragments have an opportunity to revascularize
and heal in some patients or continue to further displace in others.[91]
[92] The objective physical measures of GS and ROM are compromised to a similar moderate
extent in the course of Kienböck's reconstruction and thus do not prove very useful
to separate one procedure from another. The subjective but quantifiable results of
PROMs also tend to cluster within a range relatively equal to the minimal clinically
important difference. One potential reason for PROM clustering may be the adaptability
of human beings. After counseling by their surgeons that the reconstruction should
be preserved by avoiding excessive loading, most patients make lifestyle changes and
adapt activity patterns to accommodate their new wrists. They also learn to accept
certain symptoms and limitations because they know that few alternatives exist. The
detracting features of those alternatives are likely the reason that more patients
do not convert their initial procedures even in the face of moderate results. Although
we are unlikely to alter these fundamentals despite our efforts to advance our field,
we may be able to offer patients more individualized guidance in selecting Kienböck's
procedures.
Summary
We can best help our patients with a clear explanation of the nature of the disease
as a continuous state of flux between perfusion and architectural compromise occurring
in three phases: early, transitional, and late. This is much easier for patients to
understand than static radiographic stages that fail to elucidate key features that
correlate with specific treatment choices. Skeletally immature patients, especially
those 12 years and younger, have demonstrated a superior capacity for revascularization;
best evidence supports treatment by observation of natural history with or without
pinning. Evidence for spontaneous revascularization and resolution to a normal lunate
is lacking in adults. Unless we employ more discriminating research methods to objectively
define the status of the lunate both before and after intervention, we will remain
unlikely to discover optimal treatments for Kienböck's patients. While it is indeed
important to include plain radiographs, PROMs, ROM, and GS measures in our research,
the objective of our surgical interventions (for early and transitional Kienböck's)
is the lunate itself, specifically its vascularity and architecture. Only CT scan
adequately defines the architecture, and only MRI provides noninvasive information
regarding vascular status. Given the abundance of literature confirming that osteotomy
is no better than natural history, treatment selection should at least minimize invasiveness
and morbidity such that direct revascularization with or without core decompression
of the lunate seems appropriate for early Kienböck's. Transitional Kienböck's creates
logical pairings for treatment: replace missing core bone or replace missing bone
and cartilage with vascularized tissue. Late Kienböck's requires perhaps the most
complex decision making on the part of the patient, guided by which detracting features
the patient is willing to accept and manage over time. Limited wrist fusion has been
demonstrated no better than natural history and worse than PRC, an unlikely choice
for most patients. The historically most common operation, PRC, remains the reference
standard, but progressive loss of cartilage is assured in young patients who increasingly
seek alternatives. Unfortunately, only limited and early data exists for the prosthetic
alternatives, with only TWF remaining.
