89
Unlinked and Convertible Arthroplasty:
Design, Concept, and Technique
Graham J.W. King
INTRODUCTION
Total elbow arthroplasty (TEA) is indicated for pain and functional
limitations secondary to inflammatory arthritis, osteoarthritis, post-
traumatic arthritis, acute distal humeral fractures, distal humeral
nonunions, tumors, osteonecrosis, and dysfunctional instability. Cur-
rently available prostheses can be subclassified as linked, unlinked, or
convertible implants that can be interchanged between an unlinked or
a linked design. This chapter will focus on the design considerations,
techniques, and results of unlinked and convertible TEA.
DESIGN CONSIDERATIONS
Polyethylene wear and aseptic loosening of linked TEAs is a significant
problem with longer follow-up, particularly in younger and more
active patients.a When a linked arthroplasty reaches the limits of its
intrinsic laxity, forces are transmitted to the linkage mechanism and
thereby to the stems and the bone implant interface, contributing to
bearing wear and implant loosening.34 The concept of unlinked TEA
is that there is a sharing of load between the prosthesis, capsule, liga-
ments, and muscles in an effort to decrease the stresses in the articula-
tion and stems. A necessary prerequisite for an unlinked arthroplasty
is the presence of adequate ligaments and bone stock with which the
implant can be stabilized after insertion. Unlinked total elbows were
initially referred to as resurfacing devices primarily because most were
designed without a humeral stem. Due to the high incidence of implant
loosening with these early designs, all unlinked prostheses that became
widely used incorporated stems on the components in an effort to
improve implant longevity (Fig. 89.1). See Chapter 85 for the history
of elbow arthroplasty.
Unlinked TEAs are sometimes referred to as unconstrained devices,
whereas loose-hinge linked devices are often referred to as being
semiconstrained; however, this is a misnomer. All joint arthroplasties
have some element of intrinsic constraint by virtue of the shape and
interaction of their articular surfaces. The intrinsic constraints of the
ulnotrochlear joint of five unlinked TEA designs were compared and
found to vary markedly among implant types.22 Surprisingly, the
implants that more closely resembled the articular geometry of the
native articulation tended to less accurately replicate the intrinsic
constraint of the human elbow (Fig. 89.2). In a series of cadaveric
studies, the kinematics and stability of unlinked and convertible TEAs
have been evaluated.b The mechanical behavior of these prostheses has
been found to vary markedly depending on their articular design.
Linking an unlinked convertible total elbow has been shown to increase
the stability as would be expected.9
a
References 15, 19, 20, 36, 39, 46, 52.
b
References 2, 17, 18, 24, 40, 45, 58.
842
In an effort to improve their stability and load transfer, some
unlinked designs have incorporated a radial head replacement. The
radial head component has been shown to improve the stability of
unlinked elbow prostheses in in vitro biomechanical studies.17,40,59
While theoretically improving soft tissue balance and load transfer
across the elbow, the experience with radial head components in TEA
has been variable (Fig. 89.3). For example, the radial head component
of the capitellocondylar prosthesis was abandoned due to concerns
about radiolucent lines around the implant stem, but a subsequent
study reported that the outcome was better if a radial head component
was utilized.54 The Pritchard elbow resurfacing system (ERS) total
elbow prosthesis was reported to have a higher failure rate if the radial
head was not replaced, however, technical issues with positioning the
radial head component were common. Preservation of the native radial
head provided the longest functioning implants.38 One possible expla-
nation of the variable results with radial head prostheses in unlinked
implants to date may be that in many cases, their designs have been
suboptimal.56 Furthermore, the addition of a third component com-
plicates the surgical technique, making it more challenging to perform
correctly, particularly since the instrumentation available to accurately
position unlinked total elbows has been lacking relative to other joints
such as the hip and knee. Incorrect positioning of a radial head com-
ponent will result in eccentric motions during prosupination, and
therefore shear forces across the radiocapitellar articulation. These
abnormal kinematics and loads may cause polyethylene wear and early
loosening of the radial component. A bipolar radial head component
may be more forgiving in this setting as it will compensate for some
of the eccentric motions through the bipolar articulation but may be
prone to implant dissociation.59
Functional and balanced collateral ligaments are a key feature of a
successful unlinked TEA. Incompetent ligaments will result in articular
maltracking, joint subluxation, or dislocation. This has been demon-
strated in in vitro cadaveric studies and observed clinically.18,24 Mal-
tracking of the arthroplasty will lead to excessive polyethylene wear,
osteolysis, and aseptic loosening. During the insertion of most unlinked
designs, one or both of the collateral ligaments are usually released to
achieve adequate exposure. Careful repair and successful ligament
healing are needed if an unlinked arthroplasty is to be stable and to
articulate normally.
Optimal component positioning improves the outcome of an
unlinked TEA.47,53 Replication of the axis of motion to the native state
restores the muscle moment arms and the soft tissue tension of the
collateral ligaments, which ensures more normal articular tracking and
implant loading. Malpositioning of the humeral component has been
demonstrated to alter the in vitro kinematics and stability of the capi-
tellocondylar arthroplasty.18 It has also been reported that correct
medial–lateral and proximal–distal positioning of the Souter implant
lowered the incidence of loosening.47 Another study evaluating the
same implant, however, was unable to demonstrate a relationship
 CHAPTER 89 Unlinked and Convertible Arthroplasty: Design, Concept, and Technique
FIG 89.1 Five-and-a-half years after insertion of an unstemmed humeral
component, a fracture has occurred across the distal humeral condyles
at the site of fixation of a 39-year-old patient with rheumatoid arthritis.
(Used with permission of Mayo Foundation for Medical Education and
Research. All rights reserved.)
between humeral component position and aseptic loosening.55 Current
techniques to identify and replicate the flexion–extension axis are
prone to error. In one study, the surgeon’s ability to select the flexion–
extension axis resulted in errors in axis orientation of up to 10 degrees
in normal humeri; these errors are likely greater in clinical practice
where the anatomy is more distorted and the exposure of landmarks
is compromised.3 Computer and/or image-assisted surgery has been
shown to improve the accuracy of elbow component placement.
However, this technology is currently not available for clinical use.32
While many patients with elbow arthritis can be managed with an
unlinked elbow arthroplasty, there are patients who require a linked
prosthesis due to a lack of adequate bone stock or functional collateral
ligaments. While typically this can be predicted preoperatively, it is
always desirable to have a linkable implant immediately available in
the operating room whenever a surgeon is planning to perform an
unlinked TEA. Intraoperatively, if an unlinked implant is not stable or
is not articulating congruously, then it should be converted to a linked
design. If this is recognized before the unlinked components are
cemented then a different design of linked arthroplasty can be used,
although repeat bony preparation will be needed, which is time-
consuming. The advantage of a convertible prosthesis is significant in
this setting, as rather than changing prosthesis designs, a linkage
mechanism can simply be applied to the same prosthesis for which the
bony resection has been completed. If the stems have already been
cemented then the ability to couple an unlinked prosthesis without
removing the stems makes a convertible implant system even more
appealing. The advantage of having one implant system to manage the
majority of clinical situations is also helpful for hospitals by reducing
inventories and cost. For surgeons, a convertible design means only
one implant system needs to be mastered; the implant can be routinely
used as a linked device, and in selected younger or more active patients
with good bone stock and ligaments the device can be used in the
unlinked configuration.
843
While most patients who undergo an unlinked TEA have an excel-
lent outcome, there are some patients who develop elbow instability
postoperatively, which is often difficult to manage as discussed in
Chapter 105. When instability is recognized early, a closed reduction
and a period of immobilization may allow the ligaments to heal and
the prosthesis to function normally. However, when presenting as a
chronic problem, attempts at ligament reconstruction have not been
reliable.43 The concept of having an unlinked elbow arthroplasty that
can be linked in the setting of postoperative elbow instability is desir-
able, as removal of a well-fixed unlinked elbow arthroplasty with
revision to a linked arthroplasty is technically difficult, with a signifi-
cant risk of complications.43
Hemiarthroplasty of the distal humerus is becoming more popular
to manage comminuted articular fractures of the distal humerus,
nonunions, and avascular necrosis as discussed in Chapter 88. Should
the patient develop instability, ulnar subsidence, or have persistent
pain, the ability to convert a distal humeral hemiarthroplasty to either
a linked or unlinked TEA without removing a well-fixed humeral
component, is attractive.
By virtue of their need for adequate ligaments and bone stock, the
indications for unlinked designs are narrower than linked arthroplas-
ties and the instability rate is greater. While the loosening rate should
theoretically be lower with unlinked versus linked prostheses, this has
yet to be proven in clinical studies. With the advent of improved
linked-elbow arthroplasty designs and a lack of compelling evidence
to support the superior outcome of unlinked implants in the literature,
the popularity of unlinked implants has decreased in recent years.
Furthermore, with the development of improved biologic agents, the
incidence of patients with rheumatoid arthritis requiring elbow
replacements has decreased, which was the principal indication for
an unlinked device in the past.12,13,19 Posttraumatic conditions and
acute distal humeral fractures—disorders that usually require a linked
device—are currently the most common indications for a TEA. As a
result of these changing disease patterns and suboptimal long-term
outcomes of many older unlinked designs, most are no longer com-
mercially available.
INDICATIONS
The indications for an unlinked or convertible TEA are similar to those
for any design of elbow arthroplasty. The necessary prerequisites to
choose an unlinked implant or use the unlinked version of a convert-
ible implant are the presence of sufficient bone stock to support the
implant and intact medial and lateral collateral ligaments. The pres-
ence of functioning elbow flexor and extensor muscles is particularly
important if the implant is to be unlinked, as muscle function is
important for stability. In the absence of sufficient bone stock or liga-
ments, a linked arthroplasty or the linked version of a convertible
arthroplasty should be used. Typical indications include inflammatory
arthritis (rheumatoid, psoriatic, and hemophilic), primary or post-
traumatic osteoarthritis, osteonecrosis, periarticular tumors, and
comminuted distal humeral articular fractures in elderly patients.
CONTRAINDICATIONS
Active infection is an absolute contraindication to elbow arthroplasty.
The absence of functioning muscles to move the elbow and a nonfunc-
tional hand are relative contraindications. Poor skin quality must be
corrected before or at the same time as the arthroplasty. Unstable
elbows, with poor ligaments or deficient bone, should be managed
with a linked arthroplasty. Gross deformity is a contraindication for
an unlinked device as achieving precise soft tissue balancing is typically
844 PART VIII Joint Replacement Arthroplasty

Native Souter Pritchard

Sorbie Kudo
Capitellocondylar

A
300 Valgus 14
12
200 10

Angle displacement (degrees)

8
6
100 4
Torque (Ncm)

2
0 0
–2
–100 –4
–6
–8
–200 –10
Human Souter Kudo Capitello- Sorbie Pritchard
–12
Varus condylar
B –300 –14
FIG 89.2 (A) The intrinsic constraint of the ulnohumeral joints of five different unlinked total elbow arthroplas-
ties were compared to that of the native elbow joint in an in vitro biomechanical study. (B) The torque (bar
graphs) and angular displacement (line graphs) of the Souter and Kudo implants were most similar to the
human elbow. (From Kamineni S, O’Driscoll SW, Urban M, et al: Intrinsic constraint of unlinked total elbow
replacements—the ulnotrochlear joint, J Bone Joint Surg 87A:2019, 2005.)

not possible. As for any TEA, patients who are unwilling to live within
the activity and weight restrictions that are thought to be needed for
implant longevity should be managed with an alternative treatment.
SURGICAL CONSIDERATIONS
OF UNLINKED DESIGNS
There are a number of unlinked designs that have been developed and
employed in the past; however, only one remains widely available. The
iBP TEA (Biomet, Warsaw, IN) has evolved significantly since it was
first introduced as the Kudo device.29 Initially the Kudo prosthesis had
an all-polyethylene ulnar component and an unstemmed humeral
component with a cylindrical articulation. The humeral articulation
had a saddle-shaped design that allowed for medial–lateral translation
of the ulnar component. The prosthesis was implanted with sectioning
of the collateral ligaments and excision of the radial head. Due to a
high incidence of early loosening, both the humeral and ulnar com-
ponents were redesigned (Fig. 89.4). The Kudo type-3 prosthesis had
a stainless steel humeral component and an all-polyethylene ulnar
component. Due to ongoing concerns about humeral loosening and
stem fracture, further design modifications were made. The Kudo
type-5 prosthesis employed a cobalt–chrome humeral component with
a porous titanium coating that was typically inserted uncemented
and an all-polyethylene or metal-backed cemented ulnar component
(Fig. 89.5). The results with the metal-backed ulnar component
were better.49 The currently available iBP device has a cobalt–chrome
 CHAPTER 89 Unlinked and Convertible Arthroplasty: Design, Concept, and Technique
FIG 89.3 The theoretical advantages of distributing forces that pass
across the elbow through both columns of the humerus are obvious.
(Used with permission of Mayo Foundation for Medical Education and
Research. All rights reserved.)
A B
FIG 89.5 (A,B) The Kudo type-5 total elbow arthroplasty had a stemmed
component and a metal-backed ulnar component with better reported
outcomes.
845
FIG 89.4 The original Kudo arthroplasty had a stemless humeral
component and all-polyethylene ulnar component, which was prone to
loosening.
FIG 89.6 The iBP (Biomet, Warsaw, IN) total elbow arthroplasty has a
cobalt–chrome humeral stem and a titanium ulnar component with an
argon-packaged compression molded polyethylene bearing.
humeral stem with a roughened surface to promote cement bonding
and a titanium ulnar component with an argon-packaged compression-
molded polyethylene bearing (Fig. 89.6).10,12,25
SURGICAL CONSIDERATIONS
OF CONVERTIBLE DESIGNS
The Latitude EV system (Wright Medical, Memphis, TN) is the only
currently available convertible implant. The system is modular with
the option to be configured as a distal humeral hemiarthroplasty and
an unlinked or linked TEA (Fig. 89.7). The original Latitude system
had smooth cobalt–chrome stems; the Latitude EV has cobalt–chrome
stems proximally coated with titanium plasma spray to improve
846 PART VIII Joint Replacement Arthroplasty
A B
FIG 89.7 The Latitude EV elbow arthroplasty system (Wright Medical, Memphis, TN) has the option to be
used (A) unlinked, (B) linked, and (C) as a distal humeral hemiarthroplasty. The cobalt–chrome stems have
titanium plasma spray to improve cement fixation. (Reproduced with permission from Wright Medical,
Memphis, TN.)
cement fixation.16 The modular humeral component incorporates fins,
a rectangular shape, and an anterior flange to resist axial rotation and
posterior displacement.23 It also gains fixation on the lateral column
of the distal humerus by virtue of the capitellar portion of the articula-
tion. The axis bolt of the humeral component is cannulated to facilitate
secure stable collateral ligament repair to the implant and adjacent
bone. The ulnar component is metal-backed with thick polyethylene
and an extended coronoid process to resist dislocation. The stem has
a fin and a square cross-sectional shape to resist torsion; it also gains
fixation in the greater sigmoid notch.23 An optional bipolar radial head
is available to load share and improve stability when the radial head is
arthritic and requires excision.59 The conversion between an unlinked
and linked device is accomplished by adding a locking cap during the
initial surgery or subsequently though a minimally invasive approach
to manage postoperative instability.
SURGICAL TECHNIQUE
The precise surgical technique and approach depend on the design of
the prosthesis selected and the surgeon’s preference. It is recommended
that a linked implant system be available in the operating room when-
ever an unlinked prosthesis is planned. This will allow conversion to a
linked design if the elbow is unstable or the articulation is maltracking
following trial placement of unlinked components. Alternatively, a
convertible prosthesis can be routinely employed when an unlinked
prosthesis is planned.
The author’s preferred surgical technique for the Latitude EV TEA
is outlined.14 Place the patient supine with the arm across the chest.
Use a regional or a general anesthetic. Administer prophylactic antibi-
otics. Use a sterile tourniquet after routine skin preparation and
draping to enlarge the sterile field in an effort to reduce the risk
of infection. Make a midline posterior elbow incision just medial to
the tip of the olecranon, and elevate full thickness flaps as necessary
on the deep fascia. Identify, mobilize, and transpose the ulnar nerve
anteriorly.
C
Management of the triceps tendon varies according to the prefer-
ence of the surgeon. The implant can be placed by elevating the triceps
from medial to lateral, as in the Bryan-Morrey approach; from lateral
to medial, as in the extended Kocher approach; or through a triceps-
splitting approach where the triceps is elevated from the olecranon
both medially and laterally, using a triceps tongue, a paratricipital,1 or
a lateral paraolecranon approach.5,21,35,48,61 The design of the instru-
mentation allows for the bony preparation to be performed and the
implants to be inserted while preserving the attachment of the triceps
to the olecranon. Leaving the triceps attached to the olecranon reduces
the risk of triceps insufficiency with postoperative weakness and pain,
which can occur when the triceps tendon is detached to perform an
elbow arthroplasty.7,8,35 Unfortunately, the paratricipital approach
compromises visualization of the proximal ulna, but it can be a useful
technique, particularly when there is distal humeral condylar bone loss.
The lateral paraolecranon approach is currently preferred by the
author as it represents a compromise between visualization of the ulna
and preservation of triceps strength.48
Identify the lateral margin of the proximal ulna, and release the
anconeus, leaving a small cuff of fascia on the ulna for later repair.
Extend this along the lateral margin of the olecranon, and split the
distal triceps tendon centrally. The medial half of the triceps remains
attached to the olecranon, and the lateral half remains attached to the
fascia of the anconeus (Fig. 89.8). Detach the medial and lateral col-
lateral ligaments and their corresponding common flexor and extensor
origins from the humeral epicondyles, and tag them with a suture to
facilitate repair. Elevate the anterior and posterior capsule from the
humerus to allow dislocation of the elbow.
Select the correct implant size by matching the anatomical spools
with the articular surfaces of the proximal ulna, radial head, and
distal humerus where available. The key to sizing the Latitude EV
system is choosing the correct joint width such that with the spool
sitting in the trochlear groove, the radial head articulates congruously
with the capitellum of the spool (Fig. 89.9). This will ensure that
the native radial head (or a radial head component) will articulate
 CHAPTER 89 Unlinked and Convertible Arthroplasty: Design, Concept, and Technique 847

A B

C D
FIG 89.8 Lateral paraolecranon surgical approach. (A) Posterior skin incision just medial to olecranon.
Anterior transposition of ulnar nerve. Split the triceps tendon centrally, and extend the split along the lateral
margin of the ulna. (B) Section collateral ligaments and common flexor and extensor origins off the epicon-
dyles to allow dislocation of the elbow. (C) Rotate the forearm to allow visualization of the ulna and radius
for preparation. (D) Dislocate the humerus medial or through the triceps split to allow for humeral preparation.
(From Studer A, Athwal GS, MacDermid JC, Faber KJ, King GJ: The lateral para-olecranon approach for total
elbow arthroplasty. J Hand Surg Am 38:2219.e3, 2013.)

congruously with the capitellum of the humeral component. There are

A B
FIG 89.9 Implant sizing. (A) Size total elbow arthroplasty (TEA) using
spool in trochlear groove, capitellar portion should line up with the center
of the radial head. (B) Size TEA using width of distal humeral articula-
tion. (Reproduced with permission from Wright Medical, Memphis, TN.)
four prosthesis sizes available for the TEA and six sizes available for
anatomic distal humeral hemiarthroplasty. Short, long, and revision
length components are available.
The author prefers to prepare the radius and ulna first, to avoid
fracturing the humeral condyles while manipulating the forearm to
improve exposure. Seat the anatomic spool into the articular dish of
the radial head and the guiding ridge of the greater sigmoid notch of
the olecranon. Attach the ulnar cutting guide, and secure to the flat
posterior surface of the olecranon (Fig. 89.10). Radial head preserva-
tion is preferred if it is not too deformed. Use a sagittal saw to excise
the radial head when necessary. Using the bell saw, make the trochlear
cut from medial to lateral while protecting the ulnar nerve and radial
head. Open the ulnar canal with a burr, and enlarge it using the sup-
plied flexible reamers and rasps. Insert the trial ulnar component using
the rotational orientation of the flat spot of the posterior olecranon
and the relationship to the radial head.4,11 If the radial head is to be
replaced, broach the canal and attach the trial radial head component
to the broach. The bipolar radial head accommodates for ±10 degrees
of angular rotation but cannot substitute for malalignment of the
proximal radius with the capitellum.
848 PART VIII Joint Replacement Arthroplasty

A B C

D E F G
FIG 89.10 Ulnar preparation. (A) Insert spool into greater sigmoid notch, and align with radial head. Slide
ulnar cutting guide into spool from lateral to medial. Align cutting guide carefully with radial neck and secure
in place. (B) Excise radial head using oscillating saw if desired. (C) If radial head is to be retained, insert radial
head protector, and perform ulnar cut using bell saw. (D) Prepare medullary canal of ulna using flexible
reamers and rasps. (E) Orient rasp parallel to flat spot of the olecranon, and insert ulnar component. (F) If
radial head is to be replaced, rasp proximal canal. (G) Leave rasp in place, and attach trial bipolar head.
(Reproduced with permission from Wright Medical, Memphis, TN.)

Remove the central intercondylar portion of the distal humerus
with a saw to gain access to the proximal aspect of the olecranon fossa
(Fig. 89.11). Open the medullary canal with a burr, and enlarge the
diameter using flexible reamers. Ensure the humeral rasps are inserted
in the correct rotational orientation using an alignment rod, and to the
correct depth using a laser marking. As for any TEA, reproduction of
the humeral flexion–extension axis is crucial for a successful result.18,47
The anterior–inferior aspect of the medial epicondyle, if present, is a
reliable landmark for the rotational orientation as well as the insertion
depth of the component. If the humeral condyles are fractured or
deficient, the yoke of the component should be inserted to the depth
of the proximal olecranon fossa and internally rotated 14 degrees rela-
tive to the posterior flat spot of the distal humerus.44 Attach the
humeral cutting block to the final rasp and use a saw to complete the
distal humeral preparation. Insert the trial humeral component.
Move the elbow through a range of motion (ROM) with the trial
components in place, evaluating articular tracking and stability. An
unlinked arthroplasty may be considered if there is preservation of the
condyles and collateral ligaments, and the elbow prosthesis is tracking
well. Reposition the components if the radial head prosthesis is mal-
tracking on the capitellum. If the malalignment cannot be corrected,
a radial head replacement should not be performed, and the implant
should probably be linked, as instability may be more common if the
lateral column cannot be reconstructed. Link the implant in the setting
of bone loss or insufficient collateral ligaments.
Insert cement restrictors, irrigate the medullary canals with pulsa-
tile lavage, and carefully dry them using narrow nozzle suction. Inject
antibiotic cement into the medullary canals, insert the components,
and reduce the elbow, making sure to maintain the correct component
alignment as the cement sets. Place a cancellous bone graft behind the
 CHAPTER 89 Unlinked and Convertible Arthroplasty: Design, Concept, and Technique 849

A B C
FIG 89.11 Humerus preparation. (A) Excise center portion of trochlea to access humeral canal. Prepare
medullary canals using flexible reamers and rasps. Use alignment indicator on the rasp to insert the compo-
nent to the correct depth using the anterior–inferior portion of the medial epicondyle as a guide. (B) Orient
the rasp to align with flexion–extension axis. (C) Attach cutting block to final rasp, and resect along margins.
(Reproduced with permission from Wright Medical, Memphis, TN.)

A B C
FIG 89.12 Trial reduction. (A) Insert trial components, reduce elbow, and evaluate tracking. (B) If linking
implant, flex the elbow, and slide the locking cap in place. (C) Insert and secure screw with torque screw-
driver. (Reproduced with permission from Wright Medical, Memphis, TN.)

anterior flange of the humeral component after the cement has cured.
Evaluate the articular tracking of the components and the stability of
the elbow (Fig. 89.12). If the prosthesis is to be linked, insert the ulnar
cap, and secure the screw using the torque driver.
Repair the collateral ligaments using nonabsorbable sutures passed
through the cannulated screw that is located at the axis of rotation of
the implant (Fig. 89.13). When tied on the contralateral side of the
elbow, these sutures securely approximate the ligaments to their attach-
ment sites while avoiding problems with sutures pulling through the
often weak bone of the epicondyles. Tie the ends of the ligament
sutures over or through a drill hole in the proximal ulna to function
as a temporary “artificial ligament” to prevent instability in the post-
operative period, while the collateral ligaments are healing. Perform a
side-to-side repair of the triceps using nonabsorbable sutures with
buried knots if a lateral paraolecranon approach was employed. Reat-
tach the triceps to the olecranon using transosseous drill holes, if it
was detached. Secure the ulnar nerve in the anteriorly transposed
position. Place a suction drain overnight, and close the wound in
layers. Apply a well-padded splint with an anterior fiberglass slab in 40
degrees of extension.
850 PART VIII Joint Replacement Arthroplasty

A B
FIG 89.13 Ligament repair. (A) The medial and lateral collateral ligaments are repaired using nonabsorbable
sutures. To ensure secure fixation, the sutures are passed through the axis bolt of the implant and then
secured to the soft tissues on the other side. (B) The ends of the ligament sutures can be passed through
a drill hole in the olecranon and tied to ensure initial stability of the unlinked arthroplasty. (Reproduced with
permission from Wright Medical, Memphis, TN.)

A B C D
FIG 89.14 Anteroposterior (A) and lateral (B) radiographs of a 67-year-old woman with rheumatoid arthritis
with a painful Sorbie total elbow arthroplasty 15 years postoperatively. Valgus tilting of the ulnar component
with attenuation of the medial collateral ligament can be noted. (C,D) Radiographs 1 year postoperative
revision to a linked implant. Allograft struts were required as cortical windows were needed to remove the
well-fixed components.

REHABILITATION 6 weeks postoperatively to improve elbow extension if required.

The elbow is immobilized for 10 to 14 days postoperatively to allow
for stable wound healing and a reduction in swelling. Active ROM is
initiated after suture removal, avoiding terminal extension if an
unlinked implant was performed. A daytime 90-degree resting ortho-
sis is used between exercises to protect the collateral ligament repairs
if an unlinked implant was performed. No triceps precautions are
needed if a lateral paraolecranon or paratricipital approach is used. If
the triceps was detached from the olecranon as part of the surgical
approach, extension should be performed with gravity assistance only,
and full flexion is also avoided to prevent avulsion of the triceps.
In most cases, patients can be taught their exercises, and regular out-
patient therapy is not required. A nighttime extension splint is initiated
Strengthening begins 10 weeks postoperatively after secure ligament
healing has occurred.
COMPLICATIONS
The complications of unlinked and convertible TEAs are similar to any
arthroplasty and are discussed in Part 9. The common short-term
complications include infection, ulnar neuropathy, intraoperative
fractures, triceps avulsion, and problems with wound-tissue healing.
The complication specific to unlinked arthroplasty is instability,
which typically occurs early but can present late due to polyethylene
wear or attenuation of the collateral ligaments (Fig. 89.14). Longer
term complications include stem loosening or fracture, bearing wear,
 CHAPTER 89 Unlinked and Convertible Arthroplasty: Design, Concept, and Technique
A B
FIG 89.15 Anteroposterior (A) and lateral (B) radiographs of a well-
functioning Kudo type-5 total elbow arthroplasty.
periprosthetic fractures, and infection. If a radial head prosthesis is
employed, dissociation, wear, and loosening of this component can
also occur.
RESULTS
The reported outcome of unlinked and convertible TEAs consists
primarily of retrospective case series and registry data. There are no
prospective randomized trials comparing the outcome of different
unlinked arthroplasties.31 Little et al. compared the outcome of the
Souter, Kudo, and Coonrad-Morrey designs in a nonrandomized,
prospective cohort study design.30 The authors were unable to demon-
strate a significant difference in outcome or complications between
the unlinked Kudo and Souter designs. The linked Coonrad-Morrey
implant had a similar outcome as the unlinked designs with a lower
incidence of instability. The loosening rate did not differ between the
three designs at an average of 5 years postoperatively.
The outcome of the Kudo TEA has largely been reported by the
surgeon–inventor and has reflected the evolution of the prosthesis
design.c The early type 1 and 2 prostheses utilized an unstemmed
humeral component, which had a high incidence of loosening.26
Experience with the later designs were better with a lower incidence of
instability and loosening. The uncemented humeral component has
been more reliable than the uncemented ulnar component, where
cement and metal backing have improved the success (Fig. 89.15).
Thillemann and coworkers reported 67% survival of 17 Kudo type-3
elbows followed for an average of 9.5 years.51 Two had been revised
due to loosening of the ulnar component and one for instability.
c
References 26–29, 33, 37, 41, 42, 49, 50, 51, 57, 60.
851
A B
FIG 89.16 (A,B) Well-functioning iBP (Biomet, Warsaw, IN) total elbow
arthroplasty in a 77-year-old man with rheumatoid arthritis. The patient
had mild valgus laxity on stress testing but no clinical instability.
(Courtesy of Simon Frostick, MA, DM, FRCS, FRCS Ed, F FST RCS Ed,
University of Liverpool, UK.)
Progressive valgus tilting of the elbow at longer follow-up raised
concerns regarding polyethylene wear of the ulnar component. Tanaka
et al. reported on the outcome of 60 rheumatoid elbows treated with
a Kudo type-5 prosthesis.49 At 13 years, none of the elbows with a
metal-backed ulnar component had been revised with a Mayo Elbow
Performance Index (MEPI) of 79. Survivorship of the elbows with all
polyethylene ulnar components was 72% with an MEPI of 72. The
outcomes were worse in patients with more extensive erosive bone loss
preoperatively.
Kleinlugtenbelt et al. reported the outcome for 20 patients with
rheumatoid arthritis treated with iBP implants at a mean of 4 years
(Fig. 89.16).25 The MEPI improved from 49 to 90 points. One elbow
dislocated postoperatively; the rest were stable at follow-up. Small
lucent lines were seen around some ulnar components; however, there
were no revisions at this early follow-up. Fevang et al. reported on the
revision rate of 562 total elbows in the Norwegian Joint Registry, 111
of which were iBP and 161 Kudo implants.12 The majority of the
patients had inflammatory arthritis. The 5-year revision rate of the iBP
was 12% and the 5- and 10-year revision rates of the Kudo were 6%
and 15%, respectively. The revision rate was greater if the ulnar com-
ponents were inserted without cement. Instability was the most
common cause of revision for the iBP at a mean 2.9-year follow-up.
Loosening was the most common cause of revision of the Kudo
implants at a mean 8-year follow-up (Fig. 89.17).
There are limited data as of yet regarding the outcome of the
Latitude and none available for the newer Latitude EV convertible
elbow arthroplasty (Fig. 89.18). The functional outcome of the
Latitude total elbow was reported by Studer et al. at an average of 5
years in a study focusing on the effectiveness of the lateral paraolec-
ranon approach.48 Twenty-eight implants were linked, and 9 were
unlinked. In 28 elbows with rheumatoid arthritis the MEPI was 86.
In nine patients with acute distal humeral fractures, the MEPI was 89.
Burkhart and coworkers reported on the initial outcome of a mixed
group of 15 patients with acute fractures, nonunions, posttraumatic
arthritis, and rheumatoid arthritis using the Latitude system.6 The
implants were linked (7), unlinked (2), or a hemiarthroplasty (6).
852 PART VIII Joint Replacement Arthroplasty

A B
FIG 89.17 (A,B) A 71-year-old man with polyethylene wear 6 years postoperatively with iBP (Biomet,
Warsaw, IN) total elbow arthroplasty requiring revision to a linked implant due to pain and instability.
(Courtesy of Simon Frostick, MA, DM, FRCS, FRCS Ed, F FST RCS Ed, University of Liverpool, UK.)

A B C D
FIG 89.18 (A,B) Anteroposterior and lateral radiographs of a 44-year-old woman with pauciarticular rheuma-
toid arthritis. (C,D) Radiographs 10 years postoperatively of an unlinked Latitude (Wright Medical, Memphis,
TN) total elbow arthroplasty showing no evidence of loosening or bearing wear.

The MEPI was 89 at 13.5 months. Wagener et al. reported on the
outcome of 69 Latitude convertible total elbows at a mean follow-up of
43 months.59 Fifty-four percent had rheumatoid arthritis whereas the
remainder had primary or posttraumatic osteoarthritis. Eighty-three
percent of the implants were linked. Good functional outcomes and
patient satisfaction were reported with no implant loosening. Radial
head disengagement was noted in 31% of patients who had a radial
head component used; however, this did not correlate with clinical
symptoms (Fig. 89.19). One of the patients with an unlinked implant
had persistent instability and was revised to a linked device by placing
an ulnar cap.
In summary, the majority of unlinked implants used over the last
20 years are no longer available. As a consequence, there are few data
on the outcome of currently used unlinked and convertible TEAs. It is
likely that patient selection and surgeon experience play critical roles
in the outcome of elbow replacements in general and unlinked devices
in particular. Unlinked prostheses are less forgiving than linked designs
due to the need to accurately position the components, balance the soft
tissues, and achieve ligament healing. Randomized clinical trials and
more robust registry data are needed to accurately compare the com-
plications and longevity of different designs of currently available
elbow prostheses.
 CHAPTER 89 Unlinked and Convertible Arthroplasty: Design, Concept, and Technique 853

A B C

D E F
FIG 89.19 (A,B) Anteroposterior and lateral radiographs of a 55-year-old man with pauciarticular psoriatic
arthritis. (C,D) Postoperative radiographs show well-positioned unlinked Latitude EV (Wright Medical,
Memphis, TN) arthroplasty. (E,F) Two years postoperatively the patient had no pain and a functional range
of motion. The radial head component was noted to be disassociated from the stem, but the patient was
asymptomatic.

REFERENCES
1. Alonso-Llames M: Bilaterotricipital approach to the elbow. Its application
in the osteosynthesis of supracondylar fractures of the humerus in
children. Acta Orthop Scand 43:479–490, 1972.
2. An KN: Kinematics and constraint of total elbow arthroplasty. J Shoulder
Elbow Surg 14(1 Suppl S):168S–173S, 2005.
3. Brownhill JR, Furukawa K, Faber KJ, et al: Surgeon accuracy in the
selection of the flexion-extension axis of the elbow: an in vitro study.
J Shoulder Elbow Surg 15:451–456, 2006.
4. Brownhill JR, Ferreira LM, Pichora JE, et al: Defining the flexion-
extension axis of the ulna: implications for intra-operative elbow
alignment. J Biomech Eng 131:021005, 2009.
5. Bryan RS, Morrey BF: Extensive posterior exposure of the elbow.
A triceps-sparing approach. Clin Orthop Relat Res 166:188–192,
1982.
6. Burkhart KJ, Müller LP, Schwarz C, et al: Treatment of the complex
intraarticular fracture of the distal humerus with the Latitude elbow
prosthesis. Oper Orthop Traumatol 22:279–298, 2010.
7. Celli A, Arash A, Adams RA, et al: Triceps insufficiency following total
elbow arthroplasty. J Bone Joint Surg 87A:1957–1964, 2005.
8. Dachs RP, Fleming MA, Chivers DA, et al: Total elbow arthroplasty:
outcomes after triceps-detaching and triceps-sparing approaches.
J Shoulder Elbow Surg 24:339–347, 2015.
9. de Vos MJ, Wagener ML, Hendriks JC, et al: Linking of total elbow
prosthesis during surgery; a biomechanical analysis. J Shoulder Elbow Surg
22:1236–1241, 2013.
10. de Vos MJ, Verdonschot N, Luites JWH, et al: Stable fixation of the IBP
humeral component implanted without cement in total elbow
replacement: a radiostereometric analysis study of 16 elbows at two-year
follow-up. Bone Joint J 96B:229–236, 2014.
11. Duggal N, Dunning CE, Johnson JA, et al: The flat spot of the proximal
ulna: a useful anatomic landmark in total elbow arthroplasty. J Shoulder
Elbow Surg 13:206–207, 2004.
12. Fevang BT, Lie SA, Havelin LI, et al: Results after 562 total elbow
replacements: a report from the Norwegian Arthroplasty Register.
J Shoulder Elbow Surg 18:449–456, 2009.
13. Gay DM, Lyman S, Do H, et al: Indications and reoperation rates for total
elbow arthroplasty: an analysis of trends in New York State. J Bone Joint
Surg 94A:110–117, 2012.
14. Gramstad GD, King GJ, O’Driscoll SW, et al: Elbow arthroplasty using a
convertible implant. Tech Hand Up Extrem Surg 9:153–163, 2005.
15. Hildebrand KA, Patterson SD, Regan WD, et al: Functional outcome
of semiconstrained total elbow arthroplasty. J Bone Joint Surg 82A:
1379–1386, 2000.
16. Hosein YK, King GJ, Dunning CE: The effect of stem surface treatment
and material on pistoning of ulnar components in linked cemented elbow
prostheses. J Shoulder Elbow Surg 22:1248–1255, 2013.