Patella Fractures and Extensor Mechanism Injuries

As many as 35% of transverse patellar fractures are nondisplaced.⁶,²² While the mechanism may be multifactorial in nature, these injuries are typically associated with indirect, longitudinal

forces that fracture the patella but are insufficient to tear the
medial and lateral patellar retinaculum. As a result, the extensor
mechanism remains competent. The preserved integrity of the soft tissue
envelope helps to maintain the reduction. Approximately 80% of these
fractures occur in the middle to lower third of the patella.¹¹¹

Stellate fractures typically result from direct blow
injuries to the patella with the knee in a partially flexed position.
Approximately 65% of these injuries are nondisplaced.¹¹,²²
Active knee extension is preserved, because the medial and lateral
patellar retinacula are usually not torn with the injury. Damage to the
patellar and femoral articular surface is not uncommon given the
mechanism of injury, and careful evaluation on tangential views or MRI
is necessary to identify occult osteochondral lesions.

A vertical or longitudinal fracture pattern is not
uncommon and has been reported to account for 12% to 22% of patellar
fractures in several large series.¹⁸,⁴²,¹¹¹
The fracture line is most commonly seen involving the lateral facet and
lying between the middle and lateral third of the patella. Different
mechanisms of injury have been implicated. Bostrom et al.²³ reported that lateral avulsion was the most common mechanism in 75% of their series. Dowd,⁴²
however, reported that direct compression of the patella in a
hyperflexed knee is responsible for this pattern of injury. The
patellar retinacula are intact, preserving active knee extension by the
patient. The fracture pattern is easily missed on an AP radiograph,
emphasizing the importance of an axial view to identify this injury.

Noncomminuted, transverse fractures account for approximately 52% of displaced patellar fractures.²⁰,²¹,²²,⁸⁹
Evaluation of the integrity of the extensor mechanism is critical with
this pattern of injury. The fracture fragment separation (>3 mm) is
suggestive but not diagnostic of
retinacular and extensor mechanism disruption. A subset of these
fractures with fragment displacement but intact retinacula exists and
is characterized by preservation of full active knee extension. These
fractures may respond more favorably to nonoperative treatment. Bostrom
et al.²² has reported that preservation of the retinacula allows satisfactory healing without surgery. McMaster,¹⁰¹ however, has warned of a high risk of nonunion with nonoperative treatment in these patients.

Displaced, stellate fractures usually result from a
high-energy, direct blow to the patella. These fractures typically
demonstrate a high degree of comminution. Anterior soft tissue
contusion and/or lacerations are not uncommon, and careful evaluation
for an open fracture or traumatic arthrotomy is warranted. Transverse
fracture lines with extensive comminution may result in propagation
into the retinaculum and disruption of the extensor mechanism. However,
even if the extensor mechanism is preserved, the significant articular
incongruity may warrant operative intervention.

Fractures at the proximal pole of the patella are
typically bony avulsions of the quadriceps mechanism. Displacement is
rare and has been reported to be approximately 4% in large clinical
series.²²,¹³⁴
Active knee extension may be preserved if the medial and lateral
retinacula remain intact. The lateral radiograph may demonstrate
patella baja and a reduced Insall-Salvati ratio. Distal pole fractures
are bony avulsions of the patellar tendon. Displacement is much more
common with these injuries and has been reported to occur in up to
11.5% in large series.²²,¹³⁴
Retinacular disruption with loss of knee extension is virtually
universal with distal pole fractures. A lateral radiograph will
demonstrate patella alta and an increased Insall-Salvati ratio.⁶⁴

Osteochondral fractures of the femur or patella are also
seen in association with high-energy, stellate patellar fractures or
after patellar dislocation (Fig. 52-5). Plain
radiographs may not demonstrate these lesions. MRI with
cartilage-sensitive sequences can improve the detection of these
injuries.¹⁶² Osteochondral fracture
fragments can shear from the lateral femoral condyle or medial patellar
facet after patellar subluxation or dislocation and may warrant
surgical intervention. Kroner⁷⁹ first reported on a series of these fractures after patellar subluxation in patients 15 to 20 years of age.

Patellar fractures have been infrequently reported after
graft harvest for bone-tendon-bone anterior cruciate ligament (ACL)
reconstruction. An incidence of 0.2% has been reported in one series of
more than 1700 ACL reconstructions.⁸²
While these may occur intraoperatively secondary to technical error,
the majority of cases have been attributed to postoperative trauma from
a fall or overly aggressive rehabilitation protocols. Both transverse
and vertical fracture patterns have been reported. Rigid fixation of
these fractures, even in the setting of minimal displacement, has been
advocated to allow for early motion and avoid delayed rehabilitation of
the ACL reconstruction.

A bipartite or tripartite patella is a normal anatomic
variant and should not be misdiagnosed as a displaced patellar
fracture. A bipartite fragment typically presents as a well-corticated
fragment

in
the superolateral aspect of the patella and is the result of incomplete
fusion of ossification centers. The incidence of a bipartite patella is
approximately 8% and is almost always seen bilaterally.¹,¹¹² Radiographs of the contralateral knee will confirm the diagnosis.

Nonoperative treatment of minimally displaced fractures
has been reported with good clinical outcomes. In a large series of 422
patellar fractures reported by Bostrom et al.,²²
219 minimally displaced fractures were treated nonoperatively, and 98%
had good to excellent results at final follow-up. Only two failures
occurred with nonoperative treatment. Other series have reported low
failure rates of less than 5% with closed management of minimally
displaced fractures.¹²⁴,¹³⁷

A variety of internal fixation techniques for patellar
fracture stabilization have been described in the literature. The first
description of cerclage wiring for patellar fracture fixation was made
by Berger in 1892.⁵ Anderson⁵ discussed the use of an equatorial circumferential wire placed around the patella, and Magnuson and Payr⁹²,¹¹⁴
described successful fixation with intraosseous wires passed through
vertical drill holes. Screw fixation for longitudinal and transverse
fracture patterns with large fragments has also been described.⁴⁰,¹⁰⁵,¹⁵⁶
The potential concern of these fixation strategies, however, has been
(a) an inability to initiate early motion due to the risk of
displacement with large tensile forces from quadriceps contraction and
(b) lack of compressive forces at the articular surface. The AO/ASIF
popularized the technique of tension band wire fixation for patellar
fractures to address these concerns, based on biomechanical studies
demonstrating increased construct strength with wires placed on the
anterior, tension-side cortical surface of the patella.¹⁰⁶

Fracture patterns and examples of internal fixation are illustrated in Figure 52-9.

Biomechanics of Tension Band and Modified Tension Band
Fixation. The principle of tension band wire fixation for patellar
fractures is to convert the tensile forces generated from the
quadriceps complex at the anterior cortical surface of the patella into
compressive forces at the articular surface. With progressive knee
flexion, the passive tensile forces in the extensor mechanism in
addition to the pressure of the femoral condyles against the patella
increase interfragmentary compression at the articular surface.

A study by Benjamin et al.¹⁵
compared the strength of four different fixation strategies (tension
band wiring, modified tension band wiring over Kirschner wires
[K-wires], Lotke and Ecker longitudinal anterior banding, and
circumferential cerclage wiring) in a transverse patellar fracture and
retinacular disruption model. Cerclage wiring provided the weakest
fixation strength, with up to 20 mm of gapping at the fracture site
with tensile stress. The modified anterior tension band technique of
transosseous K-wire fixation with anterior banding demonstrated
superior strength to all other constructs.¹⁵
Isolated screw fixation with 3.5- or 4.5-mm screws may be sufficient,
particularly in the setting of simple transverse or longitudinal
fractures in patients with good bone stock. Burvant et al.,²⁷
however, also found tension banding with screws to perform
biomechanically superiorly to five other techniques of transverse
patellar fracture fixation, including screw fixation alone. Carpenter
et al.³⁰ compared a modified tension
band, parallel 4.5-mm interfragmentary lag screws, and 4.0-mm
cannulated lag screws augmented with a tension band passed through them
in a transverse patellar fracture cadaver model. The highest
load-to-failure was seen with the modified tension band and the
cannulated lag screw technique.³⁰ Regardless of which tension

band construct was selected, Baran et al.¹³
used MRI in a knee cadaveric model to advocate for placement of the
tension band wire as close to the bone as possible with minimal
interposing tendinous tissue.

The use of monofilament wire versus braided cable has also been evaluated in biomechanical studies. Scilaris et al.¹²⁹
compared anterior tension banding with 1.0-mm wire versus 1.0-mm
braided cable. The braided cable allowed for less fragment displacement
with cyclical loading. In addition, tightening of the wire by twisting
at two different sites compared with a single site has been shown to
provide greater interfragmentary compression.¹²⁸ John et al.⁶⁶ demonstrated improved stability with cyclic loading if wire twists were placed at the corners of the figure-of-eight loop.

Due to wound complications and local irritation
associated with subcutaneous wiring, braided suture tension banding has
also been evaluated. McGreal et al.⁹⁹
demonstrated that braided polyester suture was 75% as strong as wire
and performed equivalent to cerclage wire with cyclical loading.
Braided No. 5 Ethibond has also been shown to be comparable to wire
fixation with anterior tension banding or Lotke-Ecker anterior
longitudinal banding procedures for displaced, transverse

fractures.¹⁶⁰ In a clinical study, Chen et al.³³
demonstrated equivalent clinical outcomes with wire versus
biodegradable tension band fixation of patellar fractures at a mean of
2 years of follow-up.

Open Fractures. Open patellar fractures constitute a
surgical emergency and warrant immediate irrigation, débridement, and
stable fixation. The consequences of a surgical delay are not
insignificant and can include septic arthritis and patellar
osteomyelitis. A thorough débridement includes removal of all
devitalized fragments and soft tissue, followed by serial débridements,
skin grafting, and local or free tissue flaps if necessary for wound
closure. Appropriate antibiotic coverage is also essential.

Catalano et al.³²
reported on a series of 79 open patella fractures with a mean of
21-month follow-up. Almost all fractures were displaced, with 22%
transverse patterns and 39% comminuted, stellate injuries. The majority
(53%) were Gustilo and Anderson Grade II injuries. Surgical
intervention consisted of a thorough irrigation and débridement
followed by open reduction and internal fixation in 57% and partial
patellectomy in 32% of cases. Outcomes were satisfactory in most cases
with only one case of revision internal fixation for failure and no
cases of deep infection. The Mayo Clinic’s retrospective review of open
patellar fractures also reported 77% good to excellent results, with no
infections in type I or II open injuries that were treated with
immediate internal fixation after irrigation and débridement.¹⁴⁸ Anand et al.⁴
recently presented the long-term outcomes of 16 open patellar fractures
and compared their results to those of a matched group of patients with
closed injuries. The Injury Severity Score (ISS) and the incidence of
associated musculoskeletal injuries were significantly higher in the
open patellar fracture group. At a mean of 45 months of follow-up,
patients with open patellar fractures had a higher incidence of
complications, lower Knee and Osteoarthritis Outcome Scores (KOOS), and
higher visual analog scores for pain.⁴

Modified Tension Band Wiring Fixation: Surgical
Technique. The patient is positioned supine on a radiolucent table. A
tourniquet is applied high on the thigh, with care being taken to avoid
entrapment of the quadriceps complex by inflating it only after the
knee has been gently flexed beyond 90 degrees. A longitudinal, midline
incision is typically performed and allows for excellent visualization,
protects the medial cutaneous branches of the femoral nerve, and can be
extended proximally and distally along the extensor mechanism if
necessary. In the case of an open fracture or a traumatic arthrotomy,
the laceration should be incorporated into the incision if possible.
Superficial dissection should be avoided regardless of incision
orientation to preserve the blood supply and the viability of skin
flaps. After exposure of the fracture lines, all clots and devitalized
debris should be cleared. Before any fixation, the degree of injury
should be carefully assessed. The articular surfaces of the femur and
patella should be inspected, and any intra-articular loose bodies
flushed out of the joint. In addition, the integrity of the medial and
lateral retinaculum as well as the proximal and distal soft tissue
attachments of the patella must be evaluated. The fracture pattern
should be generally defined. Complex fracture patterns with moderate
comminution may be simplified with the use of interfragmentary lag
screws to create a transverse pattern that is then amenable to tension
band fixation. With the knee slightly flexed, the fracture should be
reduced and maintained with a pointed reduction forceps. The quality of
the articular reduction should be palpated through the defect in the
retinaculum. Occasionally, extension of the arthrotomy or retinacular
tear may be necessary to allow palpation of the articular surface.
Gardner et al.⁵⁰ has developed a
technique for exposure and fixation of comminuted patellar fractures
using a lateral arthrotomy and eversion of the patella. This allows
direct visualization, permits reduction of articular surfaces without
soft tissue interposition, and facilitates confirmation of articular
congruity, compared with more traditional techniques that rely on
palpation alone.⁵⁰ Intraoperative fluoroscopy with imaging in multiple planes may also be used to confirm an anatomic reduction.

Two parallel 1.6-mm K-wires are placed perpendicularly
across the fracture line to maintain the reduction, compress the
fracture site, and anchor the tension band wire. The K-wires can be
placed in an antegrade or a retrograde fashion. Using the antegrade
technique, the wires are advanced from proximal to distal at a level 5
mm below the anterior cortical surface and parallel to it. The wires
are spaced apart to divide the patella longitudinally into thirds. When
using the retrograde technique, the reduction is taken down and the
proximal fracture fragment is flexed 90 degrees to expose the fracture
surface. Starting 5 mm below the anterior cortical surface and dividing
the patella longitudinally into the thirds, the K-wires are advanced
proximally through the fracture site, exiting at the locations of the
starting points for the antegrade technique. The reduction is then
re-established and held with a pointed reduction forceps. The K-wires
are subsequently advanced from proximal to distal across the fracture
site until they exit distally at the inferior patellar margin.

The tension band wire is then passed and tightened to
complete the construct. To facilitate wire passage, a large-bore
angiocatheter (14- or 16-gauge) can be passed through the quadriceps
tendon directly adjacent to the superior patellar margin, and an
18-gauge wire is then passed through the catheter, after which it is
removed. The limbs of the wire are crossed over the anterior cortex of
the patella and one limb is passed below the patellar tendon
immediately adjacent to its inferior margin as described earlier. After
anatomic reduction is confirmed with direct palpation and fluoroscopy,
the wire is tensioned with slow twisting of the wire limbs. Twisting is
performed at two locations (one in each wire limb), as tightening at
only the ends of the wire may lead to asymmetric fracture compression
and excess slack on the contralateral side. Care must taken to avoid
overtensioning of the wires, which can result in wire breakage and loss
of reduction or malreduction secondary to iatrogenic fragment
comminution. After satisfactory wire tension is achieved, the K-wires
are cut at both ends and bent backward 180 degrees over the tension
band wire both proximally and distally. They are gently impacted and
buried into the patella to prevent migration. The arthrotomy is
copiously irrigated, and the retinacular tears and arthrotomy are
closed in a water-tight fashion with interrupted, figure-of-eight
nonabsorbable sutures. The wound is closed in a standard, layered
fashion (Table 52-3).

A slight modification of this technique uses 4.0-mm cannulated screws instead of K-wires (Fig. 52-10).
This allows for lagged interfragmentary compression and has been shown
to be biomechanically stronger than the K-wire construct.³⁰ In this technique, guide wires are placed in an identical fashion to

K-wires as just described, followed by drilling and cannulated lag
screws advancement across the fracture over the wires. Eighteen-gauge
wire is then passed in a figure-of-eight fashion through the cannulated
screws and tightened after anatomic reduction of the fracture has been
confirmed.¹⁶

Simplification of Fracture Patterns. Certain fracture
patterns may present with more than one major fracture line, dividing
the patella into more than two major fragments. In these situations,
minor fragments can be reduced and fixed together to simplify the
fracture pattern into one that is amenable to anterior tension band
wire fixation. Lag screws, K-wires, and bioabsorbable pins all may be
used to hold together smaller fragments. We have typically used 3.5-mm
interfragmentary lag screws in these situations, although screw
fixation without lag compression may be
preferable with moderate or severe comminution to avoid excessive
shortening and malreduction. If the fragment is comminuted at the point
of entry, a washer can be used to anchor the screw head.

Longitudinal Anterior Banding and Cerclage Wiring—The
Lotke-Ecker Technique. Stellate fracture patterns that are not amenable
to a modified anterior tension band technique may be treated with
longitudinal anterior banding plus cerclage wire

fixation (the Lotke-Ecker technique).⁸⁹
The fracture is exposed and the fracture fragments carefully identified
after the clot and devitalized debris are removed. Minimally displaced
fractures are fixed in situ, while severely comminuted and displaced stellate fractures are reduced through indirect techniques.⁸⁹
In these cases, a cerclage wire is first placed around the
circumference of the patella immediately adjacent to the bone with the
assistance of a 14- or 16-gauge angiocatheter. Gross manual reduction
is performed without clamps, followed by articular surface reduction
with progressive tightening of the cerclage wire.

Longitudinal anterior banding is then performed by
drilling two parallel tunnels 1 cm from the medial and lateral edges of
the patella respectively with Steinmann pins or 2.0-mm drill bits in an
antegrade fashion.⁸⁹ A large-gauge
wire (18- to 22-gauge) is then inserted into both drill holes,
preserving a closed loop distally. The distal loop is brought
anteriorly and one free proximal end of the wire is passed through this
anterior loop. The wire is then secured to its other end and tightened
with progressive twisting. This fixation results in a hybrid of
anterior tension banding and interosseous wire fixation.⁸⁹

Minimally Invasive and Arthroscopic-Assisted Fixation.
The use of arthroscopy to assist with patellar fracture reduction and
fixation has been described. Appel and Siegel⁷
presented a limited series of cases with arthroscopic-assisted
reduction followed by percutaneous screw and wire fixation with
satisfactory outcomes. Theoretical advantages include direct
visualization of the articular surface during internal fixation,
limited dissection and soft tissue stripping, and the ability to
address associated intra-articular pathology in the knee including
osteochondral fractures. While this technique may have a role in
selected cases, we do not believe that it supplants open techniques,
which allow for anatomic fracture reduction, rigid fixation, and
retinacular and extensor mechanism.

Pizarro et al.¹¹⁶
recently presented a randomized series of 53 patients treated with
percutaneous patellar osteosynthesis (PPOS) versus standard open
reduction and internal fixation for closed, displaced transverse
patellar fractures. In the PPOS group, a special device was secured via
four percutaneous portals to maintain the reduction while an anterior
tension band was placed. In the group of patients treated with PPOS,
the authors reported shorter surgical times, less pain, less
complications, and improved functional outcome scores by Knee Society
Clinical Rating Scale (KSCRS) at 1-year follow-up. Clinical outcome
scores were not statistically different between the two groups by
2-year follow-up.¹¹⁶

Outcomes. The lack of a uniform surgical technique or a
standardized assessment scale limits the utility of reported outcomes
after operative fixation of patellar fractures. As a result, the
literature provides generalizations about “good” or “excellent”
outcomes based on subjective patient complaints of pain, loss or
motion, or limitations in daily activities. Based on the combined
results of many small series, open reduction and internal fixation has
produced a good to excellent result in 73% of 361 fractures (Table 52-4).

The results of operative repair of patella fractures
need to be interpreted with some caution, however, because series are
more often reported by repair construct rather than by fracture
pattern. A compilation of published series is shown in Table 52-4. Modified anterior tension band wiring has produced the best results, with 86% good to excellent outcomes reported.⁹ Bostman et al.²¹
reported significantly better results with anterior tension banding
compared with cerclage wiring, partial patellectomy, or
interfragmentary screw fixation. While small series have reported
excellent outcomes with cerclage wiring, a review of the combined
results reveals inferior performance to tension band wire fixation,
with only 70% good to excellent results. The anterior longitudinal
banding technique of Lotke and Ecker

can also be effective, with 81% excellent results reported in their small series.⁸⁹

When comminution of the distal pole or a fragment of the
patella is extensive and cannot be stabilized with internal fixation, a
partial patellectomy should be performed. The surgical approach is the
same as for open reduction and consists of a longitudinal midline
exposure. Care is taken to preserve as many large, viable fragments as
possible. Retained fragments are anatomically reduced and secured to
one another with screws or K-wires. If the comminution primarily
involves the central patella with preserved proximal and distal
fragments, the central comminution can be excised and the remaining
fragments secured as congruously as possible with screw fixation.

With severe inferior pole comminution, resection of
fragments with patellar tendon reattachment can be performed. Most of
these fractures are extra-articular, as the distal pole is devoid of
articular cartilage. Three longitudinal drill holes are then placed
through the remaining patella to serve as tunnels for suture passage. A
tendon-grasping, woven or locking nonabsorbable suture (such as a
Krackow suture) is placed in the patellar tendon, and the sutures are
passed with a ligature passer through the tunnels and firmly tied over
bone bridges with the knee in hyperextension. Tantamount to the tendon repair, however, is a meticulous repair of the associated medial and lateral retinacular injury.¹⁵
Based on the energy of the injury and strength of the repair, the
construct may be protected with a cerclage wire, a tendon graft, or a
Mersilene tape that is passed immediately proximal to the superior pole
of the patella and inferiorly through the proximal tibia posterior to
the tibial tubercle.

The
cerclage should be tightened with the knee flexed to 90 degrees, as
tightening in extension may constrain the postoperative flexion that is
achievable by the patient. The strength of the repair should always be
evaluated intraoperatively, observing for interfragmentary motion and
the integrity of the tendon-bone interface with progressive knee
flexion. Rigid constructs may allow for early, controlled motion of the
knee. Typically, extension bracing for ambulation and partial weight
bearing with crutch assistance is maintained for 6 weeks
postoperatively.

The location of drill holes for tendon reattachment
after partial patellectomy is important and has been debated in the
literature. Studies have suggested that the holes be placed near the
articular surface to avoid abnormal tilting of the patella and creating
increased patellofemoral contact forces.⁴³ In 1958, Duthie and Hutchinson⁴³
reported tilting of the patella in five of seven patients with
postoperative arthritis and attributed these changes to malalignment
from attachment of the patellar tendon to the anterior cortex. In
contrast, Marder et al.⁹⁴ completed
contact pressure studies demonstrating improved mechanics with anterior
tendon reattachment for 20% and 40% partial patellectomy models.
Furthermore, Zhao et al.¹⁶³ reported
an increase in the force required to extend the knee with tendon
attachment to the articular surface compared with the anterior cortex.

Traditionally, comminuted fractures of the distal pole
have been managed with partial patellectomy with satisfactory outcomes.
Recently, a technique for osteosynthesis of these fractures using a
novel basket plate device has been described.⁷⁰,⁷⁸,⁹⁶,¹⁵⁰
The basket plate is contoured to fit the geometry of the patellar apex
and is augmented with several anterior and posterior flanges to collect
the comminuted fragments. Fibers of the proximal patellar tendon are
spread apart by the flanges to allow positioning it at the inferior
pole without detachment. Four holes are provided to place parallel and
oblique screws that engage the opposite fragment and resist both
tensile and shear forces.⁷⁸,⁹⁶
Biomechanical studies have demonstrated an ultimate load-to-failure of
about 420 N for this construct, approximately 70% higher than a
separate vertical wiring fixation construct.⁷⁸,¹⁶¹

Preliminary clinical outcomes with the basket plate osteosynthesis technique have been favorable. Matejcic et al.⁹⁶
reported 90% good to excellent outcomes at a mean 5-year follow-up for
51 patients who underwent basket plate osteosynthesis of comminuted
distal pole fractures. These data must be approached with caution,
however, as 61% of the study patients were lost to follow-up.⁹⁶ A retrospective cohort study by Kastelec et al.⁷⁰
compared 11 patients who had internal fixation of a distal pole
avulsion fracture to 13 patients who had a pole resection. At an
average follow-up of 4.6 years, average patellofemoral scores were
significantly better in the osteosynthesis group compared with the
resection group.⁷⁰ Furthermore,
there was a significantly greater incidence of patella baja in the pole
resection group as assessed radiographically by the method of
Blackburne and Peel.⁷⁰ Sample sizes
were small, however, and randomized control trials are necessary to
further define potential advantages of osteosynthesis compared with
resection in the surgical management of comminuted distal pole
fractures.

Outcomes. When appropriate selection criteria are used,
partial patellectomy can yield functional outcomes that are equivalent
to open reduction and internal fixation²¹,²²,¹⁰⁴,¹¹¹,¹²⁶,¹³⁴ (Table 52-5).
Multiple authors have reported nearly normal functional outcome when
large fragments of the patella and articular congruity are preserved.
Retention of small, nonviable fracture fragments or those devoid of
cartilage did not improve function, while retention of large fragments
provided a lever arm for improved extensor mechanism function. Combined
results in 138 cases have shown a good to excellent outcomes in 72% of
cases. With extensive inferior pole comminution, superior results have
been reported with partial patellectomy compared with internal fixation.²² Bostrom et al.²²
reported 88% good to excellent results with partial patellectomy for
transverse patellar fractures with inferior pole comminution, compared
with only 74% good to excellent results with internal fixation.

Total patellectomy is occasionally performed for highly
displaced, comminuted fractures in which stable fixation cannot be
achieved and when no large fragments can be retained. We have reserved
this as a salvage procedure in the rare setting of failed internal
fixation or patellar osteomyeltis. Retention of even a single fragment
is usually possible, and can substantially help to preserve the
biomechanical function of the extensor mechanism.³

If a total patellectomy is required, all bony fragments
are removed, with care being taken to preserve the retinacula and
restore the integrity of the extensor mechanism. Typically, an anterior
longitudinal incision is used, although a prior incision that will
preserve viability of skin flaps may be chosen in a revision setting.
Removal of the patella effectively lengthens the extensor mechanism,
such that end-to-end repair will result in an increased slack distance
and extension lag after surgery. This is avoided by imbrication or use
of a purse-string suturing technique that shortens the tendinous
repair. Tension should be evident on the repair at 90 degrees of
flexion when length has been adequately restored.

In certain situations, there may be an inadequate amount
of soft tissue available for primary repair of the extensor mechanism.
In general, the absence of prepatellar soft tissues is addressed with
quadriceps turndown procedures. Absence of the quadriceps tendon,
however, usually requires fascia or tendon weaving procedures.

Shorbe and Dobson¹³⁶
described the inverted V-plasty quadriceps turndown procedure. The
quadriceps tendon is exposed for approximately three inches, and a
full-thickness, V-shaped incision is made into the tendon with the apex
of the “V” located 2½ inches proximal to the former superior pole of
the patella. The V-limbs extend distally for 2 inches, creating a base
that preserves approximately ½ inch of tendon that is continuous with
the retinaculum on both the medial and lateral sides. The V-flap is
folded inferiorly, inserting it through the proximal patellar tendon
and suturing it in place with several interrupted sutures. The
quadriceps tendon defect is then closed in a side-to-side fashion with
multiple interrupted sutures.

If the quadriceps tendon is deficient, Gallie and Lemesurier⁴⁹
have described a free fascia or tendon weave procedure. After nonviable
tendon remnants have been débrided, the length of the defect is
measured with the knee in full extension. This length is doubled and 2
inches added to harvest to the appropriate-size fascial graft. Through
a separate lateral incision, a 1½-cm-wide graft of fascia lata or
iliotibial band of appropriate length is harvested and tubularized
along its long axis. Alternatively, allograft tissue of appropriate
length can be used. The graft is then woven through the quadriceps
tendon and/or muscle, passed through the patellar tendon, and sewn back
onto itself. Tension should be evident on the repair at 90 degrees of
flexion when length has been adequately restored. Excess tissue is
excised and all edges firmly sutured down.

Outcomes. The outcomes of total patellectomy are
generally inferior to those reported after internal fixation or partial
patellectomy for fracture. However, these results are difficult to
interpret as fractures treated with total patellectomy likely reflect
higher energy, comminuted injuries. The combined results of 219 cases
showed 72% good to excellent results.

Before the introduction of anterior tension band
techniques, poor methods of achieving rigid internal fixation may have
justified total patellectomy in these early series. Recent studies,
however, have reported a higher rate of fair to poor outcomes with
total patellectomy (Table 52-6). Bostrom²⁰ recommended total patellectomy only in the setting of severely comminuted fractures. Sutton and colleagues¹⁴³
evaluated quadriceps strength and functional abilities in patients who
had undergone partial or complete patellectomy, with the contralateral
normal knee serving as a control. Total excision was associated with a
49% reduction in strength of the extensor mechanism and a mean loss of
18 degrees in range of motion. Functionally, there was a high incidence
of instability, with an inability of the patients to support their
weight on the loaded knee with stair climbing.¹⁴³ Einola et al.⁴⁵ reported on 28 patients with a mean 7½ years

of follow-up and reported a good result in only 6 patients. Quadriceps
power was within 75% of the contralateral knee in only 7 patients.
Scott et al.¹³⁰
similarly reported a good outcome in only 4 of 71 patients, with 90%
complaining of chronic pain. Quadriceps atrophy, pain, and instability
in the knee during running or stair climbing were common complaints of
patients in both series. Wilkinson et al.¹⁵⁹
evaluated 31 patients with up to 13 years of follow-up and reported an
excellent outcome in less than 25% of patients. Furthermore, Sorensen
et al.¹⁴⁰ noted that no patients
regained full quadriceps strength after patellectomy and that none of
their operative reconstructions would have fared better with primary
excision. For these reasons, it is our recommendation to avoid total
patellectomy whenever possible. We have not encountered a patella so
comminuted that a single fragment could not be preserved acutely.
Maintaining even a single fragment provides a lever arm that
substantially improves the function of the extensor mechanism.⁷¹

If a total patellectomy is performed, the decision to
operate acutely compared with a delayed fashion remains controversial.
Muller et al.¹⁰⁷ retrospectively
reviewed their results of early (within 4 weeks) versus delayed
patellectomy and identified no differences in range of motion or
clinical outcome scores between the groups.

A well-padded compressive dressing is applied over the
wound. A knee immobilizer is applied initially and transitioned to a
locked hinged knee brace and the patient is allowed to bear weight as
tolerated. For the first 2 weeks, the leg is maintained in extension.
After the wound has healed, knee range of motion is permitted from 0 to
40 degrees and increased at biweekly intervals by 20 to 30 degrees with
the goal of obtaining 90 degrees of flexion by 6 to 8 weeks. The
patient is instructed in active assisted and passive range-of-motion
exercises during this time period. Active range of motion and
strengthening are begun 6 to 8 weeks postoperatively. The brace is used
until the patient has good quadriceps control and is able to perform a
straight leg raise.

Several authors have reported good to excellent results in 80% to 100% of operatively treated quadriceps tendon ruptures (Table 52-7).
Many different techniques of primary repair have been described, and no
single technique has demonstrated superiority. In early series, primary
end-to-end repair was performed with cast immobilization, while in more
recent series, heavy nonabsorbable sutures have been passed through
bone

tunnels
and early rehabilitation undertaken. The only factor that has been
associated with inferior results is delay in timing of surgical repair
of greater than 2 to 3 weeks.¹²⁵,¹³⁷ Scuderi¹³² reported good to excellent results in 85% patients treated using the turndown flap. Siwek and Rao¹³⁷
evaluated 36 ruptures and found that the 30 patients who underwent
early treatment obtained good to excellent results, while the 3
patients treated after 4 weeks had unsatisfactory results. Larsen and
Lund⁸¹ reported good to excellent
results in 15 of 18 patients. This study attempted to correlate
radiographic patellofemoral congruence with clinical outcome. Thirteen
of the 18 patients demonstrated incongruence compared with the
contralateral knee. All patients with residual pain had patellofemoral
incongruence. However, it is unclear what this finding means because 9
of the 13 patients with patellofemoral incongruence were asymptomatic.
They concluded that patellar congruence was not the only factor
contributing to persistent pain after repair. Rougraff¹²⁵
reviewed 53 ruptures treated with multiple surgical techniques and
postoperative regimens and found no differences based on repair
technique or postoperative protocol. Patients with delayed surgery had
poorer functional outcome and decreased satisfaction scores. Konrath⁷⁷
reviewed 51 quadriceps tendon ruptures in 39 patients treated with
heavy nonabsorbable suture placed through bone tunnels using functional
surveys and objective testing. Ninety-two percent were satisfied and
84% returned to previous occupations. However, 51% were unable to
return to their presurgery level of recreational activity. Objective
testing demonstrated a 12% strength loss in the quadriceps muscle and
14% strength loss in the hamstrings as well as an 8-degree loss of
range of motion.

The postoperative rehabilitation regimen is based on the
surgeon’s assessment of the repair strength. A repair with poorquality
tissue should be immobilized for 4 to 6 weeks in a cylinder cast with
weight bearing as tolerated restrictions. A sturdy repair with
reinforcement can be started on early range of motion within 7 to 10
days postoperatively with active and active assisted exercises.
Initially full flexion is not allowed, but range of motion is started
at 0 to 45 degrees and advancing 30 degrees every 2 weeks. Active knee
extension is permitted at 6 weeks. The rehabilitation program should
stress closed chain eccentric strengthening to avoid possible reinjury.
Weight bearing is allowed immediately and crutches are used for at
least 6 weeks. The hinged knee brace and crutches are discontinued when
the patient is able to ambulate with good quadriceps control.
Competitive sports should not be allowed for 4 to 6 months until the
majority of isokinetic strength has been regained.

Most series report between 70% and 100% good to excellent results (Table 52-8).
The majority of patients who undergo early primary repair achieve a
functional range of motion and normal quadriceps strength. Persistent
quadriceps muscle atrophy commonly occurs but has not been correlated
with loss of strength.⁵⁸,¹³⁷ No relationship has been demonstrated between the configuration of the rupture, the method of repair, and

clinical outcome.⁷⁷,⁸¹,¹²²,¹³⁷ An early repair within 2 to 3 weeks of injury is the only factor that has been associated with improved outcomes. Siwek and Rao¹³⁷
reviewed 25 patients treated within 7 days and found that 96% had good
to excellent results while only one third of the six patients with a
delayed repair had excellent results. They also noted a greater degree
of persistent quadriceps atrophy in patients treated with delayed
repair. Hsu et al.⁵⁸ reviewed 35
traumatic patellar tendon ruptures treated acutely with primary repair
and a neutralization wire. They reported 20 excellent (57%) and 10
(29%) good results. The majority of the patients in this series had
sustained the injury on a motorcycle and one third of the patients had
multiple injuries, which the authors concluded may have contributed to
the slightly lower success rate observed. Larsen and Lund⁸¹
performed a radiographic analysis of patellar congruence in a series of
patients who underwent operative repair of acute patellar tendon
ruptures. Seven of 10 patients demonstrated incongruity on the Merchant
and lateral radiographs. Patients with residual patellofemoral symptoms
all had incongruity, but the majority were asymptomatic. They concluded
that articular incongruity may not be the only cause of persistent knee
pain in patients who undergo patellar tendon repair.

The two most common reasons for a suboptimal outcome
after an extensor mechanism injury are the loss of knee flexion and
quadriceps weakness.²¹ In general,
these complications are more closely associated with the injury itself
than with the technical aspects of the surgical procedure. An
aggressive postoperative program emphasizing early range of motion and
quadriceps strengthening is recommended. To achieve this goal, it is
paramount to achieve an operative repair that is strong enough to allow
early postoperative rehabilitation. Most patients will ultimately be
able to achieve 90 degrees of flexion. This is critical as it allows
for sitting in tight places and also helps with arising from a sitting
position. If 90 degrees of flexion is not obtained by 8 weeks
postoperatively, a closed manipulation under anesthesia may be
considered. However, caution should be used to avoid iatrogenic
re-rupture of the tendon. Arthroscopic débridement and manipulation may
be considered to treat a flexion contracture or more severe flexion
loss. Persistent quadriceps atrophy is noted in many patients but does
not appear to compromise final function. Siwek and Rao¹³⁷
reported that 75% of their patients who underwent acute quadriceps
tendon repair had persistent quadriceps atrophy of 2 to 4 cm. However,
despite the marked atrophy, most had strength that was adequate for
normal knee function. The treatment for an extensor lag secondary to
weakness or atrophy is a vigorous rehabilitation program.

The subcutaneous location of the patella and extensor
mechanism results in a high incidence of prominent or symptomatic
hardware. The problem is more common with the addition of a cerclage or
tension band wire. Symptoms are related to irritation of the capsule
and tendons by the implants. A 15% incidence of soft tissue irritation
from hardware that necessitated removal was reported in two series.⁶¹,¹³⁹
Removal of prominent cerclage or K-wires usually alleviates symptoms
and can be performed on an elective basis. The use of nonabsorbable
suture as a reinforcing material has become more popular in an attempt
to avoid this complication.³³,⁹⁹

Infection and wound complications are infrequent in
closed injuries and often related to patient specific conditions and
the subcutaneous positioning of heavy sutures or wires used in surgical
repair. Sutures should not be placed directly in line with the incision
to prevent delayed wound healing. Placing the skin incision lateral to
the tibial tubercle may provide better soft tissue coverage. A
postoperative hematoma can result in pain and an increased risk of
wound dehiscence and infection. The use of closed suction drainage is
recommended to avoid hemarthrosis formation.

While infection after surgical management of closed
injuries is relatively uncommon, postoperative infection following open
patella fractures has been reported to be as high as 10.7%.¹⁴⁸
Surgical débridement of nonviable tissue (soft tissue or bone) and
evaluation of the stability of fixation remain the standard tenets of
care. After thorough irrigation and débridement, stable fixation can be
retained. Intraoperative cultures should be used to guide an
organism-specific, 6-week course of intravenous antibiotics. In the
vast majority of cases, early recognition of infection and aggressive
débridement of nonviable tissue will allow for preservation of the
patella. A total patellectomy, however, rarely may be necessary in
situations of intractable osteomyelitis and secondary nonunion.

The rate of loss of reduction after open reduction and
internal fixation of patellar fractures has been reported to range from
0% to 20% in clinical series.²¹,⁶¹,¹³⁹
Inadequate fixation, severe comminution, early aggressive
rehabilitation, and patient noncompliance have all been implicated as
risk factors. Smith et al.¹³⁹ noted
11 cases of lost reduction in a series of 51 patients. Five cases were
attributed to technical errors in tension band wire placement, five to
patient noncompliance, and one to sequelae of infection. Nine cases had
symptomatic hardware that ultimately required removal. Nonoperative
management with a period of immobilization may be acceptable with
minimal displacement (2 mm or less) or incongruity after loss of
reduction. However, reoperation is warranted if the extensor mechanism
is compromised or separation of fragments or incongruity of more than 3
to 4 mm has developed.

Re-rupture after tendon repair can be seen in patients
who return to activities too early before the bone-tendon junction has
healed. Timely diagnosis and revision are usually successful in
reestablishing knee motion and strength. Competitive sports should not
be allowed for 4 to 6 months to allow time for sufficient tendon
healing and remodeling. Isokinetic strength should be regained before
resuming high-risk activities.

Nonunions occur infrequently in patellar fractures with an incidence ranging between 1% and 12.5%.²⁹,¹¹¹,¹⁵⁶
Transverse fractures may have a higher incidence, but nonunions can be
seen in comminuted fractures without stable fixation. If delayed union
is noted on radiography, a period of immobilization may be beneficial
to allow time for fracture consolidation. Elderly patients may tolerate
a nonunion well with minimal functional deficits, including persistent
muscle weakness and an extensor lag. However, reoperation should be
considered if there is a large separation between the bone fragments,
or pain in a younger patient. Revision surgery with rigid internal
fixation and autogenous bone grafting should be considered. Weber et al.¹⁵⁶ achieved a 100% union rate in their series of patellar fracture nonunions using these techniques. Uvaraj et al.¹⁴⁹
recently presented outcomes of the surgical management of 22 patellar
nonunions treated at a median of 3 months after injury. Nineteen cases
were treated with tension band wiring with or without cerclage, and
three cases were managed with patellectomy. At a median of 5½-year
follow-up, 91% good to excellent results were reported using the
Bostman criteria.¹⁴⁹
Intraoperatively, if extensive cartilage injury is noted or the
fragment appears to have a poor blood supply, a partial patellectomy
may be considered. Whether revision internal fixation or partial
patellectomy is undertaken, reconstruction of the extensor mechanism
should be performed whenever possible. For long-standing nonunions, a
shortened and contracted quadriceps tendon may necessitate tendon
mobilization and quadricepsplasty.⁵⁹,⁶³

Posttraumatic osteoarthritis can be the sequela of
patellar fractures. The etiology is likely multifactorial and may
include (a) articular cartilage damage at the time of injury, (b)
inadequate reduction and restoration of articular congruity, and (c)
excessive callus formation and secondary articular incongruity. Nummi¹¹¹
reviewed more than 700 patellar fractures and found radiographic
evidence of patellofemoral arthritis in 56.4% of the patients. Most
patellofemoral pain can be managed conservatively with rest, physical
therapy, and nonsteroidal anti-inflammatory medications. Severe
debilitating patellofemoral arthritis can be treated with a
patellectomy.⁵¹ Nonanatomic
attachment of the patellar tendon to the patellar remnant after partial
patellectomy can lead to abnormal patellar tilt and has been implicated
as a cause of patellofemoral arthritis. In 1958, Duthie and Hutchinson⁴³
reported tilting of the patella in five of seven patients with
postoperative arthritis, and attributed these changes to malalignment
from attachment of the patellar tendon to the anterior cortex of the
patella.