Phẫu Thuật Gân Bàn Tay (Tendon Surgery of the Hand)

38
Section 1:  Basic Science
at days 3 and 21, not at days 7 and 14, even with the
2-0 suture. In other words, modified Pennington locks
do not improve repair strength during the second week
(days 7 to 14), when the tendon stumps usually soften
and most ruptures occur. We should be cautious in
relying upon Pennington locks to increase strength.
Recently we examined the effectiveness of the origi­
nal Pennington lock in a four-strand repair; we detected
no differences in strength between repairs with the Pen­
nington locks and those with random assignment of
grasping and locking. When a two-strand Kessler repair
with original Pennington locks was compared with the
repair with grasping anchors using 4-0 suture, our tests
indicated a modest difference (12%) in gap resistance
but no difference in ultimate strength.29 When compar­
ing two grasping anchors (open circle and open loop),
we found no difference in their gap resistance and ulti­
mate strength. Taken together, the evidence above leads
us to believe that differences between the grasping and
the original Pennington’s locking tendon–suture junc­
tions may not be substantial in a clinical setting, and
incorporating Pennington locks in multistrand repairs
does not appear to affect strength.
STRENGTHS OF CORE SUTURES
Strengths of core sutures have been tested extensively
both in vitro and in vivo and in single cycle load-to-
failure test2-14,23-28,32-34 or cyclic test setup.35-37 Using a
4-0 nylon suture, a two-strand repair has a strength of
about 20 to 25 N, a four-strand repair has a strength
of about 35 to 45 N, and a six-strand repair has a
strength of about 50 to 70 N. Thus we see the strength
is roughly proportionate to the number of strands
(Table 4-1).2-8,13,14,32-35 Strength is greatly increased when
suture caliber increases36-38 or stronger suture material is
used (see Table 4-1).38,39
The common core sutures are as follows:
1. Kessler-type repair: The two-strand modified Kessler
repair has been the most common over the past
decades. Four-strand Kessler repairs were devel­
oped, either as a double Kessler or as one Kessler
repair made with a double-stranded suture
(Figure 4-3).
2. Cruciate repair: The original cruciate repair is a
four-strand grasping repair (see Figure 4-3).4 Locks
can be incorporated to make a locking repair.14,35
3. Strickland repair: This four-strand repair consists of
a Kessler-type repair (with reinforced double-circle
locking anchors at four corners) and a simple
double right-angle suture (Figure 4-4).41,42
4. Savage repair and modifications: Savage repair is a
six-strand locking repair.32 Its modifications
include a six-strand single-cross repair, with one
cross-lock as suture anchor (see Figure 4-4).43
when the primary requirements of the repairs are met,
the differences in the strength between locking and
grasping repairs are actually narrow or insignificant.
Depending on the locks or grasps used, the strength of
repairs changes (usually less than 10 N).23-29 The increase
in strength brought about by incorporation of locking
anchors is much smaller than increases in strength
through other means—such as increasing the strength
of suture materials or the number of suture strands
passing through the repair site.2-7 From this perspective,
surgeons should pay ample attention to factors that
exert the greatest influence on strengths and ensure that
these requirements are met primarily. It is not appropri­
ate to rely on incorporating locks, hoping that this
measure alone increases strength, while neglecting other
more influential factors.
Whether various locks have different mechanical
strengths is under question. We tested the strength of
the different locks and found they have essentially the
same holding power.27 A cross-lock and a circle-lock are
identical in holding power; an exposed cross-lock and
an embedded cross-lock are the same as well.27 However,
our in vitro tests showed that the Kessler-type repairs
with the original Pennington locks are weaker than
repairs with cross- or circle-locks.28
Pennington locks are frequently incorporated in
Kessler-type repairs. Most surgeons believe that Pen­
nington locks help secure the repair. However, this con­
sideration actually lacks sufficient experimental support.
According to the definitions of Pennington30 in 1979,
in the locking configuration of the Kessler repair, “the
transverse suture component passes superficial to the
longitudinal component so that the sutures lock a
bundle of tendon fibers when tensile forces are applied
to the suture.” Pennington did not present its strength
data when describing this repair.30 In a letter communi­
cation directed to Pennington,31 Silfverskiöld com­
mented regarding this locking repair: “As to Dr.
Pennington’s claim that the locking loop design is
crucial to the tensile strength of the suture, I agree that
in theory this should be true, but in actual practice this
may not be the case. … Another practical consideration
is that even if one attempts to place the transverse part
superficial to the longitudinal part, to create the locking
design, it is often difficult to be certain of success.”
Practically, we agree that it is difficult for surgeons to
ascertain the relation of transverse and longitudinal
sutures in the repaired digital tendon clinically.
Hatanaka and Manske9,24 modified the Pennington
locking repair by placing some of the longitudinal
strands over the dorsal tendon surface, making locked
tendon–suture junctions confirmatively. The modified
repair is stronger than the grasping Kessler repair when
the suture calibers are 2-0 and 3-0 but not 4-0.24 It is
noteworthy that in an in vivo setting,9 the strength
gained by this modified locking repair was noted only

39
Chapter 4:  Biomechanics of Core and Peripheral Tendon Repairs
Table 4-1  Some Biomechanical Tests Showing Increases in the Strength With the Number of Strands and the
Strengths of Repair Devices
Investigators
Tendons
Methods
Suture Materials
Core/Epitendinous
Ultimate
Strength (N)
Gap Force (N)
(Gap Sizes)
Savage,32 1985
Procine
6-strand (Savage)
4-0 Ethibond
60–70
44 (3-mm)
2-strand (Kessler)
4-0 Ethibond
23
10 (3-mm)
Aoki et al,18 1994
Human
Tendon splint
Dacron
81–84
20–31 (initial)
2-strand Kessler
4-0 Ethibond/6-0
Prolene
26
14 (initial)
Greenwald et al,7 1995
Monkey
4-strand (MGH)
5-0 Nylon/6-0 Nylon
30
2-strand (Kessler)
5-0 Nylon/6-0 Nylon
16
Thurman et al,35 1998
Human*
6-strand (Savage)
4-0 Tricron/6-0
Surgilene
79
(0.3 mm)
4-strand
(Strickland)
4-0 Tricron/6-0
Surgilene
43
(0.3 mm)
2-strand (Kessler)
4-0 Tricron/6-0
Surgilene
34
(2.7 mm)
McLarney et al,4 1999
Human
4-strand
(cruciate)
4-0 Ethobond/6-0
Polypropylene
56
44 (2-mm)
Barrie, et al,14 2000
Human*
4-strand
(cruciate, grasp)
4-0 Ethibond/6-0
Nylon
70
49 (3-mm)
4-strand
(cruciate, lock)
4-0 Ethibond/6-0
Nylon
79
52 (3-mm)
Xie et al,6 2002
Human
6-strand (Tang)
4-0 Supramid/6-0
Nylon
60
45 (2-mm)
Wang et al,33 2003
Porcine
6-strand
(M-Tang)
4-0 Supramid/6-0
Nylon
62
46 (2-mm)
Cao and Tang,34 2005
Porcine
4-strand
(U-shaped)
4-0 Supramid/6-0
Nylon
43
37 (2-mm)
Lawrence and Davis,39 2005
Human
4-strand
(single-cross)
4-0 Fiberwire/6-0
Nylon
81
63 (initial)
Su et al,21 2005
Human
Teno Fix
Teno Fix
55
47 (2-mm)
Teno Fix/5-0
Polypropylene
67
55 (2-mm)
Hirpara et al,20 2010
Porcine
2 barbed repair
devices
Nitinol
58
30 (3-mm)
Wu et al,28 2011
Porcine
4-strand
(cross-lock)
4-0 Ethilon/6-0
Nylon
40
32 (2-mm)
*In situ test of the repair strength and gapping of flexor tendons in cadaveric hands under cyclic tension.
5. Becker repair and Massachusetts General Hospital
(MGH) repair: Becker et al44 used a series of cross-
stitches to repair a beveled tendon. The MGH
repair is its modification that uses four-strand with
a series of cross-stitches (cross-locks).7
6. Multiple looped sutures and their modifications: These
repairs are accomplished using Tsuge’s looped
sutures. Two or three groups of looped sutures can
be used. We typically used three groups of looped
sutures in repairing tendons (see Figure 4-4),45
and fewer looped sutures in modifications of such
repairs (Figure 4-5).33,34,46
7. Repairs with double-stranded sutures: These repairs
simplify surgical maneuvers (Figure 4-6). A

40
Section 1:  Basic Science
STRENGTH OF PERIPHERAL SUTURES
Interrupted, running, or locking running peripheral
sutures are common in the clinic. These peripheral
sutures are usually adequate to smooth the tendon and
to help resist gapping (Figure 4-7).
Others are more complex, which some surgeons have
advocated to increase strength, rather than simply to
approximate the tendon ends. The cross-stitch suture
designed by Silfverskiöld et al15 is among the most
famous. Silfverskiöld’s group15 had good clinical results
with this suture combined with a two-strand modified
Kessler core repair. The cross-stitch suture does provide
the tendon with great strength, but we have a concern
number of configurations can be created and have
similar strength, except that the four-strand Kessler
repair is a bit weaker.28
Core suture configurations, as detailed above, vary enor­
mously. Practically, with an identical number of suture
strands passing through the repair site, the strength of
these sutures varies within a small range. To maintain
ideal strength, two factors are crucial to all core sutures:
(1) adequate suture purchase: a 7- to 10-mm purchase is
necessary to ensure optimal performance of a core
suture, and (2) secure locks or grasps: the locks or grasps
should be of a diameter of 2 mm or more and of suf­
ficient depth. Superficial or small locks or grasps are
pulled out easily.
Figure 4-3  Methods of making a two-strand modified
Kessler repair (A), and a four-strand cruciate repair (B).
A
B
I
II
III
I
II
III
Figure 4-4  Four-strand Strickland repair (A), six-strand
original Savage repair (B), six-strand modified Savage
(Adelaide) repair (C), and six-strand Tang (or three Tsuge)
repair (D).
A
B
D
C
Figure 4-5  Methods of making a modified six-strand
Tang repair forming an M configuration (M-Tang repair).
A U-shaped four-strand repair is made with one looped
suture (A-C), and another looped suture is used to complete
the six-strand repair (D and E).
A
B
C
D
E

41
Chapter 4:  Biomechanics of Core and Peripheral Tendon Repairs
about its massive exposure of sutures over the tendon,
which may increase gliding resistance and the likeli­
hood of adhesions. Similarly, locked cross-stitch16 and
Halsted17 sutures give the tendon great strength. Com­
plexity of surgical maneuvers and suture exposures are
their major drawbacks. Nevertheless, several complex
repair methods, such as cross-stitch and locking cross-
stitches, are ideal options for repairing the thin, flat
extensor tendons.
A running peripheral suture with 2-mm purchase
generates optimal strength according to Merrell et al.47
Cross-stitch repairs usually have a purchase of 5 mm or
more. The number of runs of peripheral suture influ­
ences strength. However, increasing the number of runs
is difficult clinically, because the small diameter of the
digital tendon does not easily accommodate this
increase. The dorsal aspect of the tendon is also harder
to access.
We prefer to use multistrand core repairs to give the
tendon strength, supplemented with interrupted or
simple running peripheral stitches. We believe that
when a strong multistrand repair is used, supplementa­
tion of a complex peripheral suture such as cross-stitch
or Halsted repair is unnecessary. Giesan et al48 reported
not using any peripheral sutures with a six-strand core
repair in repairing flexor pollicis longus tendons.
FACTORS AFFECTING THE STRENGTHS
OF TENDON REPAIRS—SUMMARY
Factors affecting the strength of a surgical repair are as
follows (Figure 4-8): (1) the number of suture strands
across the repair sites—strength is roughly proportional
to the number of core sutures2-8,13,14,32-35; (2) the tension
of repairs—most relevant to gap formation and stiffness
of repairs46; (3) the core suture purchase2,3,49-51; (4) the
types of tendon–suture junction—locking or grasp­
ing14,23-29; (5) the diameter of suture locks in the
tendons—a small diameter of locks diminishes anchor
power5; (6) the suture calibers (diameter)36-38; (7) the
material properties of sutures39,40; (8) the peripheral
sutures15-17,52; (9) the curvature of tendon gliding paths—
the repair strength decreases as tendon curvature
increases53,54; and (10), above all, the holding capacity
of the tendon, affected by degrees of trauma and post-
traumatic tissue softening, is vital to strength. Following
are five important factors affecting the strength of
tendon repair.
IMPORTANT FACTOR 1—TENSION OF SUTURE
REPAIR AFFECTS GAP FORMATION
Peripheral sutures prevent tendon gapping. However,
another factor of perhaps equal importance and closely
related to gap formation has rarely been discussed:
initial tension of the core suture. In repairing a tendon,
we suggest that surgeons maintain a certain tension
across the repair site, rather than leaving the repair
Figure 4-6  Methods of making a variety of four-strand
repair using one needle carrying two suture strands (looped
or separated). These repairs achieve greater strength
with simplified surgical maneuvers. A, Cross-lock repair;
B, U-shaped repair; C, Kessler-type repair; and D, cruciate
repair. In A, B, and D, one needle carrying two separate
strands are used.
A
B
C
D
Figure 4-7  Methods of making peripheral sutures:
A, simple running; B, running locking; C, Silfverskiöld
(cross-stitch); and D, locking cross-stitch.
A
C
B
D

42
Section 1:  Basic Science
four-strand repairs (double-modified Kessler, locking
cruciate, and modified Savage) had gained significant
resistance to gapping and ultimate strength as the length
of core sutures increased from 0.4 to 1 cm. We deter­
mined that the optimal core suture purchase for a trans­
versely cut tendon is between 0.7 and 1 cm (Figure 4-9).
In obliquely cut tendons, increasing the purchase
length to 1.2 cm improved the strength of a two-strand
modified Kessler repair and a four-strand locking cruci­
ate repair.49
The importance of core suture purchase was verified
by other investigators. Kim et al50 reported progressive
increases in the strength of double Kessler repairs as
purchase lengthened from 0.3, to 0.7, to 1, and 1.3 cm
in porcine tendons; they recommend 1 cm as the
appropriate purchase length. Using cadaveric hands,
Lee et al51 studied locking cruciate repairs, randomly
assigned to placement of 0.3, 0.5, 0.7 or 1 cm from the
cut edge of the tendon, with interlocking horizontal
mattress sutures. The tendons repaired at 1 cm had the
lowest increase in work of flexion (5.2%), the highest
2-mm gap force (89.8 N) and ultimate load (111.5 N).
In contrast, the tendons repaired at 0.3 cm had the
highest increase in the work (22.1%), the lowest 2-mm
gap force (54.6 N) and ultimate load (84.6 N). In fact,
all studies suggest an optimal core suture purchase of
1.0 cm (range: 0.7 to 1.2 cm).
As stated in previous paragraphs, sizes of locks and
grasps are another important factor; they should be
equal to or greater than 2 mm to guarantee anchors of
the sutures to the tendons (see Figure 4-9).
IMPORTANT FACTOR 3—STRENGTH DOES NOT
INCREASE DURING EARLY TENDON HEALING
Only a few investigations have been conducted of surgi­
cal repair strength in vivo, using animal models.9-14 Data
on the strength of the healing tendons in human
patients are not available. Urbaniak et al55 found that
the strength of repaired canine flexor tendons decreased
at week 2 postsurgery. With simulated postsurgical
motion, Boyer et al10 found no decrease in the strength
of the canine flexor tendons. In a chicken model,
tension-free. We tested the gap formation forces of the
repair under varying tension, which resulted in 0%,
10%, or 20% shortening of the sutured tendon segment.
The tendon with tension causing 10% shortening
reduced the chance of repair gapping, with significantly
increased gap formation forces. Increasing the tension
to cause 20% tendon shortening did further increase
gap resistance forces, but by a much smaller amount. A
four-strand core repair tensioned to cause 10% tendon
shortening, regardless whether a simple peripheral
suture is added or not, had identical gap resistance.
The gap resistance of a repair closely relates to its
tension status. Adding a slight baseline tension to the
repair site when the proximal tendon is temporarily
fixed to ease the tension during surgery results in appro­
priate tension after surgery, because this counteracts the
tension of muscle pull during active motion.46 Multi­
strand repairs are placed in different areas in the tendon
cross-section, which may evenly distribute tension
throughout the tendon and may not require a periph­
eral suture. This concept of “core suture–only” multi­
strand repairs, if practiced under certain repair tension,
is proper.
IMPORTANT FACTOR 2—SUFFICIENT
SUTURE PURCHASE AND LOCK (GRASP)
SIZE ARE ESSENTIAL
Length of core suture purchase is defined as the exit/entry
distance of the core suture from the cut ends of the
tendon. We tested how the length of core suture pur­
chase affects the strength of the transversely or obliquely
cut tendon.2,3,49 We examined the strength of porcine
flexor tendons with two-strand modified Kessler core
repairs with purchase of 0.4, 0.7, 1.0, and 1.2 cm and
with four-strand circle-locking repairs with purchase of
0.4 and 1.0 cm.2 As the suture purchase increased from
0.4 to 0.7, 1.0, and 1.2 cm, the gap and ultimate strengths
increased significantly. The strengths remained constant
with the purchase from 0.7 to 1.2 cm. The four-strand
circle-locking repairs with a suture purchase of 1.0 cm
were significantly stronger than those with a suture pur­
chase of 0.4 cm. In another study,3 we found that three
Figure 4-8  Summary of factors determining the strength of surgically repaired tendons.
Material properties of sutures
TENDON REPAIR
STRENGTH
Core sutures
• Suture strands:
4 or 6 strands suggested
• Suture purchase lengths:
0.7 to 1.0 cm purchase suggested
• Tension of the repairs:
Slight tension (10% shortening
of tendon segment) suggested
• Suture calibers: 3–0 or 4–0
• Locking/grasping: locking preferred
• Sizes of the locks or grasp: >2 mm
Peripheral sutures
• Number of runs:
6–8 runs suggested
• Suture purchase lengths:
1–2 mm suggested
• Tension: slight tension
• Suture calibers: 6–0 or 5–0
Curvature of the tendon
Holding capacity of the tendon
(tendon trauma, softening, etc.)

43
Chapter 4:  Biomechanics of Core and Peripheral Tendon Repairs
smaller. These findings imply that repair strength
decreases as the finger flexes further, because both the
angles of tension on the tendon increase and the radii
of the tendon gliding curves decrease progressively. The
dorsal part of the tendon is subjected to greater tension
and may gap more easily during finger flexion. Placement
of sutures dorsally may favor gap resistance.
The clinical implications of these biomechanical
observations are that strength decreases as the fingers
are flexed progressively, and that repairs are weakest
when the finger is in marked flexion. Therefore, the final
part of finger flexion is when repairs are more vulnera­
ble to disruption. Catching of the tendon repair site by
the sheath or pulley edges during finger flexion may
trigger repair rupture.
IMPORTANT FACTOR 5—SOFTENING OF
TRAUMATIZED TENDON PARTS DECREASES
HOLDING POWER
It should be emphasized that tissue softening is crucial
to repair strength. Because most biomechanical studies
were in vitro, the influence of this factor could not be
reflected in these studies. Tissue softening renders the
repair weaker with decreases in both gap resistance and
maximal strength. As the locking or grasping anchors
move farther from the trauma area, tissue softening
decreases. Suture purchase of the core repairs should be
Hatanaka et al9 found no changes in the strength of the
healing tendons in the first 3 weeks. Our study12 using
chickens showed no significant changes in the strength
of the healing flexor tendon during the initial 4 weeks.
The strength of the digital flexor tendons in animal
models does not typically increase during the first 3
or 4 weeks postsurgery; under certain circumstances,
the strength may actually decrease slightly. This “no-
gain” lag period in strength in the initial weeks after
surgery is characteristic, reflecting the slow healing
responses of the tendon. Unfortunately, the slow bio­
logical healing response invites invasion of tissues
outside the tendon and development of adhesions,
resulting in gapping or rupture of the repairs during
early tendon mobilization.
IMPORTANT FACTOR 4—STRENGTH
DECREASES WHEN THE FINGER FLEXES
Because finger flexion is the key action of postsurgical
tendon motion exercise, the capability of a repair to resist
gapping or failure during finger flexion is important. To
understand how flexion affects the strength of tendon
repairs, we created two models to test (1) how the direc­
tion of tension on the tendon affects strength53 and (2)
how tendon curvature affects strength (Figure 4-10).54
Strength of the repairs decreased as the angles of pulling
became greater, and when the radius of the curvature is
Figure 4-9  A, Sizes of the locks affect the repair strength significantly. We determined that the diameters of locks of 2 or
3 mm generate strength greater than those of 1 mm and recommend an optimal locking size of about 2 mm. B, Our
biomechanical tests of the strength of three different types of locks indicate no significant difference in the repair strength
with the three types of locks. C, We tested the strength of two- or four-strand repairs with core suture purchase lengths
ranging from 4 to 12 mm and found that repairs with the purchase lengths between 7 to 12 mm generate strength greater
than those with the purchase of 4 mm. We recommend the optimal core suture purchase of 7 to 10 mm. The purchase
length of 10 mm is optimal.
Weak
Weak
a
b
c
Diameter of locks: 1 mm
Diameter of locks: 2 mm
Diameter of locks: 3 mm
Circle
C
B
A
Strong
-optimal
Strong
Three
locks
produce
equal
power
Stronger
Strongest
-optimal
Strongest
12 mm
12 mm
10 mm
10 mm
7 mm
7 mm
4 mm
4 mm
Exposed X
Embedded X

44
Section 1:  Basic Science
Increasing the caliber from 4-0 to 3-0 adds 10-15 N to
the ultimate strength of two- or four-strand repairs (see
Table 4-1; Figure 4-11).36-39 Increasing the suture strands
from two to four is more efficient than increasing the
caliber from 4-0 to 3-0. We suggest either a 3-0 or a 4-0
suture for a four-strand repair, and a 4-0 suture for a
six-strand repair. However, rigidity of sutures increases
along with caliber. Sutures of 2-0 become more rigid;
sutures of this or greater caliber are rarely necessary.
Lawrence and Davis39 evaluated the mechanical prop­
erties of five nonabsorbable 4-0 sutures and a four-
strand locking repair with these sutures (Table 4-2). All
repairs failed by suture rupture at the locking loop.
Fiberwire and stainless-steel sutures and repairs were
stronger and stiffer than the other sutures. The results
sufficient to minimize the negative impact of tissue soft­
ening. Clinical evidence indicates that rupture of the
repair is most likely in the second week after surgery,56
thought to be related to softening of traumatized tissues.
TWO WAYS TO INCREASE STRENGTH—
MORE SUTURE STRANDS AND LARGER
SUTURE CALIBERS
Increasing the number of suture strands across the repair
sites is a popular means to increase strength. Repair with
four or six strands are most often used to provide proper
strength for early active tendon motion. Use of double-
or triple-stranded repair simplifies surgery.26,33,34 Increas­
ing suture caliber is another easy way to augment
strength. Clinically, repairs with 4-0 sutures are common.
Figure 4-10  Left, The changes in the tendon arcs during digital flexion. A, The digital flexor tendons follow an arc of
increasing curvature as the fingers flex progressively during digital flexion. B and C, The curvature of tendon arcs is greater in
the distal (site 1) than in the proximal (site 2). As the finger flexes, the tendon curvature increases at each site close to the
joints (for example, site 1). Middle, Our test setup of the flexor tendon undergoing load to failure testing over a curve. By
changing the diameter of the pulley, we re-created different tendon curvature during tendon loading. We recorded substantial
decreases in the repair strength when the tendon curvature becomes smaller. Right, Differential loading conditions on the
tendons according to tendon curvature. A, The tendon is subjected to a linear load when pulled straight. B, The tendon is
subjected to both a linear and a bending force when pulled to glide over a pulley. The tendon is subjected to an angular
tensile load and bending forces. The bending force on the repaired tendon segment is greater in C (of a greater curvature)
than that in D (of a smaller curvature). E–G, The gliding curvature becomes increasingly greater, with increasingly greater
bending force on the repaired tendon segment; the tendon is easier to rupture. G, When the tendon arc approximates the
length of the core suture, the bending force is loaded entirely to the suture. The repair especially is prone to failure.
H, When the tendon glides over a remnant pulley with adjacent pulleys and sheath defects, tendon curvature is especially
small, and the repair is easy to disrupt.
Our model
-INSTRON
4411
Pulley
Tendon
Repair site
Clamp
F
F′
F
F′
Pulley
F
F2
F1
F
F
F
F
c
a
b
Force analysis
A
B
C
D
E
F
G
H
A
B
C
1
1′
2
2′
Finger flexion

45
Chapter 4:  Biomechanics of Core and Peripheral Tendon Repairs
achieved by Prolene and Ethibond were similar with
respect to gap and ultimate forces. The merits of Fiber­
wire have been reported by others40; knot security and
suture rigidity remain concerns. We expect to see the
development of strong yet flexible materials in the
future.
TEST SETUP AND MEASUREMENTS OF
TENDON REPAIR BIOMECHANICS
The centerpiece of the test setup is a material testing
machine. Most data on tendon repair biomechanics
were generated using this machine, under either single-
cycle load-to-failure or cyclic loading conditions. Typi­
cally, a load-to-failure curve is generated (Figure 4-12),
which allows analysis of the failure load (called ultimate
strength or maximal strength of the repair) and stiffness
of the tendons. Stiffness is obtained by measuring the
slope of the middle linear portion of the load-
displacement curve. By calculation of the area under­
neath the curve, energy to failure of the repair can be
obtained. The forces at which the gaps start to appear
or reach 2 or 3 mm are recorded as gap forces. Analyses
of the curves generated when the tendon undergoes
cyclical loads help determine which repair tends to fail
in earlier cycles. Gap sizes can be monitored and
recorded continuously over loading cycles with a video
Figure 4-11  Our test results of two-strand modified
Kessler repair and four-strand cross-lock repairs (using
double-stranded sutures) using 4-0 and 3-0 suture
(Ethilon). Increasing the suture caliber from 4-0 to 3-0 adds
about 10 N to the maximal strength. Increasing the suture
strands from two to four adds about 15 to 20 N to the
maximal strength. The gap resistance of the repairs is
increased by increasing the caliber and the number of the
repair strands.
0
10
20
30
40
50
60
Number of repair strands
Strength (N)
4-0 suture
3-0 suture
Ultimate strength
2 mm gap forces
2-strand
4-strand
2-strand
4-strand
Table 4-2  Biomechanical Tests of Different Materials and Repairs With Varying Suture Calibers
Investigators
Methods/Materials
Ultimate Strength (N)
Initial Gap Force (N)
Stiffness (N/mm)
Lawrence and Davis39
Test of material properties of
sutures
4-0, Nylon
22
2.5
4-0, Prolene
25
5.0
4-0, Ethibond
25
8.2
4-0, Stainless steel
36
14.8
4-0, Fiberwire
37
11.0
Lawrence and Davis39
4-strand (single-cross repair)
4-0, Nylon
46
36
6.0
4-0, Prolene
63
52
8.0
4-0, Ethibond
66
52
10.2
4-0, Stainless steel
87
66
11.9
4-0, Fiberwire
81
63
12.8
Taras et al38
2-strand (Kessler repair)
5-0, Ethibond
16
4-0, Ethibond
22
3-0, Ethibond
31
2-0, Ethibond
41

46
Section 1:  Basic Science
Cyclic loading has been used rather extensively to
test the strength of repairs. Because this type of test
imitates conditions of repetitive tendon loading in
patients, it has yielded valuable information regarding
gap resistance and fatigue. One concern in interpreting
the findings after cycling the tendon over hundreds and
thousands of times continuously is that the test does
not simulate situations seen clinically in healing tendons
during the early period. Clinically, the tendon usually
moves only 20 or 30 times in each session of exercise;
even ordinary daily life does not involve moving the
tendon over hundreds of cycles continuously. Testing the
repair strength or the gapping after more than 20 to 30
cycles of tension appears to make more sense. We believe
that a single cycle of motion test provides the most
useful basic information; a test of the load-to-failure
strength after cyclic loading of a limited number (e.g.,
10 to 30) of cycles may generate information relevant to
postsurgical exercise.
Gap resistance is an important measure, because a
repair with remarkable gaps is bound to fail, and
gapping impairs the healing process. Assessment of gap
forces (at the gap sizes of initial appearance, 2 mm, or
3 mm) or gap sizes under a certain load is necessary.
Nevertheless, the examiners should be aware that gap
formation forces are easily affected by the number of
runs of the peripheral sutures and tension status of the
surgical repair, two factors that should be standardized
to facilitate comparison of the gap resistance of various
repairs. The ultimate strength of a given repair, on the
other hand, is rather consistent in different studies. We
consider both parameters—gap force and ultimate
strength—essential measures of the mechanical perfor­
mance of a repair.
In our mechanical study settings, we load a 5-cm
segment of tendon in a material testing machine and
measure 2-mm gap formation force and ultimate
strength. Sometimes we also measure the initial gap
force and tendon stiffness; occasionally we include
energy to failure as well. The distraction of the tendon
during failure of the peripheral or core sutures is instruc­
tive as well. We believe that findings regarding tendon
stiffness are inconclusive. An increase in stiffness indi­
cates greater capacity to resist tendon deformation;
however, tendons normally have some elasticity.
Whether greater stiffness necessarily benefits a repaired
tendon is actually questionable.
CONSIDERATIONS IN THE DESIGN AND USE
OF REPAIR METHODS
Most surgeons use strong core sutures with simple epi­
tendinous sutures clinically; some others favor strength­
ening peripheral sutures. Rarely, surgeons try to adopt
both complex core and peripheral sutures, because the
tendon is hard to accommodate. Strength increases with
the complexity of peripheral or core sutures, but at the
camera, allowing identification of repairs vulnerable to
gapping.
In situ testing of repaired tendons using cadaveric
hands, developed by Mass et al,57 Trumble et al,35 and
Wolfe et al,14 represents another means to measure the
mechanical properties of repairs. This test simulates a
curvilinear loading condition of the tendon within the
digits. Gap formation, gap forces, ultimate strength, and
work of flexion can be measured,14,35,57 and cyclic loading
on the tendon can be simulated as well.14,35 Measuring
the resistance to tendon movement over a pulley or a
segment of the sheath using a custom-designed device
is another innovative test setup developed by the Amadio
group.58 This test generates sensitive data on the differ­
ences among different repair methods, particularly with
respect to surface friction of the repair.
Figure 4-12  A, A typical load-displacement graph
generated during a single cycle load-to-failure test. From the
graph, we can obtain ultimate strength, the energy to failure
(calculated from the area under the load-displacement
curve), and stiffness of the repairs (the slope of the linear
middle part of the load-displacement curve). B, Form this
graph, we can also determine the force at which peripheral
repairs fail completely and the tendon distraction during
peripheral or core suture failure. The initial, 1-, 2-, or 3-mm
gap formation forces are usually recorded to determine gap
resistance of the repair. These forces are measured during
tendon loading with the assistance of a video camera.
0
10
20
30
40
50
Load (N)
0
0
10
20
30
Load (N)
0
5
10
15
20
10
20
30
Stiffness (N/mm)
Linear region
Ultimate
strength
Area = Energy
to failure (J)
Displacement (mm)
Displacement (mm)
Force of complete disruption
or peripheral sutures
Ultimate strength
Tendon distraction during
peripheral suture disruption
A
B

47
Chapter 4:  Biomechanics of Core and Peripheral Tendon Repairs
In recent decades, investigations have been directed
mainly to increasing surgical repair strength. However,
as clinical reports have indicated, repair ruptures have
not been entirely eliminated by using stronger repairs.
Besides tension on the repaired tendon, resistance to
tendon gliding from the peritendinous tissue is a major
cause of repair gapping and failure. A decrease in resis­
tance to tendon gliding can be efficient in reducing
repair ruptures. If we decrease the resistance to tendon
gliding adequately, the demand for strength of surgical
repairs may no longer be as great; the strength that
present repairs provide is thus sufficient. Currently,
venting of critical parts of the pulleys, as Elliot and our
groups have practiced, or attempts to modify the gliding
surface of the tendon, as the Amadio group is exploring,
may provide alternative yet efficient ways to enhance
surgical strength, towards the goal of minimizing repair
ruptures during early active motion.
cost of more difficult surgical maneuvers, tendon bulki­
ness, and exposure of sutures or knots over the tendon.
With these considerations in mind, one should be careful
in designing and advocating complex suture techniques,
because they would not be widely accepted clinically and
the strength sometimes exceeds what is actually needed.
Assembling elements from existing methods to form a
complex (and “novel”) repair is not difficult. Such
methods may not have great practical value.
A novel and clinically acceptable repair should be
either (1) stronger than existing repairs (with an identi­
cal degree of complexity) or (2) simpler in surgical
maneuvers (without decreasing the strength). We favor
use of double-stranded sutures—either a loop at the end
or two separate strands, carried by a single needle, to
increase repair strength, with fewer needle-suture pas­
sages through the tendon and the same surgical
maneuvers.
References
1. Schuind F, Garcia-Elias M, Cooney WP 3rd, et al: Flexor
tendon forces: in vivo measurements, J Hand Surg (Am)
17:291–298, 1992.
2. Tang JB, Zhang Y, Cao Y, et al: Core suture purchase affects
strength of tendon repairs, J Hand Surg (Am) 30:1262–1266,
2005.
3. Cao Y, Zhu B, Xie RG, et al: Influence of core suture purchase
length on strength of four-strand tendon repairs, J Hand Surg
(Am) 31:107–112, 2006.
4. McLarney E, Hoffman H, Wolfe SW: Biomechanical analysis
of the cruciate four-strand flexor tendon repair, J Hand Surg
(Br) 24:295–301, 1999.
5. Xie RG, Xue HG, Gu JH, et al: Effects of locking area on
strength of 2- and 4-strand locking tendon repairs, J Hand
Surg (Am) 30:455–460, 2005.
6. Xie RG, Zhang S, Tang JB, et al: Biomechanical studies of 3
different 6-strand flexor tendon repair techniques, J Hand
Surg (Am) 27:621–627, 2002.
7. Greenwald DP, Randolph MA, Hong HZ, et al: Augmented
Becker versus modified Kessler tenorrhaphy in monkeys:
Dynamic mechanical analysis, J Hand Surg (Am) 20:267–272,
1995.
8. Hausmann JT, Vekszler G, Bijak M, et al: Biomechanical com­
parison of modified Kessler and running suture repair in 3
different animal tendons and in human flexor tendons,
J Hand Surg (Am) 34:93–101, 2009.
9. Hatanaka H, Zhang J, Manske PR: An in vivo study of locking
and grasping techniques using a passive mobilization protocol
in experimental animals, J Hand Surg (Am) 25:260–269, 2000.
10. Boyer MI, Gelberman RH, Burns ME, et al: Intrasynovial
flexor tendon repair. An experimental study comparing low
and high levels of in vivo force during rehabilitation in
canines, J Bone Joint Surg (Am) 83:891–899, 2001.
11. Cerovac S, Afoke A, Akali A, et al: Early breaking strength of
repaired flexor tendon treated with 5-fluorouracil, J Hand
Surg (Br) 26:220–223, 2001.
12. Tang JB, Cao Y, Zhu B, et al: Adeno-associated virus-2-
mediated bFGF gene transfer to digital flexor tendons signifi­
cantly increases healing strength. An in vivo study, J Bone Joint
Surg (Am) 90:1078–1089, 2008.
13. Winters SC, Gelberman RH, Woo SL, et al: The effects of
multiple-strand suture methods on the strength and excur­
sion of repaired intrasynovial flexor tendons: a biomechani­
cal study in dogs, J Hand Surg (Am) 23:97–104, 1998.
14. Barrie KA, Wolfe SW, Shean C, et al: A biomechanical com­
parison of multistrand flexor tendon repairs using an in situ
testing model, J Hand Surg (Am) 25:499–506, 2000.
15. Silfverskiöld KL, May EJ, Törnvall AH: Gap formation during
controlled motion after flexor tendon repair in zone II: A
prospective clinical study, J Hand Surg (Am) 17:539–546,
1992.
16. Dona E, Turner AW, Gianoutsos MP, et al: Biomechanical
properties of four circumferential flexor tendon suture tech­
niques, J Hand Surg (Am) 28:824–831, 2003.
17. Wade PJ, Wetherell RG, Amis AA: Flexor tendon repair: sig­
nificant gain in strength from the Halsted peripheral suture
technique, J Hand Surg (Br) 14:232–235, 1989.
18. Aoki M, Manske PR, Pruitt DL, et al: Tendon repair using
flexor tendon splints: An experimental study, J Hand Surg
(Am) 19:984–990, 1994.
19. Parikh PM, Davison SP, Higgins JP: Barbed suture tenorrha­
phy: An ex vivo biomechanical analysis, Plast Reconstr Surg
124:1551–1558, 2009.
20. Hirpara KM, Sullivan PJ, O’Sullivan ME: A new barbed device
for repair of flexor tendons, J Bone Joint Surg (Br) 92:1165–
1170, 2010.
21. Su BW, Protopsaltis TS, Koff MF, et al: The biomechanical
analysis of a tendon fixation device for flexor tendon repair,
J Hand Surg (Am) 30:237–245, 2005.
22. Su BW, Solomons M, Barrow A, et al: Device for zone-II flexor
tendon repair. A multicenter, randomized, blinded, clinical
trial, J Bone Joint Surg (Am) 87:923–935, 2005.
23. Hotokezaka S, Manske PR: Differences between locking loops
and grasping loops: effects on 2-strand core suture, J Hand
Surg (Am) 22:995–1003, 1997.
24. Hatanaka H, Manske PR: Effect of suture size on locking and
grasping flexor tendon repair techniques, Clin Orthop Relat Res
375:267–274, 2000.
25. Mashadi ZB, Amis AA: The effect of locking loops on the
strength of tendon repairs, J Hand Surg (Br) 16:35–39, 1991.
26. Tanaka T, Amadio PC, Zhao C, et al: Gliding characteristics
and gap formation for locking and grasping tendon repairs:
A biomechanical study in a human cadaver model, J Hand
Surg (Am) 29:6–14, 2004.

48
Section 1:  Basic Science
44. Becker H, Orak F, Duponselle E: Early active motion follow­
ing a beveled technique of flexor tendon repair: Report on
fifty cases, J Hand Surg (Am) 4:454–460, 1979.
45. Tang JB, Shi D, Gu YQ, et al: Double and multiple looped
suture tendon repair, J Hand Surg (Br) 19:699–703, 1994.
46. Tang JB: Indications, methods, postoperative motion and
outcome evaluation of primary flexor tendon repairs in Zone
2, J Hand Surg (Eur) 32:118–129, 2007.
47. Merrell GA, Wolfe SW, Kacena WJ, et al: The effect of increased
peripheral suture purchase on the strength of flexor tendon
repairs, J Hand Surg (Am) 28:464–468, 2003.
48. Giesen T, Sirotakova M, Copsey AJ, et al: Flexor pollicis
longus primary repair: further experience with the Tang tech­
nique and controlled active mobilisation, J Hand Surg (Eur)
34:758–761, 2009.
49. Tang JB, Tan J, Xu Y: Lengthening and locking: two ways
to improve repair strength of obliquely lacerated tendons,
J Hand Surg (Am) 28:832–837, 2003.
50. Kim JB, de Wit T, Hovius SE, et al: What is the significance
of tendon suture purchase? J Hand Surg (Eur) 34:497–502,
2009.
51. Lee SK, Goldstein RY, Zingman A, et al: The effects of core
suture purchase on the biomechanical characteristics of
a multistrand locking flexor tendon repair: A cadaveric study,
J Hand Surg (Am) 37:1165–1171, 2010.
52. Tang JB, Wang B, Chen F, et al: Biomechanical evaluation of
flexor tendon repair techniques, Clin Orthop Relat Res 386:
252–259, 2001.
53. Tang JB, Xu Y, Wang B: Repair strength of tendons of varying
gliding curvature: A study in a curvilinear model, J Hand Surg
(Am) 28:243–249, 2003.
54. Tang JB, Cao Y, Xie RG: Effects of tension direction on
strength of tendon repair, J Hand Surg (Am) 26:1105–1110,
2001.
55. Urbaniak JR, Cahill JD, Mortenson RA: Tendon suturing
methods: analysis of tensile strengths. In AAOS Symposium
on Tendon Surgery in the Hand, St Louis, 1975, CV Mosby,
pp 70–80.
56. Elliot D, Moiemen NS, Flemming AF, et al: The rupture
rate of acute flexor tendon repairs mobilized by the con­
trolled active motion regimen, J Hand Surg (Br) 19:607–612,
1994.
57. Komanduri M, Phillips CS, Mass DP: Tensile strength of flexor
tendon repairs in a dynamic cadaver model, J Hand Surg (Am)
21:605–611, 1996.
58. Momose T, Amadio PC, Zhao C, et al: Suture techniques with
high breaking strength and low gliding resistance: experi­
ments in the dog flexor digitorum profundus tendon, Acta
Orthop Scand 72:635–641, 2001.
27. Xie RG, Tang JB: Investigation of locking configurations for
tendon repair, J Hand Surg (Am) 30:461–465, 2005.
28. Wu YF, Cao Y, Zhou YL, et al: Biomechanical comparisons of
four-strand tendon repairs with double-stranded sutures:
Effects of different locks and suture geometry, J Hand Surg
(Eur) 36:34–39, 2011.
29. Wu YF, Tang JB: How much does a Pennington lock add to
strength of a tendon repair? J Hand Surg (Eur) 36:476–484,
2011.
30. Pennington DG: The locking loop tendon suture, Plast Recon­
str Surg 63:648–652, 1979.
31. Silfverskiöld K: Acknowledgement to previous work–in reply,
J Hand Surg (Am) 19:1055–1056, 1994.
32. Savage R: In vitro studies of a new method of flexor tendon
repair, J Hand Surg (Br) 10:135–141, 1985.
33. Wang B, Xie RG, Tang JB: Biomechanical analysis of a modi­
fication of Tang method of tendon repair, J Hand Surg (Br)
28:347–350, 2003.
34. Cao Y, Tang JB: Biomechanical evaluation of a four-strand
modification of the Tang method of tendon repair, J Hand
Surg (Br) 30:374–378, 2005.
35. Thurman RT, Trumble TE, Hanel DP, et al: Two-, four-, and
six-strand zone II flexor tendon repairs: An in situ biome­
chanical comparison using a cadaver model, J Hand Surg
(Am) 23:261–265, 1998.
36. Barrie KA, Tomak SL, Cholewicki J, et al: Effect of suture
locking and suture caliber on fatigue strength of flexor tendon
repairs, J Hand Surg (Am) 26:340–346, 2001.
37. Alavanja G, Dailey E, Mass DP: Repair of zone II flexor digi­
torum profundus lacerations using varying suture sizes: A
comparative biomechanical study, J Hand Surg (Am) 30:448–
454, 2005.
38. Taras JS, Raphael JS, Marczyk SC, et al: Evaluation of suture
caliber in flexor tendon repair, J Hand Surg (Am) 26:1100–
1104, 2001.
39. Lawrence TM, Davis TR: A biomechanical analysis of suture
materials and their influence on a four-strand flexor tendon
repair, J Hand Surg (Am) 30:836–841, 2005.
40. Miller B, Dodds SD, deMars A, et al: Flexor tendon repairs:
the impact of Fiberwire on grasping and locking core sutures,
J Hand Surg (Am) 32:591–596, 2007.
41. Strickland JW: Flexor tendon repair: Indiana Method, Indiana
Hand Center Newsletter 1:1–19, 1993.
42. Trumble TE, Vedder NB, Seiler JG 3rd, et al: Zone-II flexor
tendon repair: a randomized prospective trial of active place-
and-hold therapy compared with passive motion therapy,
J Bone Joint Surg (Am) 92:1381–1389, 2010.
43. Sandow MJ, McMahon MM: Single-cross grasp six-strand
repair for acute flexor tendon tenorrhaphy, Atlas Hand Clin
1:41–64, 1996.

49
CHAPTER
5 
BIOLOGY AND BIOMECHANICS
OF THE TENDON–BONE
INSERTION
Stavros Thomopoulos, PhD, and Leesa M. Galatz, MD
STRUCTURE AND FUNCTION OF THE
TENDON–BONE INSERTION
The gradual tailoring, or “functional grading,” of mate­
rial systems has been described in the engineering mate­
rials community as a way to minimize these stress
concentrations.3 A central concept to the formation of
a robust attachment between tendon and bone is that
smooth, monotonic transitions in mechanical proper­
ties are necessary for minimizing stresses. The tendon–
bone insertion site is a natural functionally graded
structure. Notably, this gradation is not maintained at
healing insertions, and the surgically repaired tissue is
prone to rerupture.4-6 Surgical technique and treatment
approaches should therefore seek to minimize stress
concentrations and thereby protect the repair site from
rupture.
MORPHOLOGY OF THE TENDON ENTHESIS
(TENDON–BONE INSERTION SITE)
Tendons insert into bone across broad fibrous transi­
tions or across short fibrocartilaginous transitions.7
Fibrous transitions are characterized by insertion sites
with large footprints (effectively distributing the forces
over large areas and reducing stresses) and by perforat­
ing mineral fibers (“Sharpey’s fibers”). Examples of
these attachments include the tibial insertion of the
medial collateral ligament and the deltoid tendon inser­
tion into the humeral head. Examples of the more
common fibrocartilaginous insertions include the
rotator cuff tendon insertions and the Achilles tendon
insertion. The structure, function, development, and
healing of fibrocartilaginous tendon–bone insertions
will be the focus of this chapter. These insertions are
characteristically short and occur across distinct fibro­
cartilage regions classically categorized into four zones
(Figure 5-1).7 The first zone consists of tendon; this
region is populated by fibroblasts and has properties
similar to those found at the tendon mid-substance. It
is composed primarily of well aligned type I collagen
OUTLINE
The repair of tendon to bone presents a major chal­
lenge due to the disparity in the mechanical properties
of the two tissues. The uninjured insertion overcomes
this mechanical mismatch by attaching the two tissues
across a functionally graded insertion. This unique
transition develops between tendon and bone postna­
tally and is driven by both biological and mechanical
factors. Muscle loading, for example, is necessary for
the development of a well-organized, mineralized
insertion. Tendon–bone healing does not recapitulate
development. Healing is characterized by bone loss
and the formation of a fibrovascular scar, leading to a
repaired attachment that is prone to rupture. Current
treatment approaches for enhancing tendon–bone
healing focus on rehabilitation. Studies in animal
models, however, have demonstrated that a fine
balance must be reached between too much load
(which can lead to microdamage) and too little
load (which can lead to a catabolic environment) to
maximize tendon–bone healing. Future treatment
approaches may include delivery of growth factors
and/or mesenchymal stem cells to stimulate regenera­
tion of a functionally graded insertion. Ultimately, a
tissue engineered scaffold with a gradation in proper­
ties and seeded with the appropriate cells and biofac­
tors may provide a solution to the clinical problem of
tendon–bone healing.
The attachment of a compliant material like tendon to
a relatively stiff material like bone is a fundamental
engineering challenge.1 As is evident from the study of
attachment of engineering materials,2 potentially dam­
aging stress concentrations can be expected to arise at
the insertion if the tissue is not tuned to this stiffness
mismatch both at the macroscopic level (i.e., the
outward splay of the tendon) and the microscopic level
(i.e., the gradual stiffening of collagen molecules by
mineral crystals).

50
Section 1:  Basic Science
tissue. There was a variation in gene expression along
the length of the insertion; tendon-specific genes were
localized to the tendon proper, bone-specific genes were
localized to the bone proper and the mineralized por­
tions of the insertion, and cartilage specific genes were
localized to the insertion.12 Collagen fibers were less
oriented at the insertion compared to the tendon.12
Mineral content increased linearly from tendon to
bone.13 Biomechanically, both the elastic and viscous
behaviors of the tissue varied from tendon to bone.12
The tensile properties of the tendon, for example, were
stiffer than those of the insertion. Structure-function
relationships were also examined at the anterior cruciate
ligament–bone insertion.14 Microcompression was per­
formed to determine the compressive properties and
energy dispersive x-ray analysis was performed to deter­
mine mineral content along the ligament enthesis. The
calcified regions of the insertion exhibited significantly
greater compressive moduli than the noncalcified
regions. Based on these results, it is apparent that there
is a gradation along the tendon–bone insertion site with
regard to collagen structure, extracellular matrix compo­
sition, mineral content, and biomechanical properties.
Composition and Microstructure of the
Insertion Site Is Optimized to Minimize
Stress Concentrations
Mechanical models of the insertion demonstrate that
the orientation of the collagen fibers and their level of
mineralization are optimized to minimize stress and
strain concentrations. In one study, a finite element
fibers.8 The second zone consists of fibrocartilage; this
region is populated by fibrochondrocytes and is com­
posed of types I, II, and III collagen and the proteogly­
can aggrecan.8-11 The third zone is also populated by
fibrochondrocytes and is composed of mineralized
fibrocartilage. Here, the predominant collagen is type II,
and there are significant amounts of type X collagen as
well as aggrecan.8-11 The fourth zone consists of bone
proper; this region is populated by osteoblasts and
osteoclasts and is composed predominantly of mineral­
ized type I collagen. While the insertion site is typically
categorized into these four zones, changes in the tissue
are gradual and continuous, minimizing stress concen­
trations between the tendon and the bone. The com­
plexity of the tendon–bone insertion, however, results
in a difficult challenge for effective healing.
Gradations in Biomechanical, Compositional,
and Structural Properties Along the Tendon–
Bone Insertion Site
The tendon–bone insertion contains gradients in bio­
mechanical, compositional, and structural properties.
Between the unmineralized and mineralized fibrocar­
tilage of the enthesis is a narrow zone that darkens
markedly during tissue staining. This “line” was tradi­
tionally thought to represent a mineralization front, or,
mechanically, a boundary between soft and hard tissue.
Studies examining the supraspinatus tendon–bone
insertions of rats, however, have demonstrated grada­
tions in properties along the enthesis rather than an
abrupt change from mineralized to unmineralized
Figure 5-1  The supraspinatus tendon–bone insertion in the rat (toluidine blue stain). Tendon attaches to bone across a
graded fibrocartilaginous transition.
Tendon
Insertion
site
Bone

51
Chapter 5:  Biology and Biomechanics of the Tendon–Bone Insertion
cell types and maturation processes are seen at both
anatomical locations.26
The developing growth plate contains cells that are
divided into distinct zones based on morphology and
mineral content.27 The zones are the reserve zone, the
proliferative zone, and the hypertrophic zone. The small
cells in the reserve zone proliferate and organize into
columns in the proliferative zone, and finally become
hypertrophic chondrocytes. It is at the leading edge of
the hypertrophic zone that calcium is deposited in the
zone of provisional calcification leading to mature
bone. Similar chondrocytes are seen at the developing
insertion site; these cells eventually mineralize to form
a tendon enthesis with a gradation in mineral.
The tendon–bone insertion site develops in the early
postnatal period.19-22 Mouse studies have elucidated the
time course of development of the tendon enthesis
using in situ hybridization.20,21 Collagens were localized
along the developing insertion. Collagen I was exam­
ined due to its prevalence in both tendon and bone.
Collagen II was examined due to its association with
chondrocytes, and hence the fibrocartilage region of the
insertion. Last, collagen X was examined due to its
reported expression by hypertrophic chondrocytes of
the growth plate. Rotator cuff tendon precursors were
juxtaposed to the humeral head at 15.5 days postcon­
ception. A transition zone between the developing
tendon and bone was not apparent until 7 days after
birth. Prior to 14 days postnatally, collagen I was con­
sistently expressed on the tendon side of the insertion
while collagen II was consistently expressed on the
humeral side, mirroring each other during these early
time periods. Collagen X was expressed at 14 days in
association with hypertrophic chondrocytes. As the
humeral head matured and mineralized, it began to
express collagen I, the main collagen type in mature
bone. A mature fibrocartilaginous insertion site was
formed by 21 days postnatally. This corresponded to the
completion of mineralization of the cartilaginous
humeral head. Similar results were seen at the develop­
ing Achilles tendon insertion of the rat21 and human.28
PTHrP has been localized to the insertion sites of
both tendons and ligaments.26 In the growth plate, it
is seen in proliferating chondrocytes, and is known to
be part of a negative feedback loop, preventing their
maturation into hypertrophic chondrocytes, which
then mineralize to form bone. Therefore, it plays an
important role not only in maintaining a population
of cells available for growth but also for preventing
inappropriate mineralization. It likely plays an analo­
gous role in maintaining a barrier between mineralized
and unmineralized tissues at the tendon to bone inser­
tion site. PTHrP works in concert with other factors.
IHH is secreted by proliferative chondrocytes before
they become hypertrophic chondrocytes. It binds to
the cell membrane receptor Patched 1 leading to the
model of the insertion was used to determine the role
of collagen orientation on stress and strain distribu­
tions.15 The idealized model showed that the micro­
structure serves to reduce stress concentrations and
material mass and to shield the insertion’s outward
splay from the highest stresses. Similarly, a model of the
medial collateral ligament femoral insertion demon­
strated that the direction of the principal tensile stresses
coincides with the direction of the collagen fibers.16 A
correlation was found between cell shape and mechani­
cal stresses, suggesting a direct relationship between
the stress field and cell activity. The effect of mineral
content on load transfer was modeled at the insertion
in a separate study.17 It was demonstrated that mineral
content and collagen fiber orientation combine to
give the tendon–bone transition a unique grading in
mechanical properties, supporting the idea for a func­
tionally graded, continuous tissue transition from
tendon to bone. An increase in mineral on collagen
fibers provided significant stiffening of the fibers. The
orientation of collagen fibers was the second major
determinant of tissue stiffness. The combination of
these two factors resulted in a variation of stiffness over
the length of the tendon–bone insertion similar to what
has been described from experimental testing.12,18 These
biomechanical models describe how tendon–bone
attachment is achieved in nature through a functionally
graded material composition.
DEVELOPMENT OF THE TENDON–BONE
INSERTION SITE
Studying the development of the tendon enthesis may
provide insights into the biological processes necessary
to form this complex anatomical site. These insights are
critical for the success of biological treatments or tissue
engineering solutions for tendon–bone repair. Animal
models indicate that the enthesis develops into a func­
tionally graded tissue postnatally,19-22 allowing a unique
opportunity to study and perturb its formation.
Role of Biological Factors
Several factors are important to the development of the
tendon–bone insertion site. Scleraxis is a transcription
factor expressed both in tenocyte progenitors and in
mature tendon cells and has been associated with
tendon and insertion site formation.23 SRY (sex deter­
mining region Y)-box 9 (SOX-9) is associated with chon­
drogenesis and has been localized in proliferating
chondrocytes.24 Parathyroid
hormone–related
protein
(PTHrP) and Indian hedgehog (IHH) drive chondrocyte
proliferation, but also form a negative feedback loop,
maintaining a population of proliferative cells at the
growth plate, available for further growth and differen­
tiation.25 These factors are seen in the growth plate of
developing bones and likely play a similar role at the
formation of the tendon enthesis, as many of the same

52
Section 1:  Basic Science
used as controls. Many effects were observed due to
paralysis in bony development, soft tissue contracture,
muscle mass, and muscle force generation. Dramatic
changes were seen at the developing tendon–bone
insertion site due to unloading.
The sensitivity of bone to its mechanical environ­
ment was highlighted in this series of experiments.
Removal of load significantly decreases amounts of
bone mineral in the humeral heads of the paralyzed
mice (Figure 5-2). These differences were not seen,
however, until after the 14-day time point; large differ­
ences were seen comparing saline to botulinum toxin
injected shoulders at 21, 28, and 56 days. In addition,
the botulinum toxin injected sides had larger numbers
of osteoclasts lining the bone compared to the controls.
Therefore, the decreased mineral content resulted not
only from decreased mineral deposition, but also due
to increased bone resorption.
Removal of load also impaired the formation of a
graded fibrocartilaginous insertion site. At 14 days, both
saline and botulinum toxin injected shoulders demon­
strated evidence of hypertrophic chondrocytes at the
insertion. However, by 21 days, the botulinum toxin
injected specimens showed a lack of fibrocartilage for­
mation, disorganized collagen fibers, and persistence of
hypertrophic chondrocytes. The saline side, on the other
hand, had a well-developed transition zone with no
evidence of hypertrophic chondrocytes. The transition
accumulation of a transmembrane protein smoothened,
which leads to the production of PTHrP. These factors
are all localized to the developing tendon insertion
site at 14 to 21 days, correlating to the onset of miner­
alization of the humeral head. These factors are also
highly mechanosensitive.29 Specifically, their expression
is decreased when load is removed.
SOX-9 and scleraxis are important factors in chon­
drogenesis and tenogenesis, respectively.30 SOX-9 is
localized to proliferating chondrocytes, but not in
hypertrophic chondrocytes, and may play some role in
insertion site development. Serving as a marker for
many connective tissues, including ligaments, tendons,
and joint capsules, scleraxis has been localized at the
developing insertion site and is required for teno­genesis.
Scleraxis likewise has been shown to be highly sensitive
to the mechanical environment, specifically, upregu­
lated in stem cells exposed to tensile forces.31
Role of Mechanical Factors
A series of experiments using mice were performed to
elucidate the effects of the mechanical environment on
the formation of the tendon–bone insertion site.19 Botu­
linum toxin A was injected to the shoulders of mice
within 24 hours of birth, simulating a neonatal brachial
plexus injury and removing muscle load from the
humeral head. Saline injections in the contralateral
shoulders and a separate group of healthy mice were
Figure 5-2  Bone volume and trabecular architecture were significantly altered in the Botox group compared to the Saline
and Normal groups at most timepoints during postnatal enthesis development (*significant difference Saline vs. Botox;
#significant difference Saline or Botox vs. Normal). (Adapted from Thomopoulos S, Kim HM, Rothermich SY, et al: Decreased
loading delays maturation of the tendon enthesis during postnatal development, J Orthop Res 25:1154–1463, 2007.)
Botox
Saline
0
0
0
0.2
0.4
0.6
0.8
0.05
0.10
0.15
0.20
0.25
0.30
TRABECULAR THICKNESS
Trabecular thickness (mm)
Trabecular spacing (mm)
TRABECULAR SPACING
0
10
20
30
40
50
60
Time (days)
0
10
20
30
40
50
60
Time (days)
0
10
20
30
40
50
60
Time (days)
*
*
*
*
*
*
*
*
*
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
Bone volume (mm3)
BONE VOLUME
Normal
Botox
Saline
Normal
Botox
Saline
Normal
Botox
Saline
1
2
3
4
5
6

53
Chapter 5:  Biology and Biomechanics of the Tendon–Bone Insertion
examined the strength of the rabbit patellar tendon–
tibia insertion site after the creation of a partial defect.35
Failure strength increased and approached normal over
a span of 12 weeks. However, the healing tissue in a
partial defect may have been stress shielded by the adja­
cent uninjured tissue. Another group found that the
strength of healing infraspinatus insertion site in goats
was only one third of the strength of the normal infra­
spinatus insertion at 12 weeks.36 This finding demon­
strates that although the histological appearance may
suggest the recreation of an insertion site, the strength
of the site may not be restored to normal. Another study
examined tendon healing in a bone tunnel.32 Strength
increased significantly between 2 and 12 weeks. After
this point, failure occurred somewhere other than the
insertion site. In a rat rotator cuff model, while the
structural properties reached two thirds of normal after
8 weeks of healing, the material properties remained an
order of magnitude weaker than control.5 The healing
tissue had a higher cross sectional area compared to
uninjured control, but was made up of poorly organized
type III collagen without a recreation of a fibrocartilagi­
nous transitional zone. Studies on canine flexor tendon
to distal phalanx repair indicated that insertion-site
healing demonstrates little improvement in repair-site
failure force from the time of suture through 42 days
following repair.37 A significant decrease was seen in
insertion site ultimate strength at 10 and 21 days
compared to time zero, indicating that a softening
zone remained disorganized and immature at 56 days
in the paralyzed specimens, whereas the saline group
had a mature, fibrocartilaginous, tendon enthesis. The
changes in fibrocartilage formation and bone formation
support the conclusion the mechanical environment
plays an important role in enthesis development.
TENDON–BONE REPAIR
Basic Science of Tendon–Bone Healing
Tendon–bone healing occurs through the generation of
a fibrovascular scar rather than regeneration of a graded
fibrocartilaginous transition (Figure 5-3).5,32-34 These
studies have focused on morphological and composi­
tional evaluations of the healing tissue. In an Achilles
tendon–bone healing study, surgical repair was impor­
tant to allow cells to reorganize the insertion site.33 At
8 weeks, the tendon was attempting to recreate the
insertion site. Low levels of type X collagen were found
in the fibrocartilage adjacent to bone, and this protein
was important in maintaining the interface between
calcified and noncalcified fibrocartilage. Similar histo­
logical findings were reported in a canine model of
tendon detachment.34 However, both models showed
that the structure and the composition of the healing
insertion did not approach normal by the longest time
points studied.
Investigators have also examined the mechanical
properties of the healing insertion site. One group
Figure 5-3  A graded fibrocartilaginous transition is not regenerated at the healing supraspinatus tendon–bone insertion in
rats (toluidine blue stain). Rather, the interface is filled with fibrovascular scar tissue, forming an abrupt transition between
tendon and bone (orange lines indicate the borders of the fibrocartilaginous transition on the left and the abrupt interface
between scar tissue and bone on the right).
Tendon
Bone
Bone
Interface
Interface
Tendon
NORMAL
HEALING

54
Section 1:  Basic Science
Increased force is beneficial to healing in a variety of
clinical settings. Early mobilization improves healing
and function after anterior cruciate ligament reconstruc­
tion in the knee.45 Early passive range of motion
decreases adhesions and improves strength after flexor
tendon repair.46 Controlled static stress is beneficial to
medial collateral ligament healing in the knee.47 On the
other hand, excessive force and motion can be cause
microdamage and/or gapping and thus be detrimental
to healing.5 Optimizing the mechanical environment in
the postoperative setting is therefore critical for improv­
ing outcomes.
In order to determine the effects of a variety of activ­
ity levels on tendon–bone healing, rotator cuff repairs
were performed in rats.5 Rat shoulders were then immo­
bilized, allowed cage activity, or exercised. Shoulders
that were immobilized demonstrated superior collagen
orientation and biomechanical properties compared to
those that were exercised. The exercised rats had a greater
quantity of tissue, but the tissue was lower in quality.
The composition of extracellular matrix generated at the
immobilized insertion better resembled a normal, unin­
jured insertion. The immobilized group was superior to
the cage activity group, and the cage activity group was
superior to the exercised group. A second study using
this animal model investigated the effect of short and
long durations of these activity levels on the healing
insertion site.48 The activity level had no effect on the
biomechanical properties of the insertion site at the
early (4 week) time point. However, decreased activity
(i.e., cast immobilization) had a positive effect on
phenomenon occurred in the early stages after repair.37
In all animal models, rather than regeneration of a func­
tionally graded transition between tendon and bone,
the interface was filled with fibrous scar tissue. These
findings demonstrate the considerable need for improve­
ment in the treatment of tendon–bone insertion site
injuries.
Bone Loss Following Tendon or
Ligament Injury
Bone loss following tendon or ligament injury has been
noted in a number of clinical studies and animal models.
In the rotator cuff, reduced bone mineral density was
seen in the humerus of patients 9 years after cuff rupture
and repair.38 These changes, however, were seen only in
patients who did not have full return of function, sug­
gesting that the bone loss was due in part to reduced
joint loading. In another clinical study of the Achilles
tendon, calcaneal bone loss was seen 4 to 12 months
after surgical repair.39 In a canine model of anterior
cruciate ligament transsection, a significant reduction in
bone mineral density was seen as early as 4 weeks
postinjury.40 Bone loss progressed through 10 weeks,
and then reached a plateau. Finally, a clinical study of
anterior cruciate ligament injury showed a 20% reduc­
tion in bone mineral density at 1-year follow-up for
patients who had surgical repair.41 Patients with less
severe injuries who were treated conservatively (i.e., no
surgery) showed only a 2% to 3% reduction in bone
mineral density. Significant bone loss was also demon­
strated in the canine flexor tendon model.42 Decreased
bone mineral density was seen at the distal phalanx
at 10, 21, and 42 days following injury and repair,
indicate that bone resorption may be a factor that con­
tributes to the low values of repair-site failure force
(Figure 5-4).42 Similar results were reported in the rat
rotator cuff model.43,44 Bone mineral density was signifi­
cantly decreased following tendon injury and repair.
A delay between injury and repair resulted in inferior
tendon–bone healing, in part due to decreased bone
quality.
Rehabilitation Based Treatments
As described earlier, tendon–bone healing is character­
ized by the formation of connective tissue with vastly
inferior biomechanical properties compared to normal,
uninjured tendon. A graded fibrocartilaginous transi­
tion between tendon and bone is not recreated. It is well
established that musculoskeletal tissues, including
tendon, bone, and cartilage, respond to their mechani­
cal environment. Similarly, as described earlier, the
development of a functional enthesis requires mechani­
cal loading. Therefore, significant efforts have been
made to enhance tendon–bone healing via rehabilita­
tion protocols (i.e., via control of the mechanical loads
across the healing insertion).
Figure 5-4  Dramatic bone loss was observed in the canine
distal phalanx after flexor tendon to bone repair. Decreases
in bone mineral density (BMD) were observed as early as 10
days after injury and repair. (Adapted from Silva MJ, Boyer
MI, Ditsios K, et al: The insertion site of the canine flexor
digitorum profundus tendon heals slowly following injury
and suture repair, J Orthop Res 20:447–453, 2002.)
0
10
20
30
40
Time (days)
–21%
–40%
–41%
*
*
*
Control
Injury and repair
0
100
200
300
400
500
600
700
BMD (mg/cm3)

55
Chapter 5:  Biology and Biomechanics of the Tendon–Bone Insertion
properties and greater range of motion. There was no
difference in bone density between loaded and unloaded
specimens. As in the rat model, cast immobilization
resulted in a better outcome than complete removal of
load. When considering these studies together, it is clear
that a balance between too much and too little load is
important to maximize tendon–bone healing. Closer
investigation suggests that high levels of motion are
detrimental but physiological muscle loading is benefi­
cial. The role of loading is consistent with the premise
that musculoskeletal tissues not only respond to, but are
dependent on, some load for development, healing,
and homeostasis.
Biologically Based Treatments
Biologically based approaches for enhancing tendon–
bone healing have focused on one or more aspects of
the tissue engineering paradigm—the use of signaling
biofactors (e.g., growth factors), responding cells (e.g.,
mesenchymal stem cells), and scaffold microenviron­
ments (e.g., collagen matrices).52 Numerous studies
have attempted to enhance tendon–bone healing by
delivering growth factors to the repair site. As significant
bone loss has been demonstrated after tendon–bone
injury and repair, one approach has been to target the
bony side of the insertion for treatment. Bone morpho­
genetic protein 2 (BMP-2) is well established as a potent
stimulator of bone formation. One experiment demon­
strated that the healing of tendon in a bone tunnel
occurred through bone ingrowth into tendon.53 The
authors demonstrated an improvement in structural
properties of the tendon healing in a bone tunnel after
application of exogenous BMP-2. Preventing bone
resorption at the healing insertion using bisphospho­
nate treatment has also been effective in improving
healing.54 Promotion of fibrocartilage at the healing
insertion has also been attempted, but with limited
success. One study demonstrated that transforming
growth factor beta is critical for tendon–bone healing,
but its modulation was ineffective in improving
healing.55 A separate study showed that application of
articular cartilage to the healing insertion increases
fibrocartilage regeneration.56 However, the partial regen­
eration of a fibrocartilaginous transition did not lead to
improvements in the biomechanical properties of the
repair.
Cell based therapies also hold great potential for
enhancing tendon–bone healing. Adult mesenchymal
stem cells (MSCs) show excellent regenerative capacity,
including the ability to proliferate rapidly in culture and
the capacity to differentiate into a wide range of cell
types. Prior reports support the use of stem cells to
enhance healing in multiple tissues. Recent work showed
that MSCs transfected with a tenogenic factor were effec­
tive in improving rotator cuff tendon–bone repair in a
rodent animal model.57 The addition of naïve MSCs to
biomechanical properties at the late (16 week) time
point. In these studies, decreasing the activity level by
immobilizing the shoulder improved tendon–bone
healing, as measured by collagen organization and bio­
mechanical properties. These results demonstrate that
increased activity can be detrimental to healing.
However, it was not determined whether this was a
result of a positive effect due to low levels of loading in
the immobilized group or a negative effect due to high
levels of motion in the exercised group.
The notion of negative effects due to motion at the
healing rotator cuff insertion site were substantiated in
a subsequent study that examined tendon biomechan­
ics and joint range of motion in rat shoulders after
rotator cuff repair.49 Continuous immobilization was
compared to two different passive range-of-motion pro­
tocols for 2 weeks followed by a 4-week remobilization
period. Both passive range-of-motion groups had less
joint range of motion compared to the continuous
immobilization group. All joints were stiffer compared
to preinjury levels. No differences were found in tendon
collagen organization or mechanical properties in the
three groups.
As both high levels of load and high levels of motion
were found to be detrimental to tendon–bone healing,
the effect of decreasing load below that seen in immo­
bilized repairs was examined in a recent study.50 After
surgical injury and repair, rat shoulders in two groups
were immobilized. One experimental group had botu­
linum toxin A injected into the supraspinatus muscle to
completely remove load from the healing insertion site.
A second group was immobilized and had saline injec­
tions into the muscle. A third group had botulinum
toxin A injections and rats were allowed cage activity
after repair. The saline/casted group had greater scar
volume and cross-sectional area of the repair tissue at
the insertion site and improved structural properties
compared to the botulinum toxin–paralyzed groups,
demonstrating that complete removal of load from the
healing insertion site was detrimental to healing.
Although reduced loading (e.g., through cast immobili­
zation) can be beneficial to healing (presumably by
eliminating excessive motion at the repair site), some
load applied to the site via normal muscle contraction
is necessary for effective healing.
The beneficial role of cast immobilization coupled
with physiological muscle loading was corroborated by
a study using a canine model of flexor tendon repair.51
This study measured biomechanical properties, bone
density (of the distal phalanx), and gap formation after
a flexor tendon–bone repair. In the experimental group,
the proximal tendon was transected to remove all load
from the repair site. All repairs were immobilized post­
operatively and subjected to a standard flexor tendon
passive range-of-motion protocol. Muscle loading across
the repair-site resulted in improved biomechanical

56
Section 1:  Basic Science
References
provide a solution to the clinical problem of tendon–
bone healing. A functionally graded material implanted
at the time of surgical repair may provide mechanical
stability and guide the repair process, leading to a suc­
cessful attachment of tendon to bone.
the healing rotator cuff insertion site did not improve
healing at the tendon attachment site. However, when
the MSCs were modified to overexpress the insertion-
site specific developmental gene membrane type 1
metalloprotease prior to delivery, rotator cuff healing
was significantly enhanced. Similar repair enhance­
ments were seen with the delivery of MSCs to a patellar
tendon defect and Achilles tendon models.58,59
A review of scaffolds currently used clinically for
rotator cuff tendon–bone repair revealed that further
work is necessary to optimize scaffold properties.60 Par­
ticularly lacking in the currently available scaffolds is an
appropriate recreation of the native tissue’s gradation in
properties. To address this lack of complexity, tissue
engineering work has focused on stratified and continu­
ously graded designs. In recent work, biphasic61 and
triphasic scaffolds62 were generated and seeded with
multiple cell types. These studies demonstrated the
importance of signaling between the various insertion-
site cell types for generation of a functional insertion.
Recent approaches have also attempted to create con­
tinuous gradients in properties to recreate the functional
grading that is seen at the natural tendon–bone inser­
tion. To this end, electrospun polymer nanofiber scaf­
folds were synthesized with gradations in mineral,
mimicking the mineral gradation seen at the native
insertion.63 The gradation in mineral content resulted in
a spatial variation in the stiffness of the scaffold (Figure
5-5). Similar results were reported using a cell-seeded
collagen scaffold with a gradient in retrovirus encoding
an osteogenic transcription factor.64 A tissue engineered
scaffold with a gradation in properties and seeded with
the appropriate cells and biofactors may ultimately
Figure 5-5  Strain maps of a nanofiber scaffold with a
gradation in mineral content under uniaxial tension.
Increasing strain is seen with increasing stress, with
localization at the compliant low mineral end.
σ = 0.29MPa
σ = 1.60MPa
σ = 2.41MPa
High mineral
Low mineral
ε =
0
0.02
0.04
1. Jones RM: Mechanics of Composite Materials, ed 2, Philadel­
phia, 1999, Taylor & Francis.
2. Williams ML: Stress singularities resulting from various
boundary conditions in angular corners of plates in exten­
sion, J Appl Mech 19:526–528, 1952.
3. Suresh S, Mortensen E: Fundamentals of Functionally Graded
Materials, London, 1998, IOM Communications.
4. Harryman DT 2nd, Mack LA, Wang KY, et al: Repairs of
the rotator cuff. Correlation of functional results with
integrity of the cuff, J Bone Joint Surg (Am) 73:982–989,
1991.
5. Thomopoulos S, Williams GR, Soslowsky LJ: Tendon to
bone healing: Differences in biomechanical, structural, and
compositional properties due to a range of activity levels,
J Biomech Eng 125:106–113, 2003.
6. Galatz LM, Ball CM, Teefey SA, et al: The outcome and repair
integrity of completely arthroscopically repaired large and
massive rotator cuff tears, J Bone Joint Surg (Am) 86:219–224,
2004.
7. Benjamin M, Kumai T, Milz S, et al: The skeletal attachment
of tendons–tendon “entheses”, Comp Biochem Physiol A Mol
Integr Physiol 133:931–945, 2002.
8. Waggett AD, Ralphs JR, Kwan AP, et al: Characterization of
collagens and proteoglycans at the insertion of the human
Achilles tendon, Matrix Biol 16:457–470, 1998.
9. Sagarriga Visconti C, Kavalkovich K, Wu J, et al: Biochemical
analysis of collagens at the ligament-bone interface reveals
presence of cartilage-specific collagens, Arch Biochem Biophys
328:135–142, 1996.
10. Kumagai J, Sarkar K, Uhthoff HK, et al: Immunohistochemi­
cal distribution of type I, II and III collagens in the rabbit
supraspinatus tendon insertion, J Anat 185:279–284, 1994.
11. Fukuta S, Oyama M, Kavalkovich K, et al: Identification of
types II, IX and X collagens at the insertion site of the bovine
achilles tendon, Matrix Biol 17:65–73, 1998.
12. Thomopoulos S, Williams GR, Gimbel JA, et al: Variation of
biomechanical, structural, and compositional properties
along the tendon to bone insertion site, J Orthop Res 21:413–
419, 2003.
13. Wopenka B, Kent A, Pasteris JD, et al: The tendon-to-bone
transition of the rotator cuff: A preliminary Raman spectro­
scopic study documenting the gradual mineralization across
the insertion in rat tissue samples, Appl Spectrosc 62:1285–
1294, 2008.

57
Chapter 5:  Biology and Biomechanics of the Tendon–Bone Insertion
14. Moffat KL, Sun WH, Pena PE, et al: Characterization of the
structure-function relationship at the ligament-to-bone inter­
face, Proc Natl Acad Sci U S A 105:7947–7952, 2008.
15. Thomopoulos S, Marquez JP, Weinberger B, et al: Collagen
fiber orientation at the tendon to bone insertion and its influ­
ence on stress concentrations, J Biomech 39:1842–1851, 2006.
16. Matyas JR, Anton MG, Shrive NG, et al: Stress governs tissue
phenotype at the femoral insertion of the rabbit MCL,
J Biomech 28:147–157, 1995.
17. Genin GM, Kent A, Birman V, et al: Functional grading of
mineral and collagen in the attachment of tendon to bone,
Biophys J 97:976–985, 2009.
18. Stouffer DC, Butler DL, Hosny D: The relationship between
crimp pattern and mechanical response of human patellar
tendon-bone units, J Biomech Eng 107:158–165, 1985.
19. Thomopoulos S, Kim HM, Rothermich SY, et al: Decreased
muscle loading delays maturation of the tendon enthesis
during postnatal development, J Orthop Res 25:1154–1163,
2007.
20. Galatz L, Rothermich S, VanderPloeg K, et al: Development
of the supraspinatus tendon-to-bone insertion: Localized
expression of extracellular matrix and growth factor genes,
J Orthop Res 25:1621–1628, 2007.
21. Fujioka H, Wang GJ, Mizuno K, et al: Changes in the expres­
sion of type-X collagen in the fibrocartilage of rat Achilles
tendon attachment during development, J Orthop Res 15:675–
681, 1997.
22. Bland YS, Ashhurst DE: Fetal and postnatal development of
the patella, patellar tendon and suprapatella in the rabbit;
changes in the distribution of the fibrillar collagens, J Anat
190:327–342, 1997.
23. Schweitzer R, Chyung JH, Murtaugh LC, et al: Analysis of the
tendon cell fate using Scleraxis, a specific marker for tendons
and ligaments, Development 128:3855–3866, 2001.
24. Akiyama H: Control of chondrogenesis by the transcription
factor Sox9, Mod Rheumatol 18:213–219, 2008.
25. Gilbert SF: Cell-cell communication in development. In
Gilbert SF, editor: Developmental Biology, ed 8, Boston, 2006,
Sinauer Associates.
26. Chen X, Macica CM, Dreyer BE, et al: Initial characterization
of PTH-related protein gene-driven lacZ expression in the
mouse, J Bone Miner Res 21:113–123, 2006.
27. Provot S, Schipani E: Molecular mechanisms of endochon­
dral bone development, Biochem Biophys Res Commun 328:
658–665, 2005.
28. Shaw HM, Vázquez OT, McGonagle D, et al: Development of
the human Achilles tendon enthesis organ, J Anat 213:718–
724, 2008.
29. Chen X, Macica C, Nasiri A, et al: Mechanical regulation of
PTHrP expression in entheses, Bone 41:752–759, 2007.
30. Asou Y, Nifuji A, Tsuji K, et al: Coordinated expression of
scleraxis and Sox9 genes during embryonic development of
tendons and cartilage, J Orthop Res 20:827–833, 2002.
31. Kuo CK, Tuan RS: Mechanoactive tenogenic differentiation of
human mesenchymal stem cells, Tiss Eng (Part A) 14:1615–
1627, 2008.
32. Rodeo SA, Arnoczky SP, Torzilli PA, et al: Tendon-healing in
a bone tunnel. A biomechanical and histological study in the
dog, J Bone Joint Surg (Am) 75:1795–1803, 1993.
33. Fujioka H, Thakur R, Wang GJ, et al: Comparison of surgically
attached and non-attached repair of the rat Achilles tendon-
bone interface. Cellular organization and type X collagen
expression, Connect Tissue Res 37:205–218, 1998.
34. Aoki M, Oguma H, Fukushima S, et al: Fibrous connection
to bone after immediate repair of the canine infraspinatus:
The most effective bony surface for tendon attachment,
J Shoulder Elbow Surg 10:123–128, 2001.
35. Waggy CA, Blaha JD, Lobosky DA, et al: Healing of tendon
to bone insertion site in rabbits: A model of the effect of
partial disruption, Trans Orthop Res Soc 40:39, 1994.
36. St Pierre P, Olson EJ, Elliott JJ, et al: Tendon-healing to corti­
cal bone compared with healing to a cancellous trough. A
biomechanical and histological evaluation in goats, J Bone
Joint Surg (Am) 77:1858–1866, 1995.
37. Silva MJ, Boyer MI, Ditsios K, et al: The insertion site of the
canine flexor digitorum profundus tendon heals slowly fol­
lowing injury and suture repair, J Orthop Res 20:447–453,
2002.
38. Kannus P, Leppälä J, Lehto M, et al: A rotator cuff rupture
produces permanent osteoporosis in the affected extremity,
but not in those with whom shoulder function has returned
to normal, J Bone Miner Res 10:1263–1271, 1995.
39. Alfredson H, Nordstrom P, Lorentzon R: Prolonged progres­
sive calcaneal bone loss despite early weightbearing rehabili­
tation in patients surgically treated for Achilles tendinosis,
Calcif Tissue Int 62:166–171, 1998.
40. Wohl GR, Shymkiw RC, Matyas JR, et al: Periarticular cancel­
lous bone changes following anterior cruciate ligament
injury, J Appl Physiol 91:336–342, 2001.
41. Leppälä J, Kannus P, Natri A, et al: Effect of anterior cruciate
ligament injury of the knee on bone mineral density of the
spine and affected lower extremity: A prospective one-year
follow-Up study, Calcif Tissue Int 64:357–363, 1999.
42. Ditsios K, Boyer MI, Kusano N, et al: Bone loss following
tendon laceration, repair and passive mobilization, J Orthop
Res 21:990–996, 2003.
43. Cadet ER, Vorys GC, Rahman R, et al: Improving bone
density at the rotator cuff footprint increases supraspinatus
tendon failure stress in a rat model, J Orthop Res 28:308–314,
2010.
44. Galatz LM, Rothermich SY, Zaegel M, et al: Delayed repair of
tendon to bone injuries leads to decreased biomechanical
properties and bone loss, J Orthop Res 23:1441–1447, 2005.
45. Beynnon BD, Johnson RJ, Fleming BC: The science of anterior
cruciate ligament rehabilitation, Clin Orthop Relat Res 402:9–
20, 2002.
46. Boyer MI, Goldfarb CA, Gelberman RH: Recent progress in
flexor tendon healing. The modulation of tendon healing
with rehabilitation variables, J Hand Ther 18:80–85, 2005.
47. Gomez MA, Woo SL, Amiel D, et al: The effects of increased
tension on healing medical collateral ligaments, Am J Sports
Med 19:347–354, 1991.
48. Gimbel JA, Van Kleunen JP, Williams GR, et al: Long dura­
tions of immobilization in the rat result in enhanced
mechanical properties of the healing supraspinatus tendon
insertion site, J Biomech Eng 129:400–404, 2007.
49. Peltz CD, Dourte LM, Kuntz AF, et al: The effect of postopera­
tive passive motion on rotator cuff healing in a rat model,
J Bone Joint Surg (Am) 91:2421–2429, 2009.
50. Galatz LM, Charlton N, Das R, et al: Complete removal of
load is detrimental to rotator cuff healing, J Shoulder Elbow
Surg 18:669–675, 2009.
51. Thomopoulos S, Zampiakis E, Das R, et al: The effect of
muscle loading on flexor tendon-to-bone healing in a canine
model, J Orthop Res 26:1611–1617, 2008.
52. Bell E: Tissue engineering in perspective. In Lanza RP, Langer
R, Vacanti J, editors: Principles of Tissue Engineering, San Diego,
2000, Academic Press, pp xxxv–xl.
53. Rodeo SA, Suzuki K, Deng XH, et al: Use of recombinant
human bone morphogenetic protein-2 to enhance tendon
healing in a bone tunnel, Am J Sports Med 27:476–488, 1999.
54. Thomopoulos S, Matsuzaki H, Zaegel M, et al: Alendronate
prevents bone loss and improves tendon-to-bone repair
strength in a canine model, J Orthop Res 25:473–479, 2007.

58
Section 1:  Basic Science
60. Derwin KA, Badylak SF, Steinmann SP, et al: Extracellular
matrix scaffold devices for rotator cuff repair, J Shoulder Elbow
Surg 19:467–476, 2010.
61. Wang IE, Shan J, Choi R, et al: Role of osteoblast-fibroblast
interactions in the formation of the ligament-to-bone inter­
face, J Orthop Res 25:1609–1620, 2007.
62. Spalazzi JP, Doty SB, Moffat KL, et al: Development of con­
trolled matrix heterogeneity on a triphasic scaffold for ortho­
pedic interface tissue engineering, Tissue Eng 12:3497–3508,
2006.
63. Li X, Xie J, Lipner J, et al: Nanofiber scaffolds with gradations
in mineral content for mimicking the tendon-to-bone inser­
tion site, Nano Lett 9:2763–2768, 2009.
64. Phillips JE, Burns KL, Le Doux JM, et al: Engineering graded
tissue interfaces, Proc Natl Acad Sci U S A 105:12170–12175,
2008.
55. Kim HM, Galatz LM, Das R, et al: The role of transforming
growth factor beta isoforms in tendon-to-bone healing,
Connect Tissue Res 52:87–98, 2011.
56. Wong MW, Qin L, Lee KM, et al: Articular cartilage increases
transition zone regeneration in bone-tendon junction
healing, Clin Orthop Relat Res 467:1092–1100, 2009.
57. Gulotta LV, Kovacevic D, Montgomery S, et al: Stem cells
genetically modified with the developmental gene MT1-MMP
improve regeneration of the supraspinatus tendon-to-bone
insertion site, Am J Sports Med 38:1429–1437, 2010.
58. Young RG, Butler DL, Weber W, et al: Use of mesenchymal
stem cells in a collagen matrix for Achilles tendon repair,
J Orthop Res 16:406–413, 1998.
59. Awad HA, Boivin GP, Dressler MR, et al: Repair of patellar
tendon injuries using a cell-collagen composite, J Orthop Res
21:420–431, 2003.

59
CHAPTER
6 
GENE THERAPY FOR TENDON
HEALING
Jin Bo Tang, MD, Ya Fang Wu, MD, Yi Cao, MD,
Chuan Hao Chen, PhD, Xiao Tian Wang, MD, and
Paul Y. Liu, MD
a recent exponential expansion in available genetic
information. We have noted that biological investiga­
tions are catching up with mechanical studies, making
an increasingly greater contribution to the efforts to
improve tendon repairs.9-29 From these biological inves­
tigations, we are rapidly accruing information regarding
the molecular mechanisms and efficacy of manipula­
tions of tendon healing, which may bear fruit in the
future.
Weak healing potential is a common problem in
tissues without a rich blood supply. At present, there is
a lack of effective treatment to ensure a prompt and
strong healing of these tissues. Finding treatments to
accelerate healing such tissues is of paramount signifi­
cance in regenerative medicine, sports medicine, reha­
bilitation, and in surgery. The treatments not only
benefit tendon repair in the hand but also benefit repair
and regeneration of tendons, ligaments, and tendon–
bone junctions throughout the body. Efforts to acceler­
ate healing of those tissues are expected to bring about
significant impacts in a number of medical specialties.
Laceration of digital flexor tendon is representative of a
wide spectrum of injuries that fail to heal reliably and
is an ideal setting for testing novel molecular therapies.
Work in this area holds promise for advancing the treat­
ment of tissue injuries due to trauma or sports or in
association with degeneration.
Of gene therapy applications in tissue healing, the
unique advantage is that this strategy necessitates
production of the encoded factors for a fairly short
period.30-32 Gene therapies traditionally aimed at curing
tumors or congenital diseases, in contrast, necessitate
long-term transgene expression, usually over the entire
lifetime. Such prolonged transgene expression is still a
difficult goal to reach reliably for chronic conditions,
which constitutes an obstacle to ultimate clinical success
in curing these diseases. However, for tissue repair, gene
therapy can establish a local, endogenous synthesis of
the authentically processed materials critical to the
healing. Long-term gene expression is neither necessary
OUTLINE
Weak intrinsic healing potential is the central issue
underlying difficulties of tendon healing and major
clinical problems associated with tendon surgery. Bio­
logical augmentation of the healing process has long
been sought. In the past decade, growth factor gene
therapy emerged as an attractive treatment modality.
Our work over the past decade has not only identified
appropriate genes as the therapeutic target and vectors
as gene-carrying vehicles but also assessed the effi­
ciency of transgene expression and sustained supply
of the growth factors in the healing tendon. In testing
the efficacy of growth factor gene therapy in vivo, we
found that injection of adeno-associated viral vectors
harboring basic fibroblast factor or vascular endothe­
lium growth factor cDNAs to the tendon during
surgery substantially increases tendon healing strength
in a chicken tendon cut and repair model. Transgenes
were expressed from postsurgical weeks 2 to 6, cover­
ing the critical healing period. In the healing tendons,
gene therapies increased the levels of growth factors,
promoted synthesis of extracellular matrix compo­
nents, decreased tenocyte apoptosis, and enhanced
cellular proliferation. These evidences show the
promise of gene therapies in the arena of tendon
repair. Our studies also indicate the possibility of
delivery of synthetic micro-RNA in limiting adhesions.
These findings may lay the groundwork for establish­
ment of methods to correct the insufficiency in intrin­
sic healing capacity of the tendon.
Improvement in tendon healing through biological
approaches, molecular manipulations in particular,1-8 is
a relatively new area of investigation related to flexor
tendon surgery. These biological manipulations are
intended to overcome a major obstacle to success in
tendon repair—low intrinsic healing potential. The
development of molecular treatment of tendon healing
has benefited from advances in biotechnology and

60
Section 1:  Basic Science
OVERALL CONCEPTS AND METHODS
OF GENE THERAPY
The central idea of gene therapy is the incorporation of
genes of interest into recipient cells to modify levels of
expression of the gene, for the purpose of treating
disease. The disorders requiring gene therapy are usually
refractory to conventional therapeutic approaches.
Although the clinical utility of these technologies
remains in its infancy, gene therapy has been used with
preliminary clinical success in some diseases. More trials
are currently under way and may generate evidence for
the utility of such techniques over a broader range of
medical problems. Gene therapy holds great promise
for incorporation into mainstream medicine.
The ideas behind gene therapy date back to the early
1970s. In 1972, Friedmann and Roblin advocated intro­
ducing exogenous genes to cure human diseases in an
article entitled, “Gene Therapy for Human Genetic
Disease?”36 Major advances were made in the technical
aspects of gene therapy in the 1990s. Over the past two
decades, investigators have taken logical steps of using
gene therapy to pursue curing diseases caused by single-
gene defects (e.g., cystic fibrosis, hemophilia, muscular
dystrophy, and sickle cell anemia) or in the treatment
of certain cancers. More recently, clinical work has
included a trial for treating Leber congenital amauro­
sis,37 a disease caused by mutations in the RPE65 gene.
The investigators proved the safety of subretinal delivery
of recombinant adeno-associated virus (AAV) carrying
the RPE65 gene and found a modest improvement in
vision of the patients.37 In another study, investigators
restored trichromatic vision to squirrel monkeys using
gene therapy to cure red-green color blindness.38 Other
investigators succeeded in treating adrenoleukodystro­
phy (ADL), a fetal brain disease, by delivering the gene
for the missing enzyme to patients via a lentiviral vector.39
Various techniques have been developed to achieve
successful gene transfer,32 including both nonviral
methods (e.g., liposomal, electroporation, sonopora­
tion, shots from a “gene gun,” etc.) and viral vectors
(e.g., adenoviruses, AAVs, herpes simplex viruses, retro­
viruses, lentiviruses, etc.).
Gene therapy is commonly considered a possible
treatment modality for patients with congenital defects
of key metabolic functions or patients with malignan­
cies. Clinical success of gene therapy for congenital
diseases and cancers is hindered by problems such as
the short-lived function of the transgenes, immune
responses, side effects with vectors, the inability to treat
multigene disorders, and the risk of insertional muta­
genesis. However, these problems rarely constitute a
barrier to the use of gene therapy strategies for tissue
healing if gene delivery methods are appropriate. The
use of gene therapy in tissue repair or regeneration may
become “dark horses” in the field of gene therapy.
nor desired.30 The required period of transgene expres­
sion for tissue healing exactly fits current possibilities
offered by gene therapy.
INNATE WEAKNESS IN INTRINSIC
TENDON HEALING AND “NO-GAIN”
PERIOD IN STRENGTH
Tissues nourished with rich blood supplies usually
acquire mechanical strength rapidly in the first and
second postsurgical weeks. However, the tendon healing
process is characterized by a “no-gain” period in strength
over the first 3 to 4 weeks (Figure 6-1)5,33-35 due to the
weak healing capacity, yielding slow healing responses
after injury. The innate weakness in tendon healing
relates to the following structural facts: (1) they lack
sufficient blood supply, (2) have sparse cellularity, and
(3) are mainly composed of tightly packed collagen
bundles. The characteristic no-gain period in strength is
the central issue underlying the difficulties and clinical
problems associated with tendon repairs. The goal of
present efforts through biological manipulations is to
elicit an earlier increase in strength, helping the tendon
“leap over” this lag period. Theoretically, this goal is
achievable through targeting major causes of weak
tendon healing, either by supplementation of cells
involved in healing or by increasing the activity levels
of growth factors. Because healing strength is structur­
ally based on collagen connections across the tendon
repair site, we expect that addition of specific cells or
growth factors will ensure speedy deposition and matu­
ration of collagens at the healing site, resulting in an
earlier gain in mechanical strength.
Figure 6-1  The tendon healing process is characterized by
a “no-gain” period in strength over the first 3 to 4 weeks.
The graph shows our data for the tendon strength obtained
from the complete cut and repair model of the FDP tendon
in chicken long toes. The tendon was repaired with a
modified Kessler method using 5-0 nylon suture. The
strength of the healing tendon can be divided into three
parts: no gain, swift gain, and further steady gain periods.
0
1
2
3
4
5
6
7
8
0
5
10
15
20
25
30
35
40
45
50
Tendon strength (N)
No gain period
in the strength
Swift gain in
the strength
Further
steady gain in
the strength
3
2
1
Weeks

61
Chapter 6:  Gene Therapy for Tendon Healing
Controlled-release drug delivery systems is another
way to provide the tendon with a prolonged supply of
growth factors.7,21 Growth factors can be incorporated
into polymers in vitro, and the impregnated polymers
are then implanted into the tissues, which may deliver
the growth factors continuously for a prolonged period
in a controlled fashion. This system holds promise for
use in tendon repairs, but research in this area is just
beginning. Growth factors (bFGF and PDGF-BB) have
been incorporated into fibrin matrices and the release
kinetics has been tested over a 10-day-period in vitro21;
bFGF and PDGF-BB were found to release gradually
from the fibrin matrices and to modulate cell prolifera­
tion and extracellular matrix synthesis in vitro. However,
this system loaded with PDGF-BB was not found to
improve the strength of the healing canine flexor
tendons at days 7, 14, and 42 after surgery.7 In a canine
model, bFGF delivered through fibrin heparin–based
controlled-release system stimulated cellular prolifera­
tion, promoted neovascularization and inflammation of
the flexor digitorum profundus (FDP) tendon in the
earliest stages after surgical repair, but it failed to improve
mechanical strength of the repair.43 This system remains
to be fully tested with respect to its effectiveness to
increase tendon strength in vivo.
GENE THERAPY APPROACHES—THE
FIRST STEP: CHOOSING GENES AND
DELIVERY METHODS
Over the past 10 years, we have directed our efforts toward
improving tendon healing through augmentation of
growth factor activity (Figure 6-2). To test the efficacy
of this approach, we faced two initial questions: (1)
what genes should be delivered, and (2) which are safe
and efficient gene-carrying vehicles for our purposes?
BIOLOGICAL METHODS TO ENHANCE
TENDON HEALING
Theoretically, tendon healing may be enhanced through
biological means other than gene therapy. Nevertheless,
none of those methods can be claimed to be of practical
value until fully tested. The initial step in the test process
involves in vitro studies. The next step is in vivo experi­
mentation, which will generate preclinical data to
support the clinical trials, the last step in the test process.
One possible way to augment tendon healing is to
provide the tendons with exogenous cells, to ameliorate
sparse cellularity. A few such attempts have been made.
Direct delivery of bone marrow stem cells to healing
rabbit Achilles tendons produced some evidence of
increases in tendon strength.40 Investigators coated sur­
gical sutures with bone marrow stem cells and delivered
those cells to the tendon in vitro,6 but the in vivo effi­
cacy of this method has not been reported.
Growth factors, such as insulin-like growth factor-1
(IGF-1),9,10 platelet-derived growth factor-B (PDGF-B),20
and basic fibroblast factor (bFGF),20,23,25 have been
reported to promote tenocyte proliferation or collagen
synthesis. Direct supplementation of growth factors to
the healing tendons has failed to reliably improve
healing strength. Chan and colleagues41 injected bFGF
to rat cut patellar tendon and detected no effects on
ultimate strength of the tendon 7 and 14 days after
surgery despite increases in cell proliferation and type
III collagen production. Surgical sutures can be coated
with growth factors such as bFGF or growth differentia­
tion factor-5 (GDF-5).4,42 Surgical repairs using such
coated sutures improved the strength of the rabbit flexor
tendon only at a single time point (postoperative week
3) but failed to improve strength at other time points.4,42
Figure 6-2  Flow chart of our investigations of growth factor gene therapy to enhance tendon healing and micro-RNA gene
therapy to inhibit tendon adhesions over the past 10 years.
FLOW CHART OF OUR INVESTIGATIONS OF TENDON GENE THERAPY
Characterization
of effects of
growth factors on
tenocytes in vitro
Tests of effects of
plasmid vectors
harboring growth
factors in vitro
Tissue reactions
caused by different
vectors in tendons
Efficiency of gene
delivery by different
vectors in tendons
Efficiency of gene
delivery by different
serotypes of AAVs
Construction of
AAV2-bFGF,
AAV2-VEGF,
and
DS-AAV2-VEGF
vectors
In vivo study
of efficacy of
AAV2-bFGF
gene therapy
In vivo study
of efficacy of
AAV2-VEGF
gene therapy
Expression of
molecules
involved in the
healing tendon
and apoptosis
Test of in vitro
engineered
miRNA gene
therapy in vitro
and in vivo
(2002–2003)
(2003–2004)
(2004–2006, 2008)
(2005–2010)
(2007–2011)
Tang, Xu
Tang, Wang
Zhu, Cao, Chen, Tang
Xin, Wang
Cao, Wu, Tang
Chen, Wu, Tang
Understand molecular biology of tendon healing
Increase the healing strength
Limit adhesions

62
Section 1:  Basic Science
than liposome-mediated gene transfer. Because more
than 10 serotypes of AAV are currently available, all with
different cellular or tissue tropism, we compared trans­
duction efficiency of AAV1, AAV2, AAV3, AAV4, AAV5,
AAV7, and AAV8 vectors containing the lacZ gene. We
found that AAV2, the best characterized and most
popular AAV vector, has the highest efficacy of gene
transfer in tenocytes.27
We then examined the duration of expression of
transgene introduced by AAV2 vectors to the healing
flexor tendon. Using an in vivo chicken model, we
injected the vectors at two sites in each stump of cut
FDP tendon through lacerated tendon surfaces. We
found that expression of transgenes lasts beyond 4
weeks and peaks during the second week. Through the
presence of enhanced green fluorescent protein (EGFP)-
positive cells, we confirmed that injected vectors were
distributed to all layers of the tendon (see Figure 6-3).
Our above examinations, coupled with desirable fea­
tures of AAV vectors (nonpathogenicity, low immunity
in human beings, etc.),44-46 and evidence of their utility
and safety from reported Phase I clinical trials37,47-49 con­
vinced us that the AAV2 vector is a proper gene transfer
vector. Although AAV vectors, which have small packag­
ing capacity, accommodate up to 4 to 5 kb of DNA
insert,46 they are sufficient to package therapeutic cDNAs
encoding growth factors (usually less than 1 kb each)
to meet our needs.
To explore how to deliver vectors into the lacerated
tendon, we decided to directly inject the vector to
tendons and tried injections at either one or two sites
on the tendon stump. We found that injection at two
sites ensures better vector distribution and greater vector
delivery. The volume of the injections without spillage
of fluid out of the tendon was also tested. Injection of
5 µl liquid to each site on the stump was found appro­
priate in rabbit or chicken digital flexor tendons. In
addition, we decided to inject the tendon through the
lacerated tendon surface, avoiding injuries to the tendon
gliding surface.
Information gathered in the above trials established
our method of vector delivery to tendons (Figure 6-4).
Typically, we inject a total of 20 µl of vectors to four sites
(5 µl/site) into two stumps. The needle is inserted from
the cut surface of the tendon at the depth of about
5 mm.
THE SECOND STEP: TESTING THE EFFECTS OF
GROWTH FACTOR GENE THERAPY IN VITRO
We constructed AAV2 vectors harboring the bFGF or
VEGF cDNAs, by inserting the therapeutic construct
between the 5′ and 3′ inverted terminal repeat (ITR)
sequences under the cytomegalovirus (CMV) immediate-
early promoter (Figure 6-5), and examined the in vitro
efficacy of therapies using these vectors. We treated
the tenocytes with AAV2-bFGF or sham AAV2 vector
To answer the first question, we examined levels of
growth factors in the healing digital flexor tendons.
Growth factor genes with low levels of expression in the
healing process are considered targets of gene therapy.
In a chicken tendon laceration model, we examined the
levels of growth factors from day 3 to week 12 after
surgery. bFGF and PDGF-BB were found to have the
lowest expression in the early period28; VEGF and IGF-1
were expressed in modest amounts.28 Consequently, we
decided to target the bFGF gene initially, and the VEGF
gene secondarily.
Our answer to the second question—how best to
deliver the genes—was based on an essential consider­
ation: the intended therapy is aimed at the healing of
normal tissues, requiring only short transgene expres­
sion, rather than treating cancers or curing inherited
diseases. In this regard, safety issues of the vectors
should be the primary concern. With these consider­
ations in mind, we first eliminated retroviral vectors.
Lentiviral vectors and herpes simplex viral vectors were
next excluded, because of concerns of biosafety of len­
tiviral vectors (e.g., potentials of generation of
replication-competent lentivirus and oncogenesis) and
the narrow applications of herpes simplex viral vectors
(i.e., in the nervous system).44 Adenoviral, AAVs, and
liposome-mediated gene transfer thus became viable
candidates.
To move our decision process along, we implemented
three sets of studies to examine (1) tissue reactions in
tendons caused by the candidate gene transfer methods;
(2) efficiency of gene delivery to tenocytes using these
methods; and (3) duration and course of transgene
expression in healing tendons by these methods.
We used both rabbits and chickens to examine reac­
tions in the tendons elicited by three candidate methods.
Histologically, we observed dramatic tissue reactions in
both epitenon and endotenon areas after liposome-
mediated gene transfer. The reactions elicited by adeno­
viral vectors were also pronounced but less severe.
Tendons treated by AAV serotype 2 (AAV2) vectors dis­
played obvious cellular proliferation and thickening of
epitenon, but the reactions were almost completely con­
fined to the epitenon (Figure 6-3). We noted striking
similarity between cellular proliferation in the healing
process of tendons and that caused by the AAV2 vectors
(see Figure 6-3).26
Efficacy of liposome-mediated transfection and trans­
duction of adenoviral and AAV2 vectors was next tested.27
We found that tenocytes are not highly permissive in
cell culture settings; all these vectors were associated
with low efficiency of gene transfer in tenocytes, even
when test at high titers of vectors. From the later in vivo
studies, we understand that transduction rates of AAV
vectors in the cells of tissues are much higher than in
cell culture conditions. AAV2 and adenoviral vectors
have similar transduction efficiency, both more efficient

63
Chapter 6:  Gene Therapy for Tendon Healing
cells treated with sham vectors or in nontreated con­
trols. We concluded that delivery of exogenous bFGF
gene to tenocytes significantly increases the levels of
expression of bFGF and type I and III collagen genes.
Similarly, we tested the effectiveness of AAV2-VEGF and
found enhancement of collagen I production and pro­
liferation of tenocytes in vitro.
(harboring the lacZ gene), with nontreated tenocytes as
the control.50 We assessed transgene expression, expres­
sion levels of the bFGF gene, and type I and III collagen
genes. Expression of the bFGF gene increased signifi­
cantly after AAV2-bFGF treatment. Expression of type I
and III collagen genes increased significantly after trans­
fer of the exogenous bFGF gene compared with that in
Figure 6-3  Tissue reactions and transduction efficiency of the AAV2 vectors. The tissue reactions during the tendon healing
(A, B, D, E) and those caused by AAV2 (C, F) are strikingly similar. The tendons treated with AAV2 vectors exhibit obvious
cellular proliferation and thickening of epitenon, but there are almost no tissue reactions in the endotenon area. Enhanced
green fluorescent protein (EGFP) expression was obvious in endotenon and epitenon areas at days 7 and 14 (G, H) after
injection of AAV2-EGFP into the tendon. *Epitenon area. (A and D, magnification ×100; B and E, ×200; all others, ×400).
Day 7
A
B
C
D
E
F
G
H
Day 7
Cut and repaired tendons
Cut and repaired tendons
AAV2 treated
AAV2 treated
AAV2 treated
AAV2 treated
Day 14
Day 14

64
Section 1:  Basic Science
In the first set of experiments, a total of 104 toes from
52 chickens underwent surgery and were randomized
into three groups, with 38 toes in each of two groups
and 28 toes in one group.5 The FDP tendon was tran­
sected completely with a sharp surgical blade and was
subjected to one of three treatments: (1) AAV2-bFGF
injection (38 toes) in which a total of 2 × l09 particles
of AAV2 containing the bFGF cDNA were injected into
both stumps of the cut tendon ends; (2) AAV2-luciferase
injection (28 toes) in which AAV2-luciferase in the same
number of viral particles was injected into the tendon
stumps; and (3) noninjection (38 toes) in which the
tendons did not receive injection. Immediately after the
above treatments, the tendon was repaired surgically
with a modified Kessler method using 4-0 suture. Post­
surgery, the toes were immobilized in semiflexion posi­
tion by adhesive tapes for the first 3 weeks and were
released to allow free motion thereafter.
The FDP tendons from 20 toes of 10 chickens were
cut and repair, for test of strength of repaired tendon
immediately after surgery, and the 10 toes from 5
chickens were not injured and were tested for work of
toe flexion, to obtain time-zero strength and work of
flexion data.
THE THIRD STEP: ASSESSMENT OF EFFECTS OF
GROWTH FACTOR GENE THERAPY IN VIVO
Over the past 5 years, we conducted four sets of in vivo
experiments to evaluate the effectiveness of growth
factor gene therapy using chickens as a model.
Figure 6-5  AAV genomes and vector construction. A
therapeutic construct was inserted into downstream of a
promoter, which replaces the Rep and Cap components of
the wild-type AAV.
AAV GENOMES AND VECTOR CONSTRUCTION
p19
p40
P5
5′
3′
Rep
Cap
ITR
ITR
Promoter
Therapeutic construct
Growth factor cDNA
Figure 6-4  Method of delivery of AAV2-bFGF to the tendon. A, How vectors were injected AAV2 to both stumps of the
tendon cut ends at the depth of 0.5 cm. The vectors were injected to two sites in each stump, at the center of the left and
right tendon bundle of the FDP tendon. B, Structures of flexor tendons and the major pulley in the area between the proximal
and distal interphalangeal joints in chicken toes, which are similar to those of the human tendons around the A2 pulley.
C shows an example of injection of vectors through the lacerated tendon cross-sectional surface into the tendon. D shows
a microinjection needle used for vector injection.
C
Micro-injection
syringe
0.5 cm
Lacerated tendons
A
D
B
FDS
FDP
Pulley

65
Chapter 6:  Gene Therapy for Tendon Healing
gene therapy, tenocyte proliferation is increased, teno­
cyte apoptosis decreased, production of extracellular
matrix components accelerated, and the rate of degrada­
tion of extracellular matrix in the tendon decreased.
These molecular changes offer a mechanistic explana­
tion for the observed effects of AAV2-bFGF gene therapy.
In the fourth set of experiments, we tested the effective­
ness of AAV2 vector harboring VEGF cDNA on improve­
ments in healing strength. Thirty-six chickens (72 toes)
were used; transversely cut FDP tendons were injected
We evaluated the outcomes at four time points from
2 to 12 weeks, with emphasis on variables reflecting
healing status and adhesions differently at early and late
healing stages. Postoperative 2 and 4 weeks correspond
to the early healing period, and 8 and 12 weeks corre­
spond to the late healing period. At 2 weeks, tendons
were tested for strength and evaluated for healing status
and adhesions. At 4 weeks, the tendons were harvested
for strength tests and to score the severity of adhesions.
At 8 and 12 weeks, when the adhesions and healing had
matured, we measured the tendon strength, work of toe
flexion, and scored adhesions.
We found that the ultimate strength of repaired
tendons treated with AAV2-bFGF was significantly
greater than that of tendons treated with the sham-
vector or noninjection at 2, 4, and 8 weeks (Figure 6-6).
The AAV2-bFGF treatment at 4 weeks increased the
strength to about 140% that of tendons treated with
sham-vector or noninjection. Statistically, the grading of
adhesions was the same among all three groups at 8
weeks and the treated tendons had less severe adhesions
at 12 weeks (see Figure 6-6). Work of toe flexion after
AAV2-bFGF treatment was not increased compared with
that of noninjection controls at 8 and 12 weeks.
In the second set of experiments, we examined the effects
of delivery of different amounts of AAV2-bFGF on
tendon healing. We examined the strength 4 weeks after
injection of 2 × l07, 1 × l08, 2 × l08, 4 × l08, 1 × l09, and
2 × l09 viral particles to each tendon and obtained a
range of vector amount (1 × l09 to 2 × l09 of viral
particles/tendon) that gives the best outcomes in terms
of healing strength.
In the third set of experiments, we expanded AAV2-
bFGF gene therapy experiments to include comprehen­
sive biomechanical evaluation of the healing strength
and molecular biological examinations at 10 postsurgi­
cal time points (day 0, weeks 1, 2, 3, 4, 5, 6, 8, 12, and
16). The purpose of this set of experiments was to obtain
a broader picture of the efficacy and mechanism of
AAV2-bFGF treatment. Three groups (AAV2-bFGF treat­
ment, sham vector control, and noninjection control)
were included, with a total of 134 chickens (268 toes)
used. The sample size ranged from 6 to 14 tendons at
each time point for each group. The results of this set
of studies are detailed in Figure 6-7. This set of experi­
ments offers further sound evidences of the effectiveness
of AAV2-bFGF: the tendons treated with AAV2-bFGF
showed an early increase in strength from postsurgical
week 1 through week 4, with percent strength gain of
40% to 70% over the non-injection controls (see Figure
6-7). The no-gain period in healing strength was effec­
tively corrected by AAV2-bFGF gene therapy. Transgene
expression was detectable from week 2 to 6, but the
expression ceased at weeks 8 or 12. The amount of bFGF
in the treated tendon returned to levels similar to those
in the noninjection control at 8 weeks. With AAV2-bFGF
Figure 6-6  A, Ultimate strength of the repaired FDP
tendon at 2, 4, and 8 weeks postsurgery. The sample
sizes were 10 to 12 tendons at each time point for each
treatment. At all the three time points, the strength of the
tendon treated with AAV2-bFGF was significantly greater
than that of the no-injection controls. B, Adhesion scores at
four time points. Sample sizes were 10 tendons at each time
point for each treatment. Adhesions were similar between
the treated tendons and no-injection controls at week 8.
The treated tendons had less severe adhesions than the
no-injection controls at week 12. Adhesions were scored
by criteria in Tang JB, Ishii S, Usui M, et al: Dorsal and
circumferential sheath reconstructions for flexor sheath
defect with concomitant bony injury, J Hand Surg (Am)
19:61–69, 1994.
0
2
4
6
8
10
12
14
Ultimate strength (N)
2
3
4
5
Scores of adhesions (points)
A
2 weeks
4 weeks
B
p <0.001
p <0.001
NS
NS
0
20
40
60
80
100
120
8 weeks
p <0.05
NS
AAV2-bFGF treated
AAV2-lucifereas
No injection
AAV2-bFGF treated
AAV2-lucifereas
No injection
p <0.05
p <0.05
p <0.05
NS
2 weeks
4 weeks
8 weeks
12 weeks

66
Section 1:  Basic Science
VEGF gene therapy is effective in correcting the insuffi­
ciency of intrinsic healing capacity and offers a thera­
peutic possibility for enhancing tendon healing.
GENE THERAPY ASSOCIATED
WITH MICRO-RNA REGULATION
IN TENDON HEALING
Adhesion formation is another major problem in
tendon healing. With the AAV2-bFGF or AAV2-VEGF
therapies described above, no increase in tendon adhe­
sions was recorded compared with sham vector-treated
or noninjection controls. However, morphologically
most treated and healed tendons still showed the forma­
tion of adhesions; the work of digital flexion did not
return to normal at postoperative 8 or 12 weeks. As an
integral part of investigatory efforts, over the past 3
years, we explored the possibility of RNA interference
(RNAi) in reducing adhesion formations of tendon.8
Synthetic RNAi is an emerging and rapidly developing
field. Gene therapy incorporating siRNA or micro-RNA
(miRNA) is a new dimension to the field of gene therapy.
Delivery of siRNA or miRNA to the healing tissues pres­
ents as an efficient tool to downregulate expression of
genes critical to tissue repair.
Because transforming growth factor (TGF)-β1 is a
master mediator and a major player in tissue fibrosis
and scar formation, we aimed at silencing the TGF-β1
gene and mitigating the effects of TGF-β1 during tendon
healing. We synthesized four pre-miRNAs according to
sequences of the chicken TGF-β1 gene and constructed
four plasmids containing these pre-miRNA sequences
and one control containing mock miRNA sequences
(Figure 6-8). Using cultured tenocytes of chicken digital
flexor tendons, we then screened the sequences that
effectively downregulate expression of the TGF-β1 gene.
We transfected cultured tenocytes with these plasmid
vectors and quantified expression of TGF-β1, type I and
III collagen, and connective tissue growth factor (CTGF)
genes using real-time polymerase chain reactions
(Figure 6-9). Subsequently, the plasmid miRNA that
most effectively silenced the TGF-β1 gene in vitro was
used to test its effectiveness on reducing adhesions in
vivo. The plasmid miRNA and the control were injected
into the injured tendons of 25 chickens. We found that
in both in vitro and in vivo settings, delivery of engi­
neered miRNA to the tendon downregulated expression
of the TGF-β1 gene but did not affect expression of the
collagen I gene.
In the healing tendon of chickens, we found that
TGF-β1 gene expression was downregulated 50% to
60% at weeks 1 and 6. At 6 weeks, collagen III gene
expression was reduced by 55%, and the CTGF gene was
downregulated by 25%. A smoother gliding surface of
the tendon with fewer adhesions was detected 6 weeks
after plasmid miRNA treatment compared with plasmid
mock miRNA treatment.
with AAV2-VEGF (2 × l09 viral particles to each tendon).
The results were evaluated at postsurgical day 0, and in
weeks 1, 2, 3, 4, 5, 6, and 8. We found that the AAV2-
VEGF treatment is basically as effective as AAV2-VEGF
treatment (see Figure 6-7).
To sum up our in vivo experiments, we found that
supplementing lacerated tendons with the bFGF or
VEGF cDNAs to produce supernormal amounts of bFGF
or VEGF (1) substantially increased growth factor activ­
ity within healing tendons, (2) promoted proliferation
(but inhibited apoptosis) of the tenocytes and produc­
tion of extracellular matrix critical to gain in strength,
and (3) significantly enhanced the strength of the healed
tendon. Moreover, production of supernormal amounts
of growth factors had ceased after healing was achieved.
Taken together, our evidence indicates that bFGF or
Figure 6-7  Effects of AAV2-bFGF and AAV2-VEGF gene
therapy on the tendon healing strength in a chicken tendon
cut and repair model. AAV2-bFGF and AAV2-VEGF gene
therapy substantially augmented tendon healing, helping the
tendon “leap over” the lag period in the early stage of the
tendon healing.
0
2
4
6
8
10
12
14
16
Ultimate strength (N)
A
0
1
2
3
4
5
Weeks
(Day)
AAV2-bFGF treated
Non-injection
Gain in the strength by
AAV2-bFGF gene therapy
AAV2-bFGF gene therapy
0
2
4
6
8
10
12
14
16
Ultimate strength (N)
B
0
1
2
3
4
5
Weeks
(Day)
AAV2-VEGF treated
Non-injection
Gain in the strength by
AAV2-VEGF gene therapy
AAV2-VEGF gene therapy

67
Chapter 6:  Gene Therapy for Tendon Healing
Figure 6-8  A, Biogenesis and biological effects of miRNAs. (Adapted from Mack GS: MicroRNA gets down to business,
Nat Biotechnol 25:631–638, 2007). B and C, Construction of a plasmid harboring in vitro engineered pre-miRNA to silence
expression of TGF-β1 gene of chickens in the tenocytes or in the healing tendon.
A
T
A
T
A
T
A
T
A
T
A
T
A
T
A
T
A
T
A
T
A
T
A
T
G
C
G
C
G
C
G
C
G
C
G
C
G
C
G
C
G
C
G
C
G
C
G
C
G
C
G
C
G
C
C
G
C
G
C
G
C
G
C
G
C
G
C
G
C
G
C
G
C
G
C
G
C
G
T
A
T
A
T
A
T
A
T
A
T
A
T
A
T
A
T
A
T
A
T
A
T
A
5′ miRNA flanking region
pCDNA 6.2-GW/
EmGFP-miR
(5699 bp)
Insertion of pre-miRNA
ds oligo
attB1
EmGFP
5′ miR flanking
region
ACGA
CAGG
3′ miR flanking
region
attB2
Dra I
Dra I
Sal I
BamH I
Bgl II
Xho I
S
p
e
c
ti
no
m
y
c
i
n
P
C
M
V
T
K

p
A
f
1

o
ri
S
V
4
0

or
i
E
M
7
B
l
a
s
ti
c
i
d
i
n
S
V
4
0
p
A
p
U
C
o
ri
3′ miRNA flanking region
Structures of engineered pre-miRNA oligo
TGCT
5′G + antisense
target sequences
Loop sequences
Sense ∆2 nt
target sequences
CAGG
pre-miRNA
ds oligo
B
C
A
Cytoplasm
Pasha
Drosha
Exportin 5
Pri-
miRNA
Nucleus
Target
transcript
or
Pre-miRNA
Dicer
Argonaute
Mature miRNA
RISC
Transcript degradation
Translational
repression
(requires only partial
complementarity)
We also found downregulation of the type III
collagen and CTGF genes after plasmid miRNA treat­
ment, but type I collagen gene expression was not
affected in cultured tenocytes and in the healing
tendons. This treatment meets an essential requirement:
type I collagen, an essential contributor to tendon
strength, was not affected, while expression of the
TGF-β1 gene was inhibited. These findings warrant
further exploration of RNAi gene therapy in preventing
adhesions.

68
Section 1:  Basic Science
1. Chang J, Thunder R, Most D, et al: Studies in flexor tendon
wound healing: neutralizing antibody to TGF-beta1 increases
postoperative range of motion, Plast Reconstr Surg 105:148–
155, 2000.
2. Khan U, Kakar S, Akali A, et al: Modulation of the formation
of adhesions during the healing of injured tendons, J Bone
Joint Surg (Br) 82:1054–1058, 2000.
References
biomechanically in vivo, though they were successful in
increasing the cellular activities or promoting healing
potential in vitro.
We believe that in vivo findings from a series of our
experiments provide support for future clinical trials,
laying a scientific ground for this treatment. Future
efforts will eventually determine the efficacy of these
gene therapy strategies in patients.
To expand the future scope of gene therapy for tendon
healing, the following needs to be considered: (1) We
tested only AAV2-bFGF and AAV2-VEGF treatments for
their ability to improve the healing strength. It is pos­
sible that therapies involving other growth factors, such
as PDGF-BB or IGF-1, could be equally effective. For
tissues that generally lack growth factors, sustained sup­
plementation of any (rather than “specific”) growth
factors (by gene therapy or other delivery methods such
as controlled release) would have a positive impact on
healing strength. Nevertheless, exogenous factors such
as TGF-β or CTGF should not be supplied, because they
are likely to exacerbate adhesions. (2) The goal of
tendon repair is to restore normal function of the
tendon, but its resolution is often hampered by two
factors: weak healing strength and adhesion formation,
which may require simultaneous biological manipula­
tion. Though we have considered the possibility of “dual
regulation” by increasing the factors favoring strength
gain, while inhibiting factors responsible for adhesions,
this idea of dual regulation remains untested. We specu­
late that such manipulation of tendon healing may be
desirable in some clinical situations such as with exten­
sive trauma. However, biological prevention of adhe­
sions may not be necessary for clean-cut tendon injuries,
because strong surgical repairs, augmented by growth
factor gene therapy, if necessary, allow rigorous active
tendon motion, which may effectively limit adhesions.
Gene therapy to inhibit adhesions may become reserved
exclusively for cases of extensive tendon or peritendi­
nous damage in which adhesions are pronounced and
unavoidable, or hand trauma precludes early tendon
movement. (3) Regarding gene therapy using miRNAs
or siRNA, our efforts have offered only preliminary evi­
dences. Comprehensive in vivo experiments examining
effectiveness are necessary. The significance of these
treatments will become clear only when further investi­
gations have generated more information.
FUTURE PERSPECTIVES
It is hard to speculate when gene therapy for tendon
healing will become a commonly available clinical
reality. Nevertheless, past investigations appear to have
identified a feasible path to overcome the slow biologi­
cal healing responses of the tendons reflected by the
“no-gain” period in strength. This is an important step
in biological augmentation of tendon healing. Previ­
ously, investigators have failed to augment tendon
healing reliably by other methods, and the effectiveness
was either small or could not be reproduced at other
time points. Most past efforts failed at increasing strength
Figure 6-9  A, Silencing effects of the miRNA on the
TGF-β1 gene expression and downregulation of the collagen
III and CTGF genes in vitro; and B, 6 weeks after delivery of
the miRNA to injured digital flexor tendons.
0
20
40
60
80
100
120
140
Gene expression (%)
TGF-beta
Collagen I
Collagen III
CTGF
p <0.001
p <0.001
p <0.001
0
20
40
60
80
100
120
140
Gene expression (%)
TGF-beta
Collagen I
Collagen III
CTGF
p <0.001
p <0.001
p <0.01
A
B
3. Lou J, Tu Y, Burns M, et al: BMP-12 gene transfer augmenta­
tion of lacerated tendon repair, J Orthop Res 19:1199–1202,
2001.
4. Hamada Y, Katoh S, Hibino N, et al: Effects of monofilament
nylon coated with basic fibroblast growth factor on endoge­
nous intrasynovial flexor tendon healing, J Hand Surg (Am)
31:530–540, 2006.

69
Chapter 6:  Gene Therapy for Tendon Healing
5. Tang JB, Cao Y, Zhu B, et al: Adeno-associated virus-2-medi­
ated bFGF gene transfer to digital flexor tendons significantly
increases healing strength. An in vivo study, J Bone Joint Surg
(Am) 90:1078–1089, 2008.
6. Yao J, Korotkova T, Riboh J, et al: Bioactive sutures for tendon
repair: assessment of a method of delivering pluripotential
embryonic cells, J Hand Surg (Am) 33:1558–1564, 2008.
7. Thomopoulos S, Das R, Silva MJ, et al: Enhanced flexor
tendon healing through controlled delivery of PDGF-BB,
J Orthop Res 27:1209–1215, 2009.
8. Chen CH, Zhou YL, Wu YF, et al: Effectiveness of microRNA
in down-regulation of TGF-β gene expression in digital flexor
tendons of chickens: in vitro and in vivo study, J Hand Surg
(Am) 34:1777–1784, 2009.
9. Abrahamsson SO, Lundborg G, Lohmander LS: Long-term
explant culture of rabbit flexor tendon: effects of recombinant
human insulin-like growth factor-I and serum on matrix
metabolism, J Orthop Res 9:503–515, 1991.
10. Abrahamsson SO, Lundborg G, Lohmander LS: Recombinant
human insulin-like growth factor-I stimulates in vitro matrix
synthesis and cell proliferation in rabbit flexor tendon,
J Orthop Res 9:495–502, 1991.
11. Gelberman RH, Amiel D, Harwood F: Genetic expression for
type I procollagen in the early stages of flexor tendon healing,
J Hand Surg (Am) 17:551–558, 1992.
12. Duffy FJ, Seiler JG, Gelberman RH, et al: Growth factors and
canine flexor tendon healing: initial studies in uninjured and
repair models, J Hand Surg (Am) 20:645–649, 1995.
13. Chang J, Most D, Stelnicki E, et al: Gene expression of trans­
forming growth factor beta-1 in rabbit zone II flexor tendon
wound healing: evidence for dual mechanisms of repair, Plast
Reconstr Surg 100:937–944, 1997.
14. Chang J, Most D, Thunder R, et al: Molecular studies in flexor
tendon wound healing: the role of basic fibroblast growth
factor gene expression, J Hand Surg (Am) 23:1052–1058, 1998.
15. Bidder M, Towler DA, Gelberman RH, et al: Expression of
mRNA for vascular endothelial growth factor at the repair site
of healing canine flexor tendon, J Orthop Res 18:247–252,
2000.
16. Boyer MI, Watson JT, Lou J, et al: Quantitative variation in
vascular endothelial growth factor mRNA expression during
early flexor tendon healing: An investigation in a canine
model, J Orthop Res 19:869–872, 2001.
17. Tsubone T, Moran SL, Amadio PC, et al: Expression of growth
factors in canine flexor tendon after laceration in vivo, Ann
Plast Surg 53:393–397, 2004.
18. Mehta V, Kang Q, Luo J, et al: Characterization of adenovirus-
mediated gene transfer in rabbit flexor tendons, J Hand Surg
(Am) 30:136–141, 2005.
19. Mehta V, Mass D. The use of growth factors on tendon inju­
ries, J Hand Ther 18:87–92, 2005.
20. Thomopoulos S, Harwood FL, Silva MJ, et al: Effect of several
growth factors on canine flexor tendon fibroblast prolifera­
tion and collagen synthesis in vitro, J Hand Surg (Am) 30:441–
447, 2005.
21. Thomopoulos S, Das R, Sakiyama-Elbert S, et al: bFGF and
PDGF-BB for tendon repair: controlled release and biologic
activity by tendon fibroblasts in vitro, Ann Biomed Eng
38:225–234, 2010.
22. Sakiyama-Elbert SE, Das R, Gelberman RH, et al: Controlled-
release kinetics and biologic activity of platelet-derived
growth factor-BB for use in flexor tendon repair, J Hand Surg
(Am) 33:1548–1557, 2008.
23. Tang JB, Xu Y, Ding F, et al: Tendon healing in vitro: promo­
tion of collagen gene expression by bFGF with NF-κB gene
activation, J Hand Surg (Am) 28:215–220, 2003.
24. Tang JB, Xu Y, Ding F, et al: Expression of genes for collagen
production and NF-κB gene activation of in vivo healing flexor
tendons, J Hand Surg (Am) 29:564–570, 2004.
25. Tang JB, Xu Y, Wang XT: Tendon healing in vitro: activation of
NIK, IKKα, IKKβ, and NF-κB genes in signal pathway and
proliferation of tenocytes, Plast Reconstr Surg 113:1703–1711,
2004.
26. Zhu B, Cao Y, Xin KQ, et al: Tissue reactions of adenoviral,
adeno-associated viral, and liposome-plasmid vectors in
tendons and comparison with early-stage healing responses
of injured flexor tendons, J Hand Surg (Am) 31:1652–1660,
2006.
27. Wang XT, Liu PY, Tang JB, et al: Tendon healing in vitro:
adeno-associated virus-2 effectively transduces intrasynovial
tenocytes with persistent expression of the transgene, but
other serotypes do not, Plast Reconstr Surg 119:227–234, 2007.
28. Chen CH, Cao Y, Wu YF, et al: Tendon healing in vivo: gene
expression and production of multiple growth factors in early
tendon healing period, J Hand Surg (Am) 33:1834–1842,
2008.
29. Wu YF, Chen CH, Cao Y, et al: Molecular events of cellular
apoptosis and proliferation in the early tendon healing
period, J Hand Surg (Am) 35:2–10, 2010.
30. Evans CH: Orthopedics: gene therapy’s dark horse, Gene Ther
11:343, 2004.
31. Evans CH, Ghivizzani SC, Robbins PD: Progress and Pros­
pects: genetic treatments for disorders of bones and joints,
Gene Ther 16:944–952, 2009.
32. Branski LK, Pereira CT, Herndon DN, et al: Gene therapy in
wound healing: present status and future directions, Gene
Ther 14:1–10, 2007.
33. Kubota H, Manske PR, Aoki M, et al: Effect of motion and
tension on injured flexor tendons in chickens, J Hand Surg
(Am) 21:456–463, 1996.
34. Hatanaka H, Zhang J, Manske PR: An in vivo study of locking
and grasping techniques using a passive mobilization proto­
col in experimental animals, J Hand Surg (Am) 25:260–269,
2000.
35. Boyer MI, Gelberman RH, Burns ME, et al: Intrasynovial
flexor tendon repair. An experimental study comparing low
and high levels of in vivo force during rehabilitation in
canines, J Bone Joint Surg (Am) 83:891–899, 2001.
36. Friedmann T, Roblin R: Gene therapy for human genetic
disease? Science 175:949–955, 1972.
37. Bainbridge JW, Smith AJ, Barker SS, et al: Effect of gene
therapy on visual function in Leber’s congenital amaurosis,
N Engl J Med 358:2231–2239, 2008.
38. Mancuso K, Hauswirth WW, Li Q, et al: Gene therapy for
red-green colour blindness in adult primates, Nature 461
(7265):784–787, 2009.
39. Cartier N, Hacein-Bey-Abina S, Bartholomae CC, et al: Hema­
topoietic stem cell gene therapy with a lentiviral vector in
X-linked adrenoleukodystrophy, Science 326 (5954):818–
823, 2009.
40. Chong AK, Ang AD, Goh JC, et al: Bone marrow-derived
mesenchymal stem cells influence early tendon-healing in a
rabbit Achilles tendon model, J Bone Joint Surg (Am) 89:74–
81, 2007.
41. Chan BP, Fu S, Qin L, et al: Effects of basic fibroblast growth
factor (bFGF) on early stages of tendon healing: a rat patellar
tendon model, Acta Orthop Scand 71:513–518, 2000.
42. Henn RF, Kuo CE, Kessler MW, et al: Augmentation of zone
II flexor tendon repair using growth differentiation factor 5
in a rabbit model, J Hand Surg (Am) 35: 1825–1832, 2010.
43. Thomopoulos S, Kim HM, Das R, et al: The effects of exo­
genous basic fibroblast growth factor on intrasynovial flexor

70
Section 1:  Basic Science
tendon healing in a canine model, J Bone Joint Surg (Am)
92:2285–2293, 2010.
44. Monahan PE, Samulski RJ: AAV vectors: is clinical success on
the horizon? Gene Ther 7:24–30, 2000.
45. Erles K, Sebökovà P, Schlehofer JR: Update on the prevalence
of serum antibodies (IgG and IgM) to adeno-associated virus
(AAV), J Med Virol 59:406–411, 1999.
46. Daly TM: Overview of adeno-associated viral vectors, Methods
Mol Biol 246:157–165, 2004.
47. Mueller C, Flotte TR: Clinical gene therapy using recombi­
nant adeno-associated virus vectors, Gene Ther 15:858–863,
2008.
48. Aitken ML, Moss RB, Waltz DA, et al: A phase I study of
aerosolized administration of tgAAVCF to cystic fibrosis sub­
jects with mild lung disease, Hum Gene Ther 12:1907–1916,
2001.
49. Kaplitt MG, Feigin A, Tang C, et al: Safety and tolerability of
gene therapy with an adeno-associated virus (AAV) borne
GAD gene for Parkinson’s disease: an open label, phase I trial,
Lancet 369(9579):2097–2105, 2007.
50. Wang XT, Liu PY, Xin KQ, et al: Tendon healing in vitro: bFGF
gene transfer to tenocytes by adeno-associated viral vectors
promotes expression of collagen genes, J Hand Surg (Am)
30:1255–1261, 2005.

71
CHAPTER
7 
TENDON TISSUE ENGINEERING
AND BIOACTIVE SUTURE
REPAIR
Brian C. Pridgen, BS, Jeffrey Yao, MD, and
James Chang, MD
plantaris, and toe extensor tendons. Limitations of
autograft harvesting include donor site morbidity and
additional anesthesia and operative time. In cases of
mutilating injuries to the hand (Figure 7-1), the demand
for graft material may outstrip the autograft supply in a
patient, leaving these patients without a good solution
for restoring hand function.
Although synthetic graft materials such as Dacron
grafts and Silastic rods have been used for tendon recon­
struction, outcomes are generally poor, and the life
span of these grafts tends to be short. Failure is typically
due to mechanical wear or rupture of the tendon–graft
interface. These synthetic materials are designed to
replace tendon tissue, rather than to regenerate it. Thus,
the resulting repair does not heal with tendon tissue
properly and does not support the ingrowth of cells
to promote biodegradation and remodeling of the
implant.
TISSUE ENGINEERING
One of the earliest articles broadly defining tissue engi­
neering was a 1993 Science article by Langer and Vacanti.1
In this article, they defined the nascent field of tissue
engineering as “an interdisciplinary field that applies
the principles of engineering and the life sciences toward
the development of biological substitutes that restore,
maintain, or improve tissue function.” They described
three general strategies of tissue engineering: (1) cells or
cell substitutes, (2) tissue-inducing substances such as
growth factors and delivery vehicles, and (3) cell-seeded
scaffolds. These principles have been applied to clinical
needs in hand surgery, resulting in commercially avail­
able products including engineered skin substitutes2
and allogeneic nerve grafts.3 However, a tissue-engineered
tendon is not yet available.
TENDON BIOLOGY
At their simplest, tendons can be thought of as thick
collagen ropes that transmit forces from muscles to act
across joints to move and stabilize them. Development
of a tissue-engineered tendon requires a much deeper
OUTLINE
Tissue engineering is a field that aims to apply our
expanding basic knowledge of tissue biology to the
design of engineered tissue substitutes. Despite
advances in tendon repair, many patients still experi­
ence poor outcomes that can lead to lifelong disability.
Tendon tissue engineering and bioactive sutures are
well suited to address this clinical need. In injuries to
the hand that require interposition tendon grafts, the
need for autograft material may outstrip the supply in
a given patient, or the available autograft material may
lead to suboptimal outcomes for the patient. A tissue-
engineered tendon would be readily available as an
off-the-shelf product designed to optimize the healing
process and to recapitulate normal function. Design
of such a tendon requires a detailed understanding of
the tendon healing process as well as normal tendon
function and biology, including its structure and com­
position, cells and their role in tendon function, and
biomechanical properties of tendons. This under­
standing can then be applied to the four primary com­
ponents of a tissue-engineered tendon: cells, scaffolds,
growth factors, and mechanical manipulation. Injuries
requiring suture repair, whether or not an interposi­
tion graft is necessary, often have poor outcomes
despite advances in suture techniques and materials.
Current research into bioactive sutures loaded with
growth factors or cells offers a new strategy for strength­
ening and improving suture repairs of tendons.
Despite recent advances in flexor tendon repair surgery,
many patients still experience poor outcomes following
repair. This is often due to severe adhesion formation,
which may require tenolysis procedures to improve
motion. Additional causes of poor outcomes include
tendon rupture or failure at the repair site, due to either
suture pull-through, suture breakage, knot slippage, or
gapping at the repair site.
When graft material is required, repairs are currently
limited to autografts such as the palmaris longus,

72
Section 1:  Basic Science
zone 2, better outcomes are achieved when reconstruc­
tion is performed with intrasynovial autografts rather
than extrasynovial autografts.
Brockis and others made early observations about the
vascular supply to flexor tendons.7 They observed that
tendons receive their blood supply through the muscu­
lotendinous junction, the paratenon surrounding the
tendon in the extrasynovial region, the vincula within
the intrasynovial region, and the osteotendinous inser­
tion. In addition to vascular nutrition, the intrasynovial
portion of the tendon receives nutrition by diffusion
through the synovial fluid.8,9
Cells
Tenocytes, the primary cell type in tendons, are spindle-
shaped fibroblast-like cells oriented along the long axis
of tendons that are responsible for the production and
maintenance of the extracellular matrix of the tendon.
As they mature, their metabolic activity and energy
demand decrease. This enables tendons to carry loads
and maintain tension for long periods of time without
becoming ischemic and necrotic. However, during
injury and repair, this slow metabolic rate leads to pro­
longed healing.
Tendon Healing
Tendons heal slowly because of their hypocellularity,
hypovascularity, and low metabolic rates.10 This pro­
longed healing process requires immobilization to
reduce the risk of tendon rupture or gap formation. The
duration of this immobilization must be balanced
against the risk of adhesion formation. Modulating this
healing process to create a strong tendon repair with
minimal adhesion formation requires an understanding
of the three stages of tendon healing: inflammatory,
proliferative, and remodeling.4,10
The inflammatory stage occurs within the first 24
hours and is characterized by recruitment of inflamma­
tory cells. The proliferative stage begins as tenocytes are
recruited to the site of injury and begin to produce
extracellular matrix components, including collagen
type III. After several weeks, the remodeling stage begins.
The randomly organized extracellular matrix produced
during the proliferative stage is organized into parallel
fibrils, while the collagen type III production decreases
and collagen type I production increases. As healing is
completed, tenocyte number and metabolism then
begin to decrease.
Extrinsic Versus Intrinsic Healing
It was originally thought that tendons must heal by an
extrinsic process in which ingrowth of new cells occurred
through vascularized adhesions to surrounding tissue.11
It was later demonstrated that intrasynovial tendons
could heal by a process mediated by cells intrinsic to
the tendon.12,13 This suggested that adhesion formation
understanding. The mechanical properties of tendons
crucial to their proper functioning rely not only on
the structure and organization of its major collagen
constituents but also on the other extracellular matrix
components. At steady state, tendons are relatively
hypocellular and have low metabolic demands. During
times of injury and healing, cells and their nutritional
pathways play a critical role, and their activity is a major
determinant of tendon healing.
Structure and Anatomy
Collagen is the basic structural unit of tendons.4,5 It is
organized into a hierarchical structure of progressively
larger and more organized subunits to form the crimped
or sinusoidal pattern visible by light microscopy.
While collagen is the major constituent of tendons,
proteoglycans are critical to the biological and mechani­
cal functioning of tendons.6 Proteoglycans promote
hydration through hydrophilic subunits and mediate
interactions between structural units of the tendon,
which are important determinants of the viscoelastic
properties of tendons. Another important function of
proteoglycans is their role in promoting cell signaling
to mediate growth, proliferation, and migration.
Synovial sheaths around tendons are found in areas
subjected to mechanical stress in which efficient gliding
is mediated by synovial lubrication. Examples of tendons
of the hand with intrasynovial portions include the
flexor digitorum superficialis and flexor digitorum pro­
fundus tendons. These tendons have an intrasynovial
region in their more distal portions within the hand and
an extrasynovial region in their more proximal portions.
In addition to assisting with gliding in areas of high
mechanical stress, the synovial sheath also provides
nutrition by diffusion through synovial fluid. Extrasyno­
vial tendons are covered by the paratenon, a loose
connective tissue that carries the blood supply. This dis­
tinction is important because although overall outcomes
are generally unsatisfactory with injuries to intrasyno­
vial tendons and graft reconstruction, particularly in
Figure 7-1  Mutilating injury to the forearm and hand with
significant soft tissue loss, including tendon damage. This
patient required extensive reconstruction and would have
benefited from a tissue-engineered tendon.

73
Chapter 7:  Tendon Tissue Engineering and Bioactive Suture Repair
tenocyte-seeded scaffolds resembled native tendons in
histologic structure, collagen arrangement, and breaking
strength.14 However, tenocytes may not be a practical
source of cells because autologous tendon harvest
required for their isolation results in donor site morbid­
ity. They also expand very slowly in culture, which may
make them further impractical due to a lengthy in vitro
expansion time following cell isolation and before
seeding of the tendon construct.
Human dermal fibroblasts are another terminally dif­
ferentiated cell type that closely resembles tenocytes and
have been widely considered as a candidate cell line for
tendon tissue engineering. Like tenocytes, dermal fibro­
blasts are derived from the mesoderm, have similar
morphology, and have similar extracellular matrix pro­
duction capacity. Harvest of autologous dermal fibro­
blasts through a simple skin biopsy would cause
minimal morbidity or expense, which is a notable
advantage over tenocytes, and they are easily and rapidly
expanded in culture. In a study comparing human teno­
cytes and human dermal fibroblasts, each cell type was
isolated and grown separately on a polyglycolic acid
scaffold.15 The authors found no difference in a number
of outcome measurements comparing tenocyte-seeded
and dermal fibroblast–seeded tendon constructs, includ­
ing gross structure, histology, collagen deposition, col­
lagen fibril diameter, and construct strength. These
results suggest that dermal fibroblasts may be a suitable
substitute for tenocytes, which will also avoid the prob­
lems with using autologous tenocytes.
Mesenchymal stem cells (MSCs) are a multipotent
cell line present in adult bone marrow and adipose
tissue that may be cultured as undifferentiated cells
or driven to differentiate into mesenchymal lineages
including tenocytes, osteoblasts, chondrocytes, and adi­
pocytes.16 Although harvest of MSCs from bone marrow
is invasive, provides low yields, and requires time con­
suming and expensive in vitro expansion in culture, it
does not carry the donor site morbidity of tenocyte
harvest. Several groups have demonstrated improved
tendon strength, remodeling, and tissue formation
when seeded with bone marrow–derived MSCs.17-19
Adipose-derived MSCs, on the other hand, are easily
obtainable in sufficient quantities with minimal mor­
bidity, and they have been shown to have a similar dif­
ferentiating potential to bone marrow–derived MSCs.20
One of the concerns with the use of MSCs is the risk for
undesirable differentiation down nontenogenic lines.
Several authors have noted ectopic bone formation with
MSCs used for tendon tissue engineering,18,21 although
another group modulated MSC differentiation by over­
expressing Smad8 to promote tenogenic differentia­
tion.22 Further research will be necessary to better
understand MSC differentiation and replication. This
may include identifying possible subpopulations of
MSCs with predispositions to differentiate down a
might not be necessary for tendon healing if cells within
the damaged tendon are available to mediate repair.
TENDON TISSUE ENGINEERING
Tissue engineering a tendon requires an understanding
of the basic science of tendon healing and tendon func­
tion. This knowledge can then be applied to designing
and choosing the components of a tissue-engineered
tendon, which can include cells, a scaffold, growth
factors, and mechanical manipulation (Figure 7-2),
some or all of which can be used to create the final
tissue-engineered product.
Cells
As the machinery responsible for tissue repair and main­
tenance, cells are a necessary component of a tissue-
engineered construct in order to achieve functional
restoration of the tendon. Despite this clear goal of a
tissue-engineered construct, the steps necessary to
achieve this goal remain uncertain. Questions research­
ers continue to face include determining the appropri­
ate cell type and cell source and determining the site of
cell seeding.
Cell Type and Cell Source
Selection of the cell type and cell source for seeding of
the tendon construct is an important decision that must
consider (1) the ease of harvest of the chosen cell type
without inflicting collateral morbidity at the donor site,
(2) the ability to culture and expand the cells in vitro,
and (3) whether the chosen cells will achieve functional
restoration of the tendon. Cell types that have been
considered include tenocytes, dermal fibroblasts, and
mesenchymal stem cells.
Tenocytes are the predominant cell type of native
tendon. Thus, they should be capable of restoring func­
tion of a tissue-engineered tendon construct. In a study
using unwoven polyglycolic acid scaffolds seeded with
hen tenocytes, Cao and colleagues demonstrated that
Figure 7-2  Components of a tissue-engineered tendon
and some of the considerations necessary for each of these
components. A final tissue-engineered tendon product may
contain some or all of these components.
Cells
Cell type
Cell source
Site of cell seeding
Autogenous vs.
allogeneic cells
Growth factors
IGF-I
TGF-β
VEGF
PDGF
bFGF
Scaffold
Biologic scaffolds
Synthetic scaffolds
Acellularized scaffolds
Mechanical manipulation
Continuous strain
Cyclic strain
Intermittent strain

74
Section 1:  Basic Science
although in vitro manipulation carries some risk of
infectious contamination.
The greatest concern with the use of allogeneic cells
is the risk of immune rejection due to incompatibility
with the recipient. However, recent work has shown that
the use of allogeneic MSCs avoids an immune response
in recipients. Possible explanations for this include low
expression of MHC-II and disruption of leukocyte
function.25
The use of allogeneic cells presents a number of
advantages over autogeneic cells. First, an existing supply
of allogeneic cells would eliminate the need for a time-
intensive and costly harvest and in vitro expansion of
autogeneic cells. Having this existing supply would
allow clinicians to more rapidly seed a scaffold and to
provide definitive repair without a staged process of cell
harvest, in vitro cell expansion, and implantation.
Second, there may be some regulatory advantages to
using allogeneic cells. By having a large bank of cultured
cells, cell samples could be removed for safety testing
without significantly reducing the cell count and neces­
sitating further in vitro expansion, as would be neces­
sary for safety testing of autogeneic cells. Last, MSCs
have been shown to decrease in number, life span, and
proliferative capacity with increasing donor age.26 This
age-related decline of MSCs may make autogeneic MSCs
an impractical option for seeding of a scaffold in older
patients. An available source of allogeneic MSCs from
younger donors would avoid this problem.
Scaffolds
Cultured cells alone are incapable of replacing the func­
tion of a tendon. They require the three-dimensional
scaffolding of the extracellular matrix in tendons. In
addition to providing an attachment site for cells, the
tendon extracellular matrix provides the mechanical
properties to the tendon and has attached signaling
molecules important for cell growth, proliferation, and
migration. The role of a scaffold in tissue engineering is
to provide a three-dimensional biocompatible and bio­
degradable construct to support growth, proliferation,
and migration of cells and to allow remodeling by
cells without the production of toxic or inflammatory
degradation products. The scaffold should provide
mechanical strength sufficient to permit postoperative
mobilization therapy without rupture of the tendon
graft. Scaffolds have been produced from natural bio­
logic materials, synthetic materials, or acellularized
cadaver tissue.
Biologic Scaffolds
Because collagen is the primary component of tendons,
it has been widely investigated as a possible scaffold for
tendon tissue engineering. As the natural major con­
stituent of tendons, collagen is highly biocompatible
with safe degradation products, and it supports cell
particular lineage and to better characterize tenocyte
markers to verify proper differentiation.
Site of Seeding (In Vivo versus In Vitro)
Much research has focused on seeding scaffolds with
cells in vitro prior to implantation. This offers a possible
advantage of providing a cell population intrinsic to the
tissue-engineered construct that may initiate the healing
process. This may mediate an intrinsically mediated
healing process that may result in fewer adhesions,
while promoting more rapid healing at the repair site,
remodeling, and engraftment. Many of the previously
mentioned studies have successfully demonstrated cell
attachment to the surface of tendon scaffolds. However,
there has been limited success with establishing cell
populations deep to the surface of tendon scaffolds.
Some groups have attributed this to the tight weave of
the scaffold matrix, which inhibits cells from migrating
deeper into the tendon.23 Another possible explanation
could be poor nutritional supply deeper into the tendon
in the absence of the native intratendinous vascular
supply or nutrition by diffusion of synovial fluid. This
may be of particular concern for newly seeded cells that
are metabolically more active than native tenocytes,
thus inhibiting them from migrating deeper from the
surface. Despite the difficulty with seeding deep to the
surface, it has been shown that host cells infiltrate and
replace seeded cell populations in a time-dependent
manner.24 Further work will be needed to understand
the importance of establishing cell populations deep to
the surface of the tendon and how to promote seeding
deeper into the core of the tendon.
Another option would be to implant an acellular
scaffold and to rely on the patient’s body to cellularize
the graft. To allow more rapid healing at the repair
site, remodeling, and engraftment, some groups have
attempted to accelerate this process of in vivo seeding
by using guided regeneration. This includes the use of
growth factors or using an acellularized tissue scaffold
that contains important biomolecules. Advantages to
this approach include avoiding in vitro cell expansion
and seeding, which would be expensive and time con­
suming, carry infectious risks to patients, and involve
regulatory hurdles. Challenges to this approach include
requirement of a detailed understanding of the biology
of tendon healing and the temporal distribution of key
growth factors. Additionally, as the seeding process
would be dependent on extrinsic cells, this may mediate
adhesion formation.
Autogenous versus Allogeneic Cell Seeding
Much of the research in tendon tissue engineering has
focused on using autogenous cells for seeding of the
scaffold. This approach avoids problems with immuno­
logic incompatibility and rejection of the grafted cells.
It also reduces concerns of infection transmission,

75
Chapter 7:  Tendon Tissue Engineering and Bioactive Suture Repair
metabolites, including glycolic acid and lactic acid,
these metabolites are acidic and may give rise to sys­
temic or local reactions.33 Such a reaction may promote
an inflammatory response that may delay healing by
killing cells, disrupting the newly forming neotendon,
and promoting adhesion formation. Another hurdle to
the implantation of synthetic scaffolds is their poor
mechanical strength. They are typically far weaker than
native tendon, and it takes several weeks of in vitro or
in vivo culture with cells before these constructs begin
to approach the strength of native tendon.14,19 If in vitro
maturation of the scaffold is performed, this will be
expensive, put the tissue-engineered construct at risk of
infection, and require the patient to wait for an extended
period of time for the repair if autogeneic cells are used.
If in vivo maturation of the scaffold is performed, this
will put the patient at risk for adhesion formation and
joint contracture with prolonged immobilization to
prevent rupture of the weak graft.
Acellularized Scaffolds
One of the disadvantages of biologic and synthetic scaf­
folds is that they are biologically inert; they lack the
signaling molecules necessary to promote cell migra­
tion, cell proliferation, and extracellular matrix produc­
tion. Although molecules may be incorporated into
these structures, tendon biology is not yet well under­
stood enough to entirely recapitulate the local
environment.
Native tendon, on the other hand, contains many of
these signaling molecules in the form of proteoglycans.
Native tendon also possesses the mechanical strength
necessary to withstand normal loading forces without
the need for lengthy in vitro or in vivo neotendon
formation, unlike many of the biological or synthetic
scaffolds. However, native tissue containing these bio­
logical and mechanical advantages cannot be grafted
from one individual to another because of the existing
immunogenic cellular material. Instead, the tissue must
be acellularized to remove the immunogenic compo­
nents without compromising the signaling molecules,
mechanical strength, and biocompatibility.
A number of groups have produced a scaffold from
fresh tendons treated with detergents to remove immu­
nogenic cellular material (Figure 7-3).34,35 They demon­
strated that these acellularized tendons retained the
biochemical signaling molecules and mechanical prop­
erties important for proper biological and mechanical
functioning of the tendon. They went on to show that
these scaffolds are capable of supporting the attachment
and growth of allogeneic cells, suggesting that these
acellularized tendon scaffolds may be suitable for clini­
cal use in tendon tissue engineering. Other tissue that
has been considered for producing an acellularized scaf­
fold for tendon tissue engineering include acellular
dermal matrix36 and human umbilical vein.37
attachment and proliferation better than synthetic
materials.27 Collagen scaffolds may be made either as a
gel or as a sponge. Collagen gels alone do not possess
inherent mechanical strength to act as a suitable scaf­
fold; they must be used with a supporting material, such
as suture.18 Collagen sponges possess superior mechani­
cal strength but have inferior cell attachment pro­
perties.28 Some groups have combined the superior
cell-seeding characteristics of collagen gels with the
superior mechanical characteristics of collagen sponges
in a joint gel–sponge construct.29 While these scaffolds
have the advantages of being easily manufacturable with
natural nonimmunogenic materials, they lack GAG sig­
naling molecules and do not have the mechanical
strength of native tendon.
Synthetic Scaffolds
The most commonly used biodegradable synthetic scaf­
folds for tendon tissue engineering are polyesters,
including polylactic acid (PLA), polyglycolic acid (PGA),
and their copolymer polylactic-co-glycolic acid (PLGA).
The ability to manufacture them synthetically permits
researchers to manipulate the material properties of
the scaffold to achieve the appropriate balance of
degradability and scaffold mechanical strength and to
create a surface environment that encourages cell attach­
ment, migration, and proliferation. A study by Lu and
coworkers compared the suitability of PLA, PGA, and
PLGA as a scaffold for ACL reconstruction.30 They found
that although PGA initially was the strongest of the
three materials, it rapidly degraded with in vitro cell
culture, which would put such a graft at risk of rupture
if used in vivo. When seeded with ACL fibroblasts, the
PLA scaffold supported the greatest number of cells
with normal morphology. In a different study by
Ouyang and colleagues, they observed that PLGA
scaffolds supported the greatest attachment and pro­
liferation of bone marrow–derived MSCs.31 Groups
continue to use a variety of scaffolds with no agreement
yet in the literature, which could attributable to a variety
of factors, including differing manufacturing specifica­
tions of the polymer scaffolds, cell types used, culture
conditions, or mechanical testing parameters. Other
observations that have been made using synthetic scaf­
folds include the use of a fibronectin coating to over­
come the surface hydrophobicity of the scaffolds to
promote cell attachment30; the use of knitted scaffolds,
rather than braided scaffolds, to promote improved cell
attachment and ingrowth19; and electrospinning of
nanofibers to promote cell attachment, proliferation,
and migration.32
Despite the advantages of having a manufacturable
and easily modified synthetic scaffold and the many
recent advances made in this field, there are a number
of disadvantages that remain. Although polyester scaf­
folds are degraded by hydrolysis into naturally occurring

76
Section 1:  Basic Science
alignment and that this collagen was composed pre­
dominantly of collagen type I typically found in tendon
rather than collagen type III typically found in skin,
suggesting a phenotype switch of the dermal fibroblasts
to a tenogenic cell type. They also noted the improved
strength of their tissue engineered constructs following
mechanical loading. A study by another group demon­
strated that intermittent cyclic strain, rather than con­
tinuous strain, promoted greater cell proliferation, type
I collagen production, and maintenance of tenocyte
morphology.40 They then went on to demonstrate that
acellular tendon scaffolds re-seeded with cells and then
subjected to intermittent cyclic strain are stronger than
re-seeded tendons that were not subjected to strain. An
example of a tendon bioreactor is shown in Figure 7-4.
BIOACTIVE SUTURE
Once a tissue-engineered tendon is developed, it must
be interposed into the defect and then sutured into
place. Following repair, patients should initiate early
postoperative mobilization therapy to reduce adhesion
formation. However, the benefits of early mobilization
must be weighed against the risk for gap formation,
which can lead to delayed healing and reduced tensile
strength at the repair site. Recent advances in suturing
Growth Factors
The temporal distribution of growth factors during the
healing process is still being investigated, but it is known
that they play an important role in mediating the tendon
repair process. Well-studied growth factors involved in
tendon healing include insulin-like growth factor-I
(IGF-I), transforming growth factor-β (TGF-β), vascular
endothelial growth factor (VEGF), platelet-derived
growth factor (PDGF), and basic fibroblast growth factor
(bFGF).4,38 These growth factors have roles in inflamma­
tion, cell migration and proliferation, collagen produc­
tion, angiogenesis, and remodeling of the tendon.
Methods of growth factor delivery include seeding of
genetically engineered cells, loading with vectors har­
boring the genes of interest, or direct delivery of growth
factors. Although much progress has been made, further
research will be required. A better understanding of
the temporal changes and role of growth factors in the
different stages of healing will be necessary. Further
research into delivery methods will also be necessary to
determine a method that is safe, efficacious, and cost
effective.
Mechanical Manipulation
Mechanical strain is known to be an important con­
tributor to tenocyte function, maintenance of a teno­
genic phenotype, and collagen production.39 One group
noted in a series of studies the role of mechanical
loading on PGA scaffolds seeded with human dermal
fibroblasts.15 They found that mechanical loading
promoted maturation of collagen fibers with parallel
Figure 7-4  A bioreactor designed to accommodate a
full-length human flexor tendon. The system can be
connected to a motor and programmable control system
that permits continuous, cyclic, and/or intermittent strain to
be applied.
Figure 7-3  Acellularization of a full-length human tendon.
A, Untreated control tendon showing nuclear staining with
a fluorescent nucleic acid stain. B, Acellularized tendon
showing removal of nuclear material.
A
B

77
Chapter 7:  Tendon Tissue Engineering and Bioactive Suture Repair
survive the trauma of passage through acellularized
tendons, proliferate, and remain metabolically active.
These cells repopulate the acellular zones shown to sur­
round suture material in repaired tendons, and may
thereby accelerate healing. Animal studies to evaluate
the use of bioactive sutures are currently ongoing.
These techniques will be beneficial both for direct
repair of tendon ends or for placement of a tissue-
engineered interposition graft. By promoting early
healing and cell proliferation at the repair site through
the use of bioactive sutures, early tensile strength may
be increased and overall healing time decreased. This
may permit earlier initiation of safe postoperative
mobilization with minimal risk of gapping or tendon
rupture.
CONCLUSION
Despite many of the recent advances in flexor tendon
repair, many patients still experience poor outcomes.
Adhesion formation leads to reduced mobility. Large
defects require autograft harvesting, which carries mor­
bidity and may be insufficient in extensive mutilating
injuries. Traditional suture repair requires delayed post­
operative mobilization due to the risk of gap formation.
Continuing research in tendon biology has increased
our understanding of tendon structure, function, and
healing. The application of this basic science research
through the use of tissue-engineering principles has led
to the development of many technologies that promise
to one day improve the repair and reconstruction of
flexor tendon injuries.
techniques and suture material have improved the
strength of tendon repairs, but the risk of gap formation
with early mobilization has not been eliminated. Addi­
tionally, the trauma from suture repair has been impli­
cated in creating acellular zones and inflammation that
delay tendon healing.41
Several groups have investigated the use of bioactive
suture to augment tendon repairs. The two primary
strategies under investigation are to modulate cells
present at the repair site through growth factors incor­
porated into the suture or to attach cells directly onto
the suture with the intention of seeding them to the
repair site.
Multiple groups have investigated the use of suture
as a drug delivery device. An in vitro study using epi­
dermal growth factor seeded onto Mersilene suture
demonstrated increased cell proliferation along the
coated suture.42In another study, bFGF was loaded onto
a nylon monofilament suture used to repair rabbit
tendon injury.43 The bioactive suture increased strength
of the repair and promoted cell proliferation at the
repair site. A third group used growth differentiation
factor-5 to coat sutures and achieved similar results to
the previous groups, demonstrating increased strength
of the tendon repair and enhanced cell proliferation.44
Another group has attempted to address directly the
problem of the acellular zones at the site of suture repair
by incorporating pluripotential cells into the suture.45
These pluripotential cells have been shown to be deliv­
ered to a tendon repair–reconstruction site (Figure 7-5)
as well as along the entire path of the suture. The cells
Figure 7-5  Seeding of mouse embryo pluripotential cells onto FiberWire suture. A, Control suture with no coating and few
adherent cells. B, Suture with poly(L-lysine) coating mediating cell adhesion. C, DAPI staining of cells adherent to the suture.
A
B
C
References
3. Whitlock EL, Tuffaha SH, Luciano JP, et al: Processed allografts
and type I collagen conduits for repair of peripheral nerve
gaps, Muscle Nerve 39:787–799, 2009.
1. Langer R, Vacanti JP: Tissue engineering, Science 260:920–
926, 1993.
2. Lou RB, Hickerson WL: The use of skin substitutes in hand
burns, Hand Clin 25:497–509, 2009.

78
Section 1:  Basic Science
4. James R, Kesturu G, Balian G, et al: Tendon: Biology, biome­
chanics, repair, growth factors, and evolving treatment
options, J Hand Surg (Am) 33:102–112, 2008.
5. Lin TW, Cardenas L, Soslowsky LJ: Biomechanics of tendon
injury and repair, J Biomech 37:865–877, 2004.
6. Doroski DM, Brink KS, Temenoff JS: Techniques for biologi­
cal characterization of tissue-engineered tendon and liga­
ment, Biomaterials 28:187–202, 2007.
7. Brockis JG: The blood supply of the flexor and extensor
tendons of the fingers in man, J Bone Joint Surg (Br) 35:131–
138, 1953.
8. Lundborg G, Holm S, Myrhage R: The role of the synovial
fluid and tendon sheath for flexor tendon nutrition. An exper­
imental tracer study on diffusional pathways in dogs, Scand
J Plast Reconstr Surg 14:99–107, 1980.
9. Abrahamsson SO, Lundborg G, Lohmander LS: Tendon
healing in vivo. An experimental model, Scand J Plast Reconstr
Surg Hand Surg 23:199–205, 1989.
10. Sharma P, Maffulli N: Biology of tendon injury: healing,
modeling and remodeling, J Musculoskelet Neuronal Interact
6:181–190, 2006.
11. Potenza AD: Critical evaluation of flexor-tendon healing and
adhesion formation within artificial digital sheaths, J Bone
Joint Surg (Am) 45:1217–1233, 1963.
12. Matthews P, Richards H: The repair potential of digital flexor
tendons. An experimental study, J Bone Joint Surg (Br)
56:618–625, 1974.
13. Lundborg G: Experimental flexor tendon healing without
adhesion formation: A new concept of tendon nutrition and
intrinsic healing mechanisms. A preliminary report, Hand
8:235–238, 1976.
14. Cao Y, Liu Y, Liu W, et al: Bridging tendon defects using
autologous tenocyte engineered tendon in a hen model, Plast
Reconstr Surg 110:1280–1289, 2002.
15. Deng D, Liu W, Xu F, et al: Engineering human neo-tendon
tissue in vitro with human dermal fibroblasts under static
mechanical strain, Biomaterials 30:6724–6730, 2009.
16. Pittenger MF, Mackay AM, Beck SC, et al: Multilineage poten­
tial of adult human mesenchymal stem cells, Science 284:143–
147, 1999.
17. Young RG, Butler DL, Weber W, et al: Use of mesenchymal
stem cells in a collagen matrix for Achilles tendon repair,
J Orthop Res 16:406–413, 1998.
18. Awad HA, Boivin GP, Dressler MR, et al: Repair of patellar
tendon injuries using a cell-collagen composite, J Orthop Res
21:420–431, 2003.
19. Ouyang HW, Goh JCH, Thambyah A, et al: Knitted poly-
lactide-co-glycolide scaffold loaded with bone marrow
stromal cells in repair and regeneration of rabbit Achilles
tendon, Tissue Eng 9:431–439, 2003.
20. Lee RH, Kim B, Choi I, et al: Characterization and expression
analysis of mesenchymal stem cells from human bone marrow
and adipose tissue, Cell Physiol Biochem 14:311–324, 2004.
21. Harris MT, Butler DL, Boivin GP, et al: Mesenchymal stem
cells used for rabbit tendon repair can form ectopic bone and
express alkaline phosphatase activity in constructs, J Orthop
Res 22:998–1003, 2004.
22. Hoffmann A, Pelled G, Turgeman G, et al: Neotendon forma­
tion induced by manipulation of the Smad8 signalling
pathway in mesenchymal stem cells, J Clin Invest 11:940–952,
2006.
23. Whitlock PW, Smith TL, Poehling GG, et al: A naturally
derived, cytocompatible, and architecturally optimized scaf­
fold for tendon and ligament regeneration, Biomaterials 28:
4321–4329, 2007.
24. Thorfinn J, Saber S, Angelidis IK, et al: Flexor tendon tissue
engineering: temporal distribution of donor tenocytes versus
recipient cells, Plast Reconstr Surg 124:2019–2026, 2009.
25. Ryan JM, Barry FP, Murphy JM, et al: Mesenchymal stem cells
avoid allogeneic rejection, J Inflamm (Lond) 2:8, 2005.
26. Stenderup K, Justesen J, Clausen C, et al: Aging is associated
with decreased maximal life span and accelerated senescence
of bone marrow stromal cells, Bone 33:919–926, 2003.
27. Qin T, Yang Z, Wu Z, et al: Adhesion strength of human
tenocytes to extracellular matrix component-modified
poly(dl-lactide-co-glycolide)
substrates,
Biomaterials
26:
6635–6642, 2005.
28. Butler DL, Juncosa-Melvin N, Boivin GP, et al: Functional
tissue engineering for tendon repair: A multidisciplinary
strategy using mesenchymal stem cells, bioscaffolds, and
mechanical stimulation, J Orthop Res 26:1–9, 2008.
29. Juncosa-Melvin N, Boivin GP, Gooch C, et al: The effect of
autologous mesenchymal stem cells on the biomechanics
and histology of gel-collagen sponge constructs used for
rabbit patellar tendon repair, Tissue Eng 12:369–379, 2006.
30. Lu HH, Cooper JA, Manuel S, et al: Anterior cruciate ligament
regeneration using braided biodegradable scaffolds: In vitro
optimization studies, Biomaterials 26:4805–4816, 2005.
31. Ouyang HW, Goh JCH, Mo XM, et al: Characterization of
anterior cruciate ligament cells and bone marrow stromal
cells on various biodegradable polymeric films, Mater Sci Eng
C 20:63–69, 2002.
32. Sahoo S, Ouyang H, Goh JC, et al: Characterization of a novel
polymeric scaffold for potential application in tendon/
ligament tissue engineering, Tissue Eng 12:91–99, 2006.
33. Böstman OM, Pihlajamäki HK: Adverse tissue reactions to
bioabsorbable fixation devices, Clin Orthop Relat Res 371:216–
227, 2000.
34. Cartmell JS, Dunn MG: Effect of chemical treatments on
tendon cellularity and mechanical properties, J Biomed Mater
Res 49:134–140, 2000.
35. Tischer T, Vogt S, Aryee S, et al: Tissue engineering of the
anterior cruciate ligament: a new method using acellularized
tendon allografts and autologous fibroblasts, Arch Orthop
Trauma Surg 127:735–741, 2007.
36. Adams JE, Zobitz ME, Reach JS, et al: Rotator cuff repair using
an acellular dermal matrix graft: an in vivo study in a canine
model, Arthroscopy 22:700–709, 2006.
37. Abousleiman RI, Reyes Y, McFetridge P, et al: Tendon tissue
engineering using cell-seeded umbilical veins cultured in a
mechanical stimulator, Tissue Eng Part A 15:787–795, 2009.
38. Molloy T, Wang Y, Murrell G: The roles of growth factors in
tendon and ligament healing, Sports Med 33:381–394, 2003.
39. Yang G, Crawford RC, Wang JH: Proliferation and collagen
production of human patellar tendon fibroblasts in response
to cyclic uniaxial stretching in serum-free conditions,
J Biomech 37:1543–1550, 2004.
40. Riboh J, Chong AKS, Pham H, et al: Optimization of flexor
tendon tissue engineering with a cyclic strain bioreactor,
J Hand Surg (Am) 33:1388–1396, 2008.
41. Wong J, Alyouha S, Kadler K, et al: The cell biology of suturing
tendons, Matrix Biology 29:525–536, 2010.
42. Rohrich RJ, Trott SA, Love M, et al: Mersilene suture as a
vehicle for delivery of growth factors in tendon repair, Plast
Reconstr Surg 104:1713–1717, 1999.
43. Hamada Y, Katoh S, Hibino N, et al: Effects of monofilament
nylon coated with basic fibroblast growth factor on endoge­
nous intrasynovial flexor tendon healing, J Hand Surg (Am)
31:530–540, 2006.
44. Dines JS, Weber L, Razzano P, et al: The effect of growth dif­
ferentiation factor-5-coated sutures on tendon repair in a rat
model, J Shoulder Elbow Surg 16(Suppl):S215–S221, 2007.
45. Yao J, Korotkova T, Smith RL: Viability and proliferation of
pluripotential cells delivered to tendon repair sites using bio­
active sutures-an in vitro study, J Hand Surg (Am) 36:252–258,
2011.

79
SECTION
2
PRIMARY FLEXOR
TENDON SURGERY

This page intentionally left blank

81
CHAPTER
8 
INDICATIONS FOR PRIMARY
FLEXOR TENDON REPAIR
Jin Bo Tang, MD
Kleinert and his colleagues on primary flexor tendon
repairs followed by early mobilization established that
the lacerated digital flexor tendon can be treated by
direct end-to-end repairs when wound conditions are
favorable.1-3
ANATOMICAL DIVISIONS
By virtue of their anatomical features, the flexor tendons
in the hand and forearm are divided into five zones,
which delineates the fundamental nomenclature for
flexor tendon anatomy and surgical repairs.1,3,6 In the
1990s, the most complex areas—the flexor tendons
within the digital sheath—were subdivided by Moiemen
and Elliot7 and by Tang.8 The zoning is described in
Box 8-1, and its relation to the locations of pulleys is
shown in Figures 8-1 and 8-2.
ETIOLOGIES AND EVALUATION
OF TENDON INJURIES
Tendons can be injured through open wounds, caused
by sharp cuts or machine injuries, as closed ruptures
after fractures, through other bony problems, or even
spontaneously, without history of injury or clear etiol­
ogy. Severe forms of tendon injury can present as part
of compound injuries due to major trauma to the
extremities. Surgically repaired tendons may disrupt
during functional exercise.
Tendon injuries often present as open injuries and
are associated with a variety of open wounds. Open
trauma with deep, narrow traumatic laceration should
raise suspicion of concomitant tendon injuries. Such a
diagnosis can be missed if surgeons do not carefully
evaluate active motion of the potentially involved digits.
When a tendon is partially cut, the fingers may still func­
tion well in many cases, easily causing such cuts to be
neglected. Some larger partial cuts of the tendon may
be detected through triggering during finger flexion.
Large open traumas deep enough to reach the bony
structures are usually accompanied by tendon injuries,
which are not difficult to diagnose. Open trauma to the
palm or wrist areas is frequently accompanied by tendon
injuries. Trauma to the wrist area may injure only a part
OUTLINE
Injuries to the flexor tendons occur in the digits, palm,
wrist, or distal and mid-forearm. Repairs of flexor
tendons in the digital sheath area are technically
demanding. When wound conditions are favorable,
primary tendon repair is preferable in terms of ease.
However, it is important that primary repairs be per­
formed by experienced surgeons. Alternatively, a well-
accepted strategy is to postpone the repair for a few
days, or even 1 to 2 weeks, to the delayed primary
stage, for an experienced surgeon to become available.
Repairs during either the primary or delayed primary
stage produce comparable clinical outcomes.
Primary and delayed primary flexor tendon repairs
are indicated in clean-cut tendon injuries with limited
damage to the peritendinous tissues or a wound that
can be converted to a clean-cut wound. Neurovascular
injury is not a contraindication to primary repairs.
Loss of soft tissue coverage over the tendon or the
presence of fractures is a borderline contraindication.
Local defects in skin and subcutaneous tissues can be
covered by flap transfer. A simple fracture limited to
the phalangeal or metacarpal shaft can be securely
fixed with screws or mini-plates and the tendons can
be repaired.
Serious crush injuries, severe wound contami­
nation, extensive loss of soft tissues, or extensive
destruction of pulleys and tendon structures are con­
traindications to primary tendon repairs.
Injuries in the flexor tendons can occur in the digits,
palm, wrist, or distal and mid-forearm. Tendon repairs
in the digital sheath area are most technically demand­
ing and controversial. The advent of primary flexor
tendon repairs within the synovial sheath region should
be credited to pioneers of nearly half a century ago,
including Verdan1 and Kleinert and colleagues.2,3 Prior
to that time, over most of the previous half century,
primary tendon repair was not advocated and surgeons
were accustomed to removing the tendons entirely and
grafting in new tendon.4,5 The reports of Verdan and of

82
Section 2:  Primary Flexor Tendon Surgery
of the wrist flexors, without interfering in normal finger
and wrist flexion. However, injuries to the majority of
the wrist flexors cause weakness of wrist flexion and/or
render active finger flexion impossible.
The levels of division of the tendons often do not lie
immediately deep to the finger skin lacerations. The
location of the distal tendon ends depends on the posi­
tion of the fingers at the time of injury. If the finger
is lacerated in flexion, the distal end of the tendon is
drawn distally as the finger is extended. If the finger is
lacerated in extension, the distal tendon end is usually
found at the site of the laceration. The proximal tendon
end often retracts to into the palm because of the pull
of the muscles. When the laceration is in the distal
half of the fingers, the vincula to the tendons may
prevent the proximal tendon from retracting. Therefore,
the proximal end of the lacerated flexor digitorum
superficialis (FDS) or flexor digitorum profundus (FDP)
tendons sometimes retracts over only a short distance
and is still found within the digital sheath. The proximal
end of the lacerated flexor pollicis longus (FPL) tendon
usually retracts to the thenar area, or even more proxi­
mally, due to the vigorous pull of the muscle.
During examination, any changes in resting position
of the fingers should be closely observed (Figure 8-3).
In assessment of function of the FDP and FDS tendons,
Figure 8-1  Zoning of the digital flexor tendons.
1
1
2
2
3
3
4
5
Figure 8-2  Pulleys and subdivisions of the digital flexor
tendons (zones 1 and 2).
Zone 2
Zone 1
2D
2C
2B
2A 1C 1B 1A
A1
A2
C1
A3 C2 A4 C3 A5
Palmar
aponeurosis
Vertical septa
Transverse metacarpal ligament
patients are asked to actively flex the fingers. Inability to
actively flex the finger proximal interphalangeal (PIP)
or distal interphalangeal (DIP) joints using the specific
tests shown in Figures 8-4 and 8-5, when passive motion
is complete, indicates that the FDP or FDS tendon (or
both) is completely severed.
In making decisions about treatment, surgeons
should consider the condition of the hand thoroughly.
Tendon injuries often present as a part of compound
injuries with open wounds. Nerve and vascular injuries
Box 8-1 
Zones of Finger Flexor Tendons (Verdan)
1
From the insertion of the FDS tendon the terminal
insertion of the FDP tendon
2
From the proximal reflection of the digital synovial
sheath to the FDS insertion
3
From the transverse carpal ligament to the digital
synovial sheath
4
Area covered by the transverse carpal ligament
5
Proximal to the transverse carpal liagment
Zones of Thumb FPL Tendon
1
Distal to the IP joint
2
From the IP joint to the A1 pulley
3
The area of the thenar eminence
Zone 1 in Fingers (Moiemen and Elliot)
1A
The very distal FDP tendon (usually <1 cm)
1B
From zone 1A to distal margin of the A4 pulley
1C
The FDP tendon within the A4 pulley
Zone 2 in Fingers (Tang)
2A
The area of the FDS tendon insertion
2B
From the proximal margin of the FDS insertion to
the distal margin of the A2 pulley
2C
The area covered by the A2 pulley
2D
From the proximal margin of the A2 pulley to the
proximal reflection of digital sheath
Nomenclature for Flexor Tendon
Anatomy in the Hand and Forearm

83
Chapter 8:  Indications for Primary Flexor Tendon Repair
are the most commonly associated injuries and soft
tissue defects may also be present. Injury to a single
artery in the digit does not need repair, but it is essential
to repair the lacerated digital nerve even when the con­
tralateral one is intact.
Wound contamination can be a serious problem. All
wounds should be thoroughly washed and débrided. If
the wounds have clear signs of contamination, thor­
ough washing of the wounds, débridement, and intra­
venous infusion of antibiotics are necessary. The tendons
should be repaired during the delayed primary stage,
after inflammation has been controlled. For such
patients, intravenous administration of antibiotics is
indicated and status of tetanus immunization should
be evaluated. If tendon injuries are accompanied by a
fracture, these can be treated on the same day or sepa­
rately on different days. The fracture can be treated first
by internal fixation, and the tendons can be repaired
after 1 week, or even later.
Closed rupture of the tendons presents as the inabil­
ity to actively flex the digits without an open wound.
Forced extension during active flexion of the finger can
cause avulsion injury of the FDP tendon at its insertion
to the distal phalanx, which is called “Jersey” finger.
Tendon rupture from chronic attrition may occur in
rheumatoid diseases, Kienböck disease, scaphoid non­
union, hamate fracture, or distal radius fracture. Tendon
rupture may also occur spontaneously without clear eti­
ology. Closed rupture of the pulleys usually presents in
special populations engaged in forceful flexion of the
fingers, such as rock climbers. Clinically, patients can
actively flex the fingers but present signs of tendon bow­
stringing of the involved fingers.
Radiographic examination should be routine for
patients with large wounds and suspected of having
compound injuries. Ultrasound or computed tomogra­
phy should be used for patients suspected of having a
closed tendon injury or closed pulley disruption.
TIMING OF TENDON REPAIR
Primary tendon repair is end-to-end repair performed
immediately after wound cleaning and débridement,
usually within several to 24 hours after trauma. Delayed
primary repair is defined as end-to-end repair performed
days to weeks after tendon injuries (usually within 3, or
even 4, weeks after tendon lacerations), after immediate
closure of the wounds without tendon repairs. When­
ever possible, acutely lacerated flexor tendons in the
hand and forearm should be treated primarily or during
the delayed primary stage. The ideal situation is that a
patient with digital flexor tendon lacerations is brought
into the clinic soon after injury, an experienced surgeon
is readily available, and surgery begins within a few
hours (Figure 8-6). In particular, a tendon injured in
critical areas (such as zone 2) should not be repaired by
an inexperienced surgeon. In such cases, tendon repair
Figure 8-3  Change of the resting position of the digits
indicates complete laceration (or disruption) of the digital
flexor tendon. The picture showing a case of complete cut of
the FDP tendon 2 weeks after primary skin closure without
tendon repairs.
Figure 8-4  Active flexion of the DIP joint indicating
integrity of the FDP tendon. Loss of active flexion of the DIP
joint indicates FDP tendon injury.
Figure 8-5  Normal function of the FDS tendon in the
middle finger is shown. Loss of active flexion of the PIP joint
while the flexion of the other fingers is blocked indicates loss
of FDS function, hence complete severance of the tendon.

84
Section 2:  Primary Flexor Tendon Surgery
wound, with limited damage to peritendinous tissues.
Simultaneous injuries to the nerves and arteries are
common and are not a contraindication for primary
repairs.
“Clean-cut” wounds, the simplest clinical situation
associated with digital tendon lacerations, are a prime
indication for primary flexor tendon repair.6,10 A “clean-
cut” wound is one in which the structures are injured
through a clean and tidy cut, usually a single transverse
or oblique laceration in the digits, palm, or distal
forearm, often produced by a knife or a piece of glass.9
The cut should also be “clean” in terms of contamina­
tion and infection. Anatomically, the tendon(s) is “only”
severed, and without tissue defect. The cut tissues may
even align well (Figure 8-7). This is the best indication
for primary repair, with the greatest likelihood of rela­
tively uncomplicated repair, rehabilitation and satisfac­
tory outcome. Such wounds are often accompanied by
divisions of the digital neurovascular structures.
Crush injuries to a very limited segment of the fingers,
palm, or wrist produce untidy skin and subcutaneous
injuries and tendon wounds (Figures 8-8 and 8-9).
Such wounds are also good candidates for primary
repairs, because the soft tissue wounds and tendons can
be converted to those associated with a clean-cut wound
through débridement of nonviable tissues and direct
wound closure. However, these injuries have a greater
potential for contamination. Primary tendon surgery is
possible, although more difficult than with a truly clean-
cut wound. Phalangeal fractures are rarely associated
with a clean-cut flexor tendon laceration but can be part
of a crush injury. A simple and stable fracture in the
phalangeal shaft can be securely fixed internally, thus
presenting no contraindication to primary tendon
repair.
should be delayed until an experienced surgeon is avail­
able (see Figure 8-6). Inadvertent damage to the tendon
structures by inexperienced operators may introduce a
second “injury” to the tendon system, which may render
the tendons difficult to repair at subsequent operations,
even in the hands of an experienced surgeon. So far, no
clinical data have shown that delayed primary surgery
within a few days to 1, or 2, weeks after injury, per­
formed by an experienced surgeon, are inferior to
primary surgery in terms of outcomes. In deciding
between primary and delayed primary repairs, I advise
to always place the expertise of the operators before the
exact timing of the repairs.
No clinical investigations have actually validated the
best time for primary or delayed primary repair. My
preferred period of deliberate delay is 4 to 7 days, when
the risk of infection can be properly addressed and
edema has reduced substantially.9,10 Delay of the repair
beyond 3 to 4 weeks may cause myostatic shortening of
the muscle–tendon unit; for these late cases, lengthen­
ing the tendon within the muscles in the forearm can
ease the tension.11
Rupture of the repaired flexor tendons after surgery
can be re-repaired if the rupture occurs within a few
weeks up to a month after surgery; secondary tendon
grafts may be the only choice for ruptured cases in the
presence of obvious retraction of the tendon end(s),
extensive scarring when both tendons ruptures or exten­
sive scarring around the intact FDS tendon when the
FDP tendon alone ruptures.12, 13
INDICATIONS
Indications for primary or delayed primary end-to-end
tendon repairs are clean-cut tendon injuries or tendon
injuries in a wound that can be converted to a clean-cut
Figure 8-6  Flow-chat of the decision-making process for primary or delayed primary tendon repairs in zone 2 flexor tendon
repair.
Primary repair
Delayed primary repair
Secondary repair
Do:
Within a few hours or one
day after injury
Experienced surgeons are
available
Not do:
Experienced surgeons are
not available
Do:
A few days or within 2 to 3
weeks after injury ideally
After edema subsided and
infection controlled
Not do:
Over 4 weeks after injury
generally
Do:
3 to 6 months later
Tendon grafting
Staged reconstruction
Pulley reconstruction
Not do:
Passive joint motion
not ample
Skin conditions
not good
Risk of infection exists
• Primary and delayed repairs have comparable outcomes.
• Avoid primary repairs by inexperienced surgeons.
• Primary wound closure followed by tendon repair in later
days by experienced surgeons is a standard practice.
When:
Lengthy loss of tendon substance, extensive pulley
destruction, serious wound contamination, or
unstable fractures or joint destruction.
Go
to

85
Chapter 8:  Indications for Primary Flexor Tendon Repair
Severity of tendon injuries ranges from isolated
partial to complete divisions of either the FDP or FDS
tendon, to complete severance of both tendons, to com­
plicated tendon injuries in multiple fingers involving
soft tissue or bone injuries. Although isolated partial
FDS tendon injuries usually do not need surgical repair,
complete FDP tendon laceration in the distal half of the
fingers do. A partial FDP tendon cut through over 80%
of its diameter is also an indication for surgical repair
and is treated similarly to a complete tendon cut.
Tendon lacerations through 50% to 80% of the tendon
diameter require surgical repair to lessen the chance of
tendon triggering during finger motion and to increase
strength. A partial cut of less than 50% of the tendon
diameter may not need surgical suture, or the cut sites
can simply be trimmed to smooth the tendon surface.
Complete lacerations of both flexor tendons in one, or
multiple, digits are typical indications for primary repair,
and tendon injuries accompanied by other tissue
damage constitute some of the borderline indications
discussed later.
BORDERLINE INDICATIONS
Borderline indications to primary or delayed primary
repairs have been less thoroughly addressed. Loss of soft
tissue coverage over the tendon or the presence of frac­
tures is a borderline indication. Five clinical situations
represent borderline indications to tendon repair.
1. Localized soft tissue injuries: crush, or compression,
injuries on the palmar aspect of the fingers some­
times lead to localized soft tissue defects. The
underlying flexor tendons may present with a
short traumatic defect, or such a defect may arise
after débridement of nonviable, ragged tendon
tissue. This situation includes no contraindica­
tions to primary repair if the soft tissue defect is
less than one-third the length of the fingers and
the tendon loss is less than 1.5 to 2 cm. In my
experience, the FDP tolerates shortening by up to
Figure 8-7  Clean-cut digital flexor tendon laceration and primary repair. The tendon was tidily severed without loss of
tendon substance. Primary tendon and nerve repairs are indicated. (Courtesy of Department of Hand Surgery, Nantong
University.)
Figure 8-8  Injuries in the distal palm area, resulting in
untidy wounds, lacerations of flexor tendons to multiple
fingers, and injuries to neurovascular bundles. Primary
tendon and nerve repairs are indicated.
Figure 8-9  Injuries to the distal forearm and wrist area,
producing lacerations of multiple wrist and digital flexor
tendons and the median nerve. Primary tendon and nerve
repairs are indicated.

86
Section 2:  Primary Flexor Tendon Surgery
tendon, even if the tendon has already been
trimmed by a similar amount at the initial surgery.
In my experience, the FDP tolerates shortening by
up to 1.5 to 2 cm. The ruptured FDS tendon
should be removed. Re-repair of both tendons is
impractical, and shortening of the FDS, particu­
larly within zone 2, is mechanically disadvanta­
geous. The digital sheath system is usually less
elastic, narrow, and inclined to collapse after
primary repair rupture. Rupture of a repair within
one month after the initial repair is always worth
an attempt at re-repair. However, by one month
after primary repair, re-repair is rarely indicated, as
ruptured tendons are likely to be surrounded by
adhesions and their healing potential is limited,
particularly if repaired under increased tension.
4. Delayed repairs: All estimates of the “best timing”
for primary flexor tendon repair suggested so far
have been empirical. I do not have a rigid “best”
time frame in mind, as previous suggestions
regarding the timing of primary repair are not con­
sistent and may not be imperative. My clinical
impression is that treatment outcomes after delay
for such a short period are almost identical to
those associated with primary repair promptly
after the trauma. Upon re-opening of the wound,
the cut tendon ends still appear fresh and no col­
lapse or fibrosis of the sheath is seen. The tendons
can be treated as if they were freshly cut. However,
when the surgery is postponed further beyond that
period, the tendon ends may be rounded, with
varying degrees of adhesions present, and the elas­
ticity of the sheath is likely to be reduced, making
repair more difficult. Although a delay of more
than 1 month would rule out direct end-to-end
repair surgery, the report of McFarlane and col­
leagues showed that direct end-to-end repair after
a delay for 1 to several months can been possible
without undue tension.14 For these late cases, the
tendon may be lengthened within the muscles in
the forearm to ease the tension on the proximal
tendon end.11 Of note is one situation in which
repair delayed more than 1 month is still feasible
(i.e., a wound around the PIP joint level in which
the FDP tendon has been cut but the long vincula
connecting to the proximal cut tendon has not
been severed). In this instance, retraction of the
FDP tendon is limited. When the wound is opened,
the retracted proximal end is found locally within
the sheath and the tendon can be repaired with
relative ease.
5. Massive soft tissue damage: Generally, this is a con­
traindication to primary flexor tendon surgery,
and I discourage primary repairs under this
circumstance. However, such situations can be
broken down further into: (1) extensive soft tissue
1.5 to 2 cm. The soft tissue can be repaired with
a local, or free, flap transfer and the tendon is
repaired by direct end-to-end suture. Flap transfer
provides fresh and vascularized tissue coverage,
roughly similar to the original subcutaneous
tissue. However, a tendon with a defect length
close to 2 cm is hard to draw together. Direct
end-to-end suture of the tendon should be accom­
panied by a procedure to reduce the tension on
the tendon when the surgeon is less experienced
with dealing with tendons with a defect. In such
cases, intramuscular tendon lengthening through
a forearm incision may release the tension.11
Lengthening of the tendons at the wrist level may
be also needed for this degree of tendon loss.
When the tendon is repaired directly, care should
be taken to prevent repair rupture when starting
early active tendon motion. Early active motion
and its progress to full active digital flexion should
be delayed.
2. Injuries including a simple and stable fracture: A
simple, stable fracture in the phalanx or metacar­
pal is not a contraindication to primary tendon
surgery. Fractures in the shafts of more than one
metacarpal bone may sometimes accompany a cut
digital flexor tendon. These injuries do not pre­
clude primary tendon surgery, providing the frac­
tures are simple, limited to the shaft and do not
involve the joints. Internal fixations in the palm
with mini-plates, screws, or K-wires usually ensure
a stable reduction, but early postoperative exercise
may have to be less aggressive. What are seen more
frequently, however, are tendon injuries associated
with fractures involving joints in more than one
phalanx, with crush, or abrasion, of the overlying
soft tissues. These skeletal injuries are contraindi­
cations to primary tendon repair, because fractures
involving joints tend to be unstable, the soft tissue
wounds are always contaminated and early post­
operative tendon mobilization is either difficult or
not feasible. However, some surgeons still perform
tendon repair and mobilize the hand after surgery
under these conditions.
3. Rupture of tendon repairs: Ruptures of primarily
repaired flexor tendons have been noted in
most case series incorporating early active finger
mobilization. I approach the ruptured tendon
repair as I would a primary tendon repair. At least
half of the segment, if not the entire segment,
encompassed by the original sutures should be
trimmed off, because the ends are softened and
ragged, which decreases the holding power of the
subsequent re-repair. The length of tendon seg­
ments that I trim off is about 0.8 to 1.0 cm (0.5 cm
or less on either end). This amount of shortening
is of no biomechanical consequence to the FDP