Keywords
rotator cuff - pressure - suture - suture techniques - tendon injuries - tendons
Introduction
The performance of arthroscopic rotator cuff repairs has been increasing constantly
in recent times.[1] Short and long-term clinical and functional results are good to excellent in most
cases;[2]
[3]
[4]
[5] however, rerupture rates are still considerable, ranging from 11% to 68% in selected
series, even reaching 94% in some studies.[6]
[7]
[8]
The surgery for rotator cuff repair seeks to establish a fibrovascular interface between
the tendon and the footprint, which is required for the healing and restoration of
the fibrocartilaginous insertion (enthesis); to do so, the construct must maximize
the pressurized contact between tendon and bone while maintaining the mechanical resistance
against a physiological load.[9]
[10]
Several anatomical factors favor healing, including a good construct tension, proper
tissue perfusion, reduced micromotion at the tendon-footprint interface, and adequate
footprint pressure and contact area.[11] The underlying principle is that a greater tendon-to-bone contact area, both in
terms of magnitude and distribution, will increase tendon healing.[12]
The double row (DR) repair increases resistance to load-related failure, improves
contact areas and pressures, and decreases gap formation at the tendon-footprint interface
compared to the single row (SR) repair.[5]
[13] However, anchors provide low resistance, in addition to being prone to loosen in
osteoporotic bone; as such, they result in poor contact at the level of the supraspinatus
tendon footprint, and may cause greater tuberosity osteolysis, making revision a challenge
and increasing costs.[14]
[15]
[16]
[17]
[18]
The transosseous (TO) technique enables the maximization of the contact area at the
tendon-footprint interface[19] and reduce movement at the tendon-bone interface.[20] In addition to this mechanical aspect, the TO technique enables the accurate preparation
of bone side of the lesion without risks or complications, including anchor removal
and/or greater tuberosity osteolysis.[21]
[22]
Transosseous suture techniques are efficient and reproducible for the arthroscopic
repair of rotator cuff tears.[23] Moreover, there is a greater potential for healing due to direct contact between
tendon and bone (without intervening anchor material) and mesenchymal stem cells from
the bone tunnels at the proximal humerus.[24]
[25]
[26]
[27]
Due to these advantages, our team designed a device capable of performing oblique
TO tunnels, enabling repairs with simple or crossed sutures.
The present study aims to compare the pressure and contact area at the tendon-footprint
interface in repairs performed with simple or crossed TO sutures. Our hypothesis is
that the crossed configuration will result in a larger contact area and a lower pressure
drop after cyclic loading.
Materials and Methods
Animal Model
A total of 12 fresh frozen shoulders from 6-month-old lambs (Ovis orientalis aries) were obtained from a local company (oyster cut, Frigorífico Simunovic Ltda., Punta
Arenas, Región de Magallanes y Antártica Chilena, Chile). They were thawed at room
temperature the night before the biomechanical tests. The infraspinatus tendon was
selected because its anatomical and functional features are equivalent to those of
the human supraspinatus tendon.[28] The specimens were dissected in a standardized way, removing all the soft tissue
adjacent to the humeral shaft, and the subscapular and the supraspinatus fossae of
the scapula to isolate the infraspinatus muscle and its tendon. No specimen had any
rotator cuff abnormalities. Next, a scapular osteotomy was performed at the level
of the spine, sparing the infraspinatus muscle attachment, to enable muscle manipulation
without tearing it apart ([Figure 1]). Lastly, 2 perforations were made 1 cm from the medial edge of the scapula, separated
by 1 cm on each side of the scapular spine, with a 5.0-mm drill bit, to enable the
osteotomized fragment to be hooked to the load cell ([Figures 2] and [3]). Specimens were irrigated intermittently with 0.9% NaCl solution throughout each
test to prevent sample dehydration.
Fig. 1 Anatomical dissection of the infraspinatus tendon of a lamb specimen. Standardized
anatomical dissection, removing all the soft tissue adjacent to the humeral shaft
and the subscapular and supraspinatus fossae of the scapula to isolate the infraspinatus
muscle and its tendon. An arrow indicates the infraspinatus tendon, the delta (δ)
shows the infraspinatus muscle, and the asterisk marks the scapular osteotomy at level
of the spine, sparing the infraspinatus muscle attachment.
Fig. 2 Scapular foramina for the fixation of the infraspinatus muscle (anterior view). Two
perforations were made 1 cm from the medial edge of the scapula, separated by 1 cm
on each side of the scapular spine, with a 5.0-mm drill bit, to hook the osteotomized
fragment to the load cell (anterior view).
Fig. 3 Scapular foramina for the fixation of the infraspinatus muscle (superior view). Two
perforations were made 1 cm from the medial edge of the scapula, separated by 1 cm
on each side of the scapular spine, with a 5.0-mm drill bit, to hook the osteotomized
fragment to the load cell (superior view). The arrow shows how the muscle remains
undamaged, with no tears.
A tailored system generated cyclic tensions at the level of the infraspinatus muscle
and tendon ([Figure 4]). The model had three fundamental parts: a modular support with adjustable height,
an adjustable support for guidance of the suture system, and a load cell digitally
regulated using the Arduino (open source) software.
Fig. 4 Cyclic stress model. The model consisted of a height-adjustable modular support,
an adjustable support for the guidance of the suture system, and a load cell. The
humeral shaft was fixated in a PVC plastic cylinder with plaster.
The humeral shaft was fixated in a polyvinyl chloride (PVC) plastic cylinder with
plaster. Then, the modular support was adjusted to ensure the tendon was parallel
to the horizon (using a level), achieving a traction angle of 0° in abduction.
Rotator Cuff Tear
In each humeral head, the orientation of the greater tuberosity was identified and
demarcated with a 1.5-mm Kirschner needle. Next, the tip of the tuberosity was identified
and, 10 mm laterally to it, a full thickness, 20-mm wide tear was made with a #15
scalpel, releasing the entire tendon attachment from the footprint, and then flattening
it with a rasp to facilitate the placement of pressure sensors ([Figure 5]).
Fig. 5 Rotator cuff tear. The arrow marks the edge of the humeral cartilage, which was palpated
and marked with a needle. Next, 10 mm lateral to it, a full thickness, 20-mm wide
tear was made.
Measurement of Pressure and Contact Area at the Tendon-Footprint Interface
The contact area was measured at the beginning of the repair, while pressure was determined
at the beginning, during and at the end of cyclic loading.
First, at time zero, the contact area at the tendon-footprint interface was measured
(percentage and mean pressure [MPa]) using a set of colorimetric, pressure-sensitive
films (Prescale Ultra Low Pressure Fuji Photo Film, C. Itoh & Co, New York, NY, US)
covered with a plastic sheet to protect them from tissue moisture. These films were
positioned on the previously-flattened surface of the footprint. Subsequently, the
repair was carried out. The films were then digitized and analyzed with a previously-calibrated
scanner and software (Fujifilm Analysis System for Prescale, Tekscan, Inc., South
Boston, MA, US).
A digital pressure sensor (Flexiforce Sensor, Tekscan) was used to measure the pressure
at the tendon-footprint interface. The sensor was positioned between the tendon and
the footprint, and remained fixed by the repair; it records pressure changes over
time and stores them in a computer for later analysis (N). The baseline pressure was
recorded at the beginning of the experiment (time zero), during cyclic loading, and
at the end of the intervention.
Repair of Rotator Cuff Tear with Transosseous Sutures and Biomechanical Testing
The repairs were performed with a #2 polyester braided, non-absorbable polymer suture
with a long-chain polyethylene core (FiberWire; Arthrex, Naples, FL, US), the most
commonly-used size in arthroscopic shoulder surgery.
The TO tunnels were prepared with a device previously designed by our team and used
in previous models to generate oblique tunnels ([Figure 6]).
Fig. 6 Device for making transosseous oblique tunnels. Inset A shows the transosseous suture
device used. Inset B shows the proper positioning of the device in relation to the
greater tuberosity. Inset C shows a section of a artifical bone model and the trajectory
of the oblique tunnel.
Two different TO repairs were performed, always by the same surgeon (JC) to reduce
interoperator variability: 1) two TO tunnels with single knots (TOS) ([Figure 7A]) on six lamb shoulders; and 2) two TO tunnels with crossed knots (TOC) ([Figure 7B]). Both repairs were made with #2 FiberWire suture, using either the “Tennessee slider”
knot for the TOS repair or the “Revo-SCOI” knot for the TOC repair. No tensiometer
was used for knotting.
Fig. 7 Repair with transosseous sutures. Inset A shows the transosseous repair with single
knots, and inset B shows the repair with crossed knots.
The repair was pretensioned with 10 N for 2 minutes. The load cell was then programmed
to 1,400 cycles, with a frequency of 2.5 Hz and a load of 5 N. These parameters were
defined based on those used in similar studies,[29]
[30] and they reflect the initial postoperative rehabilitation period (two weeks) with
passive exercises and pendular movements.
Statistical Analysis
The results were presented as means and standard deviations. Given the small sample
with non-normal distribution (as demonstrated by the Shapiro-Wilk test), the statistical
test for the non-parametric variables (that is, the Mann-Whitney test) was used. All
data were analyzed using the STATA (StataCorp LLC, College Station, TX, US) software,
version 14. Significance was set at p < 0.05.
Results
The contact area at the tendon-footprint interface at time zero was 1.4-fold greater
in the TOC repair. Pressure distribution was of 50.9 ± 12.7% for the TOS, and of 72.2 ± 5.3%
for the TOC repair (p < 0.009) ([Figure 8]).
Fig. 8 Contact area and pressure at the tendon-footprint interface at time zero. The contact
area was measured with Fujifilm. The areas in red represent a higher contact pressure
and those in green represent a lower contact pressure. The table shows quantitative
results expressed as means and standard deviations. Abbreviation: NS, not significant.
Pressure at the tendon-footprint interface at time zero (measured with a pressure-sensitive
sheet) was 1.6-fold greater in the TOC repair. The mean pressure was of 0.68 ± 0.13 MPa
in the TOS, and of 1.1 ± 0.2 MPa in the TOC repair (p < 0.007) ([Figure 8]).
Regarding pressure at the tendon-footprint interface in response to cyclic loading
(measured with a digital pressure sensor), both repair models presented a self-reinforcing
mechanism during increased cyclic stress ([Figure 9]).
Fig. 9 Example of pressure measurement under cyclic loading, demonstrating the self-reinforcement
mechanism in transosseous repair.
For the TOS repair, the pressure was of 5.3 ± 5.3 N at baseline, and of 3.8 ± 4.6 N
at the end of the intervention, with a 51.7 ± 38.0% variation after 1,400 tension
cycles. For the TOC repair, the pressure was of 10.7 ± 1.8 N at baseline, and of 12.9 ± 8.7 N
at the end of intervention, with a 114.9 ± 65.9% variation (p < 0.044; p < 0.022; and p < 0.017 respectively) ([Table 1]).
Table 1
|
Repair type
|
|
Simple transosseous repair
|
Crossed transosseous repair
|
|
Parameter (N)
|
Mean
|
SD
|
Mean
|
SD
|
|
Baseline pressure
|
5.30
|
5.30
|
10.71
|
1.78
|
|
Pressure at 25%
|
4.91
|
5.59
|
10.88
|
4.95
|
|
Peak at 25%
|
7.45
|
7.68
|
16.36
|
6.48
|
|
Pressure at 50%
|
5.18
|
5.63
|
13.12
|
8.09
|
|
Peak at 50%
|
7.34
|
6.96
|
17.71
|
7.36
|
|
Pressure at 75%
|
4.63
|
5.01
|
11.98
|
6.19
|
|
Peak at 75%
|
7.25
|
6.54
|
16.52
|
5.80
|
|
Final pressure
|
3.84
|
4.56
|
12.90
|
8.73
|
|
Final peak
|
6.42
|
6.34
|
17.57
|
8.69
|
|
Δ pressure
|
-1.46
|
1.85
|
2.19
|
7.49
|
|
Variation
|
51.71%
|
38.00%
|
114.85%
|
65.94%
|
Discussion
The main finding of the present study was that the TOC repair results in greater pressure
at the level of the tendon-bone interface at time zero, lower loss of contact force
due to cyclic loading over time, and a better distribution of force at the footprint
when compared to the TOS repair.
This is an important finding because, when using arthroscopic instruments to perform
TO sutures, the TOC technique would improve pressure and the distribution of force
by only modifying a surgical gesture. The present is the first report of such biomechanical
advantage.
These findings were expected, since the TOC repair distributes the pressure in an
area that had not been previously loaded; this is directly related to the self-reinforcement
mechanism described by Burkhart et al.[27] in 2009, in which an increased stress applied to the construct amplifies resistance
to structural failure by progressively increasing compression forces at the tendon
footprint. The compressive forces created at the footprint increase the frictional
resistance between tendon and bone, thus reducing the formation of gaps between them.[27] Wedging of the angle between the suture material and the bone is formed as the tendon
is progressively stressed; in addition, at the coronal plane, suture geometry changes
from rectangular to trapezoidal as the tensile load increases.[27] This results in an elastic deformation of the tendon, creating a compression force
perpendicular to the bone surface, which increases according to the tensile load[27] ([Figure 10]).
The high standard deviation values for baseline and final pressure levels both in
the TOS and TOC repair is probably associated with knot tension because, in the TOC
repair, the pressure at the footprint level is directly related to the tension delivered
by the knot, which was not quantitatively measured. Although all procedures were performed
by the same surgeon, there is a risk of internal variability. Clinically, the tension
delivered in the anchor repair with and without knots is a constant challenge in arthroscopic
rotator cuff repair surgery. The intraoperative use of blood pressure monitors could
improve the reproducibility of these techniques.
Our findings are comparable to those of biomechanical studies[5]
[13] with DR repairs which have shown an increase in resistance to load-induced failure,
improved contact areas and pressure, and a decrease in gap formation at the tendon-footprint
interface compared to SR repairs.[5]
[13]
Ng et al.[31] used infraspinatus tendons from a porcine model to compare the pressure distribution
in three DR configurations (suture bridge; two medial and one lateral anchors; and
one medial and two lateral anchors). These authors showed that this technique not
only results in a good footprint contact area (75%, 75%, and 73% respectively), but
that the use of a 3- or 4-anchor configuration produces a similar footprint contact
area in medium tears (no greater than 1.5 cm × 2.5 cm). These findings are consistent
with those of the present study, demonstrating at least one equivalence between the
TOC repair and the suture bridge at time zero regarding the pressure distribution
area at the tendon-bone interface. Apparently, this equivalence would only occur in
this configuration, since Caldow et al.[9] demonstrated the biomechanical inferiority of the TOS repair regarding contact area,
contact pressure, tensile strength, and stiffness compared to the Mason-Allen and
DR techniques.
Hinse et al.[32] compared the TO technique with sutures, TO with braided tape, and a TO-equivalent
(TOE) technique. Although the load at failure was not different between the braided-tape
TO and the TOE, the TO with sutures presented significantly less resistance compared
to the TOE, indicating that the type of material could be an important factor to consider.
In addition, even though significant differences were not detected, there was a trend
towards a greater loss of footprint coverage with pure TO techniques.
Park et al.[12] compared simple TO suture, SR suture, and SR with mattress suture. They demonstrated
that the TO tunnel rotator cuff repair technique generated significantly greater contact
and a greater overall pressure distribution over the defined footprint compared to
the remaining techniques. However, they did not compare TOC with TOE, which are the
most relevant techniques today. Tuoheti et al.[33] compared simple TO, and SR and DR sutures, and found out that DR was superior to
TO; however, it was a simple TO technique and a DR with mattress sutures, the same
weaknesses observed in Park et al.[12] study.
However, these studies only evaluate biomechanical properties regarding pressure magnitude
and distribution in addition to load to failure. Apparently, the TO technique would
have healing benefits in terms of the supply of the mesenchymal cells, and better
tendon vascularization.[24]
[25]
[26] Using ultrasound, Urita et al.[34] demonstrated that vascularization is superior in patients submitted to the TO arthroscopic
repair compared to the TOE repair.
A limitation of the present study is the evaluation of biomechanical aspects in an
animal model alone; therefore, the findings may be different in human beings and under
biological conditions (considering mesenchymal cells and irrigation). The use of human
cadaveric shoulders would have been better to represent these biomechanical features.
On the other hand, this model standardizes our results, because each sample is six
months old, which improves comparability. This is also true for bone mineral density,
which was not calculated for our samples, but would have been very similar since the
specimens had the same age.
Another important aspect to consider is the clinical relevance of our findings; even
though we have demonstrated significant differences in biomechanical factors, many
factors play a role in rotator cuff healing, so the clinical impact is unknown.
Conclusion
The TOC repair results in greater pressure at the tendon-bone interface, lower loss
of contact force under cyclic loading, and better force distribution at the footprint
when compared to the TOS repair.
Fig. 10 Schematic drawing of self-reinforcement in transosseous repair showing wedging of
the angle between the suture material and the bone as the tendon is progressively
stressed, and a change from rectangular to trapezoidal suture geometry in the coronal
plane as the tensile load increases.
