Biomechanical Comparison of Canine Fascia Lata and Thoracolumbar Fascia: An In Vitro Evaluation of Replacement Tissues for Body Wall Reconstruction Elisabeth R. Henderson1, BVSc, Ed J. Friend1, BVetMed, CertSAS, Diplomate ECVS, Michael J. Toscano1, Bsc, Msc, PhD, Kevin J. Parsons2, BVSc, PhD, CertSAS, Diplomate ECVS, and John F. Tarlton1, Bsc, PhD 1
Department of Clinical Veterinary Science, University of Bristol, Bristol, United Kingdom and Hospital, University of Bristol, Bristol, United Kingdom
Corresponding Author Ed J. Friend, BVetMed, CertSAS, Diplomate ECVS, School of Veterinary Sciences, Langford House, Langford, North Somerset BS40 5DU, United Kingdom. E‐mail: [email protected] Submitted June 2013 Accepted March 2014 DOI:10.1111/j.1532-950X.2014.12247.x
Langford Veterinary Services, Small Animal
Objectives: To compare the suitability of thoracolumbar fascia (TLF) and fascia lata (FL) for body wall defect repair in dogs, by examining their biomechanical properties and useable surface area. Study Design: Experimental. Animals: Dogs (n ¼ 8). Methods: Fresh TLF and FL grafts were obtained, surface area was calculated before testing to failure in 2 different modes: tensile testing and resistance to suture pullout, in 2 perpendicular orientations. Results: Useable TLF surface area was significantly greater than for FL. Maximum load, energy to break, and elastic modulus of FL was significantly greater than that of TLF in tensile testing, but no apparent difference in the ultimate stress or strain was identified. There was no overall difference in suture pullout load between TLF and FL. During tensile testing, tissue orientation had a significant influence on ultimate load, stress, and elastic modulus for both tissue types, with strain and energy to break only having significant effects for TLF and FL, respectively. Conclusions: The greater tensile strength and stiffness of FL compared to TLF was not reflected in its material properties, implying any difference was a consequence of greater thickness. Suture pullout was not significantly different between the 2 tissues, perhaps limiting the clinical significance of the tissue mechanics. Tissues were anisotropic with respect to mechanical properties, thus orientation may be an important factor.
Large body wall defects that result from surgical resection, severe trauma, infection, herniation, radiation necrosis, and congenital deformities remain challenging for reconstruction. The aim of closure is to achieve functional integrity with a tensionless repair. Full thickness defects of the body wall must ideally be reconstructed in a single‐stage procedure with the objectives to fill the defect, reduce the dead space, to provide stability, and to protect the exposed vital viscera. For thoracic wall defects, reconstruction should establish an airtight seal and be rigid enough to prevent the functional and cosmetic effects of a flail segment. In animals, various reconstructive techniques including direct closure, omental pedicle flaps1; local skin flaps, pedicled or free musculocutaneous flaps2–6; porcine small intestinal submucosa tissue grafts7–9; fascia lata (FL) tissue grafts10–14; prosthetic mesh1–4,6,15,16; and diaphragmatic advancement2,4,6,17 have been described. The type of repair used Presented in part at the 2013 BSAVA Congress, Birmingham, UK, April 2013.
126
2
depends on the size and location of the defect and presence of infection. Local or regional pedicle muscle flaps are ideal for reconstruction of chest and abdominal wall defects; however, in some cases the local tissues may have been damaged by previous operations or radiation therapy, thus the tissue may be of inadequate size to cover the defect or the tissues may have already been used. When adequate local musculofascial tissue is not present for tension‐free closure it becomes necessary to use another method of closure. Advantages of the use of autologous tissue alone when reconstructing either the thoracic or abdominal wall include the reduction in the risk of implant associated complications such as foreign body reactions, postoperative seromas, wound infections, and soft tissue erosion. Wound infection rates reported in studies of prosthetic reconstruction of chest wall defects in dogs range from 0% to 6.7%.2–4,6,15 Polypropylene mesh has been associated with persistent postimplantation wound infection and a complication rate of 66.7% when used to reconstruct the chest wall after tumor resection. The use of
Veterinary Surgery 44 (2015) 126–134 © Copyright 2014 by The American College of Veterinary Surgeons
Henderson et al.
Biomechanical Comparison of Canine Fascia Lata and Thoracolumbar Fascia
synthetic mesh is contraindicated in cases of gross contamination, infection, repeat surgeries with exposed mesh, and immunosuppression.3 A direct comparison of autologous thoracolumbar fascia (TLF) to expanded polytetrafluoroethylene in rabbits in the presence of contamination suggested that viable fascia autografts are more resistant to bacterial contamination than polytetrafluoroethylene.13 Distant autologous grafts can include the use of the FL. Though there are many reports supporting clinical use of FL as a replacement material for body wall defects in people,13,18–23 there is little veterinary literature describing its clinical use. Chest wall reconstruction using bilateral autologous FL grafts in conjunction with a single rib graft after removal of a recurrent desmoid tumor has been reported.21 FL has also been used in conjunction with its muscle; the tensor fasciae lata and overlying skin (musculofasciocutaneous free flap) for reconstruction of chest and abdominal wall defects.24 Reconstruction was required after removal of several different types of tumor or because of abdominal herniation and has been used in several sites including the thoracic wall, abdominal wall, and head and neck regions.24 In a study using autologous FL for perineal hernia repair in dogs, the graft site was anchored within the surrounding soft tissues by organized mature collagenous bands, revascularized and viable.10 However, FL grafts exist in a limited and exhaustible supply and tissue harvest can potentially result in donor‐site morbidities such as pain, seroma, and delayed ambulation.10 Fresh frozen allogeneic canine FL is available, which removes the risk of complications associated with tissue harvest. Arnold et al.25 compared fresh frozen allogeneic FL to 3 other reconstructive materials and concluded that it was a strong fascial donor tissue suitable for reconstructing abdominal and thoracic wall defects. The largest FL sheet available for purchase in this study was “at least 5 cm 5 cm” (Veterinary Transplant Services, Inc., Kent, WA). In our experience, the small size of autologous FL grafts limits it usefulness when reconstructing large body wall defects. TLF originates from the supraspinous ligament and courses over the muscles of the vertebral column between spinous and transverse processes. In the lumbar region it is divided into 2 superimposed leaves (deep and superficial) and gives origin to the caudal dorsal serratus, external abdominal oblique, internal abdominal oblique, and transverse abdominal oblique muscles.26 Our purpose was to compare the biomechanical properties and surface area of TLF and FL tissue grafts obtained from the same cadaveric dogs. Our hypotheses based on subjective gross assessments were that the biomechanical properties of TLF would be similar to that of FL, that TLF and FL would behave isotropically and that the surface area obtainable for TLF would be greater than that of FL.
MATERIALS AND METHODS Grafts were surgically removed from both sides of 8 mixed‐ breed canine cadavers (5 females, 3 males; mean weight, 10.31 kg [range, 6.635–12.005 kg]) of unknown age. Each
cadaver had 2 FL and 2 TLF collected, 1 of each from each side of the dog. TLF Collection Dogs were positioned in lateral recumbency and after 3 skin incisions, subcutaneous tissue was dissected to expose the TLF and associated musculature (Fig 1). The proposed margins for TLF harvest were (1) cranially at the level of the paralumbar fossa/caudal aspect of rib 13; (2) caudally, just cranial to the tuber coxa in alignment with the aponeurosis of the sartorius muscle; (3) dorsally, a parasagittal incision adjacent to the dorsal processes of the lumbar vertebrae; and (4) ventrally, dorsal to the aponeurosis of the TLF with the external and internal abdominal oblique muscles. The aponeurosis of the TLF with the abdominal muscles can be seen as a thinning of the TLF. However, at this location the fibers of the TLF begin to diverge and therefore the ventral incision should be made dorsal to this to ensure uniformity of the graft. Incisions were made to the depth of the overlying fascia of the epaxial and external abdominal oblique muscles (Fig 2). We found it most effective to undermine the tissue in a ventrodorsal direction. The fascia graft was then removed with Metzenbaum scissors. FL Collection A skin incision was made over the craniolateral aspect of the quadriceps musculature. Then the subcutaneous tissue was bluntly dissected to reveal the superficial layer of the FL. Margins for FL dissection were: (1) cranially, the sartorius muscle; (2) caudally, the biceps femoris muscle; (3) proximally, the tensor FL muscle; and (4) distally, the distal third of the femur. Both superficial and deep layers of the fascial graft were then incised and removed using Metzenbaum scissors (Fig 3).10
Figure 1 Photograph showing the margins for the dissection of thoracolumbar fascia. The cranial margin is at the level of the paralumbar fossa/caudal aspect of rib 13; the caudal margin, just cranial to the tuber coxa, in alignment with aponeurosis of sartorius muscle; the dorsal margin with the dorsal spinous processes of lumbar vertebrae and supraspinous ligament, and the ventral margin with the aponeurosis of the internal abdominal oblique/external abdominal oblique muscles.
Veterinary Surgery 44 (2015) 126–134 © Copyright 2014 by The American College of Veterinary Surgeons
127
Biomechanical Comparison of Canine Fascia Lata and Thoracolumbar Fascia
Henderson et al.
Biomechanical Testing
Figure 2 Photograph showing the elevation of the thoracolumbar fascia, which reveals the underlying fascia of the epaxial muscles and external abdominal oblique muscle.
Surface Area Calculation The outline of each fresh tissue specimen was traced onto paper. A known surface area of paper was weighed using a scale, accurate to 0.001 g. The traced paper area for each tissue specimen was subsequently weighed and the surface area calculated. Fresh tissues were then wrapped in saline (0.9% NaCl) solution soaked swabs and frozen until biomechanical testing.
Specimens of each tissue material were tested to failure in 2 different modes: tensile testing and resistance to suture pullout. Biomechanical testing was performed using a computer controlled electromechanical materials testing machine (Instron1 6022 Model, 6000 series, Instron European Headquarters, High Wycombe, UK) which generated load–displacement curves, from which maximum load (N), maximum displacement (mm), energy to break (J), and elastic modulus (MPa) were recorded. Before testing, tissue specimens were precut (1 cm 3 cm). Tissue thickness was calculated as an average of 5 measurements taken using digital calipers, accurate to 1 mm. Tissue thickness was measured after specimens were thawed to room temperature (naturally, without the aid of heating apparatus). Care was taken to avoid compression of the tissue with the pressure of the calipers.
Tensile Strength Eighteen (1 cm 3 cm) specimens of canine FL and TLF were tested in 2 orientations perpendicular to each other. For FL, this was cranial–caudal (FLCr–Cd) and proximal–distal (FLProx–Dist) and for TLF was cranial–caudal (TLFCr–Cd) and dorsal–ventral (TLFD–V). Specimens were adhered to sandpaper (Medium, 501884, PK25, Wickes, Northamptonshire, England) templates using cyanoacrylate adhesive (Loctite, Henkel Ltd., Hempstead, Hertfordshire, UK), which ensured that fascial layers did not slide over each other, or out of the clamp when the tissue was loaded. TLF is a multilayered structure within the lumbar region, divided into several superimposed leaves; each layer was adhered to the sandpaper, however, layers were not adhered to each other. Constructs were positioned in pneumatic clamps for loading (Fig 4).27 To account for variation in set up tension between specimens; displacement was set to 0 when 0.5 N of load was applied. Increasing load was uniaxially applied at a strain rate of 25 mm/min and maximal load was defined as the load at which the tissue failed.
Suture Pullout
Figure 3 Photograph showing the margins for dissection of fascia lata. The cranial margin is the sartorius muscle; the caudal margin, the biceps femoris muscle; the proximal margin, the tensor fascia lata muscle; and the distal margin, the distal third of the femur.
128
Eighteen (1 cm 3 cm) specimens of FL and TLF were tested in 2 orientations as described above. A single horizontal mattress suture of 2–0 polypropylene was placed through all layers of the tissue specimen at equal distance from tissue margins (3 mm) and secured with a square knot (polypropylene using a reverse cutting 26 mm, 3/8 circle, swaged on needle, Ethicon, Somerville, NJ). The specimen was then loaded onto the Instron tensiometer (Fig 5) and the pneumatic clamps positioned so that slack was removed before tissue distraction and displacement was set to 0. Increasing load was uniaxially applied at 25 mm/min and maximal load was defined as the load applied at which the fascia tore (tissue failure) or the suture pulled through the tissue.
Veterinary Surgery 44 (2015) 126–134 © Copyright 2014 by The American College of Veterinary Surgeons
Henderson et al.
Biomechanical Comparison of Canine Fascia Lata and Thoracolumbar Fascia
Statistical Analysis
Figure 4 Tensile testing. The pneumatic clamp connected to the load frame (top) was distracted from the stationary clamp at the base until failure.
Data were analyzed with MlwiN,28 a statistical package designed for hierarchical data, where each response was modeled with a structure of specimen (i) within dog (j). For analysis of the tensile testing data, dependent variables (load, stress, strain, and energy, all at structural failure, as well as elastic modulus) were modeled against the independent variables (dog mass, gender, side of dog from which specimen was taken, tissue type, and tissue orientation). Similarly for analysis of the suture pullout data, dependent variables (load, displacement, and energy, all at structural failure, as well as elastic modulus) were modeled against the independent variables (dog mass, gender, side of dog from which specimen was taken, tissue type, tissue orientation, and tissue thickness). All independent variables were initially included in the model and then removed individually by backwards stepwise regression when comparison of the respective Z‐ratio with a standard normal distribution was greater than 1.96 (P > 0.05). For each response (e.g. energy to break), individual model predictors (e.g. tissue type and tissue orientation), standard errors, and associated P values are provided. In our analysis, the model serves as a mathematical representation of the data linking model predictors (e.g. dog mass) and the response (e.g. energy at structural failure). The b0 serves as an overall reference value from which the estimates of the model predictors are compared.
RESULTS Tensile Strength Maximum Load (N). Maximum load, independent of tissue orientation was significantly greater for FL (91.7 56.1 N) than for TLF (50.2 N 24.0 N; P < 0.01) (Table 1; Fig 6). For both FL and TLF load to failure demonstrated a significant relationship with orientation. The maximum load for TLFD–V (71.0 13.7 N) was significantly greater than TLFCr–Cd (29.4 9.7 N; P < 0.0001). The maximum load for FLProx–Dist (128.2 55.1 N) was significantly greater than FLCr–Cd (55.3 24.2 N; P < 0.001). Stress (MPa). Stress was calculated using the maximum load value and each individual tissue specimen’s cross‐sectional dimensions (N/mm2). Stress, independent of tissue orientation was not significantly different when comparing FL and TLF; however, there was a significant relationship in ultimate stress with orientation. Stress for TLFD–V (25.5 5.8 MPa) was significantly greater than for TLFCr–Cd (10.3 2.9 MPa; P < 0.0001). The stress for FLProx–Dist (27.6 12.5 MPa) was significantly greater than for FLCr–Cd (9.8 3.4 MPa; P < 0.0001). Figure 5 Suture pullout testing. A single horizontal mattress suture of 2–0 polypropylene was placed through all layers of tissue at equal distance from tissue margins (3 mm) and secured with a square knot. The suture was attached by a hook to the cross head of the Instron1 machine (top).
Strain (%). When placing tissue into the construct, a consistent length of tissue (i.e. tissue length between the clamps) was subjected to loading each time (10 mm). Strain was expressed as displacement as a percentage of the original
Veterinary Surgery 44 (2015) 126–134 © Copyright 2014 by The American College of Veterinary Surgeons
129
Biomechanical Comparison of Canine Fascia Lata and Thoracolumbar Fascia
Henderson et al.
Table 1 Summary Data for Tensile Strength Testing: Reported Outcome Variable Measurements for Both Thoracolumbar Fascia and Fascia Lata, Both Independent and Dependent on Tissue Orientation
Thoracolumbar fascia Cranial–caudal Dorsal–ventral Orientation P‐value Fascia lata Cranial–caudal Proximal–distal Orientation P‐value Tissue P‐value
Stress (MPa)
Maximum Load (N)
Strain %
Energy to Break (J)
Elastic Modulus (MPa)
17.9 8.9 10.3 2.9 25.5 5.8 .05 1.3 0.5 1.1 0.4 1.6 0.6 .05
P < 0.0001) (Table 3). The surface area of useable material defined as, the tissue material that could practically be used for body wall reconstruction (e.g. after removal of thin tailed ends of both fascia), was significantly greater for TLF (49.38 14.14 cm2) than for FL (19.38 4.92 cm2; P < 0.0001). Model Predictors The elastic modulus obtained in the suture pullout test, dependent on tissue orientation, resulted in a negative correlation between both TLF and FL with the weight of the dog. Energy to break, dependent of tissue orientation, resulted in a negative correlation between FL and the weight of the dog during tensile testing. Elastic modulus, dependent on tissue orientation, resulted in higher values if the graft was taken from the right hand side of the dog versus the left hand side during tensile testing. The opposite resulted for maximum load during suture pullout.
Figure 7
Specimen Thickness FL mean thickness (450 150 mm) was significantly greater than mean TLF thickness (346 95 mm; P < 0.01).
DISCUSSION We found that in dogs, the biomechanical properties of TLF are dissimilar to those of FL, that the tissues behave anisotropically, and that the surface area obtainable for TLF is greater than that of FL. During tensile testing, the maximum load of FL was significantly greater than that of TLF; however, there was no significant difference in the material properties of ultimate stress values between FL and TLF. This suggests that it is the thickness rather than the composition of FL that results in its increased strength compared with TLF.
Representative load–displacement curves for resistance to suture pullout testing.
Veterinary Surgery 44 (2015) 126–134 © Copyright 2014 by The American College of Veterinary Surgeons
131
Biomechanical Comparison of Canine Fascia Lata and Thoracolumbar Fascia
Table 3 Mean Values for Surface Area Obtained and Surface Area of Useable Material
Thoracolumbar fascia Fascia lata P‐value
Surface Area Obtained (cm2)
Surface Area of Useable Material (cm2)
55.04 16.51 24.09 5.03
Department of Clinical Veterinary Science, University of Bristol, Bristol, United Kingdom and Hospital, University of Bristol, Bristol, United Kingdom
Corresponding Author Ed J. Friend, BVetMed, CertSAS, Diplomate ECVS, School of Veterinary Sciences, Langford House, Langford, North Somerset BS40 5DU, United Kingdom. E‐mail: [email protected] Submitted June 2013 Accepted March 2014 DOI:10.1111/j.1532-950X.2014.12247.x
Langford Veterinary Services, Small Animal
Objectives: To compare the suitability of thoracolumbar fascia (TLF) and fascia lata (FL) for body wall defect repair in dogs, by examining their biomechanical properties and useable surface area. Study Design: Experimental. Animals: Dogs (n ¼ 8). Methods: Fresh TLF and FL grafts were obtained, surface area was calculated before testing to failure in 2 different modes: tensile testing and resistance to suture pullout, in 2 perpendicular orientations. Results: Useable TLF surface area was significantly greater than for FL. Maximum load, energy to break, and elastic modulus of FL was significantly greater than that of TLF in tensile testing, but no apparent difference in the ultimate stress or strain was identified. There was no overall difference in suture pullout load between TLF and FL. During tensile testing, tissue orientation had a significant influence on ultimate load, stress, and elastic modulus for both tissue types, with strain and energy to break only having significant effects for TLF and FL, respectively. Conclusions: The greater tensile strength and stiffness of FL compared to TLF was not reflected in its material properties, implying any difference was a consequence of greater thickness. Suture pullout was not significantly different between the 2 tissues, perhaps limiting the clinical significance of the tissue mechanics. Tissues were anisotropic with respect to mechanical properties, thus orientation may be an important factor.
Large body wall defects that result from surgical resection, severe trauma, infection, herniation, radiation necrosis, and congenital deformities remain challenging for reconstruction. The aim of closure is to achieve functional integrity with a tensionless repair. Full thickness defects of the body wall must ideally be reconstructed in a single‐stage procedure with the objectives to fill the defect, reduce the dead space, to provide stability, and to protect the exposed vital viscera. For thoracic wall defects, reconstruction should establish an airtight seal and be rigid enough to prevent the functional and cosmetic effects of a flail segment. In animals, various reconstructive techniques including direct closure, omental pedicle flaps1; local skin flaps, pedicled or free musculocutaneous flaps2–6; porcine small intestinal submucosa tissue grafts7–9; fascia lata (FL) tissue grafts10–14; prosthetic mesh1–4,6,15,16; and diaphragmatic advancement2,4,6,17 have been described. The type of repair used Presented in part at the 2013 BSAVA Congress, Birmingham, UK, April 2013.
126
2
depends on the size and location of the defect and presence of infection. Local or regional pedicle muscle flaps are ideal for reconstruction of chest and abdominal wall defects; however, in some cases the local tissues may have been damaged by previous operations or radiation therapy, thus the tissue may be of inadequate size to cover the defect or the tissues may have already been used. When adequate local musculofascial tissue is not present for tension‐free closure it becomes necessary to use another method of closure. Advantages of the use of autologous tissue alone when reconstructing either the thoracic or abdominal wall include the reduction in the risk of implant associated complications such as foreign body reactions, postoperative seromas, wound infections, and soft tissue erosion. Wound infection rates reported in studies of prosthetic reconstruction of chest wall defects in dogs range from 0% to 6.7%.2–4,6,15 Polypropylene mesh has been associated with persistent postimplantation wound infection and a complication rate of 66.7% when used to reconstruct the chest wall after tumor resection. The use of
Veterinary Surgery 44 (2015) 126–134 © Copyright 2014 by The American College of Veterinary Surgeons
Henderson et al.
Biomechanical Comparison of Canine Fascia Lata and Thoracolumbar Fascia
synthetic mesh is contraindicated in cases of gross contamination, infection, repeat surgeries with exposed mesh, and immunosuppression.3 A direct comparison of autologous thoracolumbar fascia (TLF) to expanded polytetrafluoroethylene in rabbits in the presence of contamination suggested that viable fascia autografts are more resistant to bacterial contamination than polytetrafluoroethylene.13 Distant autologous grafts can include the use of the FL. Though there are many reports supporting clinical use of FL as a replacement material for body wall defects in people,13,18–23 there is little veterinary literature describing its clinical use. Chest wall reconstruction using bilateral autologous FL grafts in conjunction with a single rib graft after removal of a recurrent desmoid tumor has been reported.21 FL has also been used in conjunction with its muscle; the tensor fasciae lata and overlying skin (musculofasciocutaneous free flap) for reconstruction of chest and abdominal wall defects.24 Reconstruction was required after removal of several different types of tumor or because of abdominal herniation and has been used in several sites including the thoracic wall, abdominal wall, and head and neck regions.24 In a study using autologous FL for perineal hernia repair in dogs, the graft site was anchored within the surrounding soft tissues by organized mature collagenous bands, revascularized and viable.10 However, FL grafts exist in a limited and exhaustible supply and tissue harvest can potentially result in donor‐site morbidities such as pain, seroma, and delayed ambulation.10 Fresh frozen allogeneic canine FL is available, which removes the risk of complications associated with tissue harvest. Arnold et al.25 compared fresh frozen allogeneic FL to 3 other reconstructive materials and concluded that it was a strong fascial donor tissue suitable for reconstructing abdominal and thoracic wall defects. The largest FL sheet available for purchase in this study was “at least 5 cm 5 cm” (Veterinary Transplant Services, Inc., Kent, WA). In our experience, the small size of autologous FL grafts limits it usefulness when reconstructing large body wall defects. TLF originates from the supraspinous ligament and courses over the muscles of the vertebral column between spinous and transverse processes. In the lumbar region it is divided into 2 superimposed leaves (deep and superficial) and gives origin to the caudal dorsal serratus, external abdominal oblique, internal abdominal oblique, and transverse abdominal oblique muscles.26 Our purpose was to compare the biomechanical properties and surface area of TLF and FL tissue grafts obtained from the same cadaveric dogs. Our hypotheses based on subjective gross assessments were that the biomechanical properties of TLF would be similar to that of FL, that TLF and FL would behave isotropically and that the surface area obtainable for TLF would be greater than that of FL.
MATERIALS AND METHODS Grafts were surgically removed from both sides of 8 mixed‐ breed canine cadavers (5 females, 3 males; mean weight, 10.31 kg [range, 6.635–12.005 kg]) of unknown age. Each
cadaver had 2 FL and 2 TLF collected, 1 of each from each side of the dog. TLF Collection Dogs were positioned in lateral recumbency and after 3 skin incisions, subcutaneous tissue was dissected to expose the TLF and associated musculature (Fig 1). The proposed margins for TLF harvest were (1) cranially at the level of the paralumbar fossa/caudal aspect of rib 13; (2) caudally, just cranial to the tuber coxa in alignment with the aponeurosis of the sartorius muscle; (3) dorsally, a parasagittal incision adjacent to the dorsal processes of the lumbar vertebrae; and (4) ventrally, dorsal to the aponeurosis of the TLF with the external and internal abdominal oblique muscles. The aponeurosis of the TLF with the abdominal muscles can be seen as a thinning of the TLF. However, at this location the fibers of the TLF begin to diverge and therefore the ventral incision should be made dorsal to this to ensure uniformity of the graft. Incisions were made to the depth of the overlying fascia of the epaxial and external abdominal oblique muscles (Fig 2). We found it most effective to undermine the tissue in a ventrodorsal direction. The fascia graft was then removed with Metzenbaum scissors. FL Collection A skin incision was made over the craniolateral aspect of the quadriceps musculature. Then the subcutaneous tissue was bluntly dissected to reveal the superficial layer of the FL. Margins for FL dissection were: (1) cranially, the sartorius muscle; (2) caudally, the biceps femoris muscle; (3) proximally, the tensor FL muscle; and (4) distally, the distal third of the femur. Both superficial and deep layers of the fascial graft were then incised and removed using Metzenbaum scissors (Fig 3).10
Figure 1 Photograph showing the margins for the dissection of thoracolumbar fascia. The cranial margin is at the level of the paralumbar fossa/caudal aspect of rib 13; the caudal margin, just cranial to the tuber coxa, in alignment with aponeurosis of sartorius muscle; the dorsal margin with the dorsal spinous processes of lumbar vertebrae and supraspinous ligament, and the ventral margin with the aponeurosis of the internal abdominal oblique/external abdominal oblique muscles.
Veterinary Surgery 44 (2015) 126–134 © Copyright 2014 by The American College of Veterinary Surgeons
127
Biomechanical Comparison of Canine Fascia Lata and Thoracolumbar Fascia
Henderson et al.
Biomechanical Testing
Figure 2 Photograph showing the elevation of the thoracolumbar fascia, which reveals the underlying fascia of the epaxial muscles and external abdominal oblique muscle.
Surface Area Calculation The outline of each fresh tissue specimen was traced onto paper. A known surface area of paper was weighed using a scale, accurate to 0.001 g. The traced paper area for each tissue specimen was subsequently weighed and the surface area calculated. Fresh tissues were then wrapped in saline (0.9% NaCl) solution soaked swabs and frozen until biomechanical testing.
Specimens of each tissue material were tested to failure in 2 different modes: tensile testing and resistance to suture pullout. Biomechanical testing was performed using a computer controlled electromechanical materials testing machine (Instron1 6022 Model, 6000 series, Instron European Headquarters, High Wycombe, UK) which generated load–displacement curves, from which maximum load (N), maximum displacement (mm), energy to break (J), and elastic modulus (MPa) were recorded. Before testing, tissue specimens were precut (1 cm 3 cm). Tissue thickness was calculated as an average of 5 measurements taken using digital calipers, accurate to 1 mm. Tissue thickness was measured after specimens were thawed to room temperature (naturally, without the aid of heating apparatus). Care was taken to avoid compression of the tissue with the pressure of the calipers.
Tensile Strength Eighteen (1 cm 3 cm) specimens of canine FL and TLF were tested in 2 orientations perpendicular to each other. For FL, this was cranial–caudal (FLCr–Cd) and proximal–distal (FLProx–Dist) and for TLF was cranial–caudal (TLFCr–Cd) and dorsal–ventral (TLFD–V). Specimens were adhered to sandpaper (Medium, 501884, PK25, Wickes, Northamptonshire, England) templates using cyanoacrylate adhesive (Loctite, Henkel Ltd., Hempstead, Hertfordshire, UK), which ensured that fascial layers did not slide over each other, or out of the clamp when the tissue was loaded. TLF is a multilayered structure within the lumbar region, divided into several superimposed leaves; each layer was adhered to the sandpaper, however, layers were not adhered to each other. Constructs were positioned in pneumatic clamps for loading (Fig 4).27 To account for variation in set up tension between specimens; displacement was set to 0 when 0.5 N of load was applied. Increasing load was uniaxially applied at a strain rate of 25 mm/min and maximal load was defined as the load at which the tissue failed.
Suture Pullout
Figure 3 Photograph showing the margins for dissection of fascia lata. The cranial margin is the sartorius muscle; the caudal margin, the biceps femoris muscle; the proximal margin, the tensor fascia lata muscle; and the distal margin, the distal third of the femur.
128
Eighteen (1 cm 3 cm) specimens of FL and TLF were tested in 2 orientations as described above. A single horizontal mattress suture of 2–0 polypropylene was placed through all layers of the tissue specimen at equal distance from tissue margins (3 mm) and secured with a square knot (polypropylene using a reverse cutting 26 mm, 3/8 circle, swaged on needle, Ethicon, Somerville, NJ). The specimen was then loaded onto the Instron tensiometer (Fig 5) and the pneumatic clamps positioned so that slack was removed before tissue distraction and displacement was set to 0. Increasing load was uniaxially applied at 25 mm/min and maximal load was defined as the load applied at which the fascia tore (tissue failure) or the suture pulled through the tissue.
Veterinary Surgery 44 (2015) 126–134 © Copyright 2014 by The American College of Veterinary Surgeons
Henderson et al.
Biomechanical Comparison of Canine Fascia Lata and Thoracolumbar Fascia
Statistical Analysis
Figure 4 Tensile testing. The pneumatic clamp connected to the load frame (top) was distracted from the stationary clamp at the base until failure.
Data were analyzed with MlwiN,28 a statistical package designed for hierarchical data, where each response was modeled with a structure of specimen (i) within dog (j). For analysis of the tensile testing data, dependent variables (load, stress, strain, and energy, all at structural failure, as well as elastic modulus) were modeled against the independent variables (dog mass, gender, side of dog from which specimen was taken, tissue type, and tissue orientation). Similarly for analysis of the suture pullout data, dependent variables (load, displacement, and energy, all at structural failure, as well as elastic modulus) were modeled against the independent variables (dog mass, gender, side of dog from which specimen was taken, tissue type, tissue orientation, and tissue thickness). All independent variables were initially included in the model and then removed individually by backwards stepwise regression when comparison of the respective Z‐ratio with a standard normal distribution was greater than 1.96 (P > 0.05). For each response (e.g. energy to break), individual model predictors (e.g. tissue type and tissue orientation), standard errors, and associated P values are provided. In our analysis, the model serves as a mathematical representation of the data linking model predictors (e.g. dog mass) and the response (e.g. energy at structural failure). The b0 serves as an overall reference value from which the estimates of the model predictors are compared.
RESULTS Tensile Strength Maximum Load (N). Maximum load, independent of tissue orientation was significantly greater for FL (91.7 56.1 N) than for TLF (50.2 N 24.0 N; P < 0.01) (Table 1; Fig 6). For both FL and TLF load to failure demonstrated a significant relationship with orientation. The maximum load for TLFD–V (71.0 13.7 N) was significantly greater than TLFCr–Cd (29.4 9.7 N; P < 0.0001). The maximum load for FLProx–Dist (128.2 55.1 N) was significantly greater than FLCr–Cd (55.3 24.2 N; P < 0.001). Stress (MPa). Stress was calculated using the maximum load value and each individual tissue specimen’s cross‐sectional dimensions (N/mm2). Stress, independent of tissue orientation was not significantly different when comparing FL and TLF; however, there was a significant relationship in ultimate stress with orientation. Stress for TLFD–V (25.5 5.8 MPa) was significantly greater than for TLFCr–Cd (10.3 2.9 MPa; P < 0.0001). The stress for FLProx–Dist (27.6 12.5 MPa) was significantly greater than for FLCr–Cd (9.8 3.4 MPa; P < 0.0001). Figure 5 Suture pullout testing. A single horizontal mattress suture of 2–0 polypropylene was placed through all layers of tissue at equal distance from tissue margins (3 mm) and secured with a square knot. The suture was attached by a hook to the cross head of the Instron1 machine (top).
Strain (%). When placing tissue into the construct, a consistent length of tissue (i.e. tissue length between the clamps) was subjected to loading each time (10 mm). Strain was expressed as displacement as a percentage of the original
Veterinary Surgery 44 (2015) 126–134 © Copyright 2014 by The American College of Veterinary Surgeons
129
Biomechanical Comparison of Canine Fascia Lata and Thoracolumbar Fascia
Henderson et al.
Table 1 Summary Data for Tensile Strength Testing: Reported Outcome Variable Measurements for Both Thoracolumbar Fascia and Fascia Lata, Both Independent and Dependent on Tissue Orientation
Thoracolumbar fascia Cranial–caudal Dorsal–ventral Orientation P‐value Fascia lata Cranial–caudal Proximal–distal Orientation P‐value Tissue P‐value
Stress (MPa)
Maximum Load (N)
Strain %
Energy to Break (J)
Elastic Modulus (MPa)
17.9 8.9 10.3 2.9 25.5 5.8 .05 1.3 0.5 1.1 0.4 1.6 0.6 .05
P < 0.0001) (Table 3). The surface area of useable material defined as, the tissue material that could practically be used for body wall reconstruction (e.g. after removal of thin tailed ends of both fascia), was significantly greater for TLF (49.38 14.14 cm2) than for FL (19.38 4.92 cm2; P < 0.0001). Model Predictors The elastic modulus obtained in the suture pullout test, dependent on tissue orientation, resulted in a negative correlation between both TLF and FL with the weight of the dog. Energy to break, dependent of tissue orientation, resulted in a negative correlation between FL and the weight of the dog during tensile testing. Elastic modulus, dependent on tissue orientation, resulted in higher values if the graft was taken from the right hand side of the dog versus the left hand side during tensile testing. The opposite resulted for maximum load during suture pullout.
Figure 7
Specimen Thickness FL mean thickness (450 150 mm) was significantly greater than mean TLF thickness (346 95 mm; P < 0.01).
DISCUSSION We found that in dogs, the biomechanical properties of TLF are dissimilar to those of FL, that the tissues behave anisotropically, and that the surface area obtainable for TLF is greater than that of FL. During tensile testing, the maximum load of FL was significantly greater than that of TLF; however, there was no significant difference in the material properties of ultimate stress values between FL and TLF. This suggests that it is the thickness rather than the composition of FL that results in its increased strength compared with TLF.
Representative load–displacement curves for resistance to suture pullout testing.
Veterinary Surgery 44 (2015) 126–134 © Copyright 2014 by The American College of Veterinary Surgeons
131
Biomechanical Comparison of Canine Fascia Lata and Thoracolumbar Fascia
Table 3 Mean Values for Surface Area Obtained and Surface Area of Useable Material
Thoracolumbar fascia Fascia lata P‐value
Surface Area Obtained (cm2)
Surface Area of Useable Material (cm2)
55.04 16.51 24.09 5.03
