Research Reconstruction of a defect of the rotator cuff with ...

Research Reconstruction of a defect of the rotator cuff with polytetrafluoroethylene felt graft RECOVERY OF TENSILE STRENGTH AND HISTOCOMPATIBILITY IN AN ANIMAL MODEL A. Kimura, M. Aoki, S. Fukushima, S. Ishii, K. Yamakoshi From Sapporo Medical University and Kanazawa University, Japan

e reconstructed defects in the infraspinatus tendon using polytetrafluoroethylene (PTFE) felt grafts in 31 beagle dogs and examined the mechanical responses and histocompatibility. Except for one infected specimen, all the reconstructed infraspinatus tendons healed. We examined eight specimens each immediately after surgery and at six and 12 weeks. The ultimate tensile strength of the reconstructed tendons was 60.84 N, 172.88 N, and 306.51 N immediately after surgery and at six and 12 weeks, respectively. The stiffness of the specimens at the PTFE felt-bone interface was 9.61 kN/m, 64.67 kN/m, and 135.09 kN/m immediately after surgery and at six and 12 weeks, respectively. Six tendons were examined histologically at three, six, 12 and 24 weeks. Histological analysis showed that there was ingrowth of fibrous tissue between the PTFE fibres. Foreign-body reactions were found at the margin of the PTFE-bone interface between 12 and 24 weeks. The mechanical recovery and tissue affinity of PTFE felt to bone and to tendon support its use for reconstruction of the rotator cuff. The possible development of a foreign-body reaction should be borne in mind.

W

J Bone Joint Surg [Br] 2003;85-B:282-7. Received 17 September 2001; Accepted 9 May, 2002

A. Kimura, MD S. Fukushima, MD S. Ishii, MD, PhD Department of Orthopaedic Surgery Sapporo Medical University School of Medicine, South-1, West-16, Chuoku, Sapporo 060-8543, Japan. M. Aoki, MD, PhD Department of Physical Therapy Sapporo Medical University School of Health Sciences, South-3, West-17, Chuo-ku, Sapporo 060-8556, Japan. K. Yamakoshi, PhD Department of Human and Mechanical Systems Engineering, Faculty of Engineering, Kanazawa University. Correspondence should be sent to Dr M. Aoki. ©2003 British Editorial Society of Bone and Joint Surgery doi.10.1302/0301-620X.85B2.12823 $2.00 282

Fig. 1 Photograph showing the appearance of a PTFE felt graft 25 mm long, 7 mm wide, and 2.8 mm thick.

Non-absorbable synthetic materials have been used for many years in vascular surgery for the reconstruction of large vessels, the heart and abdominal wall. These materials include synthetic fibres such as Gore Tex, Marlex mesh, Dacron and polytetrafluoroethylene (PTFE; Teflon).1,2 In 1986, Ozaki et al3 described the use of PTFE felt for covering irreparable tears of the rotator cuff. Clinical use has been reported.3 We have been using PTFE felt graft for reconstructive surgery of irreparable defects of the rotator cuff and have obtained good clinical results. This graft was found to be very effective for the relief of pain. Kimura et al,4 however, have reported resorption of bone at the insertion of PTFE felt into the greater tuberosity in 30% of patients after two to five years. We have developed a model of PTFE felt grafting using the infraspinatus tendon of beagle dogs and have observed the biological responses of PTFE material in cancellous bone and tendon. Recovery of tensile strength and stiffness, the mode of failure and the histological response to grafted PTFE felt were examined to elucidate whether PTFE felt was applicable to reconstruction of the rotator cuff.

Materials and Methods PTFE felt graft. PTFE felt (Davol Inc, subsidiary of CR Bard Inc), is made from 100% virgin PTFE fluorocarbon fibre, and is acid-bleached to give it a bright white finish. The physical characteristics of the standard 007959 felt are as follows: water permeability 600 to 1700 ml/cm2/min; air permeability 20 to 120 FT3/mm; thickness 2.8 mm; and burst strength 211 psi minimum (Fig. 1). The cytotoxicity, THE JOURNAL OF BONE AND JOINT SURGERY

RECONSTRUCTION OF A DEFECT OF THE ROTATOR CUFF WITH POLYTETRAFLUOROETHYLENE FELT GRAFT

Fig. 2

Fig. 3a

283

Fig. 3b

Diagram of reconstruction of the infraspinatus tendon by PTFE felt. Ten millimetres of the distal infraspinatus tendon were removed. The proximal end of the PTFE felt overlapped the infraspinatus tendon by 15 mm and the distal end inserted 5 mm into a bone trough in the greater tuberosity (F, PTFE felt; SS, supraspinatus; IS, infraspinatus; T, bone trough)

Photographs of mechanical testing. The humerus was fixed in a holding jig and the intramuscular tendon was attached to a tendon clamp which was advanced along the direction of the fibres of the infraspinatus tendon. Figure 3a -- Before traction, the small arrows indicate line marks placed on the bone and PTFE felt. Figure 3b -- After traction, the PTFE felt was avulsed from bone (large arrow) (H, humerus; IS, infraspinatus; J, jig; T, transducer).

haemolysis and dissipation of ethylene oxide of PTFE felt were shown not to be significantly different from those of currently used Dacron (CR Bard Inc). The material has been approved by the Ministry of Health and Welfare in Japan as a medical tool (16000BZY01127000: artificial fibre sheet for tissue substitute). It was sterilised by ethylene oxide gas before implantation into animals. It does not have a Food and Drug Administration number because it was on the market in the USA before the enaction of the Medical Device Amendment in 1974. This material has also been widely used for vascular and abdominal surgery in Japan and has been approved by the Ministry of Health and Welfare as a medical tool. There has been no basic research on the tensile strength or biological responses of PTFE felt graft in tendon surgery, although clinical trials have been conducted.4,5 Animal model. We used 31 female beagle dogs aged between one and three years and with a mean body-weight of 10.5 kg. The animals were anaesthetised by intramuscular injection of ketamine hydrochloride (25 mg/kg) and anaesthesia was maintained by the intravenous injection of pentobarbital sodium (5 to 10 mg/kg/60 minutes). A skin incision 5 cm long was made over the left shoulder, and the infraspinatus tendon was exposed. After detachment of the tendon from the humerus, the width and thickness of each infraspinatus tendon, at 5 mm away from insertion, were measured with a digital caliper (Max-Cal; Foler, Newton, Massachusetts; accuracy, 0.05 mm) and the cross-sectional

area was calculated. The mean value of the cross-sectional areas was 9.02 ± 1.35 mm2 (SD) (n = 24). One centimetre of the infraspinatus tendon was removed at the insertion and the defect reconstructed using a PTFE felt graft 7 mm wide, 25 mm long and 2.8 mm thick (Fig. 1). The proximal end of the PTFE felt was overlapped 15 mm on the infraspinatus tendon and its distal end was inserted 5 mm into a bone trough in the greater tuberosity. Non-absorbable 2/0 sutures (Ethibond; Ethicon Inc, Somerville, New Jersey) were used for anchoring of the tendon and PTFE felt (Fig. 2). The four-strand Savage technique was used for the distal end of the PTFE felt repair and six interrupted sutures for the proximal end.6 The animals were treated according to the rules of the animal committee of our university and were allowed to move freely in a large cage after operation. For one to two weeks after operation, they avoided full weight-bearing on the left forelimb. Measurement of mechanical properties.6-8 Immediately after surgery and at six and 12 weeks, eight animals were killed using deep intravenous anaesthesia. The proximal humerus and entire infraspinatus muscle were removed. The head and proximal shaft of the humerus with the reconstructed tendons attached were fixed in a specially designed holding jig by two Kirschner wires. The wires and the shaft of the humerus were embedded in bone cement. The proximal tendinous part of the infraspinatus was stripped of surrounding muscle fibres and secured in a specially designed

VOL. 85-B, No. 2, MARCH 2003

284

A. KIMURA, M. AOKI, S. FUKUSHIMA, S. ISHII, K. YAMAKOSHI

400

Ultimate strength (Newton)

306.5 300 Fig. 4 200

172.9

100

Histogram showing temporal changes in ultimate strength after operation. The increase in ultimate strength was statistically significant (p < 0.001).

60.8

0 0

6

12

Weeks

200

135.09

Stiffness (Kn/m)

150

Fig. 5 100 Histogram showing temporal changes in stiffness after operation. The increase in stiffness was statistically significant (p < 0.001)

64.67 50 9.61 0 0

6

12

Weeks

tendon clamp during tensile testing. The clamp was advanced at a rate of 20 cm/min. The direction of distraction was aligned with the fibres of the infraspinatus tendon. To measure stiffness at the insertion of the tendon into bone, lines were painted on the surface of both the greater tuberosity and the distal end of the PTFE felt for measuring the distance between the lines. A video dimension analysing system, consisting of a computer-based dimension analyser and a charge-coupled device (CCD) camera, was used to determine the displacement between the lines. Before testing the tensile strength, traction of 4.9 N was applied ten times for preconditioning. The tensile force was measured with a load cell (LUB100KB; Kyowa Dengyo Co, Japan, capacity 100 kgf, nonlinearity, 0.03%/full scale) and recorded in Newtons. Both the displacement and force signals were entered on a personal computer (Macintosh IIci) via an analog-digital (A/D) converter (16-bit; sampling interval, 50 ms). A force-displacement curve was drawn. The slope of the linear portion of the force-displacement curve was determined in Newtons

per metre (N/stiffness) by least-squares linear regression analysis. The ultimate tensile strength at which the repair failed and the stiffness at the PTFE felt-bone interface were determined from the curve. Sites of failure in the specimens during tensile testing were also recorded and classified into three different parts: PTFE felt, PTFE-tendon interface, and PTFE-bone interface. The experiments were performed at a room temperature of 20°C and the specimens were moistened with a very fine mist of normal saline while clamped (Figs 3a and 3b). Data obtained in each postoperative group were analysed statistically by one-way ANOVA followed by Tukey’s test (i.e., post-hoc test). A p value of < 0.05 was considered to be statistically significant. Histological analysis. Two animals were sacrificed at three, six, and 12 weeks and one at 24 postoperative weeks. A specimen from one animal was discarded because of postoperative infection at three weeks. The specimens were fixed in 10% buffered formaldehyde solution. After decalcification in Decalcifying Solution A (Wako Pure Chemical THE JOURNAL OF BONE AND JOINT SURGERY


285

Table I. Site of rupture during tensile testing Site of rupture in the PTFE felt graft specimen Postoperative weeks 0 6 12 .

PTFE felt

Tendon-PTFE interface

Bone-PTFE interface

8 3 2

0 4 0

0 1 6

Industries Ltd, Tokyo, Japan; Plank Rychlo method) for seven days, longitudinal sections 5 mm in thickness were obtained from the site of repair and stained with haematoxylin and eosin. Routine light microscopy was used to determine cellular events, and views under polarised light were used to evaluate collagenous development and deterioration of PTFE fibres.9 The areas of histological observation were the PTFE-tendon interface, the central part of the PTFEbone interface (i.e. bottom of the trough) and the marginal part of the PTFE-cancellous bone interface (i.e. medial wall of the trough).

Fig. 6 Photomicrograph of a specimen of the PTFE-infraspinatus-tendon interface at three weeks. There was infiltration of granulation tissue and vessels between the PTFE fibres. The interface was bridged by loose connective tissue (haematoxylin and eosin x 60) (IS, infraspinatus tendon; F, PTFE felt; open arrows, a linear gap which was artificially made during sectioning).

Results Macroscopic observations. Except for the specimen with infection, all showed a secure connection of the graft complex which was covered with thick scar tissue with loose adhesion over the PTFE. This scar formation was most prominent at six weeks and subsided after 12 weeks. There was no sign of synovitis or degenerative change in the shoulders. Mechanical properties. The mean ultimate tensile strength of the specimens was 60.84 N immediately after surgery, 172.88 N at six weeks, and 306.51 N at 12 weeks. The increase in ultimate strength with time was statistically significant (p < 0.001) (Fig. 4). The mean stiffness of the specimens was 9.61 kN/m immediately after surgery, 64.67 kN/ m at six weeks, and 135.09 kN/m at 12 weeks. The increase in stiffness with time was statistically significant (p < 0.001) (Fig. 5). The sites of disruption of the specimens during tensile testing varied depending on the length of the postoperative period. Immediately after surgery disruption of the PTFE felt was observed in all eight specimens. At six weeks there was disruption of the PTFE felt in three specimens, at the PTFE-tendon interface in four and at the PTFE-bone interface in one. At 12 weeks disruption was seen within the PTFE felt in two specimens and at the PTFE-bone interface in six. There was no disruption at the PTFE-tendon interface at 12 weeks (Table I). Histological findings Three to six weeks. At the PTFE-tendon interface, there was infiltration of granulation tissue and vessels between the PTFE fibres. The interface was bridged by loose connective tissue (Fig. 6). At the PTFE-bone interface the distal end of the PTFE fibres was surrounded by trabecular bone with openings to the medullary canal. There was infiltration VOL. 85-B, No. 2, MARCH 2003

Fig. 7 Photomicrograph of a specimen of the PTFE-bone interface at six weeks. The distal end of the PTFE fibres was surrounded by trabecular bone with opening to the medullary canal. There was infiltration of granulation tissue and vessels between PTFE fibres and the interface was bridged by loose connective tissue (haematoxylin and eosin x 60) (F, PTFE felt; W, woven bone; M, medullary canal).

of granulation tissue and vessels between the PTFE fibres. The interface was bridged by loose connective tissue (Fig. 7). There were no wear particles of PTFE felt fibres at the marginal part in the bone trough. 12 and 24 weeks. At the PTFE-tendon interface, the interface was connected by aligned fibrous tissue. Infiltration of fibrous tissue widened the space between the PTFE fibres. The amount of granulation tissue and vessels was decreased (Fig. 8). At the central part of the PTFE-bone interface orientated fibrous tissue connecting the trabecular wall and the distal end of the PTFE developed within the graft. There were vessels in the interface which extended into the medullary canal. Some PTFE fibres had been incorporated into

286

A. KIMURA, M. AOKI, S. FUKUSHIMA, S. ISHII, K. YAMAKOSHI

Fig. 8

Fig. 9

Photomicrograph of a specimen of PTFE-tendon interface at 24 weeks. The interface was connected by aligned fibrous tissue with infiltration of fibrous connective tissue. The amount of granulation tissues and vessels had decreased (haematoxylin and eosin x 60) (F, PTFE felt; TLS, tendon-like scar).

Photomicrograph of the central part of the PTFE-bone interface of a specimen at 12 weeks. Orientated fibrous tissue connecting the trabecular wall and the distal end of the graft felt developed between the PTFE fibres. Vessels in the interface extended into the medullary canal (haematoxylin and eosin x 60) (W, woven bone; M, medullary canal; OFT, orientated fibrous tissue).

Fig. 10 Photomicrograph of the marginal part of the PTFE felt-bone interface of a specimen at 12 weeks. There was granulation tissue with macrophage-like cells containing wear particles of PTFE (small arrows). There were fractured fibres of PTFE felt (large arrows) (haematoxylin and eosin x 240).

woven bone (Fig. 9). At the marginal part of the PTFE-bone interface, there was granulation tissue with macrophage-like cells. Observation by polarised light showed that wear particles of PTFE fibres had been taken into the macrophage-like cells (Fig. 10). This finding was observed in specimens taken at both 12 and 24 weeks.

Discussion PTFE felt has both adequate thickness and flexibility because of its specially designed texture.3 There have been many studies on blood-flow dynamics and the biological response when PTFE has been used for vascular reconstruction,1,2 but there have been few on the mechanical response

when it has been used for reconstruction of tendons or ligaments. Our clinical results in 30 patients who received PTFE felt grafts for defects of the rotator cuff were acceptable.4 The mean total shoulder score (Japanese Orthopaedic Association)10 recovered from 57 points before operation to 82 points at follow-up of two to five years. Relief from pain was excellent (pain score 9.5 points before to 23.2 points after operation), but there was bone absorption at the PTFEbone interface in 30% of patients. There was no correlation between the clinical results and the occurrence of bone absorption.4 Ozaki et al5 reported similar long-term clinical and radiological results with PTFE felt. Our histological results showed that both the PTFEtendon interface and the PTFE-bone interface were filled with orientated fibrous tissue in the specimens at 12 and 24 weeks. This fibrous tissue seemed to align with mechanical forces transmitted through the reconstructed tendon. The fibrous tissue, however, which connected the PTFE felt to the infraspinatus tendon and cancellous bone was different from normal tendon fibres and was considered to be scar tissue. Foreign-body reactions were found in the marginal part of the PTFE-bone interface at 12 and 24 weeks. They seemed to occur in granulation tissue adjacent to the medullary canal where immunological responses were active. This phenomenon may explain the cause of the clinical finding of bone absorption at the PTFE-bone interface. The tensile strength of immediate repair of the infraspinatus tendon in beagle dogs has been reported to be 118.1 N.11 Therefore, the tensile strength of the PTFE felt graft immediately after surgery was smaller than that of the primary tendon repair. Aoki et al6 found that the value of a PTFE felt graft was less than that of a patellar tendon graft at 12 weeks. A long-term study to evaluate the tensile strength of PTFE grafts is therefore necessary. THE JOURNAL OF BONE AND JOINT SURGERY


In our study stiffness also recovered significantly throughout the experimental period, but small values for stiffness at the PTFE felt-bone interface during the early recovery period indicate the possibility of elongation of the construct. Therefore, when stiffness of reconstructed tendons is small relative to surrounding connective tissues, strenuous postoperative exercise should be avoided in order not to cause elongation of reconstructed infraspinatus tendons.12,13 Histocompatibility of a non-absorbable synthetic material is evaluated by both affinity with tissue and irritability against tissue. The tissue affinity of PTFE felt was observed at the PTFE felt-tendon interface between 12 and 24 weeks when fibrous tissue infiltrated between PTFE fibres with little granulation. At the PTFE felt-bone interface, aligned fibrous tissue extended between the PTFE felt fibres. Some PTFE fibres incorporated into woven bone. These phenomena indicate excellent tissue affinity of PTFE felt fibres to bone as well as to tendon. The tissue irritability induced by the PTFE fibres was demonstrated by a foreign-body reaction in the PTFE felt-bone interface between 12 and 24 weeks. Wear particles were considered to have been produced as a result of fragmentation of the PTFE material which eventually induced a foreign-body reaction. These reactions occurred in the limited area of the PTFE-bone interface. Clinically, to reinforce the interface between the PTFE felt and bone or tendon, Ozaki et al5 recommended a supplementary procedure in which PTFE felt was sutured not only to bone and tendon but also to the surrounding soft tissues, i.e. the thickened subacromial bursa or periosteum of the humerus.5 In order to cover the low initial tensile strength of the PTFE felt and to reduce a foreign-body reaction in the medullary canal, the procedure of Ozaki et al5 was considered to be useful. The mechanical recovery and tissue affinity of PTFE felt to bone and tendon observed in our study support its use for reconstruction of the rotator cuff, but the possible development of a foreign-body reaction should be borne in mind. An ongoing process at the PTFE felt-bone interface may lead to mechanical failure. Further experimental studies are needed to develop an ideal

VOL. 85-B, No. 2, MARCH 2003

287

bio-absorbable or tissue-engineered collagen material which provides effective strength and less tissue reaction for tendon to bone reconstruction. The authors wish to thank T. Yamashita, MD, PhD for editorial assistance and Dr K. Okamura, MD, for technical assistance. This study was supported in part by the Ministry of Education, Culture, Sports, Science, and Technology grants 10671370 and 11307026 for aiding scientific researches. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.

References 1. Berger K, Sauvage LR, Rao AM, Wood SJ. Healing of arterial prostheses in man: its incompleteness. Ann Surg 1972;175:118-27. 2. Kenny DA, Berger K, Walker MW, et al. Experimental comparison of the thrombogenicity of fibrin and PTFE flow surfaces. Ann Surg 1980;191:355-61. 3. Ozaki J, Fujimoto S, Masuhara K, Tamai S, Yoshimoto S. Reconstruction of chronic massive rotator cuff tears with synthetic materials. Clin Orthop 1986;202:173-83. 4. Kimura A, Okamura K, Fukushima S, Ishii S, Aoki M. Clinical results of rotator cuff reconstruction with PTFE felt augmentation for irreparable massive rotator cuff tears. The Shoulder Joint 2000;24:485-8. 5. Ozaki J, Okamura K, Kadono T, Tatsumi H. Reconstruction of chronic massive rotator cuff tears with synthetic materials: indications and limits of this procedure. The Shoulder Joint 1995;19:417-20. 6. Aoki M, Uchiyama E, Ohtera K, et al. Restoration of tensile properties at tendon insertion to bone by a patellar tendon-tibia autograft: an experimental study with canine infraspinatus. J Shoulder Elbow Surg 1999;8:628-33. 7. Noyes FR, Butler DL, Grood ES, Zernicke RF, Hefzy MS. Biomechanical analysis of human ligament grafts used in knee-ligament repairs and reconstructions. J Bone Joint Surg [Am] 1984;66-A:344-52. 8. Uchiyama E, Yamakoshi K, Sasaki T. Measurement of mechanical characteristics of tibial periosteum and evaluation of local differences. J Biomech Eng 1998;120:85-91. 9. Pierre PS-T, Olson EJ, Elliott JJ, et al. Tendon-healing to cortical bone compared with healing to a cancellous trough: a biomechanical and histological evaluation in goats. J Bone Joint Surg [Am] 1995;77A:1858-66. 10. Takagishi N, Nobuhara K, Fukuda H, et al. Manual for evaluation of the shoulder: shoulder evaluation sheet. J Jpn Orthop Assoc 1987;61:623-9. 11. Kagaya K, Aoki M, Kimura A, et al. Mechanical properties of the repaired infraspinatus tendon: comparison of immediate repair and delayed repair. Trans ORS 2001:785. 12. Woo S-LY. Biomechanics of tendons and ligaments. In: Schmid-Schonbein GW, Woo S-LY, Zweifach BW, eds. Frontiers in biomechanics. New York, etc: Springer, 1986:180-95. 13. Woo S-LY, Maynard J, Butler D, et al. Ligament, tendon, and joint capsule insertions to bone. In: Woo S-LY, Buckwalter JA, eds. Injury and repair of the musculoskeletal soft tissues. Park Ridge: The American Academy of Orthopaedic Surgeons, 1988:133-7.