Ligament.pdf

1

3

4

5

6

7

8 Q1

9101112

1314

1 6

1718192021

22 Q3232425262728

2 9

4647

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

Q2

Acta Biomaterialia xxx (2014) xxx–xxx

ACTBIO 3237 No. of Pages 9, Model 5G

2 June 2014

Contents lists available at ScienceDirect

Acta Biomaterialia

journal homepage: www.elsevier .com/locate /actabiomat

A novel silk-based artificial ligament and tricalcium phosphate/polyether ether ketone anchor for anterior cruciate ligamentreconstruction – Safety and efficacy in a porcine model

http://dx.doi.org/10.1016/j.actbio.2014.05.0151742-7061/� 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

⇑ Corresponding author at: Department Health Sciences and Technology, ETHZurich, Switzerland. Tel.: +41 443863755.

E-mail address: [email protected] (J.G. Snedeker).

Please cite this article in press as: Li X et al. A novel silk-based artificial ligament and tricalcium phosphate/ polyether ether ketone anchor for acruciate ligament reconstruction – Safety and efficacy in a porcine model. Acta Biomater (2014), http://dx.doi.org/10.1016/j.actbio.2014.05.015

Xiang Li a,b, Jiankang He c, Weiguo Bian d, Zheng Li d, Wenyou Zhang c, Dichen Li c, Jess G. Snedeker a,b,⇑a Department Health Sciences and Technology, ETH Zurich, Switzerlandb Department of Orthopaedics, Balgrist, University of Zurich, Switzerlandc State Key Lab for Manufacturing System Engineering, Xi’an Jiaotong University, People’s Republic of Chinad Department of Orthopaedics Surgery, First Hospital of Xi’an Jiaotong University, People’s Republic of China

a r t i c l e i n f o a b s t r a c t

303132333435363738394041

Article history:Received 13 January 2014Received in revised form 14 May 2014Accepted 16 May 2014Available online xxxx

Keywords:SilkTCPPEEKAnchorArtificial ACL

424344

Loss of ligament graft tension in early postoperative stages following anterior cruciate ligament (ACL)reconstruction can come from a variety of factors, with slow graft integration to bone being widely viewedas a chief culprit. Toward an off-the-shelf ACL graft that can rapidly integrate to host tissue, we have devel-oped a silk-based ACL graft combined with a tricalcium phosphate (TCP)/polyether ether ketone anchor. Inthe present study we tested the safety and efficacy of this concept in a porcine model, with postoperativeassessments at 3 months (n = 10) and 6 months (n = 4). Biomechanical tests were performed after eutha-nization, with ultimate tensile strengths at 3 months of �370 N and at 6 months of �566 N – comparableto autograft and allograft performance in this animal model. Comprehensive histological observationsrevealed that TCP substantially enhanced silk graft to bone attachment. Interdigitation of soft and hardtissues was observed, with regenerated fibrocartilage characterizing a transitional zone from silk graftto bone that was similar to native ligament bone attachments. We conclude that both initial stabilityand robust long-term biological attachment were consistently achieved using the tested construct,supporting a large potential for silk–TCP combinations in the repair of the torn ACL.

� 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

45

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

1. Introduction

Anterior cruciate ligament (ACL) ruptures are among the mostfrequent and severe ligament injuries [1], with between 250,000and 400,000 ACL disruptions being diagnosed in the United Stateseach year [2,3]. ACL reconstruction surgery is increasingly commonas well, with more than half of diagnosed complete ACL tears beingsurgically reconstructed [4,5]. While ACL reconstruction is gener-ally viewed as a low-risk procedure that provides substantialshort- and medium-term patient benefit [5,6], the long-term ben-efit of the procedure is often debated [7,8]. In addition to marginallong-term benefit over conservative treatment, various aspects ofthe procedure itself, such as donor-site morbidity related to theuse of a ligament autograft, leave substantial room for improve-ment [9,10].

Regarding functional outcome, the major issue of inadequateknee stability most likely relates to the loss of graft tension in early

80

81

82

83

postoperative stages [11]. This loss of tension in the replacementligament can come from a variety of factors, but poor/slow graftintegration to the patient’s bone is widely considered to be a pri-mary cause. To accelerate bone integration and promote ligamenttension, surgeons often employ a bone–tendon–bone (BTB) auto-graft that is typically extracted from the patient’s patellar tendon.Although transplanted bone blocks at the end of the graft robustlyincorporate at the grafting site, the graft extraction itself can bevery painful and is a major cause of patient dissatisfaction [12].For non-professional athletes, doctors thus often employ a ham-string autograft as a less painful alternative graft tissue source.While a hamstring autograft procedure can be employed with typ-ically less donor site morbidity than a BTB autograft, hamstringgrafts are often considered to provide inferior mechanical stability.This inferior stability can be attributed to the slow integration ofthe hamstring tendon within the bone tunnel, and associated graftelongation and loosening [13]. By the time the healing phase iscomplete, a ligament reconstruction using hamstring autograftmay be too slack to ensure normal joint stability.

In general, all currently available choices for ligament graftingmaterial have drawbacks. These include donor site morbidity for

nterior

http://dx.doi.org/10.1016/j.actbio.2014.05.015

mailto:[email protected]


http://www.sciencedirect.com/science/journal/17427061

http://www.elsevier.com/locate/actabiomat


84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178Q4

179

180

181

182

183

Fig. 1. The general design concept of the silk graft with TCP/PEEK anchor, andphotograph of a packaged implant ready for animal implantation.

2 X. Li et al. / Acta Biomaterialia xxx (2014) xxx–xxx


2 June 2014

autografts, disease transmission for allografts and adverse immuneresponse for allografts, xenografts and synthetic biomaterials.These aspects compound challenges associated with ligament lax-ity, and overcoming these challenges has emerged as a majorresearch objective that motivates both biomaterial and tissueengineering approaches to the problem [14–16].

Among these approaches, silk-based ACL grafts have beenincreasingly investigated [14,15,17–23], and have even reached assafety and efficacy oriented clinical trials in humans [24]. Acrossthese many studies, the regeneration of the ligament itself using silkscaffolds has emerged as a viable alternative to autografted tendon.However, these efforts have still not addressed the greatest chal-lenge – namely the integration of the ligament graft to the host bone[25,26]. Many strategies to enhance graft to bone healing have beenpursued, including the use of bone marrow derived mesenchymalstem cells [27], bioactive factors like bone morphogenetic proteins[28,29] and/or the incorporation of osteoconductive bioceramics,such as hydroxyapatite, tricalcium phosphate (TCP) and brushitecalcium phosphate cement [30–32]. However, nearly all of the abovestudies attempting to enhance graft to bone integration havefocused primarily on stimulating a biological response, whileneglecting the critical considerations of primary mechanical stabil-ity (immediate post-operative ‘‘pull-out strength’’) and confinementof bioactive compounds to the bone tunnel (restriction of suchagents from the joint space).

The present study sought to test whether a novel device for ACLreconstruction could achieve these three objectives: robust pri-mary strength, rapid integration and restriction of osteoconductivematerial from the joint space. Toward this goal, we combined aTCP/polyether ether ketone (PEEK) anchor with a silk-based ACLscaffold. We then evaluated the short- and mid-term safety andefficacy of this system using a porcine model. The results indicatethat a combination of TCP and PEEK for ACL graft fixation is notonly feasible, but also effective.

2. Materials and methods

2.1. Silk graft with TCP/PEEK anchor preparation

The general concept of the silk-based ACL graft with TCP/PEEKanchor (Fig. 1) and detailed design parameters and fabricationtechniques have been described previously [33–36]. Briefly, asilk-based artificial ACL graft was prepared with an architectureof 6(0) � 2(2) � 144(10) � 2(12), which indicates 6 fibers per bun-dle without any twist (‘‘0’’ twist reflects parallel silk fibers), 2 bun-dles per yarn with 2 mm per turn, 144 yarns per cord with 10 mmper turn and 2 cords per ACL scaffold with 12 mm per turn. All arti-ficial ACL grafts were produced from raw silk yarns (Bombyx mori,Grege 20/22, Trudel Limited, Zurich, Switzerland) with thoroughremoval of the antigenic protein sericin after graft preparation[34]. A TCP component with a diameter of 9.0 mm, length of11.5 mm and porosity of 45%, was fabricated using techniquescombining rapid prototyping and gel-casting methods [36]. ThePEEK anchor, which had a hollow cap, was machined to press-fitwith the TCP insert, having an outside diameter of 10.4 mm, aninner diameter of 6.0 mm and two arms with anchoring teeth.The TCP insert was designed with an ‘‘H’’-shaped form to accom-modate the enlacing silk scaffold around the bridge of the TCPinsert. The sub-assembled TCP-silk construct was then press-fittedwithin the external PEEK anchor housing [33]. All implants weresterilized as follows: after fabrication, each component wasimmersed in 75% ethanol for 24 h; the TCP components were fur-ther treated by ultraviolet germicidal irradiation at a dose of100 mJ cm�2. Afterwards, all parts were assembled in a sterilehood. Finally, the assembled graft was packaged and autoclavedat 121 �C for 30 min.

Please cite this article in press as: Li X et al. A novel silk-based artificial ligamcruciate ligament reconstruction – Safety and efficacy in a porcine model. Act

2.2. Study design

Animal experiments were carried out under the Rules and Reg-ulations of the Animal Care and Use Committee, First Hospital ofXi’an Jiaotong University, People’s Republic of China. The presentstudy was performed with 14 healthy adult male pigs (Chinesetri-hybrid pig: Xianyang breed) aged around 4 months and weigh-ing 55.2 ± 3.7 kg (mean ± SD) at the time of surgery. ACL recon-structions were performed on the left knee. The animals weredivided into two study groups, with 10 animals planned for sacri-fice at the 3 month time point and four animals at the 6 monthtime point. Within the 3 month group, seven of 10 animals wereused for biomechanical tests, the remaining three, plus one fromthe biomechanical test samples, being used for histological obser-vation. From the 6 month group, three of four animals were usedfor biomechanical tests, with the remaining specimen being allo-cated for histological analysis, together with one specimen fromthe three biomechanical test samples.

2.3. Preoperative treatment

The pigs were thoroughly disinfected by spraying with 0.25%didecyl dimethyl ammonium bromide solution 2 days before sur-gery. Antibiotic (800,000 U of penicillin) was given to each pig bytwo intramuscular injections the day before the operation. A 3.5%sodium pentobarbital solution was used as the anesthetic. Eachanimal was given 0.5 ml of the solution per kg body weight byabdominal injection, followed 5 min later with a 0.2 ml kg�1 doseby venous injection. The animal was then positioned supine onthe operating table in a specially designed constraint. The left hindleg was shaved and thoroughly washed with povidone–iodinesolution.

2.4. Surgical procedure

An open surgical procedure was used. First a longitudinal, med-ian skin incision was made 5 cm proximal to the superior margin ofthe patella to the tibia tubercle. The knee joint was accessed using amedial parapatellar capsular approach. The joint was then posi-tioned to 90� in flexion, and the native ACL was carefully removedat the bone surfaces. A 9.0 mm tunnel was drilled over the footprintof the native ACL, creating a tunnel of 20 mm depth (Fig. 2A). To

ent and tricalcium phosphate/ polyether ether ketone anchor for anteriora Biomater (2014), http://dx.doi.org/10.1016/j.actbio.2014.05.015


184

185 Q5

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

X. Li et al. / Acta Biomaterialia xxx (2014) xxx–xxx 3


2 June 2014

avoid damage to the articular cartilage on the medial condyle, thedrilling axis was set as 11 o’clock in the transversal plane, with a45� anterior deviation in the sagittal plane using the femoral axisas the frame of reference. A drilling sleeve was developed to preventany slippageor wobble of the drill, avoiding enlargement of the tun-nel entry and compromise of the primary stability of the implant. A7.0 mm tunnel along the same axis was drilled through the tibiawith a specially designed synchronizing sleeve (Fig. 2B). Theimplant was tapped into the tunnel using a custom insertion toolwith a hollow cylindrical cross-section to accommodate the silkgraft and secure the PEEK anchor housing. After the TCP/PEEK scaf-fold had been pushed into the femur tunnel, the free distal end of thesilk scaffold was drawn through the tibia tunnel with a speciallydesigned retractor (Fig. 2C). The knee joint was then flexed to 30�before pulling the silk graft tight and fixing with an interferencescrew (U6 � 19 mm, Anklin AG, Switzerland; Fig. 2D). Upon com-pletion of the graft fixation, the length and cross-sectional area ofthe silk graft was measured and recorded.

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

2.5. Postoperative treatment

An analgesic (100 mg pethidine) was given to each animal twicea day for 3 days following the surgery. In order to prevent infection,800,000 U of penicillin was given to each animal twice a day until5 days after the operation. A disinfectant solution (0.25% didecyl-dimethylammonium bromide) was sprayed on the animals andbedding biweekly until the end of the experiment. All pigs wererandomly assigned housing in one of three pens (5 � 8 m), andallowed unrestricted daily activity in their pen. The activity leveland degree of lameness were monitored. As planned, 10 pigs wereeuthanized by lethal injection of thiamylal sodium at a postopera-tive time point of 3 months. The remaining four pigs were eutha-nized at 6 months. After euthanization, both knees weredissected. The samples used for biomechanical test (seven of 10at 3 months, three of four at 6 months) were immediately storedat �20 �C. The remaining samples, used for histological observa-tion, were cut into small specimens and immediately fixed in a10% buffered formalin solution.

The lengths and cross-sectional areas of the regenerated ACL(left knee) and native ACL (right knee) were measured afterremoval of the surrounding soft tissues. The length was measuredthree times for each sample using calipers, with the mean valueused for statistical analysis. The cross-sectional area of the nativeACL was calculated as the width multiplied by the thickness, andthe cross-sectional area of the regenerated ACL was calculatedusing the diameter. Three different measuring points were chosensince the cross-section was not uniform, and the mean value wasadopted for statistical analysis.

Fig. 2. Schematic description of the op


2.6. Biomechanical testing

Samples used for biomechanical testing were thawed overnightat 4 �C prior to testing. All surrounding soft tissues on both kneeswere removed, leaving only the ACL graft (or native ACL in controlknees) connecting the tibia and femur (Fig. 3). The cross-sectionalarea and length were measured as described above, and recorded.Cyclic loading and tension to failure testing were performed on auniversal material testing machine mounted with a 5 kN load cell(Sans s-503, MTS Corporation., Minnesota, USA). A speciallydesigned clamp was used to fix the femur and tibia, to ensure thatthe ACL graft was fully aligned with the machine test axis. For thetests, a pre-conditioned loading of 5 N was applied, before a force-controlled cyclic loading of 0.5 mm s�1 from 100 to 250 N wasapplied over 250 cycles, representing the loads of normal walking[37]. Finally, tension to failure was applied to determine the ulti-mate tensile strength (UTS). The force–displacement curves of eachcycle were recorded. The dynamic creep of the construct was ana-lyzed by calculating the difference in machine displacement afterthe first cycle and after the final cycle. The UTS was determinedas the first peak (a more than 50% drop in attained force) in theforce–displacement curve. These data are subsequently referredto as ‘‘in vivo data’’. These values were compared to previouslyreported in vitro test data from biomechanical characterization ofthe silk graft and TCP/PEEK anchor [33,34]. These in vitro dataare assumed to reflect the implant stability immediately afterimplantation.

2.7. Histological observation

The samples used for histological observation were fixed in buf-fered formalin solution for 2 months prior to further preparation. Afirst group was embedded in resin and specimens were sectionedparallel to the longitudinal axis of the bone tunnel. The sectionslides were stained with Goldner’s trichrome for general assess-ment of bone, PEEK, new bone, TCP, cartilage, silk graft and transi-tional zone. A second group of specimens was decalcified andembedded in paraffin after removal of the PEEK or interferencescrew, since these materials cannot be sectioned with paraffinembedding. The paraffin embedded sections were stained withhematoxylin & eosin (H&E), Gomori and Masson for evaluatingthe bone, cartilage, graft, transitional zone and Sharpey’s fibers.

2.8. Statistical analysis

All data were expressed as mean ± standard deviation. Datawere compared with a two-tailed Student’s t-test, and statistically

en surgical procedure employed.



272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

Fig. 3. Images of the knee with native and reconstructed ACL mounted for biomechanical characterization. (A) Native ACL; (B) regenerated ACL at 3 months; (C) regeneratedACL at 6 months.



2 June 2014

significant values were defined as p < 0.05 (significance is indicatedwith an asterisk in all figures).

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

3. Results

3.1. General observations

After ligament reconstruction, all animals were able to stand upon three legs by postoperative day 3. All animals were walking onfour legs, though with a detectable degree of lameness, within5–7 days postoperatively. Activity levels increased gradually after1 week, until resumption of normal activity and no discerniblelameness by the second postoperative week. At the time of eutha-nization, no animal exhibited graft failure or apparent degenera-tive changes in the surrounding tissues of the knee (articularcartilage, menisci, other ligaments). Analysis of blood chemistryindicated no systemic markers of inflammation.

3.2. Biomechanical evaluations

The length of silk graft at implantation was 33.6 ± 4.2 mm(n = 14). The length of the regenerated ACL was 42.2 ± 3.4 mm at3 months (n = 7) and 43.3 ± 2.9 mm at 6 months (n = 3). The lengthof the native ACL was 37.4 ± 3.2 mm at 3 months (n = 7) and37.3 ± 2.1 mm at 6 months (n = 3). A comparison of the graft lengthwith the contralateral (native) ligament revealed the differences tobe non-significant (Fig. 4A). The cross-sectional area of the silkgraft at implantation was 30.2 ± 2.3 mm2 (n = 14). The cross-sectional area of the regenerated ACL was 57.5 ± 8.1 mm2 at3 months (n = 7) and 84.6 ± 11.5 mm2 at 6 months (n = 3). Thecross-sectional area of the native ACL was 23.6 ± 4.8 mm2 at3 months (n = 7) and 30.3 ± 4.4 mm2 at 6 months (n = 3). Compari-son of the graft cross-sectional areas indicated significant differ-ences between the area at the time of implantation and that at3 months (p < 0.01), with a further significant increase between 3and 6 months (p = 0.016; Fig. 4B).

Two regenerated ACL specimens sacrificed at 3 months failedprior to the onset of cyclic loading (with 151 and 184 N,respectively). Although the loading modes of these two failed sam-ples were different from other samples, we also included these spec-imens in the UTS statistical analysis, but excluded them from thestiffness analysis. The UTS of native ACL was 1384 ± 181 N at3 months (n = 7), and had increased, though not significantly(p = 0.14), to 1749 ± 284 N at 6 months (n = 3), similar to reports inthe literature [38]. The UTS of the regenerated ACL was311 ± 103 N at 3 months (n = 7), and had increased significantly


(p < 0.01) to 566 ± 29 N at 6 months (n = 3) (Fig. 4C). All failuresoccurred in the midsubstance of the regenerated ACL – with nopull-out failures observed at either the femoral tunnel or the tibialtunnel. The stiffness was calculated as the slope of the force–displacement curve between 100 and 250 N of the 250th cycle.The stiffness of native ACL was 192 ± 22 N mm�1 at 3 months(n = 5), and had increased significantly (p < 0.01) to259 ± 15 N mm�1 at 6 months (n = 3). The stiffness of regeneratedACL was 148 ± 19 N mm�1 at 3 months (n = 5), and had increasedsignificantly (p = 0.035) to 183 ± 10 N mm�1 at 6 months (n = 3)(Fig. 4D).

Compared to the graft length at time of implantation, there wasa significant increase (p = 0.04) in length of the regenerated ACLafter 3 months in vivo – an increase in length that was larger thanelongation observed after 100,000 cycles of in vitro testing(Fig. 5A). This parameter reflects any slippage of the anchor orinterference screw, and creep in the graft. The graft length at peakload for the regenerated ACL was 14.6 ± 6.5 mm at 3 months(n = 7), and had increased, though not significantly (p = 0.27), to18.1 ± 3.0 mm at 6 months (n = 3), which was �10% lower thanthe value of native ACL (�20 mm) (Fig. 5B). The dynamic creep ofthe native ACL was 0.74 ± 0.21 mm at 3 months (n = 5) and0.88 ± 0.30 mm at 6 months (n = 3). The dynamic creep of regener-ated ACL was 1.48 ± 0.49 mm at 3 months (n = 5), and haddecreased, though not significantly (p = 0.145), to 1.07 ± 0.25 mmat 6 months (n = 3). There was a significant decrease (p = 0.046)in dynamic creep between the TCP/PEEK anchor and the regener-ated ACL at 6 months against in vitro data and the regeneratedACL at 3 months (Fig. 5C).

3.3. Histological observations

H&E staining of longitudinal sections (along the axis of the tun-nel) and transverse sections (perpendicular to the tunnel) indi-cated substantial fibrous tissue formation surrounding the silkfibers at 6 months (Fig. 6A and B).

Goldner’s trichrome stain was adopted to observe the regener-ated tissue in the bone tunnel [39]. The TCP could still be located at3 months, with new bone tissue observed to surround the TCP(Fig. 6C). New bone tissue was increasingly present at 6 months,with fibrocartilage observed to lie between silk fibers and thenew bone tissue (Fig. 6D). The silk to bone transitional area wascharacterized with H&E stain in terms of silk, fibrous tissue, fibro-cartilage and bone (Fig. 7A at 6 months). The regenerated fibroustissue layer characterized using Masson stain at 6 months(Fig. 7D) was nearly twice as thick as that at 3 months (Fig. 7C),



Fig. 4. Comparison of the geometry and mechanical properties of implants at the time of implantation (adapted from previous in vitro test data [33,34]) with the regeneratedACL and native ACL at different time points. ⁄p < 0.05. (A) Length; (B) cross-sectional area; (C) UTS; (D) stiffness.

Fig. 5. Comparison of mechanical properties of the silk graft alone (adapted from Ref. [34]), pull-out characteristics of the TCP/PEEK anchor (adapted from Ref. [33]), theregenerated ACL and the native contralateral ACL at different time points. p < 0.05. (A) Elongation (calculated as the final graft/ligament length minus the length of the silkgraft at implantation); (B) graft length at peak load (calculated with the distance from the preload of 5 N to the failure point); (C) dynamic creep (calculated as the shiftdistance between the first cycle and the 250th cycle); (D) force–displacement loading curve (the load to failure distance and dynamic creep are indicated with red lines).



2 June 2014

Please cite this article in press as: Li X et al. A novel silk-based artificial ligament and tricalcium phosphate/ polyether ether ketone anchor for anteriorcruciate ligament reconstruction – Safety and efficacy in a porcine model. Acta Biomater (2014), http://dx.doi.org/10.1016/j.actbio.2014.05.015


357

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

402

403

404

405



2 June 2014

reflecting the regeneration of fibrous tissue surrounding the silkgraft. The connection of fibrous tissue to bone was through fibro-cartilage zones, and Gomori staining revealed interdigitated(Sharpey’s) fibers in these zones. Numerous such fibers could beseen at 6 months (Fig. 7B).

At the regions of contact between the silk, the interferencescrew (IS) and the bone in the tibial tunnel, cartilaginous tissuewas observed at 3 months. This cartilaginous layer at the silk–IS–bone corner was even more pronounced at 6 months (Fig. 8A).However, at the interface between silk and bone the transitionwas characterized by the presence of silk, fibrous tissue and bonetissue only, with no fibrocartilage layer observable at 6 months(Fig. 8B and C). There were only a few cases showing a thin, non-continuous layer of fibrocartilage at 6 months (Fig. 8D). Compari-son between the tibial and femoral tunnels revealed a relativeabsence of new bone formation in the tibial tunnel, with acorresponding lack of a cartilagenous silk to bone transition inthe tibial tunnel.

406

407

408

409

410

411

412

413

414

415

416

417

418

419

420

421

422

423

424

4. Discussion

The so-called ‘‘gold standard’’ graft choice for torn ACL recon-struction is a bone–patellar tendon–bone autograft (BPTB). Apartfrom donor-site morbidity, BPTB reconstructions have consistentlyyielded favorable clinical outcomes, with a 90–95% success rate interms of a return to pre-injury activity level [40]. We havedesigned and developed a synthetic alternative to a BPTB autograftwith analogous structural, biological and mechanical properties.The present study tested the efficacy of this concept in a largeanimal model.

Because of a uniquely advantageous combination of biocompat-ibility and robust biomechanical strength in the short and middleterms, sericin-extracted silk fibers were adopted as the ligamentgraft in the device, implemented with a wired architecture thatprovides mechanical behavior similar to that of the human ACL[34,35]. To stimulate integration of the silk within the bone tunnel,

Fig. 6. (A, B) H&E stain of the silk graft with regenerated fibrous tissues at 6 months. BlaGoldner’s trichrome stain, providing a histological overview of the femoral tunnel (Cfibrocartilage; S: silk.


we employed a porous TCP scaffold – a well-characterized US Foodand Drug Administration-approved osteoconductive biomaterial.To overcome the notorious fragility of porous TCP scaffold, weintroduced a PEEK anchor housing to provide initial fixation stabil-ity. The anchor featured a push-in design with a press-fit and saw-tooth anchorage elements to provide primary stability and limitgraft elongation and soft tissue movements at the tunnel aperture.The present study examined both the overall functional (biome-chanical) performance of the assembled construct and the host–implant integration at the various biomaterial–tissue interfaces.

4.1. Functional outcome

From a functional standpoint, this efficacy study focused on themechanical strength and stiffness of the regenerated ACL. The UTSof the regenerated ACL increased by �82% (Fig. 4C) from 3 monthsto 6 months. Although the absolute strength of the graft was stillfar from that of the native ACL, these values fall safely below typ-ical maximal ACL loads associated with normal daily activity(�250 N [37]). The UTS values we recorded at 3 months comparefavorably with other ACL reconstruction studies using porcinemodels with sacrifice after 3 months [41], although the UTS valueswe recorded at 6 months were approximately 40% lower thananother study of a similar duration [42]. Also as in previous stud-ies, failures consistently occurred in the midsubstance of thereconstructed ACL, with no incidence of tunnel pull-out failure. Itshould be noted that graft elongation at failure typically exceeded15 mm (Fig. 5B), a distance at which recruitment of other stabiliz-ing structures (muscles, other ligaments) would reasonably beexpected to prevent graft failure.

Graft slippage and elongation also play critical roles in func-tional performance, as these aspects are closely related to loss ofgraft tension and relative joint laxity. The elongation of the regen-erated ACL compared to graft length at implantation was �8.6 mmfor both 3 months and 6 months (Fig. 5A), although interpretingthese values is difficult in view of the fact that the animals were

ck arrows indicate silk fibers. (A) Transverse section; (B) longitudinal section. (C, D)) at 3 months and (D) at 6 months. T: TCP; P: PEEK; B: bone; NB: new bone; C:



425

426

427

428

429

430

431

432

433

434

435

436

437

438

439

440

Fig. 7. Histological images of the silk graft transition to bone and transitional histological zone within the femoral tunnel. (A, B, D) At 6 months; (C) at 3 months. B: bone; C:fibrocartilage; F: fibrous tissue; S: silk. (A) H&E stain; (B) Gomori stain (Sharpey’s fiber pointed with black arrows); (C, D) Masson stain.

Fig. 8. Histological images of the silk graft transition to bone and the transitional histological zone within the tibial tunnel at 6 months. IS: interference screw; B: bone; C:fibrocartilage; F: fibrous tissue; S: silk. (A) Goldner’s trichrome stain; (B) H&E stain; (C, D) Masson stain.



2 June 2014

growing over the course of the experiment. More conclusively,elongation of the silk graft during cyclic loading (dynamic creep)of the regenerated ACL decreased by �38% from 3 months to6 months, indicating that the graft became less viscoelastic in thistime frame. However, because any measure of graft elongationincludes effects from both the femoral side (silk/TCP/PEEK) andthe tibial side (silk-IS), it was not possible to assess the relativecontribution of either side to the overall function. Nonetheless,when compared to in vitro biomechanical test data, the dynamic


creep of the regenerated ACL was �35% less than the original graftafter 6 months (Fig. 5C), clearly indicating that the implant becamemore elastic (less viscoelastic) over the course of healing – and wascomparable to the native ACL.

4.2. Histological assessment

Silk-based scaffolds have been increasingly investigated as apotential graft material for tendon and ligament regeneration



441

442

443

444

445

446

447

448

449

450

451

452

453

454

455

456

457

458

459

460

461

462

463

464

465

466

467

468

469

470

471

472

473

474

475

476

477

478

479

480

481

482

483

484

485

486

487

488

489

490

491

492

493

494

495

496

497

498

499

500

501 Q6

502

503

504

505

506

507

508

509

510

511

512

513

514

515

516

517

518

519

520

521

522

523

524

525

526

527

528

529

530

531

532

533

534

535

536

537

538

539

540

541

542

543

544

545

546

547

548

549

550

551

552

553

554

555

556

557

558

559

560

561

562Q7

563Q8

564

565



2 June 2014

[15,22,43,44]. This is in part because of the advantageous biologicalproperties of silk as well as the robust biomechanical strength inthe short and middle terms. After 3 months of postoperative heal-ing, we observed that the silk scaffolds remained largely intact butwere intermixed with regenerated fibrous tissue, the cells of whichwere well aligned with and often attached to the silk fibers. After6 months the regenerated fibrous tissue that was intermixed withsilk fibers appeared to be increased, although not substantially(Fig. 6A and B), with the majority of newly generated fibrous tis-sues forming around the silk graft core. The silk graft remainedaround 70% intact even at 6 months, reflecting the characteristi-cally slow degradation of silk, which enables the biomaterial scaf-fold to continue supporting functional demands of the ligamentuntil the host tissues eventually overtake these loads. These find-ings are consistent with extensive studies by others using silkgrafts for ACL reconstruction [17–21]. The fibrous tissues coveredaround the silk graft were disorganized and can be regarded assome kind of scar tissue. There was substantial vascularity in thescar tissue (as indicated by the pink color observable in Fig. 3Band C), possibly associated with the observed increase in thicknessof the graft over the course of regeneration. The cross-sectionalarea of regenerated ACL at 6 months was �47% larger than thatat 3 months (Fig. 4B), which was mainly attributed to the growthof scar tissue. We suspect that the dense tissue around the graftmay have prevented interstitial fluid penetration within the graft,an essential factor in the silk degradation process.

The present study differentiates itself from previous studies inits use of a porous TCP scaffold (mimicking a bone block) combinedwith a PEEK anchor. From histological observation we found thatthe porous TCP scaffold substantially increased silk graft to boneattachment. There was a clear tendency toward new bone forma-tion in the femur tunnel in contrast to the tibial tunnel, whichlacked the presence of TCP (Fig. 7). At 3 months the TCP scaffoldcould still be seen clearly, while at 6 months considerably lessTCP material could be identified. Over the course of TCP remodel-ing, the enlaced silk graft was effectively incorporated within thetunnel, corresponding to apparently accelerated biological fixationby 3 months and robust incorporation into the tunnel by 6 months.In contrast, there was little new bone tissue formation observed inthe tibial tunnel, particularly at the margins of the silk graftpressed against one side of the tunnel by the interference screw.Lacking any histological transition from the silk to the host bone,it would appear that silk graft fixation remains dependent on themechanical purchase of the screw (Fig. 8A), and would thus remainsusceptible to subsequent loosening.

Based on the femoral tunnel histology, we conclude that thepresence of TCP provoked the formation of tissue transitions fromsilk into regenerated fibrous tissue, into regenerated fibrocartilageand eventually into bone (Fig. 7A). These transitions reflect thosepresent in the attachment of the native ACL to bone – a highly spe-cialized histological feature that allows the effective transmissionof forces from soft to hard tissues. Histological examination ofthe implanted constructs clearly identified such regions at3 months (Fig. 7C), with these regions becoming furtherpronounced at 6 months (Fig. 7D). Interestingly, numerous inter-digitated fibers (Sharpey’s fibers) were observed to project fromthe regenerated fibrous tissue into newly generated bone tissuethrough a regenerated fibrocartilage zone (Fig. 7B). Thus a rela-tively biomimetic attachment of the silk graft to the femoral bonetunnel was achieved. In contrast, the tibial tunnel showed compar-atively no fibrocartilage layer at the interface between the silk graftand the bone (Fig. 8B). While we attribute this lack to the absenceof TCP, other factors could potentially have played a role – forinstance, the relative mechanical stability of the different anchor-ing systems applied to each tunnel.


4.3. Limitations

The main limitations of the study regard the use of young pigs,which grew considerably over the course of the study. The surgerywas performed when the pigs were 4 months of age, correspondingto adolescence. These animals grow approximately �3 kg per weekuntil 7 months of age, then grow at �2 kg per week until10 months of age. Taking this into consideration, the results ofthe present study should be considered in relative terms, particu-larly with regard to the substantial regenerative capacity of adoles-cent mammals compared to adults. Another limitation regards thelack of strict control groups. Here the silk graft fixed by an interfer-ence screw in the tibial tunnel was used for histological compari-son. It should be noted that the regenerative capacity andbiomechanical environment of the tibia and femur are likely to dif-fer and could thus introduce a bias into our analysis. In an idealstudy, controls using a silk graft on the femoral side and a ‘‘gold-standard’’ bone–tendon–bone autograft group would be ideal formore direct comparison. We favored a reduced number of controlsto minimize the burden on the animals, relying instead on compar-ison with similar large-animal ACL reconstruction studies in theliterature. Various such studies that have investigated autograftattachment within a bone tunnel have shown relatively poor graftintegration without augmentation by TCP or bioactive compoundssuch as bone morphogenetic protein [45,46]. On this basis, as wellas broad human clinical outcomes, we would expect similarly poorbone ingrowth to autografts in the porcine model we used. A thirdlimitation imposed by our choice of animal model was the inabilityto perform functional clinical joint tests (such as the drawer test orthe Lachman test). In man, such tests represent important criteriafor postoperative evaluation of ACL reconstruction. More detailedevaluation in other animal models represents the next step inour research. These studies will include explicit microscopic inves-tigation of the joint cartilage surfaces for any abrasion that mayhave been caused by TCP particulates in the joint space. In thepresent study we assumed that, by virtue of the implant design,TCP particulates would be excluded from the joint space, and onlya macroscopic inspection of the joint surfaces was performed toconfirm this assumption.

5. Conclusion

A silk-based artificial ACL graft combined with a TCP/PEEKanchor was tested and evaluated in a porcine model of ACL recon-struction, and evaluated 3 and 6 months postoperatively. A seriesof biomechanical tests were performed and a detailed histologicalanalysis was made. We found that incorporation of the silk graftinto bone was markedly enhanced in tunnels augmented with aTCP block enlaced around the end of the silk graft. Histologicaltransitions from the silk graft to bone were similar to features ofnative ACL to bone attachment. We conclude that the concept ofa TCP/PEEK anchored silk graft performs well as a synthetic alter-native to an autograft. This study provides a basis for eventualsafety and efficacy testing in man.

Acknowledgements

The authors wish to thank Trudel Limited (Zurich, Switzerland)for providing raw silk for our research. We also thank Mr. HansruediSommer, Dr. med. Arnd Viehöfer, Prof. Dr. Marc Bohner and Dr. Xiao-jun Wang for their expertise and kind help with the experiments.This study was partly funded by the National Natural Science Foun-dation of China (No. 51105298), the Program of Leading Talent inChangshu of China, the Research Fund for the Doctoral Program ofHigher Education of China(No. 20110201120027), the Fundamental



566

567

568

569

570

571

572

573

574575576577578579 Q9580581582583584585586Q10587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624

625626627



2 June 2014

Research Funds for the Central Universities and the Bonizzi ThelerFoundation.

628629630631632633634635636637

Appendix A. Figures with essential color discrimination

Certain figures in this article, particularly Figs. 1–3 and 5–8 aredifficult to interpret in black and white. The full color images canbe found in the on-line version, at http://dx.doi.org/10.1016/j.actbio.2014.05.015.

638639640641642643644645646647648649650651652653Q11654655656657658659660661662663664665666667668669670671672673674675676677678679Q12680681682683684685686687688689690691

692

References

[1] Parkkari J et al. The risk for a cruciate ligament injury of the knee inadolescents and young adults: a population-based cohort study of 46,500people with a 9 year follow-up. Br J Sports Med 2008;42(6):422–6.

[2] Freeman JW, Kwansa AL. Recent advancements in ligament tissue engineering:the usw of various techniques and materials for acl repair. Recent Pat BiomedEng 2008;1:6.

[3] Junkin D et al. Knee ligament injuries. Orthopaedic knowledge update: sportsmedicine. Rosemont, IL: American Academy of Orthopaedic Surgeons; 2009 [p.136].

[4] Chen K et al. A hybrid silk/RADA-based fibrous scaffold with triple hierarchyfor ligament regeneration. Tissue Eng A 2012;18(13–14):1399–409.

[5] Murray MM. History of ACL treatment and current gold standard of care. TheACL handbook. Springer; 2013. p. 19–28.

[6] Hospodar MSJ, Miller MD. Controversies in ACL reconstruction: bone–patellartendon–bone anterior cruciate ligament reconstruction remains the goldstandard. Sports Med Arthrosc Rev 2009;17(4):242–6.

[7] Lind M et al. Medium to long-term follow-up after ACL revision. Knee SurgSports Traumatol Arthrosc 2012;20(1):166–72.

[8] Gifstad T et al. Long-term follow-up of patellar tendon grafts or hamstringtendon grafts in endoscopic ACL reconstructions. Knee Surg Sports TraumatolArthrosc 2012:1–8.

[9] Bernardino S. ACL prosthesis: any promise for the future? Knee Surg SportsTraumatol Arthrosc 2010;18(6):797–804.

[10] Samuelsson K et al. Trends in surgeon preferences on anterior cruciateligament reconstructive techniques. Clin Sports Med 2013;32(1):111–26.

[11] Arnold MP et al. The remains of anterior cruciate ligament graft tension aftercyclic knee motion. Am J Sports Med 2005;33(4):536–42.

[12] Sanchis-Alfonso V. Anterior knee pain after ACL reconstruction: how to avoidit. The ACL-deficient knee. Springer; 2013. p. 357–72.

[13] Baumfeld JA et al. Tunnel widening following anterior cruciate ligamentreconstruction using hamstring autograft: a comparison between doublecross-pin and suspensory graft fixation. Knee Surg Sports Traumatol Arthrosc2008;16(12):1108–13.

[14] Liu HF et al. A comparison of rabbit mesenchymal stem cells and anteriorcruciate ligament fibroblasts responses on combined silk scaffolds.Biomaterials 2008;29(10):1443–53.

[15] Fan HB et al. Anterior cruciate ligament regeneration using mesenchymal stemcells and silk scaffold in large animal model. Biomaterials 2009;30(28):4967–77.

[16] Zeugolis DI, Chan JCY, Pandit A. Tendons: engineering of functional tissues.Tissue Eng 2011:36.

[17] Altman GH et al. Silk matrix for tissue engineered anterior cruciate ligaments.Biomaterials 2002;23(20):4131–41.

[18] Altman GH et al. Silk-based biomaterials. Biomaterials 2003;24(3):401–16.[19] Horan RL et al. Yarn design for functional tissue engineering. J Biomech

2006;39(12):2232–40.[20] Altman GH et al. The use of long-term bioresorbable scaffolds for anterior

cruciate ligament repair. J Am Acad Orthop Surg 2008;16(4):177–87.[21] Horan RL et al. Biological and biomechanical assessment of a long-term

bioresorbable silk-derived surgical mesh in an abdominal body wall defectmodel. Hernia 2009;13(2):189–99.


[22] Panas-Perez E, Gatt C, Dunn M. Development of a silk and collagen fiberscaffold for anterior cruciate ligament reconstruction. J Mater Sci Mater Med2013;24(1):257–65.

[23] Teh T, Goh J, Toh S. Characterization of electrospun substrates for ligamentregeneration using bone marrow stromal cells. 13th International Conferenceon Biomedical Engineering. Springer; 2009.

[24] Horan RL. SeriACL™ device (Gen IB) trial for anterior cruciate ligament (ACL)repair. Medford, MA: Serica Technologies, Inc.; 2009 [NCT00775892].

[25] Lei J, Temenoff JS. Engineering fibrous tissues and their interfaces with bone.Structural interfaces and attachments in biology. Springer; 2013. p. 323–49.

[26] He P et al. In vitro ligament–bone interface regeneration using a trilineagecoculture system on a hybrid silk scaffold. Biomacromolecules 2012;13(9):2692–703.

[27] Soon MYH et al. An analysis of soft tissue allograft anterior cruciate ligamentreconstruction in a rabbit model – a short-term study of the use ofmesenchymal stem cells to enhance tendon osteointegration. Am J SportsMed 2007;35(6):962–71.

[28] Lee J et al. Enhanced regeneration of the ligament–bone interface using apoly(L-lactide-co-epsilon-caprolactone) scaffold with local delivery of cells/BMP-2 using a heparin-based hydrogel. Acta Biomater 2011;7(1):244–57.

[29] Kim HM et al. The role of transforming growth factor beta isoforms in tendon-to-bone healing. Connect Tissue Res 2011;52(2):87–98.

[30] Dorozhkin SV. Bioceramics of calcium orthophosphates. Biomaterials2010;31(7):1465–85.

[31] Farshad M et al. Embossing of a screw thread and TCP granules enhances thefixation strength of compressed ACL grafts with interference screws. Knee SurgSports Traumatol Arthrosc 2012;20(2):268–74.

[32] Shi P et al. Variation of the effect of calcium phosphate enhancement ofimplanted silk fibroin ligament bone integration. Biomaterials 2013.

[33] Li X et al. A novel silk–TCP–PEEK construct for anterior cruciate ligamentreconstruction: an off-the-shelf alternative to a bone–tendon–bone autograft.Biofabrication 2014;6(1):015010.

[34] Li X, Snedeker JG. Wired silk architectures provide a biomimetic ACL tissueengineering scaffold. J Mech Behav Biomed Mater 2013;22:30–40.

[35] Li X, Snedeker J. Long term mechanical behavior of silk ACL scaffolds undercyclic loading. J Biomech 2012;45:S325.

[36] Li X et al. Fabrication of porous beta-tricalcium phosphate with microchanneland customized geometry based on gel-casting and rapid prototyping. ProcInst Mech Eng [H] 2011;225(H3):315–23.

[37] Coleridge SD, Amis AA. A comparison of five tibial-fixation systems inhamstring-graft anterior cruciate ligament reconstruction. Knee Surg SportsTraumatol Arthrosc 2004;12(5):391–7.

[38] Proffen BL, Murray MM. In vivo models of ACL injury (central defect, porcine,ovine, canine). The ACL handbook. Springer; 2013. p. 139–66.

[39] Schmitz N et al. Basic methods in histopathology of joint tissues. OsteoarthritisCartilage 2010;18:S113–6.

[40] Sahoo S, Toh SL, Goh JC. A bFGF-releasing silk/PLGA-based biohybrid scaffoldfor ligament/tendon tissue engineering using mesenchymal progenitor cells.Biomaterials 2010;31(11):2990–8.

[41] Vavken P et al. Biomechanical outcomes after bioenhanced anterior cruciateligament repair and anterior cruciate ligament reconstruction are equal in aporcine model. Arthroscopy 2012;28(5):672–80.

[42] Milano G et al. Comparison of femoral fixation methods for anterior cruciateligament reconstruction with patellar tendon graft: a mechanical analysis inporcine knees. Knee Surg Sports Traumatol Arthrosc 2007;15(6):733–8.

[43] Liu HF, et al., In vivo study of ACL regeneration using silk scaffolds in a pigmodel. In: 13th International Conference on Biomedical Engineering2009;23(1–3):1512–4 [vol. 1–3].

[44] Teh TKH, Toh SL, Goh JCH. Aligned hybrid silk scaffold for enhanceddifferentiation of mesenchymal stem cells into ligament fibroblasts. TissueEng Part C Methods 2011;17(6):687–703.

[45] Huangfu X, Zhao J. Tendon–bone healing enhancement using injectabletricalcium phosphate in a dog anterior cruciate ligament reconstructionmodel. Arthroscopy 2007;23(5):455–62.

[46] Thomopoulos S et al. Effect of bone morphogenetic protein 2 on tendon-to-bone healing in a canine flexor tendon model. J Orthopaedic Res2012;30(11):1702–9.





Ligament.pdf

Documents

ligament autograft

ligament grafting83material

hamstring autograft

acl reconstruction surgery

donorsite morbidity

donor site morbidity

severe ligament injuries

btb autograft