Embed Size (px)
Citation preview
To
JMD
Tt(wt(ldrt0rawncWaaptmsspcpwpG
a
E
R
C
1d
he effect of growth differentiation factor-5–coated suturesn tendon repair in a rat model
oshua S. Dines, MD,a Lawrence Weber, MD, PhD,b Pasquale Razzano, MS,b Rita Prajapati, PhD,c
ark Timmer, PhD,d Steven Bowman, PhD,c Lawrence Bonasser, PhD,e David M. Dines, MD,a,b and
aniel P. Grande, PhD,b New York, New Hyde Park, and Ithaca, NY; Raynham, MA; and Somerville, NJwtcwtve1
Ofdprisbhmtpmwht
bficdtat
gtccap
w
endon ruptures are common injuries that are oftenreated surgically. Growth Differentiation Factor-5GDF-5) has been shown to accelerate tendon healingith varying degrees of success. We used a novel
echnique to apply recombinant human GDF-5rhGDF-5) to suture and hypothesized that controlled,ocal delivery of rhGDF-5 can be used to enhance ten-on repair. Tendons of 92 rats were transected andepaired with sutures. All researchers were blinded tohe following treatment groups (24 rats in each group):
rhGDF (control), 24 ng/cm rhGDF, 55 ng/cmhGDF, and 556 ng/cm rhGDF. Rats were euthanizedt 3 weeks (n � 48) and at 6 weeks (n � 48). Suturesere coated with rhGDF-5 using a novel dip-coat tech-ique. Enzyme-linked immunosorbent assay confirmedonsistent and reproducible delivery of rhGDF-5.ithin each group, 8 were tested biomechanically,
nd 4 were assessed histologically. Histologic gradingt 3 weeks showed improved healing in tendons re-aired with coated suture versus controls. By 6 weeks,
here were no significant differences. At 3 weeks, mini-al isolated cartilage formation was observed; 6-week
amples showed more extensive presence, typicallyurrounding suture fibers. At 3 weeks, tendons re-aired with rhGDF-5–coated sutures resulted in signifi-antly higher ultimate tensile load and stiffness com-ared with control sutures (P � .05) At 6 weeks, thereere no significant differences in the mechanicalroperties of repaired tendons. At 3 weeks, rh-DF-5 induced significant tendon hypertrophy that
From The Hospital for Special Surgery, and bDepartment ofOrthopaedic Surgery and Research, Long Island Jewish MedicalCenter; c,dJohnson and Johnson Regenerative Therapeutics, andthe eDepartment of Biomechanics, Cornell University.
quipment for this study, including the growth factors and suturesused, were donated by Depuy, Inc.
eprint requests: Joshua S. Dines, MD, Kerlan Jobe OrthopaedicClinic, Los Angeles, CA 90045 (E-mail: [email protected]).opyright © 2007 by Journal of Shoulder and Elbow SurgeryBoard of Trustees.
058-2746/2007/$32.00
roi:10.1016/j.jse.2007.03.001as more pronounced than at 6 weeks. In addition,endons repaired with rhGDF-5 showed an in-reased rate of healing versus control repairs at 3eeks. This study showed that a novel dip-coating
echnique can be used to deliver growth factors inarying concentrations to local repair sites to accel-rate tendon healing. (J Shoulder Elbow Surg 2007;6:215S-221S.)
perative repair of the rotator cuff provides success-ul clinical results; however, the clinical outcome oftenoes not reflect the structural outcome of the re-air.2,14,15,18 Studies cite the actual healing rate ofepaired tendons to be less than 50%.14,18,28 This ismportant because, despite overall good clinical re-ults, function and strength recovery are significantlyetter if the repairs heal.2,13,18,16 To this end, repairsave progressed with the evolution of better cuffobilization techniques and improved restoration of
he anatomic footprint.20,28 Currently, the major im-ediment to successful healing is biology. Enhance-ent of the repair process by using growth factorsith documented roles in the repair process couldelp regenerate functional tissue, as opposed to scarissue.
Growth Differentiation Factor-5 (GDF-5), a mem-er of the transforming growth factor-� (TGF-�) super-amily, has been shown to accelerate tendon healingn multiple animal models.1,12,25 In particular, a re-ent study by Nakase et al24 showed that cartilage-erived morphogenetic protein 1 (CDMP-1), which is
he human analogue of mouse GDF-5, is producednd activated specifically at the site of torn rotator cuff
endon.One of the main impediments precluding the use of
rowth factors in human beings is limiting delivery ofhe factors to the appropriate tissue in the properoncentrations. Thus far, bioabsorbable scaffolds,ollagen matrices, local injections, and sutures havell been used to localize growth factors and generoducts to the repair sites.1,10-12,17,23,25,29
In a previous study, we were able to coat suturesith varying concentrations of growth factors in a
eproducible fashion.5,6 In that same work, we de-
215S
sfaawsGecpv
M
wac
S
(gtrpstTwccwdp
bwpfro
tswice
R
pUuftuvw
S
twginowsstctc2ie21
ls
H
elsgccttlw
d
5
o
fi
siwe
216S Dines et al J Shoulder Elbow SurgSeptember/October 2007
cribed the pharmacokinetics of growth factor releaserom the sutures and also showed that a consistentmount of growth factor is released from the suturesfter passage through soft tissue. It was based on thisork that we hypothesized that by using this novel
uture-coating technique, different concentrations ofDF-5 could be consistently and reproducibly deliv-red to a tendon repair site. And, by using differentoncentrations of GDF-5-coated sutures, tendon re-air would be enhanced while limiting potential ad-erse reactions.
ATERIALS AND METHODS
The study was performed in 2 parts. In part 1, suturesere coated with varying concentrations of growth factor5;nd in part 2, we studied a rat model of tendon repair withoated sutures.
uture-coating processA dip-coating process was used to coat 4-0 VICRYL
Polyglactin 910) Sutures (Ethicon, Somerville, NJ) withelatin/recombinant human (rh) GDF-5. The dip-coat solu-
ion was comprised of 4 mL of gelatin solution and 2 mL ofhGDF-5 growth-factor solution. The gelatin component wasrepared by heating a 10wt% solution of medical-gradeoluble bovine collagen (Semed-S, Kensey-Nash, Exton, PA)o 80°C for 10 minutes, followed by incubation at 37°C.he rhGDF-5 (Biopharm GmbH, Heidelberg, Germany)as reconstituted with 10 mM hydrogen chloride at con-entrations of 0, 0.12, 0.6, and 3 mg/mL. The resultingoncentrations of growth factor in the dip-coating solutionsere 0, 40, 200, and 1000 �g/mL, respectively. Theip-coating solutions were kept at 37°C until use. Thereparer was blinded to the concentrations.
Before dip-coating, the sutures were pretreated with aath of 70% ethanol solution for 10 minutes, followed by aash with saline. A 12-cm segment of suture was thenlaced in the dip-coating solution and incubated at 37°Cor 30 minutes with gentle agitation. The suture wasemoved from the solution and hung vertically to air-dryvernight.
The rhGDF-5 loading on the suture was quantified usinghe previously described enzyme-linked immunosorbent as-ay (ELISA) method.5,6 The quantified results from ELISAere used to determine the amount of rhGDF-5 on the suture
n units of nanograms per centimeter of suture. Eight suturesoated with each of the 4 different dip-coat solutions werevaluated (n � 8 per solution).
at model of tendon repairThe study used 96 male Sprague-Dawley rats. The study
rotocol was approved by the Institutional Animal Care andse Committee. The operative procedures were performednder sterile conditions using a previously described modelor tendon repair.26 The tendon was immediately repairedo its bony insertion with the previously described suturessing Mason Allen stitch configurations. There was noisible shearing of the gelatin coating from the suture as it
as pulled through the tissue. gtudy designThe animals were randomized to a 3-week or 6-week
ime group. Each time group consisted of 48 rats. These ratsere randomly assigned to 1 of the following treatmentroups (all researchers were blinded to the groups, n � 12
n each group): 0 rhGDF5 (control), 24 ng/cm rhGDF5, 55g/cm rhGDF5, and 556 ng/cm rhGDF5. Note that basedn ELISA performed on the coated sutures (the results ofhich are discussed in detail subsequently), the 0 �g/mL
olution corresponded to a concentration of 0 ng/cm on theuture, the 40 �g/mL solution corresponded to a concen-ration of 24 ng/cm on the suture, the 200 �g/mL solutionorresponded to a concentration of 55 ng/cm on the su-ure, and the 1000 �g/mL solution corresponded to aoncentration of 556 ng/cm on the suture. An average of.9 cm of suture was incorporated into each repair, mean-
ng the approximate absolute dose of growth factor deliv-red was 0 ng from the 0 ng/cm suture, 72.5 ng from the4 ng/cm suture, 159.5 ng from the 55 ng/cm suture, and612.4 ng from the 556 ng/cm suture.
Of the 12 rats in each group, 4 were evaluated histo-ogically, and 8 were tested biomechanically. Figure 1ummarizes the overall study design.
istologic evaluationAt either 3 or 6 weeks postoperatively, the animals were
uthanized (48 rats at each time point). Both the right andeft limbs were fixed, processed, embedded, sectioned, andtained with hematoxylin and eosin. The specimens wereraded using a modified Soslowsky score evaluating 3omponents: collagen grade, degree of angiogenesis, andartilage induction.27 The scoring system was based on a 0o 3 score for each component and differed with relevanceo characteristics observed for each. Lower scores corre-ated with better histologic results. The scoring delineationas as follows:Collagen grade:0 � normal collagen oriented tangentially.1 � mild changes with collagen fibers, less than 25%
isorganized.2 � moderate changes with collagen fibers, less than
0% disorganized.3 � marked changes in collagen, more than 50% dis-
rganized.Degree of angiogenesis:0 � normal tendon tissue.1 � increased presence of capillaries.2 � moderate infiltration of tissue with vessels.Cartilage formation:0 � no cartilage formation.1 � isolated hyaline cartilage nodules.2 � moderate cartilage formation of 25% to 50%.3 � extensive cartilage formation, more than 50% of the
eld involved.Three independent, blinded observers assessed each
lide. Each block had an average of 10 fields. Each spec-men was assigned an average grade per variable, whichas then totaled to provide an overall histologic score forach rat. Total scores were computed for each treatment
roup and compared.
B
–osdEMalr
vtsactcpvmu
RS
wttec
R(sf
S
TvumTlr((gA
erall
T
G
C
J Shoulder Elbow Surg Dines et al 217SVolume 16, Number 5S
iomechanical testingUpon euthanasia, limbs were excised and frozen at
80°C. Separate tests comparing the mechanical behaviorf fresh and frozen samples showed no effect of frozentorage. Samples were thawed individually at 4°C on theay of testing. Limbs were mounted in the grips of annduraTec ELF 2100 test frame (Bose Corp, Minnetonka,N). Samples were pulled in uniaxial tension until failure atgrip-to-grip displacement rate of 0.025 mm/s in a closed
oop control. The resultant load was recorded to an accu-acy of 0.005 N at a sampling rate of 2 Hz.
Temporal data on load and displacement were con-erted into time-independent load-displacement curves. Ul-imate tensile strength (maximum load supported by theample normalized to the cross sectional area), stiffness,nd elastic toughness were calculated for each sample. Theross-sectional area was calculated from measurements ofhe length of the major and minor axes of the tendonross-section (eg, the width and depth of the tendon sam-les). For each calculated property, a 2-way analysis ofariance was performed to determine the effects of treat-ent (0, 24, 55, 556 ng/cm) and surgery (experimental ornoperated, contralateral control).
ESULTSuture coating
The presence of rhGDF-5 on the 4-0 VICRYL Sutureas detectable by the ELISA method. As mentioned,
he growth factor loading increased as its concentra-ion in the dip-coat solution increased. Table I delin-ates the dip-coat solution concentrations and the
Figure 1 Ov
orresponding amount of rhGDF-5 on the sutures. s
egression analysis indicates a linear relationshipR2 � 0.99) between the rhGDF-5 loading on theuture and dip-coat solution rhGDF-5 concentrationor the range evaluated in this study.
urgical model results
Histology. Histologic scores are summarized inable II. The scoring system presented here is not aalidated histology scoring scale although it has beensed previously, which makes it impossible to com-ent on the statistical significance of the findings.hat being said, grading at 3 weeks demonstratedower scores (indicating better healing) for tendonsepaired with coated sutures compared with controlsFigure 2). The low-dose (24 ng/cm) and high-dose556 ng/cm) GDF-5 exhibited improvements in colla-en orientation compared with control suture alone.t 6 weeks, regardless of treatment, results were
study design.
able I Dip-coat solution and corresponding concentration on suture
roup
Dip-coatconcentration
(�g/mL)
Concentrationon suture(ng/cm)
Absolute GrowthFactor Dose
Delivered (ng)
ontrol 0 0 01 40 24 69.62 200 55 159.53 1000 556 1612.4
imilar.
m6ittc
gv
tat6os
pogwl
Ftfs
Fsr*
T
*
Fadt�
218S Dines et al J Shoulder Elbow SurgSeptember/October 2007
Isolated hyaline cartilage formation was observedinimally at 3 weeks, regardless of treatment type. Atweeks postoperatively, there was an increased
ncidence of de novo cartilage induction regardless ofreatment compared with 3 weeks. Cartilage forma-ion could predictably be found in regions that were
able II Histologic scores*
Group (ng/cm) Histology score
Control (0) 3.5224 2.0455 2.19
556 2.45
Lower scores correspond to better healing.
igure 2 Representative slides at 3 weeks from (A) control groupnd (B) group that received 24 ng/cm recombinant human growthifferentiation factor-5 (rhGDF-5. Note the aligned collagen archi-
ecture in the group treated with rhGDF-5. (Original magnification100; hematoxylin and eosin stain.)
learly associated with suture placement (Figure 3). In c
eneral, the treated tendons exhibited greater neo-ascularization than those in the controls.
Biomechanics. At the 3-week time point, ultimateensile strength (Figure 5), ultimate load (Figure 4),nd stiffness (Figure 6) of contralateral, unoperated
endons were lower than those measured in the-week group. In general, the mechanical propertiesf the 3-week tendons were half that of the 6-weekamples.
Ultimate tensile strength data indicated that re-aired tendons were weaker than contralateral, un-perated tendons (P � .001); however, the effect ofrowth factor treatment was significant comparedith the control group (P � .01). Ultimate tensile
oads of experimental tendons were more similar to
igure 3 Example of cartilage formation at repair site in the groupreated with 55 ng/cm recombinant human growth differentiationactor-5. (Original magnification �100; hematoxylin and eosintain.)
igure 4 Ultimate load (N) at 3 weeks presented with �tandard deviation (experimental, blue; contralateral unoperated,ed.) rhGDF-5, Recombinant human growth differentiation factor-5.Statistically significant improvement compared with control.
ontralateral, unoperated tendons, although there
w.gGaerlu
etg55cudut
two
btancdso
Fwotc
Fdr*
Fwuep
Fdr*
J Shoulder Elbow Surg Dines et al 219SVolume 16, Number 5S
as still a significant difference between them (P �02). The ultimate load of experimental tendons in theroups with 0 ng/cm rhGDF-5 and 55 ng/cm rh-DF-5 were 30% lower than contralateral, unoper-ted tendons (P � .05), but the ultimate load ofxperimental tendons in the groups with 24 ng/cmhGDF-5 and 556 ng/cm rhGDF-5 were only 10%ower than the ultimate load of the contralateral,noperated tendons.
The growth factor treatment had a significantffect on tendon stiffness (P � .005). The stiffness ofhe control group tendons was 30% less than theroup with 556 ng/cm rhGDF-5 and more than0% less stiff than the groups with 24 ng/cm and5 ng/cm rhGDF-5. Experimental tendons in theontrol group were 50% as stiff as contralateral,noperated tendons, whereas experimental ten-ons were 90% to 100% as stiff as contralateral,noperated tendons. The elastic toughness of con-
igure 5 Ultimate tensile strength (MPa) at 3 weeks presentedith � standard deviation (experimental, blue; contralateralnoperated, red.) rhGDF-5, Recombinant human growth differ-ntiation factor-5. *Statistically significant improvement com-ared with control.
igure 6 Stiffness (N/mm) at 3 weeks presented with � standardeviation (experimental, blue; contralateral unoperated, red).hGDF-5, Recombinant human growth differentiation factor-5.Statistically significant improvement compared with control.
ralateral, unoperated controls and experimental t
endons did not differ significantly (P � .2). Thereas a significant effect of growth factor treatmentn toughness.
Despite the statistically significant differencesetween the experimental and control groups, when
he results of the different concentrations of rhGDF-5re isolated, the differences between the groups areot statistically significant. This indicated that with theoncentrations assessed in this study, there was noose-response effect. At 6 weeks after repair, notatistically significant differences were noted in anyf the biomechanical variables tested. This is illus-
igure 7 Ultimate tensile strength (MPa) at 6 weeks presentedith � standard deviation (experimental, blue; contralateral un-perated, red.) rhGDF-5, Recombinant human growth differentia-
ion factor-5. *Statistically significant improvement compared withontrol.
igure 8 Stiffness (N/mm) at 6 weeks presented with � standardeviation (experimental, blue; contralateral unoperated, red).hGDF-5, Recombinant human growth differentiation factor-5.Statistically significant improvement compared with control.
rated in Figures 7 and 8.
D
edrcdpilbisCf
RGaaep
oNtoissvatea
stClgtgeta
tfahtlit
tt
rrcpsaucwwbswb
rstrmssscbAroipG
lrsastfe
wctrhpofslg
220S Dines et al J Shoulder Elbow SurgSeptember/October 2007
ISCUSSION
In this study, we were successfully able to accel-rate the endogenous repair process with the localelivery of a growth factor, GDF-5. Improving theate of healing in surgically repaired rotator cuffsontinues to be a highly researched topic. Manyifferent growth factors enhance the tendon healingrocess, including platelet-derived growth factor,
nsulin-like growth factor-1, and vascular endothe-ial growth factor. Our decision to use GDF-5 wasased on previous studies in animal models show-
ng its effect on tendon repair as well as a recenttudy in humans documenting the expression ofDMP-1 at the torn edge of rotator cuff tendons in
ull-thickness tears.24
GDF-5 is a member of the TGF-� superfamily.19,21
ecent studies have elucidated the important role ofDF-5 in the establishment and maintenance of tendonsnd its ability to affect cell proliferation,22 adhesion,7,8
nd angiogenesis.30 Based on these studies, Chhabrat al4 hypothesized that GDF-5 modulates tendon re-air.
One of the first studies of the local regulating factorsf tendon injury in human beings was performed byakase et al.24 The group obtained proximal residual
endon containing the torn edge of full-thickness tearsf the rotator cuff from 7 patients. Using in situ hybrid-zation and immunohistochemical analysis, theyhowed that CDMP-1 was activated specifically at theite of the torn tendon. Because this study was obser-ational in nature, it did not elucidate whether thectivation of CDMP-1 enhanced healing of the torn
endon. It does, however, correlate well with thearlier animal studies suggesting that CDMP-1 playsrole in tendon healing.Aspenberg and Forslund1 have done multiple
tudies showing enhanced tendon healing with car-ilage-derived morphogenetic proteins, especiallyDMP1 and CDMP2, which are the human ana-
ogues of GDF-5 and GDF-6, respectively. Rickert’sroup25 tested the effects of GDF-5 on Achilles
endon repairs in rats and concluded that localrowth factor delivery showed promising beneficialffects on tendon repair and that the chondroinduc-ive capacity of GDF-5 may be modulated by dosedjustments the growth factor.Identifying the best vehicle for delivery is one of
he main impediments to the use of growth factorsor the treatment of tendon disorders. The safetynd feasibility of retroviral and adenoviral vectorsas raised concerns, as has the potential inflamma-ory effects of biodegradable scaffolds (eg, polyg-ycolic acid, poly-L-lactic acid).17 We believed thatf sutures that are normally used in rotator cuff
endon repairs could be coated with growth fac- pors, we would have a clinically relevant model forhe local delivery of growth factors.
Coating of sutures with growth factors had beeneported previously; however, the method was noteliable nor was extensive testing done to measureoncentrations on the suture both before and afterassage through the tissue.25 In a previous in vitrotudy,5 as well as in this report, we were successfullyble to impregnate VICRYL sutures with rhGDF-5 bysing a novel, dip-coating process. By adjusting theoncentration of GDF-5 in the dip-coat solution, weere able to control its concentration on the sutures,hich was then successfully quantified by ELISA. And,y comparing the concentrations of rhGDF-5 on theuture before and after passage through the tendon,e were able to prove that the growth factor hadeen delivered directly to the repair site.5
A major concern about the use of GDF-5 for tendonepair is its chondroinductive properties.3,9,12 Severaltudies, including ours, have found cartilage cell nests athe site of tendon repairs in which GDF-5 was incorpo-ated. Studies also indicate that the response to GDFsay be dependent on local factors, such as mechanical
ignals. Aspenberg and Forslund1 showed that re-ponse to GDF-6 depends on mechanical environment,uggesting that tissue differentiation depends on me-hanical environment. It is also worth noting that studiesy Rooney et al26 showed that cartilage forms in thechilles tendon of rats after transection, even without
epair. This may explain why cartilage was present inur control group repairs although no rhGDF-5 wasmplanted. We are currently investigating the effects ofostoperative loading on tendon repairs enhanced withDF-5 to better understand this issue.The results of the control group data were equiva-
ent to that previously reported for uncoated sutureepair, indicating that the gel-coat solution on theuture had no adverse effects. At 6 weeks postoper-tively, the properties of the repaired tendons wereimilar to contralateral controls, indicating that thisime period was sufficient for functional repair. There-ore the 3-week time point seems more appropriate tovaluate the effects of this growth factor.
At 3 weeks, the experimental rhGDF-5 groupsere able to withstand loads similar to unoperatedontralateral tendons. Furthermore, the control groupendon repairs failed at much lower loads. Theseesults indicate the ability of GDF-5 to accelerate theealing of tendon repairs and are consistent with ourrevious in vitro work analyzing the pharmakineticsf growth factor release from sutures. In that study weound that 50% of the growth factor had eluted off theuture at the 24-hour mark, another 50% was re-eased by the 48-hour time point, and the remainingrowth factor was released by the 72-hour time
oint.5
rretbdcdbgoms
C
aaopdtai
R
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
J Shoulder Elbow Surg Dines et al 221SVolume 16, Number 5S
The data from the biomechanical results in thiseport indicated that the varying concentrations ofhGDF-5 did not result in a significant dose-responseffect: all 3 experimental dosages were equally effec-ive in accelerating the healing of tendon. It is possi-le that if lower doses of rhGDF-5 were used, aose-response effect would have been seen. We areurrently doing in vitro studies assessing the effects ofifferent concentrations of rhGDF-5 on rat tendon fibro-last collagen synthesis, cell proliferation, and cell mi-ration. We hope this work, combined with data fromngoing animal studies, will help elucidate the ideal,inimal concentration of the growth factor necessary to
how a significant effect on tendon healing.
ONCLUSION
Future studies in larger animal models with alter-tions of the postoperative mechanical environmentre needed to better elucidate the clinical usefulnessf GDF-5 in rotator cuff tendon repairs. However, in thisreliminary work, we used a reliable method for localelivery of a growth factor. We were able to prove that
he use of rhGDF-5 accelerates both the biomechanicalnd histologic properties of surgically repaired tendons
n a rat model.
EFERENCES
1. Aspenberg P, Forslund C. Enhanced tendon healing with GDF-5and 6. Acta Orthop Scand 1999;70:51-4.
2. Boileau P, Brassart N, Watkinson D, Carles M, Hatzidakis A,Krishnan S. Arthroscopic repair of full-thickness tears of thesupraspinatus: does the tendon really heal? J Bone Joint Surg Am2005;87:1229-40.
3. Carter DR, Blenman PR, Beaupre GS. Correlations betweenmechanical stress history and tissue differentiation in initial fracturehealing. J Orthop Res 1988;6:736-48.
4. Chhabra A, Tsou D, Clark RT, et al. GDF-5 deficiency in micedelays Achilles tendon healing. J Orthop Res 2003;21:826-835.
5. Dines JS, Grande DA, Razzano P, et al. rhGDF-5–coated suturesenhance tendon healing in a rat model. Presented at: Orthopae-dic Research Society Annual Meeting; 2007.
6. Dines JS, Weber LW, Prakapati R, et al. Influence of mechanicalenvironment and rhGDF-5 on tendon repair. Presented at: TransOrthop Research Society Annual Meeting; 2006.
7. Duke J, Elmer WA. Cell adhesion and chondrogenesis in brachy-pod mouse limb mesenchyme: fragment fusion studies. J EmbryolExp Morph 1978;48:161-8.
8. Duke J, Elmer WA. Effect of the brachypod mutation on celladhesion and chondrogenesis in aggregates of mouse limbmesenchyme. J Embryol Exper Morph 1977;42:209-17.
9. Erlacher L, McCartney J, Piek E, et al. Cartilage derived morpho-genetic protein and osteogenetic protein 1 differentially regulateosteogenesis. J Bone Miner Res 1998;13:382-92.
0. Evans C, Ghivizzani S, Herndon J, Robbins P. Gene therapy for
the treatment of musculoskeletal diseases. J Am Acad Orthop Surg2005;13:230-42.1. Forslund C, Aspenberg P. Improved healing of transected rabbitAchilles tendon after a single injection of CDMP2. Am J SportsMed 2003;31:555-9.
2. Forslund C, Reuger D, Aspenberg P. A comparative dose-response study of cartilage-derived morphogenetic protein(CDMP)-1, 2, and 3 for tendon healing in rats. J Orthop Res2003;21:617-21.
3. Fuchs B, Gilbart M, Hodler J, Gerber C. Clinical and structuralresults of open repair of an isolated one-tendon tear of the rotatorcuff. J Bone Joint Surg Am 2006;88:309-16.
4. Galatz L, Ball CM, Teefey S, Middleton W, Yamaguchi K. Theoutcome and repair integrity of completely arthroscopically re-paired large and massive rotator cuff tears. J Bone Joint Surg Am2004;86:219-24.
5. Gartsman G, Khan M, Hammerman S. Arthroscopic repair offull-thickness tears of the rotator cuff. J Bone Joint Surg Am 1998;80:832-40.
6. Gerber C, Fuchs B, Hodler J. The results of repair of massive tearsof the rotator cuff. J Bone Joint Surg Am 2000;82:505-15.
7. Gerich TG, Kang R, Fu F, Robbins P, Evans C. Gene Transfer tothe rabbit patellar tendon: potential for genetic enhancement oftendon and ligament healing. Gene Ther 1996;3:1089-93.
8. Harryman D, Mack LA, Wang KY, Jackins S, Richardson M,Matsen F. Repairs of the rotator cuff. Correlation of functionalresults with integrity of the cuff. J Bone Joint Surg Am 1991;73:982-9.
9. Hogan BL. Bone morphogenetic proteins: multifunctional regula-tors of vertebrae development. Genes Dev 1996;10:1580-94.
0. Kim DH, Elattrache NS, Tibone JE, et al. Biomechanical compar-ison of a single-row versus double-row suture anchor technique forrotator cuff repair. Am J Sports Med 2006;34:407-14.
1. Kingsley DM. The TGF-Beta superfamily: new members, newreceptors, and new genetic tests of function in different organ-isms. Genes Dev 1994;8:133-46.
2. Nakamura K, Hirosawa K, Hama K. Defects in the growth plateof brachypodism mice: light and electron microscopic findingsand cell proliferation by the method of tritiated thymidine autora-diography. J Japan Orthop Assoc 1984;58:835-45.
3. Nakamura N, Shino K, Natsuume T, Horibe S, Kaneda Y, OchiT. Early biological effect of in vivo gene transfer of platelet-derived growth factor (PDGF-�) into healing patellar ligament.Gene Therapy 1998;5:1165-70.
4. Nakase T, Sugamoto K, Miyamoto T, et al. Activation of cartilagederived morphogenetic protein-1 in torn rotator cuff. Clin OrthopRel Res 2002;399:140-5.
5. Rickert M, Jung M, Adiyaman M, Richter W, Simank H. Growthand differentiation factor-5-coated suture stimulates tendon heal-ing in an achilles tendon model in rats. Growth Factors 2001;19:115-26.
6. Rooney P, Grant ME, McClure J. Endochondral ossification andde nove collagen synthesis during repair of the rat achillestendon. Matrix 1992;12:274-81.
7. Soslowsky LJ, Carpenter JE, DeBano CM, Banjeri I, Moalli M.Development and use of an animal model for investigations ofrotator cuff disease. J Shoulder Elbow Surg 1996;5:383-92.
8. Sugaya H, Maeda K, Matsuki K, Moriishi J. Functional andstructural failure of rotator cuff repair. Arthroscopy 2005;21:1307-16.
9. Uggen J, Dines JS, Uggen C, et al. Tendon gene therapymodulates the local repair environment of the shoulder. J AmerOsteopath Assoc. 105:2005;120-1.
0. Yamashita H, Shimizi A, Kato M, Nishitoh H, Ichijo H, et al.
Growth/differentiation factor-5 induces angiogenesis in vivo.Exp Cell Res 1997;235:218-26.
