Dose–Response Effect of an Intra-Tendon Application of RecombinantHuman Platelet-Derived Growth Factor-BB (rhPDGF-BB) in a RatAchilles Tendinopathy Model

Vivek Shah,1 Alison Bendele,2 Joshua S. Dines,3 Hans K. Kestler,1 Jeffrey O. Hollinger,4 Nadeen O. Chahine,5

Christopher K. Hee1

1Sports Medicine, BioMimetic Therapeutics, Inc., 389 Nichol Mill Lane, Franklin, Tennessee, 2Bolder BioPATH, Inc., Boulder, Colorado, 3SportsMedicine and Shoulder Service, Hospital for Special Surgery, New York, New York, 4Departments of Biomedical Engineering and BiologicalSciences, Bone Tissue Engineering Center, Carnegie Mellon University, Pittsburgh, Pennsylvania, 5Biomechanics & Bioengineering ResearchLaboratory, Feinstein Institute for Medical Research, North Shore Long Island Jewish Health System, Manhasset, New York

Received 29 March 2012; accepted 7 August 2012

Published online in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jor.22222

ABSTRACT: The purpose of this study was to assess whether intra-tendon delivery of recombinant human platelet-derived growthfactor-BB (rhPDGF-BB) would improve Achilles tendon repair in a rat collagenase-induced tendinopathy model. Seven days followingcollagenase induction of tendinopathy, one of four intra-tendinous treatments was administered: (i) Vehicle control (sodium acetatebuffer), (ii) 1.02 mg rhPDGF-BB, (iii) 10.2 mg rhPDGF-BB, or (iv) 102 mg rhPDGF-BB. Treated tendons were assessed for histopatholog-ical (e.g., proliferation, tendon thickness, collagen fiber density/orientation) and biomechanical (e.g., maximum load-to-failure and stiff-ness) outcomes. By 7 days post-treatment, there was a significant increase in cell proliferation with the 10.2 and 102 mg rhPDGF-BB-treated groups (p ¼ 0.049 and 0.015, respectively) and in thickness at the tendon midsubstance in the 10.2 mg of rhPDGF-BB group(p ¼ 0.005), compared to controls. All groups had equivalent outcomes by Day 21. There was a dose-dependent effect on the maximumload-to-failure, with no significant difference in the 1.02 and 102 mg rhPDGF-BB doses but the 10.2 mg rhPDGF-BB group had a signifi-cant increase in load-to-failure at 7 (p ¼ 0.003) and 21 days (p ¼ 0.019) compared to controls. The rhPDGF-BB treatment resultedin a dose-dependent, transient increase in cell proliferation and sustained improvement in biomechanical properties in a rat Achillestendinopathy model, demonstrating the potential of rhPDGF-BB treatment in a tendinopathy application. Consequently, in thismodel, data suggest that rhPDGF-BB treatment is an effective therapy and thus, may be an option for clinical applications to treattendinopathy. � 2012 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res

Keywords: platelet-derived growth factor-BB; tendinopathy; biomechanics; proliferation

Tendinopathy is defined as a painful condition thatdevelops in response to tendon degeneration1,2 and canaffect tendons throughout the body (e.g., rotator cuff,elbow, knee, foot, and ankle). Tendinopathy resultsin changes in tendon properties, including collagen de-generation, disorganization, increased mucoid groundsubstance, proliferation of capillaries and arterioles,loss of mechanical properties, and concomitant pain.In contrast to tendon ruptures, where partial or com-plete tearing of the tendon occurs, the degenerationand pain of tendinopathy occurs without macroscopictearing of the tendon. These degenerative changesmay result from a failed healing response to sub-failure injuries.1–3

Several treatment options for tendinopathy havebeen investigated with varying degrees of successincluding, but not limited to, exercise-based physicaltherapy, corticosteroid injections, non-steroidal anti-inflammatory drugs, extracorporeal shock wave thera-py, and surgical intervention.4–11 These therapies of-ten treat the symptoms associated with tendinopathy,but do not promote structural repair of the tendon andtherefore the tendinopathy persists or reoccurs.While not considered a gold-standard treatment, corti-costeroid injections are often administered for chronic

tendinopathies. However, there are concerns about thelong-term safety and effectiveness of this therapy8–10

which are supported by the observations of Jobe andCiccotti that corticosteroid injections may lead to ad-verse changes within the tendon structure.12 Conse-quently, alternative therapies have been investigatedto address the underlying pathology and improve long-term outcomes for tendinopathy.

Growth factors have also been assessed to promotetendon healing. Autologous growth factors in the formof platelet-rich plasma (PRP) have been studied; how-ever variability in the preparation and compositionof PRP makes it difficult to compare results acrossstudies.13 Additionally, variable clinical outcomesfollowing PRP treatment for tendinopathy have beenreported.14,15 As an alternative to PRP, recombinanthuman platelet-derived growth factor-BB (rhPDGF-BB) is a readily available, off-the-shelf option that pro-vides a consistent, therapeutic dose. Utilizing a varietyof delivery methods, animal models of tendon injuryhave shown that rhPDGF-BB accelerates tendonhealing by improving range of motion, mechanicalstrength, matrix remodeling, increased collagen syn-thesis, and increased cell proliferation,16–22 althoughthese all used rupture models. The impact of rhPDGF-BB on tendon healing in a non-ruptured, degenerated,tendinopathy model has not been assessed previously.In contrast to corticosteroids, rhPDGF-BB may ad-dress chronic tendinopathies by inducing proliferationand migration of progenitor cells and tenocytes,17,23–25

Grant sponsor: BioMimetic Therapeutics, Inc.Correspondence to: Christopher K. Hee (T: 615-236-4949; F: 615-236-4864; E-mail: chee@biomimetics.com)

� 2012 Orthopaedic Research Society. Published by Wiley Periodicals, Inc.

JOURNAL OF ORTHOPAEDIC RESEARCH 2012 1

which stimulate structural repair of the degeneratedtendon, as opposed to only addressing short-termsymptoms.

As a consequence of the deficiencies in clinical out-come offered by contemporary therapies, the objectiveof the study was to determine dose–response efficacyof an intra-tendon application of rhPDGF-BB in a ratcollagenase-induced tendinopathy model. We hypothe-sized that intra-tendon delivery of rhPDGF-BB wouldpromote Achilles tendon repair by up-regulating cellproliferation and restoring biomechanical strength.

METHODSExperimental ProtocolThis study was approved by the Institutional Animal Careand Use Committee at Bolder BioPATH, Inc. Male Sprague-Dawley rats weighing approximately 315 g (Charles RiverLabs, St. Constant, Quebec) were used. Rats were housedfour per cage under a 12:12-h light–dark cycle and food andwater were provided ad libitum. A diagram of the studydesign is shown in Figure 1. On Day 7, rats receivedan injection of collagenase (50 ml of 10 mg/ml Type IA inPBS, pH 7.4, 469 collagen digestion units/mg; Sigma, St.Louis, MO) into the right Achilles tendon near the osseo-tendinous junction using insulin syringes with 28.5G needleas described previously.26–30 Preliminary studies were per-formed to confirm that a percutaneous injection (collagenaseand treatment) could reliably deliver the material to the ten-don body (data not shown). Seven days following the collage-nase injection (Day 0), rats were randomized to one of fourtreatment groups (n ¼ 15/time point/group): (i) 20 mM sodi-um acetate buffer (vehicle control); (ii) 1.02 mg rhPDGF-BB;(iii) 10.2 mg rhPDGF-BB; or (iv) 102 mg rhPDGF-BB. Theseconcentrations, adjusted for animal size, are consistent withother investigations of rhPDGF-BB in tendon rupture heal-ing.22,31 Animals received an injection volume of 30 ml oftreatment into the damaged Achilles tendon. Rats were anes-thetized with isoflurane during the collagenase and subse-quent treatment injections. Rats were euthanized 7 or21 days following the rhPDGF-BB treatments. Additionalanimals (n ¼ 15) were euthanized on the day of treatment(Day 0) to establish the baseline damage due to the collage-nase treatment.

HistologyThe intact calcaneous–Achilles tendon–muscle specimen(n ¼ 6/time point/group), with skin removed, was fixed in10% neutral buffered formalin, decalcified in 10% formicacid, and the ankle trimmed medially and laterally to pre-serve the central portion of ankle with tendon attachment.Specimens were embedded in paraffin and serial parasagittalsections (5 mm) were taken. Sections were stained with he-matoxylin and eosin (H&E), Masson’s trichrome, or immuno-histochemical (IHC) staining for proliferating cell nuclearantigen (PCNA).

HistopathologyHistopathological assessment of inflammation, vasculariza-tion, collagen fiber organization/density were graded maskedto treatment by a board-certified veterinary pathologistaccording to the criteria listed in Table 1. Proliferating cellswithin the tendon, identified by PCNA immunoreactivity,were quantified using three consistent fields of an ocularmicrometer (39.4 � 197 mm; 7762 mm2). PCNA positive cellcounts were normalized by the area of the field of view.Microscopic tendon thickness was measured with an ocularmicrometer at the calcaneal attachment point and at themidsubstance of the tendon.

Biomechanical TestingThe calcaneous–Achilles tendon–muscle specimens (n ¼ 9/time point/group) were stored at �208C until the day of test-ing. Contralateral control specimens (n ¼ 9) were taken fromthe Day 21 vehicle control group. Mechanical testing wasperformed masked to treatment on an Instron testing frame(Model 5566) with samples submersed in a PBS bath. Speci-mens were mounted in hydraulic grips with roughened sur-face plates and sand paper to prevent slippage. Sampleswere tested using a 100 N load cell (load accuracy: �0.5%;sampling frequency: 10 Hz). A pre-load of 0.1 N was applied,the dimensions of the specimen at the midpoint of the gagelength were recorded using digital calipers, and then cyclicloading between 0.1 and 1 N (10 cycles at 0.133 N/s). A uni-axial tensile ramp-to-failure was then applied to the samplesunder displacement control at a strain rate of 0.1%/s untilfailure (measured load <0.05 N).31,32 Maximum load-to-fail-ure was recorded and stiffness was determined using the lin-ear slope of the load–displacement curve. Ultimate tensilestress (UTS) and modulus were determined from the stress–strain curve.

Statistical AnalysisStatistical analyses were performed using SigmaPlot 12.0.Parametric data (proliferating cell number, tendon thickness,maximum load-to-failure, stiffness) were analyzed with atwo-way ANOVA with Fisher’s LSD post hoc test to deter-mine the effect of time and treatment. Non-parametric data(histopathology scores) were analyzed using a Kruskal–Wallis test. All results are presented as mean � SEM. Thelevel of significance was set at p < 0.05.

RESULTS

Effect of rhPDGF-BB on Cell ProliferationAnimals that received 10.2 mg and 102 mg rhPDGF-BB exhibited a 1.6- and 1.7-fold increase in the num-ber of proliferating cells as compared to the vehicleFigure 1. Schematic diagram of the experimental design.

2 SHAH ET AL.

JOURNAL OF ORTHOPAEDIC RESEARCH 2012

controls at 7 days post-treatment (p ¼ 0.049 and0.015, respectively; Fig. 2). By Day 21, the proliferat-ing cell number in the 10.2 and 102 mg rhPDGF-BB-treated animals equaled the vehicle control level(Fig. 2). Cell proliferation in the 1.02 mg rhPDGF-BBdose group was not significantly increased relative tovehicle control at either 7 or 21 days post-treatment(Fig. 2).

Effect of rhPDGF-BB on Microscopic Tendon ThicknessRats treated with 10.2 mg rhPDGF-BB at 7 days post-treatment trended towards an increase in the tendonthickness at the calcaneal insertion (p ¼ 0.065) andhad a significant increase (1.97-fold; p ¼ 0.005) inthickness at the midsubstance compared to the vehiclecontrol group (Table 2). No significant differences wereobserved in the 1.02 or 102 mg rhPDGF-BB groupsrelative to vehicle controls. At 21 days post-treatmentthere were no significant differences in tendon thick-nesses among the groups, with the 10.2 mg rhPDGF-BB group returning to control levels (Table 2).

HistopathologyHistopathology scores are presented as the medianand range of six specimens evaluated by a singlemasked histopathologist. Baseline histology (day ofrhPDGF-BB treatment) had moderate inflammationand vascularization, disorganization and degenerationof the collagen (Table 3, Fig. 3). No differences in his-topathology scores were observed among groups at 7and 21 days, although there was a decrease in the me-dian inflammation score and increase in the mediancollagen organization/density scores relative to base-line (Table 3). There was neither abnormal bonegrowth nor resorption at the treatment sites for ratsthat received compositions either with or withoutrhPDGF-BB (Fig. 3). Mild to moderate inflammation,which was mononuclear in nature and transient,was observed at 7 days post-treatment in all groups,but was minimal at 21 days. At 7 and 21 days post-treatment with rhPDGF-BB, fibroplasia was concen-trated mainly in the tendon and the morphology of thecells was fibroblastic in nature.

Table 1. Description of the Histopathology Scoring System for Inflammation, Vascularization, Collagen FiberOrientation, and Collagen Fiber Density

Parameter Score Description

Inflammation 0 No inflammation1 Minimal inflammation (100% mononuclear; no neutrophils)2 Moderate inflammation (neutrophils �19%; remainder of cells mononuclear)3 Marked inflammation (neutrophils �20%; remainder of cells mononuclear)

Vascularization 0 None (no vascularization present)1 Moderate2 Abundant

Collagen Fiber Orientation 0 Collagen fibrils completely disorganized1 Some alignment of collagen fibrils (majority of collagen highly disorganized)2 Collagen fibrils highly aligned (collagen remains somewhat disorganized)3 Collagen fibrils completely aligned (no disorganization of collagen)

Collagen Fiber Density 0 Low density collagen bundles1 Medium density collagen bundles2 High density collagen bundles

Figure 2. Positive PCNA staining (dark brown nuclei) of (A) Baseline (Day 0 post-treatment), (B, F) 0 mg (vehicle control), (C,G)1.02 mg, (D, H) 10.2 mg, and (E, I) and 102 mg rhPDGF-BB-treated specimens at 7 (B–E) and 21 (F–I) days post-treatment indicatescells undergoing proliferation. Nuclei were counterstained with hematoxylin (purple). Values indicate the mean PCNA positive count �SEM. �p < 0.05 compared to vehicle control.

rhPDGF-BB IN A TENDINOPATHY MODEL 3

JOURNAL OF ORTHOPAEDIC RESEARCH 2012

Biomechanical StrengthStructural biomechanical properties (maximum load-to-failure and stiffness) are presented in Table 4 andmaterial biomechanical properties (UTS, modulus, andcross-sectional area (CSA)) are presented in Table 5.Specimen failure occurred in the tendinous region, notat the tendon–calcaneous interface. The baseline (Day0) values for the maximum load-to-failure and stiffnesswere decreased by 44% and 29%, respectively, com-pared to the contralateral controls (from Day 21 vehi-cle control animals). Maximum load-to-failure valuesincreased over time, with the baseline samples signifi-cantly lower than the 7- and 21-day specimens(p < 0.01). The mean stiffness did not increase duringthe course of the experiment. The maximum load-to-failure for the 10.2 mg rhPDGF-BB-treated group at7 days was significantly higher (143%; p ¼ 0.003) thanthe vehicle control group. Rats treated with 1.02 and102 mg rhPDGF-BB had a maximum load-to-failuresimilar to the vehicle controls. The maximum load-to-failure for the 10.2 mg rhPDGF-BB treatment group at21 days remained significantly increased (p ¼ 0.019)compared to the vehicle controls. The maximum load-to-failure in the 1.02 and 102 mg rhPDGF-BB groupswas similar to the vehicle controls. There were no sig-nificant differences in the stiffness values at either 7or 21 days among groups; however, the 1.02 and10.2 mg rhPDGF-BB groups, on average, had higherstiffness relative to the vehicle control (Table 4). The

baseline (Day 0) values for the UTS and modulus weredecreased by 46% and 33%, respectively, compared tothe contralateral controls. CSA was significantly in-creased on the Day 7 and 21 compared to the baseline(Day 0) group (p < 0.01). The CSA was significantlydecreased in the 1 mg group relative to the vehicle con-trol group (p < 0.01) on Day 21. There was no signifi-cant difference in UTS among groups at 7 days. TheUTS for the 1.02 mg rhPDGF-BB group at 21 days wassignificantly higher (175%; p ¼ 0.017) than the vehiclecontrol group. Rats treated with 10.2 and 102 mgrhPDGF-BB had a UTS similar to the vehicle controls.There were no significant differences among groups inthe modulus values at either 7 or 21 days.

DISCUSSIONThe results of this study suggest that rhPDGF-BB pro-moted cellular proliferation and improved mechanicalproperties in a dose-dependent manner. Moreover,there was an increase in early cellular proliferation(Day 7) and improved structural strength in the10.2 mg rhPDGF-BB-treated group. Additionally, his-tology suggested that a single, bolus dose of up to100 mg of rhPDGF-BB into the Achilles tendon did notresult in adverse effects on bone (e.g., osteolysis, ectop-ic intra-tendon calcification) or tendon (e.g., degenera-tion) due to rhPDGF-BB treatment. Furthermore, thedata presented in this study support the safety andutility of rhPDGF-BB intra-tendon delivery for tendon

Table 2. Microscopic Tendon Thickness (mm; n ¼ 6) at the Calcaneal Insertion and at the Tendon Midsubstance) atBaseline, 7, and 21 Days Post-Treatment With or Without rhPDGF-BB. Values are Means � SEM

rhPDGF-BB(mg)

Thickness at calcaneal insertion (mm) Thickness at tendon midsubstance (mm)

Days post-treatment Days post-treatment

0 7 21 0 7 21

0 0.5 � 0.1 0.7 � 0.1 0.6 � 0.2 1.6 � 0.3 2.2 � 0.6 2.1 � 0.61.02 0.9 � 0.2 0.6 � 0.2 2.5 � 0.5 1.6 � 0.710.2 1.6 � 0.2 1.1 � 0.5 4.4 � 0.2� 2.0 � 0.5102 1.0 � 0.3 1.0 � 0.5 2.9 � 0.4 2.2 � 0.6

�p < 0.05 relative to vehicle control.

Table 3. Histopathology Scores for Inflammation, Vascularization, Collagen Fiber Orientation, and Collagen FiberDensity at Baseline, 7, and 21 Days Post-Treatment With or Without rhPDGF-BB

rhPDGF-BB(mg)

Inflammation Vascularization Collagen fiber orientation Collagen fiber density

Days post-treatment Days post-treatment Days post-treatment Days post-treatment

0 7 21 0 7 21 0 7 21 0 7 21

0 2 [2] 1 [1–2] 1 [1–2] 1.5 [0–2] 1.5 [1–2] 1 [1–2] 1 [1–3] 3 [2–3] 2.5 [2–3] 1 [1–2] 2 [1–2] 1.5 [1–2]1.02 2 [1–2] 1 [1] 2 [1–2] 1 [0–2] 2 [2–3] 3 [2–3] 1 [1–2] 2 [2]10.2 2 [2] 1 [1] 2 [2] 1 [1–2] 2 [2] 3 [2–3] 1 [1] 2 [2]102 2 [1–2] 2 [1–2] 2 [1–2] 1.5 [1–2] 2.5 [2–3] 3 [2–3] 2 [1–2] 2 [1–2]

Values are median [range] for n ¼ 6 specimens.

4 SHAH ET AL.

JOURNAL OF ORTHOPAEDIC RESEARCH 2012

repair in the rat collagenase-induced Achilles tendin-opathy model.

In this study, a collagenase-induced model oftendinopathy was used. Because rhPDGF-BB is alsoknown to stimulate chemotaxis of neutrophils andmacrophages and promote the production and secre-tion of other growth factors by macrophages,33 a 7-daywashout period following collagenase treatment wasincluded prior to treatment in order to allow the initialinflammatory response to the collagenase to dissi-pate.26 The baseline results following this washoutperiod were evaluated to assess the early inflammato-ry response to the collagenase and the subsequentresponse to rhPDGF-BB treatment. Additionally, thepost-repair loading was not controlled in this study,compared to the regimens of loading protection pre-scribed for the human condition. The lack of control ofpost-repair loading creates a challenging environmentfor repair which would represent a non-compliantpatient, which can present clinical challenges that aredifficult to overcome with current therapies.

Tendinopathy is hypothesized to be the result of in-sufficient healing response to sub-failure injuries.1–3

The low cellular density in tendons may be a

contributing factor which leads to relatively poor heal-ing following injury, suggesting that recruitment andproliferation of tenocytes to the injured tendon mayimprove the conditions for repair.34 Therefore, thisstudy investigated the chemotactic and mitogenic fac-tor, rhPDGF-BB, as a therapy to increase the cellularresponse and ultimately improve the mechanical prop-erties of the healing tendon. Treatment of tendinop-athy with rhPDGF-BB (10.2 and 102 mg doses)stimulated early cellular proliferation of cells with afibroblastic morphology within the tendon (at 7 days)followed by a return to vehicle control levels at21 days. Given their location and morphology it wasassumed that these cells were tenocytes or fibroblasticprecursors, although their origin (extrinsic vs. intrin-sic) was not determined in this study. The decrease inproliferation from 7 to 21 days suggests a transitionfrom the proliferative phase of tendon healing to theremodeling phase, although other markers indicativeof this transition (decreased collagen and glycosamino-glycan synthesis) were not measured in this study.1,34

In addition to proliferation, an early increase in thetendon thickness in the 10.2 mg group was observed.This significant increase, observed microscopically,

Figure 3. Masson’s trichrome staining of the ankle joint of (A) Baseline (Day 0 post-treatment), (B, F) 0 mg (vehicle control), (C,G)1.02 mg, (D, H) 10.2 mg, and (E, I) and 102 mg rhPDGF-BB-treated specimens at 7 (B–E) and 21 (F–I) days post-treatment. Arrowshighlight the Achilles tendon.

Table 4. Structural Biomechanical Properties (Maximum Load-to-Failure (N) and Stiffness Values (N/mm); n ¼ 9) atBaseline, 7, and 21 Days Post-Treatment With or Without rhPDGF-BB

rhPDGF-BB(mg)

Maximum load-to-failure (N) Stiffness (N/mm)

Days post-treatment Days post-treatment

0 7 21 0 7 21

0 15.7 � 1.8 18.7 � 1.9 22.0 � 1.8 8.9 � 1.2 9.6 � 1.5 9.9 � 1.11.02 22.7 � 1.9 23.7 � 1.9 10.1 � 1.2 12.4 � 2.210.2 26.8 � 2.0� 28.6 � 1.5� 10.2 � 0.7 10.1 � 0.9102 17.1 � 1.9 21.0 � 1.9 9.4 � 1.2 9.2 � 1.6

Contralateral controls: 28.0 � 1.4 N Contralateral controls: 12.6 � 1.2 N/mm

Values are means � SEM.�p < 0.05 relative to vehicle control.

rhPDGF-BB IN A TENDINOPATHY MODEL 5

JOURNAL OF ORTHOPAEDIC RESEARCH 2012

was not observed in macroscopic measurements (CSA).This difference may be due to a number of factors,such as a histological processing effect on tissuedimensions, accuracy of the measurement techniques(calipers vs. microscopic imaging), or variations in themeasurement location between techniques. Steps weretaken to minimize these effects. For instance, the his-tology embedding protocol was standardized to mini-mize processing effects (shrinkage due to fixation ordehydration). Additionally, the repeatability of the cal-iper CSA measurement was found to range from 4% to12%, based on group, using the average of two meas-urements perfumed at the tendon midsection afterpre-tension had been applied to the specimen to reducepotential variability introduced by deformation of thesample with the calipers. Regardless, the change inthe microscopic thickness of the tendon tissue indi-cates increased biochemical activity of the tenocytes,which would contribute to the measured early struc-tural mechanical strength. A return to the controltendon thickness at 21 days, combined with the struc-tural and material mechanical properties, furthersuggests remodeling of the tendon occurred whilemaintaining the overall ability to support a load. Im-portantly, the decrease in proliferation and tendonthickness at 21 days indicates rhPDGF-BB did notpromote a sustained hyperproliferative state or uncon-trolled tissue growth. These outcomes are noteworthyfrom a safety perspective.

The tendon’s primary role as a mechanical link be-tween muscle and bone emphasizes the importance ofre-establishing tendon function for a successful repair.The results of this study indicate that a 10.2 mgrhPDGF-BB dose initiated a faster repair response,significantly increasing the maximum load-to-failureover the vehicle control group at 7 and 21 days. Fur-ther, the 10.2 mg rhPDGF-BB group reached 96% and102% of the contralateral controls at 7 and 21 days,respectively. Additionally, the material propertiesin the 1.02 mg rhPDGF-BB group were improvedat Day 21 (90% of contralaterals). Future studies withextended follow-up time points would help define the

continued durability of the repair and determine ifthere was eventually normalization of the mechanicalproperties among all groups. In contrast to the cellproliferation and tendon thickness data, the bio-mechanical strength persisted over the 21-day studyindicating that rhPDGF-BB treatment produced a last-ing repair, validating that a single rhPDGF-BB admin-istration will promote improved, sustained functionof the tendon. This outcome reinforces the potentialclinical benefit of rhPDGF-BB for treatment oftendinopathy.

In addition, this study determined the dose–response for optimizing treatment in a rat Achillestendinopathy model. A single, locally delivered 10.2 mgdose of rhPDGF-BB induced cellular proliferationand increased structural biomechanical strength. Al-though the half-life of rhPDGF-BB is short (<1 h) insystemic circulation, mechanisms including clearanceby a2-macroglobulin and binding of PDGF-BB to extra-cellular matrix proteins35–37 modulate the local tissueconcentrations. Combined with the transient up-regu-lation of proliferative response, this suggests that thelocal tissue concentrations of rhPDGF-BB available toelicit chemotactic and mitogenic effects likely remainelevated through the 7 day time point, correspondingwith the proliferative phase of tendon healing. A dose-dependent effect was observed, with the 1.02 mg and10.2 mg doses resulting in improved biomechanicalproperties. The proliferation in the 1.02 mg treatmentgroup was similar to the vehicle controls, suggestingthis dose was below the threshold for inducing cellularactivity for tendon repair, however the increase in theUTS at 21 days indicates that rhPDGF-BB was effica-cious for improving the material properties of the tis-sue, even without a detected increase in proliferation.In contrast, the 102 mg rhPDGF-BB dose significantlyincreased early cellular proliferation, although this didnot result in a corresponding improvement in bio-mechanical properties. Differences in the tendon thick-ness measurements at 7 days may provide additionalinsight into the mechanism behind this disparity. Wesurmise that the tenocytes in the 10.2 mg group were

Table 5. Material Biomechanical Properties (Ultimate Tensile Stress (MPa), Modulus (MPa), and Cross-SectionalArea (mm2)); n ¼ 9) at Baseline, 7, and 21 Days Post-Treatment With or Without rhPDGF-BB

rhPDGF-BB(mg)

Ultimate tensile stress (MPa) Modulus (MPa) Cross-sectional area (mm2)

Days post-treatment Days post-treatment Days post-treatment

0 7 21 0 7 21 0 7 21

0 2.1 � 0.3 2.5 � 0.4 2.0 � 0.2 9.1 � 1.2 9.6 � 1.5 5.5 � 1.1 7.8 � 0.9 8.4 � 0.7 12.2 � 0.61.02 2.7 � 0.4 3.5 � 0.4� 8.4 � 1.2 12.6 � 2.2 9.2 � 0.7 7.5 � 0.410.2 3.0 � 0.4 2.6 � 0.3 8.0 � 0.7 7.3 � 0.9 9.6 � 0.9 10.8 � 1.3102 1.9 � 0.3 2.1 � 0.2 7.7 � 1.2 6.1 � 1.6 9.7 � 0.9 11.7 � 1.4

Contralateral controls:3.9 � 0.3 MPa

Contralateral controls:13.6 � 1.7 MPa

Contralateral controls:7.41 � 0.3 mm2

Values are means � SEM.�p < 0.05 relative to vehicle control.

6 SHAH ET AL.

JOURNAL OF ORTHOPAEDIC RESEARCH 2012

more actively producing matrix and remodeling thetendon, thus contributing to the increase in strengthin the 10.2 mg group. Although the mechanisms arenot fully understood, the biphasic dose dependence inresponse to rhPDGF-BB treatment of tendons is con-sistent with previous studies.16,22 While it is beyondthe scope of this study, we speculate the mechanismsinvolved in the dose-dependent effect include promo-tion of cellular turnover and biochemical activity lead-ing to restoration of extracellular matrix compositionand structure. Regardless of the mechanism, suchobservations regarding dose-dependence underscorethe impact of establishing the optimal dose for futureclinical use.

In conclusion, the improved biomechanical strengthof the rhPDGF-BB-treated tendons determined in thereported study support the notion that rhPDGF-BB iseffective for stimulating repair in the rat tendinopathymodel. Future investigations will examine the clinicalimpact of rhPDGF-BB on pain abrogation and restora-tion of activity following treatment.

ACKNOWLEDGMENTSWe would like to thank Dr. Dean J. Aguiar (assistance withstudy design and interpretation of the data), Dr. YanchunLiu (aid with study execution), Jack Ratliff (histology sup-port), Dr. Luis A. Solchaga (manuscript review), and BrianOmura, Mike Bendele, and Scott Boswell (assistance duringthe study). The author’s have potential sources of conflict ofinterest due to the following relationships with BioMimeticTherapeutics, Inc.: V. Shah (employee, stock/options), A.Bendele (institutional funds to perform aspects of the study),J.S. Dines (Scientific Advisory Board, consultant), H.K. Kes-tler (employee, stock/options), J.O. Hollinger (Scientific Advi-sory Board, consultant), N.O. Chahine (institutional funds toperform aspects of the study), and C.K. Hee (employee, stock/options).

REFERENCES1. Sharma P, Maffulli N. 2005. Tendon injury and tendinop-

athy: healing and repair. J Bone Joint Surg Am 87:187–202.2. Maffulli N, Wong J, Almekinders L. 2003. Types and epide-

miology of tendinopathy. Clin Sports Med 22:675–692.3. Battery L, Maffulli N. 2011. Inflammation in overuse tendon

injuries. Sports Med Arthrosc 19:213–217.4. Buchbinder R, Green S, Bell SN, et al. 2009. Surgery for

lateral elbow pain. Cochrane Database Syst Rev 1:1–14.5. Buchbinder R, Green S, Youd JM, et al. 2005. Shock wave

therapy for lateral elbow pain. Cochrane Database Syst Rev4:CD003524.

6. Green S, Buchbinder R, Barnsley L, et al. 2002. Non-steroidal anti-inflammatory drugs (NSAIDs) for treatinglateral elbow pain in adults. Cochrane Database Syst Rev2:CD003686.

7. Smidt N, Assendelft WJJ, Arola H, et al. 2003. Effectivenessof physiotherapy for lateral epicondylitis: a systematic re-view. Ann Med 35:51–62.

8. Smidt N, Assendelft WJJ, Van der Windt DAWM, et al.2002. Corticosteroid injections for lateral epicondylitis: a sys-tematic review. Pain 96:23–40.

9. Smidt N, van der Windt D, Assendelft W, et al. 2002. Corti-costeroid injections, physiotherapy, or a wait-and-see policy

for lateral epicondylitis: a randomised controlled trial. Lan-cet 359:657–662.

10. Gosens T, Peerbooms J, van Laar W, et al. 2011. Ongoingpositive effect of platelet-rich plasma versus corticosteroidinjection in lateral epicondylitis. A double-blind randomizedcontrolled trial with 2-year follow-up. Am J Sports Med39:1200–1208.

11. Andres B, Murrell G. 2008. Treatment of tendinopathy. ClinOrthop Rel Res 466:1539–1554.

12. Jobe F, Ciccotti M. 1994. Lateral and medial epicondylitis ofthe elbow. J Am Acad Orthop Surg 2:1–8.

13. Castillo T, Pouliot M, Kim H, et al. 2011. Comparison ofgrowth factor and platelet concentration from commercialplatelet-rich plasma separation systems. Am J Sports Med39:266–271.

14. de Vos RJ, van Veldhoven PLJ, Moen MH, et al. 2010. Autol-ogous growth factor injections in chronic tendinopathy: asystematic review. Br Med Bull 95:63–77.

15. Del Buono A, Papalia R, Denaro V, et al. 2011. Platelet richplasma and tendinopathy: state of the art. Int J Immunopa-thol Pharmacol 24:79–83.

16. Chan BP, Fu SC, Qin L, et al. 2006. Supplementation-timedependence of growth factors in promoting tendon healing.Clin Orthop Relat Res 448:240–247.

17. Thomopoulos S, Zaegel M, Das R, et al. 2007. PDGF-BB re-leased in tendon repair using a novel delivery system pro-motes cell proliferation and collagen remodeling. J OrthopRes 25:1358–1368.

18. Thomopoulos S, Das R, Silva MJ, et al. 2009. Enhanced flex-or tendon healing through controlled delivery of PDGF-BB.J Orthop Res 27:1209–1215.

19. Gelberman R, Thomopoulos S, Sakiyama-Elbert S, et al.2007. The early effects of sustained platelet-derived growthfactor administration on the functional and structural prop-erties of repaired intrasynovial flexor tendons: an in vivobiomechanic study at 3 weeks in canines. J Hand Surg Am32:373–379.

20. Uggen JC, Dines J, Uggen CW, et al. 2005. Tendon genetherapy modulates the local repair environment in theshoulder. J Am Osteopath Assoc 105:20–21.

21. Uggen C, Dines J, McGarry M, et al. 2010. The effect of re-combinant human platelet-derived growth factor BB-coatedsutures on rotator cuff healing in a sheep model. Arthrosco-py 26:1456–1462.

22. Hee CK, Dines JS, Dines DM, et al. 2011. Augmentation of arotator cuff suture repair using rhPDGF-BB and a type Ibovine collagen matrix in an ovine model. Am J Sports Med39:1630–1639.

23. Costa MA, Wu C, Pham BV, et al. 2006. Tissue engineeringof flexor tendons: optimization of tenocyte proliferation usinggrowth factor supplementation. Tissue Eng 12:1937–1943.

24. Thomopoulos S, Harwood F, Silva M, et al. 2005. Effect ofseveral growth factors on canine flexor tendon fibroblast pro-liferation and collagen synthesis in vitro. J Hand Surg Am30:441–447.

25. Sakiyama-Elbert SE, Das R, Gelberman RH, et al. 2008.Controlled-release kinetics and biologic activity of platelet-derived growth factor-BB for use in flexor tendon repair.J Hand Surg Am 33:1548–1557.

26. Marsolais D, Cote CH, Frenette J. 2001. Neutrophils andmacrophages accumulate sequentially following Achilles ten-don injury. J Orthop Res 19:1203–1209.

27. Wetzel B, Nindl G, Swez J, et al. 2002. Quantitative charac-terization of rat tendinitis to evaluate the efficacy of thera-peutic interventions. Biomed Sci Instrum 38:157–162.

28. Marsolais D, Cote C, Frenette J. 2007. Pifithrin-alpha, aninhibitor of p53 transactivation, alters the inflammatory

rhPDGF-BB IN A TENDINOPATHY MODEL 7

JOURNAL OF ORTHOPAEDIC RESEARCH 2012

process and delays tendon healing following acute injury.Am J Physiol Regul Integr Comp Physiol 292:R321–R327.

29. Chen Y-J, Wang C-J, Yang KD, et al. 2004. Extracorporealshock waves promote healing of collagenase-induced Achillestendinitis and increase TGF-B1 and IGF-I expression.J Orthop Res 22:854–861.

30. Godbout C, Ang O, Frenette J. 2006. Early voluntary exer-cise does not promote healing in a rat model of Achilles ten-don injury. J Appl Physiol 101:1720–1726.

31. Cummings SH, Grande DA, Hee CK, et al. 2012. Effect ofrhPDGF-BB-coated sutures on achilles tendon healing in arat model: a histological and biomechanical study. J TissueEng 3:2041731412453577.

32. Dines JS, Weber L, Razzano P, et al. 2007. The effectof growth differentiation factor-5-coated sutures on tendonrepair in a rat model. J Shoulder Elbow Surg 16:S215–S221.

33. Heldin C-H, Westermark B. 1999. Mechanism of action andin vivo role of platelet-derived growth factor. Physiol Rev79:1283–1316.

34. James R, Kesturu G, Balian G, et al. 2008. Tendon: biology,biomechanics, repair, growth factors and evolving treatmentoptions. J Hand Surg Am 33:102–112.

35. Solchaga LA, Hee CK, Roach S, et al. 2012. Safety of recom-binant human platelet-derived growth factor-BB in Aug-ment(1) Bone Graft. J Tissue Eng 3:2041731412442668.

36. Martino MM, Hubbell JA. 2010. The 12th–14th type IIIrepeats of fibronectin function as a highly promiscuousgrowth factor-binding domain. FASEB J 24:4711–4721.

37. Gohring W, Sasaki T, Heldin CH, et al. 1998. Mapping ofthe binding of platelet-derived growth factor to distinctdomains of the basement membrane proteins BM-40 andperlecan and distinction from the BM-40 collagen-bindingepitope. Eur J Biochem 255:60–66.

8 SHAH ET AL.

JOURNAL OF ORTHOPAEDIC RESEARCH 2012