Autologous Platelets Have No Effect on the Healing of Human Achilles Tendon Ruptures

12
Autologous Platelets Have No Effect on the Healing of Human Achilles Tendon Ruptures A Randomized Single-Blind Study Thorsten Schepull,* MD, Joanna Kvist, 1 " PhD, RPT, Hanna Norrman, 1 " RPT; Marie Trinks,* BS, Gosta Berlin,* MD, PhD, and Per Aspenberg,*§ MD, PhD Investigation performed at Linkoping University, Linkoping, Sweden Background: An results from wn that local application of platelet-rich plasma (PRP) stimulates tendon repair. Preliminary retrospective case seria^^ave shown faster return to sports. Hypothesis: Autolog^ua_P^P_^tjrriulates healing of acute Achilles tendon ruptures. Study Design: Randomized controlledjjjal; Level of evidence, 2. Methods: T>irty-pa!Terrts were recruited cons^crrfjvidyHSuriflgjs^gery, tantalum beads were implanted in the Achilles tendon proximaj^and distal to the rupture. Before skin suture, randorrnzaiii^.was performed, and 16 patients were injected with 10 mL [PRP (10 times higher platelet concentration than peripheral blood) \yhereas 14 were not. With 3-dimensional radiographs (roentgeVstereophotogrammetric analysis; RSA), the distancebetvreejvHIe beads was measured at 7, 1 9, and 52 weeks while the patient mslsiSrTTtiffemHt-decsal flexion momeRte-e¥eririe"anl<l6"jolnt thereby estimating tendon strain per load. An estimate of elasticity modulus was calculated using callus dimensions from computed tomography^ At 1 year, functional outcomerwas eval- uated, including the heel raise index and Achilles Tendon Total Rupture Score. The primary effect variables were elasticity mod^ ulus at 7 weeks and heel raise index at 1 year. - --- . Results: I Tie mechanical variables showedalarge degree of variation between patients that could not be explained by measuring^^fc error. No significant group differences in elasticity modulus could be shown. There was no significant difference in heel ^^ index. The Achilles Tendon Total Rupture Score was lower in the PRP group, suggesting a detrimental effect. There was a cor- relation between the elasticity modulus at 7 and 19 weeks and the heel raise index at 52 weeks. Conclusion: The results suggest that PRP is not useful for treatment of Achilles tendon ruptures. The variation in elasticity mod- ulus provides biologically relevant information, although it is unclear how early biomechanics is connected to late clinical results. Keywords: Achilles tendon; platelet; platelet-rich plasma; tendon healing; biomechanics; roentgen stereophotogrammetric analysis Healing after an Achilles tendon rupture is a slow process. Many treatment methods have been tried to improve ten- don healing, with the aim of minimizing the risk of §Address correspondence to Per Aspenberg, Linkoping University Hospital, SE-581 85 Linkoping, Sweden (e-mail: [email protected]). 'Orthopedics, Department of Clinical and Experimental Medicine, IKE, Faculty of Health Sciences, Linkoping University, Linkoping, Sweden. tSection for Physiotherapy, Department of Medical and Health Scien- ces, Faculty of Health Sciences, Linkoping University, Linkoping, Sweden. *Department of Clinical Immunology and Transfusion Medicine, Faculty of Health Sciences, Linkoping University, Linkoping, Sweden. One or more authors has declared a potential conflict of interest: This study was supported by the Swedish Medical Research Council, the Swedish Center for Sports Medicine Research, and the King Gustav V and Queen Victoria Freemason Fund. The American Journal of Sports Medicine, Vol. 39, No. 1 DOI: 10.1177/Q363546510383515 ©2011 TheAuthor(s) rerupture and shortening the time until the patient can return to the level of activity before injury. Rerupture of the healing Achilles tendon can occur late, even 6 months after injury.9 In animal experiments, it has been possible to improve the tendon-healing process with mechanical loading1' 2 and pharmacological treatment.26 Another possi- bility is to use platelet-rich plasma (PRP).17'25 The physiol- ogy of tissue repair and scarring involves the release of a cocktail of bioactive proteins and growth factors from activated platelets. Although the efficacy of PRP in bone healing is debated,8 some studies have indicated that it may be better suited for stimulation of fibrous tissue repair. Rat models have shown that PRP increases tendon strength24 when injected locally. Increased vascularity and better tissue organization have been found in rabbit Achil- les tendons.11 Moreover, one clinical study comparing PRP- treated patients with historical controls indicated that pre- operative PRP treatment allows earlier return to sports,17 which has attracted the interest of sports physicians, 38

Transcript of Autologous Platelets Have No Effect on the Healing of Human Achilles Tendon Ruptures

Autologous Platelets Have No Effect on theHealing of Human Achilles Tendon Ruptures

A Randomized Single-Blind Study

Thorsten Schepull,* MD, Joanna Kvist,1" PhD, RPT, Hanna Norrman,1" RPT; Marie Trinks,* BS,Gosta Berlin,* MD, PhD, and Per Aspenberg,*§ MD, PhDInvestigation performed at Linkoping University, Linkoping, Sweden

Background: Anresults from

wn that local application of platelet-rich plasma (PRP) stimulates tendon repair. Preliminaryretrospective case seria^^ave shown faster return to sports.

Hypothesis: Autolog^ua_P^P_^tjrriulates healing of acute Achilles tendon ruptures.

Study Design: Randomized controlledjjjal; Level of evidence, 2.

Methods: T>irty-pa!Terrts were recruited cons^crrfjvidyHSuriflgjs^gery, tantalum beads were implanted in the Achilles tendonproximaj^and distal to the rupture. Before skin suture, randorrnzaiii .was performed, and 16 patients were injected with10 mL [PRP (10 times higher platelet concentration than peripheral blood) \yhereas 14 were not. With 3-dimensional radiographs(roentgeVstereophotogrammetric analysis; RSA), the distancebetvreejvHIe beads was measured at 7, 1 9, and 52 weeks while thepatient mslsiSrTTtiffemHt-decsal flexion momeRte-e¥eririe"anl<l6"jolnt thereby estimating tendon strain per load. An estimate ofelasticity modulus was calculated using callus dimensions from computed tomography^ At 1 year, functional outcomerwas eval-uated, including the heel raise index and Achilles Tendon Total Rupture Score. The primary effect variables were elasticity mod^ulus at 7 weeks and heel raise index at 1 year. • - --- .

Results: I Tie mechanical variables showedalarge degree of variation between patients that could not be explained by measuring^^fcerror. No significant group differences in elasticity modulus could be shown. There was no significant difference in heel ^^index. The Achilles Tendon Total Rupture Score was lower in the PRP group, suggesting a detrimental effect. There was a cor-relation between the elasticity modulus at 7 and 19 weeks and the heel raise index at 52 weeks.

Conclusion: The results suggest that PRP is not useful for treatment of Achilles tendon ruptures. The variation in elasticity mod-ulus provides biologically relevant information, although it is unclear how early biomechanics is connected to late clinical results.

Keywords: Achilles tendon; platelet; platelet-rich plasma; tendon healing; biomechanics; roentgen stereophotogrammetricanalysis

Healing after an Achilles tendon rupture is a slow process.Many treatment methods have been tried to improve ten-don healing, with the aim of minimizing the risk of

§Address correspondence to Per Aspenberg, Linkoping UniversityHospital, SE-581 85 Linkoping, Sweden (e-mail: [email protected]).

'Orthopedics, Department of Clinical and Experimental Medicine,IKE, Faculty of Health Sciences, Linkoping University, Linkoping,Sweden.

tSection for Physiotherapy, Department of Medical and Health Scien-ces, Faculty of Health Sciences, Linkoping University, Linkoping, Sweden.

*Department of Clinical Immunology and Transfusion Medicine, Facultyof Health Sciences, Linkoping University, Linkoping, Sweden.

One or more authors has declared a potential conflict of interest: Thisstudy was supported by the Swedish Medical Research Council, theSwedish Center for Sports Medicine Research, and the King Gustav Vand Queen Victoria Freemason Fund.

The American Journal of Sports Medicine, Vol. 39, No. 1DOI: 10.1177/Q363546510383515©2011 TheAuthor(s)

rerupture and shortening the time until the patient canreturn to the level of activity before injury. Rerupture ofthe healing Achilles tendon can occur late, even 6 monthsafter injury.9 In animal experiments, it has been possibleto improve the tendon-healing process with mechanicalloading1'2 and pharmacological treatment.26 Another possi-bility is to use platelet-rich plasma (PRP).17'25 The physiol-ogy of tissue repair and scarring involves the release ofa cocktail of bioactive proteins and growth factors fromactivated platelets. Although the efficacy of PRP in bonehealing is debated,8 some studies have indicated that itmay be better suited for stimulation of fibrous tissuerepair. Rat models have shown that PRP increases tendonstrength24 when injected locally. Increased vascularity andbetter tissue organization have been found in rabbit Achil-les tendons.11 Moreover, one clinical study comparing PRP-treated patients with historical controls indicated that pre-operative PRP treatment allows earlier return to sports,17

which has attracted the interest of sports physicians,

38

Vok 39, No. 1,2,011 Foot Orthoses intfi'i&Prevention of Injury 37

'31. Rauh: MJ, Macera CA, Trone DW, Shaffer RA, Brodine SK. Epidemi-ology of stress fracture and lower-extremity overuse injury in femalerecruits. Med Sci Sports Exerc. 2006;38(9):1 571 -1577.

32. RIddeir Qli- RehabjIitatiorT'of injured Royal Marine recruits. J R Nav'

33. Rome K, Gray J, Stewart F, Hannant SC, Callaghan D, Hubble J.Evaluating the clinical -effectiveness and cost-effectiveness of footorthoses in the treatment of plantar heel pain: a feasibility study. JAm Podiatr Med Assoc. 200.4;94(3):229-238.

34. Ross J. A review of lower limb overuse injuries during basic militarytraining: part 1. Types of overuse injuries. Mil Med. 1993;158(6):410-415.

35. Samet J, Chick T, Howard G. Running-related morbidity: a commu-nity survey. Annals Sports Med. 1982;1 :30-34.

36. Subotnick SI. The biornec'hanics of running; implications for the pre-vention of foot injuries. Sports Med. 1985;2(2):144-153.

37. Taunton JE, Ryan' MB, Clement DB, McKenzie DC, Lloyd-Smith DR,Zumbo BD. .A prospective study of running injuries: the VancouverSun Run *'Ih Training" clinics. Br J Sports Med. 2003;37(3):239-244.

38. Van Ginckel A, Thijs Y) Hesar NG, et al. Intrinsic gait-related risk fac-tors for Achilles tendiriopathy in novice runners: a prospective study.Gait Posture. 2009;29(3):387-391.

39. Van Middelkoop M, Ko[kraan J, Va'n Ochten J, Bierma-Zeinstra SM,Koes BW. Risk factors for Iswer extremity injuries among male mar-athon runners. Scand J Med Sci Sports. 2008;18(6):691-697.

40. WillemsTM, De Clercq D, Delbaere K, Vanderstraeten G, De Cook A,Witvrouw E. A prospective study of gait related risk factors for exer-cise-related lower leg pain. Gait Posture. 2006;23(1):91-98.

41. Willems TM, Witvrouw E, De Cook A, De Clercq D. Gait-related riskfactors for exercise-related lower-leg pain during shod running.Med So! Sports Exerc. 2007;39(2):330-339.

42. Withnall R, Eastaugh J, Freemantle N.'Do shook absorbing insolesin recruits undertaking high levels of physical activity reduce lowerlimb injury? A randomized controlled trial. J R Soc Med. 2006;99(1):32-37.

43. Yates B, White S. The incidence and risk factors in the developmentof medial tibial stress syndrome among naval recruits. Am J SportsMed. 2004;32(3):772-780. .

For reprints and pennission queries, please visit SAGE's Web site at http://www.sagepub.com/joumalsPermissions.riav

Vol. 39, No. 1, 2011 Autologous Platelets and Human Achilles Tendon Ruptures 39

especially as small centrifuges and practical kits are avail-able for production of PRP in the physician's office.

Regarding tendinosis, 2 important studies on PRP treat-ment have recently been published. Improved mechanicalproperties were found in a study on horses, where corelesions were treated with PRP, excised, and testedmechanically.5 However, a randomized double-blindplacebo-controlled clinical trial involving 54 tendinosispatients failed to show any beneficial effects on clinicaloutcome.6

In previous studies, we established the use of roentgenstereophotogrammetric analysis (RSA) with simultaneousmechanical loading as a method that can describe themechanical properties of a healing Achilles tendon.18 Wealso found that an estimate of the modulus of elasticity(E-modulus; Young's modulus, which describes an elasticproperty of the tissue) during early healing showed a corre-lation to the functional outcome at the 1-year follow-up. Wehave now used this method to measure the early mechan-ical properties of the healing tendon in a randomized trialto determine whether there is any positive effect of PRP ontendon repair. E-modulus at week 7 and the heel raiseindex at week 52 were chosen as the primary outcome var-iables because they appeared to be relevant and reasonablysensitive in a previous patient series.18

_/

21 mL (range, 16 to 29 mL) and a mean concentration of3673 ± 1051 X 109 platelets per mL.

The PRP was stored at 22°C with constant rotation forup to 20 hours before use. The swirling phenomenon(reflecting platelet viability) of the PRPs was examinedbefore use4 and found to be well maintained in all cases.

PATIENTS AND METHODS

All patients between 18 and 60 years of age presentingwith an acute rupture of the Achilles tendon (not olderthan 3 days) at our hospital were asked to participate inthe study after receiving oral and written information.Exclusion criteria were diabetes mellitus, a history of can-cer or lung or heart diseases, or any other diseases thatcould compromise the locomotory system. Between Sep-tember 2007 and April 2008, we included 30 consecutivepatients with an Achilles tendon rupture; only 1 patientrefused to participate in the study (see Figure 1). Allpatients were recreational athletes and were injured dur-ing sports or sports-related activities. The patients con-sented in writing, and the study was revised andapproved by the Regional Ethics Committee.

Preparation of Platelet Concentrate

Because randomization was planned to take place duringsurgery, autologous PRP had to be prepared from all 30patients. All patients were operated on within 5 days ofinjury; on the day before surgery, all were sent to theDepartment of Transfusion Medicine at Linkoping Univer-sity Hospital. OTTfijrnit nf whole blood (450 mL) was col-lected with citrate phosphate dextrose as anBcoagulant.

oiLblojjdJ'RP'wS obtainecLby double cen--according to accredited procedures.7 The plate-

let concentration was measured With the Micros 60hematology analyzer (ABX Diagnostics, Montpellier,France). The PRP preparations had a mean volume of

Operative Treatment and Randomization

Surgery was done using a conventional open techniquewith a dorsomedial approach. We used local anesthesiawith Carbocaine (mepivacaine hydroch^fiTTtie~HriagT7ffee^eiidoii were adapted with a resorbable

future CVicryl size 1) using the single-loop Kessler tech-nique. We implanted 2 tantalum beads (size, 0.8 mm) inthe distal part of the tendon and 2 tantalum beads of thesame size in the proximal part. We then closed the par-atenon. Then, a cannula was inserted into the rupturesite, and a randomization envelope was opened. In caseswhere a patient was allocated to PRP treatment, a syringewas filled with 10_mL of autologous PRP (with addition of" T " " n " ' idft at D ?,fi mmol/mL) and connectedto tHe cannula. About 6 mL of the platelet concentrate wasinjected into the rupture site—-the amount possible toinject without leakage before skin suture. Then the skinwas sutured with a resorbable intracutaneous suture(Monocryl 3-0). Finally, the remaining 4 ml, was injectedtransdermally into the rupture site. There was sometimesleakage from the wound, but we estimated it to be no morethan 1 mL. No injections were given to control patients.Their platelet con'centrate was discarded, but all patientsreceived their own erythrocytes as a transfusion duringsurgery.

A short-leg cast was applied with the foot in the equinusposition. After 3.5 weeks, the cast was replaced with one inwhich the ankle was in neutral position, and another 3.5weeks were allowed to pass. Fj3Jl_weightbearing_wasalkovedjas tolerated from the beginning. The cast wasremovedatter 7 weeksTn totaT7~Sli.d-1,he patients wereinstructed to use shoes with a 2-cm elevation of the heelsfor another 4 weeks. Physiotherapy started after removalof the cast, and all patients followed the same rehabilita-tion protocol. All patients received instructions for hometraining and visited the physiotherapist every 2 weeks.The training program consisted first of mobility exercises,then muscle strength exercises (2- and 1-legged toe raisesat slow and fast speed) and balance exercises (standingon 1 leg on different surfaces and wobble boards). Plyomet-ric exercises (simple 2-leg jumps) were allowed from week16, with a progression during the weeks that followed(jumping on 1 leg in different directions). If the patientcould manage 15 to 20 toe raises, jogging was allowedfrom week 18. Full activity, including sports, was allowedafter approximately 5 months.

Randomization was done with sealed envelopes, withblocks of 6 patients, and with one random exchange ofa pair of envelopes between the blocks (to preclude investi-gator deduction of treatment). The patients had earphones

40 Schepull et al The American Journal of Sports Medicine

a.

_o

£

£

31 patients with Achilles tendonrupture assessed for eligibility (n = 31)

Refused to participate (n = 1)

Randomized(n = 30)

Allocated to PRP treatment(n = 16)

Lost to follow-up 7 weeks(n = 1) due to infection

Analyzed at 7 w (n = 15)

Lost to follow-up 19 weeks(n = 1) due to rerupture

Analyzed at 19 w (n = 14)

Allocated to control group(n = 14)

Analyzed at 7 w (n = 14)

Analyzed at 19 w (n = 14)

Lost to follow-up52 weeks (n = 2)1 not interested

1 measurement failure

Analyzed at 52 w (n = 12) Analyzed at 52 w (n = 14)

Figure 1, CONSORT flow diagram. PRP, platelet-rich plasma; W, weeks.

with loud music during opening of the envelopes and dur-ing the injection to conceal the type of treatment. Thepatients were therefore not aware of whether we hadinjected the platelet concentrate or not. Sixteen patientswere randomized to the PRP group and 14 patients wererandomized to the control group.

Follow-up: Mechanical Properties

To calculate E-modulus, RSA20 was used to measure strainunder defined loading, and computed tomography (GT) wasused to measure the transverse area of the tendon at therupture site.

: Roentgen Stereophotogrammetric Analysis. RSA is a fre-quently used method in orthopaedic research. It providesthe possibility to measure the distance between tantalumbeads in 3 dimensions and with a high degree of accuracy.3A change in position (eg, ankle flexion) does not influencethe measurements if the tissue is not deformed. DuringRSA, simultaneous radiographs are taken in 2 planes usingextracorporal calibration markers in a standardized cage.

We performed RSA after 7 weeks (within 15 minutes ofcast removal), after 19 weeks (12 weeks after castremoval), and after 52 weeks (12 months). We used thesame protocol at the first 2 follow-ups and a slightly differ-ent protocol at the last follow-up. At the first 2 follow-ups,

ae patients sat on an examination table with the foot in

7

Vol. 39, No. 1, 2011

a specially designed frame and with. 8° of plantar flexion.The frame allowed us to apply a pedal to the forefoot andload it with weights. The pedal pivoted around an axiswith an adjustable distance from the posterior aspect ofthe heel, which allowed estimation of the momertt armsfrom raHaogyaphy (see below;7~The patients were thenask&d-to keep the fool in puaillun and to resist the dorsalflexion moment derived from the pedal during loading.TH e first force applied to the pedal was 25 N andJJae-seflondwas 150 N. The 25-N loading' was intrnHprl tn providea "TDaseGneP value (a reasonable relaxation of the dorsalflexor muscles), whereas 150 N was the main loading (load-ing sufficient to produce strain). The patients had to resistthe force for 15 seconds before the radiographs were taken;thereafter, the weight was immediately removedJ3etwee_nall exposures, there was a rest period of 3 minutes. Thesefirst '2 exposures were used as the control examination.Moreover, they served as preconditioning loading of_theAcEilfea tendon. AfUu1 another 3-minute rest, the mainexaSmauon was performed, again with 25 N and 150 N.Where not otherwise stated, all results refer to the second/main examination. Strain-per-force values (assuming a lin-ear relationship) were calculated with correction for thelever arms of the forefoot and the calcaneus and areexpressed as a percentage per 100 N of tendon force.

For descriptive reasons, the testing procedure at 7 weeksled to a strain of 2.0% ± 0.4%. At 19 weeks, the strain was1.0% ± 0.7%.

A final RSA was done 12 months after surgery. Thisexamination differed slightly from the first 2 examina-tions, 1 and 19 weeks after surgery. Also at this time, thepatients had to resist the applied loading for 15 secondswith 3-minute intervals, but we increased the force fromthe baseline 25 N in 100-N increments, with no repetition(ie, the forces were 25, 125, 225, 325, and 425 N). Strainper force was now calculated from the slope of the regres-sion lines for all measurements. Thirmg the course of^feestudy, we omitted^the measurements with 125 N/aad

Autologous Platelets and Human Achilles Tendon Ruptures 41

ThlTlTBeads were numbered from proximal to distal(Figure 2). The change in distance between beads 2 and 3at the second/main loading with 25 N was the value usedfor analysis.

In addition to the above, single unloaded RSA examina-tions were performed with the leg in plaster, shortly afterchange from the equinus position to the normal position at3.5 weeks and immediately before cast removal at 7 weeks.This was done to measure creep.

The RSA analysis was based on the UmRSA 4.1 system(RSA Biomedical, Umea, Sweden). Simultaneous expo-sures were done using a calibration plate designed forRSA of the hip. Rigid bodies were not calculated. Weused the RSA software to calculate distances between sin-gle beads. Tendon force was calculated from pedal force.The pedal pivoted around an axis so that the force hada defined loading point in a lateral projection. Leverarms were calculated from CT lateral radiographs, withthe center of the talar trochlea as pivot point (pedal pointto trochlea center, trochlea center to center of tendon).

•2 3 4

Figure 2. Placement of suture and numbering of markerbeads. Beads 2 and 3 were used for strain measurements,and beads 1 and 4 were used as reserve, in case beads 2or 3 were incorrectly inserted.

This was all done by a technician who was blinded topatient treatment.

Computed Tomography. Using CT, we measured thetransverse area at middistance between the proximal anddistal markers at 7, 19, and 52 weeks. We also determinedthe position of the beads within the tendon. Beads lyingoutside the tendon on the CT scans were excluded.

Follow-up: Functional Outcomes

The first 20 consecutive patients underwent functionalexamination 6 months after surgery, and all patients but1 were examined at the 1-year follow-up. All patientswere examined by a physiotherapist who was blinded totreatment and had not seen the patients before. Thepatients were still blinded regarding their treatment.

The noninjured side was always tested first. Patientswere barefoot during all tests. Passive range of motion indorsal and plantar flexion was registered with a handheldgoniometer. Calf circumference was measured with a mea-suring tape, 10 cm below the tibia! tuberosity. A reductionin calf circumference indicates muscle atrophy,12 and meas-urements have shown good test-retest reliability.13 Therelationship between calf circumference and calf musclestrength and endurance has been shown to be weak.12 Still,measurement of calf circumference is one of the most com-mon evaluation methods after Achilles tendon ruptures.Muscle performance was evaluated with maximal numberof single-limb toe raises (cadence, 30 raises per minute;height of the heel should be at least 5 cm from the floor)and maximal height of 1 single-limb toe raise, measuredin centimeters from the floor to the heel. Heel raise testinghas been recommended for evaluation of calf muscle func-tion. The number of heel raises that a person can performdepends on the height of the heel raise. Thus, we createda heel raise index, defined as the number of heel raisesthat the patient could do, multiplied by height of the heelraise and normalized as a percentage of the other side.18

Tests that combine the number of heel raises and then-height have been found to have good reliability and validityin detecting functional impairments.22 Because this vari-able was correlated to E-modulus in a previous study,19 itwas used as the primary functional outcome variable.

For analysis of force development during gait and verti-cal jump, we used the CODA mpxSO motion analysis

42 Schepull et al The American Journal of Sports Medicine

TABLE 1Mechanical Properties of Healing Achilles Tendons With Autologous Platelet-Rich Plasma (PRP) or Control Treatment"

95% Confidence Interval

PRP Control Lower Mean Upper

E-modulus, MPa7 weeks19 weeks52 weeks

Strain per force, % per 100 N7 weeks19 weeks52 weeks

Area, cm2

7 weeks19 weeks52 weeks

91 ±2990 ± 29

239 ± 69

0.7 ± 0.30.4 ± 0.10.2 ± 0.05

1.54 ± 0.363.24 ± 0.652.76 ± 0.70

80 ±2292 ± 35

260 ± 98

0.9 ± 0.20.4 ± 0.50.2 ± 0.09

1.59 ± 0.353.59 ± 0.572.75 ± 0.55

-11-29-35

-36-72-34

-20-23-18

13-2-8

-14-12

3

-3-10

" 0.4

382519

84739

143

19

"Mean ± SD and 95% confidence interval for the difference between the group means, expressed as-percentage of control mean.

TABLE 2Characteristics of Patients With Autologous Platelet-Rich Plasma (PRP) or Control Treatment

Roentgen Stereophotogrammetric Analysis, n

Age, y Sex, M:P 3.5 Weeks 7 Weeks 19 Weeks 52 Weeks

PRPControl

39.8 ± 6.239.4 ± 8.3

13:311:3

1514

1514

1414

1214

system (Charnwood Dynamics Ltd, Rothley, Leicester-shire, United Kingdom) with a force plate (Kistler HoldingAG, Winterthur, Switzerland) built into the floor. Duringgait analysis, the patients were unaware of the locationof the plate and were asked to walk with normal gait pat-tern between 2 parallel lines on the floor. For verticaljumping, the patients were instructed to stand on the forceplate, jump as high as possible on 1 leg, and land on theforce plate on the same leg. Maximal force was registeredduring the toe-off phase at vertical jump and level walking.Time in the air was analyzed for the vertical jump. Threemeasurements were carried out for each task and eachleg, and a mean value was used in the analysis. The verti-cal jump was performed at only the 1-year follow-up. Thistest has shown good test-retest reliability.21

All patients also filled in and returned the form for thevalidated Achilles Tendon Total Rupture Score (ATRS).16

The ATRS has been shown to have good validity, reliabil-ity, and sensitivity for limitations related to symptoms dur-ing various activities in individuals who have had totalruptures of the Achilles tendon.

Statistics

This study was designed to test the hypothesis that treat-ment with platelet concentrate improves mechanical prop-erties of the healing Achilles tendon. The primary outcomevariable was E-modulus at the time of cast removal. Othermechanical properties at this and other time points weresecondary variables.

A second goal was to test the hypothesis that PRP treat-ment influences the functional outcome. For this, we useda heel raise index at 1 year as the primary variable, but wealso studied the ATRS. The choice of heel raise index wasbased on our previous observation that it correlated withmechanical variables at an early time point.18 Other func-tional outcomes were regarded as descriptive.

Statistical analysis of mechanical data was performedwith the t test and simple linear regression analysis usingSPSS 17 (SPSS Inc, Chicago, IL). Group differences werefurther described by 95% confidence interval (CI) for thedifference between group means. For example, in Table1, —11 in the row for E-modulus at 7 weeks means thatmore than an 11% reduction of the modulus by PRP treat-ment could be excluded with 97.5% confidence. Functionaldata were analyzed using the Mann-Whitney test.

No a priori power analysis could be performed, becausethere was no information available about how large a dif-ference in E-modulus would be clinically relevant. Wetherefore chose 2 groups of 15 patients because this gavea power of 80% to find a difference of slightly above 1 stan-dard deviation.

RESULTS

Patient characteristics are given in Table 2. There were nocomplications related to the tantalum beads or themechanical stress of the measurement procedure. Twopatients had to be excluded from the PRP group. One of

Vol. 39, No. 1, 2011 Autologous Platelets and Human Achilles Tendon Ruptures 43

TABLESTransverse Area and Modulus of Elasticity Over Time in the Platelet-Rich

Plasma (PRP) and Control Groups: Mean ± SD

PRPControl

7 Weeks

155 ± 36159 ± 35

Transverse Area

19 Weeks

323 ± 65359 ± 57

52 Weeks

276 ± 70270 ± 55

7 Weeks

90 ±2980 ± 22

E-modulus

19 Weeks

90 + 2992 ± 35

52 Weeks

239 ± 69237 ± 97

them suffered a rerupture 2 months after removal of thecast. The measurements from this patient at 7 weekswere still included, but data after the rerupture wereexcluded. The rerupture was treated nonoperativelywith a cast in equinus position for 4 weeks and in the neu-tral position for another 4 weeks. The other patient whowas excluded suffered a deep infection that was treatedwith antibiotics. The data obtained from mechanical test-ing of this patient were excluded. Single tantalum beadsfrom 5 patients were excluded because they were outsidethe tendon substance on CT examination. However,because 2 beads were always inserted on either side ofthe rupture, the other bead could be used, and no patienthad to be excluded because of malpositioning of tantalumbeads.

One patient in the control group was unable to resist the150-N weight at 7 weeks because of pain in the operatedarea and was tested with a 100-N weight instead. At thelater examinations, there was no such problem. Onepatient in each group had a deep vein thrombosis diag-nosed by ultrasound. These patients were treated withwarfarin for 3 months. This was not cause for exclusionfrom the study. However, we analyzed the data afterexcluding them, and there was no difference in the results,which might otherwise have influenced the conclusions(unpublished results). One patient did not want to partici-pate in the RSA examination after 1 year, and in one case,the RSA radiographs at 1 year did not show the proximal 2tantalum beads, which meant exclusion of the results.

In total, values for E-modulus at 7 weeks were availableand analyzed for 15 PRP patients and 14 control patients.At 19 weeks, there were 14 PRP patients and 14 controlpatients, and at 52 weeks there were 12 PRP patientsand 14 controls. The functional outcome at 6 months wasassessed in 10 PRP patients and 10 control patients, andat 52 weeks it was assessed in 15 PRP patients and 14 con-trol patients.

Mechanical Properties

The primary variable—-E-modulus at 7 weeks—was 13%higher (95% CI, -11 to 38) with PRP treatment than withthe control treatment. At 19 weeks, the E-modulus wasreduced by 2% in the PRP group (95% CI, -29 to 25). Therewere no significant differences between PRP and controlresults regarding any of the mechanical variables or trans-verse area at any time point. There were no significant dif-ferences either in means or in variances (Table 1).

mra-450-

400-

350-

300 -

250 -

200-

150 -

100 -

50-

0 -19 52 weeks

Figure 3. Transverse area over time.

MPa450 -

400-

350 -

300 -

250 -

200 -

150 -

100 -

50 -

0 -19 52 weeks

Figure 4. E-modulus over time.

The mean transverse area of all patients doubled from 7to 19 weeks (Table 3, Figure 3). This accounted for the entirereduction in strain per force, given that the modulus wasunchanged (Table 3, Figure 4). Even at 1 year, the trans-verse area was greater than at the time of plaster removal.

Functional Outcomes

At the 6-month follow-up, 2 patients (1 in the PRP groupand 1 in the control group) were unable to do a heel raiseof more than 5 cm; therefore, they were excluded fromthe endurance test. Heel raise index could not be calcu-lated for these patients.

44 Schepull et al The American Journal of Sports Medicine

TABLE 4Functional Outcome in All Patients, With P Values for Side Differences"

6 Months (n, 20) 1 Year (n, 29)

Dorsal flexion, degPlantar flexion, degCalf circumference, % cmHeel raise index, %Peak force at gait, % NPeak force at vertical jump, % NVertical jump, % seconds

4 ± 54 ± 5

96 ± 364 ± 1596 ±4

.001

.001<.001<.001

.001

3 ±56 ± 5

. 97 ± 468 ± 24

100 ± 398 ± 691 + 9

.002<.001<.001<.001

.369

.102<.001

"Range of motion in degrees is expressed as difference from the noninjured leg. The other variables are presented as the ratio between theinjured and noninjured leg.

TABLESFunctional Outcome in the Platelet-Rich Plasma (PRP) and Control Groups0

Dorsal flexion, degPlantar flexion, degCalf circumference, % cmHeel raise index, %Peak force at gait, % NPeak force at vertical jump, % NVertical jump, % secondsATRS6

6

PRP

6.5 ± 54.5 ± 696 ±360 ± 1495 ±5

62: 52-70

Months (n, 20)

Control

2.5 ± 43.5 ±396 ± 268 ± 1498 ±2

89: 52-93

P

.079

.651

.701

.257

.139

.191

PRP

3.0 ± 58.0 ± 596 ±569 ±2399 ±398 ± 589 ± 1178: 75-85

1 Year (n, 29)

Control

2.5 ±43.5 ±397 ± 267 ± 27

100 ± 399 ± 692 ± 789: 83-92

P

.630

.011

.670

.823

.881

.727

.793

.014

"Range of motion in degrees is expressed as the difference from the noninjured leg. The other variables are presented as the ratio betweenthe injured limb and the noninjured limb.

'Achilles Tendon Total Rupture Score, presented as median with interquartile range.

Significant deficits in function of the injured leg werefound in all variables 6 months after rupture (Table 4).The largest deficiency was found for the more demandingactivities (heel raises) compared with range of motionand level walking. At the 12-month follow-up, heel raiseindex and peak force at toe-off had increased significantly(P < .05). Still, compared with the noninjured leg, theinjured leg had inferior function in all variables exceptpeak force development during toe-off in gait and verticaljumping (Table 4).

Our main variable for functional outcome—heel raiseindex at 12 months-—was not significantly differentbetween the treatment groups, and no other significant dif-ference in function could be found between the groups,except for plantar flexion at 12 months (Table 5). TheATRS at 12 months was lower (inferior function) for thePRP patients than for the controls (Table 5). This differ-ence remained significant when the 2 patients with deepvein thrombosis were excluded.

Correlation Between Mechanicaland Functional Variables .._ . ... - •

The E-modulus of all tendons at 7 and 19 weeks showeda weak but significant correlation with heel raise index

(respectively, r2 = .21, P = .02; r2 = .27, P = .005). Therewas also a correlation between strain per force at 7 and19 weeks and heel raise index (respectively, r = .19, P =.02; r2 = .16, P = .04). There was no correlation betweentransverse area at 7 and 19 weeks and heel raise index.At 52 weeks, there was a correlation between E-modulusand heel raise index (r2 = .32, P = .003). There was a nega-tive correlation between transverse area and heel raiseindex (r2 = .15, P = .04). Strain per force showed no suchcorrelation.

Tendon Elongation

The tendons elongated (median, 4.5 mm) from 3 to 7weeks (ie, between the change of foot position with thecast and cast removal). Elongation of the tendons from 7to 19 weeks was 2.9 mm (median). From 19 weeks to52 weeks, it was -0.4 mm. There was no significantdifference between the groups for any period (Table 6,Figure 5).

One patient showed a marked elongation (21 mm) from19 to 52 weeks. He also had the lowest heel raise index per-centage after 1 year (only 17%). This patient was in thecontrol group. Reanalysis after excluding him made no dif-ference to the statistical significance of the ATRS results.

Vol. 39, No. 1, 2011 Autologous Platelets and Human Achilles Tendon Ruptures 45

TABLE 6Elongation Over Time in the PRP and Control Groups: Median ± SD

3.5 to 7 Weeks 7 to 19 Weeks 19 to 52 Weeks 7 to 52 Weeks

PRPControl

4.72 ± 3.654.20 ± 2.83

2.96 ± 4.732.48 ± 7.69

0.18 ± 4.32-0.83 ± 5.96

2.08 ± 12.741.77 ± 9.96'

19 52 weeks

Figure 5. Elongation over time.

Error of Measurements

To control our settings of measurements and to confirm thatthe beads were placed correctly and that they were not loosewithin the tendon, we performed double examinations (rep-etition of RSA, 25 N and 150 N) at 7 and 19 weeks. The firstand second strain values showed correlations at 7 weeks(r2 = .54) and 19 weeks (r2 = .94). Furthermore, we comparedthe strain between the outer beads (beads 1 and 4) with thestrain between beads closer to the rupture site (beads 2 and3) in all patients where no beads had to be excluded becauseof malpositioning. Also here, there was a good correlation at7 weeks (r2 = .63) and 19 weeks (r2 = .90).

There was a correlation between the percentage changein distance between beads 2 and 3 over time from 7 to 52weeks (elongation) and the percentage change in distancefor beads I and 4 over the same period (r2 = .94).

At the 12-month follow-up, a regression coefficient wascalculated between tensile load and strain. The regressionwas based on 5 paired values for 11 patients and 3 pairedvalues for 15 patients (the number of radiographic exami-nations at 12 months was reduced after the llth patient, tominimize the radiation exposure). The r2 value was .85 orhigher for all but 1 patient; the median value was .93.

DISCUSSION

This is the first randomized controlled trial to treat Achil-les tendon ruptures with PRP. We were unable to show anybeneficial effect of the PRP treatment on Achilles tendon

healing after rupture. A small to moderate effect cannot beexcluded, but the CI for the difference between groups pre-cludes that the treatment increased the E-modulus bymore than 38%. Moreover, the treatment had a negativeeffect on the ATRS. Because this was not a predeterminedprimary variable, this negative result must be interpretedwith caution.

&ur-jgsults contrast those of Sanchez et al,17 who foundan-eaKtier]return"^o "sports after PRP treatment, compared

h'-FFo'7lTvft1Fit pfmdjP';

et al COagulated^^PRP in vitrn hsfnrp intrnHnfing rHntn

the wound. The platelet concentration in their study wasabout 3 times Ingher than thejiatient's peripheral blood.which is common in studies on~PRP. In our study, the vol-ume that was introduced^was similar, but the platelet con-cenjaation was about 17 timoo that in the .patient'sperjphegai-Mood. Both studies focused on early healing

ny+pnmo

was stored overnight. before surgery^However,we checked platelet viability by exammJngTEiT swirlingphencimgnori immediately before treatmm-it. nf the patient^in accorj.anc£_-with blood acrvicc routines. It has beenshfiwntibiat preparation and 1 day of storage of PRP leadto release of growth factor from platelets in the order of20%.10 However, the high number of platelets injected inour study makes it unlikely that the lack of effect wascaused by low growth factor content, taking into accounteven a small wound leakage and growth factor releasebefore injection. It therefore seems that PRP treatment isof little value for stimulating tendon repair under the con-ditions used in our study.

It is possible that the platelets would be efficaciousunder other conditions, such as after nonoperative treat-meirt or witb TnfifVham'nal loading, 'In^p^rl. mrmy7iri'rg°"nFinow allow postoperative loading In rat experiments, ithas been shown that platelets enhance tendon repairorjyjf_the tendojns.ji-p mpr-hani pally Inarlgj 24 However,rats and humans differ in many ways. In this case, themost important difference might be size. In the rat, whichis relatively small, a large surface area of tissue is exposedto the single drop of injected PRP in relation to the hema-toma volume, which the- cells from this -surface have toreplace with new tissue. In humans, larger spaces haveto be filled with new tissue, and the relation between vol-ume and area is less favorable. This also takes a longertime so that whatever effects the platelets initially had,the risk is higher that they will have faded with time.

This study allowed other observations regarding tendonrepair. It appears that during the first" months of healing,the most important process is callus growth. The

46 Sohepull et al The American Journal of Sports Medicine

transverse area doubled from 7 to 19 weeks, but we couldnot demonstrate any improvement in mechanical tissueproperties (E-modulus) at 19 weeks. This was seen firstat 1 year.

During later phases of healing, it appears that if thehealing has failed to provide good building material (highE-modulus), then it is compensated for by an increase intransverse area. At 1 year, there was relatively little vari-ation in tendon stiffness but a larger variation in modulus,which suggests that poor tissue properties were compen-sated for by a greater transverse area (there was a signifi-cant negative correlation between the 2 variables) andwhich might explain the negative correlation betweentransverse area at this time point and heel raise index.

Function had not been restored 12 months after surgicaltreatment of Achilles tendon rupture. Some improvementwas seen between the 6- and 12-month follow-ups. Treat-ment with PRP made no difference to the functional out-come. Despite this, the patients had a normal gaitpattern, and the self-reported function was relativelygood. These results are in agreement with the results ofprevious studies on functional outcome after Achilles ten-don rupture14'15'23 but not with results on the effect ofPRP.17 The divergence between the functional outcomeand the patient's self-rated function suggests that the dis-ability does not limit the patient in everyday activities.Even so, it may pose a risk of new injuries.

We have reported that there is a correlation betweenthe E-modulus at the time of plaster removal and functionat 18 months.18 This was the case in the present study,although the correlation was weaker, probably becausefunctional measurements were performed earlier (at 12months instead of 18). It is unclear why the properties ofearly tendon tissue should be related to late muscle func-tion. However, the relation confirms the relevance ofmechanical measurements at the time of plaster removal.

We are convinced that there was no migration of tanta-lum beads in the tendon tissue. As can be seen from thecorrelations reported in the results, the 2 beads on eachside of the rupture site behaved similarly and in a waythat is consistent with a homogeneous deformation of thetissue. However, the correlation between values obtainedfrom different beads is less than in our previous study,which is a weakness. We have reported that the strain(in/percentage) is similar regardless of whether the beadsclose to the rupture or further away are used for the mea-surement.18 Our data on elongation from one follow-up tothe next also indicate that the beads on each side of therupture site remain in an unchanged position relative toeach other.

This_study is based on the assumption that Tioalincrtertdoasjaa«ejff bclia^e. la an claotic manncav-IMs is asimpnfication,whish. Biight-be~pfeblematic. From the firstconditioning loading a certain elongation tended to remainto the second, definitive examination. Thus, there maybe an element of viscoelasticity. Our use of the termE-modulus iQu^tJJMrcfQrs-ba-regardedJLa--±ho contort ofan assumption of elastic_behavior. This simplification isnecessary" fur" pi'UCTScal reasons, and it does not influenceour group comparisons.

In conclusion, the present results show that there is nodramatic positive effect of PRP in healing of Achilles ten-don rupture, and they possibly suggest that it has a nega-tive effect on the functional 1-year result. •

ACKNOWLEDGMENT

We thank Hanna Norrman for performing the functionalmeasurements at the 52-week follow-up.

REFERENCES

1. Andersson T, Eliasson P, Aspenberg P. Tissue memory in healingtendons: short loading episodes stimulate healing. J Appi Physio!.2009;107(2):417-421.

2. Aspenberg P. Stimulation of tendon repair: mechanical loading,GDFs and platelets. A mini-review. Int Orthop. 2007;31(6):783-789.

3. Aspenberg P, Wagner P, Niisson KG, Ranstam J. Fixed or loose?Dichotomy in RSA data for cemented cups. Acta Orthop. 2008;79(4):467-473.

4. Bertolini F, Murphy S. A multicenter inspection of the swirling phe-nomenon in platelet concentrates prepared in routine practice: Bio-medical Excellence for Safer Transfusion (BEST) Working Party ofthe International Society of Blood Transfusion. Transfusion. 1996;

xf36/2):128-132.fp. Bo^ch G, van Schie HT, de Groot MW, et al. Effects of platelet-richV—plasma on the quality of repair of mechanically induced core lesions

in equine superficial digital flexor tendons: a placebo-controlledexperimental study. J Orthop Res. 2010;28(2):211-217.

6. de Vos RJ, Weir A, van Sohie HT, et al. Platelet-rich plasma injectionfor chronic Achilles tendinopathy: a randomized controlled trial.

f^-<JAMA. 2010;303(2):144-149.•' \7. Europe Co. Guide to the Preparation, Use and Quality Assurance of

^-Blood Components. 13th ed. Strasbourg, France: Council of Europe;2007.

8. Freymiller EG, Aghaloo TL Platelet-rich plasma: ready or not? J OralMaxillofac Surg. 2004;62(4):484-488.

9. Ingvar.J, Tagii M, Eneroth M. Nonoperative treatment of Achilles ten-don rupture: 196 consecutive patients with a 7% re-rupture rate.Acta Orthop. 2005;76(4):597-601.

10. Ledent E, Wasteson A, Berlin G. Growth factor release during prep-aration and storage of platelet concentrates. Vox Sang. 1995;68(4):205-209.

11. Lyras DN, Kazakos K, Verettas D, et al. The influence of platelet-richplasma on angiogenesis during the early phase of tendon healing.Foot Ankle Int. 2009;30(11):1101-1106.

12. Moller M, Lind K, Movin T, Karlsson J. Calf muscle function afterAchilles tendon rupture: a prospective, randomised study comparingsurgical and non-surgical treatment. Scand J Med Sci Sports.2002;12(1):9-16.

13. Moller M, Lind K.Styf J, Karlsson J. The reliability of isokinetic testingof the ankle joint and a heel-raise test for endurance. Knee SurgSports Traumatol Arthrosc. 2005;13(1):60-71.Moller M, Movin T, Granhed H, Lind K, Faxen E, Karlsson J. Acuterupture of tendon Achillis: a prospective randomised study of com-parison between surgical and non-surgical treatment. J Bone JointSurgBr. 2001;83(6):843-848.Mullaney MJ, McHugh MP, Tyler TF, Nicholas SJ, Lee SJ. Weaknessin end-range plantar flexion after Achilles tendon repair. Am J SportsMed. 2006;34(7):1120-1125.Nilsson-Helander K, Thomee R, Gravare-Silbernagel K, et al. TheAchilles Tendon Total Rupture Score (ATRS): development and vali-dation. Am J Sports Med. 2007;35(3):421-426.

14.

15.

16.

Vol. 39, No. 1, 2011 Autologous Platelets and Human Achilles

17. Sanchez M, Anitua E, Azofra J, Andia I, Padilla S, Mujika I. Compa-rison of surgically repaired Achilles tendon tears using platelet-richfibrin matrices. Am J Sports Med. 2007;35(2):245-251.

18. Schepull T, Kvist J, Andersson C, Aspenberg P. Mechanical proper-ties during healing of Achilles tendon ruptures to predict final out-come: a pilot roentgen stereophotogrammetric analysis in 10patients. BMC Musculoskelet Disord. 2007;8:116.

19. Schepull T, Kvist J, Aspenberg P. Early E-modulus of healing Achillestendons correlates with late function: similar results with or withoutsurgery. Scand J Med Sci Sports. 2010 Jul 29. [Epub ahead of print].

20. Selvik G. Roentgen stereophotogrammetric analysis. Acta Radio!.1990;31(2):113-126.

21. Silbernagel KG, Gustavsson A, Thomee R, Karlsson J. Evaluation oflower leg function in patients with Achilles tendinopathy. Knee SurgSports Traumatol Arthrosc. 2006;14(11):1207-1217.

22. Silbernagel KG, Nilsson-Helander K, Thomee R, Eriksson Bl, KarlssonJ. A new measurement of heel-rise endurance with the ability to detect

functional deficits in patients with Achilles tendon rupture. Knee SurgSports Traumatol Arthrosc. 2010;18(2):258.-264.

23. Suchak AA, Bostick GP, Beaupre LA, Durand DC, Jomha NM. Theinfluence of early weight-bearing compared with non-weight-bearingafter surgical repair of the Achilles.tendon. J Bone Joint Surg Am.2008;90(9):1876-1883.

24. Virchenko O, Aspenberg P. How can one platelet injection after ten-don injury lead to a stronger tendon after*4 weeks? Interplay betweenearly regeneration and mechanical stimulation. Acta Orthop.-200B;77(5):806-812. ':.--, v - ;- j , . - . - . ' " ' , : . "

25. Virchenko O, Grenegard M, Aspenberg P. Independent and addrtfvestimulation of tendon repair by thrornbin and platelets. Acta Orthop.2006;77(6):960-966.

26. Virchenko O, Skoglund B, Aspenberg P. Parecoxib impairs early tei>don repair but improves later remodeling. Am J Sports Med..2004;32(7):1743-1747.

For reprints and permission queries, please visit SAGE's Web site at httpV/www.sagepub.com/journalsPermissions.nav

Knee Rotational Laxity in a RandomizedComparison of Single- Versus Double-Bundle Anterior Cruciate LigamentReconstructionAndrea Hemmerich, *f PhD, Willem van der Merwe,* MD, Marijka Batterham,§ PhD, andChristopher L Vaughan,1" PhDInvestigation performed at the University of Cape Town, Cape Town, South Africa

Background: While single-bundle anterior cruciate ligament reconstruction reduces anterior-posterior laxity, studies have dem-onstrated residual rotational instability. Improved pivot-shift results have been shown with the double-bundle graft; however, nostudy has compared rotational laxity outcome of these surgical techniques in vivo under quantified, isolated torsional loading.

Hypothesis: The anterior cruciate ligament-deficient knee exhibits greater rotational laxity than the contralateral uninjured knee.The double-bundle reconstruction restores rotational joint stability to a greater extent than single-bundle surgery.

Study Design: Controlled laboratory study.

Methods: Rotational laxity of 32 patients with unilateral anterior cruciate ligament injury was assessed in both knees at full exten-sion and 30° of flexion using a magnetic resonance imaging-compatible torsional loading device. Patients were randomly allo-cated either a single- or double-bundle reconstruction and reassessed 5 months after surgery.

Results: The anterior cruciate ligament-deficient knees demonstrated greater laxity to internal rotational torque in the extendedposition, but not in the 30° flexed position. No significant differences in rotational laxity were found between single- and double-bundle reconstructions. In extension, excessive internal rotational laxity of injured compared with contralateral knees wasreduced by anterior cruciate ligament reconstruction. The single-bundle reconstruction did not affect internal rotation comparedwith contralateral or preoperative groups. In response to internal rotational torque in the flexed knee position, the double-bundlereconstruction reduced laxity to 10.8° from the pre-operative value of 15.3° (P = .058); postoperative rotation was also signifi-cantly less than the contralateral laxity of 16.4° (P = .022).

Conclusion: The ruptured anterior cruciate ligament resulted in increased internal rotational laxity only in the extended position.The single-bundle reconstruction did not affect rotational restraint compared with contralateral or preoperative groups. Thedouble-bundle procedure significantly reduced internal laxity in the flexed position when compared with normal.

Clinical Relevance: As the anterior cruciate ligament is not the primary restraint to rotation, its contribution to joint stability islimited under isolated torsional load. While the double-bundle graft demonstrates superior rotational constraint, this may beexcessive for isolated anterior cruciate ligament rupture.

Keywords: in vivo internal-external rotational laxity; anterior cruciate ligament (ACL) reconstruction; single-bundle graft; double-bundle graft; torsional loading

*Address correspondence to Andrea Hemmerich (e-mail: a.hemmerichOalumni.utoronto.ca).

"'"Department of Human Biology, University of Cape Town, Cape Town,South Africa.

*Sports Science Orthopaedic Clinic, Cape Town, South Africa.^Centre for Statistical and Survey Methodology, University of Wollon-

gong, Wollongong, New South Wales, Australia.One or more authors has declared a potential conflict of interest:

Financial support toward the research assistantship for Ms Hemmerichwas provided by the Natural Sciences and Engineering Research Councilof Canada, the National Research Foundation of South Africa, the Inter-national Society of Biomechanics, and Smith & Nephew inc.

The American Journal of Sports Medicine, Vol. 39, No. 1DOI: 10.1177/0363546510379333© 2011 The Author(s)

The anterior cruciate ligament (ACL), in addition to its pri-mary role restraining anterior tibial translation, has beenshown to contribute to rotational stability of the knee.3'6Conventional surgical techniques adequately restoreanterior-posterior (A-P) stability; however, subjectivesymptoms of "giving-way" and positive pivot-shift testresults reveal that rotational laxity often remains.2 Toimprove rotational stability outcome, surgical techniqueshave been modified to reconstruct not just the anterome-dial (AM) bundle, but also the posterolateral (PL) bundleof the ACL. Although the single-bundle (SB) techniquehas been shown to improve knee stability with respect tothe injured knee, several biomechanical studies haveshown superior functional outcome under anterior and

48