A comparison of outcomes following tibial plateau levelling osteotomy and cranial tibial wedge osteotomy procedures

Summary The objective of this study was to determine whether clinical outcomes were superior and complication rates were lower in dogs that had had a cranial cruciate ligament rupture treated by tibial plateau levelling osteotomy (TPLO), compared to those dogs that had been treated using the original cranial tibial wedge osteotomy (CTWO) procedure. Thirty-seven client-owned dogs with cranial cruciate ligament rupture were included in the study: 19 dogs underwent a TPLO procedure, and 18 dogs underwent a CTWO procedure. The study was retrospective, with the data being obtained from medical records and a review of radiographs. The long-term outcome was assessed by means of an owner questionnaire using a visual analogue scale. For the majority of factors that were reviewed, there was not a significant difference in outcome between the dogs that had a TPLO or those that had undergone a CTWO. All of the dogs showed a rapid return to weight bearing after surgery, and at the six week re-examination, the majority of the dogs did not have any pain on stifle palpation. They displayed a good stifle range of motion and significantly lower lameness scores than those prior to surgery. The complication rates did not differ between the procedures, however, within this small sample of dogs, complications following a CTWO were more likely to require revision surgery.

In the CTWOt echnique, plateau levelling is achievedb yr emoving ac raniallybased wedge of bone,w hich shortens the tibia cranially. This alters the femoropatellar jointb yl owering the patella relative to the femur,unlessthe stifle is relatively hyperextendedt oc ompensatef or this. Such com-pensation mayb em ore difficult in dogs with am ore straight-leggedc onformation. This is avoided in the TPLO procedure, whereap roximalr adialo steotomyo ft he tibia is performed,a nd the segment is rotatedcaudallytolevel the plateau.The more proximal position of the TPLO should also result in agreater area of bone contactatthe osteotomys ite,w hich should theoretically result in amore stable osteotomy.
TheT PLOh as largelyr eplaced the CTWO as the technique of choice,yet there is little information comparingthe twotechniques,i no rdert od etermine whethert he benefits thata re claimedf or the modificationsare reflected in improved clinical outcomes. Theo bjective of thiss tudyw as to compare the clinical outcome of theset wo techniques.

Materials and methods
Them edical records of the dogs thath ad been treated for CrCLrupturebyCTWOor TPLO at Royal Ve terinaryCollegebetween June 2002 andF ebruary2 005 were reviewed. All of the TPLO proceduresw ere performed as describedbySlocum andSlocum( 7), other than placemento fabone graft. The osteotomy wasperformedusing a Slocum sawa nd jig, ands tabilised using a 3.5m mS locum TPLO plate,w ith five 3.5 mm cortical screws, and asingle 4mmcancellouss crew placed in the mostp roximal screwh ole. All CTWOp roceduresw ere based on the technique describedbySlocum andD evine (6), althoughn eitherajig was used,n or wasm usclea dvancementp erformed. The osteotomyw as made using a standard straight oscillating saw, andstabi-lised with a3.5 mm sevenhole TPLO plate (Veterinary Instrumentation, Sheffield, UK) with seven3 .5 mm cortical screws. Only those caseswith acompleteset of medical records andradiographs taken pre-post-and sixw eeksp ost-operatively were included. The following information wase xtracted from the records:a ge,s ex,b reed,b ody weight, duration of lameness prior to surgery,lameness score,subjective assessment of stifle rangeo fm otion (normal or reduced)and responsetomanipulation,surgical technique, surgeon, surgical time,complications, presence of unilateral or bilateral cruciate disease, extent of CrCL rupture (partialorcomplete),presence or absence of meniscal damage andwhether meniscal surgery (medialmeniscectomyormeniscal release) hadbeen performed. Thepresence of anyc oncurrent orthopaedic problems,a nd whethert he dogs hadu ndergone previous CrCL surgerywas also noted.
Follow-up information on clinical outcome waso btainedf rom the owners by meansofaquestionnaire,based on the one that wasdevelopedbyInnesand Barr (9). A visuala nalogue scale wasu sed when answering the majority of the questions (Appendix A).

Statistical analysis
Descriptivestatisticsw ereused to evaluate basicpropertiesofthe data, including mean valuesand standard deviations. Quantitative data wasanalysed using eitheratwo-sample or paired t-test.For data that wasnot normally distributedorwhenvariances of the two samples were not equal, the Mann-Whitney Ut est (equivalent of at wo-sample t-test) andthe Wilcoxon signedranktest (equivalento fapaired t-test)w ereu sed instead. Qualitative data wasa nalysed using a 2-tailedP earsonC hi-squaredt est or a Fisher'se xact test if assumptions were not satisfied,i .e. if the expected frequencyi n anyo ft he cells was< 5. Fora ll analyses, a value of p<0.05 wasconsideredsignificant.
TheCTWOproceduresw ereperformed by sixdifferent surgeons,all of whom hada similarl evel of experience using the technique. In contrast,all of theTPLO procedures were performedbyasingle surgeonwho hadr ecentlya ttended the TPLO training course.Whenall of the cases in each group hadb een included,t he means urgicalt ime forthe CTWOprocedure (148 minutes)was significantlyless (p<0.01) than that for the TPLO procedure (188 minutes). When the surgical timesfor the final10cases in each group were compared, however, the mean surgical time wasnot significantlydifferent between the TPLO andC TWOp rocedures (177 and161 minutes,respectively).
VetComp OrthopTraumatol4/2007 groups were comparedatany stage. Most of the dogs hadagood range of stifle motion andminimalornodiscomfort on joint manipulation,w ith no significant difference between the groups.
Post-operative complication ratesa re showninTable 2. There wasnosignificant differenceinthe complication rate between the twogroups.Two dogs in theTPLO group suffered asingle broken screw,but recovery progressed uneventfullyw ithout intervention.I nc ontrast,a ll of the three dogs with implant failureinthe CTWOgroup required revision surgery. In twodogs, screwloosening resultedi no pening of the osteotomy, andinthe third,inlaceration of apopliteal vessel.
Ac omparison of the TPLO andC TWO groups at the time of completion of the questionnaire did notreveal anysignificant differencebetween the twogroups in activity level, disability,orabilitytojump (Fig.  4). Thed ogsi nt he CTWOg roup were reported to suffer significantlym ore from stiffness afterr esting (p=0.036) andw ere moresusceptible to cold (p=0.029).

Fig. 2
Distribution of lameness scores prior to surgeryand six weeks after surgeryfor TPLO-treated and CTWO-treated dogs. The lameness scores were notsignificantly differentbetween the two groupspre-operatively, or at six weeks. In both groups, the lameness score was significantlylower at six weeks comparedtopre-operatively(P<0.001).

Discussion
TheTPLOisamodification of the original CTWO procedure,a nd as techniques evolve, it is importanttodetermine whether the benefits thatare claimedfor the modificationsa re reflected in improved clinical outcomes. In ordert op racticet ruly evidence-based medicine,s uchaquestion should be answered in aprospective, randomised, controlled study using objective outcome measures. TheC TWOp rocedure is not widely performed now, whichm akes such ap rospectives tudyd ifficult. The popularity of each technique during differenttime periods also results in significantly longerf ollow-up periods for the CTWO's cases, whichm ay influencep arameters, such as stiffness or the effect of cold. It should also be recognisedt hati nformation on outcome gainedt hrough the widely acceptedm ethods of ownerq uestionnaire or telephonef ollow-up (10-12) is subjective, andt herefore lessr obustt hano bjective measurement, such as forcep late analysis. While acknowledging thesel imitations, it remains useful to compare the twop opulations to determine whethers ignificantly better outcomesa re obtainedf ollowing TPLO comparedtothe original CTWOprocedure.
Successful clinical practiced epends upon owners being satisfied with their pet's progress post-operatively,therefore despite being subjective,the owner'simpressions of their dog'sr ecovery arei mportant. Many studies therefore obtain follow-up information on long-termo utcome through client questionnaires, butinorderfor the results to be consideredv alid,am inimumr esponse rate of 60% is required (13,14). Thereturn rate in the present study was70%, with 92% of owners reporting that the outcome was 'good'or'excellent' without anydifference between the procedures. This agrees with other studies,w here 'good' to 'excellent' outcomeswerereported for 86% of patients following CTWO (15), and9 4% (7), and 98.7% (16) of patients following TPLO, based on client satisfaction.I nc ontrast, whereo bjective forcep late analysis has been used to assesso utcome,r esults have variedd ramatically. While ones tudyr eported that allofthe dogs hadregainednear normal peak vertical forces andvertical impulses by 18 weeksp osts urgery (17), another reported normal limb function in only 10.9% of TPLO-treated dogs by sixmonths (18). Subjective 'owner'assessmentsofoutcome were not availablei ne ithero ft he studies.
Despite being satisfied with the outcome,m osto ft he owners in the present study did not feel thattheir dogs returnedto pre-injurye xercise levels,i nc ontrast to other studies in whicht his wasr eportedly achieved ( 6,7,17,19,20). In particular, owners of the dogs that hadu ndergone a CTWO reportedt hatt heir dogs suffered from significantlym ore stiffness andw ere moreaffected by cold,comparedtoowners of the TPLO-treated dogs.This probablyreflectst he longer follow-up times of the

VetComp OrthopTraumatol4/2007
CTWO-treated dogs,a sb oth of thesep arameters arei ndicative of osteoarthritisi n humans (21,22). Despite this,the number of dogs receiving anti-inflammatorym edication did notd ifferb etween the two groups. Althoughasignificant differencew as not identified in the gradeofo steoarthritis when the groups were comparedsix weeks post-operatively,am uchl ongerf ollow-up period would be required in ordertodetermine whethereitherprocedure significantly alters the progression of osteoarthritis. Despiteoriginalclaims thatosteoarthritis does not progress following successfulT PLO (7), more recents tudiesh aver eported the progression of osteoarthritisi nm ostd ogs (16,23). The TPLO procedure maylimit or slow down the development of osteoarthritis comparedtoextra-capsularrepairs however, as ar ecents tudyr eported that dogs with greater changesi no steoarthritis scores (based on radiographs taken pre-operatively and12months post-operatively)w ere5.78 timesmore likely to have hadanextra-capsularr epairt hanaTPLO (24). Thea ssessment of osteophyte development on radiographs is very subjective,h owever,a nd radiographic signs of stifle osteoarthritis have been shownt oc orrelate poorlyw ith clinical function (25), whichp erhaps explains the 'good' to 'excellent' function that wasreported for most of the dogs following TPLO or CTWO.
As would be expected while experience is gainedwith anew technique, the surgical timesfor the TPLO procedure were initially significantlyl ongert hant hosef or the CTWO. Aftert he initial nine procedures, however, the surgical timesb ecamec om-parable. Overall, the surgical timesfor both proceduresreflect the teaching environment in whicht heyw erep erformed; in an onteaching environment, the surgical times maybeconsiderablyshorter.
In conclusion, this paperc ompared the outcomes following CTWOo rT PLO for treatmento fc ranial cruciatel igamentr upture, anddid not find anydifference in most of theparameters under study. Allofthe dogs showedarapid returntoweightbearing after surgery, and at the six week re-examination, the majorityofdogshad no pain on palpation and significantlylower lameness scores than prior to surgeryo ft he stifle. The complication rates did notd iffer between the procedures, however, within this small sample of dogs, complications following aC TWO were more likelytoresult in arevisionsurgery. It wouldb ei nterestingt od etermine whether the results of the presents tudy wouldb es upported in al arger prospective study,using objectiveoutcomemeasures.