FOOT & ANKLE INTERNATIONAL

Copyright 2008 by the American Orthopaedic Foot & Ankle SocietyDOI: 10.3113/FAI.2008.0124

Short-Term Results of Our First 49 Scandanavian Total Ankle Replacements(STAR)

B.G. Schutte, M.D.; J.W.K. Louwerens, M.D., Ph.D.Nijmegen, The Netherlands

ABSTRACT

Background: Forty-seven consecutive patients treated for anklearthritis with a Scandinavian total ankle replacement (STAR)by one surgeon were investigated retrospectively. Materialsand Methods: A modification of the Foot Function Index(FFI), which scores pain and task difficulties, was followedprospectively. Patients were assessed clinically and radiolog-ically. Failure was defined as revision of the prosthesis orarthrodesis for any reason. Results: In 47 patients (16 male, 31female) 49 total ankle replacements were carried out betweenMay 1999 and June 2004. Indication for surgery was end stagearthritis for rheumatoid arthritis in 29 cases, post-traumaticarthritis in 12, osteoarthritis in five and arthritis secondary todegenerative flatfoot in three. Mean followup time was 28 (12to 67) months. The modified FFI (range, 0 to 100, a high scoremeaning more pain and disability) improved significantly from59 before to 35 after surgery. The mean postoperative Kofoedankle score was 68. Sixteen procedures were complicated byfractures or temporary neurological damage. At the time offollowup, 45 prostheses survived, while four replacements hadfailed. Radiological examination at followup showed radiolucentlines, osteolysis, and malposition of the components in 31 cases.Conclusion: Our results are comparable with those reported inthe literature. The clinical outcome improved significantly. Dueto aseptic and septic loosening, 8.2% of the prosthesis failed.

Key Words: Total Ankle Replacement; Complications; Func-tional Outcome; Radiological Result

INTRODUCTION

Ankle fusion is still the gold standard for the surgicaltreatment of debilitating ankle arthritis. However, fusion of

Corresponding Author:J.W.K. Louwerens, M.D., Ph.D.Foot and Ankle Reconstruction UnitMaartenskliniekP.O. Box 91016500 HB NijmegenThe NetherlandsE-mail: j.louwerens@maartenskliniek.nlFor information on prices and availability of reprints, call 410-494-4994 x226

the ankle joint has disadvantages, particularly with regard tothe adjacent joints. Compensation for loss of ankle motionin the long term has a negative influence on these joints.4

Patients with pre-existing problems in the knee or forefootmight experience more pain after an ankle fusion. Fusingthe ankle in the presence of stiff or fused tarsal joints canresult in a pantalar arthrodesis which would certainly affectthe walking pattern. Replacing the ankle joint, instead offusing it, seems attractive, provided good long-term resultsand an acceptable complication rate are achieved. The resultsfor the third generation total ankle replacement, which hasan unconstrained design and consists of 2 uncemented,anatomical components and a polyethylene meniscus havebeen promising.1,10,11,13

The purpose of this study was to investigate the short-termclinical and radiological results and evaluate the compli-cations in the first group of patients from our clinic inwhich a STAR ankle replacement (Waldemar Link, Hamburg,Germany) was used.

MATERIALS AND METHODS

Between May 1999 and June 2004, 49 Scandinavian totalankle replacements (STAR) were done in 47 patients (16male, 31 female) by one surgeon (JWL). The average age atthe time of surgery was 57.1 (37.2 to 81.2) years. The leftankle was replaced 18 times, the right ankle 27 times, andboth ankles twice. The indication for surgery was debilitatingankle arthritis as a result of rheumatoid arthritis in 29 cases, apost-traumatic condition in 12, primary osteoarthritis in fiveand arthritis secondary to a degenerative flatfoot in three.

No statistically significant differences were found in age,gender, or operated side between the separate indicationgroups. An anterior approach was used and all prostheseswere placed without cement. Postoperative rehabilitationconsisted of 4 weeks lower leg cast, the first 2 weeksnonweightbearing.

A modification of the Foot Function Index used in thepresent study is nearly similar to the Ankle Osteoarthritis

124

at Selcuk Universitesi on December 21, 2014fai.sagepub.comDownloaded from

Foot & Ankle International/Vol. 29, No. 2/February 2008 SHORT-TERM STAR RESULTS 125

A

B

Fig. 1: A, Correct placement of a STAR ankle replacement on an AP x-ray of a left ankle. Tibial and talar component perpendicular to the tibia axis. Noradiolucent lines. B, Lateral x-ray shows correct alignment of both components.

Scale as developed by Domsic and Saltzman.3 This self-assessment instrument has been translated into Dutch andhas been validated for use in The Netherlands by Kuyven-hoven et al.7 Reliably measuring pain and disability dueto ankle problems, 7 items concerning pain and 8 itemsconcerning task difficulties are scored on a 5-point scale.Scores were established before surgery and at followup. Thescore runs from 0 to 100 with 0 indicating no pain andno restrictions. The Kofoed ankle score (max, 100 points:less than 70 = poor, 70 to 74 = fair, 75 to 84 = good, andgreater than 84 = excellent) was scored at followup.6 Of allpatients demographic characteristics, date of surgery, dura-tion of the procedure, additional procedures and peri- orpostoperative complications were noted. Radiographs wereexamined for the position of the prosthesis and radiolucentlines (less than 2 mm) or signs of osteolysis (greater than2 mm).12 Malposition was defined as an angulation of thecomponents greater than 5 degrees from the optimal position(Figure 1).12

Failure was revision of the prosthesis or arthrodesis.The statistical analysis the Student’s t-test was used instat analysis, with the p value less than 0.05 consideredsignificant.

RESULTS

The range of motion of the ankle preoperatively wasan average of 4 degrees dorsiflexion (±5 degrees) and 19degrees plantarflexion (±7 degrees). In order to achievenormal ankle alignment, preliminary triple arthrodesis wasperformed in 8 patients combined with an Achilles tendonlengthening in two.

At the time of ankle replacement, no additional procedurewas performed in 23 cases. The procedure was combinedwith a lengthening of the Achilles tendon 24 times, witha talonavicular, once with a subtalar arthrodesis, and twicethe forefoot was corrected additionally. Mean operationtime was 125 minutes (range, 90 to 160 minutes). In onepatient both ankles were operated in a single 165-minutesession.

In 16 patients, a major or minor complication occurredduring surgery (Table 1). The underlying pathology wasrheumatoid arthritis in 10 cases; a post-traumatic conditionin four; osteoarthritis in one; and a degenerative flatfootin the last. No statistically significant difference was foundin the occurrence of complications with regard to theindication for surgery. No difference was found in occurrenceof complications between the first 25 and the later 24patients.

The median duration of hospital stay was 5 days, rangingfrom 2 to 150 days. The 150-day hospital stay concerned apatient whose operation was complicated by a deep bacterialinfection. Despite numerous (operative) attempts to save the

Table 1: Complications during surgery

Fracture medial malleolus 6Fracture lateral malleolus 2Fracture distal tibia 3Neuropraxia of the peroneal.nerve 1Malposition talus component 2Malposition tibial component 2

at Selcuk Universitesi on December 21, 2014fai.sagepub.comDownloaded from

126 SCHUTTE AND LOUWERENS Foot & Ankle International/Vol. 29, No. 2/February 2008

prosthesis, a fusion was needed. In 2 cases, delayed woundhealing was treated successfully nonoperatively.

The mean followup time was 28 (range, 12 to 67) months.At the time of followup, 45 prostheses survived: four caseshad failed due to septic (2) and aseptic (2) loosening (Figure2).

Of these 4 patients, hematogenous infection of the anklejoint was the result of sepsis caused by a urinary tractinfection. Three fusions and 1 revision were performed. Twoother patients complaining of persistent pain underwent asecond operative procedure. In one, a cyst in the medialmalleolus was filled with cancellous bone. In the other anon-union of a fibula fracture and an anterior impingementdue to fibrosis in the lateral gutter were treated. One patientdeveloped a stress fracture of the medial malleolus 19 monthsafter her ankle replacement operation. This was successfullytreated with internal fixation and debridement of the medialgutter.

At followup, the average dorsiflexion was 7 degrees (±4degrees) and plantarflexion 20 degrees (±8 degrees). Theimprovement of the dorsiflexion was significant comparedwith the pre-operative motion. Before surgery the meantotal FFI was 59 (pain 54, activity 62). Postoperatively, theresult was 35 (pain 29, activity 40). These differences arestatistically significant (p < 0.001). The mean postoperativeKofoed score was 68 (pain 35, daily life function 20 andstatic and dynamic mobility 13). In 21 patients, the resultwas poor (less than 70), 5 moderate (70 to 74), 10 good (75to 84) and 8 excellent (greater than 84) (Table 2).

Radiological followup showed an appropriate position ofthe prosthesis and no radiolucent lines in 14 cases. In theremaining ankles, 14 tibial components were not correctlyplaced (1 varus, 4 valgus, 9 posterior slope) and 12 taluscomponents were placed in anterior slope (11) or valgus (1).Radiolucent lines (less than 2 mm) were seen around 10 tibial

Table 2: Mean functional results per indication (standarddeviation)

FFIpre-op

FFIpost-op

Kofoedpost-op

R.A. 60 (±14) 34 (±19) 69 (±20)Post-traumatic 47 (±21) 30 (±23) 70 (±18)Osteoarthritis 68 (±15) 50 (±8) 65 (±24)Degenerative 67 (±9) 39 (±8) 67 (±11)

flatfootTotal 59 (±17) 35 (±19) 68 (±19)

components and 2 talar components. Osteolysis (greater than2 mm) was seen in 10 patients around the tibial componentand twice at the talar side. In one patient, peri-articularossifications were seen and this resulted in a decreased rangeof motion.

In 5 of 8 patients with an excellent Kofoed score, x-ray images were good. Osteolysis was seen around thetibial component while the talus component was placed indorsiflexion in one case, malposition of the tibial in another,and a malposition of the talar component was seen in a third.Of the 10 cases with a good Kofoed score, three had an x-raywithout problems.

DISCUSSION

The present study was performed in order to evaluate theresults of the first 50 STAR ankle prosthesis placed in ourinstitute. Functional results found in the present study arecomparable with those described by Anderson et al.1 Theyfound a median postoperative Kofoed score of 70, while

A

B

Fig. 2: A, Severe osteolysis around tibial component with severe displacement of both components on an AP x-ray. B, Collapse of talus and marked osteolysisof the distal tibia on a lateral ankle x-ray. Displacement of both components.

at Selcuk Universitesi on December 21, 2014fai.sagepub.comDownloaded from

Foot & Ankle International/Vol. 29, No. 2/February 2008 SHORT-TERM STAR RESULTS 127

this study reports a mean total postoperative score of 69.Kofoed and Wood report better results, most likely related totheir extensive experience in total ankle replacements.5,13 Wefound that quite a large number of patients still experiencevariable degrees of discomfort and/or pain on walking. Thisresults in disappointing FFI and Kofoed scores because painis a major item in both scoring instruments. The effect ofa total ankle replacement on quality of life is not scoredsufficiently with these scores. It might be helpful to assesspatients with a test like the SF-36 in further research tobe better informed about the quality of life and not solelyto focus on ankle function.10 Recognizing that the variousdiagnostic groups are small, the outcomes between thesegroups are similar. Although more complications occur inthe group of patients with rheumatoid arthritis, no otherimportant differences were found between the groups. Thisis in accordance with the studies reported by others.1,5

The existence of a learning curve for this procedure iswell known.1,8,13 Many serious complications like fractures,early deep infections, alignment problems, septic or sterileloosening have been published in literature.1,2,8,11 The typeand amount of complications in the present study arecomparable with those reported in the recent literature. Mostcomplications were either acceptable or solvable. Secondarysurgery was performed in 8 patients (16%), in 4 cases thereplacement remained untouched. Some have reported a re-operation rate up to 25% for several reasons.1,9

Malalignment was most often found in the sagittal plane.We found it difficult to control the position of the compo-nents, in particular the talus. The instrumentation that comeswith the implant is of little help.1 This problem is partly dueto lack of experience and part of the learning curve. The50 cases following this series have been positioned better.Fluoroscopic control of component positioning intraopera-tively should be considered when one starts with this kind ofsurgery. Patients were carefully selected not to have a valgusor varus malalignment prior to placement. In 8 patients, asevere valgus deformity of the hindfoot was corrected by atriple arthrodesis prior to the placement of the prosthesis.Most often this involved patients with rheumatoid arthritiswith involvement of the tarsal joints. When analyzing theamount of involvement of the separate ankle and tarsal jointsand the quality of the bonestock, particularly in patients withsevere rheumatoid arthritis, CT scanning provides clear infor-mation.

In a single patient, medial instability was not recognizedand this resulted in failure of the prosthesis. Presentlyour experience in recognizing soft tissue imbalance andperforming additional soft tissue release when necessary isincreasing. In a relative large number of cases an Achillestendon lengthening was performed. With better positioning

of components, (placing the talar component sufficientlyposterior), this additional procedure is generally unnecessary.

CONCLUSION

It is clear that the results of total ankle replacementare not as predictable as the results of total hip and kneearthroplasty,9 although others have found results that doapproach the excellent results of hip and knee prostheses.5,13

Ankle replacements not only are technically difficult butoften involve complex deformities. As a result we feel thatthey should only be performed by experienced orthopaedicfoot and ankle surgeons.

REFERENCES

1. Anderson, T; Montgomery, F; Carlsson, A: Uncemented STAR totalankle prostheses: three to eight year followup of fifty-one consecutiveankles. J. Bone Joint Surg. Am. 85(7):1321– 9, 2003.

2. Buechel, FF; Buechel, FF; Pappas, MJ: Twenty year evaluation ofcementless mobile-bearing total ankle replacement. Clin. Orth. Rel.Res. 424:19– 26, 2004. http://dx.doi.org/10.1097/01.blo.0000132243.41419.59

3. Domsic, RT; Saltzman, CL: Ankle Osteoarthritis Scale. Foot AnkleInt. 19(7):466– 471, 1998.

4. Fuchs, S; Sandmann, C; Skwara, A; Chylarecki, C: Quality of life 20years after arthrodesis of the ankle. A study of adjacent joints. J. BoneJoint Surg. Br. 85(7):994– 998, 2003. http://dx.doi.org/10.1302/0301-620X.85B7.13984

5. Kofoed, H: Scandinavian total ankle replacement. Clin. Orth. Rel.Res. 424:73– 79, 2004. http://dx.doi.org/10.1097/01.blo.0000132414.41124.06

6. Kofoed, H; Danborg, L: Biological fixation of ankle arthroplasty. Foot.5:27– 31, 1995.

7. Kuyvenhoven, MM; Gorter, KJ; Zuithof, P; et al.: The FootFunction Index with Verbal Rating Scales (FFI-5pt): A clinimetricevaluation and comparison with the original FFI. J. of Rheumatology.29:1023– 1028, 2002.

8. McGarvey, WC; Clanton, TO; Lunz, D: Malleolar fracture after totalankle arthroplasty: a comparison of two designs. Clin. Orth. Rel. Res.424:104– 110, 2004.

9. Spirit, AA; Assal, M; Hansen, ST: Complications and failure aftertotal ankle arthroplasty. J. Bone Joint Surg. Am. 86(6):1172– 1178,2004.

10. Stengel, D; Bauwens, K; Ekkernkamp, A; Cramer, J: Efficiacyof total ankle replacement with mobile-bearing devices: a systematicreview and meta-analysis. Arch. Orthop. Trauma Surg. 125(2):109– 119,2005. http://dx.doi.org/10.1007/s00402-004-0765-3

11. Su, EP; Kahn, B; Figgie, MP: Total ankle replacement in patientswith rheumatoid arthritis. Clin. Orth. Rel. Res. 424:32– 38, 2004.http://dx.doi.org/10.1097/01.blo.0000132181.46593.82

12. Valderrabano, V; Hintermann, B; Dick, W: Scandinavian total anklereplacement; a 3.7-year average followup of 65 patients. Clin. Orth.Rel. Res. 424:47– 56, 2004.

13. Wood, PLR; Deakin, S: Total ankle replacement: the results in 200ankles. J. Bone Joint Surg. Br. 85(3):334– 341, 2003. http://dx.doi.org/10.1302/0301-620X.85B3.13849

at Selcuk Universitesi on December 21, 2014fai.sagepub.comDownloaded from