Incidence of Revision After Primary Implantation of the Salto® Mobile Version and Salto Talaris™ Total Ankle Prostheses: A Systematic Review

The Journal of Foot & Ankle Surgery 54 (2015) 311–319

Contents lists available at ScienceDirect

The Journal of Foot & Ankle Surgery journal homepage: www.jfas.org

Incidence of Revision After Primary Implantation of the SaltoÒ Mobile Version and Salto TalarisÔ Total Ankle Prostheses: A Systematic Review Thomas S. Roukis, DPM, PhD, FACFAS 1, Andrew D. Elliott, DPM, JD 2 1 2

Attending Staff, Department of Orthopaedics, Podiatry, and Sports Medicine, Gundersen Health System, La Crosse, WI Postgraduate Year 2, Podiatric Medicine and Surgery Resident, Gundersen Medical Foundation, La Crosse, WI

a r t i c l e i n f o

a b s t r a c t

Level of Clinical Evidence: 4

The incidence of revision of total ankle replacement prostheses remains unclear. We undertook a systematic review to identify the material relating to the incidence of revision after implantation of the SaltoÒ mobile version and Salto TalarisÔ total ankle prostheses. Studies were eligible for inclusion only if they had involved primary total ankle replacement with these prostheses and had included the incidence of revision. Eight studies involving 1,209 SaltoÒ mobile version prostheses, with a weighted mean follow-up period of 55.2 months, and 5 studies involving 212 Salto TalarisÔ total ankle prostheses, with a weighted mean follow-up period of 34.9 months, were included. Forty-eight patients with SaltoÒ mobile version prostheses (4%) underwent revision, of whom 24 (70.5%) underwent ankle arthrodesis, 9 (26.5%) metallic component replacement, and 1 (3%) below-the-knee amputation. Five (2.4%) Salto TalarisÔ total ankle prostheses underwent revision (3 metallic component replacement and 2 ankle arthrodeses). Restricting the data to the inventor, design team, or disclosed consultants, the incidence of revision was 5.2% for the SaltoÒ mobile version and 2.6% for the Salto TalarisÔ total ankle prostheses. In contrast, data that excluded these individuals had an incidence of revision of 2.8% for the SaltoÒ mobile version and 2.0% for the Salto TalarisÔ total ankle prostheses. We could not identify any obvious difference in the etiology responsible for the incidence of revision between these mobile- and fixed-bearing prostheses. The incidence of revision for the SaltoÒ mobile version and Salto TalarisÔ total ankle prostheses was lower than those reported through systematic review for the AgilityÔ and Scandinavian Total Ankle ReplacementÔ systems without obvious selection (inventor) or publication (conflict of interest) bias. Ó 2015 by the American College of Foot and Ankle Surgeons. All rights reserved.

Keywords: ankle implant degenerative joint disease joint arthroplasty osteoarthritis surgery

Contemporary total ankle replacement systems, compared with those available decades ago, have demonstrated promise because they include improved materials, a more precise surgical technique, and a focus on maintaining the normal anatomy and function of the ankle joint (1–10). Mobile-bearing technology has allowed for increased implant conformity, with a reduction of implant constraint. This theoretically reduces wear and loosening (8,11,12), at the expense of increased potential for ultra-high-molecularweight polyethylene (UHMWPE)-bearing fracture, dislocation, and malleolar impingement (13–15). Nevertheless, even with advancements in technology and implant engineering, total ankle

Financial Disclosure: None reported. Conflict of Interest: None reported. Address correspondence to: Thomas S. Roukis, DPM, PhD, FACFAS, Department of Orthopaedics, Podiatry, and Sports Medicine, Gundersen Health System, Second Floor Founders Building, 1900 South Avenue, La Crosse, WI 54601. E-mail address: [email protected] (T.S. Roukis).

replacement has continued to have unpredictable long-term survivorship (1–9,16). One third-generation total ankle implant is the SaltoÒ mobile version prosthesis (Tornier NV, Amsterdam, The Netherlands), which was invented by Michael Bonnin, MD, Jean-Alain Colombier, MD, Thierry Judet, MD, and Alain Tornier between 1994 and 1996 using anatomic studies of the ankle joint. The device was first implanted in January 1997 and was limited to these surgeon inventors from 1997 to 1999 (17). The first clinical results were published in 2000 (18,19). The SaltoÒ mobile version prosthesis is a 3-component, cementless, mobile-bearing anatomic resurfacing prosthesis and was initially available in 3 sizes (i.e., 1, 2, 3) until 2009, when a fourth size (i.e., 0) was added. Each tibial component initially had a flat 2-mm-thick surface intended to accommodate the superior surface of the UHMWPE mobile-bearing insert and a 3-mm medial rim intended to prevent the insert from impinging the medial malleolus. In 2013, the tibial component was redesigned to add 1 mm to the width and 2 mm to the length to improve tibial

1067-2516/$ - see front matter Ó 2015 by the American College of Foot and Ankle Surgeons. All rights reserved. http://dx.doi.org/10.1053/j.jfas.2014.05.005

312

T.S. Roukis, A.D. Elliott / The Journal of Foot & Ankle Surgery 54 (2015) 311–319

coverage, in addition to rounding off the anteromedial and anterolateral surfaces to reduce overhang. Tibial component fixation is achieved primarily with anterior cortical contact of the flat surface and a 12-mm-long central keel attached to a hollow tapered anteroposterior conical fixation plug that is impacted into the tibial metaphysis. The tibial component is designed for insertion with a 7
posterior slope relative to the long axis of the tibia. The tibial base can be the same or 1 size larger than the talar component, allowing for mismatching of the tibial and talar components, depending on the patient’s anatomy. The tibial components are interchangeable, but the talar components have dedicated left and right sides owing to the double radii (i.e., medial radius smaller than lateral radius) and biconvex articular surface, resembling the native morphology of the talar dome. The undersurface of the talar component matches 3 sharply angulated bone cuts about the posterior, anterior, and lateral talar body, affording primary stability in the anteroposterior and medial–lateral planes. Secondary talar component fixation for component sizes 1, 2, and 3 includes a posterior angled 11.6-mm-deep, 12.7-mm outer diameter medially offset hollow fixation plug, and size 0 has a 10.4-mm-deep, 8-mm outer diameter solid fixation plug. The center of the fixation plug is a constant distance from the concave lateral facet for each talar component. Talar component sizes 1, 2, and 3 remove 7 mm of talar bone height and size 0 removes 5.5 mm. Both the tibial and the talar components are made of cobalt-chromium and are dualcoated with 200-mm plasma-sprayed titanium (T40). Before being discontinued in 2012, an additional external coating of hydroxyapatite was used to promote osseous integration. The UHMWPE inserts were originally available in 3, 4, 5, 6, and 7 mm thicknesses. However, owing to fractures of the 3-mm-thick UHMWPE insert, this thickness is no longer available for use, and, in 2009, an 8-mmthick insert was added. The UHMWPE inserts’ articular surface is size matched to the talar component and have dedicated left and right sides. The contact area between the UHMWPE insert and talar component allows for 3
of freedom of motion between these components, with full congruency in dorsiflexion to plantarflexion. An optional cemented UHMWPE lateral malleolar resurfacing component was created by the surgeon design team to reduce fibular impingement; however, it was abandoned after 20 cases because no clear benefit was appreciated (20). Dedicated instrumentation, including an external alignment jig with tibial and pedal referencing, is used to resect the distal tibia and talus and insert the trial and final prosthesis components (http://www.google.com/url? sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CCQQFjAA&url=http %3A%2F%2Fwww.rpa.spot.pt%2Fgetdoc%2F6065d9a6-dd01-46cf-b5f9f6f73ed785d5%2FSalto-II-%281%29.aspx&ei=zsXlUpHpBsbsqAG_5IHo Cg&usg=AFQjCNGVxwT-iGV-bGerfA6_-lWddjBKaQ&bvme¼bv.5993 0103,d.aWM&cad=rja). The in vivo kinematics of the SaltoÒ mobile version was investigated in 20 patients using fluoroscopy with a 2-dimensional to 3-dimensional registration technique. In that study, translation between the UHMWPE mobile-bearing insert and tibial base plate averaged 1.5 mm during gait, and the insert remained in internal rotation throughout the arc of motion (21). In another study of stress lateral radiographs from 20 patients, variation of the anteroposterior translation of the UHMWPE mobile-bearing insert relative to the tibial base was not noticeable in 17 patients and was only 1 mm in 3 patients (22,23). These studies have indicated that the UHMWPE insert did not function as a mobile-bearing system but rather remained essentially fixed to the tibial component. With this realization, in conjunction with the problems associated with malleolar impingement and UHMWPE wear debris-induced osteolysis (24), the fixed-bearing Salto TalarisÔ total ankle prosthesis (Tornier, Inc, Bloomington, MN/Wright Medical Technology, Inc., Arlington, TN) was developed.

The Salto TalarisÔ total ankle prosthesis was trademarked in April 2006 (http://trademarks.justia.com/789/26/salto-talaris-78926758. html). In that same year, it was first implanted by the each of the original French inventors in June (http://www.healio.com/ orthopedics/foot-ankle/news/print/orthopedics-today/%7B8c89d9dd1a2a-4aa1-98c7-9372f3d05f9b%7D/results-by-design-researcherslook-to-future-tar-developments) received U.S. Food and Drug Administration (FDA) 510(k) clearance for use in November (http:// www.accessdata.fda.gov/cdrh_docs/pdf6/k060544.pdf) and was first implanted in the United States in December (http://www. thefreelibrary.com/TornierþReceivesþFDAþClearanceþforþtheþSaltoþ Talaris%28TM%29 þ Anatomic...-a0155628807). It has been cleared by the FDA only for use with polymethylmethacrylate cement fixation, which has been demonstrated in the technique guide (http://www. tornierdx.info/pdf/SALTO_SXT_INST2_UJAT092-1oct09-a.pdf). Compared with the SaltoÒ mobile version, the Salto TalarisÔ total ankle prosthesis’ tibial component base is 3 mm thick, does not have a medial rim because the UHMWPE insert is fixed to the tibial base, and the UHMWPE inserts are available in thicknesses of 5, 6, 7, and 8 mm (http://www.tornierdx.info/pdf/SALTO_SXT_INST2_UJAT092-1oct09a.pdf). Both the tibial and the talar components are made of cobaltchromium and are single-coated with 200-mm plasma-sprayed titanium (T40) to promote osseous integration. The talar component has been redesigned with deeper biconvex medial and lateral articular surfaces with a concave trochlear groove and a 12
apex medial frontal plane axis to allow for external rotation of the foot with dorsiflexion and internal rotation during plantarflexion. Furthermore, the mobile-bearing concept has been moved from the implant design to the instrumentation at the stage of the trial reduction. According to the surgical technique guide (http://www.tornierdx.com/pdf/SALTO_ SXT_INST2_UJAT092-1oct09-a.pdf), accurate and reproducible tibial and talar component alignment are possible by first performing a measured resection with equal implant replacement for the distal tibia and talus. Next, the trial tibial base, featuring a highly polished surface that remains mobile against the resected distal tibia, is allowed to rotate into proper position during ankle range of motion and tracking of the talar component through a securely fixed, highly conforming, articulating trial fixed-bearing insert. Only after this intrinsic ankle alignment has been achieved are the final bone cuts for the tibial keel and plug completed, fixing the tibial base and insert assembly into the optimized position. The instrumentation ensures proper positioning of the tibial implant and UHMWPE insert in relation to the talar implant. The same dedicated instrumentation, including an external alignment jig with tibial and pedal referencing, is used to resect the distal tibia and talus and insert the trial and final prosthesis components. The accuracy of tibial component alignment using this extramedullary referencing guide was tested in 83 ankles and determined to be within a mean of 1.5
and 4.1
in the coronal and sagittal planes, respectively, from the surgeon’s intended position (25). The anatomic design of the talar component is intended to reproduce normal ankle kinematics without overstressing the deltoid ligament complex. The method of fixation for the tibial base and talar implant was not altered from that used with the SaltoÒ mobile version design. The contact area between the UHMWPE insert and talar component allows 2-mm varus and valgus motion, 5
of internal and external rotation, 2 mm of anterior-to-posterior translation, and a sagittal plane arc of motion from 20
dorsiflexion to 25
plantar flexion. Total ankle replacement is a demanding procedure that can ultimately fail for a variety of reasons and require revision. As proposed by Henricson et al (26), the specific definitions for secondary procedures include revision, defined as any removal or exchange of 1 or more of the implant components, except for incidental exchange of the UHMWPE insert (i.e., metallic component replacement, custom implant usage, ankle arthrodesis, or below-the-knee amputation);

T.S. Roukis, A.D. Elliott / The Journal of Foot & Ankle Surgery 54 (2015) 311–319

Potentially relevant articles recognized through search in digital databases with duplicates removed (n=92) Articles excluded on abstract review (n=18): Technique only (n=5) Biomechanical testing only (n=5) Range of motion study only (n=3) Case report (n=3) Cadaver only (n=2) Full-text articles reviewed for detailed comparison with inclusion criteria (n=74) Articles excluded on full-text review (n=61): Total ankle replacement review only (n=17) Other implant design (n=15) Anatomical study only (n=7) Textbook chapter without data (n=6) Registry data unable to separate information (n=5) Radiographic analysis only (n=4) Duplicate data (n=5) Gait analysis only (n=2)

Articles included in qualitative synthesis for systematic review (n=13)

SALTO Mobile Version Total Ankle (n=8)

SALTO TALARIS Anatomic Ankle (n=5)

Fig. Flowchart of the articles during the selection process.

reoperation, defined as nonrevision secondary surgery involving the joint (i.e., debridement, incidental UHMWPE insert exchange, or wound treatment); and an additional procedure, defined as nonrevision secondary surgery not involving the joint (e.g., ligament reconstruction or release, adjacent joint arthrodesis, adjacent periarticular osteotomy, or tendon lengthening or transfer). Using these definitions, the incidence of revision after primary implantation of the AgilityÔ total ankle replacement system (DePuy Orthopaedics, Warsaw, IN) has been 10.2% (240 revisions of 2,353 primary implants) at a weighted mean follow-up period of 24.1 months (27,28). Specifically, 77.1% (185 of 240) of the revisions consisted of implant component replacement followed by ankle arthrodesis (44 of 240; 18.3% of revisions) and below-the-knee amputation (11 of 240; 4.6% of revisions) (27,28). Excluding the inventor, increased the incidence nearly twofold, from 7.8% (84 of 1,074) to 13% (172 of 1,320), implicating a potential selection (inventor) bias (27,28). All the studies included in the present systematic review involved an uncemented AgilityÔ total ankle replacement prosthesis, which is against the FDA requirements for the 510(k) cleared use of this prosthesis. Furthermore, the implant evaluated was the version available for use from 1998 to 2007; however, the exact version of the talar component implanted (i.e., original, posterior augmented, revision) could not be determined. Finally, data pertaining to the AgilityÔ LP total ankle replacement system, which became available for use in 2007, have not been published. However, an FDA clinical trial conducted by those other than the inventor or paid consultants, which was completed in November 2012, determined an incidence of revision of 6% (3 of 50) at a mean follow-up period of 24 months (http:// www.clinicaltrials.gov/ct2/show/results/NCT01366872?term¼Agilityþ LP&rank¼1§¼X867015#outcome3). They did note radiographic findings of talar subsidence at the final follow-up visit in 10 (20%), both talar and tibial subsidence in 5 (10%), and tibial subsidence in 1 (2%). Because metallic component subsidence is a known potential precursor to revision (27,48), the overall incidence of metallic component subsidence of 32% (16 of 50) is a cause for concern. Thus, it would be beneficial for these investigators to publish their mediumand long-term follow-up data from these patients.

313

Similarly, the incidence of revision after primary implantation of the Scandinavian Total Ankle ReplacementÔ system (STARÔ, SBi, Morrisville, PA/Stryker Orthopaedics, Mahwah, NJ) has been determined to be 10.7% (269 of 2,507) at a weighted mean follow-up period of 64 months (29,30). The specific revision performed was not clearly defined for 51.7% (139 of 269) prostheses. For the remaining prostheses, revision consisted of metallic component replacement in 50.7% (66 of 130) followed by ankle arthrodesis (62 of 130; 47.7%) and below-the-knee or above-the-knee amputation (2 of 130; 1.5% of revisions). Excluding the inventor or SBi faculty consultants increased the incidence more than twofold, from 5.6% (45 of 807) to 13.2% (224 of 1,700), implicating potential selection (inventor) and publication (conflict of interest) bias. The remaining metal-backed, fixed-bearing, cemented total ankle replacement devices that have been FDA 510(k) cleared for use (INBONEÒ I, INBONEÒ II, and INFINITYÔ Total Ankle Replacement Systems, Wright Medical Technology, Inc, Arlington, TN; Eclipse, Integra LifeSciences, Plainsboro, NJ; and Zimmer Trabecular Metal Total Ankle, Zimmer, Inc, Warsaw, IN) have no published clinical data; thus, a systematic review was not possible. Therefore, in an effort to determine the incidence of revision after primary implantation of the SaltoÒ mobile version and Salto TalarisÔ total ankle prostheses, we undertook a systematic review of electronic databases and other relevant sources. Materials and Methods We used the “Preferred Reporting Items for Systematic Reviews and Meta-Analyses” guidelines (30). Accordingly, electronic databases, including the Cochrane Database of Systematic Reviews (http://www.cochrane.org/reviews/); Cochrane Methodology Register (http://www.thecochranelibrary.com/view/0/index.html); EMBASE (Excerpta Medica database at http://www.embase.com/); and PubMedÒ (http://www. ncbi.nlm.nih.gov/pubmed) were searched from inception to December 2014, with no restriction on date or language, using an inclusive text word query for “Salto” OR “Salto Talaris” AND “total ankle replacement” OR “ankle replacement” OR “ankle implant” OR “ankle arthroplasty” OR “revision surgery” OR “failed surgery”; all capitalized words represent the Boolean operators used. The search terms were purposely broad to encompass all possibilities for applicable studies. To maximize the number of potentially useful references, every combination of text words was queried through each of the electronic databases. Additionally, the references were manually searched from all reports obtained, and all potential pertinent references were secured for review. All studies were cross-referenced to ensure that no relevant studies had been overlooked. Also, major podiatric and orthopedic foot and ankle surgery textbooks with chapters on total ankle replacement were obtained to secure any useful material not accessible from the electronic search. Finally, using various combinations of the text words listed, an Internet-based scholarly data search engine, specifically GoogleÔ Scholar (http:// scholar.google.com/) was used to identify available sources that could potentially provide useful information. The study data determined to be of interest a priori included the following: patients undergoing primary SaltoÒ mobile version or Salto TalarisÔ total ankle prostheses; evaluation of patients at a mean follow-up period of 12 months or longer; details of complications and, specifically, the incidence of revision (defined as metallic component replacement, ankle arthrodesis, or below-the-knee amputation); and revision etiologies of aseptic loosening, component subsidence, ballooning osteolysis, cystic changes, malalignment, or instability but not takedown of an ankle arthrodesis, revision of another failed ankle replacement, malleolar fracture treatment, synovectomy, arthrolysis, impingement debridement, isolated or incidental UHMWPE insert exchange, superficial or deep infection–related complications, or incision- or wound-related complications. If a reference could not be obtained through purchase, librarian assistance, or electronic mail contact with the investigator, it was excluded from consideration. If the study had not been written in English, the entire content of the reference was translated by the investigator from its native language of German or French into English using both a text- and a computer-based translation software program (i.e., BabylonÒ for MAC 3.1; Babylon.com Ltd, http://www.babylon.com/). Finally, the level of evidence of each individual study was determined according to the Evidence Based Medicine Grading System recommended by the American College of Foot and Ankle Surgeons (http://www.jfas.org/authorinfo).

Results The search for potentially eligible information for inclusion in the systematic review yielded a total of 92 references that underwent

314

T.S. Roukis, A.D. Elliott / The Journal of Foot & Ankle Surgery 54 (2015) 311–319

Table Study data included in the present systematic review Author (y)

EBM Peer Specific Reviewed Study Data

Prosthesis

Period Implants Mean Age Inserted (range) y

Follow-Up (range) mo

Total Incidence of Revision Procedures Performed

106.8 (81.6 to 133.2)

8/87 (9.2%): 1 implant component replacement, 6 arthrodeses, 1 below-knee amputation

Salto mobile version Bonnin et al (31) (2004) Weber et al (32) (2004) Bonnin et al (33) (2011)

IV IV IV

Yes Yes Yes

Inventor

87: NA R, L; NA M, F

1997 to 2000

57 (26 to 81)

Bonnin et al (34) (2009)

III

Yes

Inventor

179: 74 R, 66 L; 50 M, 90 F 1997 to 2005

60.9 (26 to 90)

Reuver et al (35) (2010)

IV

Yes

NA

59: 30 R, 29 L; 28 M, 27 F

Schenk et al (36) (2011)

IV

Yes

56.8 (NA)

Colombier et al (24) (2012) IV

No

Paid 218: 118 R, 100 L; NA M, F 2000 to 2007 consultant Inventor 33: 12 R, 21 L; 17 M, 15 F 2001 to 2003

Tomlinson et al (37) (2012) IV New Zealand Joint Registry IV 2013 (38) (2013)

No No

Registry

316: NA R, L; NA M, F

2005 to 2012

Australian Joint Registry 2012 (39) (2013)

IV

No

Registry

198: NA R, L; NA M, F

Rodrigues-Pinto et al (40) (2013) Rodrigues-Pinto et al (41) (2013)

III

Yes

NA

119: NA R, L; NA M, F

III

Yes

IV

No

Schweitzer et al (43) (2012) IV Schweitzer et al (44) (2013) IV Gaudot et al (24) (2014)

2003 to 2007

57 (34 to 77)

53.8 (12 to 125) 4/179 (4.5%): 2 implant component replacements, 2 arthrodeses 36 (12 to 65)

7/59 (11.9%): 2 implant component replacements, 4 arthrodeses, 1 below-knee amputation

42.3 (24 to 84)

12/218 (5.5%): 12 arthrodeses

114 (NA)

3/33 (3%): 3 implant component replacements

NA

NA

6/316 (1.9%): no procedures listed

2005 to 2012

NA

NA

8/198 (4.0%): no procedures listed

2005 to 2011

55.6 (24 to 81)

38.7 (18 to 72)

0/119 (0%): no procedures listed

Design team 14: 4 R, 10 L; 10 M, 4 F

2007 to 2008

53 (38 to 71)

24 (NA)

0/14 (0%): no procedures listed

No Yes

Paid 67: NA R, L; 18 M, 49 F consultant

2007 to 2009

63 (34 to 86)

III

Yes

Inventor

33: 12 R, 21 L; 18 M, 49 F

2007 to 2009

64 (38 to 79)

Nodzo et al (45) (2014)

IV

Yes

NA

75: 41 R, 34 L; 33 M, 41 F

2007 to 2011

61 (36 to 86)

Chao et al (46) (2014)

IV

Yes

NA

23: 10 R, 13 L; 6 M, 17 F

2008 to 2009

Salto TalarisÔ total ankle Mehta et al (42) (2010)

64 (38 to 79)

33.6 (24 to 53.5) 2/67 (3%): 2 implant component replacements 24 (11 to 60)

1/33 (3%): 1 arthrodesis

43.2 (24 to 74.4) 1/75 (1.3%): 1 implant component replacement 68.6 (53.2 to 85.4) 36 (24 to 49) 1/23 (4.3%): 1 arthrodesis

Abbreviations: EBM, evidence-based medicine; F, female; L, left; M, male; NA, not applicable; R, right.

stepwise analyses as detailed in the Fig. All pertinent references identified were obtained and reviewed by us between October 2013 and December 2014. After considering all the potentially eligible references, 8 studies involving 1,209 SaltoÒ mobile version prostheses with a weighted mean follow-up period of 55.2 months were included (Table) (20,24,31–33,35,36,38–41). Overlapping patients were evident for 3 data sets involving the inventors (31–33), registry data (37,38), and independent surgeons (40,41). In these instances, we considered each group to represent 1 study and abstracted the pertinent information without duplicating the data. For the studies that included it, the weighted mean patient age was 58.2 (range 24 to 90) years, with a slight preponderance of female gender (170 women versus 144 men) and right ankles (222 right versus 195 left). Five studies involving 212 Salto TalarisÔ total ankle prostheses, with a weighted mean follow-up period of 34.9 months, were included (Table) (24,42–46). Overlapping patients were evident for 1 data set involving disclosed consultants (43,44), and the data were abstracted as noted. The weighted mean patient age was 62.3 (range 34 to 86) years, with a preponderance of female gender (126 women versus 84 men) and essentially equal laterality (76 left and 67 right ankles) involved. Data isolated to the inventor, design team, or disclosed consultants had an incidence of revision of 5.2% (27 of 517) for the SaltoÒ mobile version and 2.6% (3 of 114) for the Salto TalarisÔ total ankle prostheses. In contrast, data that excluded these individuals had an incidence of

revision of 2.8% (19 of 692) for the SaltoÒ mobile version and 2.0% (2 of 98) for the Salto TalarisÔ total ankle (Table) prostheses (20,24,31–45). Forty-eight SaltoÒ mobile version prostheses (4%) underwent revision; however, specific revision information was only available for 34. Of these, 24 patients (70.5%) underwent ankle arthrodesis, 9 (26.5%) metallic component replacement, and 1 (3%) below-the-knee amputation. The metallic component replacement performed was specifically mentioned for 6 revisions, with tibial component exchange occurring 1 (16.7%) and talar component exchange 5 (83.3%) times. Five (2.4%) Salto TalarisÔ total ankle prostheses underwent revision (3 metallic component replacements and 2 ankle arthrodeses). The specific metallic component exchange was reported once and consisted of conversion to a different total ankle replacement system (44). The etiology leading to the 48 SaltoÒ mobile version prosthesis revisions were not specifically mentioned 30 times (62.5%). The specific etiology of the remaining 18 revisions included aseptic loosening of the tibial and/or talar component 11 times (61.1%), talar subsidence 4 times (22.2%), talar osteonecrosis 2 times (11.1%), and tibial component aseptic loosening 1 time (5.6%). The specific etiology leading to the 5 Salto TalarisÔ total ankle prosthesis revisions included aseptic loosening of the tibial and/or talar component, talar subsidence, tibial subsidence, talar osteonecrosis, and tibial component aseptic loosening 1 time each. Although no study for homogeneity was performed, and, likely, a number of methodologic differences were present, the incidence of

T.S. Roukis, A.D. Elliott / The Journal of Foot & Ankle Surgery 54 (2015) 311–319

revision, at face value, appeared to be consistent among the entire data cohort, regardless of the prosthesis UHMWPE-bearing type. However, the difference between the studies that did not include the inventor or paid consultants and those that only included the inventor or paid consultants revealed a decrease in the incidence of revision surgery for the studies that did not include the inventor or paid consultants and skewing of the revision procedures toward arthrodesis for the studies that only included the inventor or paid consultants. The methodologic quality of the included studies was generally poor. For the 8 studies included for the SaltoÒ mobile version prosthesis, 2 were level III (24,40,41) and 6 were level IV (20,31–33,35,36,38,39) evidence-based medicine studies. For the 5 studies included in the Salto TalarisÔ total ankle prosthesis, 1 was level III (24) and 4 were level IV (42–46) evidence-based medicine studies. Finally, none of the included reports stated that polymethylmethacrylate cement fixation was used for the Salto TalarisÔ total ankle prosthesis, despite this being included in the surgeon technique guide (http://www.tornierdx.info/pdf/SALTO_SXT_INST2_ UJAT092-1oct09-a.pdf). Discussion The purpose of the present systematic review was to evaluate the best evidence available regarding the incidence of revision after primary implantation of the SaltoÒ mobile version prosthesis and uncemented Salto TalarisÔ total ankle prosthesis. A total of 8 studies involving 1,209 SaltoÒ mobile version prostheses, with a weighted mean follow-up period of 55.2 months (20,24,31–33,35,36,38–41), and 5 studies involving 212 Salto TalarisÔ total ankle prostheses, with a weighted mean follow-up period of 34.9-months (24,42–46), could be identified that met the inclusion criteria. A review of these data allowed for some generalized statements regarding the incidence of revision after primary implantation of the SaltoÒ mobile version prosthesis and Salto TalarisÔ total ankle prosthesis. First, the overall incidence of revision was 4% for the SaltoÒ mobile version prosthesis, with 70.5% of the reported revisions consisting of ankle arthrodesis, followed by metallic implant component replacement (26.5% of revisions) and below-the-knee amputation (3% of revisions). In contrast, 5 (2.4%) Salto TalarisÔ total ankle prostheses underwent revision, consisting of 3 metallic component replacements and 2 ankle arthrodeses. The importance of these findings is fivefold. First, the incidence of revision after primary implantation of either the SaltoÒ mobile version prosthesis (4% at a weighted mean follow-up period of 55.2 months) or the Salto TalarisÔ total ankle prosthesis (2.4% at a weighted mean follow-up period of 34.9 months) was less than that reported for the AgilityÔ total ankle replacement system (10.2% at a weighted mean follow-up period of 24.1 months) (27,28) or the Scandinavian Total Ankle ReplacementÔ system (10.7% at a weighted mean follow-up period of 64 months) (29). These findings are important because patients in the United States can only undergo total ankle replacement with a limited number of prostheses. Before our current report, only 2 of the prostheses had enough data available to have undergone a detailed analysis by systematic review (27,29). Furthermore, analysis of prosthesis use and disuse trends during a 10year period for worldwide joint registries revealed a significant decline in implantation for both the AgilityÔ and Scandinavian Total Ankle ReplacementÔ systems, with a marked increase in the use of SaltoÒ mobile version prosthesis but unknown usage of the Salto TalarisÔ total ankle prosthesis (47). These are significant considerations because an understanding of the actual incidence of revision, the most common etiology leading to failure, and the revision options for each prosthesis system, is critical to optimizing patient outcome. Second, the etiology leading to revision for the SaltoÒ mobile version prosthesis included predominantly aseptic loosening. Too few data were

315

available to identify the etiology leading to revision for the Salto TalarisÔ total ankle prosthesis in the present systematic review. However, 1 radiographic complication study of 56 Salto TalarisÔ total ankle prostheses revealed that the most common etiologies leading to revision were aseptic osteolysis greater than 2 mm in 16% and metallic component subsidence in 16% (48). Although not conclusive, the location of osteolysis appears different between the SaltoÒ mobile version and Salto TalarisÔ total ankle prostheses. Specifically, the SaltoÒ mobile version prosthesis has demonstrated a greater overall incidence of osteolysis, and the pattern has been concentrated around the tibial keel and conical fixation plug and the hollow talar fixation plug (14,24). The Salto TalarisÔ total ankle prosthesis demonstrated osteolysis most commonly about the tibial base plate (24,48). This is important because the development of aseptic osteolysis with the SaltoÒ mobile version prosthesis is considered multifactorial. These factors include the initial use of a thin UHMWPE insert (i.e., 3 mm), which was susceptible to fracturing and discontinued; malleolar impingement from UHMWPE insert overhang despite the medial tibial base plate fin intended to reduce this; backside wear between the highly congruent tibial base plate and mobile UHMWPE insert; malalignment between the tibial tray and talar component causing overhanging and/or rotation imperfection between the components, leading to high contact stress and accelerated UHMWPE wear; and hydroxyapatite coating, which was discontinued in 2012. Concerns regarding UHMWPE insert fracture, subluxation, dislocation, and excessive wear are common to all modern mobilebearing total ankle replacement prostheses (49–51). UHMWPE malalignment leading to excessive wear and aseptic osteolysis has been reported most commonly with the Scandinavian Total Ankle ReplacementÔ system (52,53). The use of the hydroxyapatite coating has been implicated as the major cause of the severe aseptic osteolysis seen with the Ankle Evolutive SystemÒ (Transysteme-JMT Implants, Nimes, France) (54–58) that was subsequently withdrawn from use in 2012 (http://www.mhra.gov.uk/home/groups/dts-bs/documents/ medicaldevicealert/con174792.pdf). Similarly, a recent long-term follow-up of 77 Scandinavian Total Ankle ReplacementÔ prostheses with hydroxyapatite coating revealed a 70.7% survival (i.e., no metallic component revision) at 10 years, which had declined to 45.6% at 14 years (59). These studies raise concern regarding the use of hydroxyapatite coating for biologic osseous ingrowth in any total ankle prosthesis system. Additionally, the initial surgical technique for the SaltoÒ mobile version prosthesis left the anterior tibial cortex defect about the tibial keel or conical fixation plug and the talar drill holes for the sizing and/or cutting guides unfilled. In contrast, later techniques involved impaction bone graft filling of these defects. Leaving these osseous defects unfilled has been identified as an independent cause of aseptic osteolysis with the AgilityÔ Total Ankle Replacement system about the recessed tibial keel, the surgical half-pin path through the talar neck used for distraction, and the largely uncovered talar body surface with talar components other than the AgilityÔ LP total ankle replacement system (60). Even in a well-aligned total ankle prosthesis, leaving these osseous defects unfilled will allow access for phagocytosable UHMWPE wear debris to migrate along the bone–prosthesis interface, where they will induce a macrophage-mediated biologic cascade that will culminate in osteoclast-mediated bone loss (i.e., aseptic osteolysis) and secondary loosening and/or metallic component subsidence (61). In addition to biologically active processes as a cause for aseptic osteolysis, disruption of the talar vascular supply secondary to the anterior surgical approach and the bone preparation required to accept the metallic talar component for total ankle replacement can induce osteonecrosis and subsidence, with subsequent prosthetic malalignment (62). Although the extra- and intraosseous vascular anatomy of the talus has been studied for 65 years (63–68), it has only been recently that the risks of damage to the talar blood supply during soft tissue dissection and bone preparation requirements for total ankle replacement (69,70) have been studied

316

T.S. Roukis, A.D. Elliott / The Journal of Foot & Ankle Surgery 54 (2015) 311–319

in detail. The SaltoÒ mobile version and Salto TalarisÔ total ankle prostheses have the same anterior ankle approach to expose the ankle joint that exploits the interval between the tibialis anterior tendon and extensor hallucis longus tendon deep to which the anterior neurovascular bundle resides. During this soft tissue dissection, the medial talar head and neck blood supply from the dorsalis pedis is at high risk of injury. This pattern of arterial injury is universal for total ankle prostheses that use an anterior approach (69,70). Knowing this, limited medial gutter debridement and care taken to minimize soft tissue stripping about the medial talar head and neck would seem prudent. Damage to the posterior tibial artery at the distal part of the tibia presents a moderate risk of injury during the preparation of the tibial bone surfaces and can be minimized by using exacting technique (69). Damage to the artery of the tarsal canal was considered a moderate risk of injury and can occur during placement of the various talar resection dorsal-to-plantar metallic pins if a pin is inadvertently driven through the inferior surface of the talus (69). It is intuitive that proper metallic pin insertion technique and liberal use of intraoperative C-arm image intensification will reduce this risk pattern. This pattern is in contrast to that with the Scandinavian Total Ankle ReplacementÔ system, which uses sculpting or resurfacing of the entire talar body to accept the talar component. This adds a very high risk of injury to the posteromedial deltoid branches above the medial articular talar facet with the medial chamfer cut (69). Furthermore, the long central talar fin will damage the intraosseous artery arising from the anastomotic artery of the tarsal canal that is the main arterial supply of at least the medial two thirds of the talar body (70). Therefore, the extensive remodeling and sculpting of the talar body required for implantation of the Scandinavian Total Ankle ReplacementÔ system could be partially responsible for the high incidence of aseptic osteolysis and subsequent component subsidence seen with this prosthesis (29,52,53). Finally, unequal dispersion of the load about the subarticular bone adjacent to the metallic prosthesis components could be responsible for aseptic osteolysis and loosening. Specifically, the location of osteolysis concentrated around the tibial keel and conical fixation plug for the SaltoÒ mobile version prosthesis correlates exactly with the increased load transmission and strain during thermoelastic stress analysis (71) and finite element modeling (72). Therefore, additional refinement of the tibial component to redistribute stress for more even dispersion of load should be an area of future study. Third, according to the inventors (24), the Salto TalarisÔ total ankle prosthesis was developed after in vivo kinematics of the SaltoÒ mobile version prosthesis revealed 1 to 1.5 mm of translation between the UHMWPE insert and tibial base plate (22,23), demonstrating that the insert was not functioning as a mobile bearing but, rather, remained essentially fixed to the tibial component. During the redesign, specific attention was given to improve the tibial base plate length and width for maximum cortical coverage, alter the talar component morphology to optimize the balance between the triplane range of motion tolerance and constraint, and redesign the instrumentation to improve the accuracy of implantation. The fixed-bearing Salto TalarisÔ total ankle prosthesis is currently under clinical evaluation in Europe (24), where implantation of mobile-bearing total ankle prostheses has dominated (47). Although not proved, the inventors of the SaltoÒ mobile version prosthesis have stated that the rationale for developing the fixed-bearing Salto TalarisÔ total ankle prosthesis was to obtain improved centering of the talar component relative to the tibial component and the ability to use smaller metallic components (24). They hypothesized that both of these attributes will reduce malleolar impingement and the secondary trabecular bone changes associated with UHMWPE insert particulate wear debris (24). Their efforts have drawn some criticism, with the main concern involving the perceived difficulty in obtaining precise implantation of fixedbearing devices (62) and the ability of mobile-bearing prostheses to “.find the correct position of the components” (73). However, a

recent report evaluating active weightbearing motion of the mobilebearing FINEÔ Total Ankle System (Nakashima Medical, Okayama, Japan) in 12 rheumatoid ankles revealed anteroposterior translation between the tibial plate and UHMWPE insert of 1.6 mm (74). Similarly, the active weightbearing motion of the Scandinavian Total Ankle ReplacementÔ system in 15 ankles revealed anteroposterior translation between the tibial plate and UHMWPE insert of only 0.7 mm, with a theoretical rotation of up to 3.3
(75). The results from both of these studies indicated that the UHMWPE mobile bearing in these total ankle replacement systems was functioning as a fixed bearing. Furthermore, using biomechanical cadaveric analysis, the Scandinavian Total Ankle ReplacementÔ system (76) does not appear to tolerate malalignment any more than other mobile-bearing (77–79) or fixed-bearing (79–82) prostheses. Taken collectively, total ankle replacement requires precise implantation because mobile-bearing UHMWPE inserts or more incongruity between the fixed-bearing UHMWPE insert and talar component geometry are incapable of predictably accommodating appreciable malalignment. Finally, the € echel-Pappas total ankle system withdrawal in 2009 of the Bu (Endotec, South Orange, NJ) (83), which was the predicate mobilebearing design and had good 20-year patient outcomes and survivorship (84), combined with the good 14-year patient outcomes and survivorship with the ESKA fixed-bearing total ankle (GmbH & Co, € beck, Germany) (85), and lack of clinically significant differences in Lu outcomes or survivorship between mobile- and fixed-bearing prosthesis types (16) indicates that the bearing type should not be the main criteria for surgeon adoption of a particular prosthesis. Fourth, as implied, the vast majority of revisions for the SaltoÒ mobile version prostheses have consisted of ankle arthrodesis, much less frequently, implant component replacement, and, only in exceptionally rare catastrophic cases, below-the-knee amputation. Furthermore, when an implant component replacement has been performed, it has consisted predominantly of isolated tibial or talar replacement and only occasionally dual component replacement or conversion to another implant design. The outcomes of revised primary SaltoÒ mobile version and Salto TalarisÔ total ankle prosthesis deserves additional investigation because no meaningful data exist, despite the recent development of the SALTO XTÔ revision ankle prosthesis (Tornier, Inc, Bloomington, MN) with size 1, 2, and 3 talar components containing a flat cut and longer peg and 10-, 12-, 14-, 16-, and 18-mm-thick revision UHMWPE inserts. The tibial components are the same as the size 1, 2, and 3 SaltoÒ mobile version and Salto TalarisÔ total ankle prosthesis, but a custom, stemmed tibial component with a long central keel is available for use in the United States by surgeon prescription only. The ability for the SALTO XTÔ revision ankle prosthesis to revise failed AgilityÔ and Scandinavian Total Ankle ReplacementÔ systems is intriguing but currently unproved. Clearly, a real need exists for outcome studies to evaluate patients undergoing revision total ankle replacement for the current prosthesis systems available in the United States, and future efforts ought to be directed in this area. Finally, data isolated to the inventor, design team, or disclosed consultants had an incidence of revision of 5.2% (27 of 517) for the SaltoÒ mobile version and 2.6% (3 of 114) for the Salto TalarisÔ total ankle prostheses. In contrast, data that excluded these individuals had an incidence of revision of 2.8% (19 of 692) for the SaltoÒ mobile version and 2.0% (2 of 98) for the Salto TalarisÔ total ankle prosthesis (20,24,31–46). This was an unexpected finding because a major criticism of the studies that have involved inventors, design team members, or disclosed consultants is that they, by definition, represent a special interest group with much greater experience in using the particular implant design than the average surgeon. This effect has been clearly demonstrated to exist for the incidence of revision after primary implantation of the AgilityÔ (27) and Scandinavian Total

T.S. Roukis, A.D. Elliott / The Journal of Foot & Ankle Surgery 54 (2015) 311–319

Ankle ReplacementÔ (29) systems. As such, we expected to find a lower incidence of revision for these special interest groups compared with studies that did not involve these individuals. Instead, we found them to be twice as great. The cause for this is uncertain but could be explained by the longer term follow-up data presented by the inventors (20,24,31–34) and paid consultants (36) for the SaltoÒ mobile version prosthesis. This could also have resulted from the myriad design changes that have occurred over time for the SaltoÒ mobile version prosthesis components, surgical technique, and instrumentation. All these changes have in 1 form or another been implemented in the initially released and current Salto TalarisÔ total ankle prosthesis. The potential for a learning curve effect could also exist because the implantation of total ankle replacement systems in the United States has evolved from limited use in the late 1990s to commonplace at present (86,87). For instance, the SaltoÒ mobile version prosthesis data included this learning curve (20,24,31–34,36); however, the Salto TalarisÔ Total Ankle Prostheses data included data from surgeons experienced with the implantation of total ankle replacement systems in general (24,42–46). However, the true effect, if any, a potential learning curve has on the incidence of revision for the SaltoÒ mobile version and Salto TalarisÔ total ankle prostheses remains a matter for conjecture. The weaknesses of the present study included that, first, most studies involving the SaltoÒ mobile version and Salto TalarisÔ total ankle prosthesis were retrospective, uncontrolled, observational studies (i.e., level IV). Second, our systematic review only evaluated the incidence of revision, defined as any removal or exchange of 1 or more of the implant metallic components, except for the incidental exchange of the UHMWPE insert, and not reoperation, defined as nonrevision secondary surgery involving the joint, or additional procedures, defined as nonrevision secondary surgery not involving the joint (26). Reoperation and additional procedures represent subsequent surgical intervention separate from the primary total joint replacement, and the significance of these events must not be downplayed. However, the complexity associated with revision surgery and the largely unknown outcomes after implant component replacement specifically has made focusing attention only on the incidence of revision, rather than all secondary procedures, worthwhile. This has created much confusion over the actual survivorship of total ankle replacements and is an area in which clarity is necessary for future studies. By so doing, it will be possible to compare the incidence of revision and percentages of each type of revision surgery among implant designs in a more meaningful manner. Comparative studies evaluating the incidence of revision involving sufficiently large volumes of total ankle replacements are obviously warranted. Third, the weighted mean follow-up period for the SaltoÒ mobile version prosthesis was 55.2 months and for the Salto TalarisÔ total ankle prostheses was 34.9 months, and these were likely too brief for most ankles to develop complications requiring revision. Additionally, few of the studies included in the present systematic review provided information regarding potential or impending revisions. Furthermore, deep infection was excluded from analysis, and this would certainly have increased the incidence of revision surgery, predominantly arthrodesis and below-the-knee amputation procedures. Fourth, all the data available for the Salto TalarisÔ total ankle prosthesis involved implantation without polymethylmethacrylate cement that is against the FDA 510(k) cleared procedure. The effect this had on the incidence of revision encountered is unknown. Finally, as a result of the electronic searches, it is possible that pertinent references could have been inadvertently excluded or overlooked. However, the inclusion criteria were intentionally defined in broad terms to include as many references as possible and reduce the bias. Similarly, because we limited our search to only certain languages, we might have induced bias toward accepting reports published in English-, German-, and

317

French-speaking countries. However, our exhaustive electronic search was further supplemented by manually searching all the references. With such a detailed search, it is unlikely that pertinent references that would have altered our conclusions were overlooked. The inclusion of only published reports might have introduced a publication bias into our conclusions, because studies with poorer outcomes might have elected not to publish their data after presenting their report at a congress or similar event. We had no method of controlling for this, but we believe the importance of registry data involving total ankle replacements should be strongly considered in all countries performing these procedures (47). In conclusion, despite these weaknesses, the results of the present systematic review have made clear that the incidence of revision after primary implantation of the SaltoÒ mobile version prostheses was 4% at a weighted mean follow-up period of 55.2 months and the Salto TalarisÔ total ankle prostheses was 2.4% at a weighted mean followup period of 34.9 months. Of the failed SaltoÒ mobile version prostheses, 71% underwent revision with ankle arthrodesis, 26% implant component replacement, and 3% below-the-knee amputation. Data isolated to the inventor, design team, or disclosed consultants had an incidence of revision of 5.2% for the SaltoÒ mobile version and 2.6% for the Salto TalarisÔ total ankle prostheses. In contrast, data that excluded these individuals had an incidence of revision of 2.8% for the SaltoÒ mobile version and 2.0% for the Salto TalarisÔ total ankle prostheses. We could not identify any obvious difference in the etiology responsible for, or incidence of, revision between these fixedand mobile-bearing prosthesis systems. The incidence of revision for the SaltoÒ mobile version and Salto TalarisÔ total ankle prostheses was lower than those reported through systematic review for the uncemented AgilityÔ and Scandinavian Total Ankle ReplacementÔ systems, without obvious selection (inventor) or publication (conflict of interest) bias. However, there clearly is still a need for methodologically sound cohort studies that focus on outcomes after revision of the SaltoÒ mobile version and Salto TalarisÔ total ankle prostheses, including implant component replacement techniques that are effective and the SALTO XTÔ ankle revision prosthesis and custom stemmed components. References 1. Vickerstaff JA, Miles AW, Cunningham JL. A brief history of total ankle replacement and a review of the current status. Med Eng Phys 29:1056–1064, 2007. 2. Guyer AJ, Richardson G. Current concepts review: total ankle arthroplasty. Foot Ankle Int 29:256–264, 2008. 3. DeOrio JK, Easleu ME. Total ankle arthroplasty. AAOS Instr Course Lect 57:383–413, 2008. 4. Gougoulias NE, Khanna A, Maffulli N. History and evolution in total ankle arthroplasty. Br Med Bull 89:111–151, 2009. 5. Smith TWD, Stephens M. Ankle arthroplasty. Foot Ankle Surg 16:53, 2010. 6. Gougoulias NE, Khanna A, Maffulli N. How successful are current ankle replacements? A systematic review of the literature. Clin Orthop Relat Res 468:199–208, 2010. 7. van den Heuvel A, Van Bouwel S, Dereymaeker G. Total ankle replacement: design evolution and results. Acta Orthop Belg 76:150–161, 2010. 8. Easley ME, Adams SB, Hembree WC, DeOrio JK. Current concepts review: results of total ankle arthroplasty. J Bone Joint Surg 93-A:1455–1468, 2011. 9. Haddad SL, Coetzee JC, Estok R, Fahrbach K, Banel D, Nalysnyk L. Intermediate and long-term outcomes of total ankle arthroplasty and ankle arthrodesis. J Bone Joint Surg Am 89-A:1899–1905, 2007. 10. SooHoo NF, Zingmond DS, Ko CY. Comparison of reoperation rates following ankle arthrodesis and total ankle arthroplasty. J Bone Joint Surg Am 89-A:2143–2149, 2007. 11. Bestic JM, Bancroft LW, Peterson JJ, Kransdorf MJ. Postoperative imaging of the total ankle arthroplasty. Radiol Clin North Am 46:1003–1015, 2008. 12. Pappas MJ, Buechel FF Sr. Failure modes of current total ankle replacement systems. Clin Podiatr Med Surg 30:123–143, 2013. 13. Besse JL, Colombier JA, Asencio J, Bonnin M, Gaudot F, Jarde O, Judet T, Maestro M,  Total ankle arthroplasty in France. Orthop Trauma Lemrijse T, Leonardi C, Toullec E. Surg Res 96:291–303, 2010.  Henry M, Neron J-B, Mabit C, Brilhault J. Total ankle 14. Preyssas P, Toullec E, arthroplastydthree-component total ankle arthroplasty in western France: a radiographic study. Orthop Trauma Surg Res 98:531–540, 2012.

318

T.S. Roukis, A.D. Elliott / The Journal of Foot & Ankle Surgery 54 (2015) 311–319

15. Younger A, Penner M, Wing K. Mobile-bearing total ankle arthroplasty. Foot Ankle Clin 13:495–508, 2008. € ller AM, Paul J, Henninger HB, Barg A. Mobile16. Valderrabano V, Pagenstert GI, Mu and fixed-bearing total ankle prostheses: is there really a difference? Foot Ankle Clin 17:565–585, 2012. 17. Bonnin M, Donley B, Judet T, Colombier J-A. The Salto and Salto-Talaris total ankle arthroplasty. In: Operative Techniques in Foot and Ankle Surgery, pp. 562–575, edited by ME Easley, SW Weisel, Lippincott, Williams & Wilkins, Philadelphia, PA, 2011. se Totale de Cheville. In: Le pied en rhu18. Bonnin M, Judet T, Siguier T. La Prothe matologie, ed 2, pp. 417–431, edited by M Bouysset, Springer-Verlag, Paris, 2000. ge  ne rative 19. Piriou P, De Loubresse GC, Colombier J, Bonnin M, Judet T. Cheville de se: notre experience. In: La Cheville D post-traumatique et prothe eg en erative et Post-traumatique: La Place de la Proth ese en l’An 2000, 42, pp. 149–155, edited by risson, S Lucien, Elsevier, Masson SAS, France, 2000. C He 20. Colombier JA, Judet T, Bonnin M, Gaudot F. Techniques and pitfalls with the Salto prosthesis: our experience of the first 15 years. Foot Ankle Clin 17:587–605, 2012. 21. Leszko F, Komistek RD, Mahfouz MR, Ratron Y-A, Judet T, Bonnin M, Colombier J-A, Lin S. In vivo kinematics of the Salto total ankle prosthesis. Foot Ankle Int 29:1117– 1125, 2008. 22. Rush SM, Todd N. Salto Talaris fixed-bearing total ankle replacement system. Clin Podiatr Med Surg 30:69–80, 2013. 23. Yalamanchili P, Donley B, Casillas M, Ables M, Lin S. Salto Talaris total ankle replacement. Op Tech Orthop 18:277–281, 2008. 24. Gaudot F, Colombier J-A, Bonnin M, Judet T. A controlled, comparative study of a fixed-bearing versus mobile-bearing ankle arthroplasty. Foot Ankle Int 35:131– 140, 2014. 25. Adams SB Jr, Demetracopoulos CA, Viens NA, DeOrio JK, Easley ME, Queen RM, Nunley JA II. Comparison of extramedullary versus intramedullary referencing for tibial component alignment in total ankle arthroplasty. Foot Ankle Int 34:1624– 1628, 2013. 26. Henricson A, Carlsson  A, Rydholm U. What is revision of total ankle replacement. Foot Ankle Surg 17:99–102, 2011. 27. Roukis TS. Incidence of revision after primary implantation of the Agility total ankle replacement system: a systematic review. J Foot Ankle Surg 51:198–204, 2012. 28. Criswell BJ, Douglas K, Naik R. High revision and reoperation rates using the Agility total ankle system. Clin Orthop Rel Res 470:1980–1986, 2012. 29. Prissel MA, Roukis TS. Incidence of revision after primary implantation of the Scandinavian total ankle replacement system: a systematic review. Clin Podiatr Med Surg 30:237–250, 2013. 30. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gtzsche PC, Ioannidis JPA, Clarke M, Devereaux PJ, Kleijnen J, David Moher D. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. Ann Intern Med 151:65–94, 2009. 31. Bonnin M, Judet T, Colombier J-A, Buscayret F, Graveleau N, Piriou P. Midterm results of the Salto total ankle prosthesis. Clin Orthop Rel Res 424:6–18, 2004. 32. Weber M, Bonnin M, Columbier J-A, Judet T. [Preliminary results of the Salto total upper ankle prosthesis]. Fub Sprungg 2:29–37, 2004. 33. Bonnin M, Gaudot F, Laurent J-R, Ellis S, Colombier J-A, Judet T. The Salto total ankle arthroplasty: survivorship and analysis of failures at 7 to 11 years. Clin Orthop Rel Res 469:225–236, 2011. 34. Bonnin MP, Laurent J-R, Casillas M. Ankle function and sports after total ankle arthroplasty. Foot Ankle Int 30:933–944, 2009. 35. Reuver JM, Dayerizadeh N, Burger B, Elmans L, Hoelen M, Tulp N. Total ankle replacement outcome in low volume centers: short-term follow-up. Foot Ankle Int 31:1064–1068, 2010. 36. Schenk K, Lieske S, John M, Franke K, Mouly S, Lizee E, Meumann W. Prospective study of a cementless, mobile-bearing, third generation total ankle prosthesis. Foot Ankle Int 32:755–763, 2011. 37. Tomlinson M, Harrison M. The New Zealand joint registry: report of 11-year data for ankle arthroplasty. Foot Ankle Clin 17:719–723, 2012. 38. The New Zealand Joint Registry. Fourteen year report January 1999 to December 2012. Available at: http://www.nzoa.org.nz/system/files/NJR%2014%20Year% 20Report.pdf. Accessed April 14, 2014. 39. Australian Orthopaedic Association National Joint Replacement Registry. Demographics and outcome of ankle arthroplasty: supplementary report 2013. Available at: https://aoanjrr.dmac.adelaide.edu.au/documents/10180/127369/Demographics% 20and%20Outcome%20of%20Ankle%20Arthroplasty. Accessed April 14, 2014. 40. Rodrigues-Pinto R, Muras J, Oliva XM, Amado P. Functional results and complications after total ankle replacement: early to medium-term results from a Portuguese and Spanish prospective multicentric study. Foot Ankle Surg 19:222–228, 2013. 41. Rodrigues-Pinto R, Muras J, Oliva XM, Amado P. Total ankle replacement in patients under the age of 50: should the indications be revised? Foot Ankle Surg 19:229– 233, 2013. 42. Mehta S, Donley BG, Jockel JR, Slovenkai MP, Casillas MM, Berberian WS, Lin SS. The Salto Talaris total ankle arthroplasty system: a review and report of early results. Semin Arthro 21:282–287, 2010. 43. Schweitzer KM, Adams SB Jr, Viens NA, Queen RM, Easley ME, DeOrio JK, Nunley JA II. Early prospective results of the Salto-Talaris total ankle prosthesis. Duke Orthop J 2:23–24, 2012. 44. Schweitzer KM, Adams SB Jr, Viens NA, Queen RM, Easley ME, DeOrio JK, Nunley JA II. Early prospective clinical results of a modern fixed-bearing total ankle arthroplasty. J Bone Joint Surg Am 95:1002–1011, 2013.

45. Nodzo SR, Miladore MP, Kaplan NB, Ritter CA. Short to midterm clinical and radiographic outcomes of the Salto total ankle prosthesis. Foot Ankle Int 35:22–29, 2014. 46. Chao JC, Choi JH, Grear BJ, Tenenbaum S, Bariteau JT, Brodsky JW. Early radiographic and clinical results of Salto total ankle arthroplasty as a fixed-bearing device. Foot Ankle Surg http://dx.doi.org/10.1016/j.fas.2014.09.012. 47. Roukis TS, Prissel MA. Registry data trends of total ankle replacement use. J Foot Ankle Surg 52:728–735, 2013. 48. Lee AY, Ha AS, Petscavage JM, Chew FS. Total ankle arthroplasty: a radiographic outcome study. Am J Radiol 200:1310–1316, 2013. 49. Hoffmann AH, Fink B. Moderne 3-Komponenten-Sprunggelenkprothesen. Der Orthop€ ade 36:908–916, 2007. 50. Stengel D, Bauwens K, Ekkernkamp A, Cramer J. Efficacy of total ankle replacement with meniscal-bearing devices: a systematic review and meta-analysis. Arch Orthop Trauma Surg 125:109–119, 2005. 51. Zaidi R, Cro S, Gurusamy K, Siva N, Macgregor A, Henricson A, Goldberg A. The outcome of total ankle replacement: a systematic review and meta-analysis. Bone Joint J 95:1500–1507, 2013. 52. Thian C, Tiernan S, McEvoy F, Flavin R. An engineering evaluation of ankle prosthetics. Dublin Institute of Technology Conference Papers, 2008. Available at: http:// arrow.dit.ie/cgi/viewcontent.cgi?article¼1001&context¼ittengcon. Accessed April 4, 2014. 53. Zhao H, Yang Y, Yu G. A systematic review of outcome and failure rate of uncemented Scandinavian total ankle replacement. Int Orthop 35:1751–1758, 2011. 54. Besse J-L, Brito N, Lienhart C. Clinical evaluation and radiographic assessment of bone lysis of the AES total ankle replacement. Foot Ankle Int 30:964–975, 2009. 55. Koivu H, Kohonen I, Sipola E, Alanen K, Vahlberg T, Tiusanen H. Severe periprosthetic osteolytic lesions after the Ankle Evolutive System total ankle replacement. J Bone Joint Surg Br 91:907–914, 2009. 56. Rodriguez D, Bevernage BD, Maldague P, Deleu PA, Tribak K, Leemrijse T. Medium term follow-up of the AES ankle prosthesis: high rate of asymptomatic osteolysis. Foot Ankle Surg 16:54–60, 2010. €valko M, Tiihonen R, Kautiainen H, Belt EA. High rate of osteolytic 57. Kokkonen A, Ika lesions in medium-term follow-up after the AES total ankle replacement. Foot Ankle Int 32:168–175, 2011. 58. Besse J-L, Lienhart C, Fessy M-H. Outcomes following cyst curettage and bone grafting for management of periprosthetic cystic evolution after AES total ankle replacement. Clin Podiatr Med Surg 30:157–170, 2013. 59. Brunner S, Barg A, Knupp M, Zwicky L, Kapron AL, Valderrabano V, Hintermann B. The Scandinavian total ankle replacement long-term, eleven to fifteen-year, survivorship analysis of the prosthesis in seventy-two consecutive patients. J Bone Joint Surg Am 95:711–718, 2013. 60. Roukis TS. Management of the failed Agility total ankle replacement system. Foot Ankle Quarterly 24:185–197, 2013. 61. Gaden MTR, Ollivere BJ. Periprosthetic aseptic osteolysis in total ankle replacement: cause and management. Clin Podiatr Med Surg 30:145–155, 2013. 62. Pappas MJ, Buechel FF Sr. The ankle. In: Principles of Human Joint Replacement: Design and Clinical Application, pp. 91–152, edited by FF Buechel Sr, MJ Pappas, Springer-Verlag, Berlin, Germany, 2011. 63. Wildenauer E. Die Blutversorgung des Talus. Zeitschrift Anat Entwicklungsgeschichte 115:32–36, 1950. 64. Haliburton RA, Sullivan CR, Kelly PJ, Peterson LFA. The extra-osseous and intraosseous blood supply of the talus. J Bone Joint Surg Am 40:1115–1120, 1958. rielle de l’astragale. Lille Chir 65. Depreux MR, Hollingshausen M. Vascularisation arte 18:188–194, 1963. 66. Mulfinger GL, Trueta J. The blood supply of the talus. J Bone Joint Surg Br 52:160– 167, 1970. 67. Gelberman RH, Mortensen WW. The arterial anatomy of the talus. Foot Ankle 4:64–72, 1983. 68. Miller AN, Prasarn ML, Dyke JP, Helfet DL, Lorich DG. Quantitative assessment of the vascularity of the talus with gadolinium-enhanced magnetic resonance imaging. J Bone Joint Surg Am 93:1116–1121, 2011. 69. Tennant JN, Rungprai C, Pizzimenti MA, Goetz J, Phisitkul P, Femino J, Amendola A. Risks of blood supply of the talus with four methods of total ankle arthroplasty: a cadaveric injection study. J Bone Joint Surg Am 96:395–402, 2014. 70. Oppermann J, Franzen J, Spies C, Faymonville C, Knifka J, Stein G, Bredow J. The microvascular anatomy of the talus: a plastination study on the influence of total ankle replacement. Surg Radiol Anat 36:487–494, 2014. € ller PE, Jansson V, 71. Ficklscherer A, Wegener B, Niethammer T, Pietschmann MF, Mu Trouillier H-H. Thermoelastic stress analysis to validate tibial fixation technique in total ankle prostheses: a pilot study. Ulus Travma Acil Cerrahl Derg 19:98–102, 2013. 72. Terrier A, Larrea X, Guerdat J, Crevoisier X. Development and experimental validation of a finite element model of total ankle replacement. J Biomech 47:742–745, 2014. 73. Kofoed H. Total ankle replacement: back to the future? Clin Res Foot Ankle 1:e102, 2013. http://dx.doi.org/10.4172/2329-910X.1000e102. 74. Iwamoto K, Shi K, Tomita T, Yamazaki T, Futai K, Kunugiza Y, Yoshikawa H, Sugamoto K. In vivo kinematics of three-component mobile-bearing total ankle replacement for rheumatoid arthritis during gait. Ann Rheum Dis 72:S-895, 2013. 75. Kofoed H. Active weight bearing tibio-talar motion and meniscal translation in the S.T.A.R prosthesis: a radiographic comparative study of the S.T.A.R. toward the opposite normal ankle joint. Clin Res Foot Ankle 1:e101, 2013. http://dx.doi.org/ 10.4172/2329-910X.1000101.

T.S. Roukis, A.D. Elliott / The Journal of Foot & Ankle Surgery 54 (2015) 311–319

76. Tochigi Y, Rudert MJ, Brown TD, McIff TE, Saltzman CL. The effect of accuracy of implantation on range of movement of the Scandinavian Total Ankle Replacement. J Bone Joint Surg Br 87:736–740, 2005. 77. Affatato S, Taddei P, Leardini A, Giannini S, Spinelli M, Viceconti M. Wear behaviour in total ankle replacement: a comparison between an in vitro simulation and retrieved prostheses. Clin Biomech 24:661–669, 2009. 78. Barg A, Elsner A, Anderson AE, Hintermann B. The effect of three-component total ankle replacement malalignment on clinical outcome: pain relief and functional outcome in 317 consecutive patients. J Bone Joint Surg 93:1969–1978, 2011. 79. Espinosa N, Walti M, Favre P, Snedeker JG. Misalignment of total ankle components can induce high joint contact pressures. J Bone Joint Surg Am 92:1179–1187, 2010. 80. Saltzman CL, Tochigi Y, Rudert MJ, McIff TE, Brown TD. The effect of Agility ankle prosthesis misalignment on the peri-ankle ligaments. Clin Orthop Rel Res 424:137–142, 2004. 81. Nicholson JJ, Parks BG, Stroud CC, Myerson MS. Joint contact characteristics in Agility total ankle arthroplasty. Clin Orthop Relat Res 424:125–129, 2004.

319

82. Fukuda T, Haddad SL, Ren Y, Zhang LQ. Impact of talar component rotation on contact pressure after total ankle arthroplasty: a cadaveric study. Foot Ankle Int 31:404–411, 2010. 83. Kraal T, van der Heide HJL, van Poppel BJ, Fiocco M, Nelissen RGHH, Doets HC. Long-term follow-up of mobile-bearing total ankle replacement in patients with inflammatory joint disease. Bone Joint J 95:1656–1661, 2013. 84. Buechel FF Sr, Buechel FF Jr, Pappas MJ. Twenty-year evaluation of cementless mobile-bearing total ankle replacements. Clin Orthop Rel Res 424:19–26, 2004. 85. Rudigier J, Menzinger F, Grundei H. Total ankle replacement: 14 year experience with the cementless ESKA ankle prosthesis. Fuß Sprungg 2:65–75, 2004. 86. Pugely AJ, Lu X, Amendola A, Callaghan JJ, Martin CT, Cram P. Trends in use of TAR and ankle arthrodesis in the US Medicare population. Foot Ankle Int 35:207–215, 2014. 87. Raikin SM, Rasouli MR, Espandar R, Maltenfort MG. Trends in treatment of advanced ankle arthropathy by total ankle replacement or ankle fusion. Foot Ankle Int 35:216–224, 2014.