Abstract
Background:
Studies examining the clinical outcomes of revision total ankle arthroplasty (TAA) are sparse. Revision TAA surgery has become more common with availability of revision implants and refinement of bone-conserving primary implants. In this study, patient-reported results and clinical outcomes were analyzed for a cohort of patients who underwent both primary and revision TAA at a single high-volume institution.
Methods:
We retrospectively reviewed prospectively collected data on 29 patients with failed primary total ankle arthroplasty. Cases of isolated polyethylene exchange, infection, or extra-articular realignment procedures were excluded. Patient-reported outcome (PRO) measures and clinical results were reviewed in this longitudinal study.
Results:
Fifteen patients (51.7%) underwent revision of just the talar and polyethylene components while 13 patients (44.8%) underwent revision of all components. The most common cause was talar subsidence (51.7%). The average time to revision was 3.9 years with a follow-up of 3.2 years after revision, and 3 (10.3%) revision arthroplasties required further surgery; 2 required conversion to arthrodesis and 1 required second revision TAA. Improvements in PROs were better after primary than revision TAA.
Conclusions:
Clinical and patient-reported results of revision ankle arthroplasty after metal component failure improved significantly but never reached the improvements seen after primary ankle arthroplasty. In our series, 10.3% of revision TAAs required a second revision TAA or arthrodesis surgery.
Levels of Evidence:
Therapeutic Level III, comparative series.
Keywords
The number of total ankle replacements being performed each year continues to rise. 9 Better improvements in function and patient satisfaction, significant pain relief, and preserved hindfoot motion have led to this increase, while the rates of ankle arthrodesis remain steady.9,26-28,33 This annual increase in primary total ankle arthroplasty (TAA) has also led to an increase in the rate of revision TAA.7,20,36 The salvage procedures for failed total ankle replacement include revision with or without cementing the prosthesis, conversion to arthrodesis, cement arthroplasty, and amputation.14,22,29 Not satisfied with the traditional treatments for failed TAA including arthrodesis, foot and ankle surgeons have sought and designed revision implants that simplify the handling of large bone voids and unstable joints.6,25,27
Literature on outcomes of revision ankle arthroplasty surgery is sparse given the rates of revision TAA ranging from 8.4% to 17% reported in larger series with longer term follow-up. 8,10,11,23 The purpose of this investigation was to evaluate clinical results and improvements in patient-reported outcomes (PROs) of a cohort of patients who underwent primary TAA and subsequently required revision TAA and had both primary and revision TAA at the host institution.
Methods
This institutional review board (IRB)–approved, retrospective chart review of prospectively collected data was conducted reviewing all patients who underwent primary TAA and subsequently required revision TAA done between September 2007 and April 2016 with both primary and revision surgery done at the host institution. Patients were pulled from a prospectively collected database kept at the host institution. Four high-volume ankle replacement surgeons performed the primary and revision surgeries. Cyst debridement and bone grafting with polyethylene exchange and revision for the treatment of infection were excluded from the cohort. Patients with metal component failure and a minimum of 2 years of follow-up after revision were included in the study.
Prior to primary TAA, all patients were enrolled in a prospectively collected database at the host institution. Each patient completed PRO questionnaires preoperatively before primary TAA and then at 6 months and on an annual basis postoperatively. Patients were indicated for revision based on clinical examination as well as implant appearance on radiographs and advanced imaging (single photon emission computed tomography [SPECT] or plain computed tomography [CT]) (see Figures 1-4 for an example of the workup through revision surgery). When the patient was indicated for revision TAA after failure of the primary surgery, he or she again filled out PRO questionnaires preoperatively before revision TAA and then at 6 months and on an annual basis postoperatively. The PRO questionnaires used for this study included the American Orthopaedic Foot & Ankle Society (AOFAS) hindfoot score, visual analog scale (VAS), Short Form 36 (SF-36), and the Short Musculoskeletal Function Assessment (SMFA) questionnaires.
(A) Patient X: anteroposterior preoperative radiographs of the right ankle with retained hardware and tibiotalar arthritis. (B) Patient X: lateral preoperative radiographs of the right ankle with retained hardware and tibiotalar arthritis.
(A) Patient X: anteroposterior radiograph 6.7 years after Salto Talaris total ankle arthroplasty showing cyst formation above the tibial and below the talar prosthesis with varus subsidence of the tibial prosthesis. (B) Patient X: lateral radiograph 6.7 years after Salto Talaris total ankle arthroplasty showing cyst formation above the tibial and below the talar prosthesis.
(A) Patient X: coronal single photon emission computed tomography (SPECT-CT) scan demonstrating increased uptake surrounding the tibial and talar components of the Salto Talaris total ankle arthroplasty. (B) Patient X: sagittal single photon emission computed tomography (SPECT-CT) scan demonstrating increased uptake surrounding the tibial and talar components of the Salto Talaris total ankle arthroplasty.
(A) Patient X: anteroposterior radiograph after revision total ankle arthroplasty to an INBONE II prosthesis. (B) Patient X: lateral radiograph after revision total ankle arthroplasty to an INBONE II prosthesis.
Twenty-nine patients (16 [55.2%] males and 13 [44.8%] females) with failed primary total ankle arthroplasties requiring revision were included in this study (Please refer to Table 1 for complete demographics). The mean age (and standard deviation) of the patients was 62.4 ± 8.1 years (range, 39-76 years). Each patient had erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) blood tests to screen for infection. If either of these were mildly elevated, a frozen section was sent intraoperatively to look for white blood cells per high-powered field. Patients with suspected infection were excluded from this study. On average, revision total ankle arthroplasty occurred at 3.9 ± 2.5 years (range, 0.2-7.3 years).
Demographic Data.
The prosthesis used for primary arthroplasty included the INBONE I (Wright Medical Technology, Arlington, TN), the Scandinavian Total Ankle Replacement (STAR; Stryker, Mahwah, NJ), the Salto Talaris (Integra, Plainsboro, NJ), and the Infinity (Wright Medical Technology, Arlington, TN). The INBONE I prosthesis was the most frequently revised (15/29, 51.7%) in this study, followed by the Salto Talaris (8/29, 27.6%), STAR (5/29, 17.2%), and Infinity (1/29, 3.4%) prostheses, respectively. The most commonly used prosthesis in revision situations was the INBONE II (Wright Medical Technologies; 18/29, 62.1%), INBONE I (5/29, 17.2%), Salto XT (Integra, Plainsboro, NJ; 3/29, 10.3%), Infinity (2/29, 6.7%), and STAR (1/29, 3.4%) prosthesis, respectively. Table 2 summarizes the clinical data and prosthesis used for both primary and revision TAA.
Clinical Data.
The diagnosis at the time of primary total ankle arthroplasty was posttraumatic in 15 ankles (51.7%), recurrent instability in 5 ankles (17.2%), inflammatory or crystalline arthritis in 5 ankles (17.2%), and osteoarthritis in 4 ankles (13.8%). The most common indication for revision TAA was talar subsidence (15/29, 51.7%) followed by aseptic loosening (8/29, 27.6%), coronal talar subluxation (5/29, 17.2%), and talar malrotation (1/29, 3.4%), respectively (Table 3). Talar subsidence was diagnosed either on plain lateral radiographs demonstrating progressive distal migration of the talar component, through CT, or through SPECT-CT. Coronal talar subluxation was a phenomenon related to the “saddle” talar component in the INBONE I prosthesis. This component did not prevent subluxation medially or laterally in the coronal plane as successfully as the “sagittal sulcus” talar component present in the INBONE II prosthesis. Talar malrotation occurred intraoperatively during the primary TAA where the talar prosthesis was placed in an externally rotated position. In 15 patients (51.7%), the talar and polyethylene components were revised while 13 patients (44.8%) had tibial, talar, and polyethylene components revised. One patient (3.4%) had only the tibial and polyethylene components revised.
Indications for Primary and Revision Total Ankle Arthroplasty.
The paired Student t test was used to compare findings for preoperative and postoperative outcome scores side by side, and the 1-way analysis of variance was used for comparisons of more than 2 means. All statistical analysis was performed using the most current version of SPSS (SPSS, Inc, an IBM Company, Chicago, IL) by a professional statistician.
Results
The average length of follow-up after revision TAA was 3.3 ± 1.7 years (range, 2.0-8.3 years). At the time of most recent follow-up, 26 patients (89.7%) retained their initial revision implants, 2 patients (6.9%) went on to require conversion to arthrodesis, and 1 patient (3.4%) required a second revision TAA. The average time to second revision or conversion to arthrodesis after first revision was 18 ± 4.1 months. Furthermore, 3 other patients required additional surgery after revision TAA. Two patients required gutter debridement for impingement, and 1 underwent subtalar arthrodesis. The reason for conversion to arthrodesis for the 2 patients who failed revision TAA was talar avascular necrosis in both. The third patient who failed revision TAA required revision for coronal talar subluxation of the INBONE I prosthesis, which was used in the first revision. The overall reoperation rate in this cohort was 20.7%. None of the patients in this cohort had a periprosthetic joint infection around their revision TAA, and there were no perioperative fractures.
The average improvement in PRO scores after primary TAA was greater than those seen after revision TAA. Both procedures lead to significant improvements in all PRO measures when compared to preoperative scores, but the magnitude of improvement seen after revision TAA never reached that seen after primary TAA (see Figure 5 for PRO data). When comparing the changes in PRO scores seen after primary TAA to those seen after revision TAA, significant differences were noted in VAS, SF-36 Physical function, and SMFA function and bother. These changes are summarized in Figure 5.
Patient-reported outcomes before (solid) and after (striped) primary and revision total ankle arthroplasty (TAA). All improvements were significant when comparing postoperative scores to preoperative. AOFAS, American Orthopaedic Foot & Ankle Society; SF-36, Short Form 36; SMFA, Short Musculoskeletal Function Assessment; VAS, visual analog scale.
The time to maximal improvement also differed between the procedures. At 6 months postoperatively, the primary TAA scores had reached 93.8% of their maximal improvement in VAS while the revision group had only reached 70.3%. Furthermore, the revision scores reached 85.9% of maximal improvement by 1-year postrevision while the primary group had reached 99.4%. Similar trends were seen in SF-36 Physical, SMFA function, and AOFAS hindfoot scores. This trend is summarized in Figure 6.
This graph demonstrates the percentage of improvement over time after both primary and revision total ankle arthroplasty (TAA). In all instruments tested, the time to maximal improvement was faster after primary TAA than after revision TAA. AOFAS, American Orthopaedic Foot & Ankle Society; SF-36, Short Form 36; SMFA, Short Musculoskeletal Function Assessment; VAS, visual analog scale.
Discussion
In this series, revision total ankle arthroplasty for failed primary TAA was successful in 26 of 29 ankles (89.6%). All patients in this study underwent revision for either metal component loosening/subsidence or component malalignment. Of the 3 failures, 1 underwent successful repeat revision TAA and is currently 3 years postoperative from the second revision with prosthesis still in place. Overall, these 29 patients had significant improvements in PRO scores when comparing pre- and postoperative primary TAA scores as well as pre- and postoperative revision TAA scores. The scores reached their maximal improvement faster in the primary TAA group.
The overall reoperation rate was 20.7% (6 of 29 ankles) during an average follow-up period of 3.2 years. Two patients failed revision TAA and were converted to arthrodesis; both of these patients had talar subsidence after INBONE II revision TAA. It appeared on imaging studies and intraoperative findings during conversion to arthrodesis that talar avascular necrosis caused the failure of revision surgery. In a cadaveric study, Tennant et al 30 showed that the artery of the tarsal canal was violated in 75% (3 of 4) of the specimens by the 6-mm drill driven through the subtalar joint during the INBONE procedure. 30
Revision total ankle arthroplasty has had variable success, and salvage arthrodesis, historically, was seen as the “gold-standard” option in failed TAA.7,12,15,21,34 Kamrad et al, 15 in a review of the Swedish Ankle Registry, compared the results of 118 salvage arthrodeses with the results reported in one of the author’s previous publications, which reviewed functional outcomes after 80 revision TAAs. Both revision TAA and salvage arthrodesis led to low satisfaction and functional results, but the reoperation rate was significantly higher in the revision TAA group. The authors concluded that until studies are available that show the true benefit of revision TAA over salvage arthrodesis, they would prefer arthrodesis over revision TAA. 15
In the largest series of revision TAA, Hintermann et al 12 reviewed their experience using a nonconstrained, 3-component system that includes revision components and custom implants. The authors reviewed 117 revision TAAs, 66 of whom had primary TAA done at the host institution and 51 were referrals. The average time to revision was 4.3 years. They reported an overall reoperation rate of 15% (aseptic loosening was the most common cause). They concluded that the correlation between extent of bone loss at the resection surface and the prevalence of component failure was weak and not significant. 12
Ellington et al 7 reported on 41 revision TAAs using the Agility prosthesis with custom talus or tibial components as needed. The average time to revision was 4.2 years. In this series, 34 patients had retained their revision prosthesis at the conclusion of the study while 7 patients went on to either tibiotalar arthrodesis or below-knee amputation due to chronic infection. 7 Custom prostheses were not used for revision TAA in the current study.
Michnick et al 21 examined functional outcomes of revision total ankle arthroplasty in 24 patients. The authors created a case-matched control group for comparison and compared their revision TAA results to the control group of 43 primary TAAs. The average time to revision was 6.7 years. They demonstrated a better improvement after primary TAA (using the SF-36 and Foot and Ankle Ability Measure [FAAM]) in categories that tested the patient’s physical health. 21
The rates of failure and reoperation in this cohort were comparable with previously reported rates following both primary total ankle replacement and tibiotalar arthrodesis.2,5,13,17-19 Long considered the gold standard for salvage of failed total ankle arthroplasty, tibiotalar arthrodesis with block allograft bone grafting is fraught with complications and reoperations, and it frequently leads to patient dissatisfaction.2,4,13,19,32,34,35 Only a few series reviewing TAA conversion to arthrodesis demonstrate satisfactory results.1,15 The rate of reoperation or additional surgery after tibiotalar arthrodesis with block allograft salvage of failed TAA is notable due to nonunion rates (ranging from 8.1%-41.6%) and adjacent joint arthritis.3,4,13,16,32,34
As would be expected, the time to maximal improvement was shorter for patients undergoing primary TAA (Figure 6). The primary TAA cohort reached maximal improvement by 6 months to 1 year postoperatively while the revision group did not see this type of improvement until the 2-year postoperative visit. Many factors may be related to this trend. Operative times are generally longer in revision TAA, blood loss can be greater, and the native bone stock can be decreased in revision scenarios compared to primary TAA. In addition, the revised ankle has undergone additional trauma (failed primary TAA, additional surgery, etc), which may take longer to heal. The revision group did not achieve the same magnitude of improvement seen in the primary group, although the improvement was significant compared to prerevision scores.
There were limitations in this study. The PRO measures were collected prospectively, but the study was retrospective. There is a potential selection bias when indicating a patient with a failed TAA for revision arthroplasty vs tibiotalar arthrodesis.24,31 The time periods where PRO scores were collected postoperatively included 6 months and annually. At those intervals, it is difficult to precisely determine the time frame for maximal improvement after primary and revision TAA. The time periods were identical between groups, however, and this enabled an accurate comparison that demonstrated the above differences.
In conclusion, this is the first longitudinal series of patients at one institution who underwent primary TAA and subsequently required revision TAA. Our results indicate that revision arthroplasty for metal component failure can be done successfully with the appropriately selected patients. Revision TAA led to significant improvements in PROs, but the time to maximal improvement took longer after revision, and the magnitude of the improvement never equaled that seen after primary TAA. With an overall reoperation rate of 20.7% and a failure rate of 10.4% at an average follow-up duration of 3.2 years, case-by-case evaluation is required to determine the best revision surgery for failed total ankle arthroplasty.
Supplemental Material
FAI794956-ICMJE – Supplemental material for Patient-Reported Outcomes Before and After Primary and Revision Total Ankle Arthroplasty
Supplemental material, FAI794956-ICMJE for Patient-Reported Outcomes Before and After Primary and Revision Total Ankle Arthroplasty by James Robert Lachman, Jania Arcia Ramos, Samuel Bruce Adams, James Albert Nunley, Mark Erik Easley and James Keith DeOrio in Foot & Ankle International
Footnotes
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: James A. Nunley II, MD, reports personal fees from Wright Medical Technologies and personal fees from Stryker outside the submitted work. In addition, Dr Nunley has a patent for the Vantage Total Ankle System (Exactech) with royalties paid. Mark E. Easley, MD, reports other from Stryker Orthopaedics–STAR development team outside the submitted work. In addition, Dr Easley has a patent for the Vantage Total Ankle System (Exactech) with royalties paid. James K. DeOrio, MD, reports personal fees from Wright Medical Technologies and personal fees from Stryker outside the submitted work. In addition, Dr DeOrio has a patent for the Vantage Total Ankle System (Exactech) with royalties paid. ICMJE forms for all authors are available online.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
