Foot & Ankle International - September 2019 - 998

998	
(PROMs) are used to evaluate pain, function, and quality of
life to better assess outcomes subsequent to the treatment of
foot and ankle conditions, including joint replacements.3,4,7
Thus, operative criteria and decision-making tools should
account for patient-specific characteristics and demographics such as gender, age, preoperative deformity, smoking
status, preoperative function, and baseline comorbidities
among patients undergoing TAA. Further, TAA positioning
needs to be optimized, as TAA alignment has been shown to
affect outcomes and survival,3,8,9,21,52 with the primary
imaging modality being plain radiography. With TAA's
growth in popularity and increasing numbers of patients
being offered TAA surgery, it is important to identify
patient-specific and prosthesis positioning-specific risk
factors that may predispose patients to suboptimal results
and TAA failure.
This study had 2 main purposes. One was to analyze
TAA survival and identify risk factors for TAA failure based
on baseline patient characteristics and demographics,
PROMs, and TAA radiographic alignment parameters. The
second was to develop a predictive model that could predict
TAA failure based on the risk factors identified.

Methods
Patient Enrollment
The ankle arthritis database from a single center, based on
prospectively collected data, was retrospectively reviewed.
All skeletally mature patients enrolled between 2004 and
2015, and undergoing treatment of end-stage ankle arthritis (ESAA) with the Hintegra Total Ankle System (Integra
LifeSciences, Plainsboro, NJ) or Infinity Total Ankle
System (Wright Medical Technology, Arlington, TN) were
included. All surgeries were performed by one of 3 fellowship-trained orthopedic foot and ankle surgeons with
extensive TAA experience. The recommendation for total
ankle replacement was based on a combination of the surgeon's preoperative clinical examination, radiographic
findings, and patient preference. Institutional research ethics board approval was granted and informed consent was
obtained from all study participants. All included patients
had a minimum of 2 years' clinical follow-up.
Patients were excluded for receiving a TAA if they had
any of the following: osteonecrosis of the talus or tibia,
active infection in the ankle within 12 months prior to
surgery, medical conditions precluding safe surgery, or
severe osteoporosis. Also excluded were patients who
received a second-generation TAA given the higher rate
of revisions reported by Lefrancois previously27 and
patients without a complete series of postoperative radiographs (weightbearing anteroposterior and lateral) taken
at 6 to 12 weeks.

Foot & Ankle International 40(9)

Data Collection
Patient assessments were completed preoperatively at 6
months, 1 year postsurgery, and annually thereafter.
Baseline demographics, comorbidities, smoking status, the
Canadian Orthopaedic Foot and Ankle Society (COFAS)
preoperative classification,24 end-stage ankle arthritis etiology, and PROMs were collected as part of part of the
Musculoskeletal Outcomes Data Evaluation and
Management System questionnaire. This is currently the
only standardized, validated scoring instrument applicable
to foot and ankle surgery published by the American
Academy of Orthopaedic Surgeons.9

TAA Failure Definition
TAA failure was defined as either reoperation requiring
removal of one or both metal components of the prosthesis
or amputation above the level of the ankle.34

Patient-Reported Outcome Measures
The PROMs analyzed were the Ankle Arthritis Score
(AAS), the Ankle Osteoarthritis Scale (AOS) score, and the
the Short Form-36 physical component summary (SF-36
PCS) scores at baseline and at 6 months postoperatively.
These scores are validated, self-reported, ankle arthritis-
specific PROMs.11,30,55

Radiographic Outcomes
Seven sagittal and coronal radiographic outcomes (medial
distal tibia angle, talar tilt angle, talar center migration, sagittal distal tibial angle, lateral talar station, tibia-talus ratio,
and gamma angle) were measured according to previously
described methods,2,20,45,51-54 and 1 additional measurement
was devised for this study: posterior tibial component overor underhang. Postoperative measurements (3 coronal and 5
sagittal) were performed by 2 independent orthopedic surgeons, twice each (Figures 1-3). Normal ranges for each
radiographic parameter were defined based on previously
published literature when available, and expert opinion
when not available (Table 1). Inter- and intraobserver reliability of radiographic measurements were calculated with
the type 1 intraclass correlation coefficient (Table 2).42

Statistical Analysis
Standard summary statistics were generated followed by a
survival analysis. A univariate analysis was performed
using baseline demographics, radiographic parameters previously described, and PROMs collected preoperatively
and at 6 months postsurgery using a log-rank test. A multivariate Cox proportional hazards model was estimated to
Foot & Ankle International - September 2019

Table of Contents for the Digital Edition of Foot & Ankle International - September 2019
