Complications After Total Ankle Replacement

Victor Valderrabano

Alexej Barg

INTRODUCTION

In the past decades, substantial progresses have been made in total ankle replacement (TAR). However, TAR still remains an evolving procedure. TAR using current third-generation prosthetic designs provides substantial postoperative pain relief and good functional outcome, including preserved range of motion.¹ However, the overall failure rate observed after TAR is substantially higher than those observed after total hip replacement (THR) or total knee replacement (TKR). Labek et al.² performed a systematic review, including national registers and clinical studies with respect to revision rates after ankle joint replacement. The revision rate after TAR was 3.29 per 100 observed component years, which was significantly higher than those after TKR (1.26 revisions), medial unicompartmental replacement (1.53 revisions), or THR (1.29 revisions).² Interestingly, the revision rates published in sample-based clinical studies, many by the designers of a particular design, were less than half of the values found in national registers.³ The most common reasons for revision surgery after total ankle arthroplasty are aseptic loosening, followed by persisting pain, septic loosening, implant problems, and technical errors (Table 12.1).⁴

To date, ankle arthrodesis was considered to be the “gold standard” for salvage of failed ankle prosthesis.⁵,⁶,⁷ and ⁸ However, TAR revision surgery, that is, revision of TAR by a TAR revision system, is a therapeutic alternative to the post-TAR ankle or hindfoot fusion. However, reports on revision arthroplasty are very limited and mostly consist of case reports within large cohorts (Table 12.2).

In this chapter, we address the most common reasons for failure of TAR and describe our treatment algorithm in this patient cohort depending on the specific failure reason.

ASEPTIC LOOSENING AND SUBSIDENCE

Aseptic loosening and subsidence are the major reasons for revision of TAR.⁴³ Haddad et al.⁴⁴ performed a comprehensive literature review to analyze the intermediate- and long-term outcomes of TAR and ankle arthrodesis. Ten clinical studies, including 852 ankle arthroplasties, were reviewed. The most common reason for revision was loosening and/or subsidence (28%).⁴⁴ In 2010, Gougoulias et al.¹ performed a systematic review of the literature, including 13 level IV studies with 1,105 ankle arthroplasties. The approximate failure rate was 10% after 5 years, with a wide range between 0% and 32%. In 10 studies of this review, the presence of radiolucency and prosthesis subsidence were evaluated.¹ In two studies evaluating the clinical and radiographic outcomes of Agility prosthesis, the periprosthetic radiolucencies were observed in up to 86% of all replaced ankles.²⁶^,²⁸

Aseptic loosening and subsidence may have different reasons, including insufficient bone ingrowth (bone-implant interface), reduced quality of bone stock, mal-loading of the ankle replacement (e.g., due to malposition of prosthesis components as a consequence of technical error), and increased intra-articular shear stresses (e.g., due to high patient activity or obesity).⁴³ An important issue for appropriate ingrowth at the prosthesis-bone interface is the coating of the prosthesis components. Adding of biologic agents may help to enhance the osseous integration of prosthesis, which in turn may help to avoid the aseptic loosening. Some of the prosthesis designs (e.g., early design Scandinavian total ankle replacement [STAR], first-generation Hintegra prosthesis) used single hydroxyapatite coating on a smooth substrate in their initial manufacturing, which has been identified as a possible risk factor for aseptic loosening of prostheses. Carlsson⁴⁵ performed a clinical study with 109 consecutive ankle arthroplasties using a STAR prosthesis to investigate the difference between loosening rates of single- versus double-coated prostheses (hydroxyapatite-coated prostheses on top of a titanium spray). Revisions due to aseptic loosening were necessary in 15 of 51 single-coated ankles but only in 1 of 58 double-coated ankles.⁴⁵ Recently, Brunner et al.¹³ performed a long-term analysis of STAR prosthesis: in 29 of 77 ankles, a revision of at least one metallic component was required, while the majority of revisions (25 of 29 revisions) resulted from issues at the bone-prosthesis interface. In all patients, first-generation STAR prosthesis was implanted.¹³ Similar findings were observed in national registers.¹⁸^,⁴⁶ Barg et al.⁴⁷ performed survivorship analysis in consecutive 684 patients who underwent TAR using Hintegra prosthesis. The overall survival rates were 94% and 84% after 5 and 10 years, respectively. The most common reason for prosthesis revision was aseptic loosening, followed by cyst formation and painful arthrofibrosis with 42, 7, and 5 cases, respectively. The following independent and statistically significant risk factors for failure of the prosthesis have been identified: younger age (odds ratio [OR], 3.84), primary osteoarthritis (OR, 7.19), posttraumatic osteoarthritis (OR, 6.20), and use of first-generation prosthesis with a single hydroxyapatite coating (OR, 15.04).⁴⁷

TABLE 12.1 The Most Common Reasons for Revision Surgery After TAR in Comparison with Revision Surgery After THR and TKR⁴

Revision Cause	TAR	THR	TKR
Aseptic loosening	38	55.2	29.8
Luxation or instability	8.5	11.8	6.2
Septic loosening	9.8	7.5	14.8
Periprosthetic fracture	2	6	3
Pathologic wear	8	4.2	8.2
Pain without other cause	12	3.7	9.5
Implant breakage	5.3	2.5	4.7
Technical error	4.6	3.8	4.6
All values are presented as percentage of revision with respect to the total number of revision surgeries within 1 y.

TABLE 12.2 Literature Review Addressing the Clinical Outcomes of Patients Treated with Revision Ankle Replacement

Reference	Study	TAR	Prosthesis	FU (y)	Failures	Reasons for Failures	Time Until Revision	Failures Treated by	Results of Revision
Ali et al.⁹	RS, SC	35	Buechel-Pappas	5 (0.3-12.5)	1 (3%)	Pain (1)	3 y	Revision TAR (1)	Converted to ankle fusion 1 y later due to CRPS
Anders et al.¹⁰	RS, SC	93	AES	3.5 (1.1-6.1)	7 (8%)	Loosening (1), infection (2), instability (2), Fx (2)	NA	Revision TAR (1), ankle fusion (6)	NA
Anderson et al.¹¹	RS, SC	51	STAR	(3-8)	12	Loosening (7), PE Fx (2), others (3)	2.8 (0.1-5.3)	Revision TAR (5), ankle fusion (5), PE exchange (2)	Three revision TARs with excellent function, one with good function, one died
Bonnin et al.¹²	PS, SC	98	Salto	8.9 (6.8-11.1)	12 (12%)	Loosening (6), PE Fx (5), malposition (1)	NA	Revision TAR (1), ankle fusion (6), PE exchange (5)	NA
Brunner et al.¹³	PS, SC	77	STAR	12.4 (10.8-14.9)	29 (38%)	Loosening (9), subsidence (11), progressive cysts (5), PE Fx (1), instability (2), infection (1)	8.1 (1.8-13.4)	Revision TAR (28), ankle fusion (1)	NA
Buechel et al.¹⁴	RS, SC	50	Buechel-Pappas total ankle	5 (2-0)	2 (4%)	Malposition of talar component (1), talar subsidence (1)	NA	Revision TAR (2)	NA
Carlsson et al.¹⁵	RS, SC	69	Bath and Wessex	NA	12 (17%)	Painful loosening (12)	4.3 (2.3-8.7)	Revision TAR (6), ankle fusion (12)	NA
Christ and Hagena¹⁶	RS, SC	144	STAR	4.8	9 (6%)	Malalignment (2), loosening (1), impingement (1), instability (1), Fx (1), deep infection (2)	NA	Revision TAR (7), ankle fusion (2)	NA
Doets et al.¹⁷	PS, MC	93	LCS (19), Buechel-Pappas (74)	7.2 (0.4-16.3)	15 (16%)	Aseptic loosening (6), malalignment (6), deep infection (2), severe wound healing problem (1)	NA	Revision TAR (1), ankle fusion (14)	Revision TAR showed that loosening required eventual conversion to fusion
Fevang et al.¹⁸	RS, MC	257	Norwegian TPR (32), STAR (216), AES (3), Hintegra (6)	4 (0-2)	27 (11%)	Aseptic loosening (13), instability (3), malalignment (7), deep infection (2), Fx (1), pain (5), defect or wear PE (2), others (2)	2.3 (0.1-8)	Revision TAR (15), PE insert exchange (6), fusion (6)	NA
Giannini et al.¹⁹	PS, MC	51	BOX	2.5 (2-4)	1 (2%)	Lateral impingement (1)	2	Revision TAR (1)	NA
Henricson and Agren²⁰	RS, SC	193	STAR (109), Buechel-Pappas (62), AES (22)	4.2 (1-8)	41 (21%)	Infection (5), technical error (8), loosening (11), pain (4), instability (13)	(1.0-6.6)	Revision TAR (23), ankle fusion (15), extraction of prostheses without fusion (3)	2 good results, 19 fairly good, 2 poor with persisting pain and use of two crutches
Hobson et al.²¹	RS, SC	123	STAR	4 (2-8)	18 (15%)	NA	NA	Revision TAR (16), ankle fusion (2)	NA
Hosman et al.²²	RS, MC	202	Agility (117), STAR (45), Mobility (29), Ramses (11)	2.3 (0.6-6.3)	14 (7%)	Loosening (10), varus malalignment (1), pain (1), deep infection (2)	1.9 (0.1-5.4)	Revision TAR (10), ankle fusion (3), BKA (1)	NA
Hurowitz et al.²³	RS, SC	65	Agility	3.3 (2.0-5.9)	21 (32%)	Loosening (8), subsidence (5), malalignment (3), infection (3), osteolysis (1), postimpingement (1)	NA	Revision TAR (17), ankle fusion (2), osteochondral allograft (1), BKA (1)	NA
Karantana et al.²⁴	RS, SC	52	STAR	6.7 (5.0-9.2)	8 (15%)	Stress Fx (2), stiffness (2), PE Fx (2), talar subsidence (1), loosening (1)	NA	Revision TAR (6), ankle fusion (2)	NA
Kitaoka and Patzer²⁵	RS, SC	160	Mayo	9 (2-7)	57 (36%)	Persistent pain and loosening (all)	4.4 (0.1-13.1)	Revision TAR (10), ankle fusion (45), BKA (2)	NA
Knecht et al.²⁶	RS, SC	132	Agility	7.2 (2-4)	14 (11%)	Component Fx (2), loosening (4), deep infection (1), talar collapse (2), varus malpositioning (1), subsidence or migration (3), others (1)	5.8 (0.5-11.3)	Revision TAR (7), ankle fusion (7)	NA
Reference	Study	TAR	Prosthesis	FU (y)	Failures	Reasons for Failures	Time Until Revision	Failures Treated by	Results of Revision
Kofoed and Sørensen²⁷	PS, SC	52	STAR	9 (6-4)	11 (21%)	Loosening (10), deep infection (1)	4.5 (0.8-8.8)	Revision TAR (5), ankle fusion (6)	NA
Kopp et al.²⁸	RS, SC	43	Agility	3.7 (2.2-5.3)	1 (2%)	Aseptic loosening (1)	NA	Revision TAR (1)	NA
Kumar and Dhar²⁹	RS, SC	50	STAR	3 (1.5-5)	3 (6%)	Malalignment (2), pain (1)	NA	Revision TAR (3)	In two pat good results, in one fusion using ring fixator
Mendolia and Talus Group³⁰	RS, SC	69	Ramses	12 (0-4)	12 (10%)	Malalignment (4), loosening (3), instability (5)	NA	Revision TAR (5), ankle fusion (7)	NA
Morgan et al.³¹	RS, SC	45	AES	4.8 (4.0-6.7)	2 (4%)	Loosening (2)	NA	Revision TAR (1), ankle fusion (1)	NA
Murnaghan et al.³²	RS, SC	22	STAR	2.2 (0.7-3.8)	2 (9%)	Malalignment (2)	NA	Revision TAR (2)	Good results
Nishikawa et al.³³	RS, SC	21	TNK	6.0 (1.3-14.1)	3 (14%)	Loosening (3)	NA	Revision TAR (1), ankle fusion (2)	Revision TAR was fused 2 y after because of loosening
Reuver et al.¹³¹	RS, MC	59	Salto	3.0 (1.0-5.4)	7 (12%)	Loosening (5), deep infection (2)	NA	Revision TAR (3), ankle fusion (4)	NA
Rodriguez et al.³⁴	RS, SC	18	AES	3.3 (1.7-5.1)	1	Loosening with cysts (1)	NA	Revision TAR (1)	NA
Rudiger et al.³⁵	RS, MC	117	ESKA	(0-0)	8 (7%)	Deep infection (4), talus necrosis (1), prosthesis breakage (1), prosthesis malalignment (1), loosening with cysts (1)	NA	Revision TAR (4), fusion (4)	NA
Schutte and Louwerens³⁶	PS, SC	49	STAR	2.3 (1.0-5.6)	4 (8%)	Septic (2) and aseptic (2) loosening	NA	Revision TAR (1), ankle fusion (3)	NA
Spirt et al.³⁷	RS, SC	306	DePuy Agility TAR	2.8 (0.3-6.3)	33 (10.8%)	NA	NA	Revision TAR (24), BKA (8), ankle fusion (1)	NA
Sproule et al.³⁸	PS, MC	88	Mobility	3.3 (2.5-5)	10 (11.4%)	Aseptic loosening (6), talar migration (1), infection (1), varus edge loading (1), CRPS (1)	1-4	Revision TAR (8), fusion (1), transtibial amputation (1)	NA
Vienne and Nothdurft³⁹	RS, SC	66	Agility	2.4 (1.5-3.6)	2 (3%)	NA	NA	Revision TAR (1), ankle fusion (1)	NA
Wood and Deakin⁴⁰	PS, SC	200	STAR	3.8 (2.0-8.4)	14 (7%)	NA	NA	Revision TAR (3), ankle fusion (11)	NA
Wood et al.⁴¹	PS, SC	200	STAR	7.3 (5-3)	24 (12%)	Major delay to wound healing (1), intraop Fx (1), postop Fx (2), aseptic loosening (14), edge loading (5), PE Fx (1)	NA	Revision TAR (4), ankle fusion (20)	1 revision TAR failed after 5 y and was converted to ankle fusion
Wood et al.⁴²	PS, SC	100	Mobility	3.6 (0.3-5.3)	5 (5%)	Insert luxation (1), loosening (1), talar subsidence (1), pain (1), varus deformity (1)	2.6 (0.5-3.8)	Revision TAR (1), ankle fusion (2), insert exchange (2)	NA
FU, follow-up; RS, retrospective; SC, single center; PS, prospective; MC, multicenter; AES, Ankle Evolution System; NA, not applicable; CRPS, complex regional pain syndrome; BKA, below the knee amputation; Fx, fracture; PE, polyethylene.

A meticulous preoperative analysis of remaining bone stock is necessary for planning of revision surgery. First, standard weight-bearing radiographs in three planes and the Saltzman view⁴⁸ should be performed for standardized assessment of hindfoot alignment. We routinely suggest computed tomography (CT) or single-photon emission computed tomography (SPECT)⁴⁹ for careful evaluation of bone stock defects due to loosening of prosthesis components. With many of currently available implants, the assessment of the talus underneath the talar component using conventional radiographs is very limited. CT or SPECT may also help evaluate the bony impingement and degenerative changes of the adjacent joints. Our treatment algorithm is based on the amount of bone stock deficit (Table 12.3). In general, there are two options for revision ankle arthroplasty: one-stage versus two-stage procedure. In patients with preserved bone stock after removal of loosen prosthesis component and careful debridement, one-stage revision ankle arthroplasty can be performed (Fig. 12.1). If necessary, revision components can be used on the tibial (thicker tibial components to normalize joint line level) and/or talar (flat-cut talar component) side. The other surgical option is the twostage procedure with use of bone graft (Fig. 12.2). However, the fixation of bone graft on the remaining talar bone can be difficult. In patients with substantial degenerative changes of the subtalar joint, a subtalar arthrodesis should be performed and graft can be fixed with longer screws into the calcaneus.

PROGRESSIVE INSTABILITY OR LUXATION OF PROSTHESIS COMPONENTS

The most common etiology for end-stage ankle osteoarthritis is posttraumatic⁵⁰^,⁵¹ with previous fracture of the lower leg⁵²^,⁵³ or repetitive ligamental injuries.⁵⁴ Therefore, it is not surprising that almost half of the patients with end-stage ankle osteoarthritis present with substantial concomitant valgus or varus deformity of the hindfoot.⁵¹^,⁵³ All concomitant deformities and instabilities should be preoperatively recognized and carefully analyzed.⁵⁵ Therefore, conventional weight-bearing radiographs and Saltzman view⁴⁸^,⁵⁶ are indispensable for analysis and quantification of concomitant deformities. The clinical relevance of the Saltzman view has been addressed by Frigg et al.,⁵⁷^,⁵⁸ and it has been shown that the alignment assessment using Saltzman view is much more accurate than the clinical assessment. If necessary, all concomitant problems should be sufficiently addressed, for example, by corrective osteotomies and ligamental reconstructions.⁵⁹,⁶⁰,⁶¹,⁶² and ⁶³^,¹³² Saltzman⁶⁴ wrote in a 2000 review article addressing the state of the art of TAR. He stated that preoperative varus or valgus deformity without surgical correction at the time of implantation of ankle prosthesis may lead to pathologic loading and pathologically increased polyethylene (PE) wear.⁶⁴ In the current literature, there is still controversial debate regarding what amount of preoperative hindfoot deformity is the relative or absolute contraindication for TAR. Following cutoff points for contraindication for TAR are discussed in the literature: 10°,¹⁷^,⁴⁰^,⁴¹^,⁶⁵^,⁶⁶ 15°,⁶⁷,⁶⁸,⁶⁹,⁷⁰ and ⁷¹ and 20°.⁴²^,⁷²,⁷³ and ⁷⁴ However, Kofoed⁷⁵ described a special sculpting technique for talus preparation which allows the intraoperative correction of varus or valgus deformity of more than 45°. Kim et al.⁷⁶ compared the clinical outcomes in 23 patients with preoperative moderate-to-severe varus deformity (≥10°) with results in 22 patients with preoperative neutral alignment of the
hindfoot. The clinical outcome and failure rate were comparable in both groups, and the authors accentuated the importance of additional procedures for appropriate intraoperative correction of concomitant deformities.⁷⁶

Only gold members can continue reading. Log In or Register to continue