Revision and reoperation in total ankle arthroplasty, a multicenter study of different prosthetic models

Abstract

Background

Long term survival of different ankle prosthesis implanted by non-designer surgeons has not been reported.

Methods

Cross-sectional study in three centers. Revision and reoperation were studied as markers of survival for different prosthetic models between 2002 and 2023. Primary outcomes were analyzed using Kaplan-Meier curves with variables associated with failure reported in literature.

Results

127 prostheses were identified. The revision and reoperation rates over 22 years were 13.4 % and 9.5 %, respectively. Aseptic loosening (n = 14, p < 0.01) and septic loosening (n = 3, p < 0.01) were the most frequent causes associated with revision. Prosthetic failure was managed either with arthrodesis (n = 14, p < 0.01) or with revision and implant replacement (n = 3, p < 0.01). Female biological sex was associated higher risk of revision (PR = 6.7; 95 % CI 1.6– 28.2; p-value <0.01). Chronic pain was the only variable associated with increased risk of reoperation (PR = 8.2; 95 % CI 3.1– 22.4; p-value = <0.01).

Conclusions

Results are comparable to those reported in the literature regarding revision and reoperation. Different prosthetic models evaluated in this study provide a broader understanding of how prostheses perform when implanted by surgeons other than their creators.

1 Introduction

Total ankle arthroplasty (TAA) is a popular treatment option for ankle osteoarthritis; however, there are doubts regarding the survival of different prosthetic models ^, . Although the literature reports that new generation prostheses have survived for more than 10 years, their survival is still lower compared to other lower limb prostheses ^,^,^,^,^,^,^,^, . It is important to note that the outcome of the models is not only influenced by the implants but also by the surgeons’ experience, which has contributed to improving the survival of these prostheses ^,^,^,^,^,^,^,^,^,^,^, . However, the best and most prolonged survival reports come from publications made by the creators of prostheses ^,^,^,^,^,^,^,^,^,^,^,^,^, .

Beyond this, TAA is not without complications, which, like other prosthetic procedures, can occur both during surgery and after it. Glazebrook et al. have categorized and staged complications into a classification that, although it has limitations, is the most widely recognized to date ^, . Low-grade complications are those with a low probability of leading to prosthetic failure; medium-grade complications may fail in less than fifty percent of cases; and high-grade complications are those that lead to failure in more than fifty percent of cases . In the event of prosthetic failure, various strategies have been described, with revision and arthrodesis (whether tibio-talar or tibio-talo-calcaneal) standing out, both with satisfactory results ^,^,^, . Current evidence does not conclude that one strategy is superior to another ^, .

Additionally, the survival of different prostheses is reported in studies conducted in countries with national registries ^,^,^, . Unfortunately, it is not possible to have an adequate registry in all countries, and much of the knowledge in this area comes from high-income countries and highly specialized centers. This limits the ability to properly extrapolate the results to other regions ^, . Likewise, much of the available information today comes from studies published by those involved in the design of prostheses, rather than independent observers, raising questions about the reproducibility of the results ^,^,^,^,^,^,^,^,^,^,^, . Furthermore, few of the models used in long-term studies are available on the market today .

The development of this study arises from the lack of literature regarding the long-term outcomes of different TAA models implanted by surgeons not involved in the design of the implants, especially as there is no strong evidence that suggests good results are only associated with models and their designers.

2 Methodology

A cross-sectional study was conducted using convenience sampling of consecutive cases in three centers. Therefore, no prior sample size calculation was performed. Patients aged 18 years or older who had undergone a TAA between January 2002 and September 2023 were included. Additionally, patients with less than one year since surgery but with at least six months of follow-up were also included, considering the possibility of early revisions (less than six months). Patients undergoing surgery due to oncological causes were excluded due to the higher likelihood of reduced survival related to their underlying disease. All TAA’s were performed using anterior approach by fellowship trained orthopedic foot and ankle surgeons at each center. Implanted models were chosen according on the availability of the country’s market during the periods in which the surgeries took place.

The primary outcomes were prosthesis revision and reoperation. The concepts of prosthesis failure or revision surgery lack standardized definitions in literature and vary across studies ^,^, . Currently, most authors consider revision as any procedure requiring the removal or replacement of one or more metallic components ^, . For this study, the definition of revision was based on Henricson et al., described as the removal or exchange of one or more prosthetic components, except for an incidental polyethylene insert exchange . Similarly, reoperation was defined as any secondary surgery not intended as a revision but involving the joint .

2.1 Statistical analysis

A descriptive analysis of all variables was performed: quantitative/continuous variables were summarized using means, ranges, and standard deviations, while qualitative/categorical variables were analyzed using percentages and absolute values.

The normal distribution of variables was assessed using the Kolmogorov-Smirnov test or graphical methods. For normally distributed variables, quantitative data were compared using Student’s t -test, and qualitative data were analyzed using the Chi-squared test. Otherwise, non-parametric equivalents (Mann–Whitney and Kruskal-Wallis tests) were used. Subsequently, a survival analysis was conducted using Kaplan-Meier curves, with revision and reoperation as outcome events, comparing variables previously described in the literature as associated with failure.

A p-value < 0.05 was considered statistically significant. The statistical software R was used for data analysis.

This study was presented to and approved by the ethics committees of the participating institutions.

3 Results

One hundred and twenty-seven prostheses were identified across the three centers (sixty-two at hospital A, twenty at hospital B, and forty-five at hospital C). Mean follow up was 92.98 (SD 67.5) months. Of the 127 TAA, 17 (13.4 %) required revision, and 12 (9.5 %) required reoperation. The prostheses are presented in Fig. 1 . The population characteristics according to prosthetic failure and the need for reoperation are shown in Table 1 . Notable findings included that 67 (53 %) of the total 127 prosthesis patients had post– traumatic etiology at their initial operation, and 55 (43 %) had primary osteoarthritis at the initial operation. A higher proportion of women than men underwent surgery (67 women vs. 60 men, p < 0.01), and the BMI (Body Mass Index) of patients requiring prosthesis revision was lower than that of the group not requiring revision (X = 25.5; SD = 2.9; p = 0.07).

Table 1

Patients’ characteristics in relation to revision or reoperation.

	Revision (n = 17)		No revision (n = 110)				Reoperation (n = 12)		No reoperation (n = 115)
	n	%	n	%	Total	p value	n	%	N	%	Total	p value
Age (Years)						0,36						0,12
Mean +/- SD	64,3 (11,9)		65,2 (11,7)				69,8 (10,3)		64,2 (11,8)
Hospital						0,37						0,98
A	8	12,9	54	87,1	62		6	9,7	56	90,3	62
B	1	5,0	19	95,0	20		2	10,0	18	90,0	20
C	8	17,8	37	82,2	45		4	8,9	41	91,1	45
Gender						< 0,01*						0,77
Male	2	3,3	58	96,7	60		7	11,7	53	88,3	60
Female	15	22,4	52	77,6	67		5	7,5	62	92,5	67
BMI						0,07						0,60
Mean +/- SD	25,5 (2,9)		27,6 (4,5)				28,0 (5,7)		27,3 (4,2)
ASA						0,36						0,21
I	0	0,0	15	100,0	15		1	6,7	14	93,3	15
II	16	16,0	84	84,0	100		8	8,0	92	92,0	100
III	1	9,0	10	91,0	11		3	27,3	8	72,7	11
IV	0	0,0	1	100,0	1		0	0	1	100,0	1
Comorbidities
HBP	10	13,2	66	86,8	76	1	6	7,9	70	92,1	76	0,54
DMII	2	12,5	14	87,5	16	0,98	0	0,0	16	100	16	0,51
Smoking	3	8,1	34	91,9	37	0,39	2	5,4	35	94,6	37	0,51
RA	0	0,0	14	100,0	14	0,21	1	7,1	13	92,9	14	1
CRF	1	25,0	3	75,0	4	1	0	0,0	4	100,0	4	1
Laterality						0,79						0,82
Right	11	14,7	64	85,3	75		8	10,7	67	89,3	75
Left	6	11,5	46	88,5	52		4	7,7	48	92,3	52
Etiology						0,84						0,76
Posttraumatic	9	13,4	58	86,6	67		5	7,5	62	92,5	67
Primary	8	14,5	47	85,5	55		7	12,7	48	87,3	55
Infectious	0	0	3	100,0	3		0	0	3	100	3
Rheumatic	0	0	2	100,0	2		0	0	2	100	2

Statistically significant (*); SD (Standard Deviation), DMII (Type II Diabetes Mellitus); HBP (High Blood Pressure); BMI (Body Mass Index); RA (Rheumatoid Arthritis); CRF (Chronic Renal Failure)

Table 2 presents the characteristics of the prostheses and surgical procedures related to revisions and reoperations. The most used prosthesis was the Hintegra® model (39.4 %, n = 50), followed by the STAR® model (16 %, n = 20). Achilles tenotomy was the most frequently associated procedure with arthroplasty (18 %, n = 23). In two cases, contralateral ankle arthrodesis had been performed prior to arthroplasty (2 %, n = 2) ( Fig. 2 ). Additionally, five cases of bilateral prostheses were identified (using different models in three cases and the same model in two), none of which required revision ( Fig. 3 ).

Table 2

Implant characteristics and related procedures to revision and reoperation.

	Revision (n = 17)		No revision (n = 110)				Reoperation (n = 12)		No reoperation (n = 115)
	n	%	n	%	Total	p value	n	%	n	%	Total	p value
Implant						< 0,01*						0,87
STAR	1	5,0	19	95,0	20		2	10,0	18	90,0	20
Agility	2	50,0	2	50,0	4		1	25,0	3	75,0	4
Mobility	2	18,2	9	81,8	11		1	9,1	10	90,9	11
Infinity	0	0	10	100,0	10		1	10,0	9	90,0	10
Hintegra	9	18,0	41	82,0	50		6	12,0	44	88,0	50
Vantage	0	0	9	100,0	9		0	0	9	100,0	9
Ramses	2	100,0	0	0	2		0	0	2	100,0	2
BOX	1	9,1	10	90,9	11		0	0	11	100,0	11
Cadence	0	0	10	100,0	10		1	10,0	9	90,0	10
Previous procedures
Subtalar arthrodesis	0	0,0	5	100,0	5	1	1	25,0	4	75,0	5	0,40
Ankle fracture-orthopedic treatment	1	6,7	14	93,3	15	0,70	1	6,7	14	93,3	15	1
Ankle fracture-surgical treatment	6	16,7	30	83,3	36	1	2	5,6	34	94,4	36	1
Triple Arthrodesis	0	0,0	5	100,0	5	0,72	0	0,0	5	100,0	5	1
Contralateral ankle arthrodesis	0	0,0	2	100,0	2	1	0	0,0	2	100,0	2	1
Supramalleolar osteotomy	0	0,0	2	100,0	2	1	0	0,0	2	100,0	2	1
Calcaneal lateralizing osteotomy	1	16,7	5	83,3	6	1	2	33,3	4	66,7	6	0,08
Calcaneal medializing osteotomy	0	0,0	1	100,0	1	1	1	100,0	0	0,0	1	1
Contralateral TAA (Other institution)	0	0,0	2	100,0	2	1	0	0,0	2	100,0	2	1
Contralateral TAA (Same institution)	0	0,0	6	100,0	6	0,61	1	16,7	5	83,3	6	0,9
Chronic lateral ankle instability	0	0,0	6	100,0	6	0,65	0	0,0	6	100,0	6	1
Additional procedures to TAA
Achilles tenotomy	2	8,7	21	91,3	23	0,74	4	17,4	19	82,6	23	0,23
Lateral ligamentary complex reconstruction	2	25,0	6	75,0	8	0,73	0	0,0	8	100,0	8	1
Subtalar arthrodesis	1	33,3	2	66,7	3	0,92	0	0,0	3	100,0	3	1
Previous Ankle arthrodesis						1						1
Contralateral	0	0,0	2	100,0	2		0	0,0	2	100,0	2
None	17	13,6	108	86,4	125		12	9,6	113	90,4	125