Notre blog
Chirurgiens Othopédistes Lyon
Arthrodèse de l’arrière pied (articulation sous talienne): comment optimiser les résultats ?
Notre article dans le journal Orthopaedics & Traumatology: Surgery & Research
“Combined anterior and posterior vs isolated posterior facet fixation for subtalar arthrodesis”

Voici un résumé de notre dernier article sur l’arthrodèse (ou blocage) de l’articulation soustalienne. Le Dr Manzi a travaillé avec l’équipe du CHU de Naples (Italie) notamment le Dr Izzo Antonio. Ce dernier a effectué une période de Fellowship dans notre établissement pour améliorer sa pratique chirurgicale.
Auteur: Antonio Izzo , Giovanni Manzi , Martina D'Agostino , Massimo Mariconda , Shelain Patel , Alessio Bernasconi
Subtalar arthrodesis (SA) is a common procedure to treat end-stage degeneration of this joint when conservative measures have been exhausted. Multiple techniques have been described to achieve this (i.e., medial or lateral approaches, open or arthroscopic procedures, with or without bone graft), but near universally, fixation is achieved using large-diameter cannulated screws which leads to fusion in approximately 85–100% of cases . The majority of authors reporting on SA have used screws crossing the posterior facet of the joint in the belief that indirect stabilization of the anterior and middle facets is achieved through posterior facet fixation, making an anterior direct stabilization unnecessary. However biomechanical studies have identified that positioning screws both posteriorly and anteriorly in the subtalar joint, through the use of an additional third anterior screw, significantly increases compression and torsional resistance at the fusion site. From a clinical standpoint, in 2020 Wirth et al. retrospectively investigated 113 SAs, and found that using a third screw reduced the rate of non-union compared to fixation with two screws (14% and 35%, respectively). Similarly, in a series of 54 cases published in 2022 by Jones et al., the non-union rate was 0% and 27% for patients in which 3 or 2 screws had been used, respectively. It should be noted that these analyses were designed to compare 3-screw constructs to more traditional 2-screw ones rather than to investigate the need of fixation of the anterior subtalar facet itself. To the best of our knowledge, no comparative study has been conducted focusing on the advantages and risks of a direct stabilization of the anterior subtalar joint during subtalar fusion.
With this background, we systematically reviewed the literature in order to determine whether a combined direct fixation of both anterior and posterior facets during SA might influence the union and complication rate as compared to isolated fixation of the posterior facet. Our hypothesis was that a combined anterior and posterior fixation increases the union rate and reduces the risk of complications.
2. Synthesis of results
Continuous variables were reported as the mean and 95% confidence interval (95%CI) or standard deviations (SD) and range for normally distributed variables, and as the median and interquartile range (IQR) for non-normally distributed variables. Data were tested for normality using a Shapiro-Wilk test and were then compared using the Student-T test (for normally distributed variables) or Wilcoxon rank sum test (for non-normally distributed variables). Two groups were built based on the position of screws: a first group (ANT/POST) in which screws were positioned both in the anterior and posterior facet of the subtalar joint and a second group (ONLY POST) in which screws were placed only through the posterior facet. A proportional meta-analysis was run to pool data regarding the nonunion, the complication and reoperation rate. A subgroup analysis was led comparing the complication rate after exclusion of ‘symptomatic hardware and screw removal’. The ‘metaprop’ command was used to compute 95% confidence intervals using the score statistic and the exact binomial method and incorporate the Freeman-Tukey double arcsine transformation of proportions. Heterogeneity among studies was assessed through the Higgins’ I2 statistic and a random-effect model was applied in all cases. A meta regression (metareg module) was used to compare pooled proportions in ANT/POST vs ONLY POST group. Univariate linear regression (for normally distributed variables) or nonparametric kernel regression (for non-normally distributed variables) was used to test whether demographics (sample size, age and sex), study characteristics (level of evidence, length of follow-up and methodological quality as CMS) and technical details (use of an anterior screw or not, open or arthroscopic procedure, screws configuration as parallel or non-parallel and use or not of bone grafting) might predict a significant change in the nonunion, complication (both as overall and after exclusion of ‘symptomatic hardware and screw removal’) and reoperation rate. The significance level for the overall estimates of effect was set at p < 0.05.
3. Results
Eighteen series from fourteen studies (in four studies two different cohorts were considered) (685 feet) were selected (ANT/POST = 96, ONLY/POST = 589). The median follow-up was 28 months (IQR, 12–42). At baseline, the two groups were not significantly different concerning the number of patients (p = 0.45), number of feet (p = 0.42), percentage of female patients (p = 0.18) and length of follow-up (p = 0.70). The only demographic difference was by age since patients in ONLY POST group were younger than the ANT/POST group (46.2 ± 3.1 vs 51.6 ± 4.2 years, respectively; p = 0.006). From a surgical point of view, the number of screws used differed significantly since in the ONLY POST group a third screw was never used (p = 0.015). Similarly, the orientation of screws differed in the two groups since they were always non-parallel in the ANT/POST group and always parallel in the POST group (p < 0.001). The number of screws used in ANT/POST group was 2 and 3 in 50% (32 patients) and 50% (64 patients) of studies. The ratio of open/arthroscopic procedures (p = 0.109) and the proportion of cases in which bone graft was used (p = 0.116) was not significantly different.
3.1. Pooled estimate of nonunion, complication and reoperation rate
The pooled proportion estimate showed a similar nonunion rate in the ANT/POST group compared to the ONLY POST group (6% [95% CI, 0–11] vs 10% [95% CI, 6–15]; p = 0.46) The intergroup statistical heterogeneity was nonsignificant (p = 0.190) . The pooled proportion of complications was similar in the two groups (ANT/POST: 14% [95% CI, 4–24] vs ONLY POST: 19% [95% CI, 13–24], p = 0.47) , even after excluding ‘symptomatic hardware and screw removal’ from the list of complications (ANT/POST: 5% [95% CI, 1–11] vs ONLY POST: 9% [95% CI, 5–12], p = 0.68). The statistical heterogeneity was nonsignificant both for the overall complication rate analysis (p = 0.450) and for the same analysis after exclusion of ‘symptomatic hardware and screw removal’ (p = 0.336). Finally, the pooled proportion of reoperation was not different either (ANT/POST: 7% [95% CI, 0–15] vs ONLY POST: 10% [95% CI, 6–15], p = 0.37) , with no significant heterogeneity between groups (p = 0.514).
3.2. Regression analysis
Nonparametric Kernel regression suggested a correlation between the proportion of open/arthroscopic procedures and the nonunion rate (observed estimate, −5.8; 95% CI, −9.8 to −1.09; p = 0.025) with a median nonunion rate at 10.9% (IQR, 5.1–19.8) for open procedures vs 5.9% (IQR, 5.3–9) for arthroscopically-assisted ones. Interestingly, the Level of Evidence correlated both with the nonunion (observed estimate, −9.2; 95% CI, −14.7 to −3.3; p = 0.004) and reoperation rate (observed estimate, −10.8; 95% CI, −16.9 to −5.7; p < 0.001). In fact, the median nonunion rate was 6% (IQR, 4–6) in Level IV studies and 14% (IQR, 10–25) in Level III studies, while the median reoperation rate was 4.2% (IQR, 0–7) in Level IV studies and 17% (IQR, 12.5–18) in Level III studies.
3.3. Quality of studies
Mean CMS was 40.4 points (95% CI, 38.4–42.4). There was no significant difference (p = 0.182) in terms of methodological quality for studies included in the ANT/POST (mean 38.7 points [95% CI, 29.7–47.7] and ONLY POST (mean 40.9 points [95% CI, 38.8–43] group.
4. Discussion
The main finding of this proportional meta-analysis was that direct combined fixation of both anterior and posterior facets during SA does not significantly influence the risk of nonunion, complications or re-operation at a median 28-month follow-up compared to isolated posterior facet fixation. This confirms what hypothesized by those who only deal with the posterior facet during surgery, i.e., that isolated posterior fixation satisfactorily stabilizes the whole talocalcaneal complex in order to achieve union. However, the regression analysis identified that arthroscopically-assisted procedures might almost halve the risk of nonunion (5.9% vs 10.9%) compared to open procedures, although this does not meet what reported in previous literature. Furthermore, for studies included in this review the Level of Evidence inversely correlated with the nonunion and complication rate, confirming that non comparative studies might suffer a greater risk of bias than comparative ones. To the best of our knowledge, this is the first study to meta-analyze the role of direct anterior facet fixation during SA.
In order to correctly interpret these results some considerations have to be made. First, while our findings might apparently seem in contrast with some previous analyses where a greater mechanical compression and fusion rate had demonstrated for 3-screw constructs as compared to 2-screw ones this is not true. As a matter of fact, in some studies reporting a combined fixation of the anterior and posterior facet, only 2 screws were used meaning that a parallelism between the studies comparing the number of screws used (2 vs 3) and studies comparing the number of facets directly fixed (anterior and posterior vs only posterior) could not be made. Second, some baseline differences related to screws might reflect anatomical reasons: the fact that the two groups were not balanced in terms of number of screws used (i.e., 3 screws were used in 50% of studies of the ANT/POST group but never in the ONLY POST group) was not surprising since the small size of the posterior facet would unlikely allow to place a further large-diameter screw in the joint; also, the different orientation of screws between the two groups (always non-parallel in the ANT/POST group vs always parallel in the POST group) could be explained by the need to position the anterior screw along a vertical axis or from anterior to posterior in order to avoid a iatrogenic damage to the talonavicular joint. Third, out of 4 studies where a combined anterior and posterior fixation was documented, only 2 reported an accurate description of the preparation of the anterior facet, which means that no conclusion can be drawn regarding the necessity to surgically approach the anterior part of the subtalar joint and dedicate some time to remove the residual cartilage from this area. Fourth, although main demographic and surgical characteristics have been taken into account in this review and the proportional meta-analysis performed allowed to weigh results based on the sample size, we acknowledge that there is a number of patient- and surgeon-related variables that could not be considered in this review (lack of data in primary studies) and which may prevent from drawing a definitive conclusion on the necessity to stabilize the anterior facet. On the other side, with the studies available (which were overall of poor quality, as showed through the CMS) we believe that these findings may represent the basis for future direct comparative analyses.
The pooled proportion of nonunion cases stood at 6% for combined fixation of the anterior and posterior facet and at 10% for the posterior facet. Although the second figure was almost double the first, the difference did not achieve statistical significance. We are aware that this may be consequent to a type II error where a difference in the outcome was not found due to a not-large-enough sample size. Eleven studies were excluded from this review due to the use of 2 or 3 screws as case-by-case intraoperative decision based on the intra-operative findings of screw purchase and bone quality. Although data from over 600 ankles was pooled in this analysis, we believe that the inclusion of more studies might further strengthen our findings. We would therefore encourage authors to detail as much as possible the surgical technique is studies reporting the outcome after SA and, wherever possible, report raw data taking into account technique variations in order to allow a potential extraction and re-analysis in secondary studies.
It is worth highlighting that both cohorts suffered with secondary operations for removal of metalwork . Once this complication was excluded, however, the complication rate dropped from 14% to 5% in the ANT/POST group and from 19% to 9% in the ONLY POST group. This highlights the need for surgeons to adequately counsel patients before the index procedure on the potential need of additional surgery to remove metalwork must always be explored during the pre-operative counselling with the patient.
It should also be emphasized that in this meta-analysis arthroscopic procedures led to a better outcome in terms of nonunion rate as compared to standard open procedures. While a number of papers have been published highlighting the pros and cons of an open vs an arthroscopic fusion at the ankle not so many have dealt with the subtalar joint. In a recent systematic review on the topic, Silvampatti et al. concluded that at the moment there is no conclusive evidence to recommend an arthroscopic approach over the traditional one. Noteworthy, in our review, which was specifically aimed at investigating the value of stabilizing directly the anterior subtalar facet, we had to exclude a number of studies in which the type of fixation varied within the same cohort or was not clearly defined. This means that the results of the regression analysis should be interpreted with caution and that a prospective randomized and adequately powered trial is warranted before drawing clear conclusions on the best approach to perform a subtalar fusion.
Some limitations have to be acknowledged in this review. We included Level III and IV studies which are, by definition, not prospective and therefore potentially biased. Our choice at this level was dictated both by the will to include all the relevant evidence produced around SA and by the lack of high-quality specific analyses in the field. Additionally, it could be argued that no data on the clinical outcome after SA was reported, which would have been valuable in our analysis. Nevertheless, in the absence of prospective comparative studies, the choice of dichotomous variables allowed to run an appropriate statistical analysis and to calculate the pooled proportion estimate for nonunion and postoperative complication rate which strengthens our findings. Also, the lack of a dedicated power analysis should be considered as a potential additional drawback of our analysis. Finally, although the statistical heterogeneity found in the comparison of the two groups was nonsignificant, some inherent heterogeneity coming from the comparison of studies from different authors should be taken into account, since they will likely differ in some aspects of the treatment proposed to the patient (i.e., type of approach, type of screw, postoperative protocol).
5. Conclusion
This proportional meta-analysis and systematic review suggested that a combined direct fixation of anterior and posterior facets during subtalar arthrodesis does not significantly influence the risk of nonunion nor affects the risk of complication and reoperation as compared to isolated posterior facet fixation. High-level randomized comparative studies in this area are warranted in order to further confirm or disprove our findings.
Chirurgie orthopédique Lyon
Articles récents

Peut on marcher après une rupture du tendon d'Achille ?

L'importance de la formation des professionnels de santé
Ortho
Val d'Ouest Lyon / Ecully
Le service de chirurgie orthopédique et traumatologique Lyon - Ecully traite les maladies et les traumatismes de l'appareil locomoteur (os, articulations, ligaments, ménisques, muscles, tendons).
04 72 19 31 25
Coordonnées
• Lundi : de 8h à 17h
• Mardi : de 8h à 19h
• Mercredi : de 8h à 18h
• Jeudi : de 8h à 19h
• Vendredi : de 8h à 16h
Copyright © 2024 Clinique Orthopédique Val d'ouest