Parameters That Can Be Used to Quantify Reduction Accuracy in Talar Neck Fractures and Malunions: A PRISMA-Compliant Scoping Review and Meta-Analysis

Understanding the three-dimensional anatomy of the talar neck is essential in assessing the accuracy of reduction in talar neck fractures as well as for planning surgical correction for talar malunions. However, the geometrical parameters that describe this anatomy are sparsely reported in the orthopedics literature. We aimed to identify from the existing literature, geometrical parameters that describe the anatomy of the talar neck, determine how these are measured, and their normative values. A scoping literature review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) extension for scoping reviews (PRISMA-ScR) guidelines. The primary searches were conducted on the PubMed, Embase, and Scopus databases. Any original research study looking at the human talus neck geometry was included. Parameters that described the anatomy of the talar neck were identified, and pooled estimates were determined by the random-effects meta-analysis model. Heterogeneity was assessed by the I2 test and leave-one-out meta-analysis. Subgroup analysis was done to compare the values of parameters between the Asian and Non-Asian populations. The risk of bias was assessed by the National Institutes of Health (NIH) Case Series Tool. The combined searches yielded 6326 results, of which 21 studies were included in the review and 15 in six different sets of metanalysis. The majority of the studies (n=19, 90.5%) evaluated adult tali, and only two (9.5%) evaluated pediatric tali. In most of the studies (n=13, 61.9%), talus neck geometry was evaluated on dry bones or anatomical specimens; evaluation by imaging techniques (radiographs, CT, MRI, and radiostereometric analysis) was used in eight studies, (39.1%). A total of eight different geometrical parameters (neck length, height, width, declination angle, inclination angle, torsion angle, circumference, and cross-sectional area) were identified. Except for talar torsion, variability was noted in methods of measurement of all other parameters. Subgroup analysis revealed that Asians had a higher neck height as compared to non-Asians; other parameters were not significantly different. Although the literature reports geometrical parameters to assess the talar geometry, the methods of measurement of these parameters are variable. Most of the available literature describes measurement techniques on cadaveric tali, and there is no literature on how these parameters should be measured on conventional CT or MRI slices. Further research needs to focus on the standardization of measurement techniques for these parameters on conventional CT and/or MRI scans.


Introduction And Background
The talus is the second-largest tarsal bone, and participates in the formation of three essential lower limb joints, i.e., the ankle (talocrural), subtalar, and the talonavicular joints.The talar neck connects the talar body to the talar head, and is instrumental in load transmission from the tibia to the midfoot as well as the hindfoot [1].Alterations in talar neck geometry, which may occur after trauma, or as a result of congenital conditions, can result in altered motion as well as load transmission across these joints, and result in pain and premature arthrosis [2,3].Hence, an understanding of the three-dimensional geometry of the talar neck is vital in judging reduction after operative fixation of talar neck fractures, as well as in planning corrective surgery for talar malunions.However, the parameters that describe the three-dimensional anatomy of the talar neck are sparsely reported in the literature, and there is no consensus on which parameters should be evaluated in routine orthopedics practice.Hence, we conducted this scoping review and meta-analysis to address these gaps in the existing literature.

Review Methods
This scoping review and meta-analysis was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) extension for Scoping Review guidelines (PRISMA-ScR) [4].The review protocol was formulated following the PRISMA for systematic review protocols (PRISMA-P) guidelines [5] and published a-priori [6].

Study Objectives
The primary objective of this review was to identify the geometrical parameters that describe the threedimensional anatomy of the talar neck, the measurement techniques, and precision.As a secondary objective, the review also aimed to identify the normative values and geographic differences of these parameters.

Eligibility Criteria
Original research studies (clinical or cadaveric, prospective or retrospective, comparative or noncomparative), evaluating the geometry or morphology of the human talar neck were included.There were no restrictions on the language of publication.Studies that did not describe the talar neck geometry, animal studies, review articles, conference abstracts, and case reports were excluded [6].

Information Sources
The primary search was conducted on the PubMed, Embase, and Scopus databases, using a pre-specified search strategy.The searches were filtered to include only human studies (Table 1).For the secondary search, a manual search of references from the full text of all included articles and relevant review articles was done.There were no restrictions on the language or date of publication [6].

Data Charting Process and Data Items
Two authors (SS and KJ) extracted the data from included studies on pre-piloted data collection forms; a third author (SP) verified the data for accuracy.

Outcome Measures
The outcome measures included talar neck length (NL), neck width (NW), neck height (NH), cross-sectional area (CSA) of the talar neck, declination angle (DA), inclination angle (IA), and torsion angle (TA).Methods of measurement were ascertained.Intra-class correlation coefficients (ICC) for each parameter were recorded, if these were reported by the study authors.To determine the variability in the measurement of the geometrical parameters by different measurement techniques, the coefficient of variance was determined.The coefficient of variance is a ratio of the standard deviation to the mean; it can also be expressed as a percentage.Lower values indicate better precision in measurement, and vice-versa.For this study, the %age CV was arbitrarily graded as 'low' if it was 30% and 'high' if it was >30%.

Critical Appraisal of Individual Sources of Evidence
The methodological quality of the studies included in the review was assessed by the National Institutes of Health (NIH) Case Series tool.This is a nine-item checklist; each item of the checklist is graded as 'Yes', 'No', 'Not Reported (NR)', 'Cannot Determine (CD)' or 'Not Applicable (NA)'.For this review, a study was graded as 'Good' if the answer to all items in the checklist (excluding the not applicable items) was 'Yes'.A 'Fair' grading was given if the answer to one or two items in the checklist was 'No' or 'Not Reported'.If the answer to three or more items in the checklist was 'No' or 'Not Reported', the study was graded as 'Poor'.

Synthesis of Results
For qualitative analysis, appropriate tables and data visualization diagrams were constructed.Quantitative analysis was done for the following parameters: (a) NL, (b) NW, (c) NH, (d) IA, (e) DA, and (f) TA.Pooled values of each of these parameters were estimated by the random-effects meta-analysis model [7]; values were reported as means with 95% confidence intervals.Forest plots were constructed to visualize the results.
Statistical heterogeneity was evaluated by the I 2 test.Leave-one-out sensitivity analysis was performed if high statistical heterogeneity was identified (I 2 > 75%).No subgroup analysis or meta-regression was planned a priori, as we were uncertain about the nature of the data that would be extracted.However, after data extraction, subgroup analysis using a random-effects model was performed based on ethnicity; studies were classified as either 'Asian' or 'Non-Asian' for this purpose.Analysis of 95% confidence intervals (CI) was done to determine statistically significant differences in the geometrical parameters between the two subgroups.A parameter was deemed to be significantly different if the 95% CIs of the pooled estimate of the two groups did not overlap, and vice versa.

Literature Search
The combined search revealed 6326 results.After screening the titles and abstracts, 71 full-text articles were obtained.Of these, 21 were included in the scoping review .Fifteen articles were included in six different sets of pooled meta-analyses to calculate the normative values of talar NL, NW, NH, DA, IA, and TA (Figure 1).Of the 50 full-text articles excluded, the majority (n=49, 98%) did not report talar neck geometry as an outcome measure, and one was a conference abstract (Figure 1).

Characteristics of Studies Included
Of the 21 studies included in the review, seven (33.3%) included data from human subjects, whereas the other 14 (67.3%) were cadaveric.The majority of the studies (n=19, 90.5%) evaluated adult tali, and only two (9.5%) evaluated pediatric tali.In most of the studies (n=13, 61.9%), talus neck geometry was evaluated directly on dry bones or anatomical specimens; evaluation by imaging techniques (radiographs, CT, MRI, and radio-stereometric analysis) was used in other studies (Table 2).Of the studies, 48% (n=10) were from Asia, 24% (n=5) from Europe, 19% (n=4) from North America and Canada, and 10% (n=2) from Africa.The majority (66%, n=14) of the studies were published in the recent decade (2011-2020) (Figure 2).

CV: coefficient of variation; ICC: intra-class correlation coefficient
Leave-one-out sensitivity analysis revealed that the heterogeneity was not altered by the omission of any study.Subgroup analysis revealed that the Asian population had a marginally (insignificantly) higher NW (2.614 cm, 95%CI 2.571-2.656)than the non-Asian population (2.494 cm, 95%CI, 1.765-3.224).The coefficient of variance for this parameter was determined to be low for all studies.

Talar NH
The vertical dimension of the talar neck in the sagittal plane, termed 'neck height', or 'neck thickness', was evaluated by four studies [9,10,14,15].The pooled estimate of medial talar NH (three studies, 350 tali) was 2.62 cm (95%CI, 2.14-3.09); the heterogeneity for this estimate was high (I 2 = 99.7%)(Figure 5).Talar neck height measurement methods: (a) the measurement method for medial talar neck height as described by He et al. [15] and Harnroongroj et al. [9]; (b) the measurement method for medial talar neck height as described by Abd-Elaleem et al. [10]; (c) the measurement method for lateral talar neck height described by Ebraheim et al. [14].All meta-analysis values are in centimeters.

CV: coefficient of variation; ICC: intra-class correlation coefficient
Leave-one-out sensitivity analysis revealed that the heterogeneity was not altered by the omission of any study.Subgroup analysis revealed that the Non-Asian population had a significantly higher medial neck height (3.301 cm, 95%CI, 3.239-3.363)than the Asian population (2.274 cm, 95%CI, 2.231-2.317).The coefficient of variance for this parameter was determined to be low for all studies.Ebraheim et al. [14] determined the thickness (height) of various parts of the lateral talar neck (Figure 5).It was found to increase progressively from lateral to medial and was noted to have a maximum value of 8 mm medial to the lateral border.

DA
The angle between the body and neck of the talus in the horizontal plane, termed variably as 'declination angle', 'medial neck angle', 'horizontal angle of talar neck', 'transverse angle of talar head', 'talar bodyneck-head angle' or 'varus angle of the neck', was evaluated by nine studies [15][16][17][18][19][20][21][22][23].While the majority of these studies measured this angle between the long axis of the talar body and the long axis of the talar neck, Mahato et al. [21,24] measured this between the transverse axis of the talar body and the long axis of the talar neck.The pooled estimate of DA in adults (seven studies, 406 tali) was 19.10° (95%CI, 14.59-23.62);the heterogeneity for this estimate was high (I 2 = 98.6%) (Figure 6).

CV: coefficient of variation; ICC: intra-class correlation coefficient
Leave-one-out sensitivity analysis revealed that the heterogeneity was not altered by the omission of any study.Subgroup analysis revealed that the Asian population had a higher DA (20.873°, 95%CI, 14.216-27.530)than the Non-Asian population (16.672°, 95%CI, 12.414-20.931);however, this difference was not statistically significant.The coefficient of variance for this parameter was determined to be high for three studies [16,17,20], and low for others.Two studies evaluated the DA in the pediatric population [20,25].Ramalho et al. showed that the DA progressively decreases with the growth of the foot [25].In an MRI-based study on infants with unilateral clubfoot, Itohara et al. showed that tali in clubfeet had a significantly higher DA, lower total volume, and lower volume of the ossific nucleus as compared to tali in the contralateral, normal foot [20].
'torsion angle', or 'coronal plane angle' was evaluated by four studies [15,17,22,24].All four studies employed similar measurement techniques for this parameter.The pooled estimate of TA angle in adults (four studies, 272 tali) was 44.3° (95%CI, 31.74-56.87);heterogeneity for this estimate was high (I 2 = 99.6%)(Figure 8).Leave-one-out sensitivity analysis revealed that the heterogeneity was not altered by the omission of any study.All four studies [15,17,22,24] evaluating talar torsion used the same measurement technique.All meta-analysis values are in degrees.

CV: coefficient of variation; ICC: intra-class correlation coefficient
Subgroup analysis revealed that the Asian population had a higher TA (46.884°, 95%CI 31.668-62.1)than the Non-Asian population (36.5°, 95%CI 33.788-39.212);however, this difference was not statistically significant.The coefficient of variance for this parameter was determined to be low for all studies.

Other Parameters
He et al. determined the cross-sectional area and circumference of the talar neck [15].The mean area was 14.80 ± 2.02 cm 2 , whereas the circumference was 8.92 ± 0.6 cm.In a cadaveric study, Chan et al. compared conventional radiographs, CT scans, and radio stereometric assay (RSA) to malreduction after fixation of talar neck fractures [26].While all techniques underestimated malreduction, CT was the most accurate of all three techniques.Nozaki et al. performed a three-dimensional geometric morphometric analysis to determine differences in talar geometry in men and women [27].Women were noted to have a longer talar neck and a higher TA, as compared to men.

Critical Appraisal within Sources of Evidence
Only three of the 21 (14.3%)studies included in the review were deemed to be of 'Good' methodological quality as per the NIH Case Series tool.Eleven (52.4%) studies were graded as 'Fair' whereas the remaining seven (33.3%) were graded as 'Poor' (See Appendices).Reasons for poor methodology included lack of comparability of the study cohort (higher percentage of males as compared to females), use of measurement methods prone to high inter and intra-observer variability, as well as instrument error (goniometer and vernier calipers), and ill-defined statistical methods.

Discussion
Talar neck fractures are often associated with comminution on the medial side, which results in shortening of the medial talar neck, and consequently, a varus deformity [29][30][31][32][33][34][35].These fractures necessitate anatomical reduction, as even small amounts of varus malalignment can impact the subtalar joint and midfoot biomechanics.Sangeorzan et al. have shown that a 2 mm displacement of the talar neck can significantly affect the contact characteristics of the subtalar joint [2].Furthermore, Daniels et al. have demonstrated that a varus malalignment of the talar neck can result in varus and internal rotation deformity of the hindfoot, adduction deformity of the forefoot, and limitation of eversion [3].However, assessment of varus after fixation of talar neck fractures remains challenging, and requires bilateral visualization of the talar neck margins [32,33].In 1978, Canale and Kelly described the 'Canale view' to image the talar neck in the horizontal plane, and to detect varus malalignment [34].This view allows for visualization of the medial comminution of the talar neck and discontinuity of the lateral cortical margin indicative of a varus malreduction.This view is performed with the ankle maximally plantarflexed, the foot pronated by 15°, and the X-ray beam angled 75° cephalad.Thomas and Boyce suggested that varus malreduction may be appreciated better by using multiple Canale views, with the pronation varying from 15° to 25° [35].Given the fact that the surgeon must aim to achieve < 2mm of displacement of the talar neck after reduction, quantification of the reduction assumes paramount importance.As Bonnel et al. have shown the normal alignment of the talar neck in the horizontal plane can be varus, neutral, or even valgus, it is important to know and quantify the normal anatomy before the surgeon can plan operative fixation or correction of malunion for talar neck fractures [17].The DA can be used for this purpose, with the normal, contralateral side serving as the target.
Correction of a talar neck malunion necessitates a talar neck osteotomy and restoration of the medial neck length employing an interposition graft.The few published studies on talar neck malunion [36][37][38][39][40], while reporting the 'varus deformity', did not quantify its three-dimensional anatomy nor report the correction that was achieved post-operatively.We believe that the DA and TA are useful parameters to quantify the degree of varus and internal rotation of the neck in a talar malunion respectively.Measurement of these parameters on the normal, contralateral limb can also provide the surgeon with an estimate of the degree of correction that is needed.
The orthopedic literature also reports dorsiflexion malunion after talar neck fractures [34,40].In these cases, the distal fragment displaces or angulates in a dorsal direction.The net effect is decreased ankle dorsiflexion owing to impingement of the dorsally displaced talar head.Whereas dorsal displacement is easy to assess on lateral ankle radiographs, dorsal tilt can be hard to judge intra-operatively.Hence, we believe that assessment of the IA, and comparison with the uninjured contralateral side can serve as a reliable guide to sagittal plane reduction of the talar neck.A summary of these parameters has been provided in Table 3.When performing this review, we noted that most of the published literature on the three-dimensional geometry of the talar neck comes from anthropological studies.These studies report on a wide variety of parameters of the talar head, neck, body, and articular surfaces.The focus of these studies was to map anthropology and predict biomechanics, design implants for ankle arthroplasty or total talar replacement, or, in forensic medicine, to predict gender.We did not find any study evaluating the restoration of talar neck geometry during or after operative fixation of talar neck fractures.In contrast to other foot and ankle fractures, the orthopedic literature does not report on talar neck geometry quantitatively in talar neck fractures or malunions.We hope that this review provides the impetus for future studies that aim to quantify the post-operative reduction in talar neck fractures on radiographs and CT scans.Further research should focus on the measurement of these parameters on conventional, as well as volume-rendered CT and MRI sections.

SNo. Characteristic Synonyms
This review has several strengths such as strict adherence to the PRISMA-SCR guidelines, a well-defined search strategy covering multiple databases, sensitivity as well as subgroup analyses, and determination of the risk of bias.To the best of our knowledge, this is the first review to summarize the geometrical parameters to assess the talar neck.The pooled estimates of each geometrical parameter, estimates of the ICC, as well as the coefficient of variation of each of these parameters, are provided.
We acknowledge the inherent limitations of our review, which is based on a heterogeneous database.This is reflected in the high statistical heterogeneity of our pooled estimates.This may be attributed to geographical and ethnic differences, and also to the fact that different authors have used different imaging modalities and measurement techniques.Furthermore, many studies included in the review evaluated tali obtained from cadavers and amputation specimens, which could potentially have been afflicted by the disease.Hence, these results should be interpreted with caution.These limitations notwithstanding, the precision in our pooled estimates is reflected by the narrow CIs.

Conclusions
Although several parameters have been described to assess the talar neck geometry in three dimensions, their use to assess talus neck reduction in the orthopedic literature is sparse.Three of these parameters, the DA, IA, and TA can be used to quantify the frontal, sagittal, and coronal plane reduction of the talar neck.However, there is a need to standardize these measurements on conventional CT/MRI sections for use in routine orthopedic practice.Furthermore, our subgroup analysis has also revealed that the Asian Population has a significantly higher NH than the non-Asian population.These estimates can help orthopedic surgeons in determining the normal range of values of these parameters and the variations that might be expected.

FIGURE 1 :
FIGURE 1: PRISMA Flow Diagram for the study PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analysis

FIGURE 2 :
FIGURE 2: Baseline characteristics of studies included in the review (a) Geographical distribution, (b) Journal of publication, (c) Time trends.Ortho: orthopedics; Foren Med: Forensic Medicine; Anat: Anatomy; Radio: radiology Image Credit: Authors

FIGURE 4 :
FIGURE 4: Talar neck width: methods of measurement, variability in measurement techniques and pooled meta-analysis estimates.

FIGURE 5 :
FIGURE 5: Talar neck height: methods of measurement, variability in measurement techniques, and pooled meta-analysis estimates.

FIGURE 8 :
FIGURE 8: Talar neck torsion angle: methods of measurement, variability in measurement techniques and pooled meta-analysis estimates.

TABLE 2 : Baseline characteristics of studies included in the review
* tali were obtained from patients undergoing lower limb amputation A: adult subjects; P: pediatric subjects; C: cadaveric study; H: study conducted on alive human subjects/patients; IA: inclination angle, DA: declination angle; TA: torsion angle; NL: neck length; NW: neck width; NH: neck height; CSA: cross sectional area; RSA: radiostereometric analysis; MIND: minimum intra-articular distance of the neck

95% CI) Significance
Differential shortening of the medial neck length, as compared to the lateral neck length results in varus.Equal loss of medial and lateral neck lengths results in foreshortening of talus