Physiologic widening of the medial clear space: What's normal?

Abstract

Background

Literature has validated the use of stress radiographs for evaluation of ankle stability. However, to our knowledge no study has reported the amount of physiological widening that occurs with manual external rotation stress test in uninjured ankles. The purpose of this study was to assess the amount of medial clear space widening that occurs with a manual external rotation stress test in uninjured ankles.

Methods

A cohort of adult patients undergoing operative fixation of unstable ankle fractures were prospectively enrolled to have their contralateral ankle undergo manual external rotation stress examination. Fluoroscopic images of the unaffected ankle were performed in the OR. A non-stressed mortise view and manual external rotation stress view were obtained with a standardized marker to correct for magnification differences. The images were de-identified, presented in a randomized order and reviewers who were blinded. Each reviewer measured the medial clear space.

Results

Thirty fluoroscopic images on fifteen patients were obtained. The mean medial clear space on the non-stressed mortise view was 3.1 mm (SD-0.69; Range 1.9 to 4.2, 95% CI [2.75, 3.45]) versus a mean of 3.2 mm (SD-0.71; Range 2.0 to 4.7, 95% CI [2.94, 3.66]) in the stressed mortise view group. Inter-rater reliability was excellent between all observers for medial clear space (ICC-0.88; CI [0.78, 0.94]).

Conclusions

Our results support the previous literature and allow us to advocate for ankle fractures with >5 mm medial clear space after external rotational stress to be considered unstable. Additionally, ankles with a medial clear space between 4 and 5 mm, instability should be considered only if lateral shift is > 2 mm on stress examination. Our data shows that no physiologically healthy ankles widened beyond these established cut-offs before or after the manual external rotation stress.

1. Introduction

Ankle stability depends on both bony stability and the integrity of the soft tissue structures. Chiefly, the deep deltoid ligament has been shown to play a central role in stabilization of the ankle and prevention of lateral shift and external rotation of the talus.^{1, 2, 3, 4, 5} While stable ankle fractures have been shown to have successful long-term non-operative outcomes, unstable ankle fractures are managed most successfully with open anatomic reduction and internal fixation.^{3,6, 7, 8, 9} Early identification of instabililty is important when injury to the medial structures is suspected. External rotation stress radiograph with measurement of the medial clear space has been described as the gold standard for evaluation of deep deltoid ligament integrity.^{1,10, 11, 12}

Michelson¹³ showed that the deep deltoid ligament is most directly responsible for prevention of lateral talar shift. Use of clinical signs such as swelling, ecchymosis, or medial tenderness alone is limited in predicting ankle fracture instability and severity.^10,14,15 The use of stress radiographs has been showed to more accurately provide such information.^{11,16, 17, 18} Several studies have showed that gravity and manual stress views can be used interchangeably.^19,20 While the method of evaluation of intact medial structures has been well defined, there remain multiple ways instability is defined in the literature, including a medial clear space ≥4 mm and 1 mm greater than the superior tibiotalar space,^{1,3,10,15,16,19} a medial clear space ≥5 mm,²¹ and an increase in medial clear space of 2 mm from its baseline value.¹³

The purpose of this study was to assess the amount of medial clear space widening that occurs with a manual external rotation stress test in uninjured ankles. We hypothesized that the mean medial clear space would be less than 5 mm and that the increase in medial clear space would be less than 2 mm greater than baseline.

2. Methods

Investigators identified a cohort of adult patients undergoing operative fixation of unstable ankle fractures and prospectively enrolled these patients into an IRB approved study in which they would undergo manual external rotation stress examination of their contralateral ankle while under anesthesia. All patients signed informed consent. Exclusion criteria were age less than 18 years, prior ankle injury or known instability of the unaffected extremity, systemic musculoskeletal disorders, poly-trauma and incidental abnormal radiographic findings.

The investigators obtained a series of fluoroscopic images of the unaffected ankle in the operating room (OR) prior to fixation of the injured ankle. The image series included (1) a non-stressed mortise view, obtained with the leg internally rotated 15–20°, allowing the intermalleolar line to be parallel to the detector and (2) a manual external rotation stress view as previously described.^13,20 A standardized marker included on each image provided the ability to correct for magnification differences between studies.

After obtaining imaging, three distinct reviewers performed a series of measurements on the radiographs. The imaging was de-identified and presented in a randomized order to the reviewers who were blinded as to whether the images were stress or non-stress images. Each reviewer measured the medial clear space [Fig. 1]. A power analysis performed based on prior studies measuring medial clear space widening¹⁴ determined that 7 subjects were needed for adequate statistical power – twice this number of patients were enrolled in the study to increase this power. Additionally, the reviewers measured tibiofibular clear space, tibiofibular overlap, degree of talar tilt, and lateral talar shift before and after stress examination [Fig. 1]. Statistical analysis included a paired t-test to compare mean medial clear space, tibiofibular clear space, tibiofibular overlap, talar tilt, and talar shift between stress and non-stress examinations. Intraclass coefficient (ICC) two-way mixed model provided analysis of measurement reliability.²²

Fig. 1.

Example of mortise nonstress (#1) and stress (#2) fluoroscopic images with measurement of marker (A), Tibiofibular overlap (B), Tibiofibular clear space (C), and medial clear space (D).

3. Results

Thirty fluoroscopic images on fifteen patients were obtained. The study group demographic consisted of 5 females and 10 males, with a mean age of 42 years old (Range -19-82 years old), with imaging taken of 9 left ankles and 6 right ankles. The mean medial clear space on the non-stressed mortise view was 3.1 mm (SD-0.69; Range 1.9 to 4.2, 95% CI [2.75, 3.45]) versus a mean of 3.2 mm (SD-0.71; Range 2.0 to 4.7, 95% CI [2.94, 3.66]) in the stressed mortise view group. Mean difference was 0.10 mm (SD-0.63; Range −0.7 mm to 1.5 mm, 95% CI [-0.14 to 0.34]). No statistically significant difference was observed (p = 0.39) [Table 1].

Table 1.

Summary of measurements preformed in millimeters with mean and standard deviation.

Test	Unstressed	Stressed	Difference
Medial Clear Space	3.14 ± 0.63	3.24 ± 0.69	0.10 ± 0.65
Tibia Fibula Overlap	3.28 ± 2.13	3.64 ± 1.60	0.36 ± 1.43
Tibia Fibula Clear Space	3.90 ± 0.77	3.68 ± 0.70	−0.22 ± 0.65

Mean tibiofibular clear space on the non-stressed mortise view was 3.9 mm (SD-0.77; Range 2.3 to 5.3, 95% CI [3.5, 4.3]) versus a mean of 3.7 mm (SD-0.70; Range 2.3 to 4.8, 95% CI [3.3, 4.1]) in the stressed mortise view group. Mean difference was −0.23 mm (SD-0.65; Range −1.42 mm to 0.91, 95% CI [-0.58, 0.14]). No statistically significant difference was observed (p = 0.21). Mean tibiofibular overlap on the non-stressed mortise view was 3.3 mm (SD-2.13; Range 0.0 to 7.7, 95% CI [2.1,4.5]) versus a mean of 3.6 mm (SD-1.60; Range 0.4 to 7.2, 95% CI [2.8, 4.5]) in the stressed mortise view group. Mean difference was 0.36 mm (SD-1.43; Range −2.2 mm to 3.7 mm, 95% CI [-0.43, 1.15]). No statistically significant difference was observed (p = 0.34).

Mean talar tilt on the non-stressed mortise view was 0.8° valgus (SD-0.92; Range 0 to 3.1, 95% CI [0.28, 1.4], versus 3.0° valgus (SD-1.7; Range 1.2 to 7.1, 95% CI [2.0, 4.0]. Mean increase in valgus angulation of talus was 2.2° (SD-1.5, Range 0.3 to 4.9, 95% CI [1.3, 3.1]. Mean lateral talar shift between the non-stressed and stressed mortise views was 0.51 mm (SD-0.51; Range 0–1.6, 95% CI [0.24, 0.80]. Inter-rater reliability was excellent between all observers for medial clear space (ICC-0.88; CI [0.78, 0.94]), tibiofibular clear space (ICC-0.77; CI [0.53, 0.89]), tibiofibular overlap (ICC-0.97; CI [0.95, 0.99]), talar tilt (ICC-0.93, CI [0.87, 0.97], and talar shift (ICC 0.82, CI [0.65,0.91].

4. Discussion

External stress radiographs to evaluate for injury to the medial structures in ankle fractures provides objective information for clinical decision-making. McConnell¹⁰ evaluated 97 patients who presented with Weber type-B supination-external rotation (SER) fibular fractures and found that clinical signs of medial tenderness, ecchymosis, and swelling were not predictive of deltoid incompetence. Egol¹⁵ evaluated 101 patients with evidence of an isolated fibular fracture and an intact mortise seen on a standard ankle trauma radiograph series, finding a positive stress radiograph (defined as > 4 mm of medial clear space widening) in 65% of patients and low sensitivity of clinical signs in predicting ankle instability, highlighting the value of a stress radiograph in diagnosing ankle fractures.

However, controversy remains regarding an appropriate cutoff for ankle instability when referencing medial clear space widening. Park et al.²¹ evaluated stress radiographic views of 6 cadaveric ankles destabilized according to the SER mechanism of Lauge-Hansen and found that a medial clear space of >5 mm was superior to both >4 mm and >6 mm in specificity and positive predictive value in diagnosing deep deltoid ligament transection. Further, Michelson et al.¹³ advocated for the efficacy of >2 mm lateral shift of the talus and valgus tilt of >15° in identifying deep deltoid ligament transection. In our study, while 2/15 (13%) had a stressed medial clear space greater than 4 mm, no patients had a medial clear space greater than 5 mm. Additionally, no patients had greater than 2 mm lateral shift between non-stressed and stressed examinations, nor did they have >15° in valgus tilt of the talus.

Utilization of anatomic measurements other than medial clear space to determine ankle fracture stability has not been well defined until recently. DeAngelis et al.¹⁴ examined the correlation between measurements of the superior clear space and found significant differences between the two. Harper et al.²³ proposed the usefulness of measuring the tibiofibular clear space and overlap in ankle injuries and identified baseline measurements in non-injured ankles. More recently, Matuszewski et al.²⁴ examined the value of stress external rotation versus lateral fibular stress, comparing medial clear space, tibiofibular clear space, and tibiofibular overlap, suggesting measurement of medial clear space to be the most sensitive measurement of the three. In our study, measurements of tibiofibular clear space and tibiofibular overlap were taken in addition to measurement of the medial clear space. Similar to the measurements of medial clear space, there was no statistical difference between stressed and non-stressed views. Interestingly, while the inter-rater reliability was excellent for all three measurements and even superior in the tibiofibular overlap group, the large standard deviation in the two measurements additional to medical clear space limits their usefulness to clinicians. Our results support the use of using the medial clear space as the primary determinant of ankle instability when performing a stress examination of the ankle.

Despite these limitations, this study has several strengths. This is the first study to prospectively assess in vivo ankle stability with measurements conducted on non-stressed and stressed normal ankles. Unlike cadaveric models, this was a study of natural soft tissue integrity in physiologically normal ankles. Additionally, the stress tests were performed in an ideal situation where the patient was anesthetized. The investigators could therefore adequately stress the ankle without hindrance from patient guarding or intolerance of the stress exam. Finally, the reviewers displayed high ICC for the measures, indicating that these measurement techniques have strong external validity.

5. Conclusions

We were able to demonstrate a low likelihood of abnormal medial clear space widening. Therefore, we surmise that it is unlikely to have a false positive result from a stress exam in an injured patient. Our results support the previous literature and allow us to advocate for ankle fractures with >5 mm medial clear space after external rotational stress to be considered unstable. Additionally, ankles with a medial clear space between 4 and 5 mm, instability should be considered only if lateral shift is > 2 mm on stress examination. Our data shows that no physiologically healthy ankles widened beyond these established cut-offs before or after the manual external rotation stress. The value of >5 mm of medial clear space widening and >2 mm lateral shift are thus valid assessments of ankle stability, and the use of manual ER stress radiographs is unlikely to result in false positives using these thresholds.

Conflicts of interest

None.

Appendix A. Supplementary data

The following is the Supplementary data to this article: