- Email: [email protected]
Correlation and comparison of syndesmosis dimension on CT and MRI
The Foot 28 (2016) 36–41
Contents lists available at ScienceDirect
The Foot journal homepage: www.elsevier.com/locate/foot
Correlation and comparison of syndesmosis dimension on CT and MRI Fabian Wong ∗ , Rebecca Mills, Nadeem Mushtaq, Roland Walker, Samrendu K. Singh, Ali Abbasian Adult Foot and Ankle Reconstruction Unit, Guy’s & St Thomas’ Hospital NHS Foundation Trust, Great Maze Pond, London Bridge, SE9 2RT, London, United Kingdom
h i g h l i g h t s • Misdiagnosis of the distal tibiofibular syndesmotic injury could lead to significant long-term morbidity. • Our results show that CT scan measurements of the distal tibiofibular joint are wider than and significantly different from that measured from MRI. • We therefore recommend that when the widening found on a CT scan is minor or the diagnosis is equivocal, a contralateral comparative CT or an ipsilateral MRI scan is obtained to prevent misdiagnosis.
• To our knowledge, this is the first study to correlate CT and MRI findings of the distal tibiofibular joint.
a r t i c l e
i n f o
Article history: Received 6 October 2015 Received in revised form 25 May 2016 Accepted 15 June 2016 Keywords: Distal tibiofibular syndesmosis CT MRI Correlation
a b s t r a c t Introduction: Various methods using CT scan have been described to diagnose distal tibiofibular syndesmotic injuries. However, CT scan does not take into account the amount of cartilage within the distal tibiofibular joint and could therefore lead to false positive results. We present the first study correlating the findings of the distal tibiofibular syndesmosis on CT and MRI scans. Methods: CT and MRI scan of consecutive patients over a period of 18 months, and of a time lapsed less than 12 months between the two imaging modalities, were reviewed. Measurements of the distal tibiofibular syndesmosis were taken according to a previously published study at the level of the distal tibial physeal scar. Results: Twenty-six ankles from 25 patients were included in this study for analysis. Significant difference between CT and MRI assessments in the overall distal tibiofibular dimensions and in the posterior distal tibiofibular distance for those ankles with evidence of osteoarthritis was found. Interclass correlation coefficients suggest that such methodology was reproducible and reliable. Conclusion: When the widening found on a CT scan is minor or the diagnosis is equivocal, a contralateral comparative CT or an ipsilateral MRI scan is recommended to prevent misdiagnosis. Level of evidence: Level IV. © 2016 Elsevier Ltd. All rights reserved.
1. Introduction Accurate assessment of distal tibiofibular joint congruity is important when syndesmosis injury is suspected following trauma to the ankle. Misdiagnosis and inadequate treatment could result in significant long-term morbidity, including pain, instability, and degenerative changes of the ankle joint [2,8,11,16,21]. It is estimated that syndesmosis injuries account for 1–18% of all ankle sprains, although the true incidence is likely to be higher due to misdiagnosis in those with subtle diastasis [3,15].
∗ Corresponding author. E-mail addresses: [email protected], [email protected] (F. Wong). http://dx.doi.org/10.1016/j.foot.2016.06.001 0958-2592/© 2016 Elsevier Ltd. All rights reserved.
Currently, there are no agreed criteria when considering the distal tibiofibular separation on cross sectional imaging. A recent study by Elgafy et al. has examined the normal dimensions of the syndesmosis on Computerised Tomography (CT) [7]. However, it did not give consideration to the fact that variations exist within the distal tibiofibular joint where articulating facets could be covered with cartilage as demonstrated by previous anatomical studies, and may contribute to the false appearance of increased width of the syndesmosis. Magnetic Resonance Imaging (MRI) would allow such assessment but there is no published study comparing the findings between these imaging modalities. The aim of this study is to measure and compare anterior and posterior distances of the distal tibiofibular syndesmosis on MRI and CT imaging. To our knowledge, this is the first study to compare such measurements.
F. Wong et al. / The Foot 28 (2016) 36–41
37
Fig. 1. CT images (Coronal and Axial planes) for measuring distal tibiofibular distances at the level of physeal scar. Red lines (Coronal) demonstrating the level of physeal scar and distal tibiofibular distances measured (Axial). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 2. MRI images (Coronal and Axial planes) for measuring distal tibiofibular distances at the level of physeal scar. Red lines (Coronal) demonstrating the level of physeal scar and distal tibiofibular distances measured (Axial). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
2. Materials and methods Consecutive patients over a period of 18 months who had had both a CT scan and an MRI scan of their ipsilateral ankle, were identified from the Picture Archiving and Communication System (PACS) of a single teaching hospital. Those who had the two imaging modalities more than 12 months apart were excluded. Only patients over the age of eighteen, who had their imaging per-
formed to investigate symptoms that were unrelated to the distal tibiofibular syndesmosis were included. Patients who had imaging following trauma to the ankle in the last 3 months, or the presence of a fracture or dislocation prior to imaging, and those who had a suspected chronic distal tibiofibular injury, were excluded from this study. All CT images were obtained using a 256-slice scanner from Philips (iCT Family; Philips Healthcare (UK), Surrey, UK), while MRI
38
F. Wong et al. / The Foot 28 (2016) 36–41
Fig. 3. Hand drawing to illustrate the measurements of anterior distal tibiofibular distance (Line AB) and posterior distal tibiofibular distance (Line CD) on axial planes.
scans were performed using a 1.5T scanner from Seimens (Magnetom 1.5T; Siemens Healthcare (UK), Surrey, UK). Centricity system (GE Healthcare, Buckingshire, UK) was the viewing platform used for assessing images. Our method to compare measurements on different image modalities using PACS was validated prior to the study. The distance between the medial and lateral malleoli on the localization images of the two imaging modalities was compared. A measurement of 1 mm on a CT image corresponds to 1 mm on MRI image. Patient demographics were collected and previous radiographs were reviewed for presence of osteoarthritis of the ipsilateral ankle joint. Measurements were taken on axial images on CT scan and T1 MRI sequence by two assessors independently (FW and RM respectively). Measurements were repeated on a separate occasion with the assessors assigned to the other image modality (CT and MRI for RM and FW respectively), with previous measurements unavailable to the assessors. This was arranged to minimize the effects of measurement and recall bias. These were taken at the level of the distal tibial physeal scar, identified by correlating with coronal images (Figs. 1 and 2). We have chosen the level of distal tibial physeal scar as a landmark for correlating images for axial measurements since this is readily identifiable on both imaging modalities. The anterior and the posterior distal tibiofibular dimensions were measured using the technique previously described by Elgafy et al. for measuring aforementioned dimensions on CT images [7]. This is also demonstrated by a hand-drawing in Fig. 3. The anterior distal tibiofibular distance was taken from the anterior tubercle of the tibia to the nearest point of medial aspect of the fibula (A to B on Fig. 3), while the posterior distal tibiofibular distance was measured from the most anterior aspect of the posterior tibial tubercle to the nearest point of medial aspect of the fibula (C to D on Fig. 3). The geometrical shape of the syndesmosis was also recorded. Statistical analysis was performed using SPSS software package (IBM Corp. Released 2013. IBM SPSS Statistics for Macintosh, Version 22.0. Armonk, NY: IBM Corp). Continuous data such as the anterior and posterior distal tibiofibular distances on CT and MRI scans were compared using paired t-test, while nominal data such as the geometrical shapes were analysed using McNemar test. Inter-observer reliability was also tested using intraclass correlation coefficient. 3. Results A total of 27 ankles in 26 patients fulfilling the study criteria were identified, of which 26 ankles in 25 patients were included
in this study. One patient was excluded due to the obliquity of the axial images on the MRI scan, rendering accurate assessment of the distal tibiofibular syndesmosis impossible. Of the included patients, there were 15 males and 10 females. The mean age was 38.0 (Range, 25–60) years. The reasons for obtaining the images of the ankles were for the investigation of ankle joint osteoarthritis in 8 patients, osteochondral defect in 6 patients, ankle impingement in 3 patients, subtalar joint arthritis in 3 patients, tarsal joint pain 4 patients, and for cavus deformity in 1 patient. All patients had up-to-date plain radiographs of the ipsilateral ankle. Eight patients had osteoarthritis of the ankle joint, 1 patient had arthritis of the subtalar joint and 1 patient had talonavicular joint arthritis. One patient who was investigated for subtalar arthritis was found to have a healing navicular fracture following the CT and MRI scans. In ankles that had pre-existing ipsilateral osteoarthritis, and in those that were investigated for OCD, the degenerative disease was isolated to the tibiotalar compartment on plain radiographs and was considered mild. There was no tibiotalar tilt. The MRI scans had also confirmed intact distal syndesmosis ligaments, further supporting that the distal tibiofibular syndesmosis had not previously been injured and none of the patients had observable degenerative changes in the distal tibiofibular joint demonstrated on plain radiographs. Overall the mean (Range; SD) for the anterior distal tibiofibular distance was 2.0 mm (0.8–3.9; 0.9) on MRI and 2.7 mm (1.1–4.7; 0.9) on CT (Fig. 4), whilst the mean posterior distal tibiofibular distance was 3.0 mm (0.8–5.4; 1.1) on MRI and 4.1 mm (1.4–6.1; 1.0) on CT (Fig. 5). There were statistically significant differences between the CT and MRI measurements on the mean anterior distal tibiofibular distance (p < 0.001; 95% CI −0.98 to −0.35; paired t-test) and the mean posterior distal tibiofibular distance (p < 0.001; 95% CI −1.33 to −0.80; paired t-test). In total, there were 10 ankles in the arthritis group and 16 in the non-arthritis group. Of the ankles with known arthritis of the foot and ankle joints, the mean discrepancy of the anterior and posterior distal tibiofibular distances between CT and MRI scans were −0.54 mm (SD = 0.83) and −1.12 mm (SD = 0.77) respectively. The difference between the CT and MRI measured posterior distal tibiofibular distance was statistically significant (p = 0.001; 95% CI −1.67 to −0.56; paired T-test). The difference in anterior distal tibiofibular distance was not found to be statistically significant in patients with degeneration of the foot and ankle joints (p = 0.071; 95% CI −1.13 to 0.06; paired T-test), whereas a significant difference was detected in those without degenerative conditions (p < 0.001; 95% CI −1.34 to −0.72; paired T-test) (Fig. 4). For each imaging modality, when comparing between those feet and ankles with known arthritis to those without, there was no statistical difference in either the mean anterior distal tibiofibular distances (p = 0.284 and 0.275 for MRI and CT respectively), or the mean posterior distal tibiofibular distances (p = 0.206 and 0.342 for MRI and CT respectively). When examining the shape of the syndesmosis on MRI, 56% were crescent and 44% rectangular, this was compared to 69% and 31%, respectively, on CT. There was, however, no statistical difference in the overall shape of the syndesmosis between the two radiological modalities (p = 0.625, McNemar test) (Table 1). The intraclass correlation co-efficient between the two assessors (FW and RM) for anterior and posterior distal tibiofibular distances were 0.87 and 0.76 for MRI respectively, and 0.92 and 0.89 for CT respectively. This suggests that the method used for measuring the anterior and posterior distal tibiofibular distances on CT and MRI was reproducible and reliable between assessors.
F. Wong et al. / The Foot 28 (2016) 36–41
39
Fig. 4. Comparing CT and MRI measurements of the anterior distal tibiofibular distance.
Fig. 5. Comparing CT and MRI measurements of the posterior distal tibiofibular distance.
Table 1 Comparing the shape distribution of the distal tibiofibular joint on axial CT and MRI images.
CT MRI
OA No OA OA No OA
Cresent
Rectangle
7 5 5 5
3 11 5 11
4. Discussion The distal tibiofibular joint, also known as distal tibiofibular syndesmosis, consists of the tibiofibular contact zone and syndesmotic recess, and linked by anterior inferior tibiofibular ligament, posterior inferior tibiofibular ligament, the transverse ligament and the interosseous ligament [13]. Hyaline cartilage can be found on both tibial and fibular articulating facets of the distal tibiofibular joint and is a continuation of the cartilage covering the tibial plafond and fibular malleolus respectively, with a thickness ranging from 0.5 mm to 1.0 mm [1,6,13]. Our results have demonstrated significant differences between the CT and MRI measurements of the distal tibiofibular joint. It appears that CT consistently measures a wider posterior distal tibiofibular distance in patients with or without degenerative foot and ankle conditions, and a wider anterior distal tibiofibular distance in patients without degenerative pathology, when compared to MRI.
Diagnosis of syndesmosis injury is made based on mechanism of injury, clinical presentation and examination, as well as imaging [15]. Clinical findings such as an inability to perform a single leg hop was found to have a high sensitivity, whereas a positive “squeeze test” was found to have a high specificity [19]. Imaging modalities range from plain radiographs to CT or MRI. Harper et al. first described the use of distal tibiofibular overlap and clear space on AP and mortise views respectively to assess for distal tibiofibular diastasis and subsequently illustrated this in a cadaveric study [10,12]. However, subsequent studies have demonstrated that standard anteroposterior and mortise radiographs were often inaccurate and could lead to under-diagnosis of syndesmosis injury [18,20]. With increasing accessibility, the role of CT in the diagnosis of syndesmosis injury has been assessed with the view to improve both sensitivity and specificity. In a cadaveric study, Ebraheim et al. found that diastasis of less than 3 mm were often missed on plain radiographs, whereas diastasis with a separation of 2–3 mm was clearly appreciated on CT [5]. Normal parameters of the distal tibiofibular syndesmosis on CT were also described by Elgafy et al. [7]. They found the mean anterior and posterior distal tibiofibular distances were 2 mm and 4 mm respectively. They also found that two-third of the patients had a crescent-shaped incisura fibularis, while the remaining one-third were rectangular in shape. It is interesting to note, however, that the effect of the thickness of hyaline cartilage on the articulating facets of the tibiofibular contact zone were not mentioned or considered in these studies, as CT scan would only allow assessment of the anterior and posterior
40
F. Wong et al. / The Foot 28 (2016) 36–41
distal tibiofibular distances between the cortices of the tibia and fibula, and not between the articulating cartilages. More recently, Dikos et al. reported significant gender difference in the dimensions of the distal tibiofibular joint in 54 healthy individuals [4]. They also concluded that assessment of the contralateral syndesmosis was more accurate. However, the “Side-to-Side” variability for the tibiofibular clear space (TFCS), tibiofibular overlap (TFO) and the anterior tibiofibular interval (ATF) had a mean variability of 0.7 mm for TFCS and TFO with a range of 0–2.7 mm and 0–2.3 mm respectively, and a mean variability of 0.8 mm and a range of 0–3.0 mm for ATF distance. Furthermore, this would require CT assessment of both ankles, subjecting to patients to additional radiation, as well as relying on the assumption that the contralateral ankle was free from previous syndesmosis injury. A similar study by Nault et al. reporting a series of 100 ankles in 93 patients with an extended set of parameters had also demonstrated large variability in the anterior and posterior distal tibiofibular distances, with a mean of 4 mm and 8 mm and a range of 1.5–6.4 mm and 2.1–11.5 mm respectively [17]. The authors concluded that their results would suggest findings from previous similar studies might have over-estimated the malreduction of the distal tibiofibular joint. This would be consistent with the variability of the soft tissue thickness within the distal tibiofibular joint between individuals as described by cadaveric studies [1,6,13]. Studies by Grenier et al. and Koenig et al. which utilised fluoroscopy for the assessment of the distal tibiofibular joint would potentially encounter the aforementioned problem of overestimation [9,14]. The use of MRI has been reported in the diagnosis of distal tibiofibular syndesmosis injury, however the diagnostic focus has been placed on the associated ligament injuries; presence of bone bruising; the presence of tibiofibular joint fluid; or prolapse of interspace fat [2,22]. To the best of our knowledge, no previous study has assessed and reported the normal dimensions of the distal tibiofibular joint on MRI, in a similar fashion to the work by Ebraheim et al. on CT [5]. As such, this is the first quantitative study to describe the anterior and posterior distal tibiofibular distances on MRI and compare them to that found on CT, providing reference values for clinical application when assessing the distal tibiofibular joint on MRI. A logical explanation for the observed difference is due to the variations in the hyaline cartilage thickness on the articulating facets of the tibiofibular contact zone. As reported by previous anatomical studies, the thickness of cartilage on the facets range from 0.5 to 5.0 mm, and may be absent in some patients [1,6,13]. Therefore, it is theoretically possible that the anterior and posterior distal tibiofibular distances between the cortices could appear widened even when the distal tibiofibular articulating facets are, in fact, in contact. Analysis was also performed separately on those patients who demonstrated degenerative changes in their foot and ankle images, as we felt that thinning of the cartilage within the distal tibiofibular joint might be present even if there was no observable changes within the distal tibiofibular joint on plain radiographs, and may affect measurements. A lack of a significant difference in the anterior distal tibiofibular distance in patients with degenerative disease between the two modalities was observed. This may be explained by the previously described finding that the thickest part of the articulating cartilage was located posteriorly [1,6]. As a result, in the presence of degenerative disease the anterior thinner cartilage may be affected earlier than the thicker posterior part and thus the anterior distal tibiofibular distances on MRI can approach those on CT scans even when no observable degenerative changes is present on CT scans. There are some limitations in this study, such as the small sample size. However, this constrain is due to the retrospective
nature of this study, and required clinical indication(s) for patients to undergo both CT and MRI scans for the same ankle. A priori power analysis was not performed for the same reason. A posthoc power analysis was however performed and found sufficient power for detecting the difference in both anterior and posterior distal tibiofibular distance in all ankles (>80%), and the posterior distal tibiofibular distance in those with degenerative disease. The detection of a difference in the anterior distal tibiofibular distance in patients with degenerative disease was underpowered based on the post-hoc analysis (44%) with a minimal detectable difference of 0.58 mm, whereas our results demonstrated a difference of 0.54 mm. Therefore, the results in this study have demonstrated sufficient power to reject the null hypothesis, despite the indifference in the anterior distal tibiofibular distance on CT and MRI in patients with degenerative disease, with a logical explanation proposed. In an ideal situation, a sufficiently powered, large sample population with no known foot and ankle pathology should be recruited and prospectively assessed. In addition, there is a time lapse of up to 12 months between the obtaining the CT and MRI images. As such, it is possible to argue that the state of the ankle could have been different at the two time points. However, steps to minimize this confounding factor by ensuring that no CT or MRI images for suspected syndesmosis injury were included. In this study, the physeal scar as the level for assessment was chosen. This is due to the need to correlate level of the axial images between CT and MRI, and the physeal scar is clearly visible on both CT and MRI on the coronal and sagittal images. This is a limitation of this retrospective study as the orientation of the axial images between the two imaging modalities was not necessarily standardized and could give rise to measurement bias. However, in order to ensure that our results are comparable to that reported in the current literature, the measurements as described by Elgafy et al. [7] as their methodology took into account the variations in the shape of fibula was performed. Furthermore, this would also provide a marker for our CT results against those of the published study. The validity of this method was further supported by our interclass correlation co-efficient, which suggests good agreement on the measurements between assessors.
5. Conclusion Significant differences in the measured dimensions of the syndesmosis on MRI and CT scans were found, while the overall shapes of the distal tibiofibular syndesmosis on axial imaging were similar on either radiological modality. It is our belief that the thickness of the cartilage within the distal tibiofibular syndesmosis is variable and can affect the distance measured on a CT scan. Hence when the cartilage thickness is large a false impression of widening may be given on CT imaging which means that a single CT measurement, without comparison to the uninjured side, may over diagnose syndesmosis injury. It is therefore recommended that when the widening found on CT is minor or the diagnosis is equivocal, a contralateral comparative CT or an ipsilateral MRI scan is obtained to prevent a misdiagnosis. MRI has additional diagnostic yield over CT as it will detect low ankle ligament injury, peroneal tendon injury, peroneal reticulum tear, bone bruising and chondral injuries that are all differentials for ongoing pain following sprain that would be missed on CT. Further studies are needed to establish the reliability of MRI scans when investigating normal and abnormal syndesmoses.
F. Wong et al. / The Foot 28 (2016) 36–41
5.1. Brief summary • Misdiagnosis of the distal tibiofibular syndesmotic injury could lead to significant long-term morbidity. • Our results show that CT scan measurements of the distal tibiofibular joint are wider than and significantly different from that measured from MRI. • When the widening found on a CT scan is minor or the diagnosis is equivocal, a contralateral comparative CT or an ipsilateral MRI scan is recommended to prevent misdiagnosis. • To our knowledge, this is the first study to correlate CT and MRI findings of the distal tibiofibular joint. Conflict of interest None. References [1] Bartonicek J. Anatomy of the tibiofibular syndesmosis and its clinical relevance. Surg Radiol Anat 2003;25(5–6):379–86, http://dx.doi.org/10.1007/s00276-003-0156-4. [2] Brown KW, Morrison WB, Schweitzer ME, Parellada JA, Nothnagel H. MRI findings associated with distal tibiofibular syndesmosis injury. AJR Am J Roentgenol 2004;182(1):131–6, http://dx.doi.org/10.2214/ajr.182.1.1820131. [3] Del Buono A, Florio A, Boccanera MS, Maffulli N. Syndesmosis injuries of the ankle. Curr Rev Musculoskelet Med 2013;6(4):313–9, http://dx.doi.org/10.1007/s12178-013-9183-x. [4] Dikos GD, Heisler J, Choplin RH, Weber TG. Normal tibiofibular relationships at the syndesmosis on axial CT imaging. J Orthop Trauma 2012;26(7):433–8, http://dx.doi.org/10.1097/BOT.0b013e3182535f30. [5] Ebraheim NA, Lu J, Yang H, Mekhail AO, Yeasting RA. Radiographic and CT evaluation of tibiofibular syndesmotic diastasis: a cadaver study. Foot Ankle Int 1997;18(11):693–8. [6] Ebraheim NA, Taser F, Shafiq Q, Yeasting RA. Anatomical evaluation and clinical importance of the tibiofibular syndesmosis ligaments. Surg Radiol Anat 2006;28(2):142–9, http://dx.doi.org/10.1007/s00276-006-0077-0. [7] Elgafy H, Semaan HB, Blessinger B, Wassef A, Ebraheim NA. Computed tomography of normal distal tibiofibular syndesmosis. Skeletal Radiol 2010;39(6):559–64, http://dx.doi.org/10.1007/s00256-009-0809-4.
41
[8] Grass R, Rammelt S, Biewener A, Zwipp H. Peroneus longus ligamentoplasty for chronic instability of the distal tibiofibular syndesmosis. Foot Ankle Int 2003;24(5):392–7. [9] Grenier S, Benoit B, Rouleau DM, et al. APTF: anteroposterior tibiofibular ratio, a new reliable measure to assess syndesmotic reduction. J Orthop Trauma 2013;27(4):207–11, http://dx.doi.org/10.1097/BOT.0b013e31826623cc. [10] Harper MC. An anatomic and radiographic investigation of the tibiofibular clear space. Foot Ankle 1993;14(8):455–8. [11] Harper MC. Delayed reduction and stabilization of the tibiofibular syndesmosis. Foot Ankle Int 2001;22(1):15–8. [12] Harper MC, Keller TS. A radiographic evaluation of the tibiofibular syndesmosis. Foot Ankle 1989;10(3):156–60. [13] Hermans JJ, Beumer A, de Jong TA, Kleinrensink GJ. Anatomy of the distal tibiofibular syndesmosis in adults: a pictorial essay with a multimodality approach. J Anat 2010;217(6):633–45, http://dx.doi.org/10.1111/j.1469-7580.2010.01302.x. [14] Koenig SJ, Tornetta 3rd P, Merlin G, et al. Can we tell if the syndesmosis is reduced using fluoroscopy? J Orthop Trauma 2015, http://dx.doi.org/10.1097/BOT.0000000000000296. [15] Molinari A, Stolley M, Amendola A. High ankle sprains (syndesmotic) in athletes: diagnostic challenges and review of the literature. Iowa Orthop J 2009;29:130–8. [16] Morris MW, Rice P, Schneider TE. Distal tibiofibular syndesmosis reconstruction using a free hamstring autograft. Foot Ankle Int 2009;30(6):506–11, http://dx.doi.org/10.3113/FAI.2009.0506. [17] Nault ML, Hebert-Davies J, Laflamme GY, Leduc S. CT scan assessment of the syndesmosis: a new reproducible method. J Orthop Trauma 2013;27(11):638–41, http://dx.doi.org/10.1097/BOT.0b013e318284785a. [18] Ostrum RF, De Meo P, Subramanian R. A critical analysis of the anterior-posterior radiographic anatomy of the ankle syndesmosis. Foot Ankle Int 1995;16(3):128–31. [19] Sman AD, Hiller CE, Rae K, et al. Diagnostic accuracy of clinical tests for ankle syndesmosis injury. Br J Sports Med 2013, http://dx.doi.org/10.1136/bjsports-2013-092787. [20] Takao M, Ochi M, Naito K, et al. Arthroscopic diagnosis of tibiofibular syndesmosis disruption. Arthroscopy 2001;17(8):836–43. [21] van den Bekerom MP, de Leeuw PA, van Dijk CN. Delayed operative treatment of syndesmotic instability. Current concepts review. Injury 2009;40(11):1137–42, http://dx.doi.org/10.1016/j.injury.2009.03.011. [22] Vogl TJ, Hochmuth K, Diebold T, et al. Magnetic resonance imaging in the diagnosis of acute injured distal tibiofibular syndesmosis. Invest Radiol 1997;32(7):401–9.
Contents lists available at ScienceDirect
The Foot journal homepage: www.elsevier.com/locate/foot
Correlation and comparison of syndesmosis dimension on CT and MRI Fabian Wong ∗ , Rebecca Mills, Nadeem Mushtaq, Roland Walker, Samrendu K. Singh, Ali Abbasian Adult Foot and Ankle Reconstruction Unit, Guy’s & St Thomas’ Hospital NHS Foundation Trust, Great Maze Pond, London Bridge, SE9 2RT, London, United Kingdom
h i g h l i g h t s • Misdiagnosis of the distal tibiofibular syndesmotic injury could lead to significant long-term morbidity. • Our results show that CT scan measurements of the distal tibiofibular joint are wider than and significantly different from that measured from MRI. • We therefore recommend that when the widening found on a CT scan is minor or the diagnosis is equivocal, a contralateral comparative CT or an ipsilateral MRI scan is obtained to prevent misdiagnosis.
• To our knowledge, this is the first study to correlate CT and MRI findings of the distal tibiofibular joint.
a r t i c l e
i n f o
Article history: Received 6 October 2015 Received in revised form 25 May 2016 Accepted 15 June 2016 Keywords: Distal tibiofibular syndesmosis CT MRI Correlation
a b s t r a c t Introduction: Various methods using CT scan have been described to diagnose distal tibiofibular syndesmotic injuries. However, CT scan does not take into account the amount of cartilage within the distal tibiofibular joint and could therefore lead to false positive results. We present the first study correlating the findings of the distal tibiofibular syndesmosis on CT and MRI scans. Methods: CT and MRI scan of consecutive patients over a period of 18 months, and of a time lapsed less than 12 months between the two imaging modalities, were reviewed. Measurements of the distal tibiofibular syndesmosis were taken according to a previously published study at the level of the distal tibial physeal scar. Results: Twenty-six ankles from 25 patients were included in this study for analysis. Significant difference between CT and MRI assessments in the overall distal tibiofibular dimensions and in the posterior distal tibiofibular distance for those ankles with evidence of osteoarthritis was found. Interclass correlation coefficients suggest that such methodology was reproducible and reliable. Conclusion: When the widening found on a CT scan is minor or the diagnosis is equivocal, a contralateral comparative CT or an ipsilateral MRI scan is recommended to prevent misdiagnosis. Level of evidence: Level IV. © 2016 Elsevier Ltd. All rights reserved.
1. Introduction Accurate assessment of distal tibiofibular joint congruity is important when syndesmosis injury is suspected following trauma to the ankle. Misdiagnosis and inadequate treatment could result in significant long-term morbidity, including pain, instability, and degenerative changes of the ankle joint [2,8,11,16,21]. It is estimated that syndesmosis injuries account for 1–18% of all ankle sprains, although the true incidence is likely to be higher due to misdiagnosis in those with subtle diastasis [3,15].
∗ Corresponding author. E-mail addresses: [email protected], [email protected] (F. Wong). http://dx.doi.org/10.1016/j.foot.2016.06.001 0958-2592/© 2016 Elsevier Ltd. All rights reserved.
Currently, there are no agreed criteria when considering the distal tibiofibular separation on cross sectional imaging. A recent study by Elgafy et al. has examined the normal dimensions of the syndesmosis on Computerised Tomography (CT) [7]. However, it did not give consideration to the fact that variations exist within the distal tibiofibular joint where articulating facets could be covered with cartilage as demonstrated by previous anatomical studies, and may contribute to the false appearance of increased width of the syndesmosis. Magnetic Resonance Imaging (MRI) would allow such assessment but there is no published study comparing the findings between these imaging modalities. The aim of this study is to measure and compare anterior and posterior distances of the distal tibiofibular syndesmosis on MRI and CT imaging. To our knowledge, this is the first study to compare such measurements.
F. Wong et al. / The Foot 28 (2016) 36–41
37
Fig. 1. CT images (Coronal and Axial planes) for measuring distal tibiofibular distances at the level of physeal scar. Red lines (Coronal) demonstrating the level of physeal scar and distal tibiofibular distances measured (Axial). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 2. MRI images (Coronal and Axial planes) for measuring distal tibiofibular distances at the level of physeal scar. Red lines (Coronal) demonstrating the level of physeal scar and distal tibiofibular distances measured (Axial). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
2. Materials and methods Consecutive patients over a period of 18 months who had had both a CT scan and an MRI scan of their ipsilateral ankle, were identified from the Picture Archiving and Communication System (PACS) of a single teaching hospital. Those who had the two imaging modalities more than 12 months apart were excluded. Only patients over the age of eighteen, who had their imaging per-
formed to investigate symptoms that were unrelated to the distal tibiofibular syndesmosis were included. Patients who had imaging following trauma to the ankle in the last 3 months, or the presence of a fracture or dislocation prior to imaging, and those who had a suspected chronic distal tibiofibular injury, were excluded from this study. All CT images were obtained using a 256-slice scanner from Philips (iCT Family; Philips Healthcare (UK), Surrey, UK), while MRI
38
F. Wong et al. / The Foot 28 (2016) 36–41
Fig. 3. Hand drawing to illustrate the measurements of anterior distal tibiofibular distance (Line AB) and posterior distal tibiofibular distance (Line CD) on axial planes.
scans were performed using a 1.5T scanner from Seimens (Magnetom 1.5T; Siemens Healthcare (UK), Surrey, UK). Centricity system (GE Healthcare, Buckingshire, UK) was the viewing platform used for assessing images. Our method to compare measurements on different image modalities using PACS was validated prior to the study. The distance between the medial and lateral malleoli on the localization images of the two imaging modalities was compared. A measurement of 1 mm on a CT image corresponds to 1 mm on MRI image. Patient demographics were collected and previous radiographs were reviewed for presence of osteoarthritis of the ipsilateral ankle joint. Measurements were taken on axial images on CT scan and T1 MRI sequence by two assessors independently (FW and RM respectively). Measurements were repeated on a separate occasion with the assessors assigned to the other image modality (CT and MRI for RM and FW respectively), with previous measurements unavailable to the assessors. This was arranged to minimize the effects of measurement and recall bias. These were taken at the level of the distal tibial physeal scar, identified by correlating with coronal images (Figs. 1 and 2). We have chosen the level of distal tibial physeal scar as a landmark for correlating images for axial measurements since this is readily identifiable on both imaging modalities. The anterior and the posterior distal tibiofibular dimensions were measured using the technique previously described by Elgafy et al. for measuring aforementioned dimensions on CT images [7]. This is also demonstrated by a hand-drawing in Fig. 3. The anterior distal tibiofibular distance was taken from the anterior tubercle of the tibia to the nearest point of medial aspect of the fibula (A to B on Fig. 3), while the posterior distal tibiofibular distance was measured from the most anterior aspect of the posterior tibial tubercle to the nearest point of medial aspect of the fibula (C to D on Fig. 3). The geometrical shape of the syndesmosis was also recorded. Statistical analysis was performed using SPSS software package (IBM Corp. Released 2013. IBM SPSS Statistics for Macintosh, Version 22.0. Armonk, NY: IBM Corp). Continuous data such as the anterior and posterior distal tibiofibular distances on CT and MRI scans were compared using paired t-test, while nominal data such as the geometrical shapes were analysed using McNemar test. Inter-observer reliability was also tested using intraclass correlation coefficient. 3. Results A total of 27 ankles in 26 patients fulfilling the study criteria were identified, of which 26 ankles in 25 patients were included
in this study. One patient was excluded due to the obliquity of the axial images on the MRI scan, rendering accurate assessment of the distal tibiofibular syndesmosis impossible. Of the included patients, there were 15 males and 10 females. The mean age was 38.0 (Range, 25–60) years. The reasons for obtaining the images of the ankles were for the investigation of ankle joint osteoarthritis in 8 patients, osteochondral defect in 6 patients, ankle impingement in 3 patients, subtalar joint arthritis in 3 patients, tarsal joint pain 4 patients, and for cavus deformity in 1 patient. All patients had up-to-date plain radiographs of the ipsilateral ankle. Eight patients had osteoarthritis of the ankle joint, 1 patient had arthritis of the subtalar joint and 1 patient had talonavicular joint arthritis. One patient who was investigated for subtalar arthritis was found to have a healing navicular fracture following the CT and MRI scans. In ankles that had pre-existing ipsilateral osteoarthritis, and in those that were investigated for OCD, the degenerative disease was isolated to the tibiotalar compartment on plain radiographs and was considered mild. There was no tibiotalar tilt. The MRI scans had also confirmed intact distal syndesmosis ligaments, further supporting that the distal tibiofibular syndesmosis had not previously been injured and none of the patients had observable degenerative changes in the distal tibiofibular joint demonstrated on plain radiographs. Overall the mean (Range; SD) for the anterior distal tibiofibular distance was 2.0 mm (0.8–3.9; 0.9) on MRI and 2.7 mm (1.1–4.7; 0.9) on CT (Fig. 4), whilst the mean posterior distal tibiofibular distance was 3.0 mm (0.8–5.4; 1.1) on MRI and 4.1 mm (1.4–6.1; 1.0) on CT (Fig. 5). There were statistically significant differences between the CT and MRI measurements on the mean anterior distal tibiofibular distance (p < 0.001; 95% CI −0.98 to −0.35; paired t-test) and the mean posterior distal tibiofibular distance (p < 0.001; 95% CI −1.33 to −0.80; paired t-test). In total, there were 10 ankles in the arthritis group and 16 in the non-arthritis group. Of the ankles with known arthritis of the foot and ankle joints, the mean discrepancy of the anterior and posterior distal tibiofibular distances between CT and MRI scans were −0.54 mm (SD = 0.83) and −1.12 mm (SD = 0.77) respectively. The difference between the CT and MRI measured posterior distal tibiofibular distance was statistically significant (p = 0.001; 95% CI −1.67 to −0.56; paired T-test). The difference in anterior distal tibiofibular distance was not found to be statistically significant in patients with degeneration of the foot and ankle joints (p = 0.071; 95% CI −1.13 to 0.06; paired T-test), whereas a significant difference was detected in those without degenerative conditions (p < 0.001; 95% CI −1.34 to −0.72; paired T-test) (Fig. 4). For each imaging modality, when comparing between those feet and ankles with known arthritis to those without, there was no statistical difference in either the mean anterior distal tibiofibular distances (p = 0.284 and 0.275 for MRI and CT respectively), or the mean posterior distal tibiofibular distances (p = 0.206 and 0.342 for MRI and CT respectively). When examining the shape of the syndesmosis on MRI, 56% were crescent and 44% rectangular, this was compared to 69% and 31%, respectively, on CT. There was, however, no statistical difference in the overall shape of the syndesmosis between the two radiological modalities (p = 0.625, McNemar test) (Table 1). The intraclass correlation co-efficient between the two assessors (FW and RM) for anterior and posterior distal tibiofibular distances were 0.87 and 0.76 for MRI respectively, and 0.92 and 0.89 for CT respectively. This suggests that the method used for measuring the anterior and posterior distal tibiofibular distances on CT and MRI was reproducible and reliable between assessors.
F. Wong et al. / The Foot 28 (2016) 36–41
39
Fig. 4. Comparing CT and MRI measurements of the anterior distal tibiofibular distance.
Fig. 5. Comparing CT and MRI measurements of the posterior distal tibiofibular distance.
Table 1 Comparing the shape distribution of the distal tibiofibular joint on axial CT and MRI images.
CT MRI
OA No OA OA No OA
Cresent
Rectangle
7 5 5 5
3 11 5 11
4. Discussion The distal tibiofibular joint, also known as distal tibiofibular syndesmosis, consists of the tibiofibular contact zone and syndesmotic recess, and linked by anterior inferior tibiofibular ligament, posterior inferior tibiofibular ligament, the transverse ligament and the interosseous ligament [13]. Hyaline cartilage can be found on both tibial and fibular articulating facets of the distal tibiofibular joint and is a continuation of the cartilage covering the tibial plafond and fibular malleolus respectively, with a thickness ranging from 0.5 mm to 1.0 mm [1,6,13]. Our results have demonstrated significant differences between the CT and MRI measurements of the distal tibiofibular joint. It appears that CT consistently measures a wider posterior distal tibiofibular distance in patients with or without degenerative foot and ankle conditions, and a wider anterior distal tibiofibular distance in patients without degenerative pathology, when compared to MRI.
Diagnosis of syndesmosis injury is made based on mechanism of injury, clinical presentation and examination, as well as imaging [15]. Clinical findings such as an inability to perform a single leg hop was found to have a high sensitivity, whereas a positive “squeeze test” was found to have a high specificity [19]. Imaging modalities range from plain radiographs to CT or MRI. Harper et al. first described the use of distal tibiofibular overlap and clear space on AP and mortise views respectively to assess for distal tibiofibular diastasis and subsequently illustrated this in a cadaveric study [10,12]. However, subsequent studies have demonstrated that standard anteroposterior and mortise radiographs were often inaccurate and could lead to under-diagnosis of syndesmosis injury [18,20]. With increasing accessibility, the role of CT in the diagnosis of syndesmosis injury has been assessed with the view to improve both sensitivity and specificity. In a cadaveric study, Ebraheim et al. found that diastasis of less than 3 mm were often missed on plain radiographs, whereas diastasis with a separation of 2–3 mm was clearly appreciated on CT [5]. Normal parameters of the distal tibiofibular syndesmosis on CT were also described by Elgafy et al. [7]. They found the mean anterior and posterior distal tibiofibular distances were 2 mm and 4 mm respectively. They also found that two-third of the patients had a crescent-shaped incisura fibularis, while the remaining one-third were rectangular in shape. It is interesting to note, however, that the effect of the thickness of hyaline cartilage on the articulating facets of the tibiofibular contact zone were not mentioned or considered in these studies, as CT scan would only allow assessment of the anterior and posterior
40
F. Wong et al. / The Foot 28 (2016) 36–41
distal tibiofibular distances between the cortices of the tibia and fibula, and not between the articulating cartilages. More recently, Dikos et al. reported significant gender difference in the dimensions of the distal tibiofibular joint in 54 healthy individuals [4]. They also concluded that assessment of the contralateral syndesmosis was more accurate. However, the “Side-to-Side” variability for the tibiofibular clear space (TFCS), tibiofibular overlap (TFO) and the anterior tibiofibular interval (ATF) had a mean variability of 0.7 mm for TFCS and TFO with a range of 0–2.7 mm and 0–2.3 mm respectively, and a mean variability of 0.8 mm and a range of 0–3.0 mm for ATF distance. Furthermore, this would require CT assessment of both ankles, subjecting to patients to additional radiation, as well as relying on the assumption that the contralateral ankle was free from previous syndesmosis injury. A similar study by Nault et al. reporting a series of 100 ankles in 93 patients with an extended set of parameters had also demonstrated large variability in the anterior and posterior distal tibiofibular distances, with a mean of 4 mm and 8 mm and a range of 1.5–6.4 mm and 2.1–11.5 mm respectively [17]. The authors concluded that their results would suggest findings from previous similar studies might have over-estimated the malreduction of the distal tibiofibular joint. This would be consistent with the variability of the soft tissue thickness within the distal tibiofibular joint between individuals as described by cadaveric studies [1,6,13]. Studies by Grenier et al. and Koenig et al. which utilised fluoroscopy for the assessment of the distal tibiofibular joint would potentially encounter the aforementioned problem of overestimation [9,14]. The use of MRI has been reported in the diagnosis of distal tibiofibular syndesmosis injury, however the diagnostic focus has been placed on the associated ligament injuries; presence of bone bruising; the presence of tibiofibular joint fluid; or prolapse of interspace fat [2,22]. To the best of our knowledge, no previous study has assessed and reported the normal dimensions of the distal tibiofibular joint on MRI, in a similar fashion to the work by Ebraheim et al. on CT [5]. As such, this is the first quantitative study to describe the anterior and posterior distal tibiofibular distances on MRI and compare them to that found on CT, providing reference values for clinical application when assessing the distal tibiofibular joint on MRI. A logical explanation for the observed difference is due to the variations in the hyaline cartilage thickness on the articulating facets of the tibiofibular contact zone. As reported by previous anatomical studies, the thickness of cartilage on the facets range from 0.5 to 5.0 mm, and may be absent in some patients [1,6,13]. Therefore, it is theoretically possible that the anterior and posterior distal tibiofibular distances between the cortices could appear widened even when the distal tibiofibular articulating facets are, in fact, in contact. Analysis was also performed separately on those patients who demonstrated degenerative changes in their foot and ankle images, as we felt that thinning of the cartilage within the distal tibiofibular joint might be present even if there was no observable changes within the distal tibiofibular joint on plain radiographs, and may affect measurements. A lack of a significant difference in the anterior distal tibiofibular distance in patients with degenerative disease between the two modalities was observed. This may be explained by the previously described finding that the thickest part of the articulating cartilage was located posteriorly [1,6]. As a result, in the presence of degenerative disease the anterior thinner cartilage may be affected earlier than the thicker posterior part and thus the anterior distal tibiofibular distances on MRI can approach those on CT scans even when no observable degenerative changes is present on CT scans. There are some limitations in this study, such as the small sample size. However, this constrain is due to the retrospective
nature of this study, and required clinical indication(s) for patients to undergo both CT and MRI scans for the same ankle. A priori power analysis was not performed for the same reason. A posthoc power analysis was however performed and found sufficient power for detecting the difference in both anterior and posterior distal tibiofibular distance in all ankles (>80%), and the posterior distal tibiofibular distance in those with degenerative disease. The detection of a difference in the anterior distal tibiofibular distance in patients with degenerative disease was underpowered based on the post-hoc analysis (44%) with a minimal detectable difference of 0.58 mm, whereas our results demonstrated a difference of 0.54 mm. Therefore, the results in this study have demonstrated sufficient power to reject the null hypothesis, despite the indifference in the anterior distal tibiofibular distance on CT and MRI in patients with degenerative disease, with a logical explanation proposed. In an ideal situation, a sufficiently powered, large sample population with no known foot and ankle pathology should be recruited and prospectively assessed. In addition, there is a time lapse of up to 12 months between the obtaining the CT and MRI images. As such, it is possible to argue that the state of the ankle could have been different at the two time points. However, steps to minimize this confounding factor by ensuring that no CT or MRI images for suspected syndesmosis injury were included. In this study, the physeal scar as the level for assessment was chosen. This is due to the need to correlate level of the axial images between CT and MRI, and the physeal scar is clearly visible on both CT and MRI on the coronal and sagittal images. This is a limitation of this retrospective study as the orientation of the axial images between the two imaging modalities was not necessarily standardized and could give rise to measurement bias. However, in order to ensure that our results are comparable to that reported in the current literature, the measurements as described by Elgafy et al. [7] as their methodology took into account the variations in the shape of fibula was performed. Furthermore, this would also provide a marker for our CT results against those of the published study. The validity of this method was further supported by our interclass correlation co-efficient, which suggests good agreement on the measurements between assessors.
5. Conclusion Significant differences in the measured dimensions of the syndesmosis on MRI and CT scans were found, while the overall shapes of the distal tibiofibular syndesmosis on axial imaging were similar on either radiological modality. It is our belief that the thickness of the cartilage within the distal tibiofibular syndesmosis is variable and can affect the distance measured on a CT scan. Hence when the cartilage thickness is large a false impression of widening may be given on CT imaging which means that a single CT measurement, without comparison to the uninjured side, may over diagnose syndesmosis injury. It is therefore recommended that when the widening found on CT is minor or the diagnosis is equivocal, a contralateral comparative CT or an ipsilateral MRI scan is obtained to prevent a misdiagnosis. MRI has additional diagnostic yield over CT as it will detect low ankle ligament injury, peroneal tendon injury, peroneal reticulum tear, bone bruising and chondral injuries that are all differentials for ongoing pain following sprain that would be missed on CT. Further studies are needed to establish the reliability of MRI scans when investigating normal and abnormal syndesmoses.
F. Wong et al. / The Foot 28 (2016) 36–41
5.1. Brief summary • Misdiagnosis of the distal tibiofibular syndesmotic injury could lead to significant long-term morbidity. • Our results show that CT scan measurements of the distal tibiofibular joint are wider than and significantly different from that measured from MRI. • When the widening found on a CT scan is minor or the diagnosis is equivocal, a contralateral comparative CT or an ipsilateral MRI scan is recommended to prevent misdiagnosis. • To our knowledge, this is the first study to correlate CT and MRI findings of the distal tibiofibular joint. Conflict of interest None. References [1] Bartonicek J. Anatomy of the tibiofibular syndesmosis and its clinical relevance. Surg Radiol Anat 2003;25(5–6):379–86, http://dx.doi.org/10.1007/s00276-003-0156-4. [2] Brown KW, Morrison WB, Schweitzer ME, Parellada JA, Nothnagel H. MRI findings associated with distal tibiofibular syndesmosis injury. AJR Am J Roentgenol 2004;182(1):131–6, http://dx.doi.org/10.2214/ajr.182.1.1820131. [3] Del Buono A, Florio A, Boccanera MS, Maffulli N. Syndesmosis injuries of the ankle. Curr Rev Musculoskelet Med 2013;6(4):313–9, http://dx.doi.org/10.1007/s12178-013-9183-x. [4] Dikos GD, Heisler J, Choplin RH, Weber TG. Normal tibiofibular relationships at the syndesmosis on axial CT imaging. J Orthop Trauma 2012;26(7):433–8, http://dx.doi.org/10.1097/BOT.0b013e3182535f30. [5] Ebraheim NA, Lu J, Yang H, Mekhail AO, Yeasting RA. Radiographic and CT evaluation of tibiofibular syndesmotic diastasis: a cadaver study. Foot Ankle Int 1997;18(11):693–8. [6] Ebraheim NA, Taser F, Shafiq Q, Yeasting RA. Anatomical evaluation and clinical importance of the tibiofibular syndesmosis ligaments. Surg Radiol Anat 2006;28(2):142–9, http://dx.doi.org/10.1007/s00276-006-0077-0. [7] Elgafy H, Semaan HB, Blessinger B, Wassef A, Ebraheim NA. Computed tomography of normal distal tibiofibular syndesmosis. Skeletal Radiol 2010;39(6):559–64, http://dx.doi.org/10.1007/s00256-009-0809-4.
41
[8] Grass R, Rammelt S, Biewener A, Zwipp H. Peroneus longus ligamentoplasty for chronic instability of the distal tibiofibular syndesmosis. Foot Ankle Int 2003;24(5):392–7. [9] Grenier S, Benoit B, Rouleau DM, et al. APTF: anteroposterior tibiofibular ratio, a new reliable measure to assess syndesmotic reduction. J Orthop Trauma 2013;27(4):207–11, http://dx.doi.org/10.1097/BOT.0b013e31826623cc. [10] Harper MC. An anatomic and radiographic investigation of the tibiofibular clear space. Foot Ankle 1993;14(8):455–8. [11] Harper MC. Delayed reduction and stabilization of the tibiofibular syndesmosis. Foot Ankle Int 2001;22(1):15–8. [12] Harper MC, Keller TS. A radiographic evaluation of the tibiofibular syndesmosis. Foot Ankle 1989;10(3):156–60. [13] Hermans JJ, Beumer A, de Jong TA, Kleinrensink GJ. Anatomy of the distal tibiofibular syndesmosis in adults: a pictorial essay with a multimodality approach. J Anat 2010;217(6):633–45, http://dx.doi.org/10.1111/j.1469-7580.2010.01302.x. [14] Koenig SJ, Tornetta 3rd P, Merlin G, et al. Can we tell if the syndesmosis is reduced using fluoroscopy? J Orthop Trauma 2015, http://dx.doi.org/10.1097/BOT.0000000000000296. [15] Molinari A, Stolley M, Amendola A. High ankle sprains (syndesmotic) in athletes: diagnostic challenges and review of the literature. Iowa Orthop J 2009;29:130–8. [16] Morris MW, Rice P, Schneider TE. Distal tibiofibular syndesmosis reconstruction using a free hamstring autograft. Foot Ankle Int 2009;30(6):506–11, http://dx.doi.org/10.3113/FAI.2009.0506. [17] Nault ML, Hebert-Davies J, Laflamme GY, Leduc S. CT scan assessment of the syndesmosis: a new reproducible method. J Orthop Trauma 2013;27(11):638–41, http://dx.doi.org/10.1097/BOT.0b013e318284785a. [18] Ostrum RF, De Meo P, Subramanian R. A critical analysis of the anterior-posterior radiographic anatomy of the ankle syndesmosis. Foot Ankle Int 1995;16(3):128–31. [19] Sman AD, Hiller CE, Rae K, et al. Diagnostic accuracy of clinical tests for ankle syndesmosis injury. Br J Sports Med 2013, http://dx.doi.org/10.1136/bjsports-2013-092787. [20] Takao M, Ochi M, Naito K, et al. Arthroscopic diagnosis of tibiofibular syndesmosis disruption. Arthroscopy 2001;17(8):836–43. [21] van den Bekerom MP, de Leeuw PA, van Dijk CN. Delayed operative treatment of syndesmotic instability. Current concepts review. Injury 2009;40(11):1137–42, http://dx.doi.org/10.1016/j.injury.2009.03.011. [22] Vogl TJ, Hochmuth K, Diebold T, et al. Magnetic resonance imaging in the diagnosis of acute injured distal tibiofibular syndesmosis. Invest Radiol 1997;32(7):401–9.