571440
research-article2015
FAIXXX10.1177/1071100715571440Foot & Ankle InternationalHaskell et al
Article
Safe Zone for Placement of Talar Screws When Fusing the Ankle With an Anterior Plating System
Foot & Ankle International® 1–6 © The Author(s) 2015 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1071100715571440 fai.sagepub.com
Andrew Haskell, MD1,2, Russell Dedini, MD2, and Monara Dini, DPM3
Abstract Background: Ankle fusions fixed with anterior plates use fluoroscopic guidance to direct screws toward the subtalar joint. Special imaging views that visualize the subtalar joint are difficult to use and can be unreliable. This study evaluated whether a single lateral ankle view would provide adequate information to judge whether a screw penetrated the subtalar joint and identified strategies that would improve this technique. Methods: In 5 cadaveric ankles fixed with anterior plates, talar screws were placed up to the subtalar joint without penetration using lateral fluoroscopy to guide screw length. After dissection, the true distance from the screw tip to subchondral surface was measured. In addition, 4 readers measured the perceived distance from screw tip to subchondral surface using direct lateral, 10 degrees cephalad tilt lateral, and 10 degrees caudal tilt lateral fluoroscopic images on 2 separate occasions. Results: Nineteen (63%) of 30 screws penetrated the subchondral bone, and screw length determined using fluoroscopy was significantly longer than screw length measured directly (29.4 ± 5.5 mm vs 27.3 ± 8.5 mm, P = .014). Measurement of screw tip to bone distance demonstrated a high level of within-reader (kappa = .871, P < .001) and between-reader agreement (kappa = .807, P < .001), but poor specificity of determining screw penetration (0.50, χ2 = 22.1, P < .001) and poor correlation between radiographically measured and actual distances between screw tip and bone margin (r = .35, P < .001). Tilting the c-arm 10 degrees cephalad and directing screws toward the posterior facet improved the ability to detect screw penetration and directing screws toward the middle facet diminished it (P < .05). A safe zone of screw placement was defined by region. Conclusion: Use of a lateral fluoroscopic image to guide talar screw placement may lead to an unacceptably high rate of subtalar joint penetration. Clinical Relevance: Understanding the limitations of lateral fluoroscopy when using anterior ankle fusion plates may minimize screw penetration into the subtalar joint and diminish development of subtalar arthropathy. Keywords: ankle, fusion, arthrodesis, anterior, subtalar, screw length Tibiotalar arthrodesis remains a reliable procedure to treat ankle arthritis.7,11,12 Anteriorly placed, contoured locking plates have been used to maximize construct stability during ankle fusion.4,8 These plates are fixed with screws directed into the talus toward the subtalar joint. Since screw pull-out strength is proportional to the surface area of threads engaging bone, screws directed into the talus should be as long as possible.3 However, the desire to maximize screw length must be balanced with the risk of penetrating the subtalar joint articular surface to avoid premature subtalar joint arthropathy. Judging appropriate screw length in the talus typically is accomplished using intraoperative fluoroscopic imaging. The utility of fluoroscopy to prevent subtalar penetration during screw placement is limited by the complex anatomy of the inferior talus. Since the subtalar facets exist in
varying planes, multiple fluoroscopic views would be required to determine if any given screw had violated a facet. Specialized views have been described over the years to view select subtalar regions, including the Harris-Beath view for the middle facet,5 the Canale view for the talar neck,2 the Broden view for the posterior facet,1 and multiple 1
Department of Orthopedics, Palo Alto Medical Foundation, Palo Alto, CA, USA 2 San Francisco Department of Orthopedic Surgery, University of California, San Francisco, San Francisco, CA, USA 3 San Francisco General Hospital, Orthopedic Trauma Institute, San Francisco, CA, USA Corresponding Author: Andrew Haskell, MD, Department of Orthopedics, Palo Alto Medical Foundation, 795 El Camino Real, Palo Alto, CA 94301, USA. Email: [email protected]
Downloaded from fai.sagepub.com at UNIV OF PITTSBURGH on March 11, 2015
2
Foot & Ankle International
Broden view subtypes for the anterior and middle facets.1 However, obtaining numerous fluoroscopic views of all facets of the subtalar joint is difficult, time consuming, inaccurate, and may increase the patient’s exposure to ionizing radiation. The purpose of this cadaveric study was to determine if a single lateral ankle fluoroscopic view would provide adequate information to judge whether a screw has been placed without penetrating the subtalar joint. We hypothesized that when using anterior ankle arthrodesis plates, an intraoperative lateral fluoroscopic view of the ankle would not allow accurate assessment of screw position relative to the inferior subchondral bone of the talus. Two methods were used to test the hypothesis. First, we attempted to place screws up to but not through the subchondral bone using only a lateral image to judge screw length. Second, we tested the ability of surgeons to judge screw penetration of the subchondral bone based on saved fluoroscopic images generated during the first step. We also searched for criteria that would allow using a lateral fluoroscopic view to safely and accurately place screws directed toward the subtalar joint.
Materials and Methods Five intact fresh frozen cadaveric ankles were obtained, none of which exhibited signs of prior trauma or congenital deformity. For each specimen, a standard anterior ankle approach was performed, followed by application of anteromedial and anterolateral ankle arthrodesis plates (Tibiaxys Ankle Fusion System, Integra LifeSciences Corporation, Plainsboro, NJ, USA) with the ankle in the standard position of fusion.9 Each plate was precontoured for the left or right side and for the medial or lateral aspect of the ankle, and included 3 divergent fixed angle talus screws directed toward different regions of the subtalar joint. For each plate, the 3 locking screws placed into the talus initially were placed short. Each of the talar screws was then sequentially replaced with a fixed angle locking screw of length required to reach the subchondral bone as visualized on a lateral ankle view using a c-arm fluoroscopy unit (Philips BV Pulsera® Image Intensifier, Philips Corporation, Andover, MA, USA) and as perceived with agreement by 2 surgeons (RD, AH) (Figure 1). Since screws were available in 2 mm increments, all screws were placed with the goal of terminating within 2 mm of the subchondral bone margin without penetration. Three c-arm images per screw were then taken at prescribed angles (straight lateral, 10 degrees cephalad lateral, and 10 degrees caudad lateral) and digitally saved. A straight lateral image was acceptable if there was no overlap of the medial and lateral sides (double shadow) of the distal tibia articular surface and of the proximal talar articular surface, and if the fibula overlapped the tibia at or slightly posterior to the center of the distal tibia articular surface as measured from anterior to posterior. Cephalad lateral images were
Figure 1. A fluoroscopic lateral image of the ankle showing a talar screw directed toward the posterior facet is shown. The screw is placed as close to the inferior talar subchondral surface (black dotted line) as possible without penetration, as perceived on the image.
taken starting with a straight lateral image, then directing the fluoroscopic beam 10 degrees from inferior-lateral to superior-medial in the coronal plane. Caudad lateral images were taken starting with the straight lateral image, then directing the fluoroscopic beam 10 degrees from superiorlateral to inferior-medial in the coronal plane. In total, 90 images were generated. Each specimen was then dissected, the subtalar joint disarticulated, and the cartilage removed from the inferior surface of the talus. The distance from the tip of each previously placed screw to the inferior talar bone surface was measured at the inferior surface of the talus of each cadaveric specimen to the nearest 1 mm. This distance was defined as positive when the screw penetrated through the inferior talus into the subtalar joint or negative when the screw was entirely within the talus. The ideal screw length was calculated using the actual screw length and the screw tip to bone distance. The 90 digital images obtained were converted to JPEG format (Adobe Photoshop CS 6, Adobe Systems Corp, San Jose, CA, USA), randomly ordered, and distributed to 4 reviewers: 1 senior orthopedic surgery resident, 1 podiatrytrained surgeon, and 2 orthopedic foot and ankle fellowship trained surgeons. For each image, the reviewers were instructed to make a qualitative judgment about whether the talar screw depicted in the image had penetrated the inferior talar subchondral bone (Figure 2). In addition, the reviewers were asked to measure the distance (mm) between the screw tip and the talar subchondral bone margin (Figure 2). All images were viewed using Adobe Photoshop CS 6. Quantitative measures were made using the Adobe Photoshop ruler tool standardized to a known size marker on the image. All images were then randomly reordered, redistributed to each reviewer, and all measures were repeated at least 1 week later.
Downloaded from fai.sagepub.com at UNIV OF PITTSBURGH on March 11, 2015
3
Haskell et al Table 1. Amount of Screw Penetration Into the Subtalar Joint Based on Anatomic Region.
Mean ± Standard Deviation (mm) Number (+Subchondral Breach, 99% of –No Subchondral Confidence Screws Breach) Interval (mm) Posterior facet Sinus tarsi Middle facet Anterior facet
10 7 5 8
−1.4 ± 3.0 6.9 ± 5.1 1.0 ± 1.0 3.1 ± 2.2
−4.5 to 1.7 −0.4 to 14.1 −1.1 to 3.1 0.4 to 5.9
A logistic regression model was used to examine the influence of multiple variables on the readers’ ability to determine if a screw had penetrated the subchondral bone. The independent variables explored included the plate position (medial or lateral), the angle of incidence of the image intensifier beam (straight lateral, 10 degrees caudad lateral, 10 degrees cephalad lateral), the anatomic region of the subtalar joint into which the screw was directed as noted after disarticulating the specimens (posterior facet, sinus tarsi, middle facet, anterior facet), and the level of training of the reviewer (senior orthopaedic surgery resident, podiatry-trained, orthopaedic foot and ankle surgery fellowshiptrained). Finally, for each of 4 anatomic regions toward which a screw could be directed, a 99% confidence interval was calculated based on the measurement of screw tip to subchondral bone taken directly on the specimens. Significance was predetermined to be an alpha level of P < .05, and data are reported as mean ± standard deviation. Figure 2. (a) The method used to measure the distance from screw tip to inferior talar subchondral bone (black dotted line) is demonstrated. (b) Subjective “in” or “out” assessment of the screw tip position is made, and the distance X the screw tip is in or out of the bone is measured.
A repeated measures t-test was used to compare the screw lengths as measured during screw placement using fluoroscopy in the lateral plane with the ideal screw length calculated from the measurement of screw tip to subchondral bone taken directly from the specimen (AnalystSoft, StatPlus:mac, version 2009). Kappa statistics were calculated to determine agreement within and between each reader regarding measurement of the screw tip to subchondral bone distance as measured on the radiographs. The ability of readers to discriminate screw position as being either within the talus (in) or having penetrated the talar subchondral bone (out) was assessed using a chi-square test; sensitivity and specificity were calculated. The correlation between the readers’ radiographically measured and the directly measured distance from the screw tip to the bone margin was calculated using Pearson’s correlation coefficient.
Results Despite the stated goal of using the lateral fluoroscopy image to place screws within 2 mm of the talar subchondral bone without penetration, 19 (63%) of 30 screws were found to have penetrated the subchondral bone when the specimens were dissected. Screw length determined using lateral fluoroscopy during screw placement was significantly longer than screw length measured after dissection and direct visualization of the subtalar joint (29.4 ± 5.5 vs 27.3 ± 8.5 mm, P = .014). The length of screw tip penetration for each anatomic region of the subtalar joint is shown in Table 1. Measurement of screw-tip to bone distance demonstrated a high level of within-reader agreement (kappa = .871, P < .001) and between-reader agreement (kappa = .807, P < .001). While readers were subjectively able to distinguish subchondral bone penetration based on the lateral fluoroscopic image (χ2 = 22.1, P < .001), there was intermediate sensitivity (.81) and poor specificity (.50). Similarly, there was a significant but poor correlation between radiographically
Downloaded from fai.sagepub.com at UNIV OF PITTSBURGH on March 11, 2015
4
Foot & Ankle International
Figure 3. A scatter plot of the amount of subtalar screw penetration measured from lateral radiographs versus the true amount of screw penetration is shown. Positive values represent screw penetration into the subtalar joint. “Out” designates screws penetration into the subtalar joint; “in” designates no screw penetration. The linear regression best-fit line is shown (r = .35, P < .001).
measured and actual distances between screw tip and bone margin based on the lateral fluoroscopic image (r = .35, P < .001) (Figure 3). Three variables were identified through logistic regression analysis to have a clinically relevant influence on whether a given screw position was correctly assessed (P < .05) (Figure 4). Tilting the c-arm 10 degrees cephalad improved the ability to detect screw penetration of the inferior talar subchondral bone (Figure 5). Screws that were directed toward the posterior talar facet were more likely to be accurately categorized as penetrating or not penetrating the subchondral bone, while screws directed toward the middle facet were less likely to be accurately categorized. In contrast, the medial or lateral position of the plate was not shown to have an influence.
Discussion This study found that use of a lateral fluoroscopic image alone to guide talar screw placement may lead to an unacceptably high rate of subtalar joint screw penetration when trying to place the screw up to but not through the subchondral bone. Of screws, 63% inadvertently penetrated the inferior talar subchondral surface. Screws placed using the fluoroscopic method were, on average 2 mm longer than the ideal screw length. The specificity of determining screw penetration by fluoroscopic imaging was low, indicating frequent incorrect belief that a screw had not penetrated the inferior talar subchondral surface, when it actually had. The high intraobserver agreement in the ability to discern the presence and amount of screw penetration of the inferior talar articular surface over multiple readings a week apart demonstrates that the measurement method employed in the study was reproducible. The high interobserver agreement among multiple surgeons with varying levels and
types of training using the same technique suggests the findings of this study are generalizable. Factors were identified that may improve the ability of surgeons to identify screws that violate the subtalar joint using a simple fluoroscopic image. Changing the angle of incidence of the fluoroscopic X-ray beam 10 degrees cephalad improved the surgeons’ discriminatory ability. Furthermore, surgeons were better able to recognize subtalar joint violation in the posterior facet, and had greater difficulty in the middle facet. Recognizing the influence of anatomic region on the accuracy of assessing screw position may help surgeons avoid placing screws that penetrate the subchondral joint. A safe zone of screw placement was defined for each region of the subtalar joint by calculating the 99% confidence interval of the screw tip to subchondral bone distance measured directly on the specimens (Table 1). Of talar screws, 99% are expected to be within 2 mm of the subchondral bone without penetration if sufficient additional space is maintained between the screw tip and perceived subchondral bone on a lateral fluoroscopic image as follows: posterior facet, 1.7 mm; middle facet, 3.1 mm; and anterior facet, 5.9 mm. We have not identified other studies that report on the ability to perceive the position of the inferior talar subchondral surface based on a lateral radiographic image. However, other studies have confirmed the difficulty in interpreting radiographic views of the talus, calcaneus and subtalar joint. Broden’s radiographic views of the subtalar joint showed poor correlation (kappa = .23) with CT scan for classifying posterior facet involvement in 45 intraarticular calcaneal fractures.6 A cadaveric study of angulation in talar neck fractures found a single Canale view may not provide adequate information when compared with multiple views.10 There are several limitations to this study. First, radiographic imaging does not supplant tactile feedback and use of a depth gauge when deciding whether any given screw is safely placed. However, tactile feedback may rely on feeling the diminution of counterforce experience as the subchondral bone is perforated, leading to articular cartilage damage. In addition, the study did not assess the efficacy of Broden’s views. This was by design, since Broden’s views would be difficult and time consuming to obtain for each screw placed, and they have shown poor correlation with CT scan when evaluating the subtalar joint.6 Finally, while this study offers guidelines for a safe zone of screw placement for each anatomic region of the subtalar joint, this is based on post hoc evaluation using a 99% confidence interval of screw penetration. Future study would be needed to validate these guidelines. In conclusion, we found that using a lateral view of the ankle as the only guide to placement of screws into the talus during ankle arthrodesis using anterior plates did not have a high degree of accuracy. Nevertheless, a lateral view may
Downloaded from fai.sagepub.com at UNIV OF PITTSBURGH on March 11, 2015
5
Haskell et al
Figure 4. The effect of a variety of factors on the probability of correct assessment of a screw tip having penetrated the talar subchondral bone is displayed. Data shown are based on logistic regression, where values closer to 1 demonstrate better ability to predict screw penetration and values closer to 0 demonstrate the inability to predict screw penetration. All categories whose probability error bars do not cross 0.5 have a statistically significant effect (P < .05). *Cephalad radiographic tilt and screws directed toward the posterior facet had a clinically relevant positive effect on the ability to assess screw tip penetration into the subtalar joint, † while screws direct toward the middle facet had a clinically relevant negative effect.
be used to judge, with reasonable confidence, that a screw is safely out of the subtalar joint if a margin for error (posterior facet, 1.7 mm; middle facet 3.1 mm; anterior facet, 5.9 mm) is observed based on the region of the subtalar joint toward which the screw is directed. Declaration of Conflicting Interests The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Integra, Inc provided the plates and screws used in the study.
Funding The author(s) received no financial support for the research, authorship, and/or publication of this article.
References Figure 5. The effect of X-ray beam tilt on the ability to determine screw tip position is demonstrated. These images show a single specimen with a long screw directed toward the posterior facet that was found, after disarticulation, to be 2 mm short of the inferior subchondral surface of the talus (completely within bone). A direct lateral image (a) gives the appearance that the screws abuts the subchondral surface. (b) Directing the X-ray beam 10 degrees caudally gives the appearance that the screw has penetrated the subchondral surface. (c) Tilting the beam 10 degrees cephalad provides the best representation of the true screw tip position above the subchondral surface.
1. Broden B. Roentgen examination of the subtaloid joint in fractures of the calcaneus. Acta Radiologica. 1949;31(1): 85-91. 2. Canale ST, Kelly FB Jr. Fractures of the neck of the talus. Long-term evaluation of seventy-one cases. J Bone Joint Surg Am. 1978;60(2):143-156. 3. Chapman JR, Harrington RM, Lee KM, et al. Factors affecting the pullout strength of cancellous bone screws. J Biomech Eng. 1996;118(3):391-398. 4. Guo C, Yan Z, Barfield WR, Hartsock LA. Ankle arthrodesis using anatomically contoured anterior plate. Foot Ankle Int. 2010;31(6):492-498. doi:10.3113/FAI.2010.0492.
Downloaded from fai.sagepub.com at UNIV OF PITTSBURGH on March 11, 2015
6
Foot & Ankle International
5. Harris RI, Beath T. Etiology of peroneal spastic flat foot. J Bone Joint Surg Br. 1948;30B(4):624-634. 6. Kwon DG, Chung CY, Lee KM, et al. Revisit of Broden’s view for intraarticular calcaneal fracture. Clin Orthop Surg. 2012;4(3):221-226. doi:10.4055/cios.2012.4.3.221. 7. Mann RA, Rongstad KM. Arthrodesis of the ankle: a critical analysis. Foot Ankle Int. 1998;19(1):3-9. 8. Plaass C, Knupp M, Barg A, Hintermann B. Anterior double plating for rigid fixation of isolated tibiotalar arthrodesis. Foot Ankle Int. 2009;30(7):631-639. doi:10.3113/ FAI.2009.0631.
9. Scranton PE Jr. An overview of ankle arthrodesis. Clin Orthop Relat Res. 1991;(268):96-101. 10. Thomas JL, Boyce BM. Radiographic analysis of the Canale view for displaced talar neck fractures. J Foot Ankle Surg. 2012;51(2):187-190. doi:10.1053/j.jfas.2011.10.037. 11. Winson IG, Robinson DE, Allen PE. Arthroscopic ankle arthrodesis. J Bone Joint Surg Br. 2005;87(3):343-347. 12. Zwipp H, Rammelt S, Endres T, Heineck J. High union rates and function scores at midterm followup with ankle arthrodesis using a four screw technique. Clin Orthop Relat Res. 2010;468(4):958-968. doi:10.1007/s11999-009-1074-5.
Downloaded from fai.sagepub.com at UNIV OF PITTSBURGH on March 11, 2015
research-article2015
FAIXXX10.1177/1071100715571440Foot & Ankle InternationalHaskell et al
Article
Safe Zone for Placement of Talar Screws When Fusing the Ankle With an Anterior Plating System
Foot & Ankle International® 1–6 © The Author(s) 2015 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1071100715571440 fai.sagepub.com
Andrew Haskell, MD1,2, Russell Dedini, MD2, and Monara Dini, DPM3
Abstract Background: Ankle fusions fixed with anterior plates use fluoroscopic guidance to direct screws toward the subtalar joint. Special imaging views that visualize the subtalar joint are difficult to use and can be unreliable. This study evaluated whether a single lateral ankle view would provide adequate information to judge whether a screw penetrated the subtalar joint and identified strategies that would improve this technique. Methods: In 5 cadaveric ankles fixed with anterior plates, talar screws were placed up to the subtalar joint without penetration using lateral fluoroscopy to guide screw length. After dissection, the true distance from the screw tip to subchondral surface was measured. In addition, 4 readers measured the perceived distance from screw tip to subchondral surface using direct lateral, 10 degrees cephalad tilt lateral, and 10 degrees caudal tilt lateral fluoroscopic images on 2 separate occasions. Results: Nineteen (63%) of 30 screws penetrated the subchondral bone, and screw length determined using fluoroscopy was significantly longer than screw length measured directly (29.4 ± 5.5 mm vs 27.3 ± 8.5 mm, P = .014). Measurement of screw tip to bone distance demonstrated a high level of within-reader (kappa = .871, P < .001) and between-reader agreement (kappa = .807, P < .001), but poor specificity of determining screw penetration (0.50, χ2 = 22.1, P < .001) and poor correlation between radiographically measured and actual distances between screw tip and bone margin (r = .35, P < .001). Tilting the c-arm 10 degrees cephalad and directing screws toward the posterior facet improved the ability to detect screw penetration and directing screws toward the middle facet diminished it (P < .05). A safe zone of screw placement was defined by region. Conclusion: Use of a lateral fluoroscopic image to guide talar screw placement may lead to an unacceptably high rate of subtalar joint penetration. Clinical Relevance: Understanding the limitations of lateral fluoroscopy when using anterior ankle fusion plates may minimize screw penetration into the subtalar joint and diminish development of subtalar arthropathy. Keywords: ankle, fusion, arthrodesis, anterior, subtalar, screw length Tibiotalar arthrodesis remains a reliable procedure to treat ankle arthritis.7,11,12 Anteriorly placed, contoured locking plates have been used to maximize construct stability during ankle fusion.4,8 These plates are fixed with screws directed into the talus toward the subtalar joint. Since screw pull-out strength is proportional to the surface area of threads engaging bone, screws directed into the talus should be as long as possible.3 However, the desire to maximize screw length must be balanced with the risk of penetrating the subtalar joint articular surface to avoid premature subtalar joint arthropathy. Judging appropriate screw length in the talus typically is accomplished using intraoperative fluoroscopic imaging. The utility of fluoroscopy to prevent subtalar penetration during screw placement is limited by the complex anatomy of the inferior talus. Since the subtalar facets exist in
varying planes, multiple fluoroscopic views would be required to determine if any given screw had violated a facet. Specialized views have been described over the years to view select subtalar regions, including the Harris-Beath view for the middle facet,5 the Canale view for the talar neck,2 the Broden view for the posterior facet,1 and multiple 1
Department of Orthopedics, Palo Alto Medical Foundation, Palo Alto, CA, USA 2 San Francisco Department of Orthopedic Surgery, University of California, San Francisco, San Francisco, CA, USA 3 San Francisco General Hospital, Orthopedic Trauma Institute, San Francisco, CA, USA Corresponding Author: Andrew Haskell, MD, Department of Orthopedics, Palo Alto Medical Foundation, 795 El Camino Real, Palo Alto, CA 94301, USA. Email: [email protected]
Downloaded from fai.sagepub.com at UNIV OF PITTSBURGH on March 11, 2015
2
Foot & Ankle International
Broden view subtypes for the anterior and middle facets.1 However, obtaining numerous fluoroscopic views of all facets of the subtalar joint is difficult, time consuming, inaccurate, and may increase the patient’s exposure to ionizing radiation. The purpose of this cadaveric study was to determine if a single lateral ankle fluoroscopic view would provide adequate information to judge whether a screw has been placed without penetrating the subtalar joint. We hypothesized that when using anterior ankle arthrodesis plates, an intraoperative lateral fluoroscopic view of the ankle would not allow accurate assessment of screw position relative to the inferior subchondral bone of the talus. Two methods were used to test the hypothesis. First, we attempted to place screws up to but not through the subchondral bone using only a lateral image to judge screw length. Second, we tested the ability of surgeons to judge screw penetration of the subchondral bone based on saved fluoroscopic images generated during the first step. We also searched for criteria that would allow using a lateral fluoroscopic view to safely and accurately place screws directed toward the subtalar joint.
Materials and Methods Five intact fresh frozen cadaveric ankles were obtained, none of which exhibited signs of prior trauma or congenital deformity. For each specimen, a standard anterior ankle approach was performed, followed by application of anteromedial and anterolateral ankle arthrodesis plates (Tibiaxys Ankle Fusion System, Integra LifeSciences Corporation, Plainsboro, NJ, USA) with the ankle in the standard position of fusion.9 Each plate was precontoured for the left or right side and for the medial or lateral aspect of the ankle, and included 3 divergent fixed angle talus screws directed toward different regions of the subtalar joint. For each plate, the 3 locking screws placed into the talus initially were placed short. Each of the talar screws was then sequentially replaced with a fixed angle locking screw of length required to reach the subchondral bone as visualized on a lateral ankle view using a c-arm fluoroscopy unit (Philips BV Pulsera® Image Intensifier, Philips Corporation, Andover, MA, USA) and as perceived with agreement by 2 surgeons (RD, AH) (Figure 1). Since screws were available in 2 mm increments, all screws were placed with the goal of terminating within 2 mm of the subchondral bone margin without penetration. Three c-arm images per screw were then taken at prescribed angles (straight lateral, 10 degrees cephalad lateral, and 10 degrees caudad lateral) and digitally saved. A straight lateral image was acceptable if there was no overlap of the medial and lateral sides (double shadow) of the distal tibia articular surface and of the proximal talar articular surface, and if the fibula overlapped the tibia at or slightly posterior to the center of the distal tibia articular surface as measured from anterior to posterior. Cephalad lateral images were
Figure 1. A fluoroscopic lateral image of the ankle showing a talar screw directed toward the posterior facet is shown. The screw is placed as close to the inferior talar subchondral surface (black dotted line) as possible without penetration, as perceived on the image.
taken starting with a straight lateral image, then directing the fluoroscopic beam 10 degrees from inferior-lateral to superior-medial in the coronal plane. Caudad lateral images were taken starting with the straight lateral image, then directing the fluoroscopic beam 10 degrees from superiorlateral to inferior-medial in the coronal plane. In total, 90 images were generated. Each specimen was then dissected, the subtalar joint disarticulated, and the cartilage removed from the inferior surface of the talus. The distance from the tip of each previously placed screw to the inferior talar bone surface was measured at the inferior surface of the talus of each cadaveric specimen to the nearest 1 mm. This distance was defined as positive when the screw penetrated through the inferior talus into the subtalar joint or negative when the screw was entirely within the talus. The ideal screw length was calculated using the actual screw length and the screw tip to bone distance. The 90 digital images obtained were converted to JPEG format (Adobe Photoshop CS 6, Adobe Systems Corp, San Jose, CA, USA), randomly ordered, and distributed to 4 reviewers: 1 senior orthopedic surgery resident, 1 podiatrytrained surgeon, and 2 orthopedic foot and ankle fellowship trained surgeons. For each image, the reviewers were instructed to make a qualitative judgment about whether the talar screw depicted in the image had penetrated the inferior talar subchondral bone (Figure 2). In addition, the reviewers were asked to measure the distance (mm) between the screw tip and the talar subchondral bone margin (Figure 2). All images were viewed using Adobe Photoshop CS 6. Quantitative measures were made using the Adobe Photoshop ruler tool standardized to a known size marker on the image. All images were then randomly reordered, redistributed to each reviewer, and all measures were repeated at least 1 week later.
Downloaded from fai.sagepub.com at UNIV OF PITTSBURGH on March 11, 2015
3
Haskell et al Table 1. Amount of Screw Penetration Into the Subtalar Joint Based on Anatomic Region.
Mean ± Standard Deviation (mm) Number (+Subchondral Breach, 99% of –No Subchondral Confidence Screws Breach) Interval (mm) Posterior facet Sinus tarsi Middle facet Anterior facet
10 7 5 8
−1.4 ± 3.0 6.9 ± 5.1 1.0 ± 1.0 3.1 ± 2.2
−4.5 to 1.7 −0.4 to 14.1 −1.1 to 3.1 0.4 to 5.9
A logistic regression model was used to examine the influence of multiple variables on the readers’ ability to determine if a screw had penetrated the subchondral bone. The independent variables explored included the plate position (medial or lateral), the angle of incidence of the image intensifier beam (straight lateral, 10 degrees caudad lateral, 10 degrees cephalad lateral), the anatomic region of the subtalar joint into which the screw was directed as noted after disarticulating the specimens (posterior facet, sinus tarsi, middle facet, anterior facet), and the level of training of the reviewer (senior orthopaedic surgery resident, podiatry-trained, orthopaedic foot and ankle surgery fellowshiptrained). Finally, for each of 4 anatomic regions toward which a screw could be directed, a 99% confidence interval was calculated based on the measurement of screw tip to subchondral bone taken directly on the specimens. Significance was predetermined to be an alpha level of P < .05, and data are reported as mean ± standard deviation. Figure 2. (a) The method used to measure the distance from screw tip to inferior talar subchondral bone (black dotted line) is demonstrated. (b) Subjective “in” or “out” assessment of the screw tip position is made, and the distance X the screw tip is in or out of the bone is measured.
A repeated measures t-test was used to compare the screw lengths as measured during screw placement using fluoroscopy in the lateral plane with the ideal screw length calculated from the measurement of screw tip to subchondral bone taken directly from the specimen (AnalystSoft, StatPlus:mac, version 2009). Kappa statistics were calculated to determine agreement within and between each reader regarding measurement of the screw tip to subchondral bone distance as measured on the radiographs. The ability of readers to discriminate screw position as being either within the talus (in) or having penetrated the talar subchondral bone (out) was assessed using a chi-square test; sensitivity and specificity were calculated. The correlation between the readers’ radiographically measured and the directly measured distance from the screw tip to the bone margin was calculated using Pearson’s correlation coefficient.
Results Despite the stated goal of using the lateral fluoroscopy image to place screws within 2 mm of the talar subchondral bone without penetration, 19 (63%) of 30 screws were found to have penetrated the subchondral bone when the specimens were dissected. Screw length determined using lateral fluoroscopy during screw placement was significantly longer than screw length measured after dissection and direct visualization of the subtalar joint (29.4 ± 5.5 vs 27.3 ± 8.5 mm, P = .014). The length of screw tip penetration for each anatomic region of the subtalar joint is shown in Table 1. Measurement of screw-tip to bone distance demonstrated a high level of within-reader agreement (kappa = .871, P < .001) and between-reader agreement (kappa = .807, P < .001). While readers were subjectively able to distinguish subchondral bone penetration based on the lateral fluoroscopic image (χ2 = 22.1, P < .001), there was intermediate sensitivity (.81) and poor specificity (.50). Similarly, there was a significant but poor correlation between radiographically
Downloaded from fai.sagepub.com at UNIV OF PITTSBURGH on March 11, 2015
4
Foot & Ankle International
Figure 3. A scatter plot of the amount of subtalar screw penetration measured from lateral radiographs versus the true amount of screw penetration is shown. Positive values represent screw penetration into the subtalar joint. “Out” designates screws penetration into the subtalar joint; “in” designates no screw penetration. The linear regression best-fit line is shown (r = .35, P < .001).
measured and actual distances between screw tip and bone margin based on the lateral fluoroscopic image (r = .35, P < .001) (Figure 3). Three variables were identified through logistic regression analysis to have a clinically relevant influence on whether a given screw position was correctly assessed (P < .05) (Figure 4). Tilting the c-arm 10 degrees cephalad improved the ability to detect screw penetration of the inferior talar subchondral bone (Figure 5). Screws that were directed toward the posterior talar facet were more likely to be accurately categorized as penetrating or not penetrating the subchondral bone, while screws directed toward the middle facet were less likely to be accurately categorized. In contrast, the medial or lateral position of the plate was not shown to have an influence.
Discussion This study found that use of a lateral fluoroscopic image alone to guide talar screw placement may lead to an unacceptably high rate of subtalar joint screw penetration when trying to place the screw up to but not through the subchondral bone. Of screws, 63% inadvertently penetrated the inferior talar subchondral surface. Screws placed using the fluoroscopic method were, on average 2 mm longer than the ideal screw length. The specificity of determining screw penetration by fluoroscopic imaging was low, indicating frequent incorrect belief that a screw had not penetrated the inferior talar subchondral surface, when it actually had. The high intraobserver agreement in the ability to discern the presence and amount of screw penetration of the inferior talar articular surface over multiple readings a week apart demonstrates that the measurement method employed in the study was reproducible. The high interobserver agreement among multiple surgeons with varying levels and
types of training using the same technique suggests the findings of this study are generalizable. Factors were identified that may improve the ability of surgeons to identify screws that violate the subtalar joint using a simple fluoroscopic image. Changing the angle of incidence of the fluoroscopic X-ray beam 10 degrees cephalad improved the surgeons’ discriminatory ability. Furthermore, surgeons were better able to recognize subtalar joint violation in the posterior facet, and had greater difficulty in the middle facet. Recognizing the influence of anatomic region on the accuracy of assessing screw position may help surgeons avoid placing screws that penetrate the subchondral joint. A safe zone of screw placement was defined for each region of the subtalar joint by calculating the 99% confidence interval of the screw tip to subchondral bone distance measured directly on the specimens (Table 1). Of talar screws, 99% are expected to be within 2 mm of the subchondral bone without penetration if sufficient additional space is maintained between the screw tip and perceived subchondral bone on a lateral fluoroscopic image as follows: posterior facet, 1.7 mm; middle facet, 3.1 mm; and anterior facet, 5.9 mm. We have not identified other studies that report on the ability to perceive the position of the inferior talar subchondral surface based on a lateral radiographic image. However, other studies have confirmed the difficulty in interpreting radiographic views of the talus, calcaneus and subtalar joint. Broden’s radiographic views of the subtalar joint showed poor correlation (kappa = .23) with CT scan for classifying posterior facet involvement in 45 intraarticular calcaneal fractures.6 A cadaveric study of angulation in talar neck fractures found a single Canale view may not provide adequate information when compared with multiple views.10 There are several limitations to this study. First, radiographic imaging does not supplant tactile feedback and use of a depth gauge when deciding whether any given screw is safely placed. However, tactile feedback may rely on feeling the diminution of counterforce experience as the subchondral bone is perforated, leading to articular cartilage damage. In addition, the study did not assess the efficacy of Broden’s views. This was by design, since Broden’s views would be difficult and time consuming to obtain for each screw placed, and they have shown poor correlation with CT scan when evaluating the subtalar joint.6 Finally, while this study offers guidelines for a safe zone of screw placement for each anatomic region of the subtalar joint, this is based on post hoc evaluation using a 99% confidence interval of screw penetration. Future study would be needed to validate these guidelines. In conclusion, we found that using a lateral view of the ankle as the only guide to placement of screws into the talus during ankle arthrodesis using anterior plates did not have a high degree of accuracy. Nevertheless, a lateral view may
Downloaded from fai.sagepub.com at UNIV OF PITTSBURGH on March 11, 2015
5
Haskell et al
Figure 4. The effect of a variety of factors on the probability of correct assessment of a screw tip having penetrated the talar subchondral bone is displayed. Data shown are based on logistic regression, where values closer to 1 demonstrate better ability to predict screw penetration and values closer to 0 demonstrate the inability to predict screw penetration. All categories whose probability error bars do not cross 0.5 have a statistically significant effect (P < .05). *Cephalad radiographic tilt and screws directed toward the posterior facet had a clinically relevant positive effect on the ability to assess screw tip penetration into the subtalar joint, † while screws direct toward the middle facet had a clinically relevant negative effect.
be used to judge, with reasonable confidence, that a screw is safely out of the subtalar joint if a margin for error (posterior facet, 1.7 mm; middle facet 3.1 mm; anterior facet, 5.9 mm) is observed based on the region of the subtalar joint toward which the screw is directed. Declaration of Conflicting Interests The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Integra, Inc provided the plates and screws used in the study.
Funding The author(s) received no financial support for the research, authorship, and/or publication of this article.
References Figure 5. The effect of X-ray beam tilt on the ability to determine screw tip position is demonstrated. These images show a single specimen with a long screw directed toward the posterior facet that was found, after disarticulation, to be 2 mm short of the inferior subchondral surface of the talus (completely within bone). A direct lateral image (a) gives the appearance that the screws abuts the subchondral surface. (b) Directing the X-ray beam 10 degrees caudally gives the appearance that the screw has penetrated the subchondral surface. (c) Tilting the beam 10 degrees cephalad provides the best representation of the true screw tip position above the subchondral surface.
1. Broden B. Roentgen examination of the subtaloid joint in fractures of the calcaneus. Acta Radiologica. 1949;31(1): 85-91. 2. Canale ST, Kelly FB Jr. Fractures of the neck of the talus. Long-term evaluation of seventy-one cases. J Bone Joint Surg Am. 1978;60(2):143-156. 3. Chapman JR, Harrington RM, Lee KM, et al. Factors affecting the pullout strength of cancellous bone screws. J Biomech Eng. 1996;118(3):391-398. 4. Guo C, Yan Z, Barfield WR, Hartsock LA. Ankle arthrodesis using anatomically contoured anterior plate. Foot Ankle Int. 2010;31(6):492-498. doi:10.3113/FAI.2010.0492.
Downloaded from fai.sagepub.com at UNIV OF PITTSBURGH on March 11, 2015
6
Foot & Ankle International
5. Harris RI, Beath T. Etiology of peroneal spastic flat foot. J Bone Joint Surg Br. 1948;30B(4):624-634. 6. Kwon DG, Chung CY, Lee KM, et al. Revisit of Broden’s view for intraarticular calcaneal fracture. Clin Orthop Surg. 2012;4(3):221-226. doi:10.4055/cios.2012.4.3.221. 7. Mann RA, Rongstad KM. Arthrodesis of the ankle: a critical analysis. Foot Ankle Int. 1998;19(1):3-9. 8. Plaass C, Knupp M, Barg A, Hintermann B. Anterior double plating for rigid fixation of isolated tibiotalar arthrodesis. Foot Ankle Int. 2009;30(7):631-639. doi:10.3113/ FAI.2009.0631.
9. Scranton PE Jr. An overview of ankle arthrodesis. Clin Orthop Relat Res. 1991;(268):96-101. 10. Thomas JL, Boyce BM. Radiographic analysis of the Canale view for displaced talar neck fractures. J Foot Ankle Surg. 2012;51(2):187-190. doi:10.1053/j.jfas.2011.10.037. 11. Winson IG, Robinson DE, Allen PE. Arthroscopic ankle arthrodesis. J Bone Joint Surg Br. 2005;87(3):343-347. 12. Zwipp H, Rammelt S, Endres T, Heineck J. High union rates and function scores at midterm followup with ankle arthrodesis using a four screw technique. Clin Orthop Relat Res. 2010;468(4):958-968. doi:10.1007/s11999-009-1074-5.
Downloaded from fai.sagepub.com at UNIV OF PITTSBURGH on March 11, 2015
