Competence of the Deltoid Ligament in Bimalleolar Ankle Fractures After Medial Malleolar Fixation * BY PAUL TORNETTA, III, M.D. Investigation performed at Kings County Hospital, New York, N.Y. Abstract Background: The stability of the ankle joint is provided by the medial and lateral malleoli and ligaments. Recent studies of cadaveric ankles have demonstrated that injury to the medial structures of the ankle is necessary to allow lateral subluxation of the talus after fracture. However, cadaveric models are limited by the fracture pattern chosen for the model. We sought to investigate the competency of the deltoid ligament in vivo in patients with an operatively treated bimalleolar ankle fracture. Methods: Twenty-seven patients with a bimalleolar ankle fracture were evaluated. In each patient, the medial malleolus was anatomically reduced and fixed. A radiograph of the ankle was then made with application of an external rotation load to the joint. All lateral malleolar injuries were then reduced and fixed. The radiographs were evaluated for restoration of the competence of the deltoid ligament according to established criteria. Results: Seven (26 percent) of the twenty-seven patients had radiographically evident incompetence of the deltoid ligament after medial malleolar fixation. This finding was associated with a small medial malleolar fragment. Conclusions: In bimalleolar fractures, the medial injury may be an osseous avulsion, leaving the deltoid intact on the displaced fragment, or it may be a combination of ligamentous and osseous injury with disruption of the deep portion of the deltoid ligament. Many authors have demonstrated that an injury to the medial structures of the ankle must be present for talar subluxation to occur following a lateral malleolar fracture 1,5,6,11,14. Biomechanical studies of cadaveric ankles have supported the concept that, in patients who have a bimalleolar ankle fracture, anatomical fixation of the fracture of the medial malleolus restores medial support to the ankle mortise 1,9-11. However, these cadaveric studies may be misleading as the technique of simulating ankle *No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study. Boston Medical Center, Dowling 2 North, 818 Harrison Avenue, Boston, Massachusetts 02118. E-mail address: ptornetta@pol.net. Copyright 2000 by The Journal of Bone and Joint Surgery, Incorporated fractures ensures that the deltoid ligament remains intact. There is little clinical data available regarding the integrity of the deltoid ligament in conjunction with medial malleolar fractures. Burwell and Charnley 2, Close 3, Coonrad 4, and Mendelsohn 8 all reported that concomitant injury of the medial malleolus and the deltoid ligament may occur. It was not until Pankovich and Shivaram reported a detailed anatomical description of the origin of the superficial and deep components of the deltoid ligament that an understanding of the various patterns of medial injuries was possible 13,14. Despite these findings, the recommendation to treat syndesmotic injuries differently if the medial injury is osseous rather than ligamentous has been made 1. The purpose of this study was to assess, in vivo, the competence of the deltoid ligament following fixation of the medial malleolus in patients with a displaced bimalleolar ankle fracture and preoperative lateral talar displacement. Materials and Methods Twenty-eight consecutive patients with a bimalleolar ankle fracture treated between April 1996 and April 1997 at Kings County Hospital in New York City were evaluated. All fractures were displaced a minimum of three millimeters and were judged to be best treated by open reduction and internal fixation. None of the patients had a history of previous ankle injury. The average age of the patients was thirty-six years (range, eighteen to sixtyeight years). There were twenty-one supination-external rotation, four pronation-abduction, and three pronationexternal rotation injuries as defined with the Lauge- Hansen scheme 7. According to the Orthopaedic Trauma Association classification 12, nineteen fractures were classified as 44-B2; two, as 44-B3; five, as 44-C1.2; and two, as 44-C2.2. In each patient, the medial malleolus was reduced anatomically and fixed with 4.0-millimeter partially threaded lag screws as the first stage of the procedure. Two screws were used in twenty-six patients, and one screw and one Kirschner wire were used in two patients. The reduction and fixation was performed without stripping the attachment of either the superficial or the deep portion of the deltoid ligament from the medial malleolus. As care was taken to limit the medial dissection to the fracture site, direct examination of the deep portion of the deltoid ligament was not performed. As VOL. 82-A, NO. 6, JUNE 2000 843
844 P. TORNETTA, III TABLE I VARIABLES THAT DIFFERED SIGNIFICANTLY ACCORDING TO WHETHER THE STRESS RADIOGRAPH WAS POSITIVE OR NEGATIVE Variable Positive Radiograph* Negative Radiograph* P Value Medial malleolar width 1.8 ± 0.6 cm 2.6 ± 0.5 cm 0.003 Medial malleolar height 1.3 ± 0.2 cm 1.7 ± 0.3 cm 0.009 Medial clear space 6.0 ± 1.7 mm 3.4 ± 0.7 mm 0.0001 Talar subluxation 3.2 ± 2.0 mm 0 0.0001 *The values are expressed as the average and the standard deviation. part of the standard management of these patients, the medial malleolus was fixed first and the fixation was then assessed on an anteroposterior radiograph of the ankle made with an external rotation load applied to the joint. This radiographic examination was performed by lowering the operating table to its lowest point and elevating the portable radiography machine to its highest point. We did not try to quantitate the magnitude of the load, but an estimated load of approximately twenty pounds (9.1 kilograms) was used for each ankle. The ankle was held in neutral dorsiflexion-plantar flexion during loading. Regardless of the findings on the stress radiograph, the lateral side of the ankle was then fixed anatomically with use of lag screws only (nine ankles), a lateral one-third tubular plate (nine), an antiglide plate (nine), or only syndesmotic fixation (one). Five patients had placement of a syndesmotic screw in addition to the lateral malleolar plate because of a finding of clinical instability at the syndesmosis under direct vision. No stress radiographs were made after the lateral fixation. The ankle was then placed in a short leg cast. The preoperative radiographs were assessed for talar subluxation and the height and width of the medial malleolar fragment. The height of the fragment was defined as the maximal size of the fragment measured perpendicular to the fracture line on the lateral radiograph. The width was defined as the maximal anteroposterior length of the fracture on the lateral radiograph. The medial clear space was measured perpendicular to the inner edge of the medial malleolus to the talus, and talar subluxation was determined on the basis of the relationship of the lateral edge of the talar dome to the lateral FIG. 1-A FIG. 1-B Figs. 1-A through 1-E: A fracture-dislocation of the ankle in a twenty-eight-year-old man. Figs. 1-A and 1-B: Anteroposterior and lateral radiographs of the injury. THE JOURNAL OF BONE AND JOINT SURGERY
COMPETENCE OF THE DELTOID LIGAMENT IN BIMALLEOLAR ANKLE FRACTURES 845 Results One medial malleolar fracture in a sixty-four-yearold woman with a Lauge-Hansen 7 type-4 supinationexternal rotation fracture displaced two millimeters when stressed, and after repeat reduction and retightening of the screws the stress radiograph was not repeated. This patient was omitted from the analysis of the results. Of the twenty-seven remaining stress radiographs, seven (26 percent) were positive, indicating deltoid incompetence in conjunction with the medial malleolar fracture. These ankles demonstrated a widened medial clear space or talar subluxation, or both. The average size (and standard deviation) of the medial clear space in these seven ankles was 6.0 ± 1.7 millimeters (range, four to 8.5 millimeters), and the average amount of talar subluxation was 3.2 ± 2.0 millimeters (range, one to seven millimeters). The other twenty fractures had a stable mortise after medial malleolar fixation. These twenty ankles had a normal medial clear space with the talus con- FIG. 1-C The medial malleolar fragment, which is outlined, is of the supracollicular type. It measured 1.9 centimeters in height and 3.4 centimeters in width. edge of the plafond. The superior and medial clear spaces and talar subluxation were measured on the stress radiograph and on the final postoperative radiographs. The author made all measurements. The findings on the stress radiograph were considered positive if the medial clear space was more than four millimeters in width and at least one millimeter greater than the superior clear space or if there was talar subluxation of more than one millimeter 9 (Fig. 2-C). The fractures for which the stress radiograph was positive were compared (with regard to the above data points) with those for which it was negative with use of a paired t test with adjustment for small numbers. P < 0.05 was considered significant. FIG. 1-D External rotation stress radiograph made after anatomical reduction and fixation of the medial malleolar fracture, revealing a stable mortise. FIG. 1-E Line drawing demonstrating schematically what has occurred in this injury. The medial injury is through the bone from the anterior tibial cortex to the posterior tibial cortex, leaving the entire deep portion of the deltoid ligament intact on the displaced fragment. VOL. 82-A, NO. 6, JUNE 2000
846 P. TORNETTA, III FIG. 2-A FIG. 2-B Figs. 2-A through 2-D: A displaced ankle fracture in a thirty-four-year-old man. Figs. 2-A and 2-B: Anteroposterior and lateral radiographs of the injury. As seen on the lateral radiograph, the fracture of the medial malleolus extends only through the anterior colliculus (outlined by dots). gruent with the tibial plafond on the stress radiograph. Comparison of the fractures with a positive stress radiograph with those with a negative stress radiograph revealed significant differences (p < 0.05) in the medial malleolar height (1.3 ± 0.2 compared with 1.7 ± 0.3 centimeters) and width (1.8 ± 0.6 compared with 2.6 ± 0.5 centimeters) (Table I). There were no significant differences between the groups with regard to patient age, fracture type according to the Lauge-Hansen 7 or Orthopaedic Trauma Association scheme 12, or the amount of medial malleolar, fibular, or talar displacement on the preoperative radiographs. Because anatomical reduction and fixation of the lateral malleolus was performed in all patients, the medial clear space and the talar reduction in the mortise were restored on the final postoperative radiographs of all patients. Discussion The patterns of instability demonstrated in this study are attributable to the complex anatomy of the deltoid ligament. The deep portion of the ligament takes its origin from the posterior colliculus and variably from the intercollicular groove, and it inserts into the talar body. The deep component is the strongest portion of the ligament, and it functions to resist external rotation of the ankle when the foot is dorsiflexed 13. In contradistinction to the strong deep component, the superficial portion of the deltoid ligament takes its origin from the anterior colliculus and has a broad insertion on the talus and navicular. This portion of the ligament is not thick, is the weaker of the two portions, and is under tension during external rotation of the ankle when the foot is in plantar flexion. Thus, fixation of small fractures of the medial malleolus, to which only the superficial portion of the ligament attaches, may not be sufficient to restore medial stability. FIG. 2-C External rotation stress radiograph made after anatomical reduction and fixation of the medial malleolar fracture, revealing talar subluxation with widening of the medial clear space. THE JOURNAL OF BONE AND JOINT SURGERY
COMPETENCE OF THE DELTOID LIGAMENT IN BIMALLEOLAR ANKLE FRACTURES 847 FIG. 2-D Line drawing depicting the injury as occurring partly through the bone, anteriorly, and continuing through the deep portion of the deltoid ligament, posteriorly. Because the deep portion of the deltoid ligament is torn, fixation of the small anterior fragment does not restore medial stability. In contradistinction to the results and conclusions of cadaveric sectioning studies, other studies have shown that the deltoid ligament may be injured in conjunction with a medial malleolar fracture 2-4,8,13. Despite a precise description by Pankovich and Shivaram 13,14, in 1979, of the patterns of injury that may occur on the medial side of the ankle, little attention has been given to this issue. In the present study, only bimalleolar fractures with talar subluxation were included. This inclusion criterion eliminated four of the six types of medial injury described by Pankovich and Shivaram. In their series, fifteen patients had an isolated fracture of the anterior colliculus. The deltoid was intact in nine of those patients and disrupted in six. They reported on an additional fifty-four patients who had a supracollicular fracture but no disruption of the deltoid; thus, they reported a 10 percent prevalence of injuries in which the medial malleolus was fractured in conjunction with a disruption of the deltoid ligament. However, the competence of the deltoid ligament was not specifically tested. In the current series, the prevalence of deltoid incompetence was 26 percent. This finding may be due, in part, to the fact that only patients with bimalleolar fracture and talar subluxation were included. The size of the medial malleolar fragment was the most important variable in predicting deltoid competence. The deltoid was competent in association with all fractures in which the medial malleolar fragment was more than 2.8 centimeters wide on the lateral radiograph, and it was incompetent in all in which it was less than 1.7 centimeters wide. The former fracture would be classified as supracollicular according to the system of Pankovich and Shivaram 14. The latter fracture includes only the anterior colliculus. In several patients, the fracture entered the intercollicular groove and stability was not predictable. Thus, in all bimalleolar fractures with talar subluxation, the entire medial support of the ankle is injured, but the injury may be a combination of osseous and ligamentous disruption. In the patients with a supracollicular fracture in the present study, the deltoid was always intact and the mortise was always stable after medial fixation (Figs. 1-A through 1-E). In contrast, when the medial malleolar fragment was small, the anterior colliculus, and in some cases a portion of the intercollicular groove, was displaced and the deep portion of the deltoid ligament was likely disrupted posteriorly (Figs. 2-A through 2-D). This pattern of injury resulted in deltoid incompetence even after medial fixation. The present study demonstrated that there is an important difference between in vitro cadaveric studies and the true in vivo clinical situation. Recommendations regarding the treatment of syndesmotic injuries based on cadaveric studies in which the medial injury was either osseous or ligamentous may be flawed. These recommendations apply only to the situation in which the medial support is restored by anatomical reduction and internal fixation of the medial malleolus. The current findings demonstrated that a stress radiograph will confirm stability, which is most likely present if the fracture of the medial malleolus is of the supracollicular type. References 1. Boden, S. D.; Labropoulos, P. A.; McCowin, P.; Lestini, W. F.; and Hurwitz, S. R.: Mechanical considerations for the syndesmosis screw. A cadaver study. J. Bone and Joint Surg., 71-A: 1548-1555, Dec. 1989. 2. Burwell, H. N., and Charnley, A. D.: The treatment of displaced fractures at the ankle by rigid internal fixation and early joint movement. J. Bone and Joint Surg., 47-B(4): 634-660, 1965. 3. Close, J. R.: Some applications of the functional anatomy of the ankle joint. J. Bone and Joint Surg., 38-A: 761-781, July 1956. 4. Coonrad, R. W.: Fracture-dislocations of the ankle joint with impaction injury of the lateral weight-bearing surface of the tibia. J. Bone and Joint Surg., 52-A: 1337-1344, Oct. 1970. 5. Grath, G.-B.: Widening of the ankle mortise. A clinical and experimental study. Acta Chir. Scandinavica, Supplementum 263, 1960. 6. Harper, M. C.: Deltoid ligament: an anatomical evaluation of function. Foot and Ankle, 8: 19-22, 1987. 7. Lauge-Hansen, N.: Ligamentous ankle fractures. Diagnosis and treatment. Acta Chir. Scandinavica, 97: 544-550, 1949. 8. Mendelsohn, H. A.: Nonunion of malleolar fractures of the ankle. Clin. Orthop., 42: 103-118, 1965. 9. Michelson, J. D.; Clarke, H. J.; and Jinnah, R. H.: The effect of loading on tibiotalar alignment in cadaver ankles. Foot and Ankle Internat., 10: 280-284, 1990. VOL. 82-A, NO. 6, JUNE 2000
848 P. TORNETTA, III 10. Michelsen, J. D.; Ahn, V. M.; and Helgemo, S. L.: Motion of the ankle in a simulated supination-external rotation fracture model. J. Bone and Joint Surg., 78-A: 1024-1031, July 1996. 11. Michelson, J. D., and Waldman, B.: An axially loaded model of the ankle after pronation external rotation injury. Clin. Orthop., 328: 285-293, 1996. 12. Orthopaedic Trauma Association, Committee for Coding and Classification: Fracture and dislocation compendium. J. Orthop. Trauma, 10 (Supplement 1), 1996. 13. Pankovich, A. M., and Shivaram, M. S.: Anatomical basis of variability in injuries of the medial malleolus and the deltoid ligament. I. Anatomical studies. Acta Orthop. Scandinavica, 50: 217-223, 1979. 14. Pankovich, A. M., and Shivaram, M. S.: Anatomical basis of variability in injuries of the medial malleolus and the deltoid ligament. II. Clinical studies. Acta Orthop. Scandinavica, 50: 225-236, 1979. THE JOURNAL OF BONE AND JOINT SURGERY