International Orthopaedics (SICOT) DOI 10.1007/s00264-015-2835-2
ORIGINAL PAPER
Which AO/OTA 31-A2 pertrochanteric fractures can be treated with a dynamic hip screw without developing a lateral wall fracture? A CT-based study Gaurav Sharma 1 & Ravijot Singh 1 & Kiran Kumar GN 1 & Vaibhav Jain 1 & Ankit Gupta 1 & Shivanand Gamanagatti 2 & Kamran Farooque 1 & Vijay Sharma 1
Received: 13 May 2015 / Accepted: 21 May 2015 # SICOT aisbl 2015
Abstract Purpose To determine whether radiographic measurements derived from standard computed tomography (CT) evaluation can be used to predict likelihood of a peri-operative lateral femoral wall fracture in AO/OTA 31-A2 pertrochanteric fractures treated with a dynamic hip screw (DHS). Methods Fifty-one patients with AO/OTA 31-A2 classified pertrochanteric fractures were evaluated using a pre-operative CT scan of the pelvis with both hips. Dimensions of the lateral wall were calculated for each patient using four parameters: (1) height of the lateral wall above the vastus ridge; (2) circumference of the lateral wall 2 cm below the vastus ridge at an angle of 135°; this circumference was further divided into an anterior, lateral and posterior component; (3) cortical thickness at the centre of the lateral component of the lateral wall; and (4) cortical index. All patients were treated with a 135° DHS. Postoperative radiographs were assessed for lateral femoral wall fracture. Results Patients with a lateral wall fracture (17/51) had a smaller circumference (4.47 cm vs 5.44 cm p value 2.10 cm in circumference are not likely to sustain a lateral wall fracture when treated with a DHS. Keywords Pertrochanteric . Computed tomography . Lateral wall . Dynamic hip screw
Introduction Pertrochanteric fractures of the proximal part of the femur have been treated successfully with use of the dynamic hip screw (DHS) over the last 30 years [1]. However one of the complications of using the DHS for AO/OTA 31 A-2 fractures is a peri-operative lateral wall fracture [2]. This converts an A-2 pertrochanteric fracture into a A-3 intertrochanteric fracture leading to a deterioration in the pattern of the fracture, and has been referred to as the BPantrochanteric fracture^ [2–4]. This iatrogenic complication has been associated with a prolonged healing process [2] as well as re-operations [4]. A thin, osteoporotic lateral wall, the use of a large-bore reamer and incomplete reduction on the lateral radiograph are all felt to contribute to the genesis of this fracture [3]. There is also a higher risk of lateral wall fracture in low intertrochanteric fractures with the proximal end of fracture exiting near the vastus ridge as only a narrow bridge of bone is left proximal to the lag screw hole [5]. Finally, anterior starting of the lag screw, compared to posterior starting, has
International Orthopaedics (SICOT)
been shown to be associated with a higher incidence of lateral wall fracture in a biomechanical model [6]. Although the adverse effects of a lateral wall fracture are well described [2, 4, 7], few objective predictors of such a fracture have been studied [7]. If accurate objective criteria could identify patients at risk for such a lateral wall fracture, then these patients could be treated with either a trochanteric nail or DHS with a trochanter stabilizing plate, as these implants prevent femoral medialization even if a lateral wall fracture occurs [8, 9], while patients with a Bsafe^ lateral wall could be treated with a DHS alone, at a much lower cost [1, 10]. We conducted a CT-based study in an attempt to determine objective predictors of a lateral wall fracture, as well as to better understand the patho-anatomy of a lateral wall fracture. Although CT has been used in intertrochanteric fractures for testing the reliability of classification systems [11], we are not aware of any study which has utilized CT for measuring dimensions of lateral wall and objectively define the characteristics of a Bthin lateral wall^, i.e. one that has a higher chance of developing an iatrogenic fracture following DHS. The aims and objectives of this study were to (1) determine the incidence of lateral wall fracture in AO/OTA 31-A2 fractures after DHS in a prospective cohort, (2) objectively calculate the dimensions of the lateral femoral wall using preoperative CT, (3) compare the dimensions for the group without lateral wall fracture (group I) with the group with lateral wall fracture (group II), and (4) determine whether a prediction can be made based on preoperative CT measurements as to which AO/OTA 31-A2 fractures have a higher chance of lateral wall fracture.
Materials and methods A prospective study was undertaken at our level 1 trauma centre after obtaining clearance from the institutional review board. All patients with an AO 31-A2 pertrochanteric fracture, and more than 18 years of age, who presented to the casualty between November 2013 and May 2014 were included in the study. Pathologic fractures, periprosthetic fractures, and fractures which already had a pre-operative lateral wall fracture (A3) were excluded. Fifty-five patients were enrolled in the study and underwent DHS fixation. Two patients died within one week postoperatively during their stay in the hospital. Two patients were lost to follow-up before the stipulated follow up period. Fifty-one patients were therefore included in the analysis. CT evaluation A preoperative CT scan of the pelvis with both hips was obtained for all patients using a 40-slice CT scanner (Somatom Sensation; Siemens, Germany). A slice thickness of 1.5 mm was used and reconstructions were done at 1-mm
intervals. Measurements were made at window level of 1,500 and centring at 450 (bone window). On the CT scan DICOM viewer, the axial, coronal and sagittal views were on display on the screen together at one time. The following sequence was used for measuring the dimensions of the lateral wall (Fig. 1). First, on the coronal view, a horizontal line was drawn at the level of the vastus ridge (line 1) (Fig. 1b). The first dimension measured was the height of the lateral wall above the vastus ridge, measured proximal to line 1. Then a second line (line 2) measuring 2 cm was dropped vertically from line 1 (level of vastus ridge) onto the lateral cortex. At the point of contact of line 2 with the lateral cortex (entry point of the DHS guide wire) [12], a third line (line 3) was drawn at an angle of 135°, representing the trajectory of the guide wire (Fig. 1b). The circumference of the lateral wall along this trajectory was measured on the corresponding axial cut (Fig. 1c). This was the second dimension measured and represented the circumference available for the triple reamer. Based on the geometry of the proximal femur at this level, this circumference was further divided into an anterior component, lateral component, and posterior component. The DHS guidewire/ reamer would typically start through the lateral component of the lateral wall. While the anterior component could be easily demarcated from the lateral component as they lay at right angles to each other, the demarcation between the posterior component and the lateral component was sometimes less evident due to a more gentle bend (Fig. 1c). In case the demarcation between the posterior and lateral components was unclear, the dimension of the lateral component was determined from the contralateral uninjured side as the distance between the anterior and posterior surfaces of the neck (Fig. 2). The next measurement included the cortical thickness measured at the center of the lateral component of the lateral wall (Fig. 1c). Finally, a line measuring 7 cm was dropped vertically from the level of the vastus ridge (line 4). The cortical index (CI) was measured at this level (7 cm below the vastus ridge) (Fig. 1b). The cortical thickness as well as CI was measured to represent the degree of osteoporosis. Figure 3 shows the measurements on 3D reconstruction for better clarity. Two experienced musculoskeletal radiologists (including one of the authors, SG) who were blinded to the clinical course of the patient performed all measurements independently. The mean value of the two sets of measurements was used for statistical analyses. All patients were treated with a 135° DHS (Synthes, GmbH, Oberdorf, Switzerland) after obtaining a good reduction and correction of the posterior sag. A good reduction was defined as normal or slight valgus reduction, less than 20° of angulation on the lateral radiograph, and less than 4 mm of fracture displacement [13, 14]. The guide wire was inserted using the angled guide which was seated flush to the lateral cortex. Minimal force was
International Orthopaedics (SICOT)
Fig. 1 Screen shot showing sagittal (a), coronal (b) and axial (c) CT images of a patient with AO/OTA 31 A2 fracture. A horizontal line was drawn at the level of the vastus ridge on the coronal view (line 1). The maximum height above the vastus ridge (yellow line) was measured on the coronal view proximal to line 1 (b). Then a line was drawn 2 cm below the vastus ridge (line 2). At the point of contact of line 2 with the lateral cortex, a third line was drawn at an angle of 135° (line 3).
Circumference of the lateral wall was measured on the axial section corresponding to line 3 (c). This circumference was further divided into an anterior (red line), lateral (brown line), and posterior component (green line). Cortical thickness was measured at the centre of the lateral component of the lateral wall (small black line). Cortical index was measured 7 cm below the vastus ridge (line 4) on the coronal view (curved black arrow)
applied while drilling the lateral wall with the triple reamer to ensure a smooth transition between the portions of the triple reamer. The head screw was placed in keeping with Baumgartner’s suggestion that tip apex distance (TAD) should be
ORIGINAL PAPER
Which AO/OTA 31-A2 pertrochanteric fractures can be treated with a dynamic hip screw without developing a lateral wall fracture? A CT-based study Gaurav Sharma 1 & Ravijot Singh 1 & Kiran Kumar GN 1 & Vaibhav Jain 1 & Ankit Gupta 1 & Shivanand Gamanagatti 2 & Kamran Farooque 1 & Vijay Sharma 1
Received: 13 May 2015 / Accepted: 21 May 2015 # SICOT aisbl 2015
Abstract Purpose To determine whether radiographic measurements derived from standard computed tomography (CT) evaluation can be used to predict likelihood of a peri-operative lateral femoral wall fracture in AO/OTA 31-A2 pertrochanteric fractures treated with a dynamic hip screw (DHS). Methods Fifty-one patients with AO/OTA 31-A2 classified pertrochanteric fractures were evaluated using a pre-operative CT scan of the pelvis with both hips. Dimensions of the lateral wall were calculated for each patient using four parameters: (1) height of the lateral wall above the vastus ridge; (2) circumference of the lateral wall 2 cm below the vastus ridge at an angle of 135°; this circumference was further divided into an anterior, lateral and posterior component; (3) cortical thickness at the centre of the lateral component of the lateral wall; and (4) cortical index. All patients were treated with a 135° DHS. Postoperative radiographs were assessed for lateral femoral wall fracture. Results Patients with a lateral wall fracture (17/51) had a smaller circumference (4.47 cm vs 5.44 cm p value 2.10 cm in circumference are not likely to sustain a lateral wall fracture when treated with a DHS. Keywords Pertrochanteric . Computed tomography . Lateral wall . Dynamic hip screw
Introduction Pertrochanteric fractures of the proximal part of the femur have been treated successfully with use of the dynamic hip screw (DHS) over the last 30 years [1]. However one of the complications of using the DHS for AO/OTA 31 A-2 fractures is a peri-operative lateral wall fracture [2]. This converts an A-2 pertrochanteric fracture into a A-3 intertrochanteric fracture leading to a deterioration in the pattern of the fracture, and has been referred to as the BPantrochanteric fracture^ [2–4]. This iatrogenic complication has been associated with a prolonged healing process [2] as well as re-operations [4]. A thin, osteoporotic lateral wall, the use of a large-bore reamer and incomplete reduction on the lateral radiograph are all felt to contribute to the genesis of this fracture [3]. There is also a higher risk of lateral wall fracture in low intertrochanteric fractures with the proximal end of fracture exiting near the vastus ridge as only a narrow bridge of bone is left proximal to the lag screw hole [5]. Finally, anterior starting of the lag screw, compared to posterior starting, has
International Orthopaedics (SICOT)
been shown to be associated with a higher incidence of lateral wall fracture in a biomechanical model [6]. Although the adverse effects of a lateral wall fracture are well described [2, 4, 7], few objective predictors of such a fracture have been studied [7]. If accurate objective criteria could identify patients at risk for such a lateral wall fracture, then these patients could be treated with either a trochanteric nail or DHS with a trochanter stabilizing plate, as these implants prevent femoral medialization even if a lateral wall fracture occurs [8, 9], while patients with a Bsafe^ lateral wall could be treated with a DHS alone, at a much lower cost [1, 10]. We conducted a CT-based study in an attempt to determine objective predictors of a lateral wall fracture, as well as to better understand the patho-anatomy of a lateral wall fracture. Although CT has been used in intertrochanteric fractures for testing the reliability of classification systems [11], we are not aware of any study which has utilized CT for measuring dimensions of lateral wall and objectively define the characteristics of a Bthin lateral wall^, i.e. one that has a higher chance of developing an iatrogenic fracture following DHS. The aims and objectives of this study were to (1) determine the incidence of lateral wall fracture in AO/OTA 31-A2 fractures after DHS in a prospective cohort, (2) objectively calculate the dimensions of the lateral femoral wall using preoperative CT, (3) compare the dimensions for the group without lateral wall fracture (group I) with the group with lateral wall fracture (group II), and (4) determine whether a prediction can be made based on preoperative CT measurements as to which AO/OTA 31-A2 fractures have a higher chance of lateral wall fracture.
Materials and methods A prospective study was undertaken at our level 1 trauma centre after obtaining clearance from the institutional review board. All patients with an AO 31-A2 pertrochanteric fracture, and more than 18 years of age, who presented to the casualty between November 2013 and May 2014 were included in the study. Pathologic fractures, periprosthetic fractures, and fractures which already had a pre-operative lateral wall fracture (A3) were excluded. Fifty-five patients were enrolled in the study and underwent DHS fixation. Two patients died within one week postoperatively during their stay in the hospital. Two patients were lost to follow-up before the stipulated follow up period. Fifty-one patients were therefore included in the analysis. CT evaluation A preoperative CT scan of the pelvis with both hips was obtained for all patients using a 40-slice CT scanner (Somatom Sensation; Siemens, Germany). A slice thickness of 1.5 mm was used and reconstructions were done at 1-mm
intervals. Measurements were made at window level of 1,500 and centring at 450 (bone window). On the CT scan DICOM viewer, the axial, coronal and sagittal views were on display on the screen together at one time. The following sequence was used for measuring the dimensions of the lateral wall (Fig. 1). First, on the coronal view, a horizontal line was drawn at the level of the vastus ridge (line 1) (Fig. 1b). The first dimension measured was the height of the lateral wall above the vastus ridge, measured proximal to line 1. Then a second line (line 2) measuring 2 cm was dropped vertically from line 1 (level of vastus ridge) onto the lateral cortex. At the point of contact of line 2 with the lateral cortex (entry point of the DHS guide wire) [12], a third line (line 3) was drawn at an angle of 135°, representing the trajectory of the guide wire (Fig. 1b). The circumference of the lateral wall along this trajectory was measured on the corresponding axial cut (Fig. 1c). This was the second dimension measured and represented the circumference available for the triple reamer. Based on the geometry of the proximal femur at this level, this circumference was further divided into an anterior component, lateral component, and posterior component. The DHS guidewire/ reamer would typically start through the lateral component of the lateral wall. While the anterior component could be easily demarcated from the lateral component as they lay at right angles to each other, the demarcation between the posterior component and the lateral component was sometimes less evident due to a more gentle bend (Fig. 1c). In case the demarcation between the posterior and lateral components was unclear, the dimension of the lateral component was determined from the contralateral uninjured side as the distance between the anterior and posterior surfaces of the neck (Fig. 2). The next measurement included the cortical thickness measured at the center of the lateral component of the lateral wall (Fig. 1c). Finally, a line measuring 7 cm was dropped vertically from the level of the vastus ridge (line 4). The cortical index (CI) was measured at this level (7 cm below the vastus ridge) (Fig. 1b). The cortical thickness as well as CI was measured to represent the degree of osteoporosis. Figure 3 shows the measurements on 3D reconstruction for better clarity. Two experienced musculoskeletal radiologists (including one of the authors, SG) who were blinded to the clinical course of the patient performed all measurements independently. The mean value of the two sets of measurements was used for statistical analyses. All patients were treated with a 135° DHS (Synthes, GmbH, Oberdorf, Switzerland) after obtaining a good reduction and correction of the posterior sag. A good reduction was defined as normal or slight valgus reduction, less than 20° of angulation on the lateral radiograph, and less than 4 mm of fracture displacement [13, 14]. The guide wire was inserted using the angled guide which was seated flush to the lateral cortex. Minimal force was
International Orthopaedics (SICOT)
Fig. 1 Screen shot showing sagittal (a), coronal (b) and axial (c) CT images of a patient with AO/OTA 31 A2 fracture. A horizontal line was drawn at the level of the vastus ridge on the coronal view (line 1). The maximum height above the vastus ridge (yellow line) was measured on the coronal view proximal to line 1 (b). Then a line was drawn 2 cm below the vastus ridge (line 2). At the point of contact of line 2 with the lateral cortex, a third line was drawn at an angle of 135° (line 3).
Circumference of the lateral wall was measured on the axial section corresponding to line 3 (c). This circumference was further divided into an anterior (red line), lateral (brown line), and posterior component (green line). Cortical thickness was measured at the centre of the lateral component of the lateral wall (small black line). Cortical index was measured 7 cm below the vastus ridge (line 4) on the coronal view (curved black arrow)
applied while drilling the lateral wall with the triple reamer to ensure a smooth transition between the portions of the triple reamer. The head screw was placed in keeping with Baumgartner’s suggestion that tip apex distance (TAD) should be
