Author:Wei Chen, MD; ZekunZhang, MD; Yang Lu, Jia Li, MD; Yingze Zhang, MD; Yong Shen,

Background

Fracture of the posterior column is a common subtype of acetabularfracture. Percutaneous lag screw fixation is an alternative approach fornon-displaced or minimally displaced (<2 mm) posterior column fractures inpatients with severe soft tissue injury, burns, and an increased risk for majorsurgery. Percutaneous lag screws can also be used to fix a well-alignedacetabular fracture non-union or act as an adjunct to traditional openreduction and internal fixation. Starr et al applied this technique in thetreatment of displaced acetabular fractures after reduction was performed in aclosed or limited open fashion .

Percutaneousfixation can provide enough stability for early mobilisation . Early mobilisationis very important to avoid complications associated with prolonged bed rest. Additionally,the dense scar tissue, contractures and problematic hardware commonlyencountered after failed open operation of acetabular fractures can be avoided withthe percutaneous technique. However, this procedure is technically demandingand limited by narrow bony corridors. There are concerns about violating the lateralwall of the acetabulum, resulting in articular penetration during lag screwfixation. The lag screw may also protrude through the medial wall and jeopardiseintrapelvic organs . The sciatic nerve lies close to the posterior wall and issusceptible to injury in acetabular fractures involving the posterior wall andcolumn . The inferior gluteal neurovascular bundles also lie close to theposterior cortex of the posterior column (PCPC). The placement of lag screws into the posterior columnmay inadvertently penetrate the PCPC and damage the neurovascular structures.Therefore, it is essential to improve the accuracy of lag screw insertion into theposterior column to reduce the risk of iatrogenic injury to the hip joint,adjacent neurovascular structures and intrapelvic organs .

To guarantee safe percutaneous insertion of the lagscrew into the posterior column, a conventional image intensifier is frequentlyused during the operation. Multiple C-arm imaging at different angles isrequired for safeintraosseous placement of pelvic screws. The anteroposterior view of the pelvisand iliac oblique and obturator oblique views are commonly employed to ensurethat the guide wire does not penetrate the hip joint, medial wall or PCPC.However, misplacement of the posterior column lag screws still occurs in somecases. Thereafter, it is necessary to explore new projections forintraoperative fluoroscopic guidance of percutaneous screw insertion. Therefore,the purposes of the present study were to 1) determine the optimal fluoroscopicangles for visualisation of the PCPC and medial wall of the acetabulum incadaveric specimens, 2) confirm the effectiveness of these angles in ensuringintra-osseous positioning of retrograde percutaneous posterior column screws,and 3) evaluate whether fluoroscopic angles for visualisation of the posteriorcolumn in human subjects are similar to those identified in cadavericspecimens.

Methods

Sixteen embalmed adult cadavers were obtained from the Department ofAnatomy of Hebei Medical University (Shijiazhuang, China). The specimens wereall males with an average age of 41 years (range, 25-70 years). The specimenswere placed in the supine position on a radiolucent carbon fibre table. Thelongitudinal axis of the specimen was parallel to that of the table. CT scanswere performed on all specimens using a commercially available Siemens spiral64-slice multi-detector scanner (Siemens Medical, Nuremberg, Freistaat Bayern,Germany). The technical factors were 80-110 mAs, 120 kV, pitch 0.9 and anacquisition thickness of 0.75 mm. Axial images with a 2-mm thickness werecreated. The CT images of each specimen were observed and no bony deformity wasnoted. Previous anatomical study has shown that the smallest axialcross-section of the posterior column is at the middle height level of theacetabulum . The axial CT image at the level of the middle height of theacetabulum was selected. The PCPC was identified on the selected CT image and labelledLine A (Figure 1). The line intersecting the axial plane and the coronal planewas marked, which could be identified by drawing a line running through themost posterior points of bilateral acetabula. The angels between Line A and theline intersecting the axial plane and the coronal plane were measured using themeasurement software MB ruler (Markus Bader, Iffezheim, Germany) and labelledangle α. The medial wall of the acetabulum was also outlined and labelled LineB. The angle between Line B and the line intersecting the axial plane and the sagittalplane was labelled angle β and measured using an MB ruler (Figure 1). The line that ran through the anterior and posteriorpoints of the lateral brim of the acetabulum on the selected axial CT images waslabelled Line C. The angle between Lines A and C was marked and measured, and waslabelled angle γ (Figure 2).

Figure1 On the axial computed tomography (CT) image at the middle height levelof the acetabulum, Line A represents the posterior cortex of the posteriorcolumn. The angle between Line A and the line intersecting the axial plane andthe coronal plane is labelled angel α. Line B represents the medial wall of theacetabulum. The angle between Line B and the line intersecting the axial plane andthe sagittal plane is labelled angle β.

Figure2 On the axial CT images, Line C represents the line runningthrough the anterior and posterior points of the lateral brim of theacetabulum. The angle between Lines A and C is labelled angle γ.

Sixteen pelvic specimens, which were harvested fromthe aforementioned sixteen cadaveric specimens, were selected for posteriorcolumn retrograde lag screw insertion. All of the specimens were stripped ofsoft tissue. They were put into a radiolucent prefabricated box and placed inthe supine position on a operation table. The fluoroscopic guidance alone was usedfor screw insertion. AC-arm unit (Siemens Medical, Munchen, Germany) was used to establish the tangentialviews of both the PCPC and medial wall of the acetabulum of 16 bony pelvicspecimens by referencing the unique angles α and β measured on the CT imagesfor each specimen. The C arm was positioned according to the measured angles (Figure 3). The angular marking on the Carm was used to confirm the angles for the image intensifier. Theintra-operative correction of the C-arm position was not performed. A lag screwwas inserted retrogradely into the posterior column as close to the posteriorcortex as possible under the fluoroscopic guidance of the iliac oblique view andtwo tangential views. On the iliac oblique view, the lag screw was medial to thesubchondral bone of the acetabulum. On the tangential view of the PCPC, the lagscrew was medial and adjacent to the PCPC. On the tangential view of the medialwall of the acetabulum, the lag screw was inserted lateral to the medial wall.CT scans were then obtained to document the position of the lag screws.

Figure3 The diagrams illustrate the position of the C-arm unit and patientto obtain the tangential views of the posterior cortex of the posterior column(A) and the tangential view of the medial wall of the acetabulum of malepatient (B).

To confirm that the angles α, β and γ measured in human volunteers aresimilar to those observed in the cadaveric specimens, 138 volunteers who werescheduled to undergo CT scanning of the pelvis for suspected avascular necrosisof the femoral head were recruited. Eighty-six volunteers who were subsequently found to have evidence offemoral head pathology, bony deformity of the pelvis, or evidence of prioracetabular trauma or surgery were excluded. The remaining 52 volunteers were enrolled inthis study and provided informed consents to participate in the study. Neithermonetary nor non-monetary compensation was provided to these subjects. There were 27 males and 25 females, with a mean age of48 years (range, 31-69 years), mean height of 172 cm (range, 163-186 cm) and meanweight of 68.2 kg (range, 56-89 kg). The lead garment was used to protect thevolunteers from unnecessary radiation during CT scanning. Axial CT images of thevolunteers were obtained according to the same parameters as for the cadavericspecimens. Angles α, β and γ were measured on the axial CT images following thesame method as for the specimens. In the current study, the screw insertion itself wasonly performed on cadavers and not on the healthy volunteers. The Institutional Review Board of the Third Hospitalof Hebei Medical University approved this study after thorough examination andverification.

Statistical analysis

Statistical analyses were performed using SPSS 13.0 for Windows (SPSS,Chicago, IL, USA). Values were expressed as the mean±standard deviation (SD). Thetwo-tailed t test was applied to analysethese variables. A p value of<0.05 was considered significant.

Results

The angles α, β and γ measured onthe CT images at the level of the middle height of the acetabulum of specimens are summarised in table 1. The mean values of angles α, β and γ for specimens were 29.3 (range, 26.5-34.2), 8.1 (range, 6.9-9.8) and 93.7 (range, 85.7-103.6) degrees, respectively. Whilecollecting the obturator oblique radiographs of the acetabulum, the directionof the radiological beam was at an angle of 45 degrees with the line intersecting the axial plane and the coronalplane. Line A was at an angle of approximately 30 degrees with the lineintersecting the axial plane and the coronal plane. Accordingly, Line A was atan angle of approximately 15 degrees with the direction of the radiologicalbeam when collecting the obturator oblique view. That is to say, the tangentialprojection of the PCPC can be obtained by rotating the C-arm fluoroscopic beam 15degrees outward from the position where the obturator oblique view is collected.On this view, the projection of the PCPC appears as a nearly straight linesegment (Line segment A) between the lesser and greater sciatic notches (Figure 4). Line C was nearlyperpendicular to line A, which means that the entire outline of the acetabulumcan be almost demonstrated on the tangential views of the PCPC. In a similarway, the tangential view of the medial wall of the acetabulum can be obtainedby rotating the radiological beam angle β outward from the position where theanteroposterior view of pelvis is taken. On this view, the medial wall alsoappears as a distinct straight line (Figure5).

Figure4 On the tangential view of the posterior cortex of the posteriorcolumn, the posterior cortex appears as a nearly straight line segment betweenthe lesser and greater sciatic notches (Line segment A, the red curly brace).

Figure 5 On the tangential view of the medialwall of the acetabulum, the medial wall appears as a distinct straight line.

The lag screws were safely inserted into the posterior columns of thepelvic specimens under fluoroscopic control in the iliac oblique view and thetangential views of both the PCPC and medial wall of the acetabulum. During theprocedures, there were no failures in inserting the screws into the properposition. The CT images of the specimens confirmed that the full pathways oflag screws were all in the osseouscorridor of the posterior column. The shortest distance between the posteriorcortex and lag screw was at the level of the greater and lesser sciaticnotches, which was demonstrated on the oblique coronal reconstructed CT images(Figure 6). On the tangential viewsof the PCPC, the greater and lesser sciatic notches also seems closest to thelag screw, which was apparent in the oblique coronal CT images (Figure 4).

Figure 6 The oblique coronalreconstructed CT images shows that the full length of the lag screw is in the boneof the posterior column, and the shortest distance between the posterior cortexand lag screw was at the levels of greater and lesser sciatic notches.

The angles α, β and γ measured on the CT images at the level of the middleheight of the acetabulum of volunteers are summarised in table 1. The angles α,β and γ were 30.4±4.1 (range,25.9-36.5),9.2±1.9 (range, 6.3-11.9) and 91.6±4.8 (range, 85.2-101.4) degrees for male volunteers and 28.5±3.7 (range, 24.5-33.8), 7.7±1.8 (range, 5.2-10.3) and 93.9±5.2 (range, 86.9-103.7) degrees for female volunteers,respectively. No statistically significant differences were found between thespecimens and male volunteers for angles α, β or γ (P=0.359, P=0.067, P=0.210, respectively). For volunteers,there were no statistically significant differences between male and femalevolunteers for angles α and γ (P=0.092,P=0.111, respectively); however,angle β measured on males was significantly larger than that on females (P=0.006)

Table 1  The angles α, β and γmeasured on the selected axial CT images for both specimens and volunteers.

Discussion

In the current study, the axialCT images at the middle height level of the acetabulum were obtained for bothcadaveric specimens and volunteers. The angles between the PCPC and the lineintersecting the axial plane and the coronal plane and those between the medialwall of the acetabulum and the line intersecting the axial plane and the sagittalplane were measured. The radiological beams during collection of the obturatoroblique views of the acetabulum were at an angle of approximately 15 degreeswith the PCPC. Namely, the tangential view of the PCPC can be taken by rotatingthe C-arm unit 15 degrees outward from the position where the obturator obliqueview is taken. The tangential views of the medial wall of the acetabulum can beobtained by rotating the C-arm unit outward approximately 9.2 (for males) or7.7 degrees (for females) from the position where the anteroposterior view ofpelvis is taken. The lag screw inserted into the posterior column, under theguidance of the oblique iliac view and the tangential views of both the PCPCand medial wall of the acetabulum, can avoid penetrating the hip joint ordamaging adjacent neurovascular structures and intrapelvic organs.

Intraoperative and postoperative radiographicimaging was commonly used to evaluate possible joint penetration byperiacetabular screws.PostoperativeCT is generally accepted as the most accurate technique for the detection ofintra-articular screws . However, intraoperative CT scans are not available inthe majority of operating rooms, especially in developing countries.Fortunately, intraoperative fluoroscopy has been recently considered asaccurate as postoperative CT scans in detecting the misplacement ofperiacetabular screws.Anotherobvious advantage of fluoroscopy is that data obtained in the operating roomcan be used immediately, and corrective action can be taken before thecompletion of the surgical procedure, eliminating the need for reoperation. Fluoroscopicnavigation is a relatively new technique with numerous potential applicationsin the field of orthopaedic trauma . Recently, 2-dimensional and 3-dimensional fluoroscopicnavigation procedures were introduced for acetabular fracture surgery. Bothnavigation procedures can increase the precision of screw placement, preventintraarticular penetration during drilling, obviate the need for repeatedimaging in multiple planes and decrease radiation exposure for both the patientsand surgeons. However, fluoroscopic navigation requires specialised equipmentand instruments, and it is not available everywhere. Therefore, it is veryreasonable to apply a standard fluoroscopic technique for the safe insertion ofa lag screw to fix the posterior column fractures in most operating rooms.

Percutaneous screw fixation of a posterior column fracture has been achallenging task because of its unique and complex anatomy as well as the riskof penetration of the hip joint, damaging the adjacent neurovascular bundles orintrapelvic organs. Therefore, a technique for precise insertion ofpercutaneous screws requires knowledge of the 3-dimensional anatomy of the acetabulumand guidance with intraoperative fluoroscopy. Radiological evaluation,including an anteroposterior view, iliac oblique and obturator oblique views, isemployed during retrograde fixation of the posterior column using a lag screw.The iliac oblique view can be taken as a good reference for avoidingpenetration of the hip joint. However, there are still no specific views fordemonstrating the PCPC and the medial wall of the acetabulum, which can be usedto avoid damaging the neurovascular structures and intrapelvic organs by lagscrews.

In the current study, we introduced the tangential views of the PCPC andthe medial wall of the acetabulum. The tangential views of the PCPC can beobtained using the C-arm fluoroscopic unit during the operation. The PCPCoverlapped as a nearly straight-line segment between the lesser and greatersciatic notches as shown in figure 5. From this view, the lag screw insertedmedial to the straight line can ensure that the posterior cortex will not beprotruded and that the neurovascular bundles, including the inferior glutealnerve, inferior gluteal arteries and sciatic nerve, will not be injured. Thelag screw adjacent to Line segment A can also guarantee that the screw is awayfrom the hip joint and will not lead to intra-articular penetration. Thetangential view of the medial wall of the acetabulum can be used to ensure thatthe lag screw is placed lateral to the medial wall and leaves the intrapelvicorgans uninjured. Previousanatomical study has demonstrated that the thinnest part of posterior column is at thelevel of the middle height of the acetabulum . Therefore, appropriate positioningand directionality of lag screw at this level, as illustrated on the twotangential views and iliac oblique views, can help to estimate the pathway of thefull length of the screw. Following the three views, we inserted the lag screwsinto the posterior columns of 16 bony pelvic specimens, and subsequent CT scansconfirmed that the posterior columns safely accommodated the screws.

The limitations of this study include the small sample size of thespecimens and volunteers. The angles obtained in the study only represent the radiological features of a fraction of adult populations.Another limitation is that we have not applied this technique to treatingpatients with posterior column fractures. We plan to use these tangential views in the clinicto further confirm thetechnique’s effectiveness and validity in safe fixation of posterior columnfractures using lag screws.

Conclusions

The lag screw can be safely inserted into theposterior column under the guidance of the oblique iliac view and thetangential views of the PCPC and medialwall of the acetabulum. The angles between the PCPC and the line intersectingthe axial plane and the coronal plane and those between the medial wall of theacetabulum and the line intersecting the axial plane and the sagittal planewere measured on the axial CT images at the middle height level of theacetabulum. The tangential view of the PCPC can be taken by rotating the C-armunit approximately 15 degrees outward from the position where the obturatoroblique view is taken. The tangential views of the medial wall of theacetabulum can be obtained by rotating the C-arm unit outward approximately 9.2(for males) or 7.7 degrees (for females) from the position where theanteroposterior view of pelvis is taken.