Document Detail

From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine

 Full Text Journal Information Journal ID (nlm-ta): Clin Orthop Relat Res ISSN: 0009-921X ISSN: 1528-1132 Publisher: Springer-Verlag, New York Article Information Download PDF © The Author(s) 2010 Received Day: 11 Month: 12 Year: 2009 Accepted Day: 30 Month: 8 Year: 2010 Electronic publication date: Day: 25 Month: 9 Year: 2010 pmc-release publication date: Day: 25 Month: 9 Year: 2010 Print publication date: Month: 2 Year: 2011 Volume: 469 Issue: 2 First Page: 552 Last Page: 561 ID: 3018214 PubMed Id: 20872104 Publisher Id: 1567 DOI: 10.1007/s11999-010-1567-2
 Multilevel Measurement of Acetabular Version Using 3-D CT-generated Models: Implications for Hip Preservation Surgery Aimee C. Perreira, BSAff1 John C. Hunter, MDAff1 Thaddeus Laird, MDAff1 Amir A. Jamali, MDAff1 Address: amir.jamali@ucdmc.ucdavis.edu Dept. of Orthopaedic Surgery, University of California Davis, 4860 Y Street, #3800, Sacramento, CA 95817 USA

Introduction

OA of the hip is associated with numerous structural and morphologic abnormalities of the acetabulum, including hip dysplasia and acetabular retroversion [5, 8, 9, 16, 17, 23, 30, 35]. These disorders are characterized by abnormal acetabular bony coverage, depth, and orientation. These factors can alter load transmission across the hip, lead to instability, cause mechanical impingement, and ultimately damage the hip labrum and cartilage [79]. Some authors believe primary hip OA as an isolated entity is rare [10, 25, 31, 33] and more likely is attributable to abnormal anatomy that adversely affects the biomechanics of the hip [5]. Recognizing acetabular deformities and restoring normal anatomy may prevent the development of OA [5, 8, 17, 23, 30]. Numerous methods have been used to determine AV. These include direct manual measurements of cadaveric specimens, review of standard pelvic radiographs, measurement of 2-D axial CT cuts, and measurements from recently available 3-D models.

When using 2-D axial CT sections, positional changes of the patient in the scanner, pelvic tilt, and variability from reader to reader can introduce inconsistencies in measurement for acetabular abduction and version [1, 14, 24, 29, 34, 37]. For example, if the pelvis was tilted slightly toward one side in the coronal plane, this would introduce distortion of the axial image, with the cut plane being angled proximally on the pelvis on one side and distally on the other side. An analogous distortion of the axial images also can occur with excessive native lumbar kyphosis or lordosis. To establish quantitative, multilevel measurements of AV and measure acetabular abduction and ischial spine position not prone to this source of error, we developed a method to quantify these values using computer-generated 3-D models that would be independent of all external factors.

We (1) established quantitative measurements of AV, acetabular abduction, and ischial spine position using 3-D CT generated models from an institutional database; (2) used the models to determine the prevalence of acetabular retroversion in this series of models, and (3) determined whether retroversion was isolated to the superior acetabulum or involved the entire acetabulum and/or the pelvic segment including the acetabulum and the ischial spine.

Materials and Methods

We identified the pelvic CT scans of 50 randomly selected patients (100 acetabula) from a large database at our institution containing scans performed either for trauma or abdominal evaluation. All scans had been reviewed previously by a musculoskeletal radiologist for the original concerns. Scans with fractures, bone or soft tissue tumors, hardware, open growth plates, osteopenia, OA, or evidence of prior surgery were excluded from the study. All scans included the top of the sacrum down to the level of the lesser trochanters. The section thickness was 2.5 mm or thinner in all cases. There were 25 males and 25 females with an average age of 39 ± 12 years (range, 16–62 years). This study was performed in full compliance with the University of California at Davis Institutional Review Board.

The scans were entered in a commercially available software package (Mimics; Materialise, Ann Arbor, MI, USA), and 3-D surface models were created. The femora were removed from the 3-D models to facilitate detailed analysis of the acetabula. A specialized acetabular analysis module was created in the software. In each case, the following points were selected manually: bilateral anterior superior iliac spines (ASIS), ischial spines, the sacral midpoint (defined as the anterior midpoint of the superior endplate of S1), pubic symphysis center, and coccyx (defined as the anterior midpoint of C1) (Fig. 1).

Two specialized measurement spheres were created in a commercially available computer-assisted design software program (Solidedge; UGS, Plano, TX, USA) containing seven evenly spaced parallel planes running through each sphere with the fourth plane bisecting the sphere. A series of anatomic planes were constructed for each pelvis (Fig. 2). A summary of the definitions of each plane is provided (Table 1). The standard coronal plane (SCP) was defined as the plane formed between both ASIS points and the pubic symphysis. This plane is approximately vertical with upright standing [2, 4, 6, 20] but can vary from 10.4° ± 7.4° forward rotation in the supine position to 5.0° ± 9.4° forward rotation in the standing position [2]. The pelvic axial plane (PAP) was defined as the plane passing through both ASIS points and positioned normal to the SCP. This plane is approximately horizontal with upright standing. The sagittal plane (SP) was defined as the plane that passes through the sacral midpoint, runs through the center of the pubic symphysis, and is normal to the SCP and the PAP. The measurement spheres were sized and positioned manually to best outline the acetabular anatomy on all cuts. The measurement spheres then were rotated around their center and the center of the acetabulum such that their seven planes were exactly parallel to the PAP (Fig. 3). In this way, the equatorial acetabulum corresponded to Level 4 of the measurement sphere and defined the acetabular equatorial plane (AEP), which was defined as the plane that passes through the center of the acetabulum and is parallel to the PAP. The acetabular coronal plane (ACP) was defined as the plane passing through the acetabular center and parallel to the SCP. The point on the acetabulum where the ACP intersected the rim superiorly was labeled the 12 o’clock point and represented the most superior aspect of the acetabulum (Fig. 2). The point where the ACP intersected the acetabular rim inferiorly was labeled the 6 o’clock point and represented the inferior-most aspect of the acetabulum (Fig. 2). The acetabular abduction plane (AAP) was defined as the plane passing through the 12 and 6 o’clock points and normal to the SCP. To determine AV at multiple levels, points were plotted manually in the software where each of the seven parallel planes on the measurement sphere intersected the anterior and posterior rim of the acetabulum (Fig. 4). AV planes (AVPs) were created and defined as the plane formed between the anterior and posterior points at each level and normal to the PAP (Fig. 5). These were created at Levels 1 to 5. They were not created at Levels 6 and 7 owing to the lack of consistency of the intersection points on the anatomy of the inferior acetabulum. The AV at each level tested was calculated as the angle between the AVP and the SP as these two planes varied from one another in only one degree of freedom, namely the degree of AV. Version was given a positive value (anteversion) or a negative value (retroversion) based on the relationship between each AV and the SP (Fig. 5). Once all points had been placed, acetabular abduction was measured automatically by the software as the angle between the AAP and the AEP. The interASIS distance, interacetabular distance, and interischial spinous distances also were measured in the same manner. Interobserver and intraobserver reliability analyses were performed using intraclass correlation coefficients (ICCs) for the analysis method and were performed by two observers at two times, a minimum of 2 weeks apart for each observer on three specimens (six hips). The ICCs for the two sessions were 0.997 and 0.994 for the first (AAJ) and second observer (AP), respectively (intraobserver reliability). The ICC between the two observers was 0.996 and 0.997 for the first and second sessions, respectively (interobserver reliability).

The prevalence of acetabular retroversion at each level and overall was determined. To assess for a torsional abnormality of the pelvis at the level of the acetabulum, the distance from the ischial spine tip to the sagittal plane was divided by the acetabular center to the sagittal plane distance to calculate an ischial spine index (ISI) for each hip, thus standardizing the ischial spine position to the size of the pelvis.

To determine whether acetabular retroversion represented a torsional phenomenon of the entire pelvic segment containing the acetabulum and the ischial spine, simple linear regression was used. This analysis was first performed to determine the relationship between version at Levels 1, 2, 3, and 5 and version at the midacetabular level, Level 4. Subsequently, the same regression method was used to evaluate the relationship between ISI, as a marker of ischial spine distance from the sagittal plane, and version at Level 4. The coefficient of determination, r2, and the sample correlation coefficient, r, were obtained.

Differences in each measurement were compared based on gender using ANOVA. Post hoc testing was performed with the Bonferroni-Dunn method. Linear regression was performed using Excel® (Microsoft Corp, Redmond, WA, USA). ICC and ANOVA were performed with SPSS® (Version 9; SPSS Inc, Chicago, IL, USA) and StatView® software (SAS Institute Inc, Cary, NC, USA), respectively.

Results

The mean (± SD) anteversion at the midlevel of the acetabulum was 21.3° ± 5.8°. The anteversion decreased and was more variable at the superior levels of the acetabulum. In the uppermost level of the acetabulum, anteversion was 14.4° ± 10.5°. Acetabular abduction was 39.7° ± 4.3°. The ISI measured 0.6 ± 0.1 (Table 2).

The prevalence of acetabular retroversion overall was 7% (seven of 100 acetabula, in two pelves bilaterally, in three unilaterally). Of the seven retroverted acetabula, five occurred in males (one bilaterally, three unilaterally) and bilateral retroversion occurred in one female. The prevalence of retroversion was 7% (seven of 100 acetabula, two bilaterally, three unilaterally) at Level 1 and 2% (two of 100 acetabula, one pelvis bilaterally) at Level 2. At each level from 1 to 5, females had higher anteversion values than males (Table 2). Retroversion did not occur in this series of pelves at Levels 3, 4, or 5. Two acetabula in two separate male pelves did not have a Level 1 AV measurement because of excessive acetabular coverage caused by an abnormally deep acetabulum. We observed correlations between Level 1 and Level 4 AV (r = 0.74), the ISI and Level 4 AV (r = 0.67), Level 2 and Level 4 AV (r = 0.83), Level 3 and Level 4 AV (r = 0.95), and Level 4 and Level 5 AV (r = 0.92) (Fig. 6).

Discussion

Numerous structural deformities of the acetabulum are associated with hip OA. Current methods of evaluating acetabular anatomy are prone to inaccuracy from patient positioning and pelvic tilt. We established a quantitative measurement method using 3-D models to measure AV, abduction, and ischial spine position using an ISI. Using this method, we determined its prevalence and whether retroversion was isolated to the superior acetabulum or involved the entire acetabulum and/or the pelvic segment including the acetabulum and the ischial spine.

Using this technique, midacetabular anteversion was 21.3° ± 5.8°. This value is consistent with previous measurements reported in the literature ranging from 15° to 20° [3, 11, 21, 22, 2628, 32, 35, 36] (Table 3). We used the same software package to develop a standardized method to measure the distance from the ischial spine from the sagittal plane, termed the ISI.

We found a prevalence of acetabular retroversion of 7%, similar to published values of 5% to 6% [7, 9]. However, this prevalence was substantially lower than the 22% reported by Jamali et al. [13]. There are two potential explanations. First, in that series, the measurements were performed by hand on cadaveric specimens at a level 5 mm below the superior-most point of the acetabulum. In our study, a given distance was not used to determine the cranial acetabulum but rather a standardized measurement at a level [\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\raise0.7ex\hbox{1} \!\mathord{\left/ {\vphantom {1 8}}\right.\kern-\nulldelimiterspace} \!\lower0.7ex\hbox{8}}$$\end{document}] of the distance from the top of the acetabular cavity was used (at Level 1 of seven with Level 4 as the equator). Second, manual measurements performed in that series may have been less accurate than the computerized measurements performed here. A previous study showed the measurements performed in this software program are more accurate and precise than manual linear and angular measurements [12].

Our data are consistent with those of Giori and Trousdale [9], who suggested the radiographic findings seen in acetabular retroversion are not attributable solely to excessive bone anteriorly. However, in contrast to their suggestion that inadequate posterior coverage is solely to blame, our data would indicate that there is a torsional effect in acetabular retroversion that involves the pelvic segment that encompasses the entire acetabulum and the ischial spine. Furthermore, Jamali et al. [13] found a linear relationship between cranial and central AV in a study of cadaveric pelves measured manually, again indicating a torsional effect on AV. Finally, Kalberer et al. [15] studied 149 standardized AP pelvic radiographs and determined the presence of a visible ischial spine was highly associated with the cross-over sign and acetabular retroversion. A visible ischial spine indicates a smaller distance between the ischial spine and the sagittal plane and is a nonquantitative manifestation of the ISI we describe in this study. Our findings support theirs based on a positive linear correlation between the ISI and the central acetabular anteversion. As the ischial spine distance from the sagittal plane increases so does the degree of acetabular anteversion.

We found acetabular anatomy can be measured reproducibly using 3-D CT-generated models regardless of patient positioning in the scanner. We observed a correlation between cranial and central acetabular anteversion and between the ISI and central anteversion, suggesting acetabular retroversion is a phenomenon involving the entire pelvic segment containing the acetabulum and the ischial spine. Future effort will be directed toward automating the technique and potentially incorporating in vivo data on pelvic tilt to consider functional acetabular anteversion.

This information may be of benefit to surgeons in reproducing normal acetabular position and alignment at the time of periacetabular osteotomy or acetabular recontouring procedures. The recognition of a localized abnormal retroversion in the superior acetabulum would suggest optimal treatment with a superior anterior acetabuloplasty. In contrast if the acetabulum is abnormally retroverted at all levels, a more logical approach would be to consider reorientation of the entire socket using a periacetabular osteotomy. The nuances and degrees of such deformities and their symptomatic correlates are currently unknown and would benefit from further investigation.

Notes

Each author certifies that he or she has no commercial associations (eg, consultancies, stock ownership, equity interest, patent/licensing arrangements, etc) that might pose a conflict of interest in connection with the submitted article.

Each author certifies that his or her institution approved the human protocol for this investigation, that all investigations were conducted in conformity with ethical principles of research.

Acknowledgment

We thank Mazie Ngai for assistance with preparation of this manuscript.

References

Figures
 [Figure ID: Fig1] Fig. 1  Anatomic points are placed on the 3-D virtual model of the pelvis. Rt ASIS = right anterior superior iliac spine; Rt ACE Ctr = right acetabular center; Rt ACE 12 o’clock = right acetabulum 12 o’clock position; Rt ACE 6 o’clock = right acetabulum 6 o’clock position; Lt ASIS = left anterior superior iliac spine; Lt ACE Ctr = left acetabular center; Lt ACE 12 o’clock = left acetabulum 12 o’clock position; Lt ACE 6 o’clock = left acetabulum 6 o’clock position; Pubic Sym Ctr = pubic symphysis center. [Figure ID: Fig2] Fig. 2  The standard coronal plane (SCP) includes both anterior superior iliac spine (ASIS) points and the pubic symphysis. The pelvic axial plane (PAP) runs through both ASIS points and is normal to the SCP. Planes used in the analysis of the acetabular abduction include the acetabular equatorial plane (AEP), which is parallel to the PAP and is shown in yellow (black arrow), and the acetabular coronal plane (ACP), which is parallel to the SCP and is shown in blue (black arrow). The ACP and the AEP run through the center of the acetabulum. The acetabular abduction plane (AAP) is shown in red (black arrow) and runs through the 6 o’clock and 12 o’clock positions (white arrowheads) and is normal to the SCP. Acetabular abduction is calculated as the angle between the AAP and the AEP. L = left; R = right. [Figure ID: Fig3] Fig. 3  Anatomic planes of reference for the pelvis and acetabulum include the pelvic axial plane (PAP), the standard coronal plane (SCP), and the acetabular equatorial plane (AEP). After establishment of the AEP, the seven-level sphere (shown in red) is reoriented around its center such that its midlevel (Level 4) corresponds to the AEP. L = left; R = right; ASIS = anterior superior iliac spine. [Figure ID: Fig4] Fig. 4  A seven-layer sphere is overlaid on the left acetabulum, with Level 4 aligned with the acetabular equatorial plane (AEP) (shown in yellow). The intersection of each of the spheres with the rim of the acetabulum defines that level’s anterior and posterior version points. In this case, Level 4 Anterior, Level 5 Anterior, and Level 5 Posterior are hidden around the back of the seven-layer sphere (points shown as white open circles). Lev = level; Ant = anterior; Post = posterior. [Figure ID: Fig5] Fig. 5  An inferior view of the pelvis shows the version planes of the acetabulum. Version has a positive value (anteversion) if the AVP and SP would intersect anteriorly relative to the pelvis. Version has a negative value (retroversion) if the planes would intersect posteriorly. In this example, the “Version Level 1” refers to the Level 1 version of the right hip that intersects the SP posteriorly and thus is retroverted. In contrast, “Version Level 3” refers to the Level 3 version of the left hip that intersects the SP anteriorly and thus is anteverted. [Figure ID: Fig6] Fig. 6A–F  Simple linear regression plots are shown for (A) Level 1 versus Level 4 (midacetabular) AV, (B) Level 2 versus Level 4 AV, (C) Level 3 versus Level 4 AV, (D) Level 5 versus Level 4 AV, (E) abduction versus Level 4 AV, and (F) ISI versus Level 4 AV.

Tables
[TableWrap ID: Tab1] Table 1

Anatomic planes and their descriptions

Plane Description
Standard coronal plane (SCP) Intersects bilateral anterior superior iliac spines (ASIS) and pubic symphysis
Pelvic axial plane (PAP) Intersects bilateral ASIS and is normal to the SCP
Sagittal plane (SP) Intersects the midpoint of anterior S1 body and the pubic symphysis center and is normal to the SCP and PAP
Acetabular equatorial plane (AEP) Intersects acetabular center, is parallel to the PAP, is perpendicular to the SCP
Acetabular coronal plane (ACP) Intersects the acetabular center and is parallel to the SCP; the intersections of the plane and the superior and inferior rim of the acetabulum define the 6 o’clock and 12 o’clock points
Acetabular abduction plane (AAP) Intersects the superior and inferior acetabular points (6 o’clock and 12 o’clock points) and is normal to the SCP
Level 1 acetabular version plane (Level 1 AVP) Intersects Level 1 anterior and posterior points and is perpendicular to the PAP and AEP
Level 2 acetabular version plane (Level 2 AVP) Intersects Level 2 anterior and posterior points and perpendicular to the PAP and AEP
Level 3 acetabular version plane (Level 3 AVP) Intersects Level 3 anterior and posterior points and perpendicular to the PAP and AEP
Level 4 acetabular version plane (Level 4 AVP) Intersects Level 4 anterior and posterior points and perpendicular to the PAP and AEP
Level 5 acetabular version plane (Level 5 AVP) Intersects Level 5 anterior and posterior points and perpendicular to the PAP and AEP

[TableWrap ID: Tab2] Table 2

Quantitative results of pelvimetry, acetabular abduction/inclination, and acetabular version at five levels

Parameter Males Females Total p Value
Acetabular cup diameter (mm) 51.1 ± 2.5 (47.1–56.4) 45.4 ± 2.0 (43.1–49.9) 48.2 ± 3.7 (43.1–56.4) < 0.0001
Interanterior superior iliac spine distance (mm) 228.0 ± 14.7 (205.5–258.2) 222.4 ± 18.6 (180.5–156.7) 225.2 ± 16.9 (180.5–258.2) 0.3656
Interacetabular distance (mm) 169.3 ± 7.8 (154.0–183.0) 169.9 ± 10.0 (150.7–187.1) 169.6 ± 8.9 (150.7–187.1) 0.7502
Interischial spinous distance (mm) 90.2 ± 8.0 (74.8–109.2) 107.2 ± 6.7 (94.5–121.1) 98.7 ± 11.3 (74.8–121.1) < 0.0001
Ischial spine index 0.5 ± 0.04 (0.4–0.6) 0.6 ± 0.04 (0.5–0.7) 0.6 ± 0.1 (0.4–0.7) < 0.0001
Acetabular abduction/inclination (°) 39.4 ± 5.4 (29.4–57.0) 40.0 ± 2.9 (35.1–45.4) 39.7 ± 4.3 (29.4–57.0) 0.6904
Level 1 version (°) 11.6 ± 9.4 (−12.9–29.1) 17.0 ± 10.9 (−4.34–40.5) 14.4 ± 10.5 (−12.9–40.5) 0.0488
Level 2 version (°) 18.2 ± 7.4 (–2.4–28.57) 24.3 ± 7.8 (5.5–40.9) 21.2 ± 8.1 (−2.4–40.9) 0.0018
Level 3 version (°) 20.0 ± 4.8 (1.1–27.5) 25.1 ± 6.2 (7.5–38.8) 22.5 ± 6.1 (1.1–38.8) 0.002
Level 4 version (°) 18.9 ± 5.0 (0.7–30.47) 23.6 ± 5.5 (8.3–34.6) 21.3 ± 5.8 (0.7–34.6) 0.0014
Level 5 version (°) 19.7 ± 5.6 (1.38–32.09) 24.5 ± 6.7 (9.2–39.1) 22.1 ± 6.6 (1.38–39.1) 0.0033

Values are presented as mean ± SD, with range in parentheses.

[TableWrap ID: Tab3] Table 3

Acetabular anteversion measurements reported in the literature

Study Year Number of cases Method Anteversion (combined) Anteversion (males) Anteversion (females)
Reikeras et al. [27] 1983 47 CTs for various disorders “other than hip joint disease” (21 males, 26 females) 2-D axial CT measurement 17 ± 6° 16 ± 5° 18° ± 6°
Hoiseth et al. [11] 1989 40 (23 males, 17 females) 2-D axial CT measurement 19.4° 18.4° 20.7°
Maruyama et al. [21] 2001 200 specimens from an anthropologic collection (100 males, 100 females) Manual measurements 19.9° ± 6.6° 18.5° ± 5.8° 21.3° ± 7.1°
Jamali et al. [13] 2006 43 specimens from an anthropologic collection Manual measurements 20.1° ± 6.4° NA NA
Stem et al. [32] 2006 100 patients undergoing CT scanning for “nonorthopaedic pathology” (42 males, 58 females) 2-D axial CT measurement 23° ± 5° 22° ± 6° 25° ± 5° (age > 70 years) 23° ± 5° (age < 70 years)
Reynolds et al. [28] 1999 87 patients with hip pain, “normal radiographs per author’s evaluation” 2-D axial CT measurement 21° ± 20° NA NA
Murtha et al. [26] 2008 42 hips contralateral to THA 3-D CT-generated models 21.8° (calculated) 19.3° (8.5°–32.3°) 24.1° (14.0°–33.3°)
Bargar et al. [3] 2010 46 patients scheduled for THA (31 males, 15 females) 2-D axial CT measurement 15.1° ± 6.7° NA NA
Current study 2010 50 patients (25 males, 25 females) 3-D CT-generated models 21.3° ± 5.8° 18.9° ± 5.0° 23.6° ± 5.5°

Values are expressed as mean ± SD, with range in parentheses; 2-D = two-dimensional; 3-D = three-dimensional; NA = not available.

 Article Categories:Basic Research

Previous Document:  Is correctional osteotomy crucial in primary varus knees undergoing anterior cruciate ligament recon...
Next Document:  What factors influence long-term survivorship after hip arthroscopy?